Surface Modification of Al-Doped ZnO Transparent Conducive Thin

Jun 25, 2019 - Figure 6 shows the J–V curves of the a-Si:H solar cells deposited on ... modifiers leads to an apparent change in the series resistan...
0 downloads 0 Views 4MB Size
Research Article www.acsami.org

Cite This: ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Surface Modification of Al-Doped ZnO Transparent Conducive Thin Films with Polycrystalline Zinc Molybdenum Oxide Lei Meng,*,†,‡ Xiaoguang Yang,†,‡ Hongyu Chai,†,‡ Zunren Lv,†,‡ and Tao Yang*,†,‡ †

Downloaded via UNIV OF SOUTHERN INDIANA on July 18, 2019 at 01:11:12 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

Key Laboratory of Semiconductor Materials Science, Beijing Key Laboratory of Low Dimensional Semiconductor Materials and Devices, Institute of Semiconductors, Chinese Academy of Sciences, Beijing 100083, People’s Republic of China ‡ College of Materials Science and Optoelectronic Technology, University of Chinese Academy of Sciences, Beijing 100049, People’s Republic of China S Supporting Information *

ABSTRACT: High-work function (WF) transparent conductive thin films improve the performance of solar cells and organic light-emitting diodes by facilitating interfacial charge carrier transport. Al-doped ZnO (AZO) becomes a very promising transparent conductive material because of nontoxicity, abundant material resources, and low cost. To increase the WF of AZO without enhancing the series resistance of the device, a high-WF and low-resistance surface modifier of polycrystalline zinc molybdenum oxide (ZMO) was developed by utilizing thermal evaporation of MoO3 on the surface of AZO and a subsequent two-step annealing treatment. The first step of air annealing causes the formation of monoclinic ZnMoO4 nanocrystals in the ZMO modifier. This improves the WF of AZO from 3.83 to 4.86 eV by increasing the group electronegativity and cation oxidation state. Furthermore, the second step of N2 annealing decreases the resistivity of the polycrystalline ZMO by increasing the donor states of oxygen vacancies. The surface modification effect is verified by applying the ZMO-modified AZO to the front electrode of hydrogenated amorphous silicon thinfilm solar cells. The low-resistance polycrystalline ZMO modifier not only increases light harvesting in the solar cells by improving interfacial refractive index matching but also improves the open-circuit voltage by modifying the interfacial band alignment. In particular, the modifier increases the fill factor by ca. 13% by reducing the series resistance of the device. These enable a gain of ca. 23% in photoelectric conversion efficiency compared to the unmodified AZO. The results suggest the feasibility to tune the WF and conductivity of a material independently. KEYWORDS: Al-doped ZnO, work function, surface modification, low-resistance, thin-film solar cells including metal mesh conductors,6 silver nanowires,7 carbon nanotubes,8 graphene,9 and conductive polymers,10 have gained considerable attention because of amazing specific characteristics. However, none of these materials have been successful enough to replace ITO and FTO for many reasons, including manufacturability, cost, and stability. Noticeably, Aldoped ZnO (AZO) becomes a potent candidate to replace ITO and FTO because of its nontoxicity, abundant material resources, low cost, and high resistance to hydrogen-plasma reduction.11 The price of the zinc metal is ca. $3/kg as of December 2018 (Data from Fastmarkets). The surface work function (WF) is a very important parameter for TCO materials because it determines the band alignment and charge transport at the interface between the TCO and p-type semiconductor material in optoelectronic devices. In the field of OLED and thin-film solar cells, a high WF TCO is required because it improves the hole collection or injection efficiency by raising the energy band of the adjacent

1. INTRODUCTION Transparent conducting oxides (TCOs) are a key material in a wide variety of current technologies, including organic lightemitting diodes (OLEDs),1 solar cells,2 smart windows,3 and infrared-reflecting glass.4 At present, Sn-doped In2O3 (ITO) and F-doped SnO2 (FTO) are the most commercially available TCOs due to their excellent electrical and optical properties. However, both of them suffer drawbacks that restrict their application to next-generation optoelectronic devices. The main drawbacks of ITO are unsustainable supply and high cost. The abundance of indium in the upper continental crust of the Earth is just 0.55 ppm that of silicon; thus, ITO would suffer a shortage of material resources sooner or later. Furthermore, the production cost of ITO is very high because the production of it requires expensive indium metal, whose price is more than $300/kg as of October 2018 (Data from Kitco Metals). FTO shows a low stability in the hydrogen-rich plasma produced during plasma-enhanced chemical vapor deposition.5 The reduced tin atom diffuses into the absorption layer in solar cells, resulting in deterioration of the device performance. Consequently, there is an effort to find an ideal replacement for ITO and FTO. Emerging alternatives, © XXXX American Chemical Society

Received: May 7, 2019 Accepted: June 25, 2019 Published: June 25, 2019 A

DOI: 10.1021/acsami.9b07977 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces Table 1. Summary of Surface Modification Methods To Increase the WF of ZnO method surface cleaning

active oxygen

organic adsorption

TMO adsorptionc

key part

subject/preparation method

beforea [eV]

aftera [eV]

gain

acetone Ar+ sputtering Ar+ sputtering UV−ozone treatment UV−ozone treatment oxygen plasma treatment F13OPAb PTCDIb F4TCNQb PO3-CNb F5PPAb MoO3−x WO3−x V2O5−x

AZO thin film/RFMS ZnO thin film/RFMS ZnO thin film/ALDb ZnO thin film/RFMS AZO thin film/RFMS ZnO thin film/RFMS SC ZnO(0001)-Znc ZnO thin film/ALD ZnO(0001)-Zn/MBEb ZnO thin film/CBDb GZO thin film/RFMS AZO thin film/RFMS AZO thin film/RFMS AZO thin film/RFMS

3.83 3.74 3.7

3.94 3.95 4.5

3.7 3.74 3.9 3.6 3.7 3.9 3.29 4.1 4.1 4.1

4.03 4.21 5.6 4.5 6.5 4.5 4.87 5.55 4.7 5.0

3%19 6%20 21.6%21 0.25 eV22 8.9%23 12.6%20 42.7%24 25%26 75.7%27 15.4%28 48%29 35.4%30 14.6%32 22%32

b

The values in the “before” and “after” columns are the WF measured before and after surface modification bF13OPA: 3,3,4,4,5,5,6,6,7,7,8,8,8tridecafluorooctyl phosphonic acid, PTCDI: 3,4,9,10-perylenetetracarboxylicdiimide, F4TCNQ: 2,3,5,6-tetrafluoro-7,7,8,8-tetracyanoquinodimethane, PO3-CN: 4-cyanophenyl phosphonic acid, F5PPA: pentafluorophenyl phosphonic acid, RFMS: radio frequency magnetron sputtering deposition, ALD: atomic layer deposition, MBE: molecular-beam epitaxy, and CBD: chemical bath deposition. cSC: single-crystal, TMO: transition-metal oxides. a

p-type semiconductor material.12 The WF of AZO without a post-treatment varies from 3.7 to 4.62 eV,13,14 depending on the preparation method and WF measurement method. It is less than that of ITO (4.3−4.84 eV)15 and FTO (larger than 4.5 eV).16 Thus, increasing the WF of AZO is a very important topic. Surface modification is the usual strategy to increase the WF of TCOs.17,18 As summarized in Table 1, there are mainly four kinds of surface modification methods to increase the surface WF of AZO. First, the surface cleaning technique, including acetone solvent cleaning19 and argon ion (Ar+) sputtering cleaning,20 is a basic method to increase the WF by removing carbon-related contamination at the surface. Notably, Schulz et al. increased the WF from 3.7 to 4.5 eV by adopting Ar+ sputtering at a kinetic energy of 1 keV with 10 μA ion current on the surface of ZnO.21 Second, utilization of an UV−ozone or oxygen plasma treatment can improve the WF of ZnO by enhancing the amount of oxygen termination and activated oxygen species at the surface.22,23 This yielded a maximum increase of ca. 13% in the WF (3.74 eV → 4.21 eV).20 Third, a variety of organic surface modifiers were proposed to tune the WF through the formation of specific chemical molecule structures (interface dipoles) and binding modes on the surface.24−29 In particular, a WF of larger than 5.0 eV was achieved by adsorbing a molecular electron acceptor of F13OPA or F4TCNQ (Table 1) in the monolayer regime on a Zn-terminated single-crystal ZnO(0001) surface.24,27 However, this degree of enhancement has not been realized for polycrystalline ZnO transparent conductive thin films. Li et al. increased the WF of a Ga-doped ZnO (GZO) thin film from 3.29 to 4.87 eV through surface modification with a phosphonic acid of F5PPA;29 however, it has not been applied to any devices, and its effect on a device is still unknown. Finally, Jha et al. proposed the surface modification of AZO with physisorption of transition-metal oxides (e.g., MoO3−x, WO3−x, or V2O5−x) and initially applied them to OLEDs to improve the hole injection efficiency.30−32 Although a WF of 5.55 eV was achieved by utilizing the high WF characteristic of MoO3−x, the large resistivity (105 to 106 Ω·cm) of MoO3−x significantly increased the series resistance of OLEDs, resulting

in an increase of ca. 30% in the threshold voltage.30 This indicates that the high-resistance MoO3−x modifier increases the surface WF of AZO at the cost of increasing the series resistance of the device, resulting in the deterioration of the device performance. Consequently, it is imperative to develop a high WF and low-resistance surface modifier to improve the WF of AZO without increasing the series resistance of the device. Herein, we propose a novel surface modification method that produces a high WF and low-resistance modifier made up of zinc molybdenum oxide (ZMO), for increasing the WF of AZO without increasing the resistance of the surface layer. The ZMO modifier is prepared by thermal evaporation of MoO3 on the surface of AZO and a subsequent two-step annealing treatment. The major conclusions of this paper are as follows. (1) The crystallization of the ZMO modifier caused by an air annealing could improve the WF by increasing the group electronegativity (GE) and cation oxidation state. (2) The generation of extra oxygen vacancies in the modification layer caused by N2 annealing could reduce the resistivity of the polycrystalline ZMO modifier with a small decrease in WF. (3) The polycrystalline ZMO modifier embedded with monoclinic ZnMoO4 nanocrystals improves the performance of a-Si:H solar cells without increasing the series resistance of the device. First, the surface, structural, and optical properties of the ZMO surface modifiers were characterized in this study. Then, the surface modification effect was verified via the application of ZMO-modified AZO to hydrogenated amorphous silicon (aSi:H) thin-film solar cells. Finally, the nature of the ZMO modifiers and relevant mechanism were discussed.

2. EXPERIMENTAL SECTION 2.1. Preparation of ZMO Modifiers and a-Si:H Solar Cells. The AZO thin film was fabricated by a sol−gel process on soda-lime glass (SLG). The detailed preparation process of film coating can be found elsewhere.33 The AZO-coated substrate was first annealed in N2 at 400 °C for 120 min and then annealed in a forming gas (97% N2 + 3% H2) at 500 °C for 5 min to improve the conductivity.34 The final thickness of AZO was ca. 1500 nm. The sheet resistance of the AZO was ca. 15.6 Ω sq−1. Subsequently, a MoO3−x thin film with a thickness of ca. 30 nm was thermally evaporated onto the AZO B

DOI: 10.1021/acsami.9b07977 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces substrate from a MoO3 powder source (Sigma-Aldrich 99.97% purity) at a rate of ca. 0.1 nm/s under high vacuum (4 × 10−4 Pa). Then, the MoO3−x-coated AZO substrates received a two-step annealing treatment. In the first step, the samples were annealed in ambient air at 350 °C for 5 min. In the second step, the samples were annealed in N2 at 400 °C for 60 min. As illustrated in Table 2, three kinds of

point probe sheet resistance measurement using a Keithley 2401 source meter. Photoemission spectra were collected by a Thermo Fisher ESCALAB 250Xi system using monochromated Al Kα radiation (hν = 1486.7 eV) for X-ray photoelectron spectroscopy (XPS) measurement and nonmonochromated He Iα radiation (hν = 21.22 eV) for ultraviolet photoelectron spectroscopy (UPS) and WF measurements. In the UPS measurement, the sample was held at a negative bias of −10 V, and the analyzer was operated with an energy resolution of meV. The energy scale of the system was calibrated according to Au 4f7/2 peak at 83.96 ± 0.1 eV. The surface chemical composition and the valence state of the Mo ions in the thin films were analyzed by the XPS measurement. The current density versus voltage (J−V) characteristics of the solar cells were measured using a Keithley 2400 source meter under 1 sun (AM 1.5 G, 100 mW−1 cm2). The external quantum efficiency (EQE) was measured using an Enlitech measurement system. The J−V and EQE measurements were conducted using a mask with an illumination area of 0.09 cm2.

Table 2. Surface Modifiers for AZO Prepared in This Paper names

MoO3−x deposition

1st aira

2nd N2b

Ra [nm]

σrms [nm]

ZMO0 ZMO1 ZMO2 ZMO3

none 30 nm 30 nm 30 nm

no no yes yes

no no no yes

1.6 1.5 1.1 2.0

2.1 1.9 1.5 2.6

The first step of air annealing. bThe second step of N2 annealing.

a

ZMO (ZMO1, ZMO2, and ZMO3) modifiers were prepared. An AZO sample without deposition of MoO3−x on the top, and the twostep annealing treatment was designated as ZMO0 for convenience of discussion. In addition, a MoO3−x thin film with a thickness of ca. 30 nm was also thermally evaporated on a glass substrate with the same condition as the ZMO2. The superstrate-type a-Si:H single-junction solar cells were fabricated on these ZMO-modified AZO substrates. The structure of these solar cells was SLG/AZO/ZMO/p-a-SiC:H(10 nm)/i-aSi:H(300 nm)/n-a-Si:H(40 nm)/B-doped ZnO(1500 nm)/Ag(300 nm)/Al(600 nm). The p−i−n a-Si:H photoactive layers were prepared using the plasma-enhanced chemical vapor deposition method, while the 1500 nm B-doped ZnO was prepared using the metal organic chemical vapor deposition method. The metal electrode (Ag and Al) was prepared using a vacuum evaporation method. 2.2. Characterization. The X-ray diffractograms were measured using an X-ray diffractometer (Rigaku D/MAX-2500) with Cu Kα radiation generated at 40 kV55 mA. The grazing incidence X-ray diffraction (GAXRD) was measured with an incidence angle of 0.3°. Surface micrographs of the samples were obtained using fieldemission scanning electron microscopy (SEM) (FEI Nova NanoSEM650). The surface morphology and roughness were measured using atomic force microscopy (AFM) (Seiko NanoNavi E-Sweep). The interface microstructure and composition were studied by transmission electron microscopy (TEM) with energy-dispersive Xray (EDX) (JEM-2010F). For the cross-sectional observation, one sample comprising the a-Si:H solar cell with ZMO3 was prepared by the focused ion beam-milling technique (Zeiss Auriga). The transmittance spectra were measured using a UV−vis spectrophotometer (Cary 5000). The sheet resistance was evaluated with a four-

3. RESULTS 3.1. Characteristics of the ZMO Modifiers. Figure 1 shows the surface morphology micrographs of the ZMO modifiers on the AZO (ZMO0). All samples show very flat surfaces composed of nanoscale spherical grains. After formation of the ZMO modifiers, the size of the grains on the surface decreases and becomes uniform. The arithmetic average of the absolute values of the roughness (Ra) and rootmean-square surface roughness (σrms) measured by AFM of these samples are exhibited in Table 2. All films show a surface roughness of less than 3 nm. Both Ra and σrms decrease in the order of ZMO0, ZMO1, and ZMO2. ZMO3 shows a slightly larger Ra and σrms than ZMO0. The surface XPS analysis results of these ZMO modifiers are shown in Figure 2. The derived surface composition is shown in Table 3. ZMO1-3 modifiers mainly comprise Zn, Mo, and O elements. In Figure 2a, the presence of ZMO modifiers causes an obvious peak shift of Zn2p toward the large binding energy direction. In Figure 2b, it can be observed that the peak in the ZMO1 O 1s spectra contains two components (OI and OII). The OI peak is associated with the Zn−O bond in the wurtzite ZnO. The OII peak centered around 532 eV might be caused by the loosely bound oxygen created by the O2− ions in the oxygen-deficiency regions.35 It might also be originated from the contaminant of −OH or −CO3 groups.36 After formation of ZMO on AZO, only one peak appears in the O 1s spectra.

Figure 1. (a−d) SEM micrographs (up) and AFM images (down) of the surface of ZMO0, ZMO1, ZMO2, and ZMO3, respectively. The scale bars in (a−d) are 200 nm. C

DOI: 10.1021/acsami.9b07977 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

Figure 2. Surface XPS of (a) Zn 2p, (b) O 1s, and (c) Mo 3d spectra of the ZMO modifiers. The spectra were calibrated according to the standard C 1s peak position (284.4 ± 0.1 eV).

Table 3. Surface Composition Analyzed by the XPS Measurementa

Figure 3. X-ray diffractograms of the ZMO-modified AZO. The spectra of wurtzite ZnO (JCPDS card no. 36-1451) and Zn3Mo2O9 (JCPDS card no. 86-1771) are exhibited as a reference. The peaks marked by the unfilled rhombus are indexed to wurtzite ZnO, while the ones marked by the filled rhombus is indexed to orthorhombic MoO3. The patterns of MoO3−x, ZMO1, ZMO2, and ZMO3 are measured using GAXRD. The intensity of the ZMO0 pattern was intentionally reduced to exhibit it along with the patterns of other samples.

Mo 3d core levels (%) names

Zn/Mo

O/Zn

O/Mo

Mo6+

Mo5+

ZMO0 ZMO1 ZMO2 ZMO3

1.46 1.72 1.39

0.90 2.39 2.06 2.48

4.02 4.08 3.97

93.1 96.2 89.3

6.9 3.8 10.7

Note: the quantitative accuracy is ca. ±1 atom %.

a

ZMO3 not only show similar diffraction peaks as ZMO1 but also exhibit distinctive diffraction peaks (marked by X1 and X2) centered at approximately 27°. Figure 4 shows the transmittance spectra of the ZMOmodified AZO/glass substrates. All substrates show an average

In Figure 2c, Gauss−Lorentz deconvolution fitting was utilized to split the Mo 3d5/2 and Mo 3d3/2 peaks into two components (red and green regions). The red peaks centered at 231.2 and 233.6 eV originate from Mo5+ species, while green peaks centered at 232.4 and 234.3 eV are generated by Mo6+ species. The calculated component ratio of Mo6+ and Mo5+ is shown in Table 3. The annealing treatment causes obvious changes in the Mo6+/(Mo5+ + Mo6+) ratio among these ZMO modifiers. X-ray diffractograms of the ZMO-modified AZO are shown in Figure 3. The ZMO0 shows the (002)-oriented wurtzite ZnO structure. The difference in the crystallographic orientation between ZMO0 and wurtzite ZnO (JCPDS card no. 36-1451) should be ascribed to the film preparation process and the stoichiometry in the films. It was reported that the presence of multiple orientations is correlated to the substoichiometry in ZnO and AZO.37,38 The sol−gel-derived AZO thin films generally show (002) preferential orientation because of the minimum internal stress39 and a dipole−dipole interaction between the polar nanograins.40 The ZMO1 exhibits a GAXRD pattern that can only be indexed to the spectrum of wurtzite ZnO. The diffraction peaks on the ZMO1 should mainly originate from the underlying AZO. It should be the measuring mode that generates a difference in the diffractograms between the ZMO0 and ZMO1 samples. The diffractogram of ZMO0 was measured using the conventional θ/2θ scanning method, which generally produces a weak signal from the surface layer and an intense signal from the substrate (i.e., AZO). In addition, the diffractogram of ZMO1 was measured using a 2θ scan with a fixed grazing angle of incidence (0.3°), which can get stronger signal from the surface layer and weaker signal from the underlying AZO. The air-annealed MoO3−x, which was deposited on a glass substrate, shows a GAXRD pattern that can be indexed to the spectrum of orthorhombic MoO3.41 Notably, ZMO2 and

Figure 4. Transmittance spectra of the ZMO-modified AZO/glass substrates.

transmittance of over 85% in the wavelength region of 400− 1000 nm. The presence of ZMO modifiers has little influence on the transmittance at wavelengths larger than 400 nm. However, the transmittance of AZO/glass substrates with ZMO2 and ZMO3 at the wavelength region of 350−400 nm is lower than that of the substrates with ZMO0 and ZMO1. This reflects the change in the bandgap of the ZMO modifiers and/ or the underlying AZO. In addition, all substrates show a wavelike fluctuation of transmittance with wavelength. This is the thin-film interference phenomenon that was caused by the flat surface morphology. The surface WF of these ZMO modifiers was measured using UPS. In Figure 5a, the presence of ZMO modifiers causes an obvious shift of the secondary-electron cut-off edge D

DOI: 10.1021/acsami.9b07977 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

curves of the a-Si:H solar cells deposited on these substrates, and Table 4 exhibits the corresponding performance parameters. ZMO1 generates a bit smaller short-circuit current density (Jsc) than ZMO0. ZMO2 and ZMO3 enable larger Jsc than ZMO0 and ZMO1. Furthermore, ZMO3 significantly increases the open-circuit voltage (Voc) and fill factor (FF), resulting in a gain of ca. 23% in photoelectric conversion efficiency (η) compared to ZMO0. ZMO3 enables a Voc of ca. 0.9 eV, which is comparable to the state-of-the-art a-Si:H single-junction solar cells.42 Moreover, the presence of ZMO modifiers leads to an apparent change in the series resistance (Rs). Compared with ZMO0, ZMO1 and ZMO2 increase Rs by more than two times while ZMO3 decreases the Rs apparently. ZMO3 realizes a gain of ca. 13% in FF compared to ZMO0-2. This should be mainly ascribed to the fact that ZMO3 enabled the smallest Rs of ca. 87.2 Ω. In addition, the presence of ZMO modifiers causes little on the shunt resistance (Rsh). The EQE and “1 − reflectance” curves of the a-Si:H solar cells are shown in Figure 7. Compared to ZMO0 and ZMO1, ZMO2 and ZMO3 enable a higher EQE and lower reflectance in the wavelength region of 420−640 nm but a lower EQE at the wavelength region of 350−420 nm.

Figure 5. UPS spectra of the ZMO modifiers. (a) Secondary-electron cut-off and (b) valence band. The measurement error of WF is ca. ±0.1 eV.

toward the high energy direction, indicating a change in the WF. The derived WF values of the ZMO modifiers are shown along each spectrum. The ZMO0, that is, sol−gel-derived AZO, shows a WF of 3.83 eV, which is comparable to that of the AZO prepared by the magnetron sputtering method (Table 1).19 ZMO1 yields a gain of ca. 12% in WF with respect to ZMO0, realizing a WF of 4.30 eV. Furthermore, ZMO2 achieves a WF of 4.86 eV, which increases by ca. 27 and 13% compared to that of ZMO0 and ZMO1, respectively. Furthermore, ZMO3 shows a WF of 4.85 eV, that is, almost identical to that of ZMO2. The WF of ZMO2 and ZMO3 is comparable to that of commercialized ITO and FTO. In Figure 5b, the presence of ZMO modifiers causes new peaks on the valence band spectra, labeled d at the region near the Fermi level (EF). ZMO2 has a weaker d peak than ZMO1 and ZMO3, while ZMO3 exhibits the strongest d peak centered at approximately 2 eV. 3.2. Application to a-Si:H Solar Cells. These ZMOmodified AZO substrates were applied to the front electrode of the a-Si:H single-junction solar cells. Figure 6 shows the J−V

Figure 7. EQE and “1 − reflectance” curves of the a-Si:H solar cells deposited on the ZMO-modified AZO substrates.

Figure 8a shows the cross-sectional TEM image of the aSi:H solar cell deposited on the ZMO3-modified AZO substrate. It can be observed that very flat and uniform p− i−n a-Si:H photoactive layers were formed. The sol−gelderived AZO shows a layered structure, in which the thickness of each layer is ca. 30 nm. The layered structure was formed because of the sol−gel process. In this work, the AZO thin film is thickened by increasing the cycle numbers of spin-coating and heating. In Figure 8b, a thin layer with a thickness of ca. 30 nm was formed at the interface between the AZO and p-aSiC:H layer. Furthermore, the EDX line scan and spot analysis were conducted to determine chemical composition. In Figure 8c, the elemental profile derived along the scanning line A in Figure 8b shows an abrupt change in Mo composition at the

Figure 6. J−V curves of a-Si:H solar cells deposited on ZMOmodified AZO substrates.

Table 4. Performance Parameters of a-Si:H Solar Cells Deposited on ZMO-Modified AZO Substratesa samples

Jsc (mA/cm2)

ZMO0 ZMO1 ZMO2 ZMO3

11.17 10.88 11.90 11.94

(0.24) (0.22) (0.34) (0.14)

Voc (V) 0.88 0.88 0.88 0.90

(0.005) (0.009) (0.003) (0.003)

η (%)

FF () 0.54 0.55 0.54 0.61

(0.04) (0.04) (0.01) (0.01)

5.32 5.32 5.62 6.54

(0.50) (0.48) (0.20) (0.13)

Rs (Ω) 115.0 (11.9) 252.7 (18.3) 312.3 (10.8) 87.2 (8.0)

Rsh (Ω) 1.3 1.6 1.3 1.7

× × × ×

104 104 104 104

(5.3 (4.9 (2.6 (1.5

× × × ×

103) 103) 103) 103)

a

Note: the data outside the bracket and the data inside the bracket are the average and standard deviation over three cells deposited on the same substrate, respectively. E

DOI: 10.1021/acsami.9b07977 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

Figure 8. (a) Low-magnification cross-sectional TEM image of the a-Si:H solar cell deposited on a ZMO3-modified AZO substrate. (b) Highmagnification TEM image of the white-boxed region in (a), with the HRTEM image of region I and SAED pattern of region (II) in (b). (c) Elemental profile along line A in (b).

annealing causes interdiffusion of Zn and Mo elements at the MoO3−x/AZO interface. According to the cross-section TEMEDX elemental mapping of ZMO3 (Figure S1), the Mo element shows a diminishing distribution from the MoO3−x/ AZO interface to AZO, confirming the diffusion of the Mo element toward AZO. Interdiffusion of Zn and Mo elements might prompt an adequate reaction between ZnO and MoO 3−x . This might result in formation of ZnMoO 4 nanocrystals (Figure 8). Therefore, ZMO2 and ZMO3 are polycrystalline ZMO embedded with ZnMoO4 nanocrystals. Furthermore, the X1 and X2 diffraction peaks on the GAXRD pattern of the ZMO2 and ZMO3 in Figure 3 are probably associated with Zn3Mo2O9 (JCPDS card no. 86-1771). Moreover, according to the XPS and TEM−EDX composition analysis exhibited in Tables 3 and 5, the main component in the ZMO2 and ZMO3 modifiers is quite probably oxygendeficient Zn3Mo2O9 (Zn3Mo2O9−δ). The main difference between ZMO2 and ZMO3 probably lies in film resistivity. As exhibited in Table 4, the Rs of a-Si:H solar cells deposited on ZMO2 is ca. 3.6 times larger than that of a-Si:H solar cells deposited on ZMO3. This reflects that there is a big difference in the film resistance between ZMO2 and ZMO3, taking account of the same device structure and similar WF. As shown in Figure 5b, ZMO3 exhibits a stronger peak intensity of the d-gap state than ZMO2, implying generation of much more defect states. These d-gap states should mainly be the oxygen-vacancy defects produced during the deposition of the MoO3−x and the annealing process.43 The defect reaction is shown in eq 1 using Kröger−Vink notation

AZO/p-a-SiC:H interface. The derived elemental composition of the selected A, B, and C spots in Figure 8b is shown in Table 5. These spot compositions demonstrate that this Table 5. TEM−EDX Elemental Composition of Selected Spots Shown in Figure 8b at the AZO/a-SiC:H Interfacea spot position

Zn (%)

Mo (%)

O (%)

Zn/Mo

O/Zn

O/Mo

A B C average

16.94 11.67 20.62 16.41

11.42 6.34 11.8 9.85

41.91 31.05 39.46 37.47

1.48 1.84 1.75 1.69

2.47 2.66 1.91 2.28

3.67 4.90 3.34 3.97

Note: the measurement error of the EDX is ca. ±1 atom %.

a

interfacial layer is composed of Zn, Mo, and O elements with similar Zn/Mo, O/Zn, and O/Mo ratios as that are shown in Table 3. Besides, the TEM−EDX elemental mapping (Figure S1) shows that the distribution of Zn, Mo, and O elements in the interfacial layer is almost uniform. In addition, the highresolution TEM (HRTEM) image of region I and selected area electron diffraction (SAED) pattern of region II in Figure 8b reveal that certain nanocrystals were formed in the ZMO3 modification layer. Based on the analysis of the interplanar spacing and intersection angle, the nanocrystals are likely monoclinic ZnMoO4 (JCPDS card no. 25-1024).

4. DISCUSSION 4.1. Nature of the ZMO Modifiers. Compared to ZMO0, ZMO1-2 shows an obvious shift in the Zn 2p peaks toward a larger binding energy (Figure 2). This indicates that a slight oxidation of Zn atoms occurred. The ZMO1 should be an amorphous phase because almost no phase was detected in the modification layer during the X-ray diffraction measurement (Figure 3). At present, it is difficult to deduce that ZMO1 is a ternary ZMO compound or a composite of ZnO and MoO3. ZnO can react with MoO3 at a temperature of ca. 270 °C, producing ZnMoO4. Thus, the reaction between ZnO and MoO3−x should occur during the first step of air annealing where temperature is 350 °C. Figure S2 shows the relative change in the elemental depth profile of the ZMO-modified AZO substrates. With respect to ZMO1, there are an obvious increase in the Mo content on the AZO side and an obvious increase in the Zn content on the ZMO side in the ZMO2modified and ZMO3-modified AZO. This reflects that the

× 5+ OO + 2Mo6 + F V •• + O + 2Mo

1 O2 (g) 2

(1)

The N2 atmosphere in the second step of annealing decreases the oxygen partial pressure in the furnace chamber and prompts the reaction toward the right, resulting in the increase in the quantity of oxygen vacancies.34 Furthermore, ZMO3 shows a larger Mo5+/(Mo5+ + Mo6+) ratio than ZMO2 at the surface (Table 3), suggesting the occurrence of the above defect reaction. Consequently, the increased donor states of oxygen vacancy in ZMO3 decreases the resistivity of the modification layer by increasing the free carrier concentration. F

DOI: 10.1021/acsami.9b07977 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces 4.2. Improvement Mechanism of the WF. The ZMO modifiers effectively increase the surface WF of the AZO. Formation of ZMO at the interface enables a WF increase of ca. 12% over ZMO0, while the formation of ZnMoO4 nanocrystals in the ZMO modification layer causes a WF increase of ca. 13% over the amorphous ZMO1. As illustrated in Figure S3, the WF is defined as the energy required for moving an electron from the EF of a material to the vacuum level. According to the concept of electronegativity proposed by Mulliken, GE influences the WF value by affecting the midgap position.44 Thus, GE is a good approximation of the EF for undoped semiconductors.43 For a ternary compound with the formula AaBbCc, the GE could be calculated using the geometric mean of the electronegativity of the constituent elements, as shown in eq 2 GE = (χA a χBb χCc )1/(a + b + c)

Figure 9. Schematic band diagram of a-Si:H solar cells deposited on the AZO and ZMO3-modified AZO substrates under an illuminated open-circuit condition. The red lines refer to the band alignment at this interface after the introduction of the ZMO3 modifier. Φb, EVac, ECBM, and EVBM denote the potential barrier height at the TCO/p-aSiC:H interface, vacuum level, conduction band minimum, and valence band maximum, respectively.

(2)

where χA−χC is the electronegativity of the constituent element and A−C and a−c represent stoichiometric coefficients.45,46 Accordingly, the GE values of ZnO, Zn3Mo2O9, and ZnMoO4 are 7.7, 8.83, and 9.02 eV, respectively (χZn = 5.53 eV, χMo = 7.04 eV, and χO = 10.85 eV).49 The presence of oxygen deficiency in the ZMO modifiers does not substantially influence the GE. Thus, if the ZMO modifiers are approximatively regarded as undoped semiconductors, that is, GE is ca. equal to EF, the main improvement mechanism of WF in this paper would be the increase in the GE of the compound. In addition, the GE of an oxide increases with the cation oxidation state due to the dependence of the electronegativity of the atom on its valence state.46 As shown in Figure 2c and Table 3, air annealing causes an obvious increase in the Mo6+/(Mo5+ + Mo6+) ratio for ZMO2 compared to that of ZMO1. This also contributes to the increase in WF. 4.3. Reason Jsc Is Enhanced. ZMO2 and ZMO3 enhance the EQE of a-Si:H solar cells in the 420−640 nm wavelength region and increase the Jsc. This should be related to the improved refractive index matching at the AZO/p-a-SiC:H interface. As shown in Figure S4, there is a big refractive index mismatching between AZO and p-a-SiC:H. At present, there are few published data about the refractive index of Zn3Mo2O9 and ZnMoO4. Because derived from the reaction between AZO and MoO3−x, the ZMO2 and ZMO3 should have refractive index with values somewhere in between AZO and the air-annealed MoO3−x. It was demonstrated that the airannealed MoO3−x shows an increased refractive index compared to the as-deposited one (Figure S4). Consequently, ZMO2 and ZMO3 improved the interfacial refractive index matching at the front side of the a-Si:H solar cells. This reduces the reflectance of the solar cells and then enhances light harvesting in the absorption layer (Figure 7), resulting in the increase in the Jsc. 4.4. Improvement Mechanism of the Voc. ZMO3 clearly increases the Voc. This improvement should be attributed to the modified band alignment at the AZO/p-aSiC:H interface. As illustrated in Figure 9, the utilization of AZO without surface modification (i.e., ZMO0) causes an obvious downward band bending of the p-a-SiC:H at the interface because of the insufficient WF of 3.83 eV.47 This forms a potential barrier (Φb) at the AZO/p-a-SiC:H interface.48 The Φb impedes hole collection across the interface and then accumulates some holes at the interface.

The accumulated holes would inversely flow toward the i-a-Si layer, forming a “hole injection”. This would increase the carrier recombination loss in the p-a-SiC:H layer. The Voc increases with decreasing the carrier recombination loss.49 Compared with ZMO0, the introduction of ZMO3 with a WF of 4.85 eV at the AZO/p-a-SiC:H interface raises the band of p-a-SiC:H at the AZO/p-a-SiC:H interface and reduces the potential barrier height at the interface. This contributes to increasing hole collection across the interface and decreases hole accumulation at the interface. This results in the enhancement in the Voc through reducing the carrier recombination loss. Although ZMO2 possesses a similar WF as ZMO3, the utilization of ZMO2 does not enable an improvement in the Voc, as was the case for ZMO3. At present, the reason is not clear. One possible reason is attributed to the difference in the carrier collection mechanism across the modification layer. Assuming ZMO2 and ZMO3 are n-type semiconductors, an n−p junction might form at the interface between them and the p-a-SiC:H. Then, the built-in electric field that forms in the depletion region blocks hole collection from the p-a-SiC:H to the AZO electrode. This offsets the Voc improvement caused by the utilization of a high WF modifier (Figure 9). However, in the case of ZMO3, the blocking effect generated by the built-in electric field of the n−p junction is likely weakened because of the presence of more defect states in the modification layer. As discussed in Section 4.1, ZMO3 has more oxygen vacancy defects than ZMO2. The increased defect states might promote hole collection from the p-aSiC:H to the AZO electrode through a hopping mechanism.50 This might counteract the blocking effect generated by the built-in electric field of the n−p junction to a certain extent, enabling the high WF modifier-induced Voc improvement to work. An increased ZMO modification effect is attainable by further increasing the GE of the ZMO. This could be achieved through a more elaborate control of the composition and the cation oxidation state in the ZMO by using optimized annealing routes or advanced preparation methods, such as magnetron sputtering or pulsed laser deposition. This could further improve the performance of photovoltaic devices by G

DOI: 10.1021/acsami.9b07977 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

(4) Hamberg, I.; Granqvist, C. G. Transparent and Infraredreflecting Indium-tin-oxide Films: Quantitative Modeling of the Optical Properties. Appl. Opt. 1985, 24, 1815−1819. (5) Schade, H.; Smith, Z. E.; Thomas, J. H., III; Catalano, A. Hydrogen Plasma Interactions with Tin Oxide Surfaces. Thin Solid Films 1984, 117, 149−155. (6) Aryal, M.; Geddes, J.; Seitz, O.; Wassei, J.; McMackin, I.; Kobrin, B. Sub-Micron Transparent Metal Mesh Conductor for Touch Screen Displays. In Society for Information Display Symposium: San Diego, CA, USA, 2014; Vol. 45, pp 194−196. (7) Hu, L.; Kim, H. S.; Lee, J.-Y.; Peumans, P.; Cui, Y. Scalable Coating and Properties of Transparent, Flexible, Silver Nanowire Electrodes. ACS Nano 2010, 4, 2955−2963. (8) Wu, Z.; Chen, Z.; Du, X.; Logan, J. M.; Sippel, J.; Nikolou, M.; Kamaras, K.; Reynolds, J. R.; Tanner, D. B.; Hebard, A. F.; Rinzler, A. G. Transparent, Conductive Carbon Nanotube Films. Science 2004, 305, 1273−1276. (9) Wassei, J. K.; Kaner, R. B. Graphene, A Promising Transparent Conductor. Mater. Today 2010, 13, 52−59. (10) Vosgueritchian, M.; Lipomi, D. J.; Bao, Z. Highly Conductive and Transparent PEDOT:PSS Films with a Fluorosurfactant for Stretchable and Flexible Transparent Electrodes. Adv. Funct. Mater. 2012, 22, 421−428. (11) Kang, J.; Kim, H. W.; Lee, C. Electrical Resistivity and Transmittance Properties of Al- and Ga-codoped ZnO Thin Films. J. Korean Phys. Soc. 2010, 56, 576−579. (12) Meyer, J.; Hamwi, S.; Kröger, M.; Kowalsky, W.; Riedl, T.; Kahn, A. Transition Metal Oxides for Organic Electronics: Energetics, Device Physics and Applications. Adv. Mater. 2012, 24, 5408−5427. (13) Jiang, X.; Wong, F. L.; Fung, M. K.; Lee, S. T. Aluminumdoped Zinc Oxide Films as Transparent Conductive Electrode for Organic Light-emitting Devices. Appl. Phys. Lett. 2003, 83, 1875− 1877. (14) Kim, T. W.; Choo, D. C.; No, Y. S.; Choi, W. K.; Choi, E. H. High Work Function of Al-doped Zinc-oxide Thin Films as Transparent Conductive Anodes in Organic Light-emitting Devices. Appl. Surf. Sci. 2006, 253, 1917−1920. (15) Sharma, A.; Kippelen, B.; Hotchkiss, P. J.; Marder, S. R. Stabilization of the Work Function of Indium Tin Oxide using Organic Surface Modifiers in Organic Light-emitting Diodes. Appl. Phys. Lett. 2008, 93, 163308. (16) Robertson, J.; Falabretti, B. Electronic Structure of Transparent Conducting Oxides. In Handbook of Transparent Conductors; Ginley, D. S., Ed.; Springer: Boston, MA, USA, 2010; pp 27−50. (17) Sugiyama, K.; Ishii, H.; Ouchi, Y.; Seki, K. Dependence of indium-tin-oxide work function on surface cleaning method as studied by ultraviolet and x-ray photoemission spectroscopies. J. Appl. Phys. 2000, 87, 295−298. (18) Hotchkiss, P. J.; Li, H.; Paramonov, P. B.; Paniagua, S. A.; Jones, S. C.; Armstrong, N. R.; Brà ©das, J.-L.; Marder, S. R. Modification of the Surface Properties of Indium Tin Oxide with Benzylphosphonic Acids: A Joint Experimental and Theoretical Study. Adv. Mater. 2009, 21, 4496−4501. (19) Wang, W.; Feng, Q.; Jiang, K.; Huang, J.; Zhang, X.; Song, W.; Tan, R. Dependence of Aluminum-doped Zinc Oxide Work Function on Surface Cleaning method as Studied by Ultraviolet and X-ray Photoelectron Spectroscopies. Appl. Surf. Sci. 2011, 257, 3884−3887. (20) Kuo, F.-L.; Li, Y.; Solomon, M.; Du, J.; Shepherd, N. D. Workfunction Tuning of Zinc Oxide Films by Argon Sputtering and Oxygen Plasma: An Experimental and Computational study. J. Phys. D: Appl. Phys. 2012, 45, 065301. (21) Schulz, P.; Kelly, L. L.; Winget, P.; Li, H.; Kim, H.; Ndione, P. F.; Sigdel, A. K.; Berry, J. J.; Graham, S.; Brédas, J.-L.; Kahn, A.; Monti, O. L. A. Tailoring Electron-Transfer Barriers for Zinc Oxide/ C60Fullerene Interfaces. Adv. Funct. Mater. 2014, 24, 7381−7389. (22) Tang, W. M.; Greiner, M. T.; Lu, Z. H.; Ng, W. T.; Nam, H. G. Effects of UV-ozone Treatment on Radio-frequency Magnetron Sputtered ZnO Thin Films. Thin Solid Films 2011, 520, 569−573.

optimizing the interfacial band alignment and promoting hole collection. Nevertheless, this work opens up the potential of applying the ZMO-modified AZO substrates to OLEDs and other optoelectronic devices.

5. CONCLUSIONS The surface WF of AZO transparent conductive thin films was raised to 4.85 eV using a low-resistance polycrystalline ZMO modifier embedded with monoclinic ZnMoO4 nanocrystals. This ZMO modifier with a thickness of ca. 30 nm was prepared by the thermal evaporation of MoO3 on the surface of AZO and a subsequent two-step annealing treatment. The first step of air annealing causes the formation of monoclinic ZnMoO4 nanocrystals in the modification layer through an adequate reaction between the ZnO and MoO3−x. This not only enhances the WF but also boosts the refractive index. Subsequently, the second step of N2 annealing increases the quantity of oxygen vacancies in the ZMO modification layer and then reduces the resistivity of the modification layer by enhancing the donor states. When applied to the front electrode of a-Si:H solar cells, the ZMO-modified AZO enables a gain of ca. 23% in photoelectric conversion efficiency compared to the unmodified AZO by improving the Jsc, Voc, and FF simultaneously. This work supports the application of ZMO-modified AZO to a variety of optoelectronic devices.



ASSOCIATED CONTENT

* Supporting Information S

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsami.9b07977. Cross-sectional TEM−EDX elemental mapping of the AZO/a-SiC:H interface; relative elemental depth profile of ZMO-modified AZO substrates; schematic diagram of typical TCO band structure; and refractive index of AZO, MoO3−x, p-a-SiC:H, and i-a-Si:H thin films (PDF)



AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected] (L.M.). *E-mail: [email protected] (T.Y.). ORCID

Lei Meng: 0000-0001-9794-0629 Tao Yang: 0000-0002-2132-2442 Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS This work was supported by the National Natural Science Foundation of China (grant nos. 61804147 and 91433206) and the CAS Pioneer Hundred Talents Program, China.



REFERENCES

(1) Kim, H.; Gilmore, C. M.; Horwitz, J. S.; Piqué, A.; Murata, H.; Kushto, G. P.; Schlaf, R.; Kafafi, Z. H.; Chrisey, D. B. Transparent Conducting Aluminum-doped Zinc Oxide Thin Films for Organic Light-emitting Devices. Appl. Phys. Lett. 2000, 76, 259−261. (2) Müller, J.; Rech, B.; Springer, J.; Vanecek, M. TCO and Light Trapping in Silicon Thin Film Solar Cells. Sol. Energy 2004, 77, 917− 930. (3) Baraton, M. The Future of TCO Materials: Stakes and Challenges. In Mater. Res. Soc. Symp.: Boston, MA, USA, 2009; Vol. 1209, pp 49−58. H

DOI: 10.1021/acsami.9b07977 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces (23) Feng, Q.; Wang, W.; Jiang, K.; Huang, J.; Zhang, Y.; Song, W.; Tan, R. Effect of Deposition Condition and UV-ozone Post-treatment on Work Function of DC Magnetron Sputtered AZO Thin Films. J. Mater. Sci.: Mater. Electron. 2012, 23, 267−272. (24) Timpel, M.; Nardi, M. V.; Krause, S.; Ligorio, G.; Christodoulou, C.; Pasquali, L.; Giglia, A.; Frisch, J.; Wegner, B.; Moras, P.; Koch, N. Surface Modification of ZnO(0001)-Zn with Phosphonate-Based Self-Assembled Monolayers: Binding Modes, Orientation, and Work Function. Chem. Mater. 2014, 26, 5042−5050. (25) Wood, C.; Li, H.; Winget, P.; Brédas, J.-L. Binding Modes of Fluorinated Benzylphosphonic Acids on the Polar ZnO Surface and Impact on Work Function. J. Phys. Chem. C 2012, 116, 19125−19133. (26) Winget, P.; Schirra, L. K.; Cornil, D.; Li, H.; Coropceanu, V.; Ndione, P. F.; Sigdel, A. K.; Ginley, D. S.; Berry, J. J.; Shim, J.; Kim, H.; Kippelen, B.; Brédas, J.-L.; Monti, O. L. A. Defect-Driven Interfacial Electronic Structures at An Organic/Metal-Oxide Semiconductor Heterojunction. Adv. Mater. 2014, 26, 4711−4716. (27) Schlesinger, R.; Xu, Y.; Hofmann, O. T.; Winkler, S.; Frisch, J.; Niederhausen, J.; Vollmer, A.; Blumstengel, S.; Henneberger, F.; Rinke, P.; Scheffler, M.; Koch, N. Controlling the Work Function of ZnO and the Energy-level Alignment at the Interface to Organic Semiconductors with A Molecular Electron Acceptor. Phys. Rev. B: Condens. Matter Mater. Phys. 2013, 87, 155311. (28) Kedem, N.; Blumstengel, S.; Henneberger, F.; Cohen, H.; Hodes, G.; Cahen, D. Morphology-, Synthesis- and Dopingindependent Tuning of ZnO Work Function using Phenylphosphonates. Phys. Chem. Chem. Phys. 2014, 16, 8310−8319. (29) Li, H.; Ratcliff, E. L.; Sigdel, A. K.; Giordano, A. J.; Marder, S. R.; Berry, J. J.; Brédas, J.-L. Modification of the Gallium-Doped Zinc Oxide Surface with Self-Assembled Monolayers of Phosphonic Acids: A Joint Theoretical and Experimental Study. Adv. Funct. Mater. 2014, 24, 3593−3603. (30) Jha, J. K.; Santos-Ortiz, R.; Du, J.; Shepherd, N. D. The Influence of MoOx Gap states on Hole Injection from Aluminum Doped Zinc Oxide with Nanoscale MoOx Surface Layer Anodes for Organic Light Emitting Diodes. J. Appl. Phys. 2015, 118, 065304. (31) Jha, J. K.; Sun, W.; Santos-Ortiz, R.; Du, J.; Shepherd, N. D. Electro-optical Performance of Molybdenum Oxide Modified Aluminum Doped Zinc Oxide Anodes in Organic Light Emitting Diodes: A Comparison to Indium Tin Oxide. Mater. Express 2016, 6, 289−294. (32) Jha, J. K.; Sun, W.; Du, J.; Shepherd, N. D. Mechanisms of AZO Workfunction Tuning For Anode Use in OLEDs: Surface Dipole Manipulation with Plasma Treatments versus Nanoscale WOx and VOx Interfacial Layers. J. Appl. Phys. 2017, 121, 185304. (33) Meng, L.; Yang, X.; Yang, T. Self-Flattened ZnO:Al Transparent Conductive Thin Films Derived by Sol-Gel Process. IEEE J. Photovolt. 2018, 8, 1149−1155. (34) Meng, L.; Konagai, M.; Miyajima, S. Effect of Annealing Atmosphere on the Electrical and Optical Properties of Sol-Gel Derived Al Doped Zn1‑xMgxO Thin Film. Thin Solid Films 2015, 597, 144−150. (35) Hsu, C.-W.; Cheng, T.; Yang, C.; Shen, Y.; Wu, J.; Wu, S. Effects of Oxygen Addition on Physical Properties of ZnO Thin Film Grown by Radio Frequency Reactive Magnetron Sputtering. J. Alloy Compd. 2011, 509, 1774−1776. (36) Mickan, M.; Stoffel, M.; Rinnert, H.; Helmersson, U.; Horwat, D. Restoring the Properties of Transparent Al-Doped ZnO Thin Film Electrodes Exposed to Ambient Air. J. Phys. Chem. C 2017, 121, 14426−14433. (37) Horwat, D.; Billard, A. Effects of Substrate Position and Oxygen Gas Flow Rate on the Properties of ZnO: Al Films Prepared by Reactive Co-sputtering. Thin Solid Films 2007, 515, 5444−5448. (38) Mickan, M.; Helmersson, U.; Rinnert, H.; Ghanbaja, J.; Muller, D.; Horwat, D. Room Temperature Deposition of Homogeneous, Highly Transparent and Conductive Al-doped ZnO Films by Reactive High Power Impulse Magnetron Sputtering. Sol. Energ. Mat. Sol. Cells 2016, 157, 742−749.

(39) Bao, D.; Gu, H.; Kuang, A. Sol-gel-derived C-axis Oriented ZnO Thin Films. Thin Solid Films 1998, 312, 37−39. (40) Wang, J.; Qi, Y.; Zhi, Z.; Guo, J.; Li, M.; Zhang, Y. A selfassembly mechanism for sol-gel derived ZnO thin films. Smart Mater. Struct. 2007, 16, 2673−2679. (41) Meng, L.; Yamada, A. Low-resistance Orthorhombic MoO3‑x Thin Film Derived by Two-step Annealing. Thin Solid Films 2018, 665, 179−183. (42) Matsui, T.; Bidiville, A.; Maejima, K.; Sai, H.; Koida, T.; Suezaki, T.; Matsumoto, M.; Saito, K.; Yoshida, I.; Kondo, M. Highefficiency amorphous silicon solar cells: Impact of deposition rate on metastability. Appl. Phys. Lett. 2015, 106, 053901. (43) Greiner, M. T.; Chai, L.; Helander, M. G.; Tang, W.-M.; Lu, Z.H. Transition Metal Oxide Work Functions: The Influence of Cation Oxidation State and Oxygen Vacancies. Adv. Funct. Mater. 2012, 22, 4557−4568. (44) Mulliken, R. S. A New Electroaffinity Scale; Together with Data on Valence States and on Valence Ionization Potentials and Electron Affinities. J. Chem. Phys. 1934, 2, 782−793. (45) Nethercot, A. H. Prediction of Fermi Energies and Photoelectric Thresholds Based on Electronegativity Concepts. Phys. Rev. Lett. 1974, 33, 1088−1091. (46) Campet, G.; Portier, J.; Subramanian, M. A. Electronegativity versus Fermi Energy in Oxides: the Role of Formal Oxidation State. Mater. Lett. 2004, 58, 437−438. (47) Rached, D.; Mostefaoui, R. Influence of the Front Contact Barrier Height on the Indium Tin Oxide/Hydrogenated p-doped Amorphous Silicon Heterojunction Solar Cells. Thin Solid Films 2008, 516, 5087−5092. (48) Sánchez-Sinencio, F.; Williams, R. Barrier at the Interface Between Amorphous Silicon and Transparent Conducting Oxides and Its Influence on Solar Cell Performance. J. Appl. Phys. 1983, 54, 2757−2760. (49) Nakada, T. CIGS Solar Cell (Japanese); Nikkan Kogyo Shimbun: Tokyo, Japan, 2010; p 59. (50) Crandall, R. S.; Iwaniczko, E.; Li, J. V.; Page, M. R. A Comprehensive Study of Hole Collection in Heterojunction Solar Cells. J. Appl. Phys. 2012, 112, 093713.

I

DOI: 10.1021/acsami.9b07977 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX