Despite continued advances in mass spectrometry, our inability to identify small molecules from complex environmental samples remains a major bottleneck. In LC-MS and GC-MS, the MS information can be reliably used for compound identification, but retention times, relative retention times, and retention indices are notoriously unreliable. They are most accurate when the original experimental conditions used to measure them are rigorously reproduced, but even then, unavoidable differences between HPLC instruments (solvent misproportioning, gradient delay, gradient dispersion, etc.) or between GC instruments (flow rate errors, temperature calibration error, etc.) cause considerable error. We present a new, easy-to-use methodology that solves this problem: We spike our samples with a set of standards and use their retention times in a gradient (HPLC) or temperature program (GC), to precisely back-calculate the actual gradient or temperature program produced by the instrument. When we then use that information to calculate the retention of other compounds, the calculated retention times are extremely accurate (e.g. to �0.4 s in a 20 min run). The methodology also gives considerable flexibility in allowable experimental conditions, and in GC, even the error is predictable across labs and method, allowing automatic calculation of retention time tolerance windows without the need to have standards physically on hand. We developed a public website containing open-source software and an HPLC and GC retention database at www.retentionprediction.org. We will also briefly present several other free, open-source tools we recently developed.

Easy, accurate new way to predict HPLC and GC retention times (and other free online tools)

248th National Meeting (2014)

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.