RESUMO
Particle packed columns used for liquid chromatography (LC) can suppress a column internal band broadening (hereinafter referred to as band broadening) by packing monodisperse particles homogenously. However, a quantitative evaluation for the effects of particle shape and packed state on band broadening needs to be more investigated. In this study, we fabricated a model of particle packed bed using microfluid LC columns that have pillar array structure prepared by microfabrication technology, evaluating how structural factors inside of a column affect its band broadening. At first, microfluid LC columns was prepared using Si-quartz glass (Si-Q column) for the optimization of LC measurement system. Through the evaluation, it showed 11.6 times higher pressure tolerance compared to that of PDMS-soda lime glass (PDMS-g column). Then, an optimized LC measurement system was constructed using a microfluidic LC column made of Si-Q column, and it was confirmed that the measurement error was small enough and the LC measurement could be performed with high repeatability. Additionally, the effect of a distribution of structural size on band broadening was evaluated. It was confirmed that wide distribution of the structural size provided large band broadening in actual measurements. Comparing two columns having different structural log-normal distributions of 0 and 0.22 showed approximately 1.8 times difference in both real LC measurement. Lastly, the relationship between packed state and band broadening was evaluated. As packed state, we employed void arrangement and structural arrangement in the columns. Different location arrangements of 50 and 100 µm pillar sizes afforded different band broadening. Well-homogenized array showed approximately two times worse band broadening compared to that of delocalized array. Based on these results, the developed packed bed of particles model was able to evaluate the relation between structural factors and band broadening.
