In medium-oil alkyd syntheses where xylene reflux and partial-condenser water removal govern esterification, di-trimethylolpropane (Di-TMP) has been evaluated as a branched tetraol replacement for pentaerythritol and glycerol to achieve higher molecular weight without increasing hydroxyl excess. Di-TMP is 2,2′-[oxybis(methylene)]bis[2-ethylpropane-1,3-diol] with four primary hydroxyl groups and an equivalent weight of approximately 62.6 g/eq; its ether-linked dimer structure introduces a less compact branching geometry than pentaerythritol while retaining a hydroxyl functionality of 4. In alkyd reactors of 45,000 L capacity equipped with thermosyphon partial condensers, turbine agitators, and xylene-filled Dean-Stark traps, the polyol charge is first subjected to alcoholysis with refined soybean oil at 240–250 °C under inert-gas sparging. Lithium hydroxide or calcium hydroxide at 0.03–0.10 wt% of oil charge catalyzes ester interchange, and the formation of monoglyceride and diglyceride intermediates is monitored by solubility in anhydrous methanol. Subsequent esterification with phthalic anhydride and isophthalic acid at 230–240 °C removes water of reaction, and the acid value is titrated according to ISO 2114:2000 until the endpoint of 8–12 mg KOH/g resin solids is reached. The term hydroxyl excess is defined as [(equivalents OH − equivalents COOH)/equivalents COOH] × 100%. Conventional medium-oil formulations operate with hydroxyl excess between 15% and 25% to suppress gelation, but Di-TMP branching shifts the practical lower boundary because its primary hydroxyl groups provide a more uniform esterification profile than the mixed primary/secondary hydroxyls of glycerol and the crystalline steric hindrance of pentaerythritol.

How Does Di-TMP Branching Affect the Flory-Stockmayer Gelation Boundary?

The gelation boundary for a step-growth polyester containing Di-TMP can be estimated from the Flory-Stockmayer critical conversion α_c = 1/(f_avg − 1), where f_avg is the average functionality of the reacting polyol and polyacid mixture. When Di-TMP replaces pentaerythritol at constant oil length, f_avg remains close to 2.0 to 2.3 because the dibasic acids and fatty monobasic acids dominate the overall functionality, but the branch points become more uniform and less crystalline. The practical consequence is that the hydroxyl excess can be lowered from 15–25% to 6–10% without crossing the gelation boundary at the endpoint acid value, provided the acid value is not driven below 6 mg KOH/g and the reactor is not held above 240 °C for more than 90 min. The Carothers equation x_n = (1 + r)/(1 + r − 2rp) indicates that reducing hydroxyl excess moves the stoichiometric ratio r closer to 1.0, raising number-average degree of polymerization more rapidly as conversion p increases. In plant reactors, the corresponding increase in molecular weight appears as a rapid rise in torque at the final esterification stage. A 75 kW agitator drive may increase from 55 A to 80 A within 15 min when the acid value falls below 10 mg KOH/g, and this is the principal processing constraint rather than any limitation in Di-TMP reactivity.

Batch-to-batch variance in Di-TMP dimer distribution, free TMP content, and residual sodium sulfate from synthesis affects the onset of torque rise. Di-TMP produced by acid-catalyzed etherification of trimethylolpropane may contain 2–8 wt% residual TMP and minor oligomeric ethers; the residual TMP acts as a trifunctional chain extender and alters the effective equivalent weight. When a 45,000 L scrape-wall esterification reactor is charged with Di-TMP having 95.0 wt% assay, the hydroxyl excess calculation based on 62.6 g/eq is inaccurate. Actual equivalent weight is derived from the hydroxyl number measured by DIN 53240-2 after drying. This laboratory value should be obtained for every incoming lot, and the charge ratio should be recalculated, because a 1.0% error in Di-TMP equivalent weight at 8% hydroxyl excess can shift the Carothers r by approximately 0.008, enough to alter final viscosity by 10–15% at constant acid value. Experience with multiple campaigns in a 45,000 L reactor indicates that pre-drying Di-TMP to ≤0.10 wt% water before charging reduces esterification time and minimizes foaming in the partial condenser. Raw materials stored at relative humidity above 60% must be dried because water consumes anhydride, lowers measured acid value temporarily, and delays endpoint determination.

Polyol	Hydroxyl functionality	Equivalent weight	Hydroxyl number	Physical form
Di-trimethylolpropane	4	62.6 g/eq	890–910 mg KOH/g	Waxy crystalline solid, melting 108–115 °C
Trimethylolpropane	3	44.7 g/eq	1230–1260 mg KOH/g	Free-flowing flakes, melting 58–62 °C
Pentaerythritol	4	34.0 g/eq	1620–1660 mg KOH/g	Crystalline, melting 258–262 °C
Glycerol	3	30.7 g/eq	1800–1830 mg KOH/g	Viscous liquid, boiling 290 °C

Surface tack development in oxidatively cured alkyd enamels is assessed according to ASTM D1640-14 at 23 ± 2 °C and 50 ± 5% relative humidity using a dry film thickness of 35 μm. In these tests, the transition from a fluid, brush-marked film to a tack-free surface occurs when autoxidative polymerization of the unsaturated fatty acid side chains has generated sufficient molecular weight and polar crosslink density to resist cotton-fiber adhesion. Cobalt-based primary driers at 0.03–0.08% metal on resin solids, combined with zirconium and calcium auxiliary driers at 0.5–1.5% and 0.1–0.4% respectively, initiate decomposition of allylic hydroperoxides and accelerate oxygen uptake. The lower hydroxyl excess of Di-TMP branched alkyds influences drying through two coupled mechanisms. First, higher molecular weight at equal acid value reduces the number of chain ends and increases the number of unsaturated fatty acid chains per polymer molecule; therefore fewer oxidative crosslinks are required to immobilize the surface. Second, residual unreacted hydroxyl groups act as polar hydrogen-bonding sites that associate with cobalt and water, delaying the development of a coherently crosslinked surface. Reducing hydroxyl excess from 15% to 7% lowers the concentration of terminal hydroxyl groups and reduces moisture retention, while Di-TMP branching maintains solubility and avoids the excessive microgel formation that would occur with pentaerythritol under the same conditions. Published drying data for Di-TMP-based medium-oil alkyds at hydroxyl excess below 6% are limited, so direct tack-free comparisons must be interpreted from plant trial records and pendulum damping curves rather than from a single standard result. An operational boundary in low-hydroxyl-excess Di-TMP alkyds is the altered interaction with anti-skinning additives and amine-based neutralizing agents. Methyl ethyl ketoxime at 0.1–0.3% of resin solids is used to retard skin formation in open containers; its volatility after application must be balanced against the faster surface drying, otherwise through-dry may be delayed. Amine-based additives and ammonia-neutralized water-reducible variants should be avoided unless stoichiometrically adjusted, because primary and secondary amines can complex cobalt and reduce the available drier concentration at the film surface. For solvent-borne enamels, high-shear dispersion of cobalt drier into the letdown at 50–60 °C using a rotor-stator mixer improves drier distribution and reduces the risk of cobalt adsorption onto residual carboxylate ends at low acid value. The letdown viscosity is adjusted with white spirit to 100–120 KU at 25 °C as measured by ASTM D562, and application by HVLP requires 25–30 s in a DIN 4 cup. These conditions maintain a 35 μm dry film without sagging and allow the surface to reach stage D of ASTM D1640-14 without extending the through-dry time beyond 8 h.

When Hydroxyl Excess Drops Below 8% in a 45,000 L Reactor

Hydroxyl excess values below 8% in a 45,000 L scrape-wall esterification reactor create a processing window that is narrower than the laboratory cook suggests. At the final stage, when the acid value has fallen to 10 mg KOH/g, the reaction mass viscosity may reach 20–40 dPa·s at 60% solids in xylene, and the agitation torque climbs nonlinearly. The scrape-wall agitator with 75 kW drive must be equipped with torque interlocks set at 85% of motor nameplate to prevent a stalled agitator, because a gelation event in a 45,000 L reactor cannot be reworked once the mass reaches the Flory-Stockmayer gel point. Emergency xylene thinning through the reflux return line is used only before the torque interlock trips; after gelation begins, agitation stops and the batch must be discarded under hot xylene cleanout procedures. Maintaining the hydroxyl excess at 6–8% with Di-TMP requires endpoint control by acid value titration every 15 min, supplemented by in-line near-infrared monitoring of acid number and anhydride residual. Published data for this specific configuration is limited; plant records indicate that the acid value endpoint should not be set below 8 mg KOH/g if the Di-TMP lot contains more than 5 wt% residual TMP, because the lower effective functionality of TMP creates a broader gelation boundary but also produces lower molecular weight, and a lower endpoint may still lead to microgel domains that increase filtration pressure.

Raw material management is an additional threshold. Di-TMP absorbs moisture readily above 60% relative humidity; water in the charge above 0.10 wt% hydrolyzes phthalic anhydride to phthalic acid, altering the stoichiometric ratio and producing free acid that must be re-esterified. In plant campaigns, a 0.20 wt% water increase in Di-TMP shifted the apparent hydroxyl excess by approximately 0.7% and extended cook time by 2–3 h. Pre-drying in a vacuum tumble dryer at 80–90 °C for 4 h under 20 kPa absolute pressure is required for tropical-shore storage. The letdown solvent must also be analyzed for peroxide content if recycled from enamels; residual peroxides above 10 ppm initiate premature viscosity drift in storage and reduce the induction period after drier addition. Incompatibilities include combination with amine-based additives in the reactor or letdown; trace amines can form imide intermediates with phthalic anhydride and reduce the effective acid available for chain extension, altering the gelation boundary unpredictably. This is the principal reason why amine-neutralized water-reducible runs are isolated from solvent-borne Di-TMP alkyd production.

Viscosity, Drier Uptake, and Surface Oxygen Diffusion as Coupled Variables

The viscosity of a lower-hydroxyl-excess Di-TMP alkyd at application solids is not a simple function of molecular weight; the ether linkage in Di-TMP reduces intramolecular hydrogen bonding relative to pentaerythritol, while the compact branched architecture minimizes hydrodynamic volume at equivalent degree of polymerization. Cone-and-plate viscosity measured by ISO 2884-1:2020 at 10,000 s⁻¹ on a 60% solids xylene cut is typically lower than a pentaerythritol analog at the same oil length by 10–20%, enabling higher solids application without exceeding 120 KU at 25 °C. Drier uptake in the letdown is influenced by residual acid value, hydroxyl number, and the presence of free phthalic acid; lower hydroxyl excess reduces the polar adsorption sites for cobalt, but the higher molecular weight also slows diffusion of cobalt in the film during the first minutes after application. This diffusion coupling means that surface oxygen uptake must outpace cobalt immobilization; if cobalt drier is added at more than 0.08% metal on resin solids, excessive surface crosslinking may create an oxygen barrier that delays through-dry even as tack-free time decreases. Surface oxygen diffusion is also controlled by film thickness; at 35 μm dry film, oxygen diffusion into the lower layers is sufficient for through-dry, but above 60 μm dry film the tack-free stage may precede development of hardness, and sag resistance must be controlled with organoclay or fumed silica at 0.5–1.5 wt% of resin solids.

High-shear dispersion of driers and anti-skinning agents in the letdown is performed with a rotor-stator mixer at 3,000–5,000 rpm for 15–20 min, and the temperature is maintained below 60 °C to avoid premature drier decomposition. Batch-to-batch variation in drier uptake is monitored by drawdown on glass panels with 100 μm wet film; the tack-free stage is recorded according to ASTM D1640-14, and through-dry is confirmed by pressure with a 100 g weight under ISO 9117-3:2010. Pendulum hardness development is followed for 7 d using ASTM D4366-16; a Di-TMP-based medium-oil formulation at lower hydroxyl excess typically reaches 70–90% of its ultimate König hardness within 48 h, but published data for this specific configuration is limited and values vary with fatty acid source and drier package. The use of tung oil or linseed oil blends raises the diene and triene content, accelerating surface dry, but these oils also increase yellowing; therefore soybean oil remains the dominant fatty acid source for industrial enamels where color retention is required.

Property	Test method	Condition or reporting unit
Tack-free time	ASTM D1640-14	23 ± 2 °C, 50 ± 5% RH, 35 μm DFT
Through-dry state	ISO 9117-3:2010	100 g weight, 25 °C
Pendulum hardness	ASTM D4366-16	König pendulum, 7 d cure
Acid value	ISO 2114:2000	mg KOH/g resin solids
Hydroxyl number	DIN 53240-2	mg KOH/g polyol or resin
High-shear viscosity	ISO 2884-1:2020	10,000 s⁻¹, 25 °C
Non-volatile matter	ISO 3251:2019	105 °C, 3 h
VOC content	ASTM D2369-20	EPA Method 24 basis
Color	ASTM D1544-18	Gardner scale

In high-solids air-drying topcoats for agricultural equipment, lower-hydroxyl-excess Di-TMP alkyds are applied at 55–65% solids by volume over solvent-wiped steel primed with two-component epoxy. The faster tack-free surface reduces dust pickup in open factory bays, but the formulation must be adjusted for storage stability and sag resistance. Because lower hydroxyl excess leaves fewer terminal hydroxyl groups available for hydrogen bonding with anti-settling agents, fumed silica at 0.5–1.5 wt% is dispersed under high shear to prevent pigment settling in 200 L drums. The batch is filtered through a 20 μm bag filter after letdown; higher-molecular-weight Di-TMP alkyds at hydroxyl excess below 6% may show filter pressure rise from microgel particles formed during the final cook stages, and polishing filtration with 10 μm bags is required for airless spray applications. Incompatibility with amine-based additives is most evident during tinting; some colorant concentrates contain amine-functional wetting agents that complex cobalt and extend dry times, so tinting bases must be qualified by drawdown before color-match approval. The operational boundary for relative humidity is 60%; above this level, pre-drying of pigments and Di-TMP resin is required because surface moisture delays tack-free development and can produce haze in high-humidity spray booths. Published long-term exterior durability data for lower-hydroxyl-excess Di-TMP alkyds are limited; accelerated weathering by ISO 16474-3 exposure in QUV-B cycles shows that Di-TMP branching reduces chalking relative to pentaerythritol at the same oil length, but the faster surface dry must not be achieved by excessive cobalt levels, which can accelerate oxidative embrittlement.