Materials and methods
A variety of materials was extruded: Soluplus, Kollidon VA64, Eudragit EPO, Polycaprolactone, Poly-ethyleneoxide WSR N10, thermoplastic polyurethanes (EG72D, SP60D60, SP93A), Ethylene-vinylacetate (Ateva1070, Ateva2825A), Klucel EF (HPC EF). Ibuprofen was added as a model drug to PCL and PEO (20 % w/w, 40% w/w). Filaments were characterized using mechanical, rheological and thermal analysis. Filaments were used in 3D-printing (Prusa i3 MK3S) to determine the minimal printing temperature and maximal print speed for a variety of nozzle sizes. The effect of the material parameters on the visual quality of the end-products was assessed.
Results and discussion
Brittle filaments (SOL, KVA64, EPO) shattered on the gears of the printer. The mechanical method (based on ISO 527 – Fig. 1) could predict such behaviour based on the tensile energy to break the filament and strain at break. It was deemed impossible to print these filaments without changing the formulation by f.e. addition of a plasticizer.
For some polymers (PEO,PCL), printing conditions depended solely on the melting point (DSC) and cross-over point (G’=G”), determined by a temperature sweep in heating mode. Their behaviour was not influenced by the nozzle size of the printer. For other polymers with a more complex structure (TPUs, EVAs, HPC EF, drug loaded filaments), printing behaviour is largely influenced by the nozzle size. The following behaviour and printing failures were assessed:
- Flexible filaments buckled on the gears of the printer (SP60D60, SP93A, EVA1070, EVA2825A) in certain print conditions. The mechanical method could predict such behaviour based on the elastic modulus. For such filaments, increasing the nozzle size widened the print window, theoretically supported by the Euler buckling theory and equations describing Hagen-Poiseuille flow.
- Filaments with high viscosity (HPC EF, EG72D) blocked the nozzle of the 3D-printer at a certain temperature as it was determined that the upper viscosity limit for a Prusa i3 MK3S is 6,000 Pa.s. Increasing the nozzle size decreased the pressure drop and consecutively widened the print window as the upper viscosity limit shifted to 8,000 Pa.s. (Fig.2)
- At high nozzle size, the increased volumetric flow and characteristic low thermal conductivity of polymers counteract the beneficial effect on printability of increased pressure drop. While this is a general phenomenon, materials with a high Arrhenius activation energy seem more affected (EG72D).
- Filaments with low melting point (IBUPEO, IBUPCL, EVA2825A) softened on the gears in the enclosed printing chamber which impeded printability. An increased nozzle size enabled printing however, except for EVA2825A as this filament also possessed an extremely low Young’s modulus.
Fig. 1 The stress-strain curves of the filaments that broke during the tensile test at low displacement rate.
Fig. 2 Complex viscosity as a function of temperature during a heating sweep. Red line estimates maximal viscosity at nozzle size 0.4 for a Prusa i3 MK3S system, above which printing is not possible. With higher nozzle size, maximal viscosity is assumed to shift towards a higher value as indicated by the blue line.
It was shown that ibuprofen acted as a plasticizer for PCL and PEO by decreasing the overall viscosity and the minimal printing temperature. Either the quality of the end product was improved or over-plasticized structures were generated, depending on the ibuprofen content.
The current research showed that specific material properties determine the 3D printability and optimal process parameters for a certain formulation. Filaments should possess a high toughness and stiffness with low brittleness in order to be feedable and compatible with the printers’ gears. Secondly, if filaments are feedable, there is a complex interplay between their thermal, rheological and mechanical properties which determine the printability window. In general, enlarging the nozzle diameter of the printer reduces the minimal printing temperature, but this effect is (partially) counteracted by an increase of volumetric flow. Finally, a low melting point of the polymer could result in softening on the gears, which impedes successful feeding.
The full version of this article can be consulted at: https://doi.org/10.1016/j.ijpharm.2021.120361
(1) Moe Elbadawi, Brais Muñiz Castro, Francesca K.H. Gavins, Jun Jie Ong, Simon Gaisford, Gilberto Pérez, Abdul W. Basit, Pedro Cabalar, and Alvaro Goyanes. M3DISEEN: A novel machine learning approach for predicting the 3D printability of medicines. International Journal of Pharmaceutics, 590(August):119837, 2020.
(2) Witold Jamróz, Joanna Szafraniec, Mateusz Kurek, and Renata Jachowicz. 3D printing in pharmaceutical and medical applications. Pharmaceutical Research, 35(9):Article 176, 2018.