Optimization and manufacture of polyetheretherketone patient specific cranial implants by material extrusion: A clinical perspective

James A. Smith, Sandra Petersmann, Florian Arbeiter, Ute Schäfer

Research output: Contribution to journalReview articlepeer-review


Polyetheretherketone (PEEK) is a high performing thermoplastic that has established itself as a ‘gold-standard’ material for cranial reconstruction. Traditionally, milled PEEK patient specific cranial implants (PSCIs) exhibit uniform levels of smoothness (excusing suture/drainage holes) to the touch (
Original languageEnglish
Article number105965
Number of pages25
JournalJournal of the Mechanical Behavior of Biomedical Materials
Issue numberAugust
Early online date15 Jun 2023
Publication statusPublished - Aug 2023

Bibliographical note

Funding Information:
This work was supported by the project CAMed ( COMET K-Project 871132 ) and was funded by the Austrian Federal Ministry of Transport, Innovation and Technology (BMVIT) and the Austrian Federal Ministry for Digital and Economic Affairs (BMDW) and the Styrian Business Promotion Agency (SFG ).
Funding Information:
The dual coloration seen across the aforementioned implants is credited to the existence of both desirable semi-crystalline and undesirable amorphous polymer phases, with the latter identified through its non-uniform semi-transparent aesthetic varying from light caramel-to dark brown (Fig. 1d). The formation of these amorphous zones can be credited to the intrinsic nature of the FFF process which is more dynamically complex than the highly controlled methodologies used to produce semi-crystalline millable PEEK. As consequence, during printing, each sequential layer, as well as its length, experiences environmental temperature gradients (despite efforts to maintain a constant through heated build chambers and build platforms), which in turn, impacts the polymers internal temperature and solidification rate i.e. the bottom layer of the print in contact with a heated build platform will solidify at a different rate to layers in the middle or the top surface of the implant. The formation of amorphous PEEK typically occurs when there is insufficient energy in the system i.e. the molten polymers cools down too quickly. In turn, at the macromolecular level, the nucleation sites required to promote parallel chain alignment and densification fail to occur, generating less favorable aesthetics and performance vs. high semi-crystalline variants. This explains why the majority of amorphous PEEK tends to be mainly localized to the exterior surface or edges of implants, as these are typically further away from both the bulk of the material (which dissipates heat slower upon deposition) or other additional heat sources i.e. heated build platform. It was noted that implants printed with a surface area of 100 cm2 or greater suffered from higher levels of amorphous content opposed to smaller examples. This can be credited to the higher surface-area-to-volume of larger PSCI's which dissipate heat faster upon deposition leading to faster solidification rates. Furthermore, this phenomenon can help support the reasoning as to why conflicting reports on build orientation vs. amorphous PEEK content exist. Whereby, implants with greater surface-area-to-volume ratios benefit from being printed horizontally, as the majority of their volume is closer to a constant heat source, which enables heat to be retained in the system. For smaller specimens in a vertical build orientation, the heat generated by the heated build platform was both sufficient to permeate through the implant and maintain an ambient chamber temperature that promotes high levels of semi-crystallinity. However, in the horizontal build orientation significantly higher levels of dense support structure were required to hold the implant in-situ. As consequence, these support structures not only decreased the circulation of heat around the chamber but also took additional time to permeate into the implant itself, hence why amorphous PEEK was located on the topside and edges of the PSCIs. Studies performed by Liu et al. (2022) and Sharma et al. (2023) (Fig. 1e) appear to have overcome such dual phase appearance. However, only a single shape PSCI was printed (Liu et al., 2022). Hence, further work focusing on different implant geometries should be trialed to ensure amorphous content is limited both on the surface and within the bulk of the implant to encourage surgical acceptance.In general, the majority of literature reports PSCI surface resolution qualitatively. For example, Sharma et al. (2020a) reported increased levels of roughness in areas with amorphousness and regions where support structures had previously existed in vertically printed PSCIs. Additionally, stair-stepping phenomena was reported, attributing to a reduction in surface resolution, however both implants were considered clinically acceptable. Furthermore, when evaluating one of the PSCIs in a horizontal orientation, it was concluded that the high level of roughness caused by the removal of supports deemed the print clinically unacceptable for application (Sharma et al., 2021b). Petersmann et al. (2023) investigated the waviness of both milled and FFF printed PEEK implants. For the ‘gold-standard’ PSCI the waviness surface was considered low with negligible fluctuation (Fig. 2a). As for BO90 and BO45 specimens recorded similar values to each other, as well as the milled control, which was credited to the fact that both were printed along similar tool paths. However, far more levels of waviness were recorded for BO180 samples, which was attributed to both poorer welding between adjacent print layers and the impact of support structures on the bottom surface of the build. Increases in amorphous content rose both externally and internally when AF was decreased from the printed control (210 °C). However, negligible differences in the surface profile at the failure site existed, except for those printed without an airflow temperature which appeared far rougher. A general increase in waviness between the top and bottom of the printed samples rose when AF decreased, which is most likely credited to non-uniform heating throughout the build. In a study by Zhao et al. (2020), surface roughness was seen to rise with increasing TN (380 = 4.3; 400 = 7.8 and 420 °C = 5.4 μm. A small decrease in surface roughness was recorded as 3.9 μm when the 420 °C implants were annealed at 175 °C but this then climbed to 7.8 μm for those held at 250 °C (Fig. 2b). General implant roughness was attributed to the welding lines intrinsic to the FFF Process, whilst decreases in surface accuracy and rises in surface roughness with increasing TN and annealing was credited to stress relaxation. Whereby, higher temperatures promote greater polymeric chain reorganization in the implant causing deviations across the implants morphology.Wang et al. (2019b) reconstructed the chest cavities of 18 patients suffering from tumor resection surgery. This was achieved by implanting PEEK sternum-like structures and assessing their compatibility in-vivo. Findings from the study support the use of FFF PEEK to effectively reconstruct the anterior chest safely, however, follow-up data sets have not been reported in the literature. Liu et al. (2018) also reported similar clinical findings upon implantation of a FFF PEEK scapula into a 16-year-old male patient.This work was supported by the project CAMed (COMET K-Project 871132) and was funded by the Austrian Federal Ministry of Transport, Innovation and Technology (BMVIT) and the Austrian Federal Ministry for Digital and Economic Affairs (BMDW) and the Styrian Business Promotion Agency (SFG).

Publisher Copyright:
© 2023 The Authors


  • Additive manufacturing
  • Fused filament fabrication
  • Patient specific cranial implants
  • PEEK
  • Polyetheretherketone

Cite this