Experimental and DFT Studies of Influence of Flue Gas Components on the Interaction between CaO and As during Sludge Combustion

Abstract

The problem of As pollution emission from sludge during combustion has received widespread attention. The impact of flue gas components on the interaction with CaO and As during sludge combustion was analyzed using a series of experimental characterization methods. The strength of the activity of As₂O₃ on the CaO(001) surface as well as on the CO₂/SO₂/H₂O+CaO(001) surface with different O adsorption sites was revealed by combining with Density Functional Theory (DFT). According to the results, CO₂ in the flue gas reacted with CaO in a reversible carbonation reaction, which optimized the pore structure of the solid phase products and promoted the capture of As by CaO. SO₂ in the flue gas reacted with CaO in a sulfation reaction reaction to block the pores, which was not conducive to the capture of As by CaO. The presence of moisture led to poor pore structure collapse of the solid phase products as well as the formation of gehlenite, which reduced the enrichment of As by CaO. DFT calculations showed that the adsorption of As₂O₃ molecules on the CO₂+CaO(001) surface was affected by the position of the O active site, and the adsorption energy at the O_C1 top site was higher than that on the clean surface, which was favorable for the stable adsorption of As₂O₃ molecules. The existence of SO₂ decreased As₂O₃ molecules’ adsorption energy on the CaO(001) surface, which was unfavorable for the adsorption of As₂O₃ molecules. There were two main effects of H₂O molecules on the adsorption of As₂O₃ on the CaO(001) surface. One was the H₂O molecules weakened the interaction between the As atoms and O_surf atoms, which was unfavorable to the adsorption of As₂O₃ molecules; the other was the existence of stronger adsorption of O atoms in H₂O molecules on As atoms in As₂O₃ molecules, which made As₂O₃ molecules adsorbed at the top of O_H0 adsorbed with adsorption energies much larger than that of clean surface, and the adsorption was more stable.

1. Introduction

With accelerated cityization and industrialization, the amount of wastewater generated is on the rise year by year. According to statistics, China treated 86.21 billion tons of sewage in 2021. As a by-product of wastewater treatment, sludge contains a large amount of harmful heavy metal elements, which have a serious impact on the ecological environment and human health. The combustion method is widely used because it can achieve the reduction, stabilization, and harmlessness of sludge [1,2]. During combustion, heavy metals are not produced and destroyed; they are only migrated and transformed. As, a semi-volatile element, it is highly toxic and carcinogenic and has a non-negligible impact on humans and the environment [3,4,5,6]. Studies have shown that there exists As in flue gas of sludge after combustion, mainly in the form of As₂O₃(g) [7,8,9].

The sludge is complex in composition and contains various mineral components. During combustion, these minerals can react with heavy metals and affect their migration [10,11,12,13]. Studies have shown that As was converted to arsenate in the presence of reactive metal minerals (Ca, Fe, and Al), enhancing the enrichment of As [14,15]. Zhang et al. [16] performed As vapor adsorption tests on metallic minerals using a fixed bed reactor. The results indicated that CaO and Fe₂O₃ had better potential in terms of As removal capacity compared to Al₂O₃. Mahuli et al. [17] found after experiments using a differential bed reactor that the calcined Ca(OH)₂ had a more developed pore structure and could better immobilize As into the ash compared to kaolinite, Al₂O₃, and SiO₂. Batroňová [18] concluded from industrial and laboratory scales that limestone contributed to As capture during combustion. Chen et al. [19] found that As adsorption by CaO at low temperatures was mainly physical and chemical oxidation at high temperatures. Han et al. [20] showed that CaO could react chemically with As₂O₃ to form less volatile calcium-arsenic compounds, such as Ca(AsO₂)₂.

The components of the combustion flue gas also impact the transport and transformation of heavy metals. Li et al. [21] conducted an experimental study using a drop tube furnace and concluded that the release of As in the flue gas was facilitated by the introduction of a certain amount of SO₂ into the combustion atmosphere. Jiao et al. [22] concluded that the heavy metals collected in the cooling zone were not present in large amounts as chlorides but were converted to heavy metal sulfates deposited by introducing SO₂ or steam into a laboratory-scale electrically heated rotary kiln reactor. Miller et al. [23] investigated the effect of sulfur content on As migration by injecting SO₂ into a suspension-firing reactor; the experimental results proved that SO₂ could inhibit As precipitation and reduce As emissions. Zou et al. [24] explored the As release pattern by passing different concentrations of H₂O into the constant temperature coal combustion process. They found that the effect of H₂O on As release varied with temperature, and H₂O had a significant effect on As release in the low-temperature range, while H₂O hardly affected As release in the high-temperature range.

In addition, flue gas components interact with minerals, which affects the interaction of minerals and heavy metals. Zhao et al. [25] found that both SO₂ and NO were detrimental to the immobilization of As₂O₃ by Fe₂O₃, CaO, and γ-Al₂O₃ through simulated flue gas experiments. Among them, the inhibition effect of SO₂ was highly dependent on their concentrations. High concentrations of SO₂ weakened the inhibition effect of the three minerals. When the NO concentration was changed, the immobilization of As₂O₃ by the three minerals was largely unaffected. Yu et al. [26] found that CO₂ and SO₂ inhibited As adsorption by CaO and that SO₂ inhibition was stronger than CO₂ at 700 °C. As could be stably adsorbed to the Ca-O bond active site, and the captured As was present as AsO₂¹⁻ and AsO₄³⁻. Chen et al. [19] found that SO₂ would sulphate with CaO and occupy the reaction site of As, thus inhibiting As capture. However, at higher temperatures, the generated CaSO₄ had a certain adsorption effect on As, which was beneficial to As capture. Li et al. [21] found that SO₂ caused CaO clogging sulphation and therefore had a greater inhibitory effect on As adsorption. Therefore there was a strong competition between the flue gas components of fuel combustion and heavy metals, which was manifested by the chemical reaction of the flue gas components with minerals and thus affected the capture of heavy metals.

At present, studies on the interaction of flue gas components with minerals and heavy metals were not uniformly understood and most of them were focused on the coal combustion process, which needed to be supplemented by the study of the sludge combustion process; secondly, the microscopic mechanisms at the molecular level were not yet clear. Sludge that has been mechanically dewatered produced a certain concentration of water vapor, SO₂ and CO₂ during the actual combustion process. These three flue gas components were of interest because of their strong impact on the removal of trace heavy elements from sludge intrinsic minerals. Based on this, this paper investigated the effects of CO₂/SO₂/H₂O on the interaction of CaO with As during sludge combustion from both macro and micro aspects by combining experimental and simulation calculations, which was significant for the control of As pollution emission.

2. Materials

Sewage Sludge (SS) used in this paper was taken at a wastewater treatment plant in Taiyuan, Shanxi. Prior to the combustion test, a certain mass of SS containing about 80% moisture was placed on a tray and put into a blow-dryer and dried at a temperature of 105 °C for 48 h until there was no further change in the sludge mass. The dried sludge was then ground into a powder in a coal mill and sieved to obtain a sample of sludge with a particle size of less than 100 μm. CaO was analytically purity and no further purification was required. The proximate and ultimate analyses of the sludge samples were carried out with reference to GB/T 212-2008 and GB/T 31391-2015 [27], and their analytical results are shown in

Table 1. Characteristics of sludge.

Proximate Analysis (%)				Ultimate Analysis (%)
Mad	Aad	Vad	FCad	Cd	Hd	Od	Nd	Sd
2.53	49.67	43.03	4.77	23.45	3.51	17.14	4.35	0.58

Note: M, A, V, FC denote moisture, ash, volatile matter, and fixed carbon, respectively, and the subscript ad denotes an air-dry basis, based on a coal in which the air humidity has reached equilibrium. C, H, O, N, and S denote carbon, hydrogen, oxygen, nitrogen, and sulfur, respectively, and the subscript d denotes a dry basis, based on a coal in a hypothetical anhydrous state.

. As content and intrinsic mineral content of the sludge are shown in

Table 2. Intrinsic mineral content in sludge.

Heavy Metals (mg/kg)	Minerals (%)
As	SiO2	CaO	Al2O3	P2O5	MgO
15.025	32.732	16.992	17.14	7.680	7.639

.

3. Experimental Setup

3.1. Combustion Experiments

Figure 1 showed the test system diagram, and the system was mainly divided into three parts, namely the gas distribution system, combustion system and exhaust gas absorption system. Before the test, open the gas distribution system, adjust the flowmeter to make each gas reach the required combustion atmosphere, and keep the total gas flow rate at 2 L/min. The horizontal tube furnace was purged with a period to ensure that the atmosphere in the tube meets the test conditions. The furnace was raised until the desired combustion temperature, where a quartz boat with 2 g of material was slowly pushed into its center in order to perform the combustion test. The tail flue gas was absorbed with 5% NaOH, 5% HNO₃ solution and water, respectively, to prevent pollution of the environment. After 1 h, the temperature control system was switched off and the quartz boat was removed, and the sludge solid phase products mass was weighed after the furnace had cooled. Each test was repeated three times to reduce test errors.

Generally, the water content of sludge after mechanical dewatering is about 80% [28], so there is still a certain concentration of water vapor in the flue gas produced during the actual combustion of sludge, while the combustion flue gas usually contains 10~30% of CO₂ and 0.1~0.2% of SO₂. The experimental combustion conditions for this design are shown in

Table 3. Test conditions for sludge combustion.

Combustion Conditions	Temperature (°C)	Combustion Atmosphere (vol.%)	Samples (2 g)
T1	700 °C	20O2/80N2	100% Sludge
T2	800 °C
T3	900 °C
T4	1000 °C
TC1	700 °C		96% Sludge + 4% CaO
TC2	800 °C
TC3	900 °C
TC4	1000 °C
TCS1	700 °C		96% Sludge + 4% CaSO4
TCS2	800 °C
TCS3	900 °C
TCS4	1000 °C
C1	900 °C	10CO2/20O2/70N2	100% Sludge
C2		20CO2/20O2/60N2
C3		30CO2/20O2/50N2
C4		20CO2/20O2/60N2	96% Sludge + 4% CaO
S1		0.1SO2/20O2/79.9N2	100% Sludge
S2		0.15SO2/20O2/79.85N2
S3		0.2SO2/20O2/79.8N2
S4		0.15SO2/20O2/79.85N2	96% Sludge + 4% CaO
H1		20O2/80N2	10% H2O + 90% Sludge
H2			20% H2O + 80% Sludge
H3			30% H2O + 70% Sludge
H4			20% H2O + 76% Sludge + 4% CaO

.

3.2. Methods

The devices used and the related parameters are shown in

Table 4. Device Parameters.

Device	Make	Type	Measuring	Accuracy
ICP-MS	Agilent (Beijing, China)	Agilent 7800	ppb~ppm	1–5%
BET	Micromeritics (Beijing, China)	ASAP-2460	0.01 m2/g~1000 m2/g	1–5%
XRD	Malvern Panalytical (Beijing, China)	Panalytical Empyrean	5°~85°	1–5%
SEM-EDS	ZEISS (Beijing, China)	ZEISS Sigma 300	μm~nm	1–5%

. After microwave digestion of the sludge samples and solid phase products, an inductively coupled plasma mass spectrometer (ICP-MS) was used to measure the content of heavy metal As in them, and each measurement was repeated three times. To characterize the migration of heavy metals, the residual rate R and volatilization rate V presented the mass of heavy metals remaining in the solid phase products after sludge combustion and the mass of heavy metals volatilized into the gas phase as a percentage of the total mass of heavy metals in the original sample, respectively, i.e., R + V = 1. The residual rate was explained in Equation (1):

$$R=(\frac{C_{\mathrm{ash}}\times r_{\mathrm{ash}}}{C_{\mathrm{s}}})\times 100\%$$

(1)

where C_s was the number of heavy metals in the original sample, mg/kg; C_ash was the number of heavy metals in the solid phase products, mg/kg; r_ash was the solid phase products yield after combustion, which was the ratio of the weight of solid phase products after reaction to the weight of the original sample.

The sludge samples and solid phase products were analyzed for pore structure applying a specific surface and porosity analyzer (BET). An X-ray diffractometer (XRD) was adopted to analyze the physical composition of solid phase products with a scanning speed of 2°/min and a scanning range of 5° to 85°. The microstructure and elemental distribution of sludge samples and solid phase products were observed and analyzed using a scanning electron microscope energy spectrum analyzer (SEM-EDS).

4. Theoretical Calculations

4.1. Calculation Parameters

Based on DFT, the Cambridge Serial Total Energy Package (CASTEP) module in Materials Studio 2019 software was used for the study [29]. The exchange-correlation generalized function used the PBE (Perdew-Burke-Ernzerhof) [30] method in GGA (Generalized Gradient Approximation) [31], the electron wave function was expanded with a plane wave basis group to describe the electron-ion interaction with the help of a super-soft pseudopotential, and the optimization algorithm BFGS (Broyden-Flechter-Goldfarb-Shanno) was chosen to optimize the geometrical structure of the model. After conducting the cut-off energy test and energy convergence test, the cut-off energy was set to 450 eV with a k-point Brillouin zone of 3 × 3 × 1, taking into account the calculated efficiency and accuracy. The structural and energy convergence criteria were 1.0 × 10⁻⁶ eV/atom for the self-consistent field accuracy convergence, 1.0 × 10⁻⁵ eV/atom for the energy accuracy convergence, maximum displacement of 0.001 Å, maximum stress convergence criterion of 0.03 eV/Å, and maximum strain convergence accuracy of 0.05 GPa.

4.2. Model Parameters

The geometry of CaO cell and each molecule was optimized using the above optimization parameters and the optimized bond parameters were presented in

Table 5. The relevant data of the optimized CaO cell and each molecule.

Model	Key Length/Key Angle	Optimize Value (Å/°)	Reference Value (Å/°)	Relative Error (%)
CaO	a = b = c	4.91	4.81	2.16
	α = β = γ	90.00	90.00	0.00
CO2	C-O	1.17	1.16	0.86
	O-C-O	179.99	180.00	0.01
SO2	S-O	1.46	1.43	1.89
	O-S-O	119.48	119.50	0.02
H2O	H-O	0.98	0.96	1.88
	H-O-H	103.74	104.50	0.73
As2O3	As-O	1.87	1.86	0.43
	As-O-As	75.41	73.50	2.60

.

The relative error of each optimization value compared to the reference value [32] was less than 3%, indicating that the geometric optimization results were reliable. CaO crystal space group was FM-3M, and (001) crystal plane [33,34] was chosen as the adsorption substrate as the crystal plane was the most exposed and more common in its natural state and had the best stability. Considering the need to reduce calculation volume and improve calculation efficiency, the atomic level of CaO was chosen to be 4 layers, with 2 atomic slack on the surface and 2 layers fixed. To facilitate the adsorption of adsorbed molecules and to eliminate the impact of adjacent periodic structures on the calculation results, a (3 × 3) supercell plane was created with a 12 Å vacuum layer above the crystal plane so that there was no real interaction between the different atomic layers. Four adsorption sites on the CaO(001) surface were considered in the computational optimization process, namely the surface hollows, O top sites, Ca-O bridge sites, and Ca top sites, as shown in Figure 2 for CaO(001) periodic plate model.

4.3. Calculation of Adsorption Energy

For the purpose of characterizing how strongly the interaction between the adsorbed substance and CaO(001) surface, the adsorption energy was introduced and defined as Equation (2):

E_ads = E_finalsystem − E_slab − E_adsorbate

(2)

where E_ads energy of adsorption, kJ/mol; E_finalsystem was the total energy of the system after adsorption, kJ/mol; E_slab was the energy of the surface substrate before adsorption, kJ/mol; E_adsorbate was the energy of the adsorbed material, kJ/mol. When E_ads was positive, the reaction could not proceed spontaneously. When E_ads was negative, the reaction could proceed spontaneously, and larger values indicate stronger adsorption. In general, physical adsorption was considered to occur when the absolute value of adsorption energy was lower than 30 kJ/mol, and chemisorption was considered to occur when the absolute value was greater than 50 kJ/mol [35].

5. Results and Discussion

5.1. Migration Behaviour of As from Sludge Combustion Process

5.1.1. Impact of CaO on As Migration Pattern

Figure 3 shows the impact of CaO on the As migration pattern under different temperature conditions in the sludge combustion process. The results showed an overall decreasing trend in the residual rate of As with temperature (except 900 °C). With the combustion temperature increasing from 700 °C to 1000 °C, the residual rate of As decreased from 62.62% to 51.91%, a reduction of 10.71%. These were mainly heavy metals’ melting and boiling points in relation to volatilization characteristics. The higher the temperature, the higher the saturation vapor pressure of the heavy metals and their compounds, which made the heavy metals more easily released into the gas phase, detrimental to the immobilization of As. The residual rate of As showed an inflection point at 900 °C with a sudden increase to 67.62%. After the addition of 4% CaO, the trend of As residual rate with temperature was consistent with that of the unadded one, and it was suddenly increased to 77.61% at 900 °C. The residual rates of As increased significantly at any temperature, indicating that the presence of CaO favored the enrichment of As in the solid phase products. CaO reacted with gaseous arsenic-containing substances in the flue gas in the chemical reaction shown in Equation (3) to produce a more stable calcium-arsenic compound [13,36]. The immobilization of As by CaO was most significant at a temperature of 1000 °C, with the residual rate increasing from 51.91% to 67.08%, an increase of 15.17%.

3 CaO + O₂ + As₂O₃ → Ca₃(AsO₄)₂

(3)

5.1.2. Impact of CO₂/SO₂/H₂O on the Interaction of CaO and As

The impact of CO₂, SO₂, and H₂O on As migration pattern in the process of sludge combustion in the presence or absence of 4% CaO was presented in Figure 4.

A comparison of working conditions T3, C1, C2, and C3 showed that the residual rate of As increased when the sludge was burned in the atmosphere with CO₂. At the same time, with the increase of CO₂ concentration in the atmosphere (10% increased to 30%), the residual rate of As increased from 68.79% to 73.95% with an increase of 5.16%, and more As was enriched in the solid phase products. This outcome was determined by a combination of three factors: (a) In the case of the same air intake, the N₂ concentration in the combustion atmosphere would reduce when the CO₂ concentration increased. The specific heat capacity of CO₂ was greater than that of N₂, which lowered the combustion temperature, resulting in a decrease in the vapor pressure of the heavy metals, a decrease in the volatilization rate, and an increase in the residual rate [37]. (b) The density of CO₂ was greater than that of N₂, and compared with the 20O₂/80N₂ atmosphere, the passage of a large amount of CO₂ increased the O₂ diffusion resistance, slowed down the sludge combustion rate, and inhibited the volatilization of heavy metals. (c) When a high concentration of CO₂ was present in the combustion atmosphere, it reversed Equation (4) and converted the secondary oxides of volatile arsenic into more thermally stable arsenic oxide compounds, thus inhibiting the volatilization of As [38]. (d) At 900 °C, the high concentration of CO₂ would react with the sludge to a certain degree of gasification, which might lead to decomposition of the organic matter in the sludge and an increase in the sludge porosity, allowing As to be released more into the flue gas along with the volatile substances. Compared to TC3, the residual rate of As after sludge combustion with the addition of 4% CaO increased by 1.09% when 20% CO₂ was introduced, indicating that the presence of CO₂ facilitated As capture by CaO and inhibited As volatilization into the flue gas.

As_mO_n + CO ↔ As_mO_n−1 + CO₂

(4)

Compared to the original sludge combustion (T3), the proportion of As released into the flue gas increased with the introduction of SO₂ (working conditions S1, S2, and S3). With the increase of SO₂ concentration (0.1% to 0.2%), the residual rate of As decreased from 64.65% to 50.86%, a decrease of 13.79%, indicating that SO₂ was detrimental to the enrichment of As in the solid phase products. The large amount of SO₂ could combine with calcium-based minerals in the sludge and occupy the active sites of As in the original minerals, thus accelerating the release of As. Compared with TC3, after combustion of sludge loaded with 4% CaO in a 0.15SO₂/20O₂/79.85N₂ atmosphere, the residual rate of As decreased from 77.61% in the former to 68.65% in the latter, which was a reduction of 8.96%, suggesting that the presence of SO₂ was also detrimental to the capture of As by CaO. In addition, the As residual rate of sludge loaded with 4% CaO increased by 13.15% from 55.50% to 68.65% after combustion in 0.15SO₂/20O₂/79.85N₂ atmosphere compared to S2. There might be two reasons for this: (a) The addition of excess CaO compensated for the inhibitory impact of SO₂ presence in fixing As by CaO. (b) CaSO₄, produced by the chemical reaction between SO₂ and CaO, had a certain trapping capacity for As.

In order to investigate the strength of CaSO₄’s ability to capture As, sludge loaded with 4% CaSO₄ was combusted at different temperatures, as shown in Figure 3. As can be seen in Figure 3, the residual rate of As increased compared with the original sludge combustion after adding 4% CaO and 4% CaSO₄ to the sludge at different temperatures, and the residual rate of As increased to a greater extent after combustion of the sludge with 4% CaO, which indicated that the immobilization of As by CaO was better than that by CaSO₄. With the increase of temperature, the residual rate of As after the addition of 4% CaSO₄ first decreased and then increased, and reached a minimum value of 64.34% at 900 °C, indicating that CaSO₄ enriched As through physical adsorption in the range of 700 °C to 900 °C. With the increase in temperature, the kinetic energy of molecules increased, the physical effect weakened, the volatilization rate of As increased, and the residual rate decreased. Chen et al. [19] showed that CaSO₄ reacted with As₂O₃ to produce more stable compounds at temperatures higher than 1000 °C, which could be well explained by the fact that at 1000 °C, the immobilization of As by CaSO₄ changed from physical adsorption to chemical adsorption, resulting in an increase of As residual rate to 66.66%.

The test results of T3, H1, H2, and H3 indicated that the residual rate of As after the combustion of sludge with different water content was significantly reduced. When the moisture content was 10%, the residual rate of As was reduced by 22.29%. During the combustion process, the moisture in the sludge absorbed heat and changed into water vapor to be released from the sludge particles. At the same time, the water vapor and sludge ash coke underwent a gasification reaction in the anoxic part Equation (5), and a significant amount of gas was released from the ash particles, both of which optimized the pore structure of the sludge ash particles [39]. The developed pore structure could provide more reaction sites for heavy metals; however, it would also increase the contact area of carbon and oxygen reaction, accelerate the combustion reaction, and accelerate the volatilization of As; it would also reduce the diffusion resistance of As vapor, which was not conducive to the immobilization of As in the solid phase products. The test results also showed that the degree of reduction in the residual rate of As gradually diminished with increasing the moisture content. When the moisture content changed from 10% to 30%, the residual rate of As only reduced by 7.09% because when the moisture content increased, the high concentration of water vapor caused the collapse and closure of various micropores, forming macropores and mesopores, thus reducing the porosity of the particles. Secondly, the large amount of water vapor reacted with the sludge coke by cavitation reduced the temperature of the particle surface and was not conducive to the combustion reaction, both of which were favorable for the immobilization of As [40,41]. Compared to TC3, the residual rate of As was reduced by 8.89% in the presence of 4% CaO with a moisture content of 20% in the sludge, indicating that moisture was not conducive to As fixation by CaO. In the absence of CaO, the residual rate of As reduced by 28.00% after the combustion of sludge with 20% moisture content, which was much greater than that of 8.89%, and this indirectly suggested that the presence of CaO had a certain ability to capture As.

C + H₂O → CO + H₂

(5)

5.2. Characterization Analysis

5.2.1. Porosity Structure Analysis

The impact of CO₂/SO₂/H₂O on the pore structure of the 4% CaO added sludge solid phase products could be further investigated by N₂ adsorption/desorption method and the date were displayed in

Table 6. Pore structure under different working conditions.

Samples	Specific Surface Area (m2/g)	Pore Volume (×10−2 cm3/g)
Sludge	8.03	2.05
CaO	7.13	1.33
T3	4.82	0.84
TC3	6.32	1.11
C4	5.86	1.02
S4	2.26	0.42
H4	3.52	0.51

.

From Table 6, it can be seen that the specific surface area of the raw sludge decreased by about 40% after combustion compared with that without combustion, indicating that sludge melting occurred at 900 °C. After adding 4% CaO, there was a certain degree of improvement in the pore structure of sludge solid phase products, suggesting that adding CaO would inhibit the sintering of the sludge during combustion. With the addition of 4% CaO, the specific surface area and pore volume of the sludge solid phase products increased to a certain extent after the passage of CO₂ compared to working condition T3, which might be due to the decomposition reaction of CaCO₃ as shown in Equation (6) at 900 °C to produce more pore-like structures [42,43,44]. This reduced the diffusion resistance of heavy metal vapors but provided more reactive active sites for heavy metals, which facilitated their immobilization. Sludge solid phase products with 4% CaO were significantly worse in pore structure after the introduction of SO₂ due to the sulphation reaction of SO₂ with CaO through Equation (7), which blocked the pore structure of sludge ash particles, which was not conducive to As capture by CaO and allowed more As volatilization into the gas phase [45,46,47]. When the sludge with 4% CaO (20% moisture content) was burned, a small amount of Ca(OH)₂ underwent Equation (8) decomposition reaction, and sludge moisture evaporated to water vapor. When the water vapor underwent a gasification reaction, a large amount of gas escaped from the surface of the particles, all of which were conducive to the optimization of the pore structure [48,49,50]. On the contrary, compared to TC3, BET results showed that the specific surface area and pore volume at H4 reduced by 17.48% and 2.61% due to large amounts of gas released from the surface of the particles. The pore walls between the pores decreased until they broke and collapsed, which closed the pore volume and reduced the specific surface area, detrimental to the immobilization of heavy metals in the solid phase product.

CaCO₃ → CaO + CO₂

(6)

CaO + SO₂ + 1/2 O₂ → CaSO₄

(7)

Ca(OH)₂ → CaO + H₂O

(8)

5.2.2. Microscopic Morphological Analysis

For a more visual analyzed the impact of CO₂/SO₂/H₂O on the combustion of sludge added with 4% CaO, SEM-EDS was used to analyze the microscopic morphological structure and elemental distribution changes of the sludge solid phase products, as presented in Figure 5.

Figure 5 clearly showed that the raw sludge mainly exhibited an irregular granular distribution in terms of morphology. After combustion at 900 °C, the sludge solid phase products as a whole showed a molten surface morphology. After the addition of CaO, the surface of the sludge ash particles became more porous, and the pore channels increased significantly compared to T3, indicating that CaO enhanced the anti-sintering properties of the sludge and increased the porosity of solid phase products. The larger cavities formed on the surface at C3 were the result of CaCO₃ decomposition, where a large amount of CO₂ gas escaped from the surface of the particles, forming a more developed pore structure, which was agreed with BET test results. Ash particles surface at S4 appeared as a cubic pile of granular bodies, for which energy spectral analysis showed that the main elements in this region were Ca, S, and O, indicating that the newly generated lamellar structure was CaSO₄. A larger volume of pores appeared on the surface at H4, while the surface became smooth, which corresponded to the analysis in Section 5.2.1. After the energy spectral point sweep, in the red circled region shown in Figure 5 were significant elements of Ca, Si, Al, and O. The presence of moisture might have promoted the formation of gehlenite.

5.2.3. Phase Composition Analysis

Section 5.2.2 showed that there were some differences in the microstructure of the sludge solid phase products under different atmospheres, and XRD of the sludge ash particles under different working conditions was shown in Figure 6.

It was observed from Figure 6 that no diffraction peaks were detected in the ash particles after the combustion of the original sludge for CaO, while CaO diffraction peaks could be detected in the ash residues at TC3, C4, S4, and H4. No diffraction peak of CaCO₃ was detected in C4 since CaCO₃ could not exist stably at 900 °C and would completely decompose into CaO and CO₂, both of which could inhibit As volatilization through Equation (3) and facilitated the immobilization of As. CaSO₄ diffraction peaks were clearly detected in S4, while the intensity of the CaO diffraction peak decreased slightly compared to TC3, indicating that part of CaO reacted with SO₂ to convert into CaSO₄ and that CaSO₄ existed stably at 900 °C and did not decompose. In addition, the excess CaO had some ability to trap As, which agreed with the results in Section 5.1.2. For H4, there was a significant enhancement of the diffraction peaks of gehlenite and the intensity of the diffraction peaks of CaO was reduced because the presence of moisture promoted the conversion of the calcium-based minerals in the sludge to the more thermally stable gehlenite, as shown in Equation (9), which was consistent with EDS results.

2 CaO + Al₂O₃ + SiO₂ → Ca₂Al₂SiO₇

(9)

5.3. Calculation Results and Discussion

In order to better investigate the microscopic mechanism of CO₂, SO₂, and H₂O on the interaction of CaO with As₂O₃, the active sites of a single As₂O₃ molecule adsorbed on the CaO(001) surface in three ways, As-terminus, O-terminus and As-O bond, respectively, were firstly investigated as shown in Figure A1, and the parameters of the optimized adsorption structure are shown in

Table A1. Structural parameters of As₂O₃ optimized on the CaO(001) surface.

Adsorption Structures		Eads (kJ/mol)	As-Osurf Key Length (Å)	As-Osurf Population	q(As2O3) (e)
As-terminus	1-a	−202.533	1.836	0.32	−0.39
	2-a	−202.774	1.836	0.32	−0.38
	3-a	6.946	-	-	−0.10
	4-a	11.992	-	-	−0.07
O-terminus	1-b	−202.273	1.836	0.32	−0.38
	2-b	21.120	-	-	0.01
	3-b	−202.755	1.836	0.32	−0.32
	4-b	−6.909	-	-	−0.01
As-O bond	1-c	−31.735	-	-	−0.17
	2-c	−202.234	1.835	0.32	−0.38
	3-c	−202.427	1.837	0.32	−0.38
	4-c	−202.389	1.835	0.32	−0.38

Notes: O atoms on CaO(001) are denoted by O_surf.

. Secondly, the stable adsorption structures of the CO₂ molecule, SO₂ molecule and H₂O molecule on the CaO(001) surface were used as substrates to adsorb As₂O₃ molecules, respectively, to investigate the effects of the presence of CO₂, SO₂ and H₂O on the adsorption of As₂O₃ molecules on the CaO(001) surface. In order to ensure the completeness of the analysis, the active sites of a single CO₂/SO₂/H₂O molecule adsorbed on the CaO(001) surface in a parallel or perpendicular manner were considered, as shown in Figure A2, Figure A3 and Figure A4. The parameters of the optimized adsorption structures are shown in

Table A2. Structural parameters of CO₂ optimized on the CaO(001) surface.

Adsorption Structure		Eads (kJ/mol)	C-Osurf Bond Length (Å)	C-Osurf Population	q(CO2) (e)
C-terminus parallel	1-a	−0.341	-	-	−0.05
	2-a	−123.321	1.393	0.63	−0.64
	3-a	−123.379	1.392	0.63	−0.64
	4-a	16.863	-	-	−0.01

,

Table A3. Structural parameters of SO₂ optimized on the CaO(001) surface.

Adsorption Structure		Eads (kJ/mol)	S-Osurf Bond Length (Å)	S-Osurf Population	q(SO2) (e)
S-terminus parallel	1-a	−169.444	1.699	0.21	−0.31
	2-a	−169.415	1.700	0.21	−0.31
	3-a	−177.037	1.672	0.22	−0.30
	4-a	8.205	-	-	−0.15
S-terminus vertical	1-b	−177.056	1.671	0.22	−0.30
	2-b	−169.502	1.698	0.21	−0.31
	3-b	−176.661	1.672	0.23	−0.3
	4-b	8.195	-	-	−0.1

and

Table A4. Structural parameters of H₂O optimized on the CaO(001) surface.

Adsorption Structure		Eads (kJ/mol)	H-Osurf Bond Length (Å)	H-Osurf Population	q(H2O) (e)
O-terminus parallel	1-a	−58.875	-	-	−0.25
	2-a	11.772	-	-	−0.03
	3-a	−77.699	1.036	0.48	−0.26
	4-a	−58.537	-	-	−0.25
O-terminus vertical	1-b	18.005	-	-	0.01
	2-b	19.182	-	-	0
	3-b	−8.606	-	-	0.03
	4-b	−4.766	-	-	0.03

. For the CO₂/SO₂/H₂O+CaO(001) surface, the stabilized structures 3-a, 1-b, and 1-a (H₂O molecules intact) with the largest adsorption energies were selected as adsorption substrates, respectively. Considering the obvious periodicity and symmetry of each stabilized structure, as well as the fact that the O_surf top site on the CaO(001) surface was the active site for the adsorption of As₂O₃ molecules, the adsorption reactions at the surface vacancies, the Ca-O bridging site, and the Ca top site were not taken into account. The O top sites on the CO₂/SO₂/H₂O+CaO(001) surface that were considered to have adsorption potentials are shown in Figure 7.

As shown in Figure A1 and Table A1, the adsorption energy of As₂O₃ molecules reached a maximum after placing them vertically with As terminus at the O_surf top site on the CaO(001) surface, so a single As₂O₃ molecule was considered to adsorb in the same initial manner on the potential O-active site on CO₂/SO₂/H₂O+CaO(001) surface. The structure of As₂O₃ before and after adsorption on the CO₂/SO₂/H₂O+CaO(001) surface is shown in Figure 8, and the relevant parameters after adsorption are shown in

Table 7. Parameters of As₂O₃ after optimization on CO₂/SO₂/H₂O+CaO(001).

Adsorption Structure	EadS (kJ·mol−1)	As-OX Bond Length (Å)	As-OX Population	q(As2O3) (e)	X-Osurf Key Length (Å)	X-Osurf Population	q(X) (e)
OC0	−26.737	-	-	−0.1	1.366	0.68	−0.64
OC1	−209.972	1.857	0.31	−0.33	1.370	0.67	−0.65
OC2	−190.502	1.860	0.30	−0.34	1.384	0.65	−0.65
OC3	−51.997	2.074	-	−0.18	1.333	0.75	−0.58
OS0	−7.430	-	-	−0.04	1.659	0.23	−0.31
OS1	−192.094	1.873	0.3	−0.37	1.672	0.22	−0.27
OS2	−2.239	-	-	−0.01	1.671	0.23	−0.30
OS3	−177.109	1.877	0.3	−0.36	1.675	0.22	−0.24
OS4	80.323	-	-	0.23	1.608	0.24	−0.73
OH0	−241.446	1.788	0.34	−0.40	1.005	0.37	−0.26
OH1	−199.484	1.854	0.31	−0.34	-	-	−0.20
OH2	−236.457	1.768	0.35	−0.39	1.003	0.54	−0.24

Note: O atoms in H₂O, CO₂, and SO₂ were denoted by O_H, O_C, and O_S, respectively, and O_surf, O_H, O_C, and O_S were denoted uniformly by O_X.

. Figure 8 and Table 7 indicate:

For the CO₂+CaO(001) surface, when As₂O₃ molecules were adsorbed at the O_C0 top site, the As₂O₃ molecules would move away from the surface with an adsorption energy of −26.737 kJ/mol, which was physisorption. This might be due to the fact that the presence of CO₂ occupied the O_surf active site, and repulsion between itself and the As₂O₃ molecules would occur, making the As₂O₃ molecules unable to be stably adsorbed on the O_C0 top site. Upon adsorption at the top sites of O_C1 and O_C2, the distance between the As₂O₃ molecules and the CO₂ molecules increased, the CO₂ was deflected at an angle, and the As₂O₃ molecules formed the same stable adsorption configuration as on the clean surface. However, compared with the clean surface, the As-O_surf bond length increased, the number of bond layouts decreased, and the number of electrons transferred by As₂O₃ molecules decreased, whereas the C-O_surf bond decreased, the number of bond layouts increased, and the number of electrons transferred by CO₂ molecules increased compared with the pre-adsorption period, which suggested that the interactions between the As₂O₃ molecules and CaO(001) surface were weakened, whereas the interactions between CO₂ molecules and CaO(001) surface were enhanced. Meanwhile, the absolute magnitude of adsorption energy after stabilized adsorption: top site of the O_C1 (209.972 kJ/mol) > top site of O_surf (202.774 kJ/mol) > top site of O_C2 (190.502 kJ/mol). This might be due to the fact that, for the O_C1 top site, the strengthened interaction between CO₂ and CaO(001) surface was greater than the weakened interaction between As₂O₃ and CaO(001) surface, which overall showed a strengthened degree of interaction and increased adsorption energy. For the O_C2 top site, the weakened interactions between As₂O₃ and CaO(001) surface were greater than the strengthened interactions between CO₂ and CaO(001) surface, which overall showed a weakened degree of interactions and reduced adsorption energy. When As₂O₃ molecules were adsorbed at the top site of O_C3, the As₂O₃ molecules were transformed from a hexahedral structure to a chain structure, and the As atoms closer to the CO₂ molecules formed a 2.074 Å As-O_C bond with the O_C atoms, with an adsorption energy of −51.997 kJ/mol, which was a weak chemical adsorption. In summary, the effect of CO₂ on the adsorption of As₂O₃ molecules on the CaO(001) surface was related to the position of the O active site, and the presence of CO₂ favored the adsorption of As₂O₃ molecules at the O_C1 top site, and was detrimental to their adsorption at the O_C0, O_C2 and O_C3 top site.

For SO₂+CaO(001) surface, when As₂O₃ molecules were adsorbed on O_S0 top site and O_S2 top site, they did not react with CaO(001) surface, and the adsorption energies were −7.430 kJ/mol and −2.239 kJ/mol, which belonged to the physical adsorption. This indicated that the presence of SO₂ prevented the As₂O₃ molecules from adsorbing stably on the O_S0 top site and O_S2 top site. Upon adsorption at the O_S1 top site and the O_S3 top site, the As₂O₃ molecules formed the same stable adsorption configuration as on the clean surface. Compared with the clean surface, the As-O_surf bond lengths increased to 1.873 Å and 1.877 Å, and the As-O_surf bond layout numbers both decreased to 0.30, and the number of electrons transferred from CaO(001) surface to the As₂O₃ molecules decreased by 0.01e and 0.02e, indicating that the presence of SO₂ made the As₂O₃ molecules and CaO(001) surface interaction becomes weaker, which was unfavorable to the adsorption of As₂O₃ molecules on the CaO(001) surface. For SO₂ molecules, the number of bond layouts of As₂O₃ molecules after adsorption at the O_S1 top site remained the same as before adsorption, but the number of charges transferred by SO₂ molecules was reduced by the increase in the S-O_surf bond length. In contrast, the S-O_surf bond of the As₂O₃ molecule after adsorption at the O_S3 top site increased in bond length, the number of bond layouts remained unchanged, and the number of transferred charges decreased more compared to that before adsorption. Therefore, the absolute magnitude of adsorption energy after adsorption: top site of O_surf (202.774 kJ/mol) > top site of O_S1 (192.094 kJ/mol) > top site of O_S3 (177.109 kJ/mol). When the As₂O₃ molecule adsorbed at the O_S4 top site, one of the As-O bonds in the As₂O₃ molecule was broken, the corresponding As atom was bonded to both the S and O_S atoms in the SO₂ molecule, and this S-O_S bond was broken, and the calculated adsorption energy of the structure was 80.323 kJ/mol > 0, indicating that the reaction could not take place. In summary, SO₂’s were unfavorable for the adsorption of As₂O₃ molecules at the top sites of O_S0~O_S4.

For H₂O+CaO(001) surface, when As₂O₃ molecules were adsorbed at the O_H0 top site and O_H2 top site, both H-O_H bonds in H₂O molecules were broken, and the free H atoms formed a new chemical bond, H-O_surf bond, with Osurf atoms on the CaO(001) surface, respectively. The As₂O₃ molecule would not react with CaO(001) surface, but would interact with the free O_H atoms in the H₂O molecule, and the absolute values of adsorption energies were 241.446 kJ/mol and 236.457 kJ/mol, respectively (much larger than that of the clean surface, 202.774 kJ/mol), which were strong chemisorption. The As-O_H bond lengths were 1.788 Å and 1.768 Å, and the number of electrons transferred to the As₂O₃ molecule were 0.40e and 0.39e, respectively, suggesting that the presence of H₂O facilitated the As₂O₃ molecule to adsorb more stably on the surface. Upon adsorption at the O_H1 top site, the As₂O₃ molecule formed the same stable adsorption configuration as on the clean surface. The H₂O molecule did not decompose, and the absolute value of adsorption energy after adsorption was 199.484 kJ/mol, which was slightly reduced compared with the clean surface. The As-O_surf bond length increased by 0.018 Å, the number of As-O_surf bond layouts both decreased to 0.01, and the number of electrons transferred from the CaO(001) surface to the As₂O₃ molecule decreased by 0.04e. This indicated that H₂O had an inhibitory effect on the interactions between the As₂O₃ molecule and CaO(001) surface, which was not conducive to the adsorption of As₂O₃ molecules at the O_H1 top site on the CaO(001) surface for the adsorption of O_H1 top site.

In order to further probe the changes in the electronic structure of As₂O₃ molecules reacting on the CO₂/SO₂/H₂O+CaO(001) surface, the stable structures with the largest adsorption energies (O_surf, O_C1, O_S1, and O_H0 top sites) were subjected to partial density of state analysis, as shown in Figure 9.

As seen in Figure 9a, when As₂O₃ molecules were adsorbed on the CaO(001) surface, the s orbitals of the As atoms and the S orbitals of the O_surf atoms produced a significant overlap of the density of states peaks at −19.95 eV, −18.83 eV, −17.56 eV, −9.33 eV, −7.81 eV, and −4.84 eV, respectively. The s orbitals of the As atoms and the p orbitals of the O_surf atoms had overlapping orbitals at −17.56 eV, −9.33 eV, −7.81 eV, −5.52~0.42 eV and 2.48~14.30 eV. More orbitals of the P orbitals of the As atoms and the p orbitals of the O_surf atoms overlap at −5.52~0.42 eV and 2.48~14.30 eV, indicating that strong orbital hybridization had occurred between the As atoms and the O_surf atoms to form a new chemical bond. As shown in Figure 9b, after the adsorption of As₂O₃ molecules at the O_C1 top site, the degree of overlap of partial density of state orbitals of the reacted As atoms and O_surf atoms remained basically the same as that on a clean surface, but the overall orbitals were shifted to higher energy levels, indicating the weakening of the interactions between the As atoms and the O_surf atoms. This corresponds to the increase in As-O_surf bond length and the decrease in the number of transferred charges in Table 7. For C and O_surf atoms, the degree of overlap of partial density of state orbitals of the two atoms after adsorption remained basically the same as that before adsorption, but the overall orbitals were shifted to the lower energy level direction, indicating that the interaction between C and O_surf atoms was enhanced, which resulted in the increase of adsorption energy of As₂O₃ molecules after adsorption at the top site of O_C1. As shown in Figure 9c, When As₂O₃ molecules were adsorbed at the top site of O_S1, the partial density of state orbitals of As and O_surf atoms overlapped to a degree essentially identical to the clean surface, and their overall orbitals were similarly shifted to higher energies and were higher than the energy levels after adsorption at the top site of O_C1. It indicated that the interaction between As and O_surf atoms was weakened, which was not conducive to the stable adsorption of As₂O₃ molecules. As shown in Figure 9d, When the As₂O₃ molecule adsorbed at the O_H0 top site, the s orbitals of the As atoms and the s orbitals of the O_H atoms in the H₂O molecule underwent orbital hybridization at the density of states peaks at −19.08 eV, −17.49 eV, −9.33 eV, and −7.73 eV, and the s orbitals of the As atoms and p orbitals of the O_H atoms at −5.32~0.49 eV underwent a large area of density of states overlap, indicating that a strong interaction between As and O_H atoms occurs, forming an As-O_H bond of 1.788 Å, which was favorable for the stable adsorption of As₂O₃ molecules.

6. Discussion

In this paper, the effects of CaO and flue gas components (CO₂, SO₂, and H₂O) on As migration and transformation during sludge combustion were investigated separately using a horizontal tube furnace; moreover, the effects of flue gas components on the interaction between CaO and As were explored. The activity strength of As₂O₃ molecules adsorbed on different surfaces was revealed in combination with DFT. The results showed that:

(1)

CaO chemically reacted with gaseous arsenic-containing compounds to produce more stable calcium-arsenic compounds, and CaO immobilized As better than CaSO₄. CO₂ contributed significantly to the immobilization of As. When CO₂ concentration increased from 0% to 30%, the residual rates of As increased by 6.33%. SO₂ was detrimental to the immobilization of As. As SO₂ concentration increased from 0% to 0.2%, the residual rate of As showed a decreasing trend, decreasing by 16.76%. H₂O had a significant inhibitory effect on the immobilization of As. The moisture content in sludge increased from 0% to 30%, and the residual rate of As was decreased by 29.38%.

(2)

The residual rate of As after combustion of sludge loaded with 4% CaO in 20CO₂/20O₂/60N₂ atmosphere increased by 1.09%, and CO₂ favored the capture of As by CaO. The residual rate of As was reduced by 8.96% after combustion of sludge loaded with 4% CaO in 0.15SO₂/20O₂/79.85N₂ atmosphere, and SO₂ inhibited CaO to fix As. The residual rate of As was reduced by 8.89% at 20% water content in sludge loaded with 4% CaO, and the presence of water was unfavorable for the capture of As by CaO.

(3)

DFT calculations showed that the CaO(001) surface O top site was the active site for the adsorption of As₂O₃, CO₂, SO₂, and H₂O molecules. The adsorption energy of As₂O₃ molecules at the O_C1 top site (−209.972 kJ/mol) was higher than at the O_surf top site (−202.774 kJ/mol). CO₂ favored the stable adsorption of As₂O₃ molecules at the O_C1 top site to the detriment of their adsorption at the top sites of O_C0, O_C2, and O_C3. The adsorption energies of As₂O₃ molecules at potential active sites on SO₂+CaO(001) surface were all lower than those on clean CaO(001) surface, and SO₂ was unfavorable for the adsorption of As₂O₃ molecules at the top sites of O_S0~O_S4. For the O_H0 and O_H2 top sites, the H-O_H bonds in the H₂O molecule were broken, and the resulting O_H atoms underwent more intense orbital hybridization with the As atoms in the As₂O₃ molecule. The adsorption energies (−241.446 kJ/mol and −236.457 kJ/mol) of the stabilized structure after adsorption were higher than those of the O_surf top site, which greatly facilitated the stabilized adsorption of the As₂O₃ molecule, whereas the O_H1 top site was not conducive to the stabilized adsorption of the As₂O₃ molecule. In addition, in the theoretical calculation section only considered As₂O₃ (g), a major form of As present in combustion flue gas. In the actual combustion process, there were also singlet, chlorinated states, etc., which needed to be further investigated in the follow-up work.

(4)

Migration was a complex physicochemical reaction during the actual combustion of sludge. When exploring the migration law, not only shall we analyze the influence of physical properties such as pore structure and changes in chemical forms such as compounds on the migration of As from a macroscopic point of view by means of experimental characterization, but we shall also explain the changes in the electronic structure of atoms and molecules as well as the changes in the adsorption energy from a microchemical point of view by means of theoretical calculations. Probing the migration pattern of As was instructive for the efficient removal of heavy metals from sludge and the realization of stable and clean combustion of sludge.