Efﬁcient Removal of Azo Dye from Wastewater Using the Non-Toxic Potassium Ferrate Oxidation–Coagulation Process

: The wastewater efﬂuents from textile industries contain highly toxic metal complex dyes. For instance, azo dye has received signiﬁcant attention owing to its toxicity and environmental stability. This study investigated the oxidation and coagulation processes to effectively remove azo dye from wastewater efﬂuents. Potassium ferrate (K 2 FeO 4 ) was selected as an oxidant because it has a high oxidation potential, is environmentally stable, and does not generate toxic byproducts. Moreover, it has a combination effect of coagulation and oxidation. Its performance was compared with a single oxidation process (using NaOCl) and a single coagulation process (using FeCl 3 · 6H 2 O). Based on the jar test experiment, the optimized pH was estimated to be 3 and the optimal dosage was 56.4 mg/L for K 2 FeO 4 , and it removed nearly 100% of orange II azo dye (OD) and lissamine green B dye (LGB). However, its removal efﬁciency decreased when the pH increased to 12. In all processes, dye removal was completed in 5 min of the reaction. Overall, OD and LGB were effectively removed by K 2 FeO 4 , compared to the NaOCl and FeCl 3 · 6H 2 O. This indicates that the combination of oxidation and coagulation of K 2 FeO 4 outperformed the single treatment process without toxic byproduct production.


Introduction
The textile industry produces a significant amount of dye effluent, which contains highly toxic metal complex dyes [1]. As materials can absorb only 20-80% of the dye from dye mixtures, depending on their limited absorption capacity, approximately 10-15% of the dyes are released into effluents [2,3]. With the increasing use of synthetic chemical dyes over the past few years, a considerable amount of dye-containing industrial wastewater effluent is discharged into aquatic and soil environments [4,5]. Azo dyes, containing one or more azo groups (-N=N-) in their chemical structure, account for 50-70% of the synthetic dyes that are used in textiles, papers, food, cosmetics, and pharmaceuticals [6,7]. Given their abundant structural diversity, high molar extinction coefficient, sensitivity to light, and wetness, azo dyes are toxic and mutagenic; moreover, they lead to the discoloration of natural water.
Several approaches have been used for the removal of azo dyes from industrial wastewater. A combination of aerobic and anaerobic processes is carried out on dye effluent because it is cheap and easy to process; however, it is insufficient to remove hazardous particles from dye wastewater [8] Biological processes, including adsorption by microbial biomass and microbial degradation [9] are not effective for all dyes because of the stability, high water solubility, and high molecular weight of azo compounds [10]; furthermore, they tend to generate toxic byproducts [11,12]. In particular, azo dyes cannot be effectively treated by conventional biological processes because they are chemically stable [13]. Adsorption exhibits excellent removal efficiency, but the cost of adsorbents is significant.
Coagulation-flocculation is widely used, but it leads to the generation of sulfur and vat dye effluents. In addition, it produces a considerable amount of sludge. Ion exchange shows good dye removal efficiency, but is limited to certain dyes. Membrane filtration is applied for water recovery; however, the membrane is easily fouled and the initial investment is high [14]. Chemical oxidation, including ozonation, is also applied for chemical dye removal, but it produces toxic byproducts and has a short half-life [15][16][17]. Wastewater containing azo dyes reduces the amount of light that can penetrate aquatic environments. Moreover, it produces different amines under anaerobic conditions, resulting in serious environmental problems. A recent study reported that H 2 O 2 assisted photoelectrocatalysis for the decolorization and biodegradation of pharmaceutical wastewater [18].
Potassium ferrate (K 2 FeO 4 ), which is an iron (VI) derivative, is a powerful oxidant over a wide pH range; its standard reduction potential varies from +2.2 V to +0.7 V in acidic and basic conditions, respectively [19][20][21]. As compared to chlorine, chlorine dioxide, hydrogen peroxide, and ozone, K 2 FeO 4 works as a strong oxidant for water and wastewater treatment in acidic solutions. The ferrate ion (FeO 4 2− ) reduces rapidly and exothermally to Fe (III) and oxygen under strongly acidic conditions. Meanwhile, in strong alkali solutions, it tends to be stable and has a low reduction rate at pH 9.4-9.7. Its decomposition rate dramatically decreases as the pH increases from 7.1 to 11.9 [22]. In aqueous solution, the ferrate ion (FeO 4 2− ) is a monomer with a tetrahedral structure consisting of a high degree of four covalent character equivalent Fe-O bonds [23]. It is widely used in water treatment because it is stable for up to one year and is easy to prepare. Furthermore, it is stable in organic solvents, but highly soluble in water. Compared to other oxidants, K 2 FeO 4 generates non-toxic products with relatively low amounts of disinfection byproducts (DBPs) [21]. Moreover, during aqueous oxidation reactions, ferrate (FeO 4 2− ) is reduced to Fe (III) ion (Fe 3+ ) and hydrolyzed to form insoluble iron (III) hydroxide (Fe(OH) 2 ). This works as a conventional coagulant, depending on the pH and dosage. Therefore, K 2 FeO 4 can function as an oxidant and a coagulant in a single-step treatment. This paper hypothesized that K 2 FeO 4 can effectively remove different contaminants by multiple treatment mechanisms, namely, the oxidation and coagulation of azo dye, thus proving to be a new, reliable, and low-cost water reuse technology. It could also be ideal for small communities, given its compactly designed operational unit, economical capital cost, and less-demanding management than a conventional two-step unit process.
This study aimed to investigate the treatment efficiency of K 2 FeO 4 for dye wastewater treatment with an emphasis on removing orange II azo dye (OD) and lissamine green B dye (LGB). The treatment efficiency of K 2 FeO 4 in removing OD and LGB was evaluated by comparing its efficiency with that of the conventional NaOCl oxidant and FeCl 3 ·6H 2 O coagulant.

Experimental Procedure
The performance of K 2 FeO 4 as an oxidant and coagulant was evaluated using the jar test. It was used to simulate a full-scale coagulation-oxidation process and determine the optimum conditions for wastewater treatment. The variables of the experiment were set to K 2 FeO 4 dose, dye concentration, contact time, and pH. The jar tester had six paddles that stirred the contents of six 1 L containers.
To start, an equal volume of water sample was added to each of the six beakers. The first sample was used as a control, while the other five samples were adjusted by changing the pH or coagulant/oxidant dose to determine the optimum operating conditions. The pH was adjusted by adding predetermined volumes of 0.1 N NaOH or 0.1 N HCl solutions to the beakers, prior to adding any oxidant and coagulant. The pH was not adjusted after the coagulant/oxidant because the coagulant/oxidant could start hydrolyzing or even precipitating. After the coagulant and/or oxidant were added to each beaker, rapid mixing was conducted at 120 rpm for 1 min. Then, slow mixing was conducted for 30 min to provide the efficient contact. Next, agitation was stopped to allow the flocs to settle. After the jar testing was completed, an adequate sample volume of supernatant liquor was collected from each jar. Once collected, all samples were filtered through a 0.45 µm filter for dye concentration measurement. Each experiment was conducted in triplicates.

Performance Evaluation of Dye Removal
For pH optimization, 50 mg/L of the dye solutions were reacted with 200 mg/L K 2 FeO 4 in a jar tester at a pH range of 3-12. The dye removal efficiency was examined as a function of the K 2 FeO 4 dose (25, 50, and 100 mg/L as Fe). The dye concentration was varied to 25, 50, and 100 mg/L.
The removal efficiency of K 2 FeO 4 was then compared with that of NaOCl (oxidant) and FeCl 3 ·6H 2 O (coagulant). The concentration of NaOCl and FeCl 3 ·6H 2 O increased from 7.05 to 84.6 mg/L. In this test, the dye concentrations were 25 and 50 mg/L.
The removal efficiency was evaluated using ultraviolet-visible (UV-Vis) spectrometry, and the concentrations of OD and LGB were determined at wavelengths of 483 nm and 633 nm, respectively, using a UV-Vis spectrophotometer (HACH, DR 6000). The concentrations of the two types of dyes were derived from the calibration curves obtained through experiments using standard solutions. The dye removal efficiency (R(%)) was calculated using the following equation: where C 0 is the initial dye concentration (mg/L) and C is the final dye concentration (mg/L). All experiments were conducted in triplicate for reproducibility and reliability. In each plot, the data are presented with standard deviation (error bars). The standard deviation of the results was less than 3%, thus, error bars are not clearly shown.

Statistics
An analysis of variance was performed on the dye removal experimental group. Three groups were tested for statistical significance between different pH conditions selected as one of the process parameters.

Effect of pH on Azo Dye Removal Efficiency by Potassium Ferrate (K 2 FeO 4 )
As the oxidation process is significantly affected by pH conditions, optimized pH conditions in K 2 FeO 4 treatment (oxidation-coagulation) were investigated using the remaining azo dyes as a function of time. Figure 1 shows the remaining dye concentration (mg/L) of OD and LGB after reacting with K 2 FeO 4 as a function of time under different pH values (3, 6, 9, and 12). Overall, the removal efficiencies of both dyes by K 2 FeO 4 were enhanced under acidic conditions. The OD and LGB removal efficiencies increased from 57% to 79% when the pH decreased from 12 to 3. However, an increase in contact time did not enhance the performance of K 2 FeO 4 at all pH conditions after 5 min of reaction. Superior performance under acidic conditions was primarily achieved due to the higher reactivity of the K 2 FeO 4 . Under acidic conditions, HFeO 4 − may be generated and reacted with dyes [24]. Moreover, K 2 FeO 4 is known to be more reactive when the pH is lower than 7.3. Therefore, the optimized pH condition for the removal of dye by K 2 FeO 4 was confirmed to be pH 3. This indicates that the initial solution pH had little impact on its oxidizing power. Meanwhile, K 2 FeO 4 had a much higher oxidation potential at acidic conditions (E 0 = 2.20 V) than at basic conditions (E 0 = 0.72 V). At pH < 7, although the oxidative ability of ferrate is high, the ferrate is highly unstable. This resulted in more reactive conditions. The effect of pH on azo dye removal efficiency was verified by ANOVA. Variation of the pH from 3 to 12 led to statistically significant differences in the removal of OD by K 2 FeO 4 . For OD removal by K 2 FeO 4 , the F value (46.526) was higher than F-critical (3.239) and the p-value was 3.98E-08 (<0.05). For LGB removal by K 2 FeO 4 , the F value (55.94) was higher than F-critical (3.239) and the p-value was 1.06E-08 (<0.05). This indicates that controlling pH in the dye removal experiment led to statistically significant differences in removal.

Dye Removal Rate by K 2 FeO 4 and NaOCl
Dye removal rates by K 2 FeO 4 and NaOCl treatment were examined by obtaining the slope of ln(C t ) as a function of reaction time (t). First, the dye oxidation behavior of K 2 FeO 4 was compared with that of NaOCl. Figure 2 shows the remaining concentrations of OD and LGB after oxidation as a function of time at pH 3, 6, 9, and 12. As most of the dyes were removed in 5 min, the removal efficiency was compared based on the removal rates at 5 min using the slope (ln(C t )/t) of ln(C t ) as a function of time. Table 1 lists the reaction rates for all conditions. While K 2 FeO 4 achieved the highest OD removal efficiency (80%) at pH 3, its removal efficiency dropped to 60% as the pH increased to 12, corresponding to a decreasing reaction coefficient (−0.8489). The fastest OD removal rate (−1.1894) was at pH 3. The single oxidation process with NaOCl showed the lowest removal efficiency (~30%) at pH 3 with the slowest reaction rate (-0.6330); however, dye oxidation increased marginally with an increase in pH (to alkali). At pH 12, OD removal using NaOCl was higher than that by K 2 FeO 4 . This is because the removal efficiency of K 2 FeO 4 decreases under alkaline conditions (pH 12) [21].
Meanwhile, at pH 3, LGB was almost completely removed by NaOCl oxidation; thus, NaOCl performed better than K 2 FeO 4 at pH 3. However, the removal efficiencies of the NaOCl oxidation decreased to 60% at pH 3 with minimal differences between them. There were distinctive differences between OD and LGB, except for the molecular weight. It may be assumed that the decomposition of LGB by hydrochloric acid was proceeded by the injection of hydrochloric acid into the jar tester to adjust the pH to acidic conditions (pH 3). The removal efficiency by oxidation-coagulation using K 2 FeO 4 was slightly higher than that by single oxidation using NaOCl, except at pH 3. Moreover, the marginal decrease in the removal of LGB may be due to the larger molecular weight of LGB in comparison to OD. This indicates that LGB needs additional reaction time for complete removal with K 2 FeO 4 , compared to OD.

Effect of Potassium Ferrate (K 2 FeO 4 ) Dose on Removal Efficiency
The optimized K 2 FeO 4 dose for effective dye removal was examined at pH 3, which was determined to be the ideal pH condition in Section 3.1. Figure 3 shows the removal efficiency as a function of dosage at different dye (a: OD and b: LGB) concentrations at pH 3. Overall, the removal efficiencies of both dyes increased with increasing K 2 FeO 4 doses. In particular, the OD and LGB concentrations (initial concentration 25 mg/L) decreased rapidly when the K 2 FeO 4 dose was increased from 28 to 42 mg/L. As the K 2 FeO 4 dose was increased to 56.40 mg/L and 70.50 mg/L, the highest OD removal of 99.68% was achieved in 15 min, and the highest LGB removal of 99.68% was achieved in 30 min. This result was better than that observed in previous studies, in which ozonation coupled with a biological aerated filter removed 100% of the azo dye from industrial wastewater in 120 min [16] and the advanced oxidation process degraded 100% of the reactive azo dye in 40 min [25].
The performance of K 2 FeO 4 was compared with that of NaOCl and FeCl 3 ·6H 2 O. As shown in Figure 4a, K 2 FeO 4 exhibited superior removal efficiency at 25 mg/L of OD. Furthermore, as its dose increased to 60 mg/L, complete removal was achieved. In contrast, FeCl 3 ·6H 2 O had the lowest removal efficiency (20%), and its performance was not enhanced by an increase in the dose. The removal efficiency of NaOCl increased up to 80% when the OD dose was greater than 80 mg/L; however, it was still lower than that of K 2 FeO 4 . Moreover, its performance was reduced when treated with 50 mg/L of OD (Figure 4b).
Additionally, 92% of the OD was removed, even though its dose was increased to over 80 mg/L. In contrast, the removal efficiency of NaOCl was reduced to half, even when the dose was increased to 80 mg/L. FeCl 3 ·6H 2 O showed a marginal increase in the removal of 50 mg/L of OD; however, its removal efficiency was still less than 40%. The superior efficiency of K 2 FeO 4 for dye oxidation was mainly due to the high reactivity and oxidizing capacity of Fe (VI). Based on the oxidizing power, the reduction potential of Fe (VI) (2.20 to 0.70 V) was estimated to be higher than that of ozone (2.076 V), hydrogen peroxide (1.776 V), chlorine (1.358 V), and perchlorate (1.389 V). In addition, ferrate (VI) is quick to oxidize the dye, as compared to permanganate and chromate. Furthermore, oxidation by K 2 FeO 4 generated subsequent coagulation through its reduction. It synergistically helped in the removal of organic dyes. However, FeCl 3 ·6H 2 O had a single coagulation effect, resulting in poor removal efficiency. Similar to OD removal, the combination of oxidation and coagulation exhibited the highest removal efficiency (100%) for LGB. At 25 mg/L, LGB was almost completely removed by K 2 FeO 4 (99.97%) (Figure 5a). The removal efficiency of NaOCl was comparable to that of K 2 FeO 4 ; that is, it reached a removal efficiency of 96.21% at an LGB dose of 88.81 mg/L. A marginal decrease in removal efficiency was observed for K 2 FeO 4 and NaOCl when the LGB concentration was increased to 50 mg/L, corresponding to 93.26% and 88.57%, respectively. Overall, K 2 FeO 4 and NaOCl exhibited a similar performance for LGB removal. Similar to the OD treatment, FeCl 3 ·6H 2 O had the poorest removal efficiency (approximately 20%) when treating 25 mg/L and 50 mg/L of LGB (Figure 5b). At the dose of 30 mg/L, K 2 FeO 4 reacted more actively with 25 mg/L of LGB than with 25 mg/L of OD, corresponding to 80% of LGB and 61% of OD removal efficiencies. However, 90% removal of OD (25 mg/L) was achieved at a K 2 FeO 4 dose of 40 mg/L, while 90% of the LGB (25 mg/L) was removed at a K 2 FeO 4 dose of approximately 50 mg/L. A similar trend was observed at 50 mg/L of K 2 FeO 4 , indicating that the reaction rate with LGB was faster at the low dose of K 2 FeO 4 (<30 mg/L), while at the higher dose (>30 mg/L) the reaction rate with OD was faster. This is because the relatively large LGB with a higher molecular weight (576.6 Da) reacted more strongly to K 2 FeO 4 at low dosage, as compared to OD (350.32 Da). Meanwhile, the OD removal rate was slowly saturated as a function of the K 2 FeO 4 dose, compared to LGB.
The combination of oxidation and coagulation by K 2 FeO 4 occurred in multiple steps, particularly at pH 3, where the highest amount of OD was removed. In contrast, single coagulation by FeCl 3 ·H 2 O treatment occurred in two steps. The first step of nucleation was completed in 5 min, and most of the OD was removed by coagulation during this period. During the remaining period of nucleation growth, no further OD degradation occurred. Thus, a single coagulation process is not an efficient method for azo dye removal because it takes a longer time and a higher amount of coagulation, as compared to the combination of oxidation and coagulation by ferrate. Oxidation by NaOCl was completed in 5 min at all pH conditions (3)(4)(5)(6)(7)(8)(9)(10)(11)(12), and there was no further OD degradation during the remaining period (up to 30 min). In addition, a single coagulation process generates colored coagulated solid wastes, which may result in the toxicity of the sludge produced and an increase in total dissolved solids in the treated wastewater [26].

Conclusions
Given its high redox potential and simultaneous generation of ferric coagulating species, K 2 FeO 4 was successfully applied to the treatment of dye wastewater. The OD removal efficiency reached 99.7% when 56.4 mg/L of K 2 FeO 4 was used for 15 min of contact time. The LGB removal efficiency was 99.0% using 70.5 mg/L of K 2 FeO 4 over 30 min of contact time. The optimized pH was found to be acidic (pH 3), while contact time had no significant effect on the removal efficiency of NaOCl. Moreover, K 2 FeO 4 exhibited a superior performance in OD and LGB removal, compared to NaOCl as Cl 2 and FeCl 3 ·6H 2 O as Fe (III). K 2 FeO 4 has the combined effect of oxidation and coagulation, which is useful and important for improving water quality. In comparison with the conventional two-step unit processes (oxidation and coagulation), K 2 FeO 4 is expected to produce a reduced footprint for wastewater treatment. K 2 FeO 4 treatment for azo dye removal has the potential to replace the conventional two-step process as a cost-effective method by synthesizing the K 2 FeO 4 salts on an industrial scale. The K 2 FeO 4 process has a possible application for wastewater treatment plants to treat dye-containing effluent on a real scale.