Near-Infrared Spectroscopic Monitoring of Point-of-Care Three-Dimensional Printing of Caffeine Tablets by Direct Powder Extrusion

This study investigates the use of near-infrared (NIR) spectroscopy as a non-destructive, real-time quality control tool for 3D-printed caffeine tablets produced via direct powder extrusion (DPE). The primary objective was to evaluate whether NIR spectral data could reliably predict key quality attributes including caffeine concentration, porosity, and dissolution behavior using chemometric modeling. Caffeine-loaded hydroxypropyl cellulose (HPC) formulations ranging from 0% to 20% caffeine were prepared and printed into tablets using a DPE-equipped 3D printer. Physical characterization included rheology, differential scanning calorimetry, bulk and true density measurements, and in vitro dissolution testing. NIR spectra were collected using both a benchtop NIR system and a miniature NIR device. Partial least squares regression was employed to correlate spectral variation with tablet quality attributes, achieving strong predictive performance for caffeine content (RMSEP = 0.6%), moderate performance for porosity (RMSEP = 2.4%), and early dissolution (RMSEP = 1.1% at 15 min), and poor accuracy for later dissolution time points. The miniature NIR device yielded comparable results to the benchtop system, supporting its feasibility for point-of-care implementation. These findings highlight the potential of NIR spectroscopy, particularly when paired with chemometric modeling, to enable rapid, non-invasive quality assessment of 3D-printed pharmaceuticals. Incorporating miniature NIR devices into 3D printing workflows offers a practical and potentially automated pathway to support decentralized pharmaceutical manufacturing, enabling real-time quality assurance for each print.