Degradation of Thiol Collectors Using Ozone at a Low Dosage: Kinetics, Mineralization, Ozone Utilization, and Changes of Biodegradability and Water Quality Parameters

Abstract

Ozonation at a high O₃ dosage can achieve high efficiencies in removing flotation reagents but it has a low ozone-utilization rate. The ozonation of potentially toxic thiol collectors (potassium ethyl xanthate (EX), sodium diethyl dithiocarbamate (SN-9), O-isopropyl-N-ethyl thionocarbamate (Z-200) and dianilino dithiophoshoric acid (DDA)) was investigated in an ozone-bubbled reactor at a low O₃ dosage of 1.125 mg/(min·L). The degradation kinetics, mineralization, ozone utilization, changes of biodegradability, and water quality parameters were studied, and the degradation behaviors of four collectors were compared. Thiol collectors could be effectively degraded with a removal ratio of >90% and a mineralization ratio of 10–27%, at a low O₃ dosage. The ozonation of thiol collectors followed the pseudo first-order kinetics, and rate constants had the order of k_SN-9 > k_EX > k_Z-200 > k_DDA. The Z-200 and DDA were the refractory flotation reagents treated in the ozonation process. After ozonation, the biodegradability of EX, SN-9, and DDA solutions was remarkably raised, but the biodegradability of Z-200 only increased from 0.088 to 0.15, indicating that the Z-200 and its intermediates were biologically persistent organics. After ozonation, the solution pH decreased from 10.0 to 8.0–9.0, and both the conductivity and oxidation-reduction potential increased. The ozone utilization ratio in decomposing thiol collectors was above 98.41%, revealing almost complete usage of input O₃. The results revealed that thiol collectors could be effectively degraded by O₃, even at a low dosage, but their degradation behaviors were quite different, due to intrinsic molecular properties.

1. Introduction

Froth flotation is extensively used to separate valuable minerals from sulfide ores. Thiol collectors, including xanthates, dithiophosphates, and dithiocarbamates, are important flotation reagents which can render sulfide minerals hydrophobic and facilitate bubble attachments [1]. As a huge amount of sulfide ores are treated annually by froth flotation, the quantities of consumed collectors are extremely large. Even in the 1980s, the global xanthate consumption per year was estimated to be more than 52,000 tons [2]. Therefore, thiol collectors and derived byproducts can be frequently encountered in flotation effluents. The hazards of xanthate to humans or aquatic lives have been comprehensively reviewed [3,4].

To reduce water consumption in mineral processing, the flotation effluents should be circulated into flotation circuits. However, in most cases, residual flotation reagents and their byproducts have negative effects on mineral flotation because these compounds can randomly alter the chemistry of flotation system [5,6]. Therefore, it is necessary to remove residual reagents from flotation effluents to improve water quality of flotation feeding water.

In the past decades, numerous methods have been developed to remove organic reagents from flotation effluents, such as coagulation and precipitation [7], adsorption [8], chemical oxidation [9], advanced oxidation [10,11,12], and biodegradation [13,14]. Adsorption is an efficient and simple technique in removing flotation reagents, but further treatment of sludge with toxic reagents and high water content becomes very difficult. The chemical oxidation with oxidants such as sodium hypochlorite often results in secondary pollution [9]. The biodegradation is a low-cost process for treating flotation effluents. However, the long treatment period [14] and the toxicity of some reagents to microbes [13,15] have limited its application in treating flotation effluents.

Recently, advanced oxidation processes (AOPs), involving ozone [10,11], Fenton’s reagents [16], hydrogen peroxide [17], persulfate [18], photocatalysis [19], and photoelectrooxidation [20] have been studied to degrade organic flotation reagents, especially xanthates. As hydroxyl radicals (·OH) have an oxidation-reduction potential (ORP) of 2.8 V, the AOPs have exhibited high efficiency in decomposing flotation reagents and their intermediates. However, most of these reports have just been limited to the removal of xanthates [10,11,19,20], and little attention is paid to degrading the other thiol collectors, such as dithiophosphates and dithiocarbamates. Since the molecular structures of thiol collectors are quite different, their decomposition behaviors by the AOPs may be different from each other.

Among above mentioned AOPs, ozonation is one of the most promising processes in decomposing organic pollutants. Ozonation at a high O₃ dosage can effectively decompose xanthates with a remarkable reduction of COD and generation of SO₄²⁻ ions [10,11]. In the flotation of sulfide ores, the pulp pH is usually kept at 9–12 [21,22]. Thus, the alkalinity of flotation effluents can enhance the ozonation of organic reagents as OH⁻ ions are the catalyst in decomposing O₃ to generate·OH [23,24]. Additionally, from the viewpoints of practical applications, ozone can be generated in situ, using air or oxygen as sources, in mines, avoiding the transportation and storage of dangerous chemical oxidants such as H₂O₂ [17] and sodium hypochlorite [9]. Therefore, ozonation is considered to be an appropriate technique to treat flotation effluents in mines.

However, ozonation is somewhat disputed for high energy consumption in O₃ generation [25,26]. Thus the ozone utilization becomes very critical in economic valuation of ozonation. In our previous study, O₃ utilization in degrading n-butyl xanthate ranged from 7.2 to 51.7% at the O₃ dosage of 5.88–50.36 mg/(min·L) [11]. Even when ultraviolet radiation (UV) was combined to promote the decomposition of O₃, the O₃ utilization in UV/O₃ process only increased by 15–30%, compared to pure O₃ [11]. Liu et al. [10] observed fast degradation of butyl xanthate at high O₃ dosage of 145.59 mg/(min·L), but the O₃ utilization was not concerned. As the reagent concentration in flotation pulps is usually 10⁻³–10⁻⁴ mol/L [27], the pollutant concentration in flotation effluents is much lower than that of pharmaceutical effluents and petrochemical wastewaters [28,29]. In ozonation, a lower concentration of organics usually leads to a declined O₃ decomposition rate, resulting in lower ozone utilization. Thus, a lower O₃ dosage might achieve a higher O₃ utilization, in the treatment of flotation effluents. Therefore, it is necessary to investigate the degradation efficiencies of flotation reagents and ozone utilization at low O₃ dosage.

In this study, four thiol collectors, potassium ethyl xanthate (EX), sodium diethyl dithiocarbamate (SN-9), O-isopropyl-N-ethyl thionocarbamate (Z-200), and dianilino dithiophoshoric acid (DDA), were selected as the sulfide mineral collectors. The O₃ dosage was controlled to be as low as 1.125 mg/(min·L), much lower than that previously used in the ozonation of xanthates [10,11]. The aims of this work were, (1) to investigate degradation kinetics, mineralization, and changes of biodegradability and water quality parameters in degrading thiol collectors using O₃, (2) to compare degradation efficiencies of thiol collectors with different molecular structures, and (3) to evaluate the ozone utilization and energy consumption efficiency at a low O₃ dosage. The result can provide fundamental aspects of ozone utilization and degradation behaviors of potentially toxic sulfide mineral collectors, at a low O₃ dosage.

2. Materials and Methods

2.1. Chemicals

The EX and SN-9 with analytical grade were purchased from Shanghai Aladdin Chemical Reagent Co., Ltd., Shanghai, China. The Z-200 and industrial grade DDA were purchased from Tieling Flotation Reagents Co. Ltd., China. Their molecular formulas, molecular structures, and abbreviations used are summarized in

Table 1. Collector name, molecular formula, molecular structure, and abbreviations used in this work.

Collector Name	Molecular Formula	Molecular Structure	Abbreviation
potassium ethyl xanthate	C2H4OCS2K		EX
sodium diethyl dithiocarbamate	(C2H5)2NCSSNa		SN-9
O-isopropyl-N-ethyl thionocarbamate	(CH3)2CHOCSNHC2H5		Z-200
dianilino dithiophoshoric acid	(C6H5NH)2PSSH		DDA

. By storing the collectors in a brown vacuum desiccator, they were separated from air and radiation, to prevent the oxidation. Other chemicals, such as iodine (I₂), potassium iodide (KI), silver sulfate (Ag₂SO₄), mercury sulfate (HgSO₄), and potassium bichromate (K₂Cr₂O₇) were of analytical grade and were purchased from Sinopharm Chemical Reagent Beijing Co., Ltd., Beijing, China. In all experiments, deionized water was used.

2.2. Experimental Procedures

All the degradation experiments were conducted with a batch mode in a jacket glass bubbled reactor connected to a thermostatic bath. The schematic diagram of experimental setup is shown in Figure 1. The cylindrical reactor, with a height of 1150 mm and internal diameter of 50 mm, was installed with a porous glass plate at the bottom to distribute the O₃ stream. Ozone was generated with air as the source, by an O₃ generator (SW-004, Qingdao West Electronic Purifiers Co., Qingdao, China). The O₃ stream was introduced into the reactor with a steady-state O₃ concentration of 1347 mg/m³ and at a gas flow rate of 0.1 m³/h. The degradation experiments were carried out at 25 ± 2 °C.

Prior to degradation experiments, the collector (0.2 g) was dissolved in 2 L deionized water to prepare an EX (SN-9, Z-200 or DDA) solution of 100 mg/L concentration. The initial pH was adjusted to 10.0 with 0.05 mol/L NaOH or HCl solution. While 2 L collector solution was introduced into the reactor, bubbled with an O₃ stream, the degradation began at the O₃ dosage of 1.125 mg/(min·L). As the four different collectors had different degradation efficiencies, to achieve the collector removal ratio of 90–100%, for the comparison, the ozonation time was chosen to be 90, 90, 120, and 180 min for the EX, SN-9, Z-200, and DDA, respectively. The aqueous samples were taken at designed intervals to determine the concentrations of collector, SO₄²⁻ ions, COD, BOD₅, and TOC. The water quality parameters (pH, conductivity, oxidation-reduction potential (ORP)) were measured by immersing the electrodes into solutions as shown in Figure 1.

2.3. Analysis and Calculation

2.3.1. Determination of the Collector Concentration

The xanthate concentration has always been determined by the UV-vis spectroscopic method. However, this method has seldom been reported to measure the concentration of the SN-9, Z-200 and DDA [13]. Figure 2a showed the UV-vis absorbance spectra of EX, SN-9, Z-200, and DDA solutions recorded by a UV-vis spectrophotometer (UV-5500PC, Shanghai Metash Instruments Co. Ltd., Shanghai, China). The maximum absorption peaks appeared at 301 nm for EX, 256 nm for SN-9, 241 nm for Z-200, and 230 nm for DDA solution, respectively. The plots of the collector concentration versus absorbance at its characteristic absorption wavelength are shown in Figure 2b. The correlation coefficients (>0.99) indicated that calibration curves could be used to calculate the collector concentration by recording its absorbance.

Thus, in this work, the concentration of four collectors was determined by the UV-vis spectroscopic method. The removal ratio of the collector was calculated as the following:

$${\beta }_{\mathrm{collector}}=\frac{C_0-C_{\mathrm{t}}}{C_0}\times 100\%$$

(1)

where β_collector was the removal ratio of the collector (EX, SN-9, Z-200 or DDA), C₀ and C_t (mg/L) were the collector concentration at initial and time t, respectively.

2.3.2. Determination of COD, BOD₅, TOC, and Concentration of SO₄²⁻ ions

The chemical oxygen demand (COD) was determined by the standard dichromate method (HJ/T 399‒2007). The biochemical oxygen demand (BOD₅) was detected by the classic dilution and inoculation method (HJ 505‒2009). The aqueous samples were incubated at 20 °C, for 5 days. The total organic carbon (TOC) concentrations were measured using a Shimadzu TOC-V organic carbon analyzer. The concentration of SO₄²⁻ ions was determined by a barium chromate spectrophotometry method (HJ/T 342–2007). In this study, the removal ratio of COD was obtained by Equation (2), as follows:

$${\beta }_{\mathrm{COD}}=\frac{CO{D}_0-CO{D}_{\mathrm{t}}}{CO{D}_0}\times 100\%$$

(2)

where β_COD was the removal ratio of COD, COD₀ and COD_t (mg/L) were the COD concentration at initial and time t, respectively.

The carbon mineralization ratio was calculated by Equation (3), as follows:

$${\gamma }_C=\frac{TO{C}_0-TO{C}_{\mathrm{t}}}{TO{C}_0}\times 100\%$$

(3)

where γ_C was the carbon mineralization ratio of the collector, and TOC₀ and TOC_t (mg/L) were the TOC concentration at initial and time t, respectively. As the SO₄²⁻ ions, with the highest valence of sulfur, were the final products during the oxidation of organic sulfur in thiol collectors, the sulfur mineralization ratio was defined by Equation (4), as follows:

$${\gamma }_{\mathrm{S}}=\frac{M}{n\times 96}\times \frac{C_{{\mathrm{SO}}_4^{2-},\mathrm{t}}}{C_0}\times 100\%$$

(4)

where γ_S was the sulfur mineralization ratio of the collector, M and n were the molecular weight and number of sulfur atom in the collector molecule (EX, SN-9, Z-200 or DDA), respectively, $C_{{\mathrm{SO}}_4^{2-},\mathrm{t}}$ was the concentration of SO₄²⁻ ions at time t, and C₀ (mg/L) was initial collector concentration.

2.3.3. Analysis of Water Quality Parameters

Water quality parameters of ozonated solutions, including solution pH, conductivity and ORP, were measured by a multi-parameter water quality meter equipped with different electrodes (Bante 900-UK, Shanghai Bante Instruments Co. Ltd., Shanghai, China). The solution pH was measured by a pH combination electrode (P11 mode). Before the measurement, it was calibrated with buffer solutions of pH 4.01, 7.00 and 10.01. The conductivity was determined by a glass conductivity probe with platinum cells (K10 mode). The ORP was recorded using an epoxy ORP electrode (501 mode), which was composed of a platinum pin and Ag/AgCl reference electrode.

2.3.4. Analysis of Gaseous O₃ Concentration

The gaseous O₃ concentration was determined by the KI absorption method (CJ/T 3028.2-94). The O₃ concentration in the output O₃ gas was measured after 30 min of ozonation to meet the solubility equilibrium of O₃ in solutions. The ozone utilization ratio was calculated by Equation (5), as follows:

$${\eta }_{\mathrm{O}}{}_3=\frac{C_{\mathrm{O}3,\mathrm{in}}-C_{\mathrm{O}3,\mathrm{out}}}{C_{\mathrm{O}3,\mathrm{in}}}\times 100\%$$

(5)

where η_O3 was the ozone utilization ratio, C_O3,in and C_O3,out (mg/m³) were the O₃ concentration in input and output O₃ gas streams, respectively.

3. Results and Discussion

3.1. Degradation Kinetics of Thiol Collectors

The degradation behaviors of thiol collectors by O₃ are shown in Figure 3a. Figure 3b illustrated the logarithmic plot of ln(C_t/C₀) versus ozonation time t, and kinetic parameters were summarized in

Table 2. Kinetic equations, pseudo-first-order rate constants (k_collector), half-life time (t_1/2) and correlation coefficients (R²) in the ozonation of thiol collectors.

Collector	Kinetic Equation	kcollector (min−1)	t1/2 (min)	R2
EX	Ct = C0·e−0.0579t	0.0579	11.97	0.9903
SN-9	Ct = C0·e−0.0687t	0.0687	10.09	0.9918
Z-200	Ct = C0·e−0.0194t	0.0194	35.69	0.9835
DDA	Ct = C0·e−0.0164t	0.0164	42.27	0.9967

. As shown in Figure 3a, EX, and SN-9 were rapidly degraded by O₃, but the degradation of Z-200 and DDA were much slower. At 90 min, the removal ratio of EX and SN-9 was close to 100%, but only 87.38% of the Z-200 and 79.71% of the DDA were removed. By extending the ozonation time, 98.39% of Z-200 could be removed at 120 min, but only 90.91% of DDA was degraded even at 180 min. As shown in Figure 3b and Table 2, the removal of thiol collectors by O₃ could be well described by the pseudo first-order kinetic models. By comparing rate constant of k_collector, it could be seen that the degradation by O₃, for the four collectors, followed the order of k_SN-9 > k_EX > k_Z-200 > k_DDA. Especially, the k_collector of SN-9 and EX was almost three times higher than that of Z-200 and DDA. The results revealed that in terms of ozonation, the SN-9 and EX could be easily decomposed, but the Z-200 and DDA were refractory flotation reagents.

Due to low operation cost, the biodegradation has been tested to degrade some flotation reagents. The first-order biodegradation rate constants of SN-9 and Z-200 were reported to be 3.13 × 10⁻⁴ and 4.17 × 10⁻⁵ min⁻¹, respectively [13]. By comparing the k_SN-9 (0.0687 min⁻¹) and k_Z-200 (0.0194 min⁻¹) in Table 2, it was clear that the rate constant in ozonation was two‒three orders higher than that in the biodegradation of same flotation reagent. The ozonation even at a low O₃ dosage achieved a much higher efficiency than the biodegradation in degrading flotation reagents.

Nowadays, residual flotation reagents are usually removed in tailing ponds with a natural degradation process. The first-order rate constant of EX natural degradation has been reported by Sun et al. [4] to be 2.094 × 10⁻⁵ min⁻¹, with a half-life of 22.99 day, at pH 7 and 25 °C, and have been given by Shen et al. [30] to be 1.55 × 10⁻⁵ min⁻¹, at pH 6.8 and 20 °C. Chen et al. [31] observed almost no natural degradation of EX at a pH ≥ 5, in 4 h. However, for the EX degradation, the ozonation at a low O₃ dosage of 1.125 mg/(min·L) could achieve the k_EX of 0.0579 min⁻¹, with a half-life of 11.97 min. Therefore, compared to a natural degradation process, ozonation at low O₃ dosage can effectively degrade flotation reagents, within a much shorter treatment time.

3.2. Carbon and Sulfur Mineralization of Thiol Collectors

In the ozonation of the thiol collectors, the concentrations of COD, TOC, and SO₄²⁻ ions were measured, as illustrated in Figure 4. For all collectors investigated, the concentration of both COD and TOC decreased with the increase of SO₄²⁻ concentration while the collector was degraded. Compared to the significant decrease of COD, the decline of TOC was much slower. As shown in

Table 3. The removal ratio of the collector and the COD, the carbon and sulfur mineralization ratio, and the pseudo-first-order rate constants for COD removal, in the ozonation of thiol collectors.

Collector	Removal Ratio of Collector (%)	Removal of COD			Mineralization Ratio (%)
		Removal Ratio of COD (%)	kCOD (min−1)	R2	Carbon	Sulfur
EX	99.98 (90 min)	60.04 (90 min)	0.0111	0.9843	18.42 (90 min)	11.95 (90 min)
SN-9	99.87 (90 min)	43.47 (90 min)	0.00694	0.9838	17.84 (90 min)	21.21 (90 min)
Z-200	87.38 (90 min)	13.67 (90 min)	0.00188	0.9858	8.31 (90 min)	12.54 (90 min)
	98.37(120 min)	14.97 (120 min)			11.82 (120 min)	20.92 (120 min)
DDA	79.71 (90 min)	53.25 (90 min)	0.00779	0.9888	16.35 (90 min)	7.89 (90 min)
	90.91(180 min)	61.11 (180 min)			26.05 (180 min)	21.98 (180 min)

, at the collector removal ratio of >90%, the removal ratio of COD was approximately below 62% and the carbon mineralization ratio (γ_C) was just lower than 27%, respectively. When comparing the rate constants of collector and COD removal (summarized in Table 2 and Table 3), the k_collector value was found to be 2–10 times higher than the k_COD, for each collector. It was revealed that only a small fraction of carbon in thiol collectors was completely mineralized by O₃, although all of the above 90% of collectors were decomposed. In the ozonation of n-butyl xanthate, n-butanol was detected by UPLC/Q-TOF-MS [32], and O-butyl peroxydithiocarbonate was found [11]. Thus, it could be reasonably inferred that most organic carbon in thiol collectors were still in the forms of reductive organic intermediates, after the ozonation.

As illustrated in Figure 4, the concentration of SO₄²⁻ ions increased to 18.93 mg/L for EX, 18.07 mg/L for SN-9, 13.65 mg/L for Z-200, and 15.01 mg/L for DDA, respectively, at the end of ozonation. The generation of the SO₄²⁻ ions indicated the complete oxidation of organic sulfur in the thiol collectors. However, as given in Table 3, the sulfur mineralization ratio (γ_S) was just below 22%, indicating that some sulfur byproducts, such as CS₂, S²⁻ and organics containing sulfur, might exist in solutions [10,11,30,32]. As volatile sulfur species such as CS₂ and H₂S were generated, they may have been emitted from ozone-bubbled solutions into a gas phase, resulting in lower γ_S values. For example, Yan et al. [32] had inferred the emitting of 20.6% sulfur into the gas phase in decomposing n-butyl xanthate by O₃. Generally, in this work, the mineralization ratio of carbon and sulfur for the four collectors ranged from 10% to 27%, at the collector removal ratio of above 90%.

As given in Table 3, the γ_C of Z-200 at 90 min was just 8.31%, much lower than that of EX, SN-9, and DDA at 90 min. The k_COD of EX, SN-9, and DDA was 5.91, 3.69, and was 4.14 folds higher than that of Z-200, respectively. It clearly revealed that the intermediates derived from the Z-200 were hardly oxidized by O₃. Although the DDA was very refractory in the ozonation, the γ_C of the DDA was close to that of EX and SN-9 at 90 min, indicating that its intermediates could be readily decomposed by O₃. By considering γ_C and γ_S together, the mineralization of four collectors by O₃ had the following order: SN-9 ≈ EX > DDA > Z-200.

In general, ozonation reactions are divided into two main pathways—direct ozone oxidation, occurring at pH < 4.0 and indirect ozone oxidation, occurring at pH > 10.0. At the pH of 4.0–10.0, both pathways exist, but direct ozone oxidation becomes predominant in the neutral and weak acid medium [33,34]. As shown in Figure 5, the solution pH decreased from 10.0, initially, to 8.0–9.0 at the end of ozonation. Therefore, in this case, the direct ozone oxidation might have greatly contributed to the collector decomposition during most of the ozonation period. Unfortunately, the direct oxidation of organics by O₃ has some disadvantages of low reaction rate and strong selectivity, as compared to oxidation by free radicals [35,36,37]. Thus, the intermediates, hardly oxidized by O₃ molecules, should be accumulated in the solutions.

3.3. Variation of Biodegradability

Although biodegradation is a low-cost procedure for removing organic pollutants, some flotation reagents, such as n-butyl xanthate [10], butyl amine aerofloat [10], ethylthionocarbamate [13], turpentine [38], isopropyl xanthate [39], and aniline aerofloat [40], are hardly biodegraded due to their special molecular structures. However, the intermediates derived by O₃, from these reagents, might become biodegradable [10]. Thus, a combined ozonation and biodegradation process may be more feasible than ozonation or biodegradation alone, in treating flotation effluents.

In this study, the variation of biodegradability of thiol collectors, before and after ozonation, was investigated. As shown in Figure 6, the BOD₅/COD for EX, SN-9, Z-200, and DDA solutions before the ozonation was 0.33, 0.19, 0.088, and 0.12, respectively. In general, organic pollutants with the BOD₅/COD of >0.3 are considered to be biodegradable [10]. Thus, it was clear that excepting the EX, the SN-9, Z-200, and DDA were biologically persistent flotation reagents. After ozonation for 90 min, the BOD₅/COD for EX, SN-9, and DDA solutions increased to 0.73, 0.31, and 0.32, respectively, indicating the generation of biodegradable intermediates. However, the BOD₅/COD for Z-200 solution just increased from 0.088 to 0.15, after ozonation for 120 min, revealing that its intermediates were still hardly biodegraded. It can be seen that both the Z-200 and its intermediates were biologically persistent organics. For all four collectors investigated, the increase of the biodegradability, after the ozonation, followed the order of EX > DDA ≈ SN-9 > Z-200.

3.4. Evolution of Solution pH, ORP and Conductivity

In the flotation of sulfide minerals, solution parameters such as pulp pH and ORP are very important in determining the floatability of minerals and interaction of flotation reagents with minerals [41,42,43,44]. Since treated flotation effluents will be reused in flotation circuits as feeding water, the changes of both solution parameters and reagent concentrations should be revealed to better understand the quality of reused flotation water. However, in the previous works of degrading flotation reagents by O₃ [10,11,32], H₂O₂ [31], Fenton reagents [45], and sodium hypochlorite [9], little attention was paid to revealing the evolution of solution parameters, except for the pH value.

In this work, the solution pH, ORP, and conductivity in degrading collectors were recorded as shown in Figure 5. For four collectors studied, the solution pH was declined from 10.0 to 8.0‒9.0, and the increase of the conductivity reached 68, 86, 101, and 121 μs/cm for EX, SN-9, Z-200, and DDA solution, respectively, at the collector removal ratio of >90%. The indirect ozonation reactions were initiated with catalytic decomposition of O₃ by OH⁻ ions, as shown in Equations (6)‒(11) [23,24]. The chain reactions of O₃ decomposition could generate free radicals, such as OH, HO₂, and·O₂⁻, which were responsible for breaking the chemical bonds of organics. Thus H⁺ ions were continually dissociated from H₂O, with an observed decrease of solution pH. In addition, simple carboxylic acids, such as formic and acetic acids could be generated via indirect reactions [46,47], which also contributed to the acidity of collector solutions. Thus, the pH of collector solutions was decreased after the ozonation.

$${\mathrm{H}}_2\mathrm{O}\leftrightarrow {\mathrm{H}}^{+}{+\mathrm{OH}}^{-}$$

(6)

$${\mathrm{O}}_3{+\mathrm{OH}}^{-}\to {\mathrm{HO}}_2{+\mathrm{O}}_2^{-}$$

(7)

$${\mathrm{HO}}_2\to {\mathrm{H}}^{+}{+\mathrm{O}}_2^{-}$$

(8)

$${\mathrm{O}}_3+{\mathrm{O}}_2^{-}\to {\mathrm{O}}_3^{-}+{\mathrm{O}}_2$$

(9)

$${\mathrm{O}}_3^{-}+{\mathrm{H}}^{+}\to {\mathrm{HO}}_3$$

(10)

$${\mathrm{HO}}_3\to \mathrm{OH}+{\mathrm{O}}_2$$

(11)

In the degradation of xanthates by O₃ or H₂O₂, inorganic ions, such as S²⁻, S₂O₃²⁻, SO₄²⁻ and CO₃²⁻ were observed to be readily generated [10,11,31]. Ionized intermediates, such as O-ethyl thiocarbonate (ETC⁻) and O-ethyl peroxydithiocarbonate (EPX⁻) were also detected in the oxidation of EX by H₂O₂ [17]. Therefore, the increase of conductivity directly revealed that some ionic intermediates were generated in the ozonation of collectors. The higher conductivity in Figure 5 corresponded to a lower concentration of collector and COD, as shown in Figure 4.

In the ozonation process, it is frequently observed that the ozone consumption rate will remarkably decrease with a rapid increase of the ORP and the reduction of pollutant removal efficiency, while the pollutant concentration goes below a certain value [48,49]. So, the ORP is a useful indicator in the control of O₃ addition, to minimize the process cost [48,50,51]. For the mineral flotation, the ORP is found to determine the oxidation of collectors and speciation of metal ions [41,44]. Thus, the revelation of ORP evolution is very important for both the control of O₃ addition and the reuse of treated flotation effluents.

As shown in Figure 5, the ORP of collector solutions, before ozonation, was negative. As the ozonation reactions occurred, the ORP increased up to 68.3, 17.2, 150.3, and 48.4 mV for EX, SN-9, Z-200, and DDA solutions, at the end of ozonation, respectively. The variations of COD concentration with the ORP are shown in Figure 7. The ORP of ozonated solutions increased with the reduction of the COD for the four collectors. Thus, it can be inferred that a certain correlation should exist between the COD and ORP of ozonated solutions, which might allow the ORP, as the indicator, to control the ozonation process.

3.5. Ozone Utilization

At the O₃ dosage of 1.125 mg/(min·L), the O₃ input rate was 2.26 mg/min. As shown in Figure 8, the O₃ emitting rate for deionized water at pH 10.0 was just 0.495 mg/min, and the O₃ utilization ratio (η_O3) of 78.01% was achieved. The color of KI absorption liquid was changed from a dark red, due to the absorption of the input O₃ gas, to a light yellow for the absorption of the O₃ emitted from water. The result revealed that 78.01% of input O₃ was dissolved and decomposed in alkaline water. Since OH⁻ ions could catalytically decompose dissolved O₃ as given in Equations (6)–(11), the alkalinity of water promoted the dissolution of gaseous O₃ into water, resulting in a lower O₃ emitting rate. It suggests that the alkalinity of flotation effluents is beneficial to the O₃ utilization and generation of free radicals (OH, ·HO₂ and O₂⁻, etc.). In this work, low O₃ dosage should have also contributed to a high η_O3, for deionized water.

When the collector concentration was raised from 1 to 100 mg/L, the η_O3 shown in Figure 8b increased from 94.22% to 99.89%, for thee EX and from 88.62% to 99.97%, for the DDA solution, respectively. The color of the KI absorption liquids, for absorbing the O₃ emitted, from the 100 mg/L EX and DDA solutions was almost white, directly demonstrating nearly no escape of O₃ from the reactors. It was very clear that the addition of collectors increased the O₃ utilization, which could be attributed to the reactions of O₃ molecules, with collectors. As shown in Figure 8b, the EX had higher η_O3 than the DDA, while the concentration was <10 mg/L. The difference in η_O3 directly revealed that the EX was more sensitive in reacting with O₃ than the DDA. In this work, at a low O₃ dosage of 1.125 mg/(min·L), the η_O3 was close to 100%, in the ozonation of EX and DDA with 100 mg/L concentration. However, in the ozonation of n-butyl xanthate, the achieved η_O3 just ranged from 7.2% to 51.7%, at the O₃ dosage of 5.88‒50.36 mg/(min·L) [11]. The results revealed that the ozonation of thiol collectors at a low O₃ dosage could achieve almost 100% utilization of O₃, meaning it reduced the treating cost by adding a lesser amount of O₃.

3.6. Analysis of Energy Consumption Efficiency

In the treatment of contaminants with AOPs, the electrical energy per order (EE/O) was introduced to directly evaluate the electric efficiency and feasibility of the scale-up of AOPs [52]. This parameter was defined as “the electric energy in kWh required in degrading a contaminant by one order of magnitude in 1 m³ contaminated water”. To calculate the energy requirements for the ozonation, an average energy consumption of 15 kWh/kg for the O₃ production was assumed [53], and the energy calculations were based on a 90% removal of collectors. EE/O values (kWh/(m³·order)) were calculated for batch operations as given by Equation (12).

$$EE/O=\frac{P\times 1000}{V\times 60}\times \frac{\ln 10}{k_{\mathrm{collector}}}$$

(12)

where P was the electric power (kWh) required for producing dosed O₃, V was the volume (L) of test collector solution, and k_collector was the pseudo-first-order constant (min⁻¹).

The results of the energy calculation for the ozonation of the four collectors are summarized in

Table 4. Electrical energy per order (EE/O) in the ozonation of the four thiol collectors.

Collector	EX	SN-9	Z-200	DDA
EE/O (kWh/(m3·order))	0.45	0.33	3.18	6.06

. As shown in Table 2, the k_collector of EX and SN-9 was lower than that of Z-200 and DDA, revealing the higher efficiency in the degradation of EX and SN-9. Thus, the EE/O values for EX and SN-9 removal were much lower than that for Z-200 and DDA. However, at a low O₃ dosage of 1.125 mg/(min·L), the ozonation of all of four collectors was feasible, in terms of energy consumption, since values of EE/O, not higher than 10 kWh/(m³·order), were considered suitable for practical application [54].

4. Conclusions

Thiol collectors (EX, SN-9, Z-200 and DDA) could be effectively degraded by O₃ at a low dosage of 1.125 mg/(min·L). Both the removal of collectors and the decline of COD followed the pseudo-first-order kinetic models. The ozonation of the four collectors followed the order of k_collector: k_SN-9 (0.0687 min⁻¹) > k_EX (0.0579 min⁻¹) > k_Z-200 (0.0194 min⁻¹) > k_DDA (0.0164 min⁻¹). The k_collector in the removing-collectors was 2‒10 times higher than the k_COD of COD removal. At the collector removal ratio of >90%, the mineralization ratio of the four collectors ranged from 10% to 27%, with the order of SN-9 ≈ EX > DDA > Z-200. For the four collectors investigated, EX and SN-9 could be easily degraded by O₃, but the ozonation of Z-200 and DDA exhibited much lower efficiencies. Especially, both the Z-200 and its intermediates were hardly decomposed by O₃.

The SN-9, Z-200, and DDA were found to be biologically persistent flotation reagents as their BOD₅/COD values were below 0.2. After ozonation, the BOD₅/COD of EX, SN-9, and DDA solutions increased to 0.73, 0.31, and 0.41, respectively, with remarkable increase of the biodegradability. However, the BOD₅/COD of Z-200 solution increased just from 0.088 to 0.15, after the ozonation, revealing that its intermediates were still hardly biodegraded. After ozonation, the solution pH of four collectors decreased from 10.0 to 8.0–9.0, and an increase of conductivity (68–121 μs/cm) was observed, exhibiting the generation of various ionic intermediates. For the four collectors, the ORP rapidly increased with the decline of COD concentration, indicating that the ORP might act as the indicator to monitor the ozonation of collectors.

At low O₃ dosage of 1.125 mg/(min·L), the η_O3 reached 78.01% for ionized water at pH 10.0 due to effective O₃ decomposition initiated by OH⁻ ions. The addition of collectors significantly enhanced the O₃ decomposition. At the collector (EX and DDA) concentration of >10 mg/L, the η_O3 increased to above 98.41%, revealing an almost complete usage of input O₃ at a low O₃ dosage. EE/O values revealed that the ozonation of all of four collectors was feasible, in terms of energy consumption.