Nonwoven Separators Designed for Supercapacitors

A Comparison of Nonwoven Separators for Supercapacitors1 EXECUTIVE SUMMARYSix separators were tested from two suppliers,Dreamweaver International and each at 40, 30and 25 microns, in supercapacitors assembled ofcommercial electrodes in a pouch cell.Separator properties, scanning electronmicrographs, and capacitor performance were allmeasured for each material. The followingconclusions were reached:

SEMS: Scanning electron micrographs (SEMs) revealed thatthe competitor separators are composedprimarily of fibrillated cellulose microfiberswith diameters in the range from 1-4 microns.DWI separators have similar composition, with amuch higher population of fibers with diametersin the range of 0.2 – 0.4 microns.

Separator Properties: The separators from thedifferent companies had, on average, similar basis weights, thickness,and porosity. The pore size and bubble point for DWI separators was slightly higher, but with the following advantages:

Gurley: a much lower (63%) Gurley air resistance. Strength: 21% higher tensile strength Modulus: 129% higher modulus Moisture: 25% lower moisture content

Competit

DWI Silver

Capacitor Performance: All of the materials showed a similar 24 hour self-discharge. On average, the DWI materials showed 9% higher capacitance, and 27% lower ESR. In the most dramatic comparison, at 30 microns, the DWI Silver AR30 had 13% higher capacitance and 61% lower ESR, as shown in the graph.

2 ABSTRACTTwo separator companies have designednonwoven separators specifically forelectrolytic double layer capacitors(EDLC), including supercapacitors,ultracapacitors, and EDLC capacitor –battery hybrids. The leading

competitor uses cellulose fibrillated to 1-4 micron fibers in auniform nonwoven web, and is the leader in the industry. DreamweaverInternational (DWI) uses cellulose fibrillated to approximately 250nanometer diameter fibers, combined with microfibers of approximately5 microns diameter, also in a uniform nonwoven web. In this paper,three separators are compared from each company. They are measuredfor thickness, moisture content, porosity, mean flow pore size, bubblepoint, tensile strength, and tensile modulus and Gurley airresistance. Images were taken under a scanning electron microscope.In addition, EDLCs were made using production electrodes and measuredfor self-discharge, capacitance and internal resistance. Thesemeasurements are compared to commercial capacitors from twomanufacturers. While bothcompanies make materials thatare very suitable for use in awide variety of EDLCs, onaverage the DWI materials showed9% higher capacitance, and 27%lower ESR. They are also 21%stronger on average, and have25% lower moisture content.

Summary: DWI Silver AR separators used a higher population of nano-sized fibers to provide higher electrical conductivity andcapacitance in a separator

Capacity ESR0.05.010.015.020.025.030.035.0

Supercapacitor Comparison

Leading Competitor 30micronDreamweaver Silver AR30

Capacity (F/g) or ESR (Ohm)

3 INTRODUCTIONElectric double layer capacitors (EDLCs) are energy storage devicescapable of providing very high power, up to 100 times that of evenhigh rate lithium ion batteries. This allows applications thatotherwise could not be done, such as high power signal conditioning inthe electric grid, regenerative braking in busses and other largetransport, and energy recovery in construction where heavy lifting isinvolved. In addition, EDLCs are finding application in portableelectronics, helping to extend battery life, improve burstcommunications, and provide rapid charging.

Because the power can be so high, especially in high energyapplications of EDLCs such as regenerative braking for large vehicles,the ohmic losses due to internal resistance of the EDLC can causeconsiderable heating and loss of efficiency. One driver of the energyloss is the separator. If a separator could be provided withsignificantly lower internal resistance, it could improve theperformance of the EDLC, lowering ohmic losses, reducing operatingtemperature and increasing energy efficiency.

4 EXPERIMENTALIn this study, commercially availableseparators from two manufacturers arecompared. The first manufacturer isDreamweaver International (DWI) andthe second is a leading competitor.The tests listed in Table 1 wereperformed on each separator. All ofthe testing was done by outside testlabs, which are also listed along with the test procedure where thereis a standard available. Other procedures are described below. Table 1: All of the testing was done at outside labs, using standard test procedures where available, as listed below.

Test Procedure Test Laboratory Standard or EquipmentScanning electron microscope

Clemson University Electron Microscopy Laboratory

Hitachi Analytical Variable Pressure ScanningElectron Microscope SU6600

Basis weight Herty Advanced TAPPI T220

Materials DevelopmentCenter

Thickness Herty AMDC TAPPI T220Moisture content Herty AMDC TAPPI T220Tensile strength Herty AMDC TAPPI T220Tensile modulus Herty AMDC TAPPI T843Gurley Herty AMDC TAPPI T460Mean flow pore size Porous Materials Inc ASTM F316Bubble point Porous Materials Inc ASTM F316Self-discharge Polystor Maccor 4000, method belowCapacitance Polystor Maccor 4000, method belowInternal resistance Polystor Maccor 4000, method below

4.1 SCANNING ELECTRON MICROSCOPEImages were taken at 500x, 1000x, 2000x, 5000x and 10,000x magnification usinga Hitachi Analytical Variable Pressure Scanning Electron Microscope SU6600.Eight sets were taken with each material, for a total of 40 images.

4.2 SEPARATOR PROPERTIESSeparator properties were measured at the Herty Advanced Materials DevelopmentCenter, and at Porous Materials, Inc. according to the test methods listed above.

4.3 CAPACITANCE TESTSPolyStor built cells using the separator materials specified below andproduction EDLC electrodes. The cells were ~ 5cm x 5cm square single cells,but used double sided electrodes (most production electrodes are doublesided). The separator materials and electrodes were dried under vacuum at 120Covernight prior to cell assembly. All cells were dried at 80C under vacuumovernight after assembly, but prior to electrolyte fill. The cells usedacetonitrile with 1M TEATFB salt blended by PolyStor using high puritymaterials. The cells were slightly overfilled with electrolyte and includedextra volume for a gas pocket. They were clamped using plastic plates withbinder clips. The weight included the weight of the double-sided electrodeand current collector. Because only one side of the electrode is used in thetest, the capacitance values are low, approximately half of what would beachieved in a production cell.

The capacitors were charged to 2.70V at ~10 mA/F rate and held for 10 minutesat 2.70V. Capacity was measured in Whr/g by immediate discharge from 2.70 to0.10 V at a discharge current of ~10 mA/F. To measure D.C. ESR (equivalentseries resistance), the capacitor was charged to 2.70V, held at 2.70V for 10

minutes and then immediately discharged at ~100 mA/F. The voltage drop at theconstant current (Id) was used to calculate the DC ESR:

The reported ESR is normalized for a 1 cm2 square area of capacitor.

The 24 hour self-discharge was measured by charging the capacitors to 2.70Vand holding for 10 minutes before removing the charge voltage. The voltage wasthen recorded continuously for 24 hours. The capacitors were then charged to2.70V, held for 10 minutes at 2.70V, and capacitance and D.C. ESR weremeasured using the methods described above.

5 RESULTS

5.1 SCANNING ELECTRON MICROSCOPE PICTURESRepresentative high magnification scanning electron microscope images of the various materials are shown below.

5.1.1 Competitor’s 40 Micron

5.1.2 Competitor’s 30 Micron

5.1.3 Competitor’s 25 Micron

5.1.4 DWI Silver AR40

5.1.5 DWI Silver AR30

5.1.6 DWI Silver AR25

5.1.7 SEM Image DiscussionAll of the materials were made primarily from fibrillated cellulose,made into a nonwoven sheet that looks and feels like paper. Inaddition, all of the materials appeared uniform and homogeneous,especially at low magnification (images not shown). At highmagnification, it can be seen that the competitor’s materials areconstituted primarily of fibers that range in diameter from 1-4microns, with many fibers in the range of 1-2 microns, and very fewfibers less than 1 micron in diameter. The DWI separators, incontrast, have a much higher population of fibers below 1 micron, anda large population of fibers in the 200 – 500 nanometer diameterrange.

5.2 SEPARATOR PROPERTIESTable 2: Separator properties for each of the six separators tested, along with the average for each supplier.

Units Competitor 40Micron

Competitor30

Micron

Competitor 25Micron

AverageCompetit

or

DWI SilverAR40

DWI Silve

rAR30

DWI Silve

rAR25

AverageDWI

Basis Weight

g/m2 28 20 12 20 21 18 18 19

Thickness (7.3psi)

microns

43 32 30 35 39 33 28 33

Thickness (12.6 psi)

microns

41 31 28 33 37 32 27 32

Thickness (25 psi)

microns

38 27 22 29 35 30 25 30

Porosity

% 55% 57% 72% 61% 60% 61% 54% 58%

Pore Size

microns

0.44 0.6 0.6 0.55 0.96 1.4 1.0 1.1

Bubble Point

microns

1.5 1.7 3.4 2.2 2.9 4.3 3.2 3.4

Gurley seconds

171 80 104 118 37 39 55 44

MD Strength

kg/cm2

250 99 174.5 210 185 240 212

MD Modulus

kg/cm2

9400 6200 7800 17500 16000

20000

17833

Moisture Content

% 8% 7% 8% 8% 6% 5% 6% 6%

Separator properties are reported in Table 2, above, along with theaverage for each supplier. For Competitor 30 micron, there was notenough material available for tensile tests. There are severalsignificant results:

Compressibility: All of the materials are compressible, changingthickness by 3-8 microns over the pressure range tested. Thecompetitor materials were slightly more compressible, compressingan additional 6 microns compared to 3 for the DWI materials.

Porosity: For 30 and 40 micron materials, the DWI porosity washigher, at ~60% compared to ~56%. However, at 25 microns, thecompetitor material had significantly higher porosity.

Pore Size: Both the bubble point and pore size were higher forthe DWI materials, with the bubble points for the DWI materialsall very similar to the competitor 25 micron.

Gurley: The DWI materials had much lower Gurley air resistancethan the competitor materials. Normally, this would correspondto lower internal resistance as well. See capacitance testingbelow.

MD Strength & Modulus: All of the materials except thecompetitor 25 micron had MD strength near 200 kg/cm2. Thecompetitor 25 micron was significantly lighter weight and higherporosity than the rest of the field, which resulted in a lowertensile strength and modulus.

Moisture Content: The DWI materials has 25% lower moisturecontent, likely due to the inclusion of PVA microfibers ratherthan solely cellulosic materials.

5.3 SUPERCAPACITOR TESTINGTable 3: Performance in supercapacitors for each of the six separators tested, along with the average for each supplier.

Units

Competitor 40Micron

Competitor 30Micron

Competitor 25Micron

AverageCompetit

or

DWI Silver

AR40

DWI Silver

AR30

DWI Silve

rAR25

Average DWI

24 hr Self Discharge

% 53% 57% 56% 56% 52% 56% 52% 53%

Capacity

Ah/g 0.024 0.023 0.027 0.025 0.025

0.026

0.029 0.027

Capacity

F/g 30.4 29.4 34.5 31.4 32.4 33.2 36.8 34.1

ESR Ohm 9.5 17.6 8.6 11.9 11.2 6.8 8.1 8.7

The results of the testing of supercapacitors is shown above in Table3. There are several results:

Self-discharge: Very little difference was seen between thematerials in 24 hour self-discharge of the cells. That the self-

discharge is higher than productions cells is likely due to cellconstruction.

Capacitance Trends: On average the capacitance was higher forthinner materials, with the only exception being the competitor30 micron separator.

Capacitance Comparison: On average, the capacitors made with DWIseparators had higher capacitance, a total of 9% across the threeseparator types by each manufacturer, and higher at eachthickness.

ESR: As a whole, the ESR was 27% lower for the DWI separatorsthan the competitor materials. Most notably was the differenceat 30 microns, with the competitor 30 micron material havingmore than 150% higher ESR than DWI Silver AR30.

Lowest ESR: The lowest ESR by far was the DWI Silver AR30, withan ESR of 6.8 Ohm. The closest competitor’s material was the 25micron, with an ESR of 8.6 Ohm-cm.

5.4 COMMERCIAL SUPERCAPACITOR TESTSTwo commercial one farad supercapacitors were also tested under thesame protocol as above, and then disassembled to determine electrodeweight and area. Both commercial parts show lower ESR and lowercapacitance per gram, indicating that a higher surface area, lowerelectrode thickness strategy was taken in order to reduce ESR to thelowest possible. The Maxwell part shows much lower ESR than thatNichicon part. This compares to the parts made with the competitor’sseparators and DWI separators, where the capacitance is near thepractical limit for carbon electrodes (these were double sidedelectrodes, but only a single side was measured, which wouldeffectively double the capacitance if a bulk supercapacitors were madeusing both sides. The practical maximum is around 70 F/g.) From thiscomparison with commercial supercapacitors, the following conclusionscan be supported:

The capacity compares well, considering a high energy design. The ESR compares as would be expected given the relative

capacitances. The self-discharge is lower for the commercial supercapacitors,

which likely has to do with cell design and formation processes,neither of which were optimized for the DWI and competitor cells.

Units

Maxwell(BCAP0001P270 T(0))

Nichicon(1F, 2.7V

UM(M) 1205PET)

AverageCommercial

AverageCompetitor

AverageDWI

24 hr Self Discharge

% 29% 32% 31% 56% 53%

Capacity Ah/g

0.006 0.005 0.006 0.025 0.027

Capacity F/g 7.6 6.6 7.1 31.4 34.1ESR Ohm 1.65 6.81 4.2 11.9 8.7

6 DISCUSSION AND CONCLUSIONSAn interesting comparison can be made between the competitor’s 30micron material and the DWI 30 micron material (Silver AR30). Inthose cells, the DWI separator provided 13% higher capacitance, 61%lower ESR, and equivalent self-discharge. At 40 and at 25 microns,the products from the two companies’ performance is more similar.

SEMS: Scanning electron micrographs (SEMs) revealed that thecompetitor’s separators are composed primarily of fibrillatedcellulose microfibers with diameters in the range from 1-4 microns.DWI separators have similar composition, with a much higher populationof fibers with diameters in the range of 0.2 – 0.4 microns

Separator Properties: The separators from the different companieshad, on average, similar basis weights, thickness, and porosity. Thepore size and bubble point forDWI separators was slightlyhigher, but with the followingadvantages:

Gurley: a much lower (63%)Gurley air resistance.

Strength: 21% highertensile strength

Modulus: 129% highermodulus

Moisture: 25% lowermoisture content

Capacity ESR0.05.010.015.020.025.030.035.0

Supercapacitor Comparison

Leading Competitor 30micronDreamweaver Silver AR30

Capacity (F/g) or ESR (Ohm)

Capacitor Performance: All of the materials showed a similar 24 hourself-discharge. On average, the DWI materials showed 9% highercapacitance, and 27% lower ESR. In the most dramatic comparison, at30 microns, the DWI Silver AR30 had 13% higher capacitance and 61%lower ESR, as shown in the graph below.

In conclusion, DWI Silver AR separators used a higher population ofnano-sized fibers to provide higher electrical conductivity andcapacitance in a separator that is also stronger, with lower moisturecontent.

7 REFERENCES

7.1 TEST LABORATORIES Herty Advanced Materials Development Center: www.herty.com. Contact

Martha Simmons, [email protected], (912) 963-2641. Clemson University Electron Microscopy Facility:

http://www.clemson.edu/centers-institutes/cuadvancedmaterialscenter/electron-microscope/. Contact George Wetzel, [email protected].

Porous Materials Incorporated: www.pmiapp.com. Contact Dr. KrishnaGupta, [email protected], (607) 257-5544.

Polystor: www.polystor.com. Contact Dr. James Kaschmitter,[email protected], (925) 570-7251.

7.2 AUTHOR INFORMATIONDr. Brian Morin is co-founder, President & COO of Dreamweaver International.He may be contacted at [email protected], and at 864-968-3321.