119
Fundamentals of Electrochemical Pseudocapacitors
Recent work combining carbon black and RuO 2 showed individual cells
with 24 Wh/kg and 4 kW/kg [28]. Other recent work has shown even higher
capacitance, with RuO 2 nanoparticles with carbon supports showing 1500
F/g, 40 Wh/kg, and 2 kW/kg in aqueous electrolytes [6]. This illustrates the
high energy density and performance that are possible for pseudocapacitance induced by undissolved redox couples. The overall coupled form of
the hydration-coupled redox mechanism is seen for any hydrous metal oxide
(transition metal M of the form MO x -nH 2 O):
MO x (OH) y + aH + + ae – ↔ MO x–a (OH) y+a
(3.IV)
As noted earlier, the high cost and toxicity of RuO 2 limit its global application
to ES devices [2]. Ir is an even more expensive (and thus limited) alternative. Other highly reversible redox materials such as W and Co [2] illustrate
high capacitance at low charge rates but have smaller potential windows (<1
V). It seems that MnO 2 is a safer, low cost alternative to RuO 2 . It exhibits
a nearly double-layer-like curve through the oxide transition from Mn 3+ to
Mn 4+ which has a potential window of 1 V. However, its conductivity is much
lower and the rate of proton coupling diffusion is slower. As a result only a
small fraction of films are electroactive. Pure thick films and composite electrodes containing binder and conductive additives can only reach 150 to 250
F/g [8,29]. However, controlled thin films (smaller than a few micrometers)
on conductive substrates can achieve more comparable pseudocapacitances
of 900–1300 F/g because all the MnO 2 is active in the redox process [29].
Limitations in applications of pseudocapacitance on a practical scale continue to be tackled by attempting to combine high area and conductive carbon with thin film coatings or using controlled nanostructures to reduce
charging restrictions that limit high rate and high energy pseudocapacitive
devices. However, with increased structural importance, it is even more
important to avoid irreversible degradation caused by phase transformations, physical relaxation, or dissolution seen during cycling.
Dissolution of the oxidant or reductant during a redox reaction can greatly
reduce the reversibility of the system. The dissolution can alter an electrode
material from an undissolved state to a dissolved one. In an undissolved state,
it exhibits a controlled three-dimensional morphology that is closely linked
to the conductive transport pathways of the current collector. When a redox
reaction moves the material into a dissolved state, the charge is lost. Then the
process becomes heavily diffusion limited and if the electrode was designed
for charging based on its undissolved state, the diffusion will likely generate
an overpotential on the material and the capacitance will be irreversible.
Further, the crystal phase or morphology within the electrode structure is
irreversibly lost. An example of this situation is seen for dissolution of MnO 2
in its Mn 2+ and Mn 7+ states [30,29]. The Mn 4+ ↔Mn 2+ redox occurs at 0.47 V versus NHE, while the Mn 4+ ↔Mn 7+ reaction occurs at 1.19 V versus NHE under
neutral pH (~6.5). In a symmetric design, this restricts the potential window
Fundamentals of Electrochemical Pseudocapacitors
Recent work combining carbon black and RuO 2 showed individual cells
with 24 Wh/kg and 4 kW/kg [28]. Other recent work has shown even higher
capacitance, with RuO 2 nanoparticles with carbon supports showing 1500
F/g, 40 Wh/kg, and 2 kW/kg in aqueous electrolytes [6]. This illustrates the
high energy density and performance that are possible for pseudocapacitance induced by undissolved redox couples. The overall coupled form of
the hydration-coupled redox mechanism is seen for any hydrous metal oxide
(transition metal M of the form MO x -nH 2 O):
MO x (OH) y + aH + + ae – ↔ MO x–a (OH) y+a
(3.IV)
As noted earlier, the high cost and toxicity of RuO 2 limit its global application
to ES devices [2]. Ir is an even more expensive (and thus limited) alternative. Other highly reversible redox materials such as W and Co [2] illustrate
high capacitance at low charge rates but have smaller potential windows (<1
V). It seems that MnO 2 is a safer, low cost alternative to RuO 2 . It exhibits
a nearly double-layer-like curve through the oxide transition from Mn 3+ to
Mn 4+ which has a potential window of 1 V. However, its conductivity is much
lower and the rate of proton coupling diffusion is slower. As a result only a
small fraction of films are electroactive. Pure thick films and composite electrodes containing binder and conductive additives can only reach 150 to 250
F/g [8,29]. However, controlled thin films (smaller than a few micrometers)
on conductive substrates can achieve more comparable pseudocapacitances
of 900–1300 F/g because all the MnO 2 is active in the redox process [29].
Limitations in applications of pseudocapacitance on a practical scale continue to be tackled by attempting to combine high area and conductive carbon with thin film coatings or using controlled nanostructures to reduce
charging restrictions that limit high rate and high energy pseudocapacitive
devices. However, with increased structural importance, it is even more
important to avoid irreversible degradation caused by phase transformations, physical relaxation, or dissolution seen during cycling.
Dissolution of the oxidant or reductant during a redox reaction can greatly
reduce the reversibility of the system. The dissolution can alter an electrode
material from an undissolved state to a dissolved one. In an undissolved state,
it exhibits a controlled three-dimensional morphology that is closely linked
to the conductive transport pathways of the current collector. When a redox
reaction moves the material into a dissolved state, the charge is lost. Then the
process becomes heavily diffusion limited and if the electrode was designed
for charging based on its undissolved state, the diffusion will likely generate
an overpotential on the material and the capacitance will be irreversible.
Further, the crystal phase or morphology within the electrode structure is
irreversibly lost. An example of this situation is seen for dissolution of MnO 2
in its Mn 2+ and Mn 7+ states [30,29]. The Mn 4+ ↔Mn 2+ redox occurs at 0.47 V versus NHE, while the Mn 4+ ↔Mn 7+ reaction occurs at 1.19 V versus NHE under
neutral pH (~6.5). In a symmetric design, this restricts the potential window