Within dry-mix C2 cementitious tile adhesives classified under EN 12004-2:2017+A1:2021, redispersible acetate-ethylene copolymer powders are dry components incorporated at 2.0–6.0 wt% of the dry-mix mass to shift failure after water immersion away from brittle interfacial loss and toward energy dissipation within the cementitious matrix. The synthetic polymer is produced by spray drying an aqueous vinyl acetate-ethylene dispersion containing 10–25 wt% ethylene, a glass transition temperature between −20 °C and +25 °C, and a minimum film-forming temperature from 0 °C to 12 °C. Polyvinyl alcohol is the most common protective colloid, with residual levels of 1–8 wt% of the redispersible powder; anti-caking agents such as kaolin, talc, or precipitated silica are added at 0.5–2.0 wt% to control surface tack. In a dry-mix formulation for interior and exterior tile installation, ordinary Portland cement comprises 30–45 wt%, graded silica sand 50–65 wt%, calcium carbonate 0–20 wt%, cellulose ether 0.3–0.8 wt%, VAE powder 2–6 wt%, and set regulators or accelerators 0–1.5 wt%. The water-to-powder ratio is adjusted between 0.20 and 0.26 L/kg during site mixing; the resulting mortar has a wet density around 1.5 kg/L and a pot life of 1 to 4 h depending on cement reactivity and cellulose ether grade. C2 classification under EN 12004-2:2017+A1:2021 and ISO 13007-2:2022 requires a tensile adhesion strength of at least 1.0 MPa after standard climate, water immersion, heat ageing, and freeze-thaw conditioning. The water immersion failure mode is therefore a critical performance discriminator, because a material may exceed the dry-state threshold yet lose interfacial integrity after prolonged contact with water.

The wet-storage failure mode is a composite expression of at least three transport and chemical processes: capillary water uptake, polymer film plasticization and swelling, and hydrolytic degradation of acetate functionality. These processes operate simultaneously but at different rates depending on pore size distribution, film continuity, cement alkalinity, and the presence of soluble salts. In field installations, wet-storage conditions are often far more aggressive than the standard immersion protocol; ponded water on balcony floors, continuous contact with saturated backer board, or water trapped beneath large porcelain tiles can maintain near-saturated conditions for months. Under these conditions, the failure mode typically changes from a mixed cohesive-adhesive fracture under dry testing to a predominantly adhesive failure at the tile-adhesive or substrate-adhesive interface. The apparent wet adhesion loss may be partially reversible upon drying, but the interfacial damage caused by osmotic blistering and polymer extraction is not restored by drying alone. Consequently, wet-storage failure analysis must record not only the pull-off strength but also the fracture pattern, the location of polymer-rich transfer layers, and the chemical composition of the failed surfaces.

How Does Alkaline Water Immersion Redirect Failure in VAE-Modified C2 Mortars?

Immersion of a cured VAE-modified C2 mortar in water at 20±2 °C creates a counter-current of water ingress and ionic egress through capillary pores and microcracks. Cementitious tile adhesives are open porous composites with capillary pore radii from 5 nm to 100 nm and entrained air voids up to 1 mm. Water fills these cavities and plasticizes the VAE polymer film that had formed on cement hydrate surfaces during the initial drying period. The vinyl acetate segments of the copolymer are susceptible to alkaline hydrolysis in the cement pore solution, which has a pH typically between 12.5 and 13.8. Hydrolysis converts acetate groups to vinyl alcohol units and releases acetate ions; the resulting copolymer becomes more hydrophilic and more swollen in the wet state. In free polymer films, the water uptake of VAE copolymers can range from 5 % to 40 % by mass depending on ethylene content, protective colloid level, and film crosslinking. In the composite, this swelling generates mechanical stress near the mineral-polymer interface because the film is confined by cement hydrates and aggregate. Repeated stress cycles during prolonged immersion create microcracks at the interface, increasing hydraulic permeability and accelerating local leaching of calcium hydroxide. The residual polyvinyl alcohol protective colloid is another hydrophilic component that absorbs water and can generate osmotic pressure when soluble ions accumulate on one side of the film. The resulting blisters and microdelaminations are visible in polished cross-sections as elongated voids between the cement matrix and tile interface. After water immersion, pull-off testing may show a fracture surface that is relatively clean on the tile side, with a thin translucent film that is enriched in VAE and PVA; the adhesive bed may remain largely intact, indicating that the limiting weak plane is the polymer-cement boundary rather than the bulk mineral matrix. Published data for the spatial distribution of acetate hydrolysis products in wet-stored C2 systems is limited, but failure analyses consistently identify interfacial polymer de-anchoring rather than bulk cement disintegration.

Protective colloid chemistry influences this failure mode sharply. Polyvinyl alcohol grades with a high degree of hydrolysis, typically 88–99 mol%, provide stronger dry films and better powder storage stability, but they retain significant water affinity. Partially hydrolysed grades around 87–89 mol% are widely used for redispersible powders because they balance cold-water dispersibility, film formation, and wet-storage resistance. The wet film modulus of the polymer phase can drop from a dry value of 50–200 MPa to a saturated value below 5 MPa for hydrophilic grades, as measured by dynamic mechanical thermal analysis at 1 Hz on free films. This loss of stiffness is not merely a mechanical issue; it also permits the polymer film to shear more easily under the differential thermal expansion of tile, adhesive, and substrate during wet service. The combination of hydrolysis and modulus loss explains why low-ethylene VAE grades often produce brittle interfacial failure after water immersion, while higher-ethylene grades may produce cohesive failure within the adhesive bed but at a reduced peak stress. The fracture mode thus carries at least as much diagnostic weight as the numerical pull-off value.

When MFFT, Ethylene Content, and Protective Colloid Chemistry Intersect Below 5 wt% Addition

At VAE powder additions below 3.0 wt%, the polymer film in the cured adhesive is discontinuous, and water immersion rapidly attacks exposed cement hydrate interfaces. Between 3.0 wt% and 5.0 wt%, wet-storage performance becomes sensitive to film formation conditions. If the minimum film-forming temperature exceeds the local surface temperature during drying, full coalescence does not occur; discrete polymer particles remain that can swell independently or be removed by water ingress. Ethylene content governs both hydrophobic character and low-temperature coalescence. Copolymers with ethylene content below 10 wt% show high water uptake and rapid acetate hydrolysis, whereas copolymers above 25 wt% may lower the dry initial tensile adhesion below the C2 threshold even as wet-storage retention improves. The protective colloid also contributes ionic mobility and water affinity. A polymer powder that disperses readily in cold water may nonetheless produce a film with high water sensitivity if the colloid has low hydrolysis degree or high residual content. Table 1 summarises the directional effects of these variables on wet-storage failure behaviour.

The threshold of 5 wt% addition is not an absolute boundary for every C2 formulation, but it is a practical reference above which the polymer film is sufficiently continuous to mask moderate variability in cement composition and sand grading. However, higher polymer loading also increases wet-mix viscosity and air entrainment, often requiring a water demand increase of 0.005–0.015 L/kg per additional 1 wt% of VAE powder. Excess water enlarges capillary pores, increases water permeability, and reduces the mechanical confinement of the swollen polymer film during immersion. A formulation that reaches the 1.0 MPa C2 threshold under dry conditions can still fail the water immersion requirement if film coalescence was incomplete. Conversely, a formulation with lower initial dry strength but complete film formation may retain a higher fraction of its strength after immersion. This inversion cannot be detected by dry-state pull-off testing alone; it requires evaluation of the fracture mode and water uptake after the designated immersion protocol.

Trowelling creates a high surface-area layer that dries faster than the bulk; the VAE film forms more rapidly at the exposed surface, while the interior remains water-saturated. If the skin forms too quickly, polymer migration to the surface is enhanced, leaving a polymer-depleted zone just below the surface. During water immersion, this polymer-depleted zone becomes a fracture plane. The use of cellulose ethers with high water retention helps control film formation by maintaining water availability for film coalescence, but excessive cellulose ether can increase wet-state swelling and reduce adhesion. The balance between cellulose ether and VAE powder is therefore critical; typical cellulose ether dosages of 0.3–0.8 wt% in C2 formulations must be selected based on their water retention capacity and their interaction with the protective colloid of the redispersible powder.

On production-scale powder blending lines using twin-shaft ploughshare mixers of 500–1000 L working capacity, moisture ingress above 0.5 wt% and local temperature rise above 45 °C can cause the polyvinyl alcohol protective colloid on VAE particles to become tacky, producing agglomerates that do not fully break down during site mixing. These agglomerates act as weak inclusions after water immersion because their interior contains poorly coalesced polymer and entrapped air. Sieve residues above 2 % on a 250 µm sieve after dry mixing correlate with reduced post-immersion tensile adhesion, even when initial dry pull-off values remain acceptable. In bulk bag and silo storage, warehouse relative humidity above 70 % for periods longer than 7 d without moisture-barrier liners can raise powder moisture to 1.5 wt% or more, reducing flowability and altering the water-to-powder ratio of the wet mix. Pneumatic conveying lines with high-pressure rotary airlocks can segregate the dry blend by particle density; fine polymer powder and cellulose ether can separate from denser sand and cement, creating batch-to-batch variation in polymer content as high as ±0.5 wt% at the application point. Field audits of failed C2 applications after water exposure often localise adhesion loss near areas trowelled from the end of a batch or where additional water was applied to the trowel, conditions that dilute or redistribute the redispersible polymer at the open surface.

Dry-mix producers control wet-storage performance through moisture content, sieve residue, water demand, and powder dispersion stability. Loss-on-drying at 105 °C for 1 h is used to measure powder moisture; typical control limits are 0.3–1.0 wt%. Below 0.2 wt%, electrostatic charging reduces blending uniformity; above 1.5 wt%, caking in silo discharge increases. Commercial VAE redispersible powders have a mean agglomerate diameter between 50 µm and 250 µm, with primary latex particles in the 0.5–5 µm range. During mixing, the powder must redisperse to primary particles within the first 30 s of high-shear mixing. Inadequate redispersion leaves large polymer agglomerates that do not coalesce into a continuous film; these agglomerates absorb water during immersion and create local swelling pockets. A typical control limit for residue on a 63 µm sieve after standard redispersion is below 1 % of powder mass. When residue exceeds this limit, wet-storage failure is more likely because the coarse particles behave as hydrophilic gel domains rather than film-forming polymer. The failure surface of a polymer-lean batch after water immersion differs from that of a well-blended formulation: the polymer-lean system fails in a brittle, planar pattern at the substrate-adhesive interface, while a properly distributed VAE system tends to exhibit polymer-fibrillated cohesive failure or mixed-mode fracture with fragments of cement matrix attached to the tile. Fluorescent resin impregnation of polished cross-sections from failed samples can reveal polymer-depleted zones 50–200 µm thick at the interface when local mixing water was excessive. These processing indicators are as important as raw material properties because the redispersible polymer cannot correct a poorly blended or poorly wetted mortar.

Wet Storage Failure Begins at the Polymer-Cement Interface

Mechanical pull-off testing reports a scalar stress but does not identify the fracture locus with sufficient resolution; the fracture mode must be recorded using the visual categories described in EN 1348:2007. A pure adhesive failure between the tile and the adhesive bed after water immersion indicates that water attack occurred at the interface, where water-soluble fractions migrated and re-deposited. Often a thin polymer-rich film remains on the tile back face but is swollen and unable to transfer tensile stress to the mortar bed. Infrared spectroscopy of this transfer layer can show increased carbonyl intensity relative to the bulk adhesive, consistent with VAE segregation at the open interface during water evaporation before immersion. Cohesive failure within the adhesive bed indicates that water immersion reduced the internal tensile strength of the composite below the interfacial bond strength, which is commonly observed when the polymer film forms a continuous but highly swollen network. The practical interpretation is that wet-storage failure mode is not a single event but a sequence of interfacial water uptake, polymer plasticization, osmotic swelling, hydrolysis of polar units, and crack propagation along the weakest boundary. When the tile is non-porous porcelain with water absorption below 0.1 %, water cannot escape through the tile and remains trapped at the adhesive-tile interface, increasing the likelihood of blistering and interfacial separation. For porous clay tiles with water absorption above 10 %, capillary suction removes some water from the interface and may protect wet adhesion, although it promotes greater bulk swelling. The most severe case is dense porcelain installed over an impermeable waterproofing membrane; water migrating through the adhesive bed accumulates at two impermeable boundaries, and failure often shifts to the substrate-adhesive interface.

Scanning electron microscopy of failed surfaces after water immersion typically identifies reticular polymer films pulled from the cement hydrate matrix. In interfacial adhesive failure, the tile side carries little or no cement residue but shows carbon- and oxygen-rich patches under energy-dispersive X-ray analysis, indicating polymer film transfer. In cohesive wet failure, the tile side contains fragments of calcium silicate hydrate and embedded sand grains, indicating that the mineral phase itself failed. The absence of large calcium silicate hydrate fragments on the tile side is an indicator of interfacial adhesive loss; the presence of such fragments indicates cohesive wet failure. Formulation adjustments based on failure mode are more efficient than further increments of polymer dosage. If pull-off failure is interfacial with a clean tile side, the weak point is often the polymer-cement interface, and increasing VAE dosage may not solve the problem if the polymer film is water-sensitive or insufficiently coalesced. If the failure is cohesive, raising the polymer dosage or reducing the water-to-powder ratio may improve wet-storage strength by increasing film continuity and reducing porosity. Surface preparation of the substrate also influences failure location. A cementitious smoothing compound or a waterproofing membrane can behave as a barrier to water vapor and dissolved salts. In such configurations, water that penetrates from the tile side can condense or accumulate at the substrate interface, leading to adhesion loss even when the tile-adhesive interface remains intact. Published data comparing wet-storage failure on standard concrete slabs versus polymer-modified waterproofing membranes is limited, yet commercial technical guidance indicates that the substrate may shift the failure plane from tile-adhesive to substrate-adhesive.

Accelerated Conditioning, Film Swelling, and Hydrolytic Stability Measurements

Laboratory evaluation of wet-storage failure follows the conditioning sequence specified in EN 1348:2007 and the classification requirements of EN 12004-2:2017+A1:2021 or ISO 13007-2:2022. The immersion water is maintained at 20±2 °C; its pH, alkalinity, and dissolved solids are not usually standardised but can materially affect the outcome. Pull-off testing is performed with a tensile tester operating at a controlled loading rate, typically 250±50 N/s for 50 mm diameter steel dollies bonded to the tile surface with epoxy. Drying of the saturated specimen before pull-off, even for a few hours at 23±2 °C and 50±5 % relative humidity, can partially restore apparent strength by desorption of water from the polymer film, but this restoration does not recover the original interfacial microstructure. Dynamic water uptake tests on free VAE films show mass gains of 5–40 % after 48 h immersion depending on ethylene content and protective colloid level; low-ethylene grades can exceed 30 % mass gain. In the cementitious matrix, water uptake is lower due to pore tortuosity, but the polymer phase exhibits the same swelling tendency. Thermogravimetric analysis of wet-stored adhesive samples quantifies bound water retained up to 150 °C, while differential scanning calorimetry of extracted films detects a depression of the wet glass transition temperature of 10–30 °C relative to the dry value. These measurements explain why the standard pull-off test reports only failure stress and not the path of water attack or the chemical stability of the polymer.

In routine certification, the wet-storage failure mode of a C2 tile adhesive is reduced to a pass/fail comparison against the 1.0 MPa threshold. This approach does not distinguish between a cohesive failure that indicates uniform polymer-cement interaction and an interfacial failure that indicates polymer film de-anchoring. A material that meets the threshold after immersion but exhibits interfacial adhesive failure may still be vulnerable to progressive strength loss under additional freeze-thaw cycling or alkaline water exposure. Conversely, a cohesive failure after immersion, although representing a strength reduction, indicates that the polymer-cement composite retains a more uniform stress distribution and may be less likely to debond completely. The conversion of vinyl acetate units during immersion is measurable by infrared spectroscopy but is not reported in routine product certification. Field failures are documented in instances where initial classification testing did not include failure-mode discrimination; routine certification records only whether the pull-off stress exceeded the threshold.