Simple bonding is not enough

Battery pack adhesives must do much more than simply hold things together. Peter Donaldson reports

EV battery packs increasingly rely on adhesives to hold them together, but modern ‘glues’ have many more core functions. While that bond must maintain structural integrity against vibration, shock and potential swelling, it must also: prevent short-circuits in the presence of high voltages; efficiently transfer heat between cells and cooling plates/heating elements; withstand exposure to electrolytes, moisture and thermal cycling without degrading; accommodate different coefficients of thermal expansion between materials; and be compatible with automated dispensing and curing for high-volume manufacture. These sets of required functions and properties drive the development of adhesives with different priorities but significant overlaps.

Structural bonding adhesives provide primary structural integrity, replacing or supplementing mechanical fasteners and welds. Epoxy-based examples offer high strength, excellent chemical resistance and good thermal properties, and are widely used in module assembly. Acrylics, particularly in toughened formulations, provide fast curing along with good impact resistance and peel strength. Thirdly, polyurethanes are excellent at absorbing vibration and coping with mismatches in coefficients of thermal expansion between the materials to be bonded.

Thermal interface materials, often in the form of silicone-based soft and compliant pads or dispensed pastes, fill air gaps between cells/modules and cooling elements such as cold plates and aluminium housings, improving heat transfer. Thermally conductive adhesives (TCAs) – typically either epoxy- or silicone-based – provide bonding and heat conduction when directly attaching cells to cooling plates. Thermal conductivity is measured in Watts per metre Kelvin (W/mK).

Potting and encapsulation compounds fully encapsulate modules and components such as BMS boards and busbars for maximum protection. They provide high levels of sealing against moisture and other contaminants, superior electrical insulation, mechanical protection and, increasingly, fire retardance. These benefits come at the cost of extra weight and making repairs difficult or even practically impossible.

Finally, sealants and gasketing create environmental seals for pack housings, covers and connectors to meet IP67 dust and water ingress protection standards. They must remain flexible over a wide temperature range, and so silicones and moisture curing sealants are commonly used.

More than just a bond

In cell-to-pack architectures and structural battery designs, the adhesive has become a load-bearing, thermally active, electrically insulating interface. It is expected to manage mechanical stress, conduct heat and maintain dielectric integrity – simultaneously, continuously and for 15 years or more.

This is a profound shift that has presented adhesive formulators and design engineers with what amounts to a fundamental trilemma in balancing thermal conductivity, mechanical performance and processability.

The physics are unforgiving. To make an adhesive thermally conductive, it must be filled with ceramic or metallic particles that create a percolating heat path through the polymer matrix. But these fillers are dense, abrasive and inert. The more that are added, the less room remains for the organic binder that provides the adhesion, flexibility and toughness.

At moderate loadings – typically, around 1.5 W/mK – a skilled formulator can still preserve reasonable elongation and peel strength. The adhesive remains sufficiently compliant to accommodate cell swelling and vibration, it can be dispensed reliably and it wets out well on substrates.

At 3.0 W/mK and above, the character of the material changes fundamentally. Filler content can approach 80–90% by volume. The polymer then becomes a minority constituent and its role is reduced to that of a binding matrix holding conductive particles in proximity. Therefore, the mechanical properties inevitably suffer. The material becomes brittle, its ability to absorb strain diminishes, and its viscosity climbs – sometimes dramatically – making dispensing slower and more abrasive on pumps and nozzles.

One supplier emphasised that in a heavily filled material system, the degree of freedom in organic parts is severely limited.

This is not a failure of formulation chemistry – it is a constraint of the physical world. It forces a difficult conversation early in the design process between material suppliers and design engineers about what the adhesive must do. 

Thermal conductivity trap

It is here that a persistent knowledge gap emerges. Many design engineers, confronted with thermal simulation results, specify the highest available thermal conductivity as a margin of safety. But thermal conductivity does not exist in a vacuum. A high-W/mK material with poor wet out, thick bond lines or interfacial voids will underperform in a moderately conductive adhesive that achieves intimate, void-free contact across a thin, controlled gap.

For example, one supplier reports that a 1.2 W/mK gap filler outperformed a 3.0 W/mK material in a customer’s pack-level validation. The reason was not bulk conductivity but interface quality; the lower-conductivity material wet out the substrates more completely, displaced air more effectively and maintained consistent contact under compression.

Specifying thermal conductivity in isolation is likely to be a design error. The relevant metric is system-level thermal resistance, which is a function of conductivity, bond-line thickness, contact area and interface quality. 

Durability under stress If thermal–mechanical trade-off is the most visible challenge, long-term durability is arguably the most consequential.

A battery pack is not a static assembly; it breathes. Cells swell with state of charge and age. Temperature cycles from -40 to 85 C induce differential expansion between aluminium housings, plastic frames and steel cell cans. Vibration loads are continuous and broad-spectrum. And this environment must be withstood not for months, but for a decade and a half.

Traditional accelerated ageing tests based on exposure to stressors such as heat, humidity and thermal shock remain the baseline. But leading suppliers are now moving toward combined-stress protocols that more faithfully reproduce real-world conditions. Materials are subjected to simultaneous thermal ageing and mechanical cycling. Fatigue testing is becoming standard for structural adhesives that now bear significant load. Some programs now simulate second-life grid storage applications requiring 30 year extrapolated lifespans.

One supplier acknowledged the underlying uncertainty candidly: “EVs are so new. I don’t think OEMs have enough experience to know. We take a conservative approach and test significantly longer than required.”

This is not a weakness. It is a recognition that the industry is building reliability models on limited field data, and that material suppliers are therefore extending their safety margins, often beyond customer specifications.

One supplier emphasises that developing formulations that maintain more consistent performance across both nominal and extreme conditions is a major challenge in materials science. It notes that OEMs are increasingly looking for adhesives that deliver stable strength, flexibility and durability throughout the entire operating temperature window, not just at specific points within it. Furthermore, achieving this level of performance while reducing – or potentially eliminating – the need for surface pretreatment, requires advancements in interfacial chemistry to ensure robust bonding to diverse substrates without additional process steps.

Compatibility and harmony

There is another dimension to formulating a multifunctional adhesive. It does not exist alone, but contacts the cell casing, the cold plate, the busbar insulation, and often thermal barrier materials such as mica or aerogel composites. These substrates have different surface energies, different coefficients of thermal expansion and different chemical sensitivities.

An adhesive that bonds aggressively to aluminium may induce dangerous stress in a brittle mica sheet during thermal cycling. A formulation optimised for polypropylene may not wet out on stainless steel. A curing chemistry that is inert to standard pouch foil may corrode busbar coatings.

The consequence is that adhesive selection cannot be delegated to the procurement stage. It must occur at system architecture definition, with input from mechanical, thermal, manufacturing and materials engineering. Suppliers consistently emphasise that their most successful engagements begin before the CAD model is locked-in, when bond-line thickness, surface area and material stack-ups can still be optimised collaboratively.

From lab to gigafactory

A battery adhesive that delivers perfect thermal conductivity and structural strength in the laboratory must also support reliable, consistent and rapid dispensing across millions of units per year. In the gigafactory era, processability is a primary design constraint. Supplier’s responses emphasise that for every new formulation, the rheology and cure profile are now engineered alongside thermal and mechanical performance, rather than after them.

The central challenge is viscosity. A thermally conductive adhesive heavily loaded with ceramic or metallic fillers is inherently thick, abrasive and resistant to flow. Yet, it must be pumped through hoses, dispensed through nozzles, and deposited in precise beads or films at speeds that support takt times measured in seconds, not minutes.

The solution lies in shear-thinning behaviour. Under the high shear forces of a dispensing nozzle, the material must thin significantly to flow freely. The instant it lands on the substrate and shear is removed, it must recover viscosity rapidly enough to hold its shape, resist slump and maintain a consistent bond-line thickness.

This is not a simple formulation task. It requires careful selection of binder molecular weight, filler morphology and rheological additives. One supplier described it as engineering a material that behaves like a liquid during movement and a solid at rest – a duality that must hold across variations in temperature, humidity and batch-to-batch raw material consistency.

For potting and encapsulation materials, the requirement is different but equally demanding. Here, the goal is ultra-low viscosity to permit capillary flow into tight gaps and around complex geometries. The material must displace air completely, leaving no voids that could become sites of partial discharge or mechanical weakness. Yet, even these low-viscosity systems must remain stable; filler sedimentation during storage or dispensing is a persistent risk that suppliers mitigate through continuous agitation and tailored particle size distributions.

Cure profile compromise

If rheology governs how the material flows, cure kinetics govern how the line flows. Every production line has a takt time – the interval at which completed packs or modules exit the station. The adhesive must achieve sufficient handling strength within that window to permit immediate transfer to the next operation. Waiting minutes for a bond to cure creates bottlenecks, buffer stocks and capital inefficiency.

Yet, faster curing is not always better. A material that fixes in seconds offers almost no open time for part placement, alignment or rework. It may begin gelling in the dispensing nozzle, leading to blockages and unscheduled maintenance. The compromise between ‘fixture time’ and ‘open time’ is one of the most critical negotiations between supplier and manufacturer.

Suppliers address this through several strategies. Two-part reactive systems allow formulation tuning to accelerate or delay cure. Heat-cure materials offer latency at room temperature followed by rapid crosslinking in oven or platen contact. UV-cure systems provide near-instantaneous fixation where light can reach, although shadowed areas remain a challenge addressed by secondary moisture or thermal cure mechanisms.

Pressure-sensitive adhesives and structural tapes offer an entirely different paradigm. With no cure required, they provide instant grab and immediate handling strength. This eliminates ovens, reduces energy consumption and simplifies line integration. The trade-offs are equally real: each part geometry may require a dedicated applicator, release liners must be removed robotically and the inherent strength of a tape bond, while improving, remains below that of high-performance liquid epoxies.

Abrasion tax

There is an often-hidden cost to high thermal conductivity, taking the form of abrasive wear. At filler loadings approaching 80% by volume, the adhesive becomes a slurry of hard ceramic particles suspended in a minimal polymer matrix. This slurry is pumped through hoses, valves and nozzles at high pressure and velocity. The effect on equipment is erosive; seals wear, nozzle diameters enlarge and pump rotors lose tolerance. Consequently, preventive maintenance intervals shorten.

For a gigafactory, unplanned downtime is measured in millions of dollars per hour. Suppliers who understand this reality now optimise filler morphology. Spherical particles flow more easily and abrade less than angular fragments. Narrow particle size distributions reduce viscosity at equivalent loading. Some suppliers now offer formulations that trade a small reduction in headline thermal conductivity for substantial gains in pump life and process robustness.

In-line quality control

The final frontier of processability is assurance, meaning – ideally – that every bond in every pack is good. Vision systems have matured rapidly, and high-resolution cameras now inspect every dispensed bead for continuity, width and position. Tape release liner removal is verified, for example, and surface contamination is flagged. These systems are effective, reliable and increasingly standard.

What remains elusive is universal non-destructive evaluation of the cured bond interface.

As one supplier acknowledged: “Complete in situ direct analysis for QC is still a big challenge in the industry, which has to be solved.”

Until that challenge is met, manufacturers rely on indirect control. They monitor dispensed weight, they validate substrate cleanliness, they control environmental conditions and they perform destructive testing on sacrificial parts at regular intervals. It is a system of proxies that is rigorous but not absolute.

One supplier, for example, monitors key dispensing parameters such as component mix ratios directly in the application process. “By detecting deviations in real time and alerting operators or automation systems, it helps ensure that each dispense meets the intended formulation.” Its system can be connected to a plant level supervisory control and data acquisition platform, enabling automated monitoring and broader process oversight.

Adhesives and thermal runaway

Adhesive suppliers emphasise that their products are not primary thermal barriers in the way that mica sheets, ceramic papers or aerogel blankets are. However, they do have a foundational – if subtle – role in safety where they help avoid the conditions that might trigger thermal runaway.

For example, TCAs play a direct preventive role by providing an efficient heat path from the cell casing to the cooling plate. In this way, they reduce operating temperatures and temperature gradients across the pack. A cell that runs cooler, uniformly, is a cell less likely to initiate runaway. Suppliers report that their TCAs are now specified deliberately to eliminate hotspots in high-energy-density and fast-charging architectures.

If prevention fails and a cell enters thermal runaway, the adhesive’s primary obligation shifts: it must not make the situation worse. This is not a trivial requirement because most organic polymers are combustible. Therefore, modern battery-grade adhesives are formulated to meet stringent flame retardancy standards, eg UL 94 V-0. This rating requires that materials self-extinguish within specified time frames and not drip burning particles.

More broadly, adhesives must continue to secure auxiliary materials when exposed to the extreme heat, pressure and gas flow associated with thermal runaway. This requires formulations that maintain adhesion at elevated temperatures, resist rapid decomposition, limit smoke and byproduct generation, and provide mechanical stability long enough to help the overall system contain and redirect energy safely.

But some suppliers are going further. Ceramifying and intumescent chemistries represent the next generation of safety functionality. When exposed to extreme heat, these materials do not simply resist flame – they transform. Ceramifying silicones convert to a rigid, insulating ceramic char that maintains dielectric integrity and structural cohesion. Intumescent systems expand to form a thick, insulating carbonaceous foam that protects underlying substrates and delays thermal propagation.

Collaborative integration

Achieving system-level safety requires material compatibility and coherence across the entire thermal barrier assembly. Adhesives are often used to bond mica sheets, ceramic papers or aerogel composites to pack housings or cell frames. These are delicate, friable materials with coefficients of thermal expansion very different from the aluminium or steel structures they protect. An adhesive that is too rigid will induce stress during thermal cycling, potentially cracking the very barrier it is meant to secure.

Suppliers report extensive collaborative work with OEMs and tier suppliers to formulate adhesives with a controlled modulus and elongation specifically for these interfaces. The goal is to achieve a bond that is strong enough to survive vibration and crash loads, yet compliant enough to accommodate differential expansion without damaging the barrier material.

This is also where the distinction between adhesives and coatings becomes significant. Several suppliers explicitly separate these functions: adhesives for structural and thermal management, dedicated fire-protection coatings for the pack lid and enclosure. These coatings are engineered not to bond, but to contain – reducing heat transfer, delaying flame ejection and extending the time available for the safe exit of occupants. In some applications, they are displacing traditional mica sheets entirely, enabling lighter, simpler and more scalable pack designs. 

Testing fallacy

One of the most persistent knowledge gaps in the industry concerns flammability testing in isolation. Suppliers report a recurring scenario: an engineer specifies a flame-retardant tape or adhesive based on a datasheet showing V-0 or VTM-0 certification. The material is tested in the pack and fails. Alternatively, a non-flame-retardant material passes.

The explanation is almost always the same. A material tested alone behaves differently than a material tested in contact with its substrate. Aluminium is an excellent heat sink; therefore, a flammable tape bonded to an aluminium cooling plate may pass a vertical burn test because the metal conducts heat away faster than the flame can propagate. Conversely, a flame-retardant tape applied to a plastic carrier may fail because the plastic insulates the tape and permits sustained combustion.

The implication is clear: safety validation must occur at the system level, on representative substrates, in the assembled configuration. Datasheet certifications are necessary for material qualification but they are not sufficient for pack-level safety assurance.

Towards the circular economy

Battery developers are caught between two opposing imperatives. On the one hand, the demand for energy density, structural integration and crash safety drives the need for permanent, high-strength bonds that last 15 years without degradation. On the other hand, regulatory pressure, corporate sustainability pledges, and the economics of battery repair and reuse demand that those bonds be reversible. This is not merely a technical challenge but “the Mount Everest of material science,” as described by one supplier.

Yet, the mountain must be climbed to enable design for disassembly, which requires adhesives that can be debonded on demand. Multiple suppliers confirm that reversible adhesive technologies are a major r&d priority, with several mechanisms under active development.

With thermally triggered debonding, adhesives are formulated to undergo a controlled reduction in bond strength when exposed to a specific temperature window – typically 80–90 C, which is selected to remain below the thermal limits of lithium-ion cells. At this trigger temperature, the polymer matrix softens or undergoes a phase change, permitting clean separation of bonded components. Critically, some of these systems are reversible: when cooled, they recover their original mechanical properties, enabling reassembly.

However, whether it is feasible to ensure that adhesives with thermally triggered debonding do not debond during thermal runaway, or to design the entire pack such that localised debonding helps slow propagation of the runaway rather than accelerating it, is so far unclear.

Chemical triggers offer an alternative pathway. Solvents or reactive compounds can be introduced to the bond line to swell or dissolve the polymer interface. The challenge lies in delivering the trigger precisely to the bond line without damaging cells or electronics.

Ultrasonic and mechanical triggers are also under investigation, although these remain at earlier stages of development.

Finally, adhesives that debond in response to an electrical trigger have also been successfully demonstrated. In one example, the adhesive is applied between two electrically conductive surfaces. To debond it, a DC current applied to the bond line at a precise voltage creates a weak boundary layer that allows separation of the bonded parts.

The technical feasibility of debond-on-demand is no longer in doubt – the obstacle is cost. These systems require specialised monomers, complex formulation chemistry and often entirely new manufacturing processes. One supplier drew a direct analogy to sustainable aviation fuel: “This technology comes at a cost and will not be adopted unless the cost comes down or regulations require it.”

Sustainability hierarchy

However, a more immediate sustainability contribution that adhesives can make is to support repair and second life. Avoiding the manufacture of a replacement pack saves the full embodied carbon of that pack’s materials and production.

One important concept here is controlled adhesion, in which the material provides bonds strong enough to handle 15 years of vibration and thermal cycling, yet enables a trained technician to separate a cooling plate from a module using appropriate tools and procedures.

Preparing for solid-state

The transition to solid-state and semi-solid batteries is coming.

When it arrives, it will change adhesive requirements. Solid-state battery cells operate under higher compression loads and experience volume expansion during cycling, which increases mechanical demands on bonding materials. They also function at higher temperatures, requiring adhesives that can maintain strength, stability and reliability under such conditions. Therefore, future pack designs may need high-strength, high-temperature adhesives with greater elongation to accommodate mechanical movement while preserving structural integrity.

Pressure management is the most frequently cited challenge. Many solid-state cell designs require significant, uniform stack pressure in the realm of tens of MPa to maintain interfacial contact and suppress dendrite formation. This pressure must be maintained over the life of the battery, despite cell expansion and contraction, thermal cycling and relaxation of mechanical components.

Adhesives and elastomers in these packs will need excellent compression set resistance, which is a measure of a material’s ability to rebound after being compressed, and they must not creep or yield under sustained load.

Material compatibility is a second frontier. For example, sulphide-based solid electrolytes are hygroscopic and chemically reactive. Adhesives contacting these materials must not cause decomposition or interfacial resistance increases. Ultra-dry manufacturing environments impose additional constraints.

Thermal and mechanical requirements may shift. Several suppliers anticipate that solid-state cells, with intrinsically safer chemistries, may demand lower thermal conductivity than today’s high-power lithium-ion cells. Flame retardancy requirements may also relax. Conversely, flexibility and elongation may become more critical if solid-state cells exhibit different expansion behaviour.

A new toolbox

The complexity of modern battery adhesives has outstripped the capacity of traditional trial-and-error formulation. Fortunately, computational materials science, supply chain engineering and early-stage collaboration are converging to reshape how adhesives are developed, specified and deployed.

Although AI and machine learning play an increasingly important role, no supplier reported using AI to autonomously design a complete adhesive formulation – the complexity is just too high. Instead, AI is used as a development accelerator. Machine learning models trained on historical formulation and test data can propose promising filler combinations or polymer backbones for a given set of target properties, for example.

Molecular dynamics simulations are employed to study filler–matrix interactions. These simulations help formulators understand how surface chemistry affects dispersion, how particle shape influences rheology and how interfacial bonding evolves.

One supplier described the current state pragmatically: “By narrowing the field before physical testing begins, these tools reduce traditional trial-and-error screening and allow engineering teams to focus quickly on the most promising solutions.”

Human formulators of adhesives are not being replaced – they are being equipped with better instruments. 

Acknowledgements

The author would like to thank the following for their help with this article: Max VanRaaphorst, business development manager, e-mobility and automotive at Avery Dennison; Julian Hopf, product manager energy storage and conversion at DELO; Tyler Auvil, PhD, r&d technical leader, AMS and Tim Campbell, marketing manager, AMS at DuPont; Germaine Mariaselvaraj, technical service manager at HB Fuller; Dr Keon Lee, senior manager product development, battery solutions at Henkel; Eric Wyman, business development manager and Eric Dean, global business development manager at Parker Lord; Jean-Marc Pinel, r&d director for automotive coatings at PPG; Dr Kevin Payne, technical manager innovation silicones, at Wacker Chemical Corporation; and Tim Quinn, e powertrain segment marketing, Brandon Bartling, battery system architect and Benjamin Pütz, eAssembly B&J product development leader at 3M.

Some suppliers of battery pack adhesives

Click here to read the latest issue of E-Mobility Engineering.