This site uses cookies. By continuing to use this site you agree to our use of cookies. To find out more, see our Privacy and Cookies policy. Close this notification
NOTICE: Ukraine: Read IOP Publishing's statement.

The Japan Society of Applied Physics

The Japan Society of Applied Physics (JSAP) serves as an academic interface between science and engineering and an interactive platform for academia and the industry. JSAP is a "conduit" for the transfer of fundamental concepts to the industry for development and technological applications.

JSAP was established as an official academic society in 1946, and since then, it has been one of the leading academic societies in Japan. The society's interests cover a broad variety of scientific and technological fields, and JSAP continues to explore state-of-the-art and interdisciplinary topics.

To this end, the JSAP holds annual conferences; publishes scientific journals; actively sponsors events, symposia, and festivals related to science education; and compiles information related to state-of-the-art technology for the public.

Rapid Communication The following article is Free article

Enhancement of transparency in epitaxially-grown p-type SnO films by surface-passivation treatment in a Na2S aqueous solution

, , , and

Published 9 May 2022 © 2022 The Japan Society of Applied Physics
, , Spotlights 2022 Citation Suguri Uchida et al 2022 Jpn. J. Appl. Phys. 61 050903

1347-4065/61/5/050903

Abstract

We report on the epitaxial growth of (001)-oriented SnO films on yttria-stabilized zirconia (100) substrates by pulsed-laser deposition and the impact of surface-passivation treatment on the optical transparency. The films immersed in a Na2S aqueous solution exhibited average visible transmittance higher than that of the as-grown ones by ∼18% despite negligibly small variations in the crystalline structure, p-type conductivity, and composition. Based on these results, the enhanced visible transmittance can be attributed to the suppression of midgap states near the film surface. The extended treatment resulted in conversion to a SnS phase, demonstrating a facile anion-exchange reaction.

Export citation and abstract BibTeX RIS

As thin-film growth technology matured for the flat-panel display industry, high carrier mobility (∼102 cm2 V–1 s–1) was achieved for thin films based on n-type transparent conducting oxides (n-TCOs) such as Sn-doped In2O3 and InGaZnOx . 13) In contrast, most of p-type TCOs (p-TCOs) exhibit mobility less than 0.1 cm2 V–1 s–1 because of the localized VB characters originating from O 2p states. The lack of high-mobility p-TCOs is currently one of the central issues in the development of optoelectronic device applications. 4)

Recently, SnO has attracted much attention as a candidate of high-mobility p-TCOs. It is a widegap oxide semiconductor having a direct bandgap of 2.7–2.9 eV and a layered structure (P4/nmm) [Fig. 1(a)]. 5,6) Spatially spreading Sn 5s orbitals overlap and hybridize with O 2p orbitals strongly, giving rise to the delocalized VB characters that enhance the mobility of hole carriers. 7) However, there are several issues with the application of SnO, one of which is chemical instability. The formal valence state of Sn2+O2– tends to be disproportionate to Sn0 and Sn4+O2– 2. Therefore, the synthesis of pure SnO, especially in the form of epitaxial films, is not as easy as conventional p-TCOs. Another issue is the lowering of transparency. It is known that transparency in SnO films is very sensitive to fabrication processes and/or conditions. 8) A number of defect models have been proposed for the origin of visible-light absorption. 810) Due to the intrinsic instability, however, there was no facile solution for eliminating such defects, which made it difficult to use SnO as a practical p-TCO.

Fig. 1.

Fig. 1. (Color online) (a) Schematic crystal structure of SnO. (b) Out-of-plane XRD profiles for the films at each step. Asterisks indicate peaks originating from a sample stage. (c) Raman spectra of as-grown films, reacted films, and pristine YSZ substrates along with reference SnO and SnO2 profiles. 21,22) (d) EDX spectra for each film.

Standard image High-resolution image

In order to tackle this problem, we focused on a soft chemical treatment with Na2S aqueous solutions. This method was once employed for a post-growth treatment of semiconductors including III–V compounds such as GaAs and InP. 11) According to past studies, the Na2S treatment was effective to passivate nonradiative defects and substantially enhances the near band-edge photoluminescence. 1114) Thus, Na2S treatment can be expected to bleach a visible-light absorption band by passivating defect states while preventing metastable SnO from disproportionating or decomposing.

In this Letter, we demonstrate the epitaxial growth of single-crystalline SnO films by using pulsed-laser deposition (PLD) technique and enhancement of the internal transmittance by Na2S treatment. We also reveal no deleterious effect of Na2S treatment on structural and electronic properties. In addition, our results indicate that Na2S treatment has great potential not only for surface passivation but also for anion exchange, i.e. full conversion from SnO phase to a SnS one.

SnO films were grown on yttria-stabilized zirconia (YSZ) (100) substrates by PLD with KrF excimer laser pulses (0.9 J cm−1, 4 Hz). A polycrystalline SnO tablet was used as a laser-ablation target. The substrate temperature was set to 450 °C and the deposition was conducted under a vacuum (base pressure ∼5 × 10−8 Torr). The film thickness was regulated to be ∼90 nm as verified by using a stylus profiler.

The obtained SnO films were immersed into a 0.01 M Na2S aqueous solution at RT for 5 h and 14 h. These periods were set by visual inspection of substantial changes in film color. Each sample is hereinafter referred to as "reacted" and as "fully-converted" films, respectively. The Na2S-treated films were washed with water and ethanol and then dried by N2 blow. The structural properties were investigated by an X-ray diffraction (XRD) apparatus with Cu Kα1 radiation. The surface morphology of films was investigated by atomic force microscopy. Energy-dispersive X-ray (EDX) spectroscopy was performed to assess the composition of films. Raman spectroscopy was performed for phase identification of SnO films. The excitation wavelength was 532 nm for all measurements. The temperature dependencies of longitudinal and Hall resistance were measured by a standard four-probe method using a physical property measurement system (Quantum Design, PPMS). Ti/Au metal electrodes were vacuum-evaporated on the films for making Ohmic contacts. The optical properties were investigated by UV–visible near-IR spectroscopy at RT. Photoemission spectroscopy (PES) was performed for samples transferred ex situ at the undulator beamline of BL-2A in the Photon Factory, KEK. The PES spectra were recorded using an electron energy analyzer (SES-2002, VG Scienta) with an energy resolution of less than 100 meV at phonon energy of 600 eV at RT. The binding energies were calibrated using the Fermi level (EF) of Au. For the VB spectrum of as-grown films, there were some contributions from the Au electrode. The contributions were subtracted for clarity. Ab-initio density-functional theory (DFT) calculations were performed using the Quantum ESPRESSO package 15,16) with the reported structural parameters for SnO. We used the Perdew–Burke–Ernzerhof generalized gradient approximation (PBE-GGA) functional 17) for the exchange–correlation potentials and projector-augmented wave pseudopotentials, 18) where valence electrons at Sn 5s, 5p, and 4d, and O 2s and 2p levels were considered. Brillouin zone sampling was performed using 8 × 8 × 7 k point meshes for self-consistent field calculations. The density of states (DOS) was obtained by the tetrahedron method with 12 × 12 × 10 k point meshes. The energy cut-off was set to 60 Ry and 500 Ry for the wave function and charge density, respectively. EF was defined as the highest occupied level.

Figure 1(b) shows the out-of-plane XRD profiles for samples at each step. After testing various parameters in terms of resulting crystallinity, we found out that a set of the aforementioned parameters is suitable for the preparation of each sample. Note that under our experimental setup the use of SnO target was essential for stabilizing single-phase SnO films and eliminating different oxidation states such as Sn0 and Sn4+. As-grown films exhibited only sharp SnO 00l reflections, indicating the (001)-orientated single-phase SnO. The calculated lattice constant along the c-axis of as-grown films (c = 4.85 Å) agreed with that of bulk, 19) suggesting near stoichiometric composition. The as-grown and reacted films exhibited nearly identical X-ray reflections to each other. In addition, they exhibited nearly identical low-angle reflectivity profiles, indicating similar thickness and density. Moreover, there was no trace suggesting degradation of the crystallinity and/or segregation of a secondary phase. These results indicate that the crystalline structure of SnO remained intact during Na2S treatment for 5 h. On the other hand, fully-converted films exhibited no SnO reflection, but a new peak at 2θ ∼32, 38, and 65°. These angles are close to those of SnS 400, 311, and 502 reflections, respectively, with relatively strong intensity (JCPDS: 73-1859). This fact suggests that anion-exchange reaction occurred and SnO was converted to SnS during extended Na2S treatment. The peak intensity of SnS phase is much weaker, suggesting a distorted anion framework. Such a distortion may arise from the large structural difference between SnO and SnS. 19,20)

Figure 1(c) shows Raman spectra taken for as-grown films, reacted films, and pristine YSZ substrates, and also those reported for SnO and SnO2. 21,22) Both films indicated peaks assignable to SnO phase with similar intensity, but no trace of SnO2 phase. These results suggest that Na2S treatment has no influence on chemical bonding states in SnO.

The film composition reflected the activity of anion-exchange reaction during Na2S treatment, which was clearly captured by EDX spectroscopy. Figure 1(d) shows EDX spectra normalized by the peak intensity of Sn Lα line. The spectrum of reacted films was nearly identical to that of as-grown ones except for different contributions of substrates to O signals. Moreover, both spectra had no signal derived from S. This result is consistent with the above-mentioned structural and chemical properties and meanwhile rules out anion-exchange reaction for reacted films. As for fully-converted films, S-derived peaks emerged and O-derived peaks became much smaller reflecting only contributions of substrates. Furthermore, the signal ratio of S to Sn (S/Sn) was estimated to be 1.05 as an average of data acquired at several measurement positions. The deviation of each data from the average was less than 3.2%. These results indicate that anion-exchange reaction occurred to convert SnO into SnS during long-period Na2S treatment. It is worth mentioning that the synthesis of stoichiometric SnS films is also a big challenge and our study reveals a facile method for preparing SnS from SnO precursor. The thermal annealing is known to improve the crystallinity of as-grown SnS. 23) Therefore, the combination of Na2S and thermal treatments may pave a good synthesis route for obtaining stoichiometric and epitaxial SnS films. 23) We also note that Na-containing residue and S incorporated in YSZ substrates were not detected both in XRD profiles and in EDX spectra.

Figure 2(a) shows the temperature (T) dependence of the longitudinal resistivity (ρ) for the as-grown and reacted films. The as-grown and reacted films indicated nearly identical curves as expected from similar structural properties. The inset of Fig. 2(a) shows ln ρ versus inverse temperature and linear fits with the thermal activation model: ρ = ρ0 exp [Ea/(kB T)], where ρ0 is a preexponential constant, Ea is the activation energy, and T is the absolute temperature. Ea was calculated to be 96 meV for both of as-grown and reacted films, respectively. The fully-converted films were found to be insulating with ρ higher than the measurement limit.

Fig. 2.

Fig. 2. (Color online) (a) Temperature dependence of the resistivity of as-grown and reacted SnO films. The inset shows linear fits with the thermal activation model. (b) Temperature dependencies of carrier concentration and Hall mobility for the same films. The inset shows the magnetic field dependence of Hall resistance R xy taken at 300 K.

Standard image High-resolution image

Next, we performed Hall measurements to reveal carrier concentration p and Hall mobility μ for as-grown and reacted films. Both films showed linear dependence of Hall resistance Rxy on the applied magnetic field H, as shown in the inset of Fig. 2(b) (taken at 300 K). The positive slope verifies the p-type conductivity. Figure 2(b) shows small temperature dependencies of p (left ordinate) and μ (right ordinate) at temperatures in a range of 280–300 K. The as-grown film showed p = 7.6 × 1017 cm−3 and μ = 0.4 cm2 V–1 s–1 at 300 K. These values are comparable to those reported for epitaxial films. 9,24) It should be mentioned that in one of the previous studies μ depended strongly on p and reached ∼20 cm2 V–1 s–1 at p of high-1016 cm−3 in SnO epitaxial films grown on YSZ (100) substrates. 24) Comparing our as-grown and reacted films, it is concluded that Na2S treatment has little impact on p-type conductivity as well as structural properties.

Despite a negligibly small change in structural and electronic properties, the optical properties indicated drastic changes. Figure 3(a) shows transmittance spectra and photographs for films at each step. The first average visible transmittance (AVT, external transmittance in a range from 1.5 to 3.0 eV) of as-grown films was calculated to be 29%, which is relatively low as a p-TCO even if the substrate attenuation is considered (gray line). Such a low transmittance is not only due to the relatively rough surface, but also inherent in SnO films, which is a subject for theoretical and experimental investigations of defect states. 810,2527) On the other hand, reacted films exhibited drastic color change and great enhancement of AVT by 17.6%. Considering that the structural and electrical properties are maintained, it is deduced that the enhancement of AVT is due to modulation of the electronic states in local areas such as the film surface and grain boundaries, which will be discussed later. Fully-converted films were opaque, reflecting the narrow band gap of SnS.

Fig. 3.

Fig. 3. (Color online) (a) Optical transmittance spectra for the films at each step and a double-side polished pristine YSZ substrate. Photographs of films are shown in inset with the same magnification (sample size 5 × 2–5 mm). (b) Optical absorption spectra for the same films. (c) VB spectra of as-grown and reacted films normalized by intensity at each maximum. (d) and (e) Band dispersion and projected DOS for the ideal structure of SnO obtained by ab-initio calculations, respectively. (f) and (g) AFM images for as-grown and reacted films, respectively. (h) Schematic drawing of a model of surface-passivation mechanisms.

Standard image High-resolution image

Figure 3(b) shows the absorption spectra extracted from transmittance and reflectance spectra using the measured film thickness. The direct bandgap of as-grown films was evaluated to be 2.9 eV, which agrees with the reported value for SnO. 9) The bandgap of reacted films was identical to that of as-grown ones, as we verified that Na2S treatment had no impact on the chemical bonding state of SnO lattice. In addition, Urbach energy was estimated to be 0.24 and 0.15 eV for as-grown and reacted films, respectively. In contrast, a broad absorption band seen in as-grown films at 1.5–2.7 eV became smaller and indicated an apparent peak shift to the higher energy side. Another broad band in the near-IR range became slightly smaller. As for fully-converted films, the direct bandgap was evaluated to be 1.2 eV, which agrees with the reported value for SnS. 28)

In order to investigate the electronic structure of our SnO films, we measured VB spectra for as-grown and reacted films [Fig. 3(c)]. Both of the spectra indicated two prominent components located at ∼5 eV and 11 eV. These features are very similar to those of SnO2, 29) reflecting a tendency for surface oxidation. On the other hand, the intensity of the shoulders seen at 2‒4 eV, which is the characteristic state of SnO, increased after the reaction, suggesting the increase of SnO-like phase at the surface. The reduction of the peak at 11 eV is also reasonable. 24) Some of these results are worth seeing together with calculated band dispersion and projected DOS for the ideal structure of SnO obtained by the ab-initio DFT calculation [Figs. 3(d) and 3(e)]. The direct and indirect bandgaps are seen at Γ-point and along Γ-M line, respectively, being consistent with the previous study. 26) The absolute energies of bandgaps were underestimated as usual in DFT calculations, 30) compared with experimental values of the reported indirect bandgap and direct one of our films (0.7 eV 5,9) and 2.9 eV, respectively). Here we notice that the bandgap lies approximately above ‒2 eV and the projected DOS purely reflects bonding characters below ‒2 eV. In addition, the Sn 3d core level spectra (not shown) mainly consisted of the Sn4+ component for both films. Despite the surface oxidation layer, occupied Sn 5s and 5p states inherent to SnO could be accessible by PES because SnO2 is n-TCO being transparent electronically near EF.

The mechanism of Na2S treatment can be further deduced as follows. As we already mentioned, locations of the passivated defect states are limited to a local area near the surface and grain boundaries. We examined the surface morphology of the films before and after the Na2S treatment. As shown in Figs. 3(f) and 3(g), overall surface roughness became larger after the treatment, which might cause the reduction of transmittance. Nevertheless, AVT was substantially enhanced after the treatment. On the other hand, small corrugations that seem to distribute randomly on the as-grown films were found to be more regularly and densely rearranged on the reacted films. Such a change can be attributed to passivated surface defects. But, further characterizations are necessary to identify the crystalline structure and chemical bonding state near the surface in an atomic scale.

Aspects of surface-passivation with Na2S were once studied for III–V semiconductors, 1114) where S atoms were thought to passivate surface dangling bonds to form, for example, an As–S–As bridging structure on GaAs. 14) In the case of SnO, the atomistic structure of surface defects has been rarely studied. As for defects in bulk, however, Sn vacancies (VSn 2–) acting as shallow acceptors are suggested to be the dominant defects under O-rich conditions from the first-principle calculation. 26) This model explains the positive correlation between hole concentration and magnitude of visible-range absorption. 8,25) Therefore, VSn 2– are the most likely acceptors in our films that are uniformly distributed to generate hole carriers. We note that the tendency of surface oxidation can be associated with condensation of neutral Sn vacancies (VSn 0) near the surface (i.e. Sn1‒x O). After the Na2S treatment, the increase of SnO-like phase in VB spectra [Fig. 3(c)] and Sn 3d core-level spectra (not shown) were observed, which corresponds to the decrease of VSn 0. Assuming that dense VSn 0 near the surface mainly causes visible-light absorption (but does not contribute to generating hole carriers), a mechanism of Na2S treatment can be deduced. Figure 3(h) shows a model of surface-passivation mechanisms, where VSn 0 situated at the apical site of a square-pyramidal block can be passivated with S atoms without breaking net charge neutrality. The resulting local SO4 blocks bridge the oxygen network to protect the area near the surface from continuous S-atom intercalation. Meanwhile, the S-passivated surface eventually triggers an anion-exchange reaction: as SO4 2– ions gradually dissolve in water, S2– ions take over the space to form distorted SnS flameworks. Taking such a late-determination process into account, the observed longer waiting time for sulfurization can be explained. Taken together, Na2S treatment can provide a unique and facile solution for manufacturing high-performance devices upon surface-defect engineering. It can be noticed that surface states hinder ambipolar operations of SnO-based thin-film transistors. 31,32)especially when acceptor-like midgap states exist. 33)

In conclusion, we have successfully grown SnO epitaxial films and investigated the effects of the Na2S treatment. XRD, Raman, and EDX spectroscopy measurements revealed that (001)-oriented SnO epitaxial films remained intact after the short-period treatment while completely converted into SnS by the long-period treatment. Electrical measurements revealed nearly identical p-type conduction for as-grown and reacted films produced by the short-period treatment. Moreover, optical measurements revealed substantial enhancement of transparency despite negligibly small changes in structural and electronic properties. Based on the absorption spectra, VB spectra, ab-initio DFT calculations, and the previous studies on III–V semiconductors, a possible mechanism of this enhancement is due to passivation of midgap states near the surface by S atoms. This method can be easily and effectively used for achieving high transparency in SnO.

Acknowledgments

This work was partly supported by MEXT Elements Strategy Initiative to Form Core Research Center (JPMXP0112101001) and JSPS KAKENHI (20K15169, 21H02026). This work was the result of using research equipment (Raman spectroscopy) shared in MEXT Project for promoting public utilization of advanced research infrastructure (Program for supporting the introduction of the new sharing system, JPMXS0420900521). The work at KEK-PF was conducted with the approval of the Program Advisory Committee (Proposals No. 2021G660, and No. 2021G683) at the Institute of Materials Structure Science, KEK, Japan.

Please wait… references are loading.
10.35848/1347-4065/ac56fa