Gallium oxide semiconductor-based large volume ultrafast radiation hard spectroscopic scintillators

Scintillators play a crucial role in various applications, including nuclear security and nonproliferation, nuclear and high energy physics, medical imaging, and security screening. An ideal scintillator should simultaneously possess several key properties, such as high light yield, fast decay times, and good spectroscopic capabilities. High light yield ensures a strong signal output, while fast decay times enable precise timing resolution and application in high flux scenarios. Spectroscopic properties allow for the discrimination of different types of radiation and the identification of specific radionuclides. Despite extensive research efforts, finding a single high-density scintillator that combines all these desirable properties has remained a challenge. Many commercially available high-density inorganic scintillators suffer from long decay times, ranging from hundreds of nanoseconds to tens of microseconds, which limits their performance in applications requiring fast timing capabilities. Moreover, radiation damage in crystalline scintillators can lead to decreased light output, reduced response uniformity, and enhanced phosphorescence (afterglow), resulting in increased readout noise and recurring replacement costs.^1–5 These limitations hinder the performance of scintillators in high dose rate environments, such as those encountered in calorimetry applications.^6,7

The development of a scintillator with high light yield, fast decay rates, and spectroscopic properties is of great importance for numerous applications. In medical imaging such as positron emission tomography (PET), single-photon emission computed tomography (SPECT), and recently photon counting CT (PCCT), fast and efficient scintillators can improve the timing resolution and overall image quality, leading to better diagnostic accuracy and patient care.^8–10 In nuclear and high-energy physics experiments, scintillators with fast decay times and high light yield are essential for precise timing measurements and the detection of rare events.^11,12 Furthermore, in security applications, such as radiation portal monitors and cargo screening, scintillators with good spectroscopic capabilities can enhance the detection and identification of radioactive materials, ensuring public safety. For defense-related applications such as radiologically contaminated battlefields, the need for a radiation-hard scintillator with ultrafast counting capabilities is still unmet. β-Ga₂O₃ semiconductor has emerged as a promising scintillator material since the late 2010s.^13–16 Its wide bandgap, high electron mobility, and good thermal stability make it an attractive candidate for radiation detection applications. However, to fully exploit the potential of β-Ga₂O₃ as a scintillator, it is necessary to optimize its properties and overcome the limitations commonly associated with existing scintillator materials.

In this study, we investigate the scintillation properties of β-Ga₂O₃ and explore strategies to enhance its light yield, decay times, and spectroscopic capabilities. By addressing these key aspects, we aim to develop a high-performance scintillator that can meet the demanding requirements of various applications. The successful development of an improved β-Ga₂O₃ scintillator will represent a significant advancement in the field of radiation detection and can potentially revolutionize the performance of scintillator-based systems across multiple domains. The wide bandgap (4.9 eV) single crystal β-Ga₂O₃ is an extremely robust oxide scintillating material, not temperature- and moisture-sensitive, and is highly radiation-hard. The temperature dependence of the β-Ga₂O₃ scintillator light yield was investigated within the range of 20–50 °C. Our measurements revealed no significant variation in the light output of the β-Ga₂O₃ scintillators within this temperature range. However, it is worth noting that the light yield of β-Ga₂O₃ significantly increases at lower temperatures, as reported by Drozdowski et al.¹⁷ It has been shown that the properties of β-Ga₂O₃ electronic devices remain unchanged after irradiation with a total 1.25 MeV ⁶⁰Co gamma radiation dose of 1.6 MGy,¹⁸ indicating minimal generation of bandgap defects. In a previous study, we demonstrated the excellent radiation hardness properties of the float zone (FZ) grown β-Ga₂O₃ scintillators.¹⁹ All the β-Ga₂O₃ crystals mentioned in this study were also grown using the fast turnaround FZ technique (2 days including feed rod preparation). By optimizing the growth conditions and intentionally introducing native defects, such as oxygen vacancies (V_O) and gallium–oxygen vacancy pairs (V_Ga–V_O), we developed a high-performance scintillator that can combine the various requirements from different applications. The scintillation properties of the FZ-grown β-Ga₂O₃ crystals are discussed in Sec. III, focusing on the impact of native defects on the light yield and decay times.

The details of the FZ crystal growth process and scintillation measurements have been published previously.¹⁹ The FZ growth process is based on the relative translation of a heat source along the length of the feed rod. The heat source creates a molten zone, which translates along with it. Under proper conditions, the material in the molten zone continuously solidifies as a single crystal. At the end of the growth process, most of the material in the feed rod is transformed into a single crystal. The formation of the feed rod is a critical element in the success of the FZ process, in terms of both very low porosity and compositional uniformity. For the growth of large volume crystals, larger feed rods were fabricated with a diameter of 18 mm and a height of 20 cm. The pushdown rate of the feedrod was also varied to maintain the diameter of the β-Ga₂O₃ crystals. For some of the crystals, an FZ furnace with higher power (12 kW) was used. The increase in the β-Ga₂O₃ crystal volume and enhancements in crystal quality over the past two years is shown in Fig. 1.

The scintillation characteristics of the β-Ga₂O₃ detectors were evaluated using a fast Hamamatsu photomultiplier tube (PMT). Light yield measurements were performed by comparing the β-Ga₂O₃ scintillators with a reference bismuth germanate (BGO) detector obtained from Epic crystals, taking into account the differences in the quantum efficiency (QE) of the PMT at the emission wavelengths of BGO and β-Ga₂O₃. For scintillation decay time measurements, the PMT output signal was directly fed into a high-speed Tektronix oscilloscope, and the decay curves were obtained by averaging over 10 000 individual scintillation events. To validate the accuracy of our experimental setup and minimize measurement-related errors, we also characterized the decay times of well-established fast scintillators, including EJ-208 plastic scintillator and yttrium-doped barium fluoride (BaF₂:Y). The measured decay times of these reference materials were consistent with their reported values in the literature, confirming the reliability of our measurement technique. Well-established techniques for characterizing the light yield and decay time of scintillators were employed in this study. The free electron concentrations were measured using the Hall method.

The growth conditions including growth atmosphere and growth rates were optimized to achieve the desired native defect structure in the β-Ga₂O₃ crystals, which played a crucial role in enhancing their scintillation properties. Although the band edge transition for β-Ga₂O₃ lies in the ultraviolet region between 100 and 280 nm, fluorescence spectroscopy revealed that the emission peaked at 483 nm (∼2.6 eV) (Fig. 2). This emission wavelength exhibits an excellent spectral match with the high quantum efficiency region of certain photodetectors, enabling efficient light collection and detection. The optimized native defect structure, particularly the presence of oxygen vacancies (V_O) and gallium–oxygen vacancy pairs (V_Ga–V_O), contributed to the impressive scintillation performance of the β-Ga₂O₃ crystals. The bandgap energy levels created by these defects serve as efficient recombination centers for electron–hole pairs generated by incident radiation, leading to increased light emission. The presence of gallium–oxygen vacancy pairs (V_Ga–V_O) creates complex defect structures that can facilitate rapid charge transfer and recombination processes, contributing to ultrafast decay times observed in our β-Ga₂O₃ scintillators. The optimized concentration and distribution of these native defects help create a favorable energy level structure within the bandgap. It allows for efficient energy transfer from the host lattice to the emission centers, resulting in improved energy resolution and spectroscopic performance.

Carrier concentration is a manifestation of these traps related to the material stoichiometry (as related to growth conditions), dopants, and impurities in the crystal. Free carrier concentrations were measured by the Hall method for each grown β-Ga₂O₃ crystal. In n-type semiconductors, oxygen vacancies (V_O) act as donor defects. These defects can contribute to the free carrier concentration by donating electrons to the conduction band or by compensating acceptors. Therefore, as the concentration of oxygen vacancies increases in the material, the free carrier concentration (electron concentration) in n-type β-Ga₂O₃ is also expected to increase. The concentration of oxygen vacancies in β-Ga₂O₃ can be influenced by the growth conditions, such as the oxygen partial pressure and the growth temperature and rate. Lower oxygen partial pressures and higher growth temperatures and rates favor the formation of oxygen vacancies. Also, the presence of gallium–oxygen vacancy pairs (V_Ga–V_O) can lead to more complex behaviors in terms of free carrier concentration compared to isolated oxygen vacancies. Our focus in this study was solely on the net carrier concentration rather than the manipulation of various defects to achieve the target net-carrier concentrations. The free carrier concentration of the n-type devices in this study was targeted to be 6 × 10¹⁷ cm⁻³. The growth tuning parameters were adjusted accordingly. The oxygen content (0–90%) in the growth atmosphere was balanced with the growth rate (1–10 mm/h) and temperature (percent power of the FZ lamps) to obtain the optimized ultrafast decay rate scintillators along with high light yield and spectroscopic capabilities. Table I presents an example of the optimization process for n-type β-Ga₂O₃ growth atmosphere, along with the corresponding free carrier concentration values obtained through Hall measurements. The data reveal a clear correlation between the free carrier concentration and the scintillation light yield for these floating zone (FZ) grown crystals. However, no apparent correlation is observed in the primary decay times. These findings differ from those reported in the literature,^16,17 which we attribute to the variations in growth techniques and the presence of growth-specific defects, as discussed below. The optimized growth condition (GO4) with 20% O₂ and 80% Ar atmosphere yields the highest scintillation light output of 8000 Ph/MeV among the investigated samples while maintaining a fast primary decay time of 2 ns. This highlights the importance of fine-tuning the growth atmosphere to achieve the desired scintillation properties in β-Ga₂O₃ crystals.

From our study, it was clear that the transparency and the color of the β-Ga₂O₃ crystals are also a good indication of its scintillation properties. Scintillators grown using unoptimized crystal growth processes can produce colored crystals, as shown in 2022–2023 part of Fig. 1. The color of β-Ga₂O₃ crystals grown by different methods, such as FZ, edge-defined film-fed growth (EFG), vertical Bridgman (VB), and Czochralski (CZ) techniques, can be attributed to various defects and impurities present in the material. The specific color observed depends on the type and concentration of these defects, which can vary based on the growth method and conditions employed.

The blue color in FZ, Cz, and EFG grown β-Ga₂O₃ crystals is typically associated with the presence of excess oxygen vacancies (V_O) and gallium–oxygen vacancy pairs (V_Ga–V_O) (Fig. 1).^19,20 These defects can introduce energy levels within the bandgap of β-Ga₂O₃ that leads to the absorption of longer wavelength light and the emission of blue photons. The free carrier concentration of these blue n-type crystals is >8 × 10¹⁷ cm⁻³. The formation of these defects is favored under the high-temperature and low-oxygen partial pressure conditions often encountered in the FZ growth process. Additionally, the rapid growth rates achievable in the FZ, EFG, and Cz methods can result in a higher concentration of these defects, contributing to the intense blue color.

The gray color often observed in EFG-grown β-Ga₂O₃ crystals may be attributed to the presence of impurities, such as carbon or silicon, which can be introduced during the growth process.²⁰ These impurities can form deep-level defects within the bandgap of β-Ga₂O₃, leading to the absorption of light across a wide range of wavelengths and resulting in a gray appearance. The EFG method involves the use of a die, which can be a source of impurities if not properly cleaned or maintained. Additionally, the relatively high growth rates employed in the EFG technique may limit the ability to effectively suppress the incorporation of impurities, contributing to the gray color.

The yellowish tint in Cz and VB-grown β-Ga₂O₃ crystals may be related to the presence of gallium vacancies (V_Ga) and their complexes with other defects or impurities.²¹ Gallium vacancies can introduce energy levels near the valence band of β-Ga₂O₃, leading to the absorption of shorter wavelength light and the emission of yellowish photons. The CZ growth method typically involves slower growth rates and higher oxygen partial pressures compared to the FZ and EFG techniques, which can favor the formation of gallium vacancies over oxygen vacancies. Additionally, the use of iridium crucibles in the CZ process may introduce trace amounts of iridium impurities, which can contribute to the yellowish color.

A high light yield of approximately 8000 photons/MeV was obtained from a 2 mm thick β-Ga₂O₃ scintillator (Fig. 3) under 662 keV gamma excitation. This light yield significantly surpasses those of widely used scintillators such as PbWO₄ (100–300 photons/MeV), the ultrafast component of BaF₂ (1900 photons/MeV), and Lu₂O₃:Yb (500 photons/MeV). Moreover, the β-Ga₂O₃ scintillators demonstrated excellent spectroscopic capabilities, achieving an energy resolution of 7% at 662 keV (Fig. 4). This spectroscopic response is attributed to the high light yield and emission characteristics of the native defects, which allow for precise energy discrimination. In contrast, no 662 keV peak is visible in the spectroscopic response of the PbWO₄ and Lu₂O₃:Yb ultrafast scintillators.²² To investigate the effect of scintillator thickness on the light yield, β-Ga₂O₃ crystals with thicknesses up to 20 mm were tested. A light yield of 5200 photons/MeV was obtained for the 20 mm thick scintillator (Fig. 5), which is lower compared to the 2 mm thick sample. This reduction in light yield can be attributed to the presence of defects and subgrain boundaries within the thicker crystal, leading to increased scattering of the scintillation light. Surprisingly, this reduction (a ratio of 1:0.65) is similar to the Cz grown β-Ga₂O₃ crystals as shown in a previous study.²³ However, we believe that further optimization of the crystal growth process can mitigate these issues and enhance the light yield of thicker β-Ga₂O₃ scintillators to levels comparable to the 8000 photons/MeV achieved with the 2 mm thick sample. The combination of high light yield, good spectroscopic response, and the ability to produce scintillators with various thicknesses highlights the great potential of β-Ga₂O₃ as a versatile and high-performance scintillator material. The optimization of the native defect structure through careful control of the growth conditions has proven to be an effective strategy for achieving enhanced scintillation properties in β-Ga₂O₃ crystals.

In addition to the high light yield and good spectroscopic response, the β-Ga₂O₃ scintillators exhibited ultrafast timing response characteristics. Time-resolved spectroscopic measurements revealed that the first decay time component (>90%) of the scintillation emission for all the tested β-Ga₂O₃ detectors (over 50 detectors with various thicknesses up to 20 mm) was about 2 ns; an example is shown in Fig. 6. This best-obtained ultrafast decay time of 680 ps is close to or even surpasses that of well-known ultrafast scintillators such as BaF₂ and Yb-doped scintillators like Lu₂O₃:Yb (Fig. 7). The ultrafast response of the β-Ga₂O₃ scintillators can be attributed to the efficient radiative recombination processes associated with the native defects, particularly the oxygen vacancies (V_O) and gallium–oxygen vacancy pairs (V_Ga–V_O). These defects introduce deep energy levels within the bandgap that facilitate rapid charge carrier recombination, leading to the emission of scintillation photons with minimal trapping and recombination.

This optimized result is possible due to the unique combination of growth conditions and the resulting defect structure that is difficult to achieve using other growth techniques, such as edge-defined film-fed growth (EFG) or Czochralski (CZ) methods. One of the key factors contributing to the ultrafast scintillation in FZ-grown β-Ga₂O₃ is the ability to precisely control the oxygen partial pressure and growth rate during the growth process. The FZ method allows for the growth of crystals under a wide range of oxygen partial pressures, from near-vacuum conditions to oxygen-rich atmospheres. By carefully tuning the oxygen content and growth rate, it is possible to achieve the optimal concentration of oxygen vacancies (V_O) and gallium–oxygen vacancy pairs (V_Ga–V_O), which are crucial for ultrafast scintillation performance. The high growth rates achievable in the FZ process (up to 50 mm/h) further facilitate the formation of these defects, as rapid solidification can limit the diffusion and annihilation of vacancies. In contrast, the EFG and CZ methods typically operate under more restrictive growth conditions, with limited control over the oxygen partial pressure and growth rate. The EFG process often requires the use of a reducing atmosphere to prevent the oxidation of the molybdenum or tungsten dies, which can hinder the formation of oxygen vacancies. Similarly, the CZ method typically employs a more oxygen-rich atmosphere to maintain the stability of the iridium crucible, which may suppress the formation of oxygen vacancies and favor the creation of other types of defects, such as gallium vacancies (V_Ga) or impurity-related defects. Another critical aspect of the FZ growth process is the high purity of the starting materials and the minimization of impurities during crystal growth. The FZ method does not require the use of a crucible, as the crystal is grown from a polycrystalline feed rod, which can be purified to a high degree prior to growth. This reduces the risk of contamination from crucible materials, such as iridium in the CZ process or refractory metals in the EFG technique. Impurities can introduce deep-level defects that compete with the ultrafast scintillation centers, leading to slower decay times and reduced light output. Furthermore, the high-temperature gradients and rapid cooling rates achieved in the FZ process can promote the formation of a more homogeneous defect distribution throughout the crystal. The steep temperature gradients and fast cooling rates lead to rapid solidification of the crystal. This quick transition from liquid to solid state can “freeze in” defects, particularly native point defects like oxygen vacancies (V_O) and gallium–oxygen vacancy pairs (V_Ga–V_O), more uniformly throughout the crystal volume. The rapid cooling limits the time available for defect migration and aggregation. This helps prevent the formation of defect clusters or non-uniform distributions that might occur under slower cooling conditions. Additionally, the fast growth rates achievable in the FZ process (up to 50 mm/h) can help suppress macroscopic segregation of impurities and defects, contributing to a more homogeneous crystal composition. These are essential for maintaining consistent scintillation properties across the entire crystal volume. In contrast, the EFG and CZ methods may result in a more inhomogeneous defect distribution due to the presence of temperature gradients and the slower cooling rates, which can lead to variations in scintillation performance within the same crystal.

The ultrafast response of β-Ga₂O₃ scintillators has implications for their performance in high-count rate scenarios. The rapid decay of the scintillation emission minimizes the likelihood of pulse pileup effects, where the scintillator is unable to resolve individual events due to the overlap of scintillation pulses. The next step of this study is to estimate the coincidence timing resolution (CTR) of β-Ga₂O₃ detectors based on their rise rime, decay time, and light yield. By reducing pulse pileup and high CTR, β-Ga₂O₃ scintillators can maintain their performance and energy resolution even under high radiation fluxes, making them suitable for applications such as time-of-flight positron emission tomography (TOF-PET), ultrafast high flux x-ray counting for CT, and time stamping in physics experiments.

The FZ-grown β-Ga₂O₃ scintillators demonstrated excellent scintillation linearity over a wide range of gamma-ray energies, spanning from low-energy x rays to high-energy gamma-rays (Fig. 8). The response linearity, defined as the ratio of the scintillation yield per unit energy at a given energy to that at a reference energy (typically 662 keV), remained within ±5% across the investigated energy range. This near-linear response is essential for accurate energy calibration and high-resolution spectroscopic measurements, as it minimizes the variations in light yield that can lead to energy resolution degradation. The scintillation yield was found to increase linearly with the deposited energy up to the maximum tested energy of 1.27 MeV (²²Na), indicating the absence of significant saturation effects or nonlinear quenching mechanisms. The origins of this linear response can be attributed to several factors, including efficient charge carrier transport, minimal electron-hole trapping, and the absence of significant nonradiative recombination processes that often contribute to nonlinearity in other scintillator materials. The near-linear behavior of FZ-grown β-Ga₂O₃ scintillators was compared to that of well-established commercial scintillators, such as NaI(Tl), LaBr₃(Ce), and CsI(Tl) (Ref. 24, Fig. 3). The scintillation response linearity of β-Ga₂O₃ was found to be comparable or even superior to these benchmarks. This superior performance can be ascribed to the favorable electronic band structure and the presence of shallow donor states associated with the optimized native defects in FZ-grown β-Ga₂O₃, which facilitate efficient charge carrier capture and radiative recombination across a wide range of excitation densities. The observed 7% energy resolution at 662 keV may be influenced by factors beyond just the response linearity shown in Fig. 8, including potential nonproportionality effects related to ionization density. A more comprehensive study of true nonproportionality will require additional investigation.

The combination of high light yield, good spectroscopic response, and ultrafast decay time positions β-Ga₂O₃ as a promising scintillator material for a wide range of applications demanding both high sensitivity and fast timing capabilities as mentioned in Sec. I. The ability to tailor the native defect structure through optimization of the growth conditions offers a powerful means to further enhance the scintillation properties of β-Ga₂O₃ and adapt it to specific application requirements.

In summary, we have successfully demonstrated the exceptional scintillation properties of floating zone (FZ) grown β-Ga₂O₃ single crystals, achieving a remarkable combination of high light yield (∼8000 photons/MeV), good spectroscopic response (7% energy resolution at 662 keV), and ultrafast decay times (<2 ns) by optimizing the growth conditions and carefully controlling the native defect structure. The FZ growth technique, coupled with the careful manipulation of the growth parameters, has proven to be a powerful approach for tailoring the scintillation characteristics of β-Ga₂O₃, outperforming even higher-Z scintillators such as PbWO₄, BaF₂, and Lu₂O₃:Yb. The successful growth of large, high-quality β-Ga₂O₃ single crystals using the FZ method demonstrates the scalability and practicality of this material for real-world applications, while the ability to produce scintillators with various thicknesses highlights its versatility. Future research directions include further optimization of the FZ growth process, detailed investigation of the scintillation mechanisms, exploration of doping strategies, and the development of large-scale, cost-effective production methods. The FZ-grown β-Ga₂O₃ scintillators developed in this study represent a significant breakthrough in the quest for an ideal scintillator material, laying the foundation for a promising future in the development of high-performance, next-generation scintillators based on β-Ga₂O₃ and other semiconductors (oxide or otherwise).

The authors thank the US Department of Energy (Office of Nuclear Physics) for their support of this study.

The authors have no conflicts to disclose.

A. Datta: Conceptualization (lead); Data curation (equal); Funding acquisition (lead); Investigation (lead); Methodology (equal); Project administration (equal); Resources (lead); Supervision (lead); Validation (lead); Writing – original draft (lead). H. Mei: Investigation (equal). A. Lebedinsky: Investigation (equal). P. Shiv. Halasyamani: Investigation (equal); Supervision (supporting). S. Motakef: Funding acquisition (equal); Supervision (equal); Writing – original draft (equal).

The data that support the findings of this study are available from the corresponding author upon reasonable request.