In the manufacture of polyvinyl butyral interlayer for laminated glass by sheet extrusion, the distribution of plasticizer within the polyvinyl butyral matrix is governed less by gross solubility than by thermodynamic partitioning between residual hydroxyl-rich domains and the bulk butyral-rich amorphous phase. Polyvinyl butyral is the condensation product of polyvinyl alcohol and butyraldehyde, with residual hydroxyl content commonly controlled within 18–23 wt% and vinyl acetate content within 1–4 wt%, as determined by wet chemical or infrared methods referenced in ASTM D1396-20. Commercial interlayer formulations incorporate 20–50 phr of plasticizer, most commonly triethylene glycol bis(2-ethylhexanoate) or tetraethylene glycol di-2-ethylhexanoate, to suppress the glass transition temperature from the unplasticized range of 70–80 °C to a laminated-glass service range of 25–35 °C. On production-scale co-rotating twin-screw extruders with L/D ratios in the 40:1–52:1 range, liquid plasticizer is injected into the barrel after the resin feed section, where it is worked into the polymer under high shear and conveyed through vacuum devolatilization ports before passing through a melt pump and sheet die with thickness capability from 0.38 mm to 1.52 mm. The partition coefficient is not a single equilibrium constant; instead it reflects local free volume, hydrogen-bonding density, and thermal history. Hydroxyl residues act as transient adsorption sites for the carbonyl and ether oxygens of the plasticizer, increasing local solubility while slowing macroscopic diffusion. Butyral-rich domains, by contrast, provide lower hydrogen-bonding capacity but greater chain mobility. During cooling on the chill roll, the melt locks into a non-equilibrium distribution, and subsequent storage at room temperature allows plasticizer to migrate along the concentration gradient from the core toward the surface and from hot edges to cooler interior regions. When that migration is not controlled by formulation and roll conditioning, downstream lamination operations see variable tack, edge haze, thickness distortion, and compromised glass adhesion.

Why Does Plasticizer Migration Produce Edge Haze and Blocking in Wound Interlayer Rolls?

Edge haze and blocking in wound interlayer rolls originate from the same underlying transport process but are separated by the distance over which the plasticizer gradient acts. After extrusion and slitting, the cut edges expose a fresh cross-section. The surface concentration of plasticizer at the cut edge is lower than that of the interior by an amount determined by the partition coefficient between polymer and air at the winding temperature. Because the diffusion coefficient of plasticizer in PVB is strongly temperature-dependent, the edge region loses plasticizer more rapidly at elevated storage temperatures, leading to refractive index gradients that scatter light and produce edge haze. Simultaneously, the bulk surface exudation of plasticizer under roll winding pressure increases tack and promotes blocking between adjacent layers. Antistatic and anti-block additives, commonly based on fatty acid amides, are included in interlayer formulations, but these additives are themselves partitioned between the PVB surface and the bulk. Their effectiveness is reduced if they migrate inward or if plasticizer accumulates at the surface faster than the anti-block layer can form. In production practice, slit rolls are conditioned for 24–72 h at 10–25 °C and 30–70% relative humidity before shipment. When conditioning is omitted or when storage temperature exceeds 25 °C, blocking incidence increases and edge haze becomes visible after lamination. Published numerical diffusion coefficients for triethylene glycol bis(2-ethylhexanoate) in PVB across production melt temperatures are limited; however, the empirical consequence is well documented in edge haze and blocking complaints from laminators.

At the glass-interlayer interface after autoclave lamination, plasticizer partitioning is influenced by water uptake from the glass surface and the surrounding atmosphere. Polyvinyl butyral interlayer is deliberately conditioned to a moisture content of 0.1–0.5 wt% before lamination to control adhesion to soda-lime glass. Water molecules hydrogen-bond to the same hydroxyl and ester sites as the plasticizer; at higher moisture contents, water displaces plasticizer at the glass interface. This displacement reduces interfacial tack and can produce low-adhesion regions that fail pummel testing or show edge delamination under impact. At lower moisture contents, the interlayer retains too much plasticizer at the interface, causing excessive tack and the risk of glass frame adhesion. The practical window is narrow because adhesion control salts such as potassium formate and magnesium acetate are also partitioned between the PVB bulk and the glass surface. These salts are hygroscopic; their local concentration at the interface is affected by plasticizer type, moisture conditioning time, and autoclave temperature. Laminators running autoclave cycles at 130–150 °C and 1.0–1.2 MPa observe that interlayer rolls stored too long after unwrapping show different adhesion behavior than freshly conditioned rolls, because plasticizer migration over storage time changes the surface concentration of adhesion-control additives. Measurement of interlayer adhesion after lamination is commonly performed using pummel tests at -18 °C or through peel tests referenced in national safety-glass specifications; these results are sensitive to plasticizer distribution because the interlayer tears cohesively when the plasticizer content is uniform and adhesively when the interface is depleted.

Thermal and Shear History Governs Partitioning Uniformity Across Melt-Fracture Boundaries

Plasticizer homogeneity at the die exit is established in three zones: solid conveying, plastication, and distributive mixing. In co-rotating twin-screw extrusion, the specific energy input is the product of motor torque and screw speed divided by throughput; it controls the rate and completeness of plasticizer absorption. At low melt temperature below 150 °C, the plasticizer remains partially undissolved in the melt and forms discrete elongated pockets that appear as optical haze bands in the sheet. Above 220 °C, thermal oxidation of the butyral ring and transesterification reactions create volatile by-products that complicate devolatilization and can leave gel specks at the die lips. The melt temperature range commonly specified for sheet extrusion of plasticized PVB is therefore 180–220 °C, measured at the die adaptor. This window is not broad: a temperature excursion of ±5 °C near the upper boundary can shift volatile emissions and increase melt pump inlet pressure fluctuations. Screw designs with aggressive kneading blocks generate the shear necessary to dissolve plasticizer but also raise local melt temperature; production lines therefore use barrel cooling zones and vacuum vent ports to hold the melt within the limits. Choker bar adjustments in the sheet die are used to compensate for slight melt distribution differences, but the die cannot correct a large partition non-uniformity. Downstream, the cooling roll temperature is held low enough to freeze the surface quickly, typically in the 10–25 °C range, but rapid surface quenching traps plasticizer in a non-equilibrium state. If the sheet is not wound within a controlled time, the plasticizer concentration gradient between the quenched surface and the warmer core becomes sufficiently large to produce curl and surface tack. Tension control in winding also matters: high winding tension increases interlayer pressure, reduces the free volume available for plasticizer migration, and can produce blocking defects at the roll core.

Adhesion, Free Volume, and Hydroxyl Site Distribution

The glass-transition temperature of plasticized PVB can be described by the Fox equation, 1/Tg = w1/Tg1 + w2/Tg2, but the Fox equation assumes ideal mixing and fails to capture the composition-dependent partitioning between hydroxyl-rich and butyral-rich microdomains. A modified Gordon-Taylor equation with an interaction parameter is more appropriate. The plasticizer efficiency is defined as the depression of Tg per unit mass fraction of plasticizer; it depends on the free volume contribution of the plasticizer and on its hydrogen-bonding affinity for residual hydroxyl groups. Esters with higher polarity and lower molecular mass, such as dibutyl sebacate, may depress Tg more efficiently but have higher volatility and a greater tendency to migrate. High molecular mass ester plasticizers, such as triethylene glycol di-2-ethylhexanoate, exhibit lower volatility and slower migration but require more mechanical work to distribute. The solubility parameter of the plasticizer should lie close to that of the butyral-rich phase, typically 17–19 MPa^1/2, to avoid phase separation. Residual hydroxyl content is the principal adsorption site for polar plasticizers, and increasing hydroxyl content increases the apparent solubility but decreases diffusion rate. That is why two PVB grades with the same plasticizer loading can show different tack, haze, and adhesion after lamination. The interlayer must retain a minimum residual hydroxyl content to provide sufficient sites for glass adhesion, but excess hydroxyl content increases moisture sensitivity and can make the plasticizer distribution unstable under humid storage. Formulators therefore balance residual hydroxyl content, acetate content, plasticizer type, and plasticizer loading against the required pummel adhesion and optical clarity. The melt processing result is not simply a solution of plasticizer in PVB; it is a frozen-in distribution of plasticizer-rich and plasticizer-lean domains that evolves during storage.

Because plasticizer partitioning during sheet extrusion is not spatially uniform, in-line measurement of thickness and optical quality becomes a surrogate for distribution stability. Beta gauges and laser thickness sensors installed after the cooling roll detect thickness variation that correlates with local plasticizer concentration and die flow irregularities. Haze meters based on ASTM D1003-21 are used offline on sample sheets to quantify light scattering caused by plasticizer-rich domains or surface exudation. Melt pressure transducers at the die adaptor and melt pump inlet provide real-time evidence of unstable plastication: pressure fluctuations above ±0.2 MPa at constant throughput often indicate that plasticizer is not being absorbed uniformly. In some production lines, infrared absorption is used to measure the total plasticizer content in the sheet, but this method does not resolve spatial partitioning between edge and center. For that reason, edge samples are tested separately from center samples when qualifying a new formulation. Process engineers also monitor screw torque and specific energy consumption as indirect indicators of plasticizer distribution; a sudden torque drop at constant screw speed can indicate plasticizer slippage or melt fracture, while a torque rise can indicate resin consolidation and insufficient free volume. The response of PVB to excessive shear is different from that of polyolefins because the polymer is relatively thermally sensitive and can undergo chain scission or crosslinking depending on oxygen availability. Nitrogen blanketing of the feed hopper and vacuum vent ports reduces oxidative degradation but does not eliminate plasticizer migration after sheet winding.

When Extrusion Temperature Exceeds 220°C in High-Viscosity PVB Grades

Thermal degradation at elevated barrel temperatures produces gel specks, discoloration, and plasticizer loss. High-viscosity PVB grades with residual hydroxyl content above 22 wt% generate more shear heat and require higher torque at equivalent screw speed. At melt temperatures above 220 °C, oxidative cleavage of the butyral ring liberates butyraldehyde, while residual acetate groups may release acetic acid in the presence of moisture. The released aldehydes and acids lower melt pH and can attack the ester plasticizer, producing free acids and alcohol by-products that plate out on downstream cooling rolls. This condition is most likely during grade transitions, screen-pack changes, or extended residence time after a line stoppage. Production-scale equipment includes hydraulic screen changers and gear pumps, but even brief residence time excursions above 30 min at elevated temperature can create visible gel contamination in subsequent sheet. Mitigation requires strict control of melt temperature, vacuum level, and throughput; the use of barrel vent ports with vacuum below 50 mbar helps strip low molecular weight volatiles, but if the plasticizer itself has high vapor pressure at the vent temperature, excessive vacuum draws plasticizer out of the formulation. That conflict is one reason high molecular weight plasticizers are preferred for sheet extrusion: their lower vapor pressure permits deeper devolatilization without changing the final plasticizer content. When a formulation change to a lower molecular weight plasticizer is made, the vent vacuum and barrel temperature setpoints must be retuned to avoid plasticizer loss. The optimum vacuum is confirmed by measuring plasticizer content in the sheet before and after vent changes; values outside the specified formulation tolerance of ±1 phr require corrective action. Resin pre-drying is required when feedstock moisture exceeds 0.5 wt%; otherwise moisture-driven hydrolysis of the plasticizer and the polymer reduces molecular weight and changes the partitioning behavior.

Process limitations for additive packages arise when adhesion control salts, UV stabilizers, and anti-block agents compete for plasticizer hydrogen-bonding sites. Potassium formate, magnesium acetate, and similar adhesion control salts are compounded at low addition levels, often below 0.1 phr, because their ionic nature creates local plasticizer exclusion zones. Benzotriazole UV stabilizers and hindered amine light stabilizers are generally incorporated at 0.1–0.5 wt% of resin; their solubility in the plasticizer phase determines whether they remain uniformly distributed or migrate to the sheet surface. If the stabilizer is too soluble in the plasticizer, it migrates with the plasticizer to the surface, where it can crystallize as a surface haze or interfere with glass adhesion. If it is insoluble, it may remain as discrete particles that scatter light. Antistatic agents based on ethoxylated amines can complex with residual acetate groups and displace plasticizer from polar domains, altering the partition coefficient. The choice of antioxidant is constrained by the same partitioning behavior: phenolic antioxidants must remain in the PVB phase under prolonged thermal oxidation and must not be extracted by the plasticizer into surface layers where they can stain glass. These interactions are not described by simple binary solubility data; they require compounding trials on production-scale twin-screw extruders with defined L/D ratios, followed by lamination tests according to ISO 12543-2:2021 and optical measurements according to ASTM D1003-21. Published data for this specific configuration is limited for some additive combinations, and the operational window must be confirmed on the actual line because small variations in resin hydroxyl content, moisture level, and roll storage time change the surface concentration of plasticizer and additives enough to affect lamination yield.