Three-electrode cells are useful in studying the electrochemistry of lithium-ion batteries. Such an electrochemical setup allows the phenomena associated with the cathode and anode to be decoupled and examined independently. Here, we present a guide for construction and use of a three-electrode coin cell with emphasis on lithium plating analytics.
As lithium-ion batteries find use in high energy and power applications, such as in electric and hybrid-electric vehicles, monitoring the degradation and subsequent safety issues becomes increasingly important. In a Li-ion cell setup, the voltage measurement across the positive and negative terminals inherently includes the effect of the cathode and anode which are coupled and sum to the total cell performance. Accordingly, the ability to monitor the degradation aspects associated with a specific electrode is extremely difficult because the electrodes are fundamentally coupled. A three-electrode setup can overcome this problem. By introducing a third (reference) electrode, the influence of each electrode can be decoupled, and the electrochemical properties can be measured independently. The reference electrode (RE) must have a stable potential that can then be calibrated against a known reference, for example, lithium metal. The three-electrode cell can be used to run electrochemical tests such as cycling, cyclic voltammetry, and electrochemical impedance spectroscopy (EIS). Three-electrode cell EIS measurements can elucidate the contribution of individual electrode impedance to the full cell. In addition, monitoring the anode potential allows the detection of electrodeposition due to lithium plating, which can cause safety concerns. This is especially important for the fast charging of Li-ion batteries in electric vehicles. In order to monitor and characterize the safety and degradation aspects of an electrochemical cell, a three-electrode setup can prove invaluable. This paper aims to provide a guide to constructing a three-electrode coin cell setup using the 2032-coin cell architecture, which is easy to produce, reliable, and cost-effective.
Although the origin of lithium-batteries can be traced arbitrarily far back into the past, the large-scale production and commercialization of many of today's commonly found lithium-ion batteries began in the 1980s. Many of the materials developed during this era, one example being Lithium Cobalt Oxide (LiCoO2), are still commonly found in use today1. Many current studies have been focused towards the development of various other metal oxide structures, with some emphasis placed towards reducing or eliminating the use of cobalt in place of other lower cost and more environmentally benign metals, such as manganese or nickel2. The continuously changing landscape of materials used in lithium-ion batteries necessitates an effective and accurate method of characterizing both their performance and safety. Because the operation of any battery involves the coupled electrochemical response of both the positive and negative electrodes, typical two electrode batteries fall short of being able to characterize the electrodes independently. Poor characterization and the subsequent lack of understanding may then lead to dangerous situations or poor overall battery performance due to the presence of degradation phenomena. Previous research has been aimed at standardizing the processing techniques for typical two-electrode cells3. One method that improves upon the shortcomings of standard cell configurations is the three-electrode cell.
A three-electrode setup is one method to decouple the two electrodes' responses and provide a greater insight into the fundamental physics of the battery operation. In a three-electrode setup, a reference electrode is introduced in addition to the cathode and anode. This reference electrode is then used to measure the potential of the anode and cathode dynamically during operation. No current is passed through the reference electrode and hence, it provides a singular, and ideally stable, voltage. By using a three-electrode setup, the full cell voltage, the cathode potential, and the anode potential can be collected simultaneously during operation. In addition to potential measurements, the impedance contributions of the electrodes can be characterized as a function of the cell state of charge4.
Three-electrode setups are very useful for studying degradation phenomena in lithium-ion batteries, such as the electrodeposition of lithium metal, also known as lithium plating. Other groups have proposed three-electrode setups5,6,7,8,9,10,11,12,13 but they often use the inherently unstable lithium metal as a reference and include custom, difficult to assemble setups leading to reduced reliability. Lithium plating takes place when instead of intercalating into the host electrode structure, lithium is deposited on the surface of the structure. These deposits commonly assume the morphology of either a (relatively) uniform metallic layer (plating) or small dendritic structures. Plating can have effects ranging from causing safety issues to impeding cycling performance. From a phenomenological perspective, lithium plating occurs due to an inability of lithium to intercalate into the host electrode structure effectively. Plating tends to occur at low temperature, high charging rate, high electrode state of charge (SOC), or a combination of these three factors12. At low temperature, the solid-state diffusion inside the electrode is reduced, due to the Arrhenius diffusivity dependence on temperature. The lower solid-state diffusion results in a buildup of lithium at the electrode-electrolyte interface and a subsequent deposition of lithium. At a high charging rate, a similar phenomenon occurs. The lithium attempts to intercalate into the electrode structure very quickly but is unable to and thus is plated. At a higher SOC, there is on average less available space for the lithium to intercalate into the structure, and thus it becomes more favorable to deposit on the surface.
Lithium dendrites are of particular importance due to the safety concern they cause. If dendrites form inside a cell, there is a potential for them to grow, pierce the separator, and cause an internal short between the anode and cathode. This internal short can lead to very high-localized temperatures in the flammable electrolyte, often resulting in thermal runaway and even in an explosion of the cell. Another issue related to dendrite formation is the increased surface area of the reactive lithium. The newly deposited lithium will react with the electrolyte and cause increased solid electrolyte interphase (SEI) formation, which will lead to increased capacity loss and poor cycling performance.
One issue associated with the design of a three-electrode system is the selection of the appropriate reference electrode. Logistics relating to the location and size of the reference, positive, and negative electrodes can play an important role in acquiring accurate results from the system. One example is that the misalignment of the positive and negative electrodes during the cell construction and the resulting edge effects can introduce error in the reference reading14,15. In terms of material selection, the reference electrode should have a stable and reliable voltage and have a high non-polarizability. Lithium metal, which is often used as a reference electrode by many research groups, has a potential that depends on the passive surface film. This can produce issues because cleaned and aged lithium electrodes display different potentials16. This becomes a problem when long-term aging effects are studied. Research by Solchenbach et al. has attempted to eliminate some of these instability issues by alloying gold with lithium and using it as their reference11. Other research has looked at different materials including lithium titanate, which has been studied experimentally and shows a large electrochemical potential plateau range around 1.5 – 1.6 V17 (~50% SOC). This plateau helps to maintain a stable potential, especially in the event of accidental perturbation to the electrode's state of charge. The potential stability of LTO, including carbon-based conductive additives, is maintained even at different C-rates and temperatures.18 It is important to emphasize that the selection of the reference electrode is an important step in the three-electrode cell design.
Many research groups have proposed experimental three-electrode cell setup. Dolle et al. used thin plastic cells with a lithium titanate copper wire reference electrode to study changes in impedance due to cycling and storage at high temperatures19. McTurk et al. employed a technique whereby a lithium plated copper wire was inserted into a commercial pouch cell, with the main goal being to demonstrate the importance of noninvasive insertion techniques9. Solchenbach et al. used a modified Swagelok-type T-cell and a gold micro-reference electrode (mentioned earlier) for impedance and potential measurements.11 Waldmann et al. harvested electrodes from commercial cells and reconstructed their own three-electrode pouch cells for use in studying lithium deposition12. Costard et al. developed an in-house experimental three-electrode cell housing to test the effectiveness of different reference electrode materials and configurations13.
Most of these research groups use pure lithium metal as the reference, which can have concerns with stability and SEI growth, especially with long-term use. Other issues involve complicated and time-consuming modifications to existing or commercial setups. In this paper, a reliable and cost-effective technique for constructing three-electrode Li-ion coin cells for electrochemical tests is presented, as shown in Figure 1. This three-electrode setup can be constructed using standard coin cell components, copper wire, and lithium titanate-based reference electrode (see Figure 2). This method does not require any specialized equipment or elaborate modifications and follows standard laboratory scale electrochemical procedures and materials from commercial vendors.
1. Reference Electrode and Separator Preparation
2. Construction of the Preparation Cell
3. Lithiation Procedure
4. Construction of Working Cell
5. Electrochemical Tests
Typical results for the voltage and potential profiles for the three-electrode cell can be seen in Figure 7. In an ideal setup, the full cell voltage should be identical to that produced from a two-electrode cell using the same electrode couple. This is one method to determine whether the insertion of the reference electrode modifies the performance of the cell. If there is a significant difference between the two- and three-electrode full cell performance (for identical working and counter electrodes), then it can be assumed that the insertion of the reference electrode modifies the behavior of the cell and the results are no longer meaningful.
During the charging process, lithium moves from the cathode to the anode electrode. As lithium is being removed from the cathode microstructure, its potential with respect to Li/Li+ increases. The opposite occurs with the anode, as the structure is continually filled with lithium. During discharge, the reverse situation occurs. These changes in potential are reflected in the three-electrode potential profiles, which can be seen in Figure 7.
A powerful outcome of the three-electrode cell setup is the detection of the onset of lithium plating. Figure 8 shows an example of an anode potential profile during the fast charging of a coin cell. From the zoomed-in portion of the plot, it can be seen that the anode potential reaches negative values towards the end of the CC charging process. This is indicative of the presence of lithium plating in the cell. This measurement is not possible when using a standard two-electrode setup.
The impedance results for the three-electrode setup are shown in Figure 9. A typical impedance response consists of three characteristic regions: a high-frequency semicircle, a medium-frequency semicircle, and a low-frequency diffusion tail. The Re(Z) intercept of the plot, the radii of the semicircles, and the slope of the diffusion tail can be used to characterize important electrochemical phenomena occurring within the cell.
Another powerful use of the three-electrode tool is impedance characterization as a function of the state of charge. This impedance can be correlated to various degradation phenomena, including the electrodeposition of lithium. Figure 10 shows an example of impedance spectra collected for the full cell, cathode, and anode for a single coin cell. The changing impedance can be used to characterize the individual contributions of electrode impedance as the cell SOC is changing. For the anode, the impedance can be correlated to the various degradation phenomena, including the growth of the SEI layer and lithium plating and dendrite formation. Distorted impedance measurements including inductive loops (see Figure 11) can be correlated to two different factors. A bad sealing of the cell along with electrolyte leaking (see Figure 5) can induce an inductive impedance response. The electrode shape and the position of the reference electrode tip (see Figure 4) can also induce inductive loops on the impedance response21.
The behavior of the individual electrode potentials can be used to provide analysis, which is not available in traditional two-electrode setups. For example, plateau regions in the potential profile can represent phase changes in the electrode structure. These phase changes can be corroborated with additional electrochemical testing, such as cyclic voltammetry. Also, the value of the anode potential can be used in conjunction with other methods to determine lithium plating, which occurs once the anode potential has reached a value below 0.0 V vs. Li/Li+.
Figure 1. In-house three-electrode cell setup. (a) This panel shows a photograph of a completed three-electrode coin cell. (b) This panel shows an exploded view of the internal cell components. Please click here to view a larger version of this figure.
Figure 2. Three-electrode coin cell showing the entry point of the reference electrode as well as the internal layout. Note that in this figure, the cap is transparent and the wave spring (not shown) is located just above the top spacer. Please click here to view a larger version of this figure.
Figure 3. Reference electrode configurations. (a) This panel shows a reference electrode batch taped to a holder element (e.g., clean glass container) with the coated ends suspended for drying. The following panels show reference electrode configurations corresponding to (b) the situation immediately after the electrode casting onto a wire, (c) the use in a preparation cell, and (d) the use in a working cell. The panels are not drawn to scale. Please click here to view a larger version of this figure.
Figure 4. Various electrodes that are possible to use in the construction of three-electrode coin cells. These panels show (a) a spiral shape, (b) a central reference, (c) a keyhole shape, (d) a pizza-slice shape, (e) on the side, and (f) on the side with a small circular cutout. Please click here to view a larger version of this figure.
Figure 5. Improperly sealed three-electrode coin cells, demonstrating leaking and the resultant reaction of electrolyte with environment. Under this condition, it is recommended to remove the cell from the holder, since electrolyte can rust the electrical terminals. Please click here to view a larger version of this figure.
Figure 6. Connection to the electrochemical testing machine to measure impedance. Connection configurations are shown for (a) a full cell (ZF), (b) a cathode (ZC), and (c) an anode (ZA). A performance and cycling of the three-electrode cell can be done using the cathode connection shown in panel (b). Please click here to view a larger version of this figure.
Figure 7. Voltage measurements. These panels show the voltage measurements for an anode, a cathode, and a full cell (two- and three-electrode cell) during (a) constant current, constant voltage (CCCV) charging at C/10 and (b) constant current (CC) discharging at C/10. Please click here to view a larger version of this figure.
Figure 8. Negative anode potential during fast charging. This panel shows a negative anode potential occurring during the fast charging (1C-rate) of a three-electrode coin cell, indicating the possible presence of lithium plating. Please click here to view a larger version of this figure.
Figure 9. Impedance response. These panels show the impedance response from using a frequency response analyzer for a full cell, a cathode, and an anode showing (a) a complete frequency range and (b) a low-frequency range. Please click here to view a larger version of this figure.
Figure 10. EIS. These panels show the electrochemical impedance spectroscopy corresponding to (a) the full cell, (b) cathode, and (c) anode measurement for a three-electrode coin cell as a function of the SOC. Please click here to view a larger version of this figure.
Figure 11. Anode impedance distortion. This figure shows anode impedance distortion measured for a three-electrode coin cell, likely caused by either a misalignment of the reference electrode inside the cell or an improper sealing of the coin cell near the exit location of the wire. Please click here to view a larger version of this figure.
Cell crimping pressure plays an important part in the success rate of both the preparation and working cells. If the cell is crimped at too high a pressure (>800 psi), the reference electrode can become shorted with the cell cap due to the reference wire position in-between the cap and the gasket. Note that the wire crossing this interface is a requirement in order to connect the reference electrode reading to an external measurement device. If the cell pressure is too low (<700 psi), the cell can have issues with incomplete crimping which may cause electrolyte leakage and air penetration after the cell is removed from the inert argon environment. It was found that around 750 psi is the optimal pressure for crimping the cell to avoid leaking or shorting issues. In order to provide additional means to prevent these issues with shorting of the reference wire, a vital step in the construction process is the additional square separator that is placed along the gasket where the wire crosses the cell boundary. This separator provides an additional insulating layer that helps prevent internal shorting. In addition, slightly different crimping pressures may be required for the preparation and working cell. The preparation cell uses two lithium discs that are significantly thicker than an electrode cast on a metal foil-which are used in the working cell.
After the lithiation of the reference electrode in the preparation cell, the reference electrode must be extracted and reused in the working cell. During this process, extreme care must be taken. In general, if the reference electrode was prepared properly, there should not be any issue associated with the adhesion of the material to the flattened section of the wire. In any case, the amount of time between when the reference electrode is removed from the preparation cell and used in the working cell should be minimized. The reference electrode should not be placed on any surface or allowed to rest outside for a significant amount of time. Minimizing manipulation of the wire is ideal because it avoids possible fatiguing and breaking of the wire.
Another important consideration when constructing the three-electrode coin cell is sealing the cell properly. Because the wire is sandwiched between the cap and the gasket, there is potential for a small breach in the cell that may allow for electrolyte leakage or air penetration in the cell. If this is not rectified, distortion may be seen in the impedance measurements and the entire cell may fail due to reactions with the environment, especially after an extended period outside of the inert glovebox in which it is fabricated. In the cell construction procedure, the use of the non-conductive epoxy is vital because it completely seals the cell from the outside environment. One interesting observation is that if the cell is not crimped to a high enough pressure, the epoxy will not harden properly and will sometimes bubble up. This may be caused by the electrolyte being wicked up and mixed with the epoxy, or the higher internal pressure of the cell slowly leaking out and causing bubbles to form. Note that the epoxy, both during and after hardening, was soaked in the electrolyte and no obvious sign of any reaction was observed. If used properly, the epoxy-sealed cell should be allowed to dry for a minimum of 1 h inside the glovebox before removal. Afterward, the epoxy can harden in an atmospheric environment. Depending on the epoxy used, it may take 24 h or more for the epoxy to fully cure, and during this process, the cell should be allowed to rest. In the event that the cell is not sealed, or the sealing procedure is not sufficient, the cell will leak into the environment. After a while, the cell may begin to change colors. Some examples of poorly sealed cells can be seen in Figure 5.
When constructing the three-electrode coin cells, the shape of the host electrodes can have an influence on the performance of the cell. Various possible shapes can be seen in Figure 4. In an ideal case, the reference electrode would be placed at the center of the electrodes. Some problems that can occur involve an uneven pressure distribution within the cell due to the location of the reference electrode. Another issue is that the existence of the reference electrode between the host electrodes creates an artificial increase in the cell impedance, due to the fact that the reference is blocking a portion of the electrode area. Some configurations (Figure 4C – 4F) attempt to reduce this issue by carving out a small area in which the reference can sit. The problem is that this reduces the cell capacity as well as introduces complexity in the manufacturing process.
When connecting the three-electrode cell to the electrochemical testing measurement, the connection to the reference electrode can be very sensitive due to the small diameter of the copper wire used. Note that the wire diameter must be small in order to reduce any effects on the cell performance, one of which could be a blockage of the area between the two planar electrode disks. Because of this connection sensitivity, it is advantageous to bend the exposed end of the copper wire back on itself multiple times to increase the surface area for connection. If this is not done, the reference electrode may appear to be shorted or have failed, when in fact the cell is working as expected.
One limitation of using a three-electrode coin cell is that the entire process is done by hand. A certain amount of practice is required when constructing coin cells in order to produce consistent and reliable results. In the case of accidental shifting of the position of the reference electrode, working electrode, and/or counter electrode inside the cell, impedance and potential readings may become distorted or inaccurate. This is not as important for the preparation cell because the objective of this cell is simply to prepare the reference by partial lithiation and to determine the value of the plateau voltage (typically ~1.565 V for the lithium titanate electrodes used in this procedure).
One good method for determining the success of the cell is through the observation of impedance distortion for the anode. In the case of an improperly sealed cell, or a poor electrode alignment, inductive impedance loops are often seen when taking the anode impedance. These loops are more easily noticed when the cell is fully discharged (i.e., when the cell is first constructed), so they can be tested for prior to any cycling of the cell. An example of anode impedance spectra with the distortion present is shown in Figure 9.
The authors have nothing to disclose.
Financial support from the Texas Instruments (TI) University Research Partnership program is gratefully acknowledged. The authors also gratefully acknowledge the assistance of Chien-Fan Chen from the Energy and Transport Sciences Laboratory, Mechanical Engineering, Texas A&M University, during the initial stage of this work.
Agate Mortar and Pestle | VWR | 89037-492 | 5 in diameter |
Die Set | Mayhew | 66000 | |
Laboratory Press | MTI | YLJ-12 | |
Analytical Scale | Ohaus | Adventurer AX | |
High-Shear Mixing Device | IKA | 3645000 | |
Argon-filled Glovebox | MBraun | LABstar | |
Hydraulic Crimper | MTI | MSK-110 | |
Battery Cycler | Arbin Instruments | BT2000 | |
Potentiostat/Galvanostat/EIS | Bio-Logic | VMP3 | |
Vacuum Oven and Pump | MTI | – | |
Copper Wire | Remington | PN155 | 32 AWG |
Glass Balls | McMasterr-Carr | 8996K25 | 6 mm borosilicate glass balls |
Stirring Tube | IKA | 3703000 | 20 ml |
Celgard 2500 Separator | MTI | EQ-bsf-0025-60C | 25 μm thick; Polypropylene |
Stainless Steel CR2032 Coin Cell Kit | Pred Materials | Coin cell kit includes: case, cap, PP gasket | |
Stainless Steel Spacer | Pred Materials | 15.5 mm diameter × 0.5 mm thickness | |
Stainless Steel Wave Spring | Pred Materials | 15.0 mm diameter × 1.4 mm height | |
Li-ion Battery Anode – Graphite | MTI | bc-cf-241-ss-005 | Cu Foil Single Side Coated by CMS Graphite (241mm L x 200mm W x 50μm Thickness) |
Li-ion Battery Cathode – LiCoO2 | MTI | bc-af-241co-ss-55 | Al Foil Single Side Coated by LiCoO2 (241mm L x 200mm W x 55μm Thickness) |
Polyvinylidene Difluoride (PVDF) | Kynar | Flex 2801 | |
N-Methyl-2-Pyrrolidinone Anhydrous (NMP), 99.5% | Sigma Aldrich | 328634 | |
CNERGY Super C-65 | Timcal | ||
Electrolyte (1.0 M LiPF6 in EC/DEC, 1:1 by vol.) | BASF | 50316366 | |
Lithium Titanate (Li4Ti5O12) | Sigma Aldrich | 702277 | |
KS6 Synthetic Graphite | Timcal | ||
Lithium Metal Ribbon | Sigma Aldrich | 320080 | 0.75 mm thickness |
Epoxy Multipurpose | Loctite | ||
Electrical Tape | Scotch 3M Super 88 | ||
Isopropyl Alcohol (IPA), ACS reagent, ≥99.5% | Sigma Aldrich | 190764 |