Polybutylene Terephthalate Moisture Limitations and Comparative Selection of PPS and PA

Polybutylene terephthalate (PBT) is a semicrystalline polyester with a melting temperature of approximately 223°C, a glass transition temperature of 45°C to 60°C, and a dry-state tensile strength of 50 MPa to 60 MPa for unfilled injection molding grades. The polymer absorbs only 0.1% water after 24 h at 23°C when tested in accordance with ASTM D570. This low moisture uptake is frequently treated as evidence of hydrolytic stability; however, the ester linkage in the polymer backbone undergoes nucleophilic attack by water, and chain scission accelerates when the part is exposed to hot water, steam, or water-glycol mixtures above approximately 60°C. The following sections examine the conditions in which polybutylene terephthalate cannot take the moisture and compare polyphenylene sulfide (PPS) and polyamide (PA) as replacement materials under the same application constraints.

Hydrolytic Chain Scission Initiated by Carboxylic Acid End Groups

Hydrolysis in PBT proceeds through a reversible ester-cleavage reaction that produces a carboxylic acid end group and a hydroxyl end group. The liberated acid terminal species is itself a catalyst for further chain scission, so the process becomes autocatalytic once the concentration of carboxylic acid end groups rises. This autocatalytic behavior is monitored by end-group titration and by melt volume-flow rate testing in accordance with ISO 1133-1:2022. A rise in melt volume-flow rate from 10 cm³/10 min to 25 cm³/10 min or higher after humid aging indicates a substantial reduction in number-average molecular weight. During injection molding, moisture in the granules must therefore be kept below 0.02% before the resin enters the barrel. Industry-standard drying uses a closed-loop desiccant dryer with a dew point of −40°C and a drying temperature of 120°C for 4 h. If moist granules are plasticated on a 25 mm diameter screw with an L/D ratio of 20 at a melt temperature of 250°C, hydrolysis lowers melt viscosity, produces surface splay, and creates brittle weld lines. The upper thermal limit for melt processing is typically 270°C; above this temperature thermal degradation causes yellowing and carbon dioxide formation. On production-scale injection molding machines, the melt temperature for a 30% glass-fiber reinforced PBT is therefore controlled within ±5°C of a 250°C setpoint, while the mold is held at 60°C to 80°C. Some PBT grades are stabilized by end-capping agents or carbodiimide-based hydrolysis inhibitors, but these additives retard rather than eliminate hydrolysis above 80°C. In water-glycol mixtures, metal ions or acidic buffer components can further accelerate ester cleavage, so the hydrolytic stabilizer package must be validated in the actual coolant formulation rather than in deionized water.

When Does Moisture Stop Being a Process Concern and Become a Service Threat?

The transition from process concern to service threat occurs when the polymer is simultaneously exposed to diffusing water and a temperature above approximately 60°C. At room temperature, service moisture absorption is slow and the equilibrium concentration remains low. In hot water or condensing humidity, the diffusion rate increases and the ester cleavage rate accelerates. Qualification testing then requires long-term immersion rather than short-term cold-water absorption. A typical test sequence uses ISO 62:2008 for water absorption and ISO 527-2:2012 or ASTM D638-14 for tensile strength retention after immersion. In hot water at 80°C to 90°C, glass-filled PBT can lose more than half of its initial tensile strength over several hundred hours, depending on initial carboxyl end-group concentration, crystallinity, and glass-fiber content. Published data for specific water-glycol mixtures is limited; direct substitution of PPS or PA into a PBT part therefore requires immersion testing in the actual service fluid, not extrapolation from 24 h cold-water absorption data.

On an automotive connector production line, 30% glass-fiber reinforced PBT connectors molded on an 800 kN clamp force injection molding machine with a melt temperature of 250°C and a mold temperature of 80°C failed after 1,000 h at 85°C and 85% RH under cyclic thermal load. The observed failure mode was stress cracking of weld lines, and the recovered polymer exhibited a melt volume-flow rate more than twice that of the dry-as-molded granules. When the same housing geometry was converted to a 40% glass-filled PPS, the process required a melt temperature of 320°C and a mold temperature of 140°C, but no hydrolysis-induced embrittlement occurred in the same damp-heat exposure. A PA66 30% glass-filled variant absorbed moisture at 23°C and 50% RH, reducing the dry-state tensile strength from approximately 180 MPa to 120 MPa while increasing elongation from 2.5% to 6%. The PA66 change was acceptable where the failure mode was brittle fracture but unacceptable where terminal retention required dry-state stiffness.

Comparative Moisture Uptake of PBT, PPS, and PA66

Moisture absorption values are grade-dependent, but representative injection molding compounds show clear separation among the three polymer systems. The table below summarizes typical values for glass-fiber reinforced grades; the specimens are 60 mm × 60 mm × 2 mm plaques tested at 23°C.

The values confirm that PBT and PPS both remain dimensionally stable in humid air, but PBT retains an ester hydrolysis weakness that the low equilibrium moisture value does not eliminate. PA66 absorbs substantially more water and undergoes plasticization, which must be treated as a design parameter rather than as a hidden degradation mode.

When Polybutylene Terephthalate Cannot Take the Moisture: PPS and PA Compared

When a part is specified for continuous contact with hot water or steam, PBT is frequently eliminated during initial material screening. The replacement decision then becomes a comparison between PPS and PA. PPS enters the comparison because its hydrolytic stability is substantially higher than that of PBT. The polymer has a melt temperature near 280°C and can be compounded with 40% glass fiber for continuous service temperatures up to 200°C. Its water absorption after 24 h at 23°C is typically 0.02% to 0.05%, and it retains mechanical properties after prolonged immersion at 80°C to 120°C. The penalty is processing cost and brittleness. PPS melt must be held between 315°C and 330°C, and the mold must be maintained at 90°C to 150°C to obtain adequate crystallization. PA66 is the second comparator. Polyamide is not hydrolytically inert in the same sense as PPS, but it tolerates moisture because water is absorbed in the amide-rich amorphous phase and acts as a plasticizer rather than triggering the autocatalytic ester cleavage observed in PBT. The dry-state tensile strength of a 30% glass-filled PA66 compound is approximately 180 MPa; after conditioning at 23°C and 50% RH, tensile strength falls to approximately 120 MPa, while elongation at break increases from 2.5% to 6%. For dimensionally critical parts, this moisture response is severe. Linear dimension change from dry to equilibrium at 50% RH can exceed 0.5% for PA66, whereas PPS and PBT remain below 0.1% under the same humidity exposure. The selection therefore depends on whether the application can tolerate a ductile, moisture-plasticized polyamide with larger dimensional movement or requires the dimensionally stable but more brittle and higher-temperature-processing PPS.

Melt Processing Limits Are Narrower for PPS Than for PBT or PA66

Processors note that PBT has a relatively broad melt-processing window. Glass-filled PBT flows at melt temperatures from 245°C to 260°C, and mold temperatures from 60°C to 80°C produce a crystallized surface with acceptable dimensional stability. PPS requires a much tighter thermal profile. Below 310°C, PPS melt is too viscous for complete glass-fiber wet-out; above 340°C, the resin can degrade. Mold temperature is not optional for PPS: below 90°C, the part solidifies with a highly amorphous, low-strength surface layer. PA66 melts between 270°C and 290°C, and the mold is held at 70°C to 90°C. These differences affect tooling and cycle time. A PPS conversion often requires mold heaters capable of 150°C, additional insulation, and a longer cooling time than PBT. PA66 conversion can sometimes use existing PBT tooling if the part design absorbs the dimensional change caused by moisture conditioning, but the mold must incorporate adequate venting because PA66 releases water vapor during plastication.

In a damp-heat test performed on an industrial sensor housing at 40°C and 95% RH, PBT performed adequately when the internal heat source kept the housing below 60°C. When the same housing was placed near an engine heat source and the local wall temperature reached 85°C, the PBT component retained dimensional stability but developed surface microcracks and reduced stress retention. The PA66 alternative absorbed moisture and lost modulus under the same 95% RH exposure, causing terminal loosening after 500 h. The PPS alternative remained dimensionally and mechanically stable but required a mold temperature of 130°C and a cycle time increase of 10–15% relative to PBT. Electrical testing after damp heat was performed in accordance with IEC 60068-2-30 using a severity condition of 65°C and 93% RH for 96 h; only PPS maintained the specified insulation resistance above the lower control limit without reconditioning.

A water pump impeller molded from 30% glass-filled PBT and operated in a closed hot-water loop at 90°C failed after repeated starts because hydrolytic surface cracks initiated at the gate region and propagated along glass-fiber orientations. A replacement was evaluated by immersion of 10 mm thick plaques in 90°C deionized water for 500 h. PBT lost more than 50% of its initial tensile strength under the same immersion, while a 40% glass-filled PPS retained over 90%. Published data for this specific impeller geometry is limited, but the material difference was consistent with ester hydrolysis. When the impeller was converted to PPS, the injection molding process required a 140°C mold temperature and a 320°C melt temperature, and the finished part had lower elongation at break. When PA66 was tested, the impeller absorbed enough water to soften the hub and reduce interference fit retention; the part remained ductile but the mating assembly failed after 300 h thermal cycling because the clearance increased beyond the specified 0.05 mm.