Hydrolytic Failure in Polybutylene Terephthalate: The Six-Month Blind Spot

Hydrolytic failure in polybutylene terephthalate (PBT) is best characterised as an autocatalytic chain-scission process that remains hidden when tensile strength alone is used as the release criterion. The six-month blind spot corresponds to approximately 4,320 h of hot-wet service, during which ester bonds in the 1,4-butanediol–terephthalate backbone undergo nucleophilic attack by water. Moisture absorption of unreinforced PBT after water immersion at 23°C is normally below 0.5% according to ISO 62, but saturated surface layers in humid service can create a local hydrolysis front even when the bulk moisture content appears low. The result is not uniform erosion but brittle failure at weld lines, metal-insert boundaries, and tensile-stress concentrations. Because tensile strength measured per ISO 527-2/1A/50 is relatively insensitive to molar mass loss, components may pass 500 h qualification tests and still fracture after six months. The risk pattern is therefore defined as Hydrolytic Failure in Polybutylene Terephthalate: The Six-Month Blind Spot.

Why Does Hydrolytic Damage Remain Undetected in the First Six Months?

The detection delay is a consequence of the relationship between molecular weight, crystalline phase continuity, and the mechanical test response. Hydrolysis reduces the number of tie molecules and the amorphous-chain length long before the spherulitic superstructure loses its ability to carry stress. A tensile bar aged at 85°C/85% RH for 1,000 h can retain 80% or more of its initial tensile strength while falling below 50% of its initial nominal strain at break; notched impact strength per ISO 180/A often shows a steeper decline because crack propagation resistance depends on plastic deformation in the amorphous phase. Melt volume-flow rate measured at 250°C and 2.16 kg per ISO 1133-1:2022 is a more direct indicator, since chain scission reduces melt viscosity and raises MVR even when tensile strength remains within specification. The lag phase is not a safety margin; it is the time needed for carboxyl end-group concentration to reach a level at which autocatalysis becomes the dominant kinetic term. Reported activation energies for polyester hydrolysis in hot-water ageing fall in the range 70–90 kJ/mol, which is high enough to make room-temperature degradation slow but low enough to produce a clear nonlinear acceleration at underhood temperatures. For quality control, solution viscosity measurement in a 60/40 wt% phenol/1,1,2,2-tetrachloroethane solvent at 25°C according to DIN EN ISO 1628-5 remains the most direct routine indicator of molar mass loss, although it is slower than MVR and requires laboratory sampling.

In melt processing, the six-month blind spot is often created before the component reaches service. Virgin PBT pellets should be dried to a moisture content below 0.02% before compounding or injection moulding. A desiccant dryer with a dew point of -40°C to -30°C and a drying temperature of 120°C for 3–4 h is the standard industrial boundary; at melt temperatures of 245–270°C, residual moisture above 0.04% will cleave ester bonds in the barrel and raise the pellet MVR before the part is moulded. The damage is not always detected because the moulded part may still meet dry-as-moulded tensile requirements. On a 40:1 L/D twin-screw extruder compounding 30% glass-fibre-reinforced PBT with a side feed at zone 6, a rise in vacuum vent pressure above 50 mbar absolute signals insufficient devolatilization and correlates with retained moisture, oligomer generation, and melt viscosity loss. Injection moulding of 30% glass-fibre-reinforced PBT should use a reverse barrel temperature profile of 240–270°C, mould temperature of 60–100°C, and total melt residence time below 5 min. Screw back pressure is normally limited to 20–50 bar to avoid excessive shear heating; at barrel temperatures above 270°C, thermal degradation can mimic the hydrolytic chain-scission signal because ester elimination releases carboxylic acid groups. Dried pellets exposed to ambient air at 60% RH or higher must be rechecked by Karl Fischer titration; open-hopper residence beyond 1 h under such conditions is a common root cause of batch-to-batch MVR drift.

When 85°C/85% RH Aging Is Used as a Six-Month Accelerated Exposure

When 85°C/85% RH damp-heat ageing is used to simulate humid service, a six-month field life cannot be reduced to a simple 1,000 h pass/fail test without understanding the shape of the degradation curve. Damp-heat testing per IEC 60068-2-78 imposes a constant humidity and temperature that may not reproduce cyclic condensation, thermal expansion, or insert-interface moisture wicking. Moisture uptake measured by ASTM D570 or ISO 62 is often quoted at equilibrium, but thick PBT sections reach that equilibrium slowly, and the outer skin may be hydrolysable while the core remains dry. The hydrolytic reaction consumes water and generates acid groups, so the rate is not linear with exposure time; extrapolation from 500 h or 1,000 h data can overpredict the time to embrittlement by assuming linear kinetics. Automotive connector qualification under USCAR-2 Rev 7 includes environmental cycling and damp-heat exposure, but the pass criteria are often electrical contact resistance, insulation displacement, and terminal retention. Hydrolytic embrittlement can progress without an immediate electrical failure until a connector is handled or replugged. Published data for unreinforced PBT after 4,320 h at 85°C/85% RH are limited, but comparative studies using notched Izod per ISO 180/A show clear loss of impact toughness before tensile strength falls below specification. A qualification matrix should therefore include mechanical tests after 1,000 h, 2,000 h, and 4,320 h, with MVR and notched impact as early indicators.

Temperature, pH, and Carboxylic Acid End-Group Autocatalysis

Temperature, pH, and carboxyl end-group concentration control the autocatalytic loop that produces late-stage embrittlement. Each ester cleavage event produces one carboxyl end group and one hydroxyl end group; the carboxyl group reduces the local pH at the reaction front and accelerates further chain scission. This means that partially hydrolysed PBT can degrade faster than unhydrolysed material even when the external water concentration remains unchanged. The reaction is slow below the glass transition because segmental mobility limits water penetration, but under hot-wet service the rate becomes nonlinear and sensitive to acid scavenger depletion. Polymeric carbodiimides react with carboxyl end groups to form urea linkages and interrupt the autocatalytic cycle; typical active loadings in PBT compounds fall between 0.5 wt% and 1.5 wt%. The stabilizer is not a substitute for drying below 0.02% moisture because it cannot reverse hydrolysis that occurs in the barrel, and residual carbodiimide can alter melt viscosity. Additive systems containing free amines or strongly alkaline residues should be avoided because they can saponify terephthalate esters and increase water-soluble oligomer formation. Residual titanium or antimony catalyst residues from polycondensation can also influence hydrolysis rate by providing Lewis-acid sites, making end-group control during polymer synthesis an additional variable. Published data for stabilizer-treated PBT after 4,320 h at 85°C/85% RH are limited; therefore, acid-scavenger efficiency should be verified by MVR retention and tensile strain at break rather than tensile strength alone.

Impact-modified PBT grades can obscure the six-month blind spot further. The elastomer phase may maintain notched Izod readings even as the polyester matrix loses molar mass, because impact energy is absorbed by the dispersed phase. This creates a false negative when using only ISO 180/A; MVR and tensile strain at break should be recorded in parallel. Published data for impact-modified PBT after 4,320 h of 85°C/85% RH ageing are limited, and the stabilizer demand may differ from unreinforced compounds because the dispersed phase changes moisture transport and stress distribution.

Connector bodies moulded from 30% glass-fibre-reinforced PBT with nominal wall thicknesses of 1.5–3.0 mm show the blind spot most clearly at knit lines and metal-insert interfaces. In underhood service, intermittent electrical failures have been traced to housing cracks that open under terminal retention loads after roughly 4,000–4,500 h, while the bulk material still passes dimensional inspection. A production-scale failure analysis should compare as-moulded parts with parts aged for 4,320 h under 85°C/85% RH per IEC 60068-2-78, using notched Izod specimens per ISO 180/A and terminal pull-out force at the specified insertion depth. Tensile strength retention is insufficient because the crack originates in a constrained, high-stress region where the amorphous chain network has degraded more than the outer wall. Published data for terminal retention after 4,320 h of damp-heat ageing with field-ageing correlation are limited; therefore, draw force validation must be performed on the specific housing geometry and insert design rather than inferred from generic PBT tensile data.

Processing Stabilizer Loading Windows and the Cost of Overstabilization

Processing stabilizer loading windows and the cost of overstabilization require gravimetric control and rheological verification. Polymeric carbodiimide masterbatch is added at 0.5–2.0 wt% active carbodiimide in PBT compounds; below 0.5 wt% the acid-scavenging capacity may be exhausted before 4,320 h, while above 2.0 wt% the unreacted carbodiimide may increase melt viscosity and make thin-wall filling pressure-sensitive. The melt volume-flow rate measured per ISO 1133-1:2022 at 250°C/2.16 kg should be recorded for each lot and compared with the unstabilized control; an MVR reduction greater than 30% at 2.0 wt% indicates chain extension beyond the intended molecular weight range and may require a die temperature increase or reduced packing pressure. In a 40:1 L/D twin-screw extruder, the masterbatch should be introduced after atmospheric venting but before vacuum devolatilization to avoid hydrolysis of the additive and to permit dispersive mixing under high shear. Published data for overstabilized PBT after 4,320 h at 85°C/85% RH are limited; the absence of long-term data is itself a reason to maintain 0.5 wt% as the lower validation boundary and to avoid increasing carbodiimide loading solely on the basis of short-term tensile retention.