Trials to induce complex coacervation between two grades Gantrez-AN polymer (G-AN), and Type A gelatin were made. Physical parameters influencing the coacervation process were studied. Maximum coacervation was attained when the pH of the gelatin solution was at 6.8. Increasing the molecular weight of Gantrez decreased the ratio of combination of both polymers. The ratio for optimum coacervation was 1:4 for Gantrez-AN 119-gelatin system and 2:3 for Gantrez-AN 149-gelatin system with total colloid concentration of 2.5 g per cent w/v in both cases. High stirring speed gave almost spherical uniform coacervates. Recovery of the product as water-insoluble discrete units required the use of formaldehyde and isopropanol for coacervate denaturation and flocculation, respectively.

Nitrofurantoin crystals were encapsulated in a Gantrez-gelatin complex coacervation system. The encapsulation process was reproducible and inexpensive and the microcapsules were free flowing and directly compressible into tablets. In vitro release of nitrofurantoin from Gantrez-gelatin microcapsules was studied as a function of the core:coat ratio, the molecular weight of Gantrez and the particle size of the microcapsules. The release of the drug was significantly reduced using G149-gelatin microcapsules of core:coat ratio of 1:2. Release data were examined kinetically and were found to follow a diffusion-controlled model. In vitro release of the drug from the microcapsules filled in capsules and compressed into tablets confirmed the efficiency of the encapsulation process for preparing prolonged release formulations.

The bioavailability of marketed nitrofurantoin capsules was compared to capsules filled with Gantrez-gelatin nitrofurantoin microcapsules of a core:coat ratio of 1:2. The collective results obtained from the analysis of urine samples of five volunteers indicated that nitrofurantoin microcapsules provided a prolonged release compared with that of the control formulation. Analysis of variance showed statistically significant differences between the control formulation and the capsule filled with the microcapsule. Non-significant intersubject variation was noticed.