Suman Bhandari

Ultrawide bandgap semiconductors (UWBGs) provide great promise for optical devices operating in the near to deep ultraviolet, and recently they have become a viable semiconducting material for high power electronics. From the power grid to electronic vehicles, the intention is to replace massively awkward components with the convenience of a solid state electronic “chip.” Unfortunately, the challenges faced by wide bandgap electronic materials, such as GaN and SiC, increase as the bandgap increases. A point defect, for example, can take on more charge states and energy configurations. This perspective describes a method to investigate the many charge states and their associated transitions—photo-induced electron paramagnetic resonance (photo-EPR) spectroscopy. Although not new to the study of defects in semiconductors, photo-EPR studies can probe the entire ultrawide bandgap given the appropriate light source for excitation. Examples provided here cover specific defects in UWBGs, AlN, and Ga2O3. The discussion also reminds us how the rapid pace of discovery surrounding this newest class of semiconductors is due, in part, to fundamental research studies of the past, some as far back as a century ago and some based on very different materials systems.

Gallium oxide (Ga₂O₃), a promising candidate for high-power electronics, may be doped with Fe to control the native n-type conductivity and produce semi-insulating single-crystalline substrates. Here, intentionally as well as unintentionally Fe-doped β-Ga₂O₃ crystals grown by the Czochralski (Cz) and the floating zone (FZ) methods are studied. While the Cz samples contain Ir on the order of 10¹⁷ cm⁻³ due to the Ir crucible, the FZ samples are free of this contaminant. Photoinduced electron paramagnetic resonance (EPR) is performed by illuminating the samples with light emitting diodes from 0.7 to 4.7 eV while monitoring the intensity of the EPR signal. The following optical transitions associated with Fe are observed: one between 1 and 2 eV occurs in both Cz and FZ samples, another near 2.5 eV is unique to Cz Ga₂O₃, and a third near 3 eV appears in FZ material. The transition below 2 eV is the well-studied Fe²⁺-to-Fe³⁺ transition, intrinsic to Fe. Near 2.5 and 3 eV, the amount of Fe³⁺ decreases by capturing electrons excited from other defects present near or below the middle of the bandgap. By comparing Cz and Ir-free FZ crystals, we show that the 2.5 eV transition is primarily caused by Ir in Fe-doped Cz samples, whereas the ∼3 eV transition observed in FZ samples could be caused by isolated or complex intrinsic defects. By differentiating transitions between Cz and FZ material, the results will be relevant to epitaxial films, which do not contain the Ir present in the commonly studied Cz material.

Fe acts as an electron trap in gallium oxide (Ga₂O₃), thereby producing a semi-insulating material that can be used in device fabrication. However, such trapping can lead to negative effects when Fe is unintentionally incorporated into bulk crystals or thin films. In this work, photoinduced electron paramagnetic resonance (photo-EPR) is used to investigate carrier capture at Fe in β-Ga₂O₃. Two crystals doped with 8 × 10¹⁷ cm⁻³ and 5 × 10¹⁸ cm⁻³ Fe and one Mg-doped crystal containing 7 × 10¹⁶ cm⁻³ unintentional Fe are studied by illuminating with LEDs of photon energies 0.7–4.7 eV. Steady state photo-EPR results show that electrons excited from Ir, an unintentional impurity in bulk crystals, are trapped at Fe during illumination with photon energy greater than 2 eV. Significantly, however, trapping at Fe also occurs in the crystals where Ir does not participate. In such cases, we suggest that excitation of intrinsic defects such as oxygen or gallium vacancies are responsible for trapping of carriers at Fe. The results imply that the investigation of intrinsic defects and their interaction with Fe is necessary to realize stable and reliable Ga₂O₃:Fe devices.

Impurities such as Fe and Mg are intentionally incorporated into Ga₂O₃ to control the intrinsic n-type conductivity. This work examines the defect level of the intentional and unintentional impurities in β-Ga₂O₃ substrates grown for electronic applications. First, several dominant impurities, Fe³⁺, Mg⁰, and Ir⁴⁺, are identified using electron paramagnetic resonance spectroscopy. Then, by illuminating the crystals with selected wavelengths between 1500 and 300 nm, the photothreshold for excitation or quenching of each impurity is examined. Defect levels are estimated from the photothreshold after considering interdefect transitions and lattice relaxation. Thresholds for Ir⁴⁺ (2.0–2.3 eV) and Fe³⁺ (1.2–1.5 eV) support the reported values of the Ir^3+/4+ and Fe^2+/3+ defect levels, but the photothreshold of 1.6–1.9 eV observed for the quenching of neutral Mg requires further consideration.

This study investigates the Fe impurities believed to act as deep acceptors that contribute to electrical compensation of the n-type conductivity in as-grown Ga₂O₃. A variation of the traditional optical absorption measurement, photoinduced electron paramagnetic resonance (EPR) spectroscopy, is used to identify charge transitions in bulk Fe-doped and Mg-doped Ga₂O₃ with the support of hybrid functional calculations. Steady-state photo-EPR measurements show that the first optically induced change in Fe³⁺ occurs at 1.2 eV, significantly larger than the calculated defect levels for Fe. However, the optical cross section spectrum determined from time-dependent photo-EPR measurements compare well with a calculated cross section spectrum for the Fe²⁺-to-Fe³⁺ transition when the relaxation energy predicted from the density functional theory is folded into the model. This work explicitly demonstrates the need for an accurate accounting of electron-lattice coupling when interpreting optically induced phenomena.