Introduction

Coal mine water in Quang Ninh province is acidic and carries high levels of iron and manganese. Lime neutralisation and settling remove most of the iron. Mn ²⁺ , however, oxidises slowly at near-neutral pH and passes through the clarifier [1]. The local regulation QCĐP 3:2020/QN limits Mn in discharged wastewater to 0.5 mg/L (column A) and 1.0 mg/L (column B) [2].

Zeolites retain Mn ²⁺ by cation exchange, and the sodium form takes up more than the natural form [3]. A manganese oxide coating on the grains adds a second, oxidative route [4]. Whether zeolite is economical in this service depends on how many times it can be regenerated. Salt regeneration has been reported for clinoptilolite loaded with cadmium [5] or with lead and zinc [6], and for activated natural zeolite used to remove manganese [7]. Zeolites made from rice husk ash have mostly been tested as single-use powders. In one study, NaX from rice husk ash lost about a quarter of its methylene blue removal after four reuses [8]. We are not aware of regeneration data for such granules in real mine water.

In this work the granules were tested in a fixed-bed column with real mine water, the regeneration time in 0.5 M NaCl was determined, and Mn removal was followed over five use cycles.

Materials and methods

Zeolite granules. NaX-type zeolite was synthesised hydrothermally from rice husk ash and metakaolin and bound with sodium silicate (15 wt %). The granules were sieved to 2–5 mm, rinsed and dried at 105 ± 2 °C.

Mine water and analysis. Water was taken at the +25 Nui Nhen station (Quang Ninh) after neutralisation, coagulation and settling, just upstream of the manganese filter. It contained 8.20 mg/L Mn, and no Mn was added. Mn was determined on a Hach DR3900 spectrophotometer by the PAN method (Method 8149) [9]; samples above 0.7 mg/L were diluted.

Experiment 1 (column). Three columns were run in parallel under the conditions in Table 1, and the effluent was sampled every 60 min.

Table 1

Operating conditions of the fixed-bed columns

Experiment 2 (regeneration time). Granules taken from the columns after the run were shaken with 0.5 M NaCl (1 g: 10 mL, 150 rpm, 25 ± 2 °C) for 30, 60, 120 and 180 min, and Mn was measured in the brine.

Experiment 3 (reuse). Granules (4 g) were shaken with 200 mL of mine water (20 g/L, pH 8, 45 min, 150 rpm). After each cycle they were regenerated with 40 mL of 0.5 M NaCl for 120 min, rinsed with deionised water and used again without adding fresh material. The series comprised five cycles and four regenerations.

Removal efficiency H , the number of bed volumes treated and the manganese retained in the column between 60 and 300 min were calculated as:

H = (1 − C _t / C ₀ ) × 100 %, (1)

BV = Q · t / V _bed , (2)

q = ( Q / m ) ∫ ( C ₀ − C _t ) d t , (3)

where C ₀ and C _t are influent and effluent Mn concentrations, mg/L; t is time, min; m is zeolite mass, g. The integral in Eq. (3) was evaluated by the trapezoidal rule.

Results and discussion

Column behaviour. Mean removal rose from 18.21 ± 1.62 % at 5.65 BV to 56.10 ± 1.22 % at 28.25 BV (Fig. 1, n = 3). The SD of C _t between the three replicate columns did not exceed 0.30 mg/L.

Fig. 1. Effluent ratio C _t /C ₀ and Mn removal in the zeolite column (mean ± SD, n = 3)

The bed removed more manganese as it was loaded. A manganese oxide coating that builds up on the grains and oxidises incoming Mn ²⁺ would produce this pattern [4, 7], although the sand support layer may also have contributed. Between 60 and 300 min the bed retained 0.75 mg Mn/g. Longer runs with activated natural zeolite gave 2.38 to 5.13 mg/g [7], so our value covers only the start-up period. At an EBCT of 10.6 min the effluent still contained 3.60 mg/L Mn, well above the limit. A full-scale bed would need a longer contact time or a conditioning period before discharge.

Regeneration time. Mn in the brine reached 0.880 mg/L at 120 min and changed little by 180 min (Fig. 2, a), so 120 min was adopted. At 10 mL of brine per gram, this equals about 9 µg Mn/g, or roughly 1 % of the Mn retained in the column. The salt solution therefore returns the exchange sites to the sodium form but removes little of the manganese, and most of the retained Mn is probably present as oxide rather than as exchangeable ions. The rinse water was not analysed, so the Mn mass balance is still open.

Fig. 2. Mn released into 0.5 M NaCl versus regeneration time (a); Mn removal and residual Mn over five use cycles (b)

Reuse over five cycles. Removal fell from 97.32 % in cycle 1 to 95.00 % in cycle 5 (Fig. 2, b), with a mean of 96.0 %. Cycle 5 thus kept 97.62 % of the initial removal, a loss of about 0.6 % per cycle. Residual Mn ranged from 0.22 to 0.41 mg/L and met the 0.5 mg/L limit in every cycle. Using the same granules five times reduces the need for fresh sorbent by 80 %.

Clinoptilolite regenerated with 1.0 M NaCl also performed steadily over five cycles [5]; here half that salt concentration was enough. NaX from rice husk ash used for methylene blue kept only 76 % of its removal after four reuses [8]. A likely reason for the difference is that NaCl restores the exchange sites but leaves the oxide coating in place, so both retention routes work again in the next cycle. Two limits apply. The batch tests ran at pH 8, which favours oxidation, and at a low loading of 0.39 mg/g per cycle. Also, 0.5 M NaCl supplied about 5 meq Na ⁺ /g, whereas the retained Mn required only 0.014 meq/g. Performance at saturation and the lowest effective salt dose still have to be determined.

Conclusion

Rice husk ash zeolite granules were regenerated with 0.5 M NaCl in 120 min. After five cycles they kept 97.6 % of their initial Mn removal, and the treated water met the 0.5 mg/L limit in each cycle. Column removal increased with loading, and the brine recovered only about 1 % of the retained Mn. Both results point to surface oxides as the main retention phase. The next steps are column runs to saturation, a closed Mn balance and tests at lower salt doses.

The study was funded by the Vietnam National Coal-Mineral Industries Holding Corporation Limited (Vinacomin) under project No. KC.05.Đ32–25/21–25.

References:

1. Li Y., Xu Z., Ma H., Hursthouse A. S. Removal of manganese(II) from acid mine wastewater: a review of the challenges and opportunities with special emphasis on Mn-oxidizing bacteria and microalgae // Water. 2019. Vol. 11, No. 12. Art. 2493. DOI: 10.3390/w11122493.
2. QCĐP 3:2020/QN. Local technical regulation on industrial wastewater in Quang Ninh province. Quang Ninh Provincial People's Committee, 2020.
3. Taffarel S. R., Rubio J. On the removal of Mn ²⁺ ions by adsorption onto natural and activated Chilean zeolites // Minerals Engineering. 2009. Vol. 22, No. 4. P. 336–343. DOI: 10.1016/j.mineng.2008.09.007.
4. Taffarel S. R., Rubio J. Removal of Mn ²⁺ from aqueous solution by manganese oxide coated zeolite // Minerals Engineering. 2010. Vol. 23, No. 14. P. 1131–1138.
5. Gedik K., Imamoglu I. Removal of cadmium from aqueous solutions using clinoptilolite: influence of pretreatment and regeneration // Journal of Hazardous Materials. 2008. Vol. 155. P. 385–392.
6. Katsou E., Malamis S., Tzanoudaki M., Haralambous K. J., Loizidou M. Regeneration of natural zeolite polluted by lead and zinc in wastewater treatment systems // Journal of Hazardous Materials. 2011. Vol. 189, No. 3. P. 773–786. DOI: 10.1016/j.jhazmat.2010.12.061.
7. Sanzana S. et al. Enhancing manganese sorption: batch and fixed-bed column studies on activated zeolite // Environmental Technology & Innovation. 2024. Vol. 33. Art. 103495. DOI: 10.1016/j.eti.2023.103495.
8. Tran Nguyen Phuong Lan, Ly Kim Phung, Luong Huynh Vu Thanh, Nguyen Hong Nam, Tran Thi Bich Quyen, Le Phan Hung. Methylene blue treatment by NaX zeolite with silica derived from rice husk ash // Vietnam Journal of Science and Technology, Series B. 2022. Vol. 64, No. 6. P. 14–18. DOI: 10.31276/VJST.64(6).14–18. (In Vietnamese)
9. Hach Company. Manganese, Low Range: PAN Method 8149. Loveland, CO: Hach, 2019.