Despite an expansive selection of available thin-layer chromatography (TLC) stationary phases, almost none are well-suited for separation of polar, ionic analytes. This work demonstrated that a highly porous stationary phase layer comprised solely of nanostructured biosilica frustules, isolated from living Pinnularia sp. diatom microalgae, improved TLC separation of the polar, ionic analytes Malachite Green (MG) and Fast Green FCF (FG) relative to silica gel. Intact biosilica frustules were isolated from Pinnularia sp. cell culture via oxidation with acidified hydrogen peroxide. Diatom biosilica TLC layers were fabricated using a facile, binder-free, drop-cast technique. FT-IR spectroscopy was employed to compare surface chemistries of biosilica vs. commercial silica gel TLC layers. Pore structure parameters of the stationary phases were characterized via SEM imaging as well as experimental measurement and analytical modeling of capillary flow through the films. TLC separations of MG and FG were performed on both stationary phases using two different mobile phase mixtures (9:1:1 v/v 1-butanol:ethanol:water and 5:1:2 v/v 1-butanol:acetic acid:water). Plate height versus solvent front migration distance relationships were measured and mathematically modeled for the reference analytes MG and FG on both stationary phases to characterize TLC separation efficiencies. Although both stationary phases were composed of amorphous silica rich in silanol groups with particle size of 10–12 µm, diatom biosilica frustules were highly porous, hollow shells with surface structure dominated by 200 nm pore arrays. Diatom biosilica significantly improved the mobility of both MG and FG, enabling the resolution of these analytes. The diatom biosilica layer had a high void fraction of 96% but reduced the flow velocity and permeability constant by a factor of two relative to silica gel. TLC performance was enhanced, as evidenced by ten-fold reduction in theoretical plate height for both analytes using the 1-butanol:acetic acid:water mobile phase, and an increased difference in retardation factor between MG and FG using the 1-butanol:ethanol:water mobile phase. Analysis of plate height vs. solvent front position by the modified van Deemter equation suggested that dispersive mass transfer was reduced, leading to improved analyte resolution, and that the pore-perforated surface topology of the frustule decreased boundary layer resistance, leading to increased analyte flux. Overall, the basis for improved chromatographic performance is believed to be the unique micro- and nanostructure of the diatom biosilica frustule.