The topic of the diploma thesis is fast separation of explosives by HPLC and development of miniaturized liquid chromatograph for application in a handheld explosives detection device. In this work the retention of some nitrated explosives and selectivity in reversed phase system as a function of mobile phase composition is studied while methanol, acetonitrile and acetone as an organic solvent is used. Best selectivity and good retention can be observed in methanol mobile phase. Acetonitrile and acetone are not suitable for fast isocratic separation of mixture containing pentaerythritol tetranitrate because of its strong retention. Efficiency and permeability of monolithic column (Chromolith CapRod RP-18e) and columns filled with superficially porous particles are compared (Kinetex 2,6 µm C-18, Poroshell 120 SB-C18). Monolithic column with satisfying efficiency and high column permeability is the most suitable solution for fast separation of explosives. Assuming use of explosives detection device in different conditions the separation was optimized on temperature 50 °C. Under these conditions the 35% v/v methanol gives good retention and selectivity. For very fast scan analysis of pentaerythritol tetranitrate or other nitroaromatics use of 70% v/v acetone mobile phase is suitable. Construction of miniaturized liquid chromatograph that enables preconcentration of explosives from aqueous solutions and fast separation in less than 1 minute is described. This concept will be incorporated into the handheld explosives detection device where the explosives vapor will be absorbed into the water and after the separation detected by chemiluminescence.

Although the airbath oven is a reliable heating method for gas chromatography (GC), resistive heating is needed for higher analytical throughput and on-site chemical analysis because of size, heating rate and power requirements. In the last thirty years, a variety of resistive heating methods were developed and implemented for both benchtop and portable GC systems. Although fast heating rates and low power consumption have been achieved, losses in column efficiency and resolution, complex construction processes and difficulties experienced in recovering damaged columns have also become problematic for routine use of resistively heated columns. To solve these problems, a new resistively heated column technique, which uses metal columns and self-insulated heating wires, was developed for capillary gas chromatography. With this method, the total thermal mass was significantly less than in commercial column assemblies. Temperature-programming using resistive heating was at least 10 times faster than with a conventional oven, while only consuming 1—5% of the power that an oven would use. Cooling a column from 350 °C to 25 °C with an air fan only required 1.5 min. Losses in column efficiency and peak capacity were negligible when compared to oven heating. The major trade-off was slightly worse run-to-run retention time deviations, which were still acceptable for most GC analyses. The resistively heated column bundle is highly suitable for fast GC separations and portable GC instruments. Fabrication technologies for microelectromechanical systems (MEMS) allow miniaturization of conventional benchtop GC to portable, microfabricated GC (µGC) devices, which have great potential for on-site chemical analysis and remote sensing. The separation performance of µGC systems, however, has not been on par with conventional GC. Column efficiency, peak symmetry and resolution are often compromised by column defects and non-ideal injections. The relatively low performance of µGC devices has impeded their further commercialization and broader application. This problem can be resolved by incorporating thermal gradient GC (TGGC) into microcolumns. Negative thermal gradients reduce the on-column peak width when compared to temperature-programmed GC (TPGC) separations. This unique focusing effect can overcome many of the shortcomings inherent in µGC analyses. In this dissertation research, the separation performance of µGC columns was improved by using thermal gradient heating with simple set-ups. The analysis time was ~20% shorter for TGGC separations than for TPGC when wide injections were performed. Up to 50% reduction in peak tailing was observed for polar analytes, which significantly improved their resolution. The signal-to-noise ratios (S/N) of late-eluting peaks were increased by 3 to 4 fold. These results indicate that TGGC is a useful tool for bridging the performance gap between µGC and benchtop GC.