n- and p- Type Wide Bandgap Transparent Oxide Semiconductors via Solution Processing and Physical Vapor Deposition Methods

溶液法與物理氣相沉積法製備負型與正型的寬禁帶透明半導體氧化物

Student thesis: Doctoral Thesis

View graph of relations

Related Research Unit(s)

Detail(s)

Awarding Institution
Supervisors/Advisors
Award date26 Aug 2022

Abstract

Transition metal oxide semiconductors (TMOSs) have found multifarious applications in photovoltaic and optoelectronic devices due to their wide bandgaps, as well as high electrical conductivity and high optical transparency over a broad spectral range. For instance, Sn doped In2O3 (ITO), F doped SnO2 (FTO) and Al doped ZnO (AZO) are well developed, commercially available transparent conducting oxides (TCOs) that have been extensively used as transparent conductors in solar cells and optoelectronic devices. However, due to their rather low electron mobility (~20-50 cm2/Vs), their high conductivity is primarily achieved by the high electron concentration, which results in a strong free carrier absorption and plasma reflection at the near infrared region (~1000 nm). This severely limits the device applications of these TCOs utilizing photons from the entire solar spectrum. Among TMOSs, CdO has an extremely high electron mobility due to its small electron effective mass and a large static dielectric constant, which screens electrons and reduces scatterings from ionized impurities. It also has a high proclivity of n-type doping because of its high electron affinity of ~5.9 eV. However, its UV transparency is significantly hampered by its narrow intrinsic direct bandgap of ~2.2 eV. This dissertation first aims to widen the bandgap of CdO and further improve its transport and optical properties by alloying with In2O3 which has a much wider bandgap of 3.6 eV. In particular, instead of the commonly used physical vapor deposition method (sputtering) we explored the low-cost sol-gel method for this work.

On the other hand, only a few TMOSs exhibit p-type conductivity and p-type TMOSs with comparable performance to their n-type counterparts are still lacking. This can be attributed to the low valence band maximum position and flat O 2p-orbital derived valence band dispersion of most TMOSs. The lack of feasible p-type TMOS severely hinders the development of oxide-based transparent bipolar optoelectronic devices. Tin monoxide (SnO) with relatively less localized valence band states, is recently a popular study gaining attention to be a promising p-type TCO. The more dispersive VB arises from the hybridization of the Sn 5s and O 2p orbitals due to the proximity in their energy position. However, SnO is a metastable phase which can be easily oxidized to n-type SnO2 which has a higher valence state. Next in this dissertation, a systematic investigation was carried out to improve the p-type conductivity and phase stability of SnO by controlling the native defects and through extrinsic acceptor doping.

While various material properties and applications as well as details on the thin film growth and characterization are respectively introduced in Chapter 1 and 2, we explored the synthesis of undoped CdO films using a sol-gel solution process and compared their properties with films synthesized by the conventional sputtering technique in Chapter 3 of this thesis. By optimizing the deposition parameters such as precursor molarity, spinning rate and duration, and the drying temperature followed by spinning, highly smooth CdO films with high conductivity and high transparency are obtained. The effects of post-growth annealing at elevated temperature on the native vacancy defects and electronic and optical properties of CdO was further investigated.

In addition, the electrical and optical properties of CdO films were improved by small amount of In (up to 10%) doping. This was achieved by adding the corresponding ratio of In precursor into the CdO solution. Significant increases in the conductivity and electron concentration were demonstrated. Density functional theory calculations reveal that in both Cd-rich and O-rich growth environment, substitutional In donor defects (InCd) have the lower formation energy than both cadmium interstitials and oxygen vacancy donor defects in CdO. Nevertheless, while the electron concentration n increases with In concentration, n saturated at ~5×1020 cm-3 for In doping >5%. This low activation of In may be attributed to the high density of native defects and/or impurities incorporated in the sol-gel process. With an optimum 5% In doping, CdO can achieve a low resistivity of ρ~2.5×10-4 Ω-cm and a high mobility μ~50 cm2/Vs, comparable to conventional commercial TCOs grown by physical vapor deposition methods. Moreover, high transparency of >85% over a wide transmission window of 400 to >1650 nm was achieved.

This dissertation further explores the alloying of structurally mismatched CdO and In2O3 (or Cd1-xInxO1+δ) over the entire composition to achieve amorphous alloys with high conductivity and transparency in Chapter 4. We demonstrated that sol-gel spin-coated Cd1-xInxO alloys are amorphous within a wide composition range of 0.2≲x≲0.7 with tunable bandgaps ranging from ~2.5 to 3.5 eV. These amorphous films exhibit a low ρ in the low 10-3 Ω-cm and a μ of ~10 cm2/Vs after annealing at ~400oC. The desirable optical and electrical properties of amorphous alloy films obtained by low temperature, low-cost non-vacuum processing are promising for optoelectronic applications, especially on organic layers and/or flexible substrates.

Lastly, for the fabrication of p-n junctions, phase stability and optoelectronic properties of p-type SnO synthesized by magnetron sputtering were examined in Chapter 5 in this dissertation. In particular, the addition of a dilute amount of excess Sn was found to be an effective approach to regulate the film stoichiometry so that a low p-type resistivity can be achieved after post-growth rapid thermal annealing (RTA). The p-type resistivity of these Sn-rich SnO (SnO1-δ) films decreases with increasing δ. However, metallic β-Sn clusters were formed for films with δ>~0.05. The presence of these β-Sn clusters resulted in a degraded optical transmittance by ~10% in the visible range. With δ~0.03 in the as-grown film, a p-type pure phase SnO film with a low resistivity of ~0.5 Ω-cm, and a decent visible transparency of ~60% was achieved after 300oC RTA in N2. This p-type SnO has a wide band gap of 2.8 eV with a high valence band maximum (VBM) located in the range of 4.6-4.8 eV below the vacuum level, making it suitable for many device applications, particularly as hole transport layers in optoelectronic devices. In order to further enhance the p-type conductivity of SnO, extrinsic doping with acceptor species Ga and Na was also explored. While both species were found to be effective acceptors and enhanced the p-type conductivity of SnO, Ga acceptors (GaSn) were found to be less stable than NaSn. The lowest achievable p-type resistivity for SnO thin films was 0.2 and 0.01 Ω-cm with 1.2% Ga and 2.8% Na doping, respectively. Moreover, both NaSn and GaSn acceptors were found to be relatively shallow with respective ionization energies at 52 and 20 meV. Our experimental results are compared and discussed with first-principles calculations reported in the literature.