Microchip gas chromatographic columns dedicated for space exploration: Stationary phase coating, setup optimization and evaluation of column performances

This work is the first part of a project aiming to specifically design gas chromatographic microcolumns for space exploration. In particular, this study explored the functionalization and characterization of silicon/glass microchip gas chromatographic columns designed with a serpentine-shaped channel having an internal square cross-section. This microcolumn can be connected using a robust fluidic manifold with removable capillary connections to a conventional gas chromatograph or implemented in a prototype instrument for space exploration. First, benchtop gas chromatographic system was optimized in terms of injector liner volume, internal diameter (I.D.) of the capillary connections and data frequency of detection to ensure an optimal evaluation of column efficiency. Junction capillaries of 100 μm I.D., a liner of 1.2 mm I.D. and a detection acquisition frequency of 100 Hz were found to be the optimal set-up and parameters. Moreover, the stationary phase coating velocity was slowed-down to increase column efficiency by 100 %. Then, the performances of a square cross-section capillary column were compared with those of a conventional circular cross-section column. The coating efficiencies were estimated at 55 % and 42 % for circular and square internal cross-section geometries respectively, demonstrating that the square section geometry did not affect significantly the column performances. Stationary phase films of different thicknesses, from 0.032 to 0.260 µm, were coated on different microchips of the same production batch to assess the influence of film thickness on the chromatographic performances. Microchips performance was evaluated through Golay's plots, and some of the microchips were also studied by optical microscopy. The efficiency and retention capacity of our microchip column is shown to be highly dependent on the film thickness (9,000 plates.m^-1 with k = 0.12 for the thinnest film and up to almost 4,000 plates.m^-1 with k = 0.93 for the thickest film). Finally, the coating process repeatability was validated by producing three identical microchips with RSD of 4.8 % for the average number of theoretical plates.