Evonik INFINAM 9359 F is an unfilled polyether ether ketone filament produced for fused filament fabrication and fused deposition modeling systems. The grade is supplied in 1.75 mm and 2.85 mm diameters with a spool-level diameter tolerance of ±0.05 mm; supplier release certificates also specify a maximum moisture content of 0.02 % before vacuum sealing. The polymer is not a lower-melting PAEK copolymer. Differential scanning calorimetry on dried feedstock shows a glass transition at approximately 143 °C and a melting endotherm at approximately 343 °C, so the thermal processing envelope is set by PEEK homopolymer crystallization rather than by a depressed melting point. Because the grade contains no carbon fiber or glass fiber reinforcement, the melt is not abrasive and the printed dielectric response does not show the same anisotropic fiber-orientation effects observed in filled PEEK filament. The practical trade-off is that unfilled PEEK has a lower tensile modulus and higher creep under sustained load than filled grades.

Filament ovality and melt-flow consistency are observable batch-to-batch variables. FFF lines running 1.75 mm feedstock with a 0.4 mm nozzle at 0.15 mm layer height can develop extruder feed slip if ovality exceeds 0.03 mm, because cold-end idler pressure fluctuates. Lot release data for INFINAM 9359 F is therefore tied to diameter variance rather than visual appearance. In a production environment, a filament run-out gauge with 0.01 mm resolution mounted before the extruder inlet can detect spool-joint overruns. These checks are part of the material boundary condition because high-temperature PEEK extrusion cannot tolerate the dimensional variation permissible in lower-temperature filaments; the melt has a steep viscosity curve near the nozzle setpoint, and small dimensional variations translate directly into flow-rate error. Interlayer adhesion is governed by reptation time and surface temperature. When the second layer is deposited, the interface must remain above the melt temperature long enough for chain interdiffusion; the measured self-diffusion coefficient of PEEK at 400 °C is low compared with amorphous polymers. Chamber temperatures above 100 °C slow the quench and extend chain interdiffusion time. This is why published XZ tensile strength for unannealed unfilled PEEK FFF parts is often reported in the range of 20–40 % of XY values; after annealing, the XZ strength may improve because the crystalline lamellae reorganize, but the interface remains a structural discontinuity. INFINAM 9359 F is supplied with a controlled crystallization rate, which reduces the abrupt viscosity rise during cooling, but it does not eliminate the requirement for a heated chamber.

What Limits the Fused Filament Fabrication Processing Window for Unfilled PEEK?

Critical process limits arise from the narrow gap between the crystallization temperature and the melt temperature. For INFINAM 9359 F, nozzle setpoints of 400–430 °C are specified by the supplier. At the lower bound, melt viscosity remains high enough to require extruder torque margins above 2.0 N·m on direct-drive toolheads using a 0.6 mm nozzle and 0.2 mm layer height. At the upper bound, residence time at 430 °C must not exceed 30 min, because chain branching and gel formation can increase melt pressure and reduce weld-line integrity. The build plate setpoint should be held at 150–180 °C, while the chamber setpoint must remain between 100 °C and 150 °C. Below 100 °C, the free surface of a deposited bead cools below the crystallization onset near 288 °C too quickly; the resulting transcrystalline layer prevents molecular interdiffusion with the subsequent layer. Production-scale heated-chamber FFF lines routinely measure a Z-axis tensile strength loss of 30–50 % relative to XY when chamber control is inadequate. Above 150 °C, standard polymer machine components, including belt tensioners, guide bushings, and cooling fans, operate outside their service envelope, and unsupported overhangs sag because the printed bead remains too compliant. The hot end itself must be all-metal and rated for at least 450 °C continuous service; polytetrafluoroethylene-lined hot ends cannot be used. Nozzle material selection matters less for abrasion than for thermal stability; brass nozzles may soften at the upper processing limit over extended campaigns, so hardened steel or tungsten carbide is specified for sustained runs.

Supplier-Published Thermal and Mechanical Data for FFF Coupons

Mechanical values printed in Table 1 are not isotropic. FFF parts show orthotropic behavior: XY coupons follow the tensile axis, while ZX and XZ coupons test interlayer adhesion. The values below are supplier-published or PEEK-homopolymer typical for XY printed coupons after annealing at 200 °C for 2 h; batch-specific certificates prevail. Direct comparison to injection-molded PEEK requires identical moisture conditioning and gauge geometry, and published data for some interlayer-specific configurations is limited.

In high-temperature electrical connector bodies, unfilled PEEK is selected because the comparative tracking index and volume resistivity remain stable after moisture conditioning. Printed INFINAM 9359 F prototypes used for connector insulators should be annealed before electrical testing, because the as-printed crystalline phase may be lower and microvoids along raster boundaries can reduce dielectric strength below the annealed value. A typical acceptance test is dielectric strength according to IEC 60243-1; unfilled PEEK homopolymer can achieve values above 20 kV/mm on solid injection-molded specimens, but FFF parts may require sealing or compression during printing to reach equivalent values. In oil and gas seal stack prototypes, the material is resistant to hot aqueous brines and moderate sulfide environments, but the sealing faces must be machined after annealing to remove raster-induced roughness. The annealing step at 200 °C for 2 h raises crystallinity from as-built levels near 20–25 % to 30–35 %, which improves dimensional stability at the next thermal excursion. Continuous service above 150 °C must be assessed with creep data; unfilled PEEK tensile creep modulus at 23 °C may exceed 2.5 GPa at 1,000 h, but at 150 °C it falls below 1 GPa. Therefore, load-bearing designs at elevated temperature should either use fiber-filled PEEK or apply metal backup structures.

When Annealing Becomes the Dominant Process Variable

Annealing of printed PEEK is not a low-risk operation. The difference in coefficient of linear thermal expansion between the less-crystallized core and the already-crystallized skin means uncontrolled heating can generate internal voids. If the oven ramp rate exceeds 2 °C/min, differential expansion can cause interlaminar cracks at the raster boundaries. Production-scale trials on chamber-equipped FFF lines show that free-standing parts annealed at 200 °C for 2 h can warp by more than 1 % in the longest dimension; the same geometry supported in a packed alumina bed or clamped in an aluminum fixture warps below 0.2 %. For sections thicker than 20 mm, dwell time must be extended beyond 4 h to allow the core to reach crystallization temperature. The cooling step after annealing should be controlled at 0.5 °C/min until the part temperature falls below 140 °C, because rapid cooling reintroduces residual stress. Parts annealed in a circulating air oven should not be stacked to a mass exceeding 5 kg per shelf without validation, because internal shelf temperature gradients can exceed ±5 °C and create nonuniform crystallinity. The annealed part also shifts slightly in dimensions; post-anneal machining allowances of 0.2–0.5 % on critical features are typical when no fixed tooling is used.

Chemical Incompatibility Limits and Sterilization-Dependent Performance

Unfilled PEEK resists hot water, steam, aliphatic hydrocarbons, alcohols, and common salt solutions. The operational boundary excludes concentrated sulfuric acid and concentrated nitric acid, particularly above 50 °C; halogen gases under pressure and strong oxidizing media can also induce stress cracking. For medical or pharmaceutical use, steam autoclave cycles at 134 °C are tolerated by the base polymer, but FFF-built parts with interconnected porosity above 0.5 % may entrap moisture and fail during rapid depressurization. Gamma sterilization at 25 kGy can reduce elongation at break of unfilled PEEK; published data for INFINAM 9359 F FFF coupons under optimized chamber conditions is limited, so the sterilization method must be validated on the final printed article. Biocompatibility is not inheritable from the filament certificate alone; the finished device must be evaluated under ISO 10993-1:2018, and if radiation is used, dose-mapping must follow ISO 11137-2.

Electrical testing on FFF-built unfilled PEEK should account for void fraction. The comparative tracking index and dielectric strength measured on solid injection-molded PEEK are not transferred automatically to a part printed with 35 % infill. Under IEC 60243-1, a solid PEEK plaque can exceed 20 kV/mm; a printed part with visible raster voids may fail at 5–10 kV/mm because partial discharges occur along the interlayer boundary. Therefore, electrical insulators require dense spiral infill or post-print compression molding, and acceptance testing must be performed on the exact print parameter set. Thermal expansion also follows the printed axis: the coefficient of linear thermal expansion of unfilled PEEK is approximately 45–55×10⁻⁶/K below the glass transition and approximately 120–140×10⁻⁶/K above it, but FFF parts can show axis-dependent expansion because of residual orientation and void distribution.

Relative to carbon-fiber-filled PEEK filament, INFINAM 9359 F trades tensile modulus for cleanliness in the hot end. A filled grade can raise tensile modulus toward 7 GPa and reduce thermal expansion, but the broken carbon fibers accelerate nozzle wear, alter the melt flow index, and produce anisotropic dielectric behavior. INFINAM 9359 F retains the lower unfilled PEEK modulus near 3.6 GPa and a dielectric strength above 20 kV/mm on void-free test plaques. Against lower-melting PAEK alternatives, the product’s retention of the 343 °C melting endotherm gives higher thermal stability after annealing but requires a chamber capable of 100–150 °C; a printer designed only for 80 °C chamber operation is not suitable. Compared with general-purpose unfilled PEEK filaments, the grade is differentiated by its tighter diameter and ovality control, controlled crystallization behavior, and the supplier’s specification of moisture limits for extrusion rather than simple storage. Published data for direct comparative interlayer strength between INFINAM 9359 F and every competing unfilled PEEK filament under identical chamber temperatures is limited; material substitution should be validated on the target FFF line.

Moisture control is a critical loading condition. Unfilled PEEK absorbs less than 0.1 % moisture at 23 °C/50 % RH when tested by ISO 62, but water present at 0.05 % in filament can hydrolyze the polymer chain at 400 °C. An open spool left for 8 h at 60 % RH may exceed the 0.02 % extrusion limit. Drying in a circulating air oven at 150 °C for 3 h is the specified boundary; desiccant drying at 80 °C for 4 h alone is not sufficient for deep-moisture spools. On direct-drive FFF systems equipped with melt-pressure sensors, wet filament produces pressure deviations beyond ±0.5 MPa and a rapid pressure drop associated with steam evolution at the nozzle. This terminates the usable processing window until the spool is redried.