Research Article www.acsami.org
Efficient Upconverting Multiferroic Core@Shell Photocatalysts: Visible-to-Near-Infrared Photon Harvesting Jianming Zhang,†,⊥ Yue Huang,‡,⊥ Lei Jin,‡ Federico Rosei,‡,§ Fiorenzo Vetrone,*,‡,§ and Jerome P. Claverie*,∥ †
School of Chemistry and Chemical Engineering, Jiangsu University, Zhenjiang 212013, China Centre Énergie, Matériaux et Télécommunications, Institut National de la Recherche Scientifique, 1650 Boul. Lionel Boulet, Varennes, Québec J3X 1S2, Canada § Institute for Fundamental and Frontier Science, University of Electronic Science and Technology of China, Chengdu, China ∥ Department of Chemistry, Université de Sherbrooke, Sherbrooke, Quebec J1K 2R1, Canada
Downloaded via EASTERN KENTUCKY UNIV on January 27, 2019 at 00:53:14 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.
‡
S Supporting Information *
ABSTRACT: We report the two-step synthesis of a core@shell nanohybrid material for visible-to-near-infrared (NIR) photocatalysis. The core is constituted of NaGdF4:Er3+, Yb3+ upconverting nanoparticles (UCNPs). A bismuth ferrite (BFO) shell is assembled around the UCNPs via a hydrothermal process. The photocatalytic degradation assays of methylene orange and 4-chlorophenol reveal that these core@shell nanostructures possess remarkably enhanced reaction activity under visible and NIR irradiation, compared to the BFO powder alone and the BFO-UCNP mixture. Photo-charge scavenger tests and fluorescent assays indicate that hydroxyl radicals play a pivotal role in the photodegradation mechanism. The enhanced photoactivity of the core@shell structure is attributed to the NIR radiation which is converted into visible light by UCNPs, and which is then captured by BFO via a nonradiative luminescence resonance energy transfer process. Therefore, this core@shell architecture optimizes solar energy use by efficiently harvesting visible and NIR photons. KEYWORDS: upconversion, photocatalysis, BiFeO3, energy transfer, nanoparticle
■
activity as photocatalysts under 980 nm light irradiation.6−12,15 These results constitute a proof of principle that NIR photons can be harvested by UCNPs, leading to the formation of photogenerated charge carriers in the TiO2 photocatalyst. Yet the gain in photocatalytic activity remains low since the emission of the UCNPs in the UV range (where TiO2 is active) is usually not intense; that is, only a small portion of emitted photons possesses an energy higher than the bandgap energy, Eg, of TiO2. To our knowledge, no UCNPs can only emit UV photons without also emitting VIS ones,16 thus, a more efficient design for a broad spectral-range photocatalyst should be based on a semiconductor with a low enough Eg, so as to capture the highest possible number of photons emitted by the UCNPs (either via true radiation-reabsorption mechanism or via Förster energy transfer process). Recently, bismuth ferrite (BiFeO3, BFO), a widely studied multiferroic (ferroelectric and antiferromagnetic) material, has been shown to be a promising photocatalyst for the degradation of organic pollutants.17−23 With a bandgap of ∼2.0−2.6 eV (vs 3.0−3.2 eV for TiO2), BFO exhibits excellent response both to
INTRODUCTION The efficient removal of dissolved organic pollutants is a major challenge for wastewater treatment.1,2 Heterogeneous photocatalysis is one of the most promising depollution methods, since pollutants can be fully decomposed without subsequent manipulation or separation.2 As one of the most popular photocatalysts, TiO2 nanoparticles (NPs) have been studied extensively because of their high activity under ultraviolet (UV) illumination.3 However, pure TiO2, with its large bandgap energy (∼3.0−3.2 eV),3,4 is only active in the UV range. The abundance of visible (VIS, ∼49%) and near-infrared (NIR, ∼46%) light compared to UV light (∼5%) in the solar spectrum has thus stimulated the development of photocatalysts which are active in the longer wavelength range. Luminescent lanthanide (Ln3+)-doped upconverting nanoparticles (UCNPs) are well-known for their ability to convert lower energy NIR photons to higher energy ones via a multiphoton process commonly referred as upconversion.5 This multiphoton excitation occurs through a plethora of 4f excited electronic energy states which have long lifetimes (micro- to millisecond), resulting in the ladder-like sequential absorption of NIR photons. These outstanding optical properties have recently been exploited in the context of photocatalysis.6−15 For example, Yb/Tm codoped UCNP-TiO2 nanohybrids exhibit © 2017 American Chemical Society
Received: January 5, 2017 Accepted: February 17, 2017 Published: February 17, 2017 8142
DOI: 10.1021/acsami.7b00158 ACS Appl. Mater. Interfaces 2017, 9, 8142−8150
Research Article
ACS Applied Materials & Interfaces UV and VIS light.19,22,24 Therefore, an appropriate combination of UCNPs with BFO could potentially yield efficient photon harvesting from the VIS to NIR spectral ranges with broad applications in photocatalysis. A core@shell structure, where the semiconductor is deposited at the UCNPs surface, is expected to exhibit higher energy transfer efficiency compared to a simple mixture of both materials. Indeed, an efficient luminescence resonance energy transfer (LRET) process can occur when the two materials are in close contact with each other,6,7 in contrast with a conventional radiation-reabsorption process, which typically dominates when the two materials are separated. However, to the best of our knowledge, UCNP@ BFO core@shell structures have never been reported, probably because of the degradation of UCNPs which occurs during BFO synthesis at acidic pHs. We therefore developed a novel synthetic process for BFO where the pH is kept neutral or basic throughout the preparation. With these results in hand, we report a UCNP@BFO core@shell hybrid nanostructure for VIS-NIR photocatalysis and demonstrate that it is active for the photodegradation of organic compounds under VIS and NIR irradiation. The BFO shell is highly beneficial to capture VIS light in a wide spectral range, and close contact between the UCNP core and the BFO shell leads to efficient nonradiative energy transfer from UCNP to BFO. Due to the core@shell morphology, the entire surface of the photocatalyst is exposed to the environment, which is optimal for photocatalytic degradation of pollutants, while the NIR light can also access the UCNP core as the BFO is transparent in the NIR range. Remarkably, even when monochromatic NIR irradiation is used, hydroxyl radicals are generated by the photocatalyst, therefore indicating that this photocatalyst is putatively able to degrade persistent organic pollutants using only NIR light. This novel photocatalyst design optimally exploits upconversion of NIR photons, thus paving the way toward broad-spectrum highly active photocatalysts.
■
aqueous phase containing the citrate-coated NaGdF4:Er3+, Yb3+ UCNPs were collected. The UCNPs were precipitated with acetone (1:5 aqueous:organic volume ratio), collected by centrifugation and redispersed in 5 mL of trisodium citrate buffer (adjusted to pH 7) for removal of residual oleic acid. The dispersion was placed on a shaker for an additional 2 h. Finally, the citrate-coated UCNPs were precipitated with acetone, washed three times with acetone and water, collected by centrifugation and dispersed in 2 mL of water for further experiments. Synthesis of UCNP@BFO core@shell nanostructures. First, 0.242 g (0.5 mmol) of Bi(NO3)3·5H2O and 0.202 g of Fe(NO3)3· 9H2O (0.5 mmol) were added to 10 mL of 10% HNO3 solution, to form a clear mix. Subsequently, 0.44 g (1.5 mmol) of EDTA was dissolved to this solution and heated at 80 °C for 1 h. Water was then boiled-off and, the residue was washed with methanol and centrifuged to yield EDTA-Bi/Fe(III) complexes.28,29 The above products were dissolved in 16 mL of water, and the pH was adjusted to ∼7.5 using concentrated NaOH solution (5 M). Urea and H2O2 (30 wt %) were introduced in this aqueous solution, to reach a final concentration of 0.3 and 0.1 M, respectively. Subsequently, 0.05 g of citrate capped UCNPs were added and stirred with a magnetic bar. This mixture was transferred to a Teflon bottle sealed in a stainless steel autoclave with a capacity of 20 mL and kept at 200 °C for 18 h. Finally, the product was collected and centrifuged at 2000 rpm for 5 min and the supernatant was discarded. The deposit was washed with water/ethanol followed by a thermal treatment at 400 °C for 30 min and then cleaned with ca. 6 mL of CH3COOH at 65 °C. The final product was washed three times with water and ethanol by centrifugation and dried at 60 °C overnight in a vacuum oven. Synthesis of BFO NPs. The same recipe was used with the exception that no UCNPs were added in this process. Photocatalytic Activity Test. The photocatalytic activity of the UCNP@BFO core@shell nanostructure was investigated by monitoring the photodegradation of MO and 4-CP. Prior to photoirradiation, 20 mg of the catalyst was stirred under darkness in 20 mL of a 10 mg/ L aqueous solution of the analyte for ∼8 h to reach equilibrium adsorption. (i) Visible light catalysis. A 500 W xenon lamp (Sciencetech Inc., SS0.5 kW) with an optical filter (λ > 420 nm) was used as VIS light source. The suspension was irradiated with an intensity of ∼85 mW/cm2. (ii) Degradation under NIR light. The sample preparation was the same as for the visible light experiments. The only difference is that a 980 nm diode laser was selected as NIR light source. The power of the laser was set to ∼0.08 W/mm2. To monitor the reaction, an aliquot of reaction solution was withdrawn at varied intervals, was centrifuged to separate the catalyst, and the supernatant absorption was then measured with a spectrophotometer. (iii) Active species identification. Solutions (1 mM) of tert-butyl alcohol (t-BuOH), disodium ethylenediaminetetraacetate (Na2EDTA), or 1,4-benzoquinone (BQ) were introduced into the catalysis solution as scavengers of hydroxyl radical, hole, and superoxide radical, respectively. This test was performed under the same experimental conditions as MO photodegradation. Detection of Photogenerated OH• Radicals. The UCNP@ BFO core@shell sample (15 mg) was dispersed in 3.5 mL of a 0.1 mM coumarin aqueous solution and was irradiated in a quartz cuvette (10 × 10 × 45 mm3) with NIR light (λ = 980 nm). After irradiation for a given time, 0.5 g of KCl was added into the suspension, and then the suspension was kept in the dark to precipitate the powder. After 12 h, a clear solution was obtained, and its fluorescence was measured using an excitation wavelength of 332 nm, corresponding to the excitation wavelength of umbelliferone. Characterization. (i) X-ray diffraction (XRD). The crystal structures of the synthesized materials were examined by XRD (D8 Advance; Bruker, Billerica, MA) using Cu Kα radiation (λ = 1.5418 Å). (ii) Scanning electron microscopy (SEM). SEM images were acquired using a JEOL-JSM7600F instrument. (iii) Transmission electron microscopy (TEM) and energy dispersive spectroscopy (EDS). The morphology of the samples was analyzed using a JEOL-2100F TEM. Meanwhile, the EDS spectra were acquired on the chosen area under TEM for elemental analysis. (iv) X-ray photoelectron spectroscopy
EXPERIMENTAL SECTION
Materials. Lanthanide oxides (Ln2O3 (99.99%), Ln = Gd, Yb, and Er), trifluoroacetic acid (99%), sodium trifluoroacetate (98%), oleic acid (90%), and 1-octadecene (90%) were purchased from Alfa Aesar. Bi(NO3)3·5H2O, Fe(NO3)3·9H2O, coumarin, trisodium citrate, H2O2, ethylenediaminetetraacetic acid (EDTA), HNO3, CH3COOH, methylene orange (MO), and 4-chlorophenol (4-CP) were purchased from Sigma-Adrich and used as received. All other chemicals were used without further purification. Water was Nanopure grade (18.2 MΩ·cm at 25 °C). Synthesis of NaGdF4:Er3+, Yb3+ (UCNPs). (i) Synthesis of UCNPs in organic phase. Oleate-capped UCNPs of NaGdF4 doped with 2% mol Er3+ and 20% mol Yb3+ were prepared via a thermal decomposition process reported elsewhere.25 First, lanthanide trifluoroacetates, which are used as precursors for UCNP synthesis, were prepared from the corresponding lanthanide oxides in deionized water and trifluoroacetic acid (50/50 mixture). Second, the obtained Ln3+ precursors, together with the sodium trifluoroacetate (1:2 in mole ratio), were degassed at 125 °C under vacuum and subsequently injected at a rate of 1.5 mL min−1 in a 1:1 solution of oleic acid and octadecene at 315 °C, and aged for 2 h. The prepared β-phase NaGdF4:Er3+, Yb3+ UCNPs were then washed with a mixture of hexane and ethanol (1:4 v/v) followed by centrifugation and finally redispersed in hexane. (ii) Oleate-citrate ligand exchange. To obtain water dispersible UCNPs, ligand exchange was carried out as previously described.22 Oleate-capped NaGdF4:Er3+, Yb3+ UCNPs (60 mg) obtained from the previous step were dispersed in 5 mL of hexane and mixed with 5 mL of a 0.2 M trisodium citrate buffer (adjusted to pH 4). The mixture was kept on a shaker for 3 h and subsequently transferred to a separation funnel, from which the 8143
DOI: 10.1021/acsami.7b00158 ACS Appl. Mater. Interfaces 2017, 9, 8142−8150
Research Article
ACS Applied Materials & Interfaces (XPS). Samples for XPS measurement were prepared in an identical manner as those for XRD measurements. XPS spectra were acquired using ESCA Escalab 220i XL with a twin Al Kα X-ray source. (v) UV− vis-NIR absorption. Absorption spectra were measured using a PerkinElmer Lambda 750 spectrometer equipped with an integrating sphere. All measurements were performed at room temperature. (vi) Photoluminescence (PL). PL was measured under 980 nm excitation light (BWT Beijing Ltd.). The emission spectra were recorded by a spectrophotometer (Avaspec−2048L−USB2) using an optical fiber.
alone without BFO coating. The position and relative intensity of all diffraction peaks perfectly coincide with those of JCPDS file no. 27-0699, confirming that the pure UCNPs were crystallized in the hexagonal β-phase.25 The XRD pattern of the UCNP@BFO particles is shown in red. All peaks assigned to UCNPs are preserved during the growth of the BFO shell and additional peaks belong to the rhombohedral BFO structure (JCPDS file no. 86-1518).17 Using an identical synthetic approach but without UCNPs, pure BFO NPs were also synthesized. The XRD pattern of the pure BFO NPs (Figure S1) shows the same rhombohedral crystal structure, thus confirming that the hydrothermal-calcination approach produces highly crystalline BFO. The morphology of the UCNPs before and after BFO coating was revealed by SEM and TEM with EDS (Figure 2). The UCNPs exhibit a hexagonal plate-like structure with a diameter of ∼70 nm (from corner to corner) and a thickness of ∼30 nm (Figure 2a). Using high resolution TEM (HR-TEM, Figure 2b), lattice fringes with a spacing of ∼0.52 nm are observed. They correspond to the {100} plane of the β-phase of NaGdF4:Er3+, Yb3+ UCNPs. After the hydrothermal reaction and calcination, a rugged shell with a thickness of several nanometers was grown on the UCNPs to form the core@shell nanostructure, as illustrated in the TEM images of Figure 2c,d which respectively show particles lying flatwise or sidewise on the TEM grid. Thus, the overall hexagonal shape of the UCNP is roughly conserved in the core@shell particles. In Figure 2e (HR-TEM), the lattice fringes of the two components are visible, with a spacing of ∼0.28 nm corresponding to the {110} planes of BFO and a larger lattice spacing of 0.30 nm corresponding to the {110} crystal planes of the UCNPs. These results confirm that the structures are crystalline, in agreement with the XRD results (shown in Figure 1). The high angle grain boundary between the two materials is likely due to the kinetically driven precipitation process of the BFO on UCNPs during synthesis. Indeed, when lower amounts of BFO precursors are used, incomplete coverage of the UCNPs is observed (Figure S2). The composition of the core@shell nanostructure shown in Figure 2c was further investigated by EDX confirming the presence of Yb, Gd, Fe, and Bi in the core@shell nanostructure (Figure 2f). The chemical states of the main elements of the core@shell nanostructures were examined using XPS. Figure 3a displays high resolution XPS spectra of Gd 4d and Bi 4f. The peaks with binding energies of 144.7 and 151.5 eV can be ascribed to Gd 4d5/2 and Gd 4d3/2 peaks, respectively, while those at 161.3 and 166.7 eV can be assigned to Bi 4f7/2 and Bi 4f5/2.30,31 The asymmetric profile of the O 1s band (Figure 3b) points toward the existence of more than one kind of oxygen species. Spectral deconvolution yields two characteristic peaks at 532.0 (green curve) and 534.0 eV (blue curve), respectively, corresponding to the typical peak of O2− in BFO and hydroxyl groups (−OH).32 A scan of the Fe 2p region in the hybrid nanostructure is shown in Figure 3c (top, red). The Fe 2p3/2 band (∼713.4 eV) was deconvoluted in two peaks located at 712.6 (green) and 714.5 eV (blue) which respectively correspond to Fe2+ and Fe3+.30,31 This analysis indicates that the oxidation state of the Fe in BFO coating is composed of both Fe2+ and Fe3+ with a ratio of about 1:1.6, which is unavoidable as hydrothermal processes usually result in valence fluctuations for Fe ions.31 More importantly, as compared to the Fe 2p spectrum of UCNP-BFO mixture (bottom), the Fe 2p3/2 band of UCNP@BFO shows a distinct shift of ∼1.1 eV to
■
RESULTS AND DISCUSSION Synthesis and Structural Characterization. To design our broad spectral range photocatalyst, we selected NaGdF4:Er3+, Yb3+ UCNPs as upconverting particles since they exhibit strong emission in the VIS range when excited at 980 nm.25 The matrix (NaGdF4) was chosen because it is known to be a suitable upconversion host material for photocatalytic applications, as it has low phonon energies, high luminous intensity, well-ordered crystallite structure and excellent chemical stability.26,27 The water extinction coefficient at 980 nm is 0.482 cm−1, a value which is low enough to allow NIR light propagation. The UCNP preparation involves the thermal decomposition of lanthanide trifluoroacetates in the presence of oleic acid as stabilizing ligand. The UCNPs are further transferred to water by ligand exchange. This is a critical step since the subsequent synthesis of the BFO shell is performed in aqueous solution. Usually, BFO is synthesized by dissolving Bi(NO3)3·5H2O and Fe(NO3)3·9H2O in acidic water (pH 420 nm, including NIR), indicating that MO is degraded by the photocatalyst. By contrast, no photocatalytic activity is observed in the absence of BFO (UCNPs only; Figure 5a). In the presence of BFO alone, ∼ 62% of MO was degraded in 3.5 h. The degradation efficiency was slightly enhanced (∼66%) using a mixture of BFO and UCNPs. In the case of the UCNP@BFO nanostructure, the degradation of MO was more significant, with ∼80% degradation in 3.5 h. The source of VIS light (Xe lamp) used in this experiment contained some NIR light. Thus, the difference in activity between BFO, UCNP@BFO, and UCNP-BFO is perhaps due to the presence of NIR photons which activate the UCNP part of the photocatalyst. To verify this hypothesis, the photocatalytic activity was investigated under monochromatic NIR light (λ = 980 nm). As shown in Figure 5b, MO was not degraded by UCNPs or BFO alone, but a decrease of the MO concentration was observed when the BFO-UCNPs mixture and the UCNP@BFO nanostructures were used. Remarkably, the core@shell nanostructure exhibited higher catalytic activity than the mixture of separated UCNP and BFO: after 15 h reaction, ∼ 55% vs ∼27% degradation was 8146
DOI: 10.1021/acsami.7b00158 ACS Appl. Mater. Interfaces 2017, 9, 8142−8150
Research Article
ACS Applied Materials & Interfaces observed for the former vs the latter. An apparent pseudo-firstorder reaction rate (k) was derived from the kinetic data using a model function of the form C/C0 = exp(−kt). As shown in Figure S4, the fit of the experimental data is excellent (R2 > 0.992), and the k value for core@shell is more than two times larger than that of the UCNP-BFO mixture. A sensitization mechanism, whereby the dye absorbs the light and transfers an excited electron to the semiconductor, has been shown to occur during the photodegradation of colored organic molecules. Since 4-CP is transparent in the VIS range, a sensitization mechanism can be ruled out when 4-CP is degraded (Figure 5c). The photodegradation results of 4-CP are similar to those observed with MO, with the highest concentration decrease observed for the UCNP@BFO nanostructure, indicating that the sensitization mechanism is not at stake here. The stability/durability of the catalyst is important for its actual application; therefore, the photoactivity of the UCNP@BFO sample was examined in four successive MO degradation reactions. As presented in Figure 5d, the VIS light (λ > 420 nm) photodegradation of MO remains unchanged in these continuous cycles, indicating a stable catalytic activity. Additionally, the catalyst also demonstrated good catalytic stability under NIR light irradiation (Figure S5). To better understand the photocatalytic process under NIR light, active species generated during the reaction were identified by free radical and hole scavenging experiments. tert-Butyl alcohol (t-BuOH), the sodium salt of ethylenediamine tetraacetate (Na2EDTA), and 1,4-benzoquinone (BQ) were introduced in the catalytic solution as scavengers of hydroxyl radical (OH•), hole (h+), and superoxide radical anion (O2•−), respectively.36 Figure 6a presents the photodegradation of MO catalyzed by UCNP@BFO in the presence of these various scavengers under λ = 980 nm illumination. Compared with the scavenger-free system (black circle), the reaction in the presence of the O2•− scavenger BQ (blue triangle) is slightly slower. By stark contrast, the reaction performed in the presence of the OH• scavenger is nearly completely inhibited, with only ∼5% MO degraded after 15 h. The photocatalytic activity is also greatly reduced in the presence of the hole scavenger Na2EDTA (green star), with ∼15% MO degraded in 15 h. These results strongly suggest that hydroxyl radicals, holes and superoxide radical anion all contribute to the photodegradation, but OH• radical is a key intermediate as its trapping results in a complete suppression of catalytic activity. The trapping of O2•− and holes respectively by BQ and Na2EDTA may be less efficient because the OH• radicals may concomitantly degrade these scavengers. To further verify that UCNP@BFO generates OH• under NIR irradiation, coumarin was used as a fluorescent probe for the detection of OH•.37−39 Coumarin reacts with OH• to generate 7-OH coumarin (umbelliferone) which is a strong fluorescent emitter (Figure 6b).40 After irradiation of a coumarin solution containing UCNP@BFO for a given time, the concentration of umbelliferone is obtained via measurement of its PL intensity (Figure 6c).38 Interestingly, OH• is not observed in the absence of catalyst, or with UCNPs alone. In contrast, OH• production becomes significant in the presence of UCNP@BFO particles or of a BFO-UCNPs mixture. The OH• formation is most effective with the UCNP@BFO particles, correlating with the photodegradation tests. The above experiments confirmed that the hybridization of NIR-responsive UCNPs with VIS-active BFO leads to a photocatalyst active both in VIS and NIR range. As illustrated
Figure 6. (a) Photocatalytic degradation of MO catalyzed by UCNP@ BFO under 980 nm incident light in the presence of various scavengers. (b) Formation of 7-hydroxycoumarin (umbelliferone) by reaction of coumarin with hydroxyl radicals. (c) Concentration of umbelliferone vs NIR irradiation time in the presence of UCNP@ BFO, BFO-UCNPs mixture, and UCNPs and in the absence of catalyst. Inset shows the PL spectrum evolution of umbelliferone in coumarin solution at different irradiation times.
in Scheme 1a, the NaGdF4:Er3+, Yb3+ UCNPs show typical upconverted emissions in the visible range overlapping with the absorption spectrum of BFO, which can result in energy transfer from UCNPs to BFO. In general, an energy transfer between two materials (donor and acceptor) could proceed via a radiative or nonradiative process.6 The former conventional radiative energy transfer (RET) is based on the emission of the photons followed by the reabsorption steps (i.e., in the UCNPBFO mixture system), where inevitably the energy-loss would take place inducing fairly low energy transfer efficiency (∼10%).41,42 To improve the transfer efficiency, a nonradiative process, i.e., LRET, is highly preferable.7 The LRET process exploits a nonradiative dipole−dipole interaction between two materials43 and can prevent the unexpected energy-loss thus providing a relatively high transfer efficiency (>50%).44 The key factor dominating the LRET process is the separation distance (r) of the two materials. According to Förster energy transfer theory, the rate (K) and efficiency (E) in an energy transfer process are given by the following equations: K = (1/τD)(R 0/r )6 8147
DOI: 10.1021/acsami.7b00158 ACS Appl. Mater. Interfaces 2017, 9, 8142−8150
Research Article
ACS Applied Materials & Interfaces
The red emission at 650 nm arises from the 4F9/2 state, which is populated either by nonradiative decay from the upper 2 H11/2/4S3/2 states or direct excitation from the 4I13/2 state following Yb3+ energy transfer (see Scheme 1b). Due to its band gap of 2.0 eV, the absorption of the BFO semiconductor covers the wavelength range below ∼630 nm. Therefore, BFO can also be excited by energy transfer from the 4 S3/2 and 2H11/2 excited states of the Er3+ ions, which work as an “additional” VIS light source. As indicated in Scheme 1b, upon the absorption of the VIS light, electrons are excited to the conduction band (CB) of BFO with holes left at the valence band (VB), followed by the migration of these electron−hole (e−/h+) pairs to the surface of BFO to participate in the degradation reactions. NIR photons are absorbed by the UCNP core and upconverted into VIS photons, which are subsequently transferred to the BFO shell by an efficient LRET process. Thus, BFO is further excited by this “internal” light source to generate more charges. The remaining holes on the VB of BFO surface can react with H2O to form hydroxyl radicals, which work as the main active species to perform the photocatalytic oxidation (degradation) of organic molecules,6−11,36 as evidenced by the scavenger test. Also, the separated electrons can react with surface adsorbed O2 to produce the anion radical, O2•‑, as shown by the scavenging test with Na2EDTA. Thus, O2•‑ may also take part in dye degradation, in particular of 4-CP, as O2•‑ radical anion is known to oxygenate chlorinated organo-halides.45 Moreover, an internal electric field resulting from the spontaneous polarization of BFO shell could also contribute to the separation of the photogenerated e−/h+ pairs, thereby further enhancing photocatalytic properties.23,46−48
Scheme 1. (a) Schematic Illustration of Spectral Overlap (Violet) of the Absorption of the BFO Shell and Emission of the UCNP Core and (b) Energy Level Diagrams of Yb3+ and Er3+ Ions as well as the Upconversion Luminescence Process upon NIR Excitation; the Energy Transfer from UCNP to BFO and the Charge Generation in BFO for Photocatalysis
■
CONCLUSIONS A hybrid nanostructure photocatalyst composed of a UCNP core and a BFO shell was synthesized for VIS-NIR light photocatalysis. Although BFO or UCNP alone do not exhibit any photocatalytic activity under NIR irradiation, the UCNP@ BFO nanostructure was found to be able to photodegrade pollutants such as CP or MO using NIR light exclusively (λ = 980 nm = 1.26 eV). Remarkably, reactive oxygen species such as hydroxyl radicals were found to be generated in the reaction medium, indicating that the photocatalyst is able to oxidize water in the presence of NIR light of 1.26 eV. Therefore, the UCNP@BFO photocatalyst is able to exploit light of low energy (1.26 eV) for high energy reactions (requiring more than 1.23 eV). This remarkable feature is ascribed to the combination of an upconversion mechanism and a nonradiative energy transfer between UCNP and BFO. Thus, the core@shell design of this catalyst is highly promising for the complete exploitation of the solar energy in photocatalysis.
E = R 0 6/(R 0 6 + r 6)
where R0 and τD are the Förster critical distance (
