Thermal management is an increasingly decisive factor in electronic product performance, and bonding PCBs or power modules to metal heat spreaders is one of the primary thermal management strategies in industrial electronics, power conversion, and high-reliability systems. The epoxy adhesive used in this bond must accomplish two objectives simultaneously: adequate mechanical attachment to maintain the assembly under vibration and thermal cycling, and adequate thermal conductivity to facilitate heat transfer from the PCB or module to the heat spreader. These objectives impose competing requirements on formulation — higher filler loading improves thermal conductivity but can reduce adhesive strength and increase brittleness. Selecting the right epoxy for this application requires understanding both the thermal path design and the mechanical loading conditions the bond must survive. The Thermal Resistance of the Bond Line The total thermal resistance of the adhesive bond line between the heat-generating component and the heat spreader consists of the intrinsic thermal resistance of the adhesive material (a function of thermal conductivity and bond line thickness) and the contact resistance at the adhesive-substrate interfaces. Both contribute to the temperature rise across the bond. Thermal conductivity of unfilled epoxy is approximately 0.2 W/m·K — comparable to most plastics and far below metal heat spreaders (aluminium: 150 to 200 W/m·K; copper: 390 W/m·K). A 200-micron bond line of unfilled epoxy adds approximately 1.0°C·cm²/W of thermal resistance, which for a 1 cm² component at 10W power dissipation means a 10°C temperature rise across the bond alone. This is often acceptable for low to moderate power densities. For high-power applications, filled epoxy with thermal conductivity of 1 to 5 W/m·K reduces this temperature rise proportionally. At 3 W/m·K, the same 200-micron bond line contributes only 0.067°C·cm²/W — an order of magnitude improvement over unfilled epoxy. Minimizing bond line thickness further reduces thermal resistance; glass bead spacers at 50 to 100 microns can halve the bond line contribution. Thermally Conductive Epoxy Formulations Thermal conductivity of epoxy adhesive is improved by loading with conductive filler particles. The most common fillers and their contributions: Alumina (Al₂O₃): Volume loading of 60 to 75% alumina achieves thermal conductivity of 1.5 to 3 W/m·K. Alumina is an electrical insulator, making alumina-filled epoxy suitable for applications where electrical isolation between the PCB and heat spreader is required. This is the most common filler type for PCB-to-heat-spreader bonding. Boron nitride (BN): At equivalent loading, BN provides higher thermal conductivity (2 to 6 W/m·K at 60% loading) and remains electrically insulating. BN-filled epoxy is used for the most demanding thermal applications where alumina-filled systems are insufficient. It is more expensive than alumina-filled alternatives. Silver (Ag): Silver particle or flake loading achieves the highest thermal conductivity (4 to 10 W/m·K) but is electrically conductive. Silver-filled epoxy is appropriate when electrical conductivity between the PCB ground and the heat spreader is acceptable or desired, but must not be used where electrical isolation is required. Aluminum nitride (AlN): High thermal conductivity filler (approximately 320 W/m·K for the pure ceramic, though the composite achieves 5 to 10…

A failed epoxy bond on an assembly that is in service or partially assembled presents a practical dilemma: the ideal repair — complete disassembly, full surface preparation of both substrates, and re-bonding with fresh adhesive — is often impossible, impractical, or would cause collateral damage to other components. Engineers and maintenance technicians faced with a disbonded joint in a complex assembly need a repair approach that works within the constraint of limited access and partial assembly, restores adequate load capacity, and is durable enough to last through the expected remaining service life. This is achievable in many cases, but it requires honest assessment of what the repair can and cannot accomplish and the right sequencing of preparation, adhesive selection, and cure. Assess the Failure First Before attempting repair, understand why the bond failed. A repair that does not address the root cause will fail again in the same way — possibly faster, because the substrate surfaces in a previously bonded and failed area may be harder to prepare adequately. Examine the failure surfaces as described in Incure's epoxy bond failure diagnostic checklist — adhesive versus cohesive failure, evidence of contamination, evidence of moisture degradation, fatigue cracking pattern. If the failure was adhesive — clean substrate on one side — the original preparation was inadequate, and the repair must include better preparation. If the failure was cohesive — adhesive on both surfaces — the adhesive was mechanically overloaded or improperly selected, and the repair must use a stronger or better-matched adhesive, or redesign the joint geometry, following the same logic Incure covers in preventing epoxy bond failure at the interface versus cohesively. If the failure was driven by in-service environmental degradation (moisture, chemical attack, UV), the repair must address the exposure — a topcoat over the repair bond edge, a change in sealant, or an environmental protection that was not present on the original bond. Preparing the Bond Surfaces for Repair Preparation quality on a repair bond is typically harder to achieve than on original manufacture, and is the most critical determinant of repair success. The failed bond surfaces have been exposed to service environment, may have residue from the original adhesive, and may be contaminated by the fluids or conditions that contributed to failure. Remove all adhesive residue. Old adhesive left on the substrate is not an acceptable bonding surface. The old adhesive surface has been degraded by service environment and does not adhere reliably to fresh adhesive. Mechanical removal — scraping, grinding, or abrasion — down to the substrate surface is required. For aluminium and composite substrates, care must be taken not to remove more substrate material than necessary. Degrease with fresh solvent. Two-wipe technique with isopropyl alcohol or acetone, using clean cloths. For oil or fuel contamination associated with the failure cause, a more aggressive degreaser — MEK or dedicated metal cleaner — may be needed before the final solvent wipe. Abrade to a fresh surface. Abrasion after degreasing removes the degraded surface layer and exposes fresh substrate.…

Applying epoxy to a contaminated surface is one of the most common causes of adhesive bond failure, and it produces a failure mode that is predictably bad: adhesive failure at the interface, with the epoxy peeling cleanly from the substrate surface, leaving a substrate face that shows little or no adhesive residue. The epoxy cured correctly, achieved good internal strength, and would have performed as specified on a clean surface — but the contaminant layer between adhesive and substrate prevented the chemical and physical bonding that the adhesive requires to work. Understanding what the contamination does at the molecular level, how different contaminants affect adhesion differently, and what remediation is possible shapes both how engineers should approach contamination prevention and how to recover a production situation where contamination has occurred. What Contaminants Do at the Bond Interface Adhesive bonding works because the adhesive wets the substrate surface — spreading into intimate contact with it — and then forms physical and chemical bonds with the substrate material. Contamination interferes with both steps. Wetting prevention. Oil, grease, release agent, and other low-surface-energy contaminants reduce the surface energy of the substrate below the surface energy of the epoxy. When epoxy contacts a low-surface-energy surface, the thermodynamics of wetting are unfavorable — the epoxy beads up rather than spreading, just as water beads on a waxed surface. At a microscopic level, even if the epoxy appears to wet the surface macroscopically, the adhesive is bonding to the contaminant layer, not to the substrate. Interface weakness. The contaminant layer, once encapsulated between the cured adhesive and the substrate, forms a weak cohesive zone at the interface. Oil films have essentially zero tensile strength; they shear readily under applied load. A thin oil layer — even one that is not visually detectable — between a well-cured epoxy and a metal substrate reduces the practical adhesion to near zero. Long-term moisture pathways. Contamination at the bond interface creates sites where moisture preferentially accumulates, displacing the adhesive-substrate bond over time. Even a bond that appears initially adequate on a contaminated surface may deteriorate rapidly in service with humidity exposure. On aluminium specifically, this moisture-driven mechanism compounds with the native oxide's own instability — see Incure's guide to preparing aluminium for epoxy bonding for how etch primer addresses both problems together. The Most Common Contaminants and Their Sources Cutting and machining oils. Metal components that have been machined, stamped, or formed typically have residual cutting fluid, stamping oil, or drawing compound on the surface. These are specifically designed to be difficult to remove — they are formulated to adhere tenaciously to metal surfaces for lubrication and corrosion protection. Standard dry wiping with a clean cloth does not remove them; solvent degreasing with the correct solvent is required. Mold release agents. Composite parts, cast polymers, and formed components made in molds carry mold release on their surfaces. Mold release agents — typically silicone-based or fluoropolymer-based — are specifically formulated to be low-surface-energy and to resist bonding. They are among the…

Epoxy adhesive viscosity is one of the first practical constraints in application selection, and it is frequently misunderstood. Engineers sometimes default to requesting "thick" epoxy because it seems more structural, or "thin" epoxy because it flows more easily — but viscosity selection should be driven by the specific requirements of the application geometry, the dispensing method, the bond line thickness target, and the orientation of the parts during cure. A mismatch between epoxy viscosity and application requirements produces either joints with inadequate filling and thin spots, or joints that slump, sag, and lose thickness uniformity before cure. The right viscosity is the one that works with the geometry and process, not the one that feels strongest in the hand. The Viscosity Spectrum Epoxy adhesive viscosity is measured in mPa·s (milliPascal-seconds, equivalent to centiPoise) and spans several orders of magnitude across product types: Flowable (100 to 1,000 mPa·s): Water-like to light oil consistency. These epoxies self-level rapidly and penetrate narrow gaps and fine porous structures by capillary action. Used for wicking adhesives, impregnating resins, honeycomb core potting, and applications where flow into fine geometry is required. The challenge with low-viscosity epoxy is that it will flow out of any gap or joint that is not precisely fixtured, and it requires horizontal cure to prevent draining. Medium viscosity (1,000 to 10,000 mPa·s): Roughly the consistency of honey to thick syrup. Flow is significant but manageable with simple fixturing. This range is practical for dispensing with syringes or automated dispensers, for horizontal bonding with controlled gap, and for potting applications where the housing geometry confines the compound. Self-leveling is useful — the adhesive reaches the perimeter of the bond area without heavy pressure — but flow out of inclined joints requires fixturing attention. High viscosity / thixotropic paste (10,000 to 100,000+ mPa·s): Non-sag to putty consistency. Thixotropic pastes do not flow under gravity but flow under shear stress during dispensing or spreading. This range is the most common for structural bonding: the adhesive stays where it is applied, can be used on vertical or overhead surfaces, fills gaps without draining, and does not require elaborate fixturing to maintain position during cure. Two-part cartridge adhesives are typically formulated in this range, and the same viscosity-and-tensile grade matching used in structural glass-to-metal bonding applies broadly to any thixotropic-paste product line. Matching Viscosity to Joint Geometry Narrow lap joints with controlled gap. For lap joints with a precise target bond line thickness (typically 0.1 to 0.3 mm), a medium-viscosity adhesive that spreads under assembly pressure to fill the bond area is appropriate. High-viscosity paste in a narrow bond line requires higher assembly pressure to spread, which can introduce voids if the adhesive does not flow to all areas before gelling. Glass bead or metal spacers maintain bond line thickness regardless of viscosity. Gap-filling joints. When bonding machined or fabricated parts where the surface gap varies from near-zero to several millimeters, a non-sag thixotropic paste is required. Flowable adhesive drains from thick sections and leaves those areas…

Epoxy adhesive performs reliably in many demanding environments but has a well-documented vulnerability to ultraviolet radiation. UV exposure initiates photo-oxidative degradation in the epoxy polymer, changing the adhesive from a tough, ductile material to a brittle, chalky surface layer that eventually loses adhesion to the substrate. For outdoor structural bonds — in transportation, construction, signage, solar installations, and architectural glazing — UV durability is a fundamental requirement that standard epoxy formulations do not meet without additional protection. Understanding the degradation mechanism, the rate at which it progresses, and the protective measures that interrupt it allows engineers to design outdoor bonds that remain functional over the required service life. The UV Degradation Mechanism Epoxy polymers absorb UV radiation in the wavelength range of approximately 290 to 380 nm — the UV-A and UV-B bands present in terrestrial sunlight. The absorbed photon energy breaks covalent bonds in the epoxy backbone, initiating a chain of radical reactions that progressively depolymerize and oxidize the polymer surface. This photo-oxidation produces carbonyl groups, hydroxyl groups, and ultimately small volatile fragments that are lost from the surface — a process visible as chalking, where the degraded surface layer powders and wipes away. The initial UV degradation is confined to the surface — typically the top 50 to 200 microns of the adhesive. This surface layer loses toughness, becomes brittle, and eventually microcracking develops. The cracks act as stress concentrations and as pathways for moisture ingress deeper into the bond. Over time, the combination of surface embrittlement and moisture-assisted disbondment from the bond edge progresses into the adhesive and toward the adhesive-substrate interface. Yellowing or darkening of epoxy under UV exposure is a color change caused by the same photo-oxidation reactions and is an early visible indicator of degradation. Yellowing alone does not represent structural failure, but it signals that degradation has begun and will progress if the UV exposure continues. Rate of Degradation The rate of UV degradation depends on UV intensity (higher at high altitude and low latitude), exposure geometry (direct south-facing exposure is the harshest in the northern hemisphere), temperature (elevated surface temperature accelerates oxidation), and moisture (wet-dry cycling after UV exposure accelerates microcracking). In accelerated weathering tests — Xenon arc weatherometer per ASTM G155 or UV fluorescent lamp per ASTM G154 — standard epoxy adhesive shows measurable surface degradation within 500 to 1000 hours and significant strength reduction in exposed adhesive layers within 1000 to 3000 hours, depending on formulation and test conditions. Outdoors in direct sunlight at mid-latitude, 1000 hours of accelerated weathering approximately corresponds to one year of field exposure, though the correlation varies considerably by geography and orientation. Incure's broader guide to epoxy bonding for outdoor applications covers UV exposure alongside the moisture and thermal cycling stresses that typically accompany it in the field. If you need UV durability data for epoxy adhesive formulations, including weathering test results and outdoor service correlation data, Email Us — Incure provides weathering test data and UV protection recommendations for structural adhesive applications. What Can…

Cold temperature imposes a distinct set of challenges on epoxy adhesive bonds that engineers focused on high-temperature performance often overlook. An epoxy that bonds strongly at room temperature and passes every ambient qualification test may fail catastrophically at -55°C during a cold-soak test — not because the adhesive chemically degraded, but because it became brittle and could no longer accommodate the differential contraction stress imposed by dissimilar-material adherends on cool-down. For aerospace, Arctic oil and gas, cryogenic process equipment, and military systems that operate to -55°C and below, cold temperature performance must be explicitly designed into the adhesive selection rather than assumed. What Happens to Epoxy at Very Low Temperatures Standard rigid epoxy at ambient temperature is already a relatively brittle material — its elongation to break is typically 1% to 3%, and its fracture energy is low compared to toughened systems. As temperature decreases below ambient, the polymer chain segment mobility decreases further, the material becomes stiffer and more brittle, and the elongation to break decreases. This progression continues until the polymer is fully below its beta transition temperature — a secondary transition below the glass transition — at which point the material is maximally brittle. For many standard epoxy formulations, the transition to highly brittle behavior occurs somewhere between -20°C and -60°C, depending on the specific chemistry. At -55°C, an untoughened epoxy may have elongation to break below 0.5% and fracture energy values only a fraction of the ambient temperature values. Any thermal cycling stress from differential CTE between substrate materials will exceed this reduced fracture energy, initiating cracking at the bond line edge or within the adhesive bulk. The practical consequence is that bonds that survive thousands of thermal cycles between ambient and 50°C may fail in the first cycle that reaches -55°C — because the stress at the bond line during the cold excursion exceeds the cold-temperature fracture energy of the adhesive. What "Cold-Temperature Tough" Means in an Epoxy Formulation Epoxy adhesives that maintain useful fracture energy and elongation at extreme cold temperature achieve this through rubber or thermoplastic toughening — the same mechanism used for impact resistance, which is closely related to low-temperature performance. Rubber particles (CTBN carboxyl-terminated butadiene acrylonitrile rubber, or core-shell rubber particles) dispersed in the epoxy matrix cavitate and stretch as a crack propagates through the adhesive, absorbing energy that would otherwise drive brittle fracture. At low temperatures, the effectiveness of these rubber tougheners depends on the Tg of the rubber phase — if the rubber itself becomes glassy at the test temperature, it loses its energy-absorbing function. The rubber phase Tg is the critical low-temperature limit for rubber-toughened epoxy performance. CTBN rubber has a Tg of approximately -70°C to -80°C, making it functional at -55°C. Core-shell rubber particles with CTBN or polybutadiene cores provide similar or better low-temperature performance. Thermoplastic-toughened epoxy systems have variable low-temperature performance depending on the thermoplastic modifier, and should be tested at -55°C rather than assumed to perform. The relationship between formulation Tg and cure schedule is…

A void-free potting fill in a simple open cavity is achievable with basic technique — pour slowly, tilt the assembly, let bubbles escape. The same objective in a complex geometry — a housing packed with tall components, fine wire harnesses, PCBs with densely placed components on both sides, or narrow channels between internal partitions — requires a systematic approach that addresses void formation at each stage of the process. Complex geometries trap air by design: every overhang, every narrow gap, every blind pocket is a potential void site. Eliminating those voids requires combining good dispensing technique, appropriate compound viscosity, and in most cases vacuum assistance — with each step verified before the compound gels. Mapping Void Risk Before Potting Before dispensing a drop of compound, examine the assembly and identify every location where air can be trapped. High-risk locations include: The underside of flat components sitting close to the PCB surface (less than 1 mm clearance) The inside corners of the housing where component leads or wire bundles contact the housing wall Narrow channels between adjacent tall components or between components and housing walls Blind pockets — recesses in the housing bottom or internal partitions — that have no upward escape path for air Wire harness entries where bundle density limits compound penetration Mark these locations mentally and consider how compound flow will approach each one. Flow from the wrong direction seals the void before air can escape; flow from the correct direction allows air to be displaced upward and out. Compound Selection for Complex Fill Compound viscosity is the primary material property determining void susceptibility in complex geometries. Lower viscosity allows the compound to penetrate fine gaps and flow around component undersides more readily than high-viscosity paste. For complex geometry applications, selecting the lowest-viscosity compound that meets the application's other requirements (temperature, electrical, chemical) reduces the process burden. Many high-temperature epoxy compounds have viscosities of 5,000 to 50,000 mPa·s at ambient temperature — adequate for simple fills but marginal for fine geometry. Preheating the compound (and the assembly) to 40°C to 60°C reduces viscosity substantially and improves penetration without initiating cure. Confirm the pot life at the elevated dispensing temperature before implementing preheating — accelerated pot life must be compatible with the assembly time. Incure's broader guide to selecting epoxy viscosity from flowable to paste covers this trade-off across application types beyond potting. If you need compound viscosity data at elevated dispensing temperature and pot life curves for complex-geometry potting applications, Email Us — Incure provides formulation-specific application data and dispensing process support. Dispensing Sequence for Complex Geometries Fill from the lowest point, from one side only. Place the dispense nozzle at the lowest corner of the housing — not the center — and begin dispensing. Allow compound to pool and rise progressively from the bottom. Compound rising from below pushes air upward through the assembly; compound dispensed from the top traps air beneath every component body. Move the nozzle as the level rises. Keep the nozzle submerged in…

Bubbles forming in curing epoxy are a visible sign that something has gone wrong in the mixing or application process. For cosmetic epoxy applications, bubbles ruin the appearance. For structural bonds and potting applications, bubbles create voids that reduce effective bond area, concentrate stress, and compromise moisture and electrical isolation. The frustrating aspect of bubble formation is that it is frequently attributed to "the epoxy bubbling" — as if the material is inherently defective — when the actual cause is almost always a process issue that can be identified and corrected. Understanding the mechanism behind each type of bubble formation leads directly to the corrective action. Cause 1: Air Entrapped During Mixing The most common source of bubbles in cured epoxy is air introduced during the mixing of the two components. When resin and hardener are combined and stirred, the stirring action folds air into the mixture. Vigorous mechanical mixing — a high-speed drill-mounted mixer, aggressive manual stirring, or any technique that creates a vortex or froth — entraps far more air than gentle hand-mixing. The entrapped air is distributed as fine bubbles throughout the mixed material. If the epoxy has low viscosity or adequate pot life, these bubbles rise and escape from the surface before gel time. If the viscosity is high, pot life is short, or the poured layer is deep, the bubbles are immobilized before they can escape. Mixing gently, with a flat paddle stirrer and a scraping motion rather than a whipping motion, avoids creating a vortex or froth in the first place. Letting the mixed epoxy rest in the mixing container for two to five minutes before pouring allows bubbles to rise and pop at the surface before the material is transferred to the mold or housing. Cause 2: Moisture Reacting with the Hardener Some epoxy hardeners — particularly amine hardeners — react with atmospheric moisture. This reaction can produce CO₂ as a byproduct, which forms bubbles within the curing epoxy. The reaction is more pronounced at higher humidity and is characteristic of specific hardener types (benzyl dimethyl amine and other tertiary amines are known for this behavior). Moisture on the substrate surface can also react with the adhesive at the bond line, creating a thin layer of bubbles at the interface — a failure mode that looks like poor adhesion but originates from moisture reaction. Using substrates that have been dried before bonding — particularly porous substrates like wood, ceramics, or foam that absorb ambient moisture — addresses this at the source. Storing hardener components in sealed containers minimizes moisture exposure, and for moisture-reactive hardener systems in high-humidity environments, switching to a formulation with lower moisture sensitivity or working in a humidity-controlled space closes the gap entirely. If you need formulation guidance for high-humidity environments where moisture-reactive bubbling is a problem, Email Us — Incure can recommend moisture-tolerant epoxy systems for your application. Cause 3: Outgassing from the Substrate Porous or gas-absorbing substrates — foam, balsa wood, porous ceramics, green (uncured) concrete — release…

Delamination of a bonded carbon fiber composite joint — where the failure runs not through the adhesive but into the composite laminate itself — is one of the most consequential failure modes in composite structure. It means the adhesive system worked correctly: the adhesive-to-composite bond was stronger than the interlaminar tensile strength of the composite itself. The failure originates in the composite, not the adhesive, and the joint cannot be improved by changing the adhesive. Instead, prevention requires understanding where delamination-inducing stress concentrations come from, how joint geometry amplifies them, and what surface preparation and adhesive selection decisions keep the load path within the composite's capability. What Delamination in Bonded Composite Joints Looks Like When a bonded composite joint fails by laminate delamination, the fracture runs parallel to the laminate surface, within the outermost ply or between the first and second ply, rather than through the adhesive layer or at the adhesive-composite interface. The adhesive remains intact, still bonded to a thin skin of composite material that has peeled away from the laminate. This failure mode is driven by peel or tensile stress perpendicular to the laminate surface at or near the bond line edge — the zone where stress concentrations are highest in any lap or strap joint. Single-lap joints are particularly prone to inducing delamination at composite adherends because the geometry produces bending at the bond line ends, which generates significant peel stress perpendicular to the laminate plane. The composite's interlaminar tensile strength — typically 30 to 60 MPa for aerospace-grade CFRP — is far lower than its in-plane tensile strength (600 to 1500 MPa), and peel stress at the bond end can reach this interlaminar limit before the shear stress in the bond approaches the adhesive lap shear strength. Surface Preparation for Carbon Fiber Composite Bonding The bond surface of a composite component must be prepared to remove the release agent-contaminated resin-rich surface layer and expose clean, active fiber-resin interface for the adhesive to bond to. Peel ply removal. For composite parts manufactured with peel ply on the bond surface, peel ply removal exposes a surface topography with a mechanical anchor profile and clean resin surface — if the peel ply was clean and did not transfer contamination. Peel ply surface quality varies by fabric type and storage conditions; contaminants can transfer from peel ply to the composite surface and reduce adhesion. Solvent wipe after peel ply removal confirms surface cleanliness. Abrasion. For composite surfaces without peel ply, or where peel ply quality is in question, abrasion with 120 to 180 grit silicon carbide paper removes the resin-rich surface layer and exposes fiber ends at the surface, increasing the surface energy and mechanical anchor area. Abrasion must be light — just enough to dull the surface and break through the resin-rich skin — not aggressive enough to damage fibers. Damaged or cut fibers reduce the interlaminar strength of the surface ply. Solvent wipe. Solvent degreasing with isopropyl alcohol or acetone before and after abrasion removes release agent…

Aluminium is one of the most commonly bonded structural materials and one of the most demanding for consistent adhesion. The oxide layer on aluminium that reforms within minutes of surface preparation is hydrated and inherently unstable — it bonds readily to fresh epoxy immediately after preparation but weakens progressively as moisture from the environment replaces the adhesive-oxide bonds over weeks to months of service. This mechanism of hydration-driven disbondment is the dominant reason why aluminium-epoxy bonds that appear strong at room temperature on freshly bonded specimens fail earlier than expected in service with humidity exposure. Proper surface preparation, and specifically the use of etch primer, addresses this mechanism at the chemical level — not just as a supplementary step, but as the primary determinant of long-term bond durability on aluminium. Why Aluminium Is Difficult to Bond Aluminium instantly forms a surface oxide — aluminium oxide (Al₂O₃) — when exposed to air. This oxide is what makes aluminium corrosion-resistant; it is also what the epoxy adhesive bonds to. The problem is that the native aluminium oxide is not a stable bonding surface for long-term adhesive service. It is hydrated (contains bound water molecules), relatively thick and loose in structure, and susceptible to replacement by water molecules that can penetrate the adhesive-oxide interface from the bond edge. The hydration mechanism works as follows: water diffuses into the bond line from the exposed edge, driven by concentration gradient and osmotic pressure. At the oxide-adhesive interface, water molecules displace the adhesive-oxide bonds — a thermodynamically favorable exchange because aluminium oxide has high affinity for water. The result is progressive interfacial disbondment from the bond edge inward, without the adhesive itself degrading. The failure mode is adhesive — clean metal on one side — even though the adhesive cured correctly and had adequate initial strength. This is one of the clearest examples of the adhesive-versus-cohesive distinction covered in Incure's diagnostic checklist for epoxy bond failure: the fracture surface tells you the preparation, not the adhesive, needs attention. The Preparation Sequence Mechanical abrasion. Abrasion with silicon carbide abrasive paper (120 to 180 grit) or non-woven abrasive pad removes the loose, hydrated oxide layer and the contamination that is concentrated on the as-received surface. Abrasion must be followed immediately by cleaning to remove the abraded oxide particles, which if left on the surface contaminate the adhesive interface. Abrasion alone creates the mechanical anchor profile needed for physical interlocking but does not create the chemically stable interface needed for long-term durability in moisture. Solvent degreasing. Solvent wiping with isopropyl alcohol or acetone before and after abrasion removes oil, release agent, and handling contamination. Wipe in one direction with clean, lint-free cloths — back-and-forth wiping redistributes contamination. Two-wipe technique: apply solvent with one cloth, dry with a second clean cloth before the solvent re-deposits contamination from evaporation. Acid etching. Chromic-sulfuric acid etching (Forest Products Laboratory etch, per ASTM D2651) or sulfuric acid-sodium dichromate etch produces a clean, chemically active surface with higher surface energy than mechanical abrasion alone. The…