Skip to Content
MilliporeSigma
  • Gradient method transfer after changing the average pore diameter of the chromatographic stationary phase I - One-dimensional sample mixture.

Gradient method transfer after changing the average pore diameter of the chromatographic stationary phase I - One-dimensional sample mixture.

Journal of chromatography. A (2019-03-25)
Fabrice Gritti
ABSTRACT

Three different approaches designed to transfer gradient methods from a chromatographic column 1 packed with particles (2.7 μm 90 ÅCORTECS-C18 or CORTECS-Triphenyl) to a column 2 packed with the same particles having a larger average pore diameter (APD = 120 Åand 450 Å) are proposed for a one-dimensional sample mixture. Two approaches are based on the variation of the experimental retention plots (lnk vs. the volume fraction, φ, of the strong solvent) with increasing the APD. They lead to so-called "vertical" (vertical shift in column phase ratio, lnϕ2ϕ1) and "horizontal" (horizontal shift in eluent composition, Δφ1 →2) gradient method transfers. The third method is based on in silico predictions of the gradient retention times when considering the actual non-linearity of the retention plots. The adjusted gradient parameters (starting eluent composition, φ0, and temporal gradient steepness, β) are unambiguously determined by minimizing the distance between the calculated and targeted gradient retention times of all the analytes. The performances of the three approaches for gradient method transfer are compared for a sample mixture containing a non-retained compound (thiourea) and a series of five homologous compounds (n-alkanophenones). The ultimate goal is to keep unchanged the gradient retention times of all analytes when changing the APD of the particles. The results show that the in silico transfer systematically outperforms the "horizontal" transfer, which itself outperforms the "vertical" transfer. The first two approaches are the least successful ones because, even for a series of homologous compounds, the linear solvent strength model (LSSM) is only an approximate model and the best shifts in eluent composition that keeps retention factors unchanged are compound-dependent. In the end, the average relative deviations between the observed and targeted gradient retention times are 15.0, 1.7, and 0.4% (90 Åto 120 Åtransfer, C18 chemistry), 5.1, 5.8 and 0.8% (90 Åto 450 Åtransfer, C18 chemistry), 4.8, 0.5, and 0.4% (90 Åto 120 Åtransfer, Triphenyl chemistry), and 10.4, 7.1, and 2.2% (90 Åto 450 Åtransfer, Triphenyl chemistry) for the "vertical", "horizontal", and "in silico" gradient method transfers, respectively.