Commentary on "Trace cisplatin adsorption by thiol-functionalized sponge (TFS) and Sn/SnO₂-coated TFS: Adsorption...

Cisplatin is an environmentally relevant cytostatic contaminant, and the thiol-functionalized sponge materials reported by Han et al. represent a valuable contribution to the development of selective adsorbents for trace platinum-based pharmaceuticals. The present Commentary does not question the synthetic merit of the materials, their adsorption performance, or the surface-sensitive evidence supporting local Pt-S coordination. The scientific concern is narrower and methodological. It concerns whether the reported thermodynamic parameters and mechanistic terminology are justified by the equations, equilibrium constants, temperature dataset, and adsorption-energy descriptors used in the published article. In this revised analysis, the Dubinin-Radushkevich regression itself is not treated as a computational error. Instead, the reported mean adsorption energies, 8.17-13.98 kJ mol⁻¹, are accepted as apparent D-R descriptors and are reinterpreted within a physically consistent framework. These values are not intrinsic Pt-S bond energies, and they cannot establish that the overall hydrated solid-liquid adsorption process is dominated by covalent chemisorption. Local Pt-S coordination may occur, but it is embedded in a broader energy balance involving aquation, ligand exchange, desolvation, hydration-shell reorganization, electrostatic attraction, hydrogen bonding, and surface heterogeneity. A second and more consequential problem is the use of K_c = (C₀ - Cₑ)/Cₑ, or related distribution ratios, as if they were standard thermodynamic equilibrium constants. Although K_c is algebraically unit-free, it is not a standard-state-normalized equilibrium constant because it depends on the selected initial concentration, adsorbent dosage, phase ratio, and operating point of the isotherm. This distinction is demonstrated mathematically through the Taylor expansion of the logarithm, through dimensional analysis of transcendental functions, and empirically through a full isotherm table showing that K_c and K_d change at every initial concentration. Finally, the use of only three temperatures in a van't Hoff regression leaves one residual degree of freedom, preventing robust uncertainty estimation and curvature detection. The corrected interpretation is that the published thermodynamic conclusions require substantial qualification before they can be used for mechanistic classification, process modeling, or scale-up design.