Bioceramics have been studied extensively as drug delivery systems (DDS). Those studies have aimed to tailor the drug binding and release kinetics to successfully treat infections and other diseases. This research suggests that the drug binding and release kinetics are predominantly driven by the functional groups available on the surface of a bioceramic. The goal of the present study is to explain the role of silicate and phosphate functional groups in drug binding to and release kinetics from bioceramics.

α-cristobalite (Cris;​ SiO2) particles (90-150 µm) were prepared and doped with 0 µg (P-0), 39.1 µg (P-39.1), 78.2 µg (P-78.2), 165.5 µg (P-165.5) or 331 µg (P-331) of P₂O₅ per gram Cris, using 85% orthophosphoric (H3PO4) acid and thermal treatment. The material structure was analyzed using X-ray diffraction (XRD) with Rietveld Refinement and Fourier Transform Infrared (FTIR) spectroscopy with Gaussian fitting. XRD demonstrated an increase from sample P-0 (170.5373 ų) to P-331 (170.6466 ų) in the unit cell volume as the P₂O₅ concentration increased in the material confirming phosphate silicate substitution in Cris. Moreover, FTIR showed the characteristic bands of phosphate functional groups of ν₄ PO4/O-P-O bending, P-O-P stretching, P-O-P bending, P=O stretching, and P-O-H bending in doped Cris indicating phosphate incorporation in the silicate structure. Furthermore, FTIR showed that the ν₄ PO4/O-P-O bending band around 557.6 cm⁻¹ and P=O stretching band around 1343.9 cm⁻¹ increased in area for samples P-39.1 to P-331 from 3.5 to 10.5 and from 10.1 to 22.4, respectively due to phosphate doping. In conjunction with the increase of the ν₄ PO4/O-P-O bending band and P=O stretching band, a decrease in area of the O-Si-O bending bands around 488.1 and 629.8 cm⁻¹ was noticed for samples P-39.1 to P-331 from 5 to 2 and from 11.8 to 5.4, respectively. Furthermore, Cris samples (200 mg, n=5 for each sample) were immersed separately in DI water for 2 days and the concentrations of dissolved silicate and phosphate ions released from the surface of Cris were measured using Inductively Coupled Plasma – Optical Emission Spectrometry (ICP-OES). The phosphate ions released from the material activated the surface and exposed the silicate functional groups as indicated by the FTIR analysis. Pre-immersed Cris particles and control non-immersed samples (200 mg, n=5 for each sample) of particle size 90-150 µm were immersed in 2 mL of vancomycin (Vanc) solution (8 mg/ml) in PBS on an orbital shaker at 37°C for 24 hours. The amount of drug bound to the material was measured by High Performance Liquid Chromatography (HPLC). Control non-immersed Cris samples P-0 and P-39.1 adsorbed a comparable amount of drug. While there was a statistically significant lower amount of drug adsorbed onto P-78.2 than that adsorbed onto P-39.1 (p < 0.001), comparable amounts of drug were adsorbed onto P-78.2, P-165.5, and P-331. Releasing phosphate ions from the material surface resulted in a significant increase in drug adsorption for pre-immersed samples. Higher Vanc adsorption was noticed for all pre-immersed Cris samples compared to their corresponding control non-immersed samples. Moreover, for pre-immersed samples the amount of drug adsorbed significantly increased from P-0 to P-78.2 (P-0 < P-39.1 < P-78.2;​ p < 0.05). However, at phosphate content higher than 78.2 µg per gram of Cris there was a significant decrease in drug adsorption (P-78.2 > P-165.5 > P-331;​ p < 0.001). ICP-OES analyses showed that the percent of released phosphate ions during immersion decreased as the phosphate content in doped Cris increased (P-39.1 released 92±.08% and P-331 released 71±.05%). Therefore, the decrease in drug binding could be attributed to the presence of high phosphate content on the material surface. Comparison between the HPLC and FTIR analyses showed that ceramics that had higher content of O-Si-O bending (at ~498 cm⁻¹ and ~620 cm⁻¹) bands facilitated Vanc adsorption. On the other hand surfaces with a higher content of ν₄ PO4/OP-O bending (at ~557 cm⁻¹ ) and P=O stretching (at ~1343.9 cm⁻¹) bands did not enhance Vanc adsorption. Drug loaded pre-immersed and control non-immersed Cris samples (each 200 mg, n=5 for each sample) were immersed in 2 mL of PBS on an orbital shaker at 37°C, and a 0.5 mL aliquot was removed from the solution and replenished at 1, 3, 6, 8, 24, and 48 hour, and every 48 hour intervals to 22 days thereafter. Drug concentration released from Cris samples after each time point was measured using HPLC. The drug release kinetics demonstrated a statistically significant decrease (p < 0.05) in the cumulative and percent of Vanc released from control non-immersed Cris samples P-0 (1.521 ± .026 mg;​ 37.66 ± .89 %) to P-331 (1.276 ± .016 mg;​ 33.46 ± .77 %) of Vanc, respectively. Additionally, release kinetics also demonstrated statistically significant increase (p < 0.05) in the cumulative and percent of Vanc released from pre-immersed samples P-0 (1.505 ± .014 mg;​ 33.59 ± 1.35 %) to P-331 (1.581 ± .057 mg;​ 42.27 ± 1.51 %) of Vanc, respectively. Furthermore, in the first 4 hours, the deceleration of drug release from sample P-0 to P-331 decreased from -66.92 to -34.07 µg of Vanc/mL /hr², for control non immersed Cris and from -72.60 to -46.04 µg of Vanc/mL/hr², for preimmersed samples. Furthermore, during the first 4 hours of burst release the percentage of drug released from the total amount of drug loaded for non-immersed samples P-0 was 41 % and for P-331was 26 %. After the 4 hours of Vanc release the amount of Vanc available for release for samples P-0 and P-331 was .898 mg and .945 mg, respectively. The same relationship was found for pre-immersed samples during the first 4 hours of burst release the percentage of drug released from the total amount of drug loaded for samples P-0 was 42 % and for P-331 was 30 %. After the 4 hours of Vanc release the amount of Vanc available for release for samples P-0 and P-331 was .873 mg and 1.106 mg, respectively. These results indicated the effect of phosphate content on decreasing the drug release rate. The drug release kinetics study showed that the release of phosphate ions from the surface of Cris prior to drug loading exposed active silicate functional groups that enhanced drug binding by physisorption which in turn facilitated rapid release kinetics. On the other hand, a slower drug release rate was observed as the phosphate functional groups increased on the material surface due to chemisorption. Results from the present study indicate that it is possible to enhance the burst release stage of a bioceramic drug carrier by increasing the silicate functional groups. The sustained release profile can be engineered by controlling the phosphate content of the bioceramic drug carrier.