Open Access
Add to BibSonomy
Abstract
The unique properties of gallium oxide (GaOx) have drawn increasing interest as a material suitable for high-power electronic and optical applications. Herein, we report the demonstration of low-loss GaOx-core/SiO2-cladding waveguides on Si substrate. We present the fabrication process and annealing treatments of the waveguide devices, and we characterize the corresponding effects on optical transmission for 3 common wavelengths: 633 nm, 1064 nm, and 1550 nm. The best propagation loss achieved for these wavelengths is measured to be $-0.4\pm 0.1\: \textrm {dB/cm}$, $-0.3\pm 0.2\: \textrm {dB/cm}$, and $-2.4\pm 0.5\: \textrm {dB/cm}$, respectively. We discuss the major waveguide loss mechanisms, followed by results of pump and probe experiments using visible/IR wavelengths for waveguides treated under various post-fabrication annealing conditions. We also show nonlinear measurements for a 250 fs laser beam to offer additional insights into the loss mechanisms, which are consistent with the linear optical transmission performances. High waveguide laser-induced damage threshold (LIDT) of $ >2.5\: \textrm {J/cm}^2$ is measured at this pulse width, making GaOx a potential candidate for high-power integrated photonic devices.
© 2020 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Gallium oxide ($ \textrm {GaO}_{\textrm{x}}$) is a wide bandgap transparent conductive oxide (TCO) that exhibits many distinct properties [1–3]. Recently, due to advances in material growth and device fabrication techniques [4–9], electronics and optical applications of $ \textrm {GaO}_{\textrm{x}}$ are emerging [10–12]. The high breakdown field strength enables high-power electronics applications [13–15]. With one of the largest bandgaps among TCO, $ \textrm {GaO}_{\textrm{x}}$ is an interesting candidate material for UV detectors [16–19]. LIDT of $ \textrm {GaO}_{\textrm{x}}$ is reported to be one of the highest amongst the conductive materials, and may be comparable to other high damage threshold materials such as $ \textrm {SiO}_2$ and $ \textrm {Al}_2 \textrm {O}_3$ [20–22]. Due to the unique properties of $ \textrm {GaO}_{\textrm{x}}$ as a TCO with high LIDT, we explore its behavior as a waveguiding material for high-power integrated photonics.
In this paper we report the demonstration of low-loss $ \textrm {GaO}_{\textrm{x}}$-core/ $ \textrm {SiO}_2$-cladding waveguides on Si substrates. We present the waveguide design and device fabrication, followed by linear and nonlinear optical characterizations, including linear/nonlinear transmission, pump/probe, LIDT, and spectral measurements. We discuss different fabrication processes along with the post-fabrication annealing treatments, and their impact on the waveguide properties for three common wavelengths of 633 nm, 1064 nm, and 1550 nm.
2. Waveguide design and fabrication
We design the waveguides primarily to explore their use for high-power integrated photonic applications, especially dielectric laser accelerators (DLAs) [23–33], where short sections of waveguides (mm to cm scale) are used for high peak power delivery [34–37]. The high LIDT and relatively low $\chi _3$ nonlinearity ($n_\textrm {Kerr}=-2\sim -3\times 10^{-19}\:\textrm {m}^2/\textrm{W}$) [38] are appealing for high-power photonics, which differ from the requirements of waveguides for telecom applications [39,40]. The waveguide core is designed to be rectangular $ \textrm {GaO}_{\textrm{x}}$ with 70 nm thickness and 5.3 µm width, and is sandwiched between 5 µm of $ \textrm {SiO}_2$ on both sides as claddings. This geometry is chosen to operate in single (TE) mode for wavelengths of 1 µm and beyond (the common wavelength range for DLA applications).
The fabrication process flows are described in Fig. 1. The first step is the growth of a 5 µm thick thermal SiO$_2$ at 1150°C [Fig. 1(a)]. Then a thin film $ \textrm {GaO}_{\textrm{x}}$ layer is deposited on the top of thermal oxide by e-beam evaporation [Fig. 1(b)]. Evaporation is performed at a vacuum level of $10^{-6}$ Torr with a $\textrm {Ga}_2\textrm{O}_3$ target with 99.99% purity. The measured chemical analysis of the target is provided by the vendor [41] as listed in Table 1. Next, a 1 µm thick SPR3612 photoresist is patterned by ASML PAS5500 stepper. Straight waveguides with total length of 25.6 mm are patterned. The patterned photoresist serves as the mask to transfer the pattern to the $ \textrm {GaO}_{\textrm{x}}$ via inductively coupled plasma (ICP) etch using Oxford Instruments PlasmaPro System 100 Cobra III-V Etcher (Ox-35).
Fig. 1. Nano-fabrication process flow for the $ \textrm {GaO}_{\textrm{x}}$-core/SiO$_2$-clad waveguide. (a) Thermalization of the Si wafer surface to grow the bottom cladding layer of SiO$_2$. (b) E-beam evaporation of $ \textrm {GaO}_{\textrm{x}}$ thin film. (c) Patterning using optical lithography. (d) Transferring the photoresist pattern to the $ \textrm {GaO}_{\textrm{x}}$ thin film by plasma etch followed by cleaning. (e) Schematic illustration of cross section of the $ \textrm {GaO}_{\textrm{x}}$-core/SiO$_2$-clad planar waveguides on Si substrates after the final fabrication step, which is LPCVD SiO$_2$ to serve as the top and overall cladding layers. (f) Constituent materials of different layers.
Table 1. $ \textrm {GaO}_{\textrm{x}}$ Target chemical analysis. All values in wt. ppm.
Herein, we have developed a $ \textrm {CH}_4/\textrm{H}_2$-based plasma etching process. Hydrogen could potentially passivate defects [12], and cause less surface damage during the etching process compared to the heavy radical of ${\textrm {BCl}_3}$. It is also easier to get access to ${\textrm {H}_2}/\textrm{CH}_4$ gas in commercial foundries compared to ${\textrm {BCl}_3}$, showing a path for easier process transfer for applications that gear towards mass production. In the recipe, 50 sccm of ${\textrm {CH}_4}$ plus 20 sccm of ${\textrm {H}_2}$ at 10 mTorr pressure at 10 $^{\circ }$C with a combination of 1000 W of ICP power and 200 W of bias power have been used. The etch rate is 30 nm/min, with an improved selectivity of $ \textrm {GaO}_{\textrm{x}}$:photoresist to be 1:1.2. In later sections, we also compare the corresponding differences in linear/nonlinear transmission for waveguides that are fabricated by this newly developed ${\textrm {H}_2}/{\rm{CH}}_4$-based etching process versus the well established $\textrm {BCl}_3$-based ones [42–45].
After stripping the remaining photoresist, the final step is to deposit 5 µm of SiO$_2$ by LPCVD to cover the top and surrounding regions of the the ridge waveguide structure as shown in Fig. 1(d). This completes the fabrication of $ \textrm {GaO}_{\textrm{x}}$-core/SiO$_2$-clad waveguides on Si substrate as displayed in Fig. 1(e). A microscope image of the completed waveguide chip is shown in Fig. 2.
3. Waveguide characterization
The fabrication of these waveguide structures can be challenging due to the high requirement on surface quality resulting from the dry etching. The lack of long range order for amorphous material makes it more challenging to understand from a first-principle-based theoretical approach, compared to the single crystalline case. Therefor, we have studied the behaviors of the waveguides through a series of experiments, including linear and nonlinear transmission tests as well as spectral measurements for the waveguides fabricated under different etching/annealing conditions.
3.1 Linear measurements
For the linear transmission measurement, the experimental setup is shown in Fig. 3. We measure the linear transmission of the waveguide under three common wavelengths: For 1064 nm, the continuous wave (CW) input light sources is a Nd:YAG NPRO laser, a fiber coupled HeNe laser is used for 632 nm and an erbium-doped fiber amplified laser is used for 1550 nm. Two lenses with NA = 0.55 are used for coupling the light into and out of the waveguide. Thermal power meters are used for the power measurement.
Fig. 3. Experimental setup. IL: input lens. OL: output lens. ND wheel: continuous neutral density filter wheel. PD: photo-detector. PM: power meter. FM: flip mirror.
The linear transmission measurements have been performed for waveguides fabricated by $\textrm {BCl}_3$- and $\textrm {CH}_4/\textrm{H}_2$-based etching recipes under various annealing conditions, to investigate the optimum waveguide propagation loss at different wavelengths. Two annealing gas environments are explored independently: $\textrm {H}_2/\textrm{N}_2$ forming gas and oxygen. Each chip is annealed for 3 hours for a given condition. The annealing temperatures are chosen based on the study of thin film $\textrm {Ga}_2\textrm{O}_3$ [46], which shows the amorphous $\textrm {Ga}_2\textrm{O}_3$ deposited by chemical vapor deposition (CVD) was observed to start forming a crystalline structure in a temperature range of 700-1000 °C. Later studies suggest a crystalline temperature to form $\beta$-$\textrm {Ga}_2\textrm{O}_3$ between 750-900 °C [47]. As a result, we performed annealing in $\textrm {H}_2/\textrm{N}_2$ forming gas environment for three temperatures, 700°C, 900°C and 1050°C, and find that the ideal annealing temperature is around 800-900 °C for optimum transmissions. For higher annealing temperature at 1050°C, we observe weak guiding behavior and unguided transmission due to the diffusion of $ \textrm {SiO}_2$ into $ \textrm {GaO}_{\textrm{x}}$, similar to that reported in [46]. This leads to near zero transmission for 633 nm and 1550 nm, and $>$10 dB/cm propagation loss for 1064 nm. A summary of the results from different fabrication/annealing recipes is shown in Table 2. The absolute propagation loss is measured for a given wavelength and recipe, through image analysis of the out of plane waveguide scattered light. The lowest propagation loss obtained is $-0.4\pm 0.1$ dB/cm for 633 nm, $-0.3\pm 0.2$ dB/cm for 1064 nm, and $-2.4\pm 0.5$ dB/cm for 1550 nm. Detailed data and fitting curves can be found in the Appendix.
Table 2. Waveguide loss in dB/cm at 633 nm, 1064 nm and 1550 nm under various fabrication recipes. The unannealed $\textrm {BCl}_3$ sample has close to 0 transmission for 633 nm and 1550 nm, where the loss values can not be reliably measured.
The study of monoclinic $ \textrm {GaO}_{\textrm{x}}$ loss mechanisms is still an ongoing research effort [12,48–54], and the nature of the amorphous/polycrystalline structures of waveguides will further complicate it. To understand the results reported in Table 2, we discuss a few possible major loss mechanisms for $ \textrm {GaO}_{\textrm{x}}$ waveguide, which show qualitative consistent behaviors to the measured results. Then we provide more detailed studies for one of these mechanisms, O vacancy, through a series of linear/nonlinear measurements in the later sections.
For mechanism 1(a), since the $ \textrm {GaO}_{\textrm{x}}$ is deposited in vacuum in the absense of $\textrm {O}_2$, it is expected to have O vacancies, responsible for its n-type behavior, which can cause red/NIR absorption [53–55]. Performing $\textrm {O}_2$ annealing at 800°C leads to the highest transmission for 633 nm and 1064 nm, while the 1550 nm loss is higher due to loss mechanism 2 and 4. In general, annealing can reduce loss caused by material surface roughness (mechanism 3), dislocations and/or defects. Comparing the two annealing conditions for the waveguides fabricated by $\textrm {BCl}_3$-based etching recipes, the higher loss resulting from FGA also suggests that O vacancy is one of the dominant loss mechanisms for red/NIR wavelengths. For mechanism 1(b), it has been shown from first-principle calculations that hydrogen can contribute to n-type conductivity [50]. It is also shown that in the presence of hydrogen, vacancy-hydrogen complexes may form in $\beta$-$\textrm {Ga}_2\textrm{O}_3$ [12,52,56].
For mechanism 2, for short wavelengths (red/NIR), absorption has been partially attributed to FCA in the case of conductive n-type $ \textrm {GaO}_{\textrm{x}}$ crystal [53]. For short wavelength IR (SWIR), indirect FCA process mediated by either phonons or charged impurities start to become important. This indirect FCA has been reported for $ \textrm {GaO}_{\textrm{x}}$ [51] and other transparent conductive oxides such as $\textrm {SnO}_2$ [57], where the corresponding absorption cross section is observed to be proportional to the cube of the wavelength. At 1550 nm, the estimated phonon induced indirect FCA can contribute to $\sim 1.2-1.7$ dB/cm of loss for bulk $ \textrm {GaO}_{\textrm{x}}$ according to Ref. [51], this in combination with the charge impurity induced indirect FCA will set the fundamental limit for the propagation loss.
For mechanism 3, $\textrm {CH}_4/\textrm{H}_2$-based plasma etching results in smoother waveguide surface (compared to $\textrm {BCl}_3$-based etching recipe), which shows higher transmissions for all 3 wavelengths in the un-annealed cases. However, after annealing under the same FGA conditions, due to the complication of mechanism 1(b), the introduction of hydrogen may lead to more absorption for short wavelength (red), but longer wavelengths may be less affected.
For mechanism 4, though it is not a technical limit to deposit a thicker layer, the cladding thickness deposited is 5 µm due to the restriction on Thermco furnace reservation time at Stanford Nano Facilities (SNF). The corresponding loss due to the portion of mode field in Si is calculated to be 0.27 dB/cm, given the nominal refractive index of $\sim$1.9 for $ \textrm {GaO}_{\textrm{x}}$ at 1550 nm when deposited in oxygen-poor environment [55]. However, it is shown that thin film $\textrm {Ga}_2\textrm{O}_3$ deposited in oxygen-rich environment exhibits a lower refractive index of 1.7$\sim$1.9 [55]. In the case of $n_\textrm {G{aO}{}}$ = 1.7, the corresponding loss can be as large as 2.6 dB/cm. Detailed simulation is given in the Appendix.
To further investigate the O vacancy in mechanism 1(a), we conduct a CW pump-probe experiment using a visible pump at 633 nm (and 532 nm) and IR probe at 1064 nm (and 1550 nm). The BCl$_3$ etched 900°C FGA chip is used. The experimental setup is shown in Fig. 4(a). IR beam path of 1064 nm is combined with the visible input beam path of 633 nm, through a 50:50 beam splitter. The input coupler is set up for optimum coupling of the visible path. An additional mode matched telescope is used for optimum coupling efficiency of the IR beam. The output path is selected by a flip mirror; one path goes through a long pass filter (LPF) for IR power measurement, the other goes through short pass filter (SPF) for visible. Corresponding results are shown in Fig. 4(b), where visible wavelength induced IR absorption is observed.
Fig. 4. (a). CW pump-probe experiment setup. IR beam path is shown in orange, visible beam path is shown in green and the combined beam path is shown in purple. LPF: long pass filter. SPF: short pass filter. (b). Visible wavelength (633 nm) induced IR (1064 nm) transmission loss of the BCl$_3$ etched 900°C FGA chip. (c). 532 nm pump induced 1550 nm probe transmission loss.
As can be seen, $\sim 4\%$ 1064 nm transmission power loss (relative to with no pump light) is observed with 7 µW of pump power at 633 nm, and over 10$\%$ of 1550 nm transmission loss is induced with 1.7 mW of 532 nm pump. The visible induced IR absorption recovers to its original value when the visible pump is turned off. Green (or visible) induced IR absorption (GRIIRA) is a known effect in crystals such as $\textrm {LiNbO}_3$ and $\textrm {KTiOPO}_4$ [58,59]. In the case of $\textrm {LiNbO}_3$, the loss is mostly attributed to the small polaron located on Nb antisite defect [58]. Typically this substoichiometric defect induced GRIIRA can be largely reduced by doping, with a dopant such as MgO in the case of congruent grown $\textrm {LiNbO}_3$. The O vacancies caused by film deposition in $\textrm {O}_2$-poor environment can cause this substoichiometric defect, which leads to the visible induced IR absorption similar to the case of $\textrm {LiNbO}_3$.
To verify this assumption, we repeat the pump/probe experiment for the BCl$_3$ etched $\textrm {O}_2$ annealed chip. With up to 126 µW of pump light through the waveguide, we did not observe difference in 1064 nm light transmission. This is consistent with the improved 1064 nm transmission behavior comparing to the FGA 900°C case, indicating that the substoichiometric defect can indeed be improved by this oxygen annealing procedure.
3.2 Nonlinear measurements
We performed the nonlinear transmission and spectral measurements for these waveguides at the wavelength $\sim 1$ µm, which provides additional information on the material behavior under different fabrication recipes. We use an optical parametric amplifier (OPA) (ORPHEUS-HP) pumped by a mode locked diode-pumped Yb:KGW-based chirped pulse amplifier (PHAROS) as the input light source. The output center wavelength is tuned to 1030 nm, with 250 fs pulse width, 100 kHz repetition rate, and a maximum pulse energy up to 10 µJ. The typical transmission curves are shown in Fig. 5.
Fig. 5. Waveguide nonlinear transmission with 250 fs input pulse, 100 kHz repetition rate at 1.03 µm.
We note that the top four transmission curves in Fig. 5 are reversible within the range shown, where the same output pulse energy is preserved after turning the input power back down from the maximum. This indicates that there is no permanent damage and the saturation of the transmission is dominated by nonlinear absorption. For the unannealed chip etched by $\textrm {CH}_4/\textrm{H}_2$, the waveguides damage at $\sim 30$ nJ of input power, while the $\textrm {O}_2$ annealed chip transmission curve is still reversible (not damaged) beyond an input pulse energy of $\sim 250$ nJ. The maximum transmitted pulse energy of the $\textrm {O}_2$ annealed case is $\sim 6.8$ nJ, while the unannealed chip output pulse energy saturates at $\sim$2.1 nJ.
This reversible saturation behavior shows a clear trend of the quality of the waveguides, with the O$_2$ annealed chip containing the least amount of defects and the unannealed chip having the most for wavelengths around 1.03 µm. This result shows a consistent trend to the linear transmission behavior, where the waveguide with the least defects shows the best linear transmission at this wavelength.
Based on the nonlinear transmission result, we can expect a difference in the optical spectrum for waveguides with different nonlinear absorption, where waveguides with more defects are expected to show more/larger spectral dips compared to waveguides with less defects, under the same pulse energy. As a result, we measured the waveguide output spectrum centered around 1.03 µm for pulse energy of 1.5 nJ and 6 nJ. We choose the $\textrm {O}_2$ annealed chip etched by BCl$_3$ versus the unannealed chip etched by $\textrm {CH}_4/\textrm{H}_2$ for their large difference in linear/nonlinear transmission. As shown in Fig. 6, the unannealed chip shows two distinct dips at 1020 nm and 1034 nm, whereas the $\textrm {O}_2$ annealed chip shows much reduced spectral ripples under the same power level. We suspect that the spectral dips are associated with defects or nonlinear processes, as we observe visible white light generation along the waveguides. For the unannealed case, white light becomes visible at $\sim$ 0.4 nJ of output pulse energy. In comparison, for the $\textrm {O}_2$ annealed case, white light turns on $\sim$ 2.4 nJ of output pulse energy. As we turn up the input power for the $\textrm {O}_2$ annealed chip to 6 nJ, the spectral dip around 1020 nm becomes more prominent yet still smaller than the low power un-annealed case, while the 1034 nm dip remains absent. White light generation can be a signature of defects, which is common for III-N materials such as GaN [60] and AlN [61]. Although more detailed study on this topic is beyond the scope of this paper, the threshold difference in white light generation can be useful to indicate the relative amount of defects. Our non-linear spectrum measurements have shown consistent behaviors to the linear/nonlinear measurement results shown earlier, where the chip with more spectral dips (more defects) associated with nonlinear processes also shows lower linear and nonlinear transmission around the specific wavelength.
Fig. 6. Spectrum comparison of BCl$_3$ etched $\textrm {O}_2$ annealed chip and H$_2$ etched un-annealed chip, at a center wavelength of 1030 nm. The spectrum is arbitrarily offset from each other for clear display.
4. Conclusion
We report the fabrication and characterization of $ \textrm {GaO}_{\textrm{x}}$-core/ $ \textrm {SiO}_2$-cladding waveguides on Si substrate. We present the fabrication and annealing processes of the waveguides, and demonstrate the first CH$_4$/H$_2$-based plasma etching recipe for $ \textrm {GaO}_{\textrm{x}}$. We have studied the linear and nonlinear transmission behavior of the waveguides, and compare the effects of various annealing conditions for three common wavelengths. We show oxygen vacancy as one of the major waveguide loss mechanisms for visible/ NIR (633 nm and 1064 nm), as the 800°C $\textrm {O}_2$ annealing condition provides the lowest propagation loss at 633 nm ($-0.4\pm 0.1$ dB/cm) and 1064 nm ($-0.3\pm 0.2$ dB/cm). The 1550 nm loss ($-2.4\pm 0.5$ dB/cm) is higher possibly due to mode leakage. With thicker oxide cladding, we expect the loss at 1550 nm can be further reduced.
The propagation losses for these three common wavelengths are on par with the typical performance of Si-based waveguides on SOI wafers (0.5-2 dB/cm) [62], while the measured LIDT ($>2.5\:\textrm {J/cm}^2$ at 250 fs) is one order of magnitude higher than Si waveguides, and a factor of $\sim 4$ higher than $\textrm {SiN}_{\textrm{x}}$ waveguides [36]. We have further investigated the loss mechanisms for the wavelengths near 1 µm by pump and probe experiments, as well as spectral measurements, which show consistent results with the linear transmissions. The presented device fabrications, post-fabrication treatments, and optical characterizations of $ \textrm {GaO}_{\textrm{x}}$ as a waveguiding material have shown interesting properties that can lead to more subsequent studies. This demonstration of $ \textrm {GaO}_{\textrm{x}}$ waveguides with high LIDT may open up pathways for $ \textrm {GaO}_{\textrm{x}}$-based high-power integrated photonics and other nonlinear optical devices.
Appendix
Measurement of propagation loss
The propagation loss is characterized by image processing of out-of-plane scattered light from the waveguides. A Dino-lite CMOS digital microscope is used for the image capturing of 633 nm and 1064 nm, while an NIT (WIDY SWIR 640 V-ST) digital InGaAs camera is used for 1550 nm. The gray scale images and fittings are shown in Fig. 7. $\gamma$ correction is set to 1 for all images. The $95\%$ confidence interval is reported as the uncertainty of the fitting. The $>3\sigma$ intensity data points are removed for the fittings.
Fig. 7. Gray scale image of waveguide out-of-plane scattering light (top) and corresponding fit (bottom) for the best transmission recipes. (a). 633 nm. (b). 1064 nm. (c). 1550 nm.
Simulation of mode propagation loss
Due to the facility regulation constraints, the maximum cladding thickness deposited is 5 µm. As a result, for 1550 nm, there is an addition propagation loss due to mode leakage through finite thickness cladding into the Si substrate. In this section, We calculate this propagation loss through numerical simulation.
Specifically, we calculate the mode propagation loss of a slab waveguide with 70 nm thick $ \textrm {GaO}_{\textrm{x}}$ core and 5 µm silica cladding on both sides, atop a silicon substrate. Since the transverse width of the waveguide (5.3 µm) is much larger than the wavelength, the effective index and the propagation loss of such a slab waveguide should resemble that of the $ \textrm {GaO}_{\textrm{x}}$ waveguide studied in this paper. We set the refractive index of $ \textrm {GaO}_{\textrm{x}}$ ranging from 1.6 to 1.9 for the simulation, according to the commonly reported values [55,63]. Perfectly matched layer (PML) boundary conditions are applied on both the top and bottom boundaries to solve the fundamental TE mode of the slab waveguide. The propagation loss is inferred from the complex wave vector obtained by the finite difference eigenmode solver.
The mode propagation loss as a function of the refractive index of $ \textrm {GaO}_{\textrm{x}}$ is shown in Fig. 8. Depending on the specific fabrication recipe and subsequent annealing procedures, $n_\textrm {GaOx}$ can range from 1.7 to 1.9 [55]. The resulting mode propagation loss is as small as 0.27 dB/cm for $n_\textrm {GaOx}= 1.9$. However, as $n_\textrm {GaOx}$ approaches 1.7, the propagation loss can increase rapidly to 2.6 dB/cm given a 5 µm thick cladding layer.
Fig. 8. Mode propagation loss at 1550 nm as a function of the refractive index of $ \textrm {GaO}_{\textrm{x}}$.
Therefore, we believe that the larger propagation losses at 1550 nm (compared to 633 nm and 1064 nm) for the samples are mainly due to the mode leakage resulting from the decrease of $ \textrm {GaO}_{\textrm{x}}$ refractive index. With a commercial foundry where thicker bottom oxide layer can be deposited to ensure no leakage into the silicon substrate, we anticipate that the ultimate loss at 1550 nm can be reduced to similar level of 633 nm and 1064 nm.
Funding
Gordon and Betty Moore Foundation (GBMF4744); National Science Foundation (1535711, ECCS-1542152).
Acknowledgments
We acknowledge Gordon and Betty Moore Foundation and the ACHIP collaboration for supporting the project. Special thanks to Dr. K. J. Leedle and D. S. Black for helpful discussions.
Disclosures
The authors declare no conflicts of interest.
References
1. D. Edwards, T. Mason, F. Goutenoire, and K. Poeppelmeier, “A new transparent conducting oxide in the Ga2 O3–In2 O3–Sn O2 system,” Appl. Phys. Lett. 70(13), 1706–1708 (1997). [CrossRef]
2. M. Orita, H. Ohta, M. Hirano, and H. Hosono, “Deep-ultraviolet transparent conductive β-Ga2O3 thin films,” Appl. Phys. Lett. 77(25), 4166–4168 (2000). [CrossRef]
3. M. Rebien, W. Henrion, M. Hong, J. Mannaerts, and M. Fleischer, “Optical properties of gallium oxide thin films,” Appl. Phys. Lett. 81(2), 250–252 (2002). [CrossRef]
4. D.-S. Wuu, S.-L. Ou, R.-H. Horng, P. Ravadgar, T.-Y. Wang, and H.-Y. Lee, in “Growth and characterization of Ga2O3 on sapphire substrates for UV sensor applications,” in Oxide-based Materials and Devices III, vol. 8263 (International Society for Optics and Photonics, 2012), p. 826317
5. K. D. Chabak, N. Moser, A. J. Green, D. E. Walker Jr, S. E. Tetlak, E. Heller, A. Crespo, R. Fitch, J. P. McCandless, K. Leedy, M. Baldini, G. Wagner, Z. Galazka, X. Li, and G. Jessen, “Enhancement-mode Ga2O3 wrap-gate fin field-effect transistors on native (100) β-Ga2O3 substrate with high breakdown voltage,” Appl. Phys. Lett. 109(21), 213501 (2016). [CrossRef]
6. S. Rafique, L. Han, M. J. Tadjer, J. A. Freitas Jr, N. A. Mahadik, and H. Zhao, “Homoepitaxial growth of β-Ga2O3 thin films by low pressure chemical vapor deposition,” Appl. Phys. Lett. 108(18), 182105 (2016). [CrossRef]
7. M. Higashiwaki, K. Sasaki, H. Murakami, Y. Kumagai, A. Koukitu, A. Kuramata, T. Masui, and S. Yamakoshi, “Recent progress in Ga2O3 power devices,” Semicond. Sci. Technol. 31(3), 034001 (2016). [CrossRef]
8. S. Rafique, L. Han, and H. Zhao, “Ultrawide bandgap β-Ga2O3 thin films: growths, properties and devices,” ECS Trans. 80(7), 203–216 (2017). [CrossRef]
9. H.-C. Huang, M. Kim, X. Zhan, K. Chabak, J. D. Kim, A. Kvit, D. Liu, Z. Ma, J.-M. Zuo, and X. Li, “High aspect ratio β-Ga2O3 fin arrays with low-interface charge density by inverse metal-assisted chemical etching,” ACS Nano 13(8), 8784–8792 (2019). [CrossRef]
10. M. A. Mastro, A. Kuramata, J. Calkins, J. Kim, F. Ren, and S. Pearton, “Perspective - opportunities and future directions for Ga2O3,” ECS J. Solid State Sci. Technol. 6(5), P356–P359 (2017). [CrossRef]
11. J. Tsao, S. Chowdhury, M. Hollis, D. Jena, N. Johnson, K. Jones, R. Kaplar, S. Rajan, C. Van de Walle, E. Bellotti, and C. Chua, “Ultrawide-bandgap semiconductors: Research opportunities and challenges,” Adv. Electron. Mater. 4(1), 1600501 (2018). [CrossRef]
12. S. Pearton, F. Ren, M. Tadjer, and J. Kim, “Perspective: Ga2O3 for ultra-high power rectifiers and mosfets,” J. Appl. Phys. 124(22), 220901 (2018). [CrossRef]
13. M. Higashiwaki, K. Sasaki, A. Kuramata, T. Masui, and S. Yamakoshi, “Gallium oxide (Ga2 O3) metal-semiconductor field-effect transistors on single-crystal β-Ga2O3 (010) substrates,” Appl. Phys. Lett. 100(1), 013504 (2012). [CrossRef]
14. W. S. Hwang, A. Verma, H. Peelaers, V. Protasenko, S. Rouvimov, H. Xing, A. Seabaugh, W. Haensch, C. V. de Walle, Z. Galazka, M. Albrecht, R. Fornari, and D. Jena, “High-voltage field effect transistors with wide-bandgap β-Ga2O3 nanomembranes,” Appl. Phys. Lett. 104(20), 203111 (2014). [CrossRef]
15. J. Bae, H. W. Kim, I. H. Kang, G. Yang, and J. Kim, “High breakdown voltage quasi-two-dimensional β − Ga2O3 field-effect transistors with a boron nitride field plate,” Appl. Phys. Lett. 112(12), 122102 (2018). [CrossRef]
16. A. K. Chandiran, N. Tetreault, R. Humphry-Baker, F. Kessler, E. Baranoff, C. Yi, M. K. Nazeeruddin, and M. Gratzel, “Subnanometer Ga2O3 tunnelling layer by atomic layer deposition to achieve 1.1 V open-circuit potential in dye-sensitized solar cells,” Nano Lett. 12(8), 3941–3947 (2012). [CrossRef]
17. S.-H. Yuan, C.-C. Wang, S.-Y. Huang, and D.-S. Wuu, “Improved responsivity drop from 250 to 200 nm in sputtered gallium oxide photodetectors by incorporating trace aluminum,” IEEE Electron Device Lett. 39(2), 220–223 (2018). [CrossRef]
18. S. Cui, Z. Mei, Y. Zhang, H. Liang, and X. Du, “Room-temperature fabricated amorphous Ga2O3 high-response-speed solar-blind photodetector on rigid and flexible substrates,” Adv. Opt. Mater. 5(19), 1700454 (2017). [CrossRef]
19. D. Guo, Y. Su, H. Shi, P. Li, N. Zhao, J. Ye, S. Wang, A. Liu, Z. Chen, C. Li, and W. Tang, “Self-powered ultraviolet photodetector with superhigh photoresponsivity (3.05 A/W) based on the GaN/Sn: Ga2O3 pn junction,” ACS Nano 12(12), 12827–12835 (2018). [CrossRef]
20. H. Deng, K. J. Leedle, Y. Miao, D. S. Black, K. E. Urbanek, J. McNeur, M. Kozák, A. Ceballos, P. Hommelhoff, O. Solgaard, R. L. Byer, and J. S. Harris, “Gallium oxide for high-power optical applications,” Adv. Opt. Mater. 8(7), 1901522 (2020). [CrossRef]
21. J.-H. Yoo, A. Lange, J. Chesser, S. Falabella, and S. Elhadj, “A survey of transparent conducting films and optoelectrical materials for high optical power applications,” Phys. Status Solidi A 216(22), 1900459 (2019). [CrossRef]
22. J.-H. Yoo, S. Rafique, A. Lange, H. Zhao, and S. Elhadj, “Lifetime laser damage performance of β-Ga2O3 for high power applications,” APL Mater. 6(3), 036105 (2018). [CrossRef]
23. K. J. Leedle, A. Ceballos, H. Deng, O. Solgaard, R. F. Pease, R. L. Byer, and J. S. Harris, “Dielectric laser acceleration of sub-100 keV electrons with silicon dual-pillar grating structures,” Opt. Lett. 40(18), 4344–4347 (2015). [CrossRef]
24. M. Kozák, J. McNeur, K. J. Leedle, H. Deng, N. Schönenberger, A. Ruehl, I. Hartl, H. Hoogland, R. Holzwarth, J. S. Harris, R. L. Byer, and P. Hommelhoff, “Transverse and longitudinal characterization of electron beams using interaction with optical near-fields,” Opt. Lett. 41(15), 3435–3438 (2016). [CrossRef]
25. M. Kozák, J. McNeur, K. J. Leedle, H. Deng, N. Schönenberger, A. Ruehl, I. Hartl, J. Harris, R. Byer, and P. Hommelhoff, “Optical gating and streaking of free electrons with sub-optical cycle precision,” Nat. Commun. 8(1), 14342 (2017). [CrossRef]
26. M. Kozák, P. Beck, H. Deng, J. McNeur, N. Schönenberger, C. Gaida, F. Stutzki, M. Gebhardt, J. Limpert, A. Ruehl, A. Hartl, O. Solgaard, J. Harris, R. Byer, and P. Hommelhoff, “Acceleration of sub-relativistic electrons with an evanescent optical wave at a planar interface,” Opt. Express 25(16), 19195–19204 (2017). [CrossRef]
27. K. P. Wootton, R. J. England, S. Tantawi, R. W. Aßmann, I. Hartl, F. X. Kärtner, W. Kuropka, F. Mayet, A. Ruehl, L. Rivkin, P. Hommelhoff, J. Illmer, J. M. A. Li, A. Mittelbach, N. Schönenberger, A. Tafel, R. Ischebeck, Y.-J. Lee, M. Qi, D. S. Black, R. L. Byer, H. Deng, S. Fan, J. S. Harris, T. Hughes, N. Sapra, O. Solgaard, J. Vučković, B. M. Cowan, O. Boine-Frankenheim, T. Egenolf, U. Niedermayer, and P. Musumeci, “Towards a fully integrated accelerator on a chip: Dielectric laser acceleration (DLA) from the source to relativistic electrons,” Proceedings of the 8th International Particle Accelerator Conference, (2017), pp. 2520–2525.
28. M. Kozák, M. Förster, J. McNeur, N. Schönenberger, K. Leedle, H. Deng, J. Harris, R. Byer, and P. Hommelhoff, “Dielectric laser acceleration of sub-relativistic electrons by few-cycle laser pulses,” Nucl. Instrum. Methods Phys. Res., Sect. A 865, 84–86 (2017). [CrossRef]
29. J. McNeur, M. Kozák, N. Schönenberger, K. J. Leedle, H. Deng, A. Ceballos, H. Hoogland, A. Ruehl, I. Hartl, R. Holzwarth, O. Solgaard, J. S. Harris, R. L. Byer, and P. Hommelhoff, “Elements of a dielectric laser accelerator,” Optica 5(6), 687–690 (2018). [CrossRef]
30. D. Cesar, J. Maxson, X. Shen, K. Wootton, S. Tan, R. England, and P. Musumeci, “Enhanced energy gain in a dielectric laser accelerator using a tilted pulse front laser,” Opt. Express 26(22), 29216–29224 (2018). [CrossRef]
31. K. J. Leedle, D. S. Black, Y. Miao, K. E. Urbanek, A. Ceballos, H. Deng, J. S. Harris, O. Solgaard, and R. L. Byer, “Phase-dependent laser acceleration of electrons with symmetrically driven silicon dual pillar gratings,” Opt. Lett. 43(9), 2181–2184 (2018). [CrossRef]
32. A-Chip collaboration, “Challenges in simulating beam dynamics of dielectric laser acceleration,” Int. J. Mod. Phys. A 34(36), 1942031 (2019). [CrossRef]
33. Y. Miao, D. S. Black, K. J. Leedle, Z. Zhao, H. Deng, A. Ceballos, R. L. Byer, J. S. Harris, and O. Solgaard, “Surface treatments of dielectric laser accelerators for increased laser-induced damage threshold,” Opt. Lett. 45(2), 391–394 (2020). [CrossRef]
34. T. W. Hughes, S. Tan, Z. Zhao, N. V. Sapra, K. J. Leedle, H. Deng, Y. Miao, D. S. Black, O. Solgaard, J. S. Harris, J. Vuckovic, S. Fan, R. J. England, Y. J. Lee, M. Qi, and R. L. Byer, “On-chip laser-power delivery system for dielectric laser accelerators,” Phys. Rev. Appl. 9(5), 054017 (2018). [CrossRef]
35. Z. Zhao, T. W. Hughes, S. Tan, H. Deng, N. Sapra, R. J. England, J. Vuckovic, J. S. Harris, R. L. Byer, and S. Fan, “Design of a tapered slot waveguide dielectric laser accelerator for sub-relativistic electrons,” Opt. Express 26(18), 22801–22815 (2018). [CrossRef]
36. S. Tan, Z. Zhao, K. Urbanek, T. Hughes, Y. J. Lee, S. Fan, J. S. Harris, and R. L. Byer, “Silicon nitride waveguide as a power delivery component for on-chip dielectric laser accelerators,” Opt. Lett. 44(2), 335–338 (2019). [CrossRef]
37. N. V. Sapra, K. Y. Yang, D. Vercruysse, K. J. Leedle, D. S. Black, R. J. England, L. Su, R. Trivedi, Y. Miao, O. Solgaard, R. L. Byer, and J. Vučkovicć, “On-chip integrated laser-driven particle accelerator,” Science 367(6473), 79–83 (2020). [CrossRef]
38. H. Chen, H. Fu, X. Huang, J. A. Montes, T.-H. Yang, I. Baranowski, and Y. Zhao, “Characterizations of the nonlinear optical properties for (010) and ($\bar {2}$01) beta-phase gallium oxide,” Opt. Express 26(4), 3938–3946 (2018). [CrossRef]
39. M. Belt, M. L. Davenport, J. E. Bowers, and D. J. Blumenthal, “Ultra-low-loss Ta2O5-core/SiO2-clad planar waveguides on si substrates,” Optica 4(5), 532–536 (2017). [CrossRef]
40. J. F. Bauters, M. J. Heck, D. John, D. Dai, M.-C. Tien, J. S. Barton, A. Leinse, R. G. Heideman, D. J. Blumenthal, and J. E. Bowers, “Ultra-low-loss high-aspect-ratio Si3N4 waveguides,” Opt. Express 19(4), 3163–3174 (2011). [CrossRef]
41. Super Conductor Materials, Inc. www.scm-inc.com.
42. J. E. Hogan, S. W. Kaun, E. Ahmadi, Y. Oshima, and J. S. Speck, “Chlorine-based dry etching of β-Ga2O3,” Semicond. Sci. Technol. 31(6), 065006 (2016). [CrossRef]
43. L. Zhang, A. Verma, H. G. Xing, and D. Jena, “Inductively-coupled-plasma reactive ion etching of single-crystal β-Ga2O3,” Jpn. J. Appl. Phys. 56(3), 030304 (2017). [CrossRef]
44. A. P. Shah and A. Bhattacharya, “Inductively coupled plasma reactive-ion etching of β-Ga2O3: Comprehensive investigation of plasma chemistry and temperature,” J. Vac. Sci. Technol., A 35(4), 041301 (2017). [CrossRef]
45. S. J. Pearton, E. A. Douglas, R. J. Shul, and F. Ren, “Plasma etching of wide bandgap and ultrawide bandgap semiconductors,” J. Vac. Sci. Technol., A 38(2), 020802 (2020). [CrossRef]
46. G. Battiston, R. Gerbasi, M. Porchia, R. Bertoncello, and F. Caccavale, “Chemical vapour deposition and characterization of gallium oxide thin films,” Thin Solid Films 279(1-2), 115–118 (1996). [CrossRef]
47. S.-D. Lee, K. Akaiwa, and S. Fujita, “Thermal stability of single crystalline alpha gallium oxide films on sapphire substrates,” Phys. Status Solidi C 10(11), 1592–1595 (2013). [CrossRef]
48. H. Peelaers, J. L. Lyons, J. B. Varley, and C. G. Van de Walle, “Deep acceptors and their diffusion in Ga2O3,” APL Mater. 7(2), 022519 (2019). [CrossRef]
49. S. Bhandari, M. Zvanut, and J. Varley, “Optical absorption of fe in doped Ga2O3,” J. Appl. Phys. 126(16), 165703 (2019). [CrossRef]
50. J. B. Varley, J. R. Weber, A. Janotti, and C. G. Van de Walle, “Oxygen vacancies and donor impurities in β-Ga2O3,” Appl. Phys. Lett. 97(14), 142106 (2010). [CrossRef]
51. H. Peelaers and C. G. Van de Walle, “Phonon-and charged-impurity-assisted indirect free-carrier absorption in Ga2O3,” Phys. Rev. B 100(8), 081202 (2019). [CrossRef]
52. P. Weiser, M. Stavola, W. B. Fowler, Y. Qin, and S. Pearton, “Structure and vibrational properties of the dominant O-H center in β-Ga2O3,” Appl. Phys. Lett. 112(23), 232104 (2018). [CrossRef]
53. S. Stepanov, V. Nikolaev, V. Bougrov, and A. Romanov, “Gallium oxide: Properties and application - a review,” Rev. Adv. Mater. Sci 44, 63–86 (2016).
54. A. Luchechko, V. Vasyltsiv, L. Kostyk, and O. Tsvetkova, “Origin of point defects in β-Ga2O3 single crystals doped with Mg2+ ions,” Acta Phys. Pol., A 133(4), 811–815 (2018). [CrossRef]
55. M. Al-Kuhaili, S. Durrani, and E. Khawaja, “Optical properties of gallium oxide films deposited by electron-beam evaporation,” Appl. Phys. Lett. 83(22), 4533–4535 (2003). [CrossRef]
56. A. T. Neal, S. Mou, S. Rafique, H. Zhao, E. Ahmadi, J. S. Speck, K. T. Stevens, J. D. Blevins, D. B. Thomson, N. Moser, K. D. Chabak, and G. H. Jessen, “Donors and deep acceptors in β-Ga2O3,” Appl. Phys. Lett. 113(6), 062101 (2018). [CrossRef]
57. H. Peelaers, E. Kioupakis, and C. G. Van de Walle, “Fundamental limits on optical transparency of transparent conducting oxides: Free-carrier absorption in SnO2,” Appl. Phys. Lett. 100(1), 011914 (2012). [CrossRef]
58. Y. Furukawa, K. Kitamura, A. Alexandrovski, R. Route, M. Fejer, and G. Foulon, “Green-induced infrared absorption in MgO doped LiNbO3,” Appl. Phys. Lett. 78(14), 1970–1972 (2001). [CrossRef]
59. S. Wang, V. Pasiskevicius, and F. Laurell, “Dynamics of green light-induced infrared absorption in KTiOPO4 and periodically poled KTiOPO4,” J. Appl. Phys. 96(4), 2023–2028 (2004). [CrossRef]
60. N. Poyiatzis, M. Athanasiou, J. Bai, Y. Gong, and T. Wang, “Monolithically integrated white light leds on (11–22) semi-polar gan templates,” Sci. Rep. 9(1), 1383 (2019). [CrossRef]
61. G. Liu, C. Yan, G. Zhou, J. Wen, Z. Qin, Q. Zhou, B. Li, R. Zheng, H. Wu, and Z. Sun, “Broadband white-light emission from alumina nitride bulk single crystals,” ACS Photonics 5(10), 4009–4013 (2018). [CrossRef]
62. M. Tran, D. Huang, T. Komljenovic, J. Peters, A. Malik, and J. Bowers, “Ultra-low-loss silicon waveguides for heterogeneously integrated silicon/III-V photonics,” Appl. Sci. 8(7), 1139 (2018). [CrossRef]
63. F. Shan, G. Liu, W. Lee, G. Lee, I. Kim, and B. Shin, “Structural, electrical, and optical properties of transparent gallium oxide thin films grown by plasma-enhanced atomic layer deposition,” J. Appl. Phys. 98(2), 023504 (2005). [CrossRef]
