ELECTRON TRANSPORT LAYER FOR FLEXIBLE PEROVSKITE SOLAR CELL AND FLEXIBLE PEROVSKITE SOLAR CELL INCLUDING THE SAME

Abstract

Disclosed is an electron transport layer for a flexible perovskite solar cell. The electron transport layer includes transition metal-doped titanium dioxide particles. The titanium dioxide particles are densely packed in the electron transport layer. The electron transport layer is transparent. The use of the electron transport layer enables the fabrication of a flexible perovskite solar cell with high power conversion efficiency. Also disclosed is a flexible perovskite solar cell employing the electron transport layer.

Claims

1. A method for preparing an electron transport layer for a flexible perovskite solar cell, comprising (A) mixing a titanium precursor, a transition metal precursor, and an organic ligand to prepare transition metal-doped titanium dioxide particles and (B) irradiating the transition metal-doped titanium dioxide particles with UV.

2. The method according to claim 1, wherein, in step (A), the transition metal precursor is selected from the group consisting of scandium (III) triflate, yttrium borate, yttrium acetate hydrate, lanthanum nitrate, niobium ethoxide isopropoxide, niobium alkoxides, ammonium metavanadate, tantalum ethoxide, and mixtures thereof.

3. The method according to claim 1, wherein, in step (A), the transition metal precursor is used in an amount of 1 to 5 mol %, based on the total weight of the mixture of the titanium precursor, the transition metal precursor, and the organic ligand.

4. The method according to claim 1, wherein, in step (A), the titanium precursor and the organic ligand are mixed in a molar ratio of 1:2-5.

5. The method according to claim 1, wherein, in step (A), the mixture of the titanium precursor, the transition metal precursor, and the organic ligand is heated at 250 to 300° C. for 1 to 3 hours.

6. The method according to claim 1, wherein, in step (B), the transition metal-doped titanium dioxide particles are irradiated with UV for 40 to 120 minutes.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0032] These and/or other aspects and advantages of the invention will become apparent and more readily appreciated from the following description of the embodiments, taken in conjunction with the accompanying drawings of which:

[0033] FIGS. 1A and 1B are high-resolution transmission electron microscopy (HR-TEM) images of titanium dioxide particles capped with an organic ligand at different magnifications;

[0034] FIGS. 2A and 2B are high-resolution transmission electron microscopy (HR-TEM) images of Nb:TiO.sub.2 (Nb 1 mol %) and Nb:TiO.sub.2 (Nb 2 mol %), respectively, FIG. 2C shows powder XRD patterns of undoped TiO.sub.2, Nb:TiO.sub.2 (Nb 1 mol %), and Nb:TiO.sub.2 (Nb 2 mol %), and FIG. 2D shows details of XRD patterns of undoped TiO.sub.2, Nb:TiO.sub.2 (Nb 1 mol %), and Nb:TiO.sub.2 (Nb 2 mol %) at 2θ of around 25°;

[0035] FIG. 3 shows X-ray photoelectron spectra of undoped TiO.sub.2, Nb:TiO.sub.2 (Nb 1 mol %), and Nb:TiO.sub.2 (Nb 2 mol %);

[0036] FIG. 4A shows ATR-FTIR spectra of TiO.sub.2 nanoparticles before and after UV treatment, FIG. 4B is a TEM image of a TiO.sub.2 thin film formed by spin coating of TiO.sub.2 nanoparticles dispersed in hexane (before UV treatment), FIG. 4C is a TEM image of the TiO.sub.2 thin film after UV treatment, FIG. 4D is a HR-TEM image of FIG. 4B, and FIG. 4E is a HR-TEM image of FIG. 4C, the inset in the TEM image of FIG. 4C is the corresponding selected area electron diffraction (SAED) pattern;

[0037] FIG. 5 shows the removal of an organic ligand capped on TiO.sub.2 nanoparticles by UV treatment;

[0038] FIG. 6A is a SEM image showing the surface of HT-TiO.sub.2-coated FTO glass prepared in Comparative Example 1, FIG. 6B is a SEM image showing the surface of UV-TiO.sub.2-coated FTO glass prepared in Comparative Example 2, FIG. 6C shows cyclic voltammograms of FTO glass, HT-TiO.sub.2-coated FTO glass, and UV-TiO.sub.2-coated FTO glass, FIG. 6D shows the transmittances of FTO glass, HT-TiO.sub.2-coated FTO glass, UV-TiO.sub.2-coated FTO glass, and UV-Nb:TiO.sub.2-coated FTO glass, the inset of FIG. 6A is a SEM image showing the surface of FTO glass.

[0039] FIGS. 7A and 7B are 3D atomic force microscopy (AFM) images showing the surface morphologies of HT-TiO.sub.2-coated ITO glass and UV-TiO.sub.2-coated ITO glass, respectively;

[0040] FIG. 8 shows Tauc plots obtained from UV-Vis spectra of HT-TiO.sub.2, UV-TiO.sub.2, and UV-Nb:TiO.sub.2 thin films;

[0041] FIG. 9 shows top SEM images of CH.sub.3NH.sub.3PbI.sub.3 perovskite films coated with (a) HT-TiO.sub.2, (b) UV-TiO.sub.2, and (c) UV-Nb:TiO.sub.2 electron transport layers;

[0042] FIG. 10A is a cross-sectional SEM image of a planar PSC including a UV-Nb:TiO.sub.2 electron transport layer, FIG. 10B shows photocurrent density-voltage (J-V) curves of planar PSCs including HT-TiO.sub.2, UV-TiO.sub.2, and UV-Nb:TiO.sub.2 electron transport layers, and FIG. 10C shows incident photon-to-current conversion efficiency (IPCE) spectra of planar PSCs including HT-TiO.sub.2, UV-TiO.sub.2, and UV-Nb:TiO.sub.2 electron transport layers;

[0043] FIG. 11 shows hysteresis between J-V curves of PSCs including HT-TiO.sub.2, UV-TiO.sub.2, and UV-Nb:TiO.sub.2 electron transport layers, which were obtained from different scan directions;

[0044] FIG. 12 shows the steady-state photocurrent density of a PSC including a UV-Nb:TiO.sub.2 electron transport layer;

[0045] FIG. 13A shows UV photoelectron spectra (UPS) of a UV-Nb:TiO.sub.2 electron transport layer and FIG. 13B shows a UV-TiO.sub.2 electron transport layer;

[0046] FIG. 14A shows energy level alignment of the individual layers of a PSC, FIG. 14B shows steady-state photoluminescence (PL) spectra of perovskite layers in contact with HT-TiO.sub.2, UV-TiO.sub.2, and UV-Nb:TiO.sub.2 electron transport layers, and FIG. 14C shows time-resolved photoluminescence (TRPL) spectra of perovskite layers in contact with HT-TiO.sub.2, UV-TiO.sub.2, and UV-Nb:TiO.sub.2 electron transport layers;

[0047] FIG. 15A is a photograph of a PSC including a UV-Nb:TiO.sub.2 electron transport layer coated on an ITO/PEN substrate, FIG. 15B shows J-V curves of the PSC obtained after recovery from bending at different radii (r), and FIG. 15C shows photovoltaic parameters of the PSC after bending at r=15 mm for 1000 cycles of bending; and

[0048] FIG. 16 is a cross-sectional SEM image of the PSC of FIG. 15A.

DETAILED DESCRIPTION OF THE INVENTION

[0049] The present invention is directed to a transparent electron transport layer for a flexible perovskite solar cell in which titanium dioxide particles are densely packed and a flexible perovskite solar cell (PSC) with high power conversion efficiency employing the electron transport layer.

[0050] Particularly, the electron transport layer of the present invention can be sintered at a low temperature of 30 to 50° C. through UV treatment, unlike conventional electron transport layers that are sintered at high temperatures. The UV treatment allows dense packing of the titanium dioxide particles and makes the electron transport layer transparent. The use of the electron transport layer ensures high performance of the flexible perovskite solar cell (PSC) and allows the flexible perovskite solar cell to have advantageous characteristics in terms of short-circuit current density, open-circuit voltage, fill factor, and power conversion efficiency.

[0051] The present invention will now be described in more detail.

[0052] The present invention provides an electron transport layer for a flexible perovskite solar cell including transition metal-doped titanium dioxide particles treated with UV (for example, UV-Nb:TiO.sub.2).

[0053] The titanium dioxide particles have a diameter of 0.5 to 3 nm, preferably 1.5 to 3 nm. Outside this range, the transparency of the electron transport layer may be lowered.

[0054] The transition metal contributes to improvement of electrical conductivity and charge extraction without affecting the optical properties of the electron transport layer. Examples of suitable transition metals include Group 3 transition metals and/or Group 5 transition metals. Preferably, the transition metal is selected from the group consisting of scandium (Sc), yttrium (Y), lanthanum (La), niobium (Nb), vanadium (V), tantalum (Ta), and combinations thereof. If a metal other than the transition metal is used, an improvement in charge extraction is not expected and high performance of the electron transport layer cannot be ensured.

[0055] The transition metal precursor is used in an amount of 1 to 5 mol %, preferably 1 to 2 mol %, based on the total weight of the mixture of the transition metal and the titanium dioxide particles. If the content of the transition metal is less than the lower limit defined above, improvement of electrical conductivity and charge extraction cannot be expected. Meanwhile, if the content of the transition metal exceeds the upper limit defined above, low charge extraction may also be caused.

[0056] The UV treatment of the transition metal-doped titanium dioxide particles provides the following advantages. First, the UV treatment induces fusion of the titanium dioxide particles, allowing their dense packing. Second, the UV treatment enables the removal of an organic ligand that is necessary for the preparation of the titanium dioxide particles but causes deterioration of photovoltaic performance due to the presence of its long alkyl chain.

[0057] The present invention also provides a method for preparing an electron transport layer for a flexible perovskite solar cell.

[0058] The method includes (A) mixing a titanium precursor, a transition metal precursor, and an organic ligand to prepare transition metal-doped titanium dioxide particles and (B) irradiating the transition metal-doped titanium dioxide particles with UV.

[0059] First, in step (A), transition metal-doped titanium dioxide particles are prepared.

[0060] The transition metal-doped titanium dioxide particles are prepared by heating a titanium precursor, a transition metal precursor, and an organic ligand at 250 to 300° C., preferably 260 to 280° C., for 1 to 3 hours, preferably 1.5 to 2 hours. If the heating temperature and time are less than the respective lower limits defined above, desired effects cannot be obtained. Meanwhile, if the heating temperature and time exceed the respective upper limits defined above, a large amount of impurities is produced.

[0061] The titanium precursor is not limited to a particular kind but is preferably titanium isopropoxide.

[0062] The transition metal precursor is used in an amount of 1 to 5 mol %, preferably 1 to 2 mol %, based on the total weight of the mixture of the titanium precursor, the transition metal precursor, and the organic ligand. If the content of the transition metal is less than the lower limit defined above, improvement of electrical conductivity and charge extraction cannot be expected. Meanwhile, if the content of the transition metal exceeds the upper limit defined above, low charge extraction may also be caused.

[0063] The transition metal precursor is not particularly limited but is preferably selected from the group consisting of scandium (III) triflate, yttrium borate, yttrium acetate hydrate, lanthanum nitrate, niobium ethoxide isopropoxide, niobium alkoxides, ammonium metavanadate, tantalum ethoxide, and mixtures thereof. More preferred is niobium ethoxide isopropoxide.

[0064] The organic ligand is not particularly limited so long as it can cap and disperse titanium dioxide particles without aggregation. The organic ligand is preferably selected from the group consisting of oleic acid, oleylamine, linoleic acid, stearic acid, and mixtures thereof.

[0065] The titanium precursor and the organic ligand are mixed in a molar ratio of 1:2-5, preferably 1:2-3. If the ratio of the organic ligand to the titanium precursor is less than 2, titanium dioxide particles may aggregate, resulting in poor performance. Meanwhile, if the ratio of the organic ligand to the titanium precursor exceeds 5, the organic ligand may remain unremoved despite UV treatment, resulting in poor photovoltaic performance.

[0066] Next, in step (B), the transition metal-doped titanium dioxide particles are irradiated with UV.

[0067] The transition metal-doped titanium dioxide particles are irradiated with a UV lamp in the wavelength range of 200 to 400 nm for 40 to 120 minutes, preferably 50 to 70 minutes.

[0068] The present invention also provides a flexible perovskite solar cell including the electron transport layer.

[0069] The flexible perovskite solar cell of the present invention includes a substrate, the electron transport layer (e.g., UV-Nb:TiO.sub.2) disposed on the substrate, a perovskite layer disposed on the electron transport layer, and a hole transport layer disposed on the perovskite layer, and an electrode layer disposed on the hole transport layer.

[0070] The flexible perovskite solar cell has an open-circuit voltage (V.sub.oc) of 1.09 to 1.13 V, a short-circuit current density (J.sub.sc) of 22.0 to 23.0 mAcm.sup.−2, a fill factor (FF) of 0.75 to 0.80, and a power conversion efficiency (PCE) of 17 to 20%.

[0071] The substrate may be, for example, fluorine-doped tin oxide (FTO) glass, indium tin oxide (ITO) glass or indium tin oxide/polyethylene naphthalate (ITO/PEN). The ITO/PEN refers to a substrate in which ITO is bonded to PEN.

[0072] The following examples are provided to assist in further understanding of the invention. However, these examples are intended for illustrative purposes only. It will be evident to those skilled in the art that various modifications and changes can be made without departing from the scope and spirit of the invention and such modifications and changes are encompassed within the scope of the appended claims.

Example 1: Preparation of UV-Nb:TiO.SUB.2 .(Electron Transport Layer)

[0073] Preparation of Nb:TiO.sub.2 (Nb-Doped Titanium Dioxide Particles)

[0074] Nb-doped titanium dioxide particles were synthesized via a non-hydrolytic sol-gel reaction.

[0075] 88 mmol of oleic acid (OA, Sigma Aldrich, 93%) and niobium (V) ethoxide (Aldrich; 99.95%, 1 mol % and 2 mol %, based on the total weight of the mixture) were degassed in a three-neck flask under vacuum. 1 h later, 30 mmol of titanium (IV) isopropoxide (TTIP; Aldrich; 97%) was added to the flask. The mixture was vigorously stirred at 270° C. for 2 h. When the color of the mixture turned from transparent yellow to white, the mixture was again degassed with 99.99% argon gas, giving a precipitate (Nb:TiO.sub.2).

[0076] The precipitate was washed with excess ethanol and purified by centrifugation at 3000 rpm for 30 min.

[0077] Preparation of UV Treated UV-Nb:TiO.sub.2

[0078] The purified Nb:TiO.sub.2 was re-dispersed in hexane at a predetermined concentration. The Nb:TiO.sub.2 dispersion was dropped onto a TCO substrate (FTO glass or ITO/PEN) and immediately spin-coated at a rate of 5000 rpm. Thereafter, the Nb:TiO.sub.2-coated TCO substrate was irradiated with a UV lamp curing system (JHCI-051B, JECO) equipped with a 400 W mercury lamp (wavelength range 200-400 nm) to prepare UV-Nb:TiO.sub.2.

Comparative Example 1: Preparation of HT-TiO.SUB.2 .(Electron Transport Layer)

[0079] A titanium diisopropoxide bis(acetylacetonate) solution (75% in isopropanol, Aldrich) was spin-coated on a TCO substrate (FTO glass or ITO/PEN) and annealed at 500° C. for 30 min to prepare high temperature sintered HT-TiO.sub.2.

Comparative Example 2: Preparation of UV-TiO.SUB.2 .(Electron Transport Layer)

[0080] Preparation of TiO.sub.2

[0081] Undoped titanium dioxide particles were synthesized via a non-hydrolytic sol-gel reaction.

[0082] 88 mmol of oleic acid (OA, Sigma Aldrich, 93%) was degassed in a three-neck flask under vacuum. 1 h later, 30 mmol of titanium (IV) isopropoxide (TTIP; Aldrich; 97%) was added to the flask. The mixture was vigorously stirred at 270° C. for 2 h. When the color of the mixture turned from transparent yellow to white, the mixture was again degassed with 99.99% argon gas, giving a precipitate (TiO.sub.2).

[0083] The precipitate was washed with excess ethanol and purified by centrifugation at 3000 rpm for 30 min.

[0084] Preparation of UV Treated UV-TiO.sub.2

[0085] The purified TiO.sub.2 was re-dispersed in hexane at a predetermined concentration. The TiO.sub.2 dispersion was dropped onto a TCO substrate (FTO glass or ITO/PEN) and immediately spin-coated at a rate of 5000 rpm. Thereafter, the TiO.sub.2-coated TCO substrate was irradiated with a UV lamp curing system (JHCI-051B, JECO) equipped with a 400 W mercury lamp (wavelength range 200-400 nm) to prepare UV-TiO.sub.2.

Test Example 1: Comparison of Undoped TiO.SUB.2., Nb:TiO.SUB.2 .(Nb 1 Mol %), and Nb:TiO.SUB.2 .(Nb 2 Mol %)

[0086] FIGS. 1A and 1B are high-resolution transmission electron microscopy (HR-TEM) images of the organic ligand (OA)-capped titanium dioxide particles at different magnifications.

[0087] As shown in FIG. 1, the organic ligand-capped titanium dioxide particles are grown along the [001] direction and are highly crystalline. The spacings between the lattice fringes are 0.35 mm and 0.24 mm, which correspond to the d.sub.101 and d.sub.004 spacings of the anatase crystalline structure, respectively.

[0088] FIGS. 2A and 2B are high-resolution transmission electron microscopy (HR-TEM) images of the Nb:TiO.sub.2 (Nb 1 mol %) and the Nb:TiO.sub.2 (Nb 2 mol %), respectively, FIG. 2C shows powder XRD patterns of the undoped TiO.sub.2, the Nb:TiO.sub.2 (Nb 1 mol %), and the Nb:TiO.sub.2 (Nb 2 mol %), and FIG. 2D shows details of XRD patterns of the undoped TiO.sub.2, the Nb:TiO.sub.2 (Nb 1 mol %), and the Nb:TiO.sub.2 (Nb 2 mol %) at 2θ of around 25°.

[0089] As shown in FIGS. 2A and 2B, there is no difference in the crystal morphology of the nanoparticles despite the different Nb contents.

[0090] As shown in FIGS. 2C and 2D, the X-ray diffraction (XRD) peaks for the undoped TiO.sub.2 and the Nb:TiO.sub.2 nanoparticles showed an anatase crystal structure (JCPDS, No. 76-1940), which was maintained with increasing Nb content. The sharp (004) peaks at 37.8° show anisotropic growth of the TiO.sub.2 nanoparticles, which is in agreement with the HR-TEM image of the TiO.sub.2 nanorods in FIG. 1B. A slight shift in the (101) peak with increasing Nb content is because of the larger radius of Nb.sup.5+ (0.64 Å) than Ti.sup.4+ (0.61 Å).

[0091] Moreover, the XRD patterns reveal that the Nb:TiO.sub.2 nanoparticles were successfully prepared without by-products, such as Nb.sub.2O.sub.5.

[0092] FIG. 3 shows X-ray photoelectron spectra of the undoped TiO.sub.2, the Nb:TiO.sub.2 (Nb 1 mol %), and the Nb:TiO.sub.2 (Nb 2 mol %).

[0093] As shown in FIG. 3, the Nb:TiO.sub.2 (Nb 1 mol %) and the Nb:TiO.sub.2 (Nb 2 mol %) showed two peaks at 207.3 eV and 210.1 eV corresponding to 3d.sub.5/2 and Nb3d.sub.3/2 transitions, respectively, demonstrating the presence of Nb.sup.5+ doped into the TiO.sub.2 nanoparticles.

Test Example 2: Comparison of Examples 1 and Comparative Examples 1-2

[0094] The organic ligand capping the TiO.sub.2 nanoparticles was dispersed without aggregation (the inset of FIG. 1A) but the long alkyl chain of the organic ligand serves as a barrier to charge transport and can deteriorate photovoltaic performance of the TiO.sub.2 nanoparticles. Therefore, UV treatment was conducted to decompose the organic ligand used to introduce the photocatalytic effect of TiO.sub.2 while at the same time preparing a high performance electron transport layer (ETL) by low temperature processing. The UV treatment enables removal of the organic ligand and induces fusion of the TiO.sub.2 nanoparticles, resulting in high density of the TiO.sub.2 nanoparticles.

[0095] The decomposition of the organic ligand was confirmed by attenuated total reflection-Fourier transform infrared spectroscopy (ATR-FTIR).

[0096] FIG. 4A shows ATR-FTIR spectra of the TiO.sub.2 nanoparticles before and after UV treatment, FIG. 4B is a TEM image of the TiO.sub.2 thin film formed by spin coating of the TiO.sub.2 nanoparticles dispersed in hexane (before UV treatment), FIG. 4C is a TEM image of the TiO.sub.2 thin film after UV treatment, FIG. 4D is a HR-TEM image of FIG. 4B, and FIG. 4E is a HR-TEM image of FIG. 4C. The inset in the TEM image of FIG. 4C is the corresponding selected area electron diffraction (SAED) pattern.

[0097] As shown in FIG. 4A, the TiO.sub.2 nanoparticles showed strong peaks at 2855 cm.sup.−1 and 2925 cm.sup.−1 corresponding to the symmetric and asymmetric CH.sub.2 stretching vibration of the OA alkyl chain. Other peaks at 1425 cm.sup.−1 and 1525 cm.sup.−1 are ascribed to the symmetric and asymmetric stretching vibration of the carboxylate group of OA. The characteristic peaks of OA disappeared after UV exposure for 1 h or more, indicating the removal of OA.

[0098] As shown in FIGS. 4B to 4E, the coated TiO.sub.2 thin film shows that the intervals between the nanoparticles are small and the nanoparticles are well-aligned due to the presence of the organic ligand (OA). The exposure of the TiO.sub.2 film to UV led to the removal of the organic ligand (OA) and the fusion of the TiO.sub.2 nanoparticles, making the film smaller and denser. This was clearly observed in FIG. 4E. Moreover, the selected area electron diffraction (SAED) pattern for the UV-TiO.sub.2 film (the inset of FIG. 4C) indicates that the high degree of crystallization of the TiO.sub.2 film appeared after UV treatment.

[0099] FIG. 5 shows the removal of the organic ligand capped on TiO.sub.2 nanoparticles by UV treatment.

[0100] As shown in FIG. 5, the removal of the organic ligand (OA) after UV treatment caused a color change to darker. The TiO.sub.2 surface became more hydrophilic by removal of the hydrophobic organic ligand (OA). Accordingly, the removal of the organic ligand (OA) by UV treatment is verified by improved wettability of the perovskite precursor solution in the TiO.sub.2 layer.

[0101] FIG. 6A is a SEM image showing the surface of the HT-TiO.sub.2-coated FTO glass prepared in Comparative Example 1, FIG. 6B is a SEM image showing the surface of the UV-TiO.sub.2-coated FTO glass prepared in Comparative Example 2, FIG. 6C shows cyclic voltammograms of the FTO glass, the HT-TiO.sub.2-coated FTO glass, and the UV-TiO.sub.2-coated FTO glass, FIG. 6D shows the transmittances of the FTO glass, the HT-TiO.sub.2-coated FTO glass, the UV-TiO.sub.2-coated FTO glass, and the UV-Nb:TiO.sub.2-coated FTO glass. The inset of FIG. 6A is a SEM image showing the surface of the FTO glass.

[0102] As shown in FIGS. 6A and 6B, the FTO glass had an irregular texture and a rough surface but the HT-TiO.sub.2 and UV-TiO.sub.2 electron transport layers coated on the 50 nm thick FTO substrate had smooth surfaces.

[0103] To evaluate the blocking effect of the electron transport layer (ETL) against the FTO glass, cyclic voltammetry (CV) was performed. In an aqueous solution of Fe(CN).sub.6.sup.3−/4−, reversible redox reactions between Fe(CN).sub.6.sup.3− and Fe(CN).sub.6.sup.4− occur on the FTO surface. It was previously reported that when FTO glass is coated with an electron transport layer (ETL), the peak current density of CV decreases with increasing peak-to-peak separation (ΔE.sub.p) because the FTO surface is blocked by the electron transport layer (ETL).

[0104] As shown in FIG. 6C, the UV-TiO.sub.2 coated on the FTO glass (Comparative Example 2) exhibited an efficient blocking effect compared to the HT-TiO.sub.2 (Comparative Example 1) because of its low current density and large ΔE.sub.p. Therefore, it was confirmed that the UV-treated electron transport layer (ETL) completely blocks holes caused by a more effective charge separation and a lower charge recombination.

[0105] As shown in FIG. 6D, the FTO glass coated with the electron transport layer (ETL) showed higher transmittances than the FTO glass because of its decreased surface roughness. Particularly, the UV-treated TiO.sub.2 electron transport layers (UV-TiO.sub.2 and UV-Nb:TiO.sub.2) showed higher transmittances in the range of 300 nm to 650 nm than the HT-TiO.sub.2.

[0106] FIGS. 7A and 7B are 3D atomic force microscopy (AFM) images showing the surface morphologies of the HT-TiO.sub.2-coated ITO glass and the UV-TiO.sub.2-coated ITO glass, respectively.

[0107] The surface roughness values of the HT-TiO.sub.2 and the UV-TiO.sub.2 were confirmed by AFM.

[0108] The HT-TiO.sub.2 had an irregular surface with slightly sparse valleys, appearing as pinholes, whereas the AFM image of the UV-TiO.sub.2 showed a uniform and dense surface. The root mean square (RMS) roughness of the ITO glass was 1.0 nm and the surface roughness increased to 4.5 nm when coated with HT-TiO.sub.2.

[0109] The surface roughness of the UV-TiO.sub.2 was 2.3 nm, indicating a more smooth and dense surface than the HT-TiO.sub.2. The high uniformity and surface of UV-TiO.sub.2 were confirmed by AFM measurements. From the fact that the UV treatment allows the UV-TiO.sub.2 to have a smooth and dense surface, the UV-Nb:TiO.sub.2 (Example 1) is also believed to have a smooth and dense surface.

[0110] FIG. 8 shows Tauc plots obtained from UV-Vis spectra of the HT-TiO.sub.2, UV-TiO.sub.2, and UV-Nb:TiO.sub.2 thin films.

[0111] The absorption coefficient (α) of an indirect band gap semiconductor around the band edge is determined by αhν=A(hν−E.sub.g).sup.2, where A is a constant, hν is the photon energy, and E.sub.g is the optical band gap.

[0112] As shown in FIG. 8, E.sub.g is determined by extrapolating the tangent to the plot curve. The E.sub.g values of the HT-TiO.sub.2 and the UV-TiO.sub.2 (or the UV-Nb:TiO.sub.2) films were 3.52 eV and 3.62 eV, respectively, indicating that the UV-TiO.sub.2 (or the UV-Nb:TiO.sub.2) had a higher band gap than the HT-TiO.sub.2. Since the UV-TiO.sub.2 and the UV-Nb:TiO.sub.2 had the same optical band gap, no influence of Nb doping on the optical band gap was observed.

[0113] This can be explained by the quantum size effect of the TiO.sub.2 nanoparticles originated from the diameter (≦3 nm) of the titanium dioxide nanoparticles. As a result, the high transparency of the UV-treated TiO.sub.2 electron transport layers (UV-TiO.sub.2 and UV-Nb:TiO.sub.2) is beneficial to the light absorption of the perovskite layer due to their high photocurrent. There was no beneficial effect of Nb doping on the optical properties.

[0114] To investigate the effect of Nb doping on electrical properties, 4-point probe measurements were conducted.

[0115] The electrical conductivity of the UV-Nb:TiO.sub.2 (1 mol % Nb) was 2.65×10.sup.−4 Scm.sup.−1, which was about 3-fold higher than that of the UV-TiO.sub.2 (0.95×10.sup.−4 Scm.sup.−1) due to the increased carrier density after Nb doping. The electrical conductivity of the UV-Nb:TiO.sub.2 (1 mol % Nb) was also high compared to that of the HT-TiO.sub.2 film (0.25×10.sup.−4 Scm.sup.−1).

[0116] These results are attributed to the 1-dimensional structure of the sol-gel derived TiO.sub.2 nanoparticles in the form of a highly transparent and highly packed film for more efficient charge transport.

[0117] Planar perovskite solar cells (PSCs) including the HT-TiO.sub.2 (Comparative Example 1), UV-TiO.sub.2 (Comparative Example 2), and UV-Nb:TiO.sub.2 (Example 1) electron transport layers were fabricated on FTO glass. The photovoltaic performance of the perovskite solar cells was investigated.

[0118] FIG. 9 shows top SEM images of CH.sub.3NH.sub.3PbI.sub.3 perovskite films coated with (a) the HT-TiO.sub.2, (b) the UV-TiO.sub.2, and (c) the UV-Nb:TiO.sub.2 electron transport layers.

[0119] The highly crystalline pinhole free CH.sub.3NH.sub.3PbI.sub.3 perovskite layers were prepared by addition of a Lewis base.

[0120] FIG. 10A is a cross-sectional SEM image of the planar PSC including the UV-Nb:TiO.sub.2 electron transport layer, FIG. 10B shows photocurrent density-voltage (J-V) curves of the planar PSCs including the HT-TiO.sub.2, UV-TiO.sub.2, and UV-Nb:TiO.sub.2 electron transport layers, and FIG. 10C shows incident photon-to-current conversion efficiency (IPCE) spectra of the planar PSCs including the HT-TiO.sub.2, UV-TiO.sub.2, and UV-Nb:TiO.sub.2 electron transport layers.

[0121] In the perovskite solar cell (PSC) shown in FIG. 10A, the 50 nm thick UV-Nb:TiO.sub.2 (electron transport layer), 480 nm thick CH.sub.3NH.sub.3PbI.sub.3 (perovskite layer), and 180 nm thick spiro-OMeTAD (hole transport layer) were deposited in this order on FTO glass.

[0122] The optical parameter of the planar PSC including the HT-TiO.sub.2, UV-TiO.sub.2, and UV-Nb:TiO.sub.2 electron transport layers are shown in Table 1.

TABLE-US-00001 TABLE 1 Parameters J.sub.sc (mAcm.sup.−2) V.sub.oc (V) FF PCE (%) HT-TiO.sub.2 20.74 1.06 0.71 15.51 UV-TiO.sub.2 21.05 1.08 0.72 16.37 UV-Nb:TiO.sub.2 22.81 1.10 0.78 19.57

[0123] As shown in Table 1 and FIGS. 10B and 10C, the perovskite solar cell (PSC) including the UV-TiO.sub.2 electron transport layer showed improved short circuit current density (J.sub.sc, 21.05 mAcm.sup.−2), open-circuit voltage (V.sub.oc, 1.08 V), fill factor (FF, 0.72), and power conversion efficiency (PCE, 16.37%) compared to the HT-TiO.sub.2 perovskite solar cell (PSC). Furthermore, the J.sub.sc, V.sub.oc, FF, and PCE perovskite solar cell (PSC) including the UV-Nb:TiO.sub.2 electron transport layer were 22.81 mAcm.sup.−2, 1.10 V, 0.78, and 19.57%, respectively.

[0124] As shown in FIG. 10C, the UV-Nb:TiO.sub.2-based solar cell (PSC) showed a stronger spectral response and a higher IPCE at 300-750 nm than the UV-TiO.sub.2-based solar cell and the HT-TiO.sub.2-based solar cell. These results are attributed to the high transparency and electrical conductivity of the UV-Nb:TiO.sub.2. From the IPCE spectra, the integrated current density value of each device is in agreement with the J.sub.sc value obtained from the J-V curve.

[0125] FIG. 11 shows hysteresis between J-V curves of the PSCs including the HT-TiO.sub.2, UV-TiO.sub.2, and UV-Nb:TiO.sub.2 electron transport layers, which were obtained from different scan directions. In FIG. 11, F represents forward scans from J.sub.sc to V.sub.oc and R represents reverse scans from J.sub.sc to V.sub.oc.

[0126] As shown in FIG. 11, the perovskite solar cell (PSC) including the HT-TiO.sub.2 showed a large difference between the forward and reverse scans in the J-V curve. However, when the perovskite solar cell (PSC) including the UV-Nb:TiO.sub.2 electron transport layer showed the smallest hysteresis in the J-V curve, the hysteresis behavior was reduced.

[0127] The improved conductivity of the UV-treated electron transport layer (ETL) was proposed to reduce the hysteresis behavior by the balanced electrons and hole transport rate. The performance and photostability of the perovskite solar cell (PSC) including the UV-Nb:TiO.sub.2 electron transport layer during actual operation were evaluated by measuring the steady-state photocurrent density of the perovskite solar cell (PSC).

[0128] FIG. 12 shows the steady-state photocurrent density of the PSC including the UV-Nb:TiO.sub.2 electron transport layer.

[0129] As shown in FIG. 12, the perovskite solar cell (PSC) including the UV-Nb:TiO.sub.2 electron transport layer showed a stable power conversion efficiency (PCE) of 18.78% at the maximum power point voltage (0.91 V).

[0130] FIG. 13 shows UV photoelectron spectra (UPS) of the UV-Nb:TiO.sub.2 electron transport layer and the UV-TiO.sub.2 electron transport layer.

[0131] The Fermi level (E.sub.F) of each of the UV-TiO.sub.2 and the UV-Nb:TiO.sub.2 is obtained by subtracting the binding energy of secondary electron cut-off from the incident photon energy (21.2 eV). The E.sub.F values of the UV-TiO.sub.2 and UV-Nb:TiO.sub.2 are −4.06 eV and −4.13 eV, respectively.

[0132] As shown in FIG. 13, the energy gap between the E.sub.F and the top of valence band (E.sub.VB) was determined by the point of intersection between the bottom side and the linear portion of the spectrum around the Fermi edge. The E.sub.F-E.sub.VB values of the UV-TiO.sub.2 and the UV-Nb:TiO.sub.2 were 3.53 eV and 3.51 eV, respectively, and the bottom of conduction band (E.sub.CB) values of the UV-TiO.sub.2 and the UV-Nb:TiO.sub.2 were −3.97 eV and −4.02 eV, respectively, which were obtained by the addition of the band gap (3.62 eV) of the UV-TiO.sub.2 and the UV-Nb:TiO.sub.2 to their E.sub.VB values.

[0133] FIG. 14A is an energy level diagram of the individual layers of the perovskite solar cell (PSC), FIG. 14B shows steady-state photoluminescence (PL) spectra of the perovskite layers in contact with the HT-TiO.sub.2, UV-TiO.sub.2, and UV-Nb:TiO.sub.2 electron transport layers, and FIG. 14C shows time-resolved photoluminescence (TRPL) spectra of the perovskite layers in contact with the HT-TiO.sub.2, UV-TiO.sub.2, and UV-Nb:TiO.sub.2 electron transport layers.

[0134] As shown in FIG. 14A, the Nb—TiO.sub.2 has a deeper conduction band minimum (CBM) than TiO.sub.2 (−3.97 eV). The deep conduction band edge of the UV-Nb:TiO.sub.2 generates improved charge extraction compared to the UV-TiO.sub.2 to increase the driving force for electron injection from the perovskite conduction band to the electron transport layer (ETL).

[0135] The steady-state photoluminescence (PL) spectra of the perovskite layers were measured and time-resolved photoluminescence (TRPL) decay was performed to investigate the doping effect on charge transfer at the electron transport layer/perovskite interface.

[0136] As shown in FIG. 14B, the spectral peaks at 766 nm associated with the intrinsic fluorescence emission of CH.sub.3NH.sub.3PbI.sub.3 (perovskite) disappeared upon contact with the ETL, indicating charge transfer from the perovskite to the ETL.

[0137] Specifically, the photoluminescence (PL) density of the UV-TiO.sub.2/perovskite was lower than that of the HT-TiO.sub.2/perovskite (HT-TiO.sub.2/perov) and the PL density of the UV-Nb:TiO.sub.2/perovskite was lower than that of the UV-TiO.sub.2/perovskite. These results reveal that charge extraction from the perovskite is the most effective in the UV-Nb:TiO.sub.2 electron transport layer.

[0138] The PL lifetime was obtained by fitting the TR-PL spectra with a biexponential decay function (FIG. 14C). The fast decay is regarded as the quenching of carriers from the perovskite in contact with the electron transport layer and the slow decay is related to the radioactive decay in the perovskite layer.

[0139] Parameters for the glass/perovskite, HT-TiO.sub.2/perovskite, UV-TiO.sub.2/perovskite, and UV-Nb:TiO.sub.2/perovskite are shown in Table 2.

TABLE-US-00002 TABLE 2 Parameters a.sub.1 (%) τ.sub.1 (ns) a.sub.2 (%) τ.sub.2 (ns) τ.sub.avg Glass/perov 56.6 26.6 43.4 7.9 18.5 HT-TiO.sub.2/perov 56.1 16.0 43.9 2.5 10.1 UV-TiO.sub.2/perov 38.9 13.3 61.1 2.5 6.7 UV-Nb:TiO.sub.2/perov 37.6 12.6 62.4 1.8 5.8

[0140] As shown in Table 2 and FIG. 14C, charge extraction from the electron transport layers to the perovskite layer was dominant in the electron transport layer/perovskite structures compared to in the glass/perovskite. This fast decay decreased the decay lifetime accompanied by increased amplitude. The average lifetime of the HT-TiO.sub.2/perovskite layer was 10.1 ns. When the UV-TiO.sub.2 electron transport layer was used, the average lifetime was decreased to 6.7 ns.

[0141] The average decay lifetime of the UV-Nb:TiO.sub.2/perovskite layer was 5.8 ns, which was faster than that of the UV-TiO.sub.2/perovskite layer, and decreased from 2.5 nm to 1.8 ns with increased amplitude.

[0142] Therefore, it was confirmed that a significant improvement in the sunlight performance of the UV-Nb:TiO.sub.2 electron transport layer compared to the UV-TiO.sub.2 due to the ultraviolet photoelectron spectroscopy (UPS) and steady-state photoluminescence (PL) characteristics is attributed to improved charge extraction by a synergistic effect between energy level transfer and increased electrical conductivity. These results emphasize the importance of doping in the organic electron transport layer (ETL), which is because of the energy level alignment with perovskite in the high performance perovskite solar cell (PSC).

[0143] The novel method of the present invention is suitable for the preparation of an efficient electron transport layer (ETL) and is applicable to the fabrication of a high performance flexible perovskite solar cell (PSC). The UV-Nb:TiO.sub.2 electron transport layer of the present invention is suitable for use in a flexible perovskite solar cell (PSC).

[0144] FIG. 15A is a photograph of the PSC including the UV-Nb:TiO.sub.2 electron transport layer coated on an ITO/PEN substrate, FIG. 15B shows J-V curves of the flexible perovskite solar cell obtained after recovery from bending at different radii (r), and FIG. 15C shows photovoltaic parameters (J.sub.sc, V.sub.oc, FF, and PCE) of the flexible perovskite solar cell after bending at r=15 mm for 1000 cycles of bending.

[0145] FIG. 16 is a cross-sectional SEM image of the flexible perovskite solar cell (PSC) of FIG. 15A.

[0146] In the flexible perovskite solar cell (PSC) shown in FIG. 15A, the UV-Nb:TiO.sub.2 electron transport layer coated on an ITO/PEN substrate, a perovskite layer, spiro-OMeTAD (hole transport layer), and an Au electrode layer were deposited in this order, as shown in FIG. 16.

[0147] Photovoltaic parameters of the flexible perovskite solar cell after recovery from bending at different radii (r) are shown in Table 3.

TABLE-US-00003 TABLE 3 Parameters J.sub.sc (mAcm.sup.−2) V.sub.oc (V) FF PCE (%) r = ∞ 20.20 1.04 0.76 16.01 r = 10 mm 19.71 1.04 0.76 15.58 r = 7 mm 19.33 1.02 0.67 13.27 r = 4 mm 12.02 0.97 0.38 4.41

[0148] As shown in Table 3 and FIG. 15B, the initial power conversion efficiency (PCE), J.sub.sc, V.sub.oc, and FF of the flexible perovskite solar cell (PSC) were 16.01%, 20.20 mA cm.sup.−2, 1.04 V, and 0.76, respectively.

[0149] The power conversion efficiency (PCE) values of the PSC after bending at r=10 mm, 7 mm, and 4 mm were 15.58%, 13.27%, and 4.41%, respectively. The weak reduction in efficiency at r=7 mm is attributed to the plastic deformation of the solar cell. As previously reported, the soft ITO was broken at r=4 mm.

[0150] As shown in FIG. 15C, the initial efficiency of the flexible perovskite solar cell (PSC) was 15.42%. After 1000 cycles of bending, the efficiency was decreased to 14.45%, which corresponds to 93.7% of the initial value. The other photovoltaic parameters (J.sub.sc, V.sub.oc, and FF) showed a similar tendency to decrease, with only a ˜2% reduction after 1000 cycles of bending.

[0151] Therefore, the mechanical durability of the flexible perovskite solar cell (PSC) was demonstrated until 1000 cycles of bending at r=15 mm. All layers of the flexible perovskite solar cell (PSC) maintained their mechanical flexibility until r=15 mm.

[0152] In the present invention, the sol-gel derived TiO.sub.2 nanoparticles were treated with UV to photocatalytically decompose the organic ligand and induce the fusion of the nanoparticles, enabling the preparation of the TiO.sub.2 electron transport layer at low temperature. The TiO.sub.2 electron transport layer was used to fabricate the perovskite solar cell (PSC) with high performance.

[0153] The UV-Nb:TiO.sub.2 electron transport layer shows a higher transmittance and has a denser surface than the TiO.sub.2 film having undergone sintering at high temperature. Particularly, the Nb doping shows improved electrical conductivity and a downward shift of Fermi level, which is advantageous for charge extraction.

[0154] Specifically, the planar perovskite solar cell (PSC) employing the UV-Nb:TiO.sub.2 as an electron transport layer (ETL) showed high power conversion efficiencies (PCE) of 16.01% for a flexible ITO/PEN substrate and 19.57% for hard FTO glass.

[0155] —Fabrication of Perovskite Solar Cells—

[0156] Perovskite Layer

[0157] 1 mmol of PbI.sub.2 (99.9985%, Alfar Aesar), 1 mmol of CH.sub.3NH.sub.3I (Dyesol), and 1 mmol of dimethyl sulfoxide (DMSO, 99.9%, Aldrich) were added to 600 mg of N,N-dimethylformamide (DMF, 99.8%, Alfa Aesar) to prepare a perovskite solution. To the solution was added dropwise diethyl ether with two-stage rapid stirring. The mixture was slowly spin-coated on each of the electron transport layers prepared in Example 1 and Comparative Examples 1-2 at 1000 rpm and 4000 rpm for 20 sec. Thereafter, the coated electron transport layer was heated at 100° C. for 3 min to form a dark perovskite layer.

[0158] Hole Transport Layer

[0159] A mixture of 2,2′,7,7′-tetrakis(N,N-p-dimethoxyphenyl-amine)-9,9′-spirobifluorene (56 mg, spiro-OMeTAD, Merck), 4-tert-butylpyridine (30 mg, 96%, Aldrich), chlorobenzene (1 ml, 99.8%, Aldrich), and bis(trifluoromethane)sulfonimide lithium salt (6 mg, 99.95%, Aldrich) was coated on the perovskite layer at 2500 rpm for 20 s.

[0160] Electrode Layer

[0161] An 80 nm thick Au electrode was formed on the hole transport layer by shadow mask thermal evaporation.

[0162] —Instruments—

[0163] The morphologies of the TiO.sub.2 nanoparticles and the films were measured by field emission scanning electron microscopy (FESEM, Inspect F, FEI), high-resolution transmission electron microscopy (HR-TEM, Talos F200X, FEI), and atomic force microscopy (AFM, XE-100, Park Systems).

[0164] The crystal structure of the TiO.sub.2 nanoparticles was characterized using a powder X-ray diffractometer (XRD, D/MAX-2500, Rigaku) with Cu Kα radiation (wavelength 1.541 Å).

[0165] The compositions of the UV-treated TiO.sub.2 films were investigated using an ATR-FTIR spectrometer (Spectrum 100, PerkinElmer), and XPS and UPS measurements were conducted using a monochromator Al Kα (1486.6 eV) for XPS and a scanning XPS microprobe (PHI 5000 VersaProbe, Ulvac-PHI) with HeI (21.2 eV) for UPS, respectively.

[0166] For the measurement of electrical conductivity, a Keithley 2182A nanovoltmeter and a four-point probe connected to a Keithley 6220 precision current source were used. The optical properties of the TiO.sub.2 thin films was measured using a UV-vis spectrophotometer (Lambda 35, Perkin Elmer).

[0167] Electrochemical workstation (CHI600C, CH Instruments, Inc.) was used for cyclic voltammetry (CV) measurement. The cyclic voltammograms were recorded in a 3-electrode configuration using a Ag/AgCl reference electrode and a Pt wire counter electrode. The CV electrolyte was an aqueous solution containing 0.5 mM K.sub.4Fe(CN).sub.6, 0.5 mM K.sub.3Fe(CN).sub.6, and 0.5 M KCl.

[0168] The current density-voltage (J-V) curves were measured using a Keithley model 2400 source measurement unit and a solar simulator equipped with a 1000 W xenon lamp (Yamashida Denso, YSS-50S).

[0169] A Si solar cell calibrated by the National Renewable Energy Laboratory (NREL) was used to adjust the light intensity to the AM 1.5G 1 sun condition (100 mW cm.sup.−2). During all J-V measurements, non-reflective black masks were attached to the devices to exclude diffused light scattering.

[0170] The active area of each device (0.1-0.15 cm.sup.2) was measured using an optical microscope.

[0171] The incident photon-to-current conversion efficiency (IPCE) was measured using an IPCE measurement system (PV Measurement, Inc.). The steady-state photoluminescence (PL) spectra of perovskite films were obtained by excitation at 550 nm using a Fluorolog3 photoluminescence spectrometer system with a monochromator (iHR320, HORIBA Scientific).

[0172] Time-resolved photoluminescence (TR-PL) measurements were performed at the PL maxima of the CH.sub.3NH.sub.3PbI.sub.3 perovskite, using time-correlated single photon counting module (TCSPC, MPD-PDM Series DET-40 photon counting detector and Pendulum CNT-91 frequency counter) combined with a monochromator as the detector, and a second-harmonic generated 400 nm laser as the excitation source derived from 800 nm Ti: sapphire laser (Mai Tai, Spectra-Physics).