Bouling Chemical Co., Limited
Section

Articles

Pentaerythritol Initiation in Rigid Polyol Synthesis

Pentaerythritol, 2,2-bis(hydroxymethyl)propane-1,3-diol, enters rigid polyether polyol synthesis as a tetrafunctional starter with a theoretical hydroxyl number of 1648 mg KOH/g and a molar mass of 136.15 g/mol. The four equivalent primary hydroxyl groups undergo base-catalyzed alkoxylation with propylene oxide to generate a four-arm, short-chain polyether structure; the target hydroxyl number for rigid foam applications is generally 350–650 mg KOH/g, corresponding to an equivalent weight of 160–86 g/eq and a number-average molar mass of 640–345 g/mol at nominal functionality 4.0. Nominal functionality is 4.0, but potassium-hydroxide-catalyzed propylene oxide homopolymerization also generates allyl alcohol through a concurrent isomerization pathway; the resulting monofunctional unsaturated polyether reduces the measured average functionality to 3.6–3.9 depending on reaction severity and final chain length. Hydroxyl number is determined by ASTM D4274-21, water content by ASTM D4672-18, viscosity by ASTM D4878-15, acid number by ASTM D4662-20, and unsaturation by ASTM D4671-16. The polyol is often produced as a co-initiated starter blend containing 10–50 wt% pentaerythritol and a liquid co-starter such as glycerol or propylene glycol, which lowers the initial charge viscosity and ensures complete dissolution of the solid tetraol before propylene oxide feeding commences. The alkoxylation reaction follows the stoichiometry C(CH₂OH)₄ + 4n C₃H₆O → C[CH₂(OCH₂CH(CH₃))ₙOH]₄, with optional ethylene oxide addition at the final stage to convert secondary hydroxyl chain ends to more reactive primary hydroxyl groups. Because the starter is a solid with a melting point above 250 °C, direct melting before alkoxylation is thermally impractical and introduces oxidative color risk; production routes instead rely on dissolution into a heated liquid co-starter or a heel of previously alkoxylated polyol.

Starter Preparation and Alkoxylation Reactor Configuration

Production-scale alkoxylation of pentaerythritol is conducted in a 1.4571 stainless-steel pressure vessel designed to ASME BPVC Section VIII Division 1 or PED 2014/68/EU, with a working pressure of 1.0 MPa and a typical design pressure of 1.6 MPa; a 20 m³ reactor is representative for multi-product polyol campaigns. The vessel is equipped with a helical ribbon or anchor agitator operating at 40–90 min⁻¹, an internal cooling coil, a jacket supplied with tempered water at 90–150 °C, and a propylene oxide dip tube entering below the liquid surface. Pentaerythritol powder is charged into a liquid heel of partially alkoxylated starter or glycerol; the heel is heated to 110–125 °C and agitated for 30–60 min before catalyst addition to avoid solid accumulation on the vessel floor. Potassium hydroxide is introduced as a 45–48 wt% aqueous solution at 0.10–0.35 wt% KOH on final polyol mass; the water of solution and the water formed by alkoxide generation are removed under vacuum at 10–20 kPa absolute until the residual water is below 0.03 wt% by ASTM D4672-18. Propylene oxide is then fed at a rate controlled by the heat-removal capacity of the reactor; the ring-opening exotherm is approximately 90 kJ/mol, and failure to maintain jacket return water below 130 °C produces a temperature excursion that accelerates allyl alcohol formation and increases APHA color above 50. After complete propylene oxide addition, the batch is digested at 115–125 °C for 2–4 h until the pressure plateau remains stable within ±5 kPa for 30 min. Residual catalyst is neutralized with 85 wt% phosphoric acid or 30 wt% sulfuric acid to an acid number of 0.03–0.08 mg KOH/g, then vacuum-stripped at 10–20 kPa and 115–125 °C for 2–3 h to remove residual propylene oxide to below 10 mg/kg. The neutralized batch passes through a 10–20 µm filter before transfer to finished polyol storage, where nitrogen blanketing at 50–100 kPa prevents carbon dioxide ingress and moisture uptake.

During the initial nucleophilic ring-opening step, the apparent reaction rate is governed by catalyst concentration, propylene oxide concentration in the liquid phase, and the availability of free alkoxide rather than by simple first-order oxirane kinetics. At 115–125 °C and a total pressure of 0.35–0.55 MPa, a KOH loading of 0.20 wt% typically gives a propylene oxide addition time of 3–6 h for a 12–15 t batch; reducing catalyst to 0.10 wt% extends the feed time by 40–70% but lowers terminal unsaturation, while increasing catalyst to 0.35 wt% shortens the cycle but raises the propylene oxide isomerization rate. The processing window is narrow: below 105 °C, propylene oxide conversion falls rapidly because the alkoxide initiation of the fourth pentaerythritol arm remains incomplete, producing a polyol with an effective functionality of 3.2–3.5 and an upward-drifting hydroxyl number; above 135 °C, propylene oxide isomerization to allyl alcohol becomes kinetically relevant, generating monofunctional species that plasticize the later rigid foam and reduce crosslink density. Reactor pressure is not an independent variable but a consequence of propylene oxide feed rate, temperature, and gas-liquid mass transfer; if the pressure exceeds 0.60 MPa at 120 °C, the feed is interrupted automatically by the safety PLC, and the batch is cooled until the pressure returns to the operating envelope. The propylene oxide feed is restarted only after a 10–15 min soak confirms that the pressure decline is caused by consumption rather than by venting. Conversely, if the pressure falls below 0.15 MPa, the liquid-phase propylene oxide concentration is insufficient to maintain the target addition rate, and the reactor enters a starvation regime that broadens arm-length distribution. Batch-to-batch variance on a 20 m³ unit is typically ±2 mg KOH/g for hydroxyl number and ±10% for viscosity when the starter blend is pre-melted and the water content is controlled; plants that charge dry pentaerythritol powder without a liquid heel observe occasional hydroxyl number excursions of ±6 mg KOH/g due to incomplete initiator dissolution. Carbon dioxide ingress must be prevented because it neutralizes the alkoxide catalyst; reactors are blanketed with nitrogen at 50–100 kPa, and the propylene oxide feed is maintained oxygen-free. Dissolved oxygen above 5 mg/kg in the starter heel promotes oxidative color formation.

Does Pentaerythritol Initiation Demand Tighter Moisture Control Than Glycerol-Based Polyols?

The moisture sensitivity of pentaerythritol-initiated batches is markedly higher than that of glycerol-based polyols. In glycerol-initiated systems, water contamination produces propylene glycol and some dipropylene glycol, but the triol backbone remains the dominant functionality. With pentaerythritol, moisture introduction during starter handling is amplified by the tetraol's high melting point, which requires extended hold times at 110–125 °C under vacuum; any residual water reacts with propylene oxide to form propylene glycol, a difunctional initiator that competes with the four-arm starter. A water content of 0.10 wt% in the starter feed can lower the average functionality by 0.2–0.4 and broaden the molecular weight distribution, producing foam with lower compressive strength and higher friability. The practical limit is therefore 0.05 wt% water in the polyol, with preferred values below 0.03 wt%; plants handling pentaerythritol in humid environments above 60% RH must use closed transfer, nitrogen-purged hoppers, and heated storage at 40–60 °C to prevent surface moisture pickup. The dehydration step after KOH addition is monitored not only by pressure drop but also by the water content of the vacuum condensate; termination of dehydration before the condensate rate falls below 0.1 kg/h leaves residual water that appears later as a low-molecular-weight shoulder in the gel permeation chromatography trace. Pentaerythritol dust is hygroscopic and can form a crust on baghouse filters; published data on site-specific humidity effects are limited, but the mechanistic basis for water competition is well established in polyether polyol manufacturing.

Formulation of pentaerythritol-initiated polyols is designed around the inverse relationship between hydroxyl number and molar mass; as the hydroxyl number rises from 350 mg KOH/g to 650 mg KOH/g, the equivalent weight falls from 160 g/eq to 86 g/eq, viscosity at 25 °C increases from approximately 1,500 mPa·s to 12,000 mPa·s, and the proportion of free hydroxyl groups available for urethane formation rises. The high hydroxyl number is necessary where rigid foam crosslink density, dimensional stability, and compressive strength are governed by ASTM D1621-16; a pentaerythritol-based polyol with hydroxyl number 400 mg KOH/g typically produces rigid polyurethane foam with compressive strength of 180–280 kPa at 32 kg/m³ core density, depending on blowing agent and isocyanate index. However, the high functionality also raises the liquid-phase viscosity, and formulations must be adjusted to maintain component temperatures at 25–35 °C and viscosity below 3,000 mPa·s for high-pressure impingement mixheads. In comparison, a glycerol-initiated polyol of the same hydroxyl number has a slightly lower viscosity but lower crosslink density; a sorbitol-initiated polyol has higher functionality but more severe viscosity and color issues. A systematic property comparison is shown in Table 1.

Table 1. Comparative properties of rigid polyols based on pentaerythritol, glycerol, and sorbitol starter blends
Starter compositionHydroxyl number (mg KOH/g)Viscosity at 25 °C (mPa·s)Average functionalityCompressive strength at 32 kg/m³ (kPa)
100% pentaerythritol4203,8003.8240
50/50 pentaerythritol/glycerol4002,5003.4210
30/70 pentaerythritol/glycerol3801,2003.1170
100% sorbitol49015,0004.8280

Hydroxyl number values in Table 1 are determined by ASTM D4274-21, viscosity by ASTM D4878-15, and compressive strength by ASTM D1621-16 at a foam core density of 32 kg/m³. The pentaerythritol-based formulations show lower viscosity than sorbitol at comparable hydroxyl number but require an isocyanate index increase or a higher-functionality co-polyol to match sorbitol crosslink density and dimensional stability. The magnitude of this effect depends on the blowing agent; with water as the sole blowing agent, the water-isocyanate exotherm increases foam temperature and can accelerate gel reactions, but the lower crosslink density of pentaerythritol relative to sorbitol produces friability if the index is below 110.

When a Rigid Foam Formulation Replaces Sorbitol with Pentaerythritol Polyol

When replacing sorbitol-initiated polyol with pentaerythritol-initiated polyol in a continuous pour-in-place appliance line, the formulator must consider the lower nominal functionality and improved flowability but also the shift in gel time and catalyst demand. Sorbitol-initiated polyols have nominal functionality of 6, whereas pentaerythritol initiator gives 4; at equal hydroxyl number, the sorbitol polyol yields a higher crosslink density and faster gel, but its viscosity is often 5–10 times higher, which complicates mixing at mixhead pressures below 15 MPa. A pentaerythritol-based polyol of 400 mg KOH/g can be processed through a standard high-pressure impingement mixhead at 25 °C without preheating above 35 °C, whereas a sorbitol polyol of equivalent hydroxyl number often requires component heating to 45–60 °C and may exceed the viscosity limit of 3,000 mPa·s specified by some mixhead manufacturers. However, the pentaerythritol formulation may require a 10–20% increase in polymeric MDI index or a higher-functionality co-polyol to meet the same compressive strength and dimensional stability targets, particularly under humid aging tested by ASTM D2126-20. The processing window narrows further when water is used as the sole blowing agent: the exothermic water-isocyanate reaction increases the initial foam temperature, and the lower crosslink density of pentaerythritol relative to sorbitol can produce friability if the index is below 110. In appliance foams where thermal conductivity is measured by ASTM C518-21, pentaerythritol-based polyols can achieve similar aged k-factor values if the foam cell size is maintained below 200 µm, but the lower viscosity may reduce the shear-induced cell orientation that assists flow in narrow cavity walls.

Release of pentaerythritol-initiated rigid polyol for commercial use is anchored to a batch release matrix that includes hydroxyl number by ASTM D4274-21, viscosity by ASTM D4878-15, water content by ASTM D4672-18, acid number by ASTM D4662-20, unsaturation by ASTM D4671-16, and APHA color by ASTM D1209-19. The polyol is stored under nitrogen; if exposed to atmospheric moisture at RH above 60%, it absorbs water and must be re-dried under vacuum before use. The polyol is incompatible with strong acids and with residual isocyanates in shared lines; it should not be blended with amine-initiated polyols that contain free tertiary amine unless the acid number is below 0.10 mg KOH/g, because free alkali or acid residues can neutralize the amine catalyst in the foam formulation. For foam lines using hydrocarbons as blowing agents, the polyol must have a water content below 0.05 wt% to avoid urea formation competing with the blowing reaction. Pentaerythritol starter dust is a combustible dust hazard; handling systems must comply with NFPA 652 and ATEX 2014/34/EU. Table 2 provides the batch release compliance matrix for pentaerythritol-initiated rigid polyols.

Table 2. Batch release compliance matrix for pentaerythritol-initiated rigid polyol
PropertyTest methodTypical limit
Hydroxyl numberASTM D4274-21±3 mg KOH/g from target
Viscosity at 25 °CASTM D4878-15±10% from target
Water contentASTM D4672-18≤0.05 wt%
Acid numberASTM D4662-200.03–0.10 mg KOH/g
UnsaturationASTM D4671-16≤0.05 meq/g
APHA colorASTM D1209-19≤50

Failure to meet the acid number upper limit is commonly caused by incomplete neutralization or excessive phosphoric acid addition; acid values above 0.10 mg KOH/g reduce tertiary amine catalyst activity in the subsequent foam formulation, shifting gel time and creating surface voids in continuous lamination. Failure to meet the water limit is traced to inadequate dehydration after KOH addition, humid starter storage, or nitrogen supply interruptions during vacuum stripping. Viscosity drift outside the ±10% band is treated as a process excursion and typically triggers GPC analysis of the tail fraction, because viscosity deviations in pentaerythritol-initiated polyols correlate with incomplete four-arm initiation or with unintended diol formation from residual water.

Related Articles