Growth of α-Ga2O3 from Gallium Acetylacetonate under HCl Support by Mist Chemical Vapor Deposition

α-Ga2O3 films were grown on a c-plane sapphire substrate by HCl-supported mist chemical vapor deposition with multiple solution chambers, and the effect of HCl support on α-Ga2O3 film quality was investigated. The growth rate monotonically increased with increasing Ga supply rate. However, as the Ga supply rate was higher than 0.1 mmol/min, the growth rate further increased with increasing HCl supply rate. The surface roughness was improved by HCl support when the Ga supply rate was smaller than 0.07 mmol/min. The crystallinity of the α-Ga2O3 films exhibited an improvement with an increase in the film thickness, regardless of the solution preparation conditions, Ga supply rate, and HCl supply rate. These results indicate that there is a low correlation between the improvement of surface roughness and crystallinity in the α-Ga2O3 films grown under the conditions described in this paper.

Previous reports have shown that hydrochloric acid (HCl) contributes to the phaseselective growth and quality improvement of α-Ga 2 O 3 by HVPE [27], MOCVD [28], and mist CVD [21,22].In those reports, it is suggested that HCl present in the reaction field affects the surface energy of the sapphire substrate due to the presence of Al-Cl bonds [27] and contributes to the stability of each crystalline phase [28].With regard to the case of mist CVD with one solution chamber, it has been suggested that HCl in the gallium acetylacetonate (Ga(acac) 3 ) precursor solution contributes to the state of Ga in solution [22].Our research group examined the effect of HCl support by using mist CVD with multiple solution chambers on the use of GaCl 3 as a precursor [21].Our findings indicated that the growth rate of α-Ga 2 O 3 increased with increasing HCl supply, and that the larger Ga supply rate and HCl support suppressed the incorporation of carbon impurities into the film, resulting in improved surface roughness.However, the effect of HCl on Ga 2 O 3 growth and the underlying mechanisms remain incompletely understood and necessitate further investigation.Thus, this study aims to investigate the effect of HCl support on α-Ga 2 O 3 growth by mist CVD with multiple solution chambers when Ga(acac) 3 is employed as a precursor.It was found that HCl support resulted in a discrepancy in the trends of growth rate and surface roughness variation in comparison to the case when GaCl 3 is employed as a precursor.In addition, the effect of two starting materials, Ga(acac) 3 and GaCl 3 , on the crystallinity of α-Ga 2 O 3 was also examined.

Experimental Methods
α-Ga 2 O 3 thin films were grown on c-plane sapphire substrates using a mist CVD system, which is described in Figure 1 [38].The growth conditions are summarized in Table 1.Ga(acac) 3 was dissolved in a mixture of deionized water and hydrochloric acid (HCl), and gallium trichloride, GaCl 3 , was dissolved in deionized water.Each Ga solution was stirred until completely dissolved for a minimum of one hour.These solutions were used, respectively, as Ga precursor solutions.The Ga(acac) 3 solution was set at concentrations of 20-100 mM and GaCl 3 was set at concentrations of 50-300 mM.HCl diluted in de-ionized water at concentrations of 0.57-1.13M was used as a support to improve the quality of α-Ga 2 O 3 .Each solution was individually atomized using ultrasonics and then supplied to the fine-channel reactor with carrier and dilution gases.If multiple solutions were used, they were supplied to the reactor via the mixing chamber.Ga and HCl supply rates were controlled by solution concentrations and flow rates of carrier gas.Mist CVD with multiple solution chambers, also known as third-generation (3rd G) mist CVD, allows mist generated from multiple solutions to be supplied to the reaction field without contributing to the solution state.This process suppresses side reactions and makes it suitable for composition control [13], doping [39], and the introduction of reaction support agents [21].Details of this mist CVD system were described in Ref. [38].The growth temperature was set at 400 • C.
mist CVD [21,22].In those reports, it is suggested that HCl present in the reaction field affects the surface energy of the sapphire substrate due to the presence of Al-Cl bonds [27] and contributes to the stability of each crystalline phase [28].With regard to the case of mist CVD with one solution chamber, it has been suggested that HCl in the gallium acetylacetonate (Ga(acac)3) precursor solution contributes to the state of Ga in solution [22].Our research group examined the effect of HCl support by using mist CVD with multiple solution chambers on the use of GaCl3 as a precursor [21].Our findings indicated that the growth rate of α-Ga2O3 increased with increasing HCl supply, and that the larger Ga supply rate and HCl support suppressed the incorporation of carbon impurities into the film, resulting in improved surface roughness.However, the effect of HCl on Ga2O3 growth and the underlying mechanisms remain incompletely understood and necessitate further investigation.Thus, this study aims to investigate the effect of HCl support on α-Ga2O3 growth by mist CVD with multiple solution chambers when Ga(acac)3 is employed as a precursor.It was found that HCl support resulted in a discrepancy in the trends of growth rate and surface roughness variation in comparison to the case when GaCl3 is employed as a precursor.In addition, the effect of two starting materials, Ga(acac)3 and GaCl3, on the crystallinity of α-Ga2O3 was also examined.

Experimental Methods
α-Ga2O3 thin films were grown on c-plane sapphire substrates using a mist CVD system, which is described in Figure 1 [38].The growth conditions are summarized in Table 1.Ga(acac)3 was dissolved in a mixture of deionized water and hydrochloric acid (HCl), and gallium trichloride, GaCl3, was dissolved in deionized water.Each Ga solution was stirred until completely dissolved for a minimum of one hour.These solutions were used, respectively, as Ga precursor solutions.The Ga(acac)3 solution was set at concentrations of 20-100 mM and GaCl3 was set at concentrations of 50-300 mM.HCl diluted in de-ionized water at concentrations of 0.57-1.13M was used as a support to improve the quality of α-Ga2O3.Each solution was individually atomized using ultrasonics and then supplied to the fine-channel reactor with carrier and dilution gases.If multiple solutions were used, they were supplied to the reactor via the mixing chamber.Ga and HCl supply rates were controlled by solution concentrations and flow rates of carrier gas.Mist CVD with multiple solution chambers, also known as third-generation (3rd G) mist CVD, allows mist generated from multiple solutions to be supplied to the reaction field without contributing to the solution state.This process suppresses side reactions and makes it suitable for composition control [13], doping [39], and the introduction of reaction support agents [21].Details of this mist CVD system were described in Ref. [38].The growth temperature was set at 400 °C.A schematic image of mist CVD system with multiple solution chambers, a mixing chamber, and a fine-channel reactor [38].
Table 1.Summary of growth conditions of α-Ga 2 O 3 thin films.

α-Ga 2 O 3 Film Thicknesses Estimated from Ellipsometry and from Laue Fringes in XRD Spectra
Figure 2a shows a typical XRD spectrum of the α-Ga 2 O 3 film grown in this study.When the film thickness was less than about 250 nm, Laue fringes were clearly observed around the α-Ga 2 O 3 peak, as shown in Figure 2a. Figure 2b shows the α-Ga 2 O 3 film thickness estimated from the spacing of the Laue fringes in the XRD spectra and ellipsometry.The estimated film thickness errors were less than 3% and 1%, respectively.The α-Ga 2 O 3 films were grown on single-sided mirror c-plane sapphire substrates with 50-100 mM Ga(acac) 3 precursor solution in this study.The dashed line in Figure 2b indicates that the values on the vertical and horizontal axes are the same.Figure 2b shows similar thicknesses, as estimated from the spacing of the Laue fringes in the XRD spectra and from ellipsometry.A comparison between thicknesses determined from Laue spacings and measured by direct methods such as profilers [40]/cross-sectional transmission electron microscopy (TEM) images [41] also shows a discrepancy of <5%.This suggests that for α-Ga 2 O 3 grown on sapphire substrates, the film thickness calculated by Laue fringe spacing is close to the actual film thickness.Therefore, the thickness of α-Ga 2 O 3 films on sapphire substrates was determined based on the spacings of Laue fringes in this study.

α-Ga2O3 Film Thicknesses Estimated from Ellipsometry and from Laue Fringes in XRD Spectra
Figure 2a shows a typical XRD spectrum of the α-Ga2O3 film grown in this study.When the film thickness was less than about 250 nm, Laue fringes were clearly observed around the α-Ga2O3 peak, as shown in Figure 2a. Figure 2b shows the α-Ga2O3 film thickness estimated from the spacing of the Laue fringes in the XRD spectra and ellipsometry.The estimated film thickness errors were less than 3% and 1%, respectively.The α-Ga2O3 films were grown on single-sided mirror c-plane sapphire substrates with 50-100 mM Ga(acac)3 precursor solution in this study.The dashed line in Figure 2b indicates that the values on the vertical and horizontal axes are the same.Figure 2b shows similar thicknesses, as estimated from the spacing of the Laue fringes in the XRD spectra and from ellipsometry.A comparison between thicknesses determined from Laue spacings and measured by direct methods such as profilers [40]/cross-sectional transmission electron microscopy (TEM) images [41] also shows a discrepancy of <5%.This suggests that for α-Ga2O3 grown on sapphire substrates, the film thickness calculated by Laue fringe spacing is close to the actual film thickness.Therefore, the thickness of α-Ga2O3 films on sapphire substrates was determined based on the spacings of Laue fringes in this study.

Effect of HCl on α-Ga 2 O 3 Growth Using Ga(acac) 3 as a Starting Material
Figure 3a,b show the α-Ga 2 O 3 growth rate versus Ga supply rate and HCl supply rate when Ga(acac) 3 is used as the Ga precursor.The colors indicate the concentrations of the Ga solutions, while solid dots (•) and crosses (×) indicate data points of the α-Ga 2 O 3 growths with and without HCl support, respectively.It should be noted that HCl is added in the solvent during Ga solution preparation to dissolve Ga(acac) 3 and is also supplied additionally during growth from a separated HCl solution.Consequently, the HCl supply rate represents the total amount of HCl (i) contained in Ga solutions and (ii) supplied from the dedicated HCl source.The HCl supply rate is not zero when HCl is not supplied and the data points are plotted as crosses (×) in this case.Figure 3a clearly shows that the growth rate increased monotonically with increasing Ga supply rate.At the Ga concentration of 20 mM, the growth rates with and without HCl support were similar.However, at the Ga concentrations higher than 20 mM, the growth rates without HCl support were smaller than that with HCl support.This difference in the growth rates may be attributed to the formation of oligomers (or supermolecules) at higher Ga concentrations without HCl support, which prevented Ga molecules from reaching the substrate due to steric hindrance, as illustrated in Figure 4. Conceivably, it is expected that HCl will inhibit the formation of oligomers.In Figure 3b, it was observed that the growth rate remained constant with increasing HCl supply rate at a Ga solution concentration of 20 mM.In contrast, the growth rate seems to increase slightly with increasing HCl supply rate, despite the Ga supply rate being almost constant at Ga solution concentrations of 80 and 100 mM.This trend was similar to that observed when GaCl 3 was used as a precursor [21].Figure 5a,b show the results of plotting the TPR (thickness per surface roughness), which is the film thickness divided by the RMS surface roughness, against the Ga and HCl supply rates.It should be noted that the TPR is capable of distinguishing between 2D and 3D growth modes and can be employed in the evaluation of surface roughness of samples with different film thicknesses [21].As shown in Figure 5a, the TPR increases with decreasing Ga supply rates when the Ga supply rate is below approximately 0.07 mmol/min.Conversely, when the Ga supply rate exceeds ~0.07 mmol/min, the TPR remains constant at approximately 30. Figure 5c,d show the AFM images of samples A and B, which were marked by arrows in Figure 5a and 5b, respectively.The values of T L and RMS shown in the top and bottom of the AFM images present the thickness and RMS surface roughness, respectively.Although samples A and B exhibit nearly identical film thicknesses, the TPR value is larger for sample A due to its smaller surface roughness.However, the AFM images demonstrate that both A and B exhibit island growth.The surface morphology of all samples grown in this study was similar to that of samples A and B. When the TPR was small, the surface was composed of small surface grains, as observed in sample B (Figure 5d).Conversely, when the TPR increases, the surface grain size increases, as ob-served in sample A (Figure 5c).While sample A was, in fact, a 3D growth, the trend of surface roughness change with film thickness is analogous to that of 2D growth, indicating that it is a 2D-like growth mode.Therefore, when growing α-Ga 2 O 3 using Ga(acac) 3 , the growth mode is 3D when the TPR is about 30 and 2D-like when it is larger.According to a previous report, the film grows in a 2D-like growth mode when the impurity concentration in the film is reduced by HCl support [21].It is assumed that the carbon impurity concentration in the film is reduced by HCl support through a similar mechanism, resulting in a 2D-like growth mode when Ga(acac) 3 is used as the starting material.However, Ga(acac) 3 has 15 C atoms for each Ga atom, unlike GaCl 3 .Therefore, it is expected that HCl-supported growth did not improve the surface roughness of the films when the Ga supply rate became large enough to suppress impurity uptake in the films.

Growth Characteristics of α-Ga 2 O 3 for HCl Support in Each Starting Material
Figure 6 shows a summary of the effect of HCl support on α-Ga 2 O 3 growth when Ga(acac) 3 and GaCl 3 are employed as starting materials.Note that the experimental data of the GaCl 3 results are presented in Ref. [21].When Ga(acac) 3 was employed as the starting material, the growth rate exhibited a monotonic increase with increasing Ga supply rate.
Nanomaterials 2024, 14, 1221 6 of 9 However, when the Ga supply rate exceeded 0.1 mmol/min, the growth rate demonstrated an additional increase with increasing HCl supply rate, in addition to the Ga supply rate.In contrast, when GaCl 3 was employed, the growth rate exhibited an increase solely in response to an increasing HCl supply rate.With regard to surface roughness, it was observed that there was no dependence on the Ga and HCl supply rates when Ga(acac) 3 was employed as the starting material, provided that the Ga supply rate was larger than 0.07 mmol/min.However, when the Ga supply rate was smaller, the surface roughness was improved by HCl support.In contrast, when GaCl 3 was employed, there was no correlation between the Ga and HCl supply rates when the Ga supply rate was less than 0.1 mmol/min.However, the surface roughness exhibited an improvement with an increase in the Ga and HCl supply rates when the Ga supply rate was higher.

Growth Characteristics of α-Ga2O3 for HCl Support in Each Starting Material
Figure 6 shows a summary of the effect of HCl support on α-Ga2O3 growth when Ga(acac)3 and GaCl3 are employed as starting materials.Note that the experimental data of the GaCl3 results are presented in Ref. [21].When Ga(acac)3 was employed as the starting material, the growth rate exhibited a monotonic increase with increasing Ga supply rate.However, when the Ga supply rate exceeded 0.1 mmol/min, the growth rate demonstrated an additional increase with increasing HCl supply rate, in addition to the Ga supply rate.In contrast, when GaCl3 was employed, the growth rate exhibited an increase solely in response to an increasing HCl supply rate.With regard to surface roughness, it was observed that there was no dependence on the Ga and HCl supply rates when Ga(acac)3 was employed as the starting material, provided that the Ga supply rate was larger than 0.07 mmol/min.However, when the Ga supply rate was smaller, the surface roughness was improved by HCl support.In contrast, when GaCl3 was employed, there was no correlation between the Ga and HCl supply rates when the Ga supply rate was less than 0.1 mmol/min.However, the surface roughness exhibited an improvement with an increase in the Ga and HCl supply rates when the Ga supply rate was higher.Figure 7a,b illustrate the outcomes of plotting the thickness dependence of the full width at half maximum (FWHM) of the α-Ga2O3 (0006) peak grown on sapphire, obtained through X-ray rocking curve measurements on linear and logarithmic axes, respectively.In Figure 7a,b, the thickness of the Ga2O3 film on a quartz substrate, which was grown simultaneously with the film on sapphire, was measured by ellipsometry and utilized since the fringes are not visible when the film is thick.The FWHM decreased inversely with increasing film thickness under all growth conditions, regardless of solute species, Ga supply rate, or HCl supply rate.This indicates that the crystallinity of α-Ga2O3 is not improved by the solution preparation conditions or HCl support when grown by mist CVD.In addition, when plotted on a logarithmic axis, as shown in Figure 7b, there is a slight difference depending on the starting material, fitting y = 5503.6x− 0.795 and y = 2080.9x− 0.592 when Ga(acac)3 and GaCl3 are used as the starting materials, respectively.At this stage, there is no adequate physical explanation as to why differences in starting materials cause differences in the thickness dependence of crystallinity.This may be attributed to the amount of impurities, such as carbon and chloride atoms, incorporated into the film as impurities.It is anticipated that future growth of α-Ga2O3 using alternative starting  Figure 7a,b illustrate the outcomes of plotting the thickness dependence of the full width at half maximum (FWHM) of the α-Ga 2 O 3 (0006) peak grown on sapphire, obtained through X-ray rocking curve measurements on linear and logarithmic axes, respectively.In Figure 7a,b, the thickness of the Ga 2 O 3 film on a quartz substrate, which was grown simultaneously with the film on sapphire, was measured by ellipsometry and utilized since the fringes are not visible when the film is thick.The FWHM decreased inversely with increasing film thickness under all growth conditions, regardless of solute species, Ga supply rate, or HCl supply rate.This indicates that the crystallinity of α-Ga 2 O 3 is not improved by the solution preparation conditions or HCl support when grown by mist CVD.In addition, when plotted on a logarithmic axis, as shown in Figure 7b, there is a slight difference depending on the starting material, fitting y = 5503.6x− 0.795 and y = 2080.9x− 0.592 when Ga(acac) 3 and GaCl 3 are used as the starting materials, respectively.At this stage, there is no adequate physical explanation as to why differences in starting materials cause differences in the thickness dependence of crystallinity.This may be attributed to the amount of impurities, such as carbon and chloride atoms, incorporated into the film as impurities.It is anticipated that future growth of α-Ga 2 O 3 using alternative starting materials, such as GaBr 3 , gallium nitrate (Ga(NO 3 ) 3 ), and gallium triiodide (GaI 3 ), will elucidate the underlying cause of this trend.In any case, it can be concluded that there is a low correlation between the improvement of surface roughness and crystallinity in the α-Ga 2 O 3 film grown under the conditions described in this paper.materials, such as GaBr3, gallium nitrate (Ga(NO3)3), and gallium triiodide (GaI3), will elucidate the underlying cause of this trend.In any case, it can be concluded that there is a low correlation between the improvement of surface roughness and crystallinity in the α-Ga2O3 film grown under the conditions described in this paper.

Facts Revealed in α-Ga2O3 Grown by Mist CVD
Table 2 presents a summary of the controllability of surface roughness and crystallinity in α-Ga2O3 films.With regard to surface roughness, it was found that this property could be controlled by the type of precursor, Ga supply rate, and HCl support.In contrast, crystallinity was found to improve only with increasing film thickness, regardless of the Ga supply rate and HCl support, although there were some differences in the results depending on the type of starting material.

Conclusions
This study investigated the effect of HCl support on film quality during the growth of α-Ga2O3 by mist CVD using Ga(acac)3 as a precursor.It also examined the difference in crystallinity resulting from different starting materials.The growth rate monotonically increased with increasing Ga supply rate.However, as the Ga solution concentration increased, the growth rate further increased with increasing HCl supply rate, showing the same trend as that of GaCl3.When the Ga solution concentration exceeded 20 mM, the growth rate without HCl support was smaller than that with HCl support.This effect is believed to be caused by the formation of oligomers at high Ga solution concentrations, which inhibited Ga molecules from reaching the substrate due to steric hindrance.It is therefore anticipated that HCl support will be effective in inhibiting the formation of oligomers.Furthermore, the HCl support resulted in improved surface roughness of α-Ga2O3 due to the suppression of carbon impurities incorporated into the film, similarly to GaCl3.However, there was a difference in the amount of Ga supply between Ga(acac)3 and GaCl3 when they were employed as starting materials, respectively.In the case of Ga(acac)3, surface roughness tended to improve with smaller amounts of Ga supply.   2 presents a summary of the controllability of surface roughness and crystallinity in α-Ga 2 O 3 films.With regard to surface roughness, it was found that this property could be controlled by the type of precursor, Ga supply rate, and HCl support.In contrast, crystallinity was found to improve only with increasing film thickness, regardless of the Ga supply rate and HCl support, although there were some differences in the results depending on the type of starting material.

Conclusions
This study investigated the effect of HCl support on film quality during the growth of α-Ga 2 O 3 by mist CVD using Ga(acac) 3 as a precursor.It also examined the difference in crystallinity resulting from different starting materials.The growth rate monotonically increased with increasing Ga supply rate.However, as the Ga solution concentration increased, the growth rate further increased with increasing HCl supply rate, showing the same trend as that of GaCl 3 .When the Ga solution concentration exceeded 20 mM, the growth rate without HCl support was smaller than that with HCl support.This effect is believed to be caused by the formation of oligomers at high Ga solution concentrations, which inhibited Ga molecules from reaching the substrate due to steric hindrance.It is therefore anticipated that HCl support will be effective in inhibiting the formation of oligomers.Furthermore, the HCl support resulted in improved surface roughness of α-Ga 2 O 3 due to the suppression of carbon impurities incorporated into the film, similarly to GaCl 3 .However, there was a difference in the amount of Ga supply between Ga(acac) 3 and GaCl 3 when they were employed as starting materials, respectively.In the case of Ga(acac) 3 , surface roughness tended to improve with smaller amounts of Ga supply.Ga(acac) 3 has 15 C atoms for each Ga atom, unlike GaCl 3 .Therefore, it is expected that with high Ga supply rates, the HCl support is not sufficient to suppress the incorporation of impurities into the films, leading to no improvement in surface roughness.Crystallinity improved with increasing film thickness, regardless of the solution preparation conditions, including solute species and concentration, Ga supply rate, and HCl supply rate.This indicates

Figure 1 .
Figure 1.A schematic image of mist CVD system with multiple solution chambers, a mixing chamber, and a fine-channel reactor[38].

Figure 1 .
Figure 1.A schematic image of mist CVD system with multiple solution chambers, a mixing chamber, and a fine-channel reactor[38].

Figure 2 .
Figure 2. (a) Typical XRD spectra of α-Ga2O3 film grown on c-plane sapphire substrate and (b) thickness estimation results using different methods (vertical axis TL: thickness calculated from the

Figure 2 .
Figure 2. (a) Typical XRD spectra of α-Ga 2 O 3 film grown on c-plane sapphire substrate and (b) thickness estimation results using different methods (vertical axis T L : thickness calculated from the spacings of Laue fringes in XRD spectra; horizontal axis T E : thickness calculated from spectroscopic ellipsometry).The dashed line represents the same value on both the vertical and horizontal axes.

Figure 3 .
Figure 3. Dependence of the α-Ga2O3 growth rates on (a) Ga supply rate and (b) HCl supply rate.

Figure 4 .Figure 3 .
Figure 4.A schematic image of an oligomer acting as a steric hindrance, preventing Ga molecules from attaching to the growth surface.

Figure 3 .
Figure 3. Dependence of the α-Ga2O3 growth rates on (a) Ga supply rate and (b) HCl supply rate.

Figure 4 .
Figure 4.A schematic image of an oligomer acting as a steric hindrance, preventing Ga molecules from attaching to the growth surface.

Figure 5 .Figure 4 .
Figure 5. Dependence of the TPR on (a) Ga supply rate and (b) HCl supply rate.(c,d) show the AFM images of samples A and B, respectively.

Figure 3 .
Figure 3. Dependence of the α-Ga2O3 growth rates on (a) Ga supply rate and (b) HCl supply rate.

Figure 4 .
Figure 4.A schematic image of an oligomer acting as a steric hindrance, preventing Ga molecules from attaching to the growth surface.

Figure 5 .Figure 5 .
Figure 5. Dependence of the TPR on (a) Ga supply rate and (b) HCl supply rate.(c,d) show the AFM images of samples A and B, respectively.

Figure 6 .
Figure 6.A summary of the effect of HCl support on the growth rate and surface roughness of α-Ga2O3 thin films when Ga(acac)3 and GaCl3 are employed as starting materials.

Figure 6 .
Figure 6.A summary of the effect of HCl support on the growth rate and surface roughness of α-Ga 2 O 3 thin films when Ga(acac) 3 and GaCl 3 are employed as starting materials.

Figure 7 .
Figure 7.The results of plotting the thickness dependence of the full width at half maximum (FWHM) of the α-Ga2O3 (0006) peak on c-plane sapphire substrate measured by X-ray rocking curve on (a) linear and (b) logarithmic axes.

Figure 7 .
Figure 7.The results of plotting the thickness dependence of the full width at half maximum (FWHM) of the α-Ga 2 O 3 (0006) peak on c-plane sapphire substrate measured by X-ray rocking curve on (a) linear and (b) logarithmic axes.

Table 1 .
Summary of growth conditions of α-Ga2O3 thin films.

Table 2 .
Controllability of surface roughness and crystallinity in α-Ga 2 O 3 films.