Towards near-infrared photosensitization of tungsten trioxide nanostructured films by upconverting nanoparticles

Abstract

Upconverting materials are currently explored in the field of solar energy conversion in order to extend the light harvesting properties of semiconductors to the near-infrared (NIR) region, their absorption being generally limited to the visible region of the solar spectrum. Here, we propose to photosensitize nanostructured films of tungsten oxide (WO₃), a semiconductor widely investigated in photoelectrochemistry, photocatalysis and electrochromics, with NaGdF₄:Er³⁺, Yb³⁺ upconverting nanoparticles (UCNPs). In order to do so, we fabricate nanocomposite films of WO₃ and NaGdF₄:Er³⁺, Yb³⁺ UCNPs (indicated as UCNP/WO₃ films). Current–time measurements show that, under irradiation at λ = 980 nm, a relative increase in current of about 3% with respect to the dark current is observed in the UCNP/WO₃ films. The UCNP/WO₃ mol% ratio and the temperature of the thermal treatment of the nanocomposite films are both critical to simultaneously achieve photosensitization and charge carrier transport in the UCNP/WO₃ films.

---

Introduction

Photovoltaic (PV) solar energy conversion contributes to the fulfilment of the increasing world energy demands.¹ Different approaches have been investigated to overcome the Shockley–Queisser limit, which is the theoretical maximum efficiency attainable by a single-junction PV cell. For example, multi-junction cells are intended to match the solar spectrum by stacking materials with increasing bandgaps.² Approaches based on optical processes, such as two-photon absorption (simultaneous absorption of two photons by a material), downconversion (conversion of a high-energy photon to several lower-energy photons) or upconversion (conversion of several low-energy photons into one higher-energy photon) are also intensively pursued.^3–5

High surface area, mesoporous, transparent nanostructured films of metal oxides, such as TiO₂ and WO₃, have been widely investigated for solar energy conversion (photocatalysis and photoelectrochemistry) and energy conservation (electrochromism).^6–9 However, their large bandgaps (ca. 3.1 eV for TiO₂ and 2.5 eV for WO₃) render imperative the search for strategies to improve their solar light absorption properties, from the viewpoint of practical PV applications. In particular, the photoresponse of WO₃ extends to 500 nm, such that photons of lower energy, e.g. NIR photons, are not harvested to generate photocurrent, thus leading to considerable energy losses.^10,11 The use of upconverting (UC) materials represents a possible strategy to mitigate these losses since these materials absorb NIR photons and convert them to a frequency range that can then be absorbed by the metal oxide.

UC materials have been incorporated into different PV cell structures, including crystalline silicon,¹² amorphous silicon,¹³ dye-sensitized,^14,15 and organic¹⁶ solar cells. In addition, UC materials with different chemical compositions have been investigated, such as rare earth and transition metal ions,^17–20 semiconductor quantum dots,^21–23 and metallated macrocycles.^24–28 In UC materials, the excitation of the higher energy emitting states occurs via the sequential absorption of multiple NIR photons. In particular, in lanthanide-based UC materials, upconversion occurs via the sequential absorption of multiple NIR photons through real long-lived 4f excited electronic states. This is in stark contrast to conventional two-photon absorption (TPA) materials where two photons are absorbed simultaneously (or within less than a nanosecond), populating an excited state that is the sum of the energy of the incident photons.²⁰ Because of the extremely short lifetime of the (virtual) intermediate states, TPA can only be efficient with high-intensity photon fluxes (e.g., ultrafast laser radiation). On the other hand, the relatively long lifetime (μs to ms) of the 4f excited states of Ln³⁺ ions, such as Pr³⁺, Nd³⁺, Sm³⁺, Dy³⁺, Ho³⁺, Er³⁺, and Tm³⁺,²⁹ make them easily excited with low-power continuous wave NIR diodes^30,31 and suitable for up-converting lower intensity solar radiation. Fluoride-based crystals, in particular NaGdF₄, are commonly employed as host materials for lanthanide dopants, due to their low phonon energies, helping to limit non-radiative relaxation processes.³²

UC materials are often included in a solar energy conversion system in the form of an external layer, with different possible configurations, such as PV cell and rear converter or front converter and PV cell.^33,34 On the other hand, the use of UC nanoparticles (UCNPs) in intimate contact with nanostructured semiconductors has the potential to maximize the synergy between the two nanomaterials.

In this work, we report on the design, fabrication and characterization of nanocomposite films based on sol–gel prepared WO₃ and NaGdF₄:Er³⁺, Yb³⁺ UCNPs to improve the absorption properties of WO₃ nanostructured films in the NIR region of the solar spectrum. For the first time, UCNPs and pure metal oxides are used to produce nanocomposite films in which a photocurrent can be generated upon NIR irradiation. Er³⁺ is ideally suited for up-conversion since its electronic structure allows for many radiative transitions to occur (see Fig. S1†). Er³⁺ is often co-doped with Yb³⁺. Co-doping with the Yb³⁺ ion has been shown to improve the intensity of the up-conversion emission compared to singly doped systems since Yb³⁺ possesses only one excited state with a very high absorption cross-section in the NIR (ca. 980 nm), which leads to efficient energy transfer to the ⁴I_11/2 state of Er³⁺ also located at approximately 980 nm. Our solution to use NaGdF₄:Er³⁺, Yb³⁺ UCNPs to photosensitize WO₃ is based on the fact that, when excited at λ_exc = 980 nm, the UCNPs have two emission bands in the visible range, partly overlapping with the absorption spectrum of WO₃,^20,35 thus leading to the possibility to use them as photosensitizers. The measure of the photocurrent generated by UCNP/WO₃ nanocomposite films in a planar configuration under NIR irradiation (λ_exc = 980 nm) confirmed the interest of our strategy. Atomic Force Microscopy (AFM), X-ray diffraction (XRD) and photoluminescence (PL) studies were conducted to identify the fabrication conditions to simultaneously achieve good photosensitization and charge carrier transport properties in the UCNP/WO₃ films. Our strategy paves the way towards the NIR photosensitization of metal oxide semiconductors for PV applications by simple solution-based processing.

Experimental section

1. Microfabrication of the electrodes

ITO electrodes were patterned by photolithography. The ITO-coated glass substrates (Colorado Concept Coatings, 30 mm × 30 mm, <15 Ω square⁻¹ sheet resistance) were first sequentially cleaned in an ultrasonic bath in isopropyl alcohol (IPA), acetone, and IPA for 15 min each, and N₂-dried. Then, prior to the microfabrication steps, the substrates were exposed to a UV-ozone treatment for 15 min. After selective exposure and development of a positive-tone photoresist layer, the ITO film was chemically etched with hydrochloric acid (HCl : H₂O 65 v/v%) and the unexposed photoresist was removed with acetone. This process resulted in ITO electrodes with an interelectrode spacing, L, of 100 μm and a width, W, of 6000 μm.

2. Materials and film fabrication

The WO₃ films were prepared following a sol–gel synthesis process already reported in the literature.¹⁰ Na₂WO₄, PEG-200, ethanol and the proton-exchange resin (Dowex® 50WX2) were purchased from Sigma Aldrich. NaGdF₄:Er³⁺, Yb³⁺ (2 mol% Er³⁺, 20 mol% Yb³⁺) UCNPs were prepared via a thermal decomposition process also reported in the literature.³⁵ First, lanthanide trifluoroacetates used as precursors in the synthesis were prepared from lanthanide oxides (Alpha Aesar, Gd₂O₃, Er₂O₃, Yb₂O₃, >99.99%) in deionized water and trifluoroacetic acid (Alpha Aesar, 99%).³⁶ In the standard process, the obtained precursors, together with the required stoichiometric quantities of sodium trifluoroacetate (Sigma-Aldrich, 98%), were injected at a rate of 1.0 mL min⁻¹ in a 1 : 1 solution of oleic acid (Alpha Aesar, 90%) and octadecene (Alpha Aesar, 90%) at 315 °C, and aged for 1 h. The prepared β-phase NaGdF₄:Er³⁺, Yb³⁺ UCNPs were then washed by centrifugation and redispersed in hexane. The LiYF₄:Tm³⁺, Yb³⁺ UCNPs were synthesized using a similar technique with slight modifications (mainly, the precursors weren't injected. Instead, they were mixed with oleic acid and octadecene from the very beginning).³⁷ In this case, the oxides Y₂O₃, Tm₂O₃ were used to prepare the trifluoroacetates precursors and lithium trifluoroacetate was used in the reaction that was done at 325 °C. The synthesized UCNPs were capped with long-chained oleate molecules making them hydrophobic. To render them hydrophilic and to avoid any possible effect of the capping agent in the final material, the oleates were removed by following a process described in literature.³⁸ NaGdF₄:Er³⁺, Yb³⁺ UCNPs had a narrow size distribution, with a diameter of 45 nm (see Fig. 1a), and showed a hexagonal-nanoplate shape typical of the β-phase of NaGdF₄.

Fig. 1 (a) TEM image of β-NaGdF₄:Er³⁺, Yb³⁺ UCNPs. (b) Experimental configuration adopted in this work for the electrical characterization (photocurrent vs. time). Topographical AFM images of (c) WO₃ and (d) 8 mol% UCNP/WO₃ films, deposited on glass and treated at 385 °C. (e) PL spectrum of β-NaGdF₄:Er³⁺, Yb³⁺ UCNPs drop casted on glass and dried at room temperature (dashed line, left y axis) or thermally treated at 385 °C (full line, left y axis) and optical absorption spectrum of a WO₃ film (∼1.2 μm in thickness), deposited on glass and thermally treated at 385 °C (right y axis).

To fabricate the nanocomposite films, a mixture containing 1% weight/weight concentration of UCNPs suspension in water and ca. 0.3 M tungstic acid sol was prepared to reach different mol% UCNP/WO₃ (6, 8, and 16 mol%). The mixture was aged for about 1 h under continuous stirring before film deposition. Nanocomposite films were obtained by drop-casting 10 μL UCNP/WO₃ mixture (or pure WO₃ sol for control films) on ITO-patterned substrates, followed by a thermal treatment at 385 °C, in O₂ (160 scc min⁻¹), for 30 min.

3. Characterization techniques

Photoluminescence measurements were carried out with a Prism And Reflector Imaging Spectroscopy System (PARISS, purchased from LightForm Inc.) in ambient air conditions after excitation at λ_exc = 980 nm from a cw NIR laser diode (power density of 1 W cm⁻²). The specular and diffuse transmission/reflection properties of the WO₃ films were assessed using a PerkinElmer Lambda 1050 UV/Vis/NIR spectrophotometer equipped with a Labsphere integration sphere. The use of an integration sphere allows one to take into account light diffusion which could potentially be wrongly associated with losses due to absorption in purely specular-type measurements. The absorption (A) was calculated using the following equation: A = 100% − T − R − D where T, R and D are the specular transmission, specular reflection and total diffusion respectively. The microstructure of the samples was characterized by XRD (Bruker D-8 Advance Diffractometer), using CuKα radiation. High-resolution transmission electron microscopy (TEM, Philips CM200) was used to ascertain the phase, morphology, crystallinity and size distribution of the UCNPs. Atomic force microscopy (AFM) images were obtained in ambient air conditions using a Veeco Dimension 3100 Digital Instruments with a Si cantilever (tip radius < 10 nm, spring constant 40 N m⁻¹) at a scan rate of 1 Hz. Photocurrent–time measurements were carried out in a planar configuration (see Fig. 1b), in ambient air conditions, using a B2902A Agilent Source Measure Unit (SMU). Measurements under simulated solar irradiation (AM1.5G) were performed using a VeraSol-2 LED solar simulator (Newport).

Results and discussion

1. Film fabrication and characterization

To achieve the successful combination of UCNPs and nanostructured WO₃ in a nanocomposite film, a number of challenges had to be overcome.

First, since UCNPs are electrically insulating, it was imperative to identify the optimal UCNP/WO₃ mol% ratio to photosensitize WO₃ without hindering charge transport in the nanocomposite films. In order to do so, we prepared films with different UCNP/WO₃ mol% ratios. Films prepared at ratios above 8 mol% were discontinuous, i.e. not easily amenable for incorporation into devices (see Fig. S2†). No significant effect of the presence of UCNPs on the photocurrent was observed in films with ratios below 8 mol% (see Fig. S3†). The morphology of 8 mol% UCNP/WO₃ films (Fig. 1d) was similar to that of WO₃ control films (Fig. 1c). 8 mol% UCNP/WO₃ films, similarly to the case of WO₃ control films, showed complete substrate surface coverage. Furthermore, 8 mol% UCNP/WO₃ films showed no evidence of phase segregation and values of surface roughness (R_RMS) comparable to those of WO₃ films (0.8 nm vs. 0.9 nm). Therefore, overall, 8 mol% UCNP/WO₃ films showed to be of device quality.

Secondly, we had to identify the optimal temperature for the thermal treatment of the nanocomposite films to ensure good charge transport, while keeping a high UCNPs emission, essential for photosensitization. Achieving good transport properties in nanostructured WO₃ films synthesized by sol–gel methods usually requires thermal treatment temperatures within the range of 350–600 °C.^39,40 However, such high temperatures would affect the crystalline structure of UCNPs. Indeed, a change from the hexagonal phase (β-UCNPs) to the cubic phase (α-UCNPs) is expected to take place at temperatures as low as 400 °C.⁴¹ Most importantly, the upconverting process is reported to be one order of magnitude more efficient in β-phase NaGdF₄:Er³⁺, Yb³⁺ UCNPs than in α-UCNPs, for the nanoparticle size used here.⁴² To avoid, at least in part, this phase transformation, we treated the UCNP/WO₃ films at 385 °C. The comparison of the PL spectrum of 45 nm-sized NaGdF₄:Er³⁺, Yb³⁺ β-UCNPs treated at 385 °C with the optical absorption spectrum of ∼1.2 μm-thick WO₃ films treated at 385 °C shows the partial overlap of the WO₃ absorption and the UCNPs emission spectra (see Fig. 1e). Such an overlap represents an interesting opportunity to photosensitize WO₃ films in the NIR region and constitutes the foundation of our strategy. In addition, two emission bands located at ca. 550 nm and ca. 660 nm are observable upon NIR laser excitation (λ_exc = 980 nm, power density of 1 W cm⁻²) in the PL spectra of the UCNPs, both after room temperature drying and 385 °C thermal treatment (Fig. 1e). The green emission band located at ca. 550 nm is attributed to transitions from the ²H_11/2 and ⁴S_3/2 excited states to the ⁴I_15/2 ground state of Er³⁺ ions (see Fig. S1†). The red emission band observed at ca. 660 nm is attributed to the transition from the ⁴F_9/2 excited state to the ⁴I_15/2 ground state.²²Fig. 1e shows that the thermal treatment affects the emission properties of the UCNPs such that the green : red intensity ratio is about 1 when the UCNPs are dried at room temperature, whereas this ratio is ∼0.3 after treatment at 385 °C.

XRD measurements were then performed to gain insight on the dependence of the PL properties on the temperature of thermal treatment (see Fig. 2). When dried at room temperature (RT) on glass, the XRD peaks of NaGdF₄:Er³⁺, Yb³⁺ UCNPs (pattern 1) match the filed pattern of NaGdF₄ with a hexagonal crystal phase (β-phase).^43,44 When the UCNPs are thermally treated at 385 °C (pattern 2), UCNPs are likely to remain in the β-phase, as suggested by the presence of the peaks at 30°, 43°, and 52°. The positions of the peaks of the cubic phase (α-UCNPs) are also indicated in Fig. 2, to validate the hypothesis that UCNPs are not in the α-phase after the thermal treatment. The comparison between pattern 1 and 2 show that the relative intensity of the peaks changes from pattern 1 to pattern 2 thus suggesting that the UCNPs reorient during thermal treatment. This reorientation can account for the observed differences in the upconverted red and green emissions since NaGdF₄ is not isotropic.⁴⁵ The presence of O₂ during the thermal treatment has no noticeable effect on the structure of the UCNPs.

Fig. 2 XRD patterns of β-UCNPs, 45 nm in diameter, drop casted on glass. After room temperature drying and a thermal treatment at 385 °C, the UCNPs are still in the β-phase (hexagonal phase, patterns 1 and 2). For reference purposes, the peak positions for the α and β phases are included at the bottom of the figure.³¹

2. Electrical measurements

Current–time measurements were performed using a two-electrode, planar configuration to assess the possibility to generate a photocurrent in UCNP/WO₃ nanocomposite films (∼1 μm thick) exposed to NIR light (XRD pattern in Fig. S4†). UCNP/WO₃ nanocomposite films were systematically compared with WO₃ control film samples. During the electrical measurements, the films were initially kept in ambient light conditions for 300 s, followed by 300 s in total darkness, under a constant electrical bias (0.5 V, 0.7 V, or 1.0 V), to allow the dark current to stabilize prior to photocurrent measurements. Afterwards, the samples were illuminated by a cw NIR laser (λ_exc = 980 nm, power density of 1 W cm⁻²). Typically, during the electrical measurements the films were exposed for 30 s to light, followed by 60 s in the dark. To describe the effect of the illumination on the current measured, we used the “on/off ratio”, R_on/off, calculated as follows:

R_on/off = 100 × (i_on − i_off)/i_off

where i_on is the current when the NIR light illuminated the sample (on current) and i_off is the current in the dark (off current). When 8 mol% UCNP/WO₃ films were exposed to NIR light, an increase of the current was observed (Fig. 3a and S5†). A value of R_on/off ranging between 3.4 to 3.9% was obtained when the NIR light illuminated the nanocomposite films, depending of the applied electrical bias. This increase in the current is attributable to the absorption of the NIR light by the UCNPs followed by the harvesting of the upconverted light by the WO₃, in turn followed by the generation of a photocurrent.

Fig. 3 Current–time measurements under NIR chopped light (60 s in the dark and 30 s under irradiation, λ_exc = 980 nm, power density of 1 W cm⁻²) for films deposited on patterned ITO with an interelectrode distance of 100 μm and thermally treated at 385 °C made of (a) 8 mol% UCNP/WO₃ and (b) WO₃. Samples were kept under constant electrical bias for 300 s in ambient light conditions, followed by 300 s in the dark, before chopping the light.

To evaluate the possibility that thermal effects due to laser exposure are responsible for the current increase observed with 8 mol% UCNP/WO₃ films exposed to NIR light, we exposed WO₃ control films to the NIR laser: no noticeable current increase was observed in this case (see Fig. 3b). These results suggest that the increase in the current measured from 8 mol% UCNP/WO₃ nanocomposite films under NIR exposure are most probably the result of a photocurrent induced by the presence of UCNPs, that thermal effects likely play a minor role.

To further demonstrate the light harvesting properties of our films under more realistic lighting conditions, the samples were also exposed to simulated solar irradiation (AM1.5G); a photocurrent was also observed (see Fig. S6†) thus confirming that the NIR photosensitization process does not prevent the nanocomposite films to generate photocurrent under simulated solar light, as expected.

The hypothesis that thermal effects play a minor role in the photoresponse of thin films of UCNP/WO₃ is supported by results observed with blue-emitting UCNPs, such as LiYF₄:Tm³⁺, Yb³⁺. Here an improved overlap between the absorption spectrum of WO₃ and the emission spectrum of UCNPs with respect to the case of NaGdF₄:Er³⁺, Yb³ is paralleled by the higher values of R_on/off observed (ca. 6%) (see Fig. S7†).

Conclusions and perspectives

In conclusion, we have demonstrated an interesting strategy for the photosensitization in the NIR of metal oxide WO₃ films based on the addition of UCNPs. Specifically we observed an increase in the current measured in NaGdF₄:Er³⁺, Yb³⁺/WO₃ nanocomposite films under NIR irradiation (λ_exc = 980 nm), with respect to dark conditions. After a judicious choice of the mol% ratio in the nanocomposite film and the thermal treatment temperature, we demonstrated that nanocomposite films could be engineered so as not to hinder the transport properties in the nanocomposite films while also extending their photosensitivity to NIR light. The approach herein presented is promising for solar energy conversion applications, as it permits to prepare, by an easy solution-based technique, high surface area, nanocomposite films photosensitive to visible and NIR light.

In perspective, there remain several challenges which need to be addressed in order to optimize the use of lanthanide UCNPs as metal oxide photosensitizers in PV cells. Indeed, considering that the absorption bandwidth of the lanthanide-based UCNPs is fairly narrow, the combination of several upconverters within the metal oxide matrix has to be explored. Furthermore, increasing the upconversion quantum yield of UCNPs, which is presently relatively low,^41,46 would also lead to performance improvements. In fact, recent reports have proposed approaches to develop systems with increased yield.^47,48 Implementing the above-mentioned considerations has the potential to generate significant improvement in the power conversion efficiency of metal oxide-based PV cells in the NIR.

Acknowledgements

The authors are grateful to J. Lerner (LightForm Inc) and P. Moraille for fruitful discussions and to Y. Drolet for his technical support. Authors are grateful to Prof. G. Hanan and Dr D. Chartrand for the access to the solar simulator system. This work was financially supported by: NSERC (Discovery grants C.S. and F.V.), FRQNT Team Grant (C.S. and F.V.), MDEIE (F.V.). F. Venne acknowledges the financial support provided by CMC Microsystems (MNT program) and M. Quintanilla acknowledges the financial support of Fundacion Ramon Areces.

Notes and references

1. T. M. Razykov, C. S. Ferekides, D. Morel, E. Stefanakos and H. S. Ullal, Sol. Energy, 2011, 85, 1580–1608.
2. D. J. Friedman, Curr. Opin. Solid State Mater. Sci., 2010, 14, 131–138.
3. X. Huang, S. Han, W. Huang and X. Liu, Chem. Soc. Rev., 2013, 42, 173–201.
4. R. Naccache, F. Vetrone and J. A. Capobianco, ChemSusChem, 2013, 6, 1308–1311.
5. J. C. Goldschmidt and S. Fischer, Adv. Opt. Mater., 2015, 3, 510–535.
6. S. K. Deb, Sol. Energy Mater. Sol. Cells, 2008, 92, 245–258.
7. S. H. Baeck, K. S. Choi, T. F. Jaramillo, G. D. Stucky and E. W. McFarland, Adv. Mater., 2003, 15, 1269–1273.
8. H. Zheng, Y. Tachibana and K. Kalantar-zadeh, Langmuir, 2010, 26, 19148–19152.
9. I. Concina and A. Vomiero, Small, 2015, 11, 1744–1774.
10. C. Santato, M. Odziemkowski, M. Ulmann and J. Augustynski, J. Am. Chem. Soc., 2001, 123, 10639–10649.
11. P. Ramasamy, M. Palanisamy and J. Kim, RSC Adv., 2014, 4, 34873–34895.
12. A. Shalav, B. S. Richards, T. Trupke, K. W. Krämer and H. U. Güdel, Appl. Phys. Lett., 2005, 86, 4–7.
13. J. de Wild, J. K. Rath, A. Meijerink, W. G. J. H. M. van Sark and R. E. I. Schropp, Sol. Energy Mater. Sol. Cells, 2010, 94, 2395–2398.
14. G.-B. Shan and G. P. Demopoulos, Adv. Mater., 2010, 22, 4373–4377.
15. C. Yuan, G. Chen, P. N. Prasad, T. Y. Ohulchanskyy, Z. Ning, H. Tian, L. Sun and H. Ågren, J. Mater. Chem., 2012, 22, 16709–16713.
16. H.-Q. Wang, M. Batentschuk, A. Osvet, L. Pinna and C. J. Brabec, Adv. Mater., 2011, 23, 2675–2680.
17. M. Pollnau, D. R. Gamelin, S. R. Lüthi and H. U. Güdel, Phys. Rev. B: Condens. Matter Mater. Phys., 2000, 61, 3337–3346.
18. S. Heer, K. Kömpe, H.-U. Güdel and M. Haase, Adv. Mater., 2004, 16, 2102–2105.
19. J. C. Boyer, L. A. Cuccia and J. A. Capobianco, Nano Lett., 2007, 7, 847–852.
20. F. Auzel, Chem. Rev., 2004, 104, 139–173.
21. X. Wang, W. Yu, J. Zhang, J. Aldana, X. Peng and M. Xiao, Phys. Rev. B: Condens. Matter Mater. Phys., 2003, 68, 1–6.
22. C. Bonati, A. Cannizzo, D. Tonti, A. Tortschanoff, F. van Mourik and M. Chergui, Phys. Rev. B: Condens. Matter Mater. Phys., 2007, 76, 1–4.
23. E. Poles, D. C. Selmarten, O. I. Mićić and A. J. Nozik, Appl. Phys. Lett., 1999, 75, 971.
24. R. R. Islangulov, J. Lott, C. Weder and F. N. Castellano, J. Am. Chem. Soc., 2007, 129, 12652–12653.
25. T. N. Singh-Rachford and F. N. Castellano, Coord. Chem. Rev., 2010, 254, 2560–2573.
26. V. Yakutkin, S. Aleshchenkov, S. Chernov, T. Miteva, G. Nelles, A. Cheprakov and S. Baluschev, Chem.–Eur. J., 2008, 14, 9846–9850.
27. S. Baluschev, V. Yakutkin, G. Wegner, T. Miteva, G. Nelles, A. Yasuda, S. Chernov, S. Aleshchenkov and A. Cheprakov, Appl. Phys. Lett., 2007, 90, 1–4.
28. S. Baluschev, T. Miteva, V. Yakutkin, G. Nelles, A. Yasuda and G. Wegner, Phys. Rev. Lett., 2006, 97, 7–9.
29. F. Vetrone, J.-C. Boyer and J. A. Capobianco, Encycl. Nanosci. Nanotechnol., 2004, 10, 725–765.
30. J. Yang, L. Zhang, L. Wen, S. Dai, L. Hu and Z. Jiang, J. Appl. Phys., 2004, 95, 3020.
31. F. Vetrone, J.-C. Boyer, J. A. Capobianco, A. Speghini and M. Bettinelli, J. Phys. Chem. B, 2003, 107, 10747–10752.
32. J. F. Suyver, J. Grimm, M. K. van Veen, D. Biner, K. W. Krämer and H. U. Güdel, J. Lumin., 2006, 117, 1–12.
33. L. Li, Y. Yang, R. Fan, Y. Jiang, L. Wei, Y. Shi, J. Yu, S. Chen, P. Wang, B. Yang and W. Cao, J. Power Sources, 2014, 264, 254–261.
34. T. Trupke, M. A. Green and P. Würfel, J. Appl. Phys., 2002, 92, 4117–4122.
35. F. Vetrone, R. Naccache, V. Mahalingam, C. G. Morgan and J. A. Capobianco, Adv. Funct. Mater., 2009, 19, 2924–2929.
36. C. Rüssel, J. Non-Cryst. Solids, 1993, 152, 161–166.
37. T. Cheng, R. F. Ortiz, K. Vedantham, F. Vetrone, R. S. Marks, T. W. J. Steele and B. Sheva, Biomacromolecules, 2015, 16, 364–373.
38. N. Bogdan, F. Vetrone, G. A. Ozin and J. A. Capobianco, Nano Lett., 2011, 11, 835–840.
39. B. Yang, Y. Zhang, E. Drabarek, P. R. F. Barnes and V. Luca, Chem. Mater., 2007, 19, 5664–5672.
40. H. Zheng, J. Z. Ou, M. S. Strano, R. B. Kaner, A. Mitchell and K. Kalantar-zadeh, Adv. Funct. Mater., 2011, 21, 2175–2196.
41. N. C. Dyck, F. C. J. M. van Veggel and G. P. Demopoulos, ACS Appl. Mater. Interfaces, 2013, 5, 11661–11667.
42. A. Aebischer, S. Heer, D. Biner, K. Krämer, M. Haase and H. U. Güdel, Chem. Phys. Lett., 2005, 407, 124–128.
43. G. D. Brunton, H. Insley, T. N. McVay and R. E. Thoma, Oak Ridge Natl. Lab. Rep. ORNL (U.S.), 1965.
44. The peak (0 0 2) has been deduced using the Bragg diffraction law. It was not indexed from 43.
45. P. Chen, M. Song, E. Wu, B. Wu, J. Zhou, H. Zeng, X. Liu and J. Qiu, Nanoscale, 2015, 7, 6462–6466.
46. J.-C. Boyer and F. C. J. M. van Veggel, Nanoscale, 2010, 2, 1417–1419.
47. W. Zou, C. Visser, J. A. Maduro, M. S. Pshenichnikov and J. C. Hummelen, Nat. Photonics, 2012, 6, 2–6.
48. M. D. Wisser, M. Chea, Y. Lin, D. M. Wu, W. L. Mao, A. Salleo and J. A. Dionne, Nano Lett., 2015, 15, 1891–1897.

---

Footnote

† Electronic supplementary information (ESI) available: Energy levels diagram of UCNPs; optical images of 16 mol% UCNP/WO₃ thin films; current–time curves of 6% UCNP/WO₃ thin films; XRD measurements of WO₃ and UCNP/WO₃ thin films; current–time curves of 8% UCNP/WO₃ thin films under NIR and simulated solar light; spectroscopic and current–time measurements on LiYF₄:Tm³⁺, Yb³⁺/WO₃ thin films. See DOI: 10.1039/c5ra18320k

---