Natural and synthetic zeolites
Zeolite-N (synthetic zeolite, NanoChem Pty Ltd, Australia) was used as reference material and compared to clinoptilolite (natural zeolite, St. Cloud Mining, New Mexico), and Zeolite1-6 (synthetic zeolites, BYK Additives & Instruments Ltd, Germany). The synthetic zeolites were produced according to Mackinnon, Millar and Stolz16, granulated according to procedure described in Table 1. Clinoptilolite had a Si/Al = 4.0 (natural zeolites were reported to have a Si/Al = 3–511,32,33) while lower values of Si/Al (1.5–2.0) were measured for the synthetic zeolites (Table 1).
The zeolites were initially washed with deionized water to remove any dust from their surface and sieved to obtain the required size (1–2 mm for the synthetic zeolites and 2.5–3.5 mm for the clinoptilolite) before further tests.
Municipal wastewater characterization
Municipal wastewater was obtained from Cranfield University wastewater treatment plant in the UK (2,840 population equivalent) after the secondary treatment with trickling filters that removed organic carbon. The wastewater was filtered (filter pore diameter 1.2 µm) to prevent any residual solids to interfere with the adsorption tests34. The composition of the wastewater was: chemical oxygen demand (COD) was 37.0 ± 12.3 mg/L, ammonium (NH4+-N) was 13.6 ± 4.6 mg was NH4-N/L, orthophosphate (PO4-P) was 6.0 ± 0.25 mg PO4-P/L, calcium (Ca2+) was 25.1 ± 1.2 mg Ca2+/L, potassium (K+) was 25.4 ± 2.2 mg K+/L and the pH was 7.3 ± 0.4.
Ammonium exchange capacity (AEC) of zeolites
The maximum ammonium exchange capacity (AEC) of the zeolites was calculated in mono-component solution with an initial concentration of 1,000 mg NH4+-N/L (55.6 meq NH4+-N/L). Successively, the operational AEC in municipal wastewater was calculated starting from an initial concentration of 12.7 mg NH4+-N/L (0.7 meq NH4+-N/L).
The solutions (100 ml) were mixed with 0.5 g of media at 150 rpm for 8 h using the orbital shaker SSL1 (STUART, UK), after which, the remaining ammonium was measured. Experiments were conducted in triplicate. The AEC was calculated according to Eq. (2)35:
where AEC is the ammonium exchange capacity (meq NH4+-N/g media), Ci and Cf are the initial and final ammonium concentration in solutions (meq NH4+-N/L), Vtreated is the volume of the solution (L) and M is the mass of media (g).
To investigate the reusability of the media over multiple cycles, the zeolites were pre-treated with a fresh solution of potassium chloride (KCl) 10% w/v (the regenerant) for a period of 2 h to remove any residual ammonium from the fresh media surface. Successively, 100 mL of municipal wastewater were mixed with 1 g of the media (i.e. 10 g/L), in duplicate, mixed by agitation at 150 rpm for a period of 8 h using the orbital shaker SSL1 (STUART, UK). The average initial concentration was 12.1 ± 0.3 mg NH4+-N/L (0.7 meq NH4+-N/L). At the end of each cycle, the media were regenerated with a fresh regenerant solution for a period of 2 h for a total of 10 cycles. The ammonium exchange capacity (AEC) in wastewater was calculated as in Eq. (2) while the regeneration capacity (Qreg) was calculated as in Eq. (3) (adapted from You et al.36):
where, Qreg is the capacity of regeneration (meq NH4+-N/g media); Cr.i and Cr.f are the concentration of ammonium in the regenerant at the beginning and at the end of each cycle (meq NH4+-N/g media); Vr.treated is the volume of regenerant used (L); M is the same as in Eq. (2).
Attrition and resistance to compression tests
For the attrition tests, 3 g of fresh Zeolite were mixed at 200 rpm (in accordance to literature37,38 with 300 mL of deionized water for a period of 24 h using the orbital shaker SSL1 (STUART, UK). Samples were taken at regular intervals (each 15 min during the first hour of treatment and each 2 h from 2 to 24 h of treatment), after 1 min of settling.
The media resistance to compression was measured using the system Instron 5,965 (Instron, UK). Thirty beads of each zeolite were singularly positioned in between the compressing disks of the system and subjected to an increasing load (measured in Newton, N) until breakage. The force applied at breaking point was registered with an accuracy of 0.5%. Measurements were performed on fresh media.
Ion exchange demonstration scale plant
Zeolite-N and Zeolite6 were separately tested for ammonium removal in an ion exchange demonstration scale plant (IEX-D) (Table 2) treating 10 m3/day of municipal wastewater, with average ammonium concentrations of 13.6 ± 4.6 mg NH4-N/L (0.8 ± 0.3 meq NH4-N/L). Firstly, a column with an internal diameter of 307 mm and height of 1566 mm was filled with 69L (78 kg) of Zeolite-N. The media was first backwashed for 30 min to remove any fine particle left over from manufacturing. The wastewater was fed in down-flow operation at an empty bed contact time (EBCT) of 10 min. After media saturation, regeneration was completed using 10 bed volumes of 10% potassium chloride (KCl) that were passed through the column in up-flow operation for 2 h. At the end of the regeneration, the column was drained and the KCl was collected and stored for the next regeneration. The media was backwashed with tap water for 30 min (flow rate 500 L/h) at the end of each cycle to remove any residual solids.
Successively, a column with internal diameter of 217 mm and height 1577 mm was filled with 35L (24 kg) of Zeolite6 (Table 2). After an initial backwash, the wastewater was fed at an EBCT of 5 min. The same operation conditions as for Zeolite-N were used.
For both media, the ammonium was measured at the inlet and outlet of the columns and in the regenerant. The AEC and Qreg were obtained as in Eqs. (2) and (3), respectively. A sample of both media was analysed at the optical microscope (Optech Microscopes Ltd) for the fresh media and for the media taken from the IEX-D columns after 7 and 3 cycles for Zeolite-N and Zeolite6, respectively and resistance to attrition test were completed (“Attrition and resistance to compression tests” section). The experiment was conducted in triplicate.
Physico-chemical and statistical analysis
Turbidity was measured using the 2100 N Turbidimeter (HACH, UK) in accordance with the EPA 180.1 method39. For the characterization of the wastewater, chemical oxygen demand, calcium and potassium were analysed using Spectroquant cell tests; ammonium and phosphorus were analysed using the Smartchem200 (AMS Alliance, France). The analysis of the pH was performed using a pH meter (Jenway 3,510 pH and conductivity meter, Camlab, UK).
Statistical analyses were performed using the JMP software (SAS Institute) to identify statistical difference between the capacity of adsorption considering the solution treated (municipal wastewater and synthetic solution) and the media (natural and synthetic zeolite). The JMP tool was also used to detect any statistical difference in the ion exchange capacities of the media comparing the first and tenth cycle of batch test.