Ong, Jun Jie; (2024) Predictive modelling and novel materials for pharmaceutical 3D printing. Doctoral thesis (Ph.D), UCL (University College London).

Preview

Text thesis_jjong_final.pdf - Other Download (19MB) | Preview

Abstract

Pharmaceutical 3D printing (3DP) is revolutionising healthcare by enabling the production of bespoke medicines, tailored to an individual’s unique clinical needs, genetic makeup, physiology, and preference. Various 3DP technologies have been explored for pharmaceutical applications, such as fused deposition modelling (FDM), inkjet printing, and vat photopolymerization (e.g. digital light processing (DLP), volumetric printing). However, the clinical translation of 3DP is impeded by numerous challenges, such as the empirical process of formulation development, and material limitations for vat photopolymerization 3DP. Specifically on the latter, existing materials used in pharmaceutical vat photopolymerization 3DP are limited to insoluble matrices, which restricts the potential application of the technology. Therefore, this PhD thesis aims to (a) develop predictive models using machine learning (ML) to accelerate 3DP formulation development, and, owing to lack of material diversity and sufficient data in pharmaceutical vat photopolymerization, (b) develop a novel material that gives soluble matrices for vat photopolymerization 3DP. The major findings were as follows: (1) ML models provided accurate predictions of FDM printability and filament characteristics, and narrower temperature ranges for hot melt extrusion and FDM printing than conventional guidance and previous computational models. (2) ML models gave high prediction accuracies for inkjet printability and print quality. The ML model also provided better prediction accuracy than conventional guidance based on the ink’s Ohnesorge number. (3) Formulations based on supramolecular chemistry, comprising [2-(acryloyloxy)ethyl]trimethylammonium chloride and 1-vinyl-2-pyrrolidone, were successful used to print tablets via DLP 3DP that dissolved within 45 minutes to 5 hours in dissolution media. (4) Volumetric printing was successfully used to print water-soluble tablets within 11 seconds, and models based on their NIR spectra provided high drug loading prediction accuracies. Overall, this research demonstrates the potential for ML models to accurately predict printing outcomes, and reports the first water-soluble dosage form printed via vat photopolymerization.

Download activity - last month

Export as XMLCSVJSON

Download activity - last 12 months

Export as JSONXMLCSV

Downloads by country - last 12 months

Export as JSONCSVXML

Archive Staff Only

View Item