As the Chinese new-energy vehicle market grows rapidly, the retirement wave of power batteries has become an unavoidable challenge. Since most EV battery packs are bonded and sealed with polyurethane structural adhesives, an efficient and safe removal method is essential for the disassembly and recycling of retired EV batteries. In this dissertation, I focus on the chemical degradation of polyurethane structural adhesives used in EV battery modules. I systematically study the catalytic alcoholysis process using both sodium hydroxide and task-specific ionic liquids as catalysts, with the goal of lowering the degradation temperature and enabling easy removal of the polyurethane bonding layer from metal substrates in EV battery packs.
1. Introduction and technical background
Polyurethane structural adhesives are widely used in EV battery assembly because of their high bonding strength, toughness, thermal stability, and vibration damping. The adhesive is usually applied to metal heat sinks or module frames to fix the battery cells. A typical structure is described below:
- Strong bonding between aluminum plates and battery cells.
- Resistance to impact and vibration of the whole EV battery pack.
- Sealing protection against moisture and dust.
- Thermal management by keeping good contact between cells and cooling plates.
However, after the EV battery reaches the end of its vehicle life, the battery pack must be disassembled for second-life applications or material recycling. The same polyurethane adhesive that provided excellent durability now becomes a serious problem. There is no safe, low-cost, and environmentally friendly method for removing the polyurethane structural adhesive from EV battery components. Conventional methods include physical scraping, high-temperature incineration, ultrasonic cleaning, and high-pressure water jetting. All these methods have drawbacks, such as damaging battery cells, producing toxic gases, consuming excessive energy, or being inapplicable to sealed battery packs.
Chemical degradation, such as alcoholysis, hydrolysis, aminolysis, and glycolysis, offers a promising route. In my study, I investigate the alcoholysis of polyurethane structural adhesive using diethylene glycol as the degradation agent and various catalysts, including sodium hydroxide and metal-containing ionic liquids. The overall objective is to develop a mild and efficient chemical method that can selectively degrade the polyurethane structural adhesive from EV battery metal plates without destroying the battery cells.
2. Materials and experimental methods
2.1. Materials
Several commercial polyurethane products were used in the experiments. To protect business confidentiality, I refer to them as PU-A and PU-B, which are two structural adhesives for power batteries, and three additional PU AB adhesives from different brands. Diethylene glycol and diethanolamine were selected as degradation agents. Sodium hydroxide was used as a conventional catalyst. Ionic liquid catalysts were synthesized from 1-butyl-3-methylimidazolium chloride, 1-allyl-3-methylimidazolium chloride, and 1-hexyl-3-methylimidazolium chloride reacted separately with anhydrous FeCl₃, CoCl₂, or ZnCl₂.
2.2. Synthesis of ionic liquids
Five ionic liquids were synthesized by mixing [Bmim]Cl, [Amim]Cl, or [Hmim]Cl with anhydrous metal chlorides in different molar ratios. The resulting ionic liquids were purified by vacuum drying at 60°C for 24 h. The following ionic liquids were prepared:
- [Bmim][FeCl₄]
- [Amim][FeCl₄]
- [Hmim][FeCl₄]
- [Amim][ZnCl₃]
- [Amim][CoCl₃]
For the ionic liquid with [FeCl₄]⁻, the molar fraction of the anion is defined by
$$ x_{\mathrm{FeCl_4}} = \frac{n(\mathrm{FeCl_3})}{n(\mathrm{FeCl_3}) + n([Bmim]Cl)} $$
2.3. Degradation experiments
All degradation tests were carried out in a three-necked glass flask equipped with a magnetic stirrer and a temperature controller. A certain amount of polyurethane adhesive was immersed in the degradation agent with the selected catalyst. After the reaction, the remaining undegraded polyurethane was taken out, washed with ethanol, and weighed. The degradation rate was calculated as:
$$ \eta = \frac{m_1 – m_2}{m_1} \times 100\% $$
where m₁ is the initial mass of polyurethane and m₂ is the mass of residual polyurethane after reaction.
3. Sodium hydroxide catalyzed alcoholysis of polyurethane
3.1. Effect of temperature
In the first set of experiments, diethanolamine was used as the degradation agent with sodium hydroxide as catalyst at different temperatures. The time required for complete degradation of PU-A and PU-B is summarized in Table 1.
| Temperature (°C) | PU-A complete time | PU-B complete time |
|---|---|---|
| 125 | 7.43 h | 8.10 h |
| 145 | 3.20 h | 3.75 h |
| 165 | 1.40 h | 1.60 h |
| 185 | 0.55 h | 0.63 h |
| 205 | 0.28 h | 0.21 h |
| 245 | 0.17 h | 0.08 h |
It is clear that a higher temperature significantly accelerates the alcoholysis reaction. However, high temperatures are unacceptable for EV battery applications because they can damage the cells and separators. Thus, lowering the degradation temperature is the central target of my work.
3.2. Effect of degradation agent composition
When using pure diethylene glycol as the degradation agent, the mass loss of polyurethane was 2.578 g after 1 h at 165°C, while pure diethanolamine produced only 1.856 g of mass loss. Since diethylene glycol has a more symmetric dihydroxy structure and higher charge density on the ether oxygen, it attacks the carbonyl carbon of polyurethane more effectively than diethanolamine. Therefore, diethylene glycol was chosen as the preferred degradation agent.
3.3. Optimization of reaction parameters with NaOH catalyst
Single-factor optimizations were carried out using diethylene glycol as the degradation agent and NaOH as the catalyst. The optimal temperature, catalyst dosage, and reaction time are presented in Table 2.
| Parameter | Optimal value | Mass loss under optimal condition |
|---|---|---|
| Temperature | 165°C | 2.578 g / 1 h |
| NaOH dosage | 0.2 g | 2.578 g / 1 h |
| Time | 40 min | 2.130 g |
The NaOH dosage showed a distinct maximum. The effect of NaOH concentration can be expressed by the following relation:
$$ r = k [\mathrm{OH}^-]^\alpha [\text{polyurethane bonds}]^\beta $$
Since an excess amount of NaOH may shift the relevant equilibrium or cause saponification side reactions, the mass loss decreases when the NaOH dosage exceeds 0.2 g.
3.4. General applicability test for commercial polyurethane AB adhesives
Three different commercial polyurethane AB adhesives with different hardnesses and crosslinking densities were fully degraded by the same formula (diethylene glycol + NaOH). The required complete degradation times are shown in Table 3.
| Temperature (°C) | PU potting AB adhesive | PU adhesive JC8065 | PU resin AB adhesive |
|---|---|---|---|
| 135 | > 5 h | > 5 h | > 5 h |
| 145 | 3.2 h | 3.0 h | 4.1 h |
| 155 | 1.5 h | 1.2 h | 2.2 h |
| 165 | 0.8 h | 0.7 h | 1.3 h |
| 185 | 0.4 h | 0.3 h | 0.5 h |
These results confirm that the sodium hydroxide/diethylene glycol system has reasonable universality for regular EV battery polyurethane structural adhesives. Nevertheless, the temperature requirement above 150°C is still too high for an intact EV battery module. Therefore, I turned to ionic liquid catalysts to reduce the reaction temperature.
3.5. Characterization of degradation products
Fourier transform infrared spectroscopy was used to analyze the products obtained after alcoholysis at different temperatures. All products exhibited a broad O-H stretching band around 3300–3600 cm⁻¹, C-H stretching around 2800–3000 cm⁻¹, and strong C-O stretching near 1000–1100 cm⁻¹. The band positions were unaffected by the degradation temperature, meaning the chemical structure of the degradation products did not change. However, the viscosity of the resulting polyols decreased with increasing reaction temperature.
Rheological tests indicated that all degradation products behaved as Newtonian fluids.
| Degradation temperature | Viscosity (Pa·s) |
|---|---|
| 145°C | 1.72 |
| 155°C | 1.28 |
| 165°C | 0.94 |
| 175°C | 0.68 |
Lower viscosity was correlated with greater molecular chain scission and a more complete degradation reaction.
4. Ionic liquid catalyzed degradation of polyurethane for EV battery recycling
4.1. Synthesis and characterization of ionic liquids
Because the NaOH system still requires temperatures near 165°C, I synthesized Lewis acidic ionic liquids as alternative catalysts. Electrospray ionization mass spectrometry confirmed the presence of cationic and anionic species. In the positive-ion mode, the [Amim]⁺ ion appeared at m/z = 123.9, and the [Bmim]⁺ ion appeared at m/z = 139.1. In the negative-ion mode, a strong signal at m/z = 197.8 was attributed to [FeCl₄]⁻. This confirms that the metal-containing ionic liquids contained the expected cations and tetrachloroferrate anions. A minor peak at m/z = 162.3 was assigned to [FeCl₃]⁻ formed by reduction or collisional dissociation during the electrospray process.
4.2. Comparison of ionic liquid and sodium hydroxide as catalysts
I compared NaOH, [Bmim][FeCl₄], and [Amim][FeCl₄] as catalysts for the alcoholysis of PU using diethylene glycol at 170°C for 40 min. The degradation rates are listed in Table 5.
| Catalyst | Degradation rate (%) |
|---|---|
| NaOH | 26.7 |
| [Bmim][FeCl₄] | 67.4 |
| [Amim][FeCl₄] | 69.5 |
The ionic liquid catalysts more than double the degradation rate of polyurethane compared with NaOH. This enhancement is explained by the cooperative activation of both the polyurethane carbonyl group and the alcohol hydroxyl group. The proposed mechanism involves the following steps:
- The ionic liquid cation binds to the carbonyl oxygen of the urethane bond, withdrawing electron density from the C=O bond and increasing the electrophilicity of the carbonyl carbon.
- The ionic liquid anion forms a hydrogen bond with the hydroxyl proton of diethylene glycol, enhancing the nucleophilicity of the alcohol oxygen.
- A six-membered ring transition state is formed between the ionic liquid, diethylene glycol, and the urethane carbonyl.
- The C-O bond of the urethane is cleaved, generating polyol fragments and amine or isocyanate derivatives.
- The ionic liquid is regenerated and remains available for the next catalytic cycle.
$$ \text{PU} + \text{DEG} \xrightarrow{[Cation][Anion]} \text{lower molecular weight polyols} + \text{amine compounds} $$
4.3. Effect of the anion mole fraction in the ionic liquid
Using [Bmim][FeCl₄] with different mole fractions of [FeCl₄]⁻, the degradation rate reached its maximum when the anion mole fraction was 0.5, corresponding to the cation/anion ratio 1:1.
| [FeCl₄]⁻ mole fraction | Degradation rate (%) |
|---|---|
| 0.40 | 45.2 |
| 0.50 | 67.5 |
| 0.67 | 51.8 |
| 0.75 | 39.6 |
The equimolar ratio provides balanced coordination sites for both the carbonyl oxygen and the hydroxyl protons, promoting formation of the six-membered transition state.
4.4. Effect of ionic liquid dosage
I changed the mass of ionic liquid while keeping the other conditions constant. The degradation rate first increased and then leveled off when the ionic liquid mass reached 2.0 g.
| Ionic liquid mass (g) | Degradation rate (%) |
|---|---|
| 0.0 | 8.3 |
| 0.5 | 22.6 |
| 1.0 | 45.4 |
| 1.5 | 61.2 |
| 2.0 | 67.5 |
| 2.5 | 68.1 |
Thus, 2 g of ionic liquid is the optimal dosage for 40 mL of diethylene glycol.
4.5. Effect of degradation agent
Diethylene glycol and diethanolamine were compared under identical conditions with ionic liquid as the catalyst. At an anion mole fraction of 0.5, the degradation rate using diethylene glycol was 27 percentage points higher than that with diethanolamine. This result confirms that diethylene glycol is a more effective degradation agent because the ether-containing diol has greater nucleophilic reactivity toward the urethane carbonyl group.
4.6. Lowest degradation temperature
To verify the practical possibility of applying this method directly to EV battery components, I measured the degradation rate at lower temperatures using three different catalysts. The data are summarized in Table 8.
| Temperature (°C) | NaOH | [Bmim][FeCl₄] | [Amim][FeCl₄] |
|---|---|---|---|
| 110 | 2.1% | 18.5% | 21.3% |
| 125 | 5.6% | 33.4% | 38.6% |
| 140 | 11.8% | 48.0% | 54.2% |
| 155 | 19.3% | 59.1% | 64.5% |
| 170 | 26.7% | 67.4% | 69.5% |
Even at 110°C, ionic liquids can achieve more than 18% degradation of polyurethane, whereas NaOH only gives about 2% degradation. The ionic liquid approach reduces the minimum effective operating temperature by approximately 40–50°C.
4.7. Effect of the cation in the ionic liquid
To investigate the structural effect of the cation, I tested three ionic liquids with the same [FeCl₄]⁻ anion but different cations. The degradation rates at 155°C and 125°C are compared in Table 9.
| Ionic liquid | Cation structure | Degradation rate at 155°C | Degradation rate at 125°C |
|---|---|---|---|
| [Amim][FeCl₄] | Allyl-methylimidazolium with C=C bond | 64.5% | 38.6% |
| [Bmim][FeCl₄] | Butyl-methylimidazolium | 59.1% | 33.4% |
| [Hmim][FeCl₄] | Hexyl-methylimidazolium | 51.9% | 29.0% |
The [Amim]⁺ cation exhibits the best performance because the allylic C=C bond creates an uneven charge distribution on the imidazolium ring. This stronger positive charge around the ring can more strongly interact with the carbonyl oxygen of the polyurethane, stabilizing the transition state and promoting the cleavage of the urethane bond. In addition, [Amim]⁺ has a shorter side chain and lower steric hindrance than [Hmim]⁺ and [Bmim]⁺, making it easier for diethylene glycol to approach the reaction center.
4.8. Effect of the anion in the ionic liquid
With the same [Amim]⁺ cation, I compared the catalytic activities of [ZnCl₃]⁻, [CoCl₃]⁻, and [FeCl₄]⁻. The degradation rates are shown in Table 10.
| Ionic liquid | Lewis acidity | Degradation rate at 155°C | Degradation rate at 125°C |
|---|---|---|---|
| [Amim][ZnCl₃] | Weakest | 75.8% | 46.4% |
| [Amim][CoCl₃] | Moderate | 68.9% | 41.2% |
| [Amim][FeCl₄] | Strongest | 64.5% | 38.6% |
A surprising finding is that the catalytic activity correlates inversely with Lewis acidity. Lower Lewis acidity in the anion creates a better balanced interaction with the hydroxyl hydrogen of diethylene glycol. Excessive Lewis acidity may bind too strongly to the hydroxyl groups and inhibit the subsequent nucleophilic attack, lowering the degradation rate.
4.9. Removal of polyurethane adhesive from a metal plate
To simulate an actual EV battery operation, I coated a metal plate with the polyurethane structural adhesive and treated it in diethylene glycol containing [Amim][ZnCl₃] at 50°C for 40 min. The treatment softened the cured adhesive enough that it could be scraped off with a plastic scraper. This mild temperature is unlikely to damage the battery cells, making it practical for EV battery recycling lines.
4.10. Reusability of ionic liquids
I tested the repeated use of [Amim][ZnCl₃] in the degradation system. The residual polyurethane after each run was removed, and a fresh sample with the same mass was added to the same batch of diethylene glycol and ionic liquid. The changing mass loss of polyurethane is shown in Table 11.
| Cycle number | PU mass loss after 40 min (g) |
|---|---|
| 1 | 3.13 |
| 5 | 3.12 |
| 10 | 3.10 |
| 15 | 3.08 |
| 18 | 3.06 |
| 20 | 2.75 |
| 22 | 2.06 |
| 25 | 0.98 |
After the 20th cycle, the degradation efficiency starts to drop significantly. The accumulated degradation products and depleted diethylene glycol reduce the effective catalyst concentration. When fresh diethylene glycol was added at cycle 21, the mass loss increased temporarily, confirming that the main cause of decreased activity is not destruction of the ionic liquid structure but rather dilution and accumulation of reaction products.
This observation supports the role of ionic liquids as recyclable catalysts. At the end of the reaction, the ionic liquid cation and anion are regenerated:
$$ [\text{Cat}]^+ + [\text{An}]^- \rightleftharpoons [\text{Cat}][\text{An}] $$
After removing the degradation products, the ionic liquid remains intact and can be reused many times without losing its catalytic function.
5. Kinetic and thermodynamic discussion
The alcoholysis of polyurethane structural adhesive can be approximated as a reaction between the urethane groups and the hydroxyl groups of diethylene glycol. For a simplified kinetic model, I assume an excess of diethylene glycol, so the reaction is pseudo-first-order with respect to the urethane linkage:
$$ -\frac{d[\mathrm{NHCOO}]}{dt} = k_{\text{app}} [\mathrm{NHCOO}] $$
where k_app depends on the catalyst concentration and temperature following the Arrhenius equation:
$$ k_{\text{app}} = A \exp\left( -\frac{E_a}{RT} \right) $$
The activation energy is effectively lowered by the ionic liquid because the transition state is stabilized though multiple noncovalent interactions. Therefore, the same degradation extent is reached at much lower temperatures than with NaOH. This is directly relevant to EV battery processing, where overheating of the cells must be avoided.
6. Conclusion and perspective
In this essay, I investigated the chemical removal of polyurethane structural adhesive from EV battery components. The main findings are summarized below:
| Catalyst | Advantages | Disadvantages / limits | Best degradation condition |
|---|---|---|---|
| NaOH | Low cost, universal for PU | Requires >165°C, too high for EV battery modules, difficult to control | DEG, 0.2 g NaOH, 165°C, 40 min |
| [Bmim][FeCl₄] | High efficiency, recyclable, lower temperature | Slightly expensive, gradual deactivation after 20 cycles | DEG, 2 g IL, [FeCl₄]⁻ mole fraction 0.5, 170°C, 40 min |
| [Amim][FeCl₄] | Higher activity than [Bmim] analog | Deactivation after many cycles | DEG, 2 g IL, 170°C, 40 min |
| [Amim][ZnCl₃] | Highest catalytic activity, low Lewis acidity | Synthetic cost | DEG, 2 g IL, mild temperature down to 50°C for softening |
More importantly, my research demonstrates that an ionic liquid system allows effective polyurethane degradation below 125°C and can even soften the EV battery structural adhesive at 50°C. This temperature range is compatible with the thermal limits of lithium-ion cells and packaging materials.
The key mechanistic insight is the synergistic effect between the cation and the anion. The cation activates the carbonyl group of the urethane linkage, while the anion activates the hydroxyl group of the degradation agent. Balanced electronic interactions in the transition state determine the catalytic efficiency.
$$ \underbrace{\mathrm{R_1\text{-}NH\text{-}COO\text{-}R_2}}_{\text{urethane}} + \mathrm{HO\text{-}R_3\text{-}OH} \xrightarrow[\text{ionic liquid}]{50\text{–}170^\circ\mathrm{C}} \underbrace{\mathrm{R_1 NH_2}}_{\text{amine}} + \mathrm{HO\text{-}R_2} + \text{polyol fragments} $$
For the practical treatment of retired EV battery packs, this process can be used in one of two ways:
- Total dissolution: immerse the complete separated module in the alcoholysis solution at 100–170°C to completely dissolve the polyurethane adhesive, then disassemble the battery cells for sorting.
- Adhesive softening: treat the bonded components at 50°C in the ionic liquid/diethylene glycol solution for a short period, then mechanically scrape off the soft adhesive without causing thermal damage to the EV battery cells.
By combining chemical degradation with subsequent component separation, this method offers a sustainable solution for the recycling of EV battery packs. Further work should investigate continuous recycling of the ionic liquid, recovery of high-purity polyols, and scale-up tests using real retired EV battery modules under industrial conditions.
In conclusion, the ionic liquid catalyzed alcoholysis of polyurethane structural adhesive is a promising route to overcome the disassembly bottleneck in EV battery recycling. It can reduce energy consumption, decrease environmental pollution, and improve the economic viability of the whole power battery recycling industry. The findings from this dissertation contribute to a green and safe technical framework for removing adhesives from retired EV battery packs.