Electric vehicle drivetrains have moved rapidly toward integrated motor, reducer, and inverter architectures. In this configuration, the electric vehicle transmission fluid is no longer only a gear lubricant. It is simultaneously a dielectric medium, a thermal-management fluid, a bearing lubricant, a copper-compatibility medium, and a material-contact fluid. I therefore treated electrical conductivity as a primary formulation variable rather than as a secondary property. I selected three synthetic ester base oils because esters combine high flash point, low volatility, wide operating-temperature range, strong boundary-film formation, and favorable material compatibility. The same polar ester groups that support lubrication can, however, increase ionic conduction. My central objective was to determine how antioxidants, dispersants, corrosion inhibitors, and friction-reducing anti-wear agents shift the conductivity of ester base oils, and then to use that knowledge to formulate electric vehicle transmission fluids with low conductivity, high breakdown strength, good heat transfer, and low friction and wear.
Keywords: electric vehicle; synthetic ester; electric vehicle transmission fluid; electrical conductivity; breakdown voltage; thermal management; tribology; additive response; electric vehicle drivetrain.
1. Electric vehicle transmission fluid requirements
An electric vehicle transmission fluid operates under conditions that differ from those of a conventional automatic transmission fluid. The electric vehicle motor may rotate at very high speed, the inverter may impose fast switching transients, and the reducer and bearings may see high load and high temperature in a compact sump. The fluid must therefore satisfy several requirements at the same time. I summarized these requirements in Table 1 and used them as the screening framework for the ester formulations.
| Requirement | Why it matters in an electric vehicle | Target direction | Measured or inferred metric |
|---|---|---|---|
| Electrical insulation | High-voltage electric vehicle hardware requires low leakage and stable dielectric strength | High breakdown voltage, low conductivity | Breakdown voltage in kV; conductivity in pS/m |
| Thermal management | Integrated electric vehicle motor and reducer generate concentrated heat | High thermal conductivity, suitable specific heat | Thermal conductivity in W/(m·K); specific heat in J/(g·K) |
| Lubrication | Gears and bearings in the electric vehicle reducer require boundary and mixed-film protection | Low friction coefficient, low wear volume | COF; wear volume; wear depth |
| Oxidation stability | Electric vehicle fluids can experience high sump temperature and long service intervals | High oxidation onset temperature | Initial oxidation temperature in °C |
| Material compatibility | Copper, bearing steel, and elastomers are present in the electric vehicle drive unit | Low copper corrosion, stable seals | Copper strip corrosion grade |
| Additive compatibility | Additives must not destroy the dielectric balance of the electric vehicle fluid | Low-conductivity additive selection | Relative conductivity increase |
The electric vehicle context makes the conductivity requirement especially important. A fluid with excellent friction performance but excessive ionic conduction can compromise insulation safety. Conversely, a fluid with very low conductivity but poor boundary lubrication can increase electric vehicle gear and bearing wear. I therefore adopted a coupled electrical-thermal-tribological screening strategy rather than optimizing any single property in isolation.
2. Base oils and additive screening design
I selected three synthetic ester base oils with different molecular polarity and viscosity. Their code names are A51, RM1220, and NP451. A51 is an adipic acid di(tridecyl) ester with a linear symmetrical diester structure. RM1220 is a trimethylolpropane caprylate-caprate ester with a triester structure. NP451 is a pentaerythritol ester with multiple ester groups and hydroxyl functionality. The basic physical and chemical properties are listed in Table 2.
| Base oil code | Chemistry | Kinematic viscosity at 40 °C (mm²/s) | Kinematic viscosity at 100 °C (mm²/s) | Viscosity index | Density at 15 °C (g/cm³) | Pour point (°C) | Flash point (°C) | Polarity description |
|---|---|---|---|---|---|---|---|---|
| A51 | Adipic acid di(tridecyl) ester | 26.95 | 5.386 | 139 | 0.915 | -57 | 247 | Two ester groups; lowest polarity among the three |
| RM1220 | Trimethylolpropane caprylate-caprate ester | 19.64 | 4.417 | 140 | 0.948 | -51 | 262 | Three ester groups; intermediate polarity |
| NP451 | Pentaerythritol ester | 24.11 | 4.982 | 136 | 0.993 | -60 | 255 | Multiple ester and hydroxyl groups; highest polarity |
I screened four additive classes: dispersants, antioxidants, corrosion inhibitors, and friction-reducing anti-wear agents. The additive codes, chemistry, total base number, and screening dosage are listed in Table 3. The dispersants were selected to cover a range of total base number and steric structure. The antioxidants were selected to compare an alkyl-substituted diphenylamine with a naphthylamine. The corrosion inhibitors were selected to compare a benzotriazole derivative with an alkenyl succinic acid. The friction-reducing and anti-wear agents were selected to compare ionic phosphate ester ammonium salt, molecular phosphate esters, and a metal-organic dithiophosphate.
| Class | Code | Chemistry | Total base number (mg KOH/g) | Screening dosage (mass fraction) | Expected electrical response |
|---|---|---|---|---|---|
| Dispersant | T151 | Monoalkenyl succinimide | 45 to 55 | 3.0% | High base number and low steric hindrance; strong conductivity increase expected |
| Dispersant | T152 | Bis(alkenyl) succinimide | 15 to 30 | 3.0% | Intermediate base number; moderate conductivity increase expected |
| Dispersant | T154 | Polyisobutylene succinimide | 15 to 30 | 3.0% | High steric hindrance; lower conductivity increase expected |
| Antioxidant | L57 | Liquid octyl/butyl diphenylamine | Not applicable | 1.0% | Alkyl substitution improves ester compatibility and electron density |
| Antioxidant | T531 | N-phenyl-alpha-naphthylamine | Not applicable | 1.0% | Compact aromatic structure; lower compatibility and lower conductivity increase |
| Corrosion inhibitor | 7001 | Benzotriazole and its derivatives | Not applicable | 0.1% | Weak polarity; low conductivity increase expected |
| Corrosion inhibitor | T746 | Dodecenyl succinic acid | Not applicable | 0.1% | Carboxylic acid group; stronger conductivity increase expected |
| Friction-reducing and anti-wear agent | AW319 | Phosphate ester ammonium salt | Not applicable | 1.0% | Ionic character; strong conductivity increase expected |
| Friction-reducing and anti-wear agent | T306 | Trimethylphenyl phosphate | Not applicable | 1.0% | Molecular phosphate ester; moderate response expected |
| Friction-reducing and anti-wear agent | T203 | Zinc bis(2-ethylhexyl) dithiophosphate | Not applicable | 1.0% | Metal-organic character; moderate response expected |
| Friction-reducing and anti-wear agent | H059 | Alkyl phosphate | Not applicable | 1.0% | Predominantly molecular in ester; low conductivity increase expected |
Each screening sample contained only one base oil and one additive. I used the naming convention x additive code-base oil code, where x is the additive mass fraction. For example, a blend containing 1.0% L57 and 99.0% NP451 was designated 1.0% L57-NP451. The additive mass fraction was calculated as
$$w_i=\frac{m_i}{m_{\mathrm{total}}}\times 100\%$$
where mi is the mass of additive i and mtotal is the total blend mass. For the final electric vehicle transmission fluids, I used the low-conductivity additive response principle. I selected T154 at 3.0%, T531 at 1.0%, 7001 at 0.1%, and H059 at 1.0% and added them to A51, RM1220, and NP451. The resulting fluids were designated A51-C, RM1220-C, and NP451-C. I also tested a commercial electric vehicle transmission fluid, designated ETF, as the reference.
3. Measurement methods and data treatment
I measured electrical conductivity according to the liquid hydrocarbon conductivity method. The conductivity cell covered the range from 0.01 to 200000 pS/m, where 1 pS/m equals 10-12 S/m. I heated each sample to 150 °C, stopped heating, and then recorded conductivity at 5 °C intervals as the sample cooled to 30 °C. The conductivity is related to measured conductance and cell constant by
$$\sigma=G\,K_{\mathrm{cell}}$$
where G is conductance and Kcell is the cell constant. I used the temperature-dependent conductivity to compare base oils and additive responses. For a simple thermally activated conduction process, I used the Arrhenius form
$$\sigma(T)=\sigma_0\exp\left(-\frac{E_a}{k_B T}\right)$$
where σ0 is a pre-exponential factor, Ea is an apparent activation energy for conduction, kB is the Boltzmann constant, and T is absolute temperature. A plot of ln σ against 1/T gives a slope of –Ea/kB. I used this relation only as a comparative guide because additive-containing ester fluids can contain multiple charge carriers and cannot be treated as ideal dilute electrolytes. I also defined the relative conductivity increase produced by an additive as
$$R_{\sigma}=\frac{\sigma_{\mathrm{add}}-\sigma_{\mathrm{base}}}{\sigma_{\mathrm{base}}}\times 100\%$$
where σadd is the conductivity of the additive-containing oil and σbase is the conductivity of the corresponding base oil at the same temperature. This normalized quantity allowed me to compare additives across base oils with different absolute conductivity.
I measured breakdown voltage at 25 °C with a fully automatic insulating-oil breakdown tester. After a 3 min rest, the sample was stirred and then energized at 2.0 kV/s until breakdown. The instrument recorded six breakdown events and calculated the mean. The arithmetic mean and standard deviation were
$$\bar{V}_B=\frac{1}{N}\sum_{i=1}^{N}V_{B,i}$$
$$s_B=\sqrt{\frac{1}{N-1}\sum_{i=1}^{N}\left(V_{B,i}-\bar{V}_B\right)^2}$$
where N = 6, VB,i is the i-th breakdown voltage, and sB is the sample standard deviation. The breakdown field, if the electrode gap is known, is
$$E_B=\frac{V_B}{d}$$
where d is the gap distance. I use VB directly in this work because the gap is fixed by the standardized cell.
I measured thermal conductivity by the transient hot-wire method at 20 to 25 °C and used the average of three measurements. The underlying steady-state relation is Fourier’s law,
$$q=-k\nabla T$$
where q is heat flux, k is thermal conductivity, and ∇T is the temperature gradient. I measured specific heat at 20 °C by differential scanning calorimetry under nitrogen at a heating and cooling rate of 10 K/min from 10 to 30 °C. The specific heat was evaluated from
$$c_p=\frac{\Delta H}{m\,\Delta T}$$
where ΔH is the enthalpy change, m is sample mass, and ΔT is the temperature interval. I measured oxidation onset temperature by high-pressure differential scanning calorimetry under 3.5 MPa oxygen at a flow of 100 mL/min and a heating rate of 10 K/min. The oxidation onset temperature was taken as the extrapolated onset of the exothermic oxidation signal.
I measured tribological performance with a reciprocating ball-on-disc tribometer. The ball diameter was 10 mm, the disc diameter was 24 mm, and the disc thickness was 7.9 mm. Both ball and disc were AISI 52100 bearing steel with hardness HRC 60 ± 2 and root-mean-square roughness of about 20 nm. The test conditions were 100 N load, 25 Hz frequency, 150 °C, 1 mm stroke, and 60 min duration. The friction coefficient was calculated as
$$\mu(t)=\frac{F_t(t)}{F_n}$$
where Ft(t) is the instantaneous tangential friction force and Fn is the normal load. I averaged the friction coefficient over the steady-state period after the initial running-in stage. I measured disc wear volume by three-dimensional profilometry and defined the wear rate as
$$w_r=\frac{V_w}{F_n s}$$
where Vw is wear volume and s is total sliding distance. The maximum wear depth was taken directly from the wear-track profile.
4. Base oil conductivity and temperature response
I observed that the conductivity of all three synthetic esters increased with temperature. The order was NP451 > RM1220 > A51. NP451 showed the highest conductivity and the steepest rise with temperature, RM1220 was intermediate, and A51 was the lowest. This order follows the molecular polarity of the base oils. NP451 contains multiple ester groups and hydroxyl functionality, so its molecular polarity is the highest and its ability to support charge separation is the greatest. RM1220 contains fewer ester groups than NP451 and is therefore less polar. A51 contains two ester groups in a linear symmetrical diester structure and has the lowest polarity and the lowest conductivity. I summarized the base-oil polarity and conductivity ranking in Table 4.
| Base oil | Polar groups | Relative molecular polarity | Conductivity ranking | Temperature-response ranking | Interpretation |
|---|---|---|---|---|---|
| A51 | Two ester groups | Low | 3, lowest | Lowest slope | Linear symmetrical diester; weakest ion solvation and charge separation |
| RM1220 | Three ester groups | Intermediate | 2, intermediate | Intermediate slope | Trieste structure; moderate polar-group density |
| NP451 | Multiple ester and hydroxyl groups | High | 1, highest | Highest slope | Highest polar-group density; strongest conductivity increase with temperature |
The temperature dependence can be understood through both carrier generation and carrier mobility. As temperature rises, ester molecules gain thermal energy, viscosity falls, and ionic species can migrate more easily. I used a simplified Nernst-Einstein relation to describe the contribution of ionic carriers to conductivity,
$$\sigma=\sum_i \frac{n_i q_i^2 D_i}{k_B T}$$
where ni is the number density of carrier i, qi is its charge, and Di is its diffusion coefficient. The diffusion coefficient is related to viscosity by the Stokes-Einstein relation,
$$D_i=\frac{k_B T}{6\pi\eta r_i}$$
where η is dynamic viscosity and ri is the effective hydrodynamic radius. Combining these relations gives
$$\sigma=\sum_i \frac{n_i q_i^2}{6\pi\eta r_i}$$
This expression shows why the conductivity of an electric vehicle transmission fluid is sensitive to both polar-group concentration and viscosity. A more polar ester can support a higher carrier concentration, while a lower-viscosity fluid can increase carrier mobility. Base-oil selection therefore sets the initial conductivity floor, and additive selection determines how far the final electric vehicle transmission fluid moves above that floor.
5. Dispersant effects on conductivity
Among the four additive classes, dispersants produced the largest increase in conductivity. For all three base oils, T151 gave the highest conductivity, T152 was intermediate, and T154 gave the lowest conductivity among the dispersants. This ranking is consistent with the combination of total base number, molecular structure, and steric hindrance. T151 is a monoalkenyl succinimide with a total base number of 45 to 55 mg KOH/g and relatively low steric hindrance. Its polar imide groups and high base number increase the concentration of ionizable species and reduce the resistance to ion migration. T154 is a polyisobutylene succinimide with a total base number of 15 to 30 mg KOH/g and high steric hindrance. The bulky polyisobutylene chain suppresses molecular diffusion and ion transport, so its conductivity increase is lower. T152 is a bis(alkenyl) succinimide with an intermediate structure and an intermediate effect. I summarized the dispersant response in Table 5.
| Dispersant | Total base number (mg KOH/g) | Steric hindrance | Polar character | Conductivity impact | Mechanistic interpretation |
|---|---|---|---|---|---|
| T151 | 45 to 55 | Low | High | Highest | High base number, low steric hindrance, strong ionizable imide contribution |
| T152 | 15 to 30 | Medium | Medium to high | Intermediate | Bis-succinimide structure with moderate diffusion resistance |
| T154 | 15 to 30 | High | Medium | Lowest among dispersants | Bulky polyisobutylene chain suppresses ion mobility |
The dispersant results also reveal a base-oil interaction. Ester base oils are themselves polar, and the polar groups of the dispersant can interact with ester groups. This interaction can disrupt the local packing of ester molecules and create additional free volume, which lowers the barrier to ion migration. The effect is not simply additive. A high-base-number dispersant in a highly polar ester such as NP451 can produce a larger absolute conductivity than the same dispersant in A51, even when the relative ranking of dispersants remains the same. For electric vehicle transmission fluid design, this means that the dispersant should be selected not only for soot and sludge control but also for its electrical signature. T154 at 3.0% was the best compromise because it retained dispersancy while keeping the conductivity increase below that of T151 and T152.
I used the following simple ion-dissociation expression to describe the dispersant contribution:
$$\mathrm{AX}\rightleftharpoons \mathrm{A^+}+\mathrm{X^-}$$
$$\sigma \propto \alpha c \lambda$$
where α is the degree of dissociation, c is the nominal carrier concentration, and λ is the molar conductivity. A higher total base number can increase c and α, while a bulkier molecular structure lowers λ by reducing ion mobility. The dispersant data are consistent with this balance: T151 has the strongest carrier-generation term, while T154 has the strongest mobility-suppression term.
6. Antioxidant effects on conductivity
Antioxidants had a smaller effect on conductivity than dispersants, but the difference between L57 and T531 was measurable. L57 produced a larger conductivity increase than T531 in all three ester base oils. I attribute this to molecular structure, compatibility, and electron density. L57 is a liquid octyl/butyl diphenylamine. The alkyl substituents lower the overall polarity of the molecule and improve its compatibility with the weakly polar ester medium. Better compatibility allows L57 to disperse more uniformly, which can increase the effective concentration of polar nitrogen-containing species in the conduction pathway. The alkyl groups also act as electron-donating substituents and increase the electron density around the nitrogen atom. A higher electron density increases the tendency of the amine to undergo protonation,
$$\mathrm{R_3N}+\mathrm{H^+}\rightleftharpoons \mathrm{R_3NH^+}$$
and the resulting ionic species can contribute to conductivity. T531 is N-phenyl-alpha-naphthylamine. Its naphthalene ring and single phenyl group give a more compact and rigid molecular structure. Its compatibility with the ester is lower than that of L57, and its nitrogen electron density is also lower. The lower electron density weakens the protonation tendency and reduces the ionic contribution. I summarized the antioxidant response in Table 6.
| Antioxidant | Chemistry | Compatibility with ester | Electron density at nitrogen | Protonation tendency | Conductivity impact |
|---|---|---|---|---|---|
| L57 | Liquid octyl/butyl diphenylamine | High due to alkyl substitution | High | Higher | Larger increase |
| T531 | N-phenyl-alpha-naphthylamine | Lower due to compact aromatic structure | Lower | Lower | Smaller increase |
The antioxidant result is important for electric vehicle transmission fluids because oxidation stability is essential in a high-temperature electric vehicle drive unit. I did not want to remove the antioxidant simply to lower conductivity. Instead, I selected T531 at 1.0% for the final formulations because it provided oxidation protection with a lower conductivity penalty than L57. The final oxidation onset temperatures confirmed that this choice was effective.
7. Corrosion inhibitor effects on conductivity
Corrosion inhibitors had a relatively small effect on conductivity, but T746 produced a larger increase than 7001. T746 is dodecenyl succinic acid, which contains a carboxylic acid group. The carboxylic acid can dissociate into a carboxylate anion and a proton,
$$\mathrm{R-COOH}\rightleftharpoons \mathrm{R-COO^-}+\mathrm{H^+}$$
and the resulting ions increase the conductivity of the ester. 7001 is a benzotriazole derivative with weaker polarity. It interacts with copper surfaces through heterocyclic nitrogen atoms rather than through strong acid-base dissociation, so its contribution to ionic conduction is lower. I summarized the corrosion inhibitor response in Table 7.
| Corrosion inhibitor | Chemistry | Polar group | Ionization tendency | Conductivity impact | Copper-protection role |
|---|---|---|---|---|---|
| 7001 | Benzotriazole derivative | Heterocyclic nitrogen | Low | Lower | Forms protective copper surface complex |
| T746 | Dodecenyl succinic acid | Carboxylic acid | Higher | Higher | Acid-based corrosion inhibition with higher ionic contribution |
For the final electric vehicle transmission fluid, I selected 7001 at 0.1%. This choice kept the copper strip corrosion grade at 1b while avoiding the stronger conductivity increase associated with T746. The copper compatibility result is important because electric vehicle drive units contain copper conductors, connectors, and bearing materials that must remain stable over long service.
8. Friction-reducing and anti-wear agent effects on conductivity
The friction-reducing and anti-wear agents produced a range of conductivity responses. AW319 gave the largest increase, especially in NP451 at 150 °C. AW319 is a phosphate ester ammonium salt and is ionic in character. In an electric field, it can dissociate into charged species and increase the ionic carrier concentration. H059 gave the smallest conductivity increase among the anti-wear agents. H059 is an alkyl phosphate that exists predominantly in molecular form in the ester, so only a small fraction of ions is available for conduction. T306 and T203 were intermediate. I summarized the anti-wear response in Table 8.
| Friction-reducing and anti-wear agent | Chemistry | Ionic character in ester | Conductivity impact | Tribological role |
|---|---|---|---|---|
| AW319 | Phosphate ester ammonium salt | High | Highest | Strong boundary film and anti-wear action, but higher ionic conduction |
| T306 | Trimethylphenyl phosphate | Moderate | Intermediate | Molecular phosphate ester; moderate anti-wear |
| T203 | Zinc bis(2-ethylhexyl) dithiophosphate | Moderate | Intermediate | Metal-organic anti-wear; ash-containing chemistry |
| H059 | Alkyl phosphate | Low | Lowest | Molecular phosphate ester; low conductivity penalty |
The anti-wear results show the central trade-off in electric vehicle transmission fluid formulation. A strongly ionic anti-wear agent can improve boundary protection but can also raise conductivity. A molecular anti-wear agent can preserve low conductivity but may require a more carefully selected base oil to achieve low friction. I selected H059 at 1.0% for the final electric vehicle transmission fluids because it provided the lowest conductivity penalty while retaining phosphate-based anti-wear functionality. The final tribological results showed that this choice was compatible with low friction and low wear when combined with the A51 base oil.
9. Additive screening synthesis and formulation rules
The screening results established a clear hierarchy. Dispersants had the strongest influence on conductivity, followed by antioxidants, corrosion inhibitors, and friction-reducing anti-wear agents. Within the dispersant class, higher total base number and lower steric hindrance gave a stronger conductivity increase. Within the antioxidant class, higher compatibility and higher nitrogen electron density gave a stronger increase. Within the corrosion inhibitor class, carboxylic acid functionality gave a stronger increase than benzotriazole functionality. Within the anti-wear class, ionic phosphate ester ammonium salt gave a stronger increase than molecular alkyl phosphate. I summarized the screening hierarchy and final choices in Table 9.
| Additive class | Most conductive candidate | Least conductive candidate | Relative influence on conductivity | Final choice | Final dosage (mass fraction) | Rationale |
|---|---|---|---|---|---|---|
| Dispersant | T151 | T154 | Very high | T154 | 3.0% | Maintains dispersancy with lower base number and higher steric hindrance |
| Antioxidant | L57 | T531 | Moderate | T531 | 1.0% | Lower conductivity penalty with useful oxidation stability |
| Corrosion inhibitor | T746 | 7001 | Low to moderate | 7001 | 0.1% | Copper protection with weaker ionization |
| Friction-reducing and anti-wear agent | AW319 | H059 | Low to moderate | H059 | 1.0% | Molecular phosphate ester with low conductivity penalty |
From these results I derived several formulation rules for synthetic ester electric vehicle transmission fluids. First, the dispersant should be selected at the lowest total base number and highest steric hindrance that still meets cleanliness requirements. Second, the antioxidant should be selected for oxidation onset rather than for maximum radical-scavenging activity alone, because a highly compatible amine can raise conductivity. Third, the corrosion inhibitor should be selected for copper passivation at the lowest ionizable acid content. Fourth, the anti-wear agent should be selected for boundary film formation with the lowest ionic dissociation. Fifth, the base oil should be selected to provide a low conductivity floor while still forming a stable boundary film. A51 met this last requirement best among the three esters because its linear symmetrical diester structure gave the lowest polarity and the best friction result.
10. Physical properties of the formulated electric vehicle transmission fluids
I formulated three ester-based electric vehicle transmission fluids, A51-C, RM1220-C, and NP451-C, and compared them with a commercial electric vehicle transmission fluid, ETF. The physical properties are listed in Table 10. The ester fluids had similar kinematic viscosity to the commercial reference, but their viscosity indices were lower. A51-C had the highest 40 °C viscosity and the highest 100 °C viscosity among the three ester fluids. All four fluids had low pour points and high flash points, and all four passed the copper strip corrosion test at grade 1b.
| Property | A51-C | RM1220-C | NP451-C | ETF |
|---|---|---|---|---|
| Kinematic viscosity at 40 °C (mm²/s) | 30.71 | 23.77 | 26.82 | 25.51 |
| Kinematic viscosity at 100 °C (mm²/s) | 5.896 | 5.062 | 5.485 | 5.768 |
| Viscosity index | 139 | 146 | 147 | 180 |
| Density at 15 °C (g/cm³) | 0.914 | 0.946 | 0.990 | 0.850 |
| Pour point (°C) | -51 | -45 | -54 | -48 |
| Open flash point (°C) | 256 | 264 | 248 | 196 |
| Copper strip corrosion at 150 °C for 3 h | 1b | 1b | 1b | 1b |
The higher viscosity index of the commercial ETF indicates a more shear-stable viscosity modifier system, but the ester fluids had higher flash points and comparable low-temperature fluidity. For electric vehicle applications, the higher flash point is desirable for thermal safety, while the lower viscosity index can be compensated by formulation with a viscosity modifier if needed. I kept the ester fluids free of additional viscosity modifier in this study so that the base-oil and additive effects on conductivity and tribology could be isolated.
11. Electrical performance of the electric vehicle transmission fluids
I measured breakdown voltage after dehydrating the samples by rotary evaporation at 80 °C for 30 min. Moisture content was measured by Karl Fischer coulometric titration. The breakdown voltage and moisture results are listed in Table 11. All four fluids exceeded 45 kV. A51-C had the highest breakdown voltage at 62.81 kV, followed by RM1220-C at 59.09 kV, ETF at 56.37 kV, and NP451-C at 49.17 kV. The moisture content of A51-C was 100 ppm, while the other three fluids were 200 ppm. The higher breakdown voltage of A51-C is therefore not simply a moisture effect; it also reflects the lower conductivity and more favorable dielectric response of the A51 base oil with the selected additive package.
| Fluid | Breakdown voltage (kV) | Moisture content (ppm) | Electrical assessment |
|---|---|---|---|
| A51-C | 62.81 | 100 | Highest breakdown strength; best insulation margin |
| RM1220-C | 59.09 | 200 | High breakdown strength; good insulation margin |
| NP451-C | 49.17 | 200 | Above 45 kV; lower margin than the other ester fluids |
| ETF | 56.37 | 200 | Reference fluid; lower than A51-C and RM1220-C |
I also compared the conductivity of the four fluids over the 30 to 150 °C range. All three ester-based electric vehicle transmission fluids had lower conductivity than the commercial ETF at every measured temperature. The conductivity ranking was A51-C < RM1220-C < NP451-C < ETF. This result is significant because it shows that a synthetic ester electric vehicle transmission fluid can be formulated with lower conductivity than a commercial reference while retaining good thermal and tribological performance. The lower conductivity of A51-C is consistent with its lower base-oil polarity and its low-conductivity additive package. The higher conductivity of NP451-C is consistent with the higher polarity of the NP451 base oil, even though the same low-conductivity additives were used.
For dielectric design, the breakdown voltage and conductivity should be considered together. A high breakdown voltage indicates a large safety margin against sudden dielectric failure, while low conductivity indicates low steady-state leakage and low ionic transport. The electric vehicle transmission fluid should satisfy both conditions. I used the following normalized electrical score as a comparative index:
$$S_{\mathrm{elec}}=w_B\frac{\bar{V}_B}{V_{B,\mathrm{ref}}}+w_\sigma\frac{\sigma_{\mathrm{ref}}}{\sigma}$$
where wB and wσ are weighting factors, VB,ref is the reference breakdown voltage, and σref is the reference conductivity. A higher score indicates better electrical performance. Under equal weighting, A51-C had the highest electrical score among the four fluids because it combined the highest breakdown voltage with the lowest conductivity.
12. Thermal conductivity and oxidation stability
The thermal conductivity of all four fluids exceeded 0.15 W/(m·K), and the three ester-based electric vehicle transmission fluids had higher thermal conductivity than the commercial ETF. This is favorable for electric vehicle thermal management because the fluid must remove heat from the motor, reducer, and power electronics. The low viscosity of the ester fluids increases molecular freedom and convective heat transfer, while the polar ester groups can enhance local heat conduction. I summarized the specific heat and oxidation onset results in Table 12.
| Fluid | Thermal conductivity (W/(m·K)) | Specific heat at 20 °C (J/(g·K)) | Initial oxidation temperature (°C) | Thermal assessment |
|---|---|---|---|---|
| A51-C | > 0.15 | 1.8397 | 250.8 | Higher thermal conductivity than ETF; good oxidation stability |
| RM1220-C | > 0.15 | 1.7173 | 264.6 | Higher thermal conductivity than ETF; improved oxidation stability |
| NP451-C | > 0.15 | 1.9571 | 266.7 | Higher thermal conductivity than ETF; highest oxidation onset |
| ETF | > 0.15 | 2.0758 | 247.0 | Reference fluid; lower thermal conductivity and lower oxidation onset |
The specific heat of the ester fluids ranged from 1.7173 to 1.9571 J/(g·K), while the commercial ETF had a higher specific heat of 2.0758 J/(g·K). A higher specific heat is beneficial for thermal buffering, but thermal management also depends on thermal conductivity and viscosity. The ester fluids had higher thermal conductivity than ETF, which can partially compensate for their lower specific heat. The thermal diffusivity is
$$\alpha=\frac{k}{\rho c_p}$$
where α is thermal diffusivity, k is thermal conductivity, ρ is density, and cp is specific heat. Because the ester fluids had higher k and lower density than ETF, their thermal diffusivity can remain competitive even when their specific heat is lower. For electric vehicle thermal management, the combination of high k and low viscosity is especially useful because it supports both conduction and convection.
The oxidation onset temperatures were 250.8 °C for A51-C, 264.6 °C for RM1220-C, 266.7 °C for NP451-C, and 247.0 °C for ETF. All three ester fluids had higher oxidation onset than the commercial reference. NP451-C had the highest oxidation onset, followed by RM1220-C and A51-C. The result is consistent with the selection of T531 as the antioxidant and with the inherent oxidation stability of the ester base oils. Even A51-C, which had the lowest oxidation onset among the ester fluids, still exceeded the commercial ETF by 3.8 °C. This means that the low-conductivity additive package did not compromise oxidation stability relative to the reference electric vehicle transmission fluid.
13. Tribological performance
The friction coefficient of A51-C reached 0.1386, the lowest among the four fluids. The friction ranking was A51-C < ETF < RM1220-C < NP451-C. The wear ranking was A51-C < RM1220-C < NP451-C < ETF. A51-C had a disc wear volume of 659000 µm³ and a maximum wear depth of 8.74 µm. These results show that the A51-C electric vehicle transmission fluid combined the best friction reduction with the best anti-wear performance. I summarized the tribological results in Table 13.
| Fluid | Average friction coefficient | Friction ranking | Disc wear volume (µm³) | Wear ranking | Maximum wear depth (µm) |
|---|---|---|---|---|---|
| A51-C | 0.1386 | 1, lowest | 659000 | 1, lowest | 8.74 |
| RM1220-C | Not reported as a single value | 3 | Higher than A51-C | 2 | Higher than A51-C |
| NP451-C | Not reported as a single value | 4 | Higher than RM1220-C | 3 | Higher than RM1220-C |
| ETF | Not reported as a single value | 2 | Highest | 4 | Highest |
The excellent tribological performance of A51-C can be explained by the molecular structure of the A51 base oil. A51 is an adipic acid di(tridecyl) ester, which is a linear symmetrical diester formed from a diacid and a monohydric alcohol. Its spatial hindrance is low, and its ester bond polarity is moderate. This allows the molecule to spread closely and uniformly on the metal surface and form a boundary lubrication film that is resistant to shear. The linear structure also helps the fluid maintain a stable friction trace after the running-in period. In contrast, NP451 has a higher polar-group density and a more complex molecular structure, which can increase conductivity and may compete with anti-wear additive adsorption at the metal surface. RM1220 is intermediate. The commercial ETF had a higher friction coefficient than A51-C and the highest wear volume among the four fluids under the selected test conditions.
The friction and wear results also show that low conductivity and low friction are not mutually exclusive in ester-based electric vehicle transmission fluids. The key is to select a low-polarity base oil with good boundary-film formation and to pair it with a molecular anti-wear agent such as H059 rather than a strongly ionic anti-wear agent such as AW319. The A51-C formulation achieved this balance. Its low conductivity comes from the low polarity of A51 and the low-conductivity additive package, while its low friction and low wear come from the linear diester structure and the phosphate-based boundary film.
14. Mechanistic interpretation of additive conductivity response
The additive screening results can be interpreted through three coupled mechanisms: carrier generation, carrier mobility, and carrier recombination. Carrier generation depends on the dissociation of polar additives and on acid-base reactions with the ester medium. Carrier mobility depends on viscosity, molecular size, steric hindrance, and free volume. Carrier recombination depends on ion pairing and on the dielectric constant of the medium. I summarize these mechanisms in Table 14.
| Mechanism | Key variables | Additive examples | Effect on electric vehicle transmission fluid conductivity |
|---|---|---|---|
| Carrier generation by dissociation | Ionic character, acid strength, base number, dielectric constant | T151, T746, AW319 | Increases conductivity by producing charged species |
| Carrier generation by protonation | Nitrogen electron density, amine basicity, acid availability | L57, T531 | Increases conductivity through protonated amine species |
| Carrier mobility | Viscosity, molecular size, steric hindrance, free volume | T154, H059 | Lowers conductivity when bulky or molecular |
| Ion pairing and recombination | Dielectric constant, ion size, concentration | Ester base oil, ionic additives | Reduces effective free-ion concentration |
| Interfacial polarization | Polar groups, boundary layers, electric field frequency | Ester base oil, polar additives | Can increase dielectric loss and apparent conductivity under alternating fields |
The ester base oil plays a dual role. Its polar groups can solvate ions and increase carrier concentration, but they can also promote ion pairing and reduce the number of free ions. The net effect depends on the balance between dielectric constant, viscosity, and molecular structure. A51 has the lowest polar-group density and the lowest conductivity. NP451 has the highest polar-group density and the highest conductivity. RM1220 is intermediate. This base-oil ranking was preserved across all additive classes, which means that base-oil selection sets the conductivity floor and additive selection sets the conductivity increment.
I used a simplified conductivity model to organize the data:
$$\sigma=\sigma_{\mathrm{base}}+\Delta\sigma_{\mathrm{disp}}+\Delta\sigma_{\mathrm{ao}}+\Delta\sigma_{\mathrm{ci}}+\Delta\sigma_{\mathrm{aw}}+\Delta\sigma_{\mathrm{cross}}$$
where σbase is the base-oil conductivity, the Δσ terms are the contributions of dispersant, antioxidant, corrosion inhibitor, and anti-wear agent, and Δσcross is a cross-interaction term. The screening data show that Δσdisp is the largest term, followed by Δσao, Δσci, and Δσaw. The cross term is not negligible because polar additives can interact with ester groups and with each other. For electric vehicle transmission fluid design, the cross term should be minimized by selecting additives with compatible polarity and by avoiding high-base-number dispersants when low conductivity is the primary requirement.
15. Multi-objective performance comparison
I compared the four fluids across electrical, thermal, and tribological criteria. A51-C had the best overall balance. It had the highest breakdown voltage, the lowest conductivity, the lowest friction coefficient, the lowest wear volume, and a higher oxidation onset than the commercial ETF. RM1220-C had the second-best balance, with high breakdown voltage, good thermal stability, and intermediate friction and wear. NP451-C had the highest oxidation onset and good thermal conductivity, but its higher base-oil polarity increased conductivity and its tribological performance was lower than that of A51-C. The commercial ETF had the highest viscosity index and specific heat, but it had lower thermal conductivity, lower oxidation onset, higher conductivity, and higher wear than the ester fluids. I summarized the multi-objective comparison in Table 15.
| Fluid | Electrical performance | Thermal performance | Oxidation stability | Tribological performance | Overall assessment |
|---|---|---|---|---|---|
| A51-C | Best; highest breakdown voltage and lowest conductivity | High thermal conductivity; moderate specific heat | Good; 250.8 °C | Best; lowest friction and wear | Best overall electric vehicle transmission fluid in this study |
| RM1220-C | High breakdown voltage; low conductivity | High thermal conductivity; lower specific heat | Very good; 264.6 °C | Intermediate friction; second-lowest wear | Strong alternative with good thermal stability |
| NP451-C | Lower breakdown voltage; higher conductivity than A51-C and RM1220-C | High thermal conductivity; high specific heat | Best; 266.7 °C | Highest friction among ester fluids; third-lowest wear | Useful when oxidation stability is the dominant requirement |
| ETF | Moderate breakdown voltage; highest conductivity | Lower thermal conductivity; highest specific heat | Lowest; 247.0 °C | Second-lowest friction; highest wear | Reference fluid; lower overall balance than A51-C |
I also defined a multi-objective formulation index to compare the fluids:
$$J=w_1\tilde{\sigma}+w_2\tilde{\mu}+w_3\tilde{V}_w-w_4\tilde{k}-w_5\tilde{T}_{\mathrm{ox}}$$
where the tilde denotes normalization to the best value among the tested fluids, and w1 to w5 are weighting factors. Lower J indicates a better balance when conductivity, friction, and wear are minimized and thermal conductivity and oxidation onset are maximized. Under equal weighting, A51-C had the lowest J because it led in four of the five normalized criteria. This index is not a replacement for standardized electric vehicle fluid testing, but it provides a transparent way to compare formulation trade-offs.
16. Implications for electric vehicle transmission fluid design
The results support several practical design rules for synthetic ester electric vehicle transmission fluids. First, the base oil should be selected to provide a low conductivity floor. A linear symmetrical diester such as A51 is preferable when low conductivity and low friction are both required. Second, the dispersant should be selected at the lowest total base number and highest steric hindrance that still meets cleanliness requirements. T154 at 3.0% is a good compromise. Third, the antioxidant should be selected for oxidation onset rather than for maximum amine basicity. T531 at 1.0% provided a lower conductivity penalty than L57 while still raising the oxidation onset above the commercial reference. Fourth, the corrosion inhibitor should be selected for copper passivation with low ionization. 7001 at 0.1% maintained a 1b copper strip corrosion grade. Fifth, the anti-wear agent should be selected for boundary film formation with low ionic dissociation. H059 at 1.0% gave the lowest conductivity penalty among the anti-wear agents and still supported low friction and low wear in A51-C.
These rules are specific to electric vehicle transmission fluids because the electric vehicle environment couples electrical, thermal, and tribological requirements. In a conventional transmission fluid, conductivity is often not a primary design variable. In an electric vehicle transmission fluid, conductivity affects insulation safety, dielectric loss, and electromagnetic compatibility. A fluid that is excellent in a conventional gearbox may therefore be unsuitable for an electric vehicle drive unit if its conductivity is too high. Conversely, a fluid that is too insulating may accumulate static charge or fail to dissipate electric field stress. The goal is not the lowest possible conductivity but an optimized conductivity window that satisfies insulation, thermal management, and tribological requirements.
I also observed that additive selection can shift the conductivity ranking of base oils. A51 had the lowest base-oil conductivity, and A51-C had the lowest formulated conductivity. However, the relative differences among the formulated fluids were smaller than the differences among the base oils because the same additive package was used in all three. This means that base-oil selection is the first-order lever for conductivity, and additive selection is the second-order lever. In practice, a formulator can first select a low-polarity ester base oil and then use low-conductivity additives to fine-tune the electrical, thermal, and tribological balance.
17. Future work
Further work should examine additive synergy and antagonism in fully formulated electric vehicle transmission fluids. The screening results in this study were based on single-additive blends, and the final formulations used a combined additive package. The cross-interaction term in the conductivity model can be positive or negative depending on additive polarity, acid-base reactions, and micelle formation. A systematic design-of-experiments study should quantify these interactions and identify the combinations that minimize conductivity without sacrificing oxidation stability or anti-wear performance.
Long-term aging should also be studied under electric vehicle duty cycles. Conductivity can change with oxidation, moisture ingress, copper dissolution, and additive depletion. A fluid that has low initial conductivity may show a different conductivity after thermal cycling, electrical loading, and contact with copper and elastomers. I recommend accelerated aging tests that combine high temperature, high voltage, and mechanical shear, followed by breakdown voltage, conductivity, and tribological measurements. Such tests would provide a more realistic assessment of electric vehicle transmission fluid durability.
Electrical field effects on tribology should be investigated in more detail. In an electric vehicle drive unit, the lubricated contact may be exposed to alternating electric fields, and the field can influence additive adsorption, film formation, and wear. Tribo-electrical testing under controlled voltage and frequency would help identify whether the low-conductivity additive package remains effective under electric field stress. The interaction between electric field, surface charge, and boundary film formation is a promising area for future electric vehicle transmission fluid research.
Material compatibility should be extended to elastomers, polymers, and insulation coatings used in electric vehicle motors and inverters. Copper strip corrosion is a useful first screen, but it does not capture seal swelling, hardness change, or dielectric compatibility with magnet wire insulation. A full material compatibility matrix would help ensure that the ester-based electric vehicle transmission fluid is compatible with the entire electric vehicle drive unit.
Finally, the formulation should be tested in a full-scale electric vehicle reducer or powertrain dynamometer. Bench tribology is valuable for screening, but electric vehicle gears and bearings operate under combined rolling-sliding, high-speed, and thermal conditions that are difficult to reproduce in a simple reciprocating test. A full-scale test would provide gear efficiency, bearing life, churning loss, and cooling performance data. These data would support the scale-up of synthetic ester electric vehicle transmission fluids from laboratory formulation to electric vehicle application.
18. Conclusions
I studied the conductivity response of three synthetic ester base oils and four additive classes and used the results to formulate three electric vehicle transmission fluids. The main conclusions are as follows.
First, the conductivity of the three ester base oils increased with temperature, and the order was NP451 > RM1220 > A51. The difference was governed by molecular polarity and polar-group density. NP451 had the highest content of ester and hydroxyl groups and the highest conductivity. A51 had the lowest polar-group density and the lowest conductivity. Base-oil selection therefore sets the conductivity floor of the electric vehicle transmission fluid.
Second, dispersants had the strongest influence on conductivity among the additive classes. Higher total base number and lower steric hindrance produced a stronger conductivity increase. T151 gave the highest conductivity, T152 was intermediate, and T154 gave the lowest conductivity among the dispersants. Antioxidants, corrosion inhibitors, and anti-wear agents had smaller effects, but their molecular structure still mattered. L57 produced a larger increase than T531. T746 produced a larger increase than 7001. AW319 produced a larger increase than H059.
Third, I formulated A51-C, RM1220-C, and NP451-C using T154 at 3.0%, T531 at 1.0%, 7001 at 0.1%, and H059 at 1.0%. All three fluids had higher flash points than the commercial ETF, and all passed copper strip corrosion at grade 1b. A51-C had the best overall performance. Its breakdown voltage reached 62.81 kV, its thermal conductivity and oxidation stability were superior to the commercial ETF, and its friction coefficient was as low as 0.1386. Its disc wear volume was 659000 µm³ and its maximum wear depth was 8.74 µm, both lower than the commercial reference.
Fourth, the ester-based electric vehicle transmission fluids had lower conductivity than the commercial ETF across the measured temperature range. The conductivity ranking was A51-C < RM1220-C < NP451-C < ETF. This shows that synthetic ester base oils, when combined with low-conductivity additives, can provide a favorable electrical window for electric vehicle transmission fluids.
Fifth, the results support a coupled design strategy for electric vehicle transmission fluids. The formulator should first select a low-polarity ester base oil with good boundary-film formation, then select a dispersant with low total base number and high steric hindrance, then select an antioxidant with a low conductivity penalty, then select a corrosion inhibitor with weak ionization, and finally select a molecular anti-wear agent. This strategy allowed A51-C to achieve the best balance of electrical insulation, thermal management, oxidation stability, friction reduction, and anti-wear performance among the fluids tested.
The overall conclusion is that synthetic ester-based electric vehicle transmission fluids are a credible alternative to conventional polyalphaolefin-based fluids when conductivity is treated as a primary formulation variable. The A51-C formulation demonstrated that low conductivity, high breakdown voltage, high thermal conductivity, high oxidation onset, low friction, and low wear can be achieved in one electric vehicle transmission fluid. These results provide a practical basis for further development of ester-based electric vehicle transmission fluids and for future studies of additive synergy, long-term aging, and full-scale electric vehicle drivetrain performance.