Polybutylene terephthalate (PBT) is a semicrystalline thermoplastic polyester formed by the condensation of 1,4-butanediol with terephthalic acid or dimethyl terephthalate. The repeat unit has a molar mass of 220.22 g/mol. The thermal baseline most frequently specified for incoming resin certification is the melting endotherm peak temperature of 220–225 °C as determined by differential scanning calorimetry on second heating at 10 K/min under ISO 11357-3:2018 or ASTM D3418-21. That endotherm is not a sharp thermodynamic point; it is a distribution of lamellar thicknesses whose peak shifts with heating rate, sample mass, and prior thermal history. Alongside the melting baseline, unfilled virgin PBT at 23 °C has a density of 1.31–1.34 g/cm³ when measured by ISO 1183-1:2019 Method A or ASTM D792-20 Method B. The density spread maps predominantly to crystallinity, which by DSC enthalpy ratio using a fully crystalline reference enthalpy of 140–145 J/g falls commonly at 30–50% in molded parts.

The glass transition of unfilled PBT lies between 45 °C and 60 °C by ISO 11357-2:2020, but the glass transition does not control short-term processing latitude in the same way as the crystallization window. For unfilled PBT in a void-free two-phase approximation, bulk density is given by ρ = X_cρ_c + (1−X_c)ρ_a, where the crystalline-phase density is near 1.40 g/cm³ and the amorphous-phase density is near 1.28 g/cm³. A molded part with 40% crystallinity therefore has an expected two-phase density of approximately 1.328 g/cm³, consistent with the published baseline. Direct comparison of density-derived and DSC-derived crystallinity requires conversion between volume fraction and mass fraction; the two are not identical.

In resin certification, the melting point is commonly reported from the second heating run to erase variable processing heat history. Typical DSC specimen masses are 5–10 mg in a nitrogen purge of 20–50 mL/min. Changing the heating rate from 10 K/min to 20 K/min can shift the reported peak by several degrees Celsius; certificates of analysis should therefore state scan rate and purge conditions before a value of 223 °C from one laboratory is assumed equivalent to a value of 220 °C from another. The same constraint applies to density: specimens should be cut from the same location in a plaque or molding because gate regions, edge sections, and core regions can have different cooling histories and therefore different crystalline fractions.

What Drives the Spread Between 30% and 50% Crystallinity in Molded PBT?

Cooling rate is the dominant variable. In non-isothermal DSC experiments, the crystallization exotherm for unfilled PBT shifts to lower temperatures as cooling rate increases. At 10 K/min the crystallization peak commonly falls near 190–200 °C, while at 100 K/min it can fall below 170 °C. The suppressed crystallization temperature reduces the time available for lamellar growth before the polymer becomes diffusion-limited. In injection molding, the surface layer quenches against a cold cavity wall and develops a lower crystalline mass fraction than the more slowly cooled core. This through-thickness gradient is measurable by density: a fast-quench skin can remain at 1.30–1.31 g/cm³ while an annealed or slow-cooled core reaches 1.33–1.34 g/cm³. Bulk density measurements alone can hide that gradient, but it reappears as warpage, post-molding shrinkage, or uneven chemical resistance.

Mold temperature sets the practical floor of the crystallization window. With mold temperatures below 60 °C, the PBT surface freezes quickly and retains low crystallinity. With mold temperatures between 80 °C and 120 °C, the material remains above the glass transition long enough for crystallite perfection and higher bulk density. A 3 mm plaque molded at 40 °C versus 110 °C can show a density difference of 0.02–0.03 g/cm³ after conditioning; the offset is not universal, but the direction is consistent. Nucleating additives further widen the range. Talc or pigmentary nucleants can raise the crystallization onset to 200–205 °C even at 20 K/min, producing finer spherulites and a higher bulk density for the same nominal cooling cycle. A crystallinity specification therefore cannot be transferred from a clean unfilled grade to a nucleated or pigmented compound without reacquiring the cooling curve.

Melt viscosity must also be interpreted with the melting baseline in view. Unfilled injection-molding grades commonly exhibit a melt mass-flow rate of 10–30 g/10 min at 250 °C under 2.16 kg load by ISO 1133-1:2022. That test temperature sits well above the melting endotherm, but it does not describe behavior near the lower boundary. If the melt cools below 225 °C, recrystallization begins quickly, and the apparent viscosity increases as crystalline domains nucleate. Capillary and rotational rheology data must be coupled with crystallization kinetics when processing in the lower window.

Density values are not intrinsic constants for a PBT grade; they must be paired with filler content and conditioning history. Conditioning at 23 °C and 50% RH under ISO 291:2008 does not alter density significantly because unfilled PBT absorbs less than 0.1 wt% moisture at equilibrium. Glass-fibre reinforced PBT at 30 wt% fibre typically exhibits a compound density of 1.53–1.61 g/cm³, not 1.31–1.34 g/cm³. In such systems, crystallinity cannot be obtained by dividing the measured enthalpy of fusion by the 140–145 J/g fully crystalline reference without correcting for glass and sizing mass. The corrected PBT matrix can still fall in the 30–45% crystallinity range, but the enthalpy per gram of compound is lower. A halogenated flame-retardant PBT can shift density to 1.45–1.55 g/cm³ depending on the flame-retardant system and mineral synergist; halogen-free flame-retardant grades can exceed that range because of phosphate, boron, or mineral loadings.

The density and crystallinity baselines also interact with electrical and thermal specifications. A PBT compound used in an automotive connector may be qualified by comparative tracking index under IEC 60112 and glow-wire end-product testing under IEC 60695-2-11. These tests are not direct functions of the melting peak alone, but they are influenced by crystallinity through surface morphology, filler dispersion, and residual stress. Published data for exact crystallinity thresholds in every end-product configuration is limited; material qualification therefore uses the DSC and density baselines as incoming-control boundaries rather than as direct service-life predictions.

Alpha-to-Beta Transformation Under Tensile Load and Its Effect on the Two-Phase Density Model

PBT crystallizes in more than one crystal modification. The α-form predominates in as-molded material and contains the tetramethylene sequence in a contracted conformation; the β-form is obtained under tensile stress or annealing under strain and contains an extended methylene sequence. Wide-angle X-ray diffraction can separate the two forms by their unit-cell reflections. DSC alone cannot distinguish them cleanly because their melting endotherms overlap within the 220–225 °C baseline. The α-to-β transformation changes the crystalline unit-cell dimensions and therefore the apparent crystalline density, even though the mass fraction of crystallinity may remain unchanged. Published values for the exact density difference between α and β crystals in PBT vary among sources; the reversible transformation upon unloading is well documented and produces recoverable dimensional effects.

On the production floor, this transformation appears as a temporary density or modulus change in highly oriented parts. A connector, film hinge, or thin-wall component that has been cold-drawn or tensile-loaded can exhibit a slightly different density signature before and after annealing above 180 °C, where molecular mobility permits relaxation back to the α-dominated state. Density measured on as-molded parts with high frozen orientation should therefore not be used as a direct index of crystallinity unless the contributions of orientation and crystal form are subtracted or independently characterized.

Before any processing baseline is recorded, the PBT must be dried. Residual moisture above 0.02 wt% by Karl Fischer titration initiates hydrolytic chain scission at melt temperatures of 250–270 °C; the reduced melt viscosity then depresses the apparent melting peak and degrades crystallinity-related performance. Desiccant-dryer settings of 120 °C for 4–6 h with a dew point below -30 °C are typical for sealed resin lots, while exposed regrind or high-humidity storage may require longer residence. On an injection molding machine with a 20:1–24:1 L/D three-zone screw and compression ratio of 2.5:1–3.0:1, barrel settings from 230 °C at the feed throat to 250–260 °C at the nozzle keep the melt above the melting plateau while limiting excessive thermal history. Melt residence should remain below 10 min where possible; prolonged hold time accelerates hydrolysis, discoloration, and loss of molecular weight.

When PBT Is Processed Below 225 °C and the Melt Bank Solidifies

Running the melt below 225 °C removes the safety margin that the melting range supplies. Because the melting baseline spans 220–225 °C, zones set below 225 °C risk carrying unmelted or partially melted granules into the shot. In practice, front-zone setpoints below 230 °C are uncommon for unfilled PBT; they reduce melt fluidity and can produce screw-recovery torque spikes, particularly on machines with shallow compression ratios. If the nozzle or hot runner drops below 225 °C, the material can recrystallize in dead spots and create cold slugs or gate blockage. Accepted processing guidance places the melt-temperature window at 240–270 °C so that the melting endotherm is fully traversed and the crystallization exotherm is controlled by the mold rather than by the barrel. The density of the final part then becomes a function of cooling flux, mold temperature, packing pressure, and crystal form—not merely a readout of the resin datasheet.