In high-temperature polyester synthesis for coil and can coating resins, termination of a hydroxy-functional backbone with trimellitic anhydride (TMA) introduces terminal aromatic dicarboxylic acid functionality while simultaneously increasing the storage modulus of the subsequently crosslinked film. The ring-opening monoesterification of TMA at a primary hydroxyl chain end is usually conducted at 170–200°C under inert gas after the melt polycondensation has reached a hydroxyl number in the range of 20–60 mg KOH/g and an initial acid value below 8 mg KOH/g. Each fully reacted TMA unit contributes a 4-position carboxylic acid group plus the opened anhydride carboxylic acid group, yielding a theoretical acid value increase of approximately 5.84 mg KOH/g for every 1 wt% TMA charged on resin solids. The actual acid value gain falls between 85% and 95% of theory when the endcapping temperature is kept below 200°C, because the free aromatic acid can undergo secondary esterification and branch formation at higher temperatures. Acid value is titrated on quenched samples dissolved in a toluene–isopropanol–water mixture according to ISO 2114:2000 or ASTM D4662-20, and the value is expressed in mg KOH/g. Storage modulus of cured free films is measured by dynamic mechanical analysis in dual-cantilever or tensile mode at 1 Hz and a heating rate of 3 K/min according to ISO 6721-5:2019. The storage modulus at 25°C and the rubbery plateau modulus at T_g + 40°C are recorded because the former reflects ambient hardness and the latter reflects crosslink density by rubber elasticity theory. For densely crosslinked TMA-terminated polyester networks cured with epoxy or melamine-formaldehyde crosslinkers, the glassy storage modulus at 25°C commonly falls between 1.8 GPa and 3.2 GPa, while the rubbery plateau modulus may range from 8 MPa to 30 MPa depending on equivalent weight and functionality. The presence of the rigid trimellitic aromatic ring reduces free volume and increases cohesive energy density, so the storage modulus is not solely a function of crosslink density; monomer stiffening and hydrogen bonding from residual acid groups also contribute. These effects make TMA endcapping a dual-functional design tool for formulators that must achieve both a defined acid value for pigment wetting or water dispersibility and a higher cured-film modulus for abrasion resistance and can-lining performance.

Can Trimellitic Anhydride Endcapping Raise Acid Value Without an Undesirable Reduction in Rubbery Plateau Storage Modulus?

The relationship between acid value and storage modulus is not linear because acid value depends on the concentration of titratable carboxyl groups, whereas storage modulus depends on the molecular weight between crosslinks, chain stiffness, and the degree of cure. A TMA-capped resin with an acid value of 30 mg KOH/g has an acid equivalent weight of 56100/30 = 1870 g/eq. When this resin is cured with a stoichiometric amount of a bisphenol A diglycidyl ether epoxy resin or a phenolic resole, each acid group can form an ester or a hydroxy ester link, and the resulting network architecture is controlled by the average carboxyl functionality introduced by chain ends. The theoretical acid value contribution from TMA can be calculated from its molecular weight of 192.13 g/mol and its two titratable carboxyl groups per molecule after monoester ring opening. Because one gram of TMA provides approximately 584 mg KOH equivalents, a 1 wt% loading on resin solids corresponds to 5.84 mg KOH/g. This relationship is used in resin development to estimate the endcapping charge required to shift a resin from an acid value of 5 mg KOH/g to a target of 35 mg KOH/g; about 5.6 wt% TMA would be required at 90% efficiency, minus any acid consumed by secondary esterification or neutralization. Storage modulus in the rubbery plateau can be converted to crosslink density using ν_e = E′/3RT, where E′ is the storage modulus in Pa, R is 8.314 J/mol·K, and T is the absolute temperature in kelvin. A film with a rubbery plateau modulus of 20 MPa at 393 K would therefore have an apparent crosslink density of approximately 2040 mol/m³. TMA endcapping generally increases this apparent crosslink density because the terminal acid groups allow a hydroxy-terminated polyester to be cured as a multifunctional prepolymer rather than as a linear chain extended system. However, an acid value above a formulation-specific ceiling can reduce the rubbery plateau modulus indirectly by leaving unreacted free acid in the network, increasing water sensitivity and plasticization in humid service, and by accelerating overcure embrittlement. For this reason, TMA endcapping is evaluated not only by the final acid value but also by the ratio of the rubbery plateau modulus to the glassy modulus and by the width of the tan δ peak, which reflects network homogeneity.

Production-scale batches of TMA-capped polyester for coil primers are typically processed in stainless-steel polycondensation reactors of 15–30 m³ capacity, equipped with a helical ribbon agitator operating at 40–60 rpm, a partial condenser set to return diols to the reactor while removing water, and a vacuum system capable of less than 50 mbar for final stripping. The polycondensation is carried out at 230–245°C until the hydroxyl number reaches the target window and the acid value drops below 8 mg KOH/g. The reactor is then cooled to 170–190°C under a nitrogen sparge, and molten TMA is charged through a feed hopper over 20–40 min. The endcapping reaction is exothermic enough to require a controlled cooling jacket, and the temperature must be held within a ±5°C band around the setpoint because temperatures above 195°C promote esterification of the free 4-position carboxyl group and cause uncontrolled branching, while temperatures below 165°C slow the ring-opening reaction and can leave unreacted TMA that sublimes in the vacuum system. Acid value samples are taken from the bottom valve every 15–20 min, quenched and dissolved in a solvent mixture, and titrated potentiometrically. Batch-to-batch variation in acid value of more than ±2 mg KOH/g is observable on a production line as a shift in waterborne dispersion viscosity after neutralization and as a change in coil coating film hardness. For solventborne high-solids coil primers, the resin is cooled and discharged through a 20 µm filter; filtration pressure can exceed 3 bar when microgel particles form from localized overheating. The resin melt viscosity at 200°C is measured with a cone-and-plate viscometer, and a value above 60 Pa·s usually indicates excessive branching or TMA overdosing. The critical processing window is narrow because acid value and storage modulus are coupled through the same terminal functionality, and a deviation in TMA charge of more than 0.3 wt% can shift the acid value by more than 1.5 mg KOH/g and alter the crosslink density of the final coated film by a detectable DMA change.

When TMA Is Added as a Post-Endcapping Correction Within the Final 60 Minutes of a Resin Cook

When a polyester resin cook has already reached the target hydroxyl number and an acid value below the specification, staged addition of TMA can be used as a final correction, but the acid value ceiling is set by the number of residual hydroxyl chain ends available for ring-opening esterification. If the hydroxyl number before TMA addition is 25 mg KOH/g, the resin contains approximately 0.446 mmol hydroxyl groups per gram; complete capping of all hydroxyl groups with TMA would consume roughly 0.446 mmol TMA per gram, equivalent to 85.7 mg TMA per gram or 8.6 wt% TMA. The theoretical acid value contribution from this TMA charge is about 50 mg KOH/g, assuming both carboxyl groups remain free. In practice, the acid value does not reach the theoretical value because the free aromatic acid partially esterifies at 190–200°C and because steric hindrance at the chain end reduces the ring-opening rate. The final 60 minutes of the cook are therefore not a true correction window; they are a controlled post-polycondensation stage that must be monitored by acid value and melt viscosity every 10 min. Unreacted TMA above approximately 0.5 wt% on resin solids can remain as a dissolved crystalline monomer in the cooled resin and produce a double tan δ peak in cured films, a reduction in the storage modulus at 25°C because of plasticization, and an acid value titration overshoot that does not correspond to bound carboxyl functionality. To avoid this condition, the reactor must be held at 175–185°C until the acid value slope falls below 0.3 mg KOH/g per 10 min sampling interval. Vacuum stripping after TMA addition is generally avoided because the free acid groups can catalyze ester exchange and because molten TMA can sublime and block vacuum lines. If the acid value overshoots the specification by more than 2 mg KOH/g, reformulation with a lower functional chain extender or dilution with a non-functional polyester is required; adding more diol to consume acid groups is not recommended because it can reduce storage modulus and introduce unreacted glycol that blooms to the film surface. Free TMA in the cooled resin can be detected by differential scanning calorimetry as a melt endotherm near 163–168°C, or by Fourier-transform infrared spectroscopy as a residual anhydride carbonyl doublet at approximately 1780 cm⁻¹ and 1850 cm⁻¹. High-performance liquid chromatography with UV detection at 254 nm after dissolution in tetrahydrofuran provides a more quantitative free TMA content, but published data for that specific configuration is limited.

Waterborne dispersions based on TMA-capped resins are neutralized at 60–80°C with N,N-dimethylethanolamine at 80–100% of the titratable carboxyl groups, then dispersed under high shear in deionized water. The acid value before neutralization is usually maintained between 40 and 60 mg KOH/g for sufficient electrostatic stabilization, and the neutralized resin exhibits a translucent to opaque dispersion with a particle size below 300 nm when measured by dynamic light scattering. In continuous production, a rotor-stator disperser or a twin-screw extruder with an L/D ratio of 48:1 may be used for the inversion step, and the torque curve during water addition is used to detect batch-to-batch shifts in acid value or molecular weight. Films cast from these dispersions and cured with a waterborne epoxy or phenolic crosslinker are evaluated by DMA after preconditioning at 23°C and 50% relative humidity for 24 h. The storage modulus at 25°C typically increases from below 0.5 GPa for a low-cure control to 1.5–2.5 GPa after crosslinking, but residual neutralizer can plasticize the film and reduce the glassy modulus if the cure schedule does not allow full evaporation of dimethylethanolamine. At relative humidity above 60%, waterborne films must be pre-dried at 25–30°C under forced air for 15–30 min before reaching the peak metal temperature, otherwise microblisters form and DMA specimens show premature failure below 1% strain. High shear dispersion equipment must be cleaned between batches because dried carboxylated resin adheres to the rotor-stator gaps and can create seed particles that raise the final particle size distribution. TMA-capped waterborne resins should not be formulated with zinc oxide or calcium carbonate pigments at high acid values because zinc and calcium ions form ionic carboxylate bridges that cause a rapid viscosity rise and can gel the dispersion within hours. Published data for waterborne can coating formulations with TMA-capped polyester and zinc oxide are limited; laboratory storage stability testing at 40°C for 30 days is required before production use.

Acid Value Titration Interferences and Solvent Selection for TMA-Capped Resins

Titration of high-acid-value TMA-capped resins requires complete dissolution in a solvent mixture that can break hydrogen-bonded acid clusters without hydrolyzing the ester backbone. A mixture of toluene, isopropanol, and distilled water in a 1:1:0.2 volume ratio is commonly used for polyester resins, with a sample size of 2–5 g depending on the expected acid value. The titrant is 0.1 mol/L ethanolic potassium hydroxide standardized against potassium hydrogen phthalate, and the endpoint is detected potentiometrically rather than by phenolphthalein because the aromatic carboxylic acid groups of TMA have a buffer region that produces a weak visual endpoint. Free TMA monomer titrates differently from bound TMA, and if unreacted TMA crystallizes in the sample, the acid value can be overestimated by more than 3 mg KOH/g unless the sample is filtered and washed with isopropanol. For resins containing hydrolyzable ester linkages, the sample should be kept below 25°C during dissolution and titrated within 15 min to avoid saponification of the polyester backbone, which would liberate carboxylate groups and distort the acid value. ASTM D4662-20 and ISO 2114:2000 provide the reference procedures, but ISO 2114:2000 is preferred for unsaturated polyester resins and alkyds because it separates partial acid value from total acid value. The measured acid value is used to calculate the carboxyl equivalent weight as 56100 divided by the acid value in mg KOH/g. The storage modulus of the cured film is not measured on the same neat resin but on free films prepared by drawdown on tinplate, cured at a defined peak metal temperature, and then mechanically separated. Specimen dimensions for DMA are typically 60 mm in length, 10 mm in width, and 20–40 µm in thickness, and at least 3 specimens are tested to report an average storage modulus. The following compliance matrix summarizes the primary test methods and measurement conditions.

For DMA on free films, residual solvent and free TMA monomer are the two most common sources of storage modulus artifacts. Films prepared by drawdown over tinplate and cured at 200°C for 10 min may retain less than 2% volatile content when tested by gas chromatography according to ISO 11890-2:2020. A solvent content above 2% lowers the glassy storage modulus at 25°C by 10–20% relative to an identical formulation forced-air-dried for 15 min before cure. Free TMA monomer acts as a crystalline plasticizer below its melt endotherm near 163–168°C; above that temperature it can react with the network and distort the rubbery plateau modulus. DMA specimens must be inspected under polarized light for crystallites and must have thickness variation below 10% across the gauge length. A strain sweep at 1 Hz from 0.01% to 1.0% is performed before each temperature ramp to confirm linear viscoelastic response; nonlinear strain response above 0.2% is a sign of microphase separation, unreacted monomer domains, or network heterogeneity. The storage modulus at 25°C is taken from the resulting temperature sweep only if the storage modulus at the end of the strain sweep returns to within 2% of the initial value.

For powder coating polyester resins cured with β-hydroxyalkylamide or triglycidyl isocyanurate, TMA endcapping is used to raise the acid value to the stoichiometric window required by the crosslinker without increasing melt viscosity as much as chain extension with an aromatic diacid would. A TGIC-cured polyester typically requires an acid value of 30–45 mg KOH/g, while a β-hydroxyalkylamide-cured polyester typically requires 20–30 mg KOH/g. The resin, crosslinker, flow agent, and pigments are compounded in a co-rotating twin-screw extruder with an L/D ratio between 25:1 and 40:1, with zone temperatures maintained at 90–110°C and screw speeds between 300 and 500 rpm to disperse the formulation without initiating cure. The extrudate is cooled on a chilled roll, crushed, and milled to a median particle size of 30–40 µm before electrostatic application. Storage modulus of the cured powder film is evaluated by DMA on free films cured at 180–200°C for 10–15 min. The glassy storage modulus at 25°C for TGIC-cured TMA-capped polyester is normally in the range of 2.0–3.0 GPa, and the rubbery plateau modulus at T_g + 40°C is typically above 10 MPa when the acid value is within specification. Overcure by more than 5°C above the recommended peak metal temperature can shift the storage modulus upward but reduce reverse impact resistance because the network becomes excessively crosslinked. Undercure by more than 5°C leaves unreacted carboxyl groups and can reduce the rubbery plateau modulus by more than 30% relative to the fully cured state. Powder coatings based on TMA-capped acid-functional polyester must not be premixed with amine-catalyzed epoxy resins in a single screw extruder because the acid groups initiate premature ring-opening and can gel the extruder barrel. The processing window for these formulations is narrow, and batch records should include extruder barrel pressure, torque, and melt temperature because an increase in torque of more than 15% during compounding often indicates that the acid value is above the target range. Published data for this specific configuration is limited when non-standard pigments such as zinc oxide are included; applicator trials and dynamic mechanical analysis are required to confirm storage modulus retention after 500 h of humidity exposure according to ISO 6270-2:2017.