Removing phosphonate antiscalants from membrane concentrate solutions using granular ferric hydroxide. (October 2017)

• Record Type: Journal Article
• Authors: Chen, Yingying; Baygents, James C.; Farrell, James
• Is Part Of: Journal of water process engineering. Volume 19(2018)
• Page Start: 18
• Page End: 25
• Publication Date: 2017-10
• Subjects: Phosphonate -- Granular ferric hydroxide -- Surface complexation -- Membrane concentrate -- NTMP -- Nitrilotri(methylphosphonic) acid; Water-supply engineering -- Periodicals; Saline water conversion -- Periodicals; Seawater -- Distillation -- Periodicals; Sanitary engineering -- Periodicals; Sewage -- Purification -- Periodicals; 627
• DOI: 10.1016/j.jwpe.2017.07.002
• Languages: English
• ISSNs: 2214-7144

Abstract

Abstract: Phosphonate antiscalants are commonly used in membrane desalination to prevent fouling by mineral scale. In many circumstances, it is desirable to remove these compounds before concentrate disposal or further treatment. The goal of this research was to determine if the kinetics of phosphonate adsorption and desorption from granular ferric hydroxide (GFH) are sufficiently fast for GFH to be used in packed bed adsorption systems for antiscalant removal from membrane concentrate solutions. Well-stirred batch experiments were performed to investigate the adsorption kinetics of Permatreat 191 ® (PT191) and nitrilotri(methylphosphonic) acid (NTMP) onto GFH. Uptake of both compounds was slow and continued over the course of 6 days. Adsorption isotherms measured after 24 h elapsed showed initial concentration effects, whereby the isotherms were dependent on the initial adsorbate concentration in solution. This can be attributed to chemical adsorption reactions with faster rates of bond formation than bond breaking. Strong phosphonate adsorption in high pH solutions and high activation barriers for desorption resulted in slow kinetics for adsorbent regeneration by NaOH solutions. Desorption rates were bimodal, with 40–50% of the adsorbed phosphonate being released on a time scale of 10–24 h, while the remaining fraction was released approximately one order of magnitude more slowly. Complete regeneration could not be achieved, even after eluting the adsorbent columns with more than 300 bed volumes of 1.0 mol/L NaOH. The inability to regenerate the adsorbent in an efficient manner likely precludes its use for cost-effective antiscalant removal from membrane concentrate solutions.