Alatrista G, Pratt C, El Hanandeh A. Phosphate adsorption by metal organic frameworks: Insights from a systematic review, meta-analysis, and predictive modelling with artificial neural networks.
CHEMOSPHERE 2023;
339:139674. [PMID:
37517668 DOI:
10.1016/j.chemosphere.2023.139674]
[Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Subscribe] [Scholar Register] [Received: 04/30/2023] [Revised: 07/09/2023] [Accepted: 07/27/2023] [Indexed: 08/01/2023]
Abstract
This comprehensive study analysed 55 articles published between 2011 and 2022 on the use of metal organic frameworks (MOFs) for phosphate adsorption. The study found that the performance of MOFs in phosphate adsorption is influenced by various factors such as the type of MOF, synthesis method, modification/alteration, and operational conditions (initial concentration, adsorbent dose, pH, contact time, and temperature). Most of the MOFs have a wide range of theoretical maximum adsorption capacity for phosphate, but their long-term use in phosphorus recovery may be limited due to the adsorption mechanisms being dominated by inner sphere complexation. The study employed machine learning to construct artificial neural network (ANN) models for predicting phosphate adsorption capacity based on input features from operation and synthesis procedures. The initial phosphate concentration was the most important input from the operational features, while the modulator agent was consistently relevant during MOF synthesis. The models showed strong fitting for most MOF types recorded for the study, such as UIO-66, MIL-100, ZIF-8, Al-MOFs, La-MOFs, and Ce-MOFs. Overall, this study provides valuable insights for the design of MOF adsorbents for phosphate adsorption and offers guidance for future research in this area.
Collapse