| Autor: |
Baraldi L; Preclinical Analytics and Early Formulations Department, Chiesi Farmaceutici Spa, Largo Belloli, 43123 Parma, Italy.; Department of Food and Drug, University of Parma, Parco Area delle Scienze, 43123 Parma, Italy., Bassanetti I; Preclinical Analytics and Early Formulations Department, Chiesi Farmaceutici Spa, Largo Belloli, 43123 Parma, Italy., Mileo V; Preclinical Analytics and Early Formulations Department, Chiesi Farmaceutici Spa, Largo Belloli, 43123 Parma, Italy., Amadei F; Preclinical Analytics and Early Formulations Department, Chiesi Farmaceutici Spa, Largo Belloli, 43123 Parma, Italy., Sartori A; Department of Food and Drug, University of Parma, Parco Area delle Scienze, 43123 Parma, Italy., Venturi L; Preclinical Analytics and Early Formulations Department, Chiesi Farmaceutici Spa, Largo Belloli, 43123 Parma, Italy. |
| Abstrakt: |
Physical forms of active pharmaceutical ingredients (APIs) play a crucial role in drug discovery since 85% of API molecules exhibit polymorphism and sometimes complicated phase behavior, often resulting in important differences in the respective biochemical and physical properties. Characterization and quantitation of the different forms are becoming more and more essential in the pharmaceutical industry: once these characteristics are known, it is easier to choose the best solid form for development, formulation, manufacturing, and storage. Time domain-nuclear magnetic resonance (TD-NMR) has recently been used to develop a quantitation protocol for solid mixtures, named qSRC, based on the linear combination of T 1 saturation recovery curves (SRCs) collected on a bench-top instrument. Despite its potentials and ease of use, a limited number of application cases have been reported in the literature since its development and many aspects remain to be clarified for the technique to be adopted as a robust routinely industrial analytical tool. In the present work, the reliability of the qSRC approach has been studied by focusing on the role played by key experimental variables, including mixture composition, signal-to-noise ratio, and T 1 differences. In silico simulations were carried out for a wide range of theoretical cases to predict the expected level of accuracy obtainable for a given sample-parameter acquisition set and to clearly define the range of applicability of the method. Results of the simulation are presented alongside a comparison with three real-case studies of commercially available APIs: piroxicam, naproxen sodium, and benzocaine. |
