Polyvinyl alcohol for cement mortar modification is supplied as a water-soluble, partially hydrolyzed polyvinyl alcohol powder—most commonly a grade analogous to Kuraray Poval
17–88—with a hydrolysis degree of
87–89 mol% and a
4% aqueous solution viscosity of
20–30 mPa·s at
20°C per
ISO 3104. Unlike vinyl acetate-ethylene (VAC/E) redispersible polymer powders, PVA enters the mix water as a true solution; film formation occurs by evaporation-driven coalescence without requiring a minimum film-forming temperature, delivering high interfacial tensile strength even at ambient cure. The material is typically dosed between
0.5 wt% and
2.5 wt% of cement weight in dry-mix formulations for repair mortars, tile adhesives, self-leveling underlayments, and non-structural patching compounds. A key distinction from cellulose ethers lies in the mechanism: PVA contributes negligible water retention enhancement—often below
5% improvement over a plain cement reference measured per
ASTM C1506—but develops a tough, flexible polymer film that bridges microcracks in the range
0.1–10 µm across the cement hydrate matrix.
How does hydrolysis degree and molecular weight govern film morphology in alkaline cement matrices?
The solubility and crystalline behavior of PVA in the high-pH environment of hydrating cement is controlled by the residual acetate group content. Partially hydrolyzed grades (
87–89 mol%) retain sufficient irregularity in the polymer backbone to suppress crystallization, ensuring rapid dissolution in cold mixing water and yielding a homogeneous film with a glass transition temperature near
58°C. Fully hydrolyzed grades (
≥98.5 mol%) exhibit strong intermolecular hydrogen bonding, which increases crystallinity, reduces cold-water solubility, and shifts film formation toward a fibrillar reinforcing network rather than a continuous adhesive film. In mortar tests conducted in accordance with
ASTM C348 (flexural strength of hydraulic cement mortars), a
1.0 wt% addition of
17–88 powder has been reported to elevate
28-day flexural strength by
25–35% over an unmodified control, whereas a fully hydrolyzed grade at identical dosage delivered less than half that improvement due to incomplete dissolution. Viscosity of the aqueous phase also demonstrably shifts the fresh-state rheology. A
4% solution viscosity of
20–30 mPa·s (low molecular weight, typical
17–88) yields a slump life compatible with machine-applied repair mortars, while higher molecular weight grades above
40 mPa·s can impose a sticky consistency that reduces trowelability and increases air entrapment beyond
8 vol%, measured by gravimetric density per
EN 1015-7. Excessive air, in turn, lowers
28-day compressive strength (
ASTM C109) by
8–15% compared to an optimally dosed mix.
In dry-mix manufacturing using a horizontal twin-shaft paddle mixer with a capacity of
500–2000 kg batch, PVA powder with a bulk density of
0.4–0.6 g/cm³ can segregate toward the top of the blend if added directly to the main hopper. The powder’s low density, combined with
D₅₀ particle size typically
150–250 µm, causes a polymer concentration drift of up to
±2 wt% over a
25-tonne silo discharge cycle when the material is not pre-treated. A standard mitigation is to pre-blend the PVA with a
10–20% fraction of the fine silica sand component in a separate conical ribbon blender before metering into the main mixer. Pre-drying of the powder is mandatory at ambient relative humidity exceeding
60%; PVA powder is hygroscopic and can form lumps that block sieve screens and falsify dosage accuracy. Storage in moisture-proof big bags, with a silo residence time kept below
48 hours in facilities without dehumidified pneumatic conveying, reduces re-agglomeration failures observed on continuous production lines.
Adhesion under hydrothermal cycling and open time extension
Tensile adhesion strength of PVA-modified cementitious tile adhesives tested per
EN 1348 typically reaches
1.3–1.8 MPa after
28 days dry cure at
23°C/50% RH for a
1.5 wt% PVA dosage, compared to
0.8–1.0 MPa for an unmodified C1-type formulation. After
7 days water immersion at
20°C and subsequent
24-hour reconditioning, adhesion retention is generally
60–75%, a value inferior to that achieved with a VAC/E redispersible powder at equivalent polymer content, which often retains above
85%. This performance gap arises because PVA films lack the hydrolysable protective colloid shell and hydrophobic comonomer architecture that VAC/E RDPs possess, making them more susceptible to swelling and plasticization under prolonged wet exposure. In heat aging per
EN 1348 (
14 days at
70°C), PVA-modified mortars exhibit minimal strength loss due to the absence of ester group hydrolysis that can embrittle VAC/E films over time. Open time, characterized by the elapsed time from adhesive application to tile embedment before adhesion falls below
0.5 MPa, remains a critical differentiator: PVA does not build viscosity in the pore solution like cellulose ethers and does not form a hydrate barrier layer on the cement grain surface. Consequently, a PVA-only modified mortar can develop a surface skin within
10–15 minutes under
20°C/65% RH conditions, reducing open time compared to a cellulose ether-modified control that extends beyond
30 minutes (
EN 1346). This limitation necessitates partial formulation with a cellulose ether when extended workability is required.
When PVA partially substitutes cellulose ether in a cementitious tile adhesive
A formulation strategy commonly adopted in production-scale dry-mix plants involves a synergistic blend of low-dosage hydroxypropyl methyl cellulose (HPMC) and PVA powder. The cellulose ether provides the required water retention and anti-sagging rheology, while PVA contributes cohesive film strength without inducing excessive delayed hydration. Table 1 presents comparative data from laboratory evaluations on a C1-type tile adhesive based on
OPC 42.5R and
0–0.5 mm silica sand, mixed at a water-to-dry-mortar ratio of
0.24.
| Property |
0.3% HPMC (reference) |
0.15% HPMC + 0.5% PVA 17-88 |
Test method |
| Water retention |
98% |
96% |
ASTM C1506-17 |
| Open time (adhesion ≥0.5 MPa) |
30 min |
25 min |
EN 1346 |
| Tensile adhesion, dry 28 d |
1.2 MPa |
1.5 MPa |
EN 1348 |
| Adhesion after water immersion |
0.8 MPa |
1.0 MPa |
EN 1348 |
| Adhesion after heat aging |
0.9 MPa |
1.1 MPa |
EN 1348 |
The combination maintains water retention within
2% of the cellulose ether-only system, well above the
95% minimum specified in many project specifications for thin-bed adhesives, while dry adhesion increases by
25%. The drop in open time of
5 minutes is tolerable on substrates of moderate absorption. At the microstructural level, the HPMC component retards early hydration by adsorption onto calcium hydroxide nuclei, while the PVA remains in solution and later forms a continuous film that reinforces the interfacial zone between the adhesive mortar and the tile biscuit.
A field failure mode specific to PVA-only modified screeds and patching mortars has been recorded when the dosage exceeds
2.5 wt% and finishing is performed with multiple steel trowel passes. Under these conditions, a polymer-rich skin migrates to the surface, creating a glossy, film-like layer that delaminates as a sheet under tensile stress during later service. In one documented case on a pedestrian bridge deck repair in a temperature-cycled environment, adhesion pull-off values (
ASTM C1583) dropped to
0.3 MPa after
6 months of thermal cycling between
-10°C and
40°C, whereas the bulk mortar below the skin retained a compressive strength above
40 MPa. The resolution involved limiting PVA to
1.5 wt%, introducing a
0.05 wt% air-detraining agent to break the surface film, and restricting trowel passes to a maximum of two.
The interfacial adsorption equilibrium measured by zeta potential and its effect on porosity
The dispersion mechanism of PVA in cement paste has been characterized by electroacoustic spectrometry using a
DT-1200 analyzer. Partially hydrolyzed PVA adsorbs onto cement grain surfaces through hydrogen bonding between hydroxyl groups and the oxygen atoms of silicate phases, shifting the zeta potential from approximately
−5 mV (plain cement suspension) to
−12 to −18 mV at a PVA concentration of
0.5 g/L in the aqueous phase. This increased negative charge reduces flocculation, allowing uniform dispersion of cement particles and more complete hydration of the particle interiors. Pore-size distribution data obtained by mercury intrusion porosimetry (MIP) on mortars cured for
28 days at
23°C/95% RH consistently show a reduction in the volume fraction of capillary pores in the
50–200 nm diameter range. At a
1.5 wt% PVA addition level, the chloride ion diffusion coefficient measured per
ASTM C1556-11a is reduced by approximately
30% relative to an unmodified mix with an identical water-to-cement ratio of
0.50. Published data for this specific configuration are limited, but the magnitude of improvement aligns with the mechanism of pore refinement observed in other water-soluble polymer-latex hybrid systems. Importantly, the PVA film resists saponification in the alkaline pore solution better than polyvinyl acetate, and no loss of modified pore structure is evident after extended ponding in saturated calcium hydroxide solution for
90 days. This makes PVA a technically viable option in non-structural waterproofing screeds where film coalescence at ambient temperature is required without volatile coalescing agents.
| Grade |
Hydrolysis (mol%) |
Viscosity 4% aq. (mPa·s) |
Ash (%, max) |
Volatiles (%, max) |
Test basis |
| PVA 5-88 |
87–89 |
4.5–6.0 |
0.5 |
5.0 |
JIS K6726 |
| PVA 17-88 |
87–89 |
20–30 |
0.5 |
5.0 |
JIS K6726 |
| PVA 24-88 |
87–89 |
44–56 |
0.5 |
5.0 |
JIS K6726 |
| PVA 117 |
98–99 |
25–31 |
0.7 |
5.0 |
JIS K6726 |
