**Effective Photocatalytic Activity of Mixed Ni/Fe-Base Metal-Organic Framework under a Compact Fluorescent Daylight Lamp**

**Vinh Huu Nguyen 1, Trinh Duy Nguyen 1,\*, Long Giang Bach 1, Thai Hoang 2, Quynh Thi Phuong Bui 3, Lam Dai Tran 2,4, Chuong V. Nguyen 5, Dai-Viet N. Vo <sup>6</sup> and Sy Trung Do 7,\***


Received: 19 September 2018; Accepted: 15 October 2018; Published: 23 October 2018

**Abstract:** Mixed Ni/Fe-base metal-organic framework (Ni/Fe-MOF) with different molar ratios of Ni2+/Fe3+ have been successfully produced using an appropriate solvothermal router. Physicochemical properties of all samples were characterized using X-ray diffraction (XRD), Raman, field emission scanning electron microscopes (FE-SEM), fourier-transform infrared spectroscopy (FT-IR), N2 adsorption-desorption analysis, X-ray photoelectron spectroscopy (XPS), ultraviolet-visible diffuse reflectance spectra (UV-Vis DRS), and photoluminescence spectra (PL). The photocatalytic degradation performances of the photocatalysts were evaluated in the decomposition of rhodamine B (RhB) under a compact fluorescent daylight lamp. From XRD, IR, XPS, and Raman results, with the presence of mixed ion Fe3+ and Ni2+, MIL-88B (MIL standing for Materials of Institut Lavoisier) crystals based on the mixed metal Fe2NiO cluster were formed, while MIL-53(Fe) was formed with the presence of single ion Fe3+. From UV-Vis DRS results, Ni/Fe-MOF samples exhibited the absorption spectrum up to the visible region, and then they showed the high photocatalytic activity under visible light irradiation. A Ni/Fe-MOF sample with a Ni2+/Fe3+ molar ratio of 0.3 showed the highest photocatalytic degradation capacity of RhB, superior to that of the MIL-53(Fe) sample. The obtained result could be explained as a consequence of the large surface area with large pore volumes and pore size by the Ni2+ incorporating into the MOF's structure. In addition, a mixed metal Fe/Ni-based framework consisted of mixed-metal cluster Fe2NiO with an electron transfer effect and may enhance the photocatalytic performance.

**Keywords:** photocatalytic decomposition of rhodamine B; MIL-53(Fe); Ni/Fe-MOF; visible light irradiation

#### **1. Introduction**

Metal-organic frameworks (MOFs), a new class of high surface area and crystalline porous materials, assemble with metal clusters and organic bridging ligands [1]. These materials have received considerable attention in recent years due to their high resistance, high surface area, large pore volume, low density, and easily tunable framework. Among the MOFs, MIL-53(Fe) 88B (MIL standing for Materials of Institut Lavoisier) have attracted extensive interest for applications in gas storage [2,3], adsorption and separation of heavy metal [4], sensors [5], and in the biomedical field such as for drug delivery [6].

Recently, to eliminate organic dyes, many approaches have been suggested including adsorption [7–10] and photodegradation [11–13]. However, the latter is of interest because this process could decompose organic dyes to CO2, H2O, and harmless inorganics, while the adsorption process is only capable of removing dyes from water media. MIL-53(Fe) as a catalyst carrier or modification of MIL-53(Fe) as a catalyst for chemical reactions has received research attention [14]. MIL-53(Fe) has the chemical formula of FeIII(OH)(O2C–C6H4–CO2)·H2O, which consists of FeO6 octahedral chains connected to benzene dicarboxylate (BDC) anions, forming a three-dimensional network with a large volume and high surface area [2,14,15]. The FeO6 octahedral chains have the potential to act as a Lewis acid in many organic reactions [16]. Recently, MIL-53(Fe) with the potential use of FeO6 octahedral chains has received much attention in photocatalytic degradation of many organic dyes, such as methylene blue [11,13,17], rhodamine B (RhB) [14,16,17], and p-nitrophenol [14], and has given good decomposition results. Therefore, this is a possible application direction of MIL-53(Fe) in the removal of organic dyes.

Fe-based MOFs materials have been reported as an effective photocatalyst for decomposition of organic dyes under visible light irradiation [18–22]. However, their photocatalytic performance is not as expected because of the fast recombination of photogenerated holes (h+) and electrons (e−), resulting in the lack of h+ for degradation dyes [13]. To address this, various approaches have been proposed to depress the recombination process. For example, inorganic oxidants (e.g., H2O2, KBrO3, and (NH4)2S2O8), which act as electron acceptors, was introduced in the photocatalytic processes, significantly enhancing the photocatalytic effect of these materials. According to research by Yuan et al. [13], H2O2 is an efficient electron acceptor in the photocatalytic decomposition process of organic pigments by MIL-53(Fe) under visible light irradiation. Another approach that has been developed to enhance the photocatalytic performance of MiL-53(Fe) is the designed synthesis of composite photocatalysts containing MOFs materials such as CdS/MIL-53(Fe) [23], Ni-MOFs@GO [24], Fe3O4/MIL-53(Fe) [14], and Fe2O3/MIL-53(Fe) [25]. In addition, MIL-53(Fe) that has been doped or combined with one or more metals have also attracted much attention in recent years [26–29]. For this study, Qiao Sun et al. modified the MIL-53(Fe) by adding Mn, Co, and Ni metal into the framework of MIL-53(Fe) material, which exhibited excellent catalytic performance in liquid-phase degradation of phenol [30]. Various rare-earth or transition metals that modify MOFs structures have recently been reported such as three-dimensional Ln(III)–Zn(II) heterometallic coordination polymers [31], Fe substituted Cr MIL-101 [32], Ag-doped MOF-like organotitanium polymer (Ag@NH2-MOP(Ti)) [33], Ti-doped UiO-66 [34], Eu substituted Fe MIL-53 [35], and Zn-Ln coordination polymers (Ln = Nd, Pr, Sm, Eu, Tb, Dy) [36].

In this work, we report the synthesis of Ni/Fe-MOF with different Ni2+/Fe3+ molar ratios using the solvothermal route and their application for the degradation of RhB solution under visible light irradiation using a 40 W compact fluorescent lamp. To illustrate our method for the synthesis of Ni/Fe-MOF, we have selected the preparation of the MIL-53(Fe) structure, which consists of FeO6 octahedral chains connected to BDC anions. Thanks to the presence of Ni2+ ions in the reaction solution, MIL-88B crystals were formed with neutral mixed-metal clusters (Fe2NiO) connected via BDC anions. This structure is similar to the MIL-88B structure consisting of the trinuclear oxo-centered iron cluster (Fe3O) [27,28]. However, our bimetallic metal MOF products were expected to exhibit an excellent adsorption capacity and photocatalytic activity in comparison to the original single metal MOFs. The advantage of selecting MOF material containing Fe and Ni is due to the low cost, non-toxicity, and natural abundance of these two transition metal oxides. In addition, the MOF material is also capable of improving the separation efficiency of electron–hole pairs when Ni is incorporated into the structure of materials [37,38]. The structure, morphology, and optical properties of the obtained photocatalysts have been characterized using X-ray diffraction (XRD), Raman, field emission scanning electron microscopes and energy-dispersive X-ray spectrometer (FE-SEM/EDS), fourier-transform infrared spectroscopy (FT-IR), N2 adsorption-desorption analysis, X-ray photoelectron spectroscopy (XPS), ultraviolet-visible diffuse reflectance spectra (UV-Vis DRS), photoluminescence (PL) spectra and nitrogen physisorption measurements (BET). Besides, to obtain the optimal reaction conditions for the RhB photodecomposition, the effect of the initial RhB concentration and pH on the degradation of RhB was also investigated in detail.

#### **2. Results and Discussion**

#### *2.1. Physical Properties of MIL-53(Fe) and Ni-Doped MIL-53(Fe)*

#### 2.1.1. XRD Analysis

Figure 1 presents the XRD diffraction patterns of the MIL-53(Fe) and Ni/Fe-MOF samples isolated from dimethylformamide (DMF) and H2O. In patterns of MIL-53(Fe) samples (Figure 1A, curve a), the main diffraction peaks that appeared at 2θ of 9.1◦, 9.4◦, 14.1◦, 16.5◦, and 18.8◦ are similar to those previously reported for MIL-53(Fe) isolated from DMF [2,11,39]. In patterns of Ni/Fe-MOF samples (Figure 1A, curves b–e), the main diffraction peaks that appeared around 2θ of 7.3◦, 8.9◦, 9.3◦, 9.9◦, 16.8◦, 18.7◦, 17.7◦, 20.1◦, and 21.9◦ are similar to those previously reported for MIL-88B isolated from DMF. Notably, the diffraction peak at a 2θ of 7.3◦ observed in the XRD patterns of Ni/Fe-MOF samples increased in intensity as the molar ratio of Ni2+/Fe3+ increased from 0.1 to 0.7. With the presence of Ni2+ in the reaction solution, MIL-88B crystals were made up and the crystallinity of the material increased. This observation might be attributed to the fact that the structure formation of Ni/Fe-MOF was significantly influenced by the presence of Ni2+ in the reaction solution. In addition, no other diffraction peak associated with nickel oxides, iron oxides, or other impurities could be detected, demonstrating the high purity of the samples.

XRD patterns of the MIL-53(Fe) and Ni/Fe-MOF samples isolated from H2O (Figure 1B) showed the rugged background and weak intensities; however, the main diffraction peaks still maintained the same structure as in Reference [4]. The difference in XRD patterns of samples isolated from DMF and water may attribute to the breathing behavior of MIL-53(Fe) and MIL-88B, which has been well documented by Alhanami et al. [15]. Moreover, MIL-53(Fe)·H2O sample essentially shows a noncrystalline phase similar to those for MIL-53(Fe)·DMF. They can be explained by the effect of the synthesis temperature on the structure formation of MIL-53(Fe). Pu et al. demonstrated that iron ion and H2BDC could not coordinate successfully under a low temperature (100 ◦C), and therefore the MIL-53(Fe) crystal structure could not fully develop [40]. However, the Ni/Fe-MOF samples still show a high crystalline phase under low synthesis temperatures. Again, these results indicate that the presence of a mixed metal ion (Ni2+ and Fe3+ ion) did have a significant influence on the formation of Ni/Fe-MOF crystal structure, in which a Ni2+ and Fe3+ ion can coordinate with H2BDC to form MIL-88B crystals instead of MIL-53(Fe) crystals.

**Figure 1.** XRD patterns of as-prepared MIL-53(Fe) and Ni-MIL-53(Fe) crystals isolated from DMF (**A**,**B**) and H2O (**C**,**D**): MIL-53(Fe) (a), Ni/Fe-MOF-0.1 (b), Ni/Fe-MOF-0.3 (c), Ni/Fe-MOF-0.5 (d), and Ni/Fe-MOF-0.7 (e).

2.1.2. FT-IR Spectra

FTIR spectroscopic studies were performed for all samples in the wave range of 400–4000 cm−1, as shown in Figure 2. As shown in Figure 2A,C, strong vibrational bands around 1657, 1601, 1391, 1017, and 749 cm−1, which are attributed to υ(C=O), υas(OCO), υs(OCO), υ(C–O), and δ(C–H) vibrations confirms the presence the bridge coordination mode of metal carboxylates in the MOF structures [4,25,30]. No band at 1700 cm−<sup>1</sup> was found, implying no free H2BDC [27]. The band characteristics of DMF (1657 cm<sup>−</sup>1) and H2O (3387 cm−1) were present in the samples MIL-53(Fe)·DMF, Ni/Fe-MOF·DMF, MIL-53(Fe)·H2O and Ni/Fe-MOF-x·H2O, respectively [27].

At lower frequencies (Figure 2B), vibrational bands around 750 cm−1, 690 cm−1, and 660 cm−<sup>1</sup> represent the C–H vibration, C=C stretch, OH bend, and OCO bend, respectively, were found, implying the presence of the vibrations of the organic ligand BDC [27]. Figure 2B also shows that the strong band at 547 cm−<sup>1</sup> in all samples could be attributed to Fe–O vibrations or Ni–O vibrations [41]. The band around 625 cm−<sup>1</sup> belongs to the Fe3O vibration, which was observed in MIL-53(Fe) and Ni-Ni/Fe-MOF-0.1 samples. The weak band around 720 cm−<sup>1</sup> is related to the Fe2NiO vibration, which was observed in Ni/Fe-MOF-x samples [27]. These results reaffirmed that Ni2+ and Fe3+ ions can coordinate with H2BDC to form MIL-88B crystals.

**Figure 2.** FT-IR spectra of as-prepared MIL-53(Fe) and Ni/Fe-MOF crystals isolated from DMF (**A**,**B**) and H2O (**C**): MIL-53(Fe) (a), Ni/Fe-MOF-0.1 (b), Ni/Fe-MOF-0.3 (c), Ni/Fe-MOF-0.5 (d), and Ni/Fe-MOF-0.7 (e).

#### 2.1.3. Raman Spectra

Samples were analyzed using Raman spectroscopy using an excitation wavelength at 633 nm and spectra recorded at a wavenumber range of 100–900 cm<sup>−</sup>1, as shown in Figure 3. According to previous studies, the BDC bridge in MOFs has Raman-active modes: the symmetric vibration modes (vs. (COO)) and asymmetric vibration (vas (COO)) of the carboxylate group (1445 cm−<sup>1</sup> and 1501 cm−1), the vibration of the C–C bond between the benzene ring and the carboxylate group (1140 cm−1), and the external plane deformation of the C–H link (865 cm−<sup>1</sup> and 630 cm−1) [28]. As seen in Figure 3, the presence of a BDC linker was also observed in all samples, and no Raman signals corresponding to nickel oxides, iron oxides, or other impurities were found on any of the samples, which is consistent with the results of the XRD patterns. Notably, the Raman signal corresponding to the symmetric vibration (vs. (OCO)) of the carboxylate group showed a shift to a lower wavenumber and the peak split into two peaks corresponding to an increase of the Ni2+/Fe3+ molar ratio. This result was due to the change in the charge distribution in the organic bridge when they were coordinated with different metal ions (Figure 3B). Ionic Ni2+ has a smaller nuclear charge and a larger ionic radius than Fe3+ (*rNi*2<sup>+</sup> = 0.69 Å and *rFe*3<sup>+</sup> = 0.55 Å) [42]. Therefore, Ni2+ creates a weaker coordinated link with the OCO group on the organic bridge than Fe3+, thus the symmetric vibration (vs. (OCO)) of the carboxylate group when forming coordinated bonds with Ni2+ moves to a lower wavenumber than Fe3+ [43]. This result is commensurate with the XRD and IR results for Ni/Fe MOF.

**Figure 3.** Raman spectra of as-prepared MIL-53(Fe) and Ni/Fe-MOF crystals isolated from DMF (**A**): MIL-53(Fe) (a), Ni/Fe-MOF-0.1 (b), Ni/Fe-MOF-0.3 (c), Ni/Fe-MOF-0.5 (d), and Ni/Fe-MOF-0.7 (e), and enlarged Raman spectra around 1450 cm−<sup>1</sup> (**B**).

#### 2.1.4. FE-SEM/EDS Analysis

Figure 4 displays SEM images and EDS spectra of the as-prepared MOF samples. As shown in Figure 4, the morphologies and shapes of MOF samples varied according to the molar ratio of Ni2+/Fe3+. MIL-53(Fe) sample mostly had amorphous nanoparticles (Figure 4(a1,a2)), which is in good agreement with the results of XRD patterns with a poor crystallinity. When the molar ratio of Ni2+/Fe3+ was set to 0.1, the crystals of Ni/Fe-MOF-0.1 were not homogeneous with different shapes and sizes (Figure 4(b1,b2)). A mixture of octahedral and hexagonal bipyramidal shapes, and nanoparticles, were perceived when the molar ratio of Ni2+/Fe3+ (0.3–0.7) was increased further. However, these octahedral and hexagonal bipyramidal shapes collapsed with cracks on the crystal surface. These results, along with the XRD, IR, and Raman results above, indicate that a mixed-metal Ni/Fe-MOF was successfully synthesized using the solvothermal method.

Moreover, to confirm the molar ratio of Ni2+/Fe3+ in the Ni/Fe-MOF samples in comparison to the theoretical value, EDS was also conducted. The result from the EDS spectrum of the obtained MIL-53(Fe) sample (Figure 4(a3)) showed the coexistence of C, O, Fe, and Cl. The presence of Cl may have been due to the FeCl3 precursor, further confirming that the MIL-53(Fe) crystal structure could not fully develop at a low temperature (100 ◦C). The EDS spectra of the Ni/Fe-MOF samples (Figure 4(b3,c3,d3,e3)) revealed that these samples contained C, O, Fe, and Ni. However, the existence of Cl was still observed in the Ni/Fe-MOF-0.1 sample. The molar ratio of Ni2+/Fe3+ of Ni/Fe-MOF-0.1, Ni/Fe-MOF-0.3, Ni/Fe-MOF-0.5, and Ni/Fe-MOF-0.7, obtained using EDS analysis, was 0.16, 0.30, 0.48, and 0.66, respectively. In addition, the map of Fe, O, C, and Ni is shown in Figure S1, which indicates that they were uniformly distributed over the MOF surface.

**Figure 4.** SEM images (1, 2) and EDS patterns (3) of as-prepared MIL-53(Fe) and Ni/Fe-MOF crystals isolated from DMF: MIL-53(Fe) (**a**), Ni/Fe-MOF-0.1 (**b**), Ni/Fe-MOF-0.3 (**c**), Ni/Fe-MOF-0.5 (**d**), and Ni/Fe-MOF-0.7 (**e**).

#### 2.1.5. XPS Spectra

To analyze the chemical states of Ni and Fe in the Ni/Fe MOF structure, XPS spectroscopy was carried out. As illustrated in Figure 5A, the wide-scan XPS spectra of MIL-53(Fe)·H2O possesses the characteristic peaks of C, O, Fe, and Cl, while Ni/Fe-MOF-0.3.H2O contained C, O, Fe, and Ni. Based on the XPS analysis, the Ni/Fe-MOF-0.3 had a surface molar ratio of Ni2+/Fe3+ of 0.26, which approximates the EDS results above. Besides, N was not detected in either sample, indicating that the DMF solvent was sufficiently eliminated from the MOFs.

Figure 5B shows the C 1s XPS spectra of MIL-53(Fe)·H2O and Ni/Fe-MOF-0.3·H2O samples. Both spectra were fitted into three peaks at a binding energy (BE) of 285.01, 288.9, and 291.7 eV, which could be assigned to the carbon components on the phenyl and the carboxylate groups of the BDC linkers [30,40,44–46]. The O 1s XPS spectra (Figure 5C) could also be fitted into three peaks, which are (i) the peak at 533.8 eV corresponding to the O components on C=O/H2O, (ii) the peak at 532.3 eV attributed to the O components on the BDC linkers, and (iii) the peak at 530. 5 eV was assigned to the O components on the Fe–O bonds (for MIL-53(Fe) sample) or Fe2NiO clusters (for Ni/Fe-MOF-0.3 sample). These results further confirmed the coordination between the metal ion (Ni2+ and/or Fe3+) and BDC linkers, which is commensurate with the XRD, IR, and Raman results above.

**Figure 5.** Full scan (**A**), C1s (**B**), O1s (**C**), Fe2p (**D**), and Ni2p (**E**) XPS spectra of MIL-53(Fe) and Ni/Fe-MOF-0.3.

The Fe 2p high-resolution XPS spectrum of MIL-53(Fe) sample (Figure 5D) displays two main peaks that were indexed to Fe 2p1/2 (712.4 eV) and Fe2p3/2 (726.1 eV). The splitting energy of the 2p doublet was 13.7 eV, implying that the valence state of Fe was +3 [4,23,44]. Similarly, the valence state of Fe in the Ni/Fe MOF structure was also +3 because the splitting energy between Fe 2p1/2 (712.9 eV) and Fe 2p3/2 (726.2 eV) was 13.3 eV. To further confirm the valence state of Fe in both of these samples, the Fe 2p3/2 peak was fitted into six peaks including Gupta and Sen (GS) multiples, surface structures, and shake-up-related satellites [28,47,48]. The fitting results, as shown in Figures S6 and S7, were indexed well with Fe3+ GS multiplets, which indicated that the valence state of Fe in the MIL-53(Fe) and Ni/Fe MOF structure was +3. In the high-resolution XPS spectrum of Ni 2p (Figure 4e), we observed the BE of the Ni 2p3/2 (857.2 eV) and Ni 2p1/2 (874.8 eV) core-level peaks with the doublet separation of 17.6 eV, implying that the valence state of Ni was +2 [49,50].

#### 2.1.6. N2 Adsorption/Desorption

The specific surface area and porous structure of MIL-53(Fe) and Ni/Fe-MOF crystals isolated from DMF and H2O were determined using N2 adsorption–desorption isotherms at 77 K. The N2 adsorption–desorption isotherms, as shown in Figure 6A, displayed an intermediate mode between type I and type IV, which was associated with mesoporous and microporous materials, respectively [51]. The Brunauer–Emmett–Teller (BET) surface area, pore volume, and pore width of MIL-53(Fe) and Ni/Fe-MOF-0.3 samples are shown in Table 1. The MIL-53(Fe)·H2O, MIL-53(Fe)·DMF, Ni/Fe-MOF-0.3.H2O, and Ni/Fe-MOF-0.3.DMF had specific surface areas of 158, 300, 247, and 480 m2/g, respectively (Table 1). The mesopore size distribution curve of samples calculated using the Barrett–Joyner–Halenda (BJH) model is shown in Figure 6B. The MIL-53(Fe)·H2O and MIL-53(Fe)·DMF sample was non-porous, whereas Ni/Fe-MOF-0.3·H2O, and Ni/Fe-MOF-0.3·DMF showed a pore size centered at about 3.8 nm and 21.4 nm, respectively. Therefore, compared with MIL-53(Fe), Ni/Fe-MOF-0.3 showed a higher value in the specific surface areas. In addition, the higher surface area and micropore volume for samples isolated from DMF, as compared with samples isolated from H2O, was due to the reversible breathing behavior of these materials, which was dependent on the molecule present inside their pores, where the pores were opened in the presence of DMF and closed in the presence of H2O [27,28,52]. The formation of porous material for Ni/Fe-MOF-0.3 could be explained by the formation of Fe2NiO cluster in the Ni/Fe-MOF structure, which could affect the reversible breathing behavior of these materials. MIL-88B(Fe) crystals with trinuclear metal clusters were known as non-porous materials due to the need for compensating the anion inside their porous system [28,53]. Do and coworkers demonstrated that MOF structure with the presence of Fe2NiO cluster as nodes in the MIL-88B framework avoids the compensating anion [27,28], which results in the formation of porous material for Ni/Fe-MOF-0.3. In addition, the cracks on the crystal surface of Ni/Fe-MOF-0.3 (Figure 4) could also partly create the characteristics of microporous or mesoporous materials for this sample.


**Table 1.** Specific surface area and porosity of MiL-53(Fe) and Ni/Fe-MOF samples.

**Figure 6.** N2 adsorption–desorption isotherms (**A**) and pore size distributions (**B**) of the synthesized samples: MIL-53(Fe)·DMF (a), MIL-53(Fe)·H2O (b), Ni/Fe-MOF-0.3·DMF (c), and Ni/Fe-MOF-0.3·H2O (d).

#### 2.1.7. UV-Vis Spectra

The light absorption properties of the material were studied through the UV-Vis-DRS spectra. The UV-Vis-DRS spectrum of the material is shown in Figure 7. For washing samples with DMF (Figure 7A), MIL-53(Fe)·DMF gave strong absorption bands in the wavelength range of 200 to 400 nm. The strong absorption bands at 256 to 310 nm could be due to the transfer of the charge from the oxygen center of the organic bridge to the metal center in the octahedral FeO6 structure [17,54]. The band at 350 to 500 nm was due to the shift of d–d (6A1g → 4A1g + 4Eg (G)) of Fe3+ in the MIL-53(Fe) structure [14,27]. The main absorption edge (λ, nm) of the MIL-53(Fe)·DMF was 478 nm, corresponding to the bandgap energy Eg = 2.59 eV (Eg = 1240/λ). This result is in accordance with previous reports [44,55]. When the MIL-53(Fe) was modified with Ni, the material have the decreased absorption in the wavelength range from 200 to 500 nm, and the absorption spectrum extended in the range from 250 to 800 nm, so it was difficult to determine the absorption of the material accurately. When the material was washed with water (Figure 7B), the modified material had an increased absorption in the wavelength range from 200 to 400 nm, and the absorption intensity was higher and broader in the visible light region as compared to the modified sample washed with DMF. As the material was washed with water, there was a structural change between the large pore and the narrow pore caused by the "breathing" effect when the material absorbed the water molecules inside the pore. This phase transformation of the structure led to a change in the electronic structure [56], and subsequently, a change in the absorption spectrum of the material and decreasing Eg. For Ni/Fe-MOF-0.1·H2O, Ni/Fe-MOF-0.3·H2O, and Ni/Fe-MOF-0.5·H2O samples, the absorption intensity in the visible light region and the absorption band of the material shifted to a wavelength longer than for MIL-53(Fe)·H2O. As absorption in the visible light increased, the visible

light energy could be used more efficiently, thus contributing to the increased photocatalytic efficiency of the material. The absorption edges of MIL-53(Fe)·H2O, Ni/Fe-MOF-0.1·H2O, Ni/Fe-MOF-0.3·H2O, Ni/Fe-MOF-0.5·H2O, and Ni/Fe-MOF-0.7·H2O were 504, 553, 532, 513, and 516 nm (Figure S2), corresponding to the optical bandgap of 2.46, 2.24, 2.33, 2.42, and 2.40 eV, respectively. These results provided a potential photoreactivity of MIL-53(Fe) and Ni/Fe-MOF samples in the visible light range.

**Figure 7.** UV-Vis DRS spectra of as-prepared MIL-53(Fe) and Ni-MIL-53(Fe) crystals isolated from DMF (**A**) and H2O (**B**): MIL-53(Fe) (a), Ni/Fe-MOF-0.1 (b), Ni/Fe-MOF-0.3 (c), Ni/Fe-MOF-0.5 (d), and Ni/Fe-MOF-0.7 (e).

#### 2.1.8. PL Spectroscopy

PL spectra of MIL-53(Fe) and Ni/Fe-MOF samples were recorded at room temperature and are shown in Figure 8. When the MIL-53(Fe) sample was excited by a 320 nm laser, its emission spectrum showed a strong emission range of 350 to 500 nm and a weak emission range of 570 to 750 nm. In comparison, the intensity of Ni/Fe-MOF samples was significantly lower than that of the MIL-53(Fe) sample because of the presence of the Ni2FeO cluster in the structure of the Ni/Fe-MOF crystal. These results demonstrated that electron–hole recombination could be inhibited in the Ni/Fe-MOF, resulting in the improvement of photocatalytic performance. PL spectra, along with the UV-Vis DRS result, could satisfy the prerequisite for visible-light photocatalysis.

**Figure 8.** PL spectra of as-prepared MIL-53(Fe) (a), Ni/Fe-MOF-0.1 (b), Ni/Fe-MOF-0.3 (c), Ni/Fe-MOF-0.5 (d), and Ni/Fe-MOF-0.7 (e).

#### *2.2. Photocatalytic Activities*

#### 2.2.1. RhB Removal by MIL-53(Fe) and Ni-MIL-53(Fe)

The photocatalytic activities of MIL-53(Fe) and Ni/Fe-MOF-x photocatalysts were evaluated in the liquid-phase photodegradation of RhB dye under visible light irradiation. Figure 9 displays the changes of RhB concentrations via adsorption and photocatalytic degradation under different experimental conditions. As shown in Figure 9, a negligible degradation of RhB concentrations was observed in the several blank runs including RhB/H2O2/Dark, RhB/H2O2/Light, and RhB/Dark systems, proving the stability property of RhB under visible light irradiation of compact fluorescent light. Also, as shown in Figure 9A, after 180 min adsorption (in the dark), 16% and 51% RhB were removed in the presence of MIL-53(Fe) (MIL-53(Fe)/Dark system) and Ni/Fe-MOF-0.3 (Ni/Fe-MOF-0.3/Dark system), respectively. The higher adsorption capacity of the Ni/Fe-MOF-0.3 sample was due to its higher surface area (247 m2/g for Ni/Fe-MOF-0.3 and 158 m2/g for MIL-53(Fe)). In addition, there was no significant difference in the removal of RhB concentration in the two adsorption experiments with the presence of H2O2 (MIL-53(Fe)/H2O2/Dark and Ni/Fe-MOF-0.3/H2O2/Dark systems and the absence of H2O2 (MIL-53(Fe)/Dark and Ni/Fe-MOF-0.3/Dark systems. Therefore, our photocatalytic experiments do display the presence of a Fenton reaction.

**Figure 9.** Adsorption (**A**) and photodegradation (**B**) of RhB under different conditions over MIL-53(Fe) and Ni/Fe-MOF-0.3, and UV-Vis spectral of RhB solution separated from the Ni/Fe-MOF-0.3/ Light/H2O2 catalytic system (**C**) and MIL-53(Fe)/Light /H2O2 catalytic system (**D**).

Under visible light irradiation, the presence of MIL-53(Fe) could enhance the degradation efficiency of RhB up to 81.46% using a photolysis process in MIL-53(Fe)/Light/H2O2 catalytic system (Figure 9B). For the Ni/Fe-MOF-0.3/Light/H2O2 catalytic system, the degradation efficiency of RhB was remarkably enhanced where about 91.14% RhB removal was achieved (Figure 9B). The higher photocatalytic activity of the Ni/Fe-MOF-0.3 sample as compared with MIL-53(Fe) could also be indicated by the change of the UV-Vis absorption spectra of the solution in the course of the RhB degradation (Figure 9C,D). As seen in Figure 9C,D, the primary absorption band, which could be attributed to RhB, shifted from 554 to 500 nm in a step-wise manner. This change could be reasonably assigned to the removal of ethyl groups one by one in this reaction, which is in good agreement with the previous literature. The photodegradation of RhB over MIL-53(Fe) and Ni/Fe-MOF-0.3 photocatalysts approximately followed a pseudo-first-order kinetics model: ln(Co/C) = kobst [57–59]. The presence of Ni/Fe-MOF-0.3 promoted the photodegradation rate; the rate constants were 8.88 × <sup>10</sup>−<sup>3</sup> min−<sup>1</sup> for MIL-53(Fe) and 11.15 × <sup>10</sup>−<sup>3</sup> min−<sup>1</sup> for Ni/Fe-MOF-0.3.

To investigate the role of H2O2 on the photocatalytic performance of MIL-53(Fe) and Ni/Fe-MOF photocatalysts, the photocatalytic processes with the presence and absence of H2O2 were conducted in parallel (Figure 9B). After 180 min of irradiation, the degradation rate of RhB over MIL-53(Fe)/Light/H2O2 and MIL-53(Fe)/Light process was 81.46% and 27.60%, respectively. Only MIL-53(Fe) with the absence of H2O2 exhibits the low efficiency of RhB photodegradation due to the fast electron-hole recombination, which is in good agreement with the previous literature [13,17]. For the MIL-53(Fe)/H2O2/Light process, H2O2 acted as an electron accepter, resulting in the suppression of charge recombination; therefore, the rate for RhB decomposition could be significantly enhanced, as was demonstrated by Du et al. [13]. Similarly, Ai et al. also showed that the enhancement of MI-53(Fe) photocatalytic performance could be due to the synergistic effects of the combination of MIL-53(Fe) and H2O2 under visible light irradiation [17]. Interestingly, the effect of H2O2 on the photocatalytic performance of the Ni/Fe-MOF photocatalyst showed a considerable difference. The Ni/Fe-MOF sample could degrade more than 90% of the initial RhB content regardless of the presence or absence of H2O2.

The superior catalytic performance of the Ni/Fe-MOF sample could be explained by the formation of the mixed metal cluster Fe2NiO in the Ni/Fe-MOF framework. According to recent reports, the Fe-based framework (MIL-101, MIL-100, MIL-88, and MOF-235), containing single metal cluster Fe3-μ3-oxo clusters with small particle sizes, are proposed as a visible light photocatalyst [44,60–63]. The reaction mechanism of these materials have been reported based on semiconductor theory and previous reports [61–64]. Particularly, when the surface of MOFs material absorbs photons (Ephotons ≥ Eg), the electrons (e−) in the valence band (VB) will be excited to the conduction band (CB), leaving the holes (h+) in the VB. These photogenerated e−–h+ pairs may be further involved in the following three processes: (i) successfully migration to the surface of MOFs, (ii) being captured by the defect sites in bulk and/or on the surface region of semiconductor, and (iii) recombining and releasing the energy in the form of heat or a photon. Then, the h+ can accept electrons and induce water molecules to generate hydroxyl radicals (•OH), which exhibit a high oxidation ability to decompose the organic dyes. However, there is a recombination of excessive electrons and holes, resulting in the restricted photocatalytic activity of this material. In our study, mixed a metal Fe/Ni-based framework that consists of a mixed-metal cluster Fe2NiO with electron transfer effect may enhance the photocatalytic performance [45,61,65].

Besides, a mixed metal Fe/Ni-based framework that consists of the mixed-metal cluster Fe2NiO possesses large pores and a high surface area, as compared with a single metal Fe-based framework; therefore, Ni/Fe-MOF exhibited a high adsorption capacity of RhB and high photocatalytic activity in RhB degradation. XRD patterns of Ni/Fe-MOF-0.3 before and after reactions were shown in Figure S3 (SI file). As shown in Figure S3, there was no apparent difference in the crystal structure. This result indicated that the crystal structure of the material did not change after the photocatalytic reaction.

2.2.2. Effect of Initial Dye Concentration, Initial Solution pH, and the Molar Ratio of Ni2+/Fe3+ on the Degradation of RhB

The effect of initial dye concentration on the degradation of RhB over the Ni/Fe-MOF-0.3/Light/ H2O2 system was evaluated (Figure 10A). As shown in Figure 10A, the degradation efficiency of RhB was slightly decreased when increasing the initial dye concentration from 1 × <sup>10</sup>−<sup>5</sup> to 4 × <sup>10</sup>−<sup>5</sup> M. This was mainly because of the increase of the dye molecules around the active sites leading to inhibiting the penetration of light to the surface of the catalyst [66].

**Figure 10.** Effect of initial dye concentration (**A**), initial solution pH (**B**), and the molar ratio of Ni2+/Fe3+ (**C**) on the degradation of RhB.

The effect of the initial pH on the degradation of RhB on the degradation of RhB over Ni/Fe-MOF/Light/H2O2 system was also investigated. The pH of the initial solution was selected as follows: 3, 5 (acidic), 7 (neutral), and 9 (basic). At different pH conditions, the Ni/Fe-MOF-0.3 remained most effective when it came to removing RhB. The RhB removal efficiency peaked at the solution pH of 5 and decreased with increasing pH thereafter (Figure 10B). This result could be explained by the fact that when the pH exceeded the isoelectric point of the material, they were negatively charged. In addition, the RhB used in this experiment was a cationic color such that the material would absorb the color gradually from pH 5 to 9. As the adsorption increased, the color molecules would shield the catalytic surface, which prevented light from irradiating on the catalyst surface, thus decreasing photocatalytic activity and reducing color removal. The pH at the isoelectric point or point of zero charge-pzc of the material was an important parameter for evaluation of the acidity/basicity and the surface charge of the adsorbent in solution. The determination of pHzpc was carried out according to our previously published study [67–69], as follows: Photocatalysts (20 mg) was added to flasks containing 100 mL of KCl 0.1 M at different initial pH values (pHi = 2, 4, 6, 8, 10, and 12). The solutions were shaken in the shaker for 24 h, and then solids were removed from the mixture by centrifugation at 4000 rpm for 15 min. The final pH of the solution (pHf) is measured using a pH meter. The curve was plotted via pHf against the pHi, and the pHpzc was calculated at pHi = pHf. As shown in Figure 11A,B, the pHpzc values of the MIL-53(Fe) and Ni/Fe-MOF-0.3 were approximately equal and were within the pH range of 4.1–4.2.

**Figure 11.** Measurement of pHzpc: the initial versus final pH plot: pH initial (a), pH initial-MIL-53(Fe) (b), and pH initial-Ni/Fe-MOF-0.3 (c) (**A**) and enlarged pH initial from 3 to 5 (**B**).

The degradation results of the different molar ratios of Ni2+/Fe3+ in the samples are shown in Figure 10C, where the best performance was obtained with the Ni/Fe-MOF-0.3 sample, followed by the Ni/Fe-MOF-0.1 and Ni/Fe-MOF-0.7 samples. The Ni/Fe-MOF-0.5 sample showed the lowest catalytic activity among all the Ni/Fe-MOF catalysts. This result indicated that the different molar ratio of Ni2+/Fe3+ had a significant impact on the photocatalytic performance of Ni/Fe-MOF samples, which may be conducive to the structure and morphology formation of Ni/Fe-MOF.

#### **3. Experimental**

#### *3.1. Materials*

1,4-Benzenedioic acid (H2BDC, 98%) and RhB (≥95%) were purchased from Sigma-Aldrich (St. Louis, MO, USA). Iron(III) chloride hexahydrate (FeCl3·6H2O, 99%), nickel(II) nitrate hexahydrate (Ni(NO3)2·6H2O, 99%), *N*,*N*-dimethylformamide (DMF, 99%), ethanol, and hydrogen peroxide (H2O2, 30%) were obtained from Xilong Chemical Co., Ltd. (Guangzhou, China). All reagents were used as received without further purification.

#### *3.2. Preparation of Catalysts*

Ni/Fe-MOF samples were synthesized using a solvothermal router similar to MIL-53(Fe), according to the previous literature [39]. In a typical synthesis, 9 mmol of H2BDC, 6 mmol of FeCl3·6H2O, and a certain amount of Ni(NO3)2·6H2O were dissolved in 60 mL DMF. The obtained mixture was vigorously stirred for 30 min before being transferred into a 100 mL hydrothermal synthesis autoclave reactor 304 stainless steel high-pressure digestion tank with PTFE lining (Baoshishan Co., Ltd., Shanghai, China). The autoclave was heated at 100 ◦C in an oven (Memmert UN110, Schwabach, Germany) with a heating rate of 5 ◦C/min for three days. After being cooled to room temperature in air, the remaining H2BDC was removed using a distillation method with DMF solvent for 24 h at 100 ◦C with a heating rate of 5 ◦C/min. The obtained suspension was centrifuged at 6000 rpm for 30 min, and the orange precipitates located at the bottom of the tube were washed with DMF (three times) and water (three times), respectively. Finally, the product was dried for 24 h at 60 ◦C. The obtained MOFs samples with corresponding Ni concentration were denoted as Ni/Fe-MOF-x (x is the molar ratio of Ni2+/Fe3+, and was chosen as 0, 0.1, 0.3, 0.5, and 0.7). The specific description is shown in Table S1 and the flow chart of the synthesis method is described in Figure S4. The sample was washed with DMF and water to obtain Ni/Fe-MOF-x·DMF and Ni/Fe-MOF-x·H2O, respectively. For comparison, MIL(53) also was prepared using a similar method above without the presence of Ni(NO3)2·6H2O in the reaction solution mixture.

#### *3.3. Catalyst Characterization*

Powder X-ray diffraction (XRD) patterns were conducted on a D8 Advance Bruker powder diffractometer with a Cu Kα source (λ = 0.15405) at a scan rate of 0.04◦/s with 2θ = 2 to 30◦. The surface morphologies and particle size of Ni/Fe-MOF samples were observed using field emission scanning electron microscope (FESEM, JEOL JSM-7600F, Peabody, MA, USA) equipped with an energy dispersive X-ray spectroscope (EDS, Oxford instruments 50 mm<sup>2</sup> X-Max, Abingdon, UK). FT-IR spectra were recorded on an EQUINOX 55 spectrometer (Bruker, Germany) using the KBr pellet technique. Raman spectroscopy was carried out on the HORIBA Jobin Yvon spectrometer with a laser beam of 633 nm. To examine the existence of Ni and Fe in the samples, X-ray photoelectron spectra (XPS) of the samples was measured using MultiLab 2000 spectrometer (Thermo VG Scientific, Waltham, MA, USA). The optical absorption characteristics of the photocatalysts were determined using ultraviolet-visible (UV/Vis) diffuse reflectance spectroscopy (UV/Vis DRS, Shimazu UV-2450, Kyoto, Japan) in the range 200–900 cm−1. PL spectroscopy was performed using a Hitachi F4500 Fluorescence Spectrometer (Schaumburg, IL, USA) with the Xe Lamp Power range (700–900 V) at room temperature. The specific surface area and pore distribution of MIL-53(Fe) and Ni/Fe-MOFs

were determined using the Brunauer–Emmett–Teller (BET) method and Barrett–Joyner–Halenda (BJH) method, respectively (TriStar 3000 V6.07, Micromeritics instrument corporation, Norcross, GA, USA). The samples were kept at 200 ◦C for 5 h to degas. The pH value was measured using a pH meter (Consort-C1010, Turnhout, Belgium) at room temperature.

#### *3.4. Photocatalytic Test*

The photocatalytic activities of Ni/Fe-MOF photocatalysts were evaluated using the photodegradation of RhB under visible light irradiation with a 40 W compact fluorescent lamp (Philips) in the open air and at room temperature (Figure S5). The intensity and wavelength of the light source was 4400 lm and >400 nm, respectively (Figure S6 and Table S2). Therefore, it was suggested that the photocatalytic processes in our experiments were mainly due to the action of the visible light range [70–72]. In each run, a mixture of RhB aqueous solution (3.10−<sup>5</sup> mol/L, 100 mL), the given catalyst (20 mg), and H2O2 (10−<sup>5</sup> mol/L) was magnetically stirred in the presence or absence of light. Five milliliters of the suspension was withdrawn at the same intervals and immediately centrifuged to separate the photocatalyst particles for 15 min. The concentration of RhB was analyzed using a UV-visible spectrophotometer (Model Evolution 60S, Thermo Fisher Scientific, Waltham, MA, USA) at a maximum absorbance wavelength of λ = 554 nm. In addition, the effect of parameters including initial dye concentration and initial solution pH on the photodegradation of RhB over Ni/Fe-MOF photocatalysts was also investigated. pH levels of 3, 5, 7, and 9 were selected, whereas the concentrations of RhB were increased from 1.10−<sup>5</sup> M to 4.10−<sup>5</sup> M.

#### **4. Conclusions**

In summary, we have successfully prepared mixed Ni/Fe-base MOF with different molar ratios of Ni2+/Fe3+ via a direct solvothermal approach. The structure characterization results from XRD, Raman, XPS, and FT-IR confirmed that with the presence of mixed ionic Fe3+ and Ni2+, MIL-88B crystals based on the mixed metal Fe2NiO cluster was formed, while MIL-53 (Fe) was formed with the presence of a single ion Fe3+. The photocatalytic performance of the obtained photocatalysts was evaluated in the decolorization of RhB dye. The results indicated that the obtained Ni/Fe-MOF samples exhibited high photocatalytic activity in comparison to MIL-53(Fe). The degradation rate of Ni/Fe-MOF-0.3 could reach the highest (91.14%) after 180 min of visible light irradiation. These results suggest that the Ni/Fe-MOF, which consist mixed-metal cluster Fe2NiO with electron transfer effects, might enhance the photocatalytic performance.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/2073-4344/8/11/487/s1, Figure S1: EDS mapping of Ni/Fe-MOF-0.3 sample, Figure S2: UV-vis DRS spectra of as-prepared MIL-53(Fe) and Ni-MIL-53(Fe) crystals isolated from H2O, Figure S3: XRD patterns of Ni/Fe-MOF-0.3 before and after reactions, Table S1: Synthetic parameters of MIL-53(Fe) and Ni/Fe-MOF samples, Figure S4: The flow chart of the synthesis method, Figure S5: Illustration of the utilized photocatalytic test system, Figure S6. The spectral distribution of a 40 W compact fluorescent lamp, Figure S7: Background-subtracted Fe 2p3/2 spectrum from Ni/Fe-MOF-0.3, Figure S8: Background-subtracted Fe 2p3/2 spectrum from MIL-53(Fe), Table S2: Product data of a 40 W compact fluorescent lamp.

**Author Contributions:** T.D.N. proposed the concept and supervised the research work at Nguyen Tat Thanh University. V.H.N. and Q.T.P.B. designed the experiments and performed the experiments. T.H. and L.D.T. performed XPS and FT-IR analyses. C.V.N. performed SEM and EDS analyses. D.-V.N.V. contributed to the revision of the manucript. L.G.B. and S.T.D. analyzed the data and wrote the paper.

**Funding:** This research was funded by NTTU Foundation for Science and Technology Development under grant number 2017.01.13/HĐ-KHCN.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


© 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

*Article*

## **Photocatalytic Behavior of Strontium Aluminates Co-Doped with Europium and Dysprosium Synthesized by Hydrothermal Reaction in Degradation of Methylene Blue**

#### **Byung-Geon Park**

Department of Food and Nutrition, Kwangju Women's University, 165 Sanjung-dong, Gwangju 62396, Korea; bgpark@kwu.ac.kr or bgpark814@daum.net; Tel.: +82-62-950-0814

Received: 4 April 2018; Accepted: 22 May 2018; Published: 28 May 2018

**Abstract:** Strontium aluminates co-doped with europium and dysprosium were prepared by a hydrothermal reaction through a sintering process at lower temperatures. The physicochemical properties of the strontium aluminates co-doped with europium and dysprosium were characterized and compared with those of strontium aluminates prepared by a sol–gel method. The photocatalytic properties of the strontium aluminates co-doped with europium and dysprosium were evaluated through the photocatalytic decomposition of methylene blue dye. The strontium aluminates co-doped with europium and dysprosium prepared by the hydrothermal reaction exhibited good phosphorescence and photocatalytic activities that were similar to those prepared by the sol–gel method. The photocatalytic activity of these catalysts for methylene blue degradation was higher than that of the titanium dioxide (TiO2) photocatalyst.

**Keywords:** strontium aluminates; dye photodecomposition; hydrothermal reaction; sol–gel method; phosphorescence

#### **1. Introduction**

Alkaline earth aluminates have attracted considerable attention as long afterglow materials because of their excellent photoluminescence, radiation intensity, color purity, and good radiation resistance [1]. In particular, strontium aluminates (SAO) co-doped with europium and dysprosium (SAO; SrAl2O4: Eu2+, Dy3+) are used in many fields owing to their excellent phosphorescence [2]. SrAl2O4: Eu2+, Dy3+ is applied in emergency lighting, safe indications, signposts, graphic art, billboards, and interior design [3–5]. In addition, the material can be used to synthesize new metal compound composites [6] as well as cathoderay tubes and plasma display panels [7,8]. They also exhibit photocatalytic activity owing to their photosensitive properties [9].

The sol–gel process has attracted considerable interest in obtaining novel chemical compositions and relatively lower reaction temperatures, resulting in homogeneous products [9]. The process enables the synthesis of phosphors with a small size. The inorganic salt-based sol–gel approach has attracted greater interest than the alkoxide-based sol–gel process in the preparation of strontium aluminate luminescent materials [10,11] because inorganic salts are usually non-toxic and cheaper than alkoxides.

Many methods for preparing SAOs have been reported, such as high temperature solid-state reactions [12,13], sol–gel methods [14–16], co-precipitation methods [17], and hydrothermal reaction methods [18,19]. To prepare SAOs by sol–gel method, the mixed reactant sol process should be calcined at temperatures higher than 1000 ◦C as lower temperatures will not lead to SAOs with good crystallinity [20]. The calcination temperature should be lowered to accomplish low-cost preparation of SAOs.

This paper reports the preparation of SAOs by a hydrothermal reaction through a sintering process at lower temperatures. The physical properties and phosphorescence of SAOs were characterized and compared with those of SAOs prepared by the sol–gel method. The photocatalytic decomposition of methylene blue (MB) dye using SAO was performed to estimate its photocatalytic activities for the photocatalytic degradation of methylene blue dye.

#### **2. Results and Discussion**

#### *2.1. Physicochemical Properties of the SAOs*

Figure 1 presents X-ray diffraction (XRD) patterns of the SAOs obtained by a hydrothermal reaction and sol–gel method. Single-phase SAOs were obtained from the two methods. The positions and intensities of the main peaks of the two SAOs corresponded entirely to the standard card (No 34-0379). This suggests that the products were the SrAl2O4 phase. The XRD patterns of SAOs prepared by hydrothermal reaction showed many reflections. Small quantities of Sr3Al2O6 and SrAl4O7 were observed on the particle surface. Figure 2 presents scanning electron microscopy (SEM) images of the SAOs synthesized by the hydrothermal reaction and sol–gel methods. The SAOs were polycrystalline, and the particles were sintered into irregular shapes due to the high calcination temperature. The crystallinity of SAOs prepared by the sol–gel method was superior to that prepared by the hydrothermal reaction. In particular, SEM images of the SAOs prepared by a hydrothermal reaction showed smaller crystals on the surface of the particles. This suggests that the smaller crystals on the particle surface in the SEM images are the Sr3Al2O6 and SrAl4O7 phases.

**Figure 1.** XRD patterns of strontium aluminates (SAO) synthesized by (**a**) hydrothermal reaction and (**b**) sol–gel method.

**Figure 2.** Scanning electron microscopy (SEM) images of SAOs synthesized by (**a**) hydrothermal reaction and (**b**) sol–gel method.

Figure 3 presents the energy-dispersive X-ray (EDX) spectra of the SAOs prepared by the hydrothermal reaction and sol–gel methods. Peaks for Sr and Al were observed. The intensities of the Sr and Al peaks in SAOs prepared by a hydrothermal reaction were similar to those prepared by the sol–gel method. Figure 4 presents Fourier transform infrared (FTIR) spectra of the SAOs. SrAl2O4 belongs to a distorted stuffed tridymite structure. Tridymite is a member of the nepheline family of structures consisting of a corner-sharing tetrahedral framework that distorts to form large cation-occupying cavities. In SrAl2O4, the framework is built up by AlO4 tetrahedra and the structural channels are occupied by Sr2+ ions [21]. The XO4 molecule will have four degenerate normal modes of vibrations: Symmetric stretching (γs), symmetric bending (δs), antisymmetric stretching (γas), and antisymmetric bending (δd) [22].

**Figure 3.** Energy-dispersive X-ray (EDX) spectra of SAOs synthesized by (**a**) hydrothermal reaction and (**b**) sol–gel method.

**Figure 4.** Infrared spectra of the SAOs prepared by (**a**) hydrothermal reaction and (**b**) sol–gel method.

Figure 5 presents the N2 isotherm of SAOs prepared by a hydrothermal reaction. SAOs are composed of single crystals, as defined in the SEM images. The type of isotherm of SAOs indicated the typical adsorption pattern of nonporous particles. The hysteresis in the isotherm curve was derived from some crevices between the particles. The specific surface areas of the SAOs determined from the Brunauer–Emmett–Teller (BET) equation were 62.5 m2/g and 51.6 m2/g, respectively.

**Figure 5.** N2 isotherm of SAOs prepared by (**a**) hydrothermal reaction and (**b**) sol–gel method.

#### *2.2. Luminescent Properties of the SAO Products*

Figure 6 presents the emission spectra of the SAOs prepared by the hydrothermal reaction and sol–gel methods. The luminescence properties of the SAO particles were measured in the solid state at room temperature. Regardless of the preparation methods, the SAO particles exhibited similar emission peaks centered at approximately 615 nm under an excitation of 266 nm. The intensities of the peaks for the SAOs prepared by both synthesis methods were similar, which was associated with the typical 4f65d1→4f7 transition of the Eu2+ ion in SrAl2O4. This strongly affected the nature of the Eu2+ surroundings, where the shielding function of the electrons in the inner shell split the mixed states of 4f and 5d by the crystal field [23]. The special emissions of Dy3+ and Eu3+ were not observed in the spectra. The Eu3+ ions in the precursor were reduced to Eu2+ in a weak reducing atmosphere. The Eu2+ ions in the precursor were reduced to Eu+. The Dy3+ ions were oxidized to Dy4+ during excitation [24]. Simultaneously, thermal vibrations of the surrounding ions and local vibrations in the lattice structure resulted in luminescence spectra with broad bands [25]. The SAOs prepared by the hydrothermal reaction through lower temperature calcination exhibited a similar emission intensity to the SAOs prepared by the sol–gel method through a higher temperature calcination.

**Figure 6.** Emission spectra of the SAOs prepared by (**a**) hydrothermal reaction and (**b**) sol–gel method.

Figure 7 presents the UV-visible diffuse reflectance spectra (DRS) results of the SAOs and titanium dioxide (TiO2) converted to Kubelka–Munk units. The optical bandgap (Egap) was calculated using the method proposed by Kubelka and Munk for indirect electronic transitions [26]. The Kubelka–Munk equation is expressed as F(R) = (1 − R)2/2R = K/S, where R, K, and S are the absolute reflectance, absorption coefficient, and scattering coefficient, respectively. The optical properties of the SAOs were induced by light absorption in the photochemical reaction. The diffuse reflectance spectrum of TiO2 exhibited an adsorption edge at ca. 380 nm. The bandgap of TiO2 determined from the adsorption edge was 3.2 eV. By contrast, the DRS of the SAOs were shifted to the upper wavelength range. The SAOs exhibited a significant increase in wavelength. The bandgap of the SAOs was ca. 2.9 eV.

**Figure 7.** UV-visible diffuse reflectance spectra (DRS) of the SAOs prepared by (**a**) hydrothermal reaction and (**b**) sol–gel method.

#### *2.3. Photocatalytic Properties of the SAO*

Figure 8A presents the UV-vis spectra of a pure MB solution, MB solution containing SAO prepared by a hydrothermal reaction, and MB solution containing SAO prepared by the sol–gel method. The absorbance increased with the injection of SAOs prepared by different methods in the MB solution. Figure 8B presents the absorbance of the MB solution containing SAO prepared by hydrothermal reaction and the sol–gel method as a function of the irradiation time of UV light at 300 nm. The absorbance decreased with the injection of SAO into the MB solution despite the short irradiation time. This indicates that the injection of SAOs into the MB solution leads to rapid degradation of the MB dye, and the degradation rate was higher in the MB solution containing SAO prepared by the sol–gel method.

**Figure 8.** Results of (**A**) UV-vis spectra of as a function of wavelength of (**a**) methylene blue(MB) solution, (**b**) MB solution containing SAO prepared by hydrothermal reaction, and (**c**) MB solution containing SAO prepared by sol–gel method; (**B**) variation of absorbance of the MB solution containing (**a**) SAO prepared by hydrothermal reaction, and (**b**) SAO prepared by sol–gel method as a function of irradiation time of UV light at 300 nm.

Figure 9a presents the changes in MB concentration with different initial MB concentrations on a TiO2 photocatalyst. The MB concentrations decreased due to the photocatalytic decomposition of MB. The rate of MB degradation was faster at lower initial concentrations of MB than at higher initial concentrations because the photoefficiency increases with decreasing dye concentration. In addition, a large amount of dye might be adsorbed on the TiO2 surface, which can prevent the dye molecules from coming in contact with the free radicals and electron holes. Figure 9b shows the photocatalytic degradation of MB on TiO2 and the SAO photocatalysts at the same initial concentration of MB. The degradation of MB was faster with the SAO photocatalysts than that with the TiO2 photocatalyst. This suggests that SAOs have higher photocatalytic activity than the TiO2 photocatalyst. The higher photocatalytic activity of SAO was attributed to its higher photosensitivity, which was defined in the DRS results. The SAOs showed a lower bandgap than TiO2. A lower bandgap of SAOs led to enhanced photocatalytic activity than TiO2.

**Figure 9.** Variation of the concentration of MB by the photocatalytic decomposition on titanium dioxide (TiO2) with (**a**) various initial concentrations and (**b**) on SAOs at an initial MB concentration of 10 mg/L.

#### **3. Materials and Method**

#### *3.1. Preparation of SAOs*

The SAOs were prepared by both a sol–gel method and hydrothermal reaction. In both methods, aluminum nitrate nonahydrate (Al(NO3)3·9H2O, 99%; Duksan, Ansan-City, Korea), strontium nitrate

(Sr(NO3)3,99%; Duksan, Ansan-City, Korea), dysprosium(III) nitrate pentahydrate (Dy(NO3)3·5H2O, 99%; Alfa Aesar, Ward Hill, MA, USA), and europium(III) nitrate hexahydrate (Eu(NO3)3·6H2O, 99%; Alfa Aesar, Ward Hill, MA, USA) were dissolved in distilled water with stirring for 30 min at 90 ◦C. The chemicals of the reactants were of analytical grade and used as received. The solutions were combined according to the molar ratio of Sr:Al:Eu:Dy = 0.97:2:0.01:0.02.

The chelating reagent solution was prepared using the appropriate amount of aqueous citric acid solution and added dropwise to the above solution. A boric acid solution was then added to the chelating reagent solution. The mixture was concentrated at 80 ◦C with stirring until it changed to a high viscosity translucent gel. The mixture was then calcined in an electric muffle furnace at 1100 ◦C for 3 h in a weak reducing atmosphere using a hydrogen-containing gas mixture (Ar:H2 = 95:5).

SAOs were also prepared by a hydrothermal reaction. The mixture of SAO precursors and the chelating agent solution were poured in an autoclave. The autoclave containing the reactant was heated to 130 ◦C with stirring using a magnetic stirrer. The hydrothermal reaction conditions were maintained for 5 h. After the hydrothermal reaction, the product was dried at 110 ◦C for 12 h. The product was then calcined at 550 ◦C for 4 h in a reducing atmosphere. Commercially available TiO2 (Degussa, P25, Krefeld, Germany) with a particle size and specific area of ≈30 nm and ≈50 m2/g, respectively, was also used for photocatalytic decomposition.

#### *3.2. Photocatalytic Decomposition of Methylene Blue*

The MB solution (100 mL) was mixed with the photocatalysts as a reactant mixture. The photocatalyst loading was adjusted to 5 mL because of the significant difference in the densities between TiO2 and SAOs. The reactant mixture was stirred in the dark for 1 h to reach adsorption equilibrium. The reactant mixture was irradiated with UV light with stirring. Samples were taken at regular intervals. They were the centrifuged and the photocatalysts were separated. The concentrations of the samples were analyzed by UV-visible spectrophotometry (Shimadzu UV-2450, Tokyo, Japan). The concentrations of MB were determined from the calibrated absorbance at 665 nm using a spectrophotometer.

The photocatalytic degradation of MB was carried out using a glass reactor-installed UV lamp system. The reactor was kept in the dark to prevent the dispersion of UV light during the photoreaction. The photoreaction temperature was maintained at 25 ◦C. The UV array consisted of two 10 W UV-A lamps. The UV emission wavelength and light strength was 365 nm and 30 Lx, respectively.

#### *3.3. Characterization of the SAOs*

The phase of the SAO particles was determined by XRD (Rigaku Model D/max-II B, Texas, USA). XRD was conducted at 40 kV and 30 mA with a scan speed of 5◦/min, scanning angle from 10◦ to 60◦, and a step of 0.02◦ using Cu Kα radiation. The crystal size and morphology of the SAOs were investigated by SEM (Hitachi S-4700, Tokyo, Japan). Their chemical components were analyzed by EDX (NORAN Z-MAX 300 series, Tokyo, Japan). The N2 isotherms of the SAOs were investigated using a volumetric adsorption apparatus (Mirae SI, Porosity-X, Gwangju, Korea) at liquid N2 temperature. The sample was pretreated at 150 ◦C for 1 h before exposure to nitrogen gas. The surface areas of the samples were calculated using the BET equation [27]. Transmission electron microscopy (TEM, JEOL JEM-2100F, Tokyo, Japan) was performed using a LaB6 filament and operated at 200 kV.

The composition of the phosphors was analyzed using an EDX micro analyzer (JEOL JSM-840A, Tokyo, Japan) mounted on the microscope. The photoluminescence was evaluated by photoluminescence spectroscopy (PL, Spectrograph 500i, Acton Research Co., Acton, MA, USA) with a 0.2 nm resolution at room temperature. The samples were excited at 226 nm using a He-Cd laser. The UV–vis diffuse reflectance spectra were measured using a UV-vis spectrometer (Shimadzu UV-2450, Tokyo, Japan) in the region, 200–700 nm, with BaSO4 as the reflectance standard. The optical bandgap (*E*gap) was calculated using the Kubelka–Munk method for indirect electronic transitions.

#### **4. Conclusions**

Strontium aluminates co-doped with europium and dysprosium was prepared by a hydrothermal reaction through a sintering process at lower temperatures. The physicochemical properties of the SAOs characterized by SEM-EDX, photoluminescence, and UV-visible DRS were similar to those of the SAOs prepared by the sol–gel method. Although SAOs had been calcined at lower temperatures, their characteristics matched the standard. The photocatalytic activity for the photodecomposition of MB dye was higher than that of the TiO2 photocatalyst. The SAOs exhibited higher photocatalytic activity than the TiO2 photocatalyst. The higher photocatalytic activity of SAO was attributed to its higher photosensitivity.

**Acknowledgments:** This paper was supported by Research Funds of Kwangju Women's University in 2017. **Conflicts of Interest:** The author declares no conflict of interest.

#### **References**


© 2018 by the author. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

MDPI St. Alban-Anlage 66 4052 Basel Switzerland Tel. +41 61 683 77 34 Fax +41 61 302 89 18 www.mdpi.com

*Catalysts* Editorial Office E-mail: catalysts@mdpi.com www.mdpi.com/journal/catalysts

MDPI St. Alban-Anlage 66 4052 Basel Switzerland

Tel: +41 61 683 77 34 Fax: +41 61 302 89 18