Preparation of Fe-C porous filter
Optimization of operational parameters
Based on one-factor pre-experiment, an orthogonal experimental design of four factors (preheating temperature, preheating time, sintering temperature, sintering time) and three levels were conducted to optimize the process parameters, and removals of TP and TOC were employed as the evaluation index. Table 1 summarizes the results of L9 (34). The range analysis indicated that sintering temperature was the most important factor for TP removal, while the preheating temperature was the most important for TOC removal. Appropriate process parameters were preheating temperature 280 ℃, preheating time 15 min, sintering temperature 1080 ℃, and sintering time 11 min, separately. The corresponding TOC and TP removals were 44.4% and 75.2%, and physical properties including breaking and wear rate, porosity, water absorption, and void fraction of the filter obtained were 2.63%, 31.2%, 22.7%, and 56.5%, respectively, satisfying the Chinese Standard of CJ/T 229–2008 and GB/T 17431.2–2010.
Table 1
Performance indexes values of orthogonal test
Test case
|
Horizontal combination
|
Breaking and wear rate /%
|
Void fraction/%
|
Porosity/%
|
Water absorption/%
|
TP removal/%
|
TOC removal/%
|
1
|
A1B1C1D1
|
2.63
|
56.5
|
31.2
|
22.7
|
44.4
|
75.2
|
2
|
A1B2C2D2
|
3.56
|
57.4
|
30.8
|
25.6
|
37.0
|
75.9
|
3
|
A1B3C3D3
|
2.84
|
55.6
|
33.6
|
23.3
|
29.6
|
75.4
|
4
|
A2B1C2D3
|
2.28
|
55.9
|
35.6
|
28.8
|
34.6
|
75.0
|
5
|
A2B2C3D1
|
2.19
|
55.5
|
32.4
|
21.7
|
26.4
|
74.6
|
6
|
A2B3C1D2
|
2.26
|
56.6
|
32.0
|
24.7
|
29.7
|
70.0
|
7
|
A3B1C3D2
|
1.65
|
53.1
|
27.4
|
17.2
|
28.4
|
75.4
|
8
|
A3B2C1D3
|
1.22
|
51.9
|
36.0
|
30.9
|
25.7
|
75.5
|
9
|
A3B3C2D1
|
1.50
|
49.8
|
37.1
|
29.0
|
39.5
|
74.9
|
Breaking and wear rate /%
|
K1
|
3.01
|
2.19
|
2.04
|
2.11
|
Void fraction/%
|
K1
|
56.5
|
55.2
|
55.0
|
53.9
|
K2
|
2.24
|
2.32
|
2.45
|
2.49
|
K2
|
56.0
|
54.9
|
54.4
|
55.7
|
K3
|
1.46
|
2.20
|
2.23
|
2.11
|
K3
|
51.6
|
54.0
|
54.7
|
54.5
|
R
|
1.55
|
0.14
|
0.41
|
0.38
|
R
|
4.88
|
1.15
|
0.63
|
1.78
|
Porosity /%
|
K1
|
31.9
|
31.4
|
33.1
|
33.6
|
Water absorption/%
|
K1
|
23.9
|
22.9
|
26.1
|
24.5
|
K2
|
33.4
|
33.1
|
34.5
|
30.1
|
K2
|
25.0
|
26.0
|
27.8
|
22.5
|
K3
|
33.5
|
34.2
|
31.1
|
35.1
|
K3
|
25.7
|
25.6
|
20.7
|
27.6
|
R
|
1.64
|
2.80
|
3.40
|
4.99
|
R
|
1.81
|
3.15
|
7.04
|
5.17
|
TP removal/%
|
K1
|
37.0
|
35.8
|
33.3
|
36.8
|
TOC removal/%
|
K1
|
75.5
|
75.2
|
73.6
|
74.9
|
K2
|
30.2
|
29.7
|
37.0
|
31.7
|
K2
|
73.2
|
75.3
|
75.3
|
73.8
|
K3
|
31.2
|
32.9
|
28.1
|
30.0
|
K3
|
75.3
|
73.4
|
75.1
|
75.3
|
R
|
6.81
|
6.08
|
8.89
|
6.79
|
R
|
2.34
|
1.90
|
1.70
|
1.51
|
Additive content
Effects of additive including straw, starch, iron powder, and foam on physical characteristics, TOC and TP removals of the filter were investigated based on the optimal operational condition and results were shown in Table 2. Physical characteristics of all filter satisfied the standard of CJ/T 299–2008 and GB/T 17431.2–2010. TOC removal increased by 9.1% as the content of straw increased from 5–10%. During the pyrolysis process, the hemicellulose and cellulose in the straw were thermally decomposed and produce condensable vapors, and gas, which improved the pore structure of the filter. Besides, it yields biochar, which could form ICME with iron in the filter, enhancing the removal of organic carbon. However, TP removal did not change significantly. It may be attributed to the phosphorus content in the straw.
Table 2 Effect of mixing ratio of raw material on characteristics of the filter
|
Sediment:straw:starch:iron powder: foam
|
Physical characteristics
|
Pelletization
|
TP removal/%
|
TOC removal/%
|
straw
|
78.5:5:10:5:1.5
|
Standard
|
easy
|
49.2
|
66.2
|
76:7.5:10:5:1.5
|
Standard
|
easy
|
48.1
|
72.4
|
73.5:10:10:5:1.5
|
Standard
|
easy
|
48.9
|
75.3
|
starch
|
78.5:10:5:5:1.5
|
Standard
|
easy
|
45.3
|
76.6
|
76:10:7.5:5:1.5
|
Standard
|
easy
|
54.7
|
78.5
|
73.5:10:10:5:1.5
|
Standard
|
easy
|
48.9
|
75.3
|
iron powder
|
78.5:10:7.5:2.5:1.5
|
Standard
|
easy
|
58.2
|
73.5
|
76:10:7.5:5:1.5
|
Standard
|
easy
|
62.0
|
64.8
|
73.5:10:7.5:7.5:1.5
|
Standard
|
easy
|
68.0
|
55.8
|
71:10:7.5:10:1.5
|
Standard
|
easy
|
69.4
|
47.2
|
foam
|
76:10:7.5:5:1.5
|
Standard
|
easy
|
54.3
|
56.0
|
75.5:10:7.5:5:2
|
Standard
|
easy
|
65.1
|
64.2
|
74.5:10:7.5:5:3
|
Standard
|
easy
|
67.8
|
83.8
|
74:10:7.5:5:3.5
|
Standard
|
difficult
|
65.5
|
58.1
|
Both TOC and TP removals showed a trend of increase followed by a decrease with the increase of starch or foam. The maximum TOC and TP removals were obtained at starch 7.5% and foam 3%. The values were 78.5% and 54.7%, and 83.8% and 67.8%, separately. Starch and foam were versatile binder and pore-forming materials, favorable for the information of abundant pores. However, too much starch may lead to the enlargement of the pore diameter and decline of surface area due to the aggregation of starch-water mixture. The overabundant of foam in the raw materials would reduce the mechanical strength. The raw body was hard to form the expected shape at foam content 4% in the experiment.
Phosphorus removal increased gradually with the increase of iron powder, with the maximum value of 69.4% at iron content of 10%. It may be described to the combination of iron ion and phosphorus. However, TOC removal decreased when iron increased from 2.5–10%. This result was completely contrary to our expectation that increase of iron could enhance the ICME process in the filter. The probable reason for this was the eutectic temperature property of iron and iron oxide. They promoted the flux of liquid phase, and subsequently prevented the volatile gas from releasing and formed a denser sinter body with low porous structure. Shih et al. (2006) and Xu et al. (2009) reported that melting point of samples was lowered by the presence of higher Fe2O3 contents, enhancing the formation of liquid phases and FeO with higher viscosity. During the sintering process, we also found the leakage of iron at high iron amount, consequently reducing the iron content in the filter.
Characterization of the filter
Textural morphological characteristics
The surface of the filter was rough and irregular. Abundant pores, small grooves and cracks were observed on the surface of the material (Fig. 2). The SEM on the cross section showed pores of different sizes and shapes, and agglomerated crystals. BET surface area was 3.32 m2 g-1, and total volume of mesopore and macropore was 0.00662 cm3 g-1, with average pore diameter of 10.05 nm. Pore distribution showed that the mesopore accounted for 85.2% of whole pores.
EDS analyses
Element composition analysis by EDS revealed that the filter was essentially an association of O, Si, Al, Fe, Zn, K, Ca, C, S, and Mg atoms (Fig. 3a). The Si and Al made up the whole skeleton structure, supporting the mechanical strength of the filter. Metal ions including Fe, Zn, K, Ca, Al were effective to combine with phosphate to remove the phosphorus from the wastewater. The Fe and C may inform the ICME, consequently enhancing the degradation of organic matter.
Crystalline phase (XRD) analyses
Figure 3b shows the XRD pattern of the filter. The main minerals included quartz (SiO2), magnetite (Fe3O4), hematite (Fe2O3) and a new crystallized muscovite (KAl2(Si3Al)O10(OH)2). The muscovite was a member of the mica group of silicate minerals with strongly layered sheet like structure, thus enhancing the mechanical strength and breaking tenacity of the filter.
Leaching of potential toxic metals
In order to investigate the environmental safety of the filter, heavy metal leaching test was performed. The highest concentrations of Zn and Cu were 1.890 and 0.160 mg L− 1 among pH 2–10, separately, much lower than the standard values (Cu 100 mg L− 1, Zn 100 mg L− 1). Both Pb and Cr were not detected in the leachate.
Dynamic purification test for polluted river water
P removal
Total phosphorus concentrations in the effluents of 3# and 4# were significantly lower than those of 1# and 2#, suggesting the filter effectively improved P removal from the polluted river water. During the test, the influent TP concentration was 0.16–0.80 mg L− 1, TP concentrations in effluents of 3# and 4# were decreased to among 0-0.14 mg L− 1, respectively. Aeration could not enhance P removal effectively as effluent TP concentration difference either between 3# and 4#, or between 1# and 2# was not distinct significantly. The result indicated that P was removed mainly via adsorption by the filter. Effluent TP concentrations both in 3# and 4# did not change greatly in the test, demonstrating the filter was effective yet to remove P after 35 days of continuous operation.
Organic carbon removal
As shown in Fig. 5, effluent TOC concentrations in effluents of 3# and 4# were lower than 1# and 2#, and the differences were significant at higher influent TOC concentration. Aeration enhanced TOC removal by the filter. When influent TOC concentration was 3.41–6.89 mg L− 1, average effluent TOC concentration of 4# was 11.4% lower than 3#. This results might indicate that the organic matter removal by the filter was attributed to the degradation by microorganism adhered to the surface of the filter, and the ICME process. Furthermore, the aeration was favorable to the micro-electrolysis process. This was consistent with observations of Han et al. (2016). They reported that the aeration not only improved the micro-electrolysis process, but also protected the filter from harden and desactivation. Therefore, the aerated biofilter with the filter might be a feasible alternative for removing organic carbon matter.
N removal
The differences in concentration variations of NH4+-N, and TN in 1#, 2#, 3#, and 4# were similar with TOC (Fig. 6). Effluent NH4+-N concentrations in four treatments were 3.54–7.30, 3.32–7.08, 1.99–6.92, and 1.60–6.80 mg L− 1 separately, at influent NH4+-N concentration 3.65–8.19 mg L− 1. While effluent TN concentrations were 4.30–7.46, 3.43–7.24, 2.30–7.05, and 1.81–6.94 respectively. Thus, the filter was effective to remove nitrogen in the waters, and aeration could enhance the process. However, nitrate concentrations in all treatments were not distinct between effluent and influent, and the values were less than 0.3 mg L− 1. It may indicate that nitrogen in the water was removed mainly through adsorption by the filter. Nitrification and denitrification process were little. The reason for this may be ascribed to the low microorganism content in the raw river water, and low propagation rate of the microorganism for insufficient nutrient in the water.
Environmental safety evaluation
During the operation, effluent pH of 3# and 4# was similar with the influent, and the value was among 7.0-8.4. Concentrations of Cu, Zn, and Fe were 0.001–0.003, 0.002–0.003, 0.002–0.011 mg L− 1, separately, much lower than the class 1 standard value of environmental quality standards for surface water. Both Pb and Cr were not detected in the effluents.