Appropedia needs your support - Please Donate Today

Increasing supercapacitor performance

From Appropedia
Jump to: navigation, search

A supercapacitorW is an energy storage devices with the ability to be charged and discharged very quickly, with little to no degradation in performance with an increase in number of charge/discharge cycles. Because of this property, supercapacitors bridge the gap between long term energy storage provided by a conventional electrochemical batteryW, and short term, high current energy demands provided by a standard dielectric capacitorW.

The problem with supercapacitors is that they still have a relatively high cost per Watt-hour of energy storage potential. Depending on the design, which can rely on Carbon/Carbon electrodes or metal oxide electrodes, material costs can severely hamper the cost benefit of employing supercapacitors in these power demand situations. In the Carbon/Carbon case, this is primarily due to the use of an expensive specially prepared high surface area carbon particulate or cloth that can cost US $50-100/kg [1]. Decreasing the cost by a factor of 10 is required in order to increase the market size for supercapacitors.

Factors affecting performance[edit]

The tradeoff between energy density and the RC time constant is an important design consideration. Often, a lower energy density (W-h/kg) is required in order to lower the RC time constant, and in turn, increase the power capability (W/kg) of the supercapacitor.

The performance of a supercapacitor is often dependent on the internal resistance characteristics of the device. A good understanding of the impedance of the supercapacitor is required in order to design a system with a matched impedance load type. A lowering of the equivalent series resistanceW or ESR is critical in increasing the specific power output of the supercapacitor.

New technologies[edit]

Work is being done on new metal oxides and carbon types for supercapacitor electrodes which will potentially decrease the cost of the supercapacitors with minimal decrease in performance [1]. Two of these major studies involve the use of carbon nanotubesW (CNT) as the electrode in the supercapacitor, and the use of an Al foil to act as the current collector [2].

Al Foil Current Collector[edit]

Electrodes for supercapacitors are commonly built on Al foil in the 100 - 300 um region to act as the current collector for the electrodes. As the power delivery potential of the device depends largely on the equivalent series resistance (ESR) of the supercapacitor, any reduction in impedance of the current collector will yield immediate power delivery improvements. This is readily apparent in the equation show below which is valid for a voltage in the capacitor between full and 1/2 of its rating [1].

P_{peak} = \frac{9}{16}\times\left(1 - EF \right)\times \frac{V_{0}^{2}}{R_{ESR}}


Where P is the peak power output of the supercapacitor, EF is the efficiency of the power pulse, V is the voltage across the supercapactitor, and R is the equivalent series resistance of the supercapacitor.

Research was done to find out how to decrease this ESR resistance in the Al current collectors. The steps taken are as follows

  1. 4 cm^2, 200 um thick Al foil is bathed in 1 molar NaOH electrolyte for 10 minutes
  2. Remove and rinse with distilled water
  3. Samples were then placed in 1 molar HCl at 80 degrees Celsius with a constant 200mA/cm^2 anodic current for 20 seconds.
  4. Remove and rinse with distilled water

What was found is that the average roughness of the Al foil increased from 0.2-0.3 um to 2.5-2.6 um with channels varying from 10-15 um deep [3]. This was likely due to the nucleation sites developed by the NaOH solution, and the subsequent dissolution of Al caused by the HCl bath with the anodic current. This is substantiated by the fact that the potential of the Al foil substantially increased after the first second of current being applied, suggesting an increased dissolution of Al [3]. As a consequence of this Al surface dissolution, the specific surface area of the Al foil increased as compared to the original Al foil. When combined with the active material in the supercapicitor, the larger surface area allows for more contact between the current collector and the electrode. This decreases the ESR of the supercapacitor, and therefore increases the specific power output. However, in the study cited, active material with a D50 of 10 um were used, which meant that the deep channels that were only a few um wide were poorly suited for providing a continuous contact surface since not all of the active material particles could fit into these channels.

An alternate method to increase the surface area of the Al foil, which can also be used in conjunction with the Al etching technique, is as follows. By depositing a sol-gelW with a small percentage of carbonaceousW rich material with small diameter in the 50 nm range through dip-coating, the specific surface area of the Al foil increases. Subsequent thermal treatment then removes the polymeric sol, leaving the small diameter conductive carbon particles behind on the Al foil. This treatment yet again reduces the ESR of the capacitor, and it now approaches the theoretical Nyquist plot for a carbon-carbon supercapacitor [3]. This conductive layer of 50 nm carbon particles serves as an excellent interface between the active material and the current collector. Other benefits of using this sol-gel process is that the Al surface is now protected from the electrolyte by a layer of carbon particles [2].

The sol-gel treatment along with the Al etching allows the ESR value to remain stable at 0.5 ohm-cm^2 over 10 000 charge/discharge cycles while having a specific capacitance of 92 F/g of active material [2]. The peak power output of the was calculated to be 55 kW/kg of active material and energy capacity to be 17 Wh/kg of active material [3]. In contrast, when the untreated Al foil and the etched Al foil was used in the supercapacitor, the internal resistance was found to be 50 ohm-cm^2 and 5 ohm-cm^2 [3].

Carbon Nanotubes[edit]

Double wall carbon nanotubes (DWNTs) were synthesized using catalysed chemical vapour depositionW made with an MgO catalyst at 18 mol percent CH4 in H2. The carbon nanotubes created from this process ranged in diameter from 10-20 nm. Then, the DWNTs were added to the activated carbon in a mechanical mixing process. The DWNTs were able to electrically link activated carbon particles that were previously isolated from other activated carbon particles. Therefore, the net surface contact area increases with a larger mass concentration of DWNTs in the activated carbon - carbon nanotube mix. With an increased area, the electrical equivalent series resistance of the active material decreases, thus increasing the possible power delivered. However, this increase in power delivery comes at a cost to the capacitance of the active material. DWNTs have a specific surface area of about 985 m^2/g, significantly lower than 1500-2000 m^2/g for activated carbon. Since capacitance is directly proportional to the surface area upon which the charge carriers (electrons) can accumulate, a decrease in specific surface area of the active material decreases the capacitance of the material.

Consequently, increased DWNT concentrations increased the power density of the by decreasing the equivalent series resistance of the cell, but decreased the energy density of of the supercapacitor by lowering the capacitance. Experimentally, it was found that about 15 weight percent DWNTs yielded the best tradeoff between energy density and power density. In fact, an ESR of about 0.4 ohm-cm^2 was found possible over 10 000 charge-discharge cycles, with about a 10% degradation over the test length. Also, a capacitance 93 F/g was found, a decrease of about 2 F/g over the etched Al foil with carbonaceous sol-gel deposit without carbon nanotubes. Also, it was found that the relaxation time of the supercapacitor decreased with the addition of 15 weight percent DWNTs, thereby decreasing the time required to charge and discharge the supercapacitor, and thus increasing the power delivered. With this in mind, the specific power delivered by the supercapacitor with 15 percent carbon nanotubes was found to be 110 W/g. Therefore, a slight decrease in energy density was offset by a near doubling of the power delivered by the circuit.

Carbon Nanofiber Web[edit]

The steps required to create a carbon nanofiber web are as follows.

  1. Polyacrylonitile (PAN) of 10 weight percent is dissolved into dimethylformamide [4]
  2. Solution is spun through the positively charged capillary using an electrospinning machine at an electropotential of 10-25 kV DC. The negative electrode was connected to the drum winding up the newly formed carbon nanofiber [4]
  3. The carbon nanofiber was then stabilized by heating at a rate of 1 degree Celsius/min up to 280 degrees Celsius where it is then held for an hour in air flow [4]
  4. These stabilized fibres are then ready to be processed further to increase pore size and change surface characteristics. This is done by heating the stabilized fibers at a rate of 5 degrees Celsius per minute up to three test temperatures of 700, 750, and 800 degrees Celsius. Then, they are activated by exposure to 30 volume percent steam mixed with N2 gas [4]
  5. Finally, the activated carbon nanofibers (ACNF) is heated in an oven at 150 degrees Celsius for 2 hours in order to remove water that had been adsorbed [4]

With a diameter between 200-400 nm, the ACNF had a specific surface area of a 850-1230 m^2/g. Surface area was found to decrease with a higher activation temperature. However, the mesopore volume fraction increased form 36% at 700 degrees Celsius to 62% at 800 degrees Celsius, with a corresponding decrease in the micropore volume fraction [4]. What was found is that under small current densities (under 10 mA/g), the specific capacitance of the electrode was highly dependent on surface area. Therefore, when subjected to these small current densities, the ACNF activated at 700 degrees Celsius provided the largest capacitance at 175 F/g. However, when the current density was above 10 mA/g, the specific capacitance of the smaller surface area test cases excelled. This can be explained by the higher mesopore volume fraction generated at the increased activation temperature. As such, the ions which are solvated by water molecules in the mesopores are much quicker at responding than those in the micropores when subjected to high current densities.

Experimentally, it was shown that the 700 degree activation had a higher specific capacitance of 175 F/g at 10 mA/g, but decreased by 55% when the current density was increased to 1000 mA/g [4]. On the other hand, the 800 degree activation had a specific capacitance of 155 F/g at 10 mA/g, and only decreased by 18% when the current density was increased [4].Maintaining high capacitances under high current densities is critical for supercapacitor applications as it directly determines the current capacities of the supercapacitor.

Therefore, using ACNF as an electrode with a high activation temperature of 800 degrees Celsius maintains a high specific capacitance regardless of the current density. This large capacitance increases the energy density of the supercapacitor. However, the ESR of the electrode was not very competitive at a value 6 ohm-cm^2, therefore ACNF electrodes will have a lower power density than activated carbon and DWNT electrodes [4].


Carbon Aerogel[edit]

The process used to generate carbon aerogels used in supercapacitor electrodes is as follows.

  1. Using pyrolysis, a resorcinol-formaldehyde (RF) gel with a molar ratio of formaldehyde to resorcinol maintained at 2 [5]
  2. Next, the gel is dissolved in distilled ion-exchanged water where the mass percentage of RF was set to 40%. The ratio of resorcinol to catalyst, NA2CO3, can be controlled. In this experiment the R/C ratio varied from 500-1500 [5]
  3. The sol-gel was put into a sealed glass container where it underwent polymerization for 24 hours at 298K, then 72 hours at 333K, finally 48 hours at 353K [5]
  4. The wet gels undergo performing (????) at 50 degrees Celsius for 6 days, followed by drying at ambient conditions for three more days [5]
  5. The carbon aerogel is finally synthesized by carbonization at 1073K for three hours [5]

Since the structure, pore size, and particle size can be controlled by the amount of catalyst in the mixture, the varying R/C ratio in the sol-gel process determines the characteristics of the carbon aerogel. Also, by changing the electrolyte used in the supercapacitor, performance also changes.

By using a 6M KOH electrolyte, the carbon aerogel produced using an R/C ratio of 1500 produced the highest capacitance in the experiment with 110.06 F/g at 1 mV/s scan rate [5]. When this scan rate was increased, the capacitance dropped off quickly as the ions in the electrolyte were likely unable to diffuse into the pores as quickly as needed. Over 400 charge/discharge cycles, the specific capacitance remains relatively constant, ensuring good performance over the long run.

References[edit]

  1. 1.0 1.1 1.2 Burke, Andrew. "Ultracapacitors: why, how, and where is the technology." Journal of Power Sources 91(2000): 37-50
  2. 2.0 2.1 2.2 Portet, C., P.L. Taberna, P. Simon, E. Flahaut, C. Laberty-Robert. "High power density electrodes for Carbon supercapacitor applications." Electrochimica Acta 50(2005): 4174-4181.
  3. 3.0 3.1 3.2 3.3 3.4 Portet, C., P.L. Taberna, P. Simon, C. Laberty-Robert. "Modification of Al current collector surface by sol–gel deposit for carbon–carbon supercapacitor applications." Electrochimica Acta 49(2004): 905–912
  4. 4.0 4.1 4.2 4.3 4.4 4.5 4.6 4.7 4.8 Kim, C., K.S. Yang. "Electrochemical properties of carbon nanofiber web as an electrode for supercapacitor prepared by electrospinning." Applied Physics Letters 93(2003): 1216-1218.
  5. 5.0 5.1 5.2 5.3 5.4 5.5 Li, J., Xianyou Wang, Qinghua Huang, Sergio Gamboa, P.J. Sebastian. "Studies on preparation and performances of carbon aerogel electrodes for the application of supercapacitor." Journal of Power Sources 158(2006): 784-788.
Mat.png This page was developed as part of a project for MECH370, a Queen's University class on materials processing. It is now open edit.