Using retention index database matching for compound identification on a non-standard gas chromatography stationary phase

In gas chromatography (GC), the retention index (RI) system was developed to compare the retention times of volatilizable compounds across GC systems and to RI databases to aid in compound identification. The mid-polar 624 GC stationary phase (6% cyanopropylphenyl/94% dimethylpolysiloxane) is becoming a common choice for bioanalysis and metabolomics. However, referenceable RI data for compounds separated on the 624 stationary phase are limited, hindering the ability to identify unknowns. Comparing 624 retention indices for 252 compounds to their corresponding nonpolar and polar retention indices in the National Institute of Standards and Technology (NIST) GC RI database, we identified a strong linear relationship (R2 = 0.99) between the 624 and nonpolar values and an equation to interconvert between them. Thus, retention indices obtained on a 624 stationary phase can now be compared to the NIST GC RI database; however, there are no standardized criteria for determining whether an observed RI and a database RI value are a match. To develop RI match criteria, we evaluated the variability of the RI values in NIST for each of the 252 compounds in this study. Based on our findings, we suggest defining an observed nonpolar RI that is within one standard deviation of the median value in NIST as a strong match (i.e., within 12 units for a nonpolar RI match or 14 units for a polar RI match), and defining a weak match using looser criteria, e.g., within three standard deviations of the median RI value in NIST. Lastly, we provide an example for how to use strong and weak RI matches as part of the workflow for compound identification in a gas chromatography – mass spectrometry analysis, with the goal of assigning Metabolomics Standards Initiative confidence levels to each identification.