Purpose: The purpose of this work was to identify and prevent the formation of unknown impurity peak during RP-HPLC analysis of an opioid API. Upon introduction of an opioid active pharmaceutical ingredient (API) analytical release method, a spurious unknown peak was detected during HPLC impurity analysis. The peak was initially reported as an impurity, which approached the ICH identification threshold (0.10%) for a high dose product (>2 g daily dose). This impurity was not observed in blank injections, but was present in all samples and standards at relatively consistent levels. The peak was asymmetrical and resembled a small shoulder relative to the main analyte of interest. Extensive studies were conducted to determine whether the unknown peak was a true impurity of the API or an artifact of the chromatographic system. Once the identity and mechanism of the peak formation was determined, a strategy was developed to implement method controls to prevent in-situ formation of the impurity during chromatographic analysis.
Methods: The analytical method under investigation is a reversed phase HPLC analysis with ammonium formate/ammonium hydroxide mobile phase (A) and acetic acid diluent. Systematic evaluation of the following method components was conducted: Equipment Makes and Models, Reagents, Filtration, Gradient, Injection Volume, Mobile Phase, and Diluent. Electrospray ionization mass spectrometry on both a Waters QDA mass spectrometer and a Waters triple quadrupole detector were used to identify the impurity’s nominal mass and fragmentation patterns.
Results: The evaluation of various reagent grades and manufacturers did not impact the observed level of impurity. Different instruments from two manufacturers (Agilent 1100, 1200, 1260, Waters Alliance) generated consistent impurity results. Filtration of the mobile phase performed with different filters did not alter the impurity level. Throughout the gradient and injection volume experiments, the unknown peak remained proportional to the analyte of interest. Further review of the test method indicated that a pH mismatch between the acidic standard/sample diluent and the alkaline mobile phase may cause extraneous peaks to form. An API sample diluted entirely in mobile phase A generated a sharper, larger unknown peak. It was therefore demonstrated that the ammonia in the mobile phase was likely reacting with the sample to produce the observed impurity. Addition of ammonium hydroxide directly to the sample preparation produced a significant increase in the impurity level.
The increased level of the impurity generation upon API dilution in alkaline mobile phase A and addition of ammonium hydroxide suggested an on-column reaction between the API and the ammonia in the mobile phase. An API imine formation was proposed as the most likely source of the detected unknown peak. From a single Quadrupole detector (QDa), a difference of one Da was observed between the API (300) and the unknown impurity (299). These nominal masses were consistent and indicated that additional nitrogen was present in the impurity structure and an oxygen atom is missing. Additional mass spectrometric work was conducted using a triple quadrupole detector to establish whether additional fragmentation patterns supported the API imine structure. Observed patterns were compared to the theoretical fragmentation patterns and found to be consistent, further establishing the structure.
Since the root cause of the unidentified peak was found to be interaction with the mobile phase, steps were taken to adjust the method within the validated and established robustness parameters to minimize imine formation. Specifically, the target mobile phase pH was lowered to 9.4 from 9.6 to minimize impurity formation during the analysis by lowering the amount of ammonia present to react with the analyte of interest. This change has resulted in minimal to no in-situ imine impurity formation during analyses.
Conclusion: It was determined that the API of interest in the reversed phase HPLC impurity analysis was in-situ reacting with the ammonia in the alkaline mobile phase to form an imine impurity. This was the reason why the unknown peak was only observed in injections containing the API. The mobile phase pH was modified within the validated robustness range to minimize formation of this impurity.
Laura DeShullo
– Senior Analytical Scientist, Noramco, BearJoshua Hoerner
– Analytical Services Manager, Noramco Inc, Athens, GeorgiaJoshua Hoerner
– Analytical Services Manager, Noramco Inc, Athens, GeorgiaLaura DeShullo
– Senior Analytical Scientist, Noramco, Bear243 Views