- Email: [email protected]
Electrochemically active surface area effects on the performance of manganese dioxide for electrochemical capacitor applications
Accepted Manuscript Title: Electrochemically Active Surface Area Effects on the Performance of Manganese Dioxide for Electrochemical Capacitor Applications Author: Madeleine Dupont Anthony F. Hollenkamp Scott W. Donne PII: DOI: Reference:
S0013-4686(13)00642-7 http://dx.doi.org/doi:10.1016/j.electacta.2013.04.007 EA 20307
To appear in:
Electrochimica Acta
Received date: Revised date: Accepted date:
3-12-2012 26-2-2013 4-4-2013
Please cite this article as: M. Dupont, A.F. Hollenkamp, S.W. Donne, Electrochemically Active Surface Area Effects on the Performance of Manganese Dioxide for Electrochemical Capacitor Applications, Electrochimica Acta (2013), http://dx.doi.org/10.1016/j.electacta.2013.04.007 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Electrochemically Active Surface Area Effects on the Performance of Manganese Dioxide for Electrochemical Capacitor Applications by
Discipline of Chemistry, University of Newcastle, Callaghan NSW, 2308, Australia 2
CSIRO Energy Technology, Box 312, Clayton South, VIC 3169, Australia
cr
1
ip t
Madeleine Dupont1, Anthony F. Hollenkamp2 and Scott W. Donne1*
us
Abstract
The specific surface area, morphology and electrochemical performance of thin films of
an
electrodeposited manganese dioxide have been examined. Electrodeposition of these films was carried out using chronoamperometry, using times ranging from 10-120 s in order to obtain deposits
M
with different masses. Using a novel approach to analysing the chronoamperometric i-t data, the specific surface area of the electrodeposited material was found to range from 13 – 67 m2/g across
d
the range of deposition times, with short deposition times leading to higher specific surface areas.
te
This has implications on the electrodeposition mechanism of manganese dioxide, which favours
Ac ce p
crystallite nucleation initially, leading to a high surface area material, followed by growth of these crystallites leading to a denser, lower surface area electrode material. This is the first time that the electrochemically active surface area of porous electrode materials has been determined. This decrease in surface area with deposition time was also reflected in the specific capacitance values of the material, which decreased slightly with increased deposition time, and hence lower surface area. Keywords: electrochemically active surface area; manganese dioxide; chronoamperometry; electrochemical capacitors * Corresponding author Ph: +61 2 4921 5477; Fax: +61 2 4921 5472; Email: [email protected]
Page 1 of 27
1. INTRODUCTION 1.1. Energy Storage Devices Increasing demand for energy has required the development of high performance energy storage devices. To be commercially viable, on whatever scale, energy storage devices need to be
ip t
able to store relatively large amounts of energy, and also have this energy readily accessible.
Chemical energy storage, in the form of a fuel, is an efficient, high grade form of energy
cr
storage, particularly when the energy is released electrochemically, in which case the resultant
us
electrical energy can be produced with efficiencies approaching 100%. Electrochemical energy storage and conversion devices include batteries, supercapacitors and fuel cells, each of which has
an
complementary performance characteristics. Supercapacitors have high power, but are limited by their low energy [1, 2]. The overall focus of our research is to develop approaches to improve the
M
specific energy density of supercapacitor materials and devices, with the intent of boosting their consumer acceptance.
d
Most commercial supercapacitors, based for example on activated carbon electrodes, and also
te
known as electrochemical capacitors, store energy in the form of charge separation at an electrical
Ac ce p
double layer [3]. Unlike electrolytic capacitors, which store charge on separated metal plates, supercapacitors store charge at the interface between an electrode and an electrolyte. Furthermore, unlike batteries, conventional supercapacitors do not undergo faradaic reactions at the electrodeelectrolyte interface, and since they notionally have no composition or phase change, they have a high degree of rechargeability and hence cyclability [4]. Supercapacitors perform significantly better than conventional electrolytic capacitors, with some supercapacitors designs demonstrating capacitance values up to 104 times higher than electrolytic capacitors [5]. Electrode materials used in supercapacitors can be either in the form of thin films or cast electrodes based on powdered materials. Thin films prepared by electrodeposition for example can be deposited easily and directly onto the required substrate and hence have low resistance and good electrical conductivity. This has resulted in thin films with extremely high specific capacitance per
Page 2 of 27
unit area, compared to powdered materials [6, 7]. Powdered materials, by comparison, are more difficult to prepare and require additives to be used as an electrode. However, they can be used in bulk quantities to develop supercapacitor devices with very large capacitances, which is one of the
ip t
limits of thin film electrodes.
1.2. Pseudocapacitance
cr
As a strategy to improve the specific energy density of supercapacitors, many researchers
us
have studied materials that exhibit psuedocapacitance. This phenomenon arises when a potential is applied to an electrode and fast, highly reversible faradaic reactions occur at the electrode surface in
an
addition to double layer charging. These processes can contribute significantly to the total capacitance of an electrode, and since they are faradaic in nature, they may also involve
M
compositional and phase changes for the electrode material [4]. Of course the specific reactions
te
the nature of the electrode material.
d
occurring, and the proportion of the total capacitance to which they contribute, vary depending upon
Ac ce p
1.3. Supercapacitor Electrode Materials
A wide variety of materials have been studied for their use as supercapacitor electrodes; e.g., carbons, metal oxides and conductive polymers. Supercapacitors often utilize high surface area electrode materials since this maximises charge storage in the double layer. Materials such as activated carbon have extremely high surface area (up to 2500 m2/g [8]) , but do not exhibit pseudocapacitance, and hence their maximum capacitance has been limited to ~400 F/g [9, 10], but most observed capacitance values are ~150 F/g. Metal oxides are another commonly used class of materials, due to both a relatively high surface area (double layer charging) and pseudocapacitance contributing to the overall capacitance. The prototypical material here is hydrated amorphous ruthenium dioxide, which has been shown to have a specific capacitance of over 800 F/g [9] in an acidic electrolyte. However, despite this
Page 3 of 27
excellent performance, its cost and toxicity have limited its widespread application. Manganese dioxide has also proven to be an excellent pseudocapacitive electrode material because, unlike ruthenium dioxide, it is inexpensive, relatively abundant, non-toxic and has been shown to exhibit capacitance of up to 2000 F/g [6].
ip t
There are many different polymorphs of manganese dioxide that have been studied as supercapacitor electrodes. Typically these materials are produced hydrothermally, via either
cr
oxidation of a Mn(II) species or reduction of Mn(VII), with the conditions used (supporting
us
electrolyte, temperature, etc.) invariably affecting the properties of the resultant material. Supercapacitor electrodes can also be prepared directly by electrodeposition of manganese dioxide,
an
which is the type of electrode studied in this work. Electrodes produced by this method are extremely thin (up to 100-200 nm), and hence exhibit very little resistance, and for these reasons are
d
1.4. Surface Area Measurements
M
desirable as electrode materials.
te
Measuring the electrochemically active surface area of any electrode material has proven to
Ac ce p
be extremely difficult even though it is one of the most basic properties of an electrified interface. To be clear, the electrochemically active surface area represents the area of the electrode material that is accessible to the electrolyte that is used for charge transfer and/or storage. One such approach for determining the electrochemically active surface area involves the use of the area specific capacitance of hydrogen ad-atoms on noble metals such as platinum [11]. While this is fundamentally very significant, the approach is only workable for these types of systems, rather than being widely applicable. Similarly, analysis of voltametric data obtained from a kinetically reversible redox couple (such as [Fe(CN)6]3-/4-) can also be used to estimate the electrochemically active surface area [12]. Again, though, this approach is limited to only inert and approximately planar electrode substrates. The situation becomes even more problematic and important when the solid electrode material is porous. For the specific case of electrodeposited manganese dioxide there
Page 4 of 27
is only one report on this in the literature. Kozawa [13] reported an estimate of the electrochemically active surface area of powdered manganese dioxide based on the ion exchange reaction that occurs between Zn2+ and H+ on the surface of the material, and assuming a certain area covered by an individual Zn2+ ion. However, this method determines the electroactive surface area
ip t
accessible to a Zn2+ ion, and so is not directly transferrable to the case under study.
There are, however, some measurements of surface area that can be obtained, such as the
cr
geometric surface area and BET (after Brunauer, Emmett and Teller) surface area. These
us
measurements can be and have been used as an approximation for the electrochemically active surface area, although with some reservation. Geometric surface area, calculated from particle size
an
and shape considerations, is a very inaccurate approximation of the electrochemically active surface area, especially in porous electrode materials (such as manganese dioxide), because it does not
M
account for the surface area contributed by pores.
The BET surface area is based on measurement of the area accessible to a gaseous adsorbate
d
(typically nitrogen or carbon dioxide), at a temperature where the adsorbate can condense on the
te
surface of the solid adsorbent material; i.e., multi-layer adsorption [14]. While accessing the porous
Ac ce p
surface area, this measurement is still expected to be different than the electrochemically active surface area, as almost always the adsorbate is different in size and chemical characteristics compared to the hydrated electrolyte ion accessing the pores in a supercapacitor electrode system. However this measurement is also only applicable to powdered samples of manganese dioxide, since the electrodeposited films prepared in our previous work do not contain enough material to allow for a measurement.
1.5. This Work Much of our previous work on supercapacitor electrodes has focussed on the electrodeposition of thin manganese dioxide films using chronoamperometry [6, 15]. In this work we will capitalize on the anomalous nature of the chronoamperometric i-t data to estimate the true
Page 5 of 27
electrochemically active surface area of manganese dioxide. This is the first such report where the electrochemically active surface area has been determined directly.
2. EXPERIMENTAL
ip t
2.1. Linear Sweep Voltammetry
Linear sweep voltammetry (LSV) was used to characterize the electrochemical oxidation of
cr
Mn2+ in an acidic environment to MnO2. To accomplish this, a previously cleaned platinum
us
working electrode (geometric area = 0.785 cm2) was placed in a solution of 0.01 M MnSO4 (≥99%; Sigma Aldrich) in 0.1 M H2SO4 in a 250 mL electrochemical cell. Cleaning of the platinum was
an
achieved by initially immersing the electrode into an acidified (0.1 M H2SO4) solution of 5% H2O2 to remove any residual manganese oxides by dissolution. This electrode was then polished using a
M
moist 1 μm Al2O3 paste on a polishing cloth. After ~2 minutes polishing, the electrode was washed thoroughly with Milli-Q ultra pure water (resistivity ρ > 18.2 MΩ.cm) before being ready for use.
d
Also placed into the electrochemical cell were a saturated calomel reference electrode (SCE; against
te
which all potentials were measured and reported), and a carbon rod (area = 3.5 cm2) as the counter
Ac ce p
electrode. The LSV experiment was conducted using a Perkin Elmer VMP 16-channel potentiostat/galvanostat. The potential was swept anodically at a rate of 5 mV/s from the open circuit potential up to 2.0 V versus SCE.
2.2. Chronoamperometry
Using the results from the LSV experiments, an appropriate potential was selected to carry out the chronoamperometry experiments. In this case, a diffusion limited potential was chosen. The MnO2 films were deposited using the same electrodes, electrochemical cell and electrolyte as the LSV experiments. The protocol used was to hold the platinum working electrode at its open circuit potential for 10 s, after which the potential was stepped to the chosen value where it was held for either 10, 20, 30, 60 or 120 s. To assess the reproducibility of the procedure, each set of
Page 6 of 27
experiments was repeated eight times, with the resultant standard deviation used to determine the error in the measurements.
2.3. Electrochemical Performance Evaluation
ip t
Once electrodeposition of the MnO2 had been carried out, the platinum electrode was removed from the electrochemical cell and then washed thoroughly with Milli-Q water to remove
cr
any entrained plating electrolyte. The electrode was then patted dry with paper towel to remove any
us
excess water. The thin film MnO2 electrode was then immersed into a 0.5 M Na2SO4 electrolyte together with the same SCE reference and carbon counter electrodes as before, and allowed to
an
equilibrate for 10 minutes. After this time the thin film MnO2 electrode was cycled between 0.0-0.8
M
V versus SCE at a range of scan rates for at least 50 cycles.
d
3. RESULTS AND DISCUSSION
te
3.1. Linear Sweep Voltammetry Data
Ac ce p
Figure 1 shows an example of the LSV data collected in this work. The data consists of a voltametric wave starting at ~1.1 V, with a maximum in current at ~1.3 V, superimposed on data for the oxygen evolution reaction. This overlap of processes was to be expected given that both anodic reactions have the same standard potential; i.e,
MnO2 + 4H+ + 2e- → Mn2+ + 2H2O
(Eo = 1.23 V)
...(1)
O2 + 4H+ + 4e- → 2H2O
(Eo = 1.23 V)
...(2)
The overlapping oxygen evolution reaction is also a complicating factor in determining the active mass of manganese dioxide deposited [15]. From this LSV data a diffusion limiting potential was chosen (1.35 V vs SCE) to carry out our chronoamperometry experiments.
Page 7 of 27
3.2. Mechanism of Manganese Dioxide Deposition While the anodic reaction in Eqn (1) above appears straightforward, the underlying mechanism of oxidation is much more complicated. In sulfuric acid manganese dioxide
ip t
electrodeposition from Mn2+ is proposed to occur via the following mechanism [16]:
Mn3+ + 2H2O → MnOOH + 3H+
us
2Mn3+ → Mn2+ + Mn4+
E0 = 1.56 V
cr
Mn2+ → Mn3+ + e-
Mn4+ + 2H2O → MnO2 + 4H+
...(3)
MnOOH → MnO2 + H+ + e-
an
Disproportionation pathway
Hydrolysis pathway
M
Here, the first step is oxidation of the solvated Mn2+ to form a soluble Mn3+ intermediate. At this point the mechanism is proposed to have two alternate pathways, the choice of which is dependent
d
on the acidity of the supporting electrolyte. In more concentrated acidic electrolytes the soluble
te
Mn3+ intermediate has a greater relative stability, meaning that it has the potential to diffuse away
Ac ce p
from the electrode surface, to the edge of the double layer, where it can undergo disproportionation to form soluble Mn2+ and Mn4+ which very quickly hydrolyzes to precipitate MnO2 on the electrode surface. In less concentrated acidic electrolytes the stability of the Mn3+ intermediate is less, in which case it can hydrolyse to MnOOH which precipitates on the electrode surface. MnOOH can then undergo solid state oxidation to MnO2. While these two pathways may seem distinct, in reality the oxidation of Mn2+ to form MnO2 occurs via a combination of the two pathways, with the acid concentration shifting the reaction preference. As a further note, the conditions of electrodeposition; i.e., electrolyte composition, temperature, anodic current density, etc., also influence the morphology of the deposit on the substrate surface.
3.3. Chronoamperometry Data
Page 8 of 27
Figure 2 shows the chronoamperometric (i-t) response for a deposition time of 2 minutes. While this data is somewhat consistent with the expected result from a chronoamperometry experiment; i.e., a current pulse upon imposition of the potential step, followed by a gradually
potential step.
ip t
decreasing current, the most significant difference is the slight bump in current at ~20 s after the
Under planar, semi-infinite diffusion limited conditions, the decay in current for an ideal
us
cr
chronoamperometry experiment is defined by the Cottrell equation [12]; i.e.,
1
(πt)
1
2
...(4)
an
i d (t) =
nFAD 2 C
M
where id(t) is the current under diffusion limited conditions (A), n is the number of electrons transferred in the redox reaction, A is the electrode area (m2), D is the diffusion coefficient (m2/s),
d
and C is the electroactive Mn2+ concentration in the bulk electrolyte (mol/m3). In this case, all of the
te
parameters in Eqn (4) are constant except for time, and hence the current decays with t-½.
Ac ce p
For the chronoamperometry data collected here, the parameters D, n and C are intrinsic to the system and so remain constant throughout the deposition. As such, the increase in current seen at ~20 s in Figure 2 can only be attributed to an increase in the surface area of the electrode. Of course, this increase in area arises from the electrodeposition of MnO2 particles onto the substrate surface, increasing the electroactive surface area onto which further MnO2 can be deposited. To model this, let us first consider the initial spike in current (t < 10 s), which we have assumed to be due to the initial oxidation of Mn2+ to MnO2 without an increase in electrode area. Since we have assumed this process does not affect the surface area of the electrode, possibly as the result of it being due to the initial stages of nucleation on the substrate surface, it can be modelled directly by the Cottrell equation (Eqn (4)), as also shown in Figure 2. To carry out this modelling, it was necessary to have an appropriate estimate of the electrochemically active surface area of the
Page 9 of 27
platinum substrate. This was determined ex-situ from the manganese-containing electrolysis cell by conducting cyclic voltammetry on the platinum substrate using an electrolyte of 10 mM Fe(CN)63in 1 M KNO3. This system is considered to be quite well behaved [17], and so using the appropriate modelling for a reversible redox couple [12], the electrochemically active surface area of the
ip t
platinum substrate was determined to be 1.38 cm2, as compared to the geometric area of 0.785 cm2 (1 cm diameter). This difference can of course be attributed to microscopic roughness of the
cr
platinum surface. So now with an estimate of the true platinum area, the Cottrell equation was
us
modelled to the chronoamperometric i-t data, with the only unknown variable being the diffusion coefficient (D). Using this approach, the value of D for Mn2+ in 0.1 M H2SO4 was determined to be
an
(1.6±0.3)×10-6 cm2/s, which is consistent with other similarly sized cations in aqueous solution [17]. Using the same basic approach, a second Cottrell equation was fitted to the t > 20 s
M
chronoamperometric data, this time using the value of the Mn2+ diffusion coefficient determined above, with the only unknown variable remaining being the surface area (A). For this fitting we
d
have again assumed that n = 2, in which case we are assuming that MnO2 is being deposited onto
te
the electrode surface, and that this is the species giving rise to the increase in electrode area. Figure
Ac ce p
2 also shows the result of a typical fitting to the chronoamperometry data for t > 20 s. The resultant electrode area determined using this approach was 2.6±0.1 cm2.
3.4. Calculating the Electrode Mass
Of course to be able to determine the specific surface area of the electrodeposited material we must also know the mass of active material on the electrode surface. Based on previous work in our laboratory, we have demonstrated that the most appropriate approach is to determine the mass from integration of the chronoamperometric i-t data. This approach has been shown to be unencumbered by entrained electrolyte, as methods such as EQCM and direct dissolution and analysis of the manganese content are. Nevertheless, integration of the i-t data is complicated by the presence of competing electrochemical reactions that can take up charge that would otherwise contribute to the
Page 10 of 27
deposition of manganese dioxide. In this case, the most important competing reaction is that of oxygen evolution, for which both the MnO2/Mn2+ and O2/H2O redox couples have the same Eo value (1.23 V vs SHE). The overlap between these two processes is very apparent in the linear sweep voltammetry data in Figure 1, with the diffusion limited wave for Mn2+ oxidation quite
ip t
clearly superimposed on the growing oxygen evolution curve. To separate these two processes we have fit a high field approximation of the Butler-Volmer equation to the oxygen evolution data, as
us
cr
also shown in Figure 1; i.e.,
...(5)
an
⎛ α (V − E)F ⎞ i a = i 0 exp⎜ a ⎟ RT ⎠ ⎝
M
where ia is the current density (A/m2), i0 the exchange current density (A/m2), αa is the anodic transfer coefficient, V is the applied potential (V), E is the equilibrium potential (V), and all other
d
symbols have their usual significance. With this modelling of the oxygen evolution reaction
te
complete, the current flowing due to Mn2+ oxidation can be extracted, as is also shown in Figure 1. With this separation of processes, the fraction of current during the chronoamperometric experiment
Ac ce p
due to the electrodeposition of manganese dioxide can be determined, which in this case for a potential step to 1.35 V, is 30.3%. As an example, from the total amount of charge passed during the 120 s electrodeposition experiment ((60±7)×10-3 C), and using this charge efficiency, the mass of manganese dioxide prepared was therefore 8.2±1.0 μg. As a further example, Figure 3 shows how the electrode mass changes with chronoamperometric deposition time, in which case there is an approximately linear change in mass with deposition time. While this linearity was not expected, especially for a changing current with time (a linear response should have been expected for a constant current), it can be justified by the fact that the changes in current between subsequent sampling points was relatively small, approaching a constant current.
Page 11 of 27
3.5. Electrochemically Active Surface Area
Following on from the previous analysis of electrode mass and apparent surface area, the specific surface area was determined to be 32±5 m2/g. This result is consistent with the BET analysis of a commercial electrolytic manganese dioxide (EMD) [18], although the conditions we
ip t
have used here to carry out our electrodeposition are considerably different to commercial practices, particularly in terms of current density used, current profile (constant current versus constant
cr
potential), electrolyte composition (0.01 M Mn2+ + 0.1 M H2SO4 versus 1.0 M MnSO4 + 0.3 M
us
H2SO4) and the temperature of electrodeposition (22°C versus 98°C). The specific surface area calculated here can also be considered to be much smaller than the BET surface area of a
an
chemically prepared manganese dioxide (CMD), which are typically 80 m2/g, although there is considerable variability here due to the wide range of experimental conditions that can be used to
M
prepared CMD. It is also important to recognize that we are also measuring a different phenomena in this case, compared to the surface area assessment carried out via a BET analysis. This is
d
primarily due to the different properties of the probing molecule; i.e., a hydrated Mn2+ ion at
Ac ce p
carried out.
te
ambient temperature in this work, compared to a N2 gas molecule at 77 K when a BET analysis is
Electrodes deposited for 2 min provided the only data for which the diffusion coefficient could be fitted reliably. For the other shorter depositions, the current profile was too short to allow accurate fitting of both processes, as described previously. However, given that the chronoamperometry experiment is conducted under the same conditions; i.e., electrolyte composition and temperature, the diffusion coefficient should be a constant for all experiments. Hence, the diffusion coefficient determined from the 2 min deposition data was used in the fittings for the 1 min and 30 s deposition times, which allowed values for the specific surface area to be calculated. For deposition times of less than 30 seconds, the amount of data that could be collected over this time period was insufficient to allow an accurate fitting of the current curve. The results of this analysis are presented in Figure 4 as a plot of the specific surface area as a function of the mass
Page 12 of 27
of deposited MnO2. The general trend appears to be a negative correlation of specific surface area with deposition mass, at least for these deposition times, i.e., t > 30 s. A mechanism by which the surface area decreases with a longer deposition time was proposed by Cross et al. [6]. In the initial stages, manganese dioxide deposition occurs predominantly by crystal nucleation on the surface of
ip t
the electrode, which increases the electrode area. As further material is deposited, crystal nucleation on the surface is replaced by the continued growth of existing crystals. This mechanism acts to
cr
reduce the surface area of the material. The results obtained from these chronoamperometry
us
experiments support this proposed mechanism.
an
3.6. Deposit Morphology
The morphology of these thin film deposits was quite difficult to evaluate since very little
M
active material on the substrate is available for characterization. Conventional methods of morphology determination, including SEM and TEM, are problematic again because little material
d
is available, and also because the material must be dried from its original state which can have an
te
impact on its morphology due to surface tension effects associated with water removal from the
Ac ce p
active manganese dioxide. Furthermore, physically scraping the active material off the substrate so that analysis by TEM can be conducted is also not ideal, nor really representative, although we have reported some success in this regard in our previous work [6]. Because of these factors we have employed atomic force microscopy (AFM) in an attempt to characterize the morphology of our thin film electrodes.
Figure 5 shows an example of the morphology of the thin film electrodeposited manganese dioxide samples prepared in this study. Specifically, this figure compares the morphology of (a) the bare platinum substrate with manganese dioxide samples electrodeposited for (b) 30 s and (c) 120 s. What is of significance here is the apparent increase in surface roughness, and hence surface area, of the substrate with manganese dioxide present. This image also provides some insight into the growth mechanism of electrodeposited manganese dioxide. By comparing the bare platinum
Page 13 of 27
substrate with the electrodeposited film it is clear that the manganese dioxide crystallites nucleate in isolated regions on the substrate, which then subsequently grow into larger crystallites, and ultimately merge
ip t
3.7. Electrochemical Performance
The electrochemical performance of each electrodeposited thin film of manganese dioxide
cr
was examined using cyclic voltammetry, in which case the potential of the electrodes was scanned
us
between 0.0 – 0.8 V vs SCE for at least 50 cycles. Figure 6(a) shows a typical cyclic voltammogram for an electrode prepared in this work, in this case for an electrode deposited for 30 s, cycled using a
an
scan rate of 250 mV/s. All of the electrodes studied exhibited the typical behaviour expected for an electrochemical capacitor; i.e., a ‘box’ shaped voltammogram over the potential window of interest.
M
Furthermore, the specific capacitance of these electrodes was essentially constant over the number of cycles considered, as shown in Figure 6(b).
d
The specific capacitance reported for these materials in Figures 6(b) and 7, is certainly much
te
higher than that reported in previous studies focussed on the use of manganese dioxide as an
Ac ce p
electrode material for electrochemical capacitors [9]. In some cases the specific capacitance that we have reported (here and in previous works [6]), is above the theoretical specific capacitance for the manganese dioxide one-electron pseudo-capacitance reaction; i.e., 1386 F/g for the reaction
MnO2 + M+ + e- ↔ MnOO(M)
...(6)
where M is a cationic species from the electrolyte, such as H+, Li+, Na+ or K+, with a potential window of 0.8 V. The origin of this enhanced specific capacitance arises from a combination of sources including pseudo-capacitance (faradaic processes accessing the bulk of the manganese dioxide structure; as above in Eqn (6)) and electrical double layer charging (non-faradaic processes involving only the manganese dioxide-electrolyte interface). While the process represented by Eqn
Page 14 of 27
(6) is expected to be the main contributor to faradaic pseudo-capacitance, it is also possible that the second electron discharge of the material; i.e., Mn(III) to Mn(II) MnOO(M) + M+ + e- ↔ Mn(OM)2
...(7)
can also contribute to the specific capacitance. In various aqueous battery systems, manganese
will enable the second electron discharge; i.e., an additional 1386 F/g.
ip t
dioxide reduction occurs typically occurs only over ~0.6 V [19], and so the 0.8 V window used here
cr
Non-faradaic processes also have the potential to contribute significantly to the total specific
us
capacitance because of the high density of charge adsorption sites on the manganese dioxide surface which have the capacity to electrostatically adsorb cations from the electrolyte. Calculations
an
conducted previously [6] demonstrated that the area specific capacitance can be estimated to be ~100 μF/cm2, which is an order of magnitude higher than the area specific capacitance of a planar
M
metal electrode [11] . It was also noted here [6] that this value was expected to be dependent on the polarity of the Mn-O bond. This could also be responsible for potentially multi-layer adsorption
d
onto the manganese dioxide surface, somewhat akin to ionic liquid ‘banding’ of anions and cations
te
on a range of substrates [20]. This process is expected to facilitate an increase in charge storage at
Ac ce p
the interface. As a final point, note that the contributions made by these various charge storage processes (faradaic and non-faradaic) depends on the nature of the material being studied, and also the cycling conditions.
Voltametric data has been used to compare the specific capacitance of the electrodeposited electrodes for different deposition times and scan rates. Figure 7 shows the change in specific capacitance with increased deposition time. For deposition times shorter than 30 s, there is no apparent correlation between the specific capacitance and deposition time. However, for those deposition experiments longer than 30 s there appears to be a decrease in the specific capacitance of the active material. According to the previous data and discussion, this may be attributed to a decrease in the surface area of the electrodes deposited for longer times, and hence a lower contribution made from electrical double layer charging. The higher capacitance observed for lower
Page 15 of 27
scan rates is due to the increased contribution of proton insertion and better utilization of the active material, in addition to any non-faradaic contributions. Conversely, at higher scan rates, we see a decrease in specific capacitance for a number of reasons. Firstly, faradaic contributions to the specific capacitance decrease because the electrode kinetics (i.e., charge transfer at the manganese
ip t
dioxide-electrolyte interface, and mass transport of the intercalated cations away from the interface towards the core of the particles) are apparently too slow to allow for as great a utilization of the
cr
active material. Secondly, for non-faradaic contributions, mass transport of the intercalated ions in
us
the electrolyte through the manganese dioxide porous structure can also become limiting [21] under high cycling rates, and as such continued non-faradaic charge storage at the interface will be
an
restricted.
A number of other literature studies have observed a decrease in capacitance as the amount of
M
MnO2 deposited is increased [22-24]. This phenomenon has been attributed to a decrease in surface area; however, none of these reports were able to confirm this with specific surface area
te
Ac ce p
specific surface area.
d
measurements. Whereas the work presented here confirms this through actual calculation of the
4. SUMMARY AND CONCLUSIONS In this work, the electrochemically active surface area of thin films of electrodeposited MnO2 has been calculated using chronoamperometry. The chronoamperometry results showed an increase in the current response during deposition, which was attributed to an increase in the surface area of the electrode. These results were used to calculate the specific surface area of the MnO2 deposit. Surface areas were calculated for deposition times 30, 60 and 120 seconds, and values between 13 – 67 m2/g were obtained, with the surface area decreasing as the mass of deposited MnO2 increased. This decrease in surface area with increase in deposition time may be attributed to the mechanism of MnO2 deposition. The proposed mechanism suggests that MnO2 is initially nucleates
Page 16 of 27
as crystallites and as further MnO2 is deposited these existing crystallites grow and begin to
5. REFERENCES
ip t
agglomerate, reducing the surface area.
Simon, P. and Y. Gogotsi, Nature Materials, 7 (2008) 845-854.
2.
Shukla, A.K., et al., Electrochimica Acta, 84 (2012) 165-173.
3.
P, K., CAPACITORS | Electrochemical Double-Layer Capacitors: Carbon Materials, in
us
cr
1.
Encyclopedia of Electrochemical Power Sources, G. Editor-in-Chief: Jürgen, Editor 2009,
4.
Conway,
B.E.,
an
Elsevier: Amsterdam. p. 634-648. Electrochemical
Supercapacitors.
Scientific
Fundamentals
and
5.
M
Technological Applications1999, New York: Kluwer Academic/Plenum Publishers.
Kurzweil, P., CAPACITORS | Electrochemical Double-Layer Capacitors, in Encyclopedia
7. 8. 9. 10. 11.
Cross, A., et al., Journal of Power Sources, 196 (2011) 7847-7853.
Ac ce p
6.
te
Amsterdam. p. 607-633.
d
of Electrochemical Power Sources, G. Editor-in-Chief: Jürgen, Editor 2009, Elsevier:
Lokhande, C.D., D.P. Dubai, and O.-S. Joo, Current Applied Physics, 11 (2011) 255-270. Kötz, R. and M. Carlen, Electrochimica Acta, 45 (2000) 2483-2498.
Naoi, K. and P. Simon, Electrochemical Society Interface, 17 (2008) 34. Alonso, A., et al., Carbon, 44 (2006) 441-446.
Bockris, J.O.M. and A.K.N. Reddy, Modern Electrochemistry. Vol. 2. 1970, New York:
Plenum Press. 12.
Bard, A.J. and L.R. Faulkner, Electrochemical Methods: Fundamentals and Applications. 2nd ed2001, New York John Wiley & Sons, Inc.
13.
Kozawa, A., Journal of the Electrochemical Society, 106 (1959) 552-556.
14.
Emmett, P.H. and S. Brunauer, J. Am. Chem. Soc., 59 (1937) 1553-64.
Page 17 of 27
15.
Cross, A.D., et al., Electrochimica Acta, 87 (2013) 133-139.
16.
Clarke, C.J., G.J. Browning, and S.W. Donne, Electrochimica Acta, 51 (2006) 5773-5784.
17.
Adams, R.N., Elecrtochemistry at Solid State Electrodes1969, New York: Marcel Dekker Inc. Williams, R.P., Characterisation and Production of High Performance Electrolytic
ip t
18.
Manganese Dioxide for use in Primary Alkaline Cells, in Chemistry1996, University of
cr
Newcastle: Callaghan.
Bailey, M.R. and S.W. Donne, Electrochimica Acta, 56 (2011) 5037-5045.
20.
Atkin, R. and G.G. Warr, The Journal of Physical Chemistry C, 111 (2007) 5162-5168.
21.
Arnott, J.B., G.J. Browning, and S.W. Donne, J. Electrochem. Soc., 153 (2006) A1332-
an
us
19.
A1340.
Babakhani, B. and D.G. Ivey, J. Power Sources, 196 (2011) 10762-10774.
23.
Broughton, J.N. and M.J. Brett, Electrochim. Acta, 50 (2005) 4814-4819.
24.
Huang, Q., X. Wang, and J. Li, Electrochim. Acta, 52 (2006) 1758-1762.
Ac ce p
te
d
M
22.
Page 18 of 27
Figure Captions Figure 1. Linear sweep voltammogram of manganese dioxide electrodepsosition from an
electrolyte of 0.01 M MnSO4 + 0.1 M H2SO4 using a platinum substrate (geometric area = 0.785 cm2) and a SCE reference electrode. Scan rate = 5 mV/s. Also included (dashed line) is a fitted
ip t
curve for oxygen evolution (Butler – Volmer equation), and the fraction of current contributed by oxygen evolution and manganese dioxide electrodeposition.
cr
Figure 2. Chronoamperometric data (solid line) from the electrodeposition manganese dioxide from
us
an electrolyte of 0.01 M MnSO4 + 0.1 M H2SO4 using a platinum substrate (geometric area = 0.785 cm2; electrochemically active surface area = 1.38 cm2) and a SCE reference electrode. Step
an
potential = 1.35 V vs SCE. Also shown (dashed line) is the Cottrell equation modelling of the early and latter stages of deposition, using two expression with different surface area values.
M
Figure 3. Manganese dioxide mass as a function of electrodeposition time. Data have been
corrected for the contributions made by oxygen evolution to the total current flowing.
d
Figure 4. Specific surface area of the electrodeposited manganese dioxide as a function of electrode
te
mass.
Ac ce p
Figure 5. AFM images of (a) the bare platinum substrate, and the substrate coated with
electrodeposited manganese dioxide for (b) 30 s and (c) 120 s chronoamperometric electrodeposition at 1.35 V vs SCE from an electrolyte of 0.01 M MnSO4 + 0.1 M H2SO4. Figure 6. Sample electrochemical performance data for manganese dioxide electrodeposited using
chronoamperometry from an electrolyte of 0.01 M MnSO4 + 0.1 M H2SO4, with a step potential of 1.35 V vs SCE, for a duration of 30 s. The cyclic voltammogram in (a) was conducted in 0.5 M Na2SO4 at 250 mV/s (cycle 30), while (b) shows cycle stability. Figure 7. Specific capacitance as a function of deposition time for different scan rates for electrodes
prepared using chronoamperometry (step potential = 1.35 V vs SCE) from a solution of 0.01 M MnSO4 + 0.1 M H2SO4.
Page 19 of 27
2H 2 O → O2 + 4H + + 4e-
cr us Mn 2+ + 2H 2 O → MnO2 + 4H + + 2e-
an
0.9
MnO2 deposition
M
0.6
d
Current (mA/cm2)
1.2
ip t
1.5
O2 evolution
Ac ce p
te
0.3
0.0
1.0
1.1
1.2
1.3
1.4
1.5
1.6
Potential (V vs SCE)
Figure 1. Linear sweep voltammogram of manganese dioxide electrodepsosition from an
electrolyte of 0.01 M MnSO4 + 0.1 M H2SO4 using a platinum substrate (geometric area = 0.785 cm2) and a SCE reference electrode. Scan rate = 5 mV/s. Also included (dashed line) is a fitted curve for oxygen evolution (Butler – Volmer equation), and the fraction of current contributed by oxygen evolution and manganese dioxide electrodeposition.
Page 20 of 27
2.0
Cottrell equation:
ip t cr
Fitting with A2
1.0
us
Current (mA)
1.5
an
Mn 2+ + 2H 2 O → MnO2 + 4H + + 2e-
0.0 20
40
60
80
100
120
Time (s)
Ac ce p
te
0
d
Fitting with A1
M
0.5
Figure 2. Chronoamperometric data (solid line) from the electrodeposition manganese dioxide from an electrolyte of 0.01 M MnSO4 + 0.1 M H2SO4 using a platinum substrate (geometric area = 0.785 cm2; electrochemically active surface area = 1.38 cm2) and a SCE reference electrode. Step potential = 1.35 V vs SCE. Also shown (dashed line) is the Cottrell equation modelling of the early and latter stages of deposition, using two expression with different surface area values.
Page 21 of 27
10 9 8
ip t cr
6 5
us
Mass (μg)
7
4
an
3
M
2
0 20
40
60
80
100
120
140
te
0
d
1
Ac ce p
Deposition Time (s)
Figure 3. Manganese dioxide mass as a function of electrodeposition time. Data have been corrected for the contributions made by oxygen evolution to the total current flowing.
Page 22 of 27
80
ip t
60
cr
50
us
40
30
an
Specific Surface Area (m2/g)
70
20
M
10
2
4
te
0
d
0
6
8
10
12
14
Mass (μg)
mass.
Ac ce p
Figure 4. Specific surface area of the electrodeposited manganese dioxide as a function of electrode
Page 23 of 27
cr
ip t
(a)
Ac ce p
te
(c)
d
M
an
us
(b)
Figure 5. AFM images of (a) the bare platinum substrate, and the substrate coated with electrodeposited manganese dioxide for (b) 30 s and (c) 120 s chronoamperometric electrodeposition at 1.35 V vs SCE from an electrolyte of 0.01 M MnSO4 + 0.1 M H2SO4.
Page 24 of 27
(a) 400
ip t
300
cr us
100
0
-100
an
Current (A/g)
200
M
-200
-400 0.1
0.2
Ac ce p
0.0
te
d
-300
0.3
0.4
0.5
0.6
0.7
0.8
Potential (V vs SCE)
Page 25 of 27
(b) 900 800
ip t cr
600
us
500
an
400 300
M
Specific Capacitance (F/g)
700
200
0
10
Ac ce p
0
te
d
100
20
30
40
50
60
Cycle Number
Figure 6. Sample electrochemical performance data for manganese dioxide electrodeposited using chronoamperometry from an electrolyte of 0.01 M MnSO4 + 0.1 M H2SO4, with a step potential of 1.35 V vs SCE, for a duration of 30 s. The cyclic voltammogram in (a) was conducted in 0.5 M Na2SO4 at 250 mV/s (cycle 30), while (b) shows cycle stability.
Page 26 of 27
3000
2500
ip t cr
2000
us
1500
1000
an
Specific Capacitance (F/g)
1 mV/s
250 mV/s
0 20
40
60
80
100
120
140
te
0
d
M
500
20 mV/s
Ac ce p
Deposition Time (s)
Figure 7. Specific capacitance as a function of deposition time for different scan rates for electrodes prepared using chronoamperometry (step potential = 1.35 V vs SCE) from a solution of 0.01 M MnSO4 + 0.1 M H2SO4.
Page 27 of 27
S0013-4686(13)00642-7 http://dx.doi.org/doi:10.1016/j.electacta.2013.04.007 EA 20307
To appear in:
Electrochimica Acta
Received date: Revised date: Accepted date:
3-12-2012 26-2-2013 4-4-2013
Please cite this article as: M. Dupont, A.F. Hollenkamp, S.W. Donne, Electrochemically Active Surface Area Effects on the Performance of Manganese Dioxide for Electrochemical Capacitor Applications, Electrochimica Acta (2013), http://dx.doi.org/10.1016/j.electacta.2013.04.007 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Electrochemically Active Surface Area Effects on the Performance of Manganese Dioxide for Electrochemical Capacitor Applications by
Discipline of Chemistry, University of Newcastle, Callaghan NSW, 2308, Australia 2
CSIRO Energy Technology, Box 312, Clayton South, VIC 3169, Australia
cr
1
ip t
Madeleine Dupont1, Anthony F. Hollenkamp2 and Scott W. Donne1*
us
Abstract
The specific surface area, morphology and electrochemical performance of thin films of
an
electrodeposited manganese dioxide have been examined. Electrodeposition of these films was carried out using chronoamperometry, using times ranging from 10-120 s in order to obtain deposits
M
with different masses. Using a novel approach to analysing the chronoamperometric i-t data, the specific surface area of the electrodeposited material was found to range from 13 – 67 m2/g across
d
the range of deposition times, with short deposition times leading to higher specific surface areas.
te
This has implications on the electrodeposition mechanism of manganese dioxide, which favours
Ac ce p
crystallite nucleation initially, leading to a high surface area material, followed by growth of these crystallites leading to a denser, lower surface area electrode material. This is the first time that the electrochemically active surface area of porous electrode materials has been determined. This decrease in surface area with deposition time was also reflected in the specific capacitance values of the material, which decreased slightly with increased deposition time, and hence lower surface area. Keywords: electrochemically active surface area; manganese dioxide; chronoamperometry; electrochemical capacitors * Corresponding author Ph: +61 2 4921 5477; Fax: +61 2 4921 5472; Email: [email protected]
Page 1 of 27
1. INTRODUCTION 1.1. Energy Storage Devices Increasing demand for energy has required the development of high performance energy storage devices. To be commercially viable, on whatever scale, energy storage devices need to be
ip t
able to store relatively large amounts of energy, and also have this energy readily accessible.
Chemical energy storage, in the form of a fuel, is an efficient, high grade form of energy
cr
storage, particularly when the energy is released electrochemically, in which case the resultant
us
electrical energy can be produced with efficiencies approaching 100%. Electrochemical energy storage and conversion devices include batteries, supercapacitors and fuel cells, each of which has
an
complementary performance characteristics. Supercapacitors have high power, but are limited by their low energy [1, 2]. The overall focus of our research is to develop approaches to improve the
M
specific energy density of supercapacitor materials and devices, with the intent of boosting their consumer acceptance.
d
Most commercial supercapacitors, based for example on activated carbon electrodes, and also
te
known as electrochemical capacitors, store energy in the form of charge separation at an electrical
Ac ce p
double layer [3]. Unlike electrolytic capacitors, which store charge on separated metal plates, supercapacitors store charge at the interface between an electrode and an electrolyte. Furthermore, unlike batteries, conventional supercapacitors do not undergo faradaic reactions at the electrodeelectrolyte interface, and since they notionally have no composition or phase change, they have a high degree of rechargeability and hence cyclability [4]. Supercapacitors perform significantly better than conventional electrolytic capacitors, with some supercapacitors designs demonstrating capacitance values up to 104 times higher than electrolytic capacitors [5]. Electrode materials used in supercapacitors can be either in the form of thin films or cast electrodes based on powdered materials. Thin films prepared by electrodeposition for example can be deposited easily and directly onto the required substrate and hence have low resistance and good electrical conductivity. This has resulted in thin films with extremely high specific capacitance per
Page 2 of 27
unit area, compared to powdered materials [6, 7]. Powdered materials, by comparison, are more difficult to prepare and require additives to be used as an electrode. However, they can be used in bulk quantities to develop supercapacitor devices with very large capacitances, which is one of the
ip t
limits of thin film electrodes.
1.2. Pseudocapacitance
cr
As a strategy to improve the specific energy density of supercapacitors, many researchers
us
have studied materials that exhibit psuedocapacitance. This phenomenon arises when a potential is applied to an electrode and fast, highly reversible faradaic reactions occur at the electrode surface in
an
addition to double layer charging. These processes can contribute significantly to the total capacitance of an electrode, and since they are faradaic in nature, they may also involve
M
compositional and phase changes for the electrode material [4]. Of course the specific reactions
te
the nature of the electrode material.
d
occurring, and the proportion of the total capacitance to which they contribute, vary depending upon
Ac ce p
1.3. Supercapacitor Electrode Materials
A wide variety of materials have been studied for their use as supercapacitor electrodes; e.g., carbons, metal oxides and conductive polymers. Supercapacitors often utilize high surface area electrode materials since this maximises charge storage in the double layer. Materials such as activated carbon have extremely high surface area (up to 2500 m2/g [8]) , but do not exhibit pseudocapacitance, and hence their maximum capacitance has been limited to ~400 F/g [9, 10], but most observed capacitance values are ~150 F/g. Metal oxides are another commonly used class of materials, due to both a relatively high surface area (double layer charging) and pseudocapacitance contributing to the overall capacitance. The prototypical material here is hydrated amorphous ruthenium dioxide, which has been shown to have a specific capacitance of over 800 F/g [9] in an acidic electrolyte. However, despite this
Page 3 of 27
excellent performance, its cost and toxicity have limited its widespread application. Manganese dioxide has also proven to be an excellent pseudocapacitive electrode material because, unlike ruthenium dioxide, it is inexpensive, relatively abundant, non-toxic and has been shown to exhibit capacitance of up to 2000 F/g [6].
ip t
There are many different polymorphs of manganese dioxide that have been studied as supercapacitor electrodes. Typically these materials are produced hydrothermally, via either
cr
oxidation of a Mn(II) species or reduction of Mn(VII), with the conditions used (supporting
us
electrolyte, temperature, etc.) invariably affecting the properties of the resultant material. Supercapacitor electrodes can also be prepared directly by electrodeposition of manganese dioxide,
an
which is the type of electrode studied in this work. Electrodes produced by this method are extremely thin (up to 100-200 nm), and hence exhibit very little resistance, and for these reasons are
d
1.4. Surface Area Measurements
M
desirable as electrode materials.
te
Measuring the electrochemically active surface area of any electrode material has proven to
Ac ce p
be extremely difficult even though it is one of the most basic properties of an electrified interface. To be clear, the electrochemically active surface area represents the area of the electrode material that is accessible to the electrolyte that is used for charge transfer and/or storage. One such approach for determining the electrochemically active surface area involves the use of the area specific capacitance of hydrogen ad-atoms on noble metals such as platinum [11]. While this is fundamentally very significant, the approach is only workable for these types of systems, rather than being widely applicable. Similarly, analysis of voltametric data obtained from a kinetically reversible redox couple (such as [Fe(CN)6]3-/4-) can also be used to estimate the electrochemically active surface area [12]. Again, though, this approach is limited to only inert and approximately planar electrode substrates. The situation becomes even more problematic and important when the solid electrode material is porous. For the specific case of electrodeposited manganese dioxide there
Page 4 of 27
is only one report on this in the literature. Kozawa [13] reported an estimate of the electrochemically active surface area of powdered manganese dioxide based on the ion exchange reaction that occurs between Zn2+ and H+ on the surface of the material, and assuming a certain area covered by an individual Zn2+ ion. However, this method determines the electroactive surface area
ip t
accessible to a Zn2+ ion, and so is not directly transferrable to the case under study.
There are, however, some measurements of surface area that can be obtained, such as the
cr
geometric surface area and BET (after Brunauer, Emmett and Teller) surface area. These
us
measurements can be and have been used as an approximation for the electrochemically active surface area, although with some reservation. Geometric surface area, calculated from particle size
an
and shape considerations, is a very inaccurate approximation of the electrochemically active surface area, especially in porous electrode materials (such as manganese dioxide), because it does not
M
account for the surface area contributed by pores.
The BET surface area is based on measurement of the area accessible to a gaseous adsorbate
d
(typically nitrogen or carbon dioxide), at a temperature where the adsorbate can condense on the
te
surface of the solid adsorbent material; i.e., multi-layer adsorption [14]. While accessing the porous
Ac ce p
surface area, this measurement is still expected to be different than the electrochemically active surface area, as almost always the adsorbate is different in size and chemical characteristics compared to the hydrated electrolyte ion accessing the pores in a supercapacitor electrode system. However this measurement is also only applicable to powdered samples of manganese dioxide, since the electrodeposited films prepared in our previous work do not contain enough material to allow for a measurement.
1.5. This Work Much of our previous work on supercapacitor electrodes has focussed on the electrodeposition of thin manganese dioxide films using chronoamperometry [6, 15]. In this work we will capitalize on the anomalous nature of the chronoamperometric i-t data to estimate the true
Page 5 of 27
electrochemically active surface area of manganese dioxide. This is the first such report where the electrochemically active surface area has been determined directly.
2. EXPERIMENTAL
ip t
2.1. Linear Sweep Voltammetry
Linear sweep voltammetry (LSV) was used to characterize the electrochemical oxidation of
cr
Mn2+ in an acidic environment to MnO2. To accomplish this, a previously cleaned platinum
us
working electrode (geometric area = 0.785 cm2) was placed in a solution of 0.01 M MnSO4 (≥99%; Sigma Aldrich) in 0.1 M H2SO4 in a 250 mL electrochemical cell. Cleaning of the platinum was
an
achieved by initially immersing the electrode into an acidified (0.1 M H2SO4) solution of 5% H2O2 to remove any residual manganese oxides by dissolution. This electrode was then polished using a
M
moist 1 μm Al2O3 paste on a polishing cloth. After ~2 minutes polishing, the electrode was washed thoroughly with Milli-Q ultra pure water (resistivity ρ > 18.2 MΩ.cm) before being ready for use.
d
Also placed into the electrochemical cell were a saturated calomel reference electrode (SCE; against
te
which all potentials were measured and reported), and a carbon rod (area = 3.5 cm2) as the counter
Ac ce p
electrode. The LSV experiment was conducted using a Perkin Elmer VMP 16-channel potentiostat/galvanostat. The potential was swept anodically at a rate of 5 mV/s from the open circuit potential up to 2.0 V versus SCE.
2.2. Chronoamperometry
Using the results from the LSV experiments, an appropriate potential was selected to carry out the chronoamperometry experiments. In this case, a diffusion limited potential was chosen. The MnO2 films were deposited using the same electrodes, electrochemical cell and electrolyte as the LSV experiments. The protocol used was to hold the platinum working electrode at its open circuit potential for 10 s, after which the potential was stepped to the chosen value where it was held for either 10, 20, 30, 60 or 120 s. To assess the reproducibility of the procedure, each set of
Page 6 of 27
experiments was repeated eight times, with the resultant standard deviation used to determine the error in the measurements.
2.3. Electrochemical Performance Evaluation
ip t
Once electrodeposition of the MnO2 had been carried out, the platinum electrode was removed from the electrochemical cell and then washed thoroughly with Milli-Q water to remove
cr
any entrained plating electrolyte. The electrode was then patted dry with paper towel to remove any
us
excess water. The thin film MnO2 electrode was then immersed into a 0.5 M Na2SO4 electrolyte together with the same SCE reference and carbon counter electrodes as before, and allowed to
an
equilibrate for 10 minutes. After this time the thin film MnO2 electrode was cycled between 0.0-0.8
M
V versus SCE at a range of scan rates for at least 50 cycles.
d
3. RESULTS AND DISCUSSION
te
3.1. Linear Sweep Voltammetry Data
Ac ce p
Figure 1 shows an example of the LSV data collected in this work. The data consists of a voltametric wave starting at ~1.1 V, with a maximum in current at ~1.3 V, superimposed on data for the oxygen evolution reaction. This overlap of processes was to be expected given that both anodic reactions have the same standard potential; i.e,
MnO2 + 4H+ + 2e- → Mn2+ + 2H2O
(Eo = 1.23 V)
...(1)
O2 + 4H+ + 4e- → 2H2O
(Eo = 1.23 V)
...(2)
The overlapping oxygen evolution reaction is also a complicating factor in determining the active mass of manganese dioxide deposited [15]. From this LSV data a diffusion limiting potential was chosen (1.35 V vs SCE) to carry out our chronoamperometry experiments.
Page 7 of 27
3.2. Mechanism of Manganese Dioxide Deposition While the anodic reaction in Eqn (1) above appears straightforward, the underlying mechanism of oxidation is much more complicated. In sulfuric acid manganese dioxide
ip t
electrodeposition from Mn2+ is proposed to occur via the following mechanism [16]:
Mn3+ + 2H2O → MnOOH + 3H+
us
2Mn3+ → Mn2+ + Mn4+
E0 = 1.56 V
cr
Mn2+ → Mn3+ + e-
Mn4+ + 2H2O → MnO2 + 4H+
...(3)
MnOOH → MnO2 + H+ + e-
an
Disproportionation pathway
Hydrolysis pathway
M
Here, the first step is oxidation of the solvated Mn2+ to form a soluble Mn3+ intermediate. At this point the mechanism is proposed to have two alternate pathways, the choice of which is dependent
d
on the acidity of the supporting electrolyte. In more concentrated acidic electrolytes the soluble
te
Mn3+ intermediate has a greater relative stability, meaning that it has the potential to diffuse away
Ac ce p
from the electrode surface, to the edge of the double layer, where it can undergo disproportionation to form soluble Mn2+ and Mn4+ which very quickly hydrolyzes to precipitate MnO2 on the electrode surface. In less concentrated acidic electrolytes the stability of the Mn3+ intermediate is less, in which case it can hydrolyse to MnOOH which precipitates on the electrode surface. MnOOH can then undergo solid state oxidation to MnO2. While these two pathways may seem distinct, in reality the oxidation of Mn2+ to form MnO2 occurs via a combination of the two pathways, with the acid concentration shifting the reaction preference. As a further note, the conditions of electrodeposition; i.e., electrolyte composition, temperature, anodic current density, etc., also influence the morphology of the deposit on the substrate surface.
3.3. Chronoamperometry Data
Page 8 of 27
Figure 2 shows the chronoamperometric (i-t) response for a deposition time of 2 minutes. While this data is somewhat consistent with the expected result from a chronoamperometry experiment; i.e., a current pulse upon imposition of the potential step, followed by a gradually
potential step.
ip t
decreasing current, the most significant difference is the slight bump in current at ~20 s after the
Under planar, semi-infinite diffusion limited conditions, the decay in current for an ideal
us
cr
chronoamperometry experiment is defined by the Cottrell equation [12]; i.e.,
1
(πt)
1
2
...(4)
an
i d (t) =
nFAD 2 C
M
where id(t) is the current under diffusion limited conditions (A), n is the number of electrons transferred in the redox reaction, A is the electrode area (m2), D is the diffusion coefficient (m2/s),
d
and C is the electroactive Mn2+ concentration in the bulk electrolyte (mol/m3). In this case, all of the
te
parameters in Eqn (4) are constant except for time, and hence the current decays with t-½.
Ac ce p
For the chronoamperometry data collected here, the parameters D, n and C are intrinsic to the system and so remain constant throughout the deposition. As such, the increase in current seen at ~20 s in Figure 2 can only be attributed to an increase in the surface area of the electrode. Of course, this increase in area arises from the electrodeposition of MnO2 particles onto the substrate surface, increasing the electroactive surface area onto which further MnO2 can be deposited. To model this, let us first consider the initial spike in current (t < 10 s), which we have assumed to be due to the initial oxidation of Mn2+ to MnO2 without an increase in electrode area. Since we have assumed this process does not affect the surface area of the electrode, possibly as the result of it being due to the initial stages of nucleation on the substrate surface, it can be modelled directly by the Cottrell equation (Eqn (4)), as also shown in Figure 2. To carry out this modelling, it was necessary to have an appropriate estimate of the electrochemically active surface area of the
Page 9 of 27
platinum substrate. This was determined ex-situ from the manganese-containing electrolysis cell by conducting cyclic voltammetry on the platinum substrate using an electrolyte of 10 mM Fe(CN)63in 1 M KNO3. This system is considered to be quite well behaved [17], and so using the appropriate modelling for a reversible redox couple [12], the electrochemically active surface area of the
ip t
platinum substrate was determined to be 1.38 cm2, as compared to the geometric area of 0.785 cm2 (1 cm diameter). This difference can of course be attributed to microscopic roughness of the
cr
platinum surface. So now with an estimate of the true platinum area, the Cottrell equation was
us
modelled to the chronoamperometric i-t data, with the only unknown variable being the diffusion coefficient (D). Using this approach, the value of D for Mn2+ in 0.1 M H2SO4 was determined to be
an
(1.6±0.3)×10-6 cm2/s, which is consistent with other similarly sized cations in aqueous solution [17]. Using the same basic approach, a second Cottrell equation was fitted to the t > 20 s
M
chronoamperometric data, this time using the value of the Mn2+ diffusion coefficient determined above, with the only unknown variable remaining being the surface area (A). For this fitting we
d
have again assumed that n = 2, in which case we are assuming that MnO2 is being deposited onto
te
the electrode surface, and that this is the species giving rise to the increase in electrode area. Figure
Ac ce p
2 also shows the result of a typical fitting to the chronoamperometry data for t > 20 s. The resultant electrode area determined using this approach was 2.6±0.1 cm2.
3.4. Calculating the Electrode Mass
Of course to be able to determine the specific surface area of the electrodeposited material we must also know the mass of active material on the electrode surface. Based on previous work in our laboratory, we have demonstrated that the most appropriate approach is to determine the mass from integration of the chronoamperometric i-t data. This approach has been shown to be unencumbered by entrained electrolyte, as methods such as EQCM and direct dissolution and analysis of the manganese content are. Nevertheless, integration of the i-t data is complicated by the presence of competing electrochemical reactions that can take up charge that would otherwise contribute to the
Page 10 of 27
deposition of manganese dioxide. In this case, the most important competing reaction is that of oxygen evolution, for which both the MnO2/Mn2+ and O2/H2O redox couples have the same Eo value (1.23 V vs SHE). The overlap between these two processes is very apparent in the linear sweep voltammetry data in Figure 1, with the diffusion limited wave for Mn2+ oxidation quite
ip t
clearly superimposed on the growing oxygen evolution curve. To separate these two processes we have fit a high field approximation of the Butler-Volmer equation to the oxygen evolution data, as
us
cr
also shown in Figure 1; i.e.,
...(5)
an
⎛ α (V − E)F ⎞ i a = i 0 exp⎜ a ⎟ RT ⎠ ⎝
M
where ia is the current density (A/m2), i0 the exchange current density (A/m2), αa is the anodic transfer coefficient, V is the applied potential (V), E is the equilibrium potential (V), and all other
d
symbols have their usual significance. With this modelling of the oxygen evolution reaction
te
complete, the current flowing due to Mn2+ oxidation can be extracted, as is also shown in Figure 1. With this separation of processes, the fraction of current during the chronoamperometric experiment
Ac ce p
due to the electrodeposition of manganese dioxide can be determined, which in this case for a potential step to 1.35 V, is 30.3%. As an example, from the total amount of charge passed during the 120 s electrodeposition experiment ((60±7)×10-3 C), and using this charge efficiency, the mass of manganese dioxide prepared was therefore 8.2±1.0 μg. As a further example, Figure 3 shows how the electrode mass changes with chronoamperometric deposition time, in which case there is an approximately linear change in mass with deposition time. While this linearity was not expected, especially for a changing current with time (a linear response should have been expected for a constant current), it can be justified by the fact that the changes in current between subsequent sampling points was relatively small, approaching a constant current.
Page 11 of 27
3.5. Electrochemically Active Surface Area
Following on from the previous analysis of electrode mass and apparent surface area, the specific surface area was determined to be 32±5 m2/g. This result is consistent with the BET analysis of a commercial electrolytic manganese dioxide (EMD) [18], although the conditions we
ip t
have used here to carry out our electrodeposition are considerably different to commercial practices, particularly in terms of current density used, current profile (constant current versus constant
cr
potential), electrolyte composition (0.01 M Mn2+ + 0.1 M H2SO4 versus 1.0 M MnSO4 + 0.3 M
us
H2SO4) and the temperature of electrodeposition (22°C versus 98°C). The specific surface area calculated here can also be considered to be much smaller than the BET surface area of a
an
chemically prepared manganese dioxide (CMD), which are typically 80 m2/g, although there is considerable variability here due to the wide range of experimental conditions that can be used to
M
prepared CMD. It is also important to recognize that we are also measuring a different phenomena in this case, compared to the surface area assessment carried out via a BET analysis. This is
d
primarily due to the different properties of the probing molecule; i.e., a hydrated Mn2+ ion at
Ac ce p
carried out.
te
ambient temperature in this work, compared to a N2 gas molecule at 77 K when a BET analysis is
Electrodes deposited for 2 min provided the only data for which the diffusion coefficient could be fitted reliably. For the other shorter depositions, the current profile was too short to allow accurate fitting of both processes, as described previously. However, given that the chronoamperometry experiment is conducted under the same conditions; i.e., electrolyte composition and temperature, the diffusion coefficient should be a constant for all experiments. Hence, the diffusion coefficient determined from the 2 min deposition data was used in the fittings for the 1 min and 30 s deposition times, which allowed values for the specific surface area to be calculated. For deposition times of less than 30 seconds, the amount of data that could be collected over this time period was insufficient to allow an accurate fitting of the current curve. The results of this analysis are presented in Figure 4 as a plot of the specific surface area as a function of the mass
Page 12 of 27
of deposited MnO2. The general trend appears to be a negative correlation of specific surface area with deposition mass, at least for these deposition times, i.e., t > 30 s. A mechanism by which the surface area decreases with a longer deposition time was proposed by Cross et al. [6]. In the initial stages, manganese dioxide deposition occurs predominantly by crystal nucleation on the surface of
ip t
the electrode, which increases the electrode area. As further material is deposited, crystal nucleation on the surface is replaced by the continued growth of existing crystals. This mechanism acts to
cr
reduce the surface area of the material. The results obtained from these chronoamperometry
us
experiments support this proposed mechanism.
an
3.6. Deposit Morphology
The morphology of these thin film deposits was quite difficult to evaluate since very little
M
active material on the substrate is available for characterization. Conventional methods of morphology determination, including SEM and TEM, are problematic again because little material
d
is available, and also because the material must be dried from its original state which can have an
te
impact on its morphology due to surface tension effects associated with water removal from the
Ac ce p
active manganese dioxide. Furthermore, physically scraping the active material off the substrate so that analysis by TEM can be conducted is also not ideal, nor really representative, although we have reported some success in this regard in our previous work [6]. Because of these factors we have employed atomic force microscopy (AFM) in an attempt to characterize the morphology of our thin film electrodes.
Figure 5 shows an example of the morphology of the thin film electrodeposited manganese dioxide samples prepared in this study. Specifically, this figure compares the morphology of (a) the bare platinum substrate with manganese dioxide samples electrodeposited for (b) 30 s and (c) 120 s. What is of significance here is the apparent increase in surface roughness, and hence surface area, of the substrate with manganese dioxide present. This image also provides some insight into the growth mechanism of electrodeposited manganese dioxide. By comparing the bare platinum
Page 13 of 27
substrate with the electrodeposited film it is clear that the manganese dioxide crystallites nucleate in isolated regions on the substrate, which then subsequently grow into larger crystallites, and ultimately merge
ip t
3.7. Electrochemical Performance
The electrochemical performance of each electrodeposited thin film of manganese dioxide
cr
was examined using cyclic voltammetry, in which case the potential of the electrodes was scanned
us
between 0.0 – 0.8 V vs SCE for at least 50 cycles. Figure 6(a) shows a typical cyclic voltammogram for an electrode prepared in this work, in this case for an electrode deposited for 30 s, cycled using a
an
scan rate of 250 mV/s. All of the electrodes studied exhibited the typical behaviour expected for an electrochemical capacitor; i.e., a ‘box’ shaped voltammogram over the potential window of interest.
M
Furthermore, the specific capacitance of these electrodes was essentially constant over the number of cycles considered, as shown in Figure 6(b).
d
The specific capacitance reported for these materials in Figures 6(b) and 7, is certainly much
te
higher than that reported in previous studies focussed on the use of manganese dioxide as an
Ac ce p
electrode material for electrochemical capacitors [9]. In some cases the specific capacitance that we have reported (here and in previous works [6]), is above the theoretical specific capacitance for the manganese dioxide one-electron pseudo-capacitance reaction; i.e., 1386 F/g for the reaction
MnO2 + M+ + e- ↔ MnOO(M)
...(6)
where M is a cationic species from the electrolyte, such as H+, Li+, Na+ or K+, with a potential window of 0.8 V. The origin of this enhanced specific capacitance arises from a combination of sources including pseudo-capacitance (faradaic processes accessing the bulk of the manganese dioxide structure; as above in Eqn (6)) and electrical double layer charging (non-faradaic processes involving only the manganese dioxide-electrolyte interface). While the process represented by Eqn
Page 14 of 27
(6) is expected to be the main contributor to faradaic pseudo-capacitance, it is also possible that the second electron discharge of the material; i.e., Mn(III) to Mn(II) MnOO(M) + M+ + e- ↔ Mn(OM)2
...(7)
can also contribute to the specific capacitance. In various aqueous battery systems, manganese
will enable the second electron discharge; i.e., an additional 1386 F/g.
ip t
dioxide reduction occurs typically occurs only over ~0.6 V [19], and so the 0.8 V window used here
cr
Non-faradaic processes also have the potential to contribute significantly to the total specific
us
capacitance because of the high density of charge adsorption sites on the manganese dioxide surface which have the capacity to electrostatically adsorb cations from the electrolyte. Calculations
an
conducted previously [6] demonstrated that the area specific capacitance can be estimated to be ~100 μF/cm2, which is an order of magnitude higher than the area specific capacitance of a planar
M
metal electrode [11] . It was also noted here [6] that this value was expected to be dependent on the polarity of the Mn-O bond. This could also be responsible for potentially multi-layer adsorption
d
onto the manganese dioxide surface, somewhat akin to ionic liquid ‘banding’ of anions and cations
te
on a range of substrates [20]. This process is expected to facilitate an increase in charge storage at
Ac ce p
the interface. As a final point, note that the contributions made by these various charge storage processes (faradaic and non-faradaic) depends on the nature of the material being studied, and also the cycling conditions.
Voltametric data has been used to compare the specific capacitance of the electrodeposited electrodes for different deposition times and scan rates. Figure 7 shows the change in specific capacitance with increased deposition time. For deposition times shorter than 30 s, there is no apparent correlation between the specific capacitance and deposition time. However, for those deposition experiments longer than 30 s there appears to be a decrease in the specific capacitance of the active material. According to the previous data and discussion, this may be attributed to a decrease in the surface area of the electrodes deposited for longer times, and hence a lower contribution made from electrical double layer charging. The higher capacitance observed for lower
Page 15 of 27
scan rates is due to the increased contribution of proton insertion and better utilization of the active material, in addition to any non-faradaic contributions. Conversely, at higher scan rates, we see a decrease in specific capacitance for a number of reasons. Firstly, faradaic contributions to the specific capacitance decrease because the electrode kinetics (i.e., charge transfer at the manganese
ip t
dioxide-electrolyte interface, and mass transport of the intercalated cations away from the interface towards the core of the particles) are apparently too slow to allow for as great a utilization of the
cr
active material. Secondly, for non-faradaic contributions, mass transport of the intercalated ions in
us
the electrolyte through the manganese dioxide porous structure can also become limiting [21] under high cycling rates, and as such continued non-faradaic charge storage at the interface will be
an
restricted.
A number of other literature studies have observed a decrease in capacitance as the amount of
M
MnO2 deposited is increased [22-24]. This phenomenon has been attributed to a decrease in surface area; however, none of these reports were able to confirm this with specific surface area
te
Ac ce p
specific surface area.
d
measurements. Whereas the work presented here confirms this through actual calculation of the
4. SUMMARY AND CONCLUSIONS In this work, the electrochemically active surface area of thin films of electrodeposited MnO2 has been calculated using chronoamperometry. The chronoamperometry results showed an increase in the current response during deposition, which was attributed to an increase in the surface area of the electrode. These results were used to calculate the specific surface area of the MnO2 deposit. Surface areas were calculated for deposition times 30, 60 and 120 seconds, and values between 13 – 67 m2/g were obtained, with the surface area decreasing as the mass of deposited MnO2 increased. This decrease in surface area with increase in deposition time may be attributed to the mechanism of MnO2 deposition. The proposed mechanism suggests that MnO2 is initially nucleates
Page 16 of 27
as crystallites and as further MnO2 is deposited these existing crystallites grow and begin to
5. REFERENCES
ip t
agglomerate, reducing the surface area.
Simon, P. and Y. Gogotsi, Nature Materials, 7 (2008) 845-854.
2.
Shukla, A.K., et al., Electrochimica Acta, 84 (2012) 165-173.
3.
P, K., CAPACITORS | Electrochemical Double-Layer Capacitors: Carbon Materials, in
us
cr
1.
Encyclopedia of Electrochemical Power Sources, G. Editor-in-Chief: Jürgen, Editor 2009,
4.
Conway,
B.E.,
an
Elsevier: Amsterdam. p. 634-648. Electrochemical
Supercapacitors.
Scientific
Fundamentals
and
5.
M
Technological Applications1999, New York: Kluwer Academic/Plenum Publishers.
Kurzweil, P., CAPACITORS | Electrochemical Double-Layer Capacitors, in Encyclopedia
7. 8. 9. 10. 11.
Cross, A., et al., Journal of Power Sources, 196 (2011) 7847-7853.
Ac ce p
6.
te
Amsterdam. p. 607-633.
d
of Electrochemical Power Sources, G. Editor-in-Chief: Jürgen, Editor 2009, Elsevier:
Lokhande, C.D., D.P. Dubai, and O.-S. Joo, Current Applied Physics, 11 (2011) 255-270. Kötz, R. and M. Carlen, Electrochimica Acta, 45 (2000) 2483-2498.
Naoi, K. and P. Simon, Electrochemical Society Interface, 17 (2008) 34. Alonso, A., et al., Carbon, 44 (2006) 441-446.
Bockris, J.O.M. and A.K.N. Reddy, Modern Electrochemistry. Vol. 2. 1970, New York:
Plenum Press. 12.
Bard, A.J. and L.R. Faulkner, Electrochemical Methods: Fundamentals and Applications. 2nd ed2001, New York John Wiley & Sons, Inc.
13.
Kozawa, A., Journal of the Electrochemical Society, 106 (1959) 552-556.
14.
Emmett, P.H. and S. Brunauer, J. Am. Chem. Soc., 59 (1937) 1553-64.
Page 17 of 27
15.
Cross, A.D., et al., Electrochimica Acta, 87 (2013) 133-139.
16.
Clarke, C.J., G.J. Browning, and S.W. Donne, Electrochimica Acta, 51 (2006) 5773-5784.
17.
Adams, R.N., Elecrtochemistry at Solid State Electrodes1969, New York: Marcel Dekker Inc. Williams, R.P., Characterisation and Production of High Performance Electrolytic
ip t
18.
Manganese Dioxide for use in Primary Alkaline Cells, in Chemistry1996, University of
cr
Newcastle: Callaghan.
Bailey, M.R. and S.W. Donne, Electrochimica Acta, 56 (2011) 5037-5045.
20.
Atkin, R. and G.G. Warr, The Journal of Physical Chemistry C, 111 (2007) 5162-5168.
21.
Arnott, J.B., G.J. Browning, and S.W. Donne, J. Electrochem. Soc., 153 (2006) A1332-
an
us
19.
A1340.
Babakhani, B. and D.G. Ivey, J. Power Sources, 196 (2011) 10762-10774.
23.
Broughton, J.N. and M.J. Brett, Electrochim. Acta, 50 (2005) 4814-4819.
24.
Huang, Q., X. Wang, and J. Li, Electrochim. Acta, 52 (2006) 1758-1762.
Ac ce p
te
d
M
22.
Page 18 of 27
Figure Captions Figure 1. Linear sweep voltammogram of manganese dioxide electrodepsosition from an
electrolyte of 0.01 M MnSO4 + 0.1 M H2SO4 using a platinum substrate (geometric area = 0.785 cm2) and a SCE reference electrode. Scan rate = 5 mV/s. Also included (dashed line) is a fitted
ip t
curve for oxygen evolution (Butler – Volmer equation), and the fraction of current contributed by oxygen evolution and manganese dioxide electrodeposition.
cr
Figure 2. Chronoamperometric data (solid line) from the electrodeposition manganese dioxide from
us
an electrolyte of 0.01 M MnSO4 + 0.1 M H2SO4 using a platinum substrate (geometric area = 0.785 cm2; electrochemically active surface area = 1.38 cm2) and a SCE reference electrode. Step
an
potential = 1.35 V vs SCE. Also shown (dashed line) is the Cottrell equation modelling of the early and latter stages of deposition, using two expression with different surface area values.
M
Figure 3. Manganese dioxide mass as a function of electrodeposition time. Data have been
corrected for the contributions made by oxygen evolution to the total current flowing.
d
Figure 4. Specific surface area of the electrodeposited manganese dioxide as a function of electrode
te
mass.
Ac ce p
Figure 5. AFM images of (a) the bare platinum substrate, and the substrate coated with
electrodeposited manganese dioxide for (b) 30 s and (c) 120 s chronoamperometric electrodeposition at 1.35 V vs SCE from an electrolyte of 0.01 M MnSO4 + 0.1 M H2SO4. Figure 6. Sample electrochemical performance data for manganese dioxide electrodeposited using
chronoamperometry from an electrolyte of 0.01 M MnSO4 + 0.1 M H2SO4, with a step potential of 1.35 V vs SCE, for a duration of 30 s. The cyclic voltammogram in (a) was conducted in 0.5 M Na2SO4 at 250 mV/s (cycle 30), while (b) shows cycle stability. Figure 7. Specific capacitance as a function of deposition time for different scan rates for electrodes
prepared using chronoamperometry (step potential = 1.35 V vs SCE) from a solution of 0.01 M MnSO4 + 0.1 M H2SO4.
Page 19 of 27
2H 2 O → O2 + 4H + + 4e-
cr us Mn 2+ + 2H 2 O → MnO2 + 4H + + 2e-
an
0.9
MnO2 deposition
M
0.6
d
Current (mA/cm2)
1.2
ip t
1.5
O2 evolution
Ac ce p
te
0.3
0.0
1.0
1.1
1.2
1.3
1.4
1.5
1.6
Potential (V vs SCE)
Figure 1. Linear sweep voltammogram of manganese dioxide electrodepsosition from an
electrolyte of 0.01 M MnSO4 + 0.1 M H2SO4 using a platinum substrate (geometric area = 0.785 cm2) and a SCE reference electrode. Scan rate = 5 mV/s. Also included (dashed line) is a fitted curve for oxygen evolution (Butler – Volmer equation), and the fraction of current contributed by oxygen evolution and manganese dioxide electrodeposition.
Page 20 of 27
2.0
Cottrell equation:
ip t cr
Fitting with A2
1.0
us
Current (mA)
1.5
an
Mn 2+ + 2H 2 O → MnO2 + 4H + + 2e-
0.0 20
40
60
80
100
120
Time (s)
Ac ce p
te
0
d
Fitting with A1
M
0.5
Figure 2. Chronoamperometric data (solid line) from the electrodeposition manganese dioxide from an electrolyte of 0.01 M MnSO4 + 0.1 M H2SO4 using a platinum substrate (geometric area = 0.785 cm2; electrochemically active surface area = 1.38 cm2) and a SCE reference electrode. Step potential = 1.35 V vs SCE. Also shown (dashed line) is the Cottrell equation modelling of the early and latter stages of deposition, using two expression with different surface area values.
Page 21 of 27
10 9 8
ip t cr
6 5
us
Mass (μg)
7
4
an
3
M
2
0 20
40
60
80
100
120
140
te
0
d
1
Ac ce p
Deposition Time (s)
Figure 3. Manganese dioxide mass as a function of electrodeposition time. Data have been corrected for the contributions made by oxygen evolution to the total current flowing.
Page 22 of 27
80
ip t
60
cr
50
us
40
30
an
Specific Surface Area (m2/g)
70
20
M
10
2
4
te
0
d
0
6
8
10
12
14
Mass (μg)
mass.
Ac ce p
Figure 4. Specific surface area of the electrodeposited manganese dioxide as a function of electrode
Page 23 of 27
cr
ip t
(a)
Ac ce p
te
(c)
d
M
an
us
(b)
Figure 5. AFM images of (a) the bare platinum substrate, and the substrate coated with electrodeposited manganese dioxide for (b) 30 s and (c) 120 s chronoamperometric electrodeposition at 1.35 V vs SCE from an electrolyte of 0.01 M MnSO4 + 0.1 M H2SO4.
Page 24 of 27
(a) 400
ip t
300
cr us
100
0
-100
an
Current (A/g)
200
M
-200
-400 0.1
0.2
Ac ce p
0.0
te
d
-300
0.3
0.4
0.5
0.6
0.7
0.8
Potential (V vs SCE)
Page 25 of 27
(b) 900 800
ip t cr
600
us
500
an
400 300
M
Specific Capacitance (F/g)
700
200
0
10
Ac ce p
0
te
d
100
20
30
40
50
60
Cycle Number
Figure 6. Sample electrochemical performance data for manganese dioxide electrodeposited using chronoamperometry from an electrolyte of 0.01 M MnSO4 + 0.1 M H2SO4, with a step potential of 1.35 V vs SCE, for a duration of 30 s. The cyclic voltammogram in (a) was conducted in 0.5 M Na2SO4 at 250 mV/s (cycle 30), while (b) shows cycle stability.
Page 26 of 27
3000
2500
ip t cr
2000
us
1500
1000
an
Specific Capacitance (F/g)
1 mV/s
250 mV/s
0 20
40
60
80
100
120
140
te
0
d
M
500
20 mV/s
Ac ce p
Deposition Time (s)
Figure 7. Specific capacitance as a function of deposition time for different scan rates for electrodes prepared using chronoamperometry (step potential = 1.35 V vs SCE) from a solution of 0.01 M MnSO4 + 0.1 M H2SO4.
Page 27 of 27