Extensive Automation of the SPE-GC/MS Analysis of Opiates, Cocaine, their Metabolites and Methadone from Serum and other Matrices

Pieter Stoutjesdijk, International Sales, GERSTEL GmbH & Co KG, Muelheim an der Ruhr, Germany
Oliver Lerch, Analytical Service, Gerstel Gmbh & Co Kg, Muelheim An Der Ruhr, Germany
Oliver Temme, Forensic-toxicology, Institute For Legal Medicine/heinrich Heine Universityuniversity, Düsseldorf, Germany
Thomas Daldrup, Forensic-toxicology, Institute For Legal Medicine/heinrich Heine University, Düsseldorf, Germany
Gernot Brauers, Forensic-toxicology, Institute For Legal Medicine/heinrich Heine University, Düsseldorf, Germany

Aim: The application of an extensively automated method for the analysis of cocaine, ben­zoylecgonine, methadone, morphine, codeine, dihydrocodeine and monoacetylmorphine basing on a validated semi automated routine method was verified.

Methods: The “online” preparation of precipitated blood, serum, urine and different tissues was performed on a x-y-z robot including the solid phase extraction with modified standard cartridges (Bond Elut Certify), evaporation of the eluate, derivatization (silylation) and injec­tion into a GC/MS. Over 150 serum samples and more than 50 samples of other matrices were analyzed and compared to the analyses performed with a validated classic semi automated standalone SPE with manual evaporation and derivatization.

Results and Discussion: The analytical results of both methods are equivalent even near the limits of quantification. To proof the correlation between the two methods corresponding concentrations were compared. Coefficients of linear regressions (coefficient of slope, coeffi­cient of determination “R²”) were: 0.96, 0.9947 (cocaine); 0.99, 0.9800 (benzoylecgonin); 0.99, 0.9942 (morphine); 1.00, 0.9934 (codeine); 1.08, 0.9725 (methadone) and 0.99, 0.9957 (monoacetylmorphine). For dihydrocodeine (and also for 7-amino-flunitrazepam which is not mentioned above) there are too few positive results to determine these values. For both sub­stances accordance was proven by some control samples. The complete extraction procedure for one sample takes nearly one hour at the present state but blank injections can be inter­leaved so that more than 24 samples can be analyzed per day.

Conclusion: The automated method is suitable for routine analyses in forensic toxicology.