Open Access
Add to BibSonomy
Abstract
We present the basic scintillation properties of MgGa2O4 and ZnGa2O4 crystals grown at Leibniz-Institut für Kristallzüchtung by the Czochralski and the Vertical Gradient Freeze methods, respectively. We prove that these two Ga-based spinels do scintillate under gamma irradiation, currently displaying scintillation yields up to about 2500 ph/MeV. We also show the scintillation time profiles and their anticorrelation with the scintillation yields (the lower the yield, the faster the scintillation decay). Finally, we indicate that radioluminescence of both MgGa2O4 and ZnGa2O4 is much more efficient at low temperatures, which suggests the presence of a strong thermal quenching decreasing their yield towards room temperature.
Published by Optica Publishing Group under the terms of the Creative Commons Attribution 4.0 License. Further distribution of this work must maintain attribution to the author(s) and the published article's title, journal citation, and DOI.
1. Introduction
In the current scintillator market, being dominated by inorganic insulators, semiconductor scintillators are becoming more popular. Among many compounds, Ga-based oxides have generally attracted much attention due to a great demand for new types of ultrawide bandgap semiconductors, defining a fascinating new class of materials with a high potential for unique operability and providing applications in photovoltaics, optoelectronics and electronics systems [1–3]. Regarding the scintillation features of Ga-based oxides, first studies have been focused on β-Ga2O3 [4,5]. It has been reported that Czochralski-grown single crystals of β-Ga2O3 (undoped) with low free electron concentration (< 1017 cm-3) display scintillation yields up to 9000 photons per 1 MeV of absorbed gamma radiation energy [6]. On the other hand, Si-doped crystals with high free electron concentration (> 1018 cm-3) offer significantly faster scintillation decays, but with lower light outputs [7–9].
Once the scintillation mechanism in β-Ga2O3 has been tentatively recognized [10,11], other Ga-based oxides are also worth to be put under examination with respect to their scintillation properties. In this Communication we pay our attention to ternary systems, namely to MgGa2O4 and ZnGa2O4 crystals [12–14]. Due to their structure (inverse and normal spinel, respectively), they are referred to as Ga-based spinels. Their optical absorption edges are steep and begin at 249 nm (MgGa2O4) or 275 nm (ZnGa2O4). The cathodoluminescence spectra show a dominant band at 362 nm, which is followed by a full transparency in the visible and near-infrared spectral region. The direct optical bandgaps calculated from the absorption coefficient are about 5.0 eV for MgGa2O4 and 4.6 eV for ZnGa2O4 and [12–14].
The fundamental scintillation properties of MgGa2O4 and ZnGa2O4 crystals grown at the Leibniz-Institut für Kristallzüchtung are presented for the first time in this Communication. We demonstrate that these two Ga-based spinels scintillate when exposed to gamma radiation, exhibiting scintillation yields up to 2500 ph/MeV. Based on the pulse height, scintillation time profile and radioluminescence measurements we have found a correlation between some physical parameters, which let us focus on the physics which stands behind the data. In particular, the free electron concentration represents the most significant factor influencing the studied properties, including the scintillation yield and decay times.
2. Materials and experiment
2.1. Crystal growth and sample preparation
Bulk single crystals of Ga-based spinels were grown directly from the melt at the Leibniz-Institute für Kristallzüchtung (IKZ) in Berlin using the Czochralski (MgGa2O4) and Vertical Gradient Freeze (VGF; MgGa2O4 and ZnGa2O4) methods, as described in detail elsewhere [2,12–15]. In the Czochralski method a crystal is pulled up directly from the exposed melt surface out of a crucible, while in the VGF method a crystal is grown within a crucible, simply by melt solidification. The single crystals were grown from inductively heated Ir crucibles and oxygen-containing atmosphere to minimize thermal decomposition at growth temperatures above 1900°C [12,13]. Depending on the presence of oxygen in the growth atmosphere and the melt growth method used, the MgGa2O4 single crystals were found to be either electrical insulators (Czochralski method) or n-type semiconductors (VGF method), while ZnGa2O4 single crystals were always n-type semiconductors. The growth atmosphere for growing bulk MgGa2O4 crystals comprised 100 vol% CO2 or 80 vol% N2 + 20 vol% CO2 in the case of the Czochralski method, and N2 in the case of the VGF method. For growing ZnGa2O4 crystals by the VGF method, the growth atmosphere consisted of Ar + 2 or 20 vol% O2. Corresponding O2 concentrations for these growth atmospheres are given in Table 1. The free electron concentrations and the electron mobilities measured by the Hall effect were in the range of 3·1018–9·1019 cm-3 and 20–107 cm2 V-1 s-1 for ZnGa2O4 (VGF), and of 3·1017–2·1018 cm-3 and 4–7 cm2 V-1 s-1 for MgGa2O4 (VGF) [2]. ZnGa2O4 single crystals with such high free electron concentration are considered as degenerate semiconductors [13]. It is worth to note that the free electron concentration values of crystals grown by the VGF method vary within a crystal by one order of magnitude, i.e. crystal regions being in a direct contact with the growth atmosphere have lower values, while regions distant from the growth atmosphere have higher values of the free electron concentration (due to limited oxygen access). According to Galazka et al. [2,14], the origin of the electrical conductivity arises from antisite defects, which are formed at high temperatures during growth when spinels are random. During cooling down to room temperature, the compounds transform to inverse (MgGa2O4) or normal (ZnGa2O4) spinels with remaining some of the antisite defects giving rise to the electrical conductivity. Such transition occurs at temperatures of about 800-1000°C [12,13]. Indeed, annealing Ga-based spinels in the presence of oxygen at temperatures above 700°C for at least several hours switches them to electrically insulating state [15]. Moreover, much longer growth time by the Czochralski method as compared with the VGF method seems to be sufficient to minimize the density of the antisite defects and produce electrically insulating crystals.
Table 1. Growth details and scintillation properties of the studied MnGa2O4 and ZnGa2O4 samples (Cz – Czochralski, VGF – Vertical Gradient Freeze, O2 - oxygen concentration in the growth atmosphere, ne - free electron concentration, Y - scintillation yield, R - energy resolution at 662 keV, τi - scintillation decay time constants with their contributions, τmean - scintillation mean decay time)
In this work, (100) and (111)-oriented samples of 5 × 5 × 0.5 mm-3 were prepared by cutting and polishing semiconducting (ZnGa2O4, MgGa2O4) and electrically insulating (MgGa2O4) single crystals. All the samples were prepared from as-grown crystals, without any post-growth heat treatment. High structural quality of the crystal samples was confirmed by narrow rocking curves with the full width at half maximum of 23-47 arcsec [14].
2.2. Experimental techniques
To determine scintillation yields, pulse height spectra (PHS) were collected at room temperature under 662 keV gamma excitation from a 137Cs source. A Canberra 2005 integrating preamplifier, a Canberra 2022 spectroscopy amplifier with a shaping time set to 2 µs, and a Tukan-8K-USB [16] multichannel analyzer (MCA), processed the pulsed output signal from a Hamamatsu R878 photomultiplier tube (PMT) biased with 1250 V. The photoelectron yields, understood as the numbers of photoelectrons released from the photocathode per 1 MeV of energy deposited in each sample, were calculated using the positions of the so-called full-energy peaks in the spectra. These photoelectron yields were then translated into the scintillation yields given in numbers of photons per 1 MeV (ph/MeV), taking into the account the spectral matching of the scintillation light emitted by the crystals to the features of the PMT. The samples were adhered to the quartz glass of the PMT using Viscasil grease and wrapped with numerous layers of Teflon tape to increase the effectiveness of light gathering, creating a “reflecting umbrella”.
Scintillation time profiles (STP) were obtained using the delayed coincidence single photon counting method [17]. The samples were excited by gamma rays from a 137Cs source (the same like used for PHS), and time profiles of subsequent emission pulses were recorded by collecting many scintillation photons. Two Hamamatsu PMTs (R1104, R928) were used together with a Canberra TAC/SCA 2145 time-to-amplitude converter and a Tukan-8K-USB MCA.
For measurements of radioluminescence (i.e. X-ray excited emission spectra at various temperatures; RL) we used a single integrated custom setup. It comprises an Inel XRG3500 X-ray system with a Cu-anode tube and a 45 kV / 10 mA output, an Acton Research Corporation SpectraPro-500i monochromator, a Hamamatsu R928 photomultiplier tube (PMT), and an APD Cryogenic Inc. closed-cycle helium cooler. To prevent any possible contributions to the emission yield from the thermal release of charge carriers, we measured the RL at different temperatures ranging from 350 to 10 K, starting at the highest temperature and working our way down.
3. Results and discussion
3.1 Pulse height spectra
Figures 1–2 show the recorded pulse height spectra of the three MgGa2O4 and the three ZnGa2O4 samples, while Table 1 summarizes the values of scintillation yield and energy resolution. The sample’s IDs were given, where possible, in ascending order of the level of the free electron concentration, starting from electrical insulators or semi-insulators (Hall effect not measurable, resistivity > 108 Ω cm) and terminating with the highest values of the free electron concentration.
The first two MgGa2O4 samples (Mg1, Mg2), grown by the Czochralski method, are electrical insulators. However, their pulse height spectra differ from each other. In case of Mg1 there is no resolvable full energy peak, it is even very difficult to roughly estimate its scintillation yield using the Compton edge position. Thus, to avoid providing an incorrect value, we just denote its yield as very low. The lack of the full energy peak also precludes the determination of energy resolution. On the other hand, the insulating Mg2 sample produces a PHS with a full energy peak resolvable enough to calculate its scintillation yield (920 ph/MeV) and energy resolution (30.9%). It is possible that this sample has just a higher value of the residual free electron concentration. It should be noted, that even Hall effect was not measurable, the residual free electron concentration could be at a level of 1015 cm-3 or below in the case of oxide semiconductors. Such materials are typically called semi-insulators. The semiconducting Mg3 sample, grown by the VGF method, turns out to be the best one in this collation. It has a scintillation yield close to 2500 ph/MeV accompanied by an energy resolution of 29.9%.
For ZnGa2O4, we have three semiconducting samples with different values of free electron concentration. The first sample (Zn1) produces a PHS with a distinct full energy peak, resulting in a scintillation yield of 2120 ph/MeV and an energy resolution of 19.5%. The two remaining samples with much higher values of the free electron concentration (Zn2 and Zn3) show rather structureless PHS with no resolvable full energy peaks. Similarly to the Mg1 sample, their yield must be considered as very low.
We remind here that for β-Ga2O3 crystals the following regularities were observed: the insulating crystals are poor scintillators, whereas the semiconducting ones scintillate well and their scintillation yield decreases with increasing free electron concentration. In other words, the best scintillating samples are those with relatively low level of electrical conductivity, about mid 1016 cm-3 [7–9]. It seems that the situation is very similar for MgGa2O4 and ZnGa2O4. Neither the insulating crystals of MgGa2O4 nor the strongly conducting crystals of ZnGa2O4 are good candidates for scintillators. Only the crystals with free electron concentrations of an order of 1018 cm–3 (both MgGa2O4 and ZnGa2O4) exhibit promising scintillation properties, at least with respect to the scintillation yield. Possibly a further decrease of the free electron concentration (still maintaining the semiconducting state) would result with even higher yields, similar to β-Ga2O3. It should be noted, however, that the origin of the electrical conductivity of Ga-based spinels (antisite defects) is different from that of β-Ga2O3 (dopants or impurities).
3.2 Scintillation time profiles
The scintillation time profiles (also known as scintillation decays) of MgGa2O4 and ZnGa2O4 are shown in Figs. 3–4. The experimental curves are not distorted by afterpulses and can be used to calculate decay time constants with a good accuracy. This is done for all the studied crystals by fitting three-exponential decays, since three is the least number of components that provides a satisfactory agreement between the experimental and the fitted time profiles. Table 1 contains the values of the scintillation decay constants and their contributions for all of the samples studied. To make comparisons more manageable, we extend the number of parameters with a so-called scintillation mean decay time τmean. We calculate this value in accordance with Eq. (2) by Zatryb and Klak [18], arriving at:
From the three MgGa2O4 samples, the two insulating samples (Mg1 and Mg2) generate substantially faster scintillation than the semiconducting one (Mg3). This becomes evident by means of the 2.5 times shorter decay times (427-466 ns; Mg1-2) for the former as compared to the latter (1130 ns; Mg3), and also by the contribution of decay time components. In the “fast” Mg1 and Mg2 profiles the contribution of the two prompt components, τ1 (14-17 ns) and τ2 (86-137 ns), is above 40%, whereas in the “slow” profile it is below 30% with longer decay times (36 and 208 ns for τ1 and τ2, respectively).
In case of ZnGa2O4 all the samples are semiconducting. One can easily notice that with increasing free electron concentration the contribution of particular components shifts in favor of the shorter ones, as well as the decay times become significantly shorter. The scintillation mean decay time is 1180 ns for the sample with the lowest level of the free electron concentration. An increase of the free electron concentration by one order of magnitude shortens the mean decay time to 175-267 ns.
Analyzing the results of PHS and STP experiments comparatively we infer that the scintillation performance of both MgGa2O4 and ZnGa2O4 is greatly influenced by the free electron concentration. We point out that, in contrast to the electrically insulating crystals and those with the highest free electron concentration (> 1019 cm-3), the semiconducting crystals with high free electron concentration of order of 1018 cm-3 scintillate quite efficiently. However, with respect to the scintillation time profiles the situation looks the other way around: the samples offering the highest scintillation yield are slow scintillators, while a much faster scintillation can be achieved at the expense of decreasing the yield. Therefore, similarly to the case of β-Ga2O3 [7–9] finding such a free electron concentration that would provide both fast and efficient scintillation turns out to be a difficult task.
3.3 Radioluminescence
RL properties of MgGa2O4 and ZnGa2O4 samples were examined at various temperatures ranging from 10 to 350 K, and representative RL spectra measured at 10, 100, 200 and 300 K are presented in Figs. 5–6. To make comparisons easier, they are normalized to the same maximum intensity at 10 or 100 K. The shapes of the spectra, dominated by a band peaking at about 350 nm, are in a good agreement with cathodoluminescence spectra reported in literature for MgGa2O4 [12] and ZnGa2O4 [15]. They also resemble the radioluminescence spectra of β-Ga2O3 crystals, for which the dominant band is attributed to self-trapped excitons [19].
For MgGa2O4 all RL spectra at lower temperatures (10, 100, 200 K) look similar. The spectra at RT also do not differ much from each other, however a small satellite band at about 500 nm of unknown nature occurs for the insulating samples. The most important feature is related to the relative intensities of the bands corresponding to particular temperatures. X-ray emission is evidently quenched at RT for all samples, nevertheless the magnitude of quenching is clearly lower for the semiconducting sample (Mg3), which in fact displays the highest scintillation yield. It seems that the quenching mechanism, which needs a further investigation, is stronger for insulating crystals (Mg1-2), preventing them from being fine scintillators.
For ZnGa2O4 the case of the semiconducting sample with the lowest free electron concentration (Zn1) is quite similar to the semiconducting MgGa2O4 sample (Mg3). Although the magnitude of quenching is even smaller in Zn1 compared to Mg3, its scintillation yield is lower. One of the reasons could be the presence of nanoparticles of metallic nature in ZnGa2O4 crystals as the result of strong thermal decomposition during the growth [2,13]. Interestingly, the two remaining ZnGa2O4 samples (Zn2-3) with much higher free electron concentration, which are similar to each other with respect to pulse height spectra and scintillation time profiles, differ in radioluminescence. Zn1 and Zn2 samples show distinct spectra at all studied temperatures indicating low thermal quenching. It is really surprising that the scintillation yield is relatively low in these samples, perhaps due to above-mentioned nanoparticles. On the other hand, the spectra of the Zn3 sample are hardly visible above noise.
4. Conclusions
With the performed studies on MgGa2O4 and ZnGa2O4 single crystals we have discovered new candidates for semiconducting scintillators. Although at the moment the scintillation yield is relatively low (∼2500 and ∼2100 ph/MeV for MgGa2O4 and ZnGa2O4, respectively), we note that these materials have not been optimized for scintillation performance yet. We have also indicated the dominant role of the free electron concentration for the scintillation properties of these crystals. As the number of the investigated samples is limited, the current results should rather be taken as qualitative; much more samples with different free electron concentrations are necessary for a quantitative assessment. If, indeed, MgGa2O4 and ZnGa2O4 behave in a very similar manner as β-Ga2O3 in terms of the scintillation mechanism, we expect a much higher scintillation yield for both compounds once the free electron concentration is lowered to a level below 1017 cm-3. This is, however, a challenging task due to indirect control of the free electron concentration (caused by antisite defects) that proceeds through growth conditions, operating parameters, and post-growth heat treatment.
Funding
Narodowe Centrum Nauki (2016/23/G/ST5/04048); Deutsche Forschungsgemeinschaft (GA 2057/2-1).
Acknowledgments
This research has been partly performed in the framework of GraFOx, a Leibniz-Science Campus partially funded by the Leibniz Association-Germany. The authors would like to thank Mike Pietsch (Leibniz-Institut für Kristallzüchtung) for measurements of electrical properties and Dr. Tobias Schulz (Leibniz-Institut für Kristallzüchtung) for critical reading of the present manuscript.
Disclosures
The authors declare no conflicts of interest.
Data availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
References
1. Z. Galazka, “β-Ga2O3 for wide-bandgap electronics and optoelectronics,” Semicond. Sci. Technol. 33(11), 113001 (2018). [CrossRef]
2. Z. Galazka, K. Irmscher, M. Pietsch, S. Ganschow, D. Schulz, D. Klimm, I. M. Hanke, T. Schroeder, and M. Bickermann, “Experimental Hall electron mobility of bulk single crystals of transparent semiconducting oxides,” J. Mater. Res. 36(23), 4746–4755 (2021). [CrossRef]
3. Z. Galazka, S. Ganschow, P. Seyidov, K. Irmscher, M. Pietsch, T.-S. Chou, S. Bin Anooz, R. Grueneberg, A. Popp, A. Dittmar, A. Kwasniewski, M. Suendermann, D. Klimm, T. Straubinger, T. Schroeder, and M. Bickermann, “Two inch diameter, highly conducting bulk β-Ga2O3 single crystals grown by the Czochralski method,” Appl. Phys. Lett. 120(15), 152101 (2022). [CrossRef]
4. T. Yanagida, G. Okada, T. Kato, D. Nakauchi, and S. Yanagida, “Fast and high light yield scintillation in the Ga2O3 semiconductor material,” Appl. Phys. Express 9(4), 042601 (2016). [CrossRef]
5. M. Makowski, W. Drozdowski, M. E. Witkowski, A. J. Wojtowicz, K. Irmscher, R. Schewski, and Z. Galazka, “Tailoring the scintillation properties of β-Ga2O3 by doping with Ce and codoping with Si,” Opt. Mater. Express 9(9), 3738–3743 (2019). [CrossRef]
6. Z. Galazka, “Growth of bulk β-Ga2O3 single crystals by the Czochralski method,” J. Appl. Phys. 131(3), 031103 (2022). [CrossRef]
7. W. Drozdowski, M. Makowski, M. E. Witkowski, A. J. Wojtowicz, R. Schewski, K. Irmscher, and Z. Galazka, “Semiconductor scintillator development: pure and doped β-Ga2O3,” Opt. Mater. 105, 109856 (2020). [CrossRef]
8. W. Drozdowski, M. Makowski, M. E. Witkowski, A. J. Wojtowicz, K. Irmscher, R. Schewski, and Z. Galazka, “Recent progress in the development of β-Ga2O3 scintillator crystals grown by the Czochralski method,” Opt. Mater. Express 11(8), 2488–2494 (2021). [CrossRef]
9. W. Drozdowski, M. Makowski, A. Bachiri, M. E. Witkowski, A. J. Wojtowicz, L. Swiderski, K. Irmscher, R. Schewski, and Z. Galazka, “Heading for brighter and faster β-Ga2O3 scintillator crystals,” Opt. Mater. X 15, 100157 (2022). [CrossRef]
10. M. E. Witkowski, K. J. Drozdowski, M. Makowski, W. Drozdowski, A. J. Wojtowicz, K. Irmscher, R. Schewski, and Z. Galazka, “Low temperature thermoluminescence of β-Ga2O3 scintillator,” Opt. Mater. X 16, 100210 (2022). [CrossRef]
11. A. J. Wojtowicz, M. E. Witkowski, M. Makowski, W. Drozdowski, K. Irmscher, R. Schewski, and Z. Galazka, “Scintillation and radioluminescence mechanism in β-Ga2O3 semiconducting single crystals,” to be submitted.
12. Z. Galazka, D. Klimm, K. Irmscher, R. Uecker, M. Pietsch, R. Bertram, M. Naumann, M. Albrecht, A. Kwasniewski, R. Schewski, and M. Bickermann, “MgGa2O4 as a new wide bandgap transparent semiconducting oxide: growth and properties of bulk single crystals,” Phys. Status Solidi A 212(7), 1455–1460 (2015). [CrossRef]
13. Z. Galazka, S. Ganschow, R. Schewski, K. Irmscher, D. Klimm, A. Kwasniewski, M. Pietsch, A. Fiedler, I. Schulze-Jonack, M. Albrecht, T. Schröder, and M. Bickermann, “Ultra-wide bandgap, conductive, high mobility, and high quality melt-grown bulk ZnGa2O4 single crystals,” APL Mater. 7(2), 022512 (2019). [CrossRef]
14. Z. Galazka, S. Ganschow, K. Irmscher, D. Klimm, M. Albrecht, R. Schewski, M. Pietsch, T. Schulz, A. Dittmar, A. Kwasniewski, R. Grueneberg, S. B. Anooz, A. Popp, U. Juda, I. M. Hanke, T. Schroeder, and M. Bickermann, “Bulk single crystals of β-Ga2O3 and Ga-based spinels as ultra-wide bandgap transparent semiconducting oxides,” Prog. Cryst. Growth Charact. Mater. 67(1), 100511 (2021). [CrossRef]
15. Z. Galazka, Transparent Semiconducting Oxides - Bulk Crystal Growth and Fundamental Properties (Jenny Stanford Publishing, 2020).
16. Z. Guzik, S. Borsuk, K. Traczyk, and M. Plominski, “TUKAN - an 8K pulse height analyzer and multi-channel scaler with a PCI or a USB interface,” IEEE Trans. Nucl. Sci. 53(1), 231–235 (2006). [CrossRef]
17. L. M. Bollinger and G. E. Thomas, “Measurement of the time dependence of scintillation intensity by a delayed–coincidence method,” Rev. Sci. Instrum. 32(9), 1044–1050 (1961). [CrossRef]
18. G. Zatryb and M. M. Klak, “On the choice of proper average lifetime formula for an ensemble of emitters showing non-single exponential photoluminescence decay,” J. Phys.: Condens. Matter 32(41), 415902 (2020). [CrossRef]
19. Z. Galazka, R. Schewski, K. Irmscher, W. Drozdowski, M. E. Witkowski, M. Makowski, A. J. Wojtowicz, I. M. Hanke, M. Pietsch, T. Schulz, D. Klimm, S. Ganschow, A. Dittmar, A. Fiedler, T. Schroeder, and M. Bickermann, “Bulk β-Ga2O3 single crystals doped with Ce, Ce+Si, Ce+Al, and Ce+Al+Si for detection of nuclear radiation,” J. Alloys Compd. 818, 152842 (2020). [CrossRef]
