The experimental design approach to eluotropic strength of 20 solvents in thin-layer chromatography on silica gel

• Published in: Journal of Chromatography A Vol. 1483; pp. 138 - 141
• Main Authors: Komsta, Łukasz, Stępkowska, Barbara, Skibiński, Robert
• Format: Journal Article
• Language: English
• Published: Netherlands Elsevier B.V 03.02.2017
• Subjects: Chemical Phenomena; Chromatography, Thin Layer - methods; confidence interval; Confidence Intervals; Eluotropic strength; experimental design; Hexanes - chemistry; Linear Models; Mixture design; principal component analysis; regression analysis; Research Design; screening; Silica gel; Silica Gel - chemistry; solvents; Solvents - chemistry; Thin-layer chromatography; uncertainty; variance; Thin-layer chromatography; Silica gel; Mixture design; Eluotropic strength
• ISSN: 0021-9673; 1873-3778
• DOI: 10.1016/j.chroma.2017.01.010

•Experimental indices of eluotropic strength of 20 solvents were obtained.•The proper experimental design allowed to estimate each solvent as independent factor.•The uncertainty of estimation was studied by additional bootstrapping.•The strength measures have large cognitive value as theoretical RM difference.•A general agreement between the results and previously published series was achieved. The eluotropic strength on thin-layer silica plates was investigated for 20 chromatographic grade solvents available in current market. 35 model compounds were used as test subjects in the investigation. The use of modern mixture screening design allowed to estimate each solvent as a separate elution coefficient with an acceptable error of estimation (0.0913 of RM value). Additional bootstrapping technique was used to check the distribution and uncertainty of eluotropic estimates, proving very similar confidence intervals to linear regression. Principal component analysis proved that the only one parameter (mean eluotropic strength) is satisfactory to describe the solvent property, as it explains almost 90% of variance of retention. The obtained eluotropic data can be good appendix to earlier published results and their values can be interpreted in context of RM differences.