Critical Interface States Controlling Rectification of ... - ACS Publications

Subscriber access provided by PEPPERDINE UNIV

Article

Critical interface states controlling rectification of ultrathin NiO-ZnO p-n heterojunctions Kenneth Xerxes Steirer, Kai-Lin Ou, Neal R. Armstrong, and Erin L Ratcliff ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.7b08899 • Publication Date (Web): 23 Aug 2017 Downloaded from http://pubs.acs.org on August 24, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

ACS Applied Materials & Interfaces is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Critical interface states controlling rectification of ultrathin NiO-ZnO p-n heterojunctions K. Xerxes Steirer,1,† Kai Lin Ou,1 Neal R. Armstrong,1 and Erin L. Ratcliff 2* 1. Department of Chemistry and Biochemistry, University of Arizona, Tucson, AZ 85721. †Present address: Department of Physics, Colorado School of Mines, Golden, CO 80401. 2. Department of Materials Science and Engineering, University of Arizona, Tucson, AZ 85721. *[email protected] KEYWORDS: oxide heterojunction, nickel oxide, zinc oxide, diode, photoemission spectroscopies, interface ABSTRACT.

Herein, we consider the heterojunction formation of two prototypical metal

oxides: p-type NiO and n-type ZnO.

Elementally abundant, low-cost metal oxide/oxide‘

heterojunctions are of interest for optical sensing in the UV, gas sensing, photocatalysis, charge confinement layers, piezoelectric nanogenerators, and flash memory devices. These heterojunctions can also be used as current rectifiers and potentially as recombination layers in tandem photovoltaic stacks by making the two oxide layers ultrathin. In the ultrathin geometry, understanding and control of interface electronic structure and chemical reactions at the oxide/oxide‘ interface are critical to functionality, as oxygen atoms are shared at the interface of the dissimilar materials. In the studies presented here the extent of chemical reactions and interface band bending is monitored using X-ray and ultraviolet photoelectron spectroscopies (XPS/UPS). Interface reactivity is controlled by varying the near surface composition of nickel oxide, nickel hydroxide and nickel oxyhydroxide using standard surface-treatment procedures. A

ACS Paragon Plus Environment

1


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 24

direct correlation between relative percentage of interface hydroxyl chemistry (and hence surface Lewis basicity) and the local band edge alignment for ultrathin p-n junctions (6 nm NiO/30 nm ZnO) is observed. We propose an acid-base formulism to explain these results: the stronger the acid-base reaction, the greater the fraction of interfacial electronic states which lower the band offset between the ZnO conduction band and the NiO valence band. Increased interfacial gap states result in larger reverse current bias current of the p-n junction and lower rectification ratios. The acid-base formulism could serve as a future design principle for oxide/oxide‘ and other heterojunctions based on dissimilar materials. 1. Introduction Energy cascades comprised of metal oxide/metal oxide’ p-n heterojunctions are promising architectures for chemical, electrochemical, and photochemical systems for energy conversion processes and sensing. A simple and potentially quite useful oxide/oxide’ construct is the NiO/ZnO rectifying p-n heterojunction demonstrated by Ohta et al.1 Since its inception, the NiO/ZnO junction has shown promise in a variety of thin film technologies, including UV optical sensing,

2,3

gas sensing,4 photocatalysis,5 charge confinement in GaN light emitting

devices,6 piezoelectric nanogenerators,7 and flash memory devices.8 More recently, both NiO and ZnO have been employed individually as large band gap, charge selective interlayers in organic photovoltaics.9-13 An optimally designed NiO/ZnO thin film heterojunction could potentially be used as a recombination layer in a tandem cell, however careful control of band edge energetics would be required to guide interface free carrier transport. In the case of UV photodetection, the oxide-oxide’ junction should be highly rectifying, whereas in the case of a tandem photovoltaic cell, the oxide/oxide heterojunction should promote recombination.


2

Page 3 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


As with many electronic devices, interfacial electronic and chemical structure is expected to be important.

In the case of ultrathin junctions needed for recombination, interfacial

chemistry will be critical as one or both oxide layers may be fully depleted of its majority charge carrier. Yet, there is still a significant knowledge gap in predicting the interfacial chemistries and energy level alignment between two dissimilar metal oxides. The majority of studies on NiO/ZnO heterojunctions have been based on electronic device data and extensive analysis of the separate materials, with widely varying results.1,14-22 The approaches overlook complications that typically arise at the surface due to interfacial reaction chemistries, altering surface conductivity and carrier density, optical band gap, and density of states in the valence and conduction bands, bond strain, which ultimately affect free carrier dynamic behavior. 23-27 We propose an alternative formulism which uses acid-base properties of surfaces to predict and control the electronic properties of the oxide heterojunction. Conceptually, Brønsted and more generally Lewis acid-base reactions at metal oxide surfaces are similar regardless of whether the consideration is for gaseous molecules undergoing physical or chemical adsorption, anion or cation chemisorption in electrochemical processes, dissolution of the oxide through hydroxylation in corrosion, or here, acid-base surface reactions in the solid state. Lewis acid sites are attributed to surface metal cations in metal oxides that vary in oxidation state and/or coordination environment. Cation reactivity has been linked to the cation electronegativity and ionic radius, with low occupancy d-orbital metal oxides being acidic (MoOx and WOx) and high occupancy d-orbital metals being basic (CuO, NiO).

28-32

Conversely, Lewis base sites are

related to the lone pairs of surface oxygen atoms, with strength strongly dependent upon the local charge environments.29,33


3


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 24

Herein we consider the p-n junction NiO/ZnO as a model system to evaluate our proposed acid-base formulism. To provide clarity in the context of conventional semiconductor physics concepts, we relate the chemical states to surface electronic states within the gap and transport measurements using simple device architectures. The surface basicity of NiO is altered using surface treatments to control hydroxyl coverage,34 which allows investigation of acid-base reactions in the absence of substantial differences in lattice constants in the case of multiple oxide pairings. ZnO is well-known to be amphoteric, in that ZnO can act as a surface acid or base for electron donating and accepting, respectively. We demonstrate for the first time that in the presence of a strongly basic underlying substrate (high hydroxyl coverage on the nickel oxide surface), the ZnO interface at the heterojunction becomes more acidic. The strong acid-base reaction produced the highest fraction of interfacial electronic states and the lowest band offset (∆E) between the ZnO conduction band (ECB) and the NiO valence band (EVB). This interface showed the largest reverse current, indicating free carriers could readily leak across the interface, even under low internal electric fields. Conversely, the nickel oxide surface with the lowest basicity (lowest hydroxyl coverage) with increased acidic cation vacancies) showed the lowest fraction of interface states and largest ∆E. This interface also demonstrated the highest rectification, with decreased reverse bias and increased break-down voltage in forward bias. Overall, the proposed acid-base formulism serves as a chemistry-enabled path towards predicting oxide-oxide energy band alignment and interface transport properties. 2. Experimental NiO Preparation: ITO substrates were obtained from Thin Film Devices with surface roughness of 0.7 nm measured by AFM. ITO substrates were cleaned with Triton X-100 detergent diluted in nanopure H2O followed by three nanopure water sonication and then three ethanol sonication


4

Page 5 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


treatments for 10 min each. Synthetic air-plasma (400 mT, 10 min) was used to electronically activate the detergent cleaned substrates prior to NiO depositions. NiO was deposited at a rate of 0.3 Å/s in an excess oxygen environment (2.0x10-4 torr) with a Thermionics E-Beam evaporator having a base pressure of 1x10-6 torr. A significantly large substrate holder was used to prepare multiple NiO samples, which could be pre-treated and then loaded into the sputter chamber. Likewise, a significantly large substrate holder was used for simultaneous ZnO deposition on each of the differently pretreated NiO surfaces (three samples x three pretreatments, nine total samples each deposition) ensuring uniformity and comparability across the substrates. Thickness was monitored with a quartz crystal microbalance. NiO films were deposited uniformly (±10%) over an area of 170 cm2, as assessed using UV-vis absorption on NiO samples made in eight different locations across the holder. Substrates were cleaved in sections to further ensure that similar films were used for devices and photoemission measurements. NiO Surface Treatments: Oxygen plasma treatment of NiO films was performed in a clean quartz tube in a Harrick RF plasma system with oxygen pressure of 350 mtorr (UHP) for 2 min at high power. Argon plasma treatment was performed in the same system and settings but with Ar gas (UHP). Heat+Water treatments were performed by placing the NiO sample on a hotplate (120ºC) for 10 minutes, then rinsing the hot sample in ethanol then pure water and replacing the sample on the hotplate for 10 additional minutes. NiO samples were either placed into UHV for analysis or the sputter chamber for ZnO deposition immediately after the surface treatments. ZnO Deposition: ZnO was sputter deposited in a Lesker RF sputter system at 1.3 Å/s with sample rotation (working pressure of 7 mtorr Ar (UHP); base pressure < 1x10-6 torr) using a 3” ZnO target from Plasmaterials and a target-substrate distance of 107 mm at 100 W power. A 10


5


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 24

min burn in of the target was utilized for each deposition prior to opening the shutter. ZnO films were immediately introduced into UHV for analysis or into the Ag deposition chamber for top contacts. Devices: Ag top contacts were deposited with an ebeam gun in vacuum with a deposition rate of 0.1 Å/s to minimize device shunting. Device areas were 0.1 cm2. Working sample numbers were n=50 for O2-plamsa, n = 30 for Ar-plasma and n = 3 for damp heat. Current-Voltage sweeps were performed using a Keithley 2400 controlled by Labview software. We note that hysteresis was observed when sweep direction was switched at larger biases over 2 V, which coincided with irreversible shunting of the damp heat and Ar-plasma treated films. The only devices that survived over 2.4 V forward bias had oxygen plasma treated NiO. As such, all devices used in the statistics of this study were tested below 2V to avoid shunting and hysteresis. Thicknesses of NiO and ZnO films were determined using Field Emission Scanning Electron Microscopy (FESEM). In the case of 0.5 nm ZnO, the thickness was determined by dividing the desired thickness by the known growth rate (0.013 nm/s) to determine deposition time. Characterization: XPS studies were performed with a Kratos Axis Ultra X-ray photoelectron spectrometer with a monochromatic Al Kα source at 1486.6 eV. Binding energy calibration of the spectra was corrected using the procedure outlined by M. P. Seah, using the Cu 3p, Au 4f7/2 Ag 3d5/2, CuL3MM, Cu 2p3/2 and AgM4NN peak positions.35 XPS measurements were collected at a pass energy of 20 eV, at both 0º and 60° take off angles for n = 3 for each pretreatment. Small In 3d signal intensity from the underlying ITO substrate were near detection limits in all of the NiO measured spectra and did not contribute to the O 1s spectra, as verified in the Supplemental Information (Figure S1). For simplicity, all backgrounds were computed using a linear prescription. Only the O 1s peaks were fit and in this region, a linear background is


6

Page 7 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


appropriate. We recommend using a Shirley background for Ni 2p and Zn 2p peaks for rigorous fitting although only the relative elemental ratios were assessed in this work. A He(I) excitation source (21.22 eV) was used for UPS measurements with 5 eV pass energy and 0º take off angle. A –10.00 V bias was applied to the sample during UPS experiments to spectrally separate the lowest kinetic energy electrons and secondary electrons from the local environment. An Ar sputter-etched, atomically clean gold sample was measured before characterization of the samples to establish the Fermi edge of the spectrometer. 3. Results and Discussion 3.1 Control of surface chemistry of NiO NiO films with equivalent thicknesses (6 nm) were deposited onto freshly cleaned ITO substrates. O 1s contributions from the underlying ITO electrode were found to be minimal (In 3d < 5% of Ni 2p signal), as shown in the Supplemental Information (Figure S1). The hydroxyl coverage of the NiO thin films was varied by using post-deposition surface treatments. Postdeposition surface treatments, such as washing and plasma exposure, can strongly affect the surface chemistry, electronic properties, and photovoltaic device efficiency for hole-selective NiO interlayers, as recently shown for NiO films deposited by solution precursors, pulsed laser deposition and sputtering.13,34,36,37 X-ray and ultraviolet photoemission spectroscopies (XPS, UPS) were used to evaluate the NiO surface chemistry, valence band edge and relative Fermi level position. Figure 1A shows XPS (O 1s) spectra for each of the surface-treated NiO films described below, and includes a reference of a dried NiO (intrinsic) powder. Schematic representations of the surfaces are shown in Figure 1B; where disorder is assumed based on unobservable peaks in X-ray diffraction


7


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

measurements.

Page 8 of 24

For more precise descriptions of crystalline forms of NiO, Ni(OH)2, and

NiOOH, the reader is directed towards additional literature.38,39 In Figure 1A, all films contain a low binding energy feature at 529.4 eV corresponding to NiO-like binding environment.34 A stoichiometric NiO film is highly insulating; free carriers (holes) result from the presence of defects at the surface to due chemical reactions and deviation from stoichiometry. Specifically, the lone pairs on the oxygen atoms readily react with protons in ambient water such that NiO surfaces typically possess thin surface layers of hydroxyls (Ni(OH)2), as shown by the arrow in Figure 1A. The reaction is spontaneous - hydroxylation occurs upon removal of as-deposited films from the vacuum chamber and is evident by the higher binding energy at 531 eV of the O 1s XPS spectrum of the powder in Figure 1A for hydrated nickel hydroxyls (α-Ni(OH)2•(H2O)z) (Figure 1B).34,40-43 Leveraging the reaction of the NiO surface, Ni(OH)2 coverage was controlled using three different surface-treatments on separate films: i) 10 min exposure to water (vapor and liquid) while heating (120°C) for high coverage (most basic); ii) 2 min argon (Ar) plasma for moderate coverage; and iii) 2 min oxygen (O2) plasma, which yielded the lowest hydroxyl coverage (least basic). Higher binding energy peak at 532.5 (±0.2) eV is assigned to adventitious water present in all films. In the case of the O2-plasma treated sample, the hydrated oxy-hydroxide (γNiOOH•(H2O)z) is observed at 532.1 eV, represented in Figure 1B schematically.34,36 The O2plasma treatment further decreases the surface basicity (increase acidity) by creating higher order redox states on localized nickel atoms (average charge on Ni sites from 3.5 to 3.8) and a nearsurface region with higher energy hydrogen bonding arrangements.34,36


8

Page 9 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 1. A.) XPS of the O 1s core level for heat and water (●), argon plasma (■), and oxygen plasma (▲) treated surfaces (6 nm), as well as intrinsic NiO powder (+). The observed counts per second (y-axis) has been normalized to the low binding energy peak for each NiO layer. Key chemical shifts are identified. B.) Schematic representation of the surface structures of nickel oxide (left), hydrated nickel hydroxide (center), and hydrated nickel oxyhydroxide (right). Films are intentionally represented as disordered. C.) Energy band diagrams, as derived from UPS, for each of the surface-treatments. Arrows indicate work function (Φ), ionization energy (IE) and the energetic offset between Fermi level and occupied states.


9


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 24

The surface acid and base properties are strongly interconnected to the energetics, including electronic position of the Fermi level (EFermi) relative to the valence and conduction bands at the surface. UPS was used to evaluate the energetic positions of the Fermi level, relative to the occupied states of the valence band for each of the surface-treated films. Figure 1C shows the band diagrams taken from the UPS spectra of the three treated NiO films. Figure S2 shows the same energy band diagram, instead plotted on the same energy axes (Supplementary Information). The work function (Φ) is the energetic difference between the Fermi level (EF) and the surface vacuum level (EVAC). The ionization energy (IE) is the energetic difference between the observed onset in the occupied valence states of the oxide and the surface vacuum level. For reference, the energetic difference between the Fermi level and the onset of occupied states is also provided; a smaller value is typically associated with a more p-doped surface, although requires confirmation by measurement of carrier density In Figure 1C, there are no obvious correlations between measured work function, ionization energy, and the hydroxyl coverages given in Figure 1A, demonstrating the complexity of metal oxide surface chemistry. Changing the hydroxyl coverage and/or a change in cation vacancies can simultaneously produce interface dipole effects and variations to carrier density that alter Φ and IE in a non-systematic way. The results in Figure 1C also suggest that predicting the band alignment between the NiO and subsequent ZnO films may not be feasible based on the electronic properties of the individual materials due to a complicated near-surface environment. Alternatively, an acid-base reaction would vary in accordance with the total hydroxyl coverage and hence basicity of the surface.


10

Page 11 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


3.2 NiO/ZnO chemical interface formation The chemical compositions of NiO/ZnO heterojunctions with 0.5 nm and 18 nm of ZnO deposited on different surface-treated NiO were evaluated using XPS. The O 1s XPS spectra for each interface, along with the surface-treated NiO spectra, are shown in Figure 2; ZnO thickness increases from top to bottom. Additional data for the Zn 2p and Ni 2p are provided in the Supplementary Information section (Figures S3 and S4). The three spectra on the left are damp heat treated NiO (most basic), the center spectra are Ar-plasma treated NiO, and the right spectra

Figure 2. XPS of O 1s core level of ex-situ surface treated NiO (top spectra), 0.5 nm ZnO on NiO (middle spectra) and 18 nm of ZnO on NiO (bottom spectra). From left to right are surface-treatments of heat+water, Ar-plasma, and O2-plasma, respectively.


11


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

are O2-plasma treated NiO (most acidic). provided

in

Table

S1.

Spectral

Page 12 of 24

Percent compositions and fitting parameters are

decomposition

was

performed

with

symmetric

Gaussian/Lorentzian (70/30) line shapes. Chemical assignment of the near surface composition of the NiO surfaces follows an earlier report based on chemical standards (Figure 2, top row).34 Briefly, the lowest binding energy peak at 529.4 (±0.1) eV is assigned to lattice oxygen associated with NiO-like envirionment and the middle peak at 530.9 (±0.2) eV is associated with Ni(OH)2. The higher binding energy peak at 532.5 (±0.2) eV is assigned to adventitious water. In the case of the O2plasma treated sample, the oxy-hydroxide peak is located at 532.1 eV and can be distinguished from adventitious water using the Ni 2p spectra. The O 1s spectra for the 18 nm ZnO (bottom row Figure 2) can be fit using three peaks composed of lattice O (530.7 eV), defects commonly associated with oxygen vacancies that have undergone surface reconstruction to hydroxyls (531.8 – 532.2 eV), and free surface hydroxyls (532.7 – 532.9 eV).32,44-46 All spectra from the thick ZnO over-layers were found to be nearly indistinguishable. The middle row of Figure 2 shows O 1s spectra obtained from the underlying NiO and the thin 0.5 nm ZnO top layer. Percent atomic compositions (ratios of Ni to Zn) show similar ZnO coverage for each of the 0.5 nm over-layers (Table S1). A four peak model was required to fit the thin ZnO (0.5 nm) on NiO, with all peaks constrained by relative binding energy position and full width half maximum. Table S2 in the supplementary information provides detailed information of the fitting parameters used for each O 1s spectrum in Figure 2. It is important to note the O 1s core level spectra originate from oxygen present on both sides of the interface. Attempts to develop and compare rigorous chemical models at the interface that could be related to chemical standards did not produce statistically unique fits to


12

Page 13 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


the spectra. However, a qualitative assessment does give some insight, with the four provided peak fits to describe relative binding energy regions. No chemical change to the subsurface NiO is expected and the lattice oxide from NiO clearly distinguished in all the thin ZnO over-layer spectra as the lowest binding energy feature (529.3 -529.8 eV). This feature is not observed in the 18 nm thick ZnO samples (bottom spectra). The peak between 530.4 - 530.7 eV aligns with lattice O in emerging ZnO but is very close to the substrate derived nickel hydroxide peak at slightly higher BE. In all three surface treatment cases, the dominant spectral feature is the highest binding energy peak, which likely arises from the nucleating ZnO bound to the NiO surface through shared oxygen sites. This spectral feature coincides with oxygen vacancy defects (VO) in ZnO, which have undergone surface reconstruction to hydroxides.

These states

correspond to the formation of a shallow electron donor state in the conduction band and an increase in carrier concentration in the space-charge layer of the ZnO.32,44-46 We hypothesize that the hydroxyl coverage of the NiO surface would lead to a variation of DOS within the space charge region of an NiO/ZnO heterojunction, which would be evident in the energetic alignment and the rectification behavior in their current-voltage responses.

This hypothesis is easily

validated through energetic alignment and p-n heterojunction devices. 3.3. NiO/ZnO band edge alignment. The energetic alignment at the interface was also probed using UPS and XPS. UPS spectra are given in Figure 3, with the Φ and IE values provided in Table S3 of the supplemental information. The hash marks in Figure 3 indicate the work function (left panels), as estimated by the onset in the secondary electron edge. The center panel provides the full UPS spectrum for each film, with the Zn 3d peak featured at ~10-12 eV below the Fermi energy for the 0.5 nm and 18 nm ZnO over-layers. The O 2p peaks can be observed at ~4-7 eV below the Fermi level. In


13


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 24

the right panel, hash marks indicate the onset in the occupied states used to estimate ionization energies. The transition from a p-type oxide to an n-type oxide is readily apparent even at 0.5 nm ZnO coverage.

Figure 3: Ultraviolet photoemission spectroscopy of NiO with varying ZnO over-layers. Hash marks indicate work function and onset in occupied states used to calculate the work functions and ionization energies.


14

Page 15 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 4 shows values derived directly from the UPS data, and the relative difference between IP and Fermi level for each layer (surface treated NiO, 0.5 nm ZnO and 18 nm ZnO). In Figure 4, the surface termination chemistry of the NiO appears to strongly modulate the valence band energies of the nucleated layer of ZnO. The thin ZnO on the different NiO surfaces exhibits IE values that shift significantly towards the Fermi level for NiO films treated with water/heat (3.7 eV), argon plasma (3.3 eV), and oxygen plasma (3.0 eV).

Moreover, by

including a 3.2 eV band gap estimate for ZnO (as derived in Figure S5), large differences in the energetic alignment between the two oxides in the near surface region are readily observable. We suggest some caution should be exercised when using a bulk and optical based measurement for a surface electronic level. The water/heat treatment (most basic, highest hydroxyl coverage) shows only a small offset between the valence band onset (EVBM) of the NiO and the presumed conduction band (ECBM) of the ZnO (~0.3 eV). The argon plasma treatment (moderate basicity and hydroxyl coverage) has a moderate energetic offset (0.6 eV), while the largest offset (1.2 eV) was found for the oxygen plasma treatment (lowest basicity).

A similar, although less

Figure 4. Valence band edge positions derived from UPS data. ZnO conduction band edge positions assume Eg = 3.2 eV, derived from optical measurements.


15


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 24

pronounced trend is observed for the 18 nm thick ZnO films. However, elucidation of interface effects at 18nm thickness is greatly complicated by the low escape depth of the ultravioletgenerated photoelectrons, coupled with observed differences in vacuum level (Evac) which could be attributed to surface dipole features. 3.4 . NiO-ZnO devices Based on the band diagram provided in Figure 4, one would expect the low barrier of 0.3 eV for the water/heat, high hydroxyl covered NiO would produce the lowest rectifying junction with ZnO. Specifically, the high number of observed interface electronic states would promote free carriers to leak across the interface, even under low internal electric fields. Systematically increasing the barrier height for carriers to cross the junction (by decreasing hydroxyl coverage using the Ar- or O2-plasma treatments) are predicted to improve diode rectification. Devices were fabricated by adding sputter deposited ZnO thin films (36 nm) over the top of the surface-treated NiO films followed by e-beam deposited Ag top contacts. ZnO layers were added to all NiO films simultaneously on a rotating plate to ensure uniform ZnO thickness and properties. The current-voltage responses for these devices are shown in Figure 5, with rectification at each bias point shown in the inset (ratio of forward to reverse bias current). Device yields for the ultrathin ZnO (18 nm) were lower than devices created with thicker ZnO films (36 nm), except when using oxygen plasma treated NiO substrates where yields over 70% were obtained. Device interpretation is based on a Type II heterojunction, although with the caveat that the devices are not truly p-n junctions since the NiO is expected to be depleted of its majority charge carrier.


16

Page 17 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 5. Current density versus voltage bias for thin ITO/6 nm NiO/36 nm ZnO/Ag diodes as a function of surface-treatment: oxygen plasma (●), argon plasma (▲), and heat + water (■). Rectification (forward current/reverse current) versus voltage is shown in the inset.

In Figure 5, in reverse bias, there is a clear trend in the observed leakage current, as predicted by the energy band diagram in Figure 4. The high barrier, lowest hydroxyl O2-plasma treatment has the lowest reverse bias current. This low dark current is consistent with the rectification ratios shown in the inset, where rectification ratio increases with decreasing hydroxyl coverage: water/heat > argon plasma > oxygen plasma. The low barrier, high hydroxyl surface arising from the water/heat treatment has the highest reverse bias current. For this device, diode-like behavior (i.e. current rectification) is minimal, with a rectification ratio below 1, with recombination highly likely at the interface. When comparing the three devices in the forward bias region, the Ar-plasma treatment has a lower turn-on voltage, demonstrating a current density of 100x at 2 V over the other two devices. This observation suggests that the interface may contain more defects than detected via photoelectron spectroscopy and/or a difference in surface electronic gap between the NiO and ZnO. For all three devices, higher


17


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 24

forward biases results in irreversible shorts, likely due to the ultrathin nature of the heterojunctions (36 nm).

Figure 6. Improved rectification is shown with a 18 nm ZnO layer on 6 nm NiO layer grown on ITO. The devices using O2 plasma treated NiO reproducibly showed the highest forward bias range before damaging the diode, and the lowest current in reverse bias. Due to the lower device yields for Ar plasma (40%) and water+heat (10%) treated NiO surfaces, only the oxygen plasma determined NiO surface hydroxides was further investigated by decreasing the ZnO layer thickness, as shown in Figure 6. Device rectification reached above 104 at ± 0.8 V when using O2 plasma treated NiO films, a thin film metal oxide diode that is only 24 nm thick. These ultrathin NiO/ZnO devices are representative of thicknesses in nanostructured heterojunctions,7,47 illustrating the critical nature of the interface states in nanodevices.


18

Page 19 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Overall, the strong high BE signals in the XPS spectra from the ultrathin ZnO layer suggest that the nucleating layer is highly defective, resulting from a high density of hydroxyl species. Combining the XPS analysis with the UPS data and the device performance, we conclude that hydroxyls in the NiO surface layer react with the initially deposited ZnO layer, in what is likely a Lewis acid reaction that is mediated by the concentration of surface hydroxyls. The resulting electron donor states near the conduction band of the ZnO can then dope the interface, and affect the rectifying charge barrier of the structure. In other words, the Ni(OH)2 interaction with the ZnO results in more metallic-like interfacial states, allowing carriers to readily transport across the interface at low bias (higher leakage currents), as observed in Figure 5. Hence, by decreasing the Ni(OH)2 coverage, available carriers are reduced at the interface and rectification increases. 4. Conclusions We have shown that control of surface hydroxyl driven chemistry in ultrathin NiO/ZnO heterojunctions can significantly affect the ZnO electronic structure and performance of the heterojunction. A combined analysis of chemical and electronic features enables a closer and more accurate look at the details of these nanoscale heterojunctions, wherein the interface properties manifest in systematic changes in rectification ratios and leakage currents, with implications on device durability. The critical feature of the interface states is that the apparent magnitude of the states are chemically controllable through the basicity of the underlying substrate.

This finding is especially important as traditional p-n junctions typically have

depletion widths that can span up to microns whereas in nanoscale devices, one or both oxide layers may be fully depleted. These results suggest that hydroxyl control of functional ultra-thin oxide-oxide heterojunctions is likely to have significant impact on the electronic properties.


19


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 24

Applications can be found in sensing, catalytic and photo-catalytic systems where nanometer scale interfaces often dominate key functional properties. Supporting Information. Additional core level spectra and relevant fit parameters, band diagrams, UV-visible absorption spectroscopy, x-ray diffraction, and current-voltage curves are provided in Supplemental Information. Acknowledgements This research was supported as part of the Center for Interface Science: Solar Electric Materials (CISSEM), an Energy Frontier Research Center funded by the U.S. Department of Energy, Office of Science, Basic Energy Sciences, under Award DE-SC0001084.


20

Page 21 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Table of Contents Figure

TOC Graphic


21


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 24

REFERENCES (1) Ohta, H.; Hirano, M.; Nakahara, K.; Maruta, H.; Tanabe, T.; Kamiya, M.; Kamiya, T.; Hosono, H. Fabrication and Photoresponse of a PN-Heterojunction Diode Composed of Transparent Oxide Semiconductors, p-NiO and n-ZnO. Appl. Phys. Lett. 2003, 83, 1029-1031. (2) Kawade, D.; Moriyama, K.; Nakamura, F.; Chichibu, S. F.; Sugiyama, M. Fabrication of Visible-Light Transparent Solar Cells Composed of NiO/NixZn1-xO/ZnO Heterostructures. Phys. Status Solidi C 2015, 12, 785-788. (3) Park, N.; Sun, K.; Sun, Z.; Jing, Y.; Wang, D. High Efficiency NiO/ZnO Heterojunction UV Photodiode by Sol–Gel Processing J. Mater. Chem. C 2013, 1, 7333-7338. (4) Liu, Y.; Li, G.; Mi, R.; Deng, C.; Gao, P. An Environment-Benign Method for the Synthesis of p-NiO/n-ZnO Heterostructure with Excellent Performance for Gas Sensing and Photocatalysis. Sens. Actuators, B 2014, 191, 537-544. (5) Tian, F.; Liu, Y. Synthesis of p-Type NiO/n-Type ZnO Heterostructure and its Enhanced Photocatalytic Activity. Scr. Mater. 2013, 69, 417-419. (6) Abbasi, M.; Ibupoto, Z.; Hussain, M.; Nur, O.; Willander, M. The Fabrication of White Light-Emitting Diodes Using the n-ZnO/NiO/p-GaN Heterojunction with Enhanced Luminescence. Nanoscale Res. Lett. 2013, 8, 320. (7) Yin, B.; Qiu, Y.; Zhang, H.; Lei, J.; Chang, Y.; Ji, J.; Luo, Y.; Zhao, Y.; Hu, L. Piezoelectric Performance Enhancement of ZnO Flexible Nanogenerator by a NiO–ZnO p–n Junction Formation. Nano Energy 2015, 14, 95-101. (8) Sun, C.-E.; Chen, C.-Y.; Chu, K.-L.; Shen, Y.-S.; Lin, C.-C.; Wu, Y.-H. ZnO/NiO Diode-Based Charge-Trapping Layer for Flash Memory Featuring Low-Voltage Operation. ACS Appl. Mater. Interfaces 2015, 7, 6383-6390. (9) Cowan, S. R.; Li, J. V.; Olson, D. C.; Ratcliff, E. L. Contact-Induced Mechanisms in Organic Photovoltaics: A Steady-State and Transient Study. Adv. Energy Mater. 2015, 5, 1400549. (10) Cowan, S. R.; Schulz, P.; Giordano, A. J.; Garcia, A.; MacLeod, B. A.; Marder, S. R.; Kahn, A.; Ginley, D. S.; Ratcliff, E. L.; Olson, D. C. Chemically Controlled Reversible and Irreversible Extraction Barriers Via Stable Interface Modification of Zinc Oxide Electron Collection Layer in Polycarbazole-based Organic Solar Cells. Adv. Funct. Mater. 2014, 24, 4671-4680. (11) Ratcliff, E. L.; Garcia, A.; Paniagua, S. A.; Cowan, S. R.; Giordano, A. J.; Ginley, D. S.; Marder, S. R.; Berry, J. J.; Olson, D. C. Investigating the Influence of Interfacial Contact Properties on Open Circuit Voltages in Organic Photovoltaic Performance: Work Function Versus Selectivity. Adv. Energy Mater. 2013, 3, 647-656. (12) Ratcliff, E. L.; Zacher, B.; Armstrong, N. R. Selective Inter layers and Contacts in Organic Photovoltaic Cells. J. Phys. Chem. Lett. 2011, 2, 1337-1350. (13) Steirer, K. X.; Ndione, P. F.; Widjonarko, N. E.; Lloyd, M. T.; Meyer, J.; Ratcliff, E. L.; Kahn, A.; Armstrong, N. R.; Curtis, C. J.; Ginley, D. S.; Berry, J. J.; Olson, D. C. Enhanced Efficiency in Plastic Solar Cells via Energy Matched Solution Processed NiOx Interlayers. Adv. Energy Mater. 2011, 1, 813-820. (14) Seo, J. W.; Baik, S. J.; Kang, S. J.; Hong, Y. H.; Yang, J. H.; Lim, K. S. A ZnO Cross-Bar Array Resistive Random Access Memory Stacked with Heterostructure Diodes for Eliminating the Sneak Current Effect. Appl. Phys. Lett. 2011, 98, 233505.


22

Page 23 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(15) Xi, Y. Y.; Hsu, Y. F.; Djurisic, A. B.; Ng, A. M. C.; Chan, W. K.; Tam, H. L.; Cheah, K. W. NiO/ZnO Light Emitting Diodes by Solution-Based Growth. Appl. Phys. Lett. 2008, 92, 113505. (16) Wang, K.; Vygranenko, Y.; Nathan, A. Fabrication and Characterization of NiO/ZnO/ITO p–i–n Heterostructure. Thin Solid Films 2008, 516, 1640-1643. (17) Nel, J. M.; Auret, F. D.; Wu, L.; Legodi, M. J.; Meyer, W. E.; Hayes, M. Fabrication and Characterisation of NiO/ZnO Structures. Sens. Actuators, B 2004, 100, 270-276. (18) Auret, F. D.; Wu, L.; Meyer, W. E.; Nel, J. M.; Legodi, M. J.; Hayes, M. Electrical Characterisation of NiO/ZnO Structures. Phys. Status Solidi C 2004, 1, 674-677. (19) Wöll, C. The Chemistry and Physics of Zinc Oxide Surfaces. Prog. Surf. Sci. 2007, 82, 55-120. (20) Sato, H.; Minami, T.; Takata, S.; Yamada, T. Transparent Conducting p-type NiO Thin Films Prepared by Magnetron Sputtering. Thin Solid Films 1993, 236, 27-31. (21) Furstenau, R. P.; McDougall, G.; Langell, M. A. Initial Stages of Hydrogen Reduction of NiO(100). Surf Sci. 1985, 150, 55-79. (22) Scheidt, H.; Glöbl, M.; Dose, V. The Oxidation of Nickel: A Study of Empty Electronic States by Inverse Photoemission and Soft X-ray Appearance Potential Spectroscopy. Surf Sci. 1981, 112, 97-110. (23) Christen, J.; Grundmann, M.; Bimberg, D. Direct Imaging and Theoretical Modelling of the Atomistic Morphological and Chemical Structure of Semiconductor Heterointerfaces. Appl. Surf. Sci. 1990, 41, 329-336. (24) Grundmann, M.; Karsthof, R.; von Wenckstern, H. Interface Recombination Current in Type II Heterostructure Bipolar Diodes. ACS App. Mater. Interfaces 2014, 6, 1478514789. (25) Sah, C. t.; Noyce, R. N.; Shockley, W. Carrier Generation and Recombination in P-N Junctions and P-N Junction Characteristics. Proc. IRE 1957, 45, 1228-1243. (26) Shockley, W.; Read, W. T. Statistics of the Recombinations of Holes and Electrons. Phys. Rev. 1952, 87, 835-842. (27) Grundmann, M. The Bias Dependence of the Non-Radiative Recombination Current in p-n Diodes. Solid·State Electron. 2005, 49, 1446-1448. (28) Bañares, M. A. Supported Metal Oxide and Other Catalysts for Ethane Conversion: A Review. Catal. Today 1999, 51, 319-348. (29) Barteau, M. A. Site Requirements of Reactions on Oxide Surfaces. J. Vac. Sci. Technol., A 1993, 11, 2162-2168. (30) Eranna, G.; Joshi, B. C.; Runthala, D. P.; Gupta, R. P. Oxide Materials for Development of Integrated Gas Sensors - A Comprehensive Review. Crit. Rev. Solid State Mater. Sci. 2004, 29, 111-188. (31) Freund, H. J.; Dillmann, B.; Ehrlich, D.; Hassel, M.; Jaeger, R. M.; Kuhlenbeck, H.; Ventrice, C. A.; Winkelmann, F.; Wohlrab, S.; Xu, C.; Bertrams, T.; Brodde, A.; Neddermeyer, H. Adsorption and Reaction of Molecules on Surfaces of Metal-Metal Oxide Systems. J. Mol. Catal. 1993, 82, 143-169. (32) Henrich, V. E.; Cox, P. A.: The Surface Science of Metal Oxides; Cambridge University Press: Cambridge, UK, 1994. pp. 8, 144, 297. (33) Stair, P. C. The Concept of Lewis Acids and Bases Applied to Surfaces. J. Am. Chem. Soc. 1982, 104, 4044-4052.


23


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 24

(34) Ratcliff, E. L.; Meyer, J.; Steirer, K. X.; Garcia, A.; Berry, J. J.; Ginley, D. S.; Olson, D. C.; Kahn, A.; Armstrong, N. R. Evidence for near-Surface NiOOH Species in Solution-Processed NiOx Selective Interlayer Materials: Impact on Energetics and the Performance of Polymer Bulk Heterojunction Photovoltaics. Chem. Mater. 2011, 23, 4988-5000. (35) Seah, M. P. Post-1989 Calibration Energies for X-ray Photoelectron Spectrometers and the 1990 Josephson Constant. Surf. Interface Anal. 1989, 14, 488. (36) Widjonarko, N. E.; Ratcliff, E. L.; Perkins, C. L.; Sigdel, A. K.; Zakutayev, A.; Ndione, P. F.; Gillaspie, D. T.; Ginley, D. S.; Olson, D. C.; Berry, J. J. Sputtered Nickel Oxide Thin Film for Efficient Hole Transport Layer in Polymer-Fullerene Bulk-Heterojunction Organic Solar Cell. Thin Solid Films 2012, 520, 3813-3818. (37) Steirer, K. X.; Richards, R. E.; Sigdel, A. K.; Garcia, A.; Ndione, P. F.; Hammond, S. R.; Baker, D. J.; Ratcliff, E. L.; Curtis, C.; Furtak, T. E. Nickel Oxide Interlayer Films From Nickel Formate-Ethylenediamine Precursor: Influence of Annealing on Thin Film Properties and Photovoltaic Device Performance. J. Mater. Chem. A 2015, 3, 10949-10958. (38) Juodkazis, K.; Juodkazyte, J.; Vilkauskaite, R.; Jasulaitiene, V. Nickel Surface Anodic Oxidation and Electrocatalysis of Oxygen Evolution. J.Solid State Electrochem. 2008, 12, 1469-1479. (39) Van der Ven, A.; Morgan, D.; Meng, Y. S.; Cederc, G. Phase Stability of Nickel Hydroxides and Oxyhydroxides. J. Electrochem. Soc. 2006, 153, A210-A215. (40) Greiner, M. T.; Helander, M. G.; Wang, Z. B.; Tang, W. M.; Lu, Z. H. Effects of Processing Conditions on the Work Function and Energy-Level Alignment of NiO Thin Films. J. Phys. Chem. C 2010, 114, 19777-19781. (41) Hufner, S. Electronic Structure of NiO and Related 3D-Transistion Metal Compounds. Adv. Phys. 1994, 43, 183-356. (42) Langell, M. A.; Nassir, M. H. Stabilization of NiO(111) Thin Films by Surface Hydroxyls. J. Phys. Chem. 1995, 99, 4162-4169. (43) Oswald, S.; Bruckner, W. XPS Depth Profile Analysis of non-Stoichiometric NiO Films. Surf. Interface Anal. 2004, 36, 17-22. (44) Moormann, H.; Kohl, D.; Heiland, G. Work Function and Band Bending on Clean Zinc Oxide Surfaces. Surf Sci. 1979, 80, 261-264. (45) Nakagawa, M.; Mitsudo, H. Anomalous Temperature-Dependence of the Electrical Conductivity of Zinc Oxide Thin Films. Surf Sci. 1986, 175, 157-176. (46) Coppa, B. J.; Fulton, C. C.; Hartlieb, P. J.; Davis, R. F.; Rodriguez, B. J.; Shields, B. J.; Nemanich, R. J. In Situ Cleaning and Characterization of Oxygen- and Zinc-Terminated, ntype, ZnO{0001} Surfaces. J. Appl. Phys. 2004, 95, 5856-5864. (47) Wang, J.-Y.; Lee, C.-Y.; Chen, Y.-T.; Chen, C.-T.; Chen, Y.-L.; Lin, C.-F.; Chen, Y.-F. Double Side Electroluminescence from p-NiO/n-ZnO Nanowire Heterojunctions. Appl. Phys. Lett. 2009, 95, 131117.


24

Critical Interface States Controlling Rectification of ... - ACS Publications

Recommend Documents