Gamma sterilisation turned the polycarbonate yellow: the specification fallout began when the posted irradiation dose of 25 kGy under ISO 11137-2 was not sufficient to account for the accumulated yellowness index shift measured under ASTM E313-20. General-purpose bisphenol A polycarbonate components exposed to cobalt-60 gamma radiation develop a yellow-brown tint caused by free-radical generation at the isopropylidene unit and chain scission at the carbonate linkage. Recombination produces quinoid chromophores and conjugated unsaturation that absorb in the 400–450 nm region, reducing luminous transmittance and raising the yellowness index beyond a visual tolerance that is not governed by a single universal ISO limit. Because the specification fallout is optical before it is mechanical, first-article inspection that passes tensile testing per ASTM D638-14 can still fail incoming quality control when the post-sterilisation part is viewed in a D65 light booth. The colour shift is not a surface stain; it is a through-thickness reaction profile that integrates into the transmitted colour measurement. The implication for process control is that pre-sterilisation yellowness index alone is an incomplete predictor unless melt history and residual oxidative precursors are also held constant.

On a production-scale injection moulding line with a 1200 t clamp force and a hot-runner manifold, cavity-to-cavity yellowness variation after sterilisation is frequently traced to residence-time distribution in the manifold rather than to gamma dose heterogeneity. Molten polycarbonate held above 300 °C for a longer residence interval generates hydroperoxide precursors that accelerate chromophore formation during subsequent irradiation. The result is a batch-dependent specification fallout in which the same lot of resin and the same delivered dose produce different post-sterilisation yellowness index values across cavities. Hot-runner stagnation zones and screw-recovery delay therefore become direct variables in radiation-yellowing investigations, not peripheral moulding observations. Incoming resin with a low melt-volume-flow rate under ISO 1133-1:2022 may still fail if the moulding history exceeds the stabiliser package’s thermal budget.

What Drives the Dose-Rate and Oxygen-Diffusion Coupling in Irradiated PC?

Oxygen diffusion is a controlling variable in gamma-induced yellowing because polycarbonate has low oxygen permeability relative to the free-radical generation rate during irradiation. Under cobalt-60 exposure at dose rates commonly between 1 kGy/h and 10 kGy/h, the oxidation zone is not strictly uniform across a thick section. The surface layer consumes oxygen rapidly, and diffusion-limited oxidation creates a yellowed skin with a less oxidised core. This gradient means that haze measured by ASTM D1003-21 may remain low while the yellowness index change measured by ASTM E313-20 becomes rejectable, because diffuse light scattering and chromatic absorption do not follow the same kinetic response. At higher dose rates, oxygen is consumed before it can diffuse into deeper layers; at lower dose rates, the reaction time permits oxygen ingress and a broader oxidative profile. The practical consequence is that the same 25 kGy nominal dose can produce a different colour shift depending on conveyor dwell time, source geometry, and packaging oxygen headspace. Packaging with oxygen scavengers or nitrogen purging therefore changes the yellowing outcome not by modifying polycarbonate chemistry directly but by restricting the oxygen supply to the propagating radical pool.

Dose-rate effects are not simply linear. The competition between alkyl radical recombination, hydroperoxide formation, and phenoxy radical scavenging depends on the local oxygen partial pressure and the time spent at a given dose rate. In thick parts, a higher surface oxidation rate can create a chromophore-rich zone while the core remains optically acceptable; the integrated transmitted colour then depends on both the depth of the oxidised layer and the wavelength-dependent absorption coefficient of the chromophores. Process transfer from one gamma facility to another is therefore not fully validated by dosimeter mapping alone under ISO 11137-2. A dose-mapping film confirms delivered dose, but it does not capture the oxygen-transport history that controls the colour chemistry. For this reason, specification fallout investigations require paired dosimetry, post-sterilisation ASTM E313-20 measurements, and packaging environment records.

Pre-sterilisation melt processing is not independent from radiation response. Polycarbonate dried to below 0.02% moisture at 120 °C for 4 h per standard supplier guidance retains better molecular weight and lower pre-irradiation yellowness index than resin processed with moisture above the recommended limit. Hydrolysis at processing temperature creates chain scission and phenolic chain ends, increasing melt-volume-flow rate when measured under ISO 1133-1:2022; the same low-molecular-weight fraction is more susceptible to radiation-induced chromophore formation. Production records on twin-screw compounding and injection moulding equipment show that for a given post-sterilisation yellowness index limit, a tighter melt-volume-flow rate window is more useful than a resin grade change if the root cause is thermal history variation. The processing equipment matters: a 25:1 L/D general-purpose screw with excessive recovery time can produce more pre-sterilisation oxidative precursors than a high-output barrier screw with controlled melt temperature.

Additivation for radiation stabilisation typically includes free-radical scavengers and aromatic phosphite co-stabilisers, but the specific package is proprietary in many medical grades. These additives do not prevent yellowing completely; they shift the dose threshold at which the yellowness index increase becomes unacceptable. Extractables profiling under ISO 10993-18:2020 must be repeated if the stabiliser package is changed, because the radiolysis products of phenolic antioxidants and phosphites can alter the toxicological risk assessment. A formulation change may improve colour stability while introducing new leachables that require biological evaluation under ISO 10993-1:2018. Thus, a specification fallout triggered by yellowing cannot be resolved by simple additive substitution without reopening the regulatory file.

Quantifying the Yellowness Boundary with ASTM E313-20 and Transmission Requirements

The yellowness index is calculated from CIE XYZ tristimulus values under D65 illumination and 10° observer conditions according to ASTM E313-20. It is not a direct measure of absorbance at 420 nm; it is a colorimetric transformation that is sensitive to changes in the blue-violet region where irradiated polycarbonate chromophores absorb. A component can show a substantial drop in spectral transmission between 400 nm and 450 nm while retaining an acceptable luminous transmittance if the light source weighting masks the blue-edge loss. For this reason, single-point transmission data cannot replace a full spectral scan in specification fallout investigations. Clear polycarbonate specifications may include ASTM D1003-21 for luminous transmittance and haze, ASTM E313-20 for yellowness index, and CIELAB colour metrics calculated under D65/10°. The numerical acceptance limit is user-defined and often depends on wall thickness and application; there is no ISO standard that fixes a universal post-sterilisation yellowness index value for all devices.

Colour compensation with blue toners is sometimes used to offset yellowness in transparent polycarbonate, but it cannot restore spectral transmission loss. A blue-toned formulation may pass the yellowness index calculation while still failing a device-level light transmission function. This occurs because the toner redistributes red-green and blue-yellow opponent channel values without removing the absorbing chromophore. A specification that includes only yellowness index without a minimum luminous transmittance under ASTM D1003-21 is therefore incomplete for optical devices. In a D65 light booth evaluation, two components with the same measured yellowness index can appear different if their spectral transmission curves differ in the 420–460 nm region.

When Electron-Beam Processing Replaces Cobalt-60 Gamma in Thin-Wall Polycarbonate Devices

Electron-beam sterilisation deposits the target dose in a much shorter exposure window than cobalt-60 gamma. A 10 MeV accelerator can deliver 25 kGy in seconds, whereas gamma delivery at typical facility dose rates requires hours. The shorter time reduces oxygen diffusion into the part during irradiation, which can alter the yellowing mechanism. Free radicals are still generated, but the balance between immediate recombination and oxidative propagation shifts, and thin-wall polycarbonate devices may exhibit a smaller post-irradiation yellowness index increase than equivalent gamma-treated parts. This effect is dose-rate dependent and part-thickness dependent, and published data for specific polycarbonate formulations is limited. However, electron-beam penetration is governed by density and thickness; for a given beam energy, thick sections may receive non-uniform dose, and dose mapping under ISO 11137-2 must account for depth-dose distribution. A component that passes colour specifications after electron-beam sterilisation may not pass after gamma under the same nominal dose because the oxidation pathway differs.

The transfer between sterilisation modalities is a specification fallout risk when a product is approved on gamma data and then moved to an electron-beam line for capacity reasons. Even if the minimum sterilisation dose is delivered, the yellowing response may not be equivalent. Process equivalence under ISO 11137-1 covers sterility assurance, but colour stability is not automatically captured unless the device specification explicitly includes post-sterilisation ASTM E313-20 and ASTM D1003-21 measurements. Packaging structure also changes the oxygen and moisture environment during irradiation; a foil pouch may restrict oxygen during gamma exposure but less so during electron-beam processing if the seal and headspace differ. The practitioner should treat sterilisation modality as a material-ageing variable, not merely a logistically equivalent dose-delivery method.

Repeated sterilisation cycles are not an uncommon requirement for reusable medical devices. Each 25 kGy exposure consumes residual stabiliser capacity and generates additional radiolysis products. The accumulated yellowness shift in unstabilised polycarbonate may exceed the acceptable limit after the second cycle even if the first cycle is acceptable. This is a specification fallout mode that may be absent from single-cycle validation data. The decision to move from general-purpose polycarbonate to a radiation-stabilised grade or to an alternative polymer requires retesting of ASTM D1003-21 transmittance, ASTM E313-20 yellowness index, and ISO 10993-1:2018 biological evaluation, because colour correction cannot be assumed to be biologically neutral. Moisture conditioning before sterilisation also influences post-irradiation colour. Pre-conditioning at 40 °C and 75% RH for 48 h can raise the initial moisture content and increase hydrolytic degradation during irradiation, although the effect is smaller than thermal hydrolysis at processing temperatures. Components sealed in foil packaging with desiccant show different colour evolution than those sterilised in breathable pouch stock because the water vapour and oxygen partial pressures differ.

Where a device design requires multiple sterilisation cycles, the accumulated yellowness index shift in unstabilised polycarbonate becomes a release criterion, and batch-to-batch additive variance can be larger than the colour tolerance unless the incoming resin is controlled by melt-volume-flow rate under ISO 1133-1:2022 and pre-irradiation yellowness index under ASTM E313-20.