CHAPTER II Photocatalytic Activities of Layered Niobate Perovskite
3.3. Results and discussion
Characterization of Prepared RbLaNb2O7 and HLaNb2O7 and Hybridization with Pt Photocatalysts. Fig. 3-1 shows the crystal structure of RbLaNb2O7: a double-layer perovskite slab is interleaved by Rb; the perovskite slab is composed of
46
NbO6 octahedra sheets with La ions in the interstices of the NbO6 octahedra.
According to the diffraction line at 2θ = 7.94° in the XRD data with monochromated CuKα radiation, the interlayer distance of RbLaNb2O7 is 1.11 nm.
Figure 3-1 Crystal structure of RbLaNb2O7.
Fig. 3-2 shows the powder XRD patterns of (a) RbLaNb2O7, (b) Pt-RbLaNb2O7, (c) HLaNb2O7, and (d) Pt-HLaNb2O7. From these patterns, the samples appeared to contain single phases of RbLaNb2O7 or HLaNb2O7, with interlayer distances of approximately 1.1 nm. The interlayer distance slightly decreases by protonation and does not change by Pt deposition. However, the diffraction patterns, especially intensity ratios, apparently change by protonation because of the large decrease of the electrons within the interlayer space. Such a large decrease provides a small atomic scattering factor, which leads to a small crystal structure factor. For the hybrids with Pt, the existence of Pt-related compounds cannot be observed in these patterns.
Perovskite
←interlayer cation; A’
←cation A
←BX6 octahedra
47
Figure 3-2 XRD patterns of (a) RbLaNb2O7, (b) Pt-RbLaNb2O7, (c) HLaNb2O7, and (d) Pt-HLaNb2O7.
For determining the amount of Pt in compounds, the XRF of the samples was measured. Table 3-1 shows the results. These results show that Rb is absent in the protonated sample and Pt-HLaNb2O7, which is attributed to successful protonation.
Moreover, after Pt deposition, only marginal amounts of Pt exist in Pt-RbLaNb2O7 and Pt-HLaNb2O7. Meanwhile, the amount of Rb decreases in Pt-RbLaNb2O7. H2PtCl6·6H2O is supposed to be an acid, which can react with RbLaNb2O7. After ion exchange by H2PtCl6 to form RbCl, a part of Rb is consumed by ion exchange with protons, and the Rb/La ratio decreases.
Table 3-1 Ratio of Pt/La and Rb/La in compounds.
Cu / degree
In te n si ty ( a. u .)
a
b
c
d
10 20 30 40 50
48
RbLaNb2O7 Pt-RbLaNb2O7 HLaNb2O7 Pt-HLaNb
2O7
Pt/La 0.000 0.013 0.000 0.011
Rb/La 1.040 0.934 0.000 0.000
Fig. 3-3 shows FE-SEM micrographs of (a) RbLaNb2O7, (b) Pt-RbLaNb2O7, (c) HLaNb2O7, and (d) Pt-HLaNb2O7. The particle shapes are seemed to be formless for all samples. These micrographs cannot confirm the Pt nanoparticles on outer surface of the samples. The mean particle sizes of all samples are similar from 200nm to 500nm. Such similar size may be due to no change of shape by the protonation and/or hybridization of Pt processes. In consistence with the SEM observation, we also estimate the specific surface areas of all samples based on the corresponding geometric shape, particle size (from FESEM), and the real density. The specific surface area values are around1-3 m2/g.
49
Figure 3-3 FE-SEM micrographs of (a) RbLaNb2O7, (b) Pt-RbLaNb2O7, (c) HLaNb2O7 and (d) Pt-HLaNb2O7.
50
Figure 3-4 XPS spectra of (I) both Pt-included samples, (II) Pt-RbLaNb2O7, and (III) expanded Pt-HLaNb2O7.
XPS was employed for analyzing the chemical state of Pt in the samples. Fig. 3-4 shows the high-resolution XPS spectra of Pt 4f in Pt-RbLaNb2O7 and Pt-HLaNb2O7. These spectra seem to be composed of two states of Pt for both hybrids. In Fig. 3-4(II) Pt-RbLaNb2O7, the binding energies of both states for Pt 4f7/2 are 70.4 eV and 71.8 eV, while those for Pt 4f5/2 are 73.8 eV and 75.1 eV, respectively. Binding energies of approximately 70.4 eV and 73.8 eV are attributed to the Pt metal, while the other
80 78 76 74 72 70
Pt 4f
Intensity (a. u.)
Binding Energy / eV Pt-RbLaNb2O7
Pt-HLaNab2O7 I
80 78 76 74 72 70
700 800 900 1000 1100 1200 1300 1400
71.8 eV
Intensity / counts
Binding Energy / eV 4f7/2 70.4 eV 4f5/2
73.8 eV
Pt 4f II
75.1 eV Pt-RbLaNb2O7
80 78 76 74 72 70
680 700 720 740 760 780 800 820
4f7/2 73.6 eV 4f5/2
74.2 eV
Intensity / counts
Binding Energy / eV 4f5/2
77.0 eV
4f7/2
70.8 eV
III Pt-HLaNb2O7 Pt 4f
51
peaks at approximately 71.8 eV and 75.1 eV are attributed to the platinum oxide layer, which covered the surface of Pt metal particles. This spectrum apparently indicates that Pt metal particles are located on the RbLaNb2O7 surface in Pt-RbLaNb2O7. On the other hand, in Fig. 3-4(III) Pt-HLaNb2O7, the trimodal peak can be split into four peaks, as can be observed in the spectrum. The binding energies of 70.8 eV and 74.2 eV are attributed to Pt 4f7/2 and Pt 4f5/2 of Pt metal, respectively. On the other hand, the other peaks with binding energies of approximately 73.6 eV (4f7/2) and 77.0 eV (4f5/2) are attributed to Pt2+. By comparing Pt-RbLaNb2O7 and Pt-HLaNb2O7, the peak intensity for the Pt metal drastically decreases concurrently, and the peak of the surface oxide layer of Pt metal disappears. By contrast, peaks for Pt2+ are observed in Pt-HLaNb2O7, which are not observed in Fig. 3-4(II). In addition, the total intensity of Pt steeply decreases. This small intensity for Pt-HLaNb2O7 suggests that Pt is absent on the surface of the niobate particles. These changes in the spectra confirm that Pt possibly exists as metal particles on the surface of the niobate particles in the case of Pt-RbLaNb2O7 and as cations within the interlayer space in the case of Pt-HLaNb2O7.
At one glance, interlayer Pt2+ is interesting because Pt exists as an anionic complex, PtCl62−, in H2PtCl6. Harada [37, 38] has reported that when PtCl62− ionic complexes in an ethanol–water solution are photoirradiated by UV light, Pt4+ is reduced to Pt0 metal particles via Pt2+, as shown in eq 3-1.
𝑃𝑡𝐶𝑙62− → 𝑃𝑡𝐶𝑙42− → 𝑃𝑡0 (3-1)
52
During the process, PtCl42− can be temporarily dissociated to form Pt2+ before forming Pt0 (metallic state). The HLaNb2O7 sample still exhibits competent proton exchange and can trap temporal Pt2+ by ion exchange during UV irradiation. However, the layered perovskite RbLaNb2O7 can barely trap Pt2+ because of the lack of an exchangeable proton. The ion exchangeability for a proton is probably caused by high molar conductance. The conductance values of H+ and Rb+ are approximately 3.5 10−2 and 7.8 10−3 S m2/mol, respectively. High conductance can facilitate high ion mobility, and protonated samples may exhibit good ion exchangeability.
Further, the samples hybridized with Pt are examined by SXRD. For Pt-RbLaNb2O7, an approximately 5 mass% RbCl phase is observed as an impurity.
However, the diffraction lines of 111 and 200 for Pt metal (d111 = 2.26 and d200 = 1.95, respectively) cannot be detected because of overlap with some peaks of the mother phase RbLaNb2O7. In the case of Pt-HLaNb2O7, most of the peaks are identified as HLaNb2O7. Fig. 3-5 shows the refined SXRD pattern of Pt-HLaNb2O7. The main phase can be refined as HLaNb2O7 (𝐹𝑚3̅𝑚, a = 5.4810(4), b = 20.9305(0), c = 5.4861(8)). The refinement result confirms that a very small amount of Pt metal exists within the hybrid. However, it is difficult to ascertain this amount caused by the very weak diffraction intensity. Then, we evaluated the Pt cation within the interlayer space by the 112, 211, 042, and 240 diffraction lines of the Pt-HLaNb2O7 phase. Fig.
3-6 shows the expanded SXRD patterns of HLaNb2O7 and Pt-HLaNb2O7 and the calculated patterns of HLaNb2O7 and Pt0.5Nb2O7. In the calculated XRD patterns, the
53
presence of Pt cations within the interlayer space apparently increases the intensities of the abovementioned diffractions. From these measured patterns, an apparent peak emerges only in the case of Pt-HLaNb2O7, which possibly results from Pt2+ within the interlayer space. From SXRD and XPS data, we conclude that Pt exists not only outside of particle as metal state but also within the interlayer space as Pt2+ for Pt-HLaNb2O7.
54
Figure 3-5 Refined SXRD pattern of Pt-HLaNb2O7.
R
wp= 2 .4 9 % R
p= 1 .6 8 % S = 3 .1 0
Int en sit y / co unt s
2 / d e g re e 1 0 2 0 3 0 4 0 5 0 6 0 0
2 0 0 0 0
4 0 0 0 0
6 0 0 0 0
8 0 0 0 0
55
Figure 3-6 Expanded SXRD patterns of HLaNb2O7 and Pt-HLaNb2O7 as well as calculated patterns of HLaNb2O7 and Pt0.5Nb2O7.
Band Gap and DOS Calculation of Photocatalysts. Fig. 3-7 shows the comparison of the UV–visible spectra of (a) RbLaNb2O7, (b) Pt-RbLaNb2O7, (c) HLaNb2O7, and (d) Pt-HLaNb2O7. Sharp absorption edges are observed at approximately 330–380 nm. Meanwhile, (b) Pt-RbLaNb2O7 and (d) Pt-HLaNb2O7
exhibit more visible light absorption ability as compared to (a) and (c), which did not (211)
(112) (042) (240)
calculated Pt
0.5LaNb
2O
7observed Pt-HLaNb
2O
7calculated HLaNb
2O
7In te n si ty ( a. u .)
2 / degree
observed HLaNb
2O
711.5 12.0
56
hybridize with Pt. Such an increase in absorbance in the whole visible light range probably indicates the existence of Pt metal particles outside the perovskite grains despite the presence of very small amount of included Pt in (b) and (d). On a comparison with Pt hybrids, the amount of Pt are similar, as shown in Table 3-1, although the peak intensities of metal Pt 4f in the XPS spectra are quite different. In the UV–vis spectra of Pt-RbLaNb2O7, a broad peak is observed at approximately 450 nm, possibly attributed to interfacial charge transfer from relatively abundant Pt metal particles. Fig. 3-8 shows the Tauc plot of these materials by assuming indirect transitions. In the case of the plot calculated by direct transitions, as the curves have a short linear region, the absorption can be regarded as an indirect transition. As shown in Fig. 3-8, the X-intercept corresponds to the estimated band gap energies. The estimated band gap energies of RbLaNb2O7, Pt-RbLaNb2O7, HLaNb2O7, and Pt-HLaNb2O7 are 3.37, 3.15, 3.26, and 3.23, respectively. These band gaps imply that hybridization with Pt leads to the shrinkage of the apparent optical band gap by deposition as a metal state via IFCT. Pt2+ intercalation also marginally shrinks the band gap.
57
Figure 3-7 UV–vis absorption spectra of (a) RbLaNb2O7, (b) Pt-RbLaNb2O7, (c) HLaNb2O7, and (d) Pt-HLaNb2O7.
Figure 3-8 Tauc plot for the estimation of (a) RbLaNb2O7, (b) Pt-RbLaNb2O7, (c) HLaNb2O7, and (d) Pt-HLaNb2O7.
Wavelength / nm
Absorbance
a
b d c
300 400 500 600 700 800
0.0 0.2 0.4 0.6 0.8
h / eV
( h )
0.5a
c b d
3.0 3.5 4.0
0.0
1.0
2.0
3.0
58
Fig. 3-9 shows the DOS curve calculated by first-principles DFT simulation. The insets show simulation models used in this study. In the Pt-HLaNb2O7 model, 1/4th of the H+ position is replaced by Pt2+, while the other 1/4th of the H+ position is a vacancy. Hence, the chemical formula is Pt1/4H1/2LaNb2O7. The amount of Pt in the model is very high as compared to the actual sample; nevertheless, we can consider the ion-exchange effect by Pt using this model. For RbLaNb2O7, a band gap is observed, and valence and conduction bands are composed of O 2p and Nb 4d, respectively. For HLaNb2O7, an empty level composed of H 1s and O 2p emerges at a level slightly higher than the Fermi level, which is similar to an impurity level. In the case of Pt-HLaNb2O7, the new O 2p–Pt 5d hybridized band emerges at the top of the valence band. In this model, the conduction band, which consists of the H 1s and Nb 4d orbitals, spreads toward low energy. From this model, it can be concluded that the existence of Pt2+ should accelerate the excitation in Pt-HLaNb2O7.
Consequently, these partial DOS curves imply that the shrinking of the band gaps for Pt-HLaNb2O7 results from the spreading of the conduction band and Pt 5d new bands.
59
Figure 3-9 DOS curves simulated by first principles DFT calculation.
Rb4p La5d La4f Nb4p Nb4d H1s O2p Pt6s Pt5d total
D O S ( a .u .)
Energy / eV
Fermi level
Pt-HLaNb
2O
7HLaNb
2O
7RbLaNb
2O
7-4 -2 0 2 4
Rb La
NbO6
Pt vacancy
H
60
Figure 3-10 Photocatalytic activities of (a) RbLaNb2O7, (b) Pt-RbLaNb2O7, (c) HLaNb2O7, (d) Pt-HLaNb2O7, and (e) Pt-HLaNb2O7 without irradiation.
Photocatalytic Activity for Phenol Degradation. Fig. 3-10 shows the relationship of the normalized concentration of phenol by RbLaNb2O7, Pt-RbLaNb2O7, HLaNb2O7, Pt-HLaNb2O7, and Pt-HLaNb2O7 (without irradiation) against the elapsed time. The effect of the surface area may be ignorable because of small and similar surface area as mentioned above. From these plots, we can see that mother compound does not exhibit significant photocatalytic activity in this experiment. The normalized phenol concentrations with Pt-RbLaNb2O7 and HLaNb2O7, which have higher photocatalytic
a b c d e
N o rm a li ze d c o n c e n tr a ti o n o f p h e n o l
Irradiation time / min 0
dark
0.0 60 120 180 240 300
0.2
0.4
0.6
0.8
1.0
61
activities than mother compound RbLaNb2O7, are approximately 0.6 after 4 h irradiation by Xe light. On the other hand, phenol is completely degraded in 3 h by Pt-HLaNb2O7 under irradiation. Without irradiation (e), Pt-HLaNb2O7 did not cause a significant change in the phenol concentration. This result apparently indicates that the adsorption amounts on the surface of the samples are very small and are negligible.
The photocatalytic activity of Pt-HLaNb2O7 is superior to those of RbLaNb2O7, Pt-RbLaNb2O7, and HLaNb2O7, possibly because in Pt-RbLaNb2O7, Pt metal particles on RbLaNb2O7 serve as a promoter. Those Pt metal particles can serve as a trapping center for capturing photogenerated electrons for the effective separation of the photogenerated hole–electron pairs. In the case of HLaNb2O7, a new H 1s and O 2p level may accelerate the formation of holes, which can degrade phenol, as shown in DOS curves simulated by first principles DFT calculation (Fig. 3-9). On the other hand, in the case of Pt-HLaNb2O7 the conduction band spreads, resulting in the shrinkage of the band gap. Based on the above-mentioned result, Pt-HLaNb2O7
exhibits excellent photocatalytic ability, superior to those of RbLaNb2O7, Pt-RbLaNb2O7, and HLaNb2O7.
62