3.1. Morphological characterizations analysis
The SEM results in Fig.3 of raw fly ash in three grain size of D50(30, 45, or 60μm) supported the hypothesis of original experiment design that ball milling only change grain size and distribution by crushing and extrusion on fly ash without impacting on micromorphology. The results showed that the surface of raw fly ash particles is surrounded by flocculent and globosity with high porosity. The samples in Fig.4 of different grain size show that fly ash under ball milling is increased with bulk density. Particles change into powders from granules.
The samples of sintered fly ash are shown in Fig.5. Comparing with raw sample, it was observable that sintered fly ash bulk density grows up and the gap of particles is dropped. The color has noticeable change from gray to malachite green. It might be due to formation of new minerals. The external morphology of sintered fly ash was studied by SEM, as shown in Fig.6. The results revealed that treatment significantly affects the microstructure of the fly ash. Sintering causes particles to become lower porosity and to weld together forming a dense structure. In previous studies, sintered coal fly ash can be seen many small granular hydroxysodalite aggregates on the surface with SEM(Luo et al. 2018). Red mud-fly ash is formed a mass of bar-structure from spherical particles under sintering(Samal et al. 2015). After sintering, the spherical particles of coal fly ash disappeared(Wu et al. 2014). In addition, sintering leads to dramatic particles aggregation with bar or rod-shape in microstructure, as a result of formation and growth of crystal nucleus. The raised surfaces indicated that directional crystals grow at fracture. The upturn of crystallinity helps to improve bulk density and mechanical properties. The migration and motion rate of atomics build up by enhancing driving force during sintering.
3.2. Chemical composition and XRD analysis
The chemical compositions tested with XRF of raw fly ash and sintered fly ash under different condition in orthogonal design are given in Fig.7. It can be seen that CaO, SiO2, Al2O3, SO3 and Cl are the main compositions for sintered fly ash, which is notable differences with raw fly ash. The chemical compositions of sintered fly ash are under different group remain level. The content of CaO, SiO2, and Al2O3 all increased to 50.16%, 11.96%, and 5.25% from 41.55%, 4.81%, and 1.03%, respectively. Sintering contributes to optimize compositions of fly ash for construction use as cement, because of more similar components. CaO is beneficial to the growth of crystal with high value. While, Na2O, K2O, and Cl for sintered fly ash were all lower than raw fly ash. Particularly, the percentage of Cl nearly fell from 26.50% to 12.74%, which provides the evidence that sintering helps to reduce the content of Cl being harmful to cement(Chen et al. 2016).
Fig.8 demonstrates the XRD patterns of raw and sintered fly ash are under various orthogonal conditions. It was discerned that the main differences in the raw and sintered samples were the variation in mineral peak intensity. By comparison with raw fly ash, the relative contents of CaCO3 and CaClOH declined and even disappeared as the results that CaCO3 and CaClOH decomposing at high temperature during sintering. It can be well explained as the change of chemical composition on CaO and Cl. Whereas the peak intensity of CaTiO3, MgO, and Ca12Al14O33 increased or occurred. This phenomenon supports the analysis of SEM that crystallinity and intensity of the crystal phases increased in degree by sintering. Ca12Al14O33 is created by the solid-state reaction with CaO and Al2O3(Rudradawong et al. 2020), which belongs to orthorhombic crystal with cyan-gray Ng and blue-green Np surface, which is in good agreement with sintered fly ash images.
The content of soluble chloride salts as NaCl and KCl in sintered fly ash were reduced by 50% nearly comparing raw samples in Fig.8. Howerver, Ca5(PO4)3Cl raised practically four times under sintering. The results show that a part of Cl is immobilized as a stable mineral reducing the possibility of escaping. The other Cl is attached on the sintered fly ash particles as soluble chloride salts. Thus, sintering is limited to remove the whole Cl in this study, which may be taken off by washing for those soluble chloride salts.
Fig.9 shows the change for crystallinity of raw and sintered fly ash. The value with a mean of 89.78% under sintering, while raw fly ash only is 65.93%. this result provided specific evidence that the treatment in this study helps fly ash to improve the growing of crystals and reduce the leaching of heavy metals. It is worth notice that the value of group S8 is 99.99% which means a greatly tight structure in particles.
3.3. Leaching of toxicity metals analysis
The leaching performances of sintered fly ash were evaluated using toxicity threshold included in the standard of Identification Standards for Hazardous Wastes-Identification for Extraction Toxicity (GB5085.3-2007), Integrated wastewater Discharge Standard (GB8978-1996), and Technical Specification for Coprocessing of Solid Waste in Cement Kiln (GB30760-2014). All the results were summarized in Table 4. It is shown that leaching characteristic of toxicity elements in the most orthogonal samples are much lower than the standard limits. Specifically, the leaching values of As, Pb, Cu, Mn, Hg, Be, Se, and Fluoride were even not detected, indicating that proposed sintering methods have immobilization contribution on the most toxicity elements. In spite that the effects of sintering design have some limitations in reducing the leaching of, Cr, Cr6+ Ag, and Ba. In group S2 and S3, the immobilization of Cr is only 25.74% and 19.68%, respectively. Even the leaching of Cr6+ increased from 5.54 to 6.5mg/L. The result showed that As, Pb, Cd, Cu, Ni, Zn, Mn, Hg, Be, Se and fluoride barely indicate leaching characteristic in orthogonal test.
Table 4 Orthogonal test table for sintering experiment
No.
|
Leaching of toxicity metal elements(mg/L)
|
As
|
Pb
|
Cd
|
Cr
|
Cu
|
Ni
|
Zn
|
Mn
|
Hg
|
Cr6+
|
Be
|
Ag
|
Ba
|
Se
|
Fluoride
|
FA
|
0.41
|
0.64
|
0.04
|
8.74
|
0.54
|
0.29
|
2.4
|
1.5
|
0.07
|
5.54
|
0.008
|
0.94
|
115
|
1.65
|
0.65
|
S1
|
ND
|
ND
|
0.01
|
2.93
|
ND
|
ND
|
ND
|
ND
|
ND
|
1.85
|
ND
|
0.62
|
13.45
|
ND
|
ND
|
S2
|
ND
|
ND
|
0.03
|
2.49
|
ND
|
ND
|
0.3
|
ND
|
0.02
|
2.5
|
ND
|
0.59
|
5.29
|
ND
|
ND
|
S3
|
ND
|
ND
|
0.02
|
3.15
|
ND
|
ND
|
0.02
|
ND
|
ND
|
1.45
|
ND
|
0.41
|
4.22
|
ND
|
ND
|
S4
|
ND
|
0.29
|
ND
|
0.38
|
ND
|
0.12
|
0.01
|
ND
|
ND
|
1.612
|
ND
|
0.28
|
34.4
|
ND
|
ND
|
S5
|
ND
|
ND
|
0.02
|
0.68
|
ND
|
ND
|
0.01
|
ND
|
ND
|
0.323
|
ND
|
0.18
|
10.72
|
ND
|
ND
|
S6
|
ND
|
ND
|
ND
|
0.89
|
ND
|
0.19
|
ND
|
ND
|
ND
|
0.78
|
ND
|
0.23
|
26.8
|
ND
|
ND
|
S7
|
ND
|
ND
|
0.03
|
0.26
|
ND
|
0.23
|
ND
|
ND
|
ND
|
0.289
|
ND
|
ND
|
33.3
|
ND
|
ND
|
S8
|
ND
|
ND
|
0.03
|
0.65
|
ND
|
0.18
|
ND
|
ND
|
ND
|
0.759
|
ND
|
0.36
|
38.3
|
ND
|
ND
|
S9
|
ND
|
ND
|
0.04
|
0.28
|
ND
|
0.22
|
ND
|
ND
|
ND
|
0.254
|
ND
|
0.16
|
39.7
|
ND
|
ND
|
ST
|
0.1
|
0.3
|
0.03
|
0.2
|
0.5
|
0.2
|
1
|
1
|
0.05
|
0.5
|
0.005
|
0.5
|
100
|
1
|
100
|
Note: ND Indicates not detected, ST Indicates Standard thresholds.
3.4. Response surface of Box-Benhnken analysis
The leaching of Cr, Cr6+, Ag, and Ba were chosen to investigate with Box-Benhnken response surface method for further revealing the influence of four factors on above toxicity elements, and calculating optimal value and experiment design with sintered fly ash. It was considered that residual toxicity elements can be greatly restricted in crystal phase according to the results of orthogonal experiment. The findings as shown in Table 5 provide substantial evidence for four variables having a major impact on the immobilized toxicity metals.
Table 5 Heavy metals leaching of treated samples in Box-Benhnken experiment for Cr, Cr6+, Ag, and Ba (mg/L)
No.
|
Factor
|
Leaching characteristics of heavy metals(mg/L)
|
Grain size of D50(μm)
|
Dosage of fluxing agent(%)
|
Setting temperature(℃)
|
Setting time(min)
|
Cr
|
Ag
|
Ba
|
Cr6+
|
1
|
45
|
5
|
1050
|
240
|
0.322
|
0.1
|
26.5
|
0.803
|
2
|
30
|
2.5
|
1100
|
180
|
1.08
|
0.091
|
19.8
|
0.834
|
3
|
45
|
0
|
1100
|
180
|
2.83
|
0.47
|
5.4
|
3.21
|
4
|
60
|
5
|
1050
|
180
|
0.165
|
0.16
|
39.2
|
0.345
|
5
|
60
|
2.5
|
1000
|
180
|
0.25
|
0.58
|
32.1
|
0.732
|
6
|
45
|
2.5
|
1100
|
120
|
0.56
|
0.22
|
25.6
|
1.32
|
7
|
60
|
2.5
|
1050
|
240
|
0.34
|
0.493
|
30.4
|
1.03
|
8
|
45
|
2.5
|
1050
|
180
|
0.446
|
0.312
|
30.4
|
0.476
|
9
|
45
|
5
|
1050
|
120
|
0.22
|
0.23
|
32.9
|
0.345
|
10
|
45
|
2.5
|
1050
|
180
|
0.435
|
0.451
|
29.4
|
0.443
|
11
|
60
|
2.5
|
1100
|
180
|
0.287
|
0.21
|
32.6
|
1.98
|
12
|
30
|
2.5
|
1000
|
180
|
0.82
|
0.57
|
27.5
|
0.367
|
13
|
45
|
5
|
1100
|
180
|
0.73
|
0.05
|
27.3
|
0.892
|
14
|
45
|
0
|
1000
|
180
|
2.1
|
0.72
|
6.87
|
2.26
|
15
|
45
|
2.5
|
1050
|
180
|
0.58
|
0.45
|
26.9
|
0.623
|
16
|
45
|
0
|
1050
|
240
|
2.49
|
0.66
|
3.6
|
2.9
|
17
|
60
|
0
|
1050
|
180
|
0.9
|
0.59
|
11.2
|
2.25
|
18
|
30
|
2.5
|
1050
|
240
|
0.764
|
0.222
|
24.1
|
0.786
|
19
|
60
|
2.5
|
1050
|
120
|
0.241
|
0.47
|
34.2
|
0.98
|
20
|
45
|
2.5
|
1000
|
120
|
0.598
|
0.56
|
27.9
|
0.562
|
21
|
45
|
5
|
1000
|
180
|
0.65
|
0.283
|
34.7
|
0.432
|
22
|
30
|
0
|
1050
|
180
|
3.123
|
0.621
|
7.9
|
1.65
|
23
|
45
|
0
|
1050
|
120
|
1.42
|
0.692
|
4.12
|
1.72
|
24
|
45
|
2.5
|
1000
|
240
|
0.61
|
0.554
|
28.3
|
0.98
|
25
|
30
|
2.5
|
1050
|
120
|
0.732
|
0.481
|
21.3
|
0.34
|
26
|
30
|
5
|
1050
|
180
|
0.2
|
0.148
|
33.2
|
0.423
|
27
|
45
|
2.5
|
1100
|
240
|
0.59
|
0.23
|
20.6
|
1.45
|
3.4.1. Effect of grain size
The experiment results were analyzed with ANOVA of grain size for model in Table 6, which indicates grain size has influence on the leaching of Cr, Ba, and Cr6+. Since P-values less than 0.05 are significant yet values greater than 0.1 are not significant. Fig.10 is represented the regression model demonstrating the impact of grain size on toxicity metals when fluxing agent, setting temperature, and setting time is 2.5%, 1050℃, and 180min, respectively. It can be seen that Ag leaching remains level with grain size. However, Ba leaching was showed significant changes. The leaching of Ba, and Cr6+ increased as grain size up while Cr decreased steadily.
Table 6 ANOVA analysis of Box-Benhnken models for Cr, Cr6+, Ag, and Ba with grain size of D50
Source
|
Sum of Squares
|
Mean Square
|
F-value
|
P-value
|
|
Cr
|
1.71
|
1.71
|
27.19
|
0.0002
|
significant
|
Ag
|
0.0114
|
0.0114
|
2.89
|
0.1033
|
not significant
|
Ba
|
175.57
|
175.57
|
50.75
|
< 0.0001
|
significant
|
Cr6+
|
0.7091
|
0.7091
|
14.78
|
0.0023
|
significant
|
3.4.2. Effect of fluxing agent
Table 7 depicts that fluxing agent of CaO makes marked impression on immobilization of all toxicity metals in the experiment. As can be seen from the Fig.11, great changes have taken place in leaching of Ba, Cr, and Cr6+ with grain size, setting temperature, and setting time is D50=45μm, 1050℃, and 180min, respectively. Surprisingly, Ba leaching sharply went up with content of CaO rising between 0 to 1%. Then, the rate of increase slow down after 1%. The situation nearly reached a peak of 4%. The adding of CaO as fluxing agent helps to immobilize Ag, Cr, and Cr6+. But it is limited to Cr and Cr6+ in the number of 4.5% for CaO.
Table 7 ANOVA analysis of Box-Benhnken models for Cr, Cr6+, Ag, and Ba with fluxing agent of CaO
Source
|
Sum of Squares
|
Mean Square
|
F-value
|
P-value
|
|
Cr
|
9.32
|
9.32
|
147.81
|
< 0.0001
|
significant
|
Ag
|
0.645
|
0.645
|
163.28
|
< 0.0001
|
significant
|
Ba
|
1994.6
|
1994.6
|
576.54
|
< 0.0001
|
significant
|
Cr6+
|
9.63
|
9.63
|
200.77
|
< 0.0001
|
significant
|
3.4.3. Effect of setting temperature
As is illustrated in the Table 8, setting temperature has influence on Ag, Ba, and Cr6+ leaching when grain size, fluxing agent, and setting time is D50=45μm, 2.5%, and 180min, respectively. The tendency of heavy metals leaching is proved in Fig.12. Ag and Ba leaching were gradual decline in higher temperature which are in contrast to Cr6+. Ba is the most sensitive to temperature.
Fig.13 is used to explore interaction between fluxing agent and setting temperature on Cr6+, Cr, and Ba by 3D surface mode, which grain size of D50 is 45μm, and setting time is 180min. The leaching of Cr6+ and Cr showed an extreme influence on interaction rather than one factor. Therefore, it was decreased when adding dosage of CaO and receding the temperature at the same time. However, Ba leaching values were control with interaction approaching one factor for fluxing agent of CaO. The trends of change for three metals are similar.
Table 8 ANOVA analysis of Box-Benhnken models for Cr, Cr6+, Ag, and Ba with setting temperature
Source
|
Sum of Squares
|
Mean Square
|
F-value
|
P-value
|
|
Cr
|
0.0917
|
0.0917
|
1.45
|
0.2511
|
not significant
|
Ag
|
0.332
|
0.332
|
84.05
|
< 0.0001
|
significant
|
Ba
|
56.64
|
56.64
|
16.37
|
0.0016
|
significant
|
Cr6+
|
1.58
|
1.58
|
32.92
|
< 0.0001
|
significant
|
3.4.4. Effect of setting time
Table 9 is clearly shown that setting time only has significant influence on immobilization of Cr6+ with grain size, fluxing agent, and setting temperature being D50=45μm, 2.5%, and 1050℃, respectively. Fig.14 proves that high temperature has negative impact on immobilization of Cr6+. Especially, Cr6+ leaching showed a steep growing after 1035℃.
Table 9 ANOVA analysis of Box-Benhnken models for Cr, Cr6+, Ag, and Ba with setting time
Source
|
Sum of Squares
|
Mean Square
|
F-value
|
P-value
|
|
Cr
|
0.1508
|
0.1508
|
2.39
|
0.148
|
not significant
|
Ag
|
0.0129
|
0.0129
|
3.27
|
0.084
|
not significant
|
Ba
|
13.06
|
13.06
|
3.78
|
0.0758
|
not significant
|
Cr6+
|
0.5994
|
0.5994
|
12.5
|
0.0041
|
significant
|
Fig.15 is shown as the interaction of fluxing agent of CaO and setting time on Cr6+, Cr, and Ba in 3D model when grain size of D50 and setting temperature are set up in 45μm, and 1050℃, respectively. The tendency is like the results of the interaction between fluxing agent and setting temperature. However, the trend for change of Cr and Cr6+ was slow down.
3.4.5. Optimal prediction analysis
The model condition was predicted response by desirability function in Box-Benhnken response surface. The leaching of heavy metals as functions were set for the goal under the standard limitation for numerical optimization, which was used to calculate the highest desirability and to design the subsequent experiment. In this work, the ideal selection focused on a desirability of 1.00, and the reduction of energy consumption. Thus, the optimal condition is grain size of D50=30μm, fluxing agent of CaO=5%, setting temperature=1050℃, and setting time=180min, resulting in the leaching of Cr, Ag, Ba, and Cr6+ is 0.121, 0.131,30.25, and 0.266, respectively. The theoretical variables were compared with experiment in three tests with average results being shown in Table 10. The experimental results supported the theoretical optimization that the average is observed to be 0.129, 0.16, 31.51, and 0.23, respectively, which implies that Box-Benhnken response surface contributes to optimize the sintering process.
Table 10 Leaching of Cr, Cr6+, Ag, and Ba in optimized and actual experiment (mg/L)
Toxicity metals
|
Optimal value
|
Experimental value in average
|
Cr
|
0.121
|
0.129
|
Ag
|
0.131
|
0.16
|
Ba
|
30.25
|
31.51
|
Cr6+
|
0.266
|
0.23
|
3.5. Sintering mechanism of MSWI fly ash
The fly ash was treated with direct heating for sintering accumulating limited internal energy. Therefore, the results of crystallinity and solidification for toxicity metals was shown a low performance. crystal is formed in geometry as rules spontaneously when temperature up in a period by breaking bonds. Rising temperature provides thermal energy for internal motif to gain kinetic energy(Pan et al. 2020). The particles are formed from other aggregation structures to crystalline state, which storage the least internal energy and possess the most stable shape. At the same time, arrangement space is sharply reduced for whole atoms as the result of density decreasing as characterization. In addition, solid materials are transferred and interchanged due to the dual effects of physics and chemistry, exposing the shortage with an immediate heating.
In this study, the treatment of sintering was improved through adding two additional factors as the method to promote crystallization in fly ash and immobilization of heavy metals. The specific mechanism is shown in Fig.16 giving a rational explanation for inherent change during treatment. Atoms keep moving with certain speed and distance in particles for fly ash. It is simple for toxicity metals to depart from particles when leaching. Changing grain size with ball milling was utilized to provide energy under the mechanical work for atoms and Irregular surface. Besides, it is useful to reduce the distance of particles(Luo et al. 2017). Fly ash has saved a large energy before sintering. CaO contributes to decrease melting point in SiO2-CaO or SiO2-Al2O3-CaO system as high-melting oxide(Zha et al. 2021). It is helpful to reach the sintered phase and reduce the energy for supporting the growth of crystal. Fluxing agent was used to shorten sintering time and reduce consumption. The results of sintering were considerable with previous treatment. Particles are push to a drastic motion as temperature growing. Mineral phases, such as CaTiO3, and Ca12Al14O33 are continuously generated at setting temperature and time, occupying a large space in particle. Toxicity metals are locked and surrounded by other crystals when particles compacted. Gigantic binding force becomes the limitation on leaching of metals.
The findings of this study were restricted to other factors that may contribute to sintering, for example, sintering atmosphere, and pressure. The leaching of As, Pb, Cd, Cu, Ni, Zn, Mn, Hg, Be, Se, and fluoride hardly was detected as the values being over the lowering. Thus, they were left out to consider the relation with setting temperature, setting time, grain size of fly ash, and dosage of fluxing agent. Besides, the findings did not imply the leaching characteristic for fly ash in the environment permanently.