- Email: [email protected]
Enhanced performance of starter lighting ignition type lead-acid batteries with carbon nanotubes as an additive to the active mass
Journal of Power Sources 296 (2015) 78e85
Contents lists available at ScienceDirect
Journal of Power Sources journal homepage: www.elsevier.com/locate/jpowsour
Enhanced performance of starter lighting ignition type lead-acid batteries with carbon nanotubes as an additive to the active mass Rotem Marom a, Baruch Ziv a, Anjan Banerjee a, Beni Cahana a, b, Shalom Luski a, Doron Aurbach a, * b
Dept of Chemistry, Bar Ilan University, Ramat-Gan 5290000, Israel Vulcan Automotive Ind. Ltd, Israel
h i g h l i g h t s Enhancement of the electrical conductivity of lead-acid battery electrodes by adding CNT. Homogeneous distribution throughout the electrode matrix due to well dispersed CNT. Enhanced conductivity avoids the sulfation mechanism in lead-acid batteries. Prolonged cycle-life has been achieved by minimizing the sulfation problem.
g r a p h i c a l a b s t r a c t Improving performance of lead- acid baƩeries NegaƟve Electrode
PosiƟve Electrode
About 35000 cycles
AŌer Cycling without CNT
Without CNT (reference) Voltage / V
a
Cycling @ 1.3 % DOD
About 90000 cycles NegaƟve Electrode
PosiƟve Electrode
With CNT AŌer Cycling with CNT Cycle Number
a r t i c l e i n f o
a b s t r a c t
Article history: Received 2 March 2015 Received in revised form 1 July 2015 Accepted 5 July 2015 Available online xxx
Addition of various carbon materials into lead-acid battery electrodes was studied and examined in order to enhance the power density, improve cycle life and stability of both negative and positive electrodes in lead acid batteries. High electrical-conductivity, high-aspect ratio, good mechanical properties and chemical stability of multi-wall carbon nanotubes (MWCNT, unmodified and mofified with carboxylic groups) position them as viable additives to enhance the electrodes' electrical conductivity, to mitigate the well-known sulfation failure mechanism and improve the physical integration of the electrodes. In this study, we investigated the incorporation-effect of carbon nanotubes (CNT) to the positive and the negative active materials in lead-acid battery prototypes in a configuration of flooded cells, as well as gelled cells. The cells were tested at 25% and 30% depth-of-discharge (DOD). The positive effect of the carbon nanotubes (CNT) utilization as additives to both positive and negative electrodes of lead-acid batteries was clearly demonstrated and is explained herein based on microscopic studies. © 2015 Elsevier B.V. All rights reserved.
Keywords: Lead-acid batteries Carbon nanotubes (CNT) Multi-wall carbon nanotubes (MWCNT) Pb electrodes PbO2 electrodes
1. Introduction Lead-acid batteries were invented during the 19th century [1,2]. These batteries are the most widely used rechargeable systems.
* Corresponding author. E-mail address: [email protected] (D. Aurbach). http://dx.doi.org/10.1016/j.jpowsour.2015.07.007 0378-7753/© 2015 Elsevier B.V. All rights reserved.
Even today in the 21st century, lead-acid batteries share about 50% of the rechargeable batteries market. Though lead-acid systems were studied and developed over 150 years, research continues to enhance their performance in terms of rate capability, stability, cycle life and durability [3,4]. In this study, we demonstrate modifications in flooded lead-acid start light ignition (SLI) batteries. The standard use of SLI flooded batteries is based on very shallow depth
R. Marom et al. / Journal of Power Sources 296 (2015) 78e85
of discharge. The use of SLI batteries in deeper depth-of-discharge (DOD) leads to a pronounced capacity fading and too short life. The main failure modes of SLI type batteries are sulfation and shedding [5e7]. Sulfation is related to the formation of too large crystals (or crystal agglomerates) of PbSO4. This salt is the product of both lead metal oxidation at the negative electrode and lead oxide reduction at the positive electrode. PbSO4 is not conductive and the larger crystals are not soluble during cycling. Evolution of large crystals of PbSO4 reduces the recyclable lead content, what degrades cell performance [8,9]. Due to physical, morphological and chemical changes ðPb⇔PbSO4 ⇔PbO2 Þ at the positive and negative electrodes, they are mechanically degrade, losing contact and causing short-circuits. Lead-acid battery chemistry is based on the following reactions [10]:
Negative electrode side: þ PbðsÞ þ HSO 4 ðaqÞ ⇔ PbSO4 ðsÞ þ HðaqÞ þ 2e
Positive electrode side: þ PbO2 ðsÞ þ HSO 4 ðaqÞ þ 3HðaqÞ þ 2e ⇔ PbSO4 ðsÞ þ 2H2 OðlÞ
Over all reaction: PbðsÞ þ PbO2 ðsÞ þ 2H2 SO4 ðaqÞ ⇔ 2PbSO4 ðsÞ þ 2H2 OðlÞ Other failure modes in flooded cells are related to grid corrosion and acid stratification which are not discussed in the present paper. The increasing demand from lead-acid batteries to operate in deeper depth-of-discharge namely 20e30% DOD, emphasizes further the above described problem of sulfation. Due to sulfation, SLI batteries are planned to work at a shallow DOD. Their cycle life becomes very limited when they are forced to operate at 20e30% DOD. In order to solve the sulfation problem in deep cycling of SLI batteries, carbon materials were added into the active-masses of their electrodes [5]. It is universally accepted that carbon materials as additives enhance the performance of these batteries in terms of higher charge-acceptance, which is the key metric for addressing the sulfation problem. However, controversy arises regarding the preference of carbon allotrops, which renders the best performance. Graphitic powder, carbon black, various activated carbons are already introduced by several industries and their use as additives indeed enhances performance. Apart from that, carbonbased nano materials like single-walled carbon nanotubes (SWCNT), multi-walled carbon nanotubes (MWCNT) and graphenes are envisaged as most efficient materials, because of their ordered structures and consequently high intrinsic electrical conductivity. However, exploring the use of carbon-based nanomaterials in lead-acid batteries was very limited so far. In any event, it is clear that extending battery life cycling at both low and high DOD values definitely remains one of the most important challenges for this battery technology. We report here for the first time the use of CNT in both negative and positive electrodes of lead acid batteries in order to improve the cycle life of these batteries, especially at relatively high DOD. In this publication we describe how the use of carbon nanotubes (CNT) as a component in the composite electrodes improves the performances of flooded lead-acid cells. The use of CNT in these systems was reported already by several groups [11e14]. For instance, reference [14] reports on the use of CNT called “Molecular rebar” in both positive and negative plates of practical lead acid batteries produced by a commercial company. It is clear that the CNT used in the present work (in both negative and positive electrodes), are different than those used in the studies described before. Hence, we accumulate results coming from
79
several groups, showing clearly the improvement that can be reached for this important battery technology, by adding CNT to the active mass of both anodes and cathodes of L-A battery systems. It was important to realize that the results presented herein, obtained with new types of commercial CNT, were fully coherent with previous findings (reported in references [11e14]). The presence of CNT in the electrodes improves the electrical conductivity between the active mass particles though preventing thickening and the growth of large PbSO4 particles [15]. This improvement is naturally attributed to the formation of a stable conductive active mass matrix that enables the delivery and distribution of current to all the active material homogeneously. This modified active mass consists of composites made of CNT conformably coated with lead salts. By enabling a uniform current distribution, and subsequently well distributed electrochemical redox reactions throughout the electrode matrix, arrested the formation of too large PbSO4 particles [16]. The presence of CNT is supposed to improve both the mechanical stability and electrical integrity of the electrodes [17,18] and to induce uniform changes in the active mass during the complicated conversion reactions during cycling. We expect that the CNT in the electrodes will affect the electrical integrity of the active mass and should lead to the formation and existence of small enough particles, thus facilitating effective charge-transfer interactions [19e21]. 2. Experimental Two sets of SLI type electrodes were fabricated by Vulcan Automotive Ind. using their standard processes (see a description below). The electrodes were in two sizes: 1 cm2 for laboratory cells and 150 cm2 for full plate electrodes as are used in commercial batteries. The laboratory cells' capacity varied from 60 to 100 mAh while the large, industrial electrodes had capacities in the range 13e16 Ah. Reference sets of electrodes were based on standard SLI positive and negative electrodes. Other sets of positive and negative electrodes were modified with CNT. We tested the following cells combinations: (1) positive and negative electrodes without CNT as reference, (2) Positive electrode with CNT and negative electrode without CNT, (3) Positive electrode without CNT and negative electrode with CNT, (4) Positive electrode with CNT and negative electrode with CNT. All sets were cured and aged in similar curing processes and were confirmed to be tri-basic lead sulfate (3BS). The CNT which were used during the tests were: i MWCNT from Arkema Inc., having a diameter of 10e15 nm, 1e10 microns long. ii MWCNT having diameter of 40e60 nm, about 15 microns long, purchased from Hongwu Nanometer Inc., China. The same MWCNT as in (ii), modified with carboxylic groups (eCOOH). These CNT contained carboxylic groups that where bound to their surface, obtained via oxidation process with an aqueous nitric acid solution (by the manufacturer). The concentration of the carboxylic group was estimated as 6% by weight. In general, immersing the CNT in a concentrated solution of HNO3 at elevated temperature (e.g. 80 C), leads to an obvious oxidation of their surface, thus forming surface carboxylic groups. Their concentration depends on the immersion time and temperature. Laboratory electrodes contained 0.008e0.02% CNT by weight. Large format electrodes also contained 0.008e0.02% of CNT by weight. It was found that dispersing CNT in the electrodes' active mass in this range of concentration was suitable for reaching a desirable improvement in performance. With such a low amount of CNT required, the effect on the price of the batteries is negligible.
80
R. Marom et al. / Journal of Power Sources 296 (2015) 78e85
For the large cells, we used mainly CNT type (i). The specific gravimetric density of the electrolyte solution was 1.225 g/ml. Gel type sulfuric acid-based electrolytes were also prepared by addition of 5% fumed silica (Strem Chemicals) to the aqueous sulfuric acid solution [18]. The electrodes were prepared as follows: Pastes were prepared by dry mixing of metal lead 23e27% Pb and 73e77% PbO with glass fibers (Modacrylic) 0.1% in positive electrode and 0.2% in negative electrode, 0.75% proprietary surfactant (all by weight as a first stage). Then, water is added, 9% for positive electrodes and 0.85% for negative electrodes, and the mixtures were homogenously mixed. In the next stage, 9.8% (positive electrode) and 8.9% (negative electrode) sulfuric acid solution (specific gravity of 1.325) is gradually added to the paste mixtures. The differences among the various paste mixtures included the presence of CNT (or not, for reference electrodes), the type of CNT used and the method of dispersing and mixing the CNT in pastes. The pastes thus prepared were spread onto leadeantimony (positive electrode) and leadecalcium (negative electrode) grids by a custom-made pasting device. The plates are thus formed are then being covered by a special paper (Conpast Inc. pasting paper) and then are introduced into the curing chambers where they are cured at 40 C and 90% relative humidity for 24 h, and then dried at 60 C for 24 h. Cells were assembled with these electrodes, sulfuric acid electrolyte solutions were added and the cell underwent a formation process. The formation process produces in fact the lead-acid batteries by creating lead metal negative electrodes and lead dioxide positive electrodes. This initial formation cycle is proceeds in 3 steps: Step 1: charging at 0.1C rate for 3 h. Step 2: charging at 0.14C rate for 20e24 h. Step3: charging further at 0.1C rate for 3 more hours. This procedure was developed for commercial lead acid batteries. It ensures almost completed and successful conversion of the active mass comprising lead and lead mono-oxide to spongy lead metal negative electrode and lead dioxide positive electrode. The final step was a full discharge process at a slow rate until the cells reached 1.75 V following 0.1C rate charging till the cells reached 2.4 V. Then, the cells could be continuously cycled. Cycling included discharge at 30% DOD for small cells and at 25% depth of discharge for large cells at around 0.25C rates. When the charging (at around 0.5C rates) process was completed, the cells usually reached a potential of 2.4 V. During these cycling the indication for failure is the voltage dropped below 1.75 V. Hence, the indications for failure and the end of experiments are clear. HRTEM characterization of materials was carried out with a JEOL-JEM 2100 electron microscope with LaB6 emitter operating at 200 kV. Samples for the TEM studies were prepared by scraping some material from cycled electrodes followed by sonicating the powdered samples in ethanol and adding a few drops of the resulting suspension to a copper grid which is used as a substrate for the TEM measurements. Particles physically attached to the copper grid are introduced into the microscope for imaging and related studies. 3. Results and discussion Fig. 1 shows representative cycling data of standard cells and cells which positive electrodes contained CNT. Cycling was proceeded galvanostatically with a constant capacity, related to 30% DOD of the cells. The charging voltage was fixed to 2.4 V. Discharging to 30% DOD lowered the cells' voltage to values around 2 V. When a cell fails it's voltage at the end-of-discharge drops to low values. Failure of cells in this work was determined when the discharge voltage dropped below 1.75 V. Typical voltages at the end-of-discharge of modified and reference cells are plotted as a function of cycle number. The improvement by adding CNT to the positive electrodes in these cells is spectacular. Fig. 1B indicates 2
Fig. 1. (A) Typical data (voltage at the end of discharge and end of charge vs. cycle number) from a cycling experiment, in which repeated voltage profiles are recorded upon cycling around 30% of the total capacity at around C/4 rates. The cells are always charges at around C/2 rates up to 2.4 V (what bring them back to their full capacity). The indicator in these experiments is the voltages recorded at the end of discharge. The failure is clearly indicated in this chart as a gradual decrease of the voltage to values below 1.75 V, and (B) Typical cycling data (constant capacity, corresponding to 30% DOD) from lead-acid cells containing positive electrodes with and without CNT. The voltage at the end of discharge at each cycle is plotted as a function of cycle number. Failure (and the end of measurements) is indicated by the sharp drop in potentials (below 1.75 V as seen in the figure). In the chart, data from 3 cells are compared as indicated: a standard cells and cells with positive electrodes contain CNT, liquid and gel electrolyte solutions (as marked). The negative electrodes' active mass was in excess.
and 3 fold improvement in cycle life for cells with positive electrodes containing CNT based on liquid and gel electrolyte solutions, respectively. Standard cells with gel electrolyte solutions cycled at the same conditions, usually exhibited a few tens of cycles beyond those obtained with regular electrodes. In the experiments represented by Fig. 1, in which we concentrated on the positive electrode side, the active mass of the negative electrodes was in a large excess. As a next step, cells with positive electrode containing modified CNT to which surface carboxylic groups were bonded. We conducted similar sets of experiments, as those related to Fig. 1 (negative electrodes' active mass in large excess, cycling at 30% DOD of positive electrodes' capacity etc.). In these experiments only cells with liquid electrolyte solutions were tested. The beneficial effect of CNT with carboxylic groups bound to their surface is demonstrated in Fig. 2, which is similarly presented as Fig. 1B. More than 3 fold improvement is demonstrated for cells which positive electrodes included CNT which contained carboxylic groups. The performance of cells with positive electrodes containing CNT which included carboxylic groups, in flooded cells is superior (>750 cycles) compared to cells containing positive electrodes with unmodified CNT, which demonstrated 450 cycles (See Fig. 1). The improvement due to the CNT surface modification can be explained by a better wetting of the active mass which contains hydrophilic CNT. The presence of the carboxylic group may also improve the adhesive interactions between the modified CNT and the lead oxide in the positive electrodes' active mass. The carboxylic groups on the
R. Marom et al. / Journal of Power Sources 296 (2015) 78e85
Fig. 2. Typical cycling data (constant capacity, corresponding to 25% DOD) from leadacid cells containing positive electrodes with and without CNT. The voltage at the end of discharge of each cycle is plotted as a function of cycle number. Failure (and the end of measurements) is indicated by the sharp drop in voltages (below 1.75 V as seen in the figure). In the chart, data from two cells are compared as indicated: a standard cell (red) and a cell which positive electrode contains CNT which had carboxylic groups bound to their surface (blue), liquid electrolyte solutions. The negative electrodes mass was in excess. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
CNT surface may also stabilize the active mass upon cycling by chelation of Pb2þ ions [22], thus preventing growth of large PbSO4 crystals. The effect of CNT as additives to the negative electrodes in these systems was studied as well. CNT from Arkema and Hongwu were added to negative electrodes. In the cells assembled for these experiments the active mass of the positive electrodes was of course in large excess. The cells' cycling was limited to 30% DOD of the negative electrodes. In general, the negative electrode limited cells showed lower cycle life than the positive electrode limited cells. However, the cells with negative electrodes containing CNT also performed better than reference cells with no CNT. Fig. 3 compares the representative performance of cells in which both positive and negative electrodes contained CNT to that of similar CNT free reference cells. As presented in this figure, cells with CNT from the two sources we tested were compared. The cells which electrodes contained 15 micron long CNT demonstrated in average shorter cycle life compared to cells which electrodes contained CNT with length varied from 1 to 10 microns. Understanding the effect of the CNT length requires much more studies (on-going), beyond the scope of the present paper. Our first indication connects the difference in performance to the
Fig. 3. Typical cycling data (constant capacity, corresponding to 25% DOD) from leadacid cells containing negative and positive electrode with and without CNT. The voltage at the end of discharge of each cycle is plotted as a function of cycle number. Failure (and the end of measurements) is indicated by the sharp drop in voltages (below 1.7 V as seen in the figure). In the chart, data from 3 typical cells are compared as indicated: a standard cells and cells which electrodes contained two types of CNT as indicated: short (1e10 microns from Arkema) and long (15 microns from Hongwu Nanometer, China) as marked.
81
effectiveness of mixing and dispersion, which is better achieved with the shorter CNT. The beneficial effect of CNT for large, practical industrially made electrodes in real lead acid batteries was confirmed at Vulcan Batteries Ltd. Vulcan Batteries Ltd., used CNT from Arkema in electrodes for SLI lead acid batteries. The suitable amount of CNT in the large size electrodes was in the range of 0.008%e0.02% by weight. The CNT were introduced into the active mass by a unique patented process. Table 1 summarizes results of tests with large size electrodes containing CNT in the positive and negative electrodes (vs. reference experiments). In all of these tests, CNT from Arkema were examined. The tests included reference batteries (no CNT), batteries with either negative or positive electrodes containing CNT and batteries which both electrodes contained CNT. The batteries included single cells at capacity of 13e16 Ah which were cycled to 25% DOD. The indication for failure was the drastic drop of the endof-discharge voltage below 1.75 V. The results reported, in Table 1 confirmed well the above described laboratory tests. A clear improvement in specific capacity and cycle life was found in all cases were the electrodes (any or both) contained CNT. The improvement recorded has definitely important practical and commercial consequences. The most important finding from these experiments, as reflected from the data presented in the table was that addition of CNT into both positive and negative electrodes doubled the cycle life of practical SLI type lead-acid batteries. It should be noted that this result is especially important in light of the operation at 25% DOD, which can be considered as high, for this type of lead-acid batteries. It is important to note that there are commercial lead-acid batteries which are designed to operation at high DOD and very prolonged cycle life (thousands). However, these systems require special structure, geometry and complicated electrodes compositions. We assume that judicious and optimized use of CNT in all types of lead-acid batteries is expected to improve their performance beyond the state-of-the-art. In order to understand the beneficial effect of the presence of CNT in the electrodes, we conducted postmortem analyses of selected, representative cells after their end of life. Fig. 4 presents cycle life profiles of the cells on which teardown analysis was performed. The Figure shows the response of a cell of electrodes without CNT (red line) and the response of a cell, which electrodes contained CNT from Arkema (blue line). The cells capacity was 94.8 Ah/kg and they were cycled at 25% DOD. As usual, the drop of discharge potential below 1.75 V marked their end of life. In both figures the voltage at the end-of-discharge is plotted vs. number of cycles. These figures reflect the advantage of the presence of CNT, dispersed in the electrodes' active mass, in terms of increasing the cycle life of these batteries. Electrodes from the batteries which cycling data are presented in Fig. 4, were examined by scanning electron microscopy (SEM). Representative images are presented below. Fig. 5(AeD) show typical SEM images of electrodes have taken from lead-acid batteries after their failure. These figures comparing the morphology of electrodes from reference cells which did not contain CNT, with electrodes containing CNT that underwent the same cycling experiments as the reference systems (but failed after many more cycles, see examples in Table 1). The main features in these images are the size variation of PbSO4 crystals, the product of lead negative electrode's reduction and lead oxide positive electrode's oxidation. Their irreversible accumulation during cycling and growth to larger dimensions that avoid their accessibility into the electrochemical processes, are considered as the major failure mechanism of electrodes upon cycling lead-acid batteries. The life extension of
82
R. Marom et al. / Journal of Power Sources 296 (2015) 78e85
Table 1 Performances of industrial SLI type batteries: positive & negative electrodes containing 0.008e0.02% by weight CNT from Arkema as indicated. Cells type 1 (reference) are based on standard electrodes in sulfuric acid (specific gravity: 1.215). Cells type 2e6 included positive electrode modified with CNT. Cells type 7e8 included both positive and negative electrodes modified with CNT. The cells were discharged to DOD of 25% and the indication for end of life was a drop of the end-of-discharge potential to 1.7 V. Cell CNT source
CNT length (microns)
Capacity Measured Ah/kg
% Of effective capacity over 115 Ah/kg (115 ¼ 100%)
1 2 3 4 5 6 7 8
0 1e10 1e10 1e10 1e10 1e10 1e10 1e10
94.8 128 126.5 124.4 131.8 128.7 132.2 136.5
82 112 110 108 114.5 112 115 119
Reference Arkema Arkema Arkema Arkema Arkema Arkema Arkema
Number of cycles to end of life Enhanced cycle life compared to the (1.7 V) reference (%)
170 257 276 298 244 276 335 387
e 51.1 62.3 75.3 43.0 62.3 97 127.6
Fig. 4. Plots of the voltage of end of discharge vs. cycle number, for a reference flooded lead-acid battery (single cell) 13.6 Ah (red line), and a similar battery which electrodes contained CNT from Arkema (see details in the Experimental section) (blue line), that were cycled at 25% DOD. The electrodes were manufactured by a fully industrial process. The end of life for teardown analysis was the drop of potential below 1.75 V, as shown here. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
electrodes due to addition of CNT to their active mass described above, correlate very well to the morphological studies. The electrodes containing CNT develop much smaller crystals of PbSO4 because the presence of CNT facilitates their reversible precipitationedissolution during cycling, by maintaining a good electrical integration of the active mass. It was important to detect by appropriate imaging, the presence of CNT in the active mass. Only a relatively small amount of CNT, dispersed initially in the lead oxides slurry from which the electrodes are prepared, was needed, in order to obtain their pronounced positive effect on the performance. Hence, finding good microscopic images of CNT embedded in the active mass of the electrodes, especially after cycling, was not trivial. Nevertheless, it was possible to obtain reasonably good images of CNT embedded in the electrodes' active mass, by TEM measurements of particles scraped from pristine and cycled electrodes. The dispersion of the CNT in the electrodes is very uniform, thereby, any particle scraped from the CNT containing electrodes, is supposed to contain CNT that can be detected by a rigorous TEM analysis. Fig. 6(a) shows a typical bright field TEM image showing a single CNT embedded in the active mass of a cycled cathode. In Fig. 6(b), we show HRTEM image of such a CNT, in which its fine morphology (the walls) is well seen. Fig. 6(d) shows the intensity profile taken
from the area marked by white rectangular in Fig. 6(c) showing the measurements for 10 carbon layers at the CNT walls. These high resolution measurements of a single CNT embedded in the active mass are in a very good agreement with literature data reporting that the d-spacing in multi-walled CNT is around 0.34 nm [23]. From the studies described above, it is clear that the presence of uniformly dispersed CNT in the active mass of negative and positive electrodes of lead acid batteries extend their cycle life. A-priori, it was expected that their presence should enable a better electrical integrity of the active mass, thus facilitating more effective conversion reactions. From a previous work [16] it was known that CNT remain robust in positive electrodes, despite their relatively high upper voltage. The morphological studies described above connect the electrochemical improvement in performance to the formation of small PbSO4 particles upon discharge. Hence, these studies converge well to the following conclusion: We suggest that the mechanism in which the presence of CNT in the electrodes active mass improves performance, is via the formation of networks (mostly local, well distributed) by CNT that provide a better uniform current-distribution within the active mass. Consequently, small product particles are formed and their electrochemical accessibility is good enough to enable improvement in the electrodes cycle life. There is an open question about the interactions of
R. Marom et al. / Journal of Power Sources 296 (2015) 78e85
83
Fig. 5. Representative SEM images (<10 microns resolution) of electrodes have taken from the cells that underwent cycling until they failed. (A) Negative electrode with CNT, (B) negative electrode without CNT, (C) positive electrode with CNT and (D) positive electrode without CNT.
Fig. 6. (a) A typical bright field image showing a CNT embedded in the active mass of a cycled cathode (b) HRTEM image of such a CNT (12 nm thick) showing well resolved lattice fringes of the CNT's walls; (c) HRTEM image of a similar CNT displaying the lattice-fringes of the CNT; (d) Intensity profile taken from the area marked by a white rectangular in (c) showing the measurements for the 10 lines. These measurements are in very good agreement to the literature data of 0.34 nm for the d-spacing of CNT [21].
the CNT with the various moieties existing in the electrodes: electrically conductive metallic lead and lead oxides, nonconductive lead sulfate. The microscopic studies cannot provide insights about the intimate interactions between the CNT and electrodes' active mass components. However, from the clearly visible morphological effect, we conclude that the CNT are bound to all these moieties, during their reactions and transformations. Due to the way in which CNT are mixed with the active mass, they are not embedded within the particles, but rather they are well dispersed, maintaining a very good physical contact with most particles of the active mass. As a final stage of these studies, it was important to demonstrate their practical aspects, by measuring full SLI battery prototypes which both electrodes contain CNT. Fig. 7 shows a significant improvement in practical automotive batteries upon cycling at 1.3% DOD. Adding Arkema's CNT to the paste from which the electrodes of these practical batteries are prepared, led to a dramatic improvement in the cycle life of the batteries. In this figure, typical voltageecycles curves of reference (A) and CNT containing battery (B) are compared. As presented
herein, an extension of nearly 3-fold in the batteries cycle life could be obtained due to the presence of CNT. 4. Conclusions We have demonstrated that addition of CNT produced by commercial companies to both lead negative electrodes and lead dioxide positive electrodes in lead acid batteries pronouncedly improve the cells performance and doubling the cycle life of practical batteries. The introduction of the CNT is done in the early stage of electrodes preparation, mixed with the lead precursors in laboratory cells. Hence, formation leads to a uniform integration of the CNT in both the lead and the lead dioxide particulate matrices of the electrodes, at their final pristine form (before cycling) and during the periodic chargeedischarge cycling. We also demonstrate in small scale experiments that using modified CNT with carboxylic groups can lead to further improvement in performance. Cells comprising electrodes containing MWCNT functionalized with carboxylic groups exhibited
84
R. Marom et al. / Journal of Power Sources 296 (2015) 78e85
Fig. 7. Typical data of full 12.6 V batteries (voltage at the end of discharge and end of charge vs. cycle number) from cycling experiments, in which repeated voltage profiles are recorded upon cycling at 1.3% DOD at HRPSOC (based on SBA protocol). (a) Shows the response of a standard battery without CNT that could performed 37,000 cycles until failure. (b) Shows the response of a standard battery with Arkema's CNT that performed more than 89,000 cycles until failure.
longer cycle life than cells which electrodes contained usual CNT from commercial sources. Addition of MWCNT into the active mass of industrial manufactured electrodes (both negative and positive electrode) clearly extended the cycle life of SLI type flooded batteries: standard cells with commercial electrodes demonstrated in average 170 cycles (25% DOD) while cells with CNT modified electrodes exhibited averagely 360 cycles at 25% DOD. The improvement observed is well correlated with teardrop analysis of cycled cells that showed cycled electrodes containing CNT develop much smaller crystals of lead sulfate compared to cycled (CNT free) reference electrodes. Hence, it is clear that the presence of CNT enable much better accessibility of the active mass to the electron transport processes during cycling, due to their physical properties (high aspect-ratio, thin, flexible, robust and stable electronically conducting agents). Initial studies demonstrated that CNT derivatized with hydrophilic functional groups such as carboxylic groups can serve as better additives to lead-acid batteries electrodes than regular CNT, probably due to a better adhesion to their active mass. We demonstrated a significant improvement in practical SLI automotive batteries upon cycling at 1.3% DOD at HRPSOC (based on SBA protocol, as usual and acceptable for these practical battery systems), due to the presence of CNT in the electrodes. Excellent and highly reliable results were obtained with CNT from Arkema. It is highly important that improving the performance of LA batteries
could be achieved with commercial CNT. Commercial battery prototypes containing CNT dispersed in their electrodes, demonstrated and improvement of more than 100% in their cycle life, compared to reference batteries, prepared identically but without CNT. As mentioned in the introduction section, there are previous reports on the improvement of lead-acid batteries achieved by adding CNT to their electrodes. Accumulation of such positive coherent results coming from different groups substantiates their validity. Acknowledgment A partial support for this work was obtained from the ISF, Israel Science Foundation, in the framework of the INREP project and from the Chief Scientist of Israel Ministry of Economics. References , C.R. Acad. Sci. Paris 50 (1860) 640e642. [1] G. Plante [2] D.A.J. Rand, Valve-regulates Lead-acid Batteries, Elsevier, Amsterdam; Boston, 2004. [3] B. Culpin, D.A.J. Rand, J. Power Sources 36 (1991) 415e438. [4] R.D. Prengaman, J. Power Sources 95 (2011) 224e233. [5] D. Pavlov, Lead-acid Batteries: Science and Technology, Elsevier, Amsterdam, 2011. [6] P.T. Moseley, J. Power Sources 191 (2009) 134e138. [7] A.H. Catherino, F.F. Feres, F. Trinidad, J. Power Sources 129 (2004) 113e120. [8] D. Pavlov, P. Nikolov, ECS Trans. 41 (2012) 71e82. [9] P. Ruetschi, J. Power Sources 2 (1977) 3e24.
R. Marom et al. / Journal of Power Sources 296 (2015) 78e85 [10] D. Linden, T.B. Reddy, Handbook of Batteries, third ed., McGraw-Hill, NewYork, 2002, p. 23. [11] Y. Yin, C. Liu, S. Fan, RSC Adv. 4 (2014) 26378e26382. [12] M. Saravanan, P. Sennu, M. Ganesan, S. Ambalavanan, J. Electrochem. Soc. 160 (2013) A70eA76. [13] S.W. Swogger, P. Everill, D.P. Dubey, N. Sugumaran, J. Power Sources 261 (2014) 55e63. [14] N. Sugumaran, P. Everill, S.W. Swogger, D.P. Dubey, J. Power Sources 279 (2015) 281e293. [15] L. Vaisma, H.D. Wagner, G. Marom, Adv. Colloid Interface Sci. 128e130 (2006) 37e46.
85
[16] R. Shapira, G.D. Nessim, T. Zimrin, D. Aurbach, Energy Environ. Sci. 6 (2013) 587e594. [17] S. Logeshkumar, R. Manoharan, Electrochim. Acta 144 (2014) 147e153. [18] D. Pavlov, P. Nikolov, J. Power Sources 242 (2013) 380e399. [19] S. Takenaka, Carbon 47 (2009) 1251e1257. [20] S.S. Misr, J. Power Sources 168 (1) (2006) 40e48. [21] S.M. Kumar, S. Ambalavanan, S. Mayavan, RSC Adv. 4 (2014) 36517e36521. [22] X. Qin, L. Ni-Na, Z. Jun-Jie, Chin. J. Chem. 23 (2005) 1510e1514. [23] G. Maurin, I. Stepanek, P. Bernier, J.F. Colomer, J.B. Nagy, F. Henn, Carbon 39 (2001) 1273e1278.
Contents lists available at ScienceDirect
Journal of Power Sources journal homepage: www.elsevier.com/locate/jpowsour
Enhanced performance of starter lighting ignition type lead-acid batteries with carbon nanotubes as an additive to the active mass Rotem Marom a, Baruch Ziv a, Anjan Banerjee a, Beni Cahana a, b, Shalom Luski a, Doron Aurbach a, * b
Dept of Chemistry, Bar Ilan University, Ramat-Gan 5290000, Israel Vulcan Automotive Ind. Ltd, Israel
h i g h l i g h t s Enhancement of the electrical conductivity of lead-acid battery electrodes by adding CNT. Homogeneous distribution throughout the electrode matrix due to well dispersed CNT. Enhanced conductivity avoids the sulfation mechanism in lead-acid batteries. Prolonged cycle-life has been achieved by minimizing the sulfation problem.
g r a p h i c a l a b s t r a c t Improving performance of lead- acid baƩeries NegaƟve Electrode
PosiƟve Electrode
About 35000 cycles
AŌer Cycling without CNT
Without CNT (reference) Voltage / V
a
Cycling @ 1.3 % DOD
About 90000 cycles NegaƟve Electrode
PosiƟve Electrode
With CNT AŌer Cycling with CNT Cycle Number
a r t i c l e i n f o
a b s t r a c t
Article history: Received 2 March 2015 Received in revised form 1 July 2015 Accepted 5 July 2015 Available online xxx
Addition of various carbon materials into lead-acid battery electrodes was studied and examined in order to enhance the power density, improve cycle life and stability of both negative and positive electrodes in lead acid batteries. High electrical-conductivity, high-aspect ratio, good mechanical properties and chemical stability of multi-wall carbon nanotubes (MWCNT, unmodified and mofified with carboxylic groups) position them as viable additives to enhance the electrodes' electrical conductivity, to mitigate the well-known sulfation failure mechanism and improve the physical integration of the electrodes. In this study, we investigated the incorporation-effect of carbon nanotubes (CNT) to the positive and the negative active materials in lead-acid battery prototypes in a configuration of flooded cells, as well as gelled cells. The cells were tested at 25% and 30% depth-of-discharge (DOD). The positive effect of the carbon nanotubes (CNT) utilization as additives to both positive and negative electrodes of lead-acid batteries was clearly demonstrated and is explained herein based on microscopic studies. © 2015 Elsevier B.V. All rights reserved.
Keywords: Lead-acid batteries Carbon nanotubes (CNT) Multi-wall carbon nanotubes (MWCNT) Pb electrodes PbO2 electrodes
1. Introduction Lead-acid batteries were invented during the 19th century [1,2]. These batteries are the most widely used rechargeable systems.
* Corresponding author. E-mail address: [email protected] (D. Aurbach). http://dx.doi.org/10.1016/j.jpowsour.2015.07.007 0378-7753/© 2015 Elsevier B.V. All rights reserved.
Even today in the 21st century, lead-acid batteries share about 50% of the rechargeable batteries market. Though lead-acid systems were studied and developed over 150 years, research continues to enhance their performance in terms of rate capability, stability, cycle life and durability [3,4]. In this study, we demonstrate modifications in flooded lead-acid start light ignition (SLI) batteries. The standard use of SLI flooded batteries is based on very shallow depth
R. Marom et al. / Journal of Power Sources 296 (2015) 78e85
of discharge. The use of SLI batteries in deeper depth-of-discharge (DOD) leads to a pronounced capacity fading and too short life. The main failure modes of SLI type batteries are sulfation and shedding [5e7]. Sulfation is related to the formation of too large crystals (or crystal agglomerates) of PbSO4. This salt is the product of both lead metal oxidation at the negative electrode and lead oxide reduction at the positive electrode. PbSO4 is not conductive and the larger crystals are not soluble during cycling. Evolution of large crystals of PbSO4 reduces the recyclable lead content, what degrades cell performance [8,9]. Due to physical, morphological and chemical changes ðPb⇔PbSO4 ⇔PbO2 Þ at the positive and negative electrodes, they are mechanically degrade, losing contact and causing short-circuits. Lead-acid battery chemistry is based on the following reactions [10]:
Negative electrode side: þ PbðsÞ þ HSO 4 ðaqÞ ⇔ PbSO4 ðsÞ þ HðaqÞ þ 2e
Positive electrode side: þ PbO2 ðsÞ þ HSO 4 ðaqÞ þ 3HðaqÞ þ 2e ⇔ PbSO4 ðsÞ þ 2H2 OðlÞ
Over all reaction: PbðsÞ þ PbO2 ðsÞ þ 2H2 SO4 ðaqÞ ⇔ 2PbSO4 ðsÞ þ 2H2 OðlÞ Other failure modes in flooded cells are related to grid corrosion and acid stratification which are not discussed in the present paper. The increasing demand from lead-acid batteries to operate in deeper depth-of-discharge namely 20e30% DOD, emphasizes further the above described problem of sulfation. Due to sulfation, SLI batteries are planned to work at a shallow DOD. Their cycle life becomes very limited when they are forced to operate at 20e30% DOD. In order to solve the sulfation problem in deep cycling of SLI batteries, carbon materials were added into the active-masses of their electrodes [5]. It is universally accepted that carbon materials as additives enhance the performance of these batteries in terms of higher charge-acceptance, which is the key metric for addressing the sulfation problem. However, controversy arises regarding the preference of carbon allotrops, which renders the best performance. Graphitic powder, carbon black, various activated carbons are already introduced by several industries and their use as additives indeed enhances performance. Apart from that, carbonbased nano materials like single-walled carbon nanotubes (SWCNT), multi-walled carbon nanotubes (MWCNT) and graphenes are envisaged as most efficient materials, because of their ordered structures and consequently high intrinsic electrical conductivity. However, exploring the use of carbon-based nanomaterials in lead-acid batteries was very limited so far. In any event, it is clear that extending battery life cycling at both low and high DOD values definitely remains one of the most important challenges for this battery technology. We report here for the first time the use of CNT in both negative and positive electrodes of lead acid batteries in order to improve the cycle life of these batteries, especially at relatively high DOD. In this publication we describe how the use of carbon nanotubes (CNT) as a component in the composite electrodes improves the performances of flooded lead-acid cells. The use of CNT in these systems was reported already by several groups [11e14]. For instance, reference [14] reports on the use of CNT called “Molecular rebar” in both positive and negative plates of practical lead acid batteries produced by a commercial company. It is clear that the CNT used in the present work (in both negative and positive electrodes), are different than those used in the studies described before. Hence, we accumulate results coming from
79
several groups, showing clearly the improvement that can be reached for this important battery technology, by adding CNT to the active mass of both anodes and cathodes of L-A battery systems. It was important to realize that the results presented herein, obtained with new types of commercial CNT, were fully coherent with previous findings (reported in references [11e14]). The presence of CNT in the electrodes improves the electrical conductivity between the active mass particles though preventing thickening and the growth of large PbSO4 particles [15]. This improvement is naturally attributed to the formation of a stable conductive active mass matrix that enables the delivery and distribution of current to all the active material homogeneously. This modified active mass consists of composites made of CNT conformably coated with lead salts. By enabling a uniform current distribution, and subsequently well distributed electrochemical redox reactions throughout the electrode matrix, arrested the formation of too large PbSO4 particles [16]. The presence of CNT is supposed to improve both the mechanical stability and electrical integrity of the electrodes [17,18] and to induce uniform changes in the active mass during the complicated conversion reactions during cycling. We expect that the CNT in the electrodes will affect the electrical integrity of the active mass and should lead to the formation and existence of small enough particles, thus facilitating effective charge-transfer interactions [19e21]. 2. Experimental Two sets of SLI type electrodes were fabricated by Vulcan Automotive Ind. using their standard processes (see a description below). The electrodes were in two sizes: 1 cm2 for laboratory cells and 150 cm2 for full plate electrodes as are used in commercial batteries. The laboratory cells' capacity varied from 60 to 100 mAh while the large, industrial electrodes had capacities in the range 13e16 Ah. Reference sets of electrodes were based on standard SLI positive and negative electrodes. Other sets of positive and negative electrodes were modified with CNT. We tested the following cells combinations: (1) positive and negative electrodes without CNT as reference, (2) Positive electrode with CNT and negative electrode without CNT, (3) Positive electrode without CNT and negative electrode with CNT, (4) Positive electrode with CNT and negative electrode with CNT. All sets were cured and aged in similar curing processes and were confirmed to be tri-basic lead sulfate (3BS). The CNT which were used during the tests were: i MWCNT from Arkema Inc., having a diameter of 10e15 nm, 1e10 microns long. ii MWCNT having diameter of 40e60 nm, about 15 microns long, purchased from Hongwu Nanometer Inc., China. The same MWCNT as in (ii), modified with carboxylic groups (eCOOH). These CNT contained carboxylic groups that where bound to their surface, obtained via oxidation process with an aqueous nitric acid solution (by the manufacturer). The concentration of the carboxylic group was estimated as 6% by weight. In general, immersing the CNT in a concentrated solution of HNO3 at elevated temperature (e.g. 80 C), leads to an obvious oxidation of their surface, thus forming surface carboxylic groups. Their concentration depends on the immersion time and temperature. Laboratory electrodes contained 0.008e0.02% CNT by weight. Large format electrodes also contained 0.008e0.02% of CNT by weight. It was found that dispersing CNT in the electrodes' active mass in this range of concentration was suitable for reaching a desirable improvement in performance. With such a low amount of CNT required, the effect on the price of the batteries is negligible.
80
R. Marom et al. / Journal of Power Sources 296 (2015) 78e85
For the large cells, we used mainly CNT type (i). The specific gravimetric density of the electrolyte solution was 1.225 g/ml. Gel type sulfuric acid-based electrolytes were also prepared by addition of 5% fumed silica (Strem Chemicals) to the aqueous sulfuric acid solution [18]. The electrodes were prepared as follows: Pastes were prepared by dry mixing of metal lead 23e27% Pb and 73e77% PbO with glass fibers (Modacrylic) 0.1% in positive electrode and 0.2% in negative electrode, 0.75% proprietary surfactant (all by weight as a first stage). Then, water is added, 9% for positive electrodes and 0.85% for negative electrodes, and the mixtures were homogenously mixed. In the next stage, 9.8% (positive electrode) and 8.9% (negative electrode) sulfuric acid solution (specific gravity of 1.325) is gradually added to the paste mixtures. The differences among the various paste mixtures included the presence of CNT (or not, for reference electrodes), the type of CNT used and the method of dispersing and mixing the CNT in pastes. The pastes thus prepared were spread onto leadeantimony (positive electrode) and leadecalcium (negative electrode) grids by a custom-made pasting device. The plates are thus formed are then being covered by a special paper (Conpast Inc. pasting paper) and then are introduced into the curing chambers where they are cured at 40 C and 90% relative humidity for 24 h, and then dried at 60 C for 24 h. Cells were assembled with these electrodes, sulfuric acid electrolyte solutions were added and the cell underwent a formation process. The formation process produces in fact the lead-acid batteries by creating lead metal negative electrodes and lead dioxide positive electrodes. This initial formation cycle is proceeds in 3 steps: Step 1: charging at 0.1C rate for 3 h. Step 2: charging at 0.14C rate for 20e24 h. Step3: charging further at 0.1C rate for 3 more hours. This procedure was developed for commercial lead acid batteries. It ensures almost completed and successful conversion of the active mass comprising lead and lead mono-oxide to spongy lead metal negative electrode and lead dioxide positive electrode. The final step was a full discharge process at a slow rate until the cells reached 1.75 V following 0.1C rate charging till the cells reached 2.4 V. Then, the cells could be continuously cycled. Cycling included discharge at 30% DOD for small cells and at 25% depth of discharge for large cells at around 0.25C rates. When the charging (at around 0.5C rates) process was completed, the cells usually reached a potential of 2.4 V. During these cycling the indication for failure is the voltage dropped below 1.75 V. Hence, the indications for failure and the end of experiments are clear. HRTEM characterization of materials was carried out with a JEOL-JEM 2100 electron microscope with LaB6 emitter operating at 200 kV. Samples for the TEM studies were prepared by scraping some material from cycled electrodes followed by sonicating the powdered samples in ethanol and adding a few drops of the resulting suspension to a copper grid which is used as a substrate for the TEM measurements. Particles physically attached to the copper grid are introduced into the microscope for imaging and related studies. 3. Results and discussion Fig. 1 shows representative cycling data of standard cells and cells which positive electrodes contained CNT. Cycling was proceeded galvanostatically with a constant capacity, related to 30% DOD of the cells. The charging voltage was fixed to 2.4 V. Discharging to 30% DOD lowered the cells' voltage to values around 2 V. When a cell fails it's voltage at the end-of-discharge drops to low values. Failure of cells in this work was determined when the discharge voltage dropped below 1.75 V. Typical voltages at the end-of-discharge of modified and reference cells are plotted as a function of cycle number. The improvement by adding CNT to the positive electrodes in these cells is spectacular. Fig. 1B indicates 2
Fig. 1. (A) Typical data (voltage at the end of discharge and end of charge vs. cycle number) from a cycling experiment, in which repeated voltage profiles are recorded upon cycling around 30% of the total capacity at around C/4 rates. The cells are always charges at around C/2 rates up to 2.4 V (what bring them back to their full capacity). The indicator in these experiments is the voltages recorded at the end of discharge. The failure is clearly indicated in this chart as a gradual decrease of the voltage to values below 1.75 V, and (B) Typical cycling data (constant capacity, corresponding to 30% DOD) from lead-acid cells containing positive electrodes with and without CNT. The voltage at the end of discharge at each cycle is plotted as a function of cycle number. Failure (and the end of measurements) is indicated by the sharp drop in potentials (below 1.75 V as seen in the figure). In the chart, data from 3 cells are compared as indicated: a standard cells and cells with positive electrodes contain CNT, liquid and gel electrolyte solutions (as marked). The negative electrodes' active mass was in excess.
and 3 fold improvement in cycle life for cells with positive electrodes containing CNT based on liquid and gel electrolyte solutions, respectively. Standard cells with gel electrolyte solutions cycled at the same conditions, usually exhibited a few tens of cycles beyond those obtained with regular electrodes. In the experiments represented by Fig. 1, in which we concentrated on the positive electrode side, the active mass of the negative electrodes was in a large excess. As a next step, cells with positive electrode containing modified CNT to which surface carboxylic groups were bonded. We conducted similar sets of experiments, as those related to Fig. 1 (negative electrodes' active mass in large excess, cycling at 30% DOD of positive electrodes' capacity etc.). In these experiments only cells with liquid electrolyte solutions were tested. The beneficial effect of CNT with carboxylic groups bound to their surface is demonstrated in Fig. 2, which is similarly presented as Fig. 1B. More than 3 fold improvement is demonstrated for cells which positive electrodes included CNT which contained carboxylic groups. The performance of cells with positive electrodes containing CNT which included carboxylic groups, in flooded cells is superior (>750 cycles) compared to cells containing positive electrodes with unmodified CNT, which demonstrated 450 cycles (See Fig. 1). The improvement due to the CNT surface modification can be explained by a better wetting of the active mass which contains hydrophilic CNT. The presence of the carboxylic group may also improve the adhesive interactions between the modified CNT and the lead oxide in the positive electrodes' active mass. The carboxylic groups on the
R. Marom et al. / Journal of Power Sources 296 (2015) 78e85
Fig. 2. Typical cycling data (constant capacity, corresponding to 25% DOD) from leadacid cells containing positive electrodes with and without CNT. The voltage at the end of discharge of each cycle is plotted as a function of cycle number. Failure (and the end of measurements) is indicated by the sharp drop in voltages (below 1.75 V as seen in the figure). In the chart, data from two cells are compared as indicated: a standard cell (red) and a cell which positive electrode contains CNT which had carboxylic groups bound to their surface (blue), liquid electrolyte solutions. The negative electrodes mass was in excess. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
CNT surface may also stabilize the active mass upon cycling by chelation of Pb2þ ions [22], thus preventing growth of large PbSO4 crystals. The effect of CNT as additives to the negative electrodes in these systems was studied as well. CNT from Arkema and Hongwu were added to negative electrodes. In the cells assembled for these experiments the active mass of the positive electrodes was of course in large excess. The cells' cycling was limited to 30% DOD of the negative electrodes. In general, the negative electrode limited cells showed lower cycle life than the positive electrode limited cells. However, the cells with negative electrodes containing CNT also performed better than reference cells with no CNT. Fig. 3 compares the representative performance of cells in which both positive and negative electrodes contained CNT to that of similar CNT free reference cells. As presented in this figure, cells with CNT from the two sources we tested were compared. The cells which electrodes contained 15 micron long CNT demonstrated in average shorter cycle life compared to cells which electrodes contained CNT with length varied from 1 to 10 microns. Understanding the effect of the CNT length requires much more studies (on-going), beyond the scope of the present paper. Our first indication connects the difference in performance to the
Fig. 3. Typical cycling data (constant capacity, corresponding to 25% DOD) from leadacid cells containing negative and positive electrode with and without CNT. The voltage at the end of discharge of each cycle is plotted as a function of cycle number. Failure (and the end of measurements) is indicated by the sharp drop in voltages (below 1.7 V as seen in the figure). In the chart, data from 3 typical cells are compared as indicated: a standard cells and cells which electrodes contained two types of CNT as indicated: short (1e10 microns from Arkema) and long (15 microns from Hongwu Nanometer, China) as marked.
81
effectiveness of mixing and dispersion, which is better achieved with the shorter CNT. The beneficial effect of CNT for large, practical industrially made electrodes in real lead acid batteries was confirmed at Vulcan Batteries Ltd. Vulcan Batteries Ltd., used CNT from Arkema in electrodes for SLI lead acid batteries. The suitable amount of CNT in the large size electrodes was in the range of 0.008%e0.02% by weight. The CNT were introduced into the active mass by a unique patented process. Table 1 summarizes results of tests with large size electrodes containing CNT in the positive and negative electrodes (vs. reference experiments). In all of these tests, CNT from Arkema were examined. The tests included reference batteries (no CNT), batteries with either negative or positive electrodes containing CNT and batteries which both electrodes contained CNT. The batteries included single cells at capacity of 13e16 Ah which were cycled to 25% DOD. The indication for failure was the drastic drop of the endof-discharge voltage below 1.75 V. The results reported, in Table 1 confirmed well the above described laboratory tests. A clear improvement in specific capacity and cycle life was found in all cases were the electrodes (any or both) contained CNT. The improvement recorded has definitely important practical and commercial consequences. The most important finding from these experiments, as reflected from the data presented in the table was that addition of CNT into both positive and negative electrodes doubled the cycle life of practical SLI type lead-acid batteries. It should be noted that this result is especially important in light of the operation at 25% DOD, which can be considered as high, for this type of lead-acid batteries. It is important to note that there are commercial lead-acid batteries which are designed to operation at high DOD and very prolonged cycle life (thousands). However, these systems require special structure, geometry and complicated electrodes compositions. We assume that judicious and optimized use of CNT in all types of lead-acid batteries is expected to improve their performance beyond the state-of-the-art. In order to understand the beneficial effect of the presence of CNT in the electrodes, we conducted postmortem analyses of selected, representative cells after their end of life. Fig. 4 presents cycle life profiles of the cells on which teardown analysis was performed. The Figure shows the response of a cell of electrodes without CNT (red line) and the response of a cell, which electrodes contained CNT from Arkema (blue line). The cells capacity was 94.8 Ah/kg and they were cycled at 25% DOD. As usual, the drop of discharge potential below 1.75 V marked their end of life. In both figures the voltage at the end-of-discharge is plotted vs. number of cycles. These figures reflect the advantage of the presence of CNT, dispersed in the electrodes' active mass, in terms of increasing the cycle life of these batteries. Electrodes from the batteries which cycling data are presented in Fig. 4, were examined by scanning electron microscopy (SEM). Representative images are presented below. Fig. 5(AeD) show typical SEM images of electrodes have taken from lead-acid batteries after their failure. These figures comparing the morphology of electrodes from reference cells which did not contain CNT, with electrodes containing CNT that underwent the same cycling experiments as the reference systems (but failed after many more cycles, see examples in Table 1). The main features in these images are the size variation of PbSO4 crystals, the product of lead negative electrode's reduction and lead oxide positive electrode's oxidation. Their irreversible accumulation during cycling and growth to larger dimensions that avoid their accessibility into the electrochemical processes, are considered as the major failure mechanism of electrodes upon cycling lead-acid batteries. The life extension of
82
R. Marom et al. / Journal of Power Sources 296 (2015) 78e85
Table 1 Performances of industrial SLI type batteries: positive & negative electrodes containing 0.008e0.02% by weight CNT from Arkema as indicated. Cells type 1 (reference) are based on standard electrodes in sulfuric acid (specific gravity: 1.215). Cells type 2e6 included positive electrode modified with CNT. Cells type 7e8 included both positive and negative electrodes modified with CNT. The cells were discharged to DOD of 25% and the indication for end of life was a drop of the end-of-discharge potential to 1.7 V. Cell CNT source
CNT length (microns)
Capacity Measured Ah/kg
% Of effective capacity over 115 Ah/kg (115 ¼ 100%)
1 2 3 4 5 6 7 8
0 1e10 1e10 1e10 1e10 1e10 1e10 1e10
94.8 128 126.5 124.4 131.8 128.7 132.2 136.5
82 112 110 108 114.5 112 115 119
Reference Arkema Arkema Arkema Arkema Arkema Arkema Arkema
Number of cycles to end of life Enhanced cycle life compared to the (1.7 V) reference (%)
170 257 276 298 244 276 335 387
e 51.1 62.3 75.3 43.0 62.3 97 127.6
Fig. 4. Plots of the voltage of end of discharge vs. cycle number, for a reference flooded lead-acid battery (single cell) 13.6 Ah (red line), and a similar battery which electrodes contained CNT from Arkema (see details in the Experimental section) (blue line), that were cycled at 25% DOD. The electrodes were manufactured by a fully industrial process. The end of life for teardown analysis was the drop of potential below 1.75 V, as shown here. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
electrodes due to addition of CNT to their active mass described above, correlate very well to the morphological studies. The electrodes containing CNT develop much smaller crystals of PbSO4 because the presence of CNT facilitates their reversible precipitationedissolution during cycling, by maintaining a good electrical integration of the active mass. It was important to detect by appropriate imaging, the presence of CNT in the active mass. Only a relatively small amount of CNT, dispersed initially in the lead oxides slurry from which the electrodes are prepared, was needed, in order to obtain their pronounced positive effect on the performance. Hence, finding good microscopic images of CNT embedded in the active mass of the electrodes, especially after cycling, was not trivial. Nevertheless, it was possible to obtain reasonably good images of CNT embedded in the electrodes' active mass, by TEM measurements of particles scraped from pristine and cycled electrodes. The dispersion of the CNT in the electrodes is very uniform, thereby, any particle scraped from the CNT containing electrodes, is supposed to contain CNT that can be detected by a rigorous TEM analysis. Fig. 6(a) shows a typical bright field TEM image showing a single CNT embedded in the active mass of a cycled cathode. In Fig. 6(b), we show HRTEM image of such a CNT, in which its fine morphology (the walls) is well seen. Fig. 6(d) shows the intensity profile taken
from the area marked by white rectangular in Fig. 6(c) showing the measurements for 10 carbon layers at the CNT walls. These high resolution measurements of a single CNT embedded in the active mass are in a very good agreement with literature data reporting that the d-spacing in multi-walled CNT is around 0.34 nm [23]. From the studies described above, it is clear that the presence of uniformly dispersed CNT in the active mass of negative and positive electrodes of lead acid batteries extend their cycle life. A-priori, it was expected that their presence should enable a better electrical integrity of the active mass, thus facilitating more effective conversion reactions. From a previous work [16] it was known that CNT remain robust in positive electrodes, despite their relatively high upper voltage. The morphological studies described above connect the electrochemical improvement in performance to the formation of small PbSO4 particles upon discharge. Hence, these studies converge well to the following conclusion: We suggest that the mechanism in which the presence of CNT in the electrodes active mass improves performance, is via the formation of networks (mostly local, well distributed) by CNT that provide a better uniform current-distribution within the active mass. Consequently, small product particles are formed and their electrochemical accessibility is good enough to enable improvement in the electrodes cycle life. There is an open question about the interactions of
R. Marom et al. / Journal of Power Sources 296 (2015) 78e85
83
Fig. 5. Representative SEM images (<10 microns resolution) of electrodes have taken from the cells that underwent cycling until they failed. (A) Negative electrode with CNT, (B) negative electrode without CNT, (C) positive electrode with CNT and (D) positive electrode without CNT.
Fig. 6. (a) A typical bright field image showing a CNT embedded in the active mass of a cycled cathode (b) HRTEM image of such a CNT (12 nm thick) showing well resolved lattice fringes of the CNT's walls; (c) HRTEM image of a similar CNT displaying the lattice-fringes of the CNT; (d) Intensity profile taken from the area marked by a white rectangular in (c) showing the measurements for the 10 lines. These measurements are in very good agreement to the literature data of 0.34 nm for the d-spacing of CNT [21].
the CNT with the various moieties existing in the electrodes: electrically conductive metallic lead and lead oxides, nonconductive lead sulfate. The microscopic studies cannot provide insights about the intimate interactions between the CNT and electrodes' active mass components. However, from the clearly visible morphological effect, we conclude that the CNT are bound to all these moieties, during their reactions and transformations. Due to the way in which CNT are mixed with the active mass, they are not embedded within the particles, but rather they are well dispersed, maintaining a very good physical contact with most particles of the active mass. As a final stage of these studies, it was important to demonstrate their practical aspects, by measuring full SLI battery prototypes which both electrodes contain CNT. Fig. 7 shows a significant improvement in practical automotive batteries upon cycling at 1.3% DOD. Adding Arkema's CNT to the paste from which the electrodes of these practical batteries are prepared, led to a dramatic improvement in the cycle life of the batteries. In this figure, typical voltageecycles curves of reference (A) and CNT containing battery (B) are compared. As presented
herein, an extension of nearly 3-fold in the batteries cycle life could be obtained due to the presence of CNT. 4. Conclusions We have demonstrated that addition of CNT produced by commercial companies to both lead negative electrodes and lead dioxide positive electrodes in lead acid batteries pronouncedly improve the cells performance and doubling the cycle life of practical batteries. The introduction of the CNT is done in the early stage of electrodes preparation, mixed with the lead precursors in laboratory cells. Hence, formation leads to a uniform integration of the CNT in both the lead and the lead dioxide particulate matrices of the electrodes, at their final pristine form (before cycling) and during the periodic chargeedischarge cycling. We also demonstrate in small scale experiments that using modified CNT with carboxylic groups can lead to further improvement in performance. Cells comprising electrodes containing MWCNT functionalized with carboxylic groups exhibited
84
R. Marom et al. / Journal of Power Sources 296 (2015) 78e85
Fig. 7. Typical data of full 12.6 V batteries (voltage at the end of discharge and end of charge vs. cycle number) from cycling experiments, in which repeated voltage profiles are recorded upon cycling at 1.3% DOD at HRPSOC (based on SBA protocol). (a) Shows the response of a standard battery without CNT that could performed 37,000 cycles until failure. (b) Shows the response of a standard battery with Arkema's CNT that performed more than 89,000 cycles until failure.
longer cycle life than cells which electrodes contained usual CNT from commercial sources. Addition of MWCNT into the active mass of industrial manufactured electrodes (both negative and positive electrode) clearly extended the cycle life of SLI type flooded batteries: standard cells with commercial electrodes demonstrated in average 170 cycles (25% DOD) while cells with CNT modified electrodes exhibited averagely 360 cycles at 25% DOD. The improvement observed is well correlated with teardrop analysis of cycled cells that showed cycled electrodes containing CNT develop much smaller crystals of lead sulfate compared to cycled (CNT free) reference electrodes. Hence, it is clear that the presence of CNT enable much better accessibility of the active mass to the electron transport processes during cycling, due to their physical properties (high aspect-ratio, thin, flexible, robust and stable electronically conducting agents). Initial studies demonstrated that CNT derivatized with hydrophilic functional groups such as carboxylic groups can serve as better additives to lead-acid batteries electrodes than regular CNT, probably due to a better adhesion to their active mass. We demonstrated a significant improvement in practical SLI automotive batteries upon cycling at 1.3% DOD at HRPSOC (based on SBA protocol, as usual and acceptable for these practical battery systems), due to the presence of CNT in the electrodes. Excellent and highly reliable results were obtained with CNT from Arkema. It is highly important that improving the performance of LA batteries
could be achieved with commercial CNT. Commercial battery prototypes containing CNT dispersed in their electrodes, demonstrated and improvement of more than 100% in their cycle life, compared to reference batteries, prepared identically but without CNT. As mentioned in the introduction section, there are previous reports on the improvement of lead-acid batteries achieved by adding CNT to their electrodes. Accumulation of such positive coherent results coming from different groups substantiates their validity. Acknowledgment A partial support for this work was obtained from the ISF, Israel Science Foundation, in the framework of the INREP project and from the Chief Scientist of Israel Ministry of Economics. References , C.R. Acad. Sci. Paris 50 (1860) 640e642. [1] G. Plante [2] D.A.J. Rand, Valve-regulates Lead-acid Batteries, Elsevier, Amsterdam; Boston, 2004. [3] B. Culpin, D.A.J. Rand, J. Power Sources 36 (1991) 415e438. [4] R.D. Prengaman, J. Power Sources 95 (2011) 224e233. [5] D. Pavlov, Lead-acid Batteries: Science and Technology, Elsevier, Amsterdam, 2011. [6] P.T. Moseley, J. Power Sources 191 (2009) 134e138. [7] A.H. Catherino, F.F. Feres, F. Trinidad, J. Power Sources 129 (2004) 113e120. [8] D. Pavlov, P. Nikolov, ECS Trans. 41 (2012) 71e82. [9] P. Ruetschi, J. Power Sources 2 (1977) 3e24.
R. Marom et al. / Journal of Power Sources 296 (2015) 78e85 [10] D. Linden, T.B. Reddy, Handbook of Batteries, third ed., McGraw-Hill, NewYork, 2002, p. 23. [11] Y. Yin, C. Liu, S. Fan, RSC Adv. 4 (2014) 26378e26382. [12] M. Saravanan, P. Sennu, M. Ganesan, S. Ambalavanan, J. Electrochem. Soc. 160 (2013) A70eA76. [13] S.W. Swogger, P. Everill, D.P. Dubey, N. Sugumaran, J. Power Sources 261 (2014) 55e63. [14] N. Sugumaran, P. Everill, S.W. Swogger, D.P. Dubey, J. Power Sources 279 (2015) 281e293. [15] L. Vaisma, H.D. Wagner, G. Marom, Adv. Colloid Interface Sci. 128e130 (2006) 37e46.
85
[16] R. Shapira, G.D. Nessim, T. Zimrin, D. Aurbach, Energy Environ. Sci. 6 (2013) 587e594. [17] S. Logeshkumar, R. Manoharan, Electrochim. Acta 144 (2014) 147e153. [18] D. Pavlov, P. Nikolov, J. Power Sources 242 (2013) 380e399. [19] S. Takenaka, Carbon 47 (2009) 1251e1257. [20] S.S. Misr, J. Power Sources 168 (1) (2006) 40e48. [21] S.M. Kumar, S. Ambalavanan, S. Mayavan, RSC Adv. 4 (2014) 36517e36521. [22] X. Qin, L. Ni-Na, Z. Jun-Jie, Chin. J. Chem. 23 (2005) 1510e1514. [23] G. Maurin, I. Stepanek, P. Bernier, J.F. Colomer, J.B. Nagy, F. Henn, Carbon 39 (2001) 1273e1278.