A collaborative study to investigate the retention reproducibility of barbiturates in HPLC with a view to establishing retention databases for drug identification.

A collaborative study among 10 laboratories has been undertaken to investigate the interlaboratory reproducibility of retention measurements in a high performance liquid chromatographic (HPLC) system for separating barbiturates. The system involved an ODS-Hypersil column with an eluent of methanol:phosphate buffer (40:60, v/v) at pH 8.5; all laboratories used the same batch of packing material. The conventional methods of recording retention properties (retention times and capacity factors) gave poor interlaboratory reproducibility. Better results were obtained by expressing the retentions relative to a standard barbiturate (quinalbarbitone); relative adjusted retention times proved to be more effective than straightforward relative retention times. Retention indices based on the alkylarylketone scale were not as reproducible as the methods based on a single closely related compound. The best reproducibility was obtained using corrected capacity factors based on the retention of four barbiturates in a standard mixture. The results of the study are discussed with a view to selecting the best methods of recording retention in HPLC when considering the establishment of databases for drug identification.