TiO2 recovery from spent SCR catalyst
Table 1
The composition of R-TiO2 at different alkaline leaching temperatures
|
TiO2
|
WO3
|
SiO2
|
SO3
|
Al2O3
|
CaO
|
P2O5
|
others
|
60 ℃
|
94.910
|
1.763
|
1.563
|
0.320
|
0.463
|
0.744
|
0.106
|
0.133
|
80 ℃
|
95.945
|
1.699
|
0.958
|
0.615
|
0.251
|
0.218
|
0.136
|
0.178
|
100 ℃
|
97.616
|
1.376
|
0.179
|
0.509
|
0.120
|
0.000
|
0.141
|
0.058
|
120 ℃
|
97.774
|
1.008
|
0.177
|
0.658
|
0.059
|
0.000
|
0.238
|
0.086
|
140 ℃
|
97.602
|
1.001
|
0.234
|
0.770
|
0.080
|
0.000
|
0.232
|
0.081
|
160 ℃
|
97.604
|
0.997
|
0.346
|
0.775
|
0.072
|
0.000
|
0.133
|
0.073
|
The composition of R-TiO2 at different alkaline leaching temperatures was explored due to alkali leaching temperature as the key factor of the recovery process. The chemical composition of R-TiO2 was investigated by XRF, the results are shown in the Table 1. Meanwhile, the effects of the alkaline leaching temperatures on the purity of R-TiO2 are shown in Fig. 1. The purity of R-TiO2 increases with the increase of alkali leaching temperature. At the outset, there is a relatively rapid increase in the purity of R-TiO2 when the alkali leaching temperature span extends from 60 to 100°C. The purity of R-TiO2 increases from 94.910–97.774%. However, the purity of R-TiO2 keeps stable and the effect of temperature is minimal when the alkali leaching temperature varies from 100 to 160°C. It may be referred to that TiO2 reacts with NaOH almost completely after the temperature reaches 100°C. Thus, RT60, RT80 and RT100 with a large purity gap were selected for further research.
The crystal structure of the R-TiO2 is shown in Fig. 2(a). The R-TiO2 was obtained from spent SCR catalyst under the alkali temperature was 100°C (RT100). The major peaks at 25.3, 37.8, 48.0, 53.7, 55.0, 62.5, 68.6, 70.2, 75.0 ° and 82.6 ° are consistent with the standard anatase TiO2 phase (JCPDS No. 21-1272) and no other miscellaneous peaks. Meanwhile, the crystal structure of alkali leaching residues and the spent SCR catalysts appear the anatase TiO2 phase, indicating that the R-TiO2 remains the crystal structure intact all along during the recovery process. Anatase TiO2 as photocatalyst with a better photocatalytic effect. The morphology of R-TiO2 were analyzed by SEM and TEM, as shown in Fig. 2(b, c). The R-TiO2 is obviously aggregated with diameters of 50-100 nm and shows homogeneous particles. Besides, the R-TiO2 particles are seriously agglomerated, which might be due to the high temperature sintering of SCR catalysts during the long-term denitrification catalytic process.
Characterization of the R-TiO2/g-C3N4 photocatalyst
The crystal structure of g-C3N4, R-TiO2 and TCNX samples are shown in Fig. 3. The diffraction peak at 13.2 ° corresponds to the (100) plane of g-C3N4 which is attributed to the planar repeating of trisulfide triazine units, the other peak appear at 27.4 ° matches with the (002) planes of g-C3N4 which ascribes to the interlayer stacking of conjugated aromatic system [32, 33]. After the same amount of g-C3N4 is combined with R-TiO2 with different purity respectively, there are diffraction peaks at 27.4 °, while no diffraction peaks at 13.2 ° are visibly detected in the XRD patterns of TCNX samples. It may be caused by the low loading rate and crystallinity of g-C3N4. In the meantime, the peak shifts and phase purities of TCNX samples are no obvious change, which implies that the lattice structure of g-C3N4 and R-TiO2 samples is not affected.
The morphologies and structure of g-C3N4 and TCN100 samples were characterized by SEM and TEM, as shown in Fig. 4. SEM images reveal the smooth lamellar structure of g-C3N4 (Fig. 4a), and for TCN100, the R-TiO2 is loading on the g-C3N4 sheet layer (Fig. 4c). Similarly, TEM images describe the lamellar structure of g-C3N4 layered on top of each other in detail (Fig. 4b). As we can see from Fig. 4d, the R-TiO2 particles are loaded on the g-C3N4 thin sheet to successfully produce the R-TiO2/g-C3N4 photocatalysts via self-assembly method.
The FTIR spectra of g-C3N4, R-TiO2 and TCNX samples are displayed in Fig. 5. All samples are observed to adsorb H2O molecules and physical hydroxyl groups on the surface of the materials at the peak positions of 1642 cm−1 and 3000-3400 cm−1 [22, 34]. For pure g-C3N4, the distinct bands at 1241, 1323, 1409, 1462 and 1568 cm−1 are caused by the heterocycle stretching modes of aromatic sp2 C-N bond, and the characteristic mode of triazine units lead to the presence of the peak at 808 cm−1 [35, 36]. For R-TiO2, the absorption band at 400-700 cm−1 is assigned to Ti-O-Ti bridging stretching vibration modes [37]. For samples of TCN60, TCN80 and TCN100, the characteristic stretching modes of g-C3N4 and R-TiO2 can be observed, no apparent new stretching vibration spectrum appears. This strongly confirms the formation of R-TiO2/g-C3N4 composite materials and corresponds to the XRD analysis results of Fig. 3.
The surface chemical states of the R-TiO2, g-C3N4, and TCN100 samples are measured by XPS which were shown in Fig. 6. The survey spectra in Fig. 6a show that the g-C3N4 contains C, N, O elements, R-TiO2 contains C, Ti, O elements and TCN100 includes all the expected elements (C, N, O and Ti). As shown in Fig. 6b, the C 1s spectrum of g-C3N4 can be deconvoluted into three distinct peaks at 284.8, 285.9 and 288.1 eV by the Gaussian analysis method. The peak at 284.8 is the signal of the sp2 C-C bonds, the second peak located at 285.9 eV is owed to the C=N bonds, and another peak at 288.1 eV indicates the C-N or C-(N)3 groups [38]. The peak positions of TCN100 in the C 1s spectrum are closed to that of g-C3N4, both the peak of C=N shifts 0.4 eV and the peak of C-N or C-(N)3 shifts 0.3 eV towards higher binding energy. The N 1s spectrum of g-C3N4 in Fig. 6c is composed of three peaks at 398.2, 399.4, and 400.6 eV, which correspond to sp2-hybridized nitrogen (C-N=C), tertiary nitrogen (N-(C)3) and secondary amino groups (C-N-H), respectively [39]. Compared with g-C3N4, these peaks of TCN100 in the N 1s spectrum have a shift of 0.2, 0.8 and 0.6 eV towards higher binding energy. Fig. 6d illustrates the O 1s spectra of R-TiO2 and TCN100. The peaks at approximately 531 and 529 eV put down to Ti−O band and the surface -OH groups [40, 41]. Additionally, the two peaks in O 1s spectrum of TCN100 move 0.1 eV towards lower binding energy in comparison to g-C3N4. In the Ti 2p spectrum (Fig. 6e), the two obvious peaks of R-TiO2 at 458.3 due to the Ti 2p3/2 and 464.1 eV ascribe to the Ti 2p1/2, respectively [42]. However, these peaks of TCN100 all shift 0.1eV towards lower binding energy. Besides, there are no peaks of Ti-C and Ti-N bond in the Ti 2p spectrum of TCN100 which indicate that g-C3N4 does not enter into the lattice structure of R-TiO2. Distinctly, for TCN100, the binding energy of C 1s and N 1s peaks become higher, while the binding energy of O 1s and Ti 2p peaks become lower, which demonstrates the TiO2/g-C3N4 heterojunction was formed [43].
Optical properties of the R-TiO2/g-C3N4 photocatalyst
The PL spectra of g-C3N4, R-TiO2 and TCNX samples in the range of 400−600 nm under the excitation wavelength of 360 nm are presented in Fig. 7. Generally speaking, the low PL intensity means the low the recombination rate of the photon-generated carriers [30]. For pure g-C3N4, the position of the distinct emission peak at about 462 nm could put down to the recombination of the photon-generated carriers produced by g-C3N4 [44]. Compared with g-C3N4, the PL intensity of TCNX samples is much lower, which means that TiO2 inhibits the recombination of photo-generated carriers of g-C3N4. Meanwhile, the PL intensity of TCN100 decreases most obviously. The R-TiO2 shows low PL intensity in the visible light range because it hardly absorbs visible light, and the combination with g-C3N4 strengthens the absorption ability of visible light of the R-TiO2. Therefore, it could be expected that low recombination rate of photo-generated carriers in the TCNX samples could lead to enhancement in photocatalytic performance.
The UV-vis DRS spectra of R-TiO2, g-C3N4 and TCNX samples are shown in Fig. 8. As shown in Fig. 8(a), the absorption edges for g-C3N4 and R-TiO2 are about 468 and 420 nm, indicating that pure g-C3N4 shows an extensive visible light absorption and the R-TiO2 displays absorption in the ultraviolet region. Under the modification of g-C3N4, the visible absorption ability of R-TiO2 is enhanced, and the absorption edge of TCN100 reaches approximately 475 nm. According to the plot of (Ahν)2 versus photo energy (hν), as shown in Fig. 8(b), it can be obtained that the bandgap energies of R-TiO2 and g-C3N4 are approximately 3.1 and 2.71 eV, separately. The band gap energy of R-TiO2 which is slightly narrower than the anatase TiO2. The band gap energies of TCN60, TCN80, and TCN100 samples are nearly 2.69, 2.67, 2.63 eV, respectively. Their band gap energies are significantly narrower than R-TiO2, which implies that TCNX samples can absorb more visible light and have better photocatalytic activity. Their band gap energies are significantly narrower than R-TiO2, which implies that TCNX samples can absorb more visible light and have better photocatalytic activity.
The Photocatalytic Activity Evaluation
The photocatalytic activities of R-TiO2, g-C3N4 and TCNX samples were further studied by RhB photodegradation under the irradiation of visible light, as shown in Fig. 9. The photodegradation RhB curves of different catalysts are shown in Fig. 9(a). The photodegradation efficiencies of RhB are 50.63% and 94.07% over R-TiO2 and g-C3N4 when the systems are irradiated for 80 min, respectively. Meanwhile, the photodegradation efficiencies of RhB over TCNX samples are higher than R-TiO2 and g-C3N4. Especially, the degradation activities of TCNX samples are gradually enhanced with the increase in the purity of R-TiO2, which shows an order of TCN100>TCN80>TCN60. The TCN100 appears the best photocatalytic activity, the degradation efficiency reaches 97.11% when the systems are irradiated with visible light for only 40 min.
The process of photodegradation of RhB over different samples accorded with the pseudo-first-order kinetic model (lnC0/Ct = kt). As shown in Fig. 9(b). The RhB degradation rate constant from R-TiO2 is 0.007(k = 0.007 min−1), which is caused by wide band gap energy (3.1 eV) and relatively weak absorption of visible light. The RhB degradation rate constant from g-C3N4 is 0.032 (k = 0.032 min−1), which is about 4.5 times that of R-TiO2. Compared with R-TiO2, g-C3N4 has an excellent absorption of visible light. The RhB degradation rates are obviously improved as the increase in the purity of R-TiO2, For the TCNX samples and the order of the RhB degradation rate constant is TCN100>TCN80>TCN60. The RhB degradation rate constant from TCN100(k = 0.072 min−1) is about 10.2 times that of R-TiO2, which indicates that the photocatalytic activity of R-TiO2 modified by g-C3N4 is markedly improved.
For comparison, tetrabutyl titanate was selected as the TiO2 precursor to prepare TiO2/g-C3N4 photocatalyst under the same conditions, which is named PTCN. The PTCN can degrade 97.64% RhB within 80 min under the irradiation of visible light. Meanwhile, the degradation rate constant from PTCN is 0.038(k = 0.038min−1). Although PTCN with better photodegradation efficiency, its degradation efficiency is still lower than that of TCNX photocatalysts. The RhB degradation rate constant from TCN100 is about 1.9 times that of PTCN, which suggests it is quite feasible for TiO2/g-C3N4 photocatalyst to be prepared with TiO2 recovered from spent SCR catalysts.