3.1. Adsorption Kinetics
The kinetics of adsorption of Mo(VI) by NZVI/AC, as illustrated in Figure 2
, comprises two steps: initial fast sorption followed later by a relatively slower adsorption event. Roughly 89.5% of Mo(VI) was eliminated from simulation water during the first 1.5 h, whereas 74.0% was removed from raw water during the first 14 h; meanwhile adsorption equilibrium was attained in ~9 h and 72 h, respectively. Thus 9 h and 72 h equilibration times were applied in further investigations carried out with SW and RW. Quite a similar phenomenon was manifested in the case of the adsorption of metal cations (e.g.,
) on NZVI/AC and the fast initial adsorption was ascribed to the metal ions transferring initially at a fast rate to adsorbent particles’ surface, whereas the slow adsorption following later was due to metal ions slowly diffusing into the pores of the intra-particle adsorbent [15
We have used many models to analyze the adsorption kinetic data, but only the intraparticle diffusion model fitted well (R2
> 0.9) in this study. In an attempt to devise the adsorption-based treatment systems, the adsorption kinetic data of liquid-phase adsorption was analyzed using many models; however, in this study, the best explanation of the experimental data is given by the intraparticle diffusion model (R2
> 0.99) (see Table 3
It has been illustrated by Weber and Morris that the rate-determining step in an adsorption system is intraparticle diffusion, and the amount of the substrate adsorbed (qt
) varies as a function of the square root of time (t0.5
) in a linear fashion. The speeds of adsorption are then calculated using these data [9
]. A linear relationship was observed in a plot of qt
in two separate stages (Figure 2
b). Therefore, to both these stages, Equation (1) was applied individually. The adsorption on the NZVI situated in the macropores corresponds to the first linear section of the graph, while the diffusion of Mo(VI) corresponds to diffusion into meso- and/or micropores. The adsorption was quick on the NZVI particles in the AC channels or macropores; however, their diffusion was rather slow into micro- and mesopores due to the blockage of most pores. Additionally, there was the involvement of the corrosion of the NZVI surface, diffusion, and adsorption in the corrosion layers. For the 1st stage, the kid
values were higher than the 2nd stage, in fact, 11.99 and 22.40 times higher, respectively, in SW and RW, suggesting a higher velocity of the stage 1 reaction compared to the stage 2, thus corroborating with Figure 2
A similar phenomenon was manifested during adsorption of acidic dye [16
], and Cu(II) and Cd(II), on activated palm ash and rice/modified rice husk [17
], on NZVI/AC. However, when using clay-impregnated nanoscale zero-valent iron for the degradation and adsorption of Cu2+
from wastewaters, the calculated qe
values and the rate constants were observed to lie close to the experimental qe
values for the pseudo-second-order (PSO) model as compared to when the pseudo-first-order (PFO) expression was taken into consideration.
The mechanism of removal of Mo(VI) by NZVI is immature. However, it was suggested by Jingge Shang that, for Cr(VI) elimination by nanoscale zerovalent iron particles supported on herb-residue biochar, the PSO model provides a better fit with the kinetic data. Hence, the kinetic parameters are regulated by a chemical process, which implies that the processes occurring during Cr(VI) removal are reduction and adsorption/coprecipitation [18
]. Considering that
are both anions (−2 valence) and have the same structure. The radius of
are 0.240 nm and 0.246 nm, respectively. We therefore speculate the similarity in the adsorption/coprecipitation mechanism of Cr(VI) and Mo(VI) in solution by nanoscale zero-valent iron.
3.2. Effect of pH
The adsorption envelopes of Mo(VI) from solutions on NZVI/AC in a pH range of 3.5–9.5 are shown in Figure 3
. Apparently, media pH strongly impacts Mo(VI) removal efficiency, and increased removal of Mo(VI) from Mo(VI) solution was observed from pH 3.5 to pH 4.5, whereas a decrease was observed from pH 4.5 to pH 9.5. The previous work also reports a decrease in Mo(VI) adsorption on nanosized zero-valent iron after pH 5, while adsorption of Mo(VI) exhibited an increase in acidic solution but a decrease in alkaline media [5
]. At pH ~5, the adsorption profiles of arsenate and arsenate onto iron oxyhydroxide supported by bead cellulose were observed to intersect [19
A strong acid will dissolve and lose iron from NZVI/AC, so we investigated the pH of the simulation and raw water from 3.5 to 9.5, while J.J. Lian studied the pH 2–10 using nano zero-valent iron supported on biochar [5
The pH dependency in the adsorption of Mo(VI) onto NZVI/AC is the impact of numerous factors that compete with one another in regulating adsorption. We suggest that there are three steps in the adsorption of Mo(VI) onto the NZVI surface: (1) migration onto the surface; (2) deprotonation or dissociation of an aqueous complex of Mo(VI); and (3) surface complexation [15
]. As a prerequisite to the adsorption reaction, Step 1 is largely controlled by electrostatic interaction (attractive or repulsive) among the surfaces of the adsorbent and aqueous Mo(VI) species. Hence, the speciation of aqueous Mo(VI) and the pH of zero-point charge (pHZPC
) of the adsorbent are the governing factors. As is already known, there exists a negative charge on solid surfaces at pH above pHZPC
, whereas a positive charge exists at pH below pHZPC,
leading to a rise in electrostatic repulsion or attraction with anionic Mo(VI) species, thereby resulting in an increased or decreased efficiency of adsorption. When the pH increased to 9.0, the percentage of adsorption decreased rapidly toward a negligible efficiency of removal by the end of the test (<30%). Other studies reported similar observations [5
3.3. Effect of Coexisting Ions
Various anions and cations are commonly present in drinking water which may positively or negatively impact the adsorption of Mo(VI). The influence of some common anions (
), cations (Fe3+
), and HA on Mo(IV) adsorption by NZVI/AC were examined in the current study (see Figure 4
). The presence of humic acid and anions adversely affected the removal of Mo(VI). Silicate, arsenate, and phosphate caused a more adverse effect; however, the HA caused a less-adverse effect on Mo(VI) removal amongst all the oxyanions investigated in this work.
Arsenate, molybdenum, phosphate, and silicate can form inner-sphere complexes with an iron oxide surface. Due to their competition for capturing the similar binding sites, they would decrease the sorption of molybdenum. Both non-specific and specific sorption is observed for sulfate ions. However, they have a much weaker strength of bonding with iron (hydr)oxide than molybdate [20
]. While earlier reports suggest that arsenic removal by zero-valent iron improved at an increased concentration of sulphate [21
], the presence of sulfate at a level relevant to the environment manifested a very little negative effect.
Divalent/multivalent metallic cations of the Fe3+
, and Al3+
notably enhanced Mo(VI) adsorption on NZVI/AC. The positive impact remarkably increased upon the increase in pH, particularly for ferrous iron Fe3+
. This is particularly useful for the treating reservoir water contaminated with Mo(VI), contamination which normally contains copious amounts of Al3+
dissolved in it. The augmenting impact of metal cationic species on Mo(IV) adsorption has been mentioned in earlier reports for iron (hydr)oxides [22
]. Presumably, this is because the metal cations present in the solution transformed the adsorbent surface to a positively charged nature, which in turn led to the adsorbent showing a higher affinity for Mo(VI) anions.
According to the discussion above, the adsorption of Mo(VI) is majorly regulated by H2
deprotonating at pH < 9. Fe3+
ions presumably form complexes with Mo(VI) in aqueous media. Consequently, the extent of deprotonation/dissociation suppresses and the net adsorption reduces. The adsorption of metal cations (e.g.,
) on NZVI/AC also manifested a similar phenomenon in our previous research [15
3.4. Fixed-Bed Column Runs
In view of the treatment efficiency, further investigation was based on examining the operational parameters like EBCT. EBCT is undoubtedly a crucial attribute to be taken into account in adsorption studies, because the efficiency of removal efficiency is strongly dependent on the duration of contact between the adsorbate (Mo(VI)) and the adsorbent (NZVI/AC). The empty-bed contact time of the influent was kept generally at 1.5–12 min when adsorption was employed as a potential technique to eliminate the pollutants from aqueous systems [23
]. Figure 5
shows breakthrough behavior during the fixed-bed column runs from deionized water supplemented with Mo(VI).
As anticipated, EBCT essentially impacted the breakthrough behavior, and the increasing EBCT resulted in an increase in adsorptive efficiency due to the increase in contact time. At 6.0 min, even when the effluent volume was as high as 1100 BV, the RW [Mo(VI)] concentration in the remaining solution remained well below 70 μg/L, whereas at an EBCT of 3.0 min breakthrough (MCL, 70 μg/L) was found to be 800 BV.
In accordance with WHO’s Mo(VI) maximum contaminant level (MCL, 70 μg/L), at a 3.0 min EBCT (empty-bed contact time), the SW and RW Mo(VI) effluent concentrations were under 70 μg/L till the empty-bed volume reached 1600 and 800 BV, respectively. The SW [Mo(VI)] and RW [Mo(VI)] concentration in the effluent solution was above 70 μg/L, at an EBCT of 6.0 min when the effluent volume reached 2300 BV and 1300 BV. The RW’s EBCTs of neither 3.0 nor 6.0 min are less than the SWs attributed to its complicated water quality composition
As the solute concentration in the effluent reaches up to 95% of the influent value, the point of column exhaustion is attained, where the solid-phase concentration reaches a maximum value. It is noteworthy that in comparison to the obtained maximum adsorption capacities, the maximal retention capacities were much lower in batch experiments, which can be attributed to the comparatively high flow rate in the column experiments, which might result in insufficient contact time between Mo(IV) and NZVI/AC. Breakthrough empty-bed volume depends primarily on NZVI/AC and influent Mo(VI) concentration. It should hence have a bigger value at longer EBCTs and lower initial Mo(VI) concentrations.
The batch experiments described above exhibit that the adsorption capacity for Mo(VI) by NZVI/AC was about 3 times compared to that of Mo(VI) at neutral pH. This shows that in batch experiments, NZVI/AC manifested a higher removal performance for Mo(VI) (neither SW nor RW) than in fixed-bed column; however, the fixed-bed column is more suitable for continuous industrial production than batch experiments.
The exhausted NZVI/AC was regenerated in 0.5 mol/L NaOH solution 5 times, and then it was eluted by deionized water a few times. The alkaline solutions have been analyzed prior to handing over to the special laboratory hazardous waste treatment center according to Chinese policy. The results displayed that more than 90% of adsorbed Mo(IV) was recovered and iron showed almost no shedding when Mo(IV)-saturated NZVI/AC was regenerated with 0.5 mol/L NaOH. The regeneration process could be interpreted as follows:
Fe-Mo(IV) + OH−⇄Fe-OH + Mo(IV) (Kdes)
3.5. Mechanism of Removing Mo(VI) from Water by NZVI/AC
The mechanism regulating the removal of molybdenum from water through the use of zero-valent iron is mainly adsorption, surface complexation, Mo(OH)3 precipitation, and Mo-Fe mineral (FeMoO4) formation.
In the current work, the removal rate of molybdenum is faster in the initial stage of the reaction, indicating that adsorption is the major process during the initial stages of reaction. The adsorptive removal of molybdenum can take place through complexation or electrostatic interaction.
The main mechanism enabling zero-valent iron to remove molybdenum from water is the combined action of adsorption and chemical precipitation.
Once zero-valent nano-iron enters the solution, it will be oxidized by water to form Fe2+. Fe2+ is affected by the pH of the solution, oxidation-reduction potential, and other factors to further form iron (hydrated) oxide and various iron (hydrated) oxides in solution. The molybdenum in the liquid tends to form compounds such as FeMoO4·xH2O, thus molybdenum in the liquid phase is transferred to the adsorbent and removed.
In this experiment, the loaded iron was found to play a major role in Mo(VI) elimination. The nano-iron will react with water and trace oxygen dissolved in water and corrode [7
]. According to the results obtained by Bruce et al. [24
] using EXAFS research, nano-iron first generates intermediate products like ferrous iron (Hydrate) oxides, and later produces iron (hydrated) oxides. The final product may include maghemite (γ-Fe2
), magnetite (Fe3
), lepidocrocite (γ-FeOOH), etc. Following the series of multiphase complex reactions on the iron surface as mentioned above, a variety of hydrated oxides having a strong adsorption capacity for Mo(IV) are finally formed (see Figure 6
The above reactions can be expressed as follows [20
Fe0 reacts with water or dissolved oxygen to form Fe2+:
Fe0 + 2H2O→Fe2+ + H2 + 2OH−
Fe0 + O2 + 2H2O→Fe2+ + 4OH−
Fe2+ is further transformed into iron (hydrated) oxides by the pH of the solution and the oxidation-reduction potential and other factors:
6Fe2+ + O2 + 6H2O→2Fe3O4(s) + 12H+
Fe2+ + 2OH−→2Fe(OH)2 (s)
6Fe(OH)2(s) + O2→2Fe3O4(s) + 6H2O
Fe3O4(s) + O2(aq) + 18H2O⇌12Fe(OH)3(s)
Fe2− + MoO42− + xH2O→FeMoO4·xH2O
Hydrated iron oxide has a large number of active sites (-OH) on the surface, so it has a strong adsorption capacity for Mo(VI).
It is generally believed that the adsorption of Mo(VI) on the surface of hydrated iron oxide takes place as a bidentate binuclear chelate [9
]. Recent studies have shown that in neutral to weakly alkaline media, this form is adsorbed on the surface of hydrated iron oxide; however, it forms an amorphous iron molybdate surface precipitate (FeMoO4
O) in a weakly acidic medium [25
]. Yong H. Huang et al. found a conventional ZVI-only system or a ZVI/Fe(II) system could not be efficiently removed molybdate from water. The hybridized ZVI/Fe3
/Fe(II) system can achieve rapid and sustainable reduction and immobilization of molybdate from water. Molybdate can be rapid and sustainable reduction and immobilization from water by the hybridized ZVI/Fe3
/Fe(II) system [26