Simultaneous removal of SO₂ and NO_x by potassium-modified carbide slag

2.1 Adsorbent preparation

In this experiment, the modified CS was prepared by ultrasonic impregnation, and its visual preparation flowchart is shown in Fig. 1(a). The CS samples were immersed in the solutions of KOH, K₂CO_3, and KHCO₃, each at a concentration of 1.0 mol·L^-1, respectively, and sonicated at 100 W and 59 Hz for 15 min. Subsequently, the resulting solution was dried in a blast drying oven at 100 °C for 12 h. Finally, the prepared modified CS was sieved using a particle size range of 40–60 mesh. The initial carbide slag was recorded as CS, the potassium-modified calcium carbide slag was recorded as KCS, and the different potassium-modified carbide slags were recorded as CS-KOH, CS-K₂CO₃, and CS-KHCO₃. Following the reaction, the modified carbide slags were further recorded as UCS-KOH, UCS-K₂CO₃, and UCS-KHCO₃.

Close

Fig. 1

(a) Preparation process of adsorbents; (b) schematics of the experimental setup: (1) Gas cylinder; (2) mass flowmeter;(3) gas mixing tank; (4) control valve; (5) resistance tube furnace; (6) quartz tube; (7) temperature controller; (8) flue gas analyzer; (9) K₂MnO₄ solution; (10) NaOH solution.

Export to PPT

2.2 Activity evaluation

Desulfurization and denitrification reactions were conducted in a fixed bed using 1.0 g of modified CS within a quartz tube reactor with an inner diameter of 6 mm. The performance of the modified CS in terms of SO₂/NO removal was assessed using the experimental setup illustrated in Fig. 1(b). To simulate real conditions, a dynamic gas distribution was adopted, setting the flow rate of the simulated flue gas was 200 mL·min⁻¹, and the gas hour space velocity (GHSV) at 8488 h⁻¹. The flue gas composition consisted of 1500 mg·m⁻³ of SO₂, 700 mg·m⁻³ of NO, and 5 % of O₂, with N₂ serving as the equilibrium gas. To monitor the SO₂/NO concentration in the flue gas, an online MRU infrared flue gas analyzer was employed. Any escaping SO₂/NO from the experiment was subsequently oxidized and absorbed by an absorption solution comprising NaOH and K₂MnO₂ before discharge. Eqs. (1) and (2) represent the removal efficiency and adsorption capacity of SO₂/NO, respectively. In accordance with the national standard (Emission standard, 2014), a denitrification rate of 45 % and desulfurization rate of 75 % were employed as benchmarks for measuring effective removal. Denitrification efficiency was evaluated using T_NO45%, indicating the duration of denitrification with a denitrification rate exceeding 45 %. Similarly, T_SO275% was utilized to evaluate the desulfurization efficiency signifying the duration of desulfurization with a desulfurization rate above 75 %.

where η denotes the SO₂/NO removal rate (%), C_in denotes the SO₂/NO concentration before the reaction (mg·m⁻³), C_out denotes the SO₂/NO concentration after the reaction (mg·m⁻³), q_e denotes the adsorption capacity of SO₂/NO (mg·g⁻¹), Q is the gas flow rate (mL·min⁻¹), t₀ and t_b denotes the onset and end time of adsorption (min), respectively, and m denotes the material mass (g).

2.3 Characterization methods

The detailed characterization methods are shown in Supplementary Text.

3.1 Results of adsorbent activity evaluation

In the conducted experiment, the reaction temperature emerged as a critical factor affecting the activity of the modified CS. The impact of different reaction temperatures on the simultaneous desulfurization and denitrification activities of the modified CS is presented in Fig. 2. The adsorption capacities of potassium-modified CS for SO₂/NO at different temperatures are shown in Fig. 3. As shown in Figs. 2 and 3, the SO₂/NO removal activity and adsorption capacity of different modified CS exhibited temperature dependent variations. Notably, the denitrification efficiency of the modified CS was significantly higher than that of CS alone. The highest denitrification rate of CS-KHCO₃ was 62 %, whereas the denitrification performance of CS-KOH and CS-K₂CO₃ was improved to 50 % and 48 %, respectively. From the data in Fig. 2(a), the effect of the reaction temperature on the NO removal rate of the CS exhibited relatively minor fluctuations. In the temperature range of 100–250 °C, the NO removal rate remained below 45 %, and the NO adsorption capacity was limited. As the reaction temperature increased from 100℃ to 200 °C, the value of T_NO45 % increased with increasing temperature at 120 min for CS-KOH, 75 min for CS-K₂CO₃, and 120 min for CS-KHCO₃. During this period, the NO adsorption capacities of CS-KOH, CS-K₂CO₃, and CS-KHCO₃ were 2.48, 2.24, and 3.05 times higher than those of CS. However, as time progressed, NO removal decreased, possibly due to the coverage of active sites by the products after SO₂/NO adsorption (Guo et al., 2022). However, when the reaction temperature reached 250 °C, the NO removal rate of different modified CS decreased. This is because SO₂ has a higher boiling point than NO, making NO more prone to diffusion than SO₂ at lower temperatures (100 °C) (Silas et al., 2018). Molecules with higher boiling points have stronger van der Waals forces, whereas those with lower boiling points exhibit weaker intermolecular interactions. Consequently, NO can be readily displaced and desorbed by SO₂ at higher temperatures. Therefore, the feasible denitrification temperature of the modified CS was 200 °C.

Close

Fig. 2

Efficiency of SO₂/NO removal at different temperatures: (a)(b) CS; (c)(d) CS-KOH; (e)(f) CS-K₂CO₃; (g)(h) CS-KHCO₃. Reaction conditions: gas flow rate of 200 mL·min⁻¹; initial concentrations of SO₂ at 1500 mg·m⁻³, NO at 700 mg·m⁻³, and O₂ at 5 %.

Export to PPT

Close

Fig. 3

Adsorption capacity of potassium-modified CS for SO₂/NO at various temperatures.

Export to PPT

As illustrated in Fig. 2(b) and 3(a), the desulfurization rate of CS exhibited relatively minor fluctuations within the range of 100–250 °C, which proved that the effect of reaction temperature on the performance of CS for SO₂ removal was not predominantly evident. The optimal performance of CS in terms of SO₂ removal remained consistent with a 100 % desulfurization rate within 120 min, and its SO₂ adsorption capacity reached 36 mg·g⁻¹. From Fig. 2(d), (f), (h), and 3(a), the removal performance of the modified CS for SO₂ decreased slightly at temperatures below 200 °C. However, the T_SO275 % values for all three modified CS continued to increase with increasing in temperature. At 200 °C, the SO₂ adsorption capacities of the three materials are essentially equal. This can be attributed to the gradual increase in temperature, which reduced the mass transfer resistance and accelerated the reaction rate, leading to an enhanced SO₂ adsorption capacity (Li et al., 2021).

Table 1 summarizes the optimal reaction temperatures, removal efficiencies, and durations of the potassium-modified calcium carbide slag. It is evident that potassium-modified cs at 200 °C showed the most efficient and stable SO₂/NO removal. In summary, based on the desulfurization and denitrification rates, the removal capacity of the modified CS can be derived as follows: CS-KHCO₃ > CS-KOH > CS-K₂CO₃ > CS.

3.2 Activation energy of adsorbents

From the above experimental results, the desulfurization performance of the different modified CS exhibited slight variations. However, the denitrification performance was significantly different. Therefore, the reaction activation energy of the modified CS for SO₂/NO removal was determined. The reaction rate constant k in the rate equation of the catalytic oxidation can be expressed using the Arrhenius equation (Dou et al., 2020).

as follows:

where k denotes the reaction rate constant (mol·s⁻¹·g⁻¹); A is denotes the pre-exponential factor; R denotes the gas constant (8.314 J·mol⁻¹·K⁻¹); T denotes the reaction temperature (K); and Ea denotes the activation energy (kJ·mol⁻¹).

The rate equation of the NO catalytic oxidation of the modified CS is represented by Equation (4). Subsequently, Equation (4) is incorporated into Equation (3), resulting in Equation (5). Simultaneously, the logarithm of both sides of Equation (5) is taken, yielding Equation (6). The conversion rate of NO at different temperatures was measured while maintaining the inlet concentration constant, and the reaction rate r was calculated using Equation (7) (Lei et al., 2020). Using Equation (6), a plot can be constructed with T⁻¹ as the x-axis and lnr as the y-axis. This plot facilitates the derivation of a linear regression equation. The reaction activation energy (Ea) was determined by multiplying the slope of the fitted regression equation by with gas constant R (Zheng et al., 2018).

where r denotes the reaction rate (mol·s⁻¹·g⁻¹); η_NO denotes the NO conversion rate (%); C_NO denotes the NO inlet concentration (mg·m⁻³); Q denotes the total gas flow (L·s⁻¹); mdenotes the mass of the adsorbents (g); and V_m denotes the molar volume of the gas.

As shown in Fig. 4(a)–(d), the correlation coefficients R² of the linear regression equations for the NO activation energies of CS, CS-K₂CO₃, CS-KOH, and CS-KHCO₃ were 0.9692, 0.9385, 0.9429, and 0.9554, respectively. The calculated activation energies of NO for CS, CS-K₂CO₃, CS-KOH, and CS-KHCO₃ were −7.2446 kJ·mol⁻¹, 15.9048 kJ·mol⁻¹, 8.7301 kJ·mol⁻¹, and 6.7838 kJ·mol⁻¹, respectively. These results demonstrated a significant reduction in the activation energy of the modified CS, indicating higher reaction rates and enhanced catalytic oxidation of NO (Lei et al., 2020). In addition, the negative activation energy of CS suggested a decrease in the reaction rate r with increasing temperature, leading to a weakening of the catalytic oxidation of NO (Valverde 2015, Jayakumar et al., 2017, Joshi et al., 2018). These findings suggest that modified CS materials promote favorable reactions by effectively reducing energy barriers. Hence, the catalytic mechanism of the modified CS was further changed to improve the catalytic efficiency of NO. As shown in Fig. 4(e)–(h), the activation energy for the SO₂ reaction in CS- K₂CO₃ was 10.10 kJ·mol⁻¹, and it remained at 0 kJ·mol⁻¹ for all other materials. This indicates that K₂CO₃ modification slightly increased the activation energy for CS desulfurization, leading to a reduced reaction rate. Combined with Fig. 2, it became evident that the lower activation energy of the modified CS reaction promoted easier occurrence of the catalytic process and improved desulfurization and denitrification. The catalytic activities of the different modified CS materials can be ranked as follows: CS-KHCO₃ > CS-KOH > CS-K₂CO₃, which is consistent with their respective SO₂/NO removal performance.

Close

Fig. 4

Linear regression diagram of SO₂/NO surface activation energy for different materials: (a) (e) CS; (b) (f) CS-K₂CO₃; (c) (g) CS-KOH; (d) (h) CS-KHCO₃.

Export to PPT

3.3 Characterization of materials

According to the XRF characterization results, the main compositions of CS are shown in Table S1 (as oxides, wt %), such as CaO (96.96 %), K₂O (0.05 %), SiO₂ (1.20 %), Fe₂O₃ (0.80 %), and Al₂O₃ (0.44 %). To facilitate the comparison, the K₂O content was used to express the total potassium oxide (KO_x) content. Loading potassium onto the CS surface increased its potassium content, with K₂O reaching 22.92 %, 10.34 %, and 12.20 % for CS-K₂CO₃, CS-KOH, and CS-KHCO₃, respectively, whereas SiO₂ and Al₂O₃ experienced slight decreases. Combining these findings with Fig. 2, the addition of potassium can promote the oxidation efficiency observed in NO, which likely explained the increased denitrification efficiency observed in the modified CS. However, CS-K₂CO₃ exhibited poor effectiveness in desulfurization and denitrification, which can be attributed to the excessive blockage of pore channels and surface adsorption sites on the surface caused by excessive KO_x concentration resulting from K₂CO₃ addition(Yang et al., 2019). The SEM images in Fig. 5(a-d) illustrated the significant changes in the microstructure of CS before and after modification. Ultrasonic impregnation disrupted the CS structure, facilitating accelerated potassium loading on the CS surface. In addition, different potassium modifications yielded distinct microstructures. Fig. 5(a) displayed layered bulk particles of CS with a rough surface, few cracks, and no noticeable pore structure. Fig. 5(b) and 5(c) indicated irregular porous layered surfaces and bulk structures of CS-KOH and CS-K₂CO₃, respectively, after ultrasonic impregnation and potassium modification. The CS-KOH particles exhibited small honeycomb pores on the surface and a significant number of channels within the particles. In contrast, CS-K₂CO₃ displayed fewer small surface pores than CS-KOH, potentially due to pore channel blockage caused by excess KO_x (Yang et al., 2019). The microstructure of CS-KHCO₃ shown in Fig. 5(d) revealed a unique flower cluster-like ordered structure at high resolution, characterized by a loose and uniform appearance. This flower cluster-like microstructure could provide abundant volume and active sites to facilitate the mass transfer of desulfurization and denitrification reaction intermediates (Peng et al., 2021). These microstructural observations were closely linked to CS composition modifications and adherence of potassium to the adsorbent surface (Ding and Liu 2020).

Close

Fig. 5

(a–d) SEM, (e) XRD, (f) FTIR spectra, (g) O 1 s, (h) C 1 s, and (i) K 2p spectra of CS, CS-K₂CO₃, CS-KOH, and CS-KHCO₃.

Export to PPT

In Fig. 5(e), the XRD spectra of different potassium-modified CS exhibited characteristic peaks of Ca(OH)₂ and some CaCO₃. The modified CS also indicated certain diffraction peaks corresponding to K₂Ca(CO₃)₂, K₂O, and KO₂. From Fig. 5(e), the intensity of the Ca(OH)₂ diffraction peaks in CS-KHCO₃ and CS-K₂CO₃ was significantly reduced, whereas the intensity of the Ca(OH)₂ diffraction peaks in CS-KOH was higher than that in CS. This observation confirmed that the addition of KHCO₃ and K₂CO₃ reduced the crystallinity of CS, resulting in a more dispersed Ca(OH)₂ in the modified CS (Tang et al., 2015). Moreover, Ca(OH)₂ in CS played a crucial role in the reaction process and promoted the formation of desulfurization and denitrification products (Guo et al., 2022). However, the doping of KOH improved the crystallinity of CS. Notably, the XRD spectra in Fig. 5(e) revealed that the characteristic peaks of KO_x in CS-KHCO₃ were significantly more than those of CS-K₂CO₃ and CS-KOH, indicating that doped KHCO₃ can facilitate the formation of more KO_x, which played a critical role in the denitrification process of desulfurization (Li et al., 2022). However, in the modified CS, CaCO₃ easily combined with KO_x, to form a stable potassium phase containing calcium carbonate and K⁺. This interaction affected the activity of KO_x, thereby reducing the catalytic capacity of KO_x and the adsorption capacity of Ca(OH)₂ on sulfur oxides (Li et al., 2022). In Fig. 5(e), CS-K₂CO₃ exhibited numerous characteristic peaks of K₂Ca(CO₃)₂, resulting in a reduction in KO_x content, which may explain the poor desulfurization and denitrification performance of CS-K₂CO₃. As shown in Fig. 5(f), the FTIR spectra were employed to investigate the surface functional groups of CS before and after modification(CS, CS-KOH, CS-K₂CO₃, and CS-KHCO₃). The FTIR absorption peaks of the modified CS did not exhibit new appearances or disappearances compared to CS; instead, only the intensity of the modified CS peaks changed in the infrared spectrum, without the formation of new functional groups. A strong narrow peak at 3643 cm⁻¹ was observed, representing the O–H bond created by the vibrational pulse. This peak is attributable to the stretching vibration of O–H caused by surface hydroxyl groups and chemically adsorbed potassium water on the adsorbent surface (Cong and Mei 2021). The stretching vibration of the carbonyl C=O surface functional group was represented by an absorption peak at 1430 cm⁻¹ (Tang et al., 2013, Ding and Liu 2020). The O-CO- band in CO₃^2-, demonstrated peaks at 710 cm⁻¹ and 869 cm⁻¹, corresponding to its in-plane and out-of-plane bending vibrations, respectively (Li and Yi 2020, Ma et al., 2021). After potassium modification, the intensity of the C = O peaks increased due to the activation and oxidation properties of potassium, resulting in the generation of numerous oxygen-containing functional groups (Ding and Liu 2020). Zhao et al. (1994), demonstrated that an increase in oxygen-containing functional groups improved the adsorption capacity of SO₂ and NO. Interestingly, although CS and modified CS shared similar characteristics, CS-KOH and CS-KHCO₃ exhibited a stronger C=O stretching pattern than CS and CS-K₂CO₃, which indicated a higher concentration of oxygen functional groups in CS-KOH and CS-KHCO₃. In summary, the adsorption capacity of modified CS was closely related to the concentration of its surface organic functional groups.

The high-resolution XPS spectra of O 1 s, C 1 s, and K 2p for CS, CS-KOH, CS-K₂CO₃, and CS-KHCO₃ are shown in Fig. 5(g)–5(i). Table S2 lists the corresponding binding energies and their relative contents. In Fig. 5(g), the O 1 s spectra of the modified CS and the corresponding fitted curves were illustrated, which can be deconvoluted into two peaks. A unique peak at a binding energy of 531 eV corresponds to the formation of lattice oxygen (denoted as O_lat, such as O^2–) by the defective oxides present in the modified CS. Surface adsorbed oxygen (denoted as O_ads, such as O₂^2-, O^-, OH^–, CO₃^2-) in modified CS exhibited a characteristic peak in the range of 531.20 eV–532.90 eV (Panov et al., 2006, Liu and He 2010). It is widely recognized that O_ads possesses higher mobility and activity than O_lat, thereby accelerating the oxidation process through redox reactions (Liu et al., 2021). In addition, low coordination O_ads greatly facilitated the formation of surface oxygen groups, such as carbonyl (–COOH), surface-active chemisorbed oxygen (O₂^-, O^-, etc.), and surface hydroxyl groups (–OH). Table S2 illustrates that the O_ads content of CS was 14.93 %, and the loading of potassium resulted in higher peak area ratios of O_ads than CS for CS-KOH and CS-KHCO₃, which reached 18.54 % and 47.47 %, respectively. In contrast, the O_ads content of CS-K₂CO₃ was only 2.69 %, which could be attributed to the fact that K₂Ca(CO₃)₂ covered the CS surface and blocked the pore channels, consistent with the SEM results of CS-K₂CO₃. The content of O_ads followed the order: CS-KHCO₃ > CS-KOH > CS > CS-K₂CO₃, with CS-KHCO₃ showing the highest percentage of O_ads. This indicated that CS-KOH and CS-KHCO₃ possessed a higher content of surface hydroxyl and/or surface active chemisorbed oxygen. The increased O_ads content enhanced the oxidation ability of the modified CS at low temperatures, which aligned with the previous analysis (Ding and Liu 2020). Tiwari et al. (2018) concluded that O_ads indicated alkaline properties; thus, high levels of O_ads can be more favorable for the adsorption of acidic gases. The measured C 1 s spectra of the modified CS are shown in Fig. 5(h). The C 1 s peak was deconvoluted into three functional groups, including graphitic carbon (C–C, 284.8 ± 0.1 eV), carbonyl (C=O, 287.7 ± 0.2 eV), and carboxyl and/or ester groups (–COO, 289.6 ± 0.2 eV) (Li et al., 2021, Li et al., 2022). As shown in Table S2, CS- KHCO₃, CS-KOH, and CS-K₂CO₃ demonstrated lower C-O contents than CS, but higher C=O contents. The C=O content of CS was 11.17 %, and after modification, the C=O contents of CS-KOH, CS-K₂CO₃, and CS-KHCO₃ increased to 23.46 %, 14.41 %, and 21.35 %, respectively. This increase may be attributed to the basicity of potassium, which introduces defective or unpaired electrons on the CS surface. These electrons promote the formation of surface functional groups via the absorption of oxygen from the environment (Liu et al., 2014). This was particularly evident on the CS-KOH surface. Combining these observations with Fig. 2, it was evident that the adsorption capacities of CS-KHCO₃, CS-KOH, and CS-K₂CO₃ for SO₂/NO were superior to that of CS. This can be attributed to the higher C = O content in CS-KHCO₃, CS-KOH, and CS-K₂CO₃, which was favorable to improve the adsorption capacity of SO₂/NO. According to the literature(Zhang et al., 2017), the essential property of C = O Brønsted is the active center of SO₂/NO oxidation. As shown in Fig. 5(i), the K 2p spectrum presented two components at 292.72 eV and 295.42 eV, confirming the presence of K ions. The peaks of K 2p1/2 were observed at 295.40 eV and 295.71 eV, while the peaks of K 2p3/2 appeared at 292.47 eV and 293.14 eV, respectively, with evident differences in the relative intensities. The binding energies of the different valence states of K cations (K⁴⁺ and K⁺) with K 2p electrons were different. As observed in Fig. 5(d) and Table S2, the relative K⁴⁺ content on the surface of CS-KOH and CS-KHCO₃ was higher than that of CS-K₂CO₃. This is in accordance with the XRD results, indicating that CS-KOH doped with KHCO₃ promoted the formation of more KO₂. KHCO₃ loading led to the generation of oxygen vacancies on the modified CS surface and increased the O_ads content by absorbing oxygen from the environment, which was consistent with the O_ads content in the O 1 s spectrum. This may account for the higher effectiveness of CS-KHCO₃ in desulfurization and denitrification. It can be inferred that a certain amount of K⁴⁺ can potentiate the oxidation of NO and SO₂ during the redox process, thereby promoting the redox reaction and increasing the adsorption activity of CS. These results are consistent with those presented in Figs. 2 and 3.

3.4 Mechanism analysis

To gain deeper understanding of the mechanisms underlying desulfurization and denitrification, the reaction materials were examined by XRD, CO₂-TPD, and XPS. The obtained results are presented in Fig. 6. Fig. 6(a) displayed the XRD patterns of several modified CS samples after desulfurization and denitrification. Following these processes, multiple distinctive diffraction peaks corresponding to K₂SO₄, KNO₃, Ca(NO₃)₂, CaSO_4, and CaSO₃ were observed (Guo et al., 2022). The production of Ca(NO₃)₂, CaSO_4, and CaSO₃ suggested the participation of Ca(OH)₂ in the denitrification reaction during desulfurization. In comparison with Fig. 4(e), the intensity of the diffraction peaks related to KO_x decreased after desulfurization and denitrification, accompanied by K₂SO₄ and KNO₃ diffraction peaks, indicating the involvement of KO₂ in the reaction. Meanwhile, a new K₂Ca(CO₃)₂ diffraction peak was detected, indicating the formation of stable K₂Ca(CO₃)₂ during the desulfurization and denitrification. This formation can lead to a decline in the activity of CS after modification. Especially, the intensity of the diffraction peak of K₂Ca(CO₃)₂ followed the CS-K₂CO₃ reaction, which generated a substantial amount of K₂Ca(CO₃)₂. The combined analysis in Fig. 2 further demonstrated that K₂Ca(CO₃)₂ hindered the denitrification and desulfurization.

Close

Fig. 6

(a) XRD patterns, (b) CO₂-TPD profiles, (c) C 1 s, (d) O 1 s, (e) S 2p, and (f) N 1 s spectra of UCS-KHCO₃, UCS-KOH, and UCS-K₂CO₃.

Export to PPT

To observe the variation in the alkaline sites on the surfaces of the materials, CO₂-TPD experiments were conducted on both the pre-modified and post-modified materials. The experimental results, according to the experimental findings illustrated in Fig. 6(b), revealed the presence of distinct desorption peaks at weak basic sites (0–200 °C), medium basic sites (200–400 °C), and strong basic sites (>400 °C) (de Oliveira et al., 2022). Compared with CS, the modified CS exhibited the following changes: (i) a slight reduction in the intensity of weakly basic sites, (ii) relatively unchanged medium basic sites, and (iii) a significant increase in the intensity of strongly basic sites for CS-KOH and CS-KHCO₃. This further indicated that potassium addition had a substantial impact on the distribution of the basic sites in the CS. From Fig. 6(b) and 2, CS exhibited a larger peak area in the weakly basic site, implying a greater number of basic sites and enhanced desulfurization efficiency. This indicated that the removal of SO₂ occurred mainly through chemisorption and an acid-base neutralization reaction. However, the weak base sites of the modified CS were reduced, whereas the denitrification effect of the modified adsorbents was significantly enhanced, which can be attributed to the promotion of denitrification by KO_x.

The C 1 s spectrum of modified CS obtained after the reaction is shown in Fig. 6(c). Upon comparison with Fig. 4(h), it was evident that the surface functional groups underwent significant changes following the utilization of the modified CS in the desulfurization and denitrification processes. The contents of C = O in UCS-KOH, UCS-K₂CO_3, and UCS-KHCO₃ were reduced by 23.46 %, 14.41 %, and 11.09 %, respectively. Conversely, the C-O content in UCS-KOH increased to 45.5 %, and typical C-O (285.0 ± 0.5 eV) peaks were observed for UCS-K₂CO₃ and UCS-KHCO₃ with 55.13 % and 33.89 %, respectively. This indicated the involvement of C = O in the oxidation of SO₂/NO and its consumption in the SO₂/NO removal process. The presence of C = O, influenced the adsorption capacity of the modified CS surface by facilitating electron transfer to the adsorbed oxygen species (Fang et al., 2017). Considering the analysis in Figs. 2 and 4(h), C = O, serving as the main active site for acid gas chemisorption (Li et al., 2003, Wang et al., 2020), may play a crucial role in the effective simultaneous desulfurization and denitrification achieved by the modified CS. The O 1 s spectra of UCS-KOH, UCS-K₂CO_3, and UCS-KHCO₃, and their fitted curves are presented in Fig. 6(d). Comparing Tables S2 and S3, a significant decline of the O_ads content on UCS-KHCO₃ was observed, dropping from 47.47 % to 38.34 %. Only O_lat characteristic peaks existed for UCS-KOH and UCS-K₂CO₃, with the disappearance of the characteristic peaks of O_ads. The results indicated a significant decrease in the value of O_ads after the reaction, which further determined the involvement of adsorbed oxygen in the reaction and possibly played a key role in the oxidation reaction of the modified CS. Fig. 6(d) illustrated the XPS results of S 2p after desulfurization and denitrification of different modified CS. The S 2p in the XPS spectrum demonstrated two peaks at 168.67 eV and 169.81 eV for sulfate SO₄^2- and sulfite SO₃^2-2-3 (Li et al., 2016, Lian et al., 2017, Arfaoui et al., 2018, Li et al., 2018), respectively. This suggested that Ca²⁺ readily formed sulfur-containing substances with SO₄^2-and SO₃^2- on the surface of the modified CS. These sulfates and sulfites adhered to the material surface, obstructing the pore channels and covering the active sites, which ultimately led to the deactivation of the modified CS due to SO₂ poisoning(Wu et al., 2016, Silas et al., 2018). The formation of NO₃^- was responsible for the peak in the N 1 s XPS spectrum of UCS-KHCO₃ at 407 eV (Hao et al., 2017), as shown in Fig. 6(f). In contrast, UCS-KOH and UCS-K₂CO₃ failed to form characteristic peaks, probably due to the partial occupation of surface-active sites by the formation of sulfate, resulting in a low level of surface-bound N. The XPS data further confirmed the production of sulfate, sulfite, and nitrate during the reaction, which is consistent with Fig. 6(a).

Based on the above process, the successive steps involved in the desulfurization of the modified CS can be postulated through the following equations (Eq. (8) to Eq. (13)) (Li et al., 2022). Firstly, SO₂ in the flue gas was converted to the adsorbed state SO₂(ads). The Brønsted basic sites on the Ca(OH)₂ surface react with SO₂(ads) to form CaSO₃. Furthermore, the modified CS containing KO₂, which has potent oxidizing properties, supplied active oxygen (O*) and facilitated the desulfurization process. As a result, O* originating from the modified CS and O₂ in the flue gas oxidized SO₂(ads) into SO₃(ads). Finally, the formed SO₃(ads) was combined with Ca(OH)₂ and K₂O on the modified CS to form CaSO₄ and K₂SO₄. Meanwhile, the reaction mechanism of the denitrification process was speculated and summarized as the following equations (Eq. (14) to Eq. (18)). In the process of denitrification, NO was converted to the adsorbed state and subsequently oxidized by O* to generate NO₂(ads), and NO₂(ads) dimerizes to N₂O₄(ads). In addition, the formation of N₂O₄ was also promoted by O₂. N₂O₄ isomerizes to NO⁺NO₃^– at low temperatures (Wu et al., 2012) and then reacts with Ca(OH)₂ in the presence of O* to form Ca(NO₃)₂ according to an acid-base reaction (Liu et al., 2011). The reaction intermediate products NO₂(ads) and K₂O combine to form KNO₃. Deactivation of the modified CS after the reaction was due to the production of sulphate, sulphite, and nitrate. These substances covered the surface of the modified CS, obstructing the access of SO₂/NO to the active sites, thus resulting in a decrease in the adsorption activity of the modified CS(Li et al., 2022).