ARSENIC REMOVAL FROM WATER USING COMBINED ADSORPTION-MEMBRANE FILTRATION PROCESS

Abstract

Arsenic contamination of water sources is a global issue that puts the health of millions of people worldwide at risk. WHO has recommended a permissible limit of 10 μg/L for arsenic in drinking water. Given the carcinogenicity and toxicity potential of arsenic even at low concentrations, some developed nations even strive to provide water with arsenic concentration below 1 μg/L. Extensive research is available on the remediation of arsenic-contaminated water using iron or aluminium-based adsorbents. For real field applications where the treatment of large volumes of contaminated water is concerned, a conventional standalone adsorption process is performed using fixed columns or deep filter beds containing the adsorbent media. However, these conventional columns or filter beds have certain limitations such as the issue of channeling or large pressure drop, the threat of escape of reactive nanoadsorbents into surroundings, low effluent flow rate, etc. As an alternative, combined or hybrid adsorption-membrane processes seem promising for large-scale treatment of arsenic-contaminated water. In the present study, we have investigated the use of a combined adsorption-ultrafiltration process for arsenic removal from groundwater. Ultrafiltration (UF) is a low-pressure membrane filtration process which although cannot remove arsenic owing to its higher pore size but has reasonable energy demand and can removal particulate as well as some portions of colloidal impurities, bacteria, and virus from the contaminated feed. Therefore, when used in combination with efficient adsorbents, the combined adsorption-UF process can be used for upscaling the adsorption-based water treatment. In this investigation, we have developed two forms of adsorbents. One, is in the form of an iron (oxy)hydroxide containing biopolymer beads and the other is a powdered laterite-derived adsorbent (LDA) containing mixed metal (oxy)hydroxides using simple chemical treatments. The bionanocomposite beads were prepared using in-situ precipitation of metals technique to achieve optimal distribution of adsorbent particles within the polymer support matrix. Comparative analyses of the arsenic adsorption potential of these beads prepared with two different biopolymers i.e. chitosan and alginate have been presented. The in-situ iron chitosan beads (IICB) were found to have rougher surface morphology than the in-situ iron alginate beads through the FESEM image analysis. Batch studies were carried out to assess the adsorption potential of IICB and IIAB for arsenate as well as arsenite. Maximum adsorption capacities obtained via Langmuir isotherm model fitting were in the following order: IICB for As (V) > IICB for As (III) > IIAB for As (V) > IICB for As (III). FT-IR and XPS analysis revealed ion exchange and inner surface complexation as the arsenic removal mechanism. The shrinking core model was used for the analysis of adsorption kinetics on these macroporous beads. Simulated dynamic non-dimensional liquid phase arsenic concentration profiles fit well with the experimental data (SSE = 0.018-0.025), suggesting that intra particle diffusion is the limiting step in the process of arsenic adsorption on these beads. External mass transfer coefficient (kf) and effective diffusivity coefficient (De) for both kinds of beads at the different initial arsenic concentrations (C0~ 1 mg.L-1 and 5 mg.L-1) were derived. By comparison of the diffusivity coefficient values and the simulated concentration front profiles, it was observed that chitosan provided a better porous polymer template with more easily accessible adsorption sites covering upto 85% of the volume of the bead. IICB was further used in combination with a cross-flow UF process using a commercial flat sheet polyethersulfone membrane at different TMP (0.2-0.6 MPa) for arsenic removal from water. Membrane fouling is the most important operational aspect of any membrane treatment process. The analysis of permeate flux profiles and fouling was done using combined Hermia‘s pore-blocking model for cross-flow UF and resistance in series model. For arsenic-spiked groundwater, at a dose of 2 g/L IICB, the combined UF process could bring down arsenic to permissible limits (<10 μg/L) and reduce irreversible fouling by up to 32±2 %. The reduction of free energy of adhesion (based on XDLVO theory analysis) and UV254 absorbance values indicated an overall decrease in the fouling potential of contaminated feed. Similar arsenic removal studies were also performed using the fine powdered LDA. The effect of adsorbent particles on permeate flux has been assessed at different transmembrane pressure (0.2-0.6 MPa) through the use of an intermediate sand filtration (SF) step prior to UF. Two different hybrid configurations, with and without intermediate sand filtration (SF), i.e. Ads-SF-UF and Ads-UF, were considered. In the case of arsenic-spiked groundwater, it was observed that flux decline, at 0.6 MPa, was 28% higher with Ads-UF during a 12 h run compared to Ads-SF-UF. Spent LDA retrieved from the sand column was found to retain the elemental composition as that of the unused LDA (as per FT-IR and EDX) and was considered safe for disposal based on Toxicity Characteristic Leaching Procedure (TCLP). In the end, a techno-economic assessment of the three combined UF processes studied namely IICB-UF, LDA-UF, and LDA-SF-UF have been presented. LDA-SF-UF was considered the best among the other alternatives investigated for arsenic remediation. Therefore, the cost estimation of a treatment facility producing 200 m3 of safe drinking water per day from arsenic-contaminated groundwater using the LDA-SF-UF process has been provided.