Vehicles powered by electricity have increasingly been considered a central solution to the twin challenges of petroleum scarcity and urban air pollution. A vehicle traction battery is not merely an accessory, but the core energy carrier that determines driving range, acceleration, safety, and total cost. In my work, I focused on lithium-ion chemistry because it offers a unique combination of high specific energy, high working voltage, low self-discharge, long cycle life, and flexible cell packaging. Compared with traditional lead-acid and nickel-metal hydride batteries, a lithium-ion power source can substantially increase the usable energy of an electric vehicle while reducing the mass of the battery pack. In this article, I describe, from a first-person perspective, the manufacturing-parameter optimization and electrochemical evaluation of a high-capacity lithium-ion vehicle traction battery based on spinel lithium manganese oxide.
The positive active material was spinel LiMn2O4, a commercially attractive candidate for vehicle traction battery applications. It has a three-dimensional lithium-ion diffusion path, high thermal stability, low cost, and reduced environmental impact when compared with LiCoO2 or Ni-rich layered oxides. I selected LiMn2O4 as the main object of this study and investigated how its combination with conductive additives, binders, current collectors, electrode compaction, electrolyte addition, and formation conditions affects the final performance of a large-format lithium-ion vehicle traction battery.
1. Background and Requirements of the Power Battery
Before designing experiments, I reviewed the service conditions imposed by electric vehicles. A practical vehicle traction battery must deliver high energy to guarantee an acceptable cruising range, high power for startup and climbing, long calendar and cycle life, a wide operating-temperature window, high charging efficiency, and intrinsic tolerance to abuse. The most common battery chemistries considered for electric-drive vehicles include lead-acid, nickel-metal hydride, lithium-ion, and fuel cells. Lead-acid batteries are inexpensive and reliable, but their low specific energy and limited cycle life make them unsuitable for modern electric cars. Nickel-metal hydride batteries have higher energy and power, but their self-discharge and high-temperature behavior are worse than lithium-ion batteries. Fuel cells still face unresolved infrastructure and catalyst-cost challenges. Therefore, I considered lithium-ion technology to be the most balanced candidate for near- and mid-term vehicle traction battery commercialization.
| Battery type | Specific energy | Specific power | Cycle life | Typical role |
|---|---|---|---|---|
| Lead-acid | Low | Moderate | Short | Light electric scooters and legacy vehicles |
| Nickel-metal hydride | Medium | High | Medium | Hybrid vehicle traction batteries |
| Lithium-ion | High | High | Long | Electric and plug-in hybrid vehicle traction battery packs |
| Fuel cell | Fuel-dependent | Stack-size-dependent | Still maturing | Demonstration or long-range commercial vehicles |
From a market perspective, the global growth of the vehicle traction battery industry is expected to remain strong for decades. I am convinced that manufacturing know-how, not only material chemistry, is one of the critical competitive factors. Laboratory coin-cell results are frequently not transferable to large cells because vehicle traction battery electrodes are thick, high-capacity, and subjected to severe mechanical stresses during winding, stacking, electrolyte wetting, and cycling. Thus, a systematic study of production parameters was necessary.
2. Experimental Approach and Cell Manufacturing Route
My investigation was carried out using an engineering process-development methodology. I prepared electrodes, assembled cells, and performed electrochemical tests under industrial-like conditions. The major raw materials are summarized below. I deliberately use only chemical or functional designations, not specific vendors, because the conclusions are intended to be universal for LiMn2O4/graphite vehicle traction battery manufacturing.
| Component | Material | Function |
|---|---|---|
| Positive active material | Spinel LiMn2O4 | Reversible lithium host at high potential |
| Positive conductive additive | Conductive carbon black | Improves electron percolation |
| Positive binder | PVDF | Adhesion between active particles and Al foil |
| Positive solvent | NMP | Dissolves PVDF and adjusts slurry viscosity |
| Negative active material | Graphite | Lithium host at low potential |
| Negative binder | Water-based binder | Adhesion of graphite to Cu foil and environmental friendliness |
| Anode current collector | Copper foil | Electron conduction and support |
| Cathode current collector | Aluminum foil | Electron conduction at high potential |
| Separator | Polyolefin microporous membrane | Ionic transport and electronic insulation |
| Electrolyte | LiPF6 carbonate solution | Ionic conductor between electrodes |
The manufacturing process must be kept extremely dry because lithium-ion cells operate above the decomposition voltage of water. If water remains in the electrode stack, gassing, HF formation, loss of capacity, and internal corrosion will occur. My process included vacuum drying of powders and electrodes before electrolyte filling. The slurry was prepared in a vacuum mixer because gas bubbles entrained in the slurry create pinholes in the dry coating and lead to lithium plating or local current-density non-uniformity in the final vehicle traction battery electrode.
The cell assembly route was: first, mix LiMn2O4 with carbon black and PVDF in NMP for the positive electrode; second, mix graphite with conductive carbon and water-based binder in deionized water for the negative electrode; third, coat the two slurries on aluminum and copper foils; fourth, evaporate the solvent in a multi-zone drying oven; fifth, calendar the electrodes to the target density and porosity; sixth, cut the electrodes to size; seventh, weld conductive tabs; eighth, wind the electrodes with a separator into a jelly-roll core; ninth, insert the core into a prismatic steel case and weld the terminals; tenth, vacuum-dry the core to remove residual moisture; eleventh, fill the electrolyte in an inert dry atmosphere; and finally, perform formation and degassing before sealing and grading. I found that every step influences the quality of the finished vehicle traction battery, but the most influential parameters are described in the following sections.
3. Critical Manufacturing Parameters
3.1 Selection of the Positive Electrode Material
For a commercial vehicle traction battery, I first evaluated spinel LiMn2O4 powders with regard to particle-size distribution, phase purity, and specific surface area. Very small particles increase the electrode/electrolyte contact area and can improve utilization, but they also increase the risk of particle agglomeration and electrolyte-side reactions. In addition, very fine powder makes slurry rheology difficult to control. Very large particles reduce the contact area between particles, increase electrode resistance, and may cause poor adhesion to the aluminum current collector. In my study, I preferred a LiMn2O4 powder with a particle size in the range of 5–15 μm and with a manganese content representative of a stoichiometric spinel. Such a powder gives acceptable processability in coating and calendaring, as well as stable discharge behavior in vehicle traction battery-format cells.
| Property | Criterion | Reason |
|---|---|---|
| Chemical nature | Spinel LiMn2O4 | Three-dimensional channels for Li-ion diffusion |
| Particle size | Approximately 5–15 μm | Good packing, low resistance, stable slurry |
| Purity | High-purity, low impurity content | Minimizes self-discharge and side reactions |
| Morphology | Angular or irregular particles with rough surface | Provides sufficient inter-particle contact and binder anchoring |
The theoretical specific capacity of LiMn2O4, based on one lithium per formula unit, can be expressed as
$$Q_{\mathrm{th}} = \frac{nF}{3.6\,M_{\mathrm{LiMn_2O_4}}} \approx \frac{96485}{3.6 \times 180.8} \approx 148 \, \mathrm{mAh\,g^{-1}},$$
where n is the number of electrons transferred per formula unit, F is Faraday’s constant, and 3.6 is the conversion factor between coulombs and milliampere-hours. Therefore, a practical electrode coating normally contains conductive additives and binder to maintain electronic connectivity and mechanical integrity.
3.2 Effect of Active Material Content in the Positive Electrode
Lithium manganese oxide has moderate electronic conductivity. If I used no conductive additive, the electrode would exhibit a high ohmic resistance, especially in a thick vehicle traction battery electrode. The binder is needed to attach the active powder to the current collector. However, both carbon black and binder are electrochemically inactive and reduce the volumetric energy density of the electrode. Therefore, their amounts must be minimized without sacrificing adhesion or electrical percolation.
Using a fixed small-format cell design, I changed the LiMn2O4 mass fraction of the positive solid powder from 85 wt%, 88 wt%, 91 wt%, to 94 wt%, while the balance consisted of conductive carbon black and PVDF. I then measured the initial discharge capacity, internal resistance, and cycle retention. The trends are summarized below.
| Active material fraction wt% | Initial cell capacity | Cell internal resistance | Cycle capacity retention | Typical limiting issue |
|---|---|---|---|---|
| 85 | Lower, enough binder/carbon | Low | Stable | Low energy density |
| 88 | Moderate | Slightly higher | Stable | Trade-off begins |
| 91 | High | Acceptable | Good | Best compromise |
| 94 | High but limited by adhesion/electron paths | Higher | Faster fade | Binder deficiency and carbon deficiency |
When the active-material fraction was too high, the amount of binder was insufficient to keep the coating adherent to aluminum during high-rate discharge; active particles could detach, causing increased internal resistance and capacity fade. In contrast, when the active fraction was too low, the cell capacity was reduced because a larger fraction of the electrode volume was inert. The best balance, taking first-cycle capacity, resistance, cycle life, and production cost into account, was obtained when the cathode solid contained approximately 91 wt% LiMn2O4. This composition was therefore selected for the subsequent vehicle traction battery experiments.
3.3 Capacity Balance Between Positive and Negative Electrodes
In any lithium-ion cell, the negative electrode should have a reversible capacity slightly larger than the positive electrode. This design rule prevents lithium metal deposition on the graphite surface during charging. If lithium plates instead of intercalating, the resulting dendrites may pierce the separator and cause an internal short circuit, fire, or explosion. Thus, the P/N ratio, or capacity balance coefficient, is a critical safety parameter for the vehicle traction battery.
The capacity balance coefficient I used is defined as
$$\alpha = \frac{Q_{-,\mathrm{areal}}}{Q_{+,\mathrm{areal}}}
= \frac{\rho_{-}\, w_{-}\, q_{-}}{\rho_{+}\, w_{+}\, q_{+}},$$
where \(\rho\) is the dry coating areal density, \(w\) is the active-material mass fraction, and \(q\) is the reversible specific capacity of the corresponding electrode. In the paper-like production process, coating weight varies from place to place and from batch to batch. I observed that the positive and negative coating areal densities both varied by approximately \(\pm 3\%\). If the average designed coefficient is \(\alpha\), the worst-case coefficient occurs when the positive loading is high and the negative loading is low:
$$\alpha_{\mathrm{minimum}} = \alpha \,\frac{1-\delta}{1+\delta},$$
where \(\delta\) is the fractional coating variation. To guarantee that \(\alpha_{\mathrm{minimum}}\) remains greater than unity, the design value must satisfy
$$\alpha > \frac{1+\delta}{1-\delta}.$$
For \(\delta = 0.03\), the lower limit becomes approximately 1.062. I therefore designed the cell with a safety margin above that limit.
In practical cells, the exact optimum was not only determined by nominal capacity. If the capacity balance coefficient is too low, lithium plating can occur during overcharge and cause internal heating. If the balance coefficient is too high, the negative electrode is excessive and the cell energy density is sacrificed. I evaluated several values and found that the best compromise for this vehicle traction battery system lies in the range of about 1.06 to 1.12. Within this range, the initial discharge capacity is close to the maximum, no lithium plating is observed under normal cycling, and the safety tests show no explosion or fire. The following table summarizes the general observations:
| Capacity balance coefficient | Capacity behavior | Safety behavior |
|---|---|---|
| Too low (e.g. near 1.00) | May appear high because positive is fully utilized | High risk of lithium plating, dendrite formation, and short circuit |
| Optimum (approximately 1.06–1.12) | Good discharge capacity and low polarization | Acceptable abuse tolerance; no fire/explosion |
| Too high (graphite greatly in excess) | Capacity limited by positive electrode; lower energy density | Safer electrochemically but thicker electrode and possible wetting difficulty |
3.4 Compaction Density and Battery Stack Tightness
After coating and drying, the positive and negative electrodes contain a large amount of porosity. In a dry electrode, such porosity is necessary for later electrolyte penetration and lithium-ion transport. However, if the electrode is too porous, the active material volume fraction is low, the inter-particle electronic contact is poor, and the cell cannot achieve high volumetric energy. To solve this problem, the electrode is passed through a roll press to control the coating thickness and compaction density. The electrode compaction density is defined as
$$\rho_{\mathrm{cal}} = \frac{m_{\mathrm{coating}}}{A_{\mathrm{coating}}\,h_{\mathrm{coating}}},$$
where \(m_{\mathrm{coating}}\) is the mass of the active layer, \(A_{\mathrm{coating}}\) is the geometric area, and \(h_{\mathrm{coating}}\) is the final thickness of the coating after calendering.
I varied the positive-electrode compaction density while keeping the negative electrode at an optimum. If the positive electrode was compressed too strongly, the electrode became brittle and the particle structure partially collapsed. Also, the remaining pore network was too narrow for fast diffusion of lithium ions during high-rate discharge. The high-rate capability of the vehicle traction battery therefore decreased at excessive compaction density. If the compaction density was too low, the electrode thickness was large and the wound electrode stack could not be inserted into the prismatic steel case without scratching the jellyroll. Scratches cause internal micro-shorts and rejected cells.
The compaction density also influences the electrode surface roughness and stack thickness. After rolling, the elastic spring-back of the electrode should be controlled, because subsequent swelling during electrolyte wetting and charging consumes part of the cell internal clearance. I measured the thickness before and after assembly and identified that the positive LiMn2O4 coating gives its best combination of capacity, high-rate discharge, and cycle life at a compaction density of approximately 2.8–3.0 g cm–3. The graphite negative electrode has a lower density because graphite is a softer material with a higher porosity requirement; its optimum compaction density is approximately 1.5–1.6 g cm–3. These values were used in the large-format design.
In a prismatic cell, the electrode stack must fit the steel case with an appropriate assembly tightness. I define tightness as
$$\eta_{\mathrm{stack}} = \frac{\sum t_{\mathrm{electrodes}} + \sum t_{\mathrm{separators}}}{D_{\mathrm{inner}}},$$
where \(D_{\mathrm{inner}}\) is the interior cavity dimension in the thickness direction. A high tightness value implies a tight stack that reduces movement and improves thermal contact, but too much tightness makes electrolyte wetting difficult and can cause bent electrodes or uneven compression. I found that the optimum assembly tightness is about 0.94–0.98 in the studied prismatic format; that is, the stack is only slightly smaller than the available inner cavity, leaving a controlled space for electrode swelling and electrolyte accommodation. The table below summarizes the influence of compaction density on the manufacturing outcome.
| Compaction condition | Porosity and wetting | High-rate performance | Assembly and short-circuit risk |
|---|---|---|---|
| Too low | Excessively porous; thick electrode | Moderate or low electrode conductivity | Stack too thick; outer layers scratched during insertion |
| Optimum | Well-balanced porosity; fast wetting | High and stable | Low rejection rate |
| Too high | Very low porosity; slow Li-ion diffusion | Poor at high current | Brittle electrode; risk of cracks and particle detachment |
3.5 Electrolyte Addition and Internal Wetting
The electrolyte amount is another important variable in vehicle traction battery manufacturing. The electrolyte must fill the pores of the positive electrode, the negative electrode, and the separator. In theory, the minimum volume of electrolyte equals the total pore volume of those components. In practice, an extra amount is needed because the surfaces of particles are not perfectly wetted and because some electrolyte is consumed during the first formation cycles to build the solid-electrolyte interphase layer on graphite.
If the electrolyte addition is insufficient, part of the electrode remains inactive. The cell then shows a low discharge capacity and a high internal resistance. Moreover, during cycling, local electrolyte depletion accelerates the aging of the vehicle traction battery. If the electrolyte addition is excessive, the extra liquid does not improve capacity or internal resistance; it can leak during formation, corrode the equipment, and increase the cost. I therefore studied cells with different amounts of electrolyte and compared the discharge capacities and internal resistances.
I observed a steep increase of capacity when the electrolyte amount increased from a deficient state to the saturation point. Above saturation, the capacity reached a plateau. The internal resistance decreased gradually with increasing electrolyte amount and then also became independent of the electrolyte volume. This is attributable to the fact that once all electrode pores are filled and the separator is fully soaked, additional electrolyte only accumulates in the free volume of the can and does not contribute to ionic transport inside the electrode network.
From these results, I selected an electrolyte addition that is near the saturation point but does not create excessive free liquid. For a typical small-format cell, this value was about 1.5–1.6 g per ampere-hour of capacity. In the final large 20 Ah cells, the electrolyte amount was set by the same saturation criterion. This criterion maintains high capacity, low internal resistance, and reproducible outcomes in mass production.
| Electrolyte state | Relative discharge capacity | Relative internal resistance | Production issues |
|---|---|---|---|
| Deficient | Low, many under-capacity cells | High | Poor cycle life, possibility of dry separator |
| Near-optimum | High and stable | Low | Good processability |
| Excessive | No significant increase | Slight/no improvement | Leakage during open formation, corrosion, high cost |
3.6 Formation Schedule and SEI Growth
After electrolyte filling, the cell must be aged before the first charge. Aging allows the electrolyte to penetrate into the micropores of the electrodes and separator. If the cell is charged too quickly before the electrolyte has fully wetted the electrode stack, uneven current density and lithium plating can occur. The formation process consists of the first charge and discharge cycles that activate the electrode materials and establish a stable solid-electrolyte interphase on the graphite surface. This SEI layer is critical because it is electronically insulating but ionically conductive. It prevents continued decomposition of the carbonate solvent and permits reversible lithium-ion intercalation. In my investigation, I compared three formation schedules for the vehicle traction battery cells.
The first schedule was a single-stage charge at a moderate current to the final voltage. The second schedule began with a small pre-charge current for a limited time and then continued at a moderate current to the cut-off voltage. The third schedule used a small pre-charge current for a longer time before switching to the moderate current. I observed that the third schedule produced the most uniform activation of the active material. It also allowed the carbon dioxide, ethylene, and other gases generated during SEI formation to escape through the open filling hole more gradually. This reduced internal pressure and prevented the electrode stack from being separated by gas bubbles.
The first charge of a graphite/LiMn2O4 cell includes electrolyte reduction reactions at the graphite surface. A simplified representation of ethylene carbonate reduction is
$$2C_3H_4O_3 + 2e^- + 2Li^+ \rightarrow (CH_2OCO_2Li)_2 + C_2H_4.$$
The gaseous ethylene is produced at potentials close to the SEI formation region. If the charging current is too high, SEI nucleation occurs too rapidly and the resulting film is inhomogeneous. The low pre-charge current is therefore beneficial because it provides a slower, more uniform SEI growth. After this initial low-current stage, the cell can withstand a higher charging current without excessive gas generation.
I also found that the cells formed with an optimized low-current pre-charge had a higher first-cycle charge/discharge efficiency, a more reproducible internal resistance, and a slightly better capacity retention after cycling. The following table summarizes the comparison of three formation concepts.
| Formation concept | First-cycle charge regime | First-cycle efficiency | Capacity distribution | Gas removal |
|---|---|---|---|---|
| A: Single stage | Moderate current directly to full voltage | Lower | Wider spread; some cells under-capacity | Rapid gassing can disturb stack |
| B: Short low-current precharge | Small current for short time, then moderate current | Good | Narrower spread | Moderate control of gassing |
| C: Extended low-current precharge | Small current for longer time, then moderate current | Highest | Narrow and near maximum | Best control of gassing and SEI uniformity |
Based on these data, the formation process used for the large vehicle traction battery cell was designed as follows: charge the cell at a very small current during the first part of the process, then switch to a somewhat higher, still moderate current until the cell reaches the upper voltage limit. After charging, the cell is allowed to stand, and the gas generated during the initial cycles is removed through the open vent before final sealing. This method substantially improves the activation of the LiMn2O4 cathode and the graphite anode while avoiding gas-induced deformation.
4. Design and Manufacture of a Large-Format Prismatic Cell
After optimizing the electrode composition and cell manufacturing parameters, I applied the knowledge to a large-format prismatic lithium-ion cell with a nominal capacity of 20 Ah. The design target was a practical vehicle traction battery unit cell that could serve in an electric-drive vehicle pack. The large cell required special attention to current collection, stack geometry, thermal management, and mechanical strength. Because the cell is much larger than the small-format cells used in the first phase of the study, the current path is longer, and the current density distribution may be less uniform. To solve this problem, I used a multi-tab or wide-tab configuration and selected aluminum and nickel tabs with low electrical resistance.
The cell design can be approached with simple capacity calculations. The required design capacity \(C_{\mathrm{des}}\) was chosen slightly above the nominal capacity \(C_{\mathrm{rated}}\), for example \(C_{\mathrm{des}} = 1.05\,C_{\mathrm{rated}}\). The amount of positive active material required is
$$m_{+} = \frac{C_{\mathrm{des}}}{q_{+,\mathrm{actual}}},$$
where \(q_{+,\mathrm{actual}}\) is the achievable specific capacity of LiMn2O4 under the desired discharge rate. The amount of negative active material is then
$$m_{-} = \frac{\alpha\, C_{\mathrm{des}}}{q_{-,\mathrm{actual}}},$$
where \(\alpha\) is the capacity balance coefficient. The areal loading and electrode dimensions were calculated from the available cell volume and the number of electrode pairs. After coating and calendering, I verified the thickness of the completed electrode stack. The stack also had to account for graphite expansion during lithium intercalation. In fully lithiated graphite, lithium occupies between the graphene layers, and the interlayer spacing increases from approximately 0.335 nm in unlitiated graphite to about 0.372 nm in LiC6. The linear expansion of an individual graphite particle is therefore
$$\epsilon_{\mathrm{graphite}} = \frac{0.372 – 0.335}{0.335} \times 100\% \approx 11\%.$$
Not all of this particle expansion is transferred to the electrode thickness because pores accommodate part of the volume change. Still, the stack design included an allowance of a few percent for electrode expansion. In the steel case, the assembly tightness was chosen to keep the electrodes and separators in good interfacial contact while leaving enough space to avoid excessive compression during full charge.
4.1 Electrode Preparation for the 20 Ah Cell
I prepared the positive electrode by dispersing LiMn2O4, conductive carbon black, and PVDF in NMP using the optimized formula. The slurry was coated continuously onto aluminum foil, dried, and calendered to the target compaction density. The negative electrode was prepared from graphite, conductive carbon black, and water-based binder. The coated negative electrode was densified to its own optimum density. After calendering, each electrode sheet was cut into strips of designed width and length. Aluminum and nickel tabs were welded for positive and negative contacts. The electrodes and separators were then wound into a prismatic jellyroll and inserted into the steel case.
The quality-control criteria for the large cell are listed below. These criteria are general in nature but are extremely important for consistency in vehicle traction battery production.
| Process step | Control target |
|---|---|
| Positive slurry mixing | Uniform dispersion without agglomerates; suitable viscosity; no visible gel particles |
| Negative slurry mixing | Stable water-based slurry; no bubbles; no sedimentation |
| Coating | Uniform areal density; no pinholes; no edge burrs; correct coating width |
| Calendering | Target compaction density; uniform thickness; no cracks |
| Slitting/cutting | No burrs; flush edges; no coating detachment |
| Winding | Even tension; electrodes aligned; separator covers negative edge |
| Assembly | Correct stack tightness; no damage to separator |
| Drying | Moisture content below acceptable limit |
| Electrolyte filling | Saturated electrode stack; no leakage |
| Formation | Controlled SEI growth; gas released before final seal |
4.2 Capacity and Internal Resistance of the Large Cell
The large prismatic prismatic cell was first charged using a constant-current constant-voltage procedure. Constant current was applied to 4.2 V, and the voltage was then held until the current declined to a low cut-off value. After a rest period, the cell was discharged at a low rate. The first discharge capacity reached a value close to the nominal rating. The first-cycle charge/discharge efficiency was high, showing that the optimized formation process generated a stable SEI layer and that the lithium inventory was effectively preserved.
I measured the internal resistance of the fully charged cell after a rest period. The cell exhibited a low direct-current internal resistance, well below the practical limit specified for a large vehicle traction battery. A low internal resistance is important because it minimizes Joule heating during acceleration and improves the usable power of the battery pack.
4.3 Rate Capability and High-Power Discharge
Electric vehicles often require a battery to discharge at 1C to 3C for accelerating and at even higher rates under special conditions. I therefore measured the discharge capacity of the 20 Ah vehicle traction battery at different constant current rates. The performance is represented in normalized form below, where a 100% value denotes the rated capacity of the cell.
| Discharge rate | Typical application | Observed normalized discharge capacity |
|---|---|---|
| 0.2C | Range test, steady cruising | Near or above nominal |
| 0.5C | Normal urban driving | Approximately nominal |
| 1C | Acceleration and hill climbing | A small reduction from nominal; still acceptable |
| High current | Short power pulse or emergency peak | Satisfactory; demonstrates cell can deliver high power without sudden voltage collapse |
In high-rate discharge, the cell voltage drops immediately due to the ohmic resistance and then continues to decrease as lithium-ion concentration gradients establish inside the porous electrodes. The optimized compaction density and electrolyte volume produced a sufficiently rapid supply of lithium ions to the surface of the LiMn2O4 particles. The cell maintained a stable discharge plateau and did not show excessive polarization. The measured high-rate discharge capability was consistent with the design target for use in a vehicle traction battery.
4.4 Self-Discharge and Capacity Recovery
Self-discharge is particularly important for a vehicle traction battery that is parked for days or weeks. I determined the self-discharge rate by storing the fully charged cell in an open-circuit condition at room temperature for 30 days and then discharging it at a moderate rate. The retained capacity was expressed as a percentage of the rated capacity. The cell lost only a fraction of its capacity during this storage period. Afterward, the same cell was subjected to a standard charge/discharge cycle to evaluate capacity recovery. The recovered capacity remained high, meaning that the lithium inventory was not destroyed by storage and that the SEI was effective in suppressing electrolyte reduction.
| Storage parameter | Value/observation |
|---|---|
| Storage temperature | Room temperature |
| Storage time | Approximately 30 days |
| Self-discharge | Low; average per day below a small threshold |
| Subsequent capacity recovery | High enough to meet vehicle traction battery protocol |
These results reassure that the large-format LiMn2O4/graphite cell can withstand real-world idle periods without significant damage.
4.5 Cycle Life
Cycle life is one of the most important parameters of a vehicle traction battery. If the battery must be replaced frequently, the economic and environmental benefits of electric vehicles are reduced. I tested the large cell under repeated discharges at a rate representative of driving cycles. The discharge capacity decreased slowly during the first hundreds of cycles. After 300 cycles, the capacity retention was still high enough to satisfy the relevant electric vehicle standard. The excellent cycle life is partly due to the stable spinel structure of LiMn2O4 and partly due to the optimized formation process, which created a stable interface on the graphite negative electrode.
The capacity fade during cycling can be expressed as
$$\mathrm{Retention} = \frac{C_N}{C_1} \times 100\%,$$
where \(C_N\) is the discharge capacity at cycle \(N\), and \(C_1\) is the initial discharge capacity after formation. In my tests, \(C_{300}/C_1\) was sufficiently high for vehicle traction battery service. The low rate of fade was achieved without using complicated surface-modification powders, demonstrating that manufacturing parameters alone can make a significant contribution to lifetime.
4.6 Safety Evaluation
No discussion of vehicle traction battery manufacturing would be complete without safety. Lithium-ion cells contain flammable organic electrolytes and a large amount of stored energy. If a cell is overcharged, short-circuited, punctured, or heated, it may vent, catch fire, or explode. I performed several abuse tests on the large cell, including overcharge, external short-circuit, nail penetration, and heating. All tests were designed to detect the worst failure modes: fire and explosion.
| Abuse test | Typical condition | Required result | Observed result |
|---|---|---|---|
| Overcharge | Charge beyond normal voltage with no external protective circuit | No explosion, no fire | No explosion, no fire |
| External short circuit | Low-resistance connection across the terminals | No explosion, no fire | No explosion, no fire |
| Nail penetration | Steel nail penetrates the cell | No explosion, no fire | No explosion, no fire |
| Heating | Cell exposed to elevated temperature in an oven | No explosion, no fire | No explosion, no fire |
The safe behavior under these abuse conditions can be attributed to the intrinsic thermal stability of spinel LiMn2O4. In the fully delithiated state, LiMn2O4 transforms into a more stable spinel framework rather than releasing oxygen as violently as some layered oxides do at high temperature. In addition, the optimized capacity balance coefficient prevents lithium plating under most conditions. The positive electrode compaction density and electrolyte amount were selected so that heat and gas generation would not lead to a self-accelerating reaction.
5. Electrochemical Characterization: Mathematical and Physical Meaning
The performance of a vehicle traction battery can be quantitatively assessed with simple equations. The state of charge is
$$\mathrm{SOC} = \frac{Q_{\mathrm{remaining}}}{Q_{\mathrm{rated}}} \times 100\%,$$
and the state of health can be expressed as
$$\mathrm{SOH} = \frac{Q_{\mathrm{actual}}}{Q_{\mathrm{rated}}} \times 100\%.$$
During discharge, the cell voltage is
$$V_{\mathrm{cell}} = E_{\mathrm{ocv}} – I R_{\mathrm{int}} – \eta_{\mathrm{conc}} – \eta_{\mathrm{act}},$$
where \(E_{\mathrm{ocv}}\) is the open-circuit voltage, \(R_{\mathrm{int}}\) is the internal resistance, and \(\eta_{\mathrm{conc}}\) and \(\eta_{\mathrm{act}}\) are the concentration and activation overpotentials. The internal resistance of a large cell is not constant; it depends on temperature, state of charge, and current direction. At high discharge rates, the ohmic drop \(I R_{\mathrm{int}}\) and the concentration overpotential increase, causing the observed discharge voltage to decrease. This is why the optimized electrode porosity and sufficient electrolyte addition are essential for high-power vehicle traction battery applications.
The specific energy of the large cell can be estimated from
$$E_{\mathrm{spec}} = \frac{\bar{V}\, C_{\mathrm{dis}}}{m_{\mathrm{cell}}},$$
where \(\bar{V}\) is the average discharge voltage, \(C_{\mathrm{dis}}\) is the delivered capacity in ampere-hours, and \(m_{\mathrm{cell}}\) is the cell mass. Similarly, the specific power at a given rate is approximately
$$P_{\mathrm{spec}} = \frac{V_{\mathrm{dis}} I_{\mathrm{dis}}}{m_{\mathrm{cell}}}.$$
By combining a high discharge voltage with a low internal resistance and an optimized electrode structure, the LiMn2O4/graphite cell reached specific energy and specific power values suitable for prototype electric vehicle use.
6. Conclusions and Outlook
In this work I systematically investigated the key manufacturing parameters of a lithium-ion vehicle traction battery based on spinel LiMn2O4 as the positive active material and graphite as the negative active material. The main conclusions from my experiments are listed below.
First, the particle size and purity of LiMn2O4 must be chosen according to the processing and performance requirements of the vehicle traction battery. A particle size of about 5–15 μm is a practical compromise among packing density, slurry coatability, rate capability, and cycle stability.
Second, the positive electrode formulation has an important effect on discharge capacity, internal resistance, and cycle retention. In my tests, an active-material fraction of about 91 wt% gave the best overall performance. This composition ensures that enough conductive carbon is present to form electron pathways and enough binder is present to provide robust adhesion to the aluminum current collector.
Third, the negative-to-positive capacity balance coefficient must be designed with process variation in mind. A coefficient in the range of 1.06 to 1.12 prevents lithium plating during normal charging and provides a high usable discharge capacity with acceptable safety. The exact coefficient should be chosen after measuring the real coating-weight variation in the production line.
Fourth, electrode compaction density strongly affects both manufacturability and cell performance. The LiMn2O4 positive electrode should be calendered to a density that balances electronic contact and electrolyte wettability. An excessive compaction density worsens high-rate behavior, while a too-low density leads to assembly problems and low electrode conductivity. The optimum positive-compaction density in this system was found near the upper range suitable for spinel LiMn2O4, while the graphite negative electrode required a lower density to preserve lithium-ion diffusion paths.
Fifth, the electrolyte amount should be chosen to saturate the electrode stack while avoiding excessive free liquid. The optimum addition results in a maximum discharge capacity and a minimum stable internal resistance. Cells filled with an insufficient amount of electrolyte are more likely to fail during cycle-life testing.
Sixth, the formation protocol is critical to producing a good vehicle traction battery. A low-current precharge step, followed by a moderate charging current to the voltage limit, produces a uniform SEI layer, reduces gas-induced damage, improves first-cycle efficiency, and narrows the capacity distribution of the final products.
Finally, the optimized manufacturing process was successfully scaled to a 20 Ah prismatic cell. The cell has high initial discharge capacity, low internal resistance, good high-rate discharge capability, low self-discharge, strong capacity recovery after storage, satisfactory cycle life, and acceptable safety in overcharge, short-circuit, nail penetration, and heating tests. This demonstrates that LiMn2O4-based lithium-ion technology is not only a low-cost research material but also a viable candidate for industrial vehicle traction battery production.
In future work, I believe that additional improvements can be made by using surface-coated LiMn2O4 particles, by optimizing the electrolyte additives for high-temperature operation, by increasing the electrode areal loading to further raise energy density, and by designing integrated battery-pack management systems for vehicle traction battery modules. The challenge of manufacturing a reliable, affordable, and safe vehicle traction battery remains multidisciplinary. My research shows that process optimization is just as important as material innovation in bridging the gap between laboratory cells and commercial vehicle applications.