In ﬂ uence of nitrite on the removal of Mn(II) using pilot-scale bio ﬁ lters

Two pilot-scale bio ﬁ lters were used to systematically investigate the in ﬂ uence of nitrite on biological Mn(II) removal. Gibbs free energy change ( Δ G) of the redox reaction between MnO 2 and NO 2 – was 122.28 kJ mol – 1 in 298 K, suggesting that MnO 2 could not react with NO 2 – . When nitrite in the in ﬂ uent was increased from 0.05 to 0.5 mg L – 1 , manganese oxides did not react with nitrite in anaerobic conditions; nitrite was quickly oxidized and biological Mn(II) removal was slightly affected in 2 h in aerobic conditions. When nitrite was accumulated in the bio ﬁ lter by increasing ammonia concentration, nitrite existed for more than 3 d and biological Mn(II) removal was affected in 3 d. When Mn(II) and ammonia in the in ﬂ uent were about 2 and 1.5 mg L – 1 , respectively, both of them were completely removed and the oxidation-reduction potential was increased with the depth of the ﬁ lter from 16 to 122 mV. Biological Mn(II) removal followed the ﬁ rst-order reaction, and the k -value was 0.687 min – 1 .


INTRODUCTION
Groundwater is often mildly acidic and devoid of dissolved oxygen (DO) (Azher et al. ), so when groundwater flows through soils, minerals and rocks, soluble Fe(II) and Mn (II) are present ( Jusoh et al. ), either in dissolved mineral form, or associated with various organics, minerals or chelating agents. In addition, the predominant form of Mn(II) at low or neutral pH values is Mn 2þ , which occurs primarily as a free cation in natural waters (Nealson et al. ).

Continuously increasing ammonia concentration in ground-
water has been observed in the past years, owing to the discharge of waste from both industry and bank-side residents without adequate pre-treatment and sub-optimal in water distribution systems, Fe(II) and Mn(II) are substrates for bacteria growth (Azher et al. ), hence when the bacteria die and slough off, bad odors and unpleasant tastes may be produced (Kontari ; Gouzinis et al. ). In addition, when Mn(II) exceeds the permitted limit, Mn(II) has been found to affect the central nervous system (Sharma et al. ). The presence of ammonia in drinking water treatment could affect the chlorination process and Mn(II) biofiltration system (Hasan et al. ).
Ammonia will react with chlorine to form disinfection by-products (Richardson & Postigo ), which could damage the human nervous system (Nieuwenhuijsen et al. ), cause a deterioration in the taste and odor of water (Richardson et al. ), and reduce the disinfection efficiency (WHO ). Furthermore, ammonia can interfere with the Mn(II) biofiltration process by consuming excessive oxygen during nitrification, resulting in moldy and earthy tasting water (WHO ). So groundwater which contains high concentrations of Fe(II), Mn(II) and ammonia needs to be treated before it used for industry and humans. Chemical methods could be used to oxidize Fe(II), Mn(II) and/or ammonia, but chemical oxidation may produce potential hazardous by-products and/or introduce other pollutants into the produced water. Moreover, it is difficult to simultaneously remove ammonia and manganese with only one chemical (Han et al. ). The advantages of the biological oxidation process over the conventional physicochemical methods include high filtration rates, and low operation and maintenance costs. As single stage filtration is used, it is not necessary to provide additional chemicals, and the volume of the generated sludge is appreciably smaller and easier to handle (Pacini et al. ; Han et al. ). In this study, the ΔG of the redox reaction between MnO 2 and NO 2 was calculated. Two pilot-scale biofilters were established to investigate the redox reaction between MnO 2 and HNO 2 in anaerobic conditions, and the influence of added and generated nitrite on biological Mn(II) removal in a biofilter. The main objectives of this study were to verify whether the reaction between MnO 2 and NO 2 could occur, and find the reason why the start-up period of biofilters for Mn(II) and ammonia removal was much longer than the biofilters for Mn(II) removal.   The influence of added and generated nitrite on Mn(II)

removal in a biofilter
In order to investigate the influence of added nitrite on Mn(II) removal using a biofilter, the experiment was operated as in the previous section except that the DO in the influent of filter 2 was above 10 mg L -1 . In order to investigate the influence of generated nitrite on Mn(II) removal using a biofilter, the concentration of ammonia in influent of filter 1 was increased from approximately 1.1 to 1.5 mg L -1 by adding the stock solution of ammonia in tank 5 (DO in the influent of filter 1 was approximately 11 mg L -1 ).

The concentration profiles of ORP in simultaneous
Mn(II) and ammonia removal The aerated raw groundwater in tank 1 was pumped into filter 1, and the concentration of total iron, Mn(II) and ammonia in effluent water of filter 1 was lower than 0.1, 0.05 and 0.1 mg L -1 , respectively. The effluent water of filter 1 in tank 2 was aerated and DO was increased to approximately 11 mg L -1 , then the effluent water in tank 2 and the stock solutions of Mn(II) and ammonia in tanks 3 and 5, respectively, were pumped to filter 2. The concentration of Mn(II) and ammonia in the influent of filter 2 was approximately 2 and 1.5 mg L -1 , respectively.

Kinetics of biological Mn(II) oxidation
The effluent water in tank 2 (as in the previous section) and the stock solution of Mn(II) in tank 3 were pumped into

RESULTS AND DISCUSSION
The redox reaction between nitrite and Mn(IV) An equation of Gibbs free energy change: where ΔG is Gibbs free energy change (kJ mol -1 ), ΔH is free enthalpy change (kJ mol -1 ), T is thermodynamic temperature (K), and ΔS is entropy change (kJ mol -1 K -1 ).
When the thermodynamic temperature was 298 (25 W C) and 281 K (8 W C), the ΔG was 122.28 and 120.92 kJ mol -1 in 100 kPa, respectively, which suggested that MnO 2 cannot react with NO 2 under these conditions. When nitrite was generated in the filter, Mn(II) removal was affected in 3 days; however, when nitrite was added to the filter, even the concentration of nitrite was much higher, Mn(II) removal was affected in only 2 h. The reason was because the added nitrite was quickly   removal, and found that ORP increased along the filter depth from 150 to 600 mV, depending on the feeding concentrations. In their investigation, ORP was much higher than in filter 2, because DO in their filter was 7-8 mg L -1 ; however, DO in the effluent of filter 2 was lower than 1 mg L -1 .

Kinetics of biological manganese oxidation
The removal kinetics of contaminants during water treatment is considered an important issue, because it can provide information about the required time that the specific contaminant needs to be removed efficiently, which is necessary in sizing treatment units (  ammonia removal being much longer than biofilters for Mn(II) removal. When Mn(II) and ammonia in influent were 2 and 1.5 mg L -1 , respectively, ORP increased along the filter depth from 16 to 122 mV. Biological Mn(II) removal followed the first-order reaction, the k-value was 0.687 min -1 and the halflife time for the depletion of Mn(II) was 1.010 min.