Background
Figure 1: Bar graph showing purity determinations of a variety of blind-spiked samples of ibuprofen. The samples were spiked with a known weight of either tetrabutylammonium fluoride (TBAF), theophylline, or vanillin by one analyst and then HPLC/qNMR purity analyses performed by separate analysts who were unaware of the theoretical purity.
The quantitative determination of purity in active pharmaceutical ingredients (APIs) and their intermediates is an integral part of the research, manufacture and clinical evaluation of drugs at all phases of development.
Quantitating the amount of API in a sample helps assess reaction yield, estimate purity, estimate impurities, and calibrate dosages. The quantitative analysis of APIs has traditionally been done indirectly, using a combination of high-performance liquid chromatography (HPLC) for relative purity and impurities; gas chromatography (GC), thermal gravimetric analysis (TGA) or Karl Fischer (KF) titration for residual volatiles and water; and residue on ignition (ROI) for inorganic contaminants. The calculated quantitative purity is typically the HPLC relative purity less the results from the GC, TGA, KF, and ROI experiments.
Though this has become the gold standard of quantitative API analysis, this procedure requires significant amounts of material to perform the myriads of experiments. Additionally, HPLC methods must be developed and validated for each compound along the synthetic route, taking several weeks and increasing product development timelines, budget and additional material necessary for the characterization assays.
Nuclear magnetic resonance (NMR) has the capability to identify not only the product of interest but also residual solvents and other organic impurities. Using a known amount of almost any commercially available reference standard, quantitative information about the purity of the analyte can be obtained without needing a reference standard of the exact same compound to compare against. Studies have shown qNMR to be extremely useful in determining the weight-for-weight purity of a variety of organic compounds, but data within the pharmaceutical context for qNMR in direct comparison to HPLC has been sparse.
Approach
We chose to study the commercially available compounds ibuprofen, caffeine and hydroxychloroquine, which were chosen based on their structural similarity to other APIs that have been investigated at SwRI. Throughout the course of the study, standard procedures for optimizing NMR parameters, such as T1 relaxation delay and acquisition time were developed using USP standards of these compounds. After parameter optimization for each compound, qNMR results were compared to results obtained by HPLC. To better assess the strengths and weaknesses of each technique, blind-spiked samples of each standard compound were analyzed by both qNMR and HPLC. Because qNMR provides a weight-for-weight assessment of purity, we compared the qNMR results to both HPLC % by area, as would be used for early phase work, as well as HPLC % recovery, determined via comparison to a standard curve generated using various known concentrations of each reference standard. Figure 1 shows the results of this blind-spiked study for ibuprofen.
Accomplishments
This study has provided clear evidence that qNMR can be a powerful tool in drug development when used in conjunction with HPLC. For early-phase work, without a fully developed analytical method, qNMR has shown that it can provide more reliable data than HPLC because it is a direct method of analyzing purity and does not rely solely on the physical properties of the analyte or impurities, namely UV absorption. This will allow for more accurate purity determination of compounds in early phase work as well as allowing for analysis of compounds that do not contain strong chromophores and produce weak signals in HPLC or are essentially invisible in an HPLC chromatogram. In such cases, it could be impossible to determine accurate dosing of the API for drug efficacy and toxicity studies. For late-stage development, once a full analytical method has been developed, qNMR has shown that it can be used as an orthogonal method to HPLC by confirming the weight-for-weight purity of the target compound, providing more confidence in dose calculations, and leading to a better overall drug product.