Does the low optical band gap of yellow Bi3YO6 guarantee the photocatalytical activity under visible light illumination?
- 435 Downloads
Bi3YO6, which is known as an ionic conductor, was tested here as an electrode and photoanode in contact with aqueous electrolytes. Bi3YO6 was deposited onto the Pt substrate and the such prepared electrode was polarized in various aqueous electrolytes. The optical energy band gap of the material equal to 1.89 eV was determined using the Kubelka-Munk function resulting from the UV-Vis spectrum (allowed indirect transition) and also was calculated using the semi-empirical PM7 method (3.38 eV of HOMO-LUMO energy gap). Despite the yellow color of Bi3YO6, the tested material exhibits photoelectroactivity only in the UV range of electromagnetic radiation. The anodic photocurrent characteristic for n-type metal oxide semiconductors was recorded. The electrode exhibits diffusion-controlled cathodic activity while polarized in chloride-free aqueous electrolytes.
KeywordsBi3YO6 Semiconductor Aqueous electrolyte Photoactivity Photoanode
Metal oxides are known to be used in many electrochemical-based devices such as gas sensor, biosensor, and chemical sensor , optical sensors , lithium ion batteries , photoanodes , and environmental remediation photocatalysts . A group of bismuth inorganic compounds such as Bi2O3, BiVO4, BIMEVOX, BiMoO6, BiOCl, and Bi2WO6 are known to act as electrode materials, exhibiting high electrical capacity, chemical stability under multiple polarization cycles, photoelectroactivity, and photocatalytical properties [6, 7, 8, 9, 10, 11]. Their potential applications spread from supercapacitors to photoelectrochemical cells for water splitting or water pollutant degradation. Some of these compounds belong to high temperature solid state electrolytes, having a structure allowing oxygen ions mobility to occur.
The δ-Bi2O3 is a high temperature polymorph of bismuth oxide . The significant decrease of ionic conductivity at lower temperatures is observed due to the phase transition of cubic δ-phase to α, β, or γ polymorphs. Great effort has been made to preserve the high-conducting structure at lower temperatures. It can be achieved via partial substitution of Bi atoms in Bi2O3 by, e.g., rare earth metals , which leads to the formation of a series of new compounds with new properties. One of them is Bi3YO6 that occurs in the δ-phase even at room temperature . The defect structure, vacancy ordering, and oxygen ion transport in the Bi3YO6 δ-phase were studied using ab initio molecular dynamic, as well as total neutron scattering analysis [15, 16, 17]. Ionic transport in a material and its defect structure are commonly tested in Bi3NbO7-Bi3YO6 systems [18, 19] and in tungsten-doped Bi3YO6 .
Some of bismuth-containing solid electrolytes which are characterized by high ionic conductivity may exhibit also gas sensing properties  or be tested as solid membranes for gas separation . Such materials in contact with aqueous electrolytes may show completely new features. The materials from the Bi2O3-V2O5-MexOy system (BIMEVOX) are a good example. It was reported that BIMEVOX layers deposited onto the conductive substrate exhibit photoelectrochemical activity and may be used as photoanodes [8, 23, 24]. The unique electronic structure of Bi-containing oxides characterized by a well spread valence band consisting of Bi 6s and O 2p orbitals makes them good candidates for being visible-light active photocatalysts . Materials in the form of powder were tested also as catalysts active under visible light illumination, able to photodegrade organic contaminations [26, 27] and photoreduce Cr (VI) . Thus, materials which are known as solid electrolytes and exhibit outstanding ionic conductivity should be tested and exploited as semiconductors at low temperatures in contact with aqueous electrolytes. We focus on these compounds which exhibit optical properties, suggesting their activity under visible light illumination in respect to photoelectroactivity—semiconductors with a narrow energy band gap. Such a material is tested here—the yellow in color Bi3YO6 double oxide known as a high temperature solid state electrolyte .
In the present work, Bi3YO6 powder was deposited onto a conductive substrate and tested as an electrode. The influence of electromagnetic radiation on electrochemical performance of prepared electrodes was studied. The films of Bi3YO6 on Pt substrate were used as photoanodes for photoelectrocatalytical oxidation of water. The structure, surface, and optical properties were tested using XRD, FT-IR, XPS, and UV-Vis, respectively.
Ambient temperature X-ray powder diffraction data were collected on a Philips X’Pert Pro X-ray diffractometer fitted with an X’Celerator detector, using Ni filtered Cu-Kα radiation (λ1 = 1.54056 Å and λ2 = 1.54439 Å), in flat plate θ/θ geometry on a spinning sample holder. Data collection was carried out the range 5–125° 2θ, in steps of 0.0167°, with an effective scan time of 50 s per step. Calibration was carried out with an external LaB6 standard.
The UV-Vis spectra of Bi3YO6 were recorded using a dual beam UV-Vis spectrophotometer (Lambda 35, Perkin-Elmer) equipped with a diffuse reflectance accessory. FT-IR analyses were carried out by using a Nicolet 8700 FT-IR spectrometer equipped with an attenuated total reflectance (ATR) accessory. The morphology of the Bi3YO6 layers was investigated by the Schottky field emission scanning electron microscope (FEI Quanta FEG 250) with an ET secondary electron detector. The beam accelerating voltage was kept at 15 kV. Electrochemical and photoelectrochemical experiments (cyclic voltammetry, electrochemical impedance spectroscopy, chronoamperometry) were performed using the AutoLab PGStat10 potentiostat–galvanostat system under GPES 4.9 software control. Electrochemical impedance spectroscopy (EIS) measurements were recorded in 0.2 M KCl and 0.1 M K2SO4 in the 20 kHz–0.1 Hz frequency range at rest potential. The fitting procedure was performed using EIS spectrum analyzer software. Three types of elements were utilized to prepare the electrical equivalent circuit: R—resistor, CPE—constant phase element, and Wo—Warburg open element (the impedance of finite-length diffusion with reflective boundary).
Measurements carried out under illumination were performed in a photoelectrochemical cell equipped with a quartz window. The geometrical surface area of the electrodes was equal to ~ 0.5 cm2. A high-pressure 150-W xenon lamp (Osram XBO 150) with and without the AM1.5 filter was used as a source of electromagnetic radiation. The light intensity was adjusted to 100 mW cm−2 (with the AM1.5 filter) and 160 mW cm−2 (without the AM1.5 filter) and was controlled by an Ophir power meter. All electrochemical measurements were carried out at room temperature 21 °C. XPS analysis was performed for three Pt/Bi3YO6 samples before and after polarization in K2SO4 electrolyte using Escalab 250Xi from Thermo Fisher Scientific. In order to normalize spectroscopic measurements, the x axis (binding energy) from the XPS spectrum was calibrated for peak characteristics for carbon 1 s (284.6 eV). Data analysis was performed using Avantage software provided by the manufacturer. One sample was electrooxidized (60 min at E = 0.9 V), the second one was electroreduced (60 min at E = − 0.85 V), and the last one was not polarized.
Li2SO4, Na2SO4, K2SO4, Cs2SO4, KCl, and KNO3 used as electrolytes were of analytical grade and were supplied by POCH. Bi2O3 (99.9%) and Y2O3 (99.99%) used for material synthesis were supplied by Sigma Aldrich.
Bi3YO6 was synthesized by conventional solid state reaction technique. Sample of Bi3YO6 was prepared using stoichiometric amounts of Bi2O3 and Y2O3. The starting mixtures were ground in ethanol using a planetary ball mill. The dried mixtures were heated at 740°C for 24 h, then slow cooled and reground. The sample was then reheated at 800 °C for further 24 h before slow cooling in air to room temperature, over a period of approximately 5 h.
Bi3YO6 powder was deposited onto the platinum foil using the dip-coating method. First, 0.2 g of material and about 0.1 g of poly(ethylene oxide—PEO) (M = 300,000, Aldrich) were mixed with 1 ml of water. The platinum foil was immersed in the resulting suspension, pulled out, dried, and heated for 5 h at 400 °C in an air atmosphere (with a heating rate of 1 °C min−1). The annealing procedure was performed in a quartz tube using a Czylok PRC 55 L/1300 M furnace.
To avoid contact of the studied material with PEO, some of the electrodes were prepared using a glass capillary filled with Bi3YO6, where the Pt wire was used as an electrical contact. To test the influence of binder on the photoelectrochemical properties, some of the electrodes were prepared by drop-casting of the Bi3YO6 water suspension on the platinum foil and then dried at 100 °C.
Powder and electrode characterization
The energy band gap was determined by extrapolation of the linear region of (f(KM) hν) n vs. hν and taking an intercept on the x axis. The power “n” is dependent on the type of electron transition (n = 2—direct allowed (d.a.), n = 0.5—indirect allowed (i.a.), n = 2/3—direct forbidden (d.f.), n = 1/3—indirect forbidden (i.f.)) . All possibilities (direct and indirect, allowed and forbidden) for the material in the form of powder are shown in Fig. 4b, c (differences between powders and layers were negligible as presented in Fig. S1 in supporting information). The linear region of (f(KM)·hν) n vs. hν function can be found in all cases. The estimated values of energy band gaps for each case are marked with an appropriate color on the UV-Vis spectrum in Fig. 4a. There are materials which exhibit two types of transitions [35, 36]. However, taking into account the UV-Vis spectrum in Fig. 4a, absorption starts to rise near 656 nm, which is the value that corresponds to 1.89 eV. Thus, it is very likely that the optical band gap of tested material is related to the allowed indirect transition. The determined value of an energy band gap allowing visible light absorption makes Bi3YO6 interesting from the photocatalytical point of view. Nevertheless, optical properties of the studied material should be further examined to describe them in more detail.
The energy band gap of tested material has been also determined using semi-empirical PM7 calculations under LS (Singlet) Born–von Kärmän periodic boundary conditions  implemented in MOPAC2016, Version: 17.119W package by James J. P. Stewart  at experimental geometry of cubic δ-type phase of a = 5.496 Å. Surprisingly, the calculated value of Eg of Bi3YO6 is equal to 3.38 eV. This value is 1.49 eV higher than the value of the indirect gap determined from the UV-Vis spectrum. The average unsigned error of PM7 calculated ionization energy in sets of reference compounds (http://openmopac.net/PM7_accuracy/molecules.html) is 0.55 eV. Such a big difference between determined values may then result from the surface properties of Bi3YO6. Exclusion of the influence of absorption by impurities or surface electronic states cannot be unambiguously done . For example, as it was previously shown for yttrium-doped BiVO4 (BixY1-xVO4), a surface of bismuth-containing metal oxide is mainly built by BiOy units . Their presence allows the part of visible light to be absorbed; however, the bulk material is characterized by a different electronic structure and absorbs only UV light.
The electroactivity recorded in sulfate- and nitrate-containing electrolytes is not related to the surface groups observed on the FT-IR spectrum and formed during layer preparation. An additional cyclic voltammetry curve was recorded for the electrode prepared without PEO (using a glass capillary filled with Bi3YO6 and a platinum wire as the electrical contact). CV is presented in Fig. 5f. As it can be seen, the shape of the curve is the same as in the case of the electrodes prepared by the dip-coating method; thus, the used method of film deposition does not affect Bi3YO6 electroactivity.
The Pt/Bi3YO6 was tested as an potential electrode for energy storage devices. The capacitance was calculated from multiple galvanostatic (1 mA cm−2) charge/discharge cycles recorded in 0.1 M K2SO4 presented in Fig. 5g. Multiple chronopotentiometry curves allow the electrochemical stability to be tested. The capacitance is relatively stable during the first 6000 cycles and then surprisingly increases reaching the value of 10.5 mF cm−2 (~ 500 mF g−1, ~ 2100 mF cm−3) as it is shown in Fig. 5h. It may be related to the penetration of the layer pores with electrolyte. Material started to lose its capacitance after 12,500 cycles which indicates high electrostability of the tested Bi3YO6 electrode.
The doublet characteristic for yttrium 3d orbital overlaps with the signal from Bi 4f, as it is shown in Fig. 7c. After deconvolution of all samples’ spectra, positions of peaks at 155.7 eV (A) and 158.1 eV (B) were shifted slightly (± 0.4 eV) between the samples and no new peak characteristic for yttrium appeared after electrochemical treatment. More detailed analysis of the Y 3d orbital is very difficult due to the low intensity of peaks in comparison with overlapping Bi 4f peaks. Polarization of Pt/Bi3YO6 does not clearly affect Y atoms in the samples. However, changes of the peaks coming from the Bi 4f orbital can be found. The intensity and area ratios of the peaks marked as C1 (158.3 ± 0.2 eV), C2 (163.6 ± 0.1 eV), D1 (160.3 ± 0.1 eV), and D2 (165.5 eV) change as it is shown for the samples electrooxidized and electroreduced in K2SO4 electrolyte (see. Fig. 7c, d). Two doublets of the Bi 4f orbital were already reported, e.g., for Bi2O2.33  and Bi2WO6 , and were interpreted as coexistence of Bi3+ and Bi in the lower oxidation states. It may be concluded that electroactivity recorded on the cyclic voltammetry curves is related to the changes of Bi species on the surface of the electrode. Thus, cathodic polarization of the electrode leads to the formation of the “Bi suboxides” [57, 58] as the increase of the intensity of the peak related to the reduced form of Bi is observed.
Photoelectrochemical properties of prepared electrodes were examined using the chronoamperometry technique recorded under intermittent illumination. First, the electrode was polarized (E = const) in dark conditions to achieve a steady state current. Then, the electrode was illuminated and the current was measured. In the case of simulated solar light (AM 1.5 filter), the photocurrent was not generated. Despite the material is yellow and exhibits absorption in the visible range of electromagnetic radiation, absorbed photons were not converted to electrical energy. Thus, it is very likely that UV-Vis spectroscopy shows the surface electronic states, but the real, “bulk” energy band gap is higher as it follows from the calculations.
The photocurrent recorded when the electrode was immersed in KCl and K2SO4 is comparable and is related to water and/or chloride oxidation. Illumination of the electrode which is in contact with KOH leads to an over 20 times higher photocurrent. OH− anions may act as “hole scavengers” because of a lower potential of O2 evolution in comparison with the water oxidation reaction (2H2O ➔ O2 + 4H+ + 4e−, E = 1.23 V and 4OH−➔ O2 + 2H2O + 4e−, E = 0.4 V). The chronoamperometry curves of the Pt/Bi3YO6 electrode at E = − 0.75 V vs. Ag/AgCl (0.1 M KCl) are presented in Fig. 8b. In the case of the KOH and KCl electrolyte, the effect of photocurrent generation is not clearly observed. However, the electrode immersed in K2SO4 electrolyte generates an anodic photocurrent even at cathodic potential. The n-type semiconductor can act as a photoanode only when the applied potential is higher (more anodic) than the flat band potential (Efb). It means that the Efb is more cathodic than − 0.75 V vs. Ag/AgCl (0.1 M KCl); however, the precise value was not evaluated. A routine procedure for Efb evaluation using Mott-Schottky plot is found in the studied case to be ambiguous due to complexity of the electrode/electrolyte interface in a broad potential range. Surface active species give rise to electrochemical capacitance of Pt/Bi3YO6 electrodes as one may see on CV curves (see Fig. 5).
It was reported that photoactivity of bismuth-containing catalysts may be inhibited due to the BiOy clusters on the surface of the oxide. Differences in the electronic structure of the bulk material and the BiOy-rich surface lead to the adverse “self-heterojunction” formation . The authors proposed a method of BiOy removing from the surface using diluted HNO3. Almost four times enhancement of photocatalytic water splitting efficiency was observed after HNO3 treatment . In the present work, the same method has been utilized to improve photoelectrocatalytical performance of Bi3YO6. The comparison of chronoamperometry curves recorded during illumination of the Pt/Bi3YO6 photoanode before and after acid treatment is presented in Fig. 8c (E = 0.8 V, 0.1 M K2SO4). It is clearly seen that HNO3 treatment leads to photocurrent increase from 0.19 to 1.14 μA cm−2. Thus, acid treatment enhanced not only photocatalytic, but also photoelectrocatalytic properties of Bi-containing metal oxide semiconductors.
The influence of the deposition procedure on the Bi3YO6 structure and optical properties was presented. The surface of the deposited layer was slightly changed in comparison with the bulk material, as it was presented in IR spectra. The indirect energy band gap of the tested material was estimated to be 1.89 eV using the Kubelka-Munk function, which is a 1.49 eV lower value than that calculated using semi-empirical PM7 calculations. Photoelectrochemical tests under UV-Vis illumination show that the Bi3YO6 film deposited onto Pt foil may act as a photoanode. An anodic photocurrent was generated in a wide range of applied potentials, proving that the tested material is an n-type semiconductor. It was shown that despite optical energy band gap equals to 1.89 eV, tested material does not generate photocurrent when illuminated with visible light. The electron transition observed on the reflectance spectrum cannot be converted to the photoelectrochemical water oxidation.
The electrode Pt/Bi3YO6 exhibited electrochemical activity related to the changes of the Bi oxidation state when polarized in aqueous electrolytes. However, the reversible cathodic process at E = − 0.75 V vs. Ag/AgCl (0.1 M KCl), which is controlled by diffusion, was registered only in Cl−-free electrolytes. The influence of electrolyte (KCl and K2SO4) on electrical properties of the electrode/electrolyte interface was investigated using electrochemical impedance spectroscopy. The lower electrical capacitance was observed in the KCl electrolyte.
The financial support provided by the Gdańsk University of Technology DS 032406 is gratefully acknowledged.
- 23.dos Santos WS, Rodriguez M, Afonso AS, Mesquita JP, Nascimento LL, Patrocinio AOT, Silva AC, Oliveira LCA, Fabris JD, Pereira MC (2016) A hole inversion layer at the BiVO4/Bi4V2O11 interface produces a high tunable photovoltage for water splitting. Sci Rep 6(1):31406CrossRefPubMedPubMedCentralGoogle Scholar
- 29.Larson AC, Von Dreele RB (1987) Los Alamos National Laboratory report LA-10160-MS. No. LAUR-86-748Google Scholar
- 38.Stewart Computational Chemistry. http://openmopac.net
- 40.Fang W, Liu J, Yang D, Wei Z, Jang Z, Shangguan W (2017) Effect of surface self-heterojunction existed in BixY1-xVO4 on photocatalytic overall water splitting effect of surface self-heterojunction existed in BixY1-xVO4 on photocatalytic overall water splitting. ACS Sustain Chem Eng 5(8):6578–6584CrossRefGoogle Scholar
- 59.Bott AW (1998) Electrochemistry of semiconductors. Curr Sep 3:87–91Google Scholar
Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.