Liquid-phase adsorption: Common problems and how we could do better

Adsorption plays a vital role in many applications from adsorbents for concentrating valuable compounds or removing pollutants to catalysts. In gas and liquid phases, the adsorption phenomena may look similar and the results are often transferred. But solvents play a role and may change the adsorption behaviour even for strong adsorbates – liquid phase adsorption is different! The review covers kinetics and thermodynamics of adsorption processes and focuses on several areas that receive only minor attention despite being crucial for obtaining reliable results. Such underappreciated areas include the analysis of how to maximise experimental accuracy of adsorption studies and analyse the model parameters and their confidence intervals; the effect of the mathematical representation and model linearization on the results; the possibility of processes other than adsorption during the experiments. The experiments based on disappearance of the adsorbate from the equilibrium solution shall be performed to ensure at least 10% decrease in concentration for reasonable accuracy. Regression analysis and analysis of the confidence intervals of the parameters merit particular attention as well as an independent validation of the model assumptions. Even an excellent data fit may provide results differing by several times from the correct values. Adding to the dispute on dimensions in the adsorption constants in van't Hoof equation, the review adds arguments in favour of using constants in L/mol units. The review concludes with the proposed workflow in the analysis of liquid-phase adsorption data from the data acquisition to data analysis and modelling and offers a Matlab app for Langmuir adsorption data analysis. •Linearization of the (adsorption) model affects the model parameters obtained.•Narrow confidence intervals in parameters are required, not sufficient, for valid model.•A workflow for experiment and analysis is proposed for adsorption studies.