- Email: [email protected]
The role of carbon in the negative plate of the lead–acid battery
Journal of Energy Storage 1 (2015) 15–21
Contents lists available at ScienceDirect
Journal of Energy Storage journal homepage: www.elsevier.com/locate/est
The role of carbon in the negative plate of the lead–acid battery Abhishek Jaiswal 1, Subhas C. Chalasani * R&D Centre, Exide Industries Ltd. 217 Nazrul Islam Avenue, Kolkata, West Bengal 700059, India
A R T I C L E I N F O
A B S T R A C T
Article history: Received 20 March 2015 Received in revised form 21 April 2015 Accepted 1 May 2015
Inclusion of an appropriate form of carbon as an additive in the negative active material (NAM) improves the performance of lead–acid batteries in high rate partial state of charge (HRPSoC) cycling. We have used cyclic voltammetry to evaluate the performance of different carbons – carbon black, acetylene black and graphite – in this study and found the technique to be useful in separating the capacitive and the Faradaic contributions to the total charge. It is noted that the presence of carbon at the Pb-interface substantially enhances the electrochemical activity and increases the Faradaic charge, while high surface area carbons add to the capacitive charge. The technique is further used to study combinations of carbons as physical mixtures and bilayers. ã 2015 Published by Elsevier Ltd.
Keywords: Carbon Negative active material Lead–acid battery Cyclic voltammetry
1. Introduction Lead–acid battery technology, although over 150 years old, plays a leading role in the rechargeable battery market globally and finds wide-scale application in automotive and industrial markets. In new applications requiring high rate partial state of charge (HRPSoC) operation, such as hybrid vehicles and certain grid storage applications, the performance and life of lead–acid battery are severely limited due to negative plate sulfation. These shortcomings are overcome by the inclusion of an appropriate form of carbon as an additive in the negative plate [1–4]. This battery technology is commonly referred to as the lead–carbon battery or the carbon lead–acid battery (CLAB) and is currently the only mass produced and viable technology available for start–stop and basic micro-hybrid vehicles [5]. It is expected that the CLAB technology will play a major role in grid storage applications in the future [6,7]. Carbon additives have been experimentally observed to suppress hard sulfation on the surface of the negative plate, which has been the main failure mode of lead–acid batteries under PSoC operation [8]. Different types of carbons – carbon black, acetylene black, activated carbon and graphite – have been looked at by various research groups and have resulted in varied degrees of performance enhancement [1,2,4,7,9,10]. Much of the research literature available on these carbon additives is based primarily on
* Corresponding author. Present address: R&D, East Penn Manufacturing Co., Lyon Station, PA 19536, United States. Tel.: +1 7242527264; fax: +1 6106826371. E-mail address: [email protected] (S.C. Chalasani). 1 Present address: Department of Materials Science and Engineering, University of Maryland, College Park, MD, USA. http://dx.doi.org/10.1016/j.est.2015.05.002 2352-152X/ ã 2015 Published by Elsevier Ltd.
empirical studies showing cycle life improvements. The reaction mechanism and the key properties (physical, chemical) of carbon which benefit the negative plate are yet not properly understood and a number of possible reasons for the improved performance of the negative plate have been put forward. Moseley [11] proposed several potential mechanisms including that carbon improves electronic conductivity of negative active material (NAM) in PSoC condition, that carbon introduces double layer capacitance which acts as a buffer to the charge and the discharge cycles of the negative plate, that carbon restricts PbSO4 crystallite growth in NAM and that carbon impedes hydrogen evolution due to the presence of impurities. In addition, Pavlov et al. [12] have proposed another mechanism in which carbon promotes lead deposition electrocatalytically in NAM and improves HRPSoC performance. Recently, Pavlov and Nikolov [13] and Xiang et al. [14] have highlighted the key processes in the capacitive behavior of the carbon additives in high rate charge and discharge conditions. They have studied the effect of carbon particle size on the porous structure of NAM and the consequent effect on HRPSoC performance and they have proposed models based on double layer capacitance arising from carbons with different particle sizes. Their studies concluded that nano-sized carbon particles result in a NAM structure with pores of less than a micron size, which severely restrict ion movement resulting in poor electrode performance. On the other hand, larger carbon particles of several tens of microns produce a highly porous structure, which results in improved HRPSoC performance. The key finding from these studies is that for the double layer capacitance of carbon to be effective as a buffer in high rate charge and discharge cycles of negative plate, a highly porous NAM structure with embedded carbon particles in the NAM’s backbone is required.
16
A. Jaiswal, S.C. Chalasani / Journal of Energy Storage 1 (2015) 15–21
In the same study, Pavlov and Nikolov [13] showed that the dominating process in the electrode reaction is the double layer capacitance (non-Faradaic process) when the charge and the discharge cycles are limited to 5 s of duration. If the charge and the discharge duration are between 30 and 50 s, then the electrochemical reactions (Faradaic processes), related to lead sulfate dissolution or lead deposition, dominate which limit the life of the battery. In order to increase the HRPSoC life of a lead–acid battery, the capacitive contribution needs to be higher such that the cycling of the Faradaic processes is kept to the minimum. The capacitance of the electrode is determined not only by the form, size and loading of the carbon but more importantly by how the carbon is incorporated into the negative plate matrix. The carbon content is limited to around 2% of the NAM, largely due to the electrode processing parameters. Given this limitation on the carbon loading, the effectiveness of the resultant capacitance is limited to high rate charge and discharge cycles of about 5 s duration as shown by Pavlov and Nikolov [13]. Thus the use of CLAB is expected to be mainly limited to start–stop and micro-hybrid applications, where the key criterion is the short duration of high rate charge and discharge cycles. Following the CLAB approach, further increment in the capacitance of the negative plate would allow the lead–acid battery to go beyond micro-hybrid applications. The batteries then would be well suited for mild-hybrids where the battery is cycled deeper than in micro-hybrid applications. Increased capacitance without impacting the energy density has been achieved in the UltraBattery1 by incorporating a supercapacitor layer over the lead negative-electrode of a CLAB [15–17]. It is evident from earlier works [13,14] that an appropriate carbon should promote both the Faradaic and the capacitive processes to improve the performance of the lead–acid battery in HRPSoC operation. Understanding of the differences between the Faradaic and the capacitive processes due to the addition of carbon to the negative plate is limited to a few studies in the literature [18–20]. Pavlov et al. [18] employed cyclic voltammetry (CV) technique to demonstrate the behavior of different carbons with respect to electrocatalytic activity toward Pb2+ reduction and capacitive contribution. The usefulness of the same technique for screening various additives for the negative-electrode was shown by one of the authors [19]. In this study, we have employed CV to study the electrochemical properties of different carbons at the Pb interface in the negative plate of a lead–acid battery and have separated the Faradaic and the capacitive processes in the electrode reaction.
Fig. 1. Electrode assembly used in the study.
poured on the AGM layer and then spread uniformly by tapping the Teflon-cap. A few drops of the electrolyte were added to wet the carbon layer and to ensure a good contact with the electrode. The Teflon-holder with electrode (Pb or C) was then finally added to complete the assembly. Cyclic voltammetry studies were conducted in a three-electrode configuration with Hg/Hg2SO4 (MSE) as reference-electrode and Pt-foil as counter-electrode using Metrohm mAutolab Type III Potentiostat/Galvanostat. Scans were done at 20 mV/sec between 1.3 and 0.7 V with respect to MSE; duplicate measurements were done to verify the repeatability. The charge capacity (coulomb) using the Pb-electrode was broken down into contributions from Faradaic-Pb (Pb/PbSO4 reaction), capacitive (double layer adsorption/desorption) and Faradaic-H2 (gas evolution) as illustrated in Fig. 2. Sweep direction is indicated by arrow-heads in the Fig. 2, which is the same for all CVs presented in this study. The capacitive contribution was estimated from the CV curve using the carbon-electrode, while the FaradaicPb contribution was estimated from the area under the peak using the Pb-electrode with base line fitting. Balance in the CV curve using the Pb-electrode gave the Faradaic-H2 contribution. Physical mixtures and bilayers of two different carbon types were also studied in a 50:50 weight ratio and the performance was compared to that of individual carbons. The total weight of carbon was 10 mg in all cases. Physical mixture between carbon “X” and “Y” is identified as “X + Y”, while bilayer with carbon “X” layer
2. Materials and methods Three types of carbon were looked at in this study – carbon black (CB-1, CB-2, CB-3), acetylene black (AB) and graphite (G). Specific surface area (SSA), apparent density and tap density of the carbons were measured using Thermo Electron Sorptomatic 1990, Scott volumeter and a home-made device, respectively. Electrical resistance of compressed carbon powders was measured in a home-made die in a 2-probe configuration and the resistance was normalized against weight. Electrochemical properties of the carbons were studied in a test fixture, shown in Fig. 1, based on earlier works [12,19,20]. Two type of electrodes in rod shape were used – Pb and carbon. The Pbelectrode was cast from 99.999% pure metal, while the carbonelectrode was taken from a new dry AA pencil-cell. The electrode was encapsulated in a Teflon holder such that only the end (0.2 cm2 area) of the rod is exposed to electrolyte. All experiments were conducted in H2SO4 electrolyte with specific gravity of 1.28 g/cc. Two layers of AGM separator were soaked in the electrolyte and placed into a Teflon-cap. Carbon powder sample (5–50 mg) was
Fig. 2. Separation of Faradaic-Pb (Pb/PbSO4 reaction), capacitive (double layer adsorption/desorption) and Faradaic-H2 (gas evolution) contributions from CV curves with Pb-electrode and carbon-electrode.
A. Jaiswal, S.C. Chalasani / Journal of Energy Storage 1 (2015) 15–21
17
between Pb and carbon “Y” layer in configuration Pb/X/Y/AGM is identified as “X/Y”. The contribution of Faradaic-Pb to the total peak current was separated using base line fitting. 3. Results The physical properties (SSA, SSA equivalent size, apparent density, tap density) and the electrical resistance of the studied carbons are presented in Table 1. SSA of the carbons varied from 8.4 m2/g (G) to 133.1 m2/g (CB-3). AB showed low apparent density and tap density and it was found to be more difficult to pack, compared to the other carbon powders. In addition, the electrical resistance of AB was the highest amongst the carbons studied. Representative CV curves at the 20th cycle for Pb, Pb/CB-2 and C/CB-2 electrodes are shown in Fig. 3. The curve of the bare Pbelectrode shows the peaks corresponding to the classic Pb/PbSO4 redox couple in H2SO4. Oxidation (Pb to PbSO4) and reduction (PbSO4 to Pb) peaks represent the discharge and the charge processes at the negative-electrode of the lead–acid battery, respectively. The reduction (cathodic) peak is smaller than the oxidation (anodic) peak due to the low concentration of PbSO4 compared to that of Pb and the partial conversion of PbSO4 to Pb. The CV curve of bare Pb-electrode is smaller compared to the rest of the curves that show the typical rectangular type-shape corresponding to the capacitive behavior from CB-2. It can also be noticed that the cathodic and the anodic current peaks are broad and enhanced in the presence of CB-2. These peak positions correspond to the Pb/PbSO4 redox potentials [18]. Thus, the CV of Pb/CB-2 is a combination of the capacitance from the carbon and the enhanced activity of the Pb-electrode, suggesting that the presence of carbon on the surface of the Pb-electrode is accelerating the rate of oxidation and reduction reactions for the Pb/PbSO4 couple. This supports the hypothesis of the electrocatalytic activity of carbon on Pb as proposed by Pavlov et al. [12]. It is noticeable that the oxidation and the reductions peaks at the 200th cycle for Pb/CB-2 have increased while the capacitance part has remained almost the same. As expected, the curve for the C/CB-2 did not show any redox peaks corresponding to the Pb/PbSO4 couple. CV curves for Pb/CB-2 with varying amounts of CB-2 are presented in Fig. 4. As expected, the capacitive contribution of the carbon-electrode increased with the carbon loading. It can be seen that the CVs tilt as the carbon weight is increased. This behavior is expected as higher weight and thickness of the carbon layer can result in a substantial increase in the internal resistance to electrolyte migration, resulting in delayed current response. On the other hand, a rectangular shape of the CV curve suggests an unrestricted transport of the electrolyte in the carbon pores [21]. The capacitive contribution to the anodic and the cathodic charge capacities, derived from C/CB-2 electrode, increases proportionally with carbon weight, as shown in Fig. 5. It can be observed in Fig. 4 that the peaks corresponding to the Faradaic processes decrease as the carbon loading is increased. This can be attributed to restricted electrolyte access to the Pb-surface as the thickness of CB-2 is increased by about 5 times. It can be
Fig. 3. Cyclic voltammetry curves of 10 mg of CB-2 powder using Pb-electrode (Pb/CB-2) and carbon-electrode (C/CB-2) at 20th cycle. Curves of bare Pb-electrode at 20th cycle and of Pb/CB-2 at 200th cycle are also presented.
Fig. 4. Cyclic voltammetry curves of 10, 20 and 50 mg of CB-2 powder using Pb-electrode (Pb/CB-2) at 20th cycle.
expected that further increase in the carbon loading would prevent the electrolyte from reaching the Pb-surface and would result in the complete disappearance of the Faradic processes; this kind of behavior was observed by Pavlov et al. with Pb/AC1 electrode in their study [18]. In this case, carbon is divided into two layers: one closer to the Pb-surface with restricted access to the electrolyte and the other on the electrolyte side. The carbon layer closer to the Pb-surface is dry and acts as an electronic conductor supporting electron flow between the Pb-surface and the carbon layer wetted by the electrolyte. The Pb/carbon electrode then behaves like a capacitor without any Faradaic contribution. We further studied different grades of carbon black to evaluate the effect of carbon SSA on the voltammetry curves. The results are presented in Fig. 6, where the curve size increases as SSA is increased. The capacitive contributions to the anodic and the
Table 1 Specific surface area (SSA), SSA equivalent size, apparent density, tap density and electrical resistance of various carbons. Carbon type
SSA (m2/g)
SSA equivalent size (nm)
App. density (g/cc)
Tap density (g/cc)
Resistance (V/g)
G CB-1 AB CB-2 CB-3
8.4 29.5 96.9 108.0 133.1
170 48 15 13 11
0.21 0.12 0.03 0.18 0.18
0.62 0.44 0.07 0.38 0.39
2.5 3.7 14.9 5.0 6.5
18
A. Jaiswal, S.C. Chalasani / Journal of Energy Storage 1 (2015) 15–21
Fig. 5. Capacitive charge capacity of CB-2 powder using carbon-electrode (C/CB-2) as a function of weight.
Fig. 6. Cyclic voltammetry curves of 10 mg of CB-1, CB-2 and CB-3 powders using Pb-electrode (Pb/CB) at 20th cycle.
cathodic charge capacities, derived from C/CB electrodes, are plotted as a function of SSA in Fig. 7. The capacitive contributions of G and AB powders are lower than that of CB powders and also shown in the figure. Performance of G and AB powders is discussed later in detail. The fact that the data points for CB in Figs. 5 and 7 lie on a straight line emphasizes the accuracy of the methodology. Unlike the capacitive behavior of carbon black powders on the Pb-electrode, the Faradaic-Pb contribution does not show a correlation with SSA, suggesting that the Pb/PbSO4 oxidation and reduction reactions are largely limited to the Pb/carbon interface with H2SO4. Surface electronic conductivity is greatly
Fig. 7. Capacitive charge capacity of G, AB, CB-1, CB-2 and CB-3 powders using carbon-electrode as a function of SSA.
enhanced by the presence of carbon at the Pb-electrode and the carbon plays a key role in increasing the Faradaic current. Pavlov et al. [12,13] have proposed a parallel mechanism of charge and discharge in the presence of carbon – during discharge cycle, PbSO4 crystal deposition and growth happens on both Pb- and carbonsurfaces, while during charge cycle, carbon-surface provide a lower resistive path for Pb2+ reduction and for Pb crystal growth compared to Pb-surface. The current results are in good agreement with the parallel mechanism proposed by Pavlov et al. [12,13]. Similar CVs collected for acetylene black and graphite are shown in Figs. 8 and 9 , respectively. As observed for CB-2, Faradaic-Pb current with the carbons is higher than in the case of bare Pb-electrode and the increase is comparatively higher in the oxidation cycle than in the reduction cycle. At the 20th cycle, Faradaic-Pb peak current for CB-2, AB, G and bare-Pb during the oxidation cycle is 4.0, 3.7, 4.9 and 1.2 mA and during the reduction cycle is 0.9, 0.9, 1.1 and 0.4 mA, respectively. CV studies on the Pb/C electrode have also been done by Pavlov et al. [18] in a similar configuration; however, the amount of carbon loading in their Pb/AC1 electrode was not mentioned. It is interesting to note that the Faradiac component due to the Pb/PbSO4 redox couple was found to be missing in the CV curves in that study. It is possible that the capacitive component of the Pb/AC1 electrode is so large that the Faradaic peaks are not visible. In our experience, incomplete wetting of the Pb/C interface with the electrolyte can also result in the absence of the Faradaic-Pb peak. The charge (cathodic) capacity of the three carbons using the Pb-electrode comprising of Faradaic-Pb, capacitive and FaradaicH2 contributions at the 20th cycle is shown in Fig. 10. Compared to the bare Pb-electrode, all carbons provide significantly higher charge capacity. The cathodic charge capacity of Faradaic-Pb for CB-2, AB, G and bare-Pb during reduction cycle is 8.9, 3.5, 6.9 and 2.3 mC, respectively. In addition, CB-2 and AB provide substantial charge capacity from the capacitive processes due to adsorption/ desorption on the carbon surface. Although SSA of CB-2 and AB are similar, it is interesting to note that the shape profile and the magnitude of capacitance of the two carbons are very different; capacitive charge capacity of CB-2 (44.9 mC) is about 2 times larger than that of AB (22.6 mC). AB powder was more difficult to pack and it has 3 times higher resistance than CB-2 powder. The CV behavior indicates that the conduction path within the AB layer and at the Pb/AB interface is inhibited. In comparison, G powder does not show any capacitive behavior due to its much lower surface area but it shows significant gassing characteristics. With further cycling to 200 cycles, all three carbons (CB-2, AB, G) showed a substantial increase in the Faradaic-Pb current, both oxidation and reduction, as shown in Figs. 3, 8 and 9. This suggests
Fig. 8. Cyclic voltammetry curves of 10 mg of AB powder using Pb-electrode (Pb/AB) and carbon-electrode (C/AB) at 20th cycle. Curves of bare Pb-electrode at 20th cycle and of Pb/AB at 200th cycle are also presented.
A. Jaiswal, S.C. Chalasani / Journal of Energy Storage 1 (2015) 15–21
Fig. 9. Cyclic voltammetry curves of 10 mg of G powder using Pb-electrode (Pb/G) and carbon-electrode (C/G) at 20th cycle. Curves of bare Pb-electrode at 20th cycle and of Pb/G at 200th cycle are also presented.
that a conditioning effect on the carbon/Pb-PbSO4 interface or a Pb dissolution/precipitation reaction is coming into play, which is further enhancing the rate of the reaction with cycling. The latter process can occur in parallel on both Pb and carbon surfaces if some of the Pb-ions, dissolved during the oxidation cycle from the Pb-surface, get reduced during the reduction cycle on carbon-sites which are in close vicinity of Pb. As a result, the Pb-surface will have a net deficit of Pb after deposition. However, this deficit would not hamper the Pb oxidation on subsequent cycles, as the Pbelectrode has no shortage of elemental Pb. This process will continue until all the active carbon-sites are saturated with Pb and the diffusion of HSO4– becomes restricted in the porous matrix of the carbon. In fact, the observed peak shift suggests such diffusion limitations. The conductivity of carbon and its connectivity with the Pb-electrode surface will determine the observed increase in Faradaic-Pb current with cycling. Mixtures of different carbons in NAM have been earlier studied to improve the battery performance [4,22]. To simulate such a system, we have looked at the performance of combinations of CB-2, AB and G as physical mixtures and bilayers with total weight of 10 mg in a 50:50 ratio. Voltammetry results at the 200th cycle are presented in Fig. 11. It is observed that the capacitive behavior can be readily modified by employing a combination of carbons, not just in the magnitude but also in the shape profile. For example, combinations of G with CB-2 and AB show much higher capacitance compared to the negligible capacitance of G itself.
19
While the shape profile of combinations of CB-2 with G and AB show a flat profile compared to high slope profile of CB-2 itself. CB2 and G combinations show intermediate capacitance between individual CB-2 and G, as would be expected with 50:50 ratio. Interestingly, AB and G combinations show the capacitance of AB. The capacitance of AB (single) layer was found to be about the same at 5 mg and 10 mg, suggesting that the performance of AB is limited by low conductivity and packing density. On the other hand, CB-2 and AB combinations show higher capacitance than individual CB-2 and AB, indicating the effectiveness of high conductivity CB-2 to provide connectivity to AB. There is no marked trend in the capacitive performance of physical mixture vs. bilayer and it can be concluded that the capacitance is essentially a cumulative sum of individual components, irrespective of their physical locations. Unlike the capacitive behavior, the Faradaic-Pb contribution for the combinations is not a simple sum of components as shown in Fig. 12. We had initially hoped to combine a carbon with the highest Faradaic-Pb current with another carbon with the highest capacitive current to develop a combination with highest total current e.g., a combination of G with CB-2. One of the possible reasons for this behavior of the Faradaic current is that the Pb/ PbSO4 oxidation and reduction reactions are largely limited to the vicinity of the Pb-carbon interface, irrespective of the carbon type and combination. It is interesting to note that the physical mixtures of G with CB-2 and AB showed worse performance compared to that of the bilayers and even individual carbons. This indicates that the local order and packing of the two carbons – low SSA graphite and high SSA carbon in this case – are important factors, which have practical significance as typical paste processing will result in physical mixtures of carbons. Among the carbons studied, combinations of CB-2 and AB gave the best performance in terms of the Faradaic-Pb and total peak current. The AB/CB-2 bilayer gave the highest peak cathodic current of 4.6 mA with Faradaic-Pb contribution of 2.0 mA at the 200th cycle. Within individual carbons, CB-2 gave the highest peak cathodic current of 3.5 mA with Faradaic-Pb contribution of 1.4 mA at the 200th cycle. 4. Discussion Cyclic voltammetry was employed to study different carbons using Pb- and carbon-electrodes in order to understand the role of carbon in the negative-electrode of the lead–acid battery at high rate cycling. It is experimentally shown that the capacitive contribution can be separated from the Faradaic processes using CV. The capacitive part can be further divided into discharge and charge capacities. The latter is important in general for the lead–acid batteries and particularly for CLAB, where the charge acceptance is of critical importance. The following are the highlights of the current study: 4.1. Capacitive contribution from carbon
Fig. 10. Cathodic charge capacity using data from Figs. 3, 8 and 9 for Pb/CB-2, Pb/AB and Pb/G at 20th cycle; capacity of bare Pb-electrode is also shown.
In this study, we have examined the capacitive charge capacity for 3 different carbons. SSA of CB-2 and AB powders are similar, but the capacitive charge capacity of CB-2 was found to be about 2 times larger than that of AB. Low performance of AB is due to its poor conductivity, which makes the surface area less effective. This result suggests that high SSA of carbon does not always correspond to high capacitance unless it is supported by high conductivity. It is evident from recent publications [13,14] that the capacitive contribution from the carbon is important for prolonging the life of CLAB in HRPSoC conditions. Carbon in NAM provides capacitive charge capacity due to double layer adsorption/desorption processes, which is 10 times faster than the Faradaic current due to the Pb/PbSO4 reaction under HRPSoC conditions [18].
20
A. Jaiswal, S.C. Chalasani / Journal of Energy Storage 1 (2015) 15–21
Fig. 11. Cyclic voltammetry curves of combination of (a) G and CB-2, (b) G and AB and (c) CB-2 and AB powders using Pb-electrode at 200th cycle. Curves for individual carbon powders at 200th cycle using Pb-electrode are also presented. Total mass of carbon in each case is 10 mg.
Improved HRPSoC performance of CLAB was observed by combining conductive carbons and graphite [4,22]. Results observed in the current study suggest that the capacitive behavior can be readily modified by employing a combination of carbons, not just in the magnitude but also in the shape profile.
Faradaic contribution can be easily enhanced at a planar Pbelectrode, which was used in the current study, as the carbon content is up to an order of magnitude higher than that of NAM in CLAB, where the carbon content is typically limited to about 2 wt%. 4.3. Good contact between carbon and Pb
4.2. Faradaic contribution enhancement by carbon All three types of carbon used in the study showed enhanced Faradaic activity corresponding to the Pb/PbSO4 couple. This observation is in good agreement with the parallel mechanism proposed by Pavlov et al. [12,13] – carbon hosts the charge and the discharge of Pb. Mixtures of different carbons showed interesting results with the best performance observed with carbon black and acetylene black mixtures. However, it should be noted that the
The current results emphasize the need of a good contact between the conductive carbon and the Pb-surface, as the electrochemical reactions are largely limited to the Pb/carbon interface with H2SO4 electrolyte. This observation is in good agreement with the studies done recently [13,14], where improved HRPSoC performance is observed with large carbon particles (several tens of microns) that are well connected in the backbone structure of the NAM rather than in the energy portion of the NAM. Accessibility of the H2SO4 electrolyte to the Pb/carbon interface is also critical for high performance in HRPSoC applications and is determined by how well the carbon is dispersed in the NAM. Carbon dispersion in NAM depends not only on the type and size of carbon but also on the manufacturing conditions such as paste mixing and plate processing which varies from one battery manufacturer to another. Due to these variations, it is not surprising to see different results from different groups. 5. Conclusion
Fig. 12. Cathodic peak current of individual carbons and combinations using Pb-electrode at 200th cycle.
Cyclic voltammetry was found to be a good semi-quantitative technique to evaluate and characterize different carbons as an additive for the negative plate of the lead–acid battery, prior to time-intensive battery trials. Carbon black, acetylene black and graphite powders show different behaviors in terms of the Faradaic and the capacitive contributions. All carbons used in this study showed much higher Faradaic current, compared to the bare Pbsurface, which further increased with cycling. This indicated that
A. Jaiswal, S.C. Chalasani / Journal of Energy Storage 1 (2015) 15–21
carbon plays a key role in enhancing the electrochemical activity at the Pb-interface with the H2SO4 electrolyte and that conditioning of the interface is taking place with cycling. Mixtures of two different types of carbons can be used to generate new signature CV-profiles with different Faradaic and capacitive contributions, which may be difficult to achieve using a single carbon form. Combination of carbon black and acetylene black was found to give the best performance in terms of the Faradaic and total peak current. Further work on the carbons is in progress and will be reported in the future. Acknowledgements Authors would like to thank Mr. T.V. Ramanathan for his encouragement and support for the study and Mr. Sourav Chakraborty for his help with the experiments. References [1] D. Pavlov, et al., ALABC Project C2.3, Identification of the mechanism(s) by which certain forms of carbon, when included in the negative active material of a valve-regulated lead–acid battery exposed to high-rate partial-state-ofcharge operation, are able to resist sulfation, Final Report, 2009. [2] D.P. Boden, et al., ALABC Project N4.4, Optimization of Additives to the Negative Active Material for the Purpose of Extending the Life of VRLA Batteries in Highrate PSOC Operation, Final Report, 2005. [3] P.T. Moseley, D.A.J. Rand, ECS Trans. 41 (2012) 3–16. [4] P.S. Walmet, Sandia Report SAND 2009-5537, Evaluation of lead/carbon devices for utility applications, 2009. [5] Eurobat Publication, A Review of Batteries for Automotive Applications,
[6] [7]
[8] [9] [10]
[11] [12] [13] [14] [15] [16] [17] [18] [19]
[20] [21] [22]
21
a_review_of_batteries_for_automotive_applications_-_full_report_0pdf>. (accessed 15.02.15).. (accessed on 15.02.15). D.G. Enos, S.R. Ferreira, R. Shane, Sandia Report SAND 2011-8263, Understanding the function and performance of carbon-enhanced lead– acid batteries, 2011 M. Shiomi, T. Funato, K. Nakamura, K. Takahashi, M. Tsubota, J. Power Sources 64 (1997) 147–152. D. Pavlov, Lead–Acid Batteries: Science and Technology, first ed., Elsevier, 2011, pp. 325–343. A.F. Hollenkamp, Secondary batteries –lead–acid systems: carbon additives, in: J. Garche (Ed.), Encyclopaedia of Electrochemical Power Sources, Elsevier Science, 2009, pp. 638–647. P.T. Moseley, J. Power Sources 191 (2009) 134–138. D. Pavlov, T. Rogachev, P. Nikolov, G. Petkova, J. Power Sources 191 (2009) 58–75. D. Povlov, P. Nikolov, J. Power Sources 242 (2013) 380–399. J. Xiang, P. Ding, H. Zhang, X. Wu, J. Chen, Y. Yang, J. Power Sources 241 (2013) 150–158. L.T. Lam, R. Louey, J. Power Sources 158 (2006) 1140–1148. L.T. Lam, R. Louey, N.P. Haigh, O.V. Lim, D.G. Vella, C.G. Phyland, L.H. Vu, J. Furukawa, T. Takada, D. Monma, T. Kano, J. Power Sources 174 (2007) 16–29. A. Cooper, J. Furakawa, L. Lam, M. Kellaway, J. Power Sources 188 (2009) 642–649. D. Pavlov, P. Nikolov, T. Ragachev, J. Power Sources 195 (2010) 4444–4457. S.S. Gupta, D. Sarkar, S. Goswami, K. Adhikary, B. Chakrabarty, S. Chalasani, Electrochemical techniques to aid in the selection of appropriate additives for the negative plate in lead–acid battery, in: G. Papazov (Ed.), Proceedings of the 8th International Conference on Lead–Acid Batteries LABAT'2011, Bulgaria, 2011, pp. 273–276. B. Monahov, ECS Trans. 41 (2012) 45–69. H.Y. Liu, K.P. Wang, H.S. Teng, Carbon 43 (2005) 559–566. M. Fernandez, ALABC Project C1.2A, Supercapacitor–VRLA Spiral Wound battery –Determination of the amount of high SSA Carbon Black to be added to negative mix to implement supercapacitor effect inside the battery, 2nd Project Report, 2009.
Contents lists available at ScienceDirect
Journal of Energy Storage journal homepage: www.elsevier.com/locate/est
The role of carbon in the negative plate of the lead–acid battery Abhishek Jaiswal 1, Subhas C. Chalasani * R&D Centre, Exide Industries Ltd. 217 Nazrul Islam Avenue, Kolkata, West Bengal 700059, India
A R T I C L E I N F O
A B S T R A C T
Article history: Received 20 March 2015 Received in revised form 21 April 2015 Accepted 1 May 2015
Inclusion of an appropriate form of carbon as an additive in the negative active material (NAM) improves the performance of lead–acid batteries in high rate partial state of charge (HRPSoC) cycling. We have used cyclic voltammetry to evaluate the performance of different carbons – carbon black, acetylene black and graphite – in this study and found the technique to be useful in separating the capacitive and the Faradaic contributions to the total charge. It is noted that the presence of carbon at the Pb-interface substantially enhances the electrochemical activity and increases the Faradaic charge, while high surface area carbons add to the capacitive charge. The technique is further used to study combinations of carbons as physical mixtures and bilayers. ã 2015 Published by Elsevier Ltd.
Keywords: Carbon Negative active material Lead–acid battery Cyclic voltammetry
1. Introduction Lead–acid battery technology, although over 150 years old, plays a leading role in the rechargeable battery market globally and finds wide-scale application in automotive and industrial markets. In new applications requiring high rate partial state of charge (HRPSoC) operation, such as hybrid vehicles and certain grid storage applications, the performance and life of lead–acid battery are severely limited due to negative plate sulfation. These shortcomings are overcome by the inclusion of an appropriate form of carbon as an additive in the negative plate [1–4]. This battery technology is commonly referred to as the lead–carbon battery or the carbon lead–acid battery (CLAB) and is currently the only mass produced and viable technology available for start–stop and basic micro-hybrid vehicles [5]. It is expected that the CLAB technology will play a major role in grid storage applications in the future [6,7]. Carbon additives have been experimentally observed to suppress hard sulfation on the surface of the negative plate, which has been the main failure mode of lead–acid batteries under PSoC operation [8]. Different types of carbons – carbon black, acetylene black, activated carbon and graphite – have been looked at by various research groups and have resulted in varied degrees of performance enhancement [1,2,4,7,9,10]. Much of the research literature available on these carbon additives is based primarily on
* Corresponding author. Present address: R&D, East Penn Manufacturing Co., Lyon Station, PA 19536, United States. Tel.: +1 7242527264; fax: +1 6106826371. E-mail address: [email protected] (S.C. Chalasani). 1 Present address: Department of Materials Science and Engineering, University of Maryland, College Park, MD, USA. http://dx.doi.org/10.1016/j.est.2015.05.002 2352-152X/ ã 2015 Published by Elsevier Ltd.
empirical studies showing cycle life improvements. The reaction mechanism and the key properties (physical, chemical) of carbon which benefit the negative plate are yet not properly understood and a number of possible reasons for the improved performance of the negative plate have been put forward. Moseley [11] proposed several potential mechanisms including that carbon improves electronic conductivity of negative active material (NAM) in PSoC condition, that carbon introduces double layer capacitance which acts as a buffer to the charge and the discharge cycles of the negative plate, that carbon restricts PbSO4 crystallite growth in NAM and that carbon impedes hydrogen evolution due to the presence of impurities. In addition, Pavlov et al. [12] have proposed another mechanism in which carbon promotes lead deposition electrocatalytically in NAM and improves HRPSoC performance. Recently, Pavlov and Nikolov [13] and Xiang et al. [14] have highlighted the key processes in the capacitive behavior of the carbon additives in high rate charge and discharge conditions. They have studied the effect of carbon particle size on the porous structure of NAM and the consequent effect on HRPSoC performance and they have proposed models based on double layer capacitance arising from carbons with different particle sizes. Their studies concluded that nano-sized carbon particles result in a NAM structure with pores of less than a micron size, which severely restrict ion movement resulting in poor electrode performance. On the other hand, larger carbon particles of several tens of microns produce a highly porous structure, which results in improved HRPSoC performance. The key finding from these studies is that for the double layer capacitance of carbon to be effective as a buffer in high rate charge and discharge cycles of negative plate, a highly porous NAM structure with embedded carbon particles in the NAM’s backbone is required.
16
A. Jaiswal, S.C. Chalasani / Journal of Energy Storage 1 (2015) 15–21
In the same study, Pavlov and Nikolov [13] showed that the dominating process in the electrode reaction is the double layer capacitance (non-Faradaic process) when the charge and the discharge cycles are limited to 5 s of duration. If the charge and the discharge duration are between 30 and 50 s, then the electrochemical reactions (Faradaic processes), related to lead sulfate dissolution or lead deposition, dominate which limit the life of the battery. In order to increase the HRPSoC life of a lead–acid battery, the capacitive contribution needs to be higher such that the cycling of the Faradaic processes is kept to the minimum. The capacitance of the electrode is determined not only by the form, size and loading of the carbon but more importantly by how the carbon is incorporated into the negative plate matrix. The carbon content is limited to around 2% of the NAM, largely due to the electrode processing parameters. Given this limitation on the carbon loading, the effectiveness of the resultant capacitance is limited to high rate charge and discharge cycles of about 5 s duration as shown by Pavlov and Nikolov [13]. Thus the use of CLAB is expected to be mainly limited to start–stop and micro-hybrid applications, where the key criterion is the short duration of high rate charge and discharge cycles. Following the CLAB approach, further increment in the capacitance of the negative plate would allow the lead–acid battery to go beyond micro-hybrid applications. The batteries then would be well suited for mild-hybrids where the battery is cycled deeper than in micro-hybrid applications. Increased capacitance without impacting the energy density has been achieved in the UltraBattery1 by incorporating a supercapacitor layer over the lead negative-electrode of a CLAB [15–17]. It is evident from earlier works [13,14] that an appropriate carbon should promote both the Faradaic and the capacitive processes to improve the performance of the lead–acid battery in HRPSoC operation. Understanding of the differences between the Faradaic and the capacitive processes due to the addition of carbon to the negative plate is limited to a few studies in the literature [18–20]. Pavlov et al. [18] employed cyclic voltammetry (CV) technique to demonstrate the behavior of different carbons with respect to electrocatalytic activity toward Pb2+ reduction and capacitive contribution. The usefulness of the same technique for screening various additives for the negative-electrode was shown by one of the authors [19]. In this study, we have employed CV to study the electrochemical properties of different carbons at the Pb interface in the negative plate of a lead–acid battery and have separated the Faradaic and the capacitive processes in the electrode reaction.
Fig. 1. Electrode assembly used in the study.
poured on the AGM layer and then spread uniformly by tapping the Teflon-cap. A few drops of the electrolyte were added to wet the carbon layer and to ensure a good contact with the electrode. The Teflon-holder with electrode (Pb or C) was then finally added to complete the assembly. Cyclic voltammetry studies were conducted in a three-electrode configuration with Hg/Hg2SO4 (MSE) as reference-electrode and Pt-foil as counter-electrode using Metrohm mAutolab Type III Potentiostat/Galvanostat. Scans were done at 20 mV/sec between 1.3 and 0.7 V with respect to MSE; duplicate measurements were done to verify the repeatability. The charge capacity (coulomb) using the Pb-electrode was broken down into contributions from Faradaic-Pb (Pb/PbSO4 reaction), capacitive (double layer adsorption/desorption) and Faradaic-H2 (gas evolution) as illustrated in Fig. 2. Sweep direction is indicated by arrow-heads in the Fig. 2, which is the same for all CVs presented in this study. The capacitive contribution was estimated from the CV curve using the carbon-electrode, while the FaradaicPb contribution was estimated from the area under the peak using the Pb-electrode with base line fitting. Balance in the CV curve using the Pb-electrode gave the Faradaic-H2 contribution. Physical mixtures and bilayers of two different carbon types were also studied in a 50:50 weight ratio and the performance was compared to that of individual carbons. The total weight of carbon was 10 mg in all cases. Physical mixture between carbon “X” and “Y” is identified as “X + Y”, while bilayer with carbon “X” layer
2. Materials and methods Three types of carbon were looked at in this study – carbon black (CB-1, CB-2, CB-3), acetylene black (AB) and graphite (G). Specific surface area (SSA), apparent density and tap density of the carbons were measured using Thermo Electron Sorptomatic 1990, Scott volumeter and a home-made device, respectively. Electrical resistance of compressed carbon powders was measured in a home-made die in a 2-probe configuration and the resistance was normalized against weight. Electrochemical properties of the carbons were studied in a test fixture, shown in Fig. 1, based on earlier works [12,19,20]. Two type of electrodes in rod shape were used – Pb and carbon. The Pbelectrode was cast from 99.999% pure metal, while the carbonelectrode was taken from a new dry AA pencil-cell. The electrode was encapsulated in a Teflon holder such that only the end (0.2 cm2 area) of the rod is exposed to electrolyte. All experiments were conducted in H2SO4 electrolyte with specific gravity of 1.28 g/cc. Two layers of AGM separator were soaked in the electrolyte and placed into a Teflon-cap. Carbon powder sample (5–50 mg) was
Fig. 2. Separation of Faradaic-Pb (Pb/PbSO4 reaction), capacitive (double layer adsorption/desorption) and Faradaic-H2 (gas evolution) contributions from CV curves with Pb-electrode and carbon-electrode.
A. Jaiswal, S.C. Chalasani / Journal of Energy Storage 1 (2015) 15–21
17
between Pb and carbon “Y” layer in configuration Pb/X/Y/AGM is identified as “X/Y”. The contribution of Faradaic-Pb to the total peak current was separated using base line fitting. 3. Results The physical properties (SSA, SSA equivalent size, apparent density, tap density) and the electrical resistance of the studied carbons are presented in Table 1. SSA of the carbons varied from 8.4 m2/g (G) to 133.1 m2/g (CB-3). AB showed low apparent density and tap density and it was found to be more difficult to pack, compared to the other carbon powders. In addition, the electrical resistance of AB was the highest amongst the carbons studied. Representative CV curves at the 20th cycle for Pb, Pb/CB-2 and C/CB-2 electrodes are shown in Fig. 3. The curve of the bare Pbelectrode shows the peaks corresponding to the classic Pb/PbSO4 redox couple in H2SO4. Oxidation (Pb to PbSO4) and reduction (PbSO4 to Pb) peaks represent the discharge and the charge processes at the negative-electrode of the lead–acid battery, respectively. The reduction (cathodic) peak is smaller than the oxidation (anodic) peak due to the low concentration of PbSO4 compared to that of Pb and the partial conversion of PbSO4 to Pb. The CV curve of bare Pb-electrode is smaller compared to the rest of the curves that show the typical rectangular type-shape corresponding to the capacitive behavior from CB-2. It can also be noticed that the cathodic and the anodic current peaks are broad and enhanced in the presence of CB-2. These peak positions correspond to the Pb/PbSO4 redox potentials [18]. Thus, the CV of Pb/CB-2 is a combination of the capacitance from the carbon and the enhanced activity of the Pb-electrode, suggesting that the presence of carbon on the surface of the Pb-electrode is accelerating the rate of oxidation and reduction reactions for the Pb/PbSO4 couple. This supports the hypothesis of the electrocatalytic activity of carbon on Pb as proposed by Pavlov et al. [12]. It is noticeable that the oxidation and the reductions peaks at the 200th cycle for Pb/CB-2 have increased while the capacitance part has remained almost the same. As expected, the curve for the C/CB-2 did not show any redox peaks corresponding to the Pb/PbSO4 couple. CV curves for Pb/CB-2 with varying amounts of CB-2 are presented in Fig. 4. As expected, the capacitive contribution of the carbon-electrode increased with the carbon loading. It can be seen that the CVs tilt as the carbon weight is increased. This behavior is expected as higher weight and thickness of the carbon layer can result in a substantial increase in the internal resistance to electrolyte migration, resulting in delayed current response. On the other hand, a rectangular shape of the CV curve suggests an unrestricted transport of the electrolyte in the carbon pores [21]. The capacitive contribution to the anodic and the cathodic charge capacities, derived from C/CB-2 electrode, increases proportionally with carbon weight, as shown in Fig. 5. It can be observed in Fig. 4 that the peaks corresponding to the Faradaic processes decrease as the carbon loading is increased. This can be attributed to restricted electrolyte access to the Pb-surface as the thickness of CB-2 is increased by about 5 times. It can be
Fig. 3. Cyclic voltammetry curves of 10 mg of CB-2 powder using Pb-electrode (Pb/CB-2) and carbon-electrode (C/CB-2) at 20th cycle. Curves of bare Pb-electrode at 20th cycle and of Pb/CB-2 at 200th cycle are also presented.
Fig. 4. Cyclic voltammetry curves of 10, 20 and 50 mg of CB-2 powder using Pb-electrode (Pb/CB-2) at 20th cycle.
expected that further increase in the carbon loading would prevent the electrolyte from reaching the Pb-surface and would result in the complete disappearance of the Faradic processes; this kind of behavior was observed by Pavlov et al. with Pb/AC1 electrode in their study [18]. In this case, carbon is divided into two layers: one closer to the Pb-surface with restricted access to the electrolyte and the other on the electrolyte side. The carbon layer closer to the Pb-surface is dry and acts as an electronic conductor supporting electron flow between the Pb-surface and the carbon layer wetted by the electrolyte. The Pb/carbon electrode then behaves like a capacitor without any Faradaic contribution. We further studied different grades of carbon black to evaluate the effect of carbon SSA on the voltammetry curves. The results are presented in Fig. 6, where the curve size increases as SSA is increased. The capacitive contributions to the anodic and the
Table 1 Specific surface area (SSA), SSA equivalent size, apparent density, tap density and electrical resistance of various carbons. Carbon type
SSA (m2/g)
SSA equivalent size (nm)
App. density (g/cc)
Tap density (g/cc)
Resistance (V/g)
G CB-1 AB CB-2 CB-3
8.4 29.5 96.9 108.0 133.1
170 48 15 13 11
0.21 0.12 0.03 0.18 0.18
0.62 0.44 0.07 0.38 0.39
2.5 3.7 14.9 5.0 6.5
18
A. Jaiswal, S.C. Chalasani / Journal of Energy Storage 1 (2015) 15–21
Fig. 5. Capacitive charge capacity of CB-2 powder using carbon-electrode (C/CB-2) as a function of weight.
Fig. 6. Cyclic voltammetry curves of 10 mg of CB-1, CB-2 and CB-3 powders using Pb-electrode (Pb/CB) at 20th cycle.
cathodic charge capacities, derived from C/CB electrodes, are plotted as a function of SSA in Fig. 7. The capacitive contributions of G and AB powders are lower than that of CB powders and also shown in the figure. Performance of G and AB powders is discussed later in detail. The fact that the data points for CB in Figs. 5 and 7 lie on a straight line emphasizes the accuracy of the methodology. Unlike the capacitive behavior of carbon black powders on the Pb-electrode, the Faradaic-Pb contribution does not show a correlation with SSA, suggesting that the Pb/PbSO4 oxidation and reduction reactions are largely limited to the Pb/carbon interface with H2SO4. Surface electronic conductivity is greatly
Fig. 7. Capacitive charge capacity of G, AB, CB-1, CB-2 and CB-3 powders using carbon-electrode as a function of SSA.
enhanced by the presence of carbon at the Pb-electrode and the carbon plays a key role in increasing the Faradaic current. Pavlov et al. [12,13] have proposed a parallel mechanism of charge and discharge in the presence of carbon – during discharge cycle, PbSO4 crystal deposition and growth happens on both Pb- and carbonsurfaces, while during charge cycle, carbon-surface provide a lower resistive path for Pb2+ reduction and for Pb crystal growth compared to Pb-surface. The current results are in good agreement with the parallel mechanism proposed by Pavlov et al. [12,13]. Similar CVs collected for acetylene black and graphite are shown in Figs. 8 and 9 , respectively. As observed for CB-2, Faradaic-Pb current with the carbons is higher than in the case of bare Pb-electrode and the increase is comparatively higher in the oxidation cycle than in the reduction cycle. At the 20th cycle, Faradaic-Pb peak current for CB-2, AB, G and bare-Pb during the oxidation cycle is 4.0, 3.7, 4.9 and 1.2 mA and during the reduction cycle is 0.9, 0.9, 1.1 and 0.4 mA, respectively. CV studies on the Pb/C electrode have also been done by Pavlov et al. [18] in a similar configuration; however, the amount of carbon loading in their Pb/AC1 electrode was not mentioned. It is interesting to note that the Faradiac component due to the Pb/PbSO4 redox couple was found to be missing in the CV curves in that study. It is possible that the capacitive component of the Pb/AC1 electrode is so large that the Faradaic peaks are not visible. In our experience, incomplete wetting of the Pb/C interface with the electrolyte can also result in the absence of the Faradaic-Pb peak. The charge (cathodic) capacity of the three carbons using the Pb-electrode comprising of Faradaic-Pb, capacitive and FaradaicH2 contributions at the 20th cycle is shown in Fig. 10. Compared to the bare Pb-electrode, all carbons provide significantly higher charge capacity. The cathodic charge capacity of Faradaic-Pb for CB-2, AB, G and bare-Pb during reduction cycle is 8.9, 3.5, 6.9 and 2.3 mC, respectively. In addition, CB-2 and AB provide substantial charge capacity from the capacitive processes due to adsorption/ desorption on the carbon surface. Although SSA of CB-2 and AB are similar, it is interesting to note that the shape profile and the magnitude of capacitance of the two carbons are very different; capacitive charge capacity of CB-2 (44.9 mC) is about 2 times larger than that of AB (22.6 mC). AB powder was more difficult to pack and it has 3 times higher resistance than CB-2 powder. The CV behavior indicates that the conduction path within the AB layer and at the Pb/AB interface is inhibited. In comparison, G powder does not show any capacitive behavior due to its much lower surface area but it shows significant gassing characteristics. With further cycling to 200 cycles, all three carbons (CB-2, AB, G) showed a substantial increase in the Faradaic-Pb current, both oxidation and reduction, as shown in Figs. 3, 8 and 9. This suggests
Fig. 8. Cyclic voltammetry curves of 10 mg of AB powder using Pb-electrode (Pb/AB) and carbon-electrode (C/AB) at 20th cycle. Curves of bare Pb-electrode at 20th cycle and of Pb/AB at 200th cycle are also presented.
A. Jaiswal, S.C. Chalasani / Journal of Energy Storage 1 (2015) 15–21
Fig. 9. Cyclic voltammetry curves of 10 mg of G powder using Pb-electrode (Pb/G) and carbon-electrode (C/G) at 20th cycle. Curves of bare Pb-electrode at 20th cycle and of Pb/G at 200th cycle are also presented.
that a conditioning effect on the carbon/Pb-PbSO4 interface or a Pb dissolution/precipitation reaction is coming into play, which is further enhancing the rate of the reaction with cycling. The latter process can occur in parallel on both Pb and carbon surfaces if some of the Pb-ions, dissolved during the oxidation cycle from the Pb-surface, get reduced during the reduction cycle on carbon-sites which are in close vicinity of Pb. As a result, the Pb-surface will have a net deficit of Pb after deposition. However, this deficit would not hamper the Pb oxidation on subsequent cycles, as the Pbelectrode has no shortage of elemental Pb. This process will continue until all the active carbon-sites are saturated with Pb and the diffusion of HSO4– becomes restricted in the porous matrix of the carbon. In fact, the observed peak shift suggests such diffusion limitations. The conductivity of carbon and its connectivity with the Pb-electrode surface will determine the observed increase in Faradaic-Pb current with cycling. Mixtures of different carbons in NAM have been earlier studied to improve the battery performance [4,22]. To simulate such a system, we have looked at the performance of combinations of CB-2, AB and G as physical mixtures and bilayers with total weight of 10 mg in a 50:50 ratio. Voltammetry results at the 200th cycle are presented in Fig. 11. It is observed that the capacitive behavior can be readily modified by employing a combination of carbons, not just in the magnitude but also in the shape profile. For example, combinations of G with CB-2 and AB show much higher capacitance compared to the negligible capacitance of G itself.
19
While the shape profile of combinations of CB-2 with G and AB show a flat profile compared to high slope profile of CB-2 itself. CB2 and G combinations show intermediate capacitance between individual CB-2 and G, as would be expected with 50:50 ratio. Interestingly, AB and G combinations show the capacitance of AB. The capacitance of AB (single) layer was found to be about the same at 5 mg and 10 mg, suggesting that the performance of AB is limited by low conductivity and packing density. On the other hand, CB-2 and AB combinations show higher capacitance than individual CB-2 and AB, indicating the effectiveness of high conductivity CB-2 to provide connectivity to AB. There is no marked trend in the capacitive performance of physical mixture vs. bilayer and it can be concluded that the capacitance is essentially a cumulative sum of individual components, irrespective of their physical locations. Unlike the capacitive behavior, the Faradaic-Pb contribution for the combinations is not a simple sum of components as shown in Fig. 12. We had initially hoped to combine a carbon with the highest Faradaic-Pb current with another carbon with the highest capacitive current to develop a combination with highest total current e.g., a combination of G with CB-2. One of the possible reasons for this behavior of the Faradaic current is that the Pb/ PbSO4 oxidation and reduction reactions are largely limited to the vicinity of the Pb-carbon interface, irrespective of the carbon type and combination. It is interesting to note that the physical mixtures of G with CB-2 and AB showed worse performance compared to that of the bilayers and even individual carbons. This indicates that the local order and packing of the two carbons – low SSA graphite and high SSA carbon in this case – are important factors, which have practical significance as typical paste processing will result in physical mixtures of carbons. Among the carbons studied, combinations of CB-2 and AB gave the best performance in terms of the Faradaic-Pb and total peak current. The AB/CB-2 bilayer gave the highest peak cathodic current of 4.6 mA with Faradaic-Pb contribution of 2.0 mA at the 200th cycle. Within individual carbons, CB-2 gave the highest peak cathodic current of 3.5 mA with Faradaic-Pb contribution of 1.4 mA at the 200th cycle. 4. Discussion Cyclic voltammetry was employed to study different carbons using Pb- and carbon-electrodes in order to understand the role of carbon in the negative-electrode of the lead–acid battery at high rate cycling. It is experimentally shown that the capacitive contribution can be separated from the Faradaic processes using CV. The capacitive part can be further divided into discharge and charge capacities. The latter is important in general for the lead–acid batteries and particularly for CLAB, where the charge acceptance is of critical importance. The following are the highlights of the current study: 4.1. Capacitive contribution from carbon
Fig. 10. Cathodic charge capacity using data from Figs. 3, 8 and 9 for Pb/CB-2, Pb/AB and Pb/G at 20th cycle; capacity of bare Pb-electrode is also shown.
In this study, we have examined the capacitive charge capacity for 3 different carbons. SSA of CB-2 and AB powders are similar, but the capacitive charge capacity of CB-2 was found to be about 2 times larger than that of AB. Low performance of AB is due to its poor conductivity, which makes the surface area less effective. This result suggests that high SSA of carbon does not always correspond to high capacitance unless it is supported by high conductivity. It is evident from recent publications [13,14] that the capacitive contribution from the carbon is important for prolonging the life of CLAB in HRPSoC conditions. Carbon in NAM provides capacitive charge capacity due to double layer adsorption/desorption processes, which is 10 times faster than the Faradaic current due to the Pb/PbSO4 reaction under HRPSoC conditions [18].
20
A. Jaiswal, S.C. Chalasani / Journal of Energy Storage 1 (2015) 15–21
Fig. 11. Cyclic voltammetry curves of combination of (a) G and CB-2, (b) G and AB and (c) CB-2 and AB powders using Pb-electrode at 200th cycle. Curves for individual carbon powders at 200th cycle using Pb-electrode are also presented. Total mass of carbon in each case is 10 mg.
Improved HRPSoC performance of CLAB was observed by combining conductive carbons and graphite [4,22]. Results observed in the current study suggest that the capacitive behavior can be readily modified by employing a combination of carbons, not just in the magnitude but also in the shape profile.
Faradaic contribution can be easily enhanced at a planar Pbelectrode, which was used in the current study, as the carbon content is up to an order of magnitude higher than that of NAM in CLAB, where the carbon content is typically limited to about 2 wt%. 4.3. Good contact between carbon and Pb
4.2. Faradaic contribution enhancement by carbon All three types of carbon used in the study showed enhanced Faradaic activity corresponding to the Pb/PbSO4 couple. This observation is in good agreement with the parallel mechanism proposed by Pavlov et al. [12,13] – carbon hosts the charge and the discharge of Pb. Mixtures of different carbons showed interesting results with the best performance observed with carbon black and acetylene black mixtures. However, it should be noted that the
The current results emphasize the need of a good contact between the conductive carbon and the Pb-surface, as the electrochemical reactions are largely limited to the Pb/carbon interface with H2SO4 electrolyte. This observation is in good agreement with the studies done recently [13,14], where improved HRPSoC performance is observed with large carbon particles (several tens of microns) that are well connected in the backbone structure of the NAM rather than in the energy portion of the NAM. Accessibility of the H2SO4 electrolyte to the Pb/carbon interface is also critical for high performance in HRPSoC applications and is determined by how well the carbon is dispersed in the NAM. Carbon dispersion in NAM depends not only on the type and size of carbon but also on the manufacturing conditions such as paste mixing and plate processing which varies from one battery manufacturer to another. Due to these variations, it is not surprising to see different results from different groups. 5. Conclusion
Fig. 12. Cathodic peak current of individual carbons and combinations using Pb-electrode at 200th cycle.
Cyclic voltammetry was found to be a good semi-quantitative technique to evaluate and characterize different carbons as an additive for the negative plate of the lead–acid battery, prior to time-intensive battery trials. Carbon black, acetylene black and graphite powders show different behaviors in terms of the Faradaic and the capacitive contributions. All carbons used in this study showed much higher Faradaic current, compared to the bare Pbsurface, which further increased with cycling. This indicated that
A. Jaiswal, S.C. Chalasani / Journal of Energy Storage 1 (2015) 15–21
carbon plays a key role in enhancing the electrochemical activity at the Pb-interface with the H2SO4 electrolyte and that conditioning of the interface is taking place with cycling. Mixtures of two different types of carbons can be used to generate new signature CV-profiles with different Faradaic and capacitive contributions, which may be difficult to achieve using a single carbon form. Combination of carbon black and acetylene black was found to give the best performance in terms of the Faradaic and total peak current. Further work on the carbons is in progress and will be reported in the future. Acknowledgements Authors would like to thank Mr. T.V. Ramanathan for his encouragement and support for the study and Mr. Sourav Chakraborty for his help with the experiments. References [1] D. Pavlov, et al., ALABC Project C2.3, Identification of the mechanism(s) by which certain forms of carbon, when included in the negative active material of a valve-regulated lead–acid battery exposed to high-rate partial-state-ofcharge operation, are able to resist sulfation, Final Report, 2009. [2] D.P. Boden, et al., ALABC Project N4.4, Optimization of Additives to the Negative Active Material for the Purpose of Extending the Life of VRLA Batteries in Highrate PSOC Operation, Final Report, 2005. [3] P.T. Moseley, D.A.J. Rand, ECS Trans. 41 (2012) 3–16. [4] P.S. Walmet, Sandia Report SAND 2009-5537, Evaluation of lead/carbon devices for utility applications, 2009. [5] Eurobat Publication, A Review of Batteries for Automotive Applications,
[6] [7]
[8] [9] [10]
[11] [12] [13] [14] [15] [16] [17] [18] [19]
[20] [21] [22]
21
a_review_of_batteries_for_automotive_applications_-_full_report_0pdf>. (accessed 15.02.15).