Quantification of 2D Printed Medicine Using Near Infrared Spectroscopy

In this work three “inks” consisting of sodium picosulfate solved in water, carvedilol solved in ethanol and metformin hydrochloride solved in water were formulated and printed on a number of substrates namely: EVA, HPMC, HGC, icing sheet, potato starch wafer and HGC with 2% titanium dioxide using the pico system of Scienion’s sciFLEXXARRAY printer.
After a drying period of one day the samples were measured using EVK hyperspectral imaging system, which measures wavelengths ranging from 931 - 1700 nm. The samples were then displayed in MATLAB. After comparing all samples, the combination of metformin hydrochloride printed on gelatin with 2% titanium dioxide was chosen to investigate the possibility of implementing NIR spectroscopy as PAT technology for printed medicine.
A calibration model was created with a number of spots with increasing drops per spot using projection to latent structures as implemented in MATLAB using the SIMPLS algorithm.
In order to create an appropriate response matrix Raman measurements were conducted, by revealing the distribution of API within the sample drops.
To determine the viability of the model HPLC measurements were conducted on both the calibration and the test samples.