Assembly of β-Cu2V2O7/WO3 heterostructured nanocomposites and the impact of their composition on structure and photoelectrochemical properties

Mariateresa Scarongella a, Chethana Gadiyar a, Michal Strach a, Luca Rimoldi b, Anna Loiudice a and Raffaella Buonsanti *a
aLaboratory of Nanochemistry for Energy, Institute of Chemical Sciences and Engineering, EPFL, Sion, 1050, Switzerland. E-mail:
bDepartment of Chemistry, Università degli Studi di Milano, 20133 Milan, Italy

Received 12th June 2018 , Accepted 11th September 2018

First published on 20th September 2018

Multinary metal oxides and their heterostructures play a key role as light absorbers in the production of solar chemicals. Synthetic tunability is crucial to understand the impact of composition and structure on the photoelectrochemical performance. Here, we assemble β-Cu2V2O7/WO3 heterostructured nanocomposites using a novel seeded-growth approach which allows an unprecedented compositional tunability. A 10 fold increase in the net photocurrent density towards sulfite oxidation was measured for the nanocomposite with the lowest loading of WO3 (β-Cu2V2O7[thin space (1/6-em)]:[thin space (1/6-em)]WO3 = 1[thin space (1/6-em)]:[thin space (1/6-em)]0.1) as compared to the bare β-Cu2V2O7 counterpart. This improvement is attributed to the formation of an intimate junction between the two metal oxides which favors charge transfer and separation. An increase in the WO3 content results in the formation of macroscopic phase segregated domains which reduce these interfacial areas, thus degrading the phototoelectrochemical performance of the nanocomposites. While highlighting the effectiveness of heterostructuring and the importance of compositional tunability, this study points at the emerging need of techniques to control and to probe the intrinsic inhomogeneity of these complex inorganic heterojunctions.


Heterostructured nanocomposites are attractive material platforms where the interface between the components plays a key role in conferring improved functionalities compared to those of the isolated counterparts. In solar energy conversion, organic bulk-heterojunction solar cells exemplify the importance of interfaces for enhanced photoconversion efficiencies as the donor/acceptor blend overcomes the charge transport limitations of light absorbing polymers.1–6 The impact of the morphology of the organic blends on charge carrier separation and dynamics has been widely investigated, and has been shown to be crucial for performance optimization.1–6 Photoelectrochemical cells (PEC) are a promising technology to convert sunlight into renewable chemicals (hydrogen from water or hydrocarbons from CO2).7 At the present, these devices suffer from the lack of optimal n-type inorganic semiconductors to carry out the water oxidation reaction, referred to as photoanodes. Metal oxides (MOs) are typically chosen because of their higher stability in alkaline aqueous electrolytes compared to silicon, III–V and II–VI semiconductors. So far the most studied photoanodes are α-Fe2O3 and BiVO4.8–10 One of the common limitations of MOs is the incompatibility between light absorption capabilities and carrier transport length.8–10 The combination of smaller band gap and larger band gap MO semiconductors has been demonstrated to enhance charge separation and thus water oxidation photocurrents (e.g. BiVO4/WO3,11–14 BiVO4/TiO2,15–18 α-Fe2O3/WO3,19 α-Fe2O3/TiO220,21). Yet, the detailed mechanisms of charge separation at the semiconductor/semiconductor interface are not often fully elucidated and the understanding of how the structure/composition of the MO nanocomposites correlates with their photoelectrochemical properties is at its infancy when compared to organic heterojunctions.1–6 In the case of BiVO4, the combination of doping, nanostructuring and heterostructuring have pushed the photoelectrochemical performance very close to the theoretical limit.22,23 Nevertheless, the band gap of BiVO4 of around 2.5 eV is still too wide for optimal solar harvesting. Recently, copper vanadates with their band gaps around 2 eV have emerged as a novel class of photoanode materials.24–34 The two major limitations of copper vanadates for solar water oxidation are the position of their conduction band minimum (which is a few hundreds of millivolts more positive than the water reduction potential) and their short carrier-diffusion lengths (around 20–40 nm).25,33 We have recently demonstrated that some photocurrent increase can be achieved by nanostructuring of β-Cu2V2O7 or by creating the β-Cu2V2O7/CuV2O6 heterojunction, which supposedly favors charge separation.32

In this work, first we tailor-make β-Cu2V2O7/WO3 (CVO/WO3) heterostructured nanocomposites in a wide compositional range by exploiting our recently developed nanocrystal-seeded method for nanocrystalline β-Cu2V2O7.32 While it is difficult to make conclusive remarks based solely on the band alignment of bulk materials and the actual interfacial energetics need to be carefully probed, we choose WO3 as the wide band gap MO in the heterojunction because of its more favorable energetic alignment for charge transfer from β-Cu2V2O7 compared to TiO2.11,14 The morphology is selected based on our previous studies.11 After assembling the CVO/WO3 nanocomposites, we study the impact of the different compositions on the photoelectrochemical performance of the nanocomposites through photocurrent measurements, and compare those to the pure counterparts. Finally, the investigation of the charge carrier dynamics of the best performing nanocomposite by means of transient absorption spectroscopy enables us to suggest the possible mechanism behind the observed behavior.

Results and discussion

For the assembly of the CVO/WO3 heterostructured nanocomposites, a solution containing 6 nm surface oxidized Cu nanocrystals (NCs), VO(acac)2 and 4 nm × 15 nm WO2.7 nanorods (NRs) in dimethylformamide was drop-cast on a substrate and annealed at 350 °C for 8 h. These synthesis parameters were chosen based on our previous results for pure nanocrystalline CVO.32 Nanocomposites with three different mass ratios of the two components CVO[thin space (1/6-em)]:[thin space (1/6-em)]WO3 = 1[thin space (1/6-em)]:[thin space (1/6-em)]0.25, 1[thin space (1/6-em)]:[thin space (1/6-em)]0.5 and 1[thin space (1/6-em)]:[thin space (1/6-em)]1 were prepared by adding different amounts of WO2.7 NRs in the precursor solution. Fig. 1A and B show the transmission electron microscopy (TEM) images of the Cu NCs and WO2.72 NRs used as building blocks for the assembly of the nanocomposites. A representative top-view scanning electron microscopy (SEM) image and the corresponding compositional mapping by energy dispersive X-ray (EDX) spectroscopy of one of the CVO/WO3 nanocomposites (mass ratio 1[thin space (1/6-em)]:[thin space (1/6-em)]0.25) are reported in Fig. 1C and D, respectively. SEM reveals a continuous film composed of a network of nanograins. The limited resolution does not allow one to distinguish between copper vanadate and WO3, yet the EDX mapping evidences a spatially homogeneous distribution of Cu, V, and W both in plane (Fig. 1D) and in cross-section (Fig. S1, ESI). Similar results were obtained for the nanocomposites with different compositions (Fig. S2, ESI) and elemental analysis confirmed the CVO[thin space (1/6-em)]:[thin space (1/6-em)]WO3 ratio (Table S1, ESI).
image file: c8tc02888e-f1.tif
Fig. 1 (A) TEM image of the 6 nm Cu NC seeds, (B) TEM image of the 4 nm × 15 nm WO2.72 NRs, (C) top-down SEM image and (D) EDX mapping of a typical CVO/WO3 nanocomposite film with mass ratio 1[thin space (1/6-em)]:[thin space (1/6-em)]0.25.

To get more insights into the CVO/WO3 interface of the synthesized nanocomposite, transmission electron microscopy (TEM) was employed (Fig. 2 and Fig. S3, S4, ESI). Fig. 2A shows the high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) image of one of the nanocomposites. The stark difference in contrast and in morphology between β-Cu2V2O7 (irregular shaped areas) and WO3 (rod shaped areas) clearly identifies the two different domains in the nanocomposite. EDX-TEM analysis in Fig. S3 (ESI) provides evidence of the uniform coverage of WO3 nanorods by β-Cu2V2O7. The interface between the two materials in the nanocomposite was further investigated by high resolution TEM (HR-TEM). Fig. 2B and C evidence a continuous interface between the two domains, which might suggest a hetero-epitaxial growth of β-Cu2V2O7 on WO3 along the a crystallographic direction (lattice spacing 7.687 Å) of the monoclinic β-Cu2V2O7 structure and one of the monoclinic WO3 directions (a = 7.3271(2) Å, b = 7.5644(2) Å, c = 7.7274(3) Å, α = 90°, β = 90.488(3)°, γ = 90°).

image file: c8tc02888e-f2.tif
Fig. 2 (A) HAADF-STEM image of a typical CVO/WO3 nanocomposite film with ratio 1[thin space (1/6-em)]:[thin space (1/6-em)]0.25, indicating with the yellow arrows the β-Cu2V2O7 and WO3 nanorod domains. (B) HR-TEM image of the same nanocomposite evidencing the interface zone. (C) Zoom of (B) showing a continuous interface between β-Cu2V2O7 and the WO3 nanorod.

The structural and compositional characterization of the nanocomposites was completed by X-ray diffraction (XRD) (Fig. 3) and Raman spectroscopy (Fig. S5, ESI). The XRD data were acquired in grazing-incident mode using synchrotron radiation. Fig. 3A reports the XRD data of the nanocomposites along with those of the pure components for comparison. The broadening of the peaks in all patterns is consistent with the formation of nano-sized crystalline domains and the patterns are in agreement with the previously reported data for the WO3 NR thin film and for the nanocrystalline CVO.11,32,35 The main reflections of the WO3 phase are observed in all composite samples with the intensity increasing proportionally to the amount of WO3. The preservation of the peak width and relative peak ratio for WO3 in the nanocomposites indicates that the NR size and shape are maintained upon annealing at 350 °C for 8 hours, which is consistent with the TEM data (Fig. S4, ESI). This observation is not surprising, considering that we did not observe any morphological change even in our previous work when the same NRs were annealed at a much higher temperature.11 The CVO peaks are distinguishable in the 1[thin space (1/6-em)]:[thin space (1/6-em)]0.25 and 1[thin space (1/6-em)]:[thin space (1/6-em)]0.5 samples; instead in the 1[thin space (1/6-em)]:[thin space (1/6-em)]1 nanocomposite they are completely covered by the more strongly scattering WO3. XRD patterns were also acquired while heating the deposited precursor solution in order to get deeper insights into the formation of the nanocomposites (Fig. 3B). The data relative to the CVO/WO3 sample with mass ratio 1[thin space (1/6-em)]:[thin space (1/6-em)]0.25 are reported as one representative example, taking into account that all the nanocomposites behaved similarly (Fig. S6, ESI). In the initial scans at room temperature, the most intense peaks of the WO2.72 phase together with those of the vanadium molecular precursor are detected. Cu reflections are not observed due to peak broadness and low intensity associated with the small size of the NCs (6 nm). During the heating, the sharp reflections of VO(acac)2 disappear at around 150 °C, which corresponds to the melting temperature of the salt. Thermal dilatation of the WO2.72 lattice is detected until about 230 °C, when the main (010) peak at 10.4° starts shifting at higher angles, which is indicative of a WO2.72 → δ-WO3 (triclinic P[1 with combining macron]) phase transition.36 At 270 °C the CVO reflections become evident, as the Cu NCs and the VOx species start reacting. At this temperature we expect the γ-WO3 (monoclinic P21/n) to be the stable polymorph, but the quality of the data does not allow us to discern this phase transition. The reflections grow in intensity until the temperature reaches 350 °C, indicating improved crystallinity, and then remain constant over time. The same behavior is observed in the binary mixture of surface oxidized Cu NCs and VO(acac)2 (Fig. S7, ESI) indicating that the presence of NRs does not impact significantly the kinetics of the chemical transformations leading to the formation of CVO. The reactivity of the binary mixtures WO2.72 NRs + VO(acac)2 and WO2.72 NRs + surface oxidized Cu NCs was also investigated by in situ experiments (Fig. S7, ESI). Significant structural and compositional changes were observed only above 400 °C, thus confirming that in the synthesis conditions chosen for the nanocomposites bulk doping or alloying of WO3 by Cu- or V- and of CVO by W- can be most likely excluded.

image file: c8tc02888e-f3.tif
Fig. 3 (A) XRD patterns of the CVO/WO3 nanocomposites, obtained by annealing at 350 °C for 8 hours the precursor mixture containing surface oxidized Cu NCs, VO(acac)2 and different amounts of WO2.72 NRs, along with the pattern of a pure WO3 thin film, obtained by annealing at 350 °C for 8 hours the as-synthesized WO2.72 NRs, and of a pure CVO, obtained by annealing at 350 °C for 8 hours the 6 nm surface oxidized Cu NCs and VO(acac)2. Reference patterns for WO3 and β-Cu2V2O7 are reported at the bottom. (B) In situ XRD measurements during the formation of the CVO/WO3 nanocomposites with mass ratio 1[thin space (1/6-em)]:[thin space (1/6-em)]0.25. These data were acquired with synchrotron radiation at a wavelength of 0.6897 Å.

The photoelectrochemical (PEC) performance of the nanocomposites were evaluated by measuring the photocurrent density (J) of the films versus the applied potential (E) in a borate buffer electrolyte (pH = 8.2) with Na2SO3 as the hole scavenger (Fig. 4A). The factors contributing to the photocurrent density are light absorption, charge transport within the photoelectrode and charge transfer at the semiconductor/electrolyte interface. The slow water oxidation kinetics on the CVO as well as on WO3 electrodes lead to accumulation of holes at the semiconductor/electrolyte interface.25,29,37 Hence, in order to mitigate this issue, a hole scavenger is utilized to accelerate the charge transfer at the semiconductor/electrolyte interface and thus independently study the charge generation and charge transport properties in the nanocomposites. The large dark currents for sulfite oxidation have already been observed in various reports on copper vanadates and it might be an indication of the electrochemical activity of these materials towards the sulfite oxidation.25,26,29,32–34 The dark currents decrease as the amount of WO3 increases, thus confirming that they originate from the CVO component. A new nanocomposite CVO/WO3 = 1[thin space (1/6-em)]:[thin space (1/6-em)]0.1 was prepared with lower WO3 mass content in order to follow the trends of photocurrent over a broader range of compositions. The onset potential of all the nanocomposites cathodically shifts as the amount of WO3 decreases (Fig. S8, ESI). Such a cathodic shift has been previously observed in BiVO4/WO3 nanocomposites and attributed to reduced charge recombination in comparison to the individual metal oxides.12 To facilitate the comparison between the different samples, the net photocurrent density, obtained by subtracting the dark current density from the total photocurrent density, is reported in Fig. 4B. The potential of 1.23 V versus the reversible hydrogen electrode (RHE) is used as a reference to ease the comparison with the literature. The highest photocurrent value of 0.45 mA cm−2 at 1.23 V vs. RHE is observed for the nanocomposite with a CVO[thin space (1/6-em)]:[thin space (1/6-em)]WO3 mass ratio of 1[thin space (1/6-em)]:[thin space (1/6-em)]0.1 which corresponds to a 10 fold increase compared to the bare CVO counterpart (0.05 mA cm−2). The increased photocurrent does not correspond to increased light absorption (Fig. S9, ESI), which suggests that charge separation and transport drive the performance of the nanocomposites.

image file: c8tc02888e-f4.tif
Fig. 4 (A) JE curves under dark and illuminated conditions for WO3, CVO/WO3 nanocomposites with different mass ratios and CVO measured in a 0.1 M sodium borate buffer (pH 8.2) containing 0.1 M Na2SO3 as the hole scavenger using AM 1.5G (100 mW cm−2) illumination. The dark currents are shown as dashed lines and the photocurrents are shown as solid lines. (B) Net photocurrent density (Jnet = JlightJdark) at 1.23 V vs. RHE plotted against the mass fraction of WO3 (xWO3) in the nanocomposites, given by the formula image file: c8tc02888e-t1.tif

Optical spectroscopy was utilized to get insights into the PEC results. Steady-state absorption spectra of pure WO3, pure CVO and CVO/WO3 nanocomposites are reported in Fig. 5. WO3 has an absorption band edge at 420–430 nm, typical of nanostructured samples,38 and presents also a characteristic tail which starts to rise at 600 nm and extends in the near IR-range. Previous studies have attributed this tail to the absorption from oxygen vacancies and defects, which here persist even upon annealing.39,40 As for the pure nanocrystalline CVO, we observe a very specific feature of copper vanadates, which is the symmetric shoulder peaking between 800 and 1000 nm very similar to the one reported by Sharp et al., which the authors attributed to the ligand field effect of Cu2+ cations.25,33 In the nanocomposites, the absorption edge blue-shifts as the amount of WO3 increases similarly to that observed by Lee et al. for BiVO4/WO3 heterojunctions with different amounts of WO3.12

image file: c8tc02888e-f5.tif
Fig. 5 Steady-state absorption spectra of the CVO/WO3 nanocomposites along with the spectra of pure WO3 and CVO for comparison.

Photoluminescence (PL) spectroscopy is one simple yet powerful tool to investigate the efficiency of charge carrier generation which accompanies the suppression of the electron–hole pairs in the heterojunctions.15,16,41 Usually, metal oxides do not have strong emission at room temperature and its origin is usually attributed to the radiative recombination from surface trapped-excitons or from intra-band defects, such as oxygen vacancies.15Fig. 6 shows the PL spectra of pure WO3, pure CVO and CVO/WO3 nanocomposites with different mass ratios. To avoid misinterpretation, the emission spectra have been normalized and compared with respect to the shape; thus we will not discuss the changes in amplitude which can only provide meaningful information for very similar samples. The spectra were recorded upon excitation at 457 nm (2.7 eV) at room temperature. In the pure WO3, one major peak at 492 nm is observed. Since the excitation wavelength does not allow reaching the conduction band of WO3, this emission peak must come from intra-band defects such as oxygen vacancies.40,42 The pure CVO presents a pronounced emission peak at 496 nm. The fact that we observe the emission at room temperature is an indication that the electron–hole pairs might be very strongly bound. In the nanocomposite samples with a mass ratio of 1[thin space (1/6-em)]:[thin space (1/6-em)]1 and 1[thin space (1/6-em)]:[thin space (1/6-em)]0.5, the emission peaks of both WO3 and CVO are clearly evident; instead the nanocomposite with mass ratio 1[thin space (1/6-em)]:[thin space (1/6-em)]0.25 exhibits a much broader PL emission peak indicating that the emission from WO3 and CVO is strongly suppressed. This finding is consistent with the interpretation that in the blended nanocomposite with a 1[thin space (1/6-em)]:[thin space (1/6-em)]0.25 proportion of the two domains, the electron–hole pairs are more efficiently separated. To further elucidate the role of the bulk heterojunction in the charge separation efficiency and to monitor the charge recombination time scale of the generated species to the ground state, transient absorption (TA) in the μs time scale was performed and the results comparing the pure CVO and one of the nanocomposites are reported in Fig. 7. In these measurements, the excitation energy is higher than the band gaps of both WO3 and β-Cu2V2O7 in order to better simulate the solar illumination conditions wherein both components of the heterojunction are excited. The charge carrier dynamics are faster in the pure CVO compared to the CVO/WO3 nanocomposite with a time constant τ50% (which represents the time to reach half of the initial absorption43) equal to 250 ns and 730 ns, respectively. The longer time decay for the nanocomposite is the signature that the recombination of electrons and holes has been suppressed. If we assume that the transient absorption signal originates from photo-excited holes as in the case of BiVO4, then this observation suggests the occurrence of electron transfer from CVO to WO3, which results in longer lived holes. Hence, this novel heterojunction behaves consistently with a type II band alignment, similarly to previously reported systems (e.g. BiVO4/TiO2 and BiVO4/WO3).11,15 While estimating the band gap alignment of the CVO/WO3 junction from the values of the onset potential and of the optical band gap (Fig. S10, ESI) supports this hypothesis, further proofs are needed to confirm a type II band alignment in the CVO/WO3 nanocomposites as we showed for BiVO4/TiO2.15

image file: c8tc02888e-f6.tif
Fig. 6 Normalized room-temperature PL emission spectra of pure WO3, CVO, and CVO/WO3 nanocomposites deposited on a quartz substrate upon excitation of 457 nm.

image file: c8tc02888e-f7.tif
Fig. 7 Normalized transient absorption decay response in the μs timescale for the pure CVO and the CVO/WO3 nanocomposite with mass ratio 1[thin space (1/6-em)]:[thin space (1/6-em)]0.25. The pump wavelength was 355 nm, the transient signal was probed at 700 nm and the fluence was 3.5 mJ cm−2 to avoid physical damage to the sample.

Finally, it is interesting to comment on the finding that a lower loading of WO3 is associated with higher PEC performance in the CVO/WO3 nanocomposites. To explain this behavior, NC assembly must be discussed. Periodic 2D and 3D binary NC solids are obtained only when two species of highly monodisperse NCs with a properly chosen relative size self-assemble upon drying of the solvent in controlled conditions.44–46 Being far from these ideal conditions, the Cu NCs and WO2.7 NRs used as building blocks in this work tend to phase segregate, as commonly observed also in other systems.47Fig. 8 shows TEM images of the precursor solutions utilized for the synthesis of the CVO/WO3 nanocomposites. Upon the removal of the native ligands, which enable the dispersion of the NRs and NCs in polar solvents, the WO2.7 NRs tend to aggregate and segregate in domains which are a few hundreds of nanometers in size on the TEM grid. Low resolution SEM and optical microscopy suggest that such phase segregation at the nanoscale is accompanied by the formation of mesoscale features with a branched morphology in the order of 100 μm in size (Fig. S11, ESI). As the amount of the WO2.7 NRs in the precursor solution increases, the domains become bigger (Fig. 8B and C). Bigger domains imply a less extended interfacial area between CVO and WO3 in the resulting nanocomposite, thus less efficient charge injection and worse PEC performance.11

image file: c8tc02888e-f8.tif
Fig. 8 (A–C) TEM images of the precursor solution for CVO/WO3 with mass ratio 1[thin space (1/6-em)]:[thin space (1/6-em)]0.25, 1[thin space (1/6-em)]:[thin space (1/6-em)]0.5 and 1[thin space (1/6-em)]:[thin space (1/6-em)]1, respectively, containing the Cu NCs and WO2.7 NRs in dimethylformamide, along with the corresponding schematics of the composition-dependent structure in the resulting CVO/WO3 nanocomposites.

In conclusion, CVO/WO3 heterostructured nanocomposites with tunable composition were assembled through a novel solution-based approach from a precursor ink containing Cu NCs, WO2.7 NRs, and VO(acac)2. The WO3 loading was found to be the key parameter controlling the PEC performance of the nanocomposites, with the highest photocurrent density towards sulfite oxidation measured for the CVO/WO3 with a 1[thin space (1/6-em)]:[thin space (1/6-em)]0.1 mass ratio. Optical spectroscopy provided insights into the charge injection mechanism across the interface and a type-II heterojunction is finally suggested. The better performance of the nanocomposites with lower WO3 content is attributed to the more extended contact area between the two components, which is beneficial for charge injection. While the segregation between CVO and WO3 is detrimental in our case, this study reveals an interesting correlation between the nanoscale assembly and the mesoscale structure in inorganic nanocomposites (Fig. 8 and Fig. S11, ESI), in a similar manner in which the molecular interaction in organic bulk heterojunctions impacts the macroscale morphology of the polymeric blends.1–6 Together with a few others,11,18 this work strengthens the importance of aiming at a superior structural tunability and control of the intrinsic inhomogeneity in inorganic bulk heterojunctions, and motivates researchers to explore ordered NC binary superlattices wherein the number of contact points between the individual NCs can be increased and finely tuned.45,46,48

Conflicts of interest

There are no conflicts to declare.


This work was supported by the Swiss National Science Foundation (AP Energy Grant, project number PYAPP2_166897/1). M. S. and A. L. acknowledge the H2020-Marie Curie Individual Fellowships with grant numbers 753124 and 701745, respectively. L. R. acknowledges Prof. Silvia Ardizzone for the helpful discussions and the PhD course in Chemistry of the Università degli Studi di Milano for the economic support. The XRD experiments were performed on the Swiss-Norwegian beamline BM01 at the European Synchrotron Radiation Facility (ESRF), Grenoble, France (proposal number: 01-02 1176). We are grateful to Dr Dmitry Chernyshov at the ESRF for providing great assistance in using beamline BM01.


  1. J. J. M. Halls, K. Pichler, R. H. Friend, S. C. Moratti and A. B. Holmes, Appl. Phys. Lett., 1996, 68, 3120–3122 CrossRef.
  2. A. C. Morteani, P. Sreearunothai, L. M. Herz, R. H. Friend and C. Silva, Phys. Rev. Lett., 2004, 92, 247402 CrossRef PubMed.
  3. W. Zhang, M. Saliba, D. T. Moore, S. K. Pathak, M. T. Hörantner, T. Stergiopoulos, S. D. Stranks, G. E. Eperon, J. A. Alexander-Webber, A. Abate, A. Sadhanala, S. Yao, Y. Chen, R. H. Friend, L. A. Estroff, U. Wiesner and H. J. Snaith, Nat. Commun., 2015, 6, 6142 CrossRef PubMed.
  4. R. H. Friend, R. W. Gymer, A. B. Holmes, J. H. Burroughes, R. N. Marks, C. Taliani, D. D. C. Bradley, D. A. Dos Santos, J. L. Brédas, M. Lögdlund and W. R. Salaneck, Nature, 1999, 397, 121–128 CrossRef.
  5. K. Wojciechowski, S. D. Stranks, A. Abate, G. Sadoughi, A. Sadhanala, N. Kopidakis, G. Rumbles, C. Z. Li, R. H. Friend, A. K. Y. Jen and H. J. Snaith, ACS Nano, 2014, 8, 12701–12709 CrossRef PubMed.
  6. Y. S. Huang, S. Westenhoff, I. Avilov, P. Sreearunothai, J. M. Hodgkiss, C. Deleener, R. H. Friend and D. Beljonne, Nat. Mater., 2008, 7, 483–489 CrossRef PubMed.
  7. Y. Tachibana, L. Vayssieres and J. R. Durrant, Nat. Photonics, 2012, 6, 511–518 CrossRef.
  8. A. Loiudice, J. Ma, W. S. Drisdell, T. M. Mattox, J. K. Cooper, T. Thao, C. Giannini, J. Yano, L. W. Wang, I. D. Sharp and R. Buonsanti, Adv. Mater., 2015, 27, 6733–6740 CrossRef PubMed.
  9. I. D. Sharp, J. K. Cooper, F. M. Toma and R. Buonsanti, ACS Energy Lett., 2017, 2, 139–150 CrossRef.
  10. K. Sivula, F. Le Formal and M. Grätzel, ChemSusChem, 2011, 4, 432–449 CrossRef PubMed.
  11. A. Loiudice, J. K. Cooper, L. H. Hess, T. M. Mattox, I. D. Sharp and R. Buonsanti, Nano Lett., 2015, 15, 7347–7354 CrossRef PubMed.
  12. S. J. Hong, S. Lee, J. S. Jang and J. S. Lee, Energy Environ. Sci., 2011, 4, 1781 RSC.
  13. I. Grigioni, K. G. Stamplecoskie, D. H. Jara, M. V. Dozzi, A. Oriana, G. Cerullo, P. V. Kamat and E. Selli, ACS Energy Lett., 2017, 2, 1362–1367 CrossRef.
  14. J. Su, L. Guo, N. Bao and C. A. Grimes, Nano Lett., 2011, 11, 1928–1933 CrossRef PubMed.
  15. L. H. Hess, J. K. Cooper, A. Loiudice, C. M. Jiang, R. Buonsanti and I. D. Sharp, Nano Energy, 2017, 34, 375–384 CrossRef.
  16. S. Ho-Kimura, S. J. A. Moniz, A. D. Handoko and J. Tang, J. Mater. Chem. A, 2014, 2, 3948 RSC.
  17. J. Resasco, H. Zhang, N. Kornienko, N. Becknell, H. Lee, J. Guo, A. L. Briseno and P. Yang, ACS Cent. Sci., 2016, 2, 80–88 CrossRef PubMed.
  18. M. Xie, X. Fu, L. Jing, P. Luan, Y. Feng and H. Fu, Adv. Energy Mater., 2014, 4, 1300995 CrossRef.
  19. K. Sivula, F. Le Formal and M. Grätzel, Chem. Mater., 2009, 21, 2862–2867 CrossRef.
  20. M. Bartsch, M. Sarnowska, O. Krysiak, C. Willa, C. Huber, L. Pillatsch, S. Reinhard and M. Niederberger, ACS Omega, 2017, 2, 4531–4539 CrossRef.
  21. D. Barreca, G. Carraro, A. Gasparotto, C. Maccato, M. E. A. Warwick, K. Kaunisto, C. Sada, S. Turner, Y. Gönüllü, T. P. Ruoko, L. Borgese, E. Bontempi, G. Van Tendeloo, H. Lemmetyinen and S. Mathur, Adv. Mater. Interfaces, 2015, 2, 1500313 CrossRef.
  22. Y. Qiu, W. Liu, W. Chen, G. Zhou, P. C. Hsu, R. Zhang, Z. Liang, S. Fan, Y. Zhang and Y. Cui, Sci. Adv., 2016, 2, e1501764 Search PubMed.
  23. X. Shi, I. Y. Choi, K. Zhang, J. Kwon, D. Y. Kim, J. K. Lee, S. H. Oh, J. K. Kim and J. H. Park, Nat. Commun., 2014, 5, 4775 CrossRef PubMed.
  24. Q. Yan, J. Yu, S. K. Suram, L. Zhou, A. Shinde, P. F. Newhouse, W. Chen, G. Li, K. A. Persson, J. M. Gregoire and J. B. Neaton, Proc. Natl. Acad. Sci. U. S. A., 2017, 114, 3040–3043 CrossRef PubMed.
  25. C. M. Jiang, G. Segev, L. H. Hess, G. Liu, G. Zaborski, F. M. Toma, J. K. Cooper and I. D. Sharp, ACS Appl. Mater. Interfaces, 2018, 10, 10627–10633 CrossRef PubMed.
  26. M. A. Lumley and K. S. Choi, Chem. Mater., 2017, 29, 9472–9479 CrossRef.
  27. P. F. Newhouse, D. A. Boyd, A. Shinde, D. Guevarra, L. Zhou, E. Soedarmadji, G. Li, J. B. Neaton and J. M. Gregoire, J. Mater. Chem. A, 2016, 4, 7483–7494 RSC.
  28. D. Cardenas-Morcoso, A. Peiro-Franch, I. Herraiz-Cardona and S. Gimenez, Catal. Today, 2017, 290, 65–72 CrossRef.
  29. W. Guo, W. D. Chemelewski, O. Mabayoje, P. Xiao, Y. Zhang and C. B. Mullins, J. Phys. Chem. C, 2015, 119, 27220–27227 CrossRef.
  30. L. Zhou, Q. Yan, A. Shinde, D. Guevarra, P. F. Newhouse, N. Becerra-Stasiewicz, S. M. Chatman, J. A. Haber, J. B. Neaton and J. M. Gregoire, Adv. Energy Mater., 2015, 5, 1500968 CrossRef.
  31. L. Zhou, Q. Yan, J. Yu, R. J. R. Jones, N. Becerra-Stasiewicz, S. K. Suram, A. Shinde, D. Guevarra, J. B. Neaton, K. A. Persson and J. M. Gregoire, Phys. Chem. Chem. Phys., 2016, 18, 9349–9352 RSC.
  32. C. Gadiyar, M. Strach, P. Schouwink, A. Loiudice and R. Buonsanti, Chem. Sci., 2018, 9, 5658–5665 RSC.
  33. C.-M. Jiang, M. Farmand, C. H. Wu, Y.-S. Liu, J. Guo, W. S. Drisdell, J. K. Cooper and I. D. Sharp, Chem. Mater., 2017, 29, 3334–3345 CrossRef.
  34. J. A. Seabold and N. R. Neale, Chem. Mater., 2015, 27, 1005–1013 CrossRef.
  35. T. M. Mattox, A. Bergerud, A. Agrawal and D. J. Milliron, Chem. Mater., 2014, 26, 1779–1784 CrossRef.
  36. P. M. Woodward, A. W. Sleight and T. Vogt, J. Solid State Chem., 1997, 131, 9–17 CrossRef.
  37. J. A. Seabold and K. S. Choi, Chem. Mater., 2011, 23, 1105–1112 CrossRef.
  38. H. Zheng, J. Z. Ou, M. S. Strano, R. B. Kaner, A. Mitchell and K. Kalantar-Zadeh, Adv. Funct. Mater., 2011, 21, 2175–2196 CrossRef.
  39. Z. Xu, X. Li, J. Li, L. Wu, Q. Zeng and Z. Zhou, Appl. Surf. Sci., 2013, 284, 285–290 CrossRef.
  40. F. Zheng, M. Guo and M. Zhang, CrystEngComm, 2013, 15, 277–284 RSC.
  41. H. Li, H. Yu, X. Quan, S. Chen and H. Zhao, Adv. Funct. Mater., 2015, 25, 3074–3080 CrossRef.
  42. M. Ghiyasiyan-Arani, M. Masjedi-Arani and M. Salavati-Niasari, J. Mater. Sci.: Mater. Electron., 2016, 27, 4871–4878 CrossRef.
  43. A. Kafizas, X. Wang, S. R. Pendlebury, P. Barnes, M. Ling, C. Sotelo-Vazquez, R. Quesada-Cabrera, C. Li, I. P. Parkin and J. R. Durrant, J. Phys. Chem. A, 2016, 120, 715–723 CrossRef PubMed.
  44. Z. Chen, J. Moore, G. Radtke, H. Sirringhaus and S. O’Brien, J. Am. Chem. Soc., 2007, 129, 15702–15709 CrossRef PubMed.
  45. X. Ye, J. A. Millan, M. Engel, J. Chen, B. T. Diroll, S. C. Glotzer and C. B. Murray, Nano Lett., 2013, 13, 4980–4988 CrossRef PubMed.
  46. Y. Kang, X. Ye, J. Chen, L. Qi, R. E. Diaz, V. Doan-Nguyen, G. Xing, C. R. Kagan, J. Li, R. J. Gorte, E. A. Stach and C. B. Murray, J. Am. Chem. Soc., 2013, 135, 1499–1505 CrossRef PubMed.
  47. A. Brumberg, B. T. Diroll, G. Nedelcu, M. E. Sykes, Y. Liu, S. M. Harvey, M. R. Wasielewski, M. V. Kovalenko and R. D. Schaller, Nano Lett., 2018, 18, 4771–4776 CrossRef PubMed.
  48. A. Dong, J. Chen, P. M. Vora, J. M. Kikkawa and C. B. Murray, Nature, 2010, 466, 474–477 CrossRef PubMed.


Electronic supplementary information (ESI) available. See DOI: 10.1039/c8tc02888e
These authors have contributed equally to the work.

This journal is © The Royal Society of Chemistry 2018