Treatment of complex sulfur-containing solutions in ammonia desulfurization ammonium sulfate production by ultrasonic-assisted ozone technology

Highlights

• •The complex transformation law of sulfur-containing ions in solution was studied.
• •The existence and cause of sulfur impurities in complex sulfur-containing solutions were proven.
• •US/O₃ is exploited to simultaneously remove thiosulfate and sulfite to address the yellowing of (NH₄)₂SO₄.
• •The degree of involvement of the oxidation component is O₃ ≫ •OH > ¹O₂ > •O₂⁻.

Abstract

In this work, the cause of abnormal color in ammonium sulfate products formed by flue gas desulfurization is revealed by investigating the conversion relationship between different sulfur-containing ions and their behavior in a sulfuric acid medium. Both thiosulfate (S₂O₃²⁻) and sulfite (SO₃²⁻ ＆ HSO₃⁻) impurities affect the quality of ammonium sulfate. The S₂O₃²⁻ is the main reason for the yellowing of the product due to the formation of sulfur impurities in concentrated sulfuric acid. To address the yellowing of ammonium sulfate products, a unified technology (US/O₃), using ozone (O₃) and ultrasonic waves (US) simultaneously, is exploited to remove both thiosulfate and sulfite impurities from the mother liquor. The effect of different reaction parameters on the degree of removal of thiosulfate and sulfite is investigated. The synergistic effect of ultrasound and ozone on ion oxidation is further explored and demonstrated by the comparative experiments with O₃ and US/O₃. Under the optimized conditions, the thiosulfate and sulfite concentration in the solution is 2.07 and 5.93 g/L, respectively, and the degree of removal is 91.39 and 90.83%, respectively. The product obtained after evaporation and crystallization is pure white and meets the national standard requirements for ammonium sulfate products. Under the same conditions, the US/O₃ process has apparent advantages, such as saving reaction time compared with the O₃ process alone. Introducing an ultrasonically intensified field improves the generation of oxidation radicals ·OH, ¹O₂, and ·O₂^– in the solution. Furthermore, the effectiveness of different oxidation components in the decolorization process is studied by adding other radical shielding agents using the US/O₃ process supplemented with EPR analysis. The order of the different oxidation components is O₃(86.04%) > ¹O₂(6.53%) > •OH(4.45%) > •O₂^–(2.97%) for the oxidation of thiosulfate, and it is O₃(86.28%) > •OH(7.49%) > ¹O₂(4.99%) > •O₂^–(1.25%) for the oxidation of sulfite.

1. Introduction

Sulfur dioxide (SO₂) is one of the primary pollutants in flue gas produced by the metallurgical and chemical industries [1]. In some metal smelters, ammonia desulfurization to prepare ammonium sulfate ((NH₄)₂SO₄) is an effective way to realize flue gas desulfurization. The obtained product (NH₄)₂SO₄ is a necessary chemical by-product with enormous application demand in many fields, such as agriculture, textiles, leather, and medicine [2]. The specific steps involved in this process are as follows: (i) passing the cooled and purified sulfur dioxide flue gas through the absorption tower to produce a mixed solution of (NH₄)₂SO₄ and sulfites ((NH₄)₂SO₃ and NH₄HSO₃), called the mother liquor, (ii) adding an excess of H₂SO₄ to oxidize all the sulfites in the mother liquor, (iii) adding NH₃·H₂O to neutralize excess concentrated sulfuric acid, and (iv) obtaining the ammonium sulfate product through evaporation and crystallization [3]. However, in actual production, yellow foam is formed after adding concentrated sulfuric acid to oxidize sulfites, which subsequently gets precipitated into yellow solids in the solution. Yellow ammonium sulfate products are produced upon further evaporation and crystallization. The yellow ammonium sulfate can only be stockpiled or returned to the desulfurization process for repurification because its quality does not meet the chromaticity requirement of the national standards. This process brings substantial economic loss to the enterprise and reduces the productive recycling of sulfur elements. For example, the annual loss of the yellow ammonium sulfate for a non-ferrous metal smelting enterprise that uses flue gas desulfurization is as high as 1.856 million U.S.D. to produce 8000 tons of ammonium sulfate.

Many reasons have been proposed for the abnormal coloration of ammonium sulfate products. Mohod et al. [4] showed that the crystallization conditions of ammonium sulfate, such as temperature, stirring rate, and pH value, also affect the color of solid ammonium sulfate. For sulfur impurity, some physical methods like filtration [5] and membrane separation [6] have been reported in the literature and adopted by enterprises in the production process to reduce the impact of sulfur on ammonium sulfate products. However, the color of ammonium sulfate products could improve from clay bank or yellow to light yellow instead of white even after using these methods. In addition, the production process must frequently pause to replace filter cloths due to clogging. These physical methods failed to solve the problem at the source, increasing the operation's difficulty and cost. A white solid product can be obtained by reducing the amount of concentrated sulfuric acid in the oxidation step. However, the product is in a fine powder or sticky state or does not crystallize because of the high sulfite concentration after oxidation. Also, this product can not meet the requirements of commercial standards. Therefore, the most critical work is to analyze the cause of sulfur production at the front end and solve the problem from the source. Kelemen et al. [7] reported that S₂O₃²⁻ generates sulfur and other sulfur oxides by disproportionation reactions under acidic conditions. Lara et al. [8] showed that sulfur ions and SO₃²⁻ also produce sulfur precipitates under acidic conditions (2S²⁻ + 6H⁺ + SO₃²⁻ → 3S + 3H₂O). Liu et al. [9] showed sulfur ions have a complex conversion relationship. Hence, it is of great scientific and industrial significance to determine the cause of the abnormal coloration of ammonium sulfate products and find corresponding solutions due to the diversity of sulfur-containing substances and their complex transformation relationship.

To address the above issues, we perceived and discussed the transformation between sulfur-containing ions in the mother liquor and their changes under the action of concentrated sulfuric acid by preparing different kinds of synthetic solutions. We examined the reasons for the formation of sulfur by employing analytical characterizations, such as XRD and XPS. Subsequently, an advanced oxidation technology combining ultrasonic irradiation (US) and ozone (O₃) is proposed to remove impurities in the inadequate mother liquor. Further, the optimal process conditions are determined by systematically studying the influence of each reaction parameter. Finally, the oxidation mechanism of inadequate mother liquor by the advanced oxidation technology is discussed with the help of EPR studies carried out by adding free radical shielding agents, and the degree of participation of different oxidation components is revealed. In this work, for the first time, the details of ultrasonic-assisted ozone processing for treating complex sulfur-containing solutions are systematically studied, focusing on the reason for product discoloration, solution provision, and mechanism. The results of this study provide crucial theoretical guidance for improving the quality and efficiency of ammonium sulfate produced by the ammonia method.

2. Experimental

2.1. Materials

In this work, the solution obtained in step (i) of the ammonium sulfate production process by ammonia desulfurization is used as the test solution. A solution with S₂O₃²⁻ concentration of 24.59 g/L, HSO₃⁻&SO₃²⁻ concentration of 64.70 g/L, and pH of 5.0 was obtained from the flue gas desulfurization process of a lead–zinc smelting company in Yunnan, China. I₂ (0.1 mol/L, AR grade), Na₂S₂O₃ (0.1 mol/L, AR grade), CH₃COOH (95%, AR grade), HCHO (AR grade), and Na₂SO₃ (AR grade) were purchased from Shanghai Sinopharm Chemical Reagent Co. Ltd., China. O₂ (>99.99%) and SO₂ gases (>99%) were obtained from Kunming Gas Plant, China. Tert-butyl alcohol (AR grade, 99%) was obtained from the Chengdu Institute of the Joint Chemical Reagent, China. Superoxide dismutase SOD (AR grade, 99%) was obtained from Shanghai Yuanye Biological Technology Co. Ltd., China.

2.2. Experimental procedure

Ultrasonic equipment with a low frequency of 20 kHz and a maximum power of 1200 W developed in-house by the Key Laboratory of Unconventional Metallurgy, Kunming University, Ministry of Education, was used in this study. Ozone was produced from pure oxygen through an ozone generator (Model JW-30A, Xuzhou Jiuzhoulong Ozone Equipment Manufacturing Co., Ltd.). During the experiments, ozone was continuously bubbled with a suitable flow rate controlled by adjusting the glass rotameter and then dispersed into the solution through a diffusion aerator.

Typically, 500 mL of solution was taken in a three-necked flask. The stirring rotor, ozone gas pipeline, and ultrasonic generator were inserted into the flask before placing it in the water bath with continuous magnetic stirring. In the preliminary experiment, we studied the effect of mechanical stirring on the oxidation effect of the mother liquor and found that the optimal mechanical stirring rate was 800 RPM. Therefore, the rate of mechanical stirring was fixed at 800 RPM in this work. During the experiment, numerical control equipment was employed to monitor and regulate parameters such as ultrasonic power density and temperature. The concentration of S₂O₃²⁻ and sulfite (SO₃²⁻＆HSO₃⁻) in the samples subjected to various reaction conditions was measured. The effect of reaction temperature (20, 30, 40, and 50 °C), reaction time (0.5, 1, 1.5, 2, 2.5, 3, 3.5, and 4 h), ozone gas flow (0.4, 0.6, 0.8, and 1.0 L/min), and ultrasonic power density (200, 400, 600, 800, and 1000 W·L⁻¹) on decolorization was investigated.

Subsequently, ammonia was used to adjust the pH value of the treated solution to 8. The solution was evaporated and crystallized at 85 °C with a negative pressure of −0.07 MPa by a rotational vacuum device. Evaporation could be stopped when the solid content of the mother liquor was 20%. The solid and liquid were separated, and the solid was dried at 60 °C for 5 h to obtain the ammonium sulfate product.

2.3. Analytical methods

In this work, sulfite was used to represent the sum of SO₃²⁻ and HSO₃⁻, and its content was used to describe the content of SO₃²⁻and HSO₃⁻. The iodometric method determined the S₂O₃²⁻ and sulfite contents [10], and the details are given in Text S1 in the Supplementary Material.

The generation of •OH, •O₂⁻, and ¹O₂ during the reaction was characterized by electron paramagnetic resonance spectrometry (EPR, Bruker EMX PLUS, Germany) using 5,5-dimethyl-1-pyrroline (DMPO) as the spin trapping agent. The phase of the ammonium sulfate product was investigated by a powder X-ray diffractometer (Rigaku dX 2000, Cu-Ka radiation, operating at 40 kV and 25 mA), and the XRD data was analyzed using MDI Jade (Materials Data, Inc., version 6.0). X-ray photoelectron spectroscopy determined the change in the mother liquor's sulfur valence state (PHI-5300, PHI, Waltham, MA, USA).

3. Results and discussion

3.1. Reasons for the abnormal color of ammonium sulfate products

This section discusses the causes of the yellow color of ammonium sulfate. By preparing a series of synthetic solutions with the same concentration as the real solution, we discuss the transformation relationship of sulfur-containing ions and provide a theoretical basis for solving the problem of abnormal coloration.

3.1.1. Effect of sulfur on the color of ammonium sulfate products

In this work, the yellow foam was generated after adding concentrated sulfuric acid to adjust the pH value of the inadequate mother liquor to 1–2, different from the transparent solution obtained for the competent mother liquor. The oxidation step could not be entirely completed as the liquid level went beyond its standard due to the influence of the yellow foam in the actual production process. When the mixture of mother liquor and acid was kept for 1–2 days, yellow precipitates (marked 1# yellow solid) were formed and accumulated at the bottom and on the walls of the glass bottle. Subsequently, the neutralized solution was evaporated and crystallized in a water bath connected to a rotational vacuum device to obtain a yellow ammonium sulfate product (marked as 1# yellow ammonium sulfate product), different from the white product obtained under normal production conditions. The S2p XPS results of 1# yellow solid, 1# yellow ammonium sulfate, and white ammonium sulfate with AR grade are shown in Fig. 1a. The peak at 168.38 eV in the S2p XPS spectra of both white and yellow ammonium sulfate products corresponds to SO₄²⁻ [11]. The peak with the binding energy of 163.18–163.88 eV, corresponding to elemental sulfur [12], is the only peak noticed in the S2p XPS spectrum of 1# yellow solid, which is also evident in the XPS spectrum of the yellow ammonium sulfate product.

Fig. 1.

Characterization of different solids (1# yellow solid, 2# yellow solid, 1# yellow ammonium sulfate, and 2# yellow ammonium sulfate). (a) XPS spectra; (b) photos of solid products; (c) XRD patterns.

For further verification, the obtained 1# yellow ammonium sulfate was dissolved in excess water and filtered by a Buchner funnel suction device to get a filtrate and a yellow solid (marked as 2# yellow solid). The filtrate obtained in the preceding steps was evaporated and crystallized again to obtain 2# yellow ammonium sulfate. The 2# yellow ammonium sulfate product exhibits a light yellow color (shown in Fig. 1b). It can be seen that the color of the 2# yellow ammonium sulfate product is improved compared with the 1# ammonium sulfate. The XRD patterns of 2# yellow solid, 1# yellow ammonium sulfate product, and white ammonium sulfate with AR grade are shown in Fig. 1c. The XRD analysis confirms that 2# yellow solid is also a sulfur impurity. The results suggest that the sulfur impurity produced by mother liquor and concentrated sulfuric acid is the direct cause of the yellowing of the ammonium sulfate product. In addition, the results show that the chromaticity of the ammonium sulfate can only be slightly improved by filtration after dissolution but cannot meet the white chromaticity requirement of the national standard.

3.1.2. Reasons for the existence of sulfur

The amount of sulfur-containing anions in the inadequate mother liquor, including sulfate (SO₄²⁻), sulfites (HSO₃⁻ & SO₃²⁻), thiosulfate (S₂O₃²⁻), and sulfide (S²⁻) is 392.41, 64.70, 24.03, and 0.17 g/L, respectively. Among these, SO₄²⁻ is very stable in concentrated sulfuric acid and can not generate sulfur. Therefore, we performed the following experiments on HSO₃⁻, SO₃²⁻, S₂O₃^2−, and S²⁻. Synthetic solutions containing an individual ion or a combination of different ions were prepared as per the ionic concentrations of the real solutions. Then concentrated sulfuric acid was added to these solutions to adjust the pH to 1–2. The composition of different synthetic solutions and their changes under the influence of concentrated sulfuric acid are shown in Table 1 and Fig. 2a-2d. The yellow solid produced in some solutions due to the reaction with concentrated sulfuric acid was filtered and dried, and its composition was analyzed by XPS, as shown in Fig. 2e.

Table 1.

Color changes of synthetic solutions with different combinations of sulfur-containing components under concentrated sulfuric acid with varying aging times.

No.	Combination	After the preparation of the synthetic solution	After adding concentrated sulfuric acid and no aging	After adding concentrated sulfuric acid and aging for 2 days	After adding concentrated sulfuric acid and aging for 14 days
1	S2−	transparent colorless solution	transparent colorless solution	transparent colorless solution	transparent colorless solution
2	SO32−	transparent colorless solution	transparent colorless solution	transparent colorless solution	transparent colorless solution
3	HSO3−	transparent colorless solution	transparent colorless solution	transparent colorless solution	transparent colorless solution
4	S2O32−	transparent colorless solution	turbid yellow solution	yellow solid + yellow solution	yellow solid + transparent colorless solution
5	S2− + SO32−	transparent colorless solution	turbid yellow solution	yellow solid + yellow solution	yellow solid + transparent colorless solution
6	S2− + HSO3−	transparent colorless solution	turbid yellow solution	yellow solid + yellow solution	yellow solid + transparent colorless solution
7	S2− + S2O32−	transparent colorless solution	turbid yellow solution	yellow solid + yellow solution	yellow solid + transparent colorless solution
8	SO32− + HSO3−	transparent colorless solution	transparent colorless solution	transparent colorless solution	transparent colorless solution
9	SO32− + S2O32−	transparent colorless solution	turbid yellow solution	yellow solid + yellow solution	yellow solid + transparent colorless solution
10	HSO3− + S2O32−	transparent colorless solution	yellowish transparent solution	yellow solid + yellow solution	yellow solid + transparent colorless solution
11	S2− + SO32− + HSO3−	transparent colorless solution	turbid yellow solution	yellow solid + yellow solution	yellow solid + transparent colorless solution
12	S2− + SO32− + S2O32−	transparent colorless solution	turbid yellow solution	yellow solid + yellow solution	yellow solid + transparent colorless solution
13	SO32− + HSO3− + S2O32−	transparent colorless solution	yellowish transparent solution	yellow solid + yellow solution	yellow solid + transparent colorless solution
14	S2− + HSO3− + S2O32−	transparent colorless solution	turbid yellow solution	yellow solid + yellow solution	yellow solid + transparent colorless solution
15	S2− + SO32− + HSO3− + S2O32−	transparent colorless solution	yellowish transparent solution	yellowish transparent solution	yellow solid + transparent colorless solution

Fig. 2.

The colors (a-c) of synthetic solutions with different sulfur-containing components before (above) and after (below) adding sulfuric acid (1#) S²⁻, (2#) SO₃²⁻, (3#) HSO₃⁻, (4#) S₂O₃²⁻, (5#) S²⁻ + SO₃²⁻, (6#) S²⁻ + HSO₃⁻, (7#) S²⁻ + S₂O₃²⁻, (8#) SO₃²⁻ + HSO₃⁻, (9#) SO₃²⁻ + S₂O₃²⁻, (10#) HSO₃⁻ + S₂O₃²⁻, (11#) S²⁻ + SO₃²⁻ + HSO₃⁻, (12#) S²⁻ + SO₃²⁻ + S₂O₃²⁻, (13#) SO₃²⁻ + HSO₃⁻ + S₂O₃²⁻, (14#) S²⁻ + HSO₃⁻ + S₂O₃²⁻,(15#) S²⁻ + SO₃²⁻ + HSO₃⁻ + S₂O₃²⁻; (d) of the samples after adding sulfuric acid and aging for different times; (e) XPS spectrum of the yellow precipitate.

Fig. 2a shows the case of the four individual ions. Although S²⁻ could be converted to sulfur impurities by the oxidation effect of concentrated sulfuric acid, the S²⁻ synthetic solution (1#) remained colorless after adding concentrated sulfuric acid. As the concentration of S²⁻ in the solutions used in this study is far less than that of concentrated sulfuric acid, S²⁻ is more likely to be converted into higher valence sulfur oxide species such as SO₃²⁻, SO₄²⁻, and SO₂ instead of producing sulfur. In the presence of concentrated sulfuric acid, SO₃²⁻ or HSO₃⁻ forms SO₄²⁻ through the oxidation effect and SO₂ through the acidity effect. Therefore, the chromaticity of SO₃²⁻ or HSO₃⁻ synthetic solutions (2# and 3#) has no noticeable change before and after adding concentrated sulfuric acid. The only sulfur impurity reacting with concentrated sulfuric acid to form yellow precipitates is S₂O₃²⁻, as confirmed by the XPS (Fig. 2e). As reported in other studies [8], [13], the possible reactions between thiosulfate and concentrated sulfuric acid are given below (Eqs. (1)-(3)).

$$3S_2{O}_3^{2-}+2H^{+}\to 4S+2S{O}_4^{2-}+H_{2}O$$ (1)

$$2S_2{O}_3^{2-}+H^{+}\to 2S+S{O}_3^{2-}+HS{O}_3^{-}$$ (2)

$$S_2{O}_3^{2-}+2H^{+}\to S+S{O}_2+H_{2}O$$ (3)

It is observed that the process becomes more complex when concentrated sulfuric acid is added to the synthetic solutions containing S₂O₃²⁻ such as S₂O₃²⁻ (4#), S²⁻+S₂O₃²⁻ (7#), SO₃²⁻+S₂O₃²⁻ (9#), and S²⁻+SO₃²⁻+S₂O₃²⁻ (12#).

The phenomenon and principle of the whole conversion process are explained below:

(i) Adding concentrated sulfuric acid (98.3%) dropwise to the transparent synthetic solution to adjust the pH value of the mixture to 1.5–2 resulted in heat generation and evolution of SO₂ with some H₂S and the formation of a yellow turbid solution. The observed turbid yellow solution is a micelle of sulfur sol composed of many colloidal particles. The structure of each colloidal particle is such that the sulfur element S_n is in the core of the colloidal particle, whereas the hydrophilic long-chain polythionates S_mO₆²⁻ (m = 11–40, even 40–140) cover the surface [14], [15]. Sulfur sol, represented by -SO₃-S_n-SO₃- or (S_n-S_mO₆)²⁻, can be formed as per Eq. (4). In addition to hydrophobic S₈, elemental sulfur S_n also exists in the colloidal core as hydrophilic sulfur elements such as S₆, S₇, S₉, S₁₀, and S₁₂ [16], making the total elemental sulfur being stable in the colloid without precipitation. However, these hydrophilic sulfur elements are unstable; they are transformed to more stable hydrophobic S₈ and subsequently, precipitate if eroded by the surrounding water as per Eq. (5). The long-chain polythionates S_mO₆²⁻ on the surface of the colloidal particles play a crucial role in keeping the overall elemental sulfur, including hydrophobic and hydrophilic sulfur, stable in the core of the colloidal particles without precipitation. This is the cause of the entire colloidal particles' overall hydrophilicity and negative ionic charge. The chain structure of S_mO₆²⁻ with 11–40 S atoms can make these colloidal particles slightly soluble, thus resulting in a turbid yellow colloidal solution.

(ii) After aging for 2 days, yellow solid precipitates are formed at the bottom of the glass bottle. The solution gradually turns from a turbid yellow to a light yellow turbid to a yellowish transparent one. The long-chain polythionate continuously decomposes to form medium-chain polythionate S_mO₆²⁻ (m = 6–10) or short-chain polythionate S_mO₆²⁻ (m < 6) as the aging time is prolonged, leading to the precipitation of elemental sulfur as per Eqs. (6)-(8). Steudel [16] also observed that short-chain polythionates were absent in fresh sulfur sol obtained by the reaction between sodium thiosulfate and concentrated sulfuric acid, but then the concentration of short-chain polythionate increased while the concentration of long-chain polythionate decreased with prolonged aging time.

(iii) After aging for 14 days, more yellow solid residue is obtained, and the color of the solution changes from yellowish to transparent. This is because the long-chain polythionates are decomposed into sulfur elements and short-chain polythionates that are easily soluble in water. The short-chain polythionates may also be oxidized by concentrated sulfuric acid to generate SO₂, SO₃²⁻ or stable SO₄²⁻ with higher sulfur valence, as given in Eqs. (1)-(3). The final solution is transparent and colorless, while sulfur impurities precipitate as yellow sediments.

The morphology of the remaining samples after aging for different times is shown in Fig. S1 in the Supplementary Material.

nS₂O₃²⁻ + (n-2)H⁺ → ⁻SO₃-S_n-SO₃⁻ + (n-2)HSO₃⁻ (4)

4/3S₆ = 8/7S₇ = S₈ (5)

⁻SO₃-S_n-SO₃⁻ + H₂O → HSO₄⁻ + HS_n-SO₃⁻ (6)

HS_n-SO₃⁻ → S_n + HSO₃⁻ (7)

S_mO₆²⁻ = S_m-xO₆²⁻ + x/nS_n (8)

However, HSO₃⁻+S₂O₃²⁻ (10#), SO₃²⁻+HSO₃⁻+S₂O₃²⁻ (13#), and S²⁻+SO₃²⁻+HSO₃⁻+S₂O₃²⁻ (15#) samples react with concentrated sulfuric acid to produce a yellowish transparent solution instead of a turbid yellow solution. As shown above, the turbid yellow sulfur sol is formed by wrapping sulfur with long-chain polythionate. HSO₃⁻ may inhibit the formation of long-chain polythionates from S₂O₃²⁻ and concentrated sulfuric acid, as given in Eq. (9). Short-chain polysulfates such as S₃O₆², which are readily soluble in water, may still form [15], [17]. Therefore, the obtained yellowish transparent solution signifies the formation of short-chain polythionates and hydrophilic sulfur S_n. However, short-chain polythionates were also unstable [18] and might have decomposed into sulfur or oxidized to generate sulfur-containing species with higher sulfur valence after reacting with concentrated sulfuric acid. Therefore, the result after aging is still the same as for the synthetic solution containing S₂O₃²⁻, where more solid sulfur residues are seen at the bottom of the glass bottle, while the solution is colorless.

4HSO₃⁻ + S₂O₃²⁻ + 2H⁺ = 2S₃O₆²⁻ + 3H₂O (9)

Interestingly, the colors of individual S²⁻, SO₃²⁻, and HSO₃⁻ synthetic solutions do not change after reacting with concentrated sulfuric acid under the used concentration in this study. But the S²⁻ + SO₃²⁻(5#), S²⁻ + HSO₃⁻(6#), and S²⁻ + SO₃²⁻ + HSO₃⁻(11#) samples produce turbid yellow sulfur sol and then form solid sulfur elements after reacting with concentrated sulfuric acid. It can be seen from Eqs. (9), (10) that S²⁻ reacts with sulfite (SO₃²⁻ or HSO₃⁻) to generate S₂O₃²⁻ and S₂O₃²⁻ then reacts with the subsequently added concentrated sulfuric acid to generate sulfur. Using the synthetic solution, we verified that S₂O₃²⁻ could be produced from the reaction of S²⁻ and sulfite (S₂O₃²⁻ or HSO₃⁻). Specific reaction conditions are given in the Supplementary Material. When 20 g/L S²⁻ reacted with 20 g/L HSO₃⁻ or 20 g/L SO₃²⁻ for 1 h, the S₂O₃²⁻ content is found to be 15.65–24.03 g/L. The results of Siu and Jia [15] also provided the same conclusion, which showed that the concentration of S₂O₃²⁻ was as high as 0.00053–0.00054 M when 0.0121 M Na₂S and 7.94 × 10⁻⁴ M Na₂SO₃ reacted at pH 7 and 20 °C for 600–800 s. Steudel [16] also reported that the reaction of Na₂S and Na₂SO₃ produced sulfur sol under the action of concentrated sulfuric acid and proved that sulfur sol contained S₂O₃²⁻ and long-chain polythionates S_mO₆²⁻ in addition to sulfur elements. This explains why the S²⁻ + HSO₃⁻+ S₂O₃²⁻ sample (14#) turns into a turbid yellow solution instead of a transparent yellow solution after adding concentrated sulfuric acid, although it contains both HSO₃⁻ and S₂O₃²⁻.

2S²⁻ + 4SO₃²⁻ + 6H⁺ = 3S₂O₃²⁻ + 3H₂O (10)

2S²⁻ + 4HSO₃⁻ + 2H⁺ = 3S₂O₃²⁻ + 3H₂O (11)

In general, S₂O₃²⁻ reacts with concentrated sulfuric acid to form sulfur impurity, which is the direct cause of the yellowing of the ammonium sulfate product. Sulfite (SO₃²⁻, HSO₃⁻) causes the formation of S₂O₃²⁻ by reacting with S²⁻ or the generation of polythionates by reacting with S₂O₃²⁻. The intermediate products, thiosulfate and polysulfate, are unstable in acidic media resulting in the precipitation of sulfur. In addition, if the oxidation effect in the ammonia desulfurization process is not optimal, a high sulfite concentration in the mother liquor is absorbed after oxidation, causing concerns such as limited output or an inadequate amount of the ammonium sulfate product. Therefore, it is necessary to find a way to efficiently purify S₂O₃²⁻ impurities in the mother liquor while removing sulfite (SO₃²⁻, HSO₃⁻), which is an effective measure to solve the abnormal color of ammonium sulfate products. Because the S²⁻ content in the absorption solution is 140–380 times lower than that of S₂O₃²⁻ or sulfite (SO₃²⁻, HSO₃⁻), the change in S²⁻ is not studied in the following purification experiments. The subsequent experiments focus on the purification effect of S₂O₃²⁻ and sulfite.

3.2. Effect of different factors on ion degradation efficiency in the solution under ultrasonic-assisted ozone treatments

Based on the observations of studies presented in section 3.1, it is learned that we need to ensure that S₂O₃²⁻ and sulfite (SO₃²⁻, HSO₃⁻) are removed to the possible extent, which is an essential factor in avoiding sulfur generation. Smelters oxidize sulfite when concentrated sulfuric acid is added to the actual production process, which is not feasible due to sulfur production. It is better to convert thiosulfate and sulfite in the inadequate mother liquor into stable sulfite using a non-acidic oxidizing agent. Gaseous oxidizing agents such as oxygen (O₂) or ozone (O₃) are suitable choices. However, oxygen is inefficient as an oxidizer, resulting in a poor-quality ammonium sulfate product [19]. Some studies reported that using metal catalysts can improve the oxidation effect of oxygen on sulfite but cannot meet the needs of the industry, especially for solutions containing high concentrations of sulfite [20], [21].

Ozone, with a redox potential of 2.08 eV [22], is possibly a gas oxidant with extreme oxidizing performance. Its oxidizing ability exceeds potassium permanganate, hydrogen peroxide, and oxygen. Zheng et al. [23] studied the oxidation effect of O₃ and O₂ on sulfite and proved the superiority of O₃. The oxidizing effect of O₃ relies on the O₃ molecule itself and the oxidizing radicals it generates [24]. However, utilizing ozone alone to oxidize refractory or high-concentration pollutants shows low oxidation efficiency and is time-consuming because of its low solubility in solution, low mass transfer efficiency, and low free radical yield [25]. Increasing the flow rate of ozone gas can improve its oxidation effect while increasing the cost. Some studies on the O₃ oxidation of sulfites showed that O₃ oxidation efficiency could be enhanced by generating more oxidizing radicals through electrochemical catalysis [26] and photocatalysis [27]. Still, the economics of these methods and the stability of results from large-scale tests are not suitable for industrial applications. Ultrasound (US) has been shown to have an enhanced effect when combined with O₃ for oxidation [28]. The cavitation effect of ultrasound and mechanical agitation enhances the oxidation effect by strengthening gas–liquid mass transfer and promoting the production of oxidizing radicals [29].

The advanced oxidation method employing O₃ and US together is commonly used in research on treating organic pollutants in water and has achieved remarkable results [30]. However, no study carried out the simultaneous oxidative removal of S₂O₃²⁻ and sulfites from solution employing ultrasound combined with ozone and given a detailed analysis of the enhancement mechanism. Therefore, we performed an experimental investigation on the ability of ultrasound and ozone to oxidize S₂O₃²⁻ and sulfites in real solutions simultaneously.

3.2.1. Effect of reaction temperature

The experiments to oxidize the real solution of mother liquor employing the ultrasonic-assisted ozone method were conducted with 500 mL solution, 800 RPM mechanical stirring rate, 0.8 L/min ozone gas flow, 400 W·L⁻¹ ultrasonic power density, and 20 kHz ultrasonic frequency. Titration was used to determine the amount of S₂O₃²⁻ and sulfites in the samples collected at various temperatures (20, 30, 40, and 50 ℃) for different times (0, 0.5, 1, 1.5, 2, and 4 h). The results are shown in Fig. 3.

Fig. 3.

Effect of reaction temperature on the removal degree of (a) S₂O₃²⁻ and (b) sulfite. (500 mL solution, 800 RPM mechanical stirring rate, 0.8 L/min ozone gas flow, 400 W·L⁻¹ ultrasonic power density, 20 kHz ultrasonic frequency).

Under the ultrasonic-assisted ozone process, raising the temperature improves S₂O₃²⁻ and sulfite removal efficiency slightly. This effect is more significant for sulfite, as shown in Figs: 3a and 3b. At a low temperature (20 °C), the high viscosity of the solution, the decelerated mass transfer of ozone, and the lack of free radicals produced by chemical mechanisms inside cryogenic bubbles [31] result in low oxidative capacity. Increasing the temperature reduces the viscosity of the solution, increases the ozone mass transfer, and expedites the transformation of ozone molecules into oxidizing radicals, thereby enhancing the oxidation effect [28]. As a result, the removal degree of S₂O₃²⁻ increases from 62.79 to 74.42%, and that of sulfites increases from 36.69 to 80.22% when the reaction temperature rises from 20 to 40 °C for a constant reaction time of 2 h. However, when the temperature increases from 40 to 50 °C, the effect of temperature on the degree of S₂O₃²⁻ and sulfite removal is insignificant; instead, their removal degree is slightly reduced. This is because ozone solubility strongly depends on temperature. The ozone concentration in water decreases from 6.1 to 3.8 mM when the temperature increases from 40 to 50 °C [32], [33], but the peak temperature of cavitation bubbles decreases [34], [35]. Dehane et al. [36] showed that the rate of oxidative species produced by ultrasound slowed down or decreased after the solution temperature was >40 °C. Therefore, the optimal reaction temperature is chosen to be 40 °C.

3.2.2. Effect of ultrasonic power density

Here, the ultrasonic power density denotes the magnitude of ultrasonic power acting per unit volume of solution. To study the effect of ultrasonic power density, the frequency of ultrasound was fixed at 20 kHz. Ultrasonic-assisted ozone oxidation experiments were carried out on 500 mL of real ammonium sulfate solution at 40 °C for 2 h with a gas flow rate of 0.8 L/min and a mechanical stirring speed of 800 RPM at various ultrasonic power densities (200, 400, 600, 800, and 1000 W·L⁻¹). The sample was collected at 2 h of the oxidation reaction, and titration was used to determine the S₂O₃² and sulfite contents and the corresponding removal degree. The results are shown in Fig. 4.

Fig. 4.

Effect of ultrasonic power density on the removal degree of S₂O₃²⁻ and sulfites from the real solution (500 mL solution, 800 RPM mechanical stirring rate, 0.8 L/min ozone gas flow, 2 h reaction time, 40 °C reaction temperature, 20 kHz ultrasonic frequency).

Increased ultrasonic intensity improves the reaction system's mechanical stirring and cavitation effects. The mass transfer efficiency for ozone [37] and the reaction rate constantly increased with an increase in ultrasonic density due to the mechanical impact of stirring [38]. The primary mechanism of the sonochemical reaction is the generation of highly active free radicals by the instantaneous collapse of cavitation bubbles under ultrasound irradiation [39]. The increase of ultrasonic power density can increase pressure and temperature in the bubble collapse process [40], increasing the content of oxidation free radicals in the solution due to the cavitation effect [41]. This will also improve the oxidation and removal degree of S₂O₃²⁻ and sulfite per unit time [42], [43]. In addition, the increase in ultrasonic intensity enhances the solubility and dispersion of ozone in an aqueous solution, further improving the mass transfer efficiency of ozone [38]. Therefore, when the ultrasonic power density increases from 200 to 800 W·L⁻¹, the degree of S₂O₃²⁻ and sulfite removal rises rapidly at the ultrasonic power density of 200–400 W·L⁻¹ and then increases slowly at 400–800 W·L⁻¹. However, increasing the ultrasonic power density has a limited advantage. When the ultrasonic power density reaches a threshold value, its intensifying effect in the solution reaches saturation, and the number of free radicals created and dissolved from ozone gas in the solution reaches a saturation point. At this stage, the impact of increasing ultrasonic power density on the oxidation process becomes invariable, and the degree of S₂O₃²⁻ and sulfite removal no longer rises rapidly. Also, excessive ultrasonic power increases the temperature, thereby reducing ozone solubility and increasing ozone bubble coalescence, which is unfavorable for ion removal [44]. Moreover, the cavitation bubbles grow very large in the negative phase of the sound wave, resulting in the insufficient collapse of the cavitation bubbles and acoustic screen; the degassing rate of O₃ gas becomes higher, giving rise to lower concentrations of O₃ or radicals [38], [45]. Therefore, with an increasing ultrasonic density from 800 to 1000 W·L⁻¹, the degree of S₂O₃²⁻ removal grows gradually, and the degree of sulfite removal tends to decrease. If only the oxidation effect is considered, controlling the ultrasonic intensity in the 400–800 W·L⁻¹ range is appropriate.

In this work, the effective calorimetric power ($P_{\mathit{cal}}$) [46], [47], [48] is used to quantify the energy transfer efficiency of the sonochemical reactor. The formula for calculating $P_{\mathit{cal}}$ is given in Eq. (12).

$$P_{\mathit{cal}}=C_P\times M\times dT/dt$$ (12)

Where, $C_P$ is the specific heat capacity of the corresponding substance, $M$ is the mass of the substance, and $\mathit{dT}/dt$ is the calorimetric calibration curve.

Without any cooling, 500 mL of the real ammonium sulfate solution (equivalent to 1.25 kg of ammonium sulfate mass) was subjected to different ultrasonic intensities (200, 400, 600, 800, and 1000 W·L⁻¹) for 10 min. The temperature rise was recorded every 30 s. Fig. 5a shows the calorimetric calibration curves of ultrasound in the real ammonium sulfate solution used to obtain the heating rate of ammonium sulfate at different ultrasound intensities. The specific heat capacity $(C_P)$ of the ammonium sulfate solution is 1.417 kJ kg⁻¹ °C⁻¹. The power determined from the calorimetric method can be used to find the system's energy efficiency by taking the ratio between the power dissipated in the system and the supplied electrical power [49]. According to the values of $\mathit{dT}/dt$ recorded under various ultrasonic intensities, the energy efficiency ($\eta =\frac{P_{\mathit{cal}}}{P_E}\times 100\%$) and energy loss ($1-\eta$) of the associated ultrasonic power density are computed using Eq. (12) and shown in Fig. 5b. The results show that the energy loss is the lowest when the ultrasonic power density is 200 W·L⁻¹, but the oxidation rate of S₂O₃²⁻ and sulfite is low (Fig. 4). When the ultrasonic power density increases from 400 to 800 W·L⁻¹, the oxidation effect enhances slowly, but at the same time, the energy loss also increases. In other words, the improvement in ultrasonic energy utilization efficiency is not directly proportional to the improvement in the oxidation effect within the ultrasonic intensity range of 400–800 W·L⁻¹. Therefore, comprehensively considering the oxidation effect and energy efficiency, the optimized ultrasonic power density is selected to be 400 W·L⁻¹.

Fig. 5.

Calorimetric method to determine the energy transfer efficiency. (a) Calorimetric calibration curves of ultrasound in ammonium sulfate medium; (b) energy efficiency and loss under different ultrasonic power densities.

3.2.3. Effect of ozone gas flow rate

To study the effect of the ozone gas flow rate in oxidizing the real solution of mother liquor using the ultrasonic-assisted ozone method, experiments were conducted at a solution volume of 500 mL, a mechanical stirring speed of 800 RPM, an ultrasonic power density of 400 W·L⁻¹, and an ultrasonic frequency of 20 kHz. Titration was used to determine the amount of S₂O₃²⁻ and sulfites in the samples collected at various reaction durations (0.5, 1, 1.5, 2, and 4 h) and ozone gas flow rates (0.4, 0.6, 0.8, and 1.0 L/min). The experimental findings are shown in Fig. 6.

Fig. 6.

Effect of ozone gas flow rate on the removal degree of (a) S₂O₃²⁻ and (b) sulfite. (500 mL solution, 800 RPM mechanical stirring rate, 40 °C reaction temperature, 400 W·L⁻¹ ultrasonic power density, 20 kHz ultrasonic frequency).

An increase in the ozone gas flow rate raises the concentration of ozone molecules in the solution, and correspondingly, the concentration of oxidative radicals created in the solution is enhanced. In addition, a higher gas flow rate might intensify the stirring action when the gas stream penetrates the solution. The introduction of ultrasound can promote the formation of cavitation bubbles and create exceptional reaction conditions like high temperature and pressure when the bubbles collapse [50], [51], thereby accelerating the mass transfer of ozone to the solution and enhancing the removal of S₂O₃²⁻ and sulfite. As shown in Fig. 6a and 6b, when the reaction time is 2 h and the ozone gas flow rate is increased from 0.4 to 0.8 L/min, the S₂O₃²⁻ removal degree increases from 51.16 to 74.4%, and the sulfite removal degree increases from 32.74 to 80.2%. Under the influence of ultrasound, ozone gas diffuses into the mother liquor as microbubbles at a proper flow rate. However, a high gas flow rate encourages the development of large ozone bubbles and generates severe turbulence in the solution. The gas–liquid contact area of large bubbles is lower than that of microbubbles.

Weavers [52] demonstrated that ultrasound could crush O₃-containing bubbles of size 0.5 to 1.0 cm into “microbubbles” of size 0.2 to 0.3 μm in aqueous solutions, increasing the total reaction surface area by a factor of 10³ to 10⁴. Under intense turbulence, large bubbles overflow from the liquid phase, resulting in a short ozone-liquid contact period and a poor mass transfer rate [44], [52]. Pflieger et al. [53] found continuous bubbling strongly perturbs the solution at high gas flow conditions. Wood et al. [54] found that the collapse strength of bubbles would increase with the increase of flow, thus improving the experimental effect. However, when the flow continues to increase, it has a negative impact on the experimental results. It has been noted that excessive ozone flow is adverse to the elimination effect of contaminants [55], and the same phenomenon is observed in our study. When the ozone gas flow rate increases to 1.0 L/min, the excessive ozone gas flow rate lowers the residence time of the ozone gas in the solution, shortens the contact time between ozone and sulfur-containing ions, and reduces the mass transfer effect. Therefore, the removal degree of S₂O₃²⁻ and sulfites are reduced. Thus, the optimal ozone gas flow is determined to be 0.8 L/min.

When the ozone gas flow is 0.8 L/min, although the removal effect of sulfite reaches the requirements of actual production, the removal degree of S₂O₃²⁻ is only 74.4% after treatment for 2 h. After 2.5 h of ultrasonic-assisted ozone oxidation, the remaining concentration of S₂O₃²⁻ and sulfites in the solution is 2.07 and 5.93 g/L, respectively, and the degree of removal are 91.39 and 90.83%, respectively. The quality of the mother liquor fully meets the production requirements. Therefore, the optimal experimental conditions are 40 °C temperature, 0.8 L/min ozone flow rate, 400 W·L⁻¹ ultrasonic power density, 20 kHz ultrasonic frequency, and 2.5 h reaction time.

3.2.4. Characterization of solid products

The pH of the solution (Fig. 7a) treated under optimal experimental conditions was adjusted to 8 and then evaporated and crystallized at 85 °C with a negative pressure of −0.07 MPa by a rotational vacuum device to obtain a white product, as shown in Fig. 7b. The XPS analysis result of the white product (Fig. 7c) displays a prominent S2p peak at 167.48–168.48 eV correspond to SO₄²⁻ and the peak intensity is more significant than that in the yellow ammonium sulfate product (Fig. 1a). It can be noted that, compared with the concentrated sulfuric acid oxidation method, the ultrasonic synergistic ozonation technology not only solves the yellowing of ammonium sulfate product but also obtains products with higher purity. It is worth noting that the S2p peak is also detected at a binding energy of 162.8–163.8 eV, which corresponds to polysulfate S₄O₆²⁻. In the case of insufficient oxidant, S₄O₆²⁻ is the primary intermediate in the oxidation reaction of S₂O₃²⁻. The presence of S₄O₆²⁻ is detected in the white product obtained because the amount of ozone is insufficient to convert all S₂O₃²⁻ into SO₄²⁻, resulting in the partial conversion of S₂O₃²⁻ to the intermediate product S₄O₆²⁻. The chemical reactions might be as follows:

S₂O₃²⁻ + 2O₃ = SO₄²⁻ + SO₂ + 1.5O₂ (13)

S₂O₃²⁻ + 2O₃ + H₂O = 2SO₄²⁻ + O₂ + 2H⁺ (14)

S₂O₃²⁻ + O₃ = 2/3SO₄²⁻ + 1/3S₄O₆²⁻ + 2/3O₂ (15)

SO₃²⁻ + O₃ = SO₄²⁻ + O₂ (16)

Fig. 7.

Characterization of samples treated under optimal conditions. (a) The solution obtained after ultrasonic-assisted ozone treatment; (b) the solid product obtained after crystallization; (c) the XPS of the solid product.

The stability of S₄O₆²⁻ in acidic media and at high temperatures is maximum for polysulfates [56], and S₄O₆²⁻ was stable when the pH of the treated solution was adjusted to 1 in our experiments, as shown in Fig. S2. In addition, the pH of the mother liquor was only reduced from the initial value of 5 to 4–4.5 after ultrasonic-assisted ozone treatment. Therefore, sulfur impurities are not observed after treatment with ultrasound-assisted ozone, resulting in a transparent colorless solution and a white ammonium sulfate product. The decrease in pH value after ultrasonic-assisted ozone treatment is because the ozone reaction with S₂O₃²⁻ produces protons (H⁺) or acidic gas (SO₂), as per reactions (13), (14). In addition, the XPS spectrum (Fig. S3) exhibits a peak at 162.88 eV designated to the analytically pure (NH₄)₂S₄O₆ supports the above idea [57]. The presence of sulfate is also seen in Fig. S3, which may be produced by the oxidation of (NH₄)₂S₄O₆ to (NH₄)₂SO₄ during the XRD analysis [58].

3.3. The enhancement mechanism of the ultrasonic field

The advantages of ultrasonic-assisted ozone technology can be comprehended by comparing it with the ozone technology alone. In addition, the strengthening mechanism of ultrasound is also studied and discussed in this section.

3.3.1. Comparison of the O₃ alone and US/O₃ techniques

The effects of O₃ alone, ultrasound (US) alone, and US/O₃ on the oxidation of thiosulfate and sulfites are compared in Fig. 8. These experiments were conducted at a solution volume of 500 mL, a temperature of 40 °C, an ozone flow rate of 0.8 L/min, a mechanical stirring speed of 800 RPM, an ultrasonic power density of 400 W·L⁻¹ and an ultrasonic frequency of 20 kHz for different reaction times (0.5, 1, 1.5, 2, 2.5, 3, 3.5 and 4 h).

Fig. 8.

Comparing the O₃ alone, US alone, and US/O₃ on the oxidation of thiosulfate and sulfites.

It can be seen from Fig. 8 that the oxidation effect of ultrasound (US) alone on thiosulfate and sulfites was negligible, which may be due to the formation of very few hydroxyl radicals from water. But using ultrasound enhances the oxidation effect of ozone, which is more prominent for both S₂O₃²⁻ and sulfites. The oxidizing power of the ozone molecule and the oxidizing radicals produced by ozone breakdown determine the oxidizing effect of ozone. The mechanical effect of ultrasound and the cavitation effect simultaneously promote ozone oxidation in two aspects. The mechanical effect can efficiently and uniformly disperse the ozone gas in the solution throughout the reactor and encourage the creation of ozone microbubbles. This increases the gas–liquid contact area and, in turn, boosts the mass transfer and enhances the oxidation effect of ozone molecules. In addition, microbubble creation can be promoted, thereby increasing the cavitation effect compared to large bubbles produced in the absence of ultrasound. The cavitation effect of ultrasound also encourages the formation of more oxidative radicals from ozone through chain reactions (Eqs. (17)-(23)) [33], [59], [60]. Together, these two advantages of ultrasonography aid in improving the oxidation effect of ozone. After treatment for 2.5 h, the removal degree of S₂O₃²⁻ and sulfite are as high as 91.39 and 90.83%, respectively, when ultrasonic-assisted ozone is used, and only 80.36 and 82.31%, respectively, when ozone alone is used. The ultrasonic-assisted ozone process enhanced the removal degree of S₂O₃²⁻ and sulfites by 11.03 and 8.52%, respectively, compared to the ozone process alone. In addition, with a reaction time of 2.5 h, the concentration of the two remaining ions in the solution obtained by ultrasonic co-ozone technology meets the appropriate requirements of the actual production of the enterprise. In contrast, ozone technology alone requires at least 30 min of further treatment to meet identical conditions.

O₃(g) → O₃(aq) (17)

H₂O + ))) → H•+•OH (18)

O₃ + ))) → ¹O₂ + O₂ (19)

H•+O₂ + ))) → HO₂• (20)

HO₂•+))) → •O₂⁻+H⁺ (21)

¹O₂ + H₂O + ))) → 2•OH (22)

¹O₂ + O₃ + ))) → 2O₂ (23)

3.3.2. Analysis of different oxidation components

The DMPO spin-trapping EPR technique was employed to verify further the generation of free radicals by using 5,5-dimethyl-1-pyrroline N-oxide (DMPO) and 2,2,6,6-tetramethylpiperidine (TMP) as spin traps [61]. •OH and •O₂⁻ can react with DMPO to form DMPO-•OH and DMPO-•O₂⁻ adducts, and ¹O₂ can react with TEMP to form TEMP-¹O₂ adducts [62]. Fig. 9 displays the EPR results of the free radical generation in the real and synthetic solutions under the ultrasonic-assisted and ozone-alone processes. The real solution is the solution to be treated, and the synthetic solution is prepared with the same concentration of ammonium sulfate but without the S₂O₃²⁻ and sulfite components.

Fig. 9.

EPR analysis of solutions under different conditions. (a) Analysis results of •OH; (b) analysis results of •O₂⁻; (c) analysis results of ¹O₂.

For both solutions, the distinctive peaks corresponding to the three radicals under ozone-alone treatment are weak, while these characteristic peaks are significant under ultrasound irradiation. This phenomenon is quite evident in the synthetic solution, suggesting that the cavitation action of ultrasonic waves stimulates the production of free radicals in the solution. By comparing the spectra of the real solution with those of the synthetic solution, it is found that the peaks of the three radicals detected in the real solution are considerably smaller than those in the synthetic solution. This may be because the oxidizing radicals generated under either the ultrasonic-assisted ozone process or the ozone-alone process are consumed in the oxidation of S₂O₃²⁻ and sulfites in the real solution. In contrast, the synthetic solution has a higher content of radicals due to the absence of S₂O₃²⁻ and sulfites. Moreover, qualitatively, the characteristic peak of ¹O₂ is the most prominent among the three types of radicals. The difference in the height of the ¹O₂ peak observed for the real solution and the synthetic solution is apparent, which suggests that ¹O₂ participates in the actual oxidation process more than the other two radicals. This observation will be further manifested in the following free radical shielding experiments.

The shielding experiments of free radicals are used to quantitatively analyze the degree of contribution of different oxidation components in the solution. Tert-butanol (TBA), furfuryl alcohol (FFA), and superoxide dismutase (SOD) were utilized as scavengers for •OH, ¹O₂, and •O₂⁻ [63], [64]. Experiments were conducted under previously optimized conditions, i.e., 40 °C reaction temperature, 800 RPM mechanical stirring rate, 0.8 L/min ozone gas flow rate, 400 W·L⁻¹ ultrasonic power density, and 20 kHz ultrasonic frequency (no ultrasonic waves were introduced in the ozone treatment alone experiment). Various combinations of radical shielding were added to the reaction solution to shield different kinds of radicals. The solutions were titrated for a reaction time of 2 h to examine and calculate the S₂O₃²⁻ and sulfite contents. The results obtained are tabulated in Table S1 in the Supplementary Material. From Table S1, the degree of involvement of each oxidation component in the oxidation process under ultrasonic-assisted ozone and ozone-alone treatments is computed, and the findings are shown in Fig. 10. The figure makes it abundantly clear that the process of oxidation of ozone can be broken down into two distinct categories: the first is the direct oxidation of ozone molecules, and the second is the inductive oxidation of ozone molecules, which relies on chain reactions to produce a wide range of oxidizing radicals. These two conclusions apply to both the ultrasonic-assisted ozone and the ozone-alone treatments. The direct oxidation of ozone is always the most significant in either the single ozonation or the ultrasonic-assisted ozonation process because the concentration of ozone molecules is significantly more than that of free radicals during the entire oxidation process. In addition, the half-life of ozone in aqueous solution at 15-35℃ is reported to be approximately 8–30 min [65], while the half-life times of •OH, ¹O₂, and •O₂⁻ in aqueous solution are only 10⁻¹⁰, 10⁻⁶, and 10⁻⁶ s, respectively [66]. The involvement of each radical in the oxidation process is more in the ultrasonic-assisted ozone process than in the azone-alone process, as found from the participation experiments of various oxidation components. This again illustrates that ultrasound's reinforcing action boosts the ozone bubbles' cavitation effect, ultimately generating additional free radicals. Throughout the reaction, the degree of involvement of free radicals is essentially the same as the ordering of the magnitude of their standard oxidation potential. The order is found to be •OH(2.7 V) > ¹O₂(2.2 V) > •O₂⁻(0.89 V) [67], [68].

Fig. 10.

The participation order of oxidation components in the removal of (a) S₂O₃²⁻ in the US + O₃ process; (b) Sulfites in the US + O₃ process; (c) S₂O₃²⁻ in the O₃ process; (d) Sulfites in the O₃ process.

In conclusion, we can see more clearly from Fig. 11 how ultrasound can strengthen the whole oxidation process, thus improving the removal degree of S₂O₃²⁻ and sulfites.

Fig. 11.

Mechanistic diagram of the enhanced ozone oxidation process by ultrasound.

3.4. Economic evaluation

We know from the actual production process that the main cost of treating ammonium sulfate solution with ultrasonic-assisted ozone technology comprises gas source cost (for ozone production) and electricity cost (including ozone machine and ultrasonic generator) without considering the initial equipment cost. According to the actual data in China, the selling price of industrial-grade ammonium sulfate is 186 USD/t, the cost of oxygen is 0.055 USD/Nm³, and the cost of industrial electricity is 0.05 USD/kW•h. After the above-optimized conditions, the treatment cost per ton of ammonium sulfate is 176 USD. For the annual output of 8,000 tons of ammonium sulfate from non-ferrous smelting enterprises, the yearly profit can reach 80,000 USD, with considerable economic benefits. In addition, the cost of adding concentrated sulfuric acid and ammonia in the traditional process can also be eliminated.

The treatment of ammonium sulfate solution by ultrasonic-assisted ozone technology can not only save the economy but also improve the operating environment and simplify the operation process. Therefore, it has vast application prospects.

4. Conclusion

Ammonium sulfate is one of the leading products of sulfur dioxide flue gas, which is vital in realizing resource recovery, utilization of sulfur resources, and environmental protection. In this study, given the abnormal chromaticity of yellow ammonium sulfate produced by smelting flue gas, we studied the real mother liquor and synthetic solution. It was revealed that the presence of S₂O₃²⁻ and sulfite in the mother liquor reduced the quality of ammonium sulfate products. Mainly, S₂O₃²⁻ produced sulfur impurities under the action of concentrated sulfuric acid, an oxidant used in actual production, thus resulting in yellow ammonium sulfate products. Subsequently, we provided a highly efficient oxidation technology using ultrasound-assisted ozone. The experimental results showed that after 2.5 h of the oxidation reaction at 40 °C, 0.8 L/min ozone flow rate, 400 W·L⁻¹ ultrasonic power density, and 20 kHz ultrasonic frequency, the concentrations of S₂O₃²⁻ and sulfites in the solution were 2.07 and 5.93 g/L, respectively. The degree of removal was 91.39 and 90.83%, respectively, for the S₂O₃²⁻ and sulfites. The quality of the treated solution met the production requirements, and a white ammonium sulfate product was obtained after evaporation and crystallization. The comparison experiments showed that the decolorization effect of US/O₃ was significantly better than that of O₃; it was more advantageous in reducing the reaction time. In addition, O₃ was substantially more effective for decolorization than the free radicals in the US/O₃ process. The degree of involvement of the oxidation component in the whole oxidation process was O₃ ≫ •OH > ¹O₂ > •O₂⁻. Ultrasonic-assisted ozone technology can bring good economic benefit to the upgrading of ammonium sulfate product.

In summary, this work improved the quality of the ammonium sulfate product by advanced oxidation of the ammonium sulfate solution by the US/O₃ method and avoided sulfur impurities produced by the traditional concentrated sulfuric acid oxidation method. The proposed methodology in this work is of practical significance for the sustainable development of the ammonium sulfate industry and the metal smelting industry, increasing the value of sulfur resource recovery and realizing the concept of environmental protection.

CRediT authorship contribution statement

Tian Wang: Formal analysis, Writing – original draft. Hongtao Qu: Writing – review & editing. A.V. Ravindra: Writing – review & editing. Shaobin Ma: Writing – review & editing. Jue Hu: Validation. Hong Zhang: Formal analysis. Thiquynhxuan Le: Conceptualization, Project administration, Writing – review & editing. Libo Zhang: Resources, Supervision.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following are the Supplementary data to this article:

Data availability

Data will be made available on request.