Application of 2D-HPLC coupled with principal component analysis to study an industrial opiate processing stream.

Principal component analysis (PCA) loading plots were used to elucidate key differences between two-dimensional high performance liquid chromatography (2D-HPLC) fingerprint data from samples collected from stages along the Papaver somniferum industrial process chemistry workflow. Data reduction was completed using a 2D-HPLC peak picking approach as a precursor to chemometric analysis. Using comparisons of the final stages of product extraction as an example, PCA analysis of characteristic 2D-HPLC fingerprints accounted for 84.9% of variation between the two sample sets measured in triplicate, with 64.7% explained by PC1. Loadings plots of PC1 on each sample set identified where the significant changes were occurring and normalised bubble plots provided an indication of the relative importance of each of these changes. These findings highlight 2D-HPLC with appropriate chemometric analysis as a useful tool for the exploration of bioactive molecules within biomass.

In-silico optimisation of two-dimensional high performance liquid chromatography for the determination of Australian methamphetamine seizure samples.

In-silico optimisation of a two-dimensional high performance liquid chromatography (2D-HPLC) separation protocol has been developed for the interrogation of methamphetamine samples including model, real world seizure, and laboratory synthesised samples. The protocol used Drylab(®) software to rapidly identify the optimum separation conditions from a library of chromatography columns. The optimum separation space was provided by the Phenomonex Kinetex PFP column (first dimension) and an Agilent Poroshell 120 EC-C18 column (second dimension). To facilitate a rapid 2D-HPLC analysis the particle packed C18 column was replaced with a Phenomenex Onyx Monolithic C18 withought sacrificing separation performance. The Drylab(®) optimised and experimental separations matched very closely, highlighting the robust nature of HPLC simulations. The chemical information gained from an intermediate methamphetamine sample was significant and complimented that generated from a pure seizure sample. The influence of the two-dimensional separation on the analytical figures of merit was also investigated. The limits of detection for key analytes in the second dimension determined for methamphetamine (4.59×10(-4)M), pseudoephedrine (4.03×10(-4)M), caffeine (5.16×10(-4)M), aspirin (9.32×10(-4)M), paracetamol (5.93×10(-4)M) and procaine (2.02×10(-3)M).

Blind column selection protocol for two-dimensional high performance liquid chromatography.

The selection of two orthogonal columns for two-dimensional high performance liquid chromatography (LC×LC) separation of natural product extracts can be a labour intensive and time consuming process and in many cases is an entirely trial-and-error approach. This paper introduces a blind optimisation method for column selection of a black box of constituent components. A data processing pipeline, created in the open source application OpenMS®, was developed to map the components within the mixture of equal mass across a library of HPLC columns; LC×LC separation space utilisation was compared by measuring the fractional surface coverage, fcoverage. It was found that for a test mixture from an opium poppy (Papaver somniferum) extract, the combination of diphenyl and C18 stationary phases provided a predicted fcoverage of 0.48 and was matched with an actual usage of 0.43. OpenMS®, in conjunction with algorithms designed in house, have allowed for a significantly quicker selection of two orthogonal columns, which have been optimised for a LC×LC separation of crude extractions of plant material.