Next Article in Journal
An Analysis of the Training Data Impact for Domain-Adapted Tokenizer Performances—The Case of Serbian Legal Domain Adaptation
Previous Article in Journal
Optimization of Arrangements of Heat-Storage Bricks in a Regenerative Combustion System by Tree Search
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Applied Sciences
Volume 15
Issue 13
10.3390/app15137487
applsci-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

DFT Study of Initial Surface Reactions in Gallium Nitride Atomic Layer Deposition Using Trimethylgallium and Ammonia

by
P. Pungboon Pansila
1,
Seckson Sukhasena
2,
Saksit Sukprasong
1,
Worasitti Sriboon
1,
Wipawee Temnuch
1,
Tongsai Jamnongkan
1,* and
Tanabat Promjun
1,*
1
Faculty of Science at Si Racha, Kasetsart University, Sriracha Campus, Chonburi 20230, Thailand
2
Department of Physics, Faculty of Science, Naresuan University, Phitsanulok 65000, Thailand
*
Authors to whom correspondence should be addressed.
Appl. Sci. 2025, 15(13), 7487; https://doi.org/10.3390/app15137487
Submission received: 23 May 2025 / Revised: 29 June 2025 / Accepted: 30 June 2025 / Published: 3 July 2025
(This article belongs to the Section Surface Sciences and Technology)
Browse Figures
Versions Notes

Abstract

The initial surface reaction of gallium nitride (GaN) grown by atomic layer deposition (GaN-ALD) was investigated using density functional theory (DFT) calculations. Trimethylgallium (TMG) and ammonia (NH3) were used as gallium (Ga) and nitrogen (N) precursors, respectively. DFT calculations at the B3LYP/6-311+G(2d,p) and 6-31G(d) levels were performed to compute relative energies and optimize chemical structures, respectively. TMG adsorption on Si15H18–(NH2)2 and Si15H20=(NH)2 clusters was modeled, where –NH2 and =NH surface species served as adsorption sites. The reaction mechanisms in the adsorption and nitridation steps were investigated. The results showed that TMG can adsorb on both surface adsorption sites. In the initial adsorption stage, TMG adsorbs onto =NH- and –NH2-terminated Si(100) surfaces with activation energies of 1.11 and 2.00 eV, respectively, indicating that the =NH site is more reactive. During subsequent NH3 adsorption, NH3 adsorbs onto the residual TMG on the =NH- and –NH2-terminated surfaces with activation energies of approximately 2.00 ± 0.02 eV. The reaction pathways indicate that NH3 adsorbs via similar mechanisms on both surfaces, resulting in comparable nitridation kinetics. Furthermore, this study suggests that highly reactive NH2− species generated in the gas phase from ionized NH3 may help reduce the process temperature in the GaN-ALD process.

1. Introduction

Atomic layer deposition (ALD) is a technique based on sequential, self-limiting surface reactions, in which gas-phase precursors are alternately introduced into the ALD reaction chamber to form a single layer. Each cycle includes adsorption and reaction steps [1]. The thickness of the film can be controlled by varying the number of deposition cycles. ALD enables the deposition of very high-quality films at the atomic scale. Films prepared using this method are extremely smooth and uniform, without pinholes, even on complex substrates. Both the thickness and composition of the films can be precisely controlled at the atomic level [2]. ALD plays a key role in a wide range of high-quality thin-film coating industries, such as semiconductors, microelectronics, electroluminescent displays, photovoltaics, and medical devices [3,4,5,6,7,8,9]. In particular, the semiconductor industry today relies heavily on ALD technology due to ongoing miniaturization, as the size of modern semiconductor chips continues to decrease. Films must be uniformly deposited over increasingly larger wafer sizes while also conforming to the complex three-dimensional (3D) structures of integrated circuit (IC) devices. The uniformity, conformality, and composition of thin films have been successfully achieved using ALD. Currently, no other technology can match ALD in this regard, and it holds great potential for applications in the coating industry [10].
Gallium nitride (GaN) belongs to the group of III–V semiconductors and has a direct band gap, high breakdown field, excellent electron mobility, and good thermal stability [11]. GaN plays a crucial role in a wide range of applications, including light-emitting diodes (LEDs) [12,13], heterojunction field-effect transistors (FETs) [14], p–i–n rectifiers [15,16], and photodiodes [17]. Although GaN offers many advantages over silicon (Si), its high production cost limits large-scale commercialization. Producing GaN in large quantities is expensive, making mass production challenging [11]. To reduce production costs, silicon is commonly used as a substrate in the GaN deposition process [18], as it also offers better thermal conductivity and good electrical conductivity [19,20]. Despite these advantages, the mismatch in thermal expansion coefficients between GaN and silicon substrates can lead to film cracking and structural failure, particularly when the growth temperature exceeds 1000 °C [21,22]. Several conventional methods have been developed for GaN deposition, such as molecular beam epitaxy (MBE) [23], sputtering [24], supersonic gas jet-assisted vapor deposition [25,26], metalorganic chemical vapor deposition (MOCVD) [27,28], and hydride vapor phase epitaxy (HVPE) [29]. Although these techniques can deposit GaN films at temperatures below 1000 °C, the resulting film quality is often relatively low. Additionally, most of these methods still require high deposition temperatures (typically above 700 °C), which are unsuitable for heat-sensitive substrates, such as flexible substrates used in light-emitting applications [30,31].
ALD is a method that was developed for depositing conformal metal thin films at low temperature. The ALD technique can grow a high quality GaN film. At the temperatures of 800–1100 °C, GaN film was grown based on MOCVD using trimethylgallium (TMG) and NH3 [27,28]. A thermal ALD process has been reported at a processing temperature of 375–425 °C using TMG and NH3 sources [11]. Additionally, the plasma and electron-enhanced ALD processes make it possible to reduce the process temperature to near room temperature (RT) [30,32]. This allows for the GaN films to be grown on heat-sensitive materials and expands their applications as well as reducing the thermal budget in mass production. Amorphous GaN films grown at low temperatures can be crystallized through temperature-programmed rapid annealing at minimal cost [32]. Pansila et al. reported the use of plasma-excited NH3 to deposit GaN films with TMG and NH3 precursors at near RT in 2015. The process temperature of 115 °C was in the nitridation step. Adsorption and nitridation models at that process temperature were proposed [32]. Jaclyn et al. reported electron-enhanced ALD grown GaN film using the same precursors in 2016. Crystalline GaN thin film can be grown on Si substrate at RT using electron enhancement [30]. ALD techniques using plasma [33,34,35,36] or a hot filament [37] were developed to fabricate the GaN film at low temperature [11]. Although the processing temperature could reduce to near RT, the quality of the film needs to be considered for qualities such as conformality, which is a very complicated area dependent on the recombination tendency. A new precursor is key to developing a process. The reaction mechanism in the ALD process with any precursor must be understood. A Ga source of GaCl3 had several drawbacks, such as low vapor pressure, releasing of corrosive by-product (HCl), and residual Cl in the film [35]. These problems can be solved by utilizing TMG or TEG. In the current study, TMG was used as the Ga source because of its higher vapor pressure [30].
Quantum mechanical simulations serve as a powerful tool for gaining profound insights into the underlying mechanisms of the ALD process. A report of ALD simulations in 2022 has been produced [38,39,40]. For theoretical study, the growth processes were investigated using computational fluid dynamics and kinetic Monte Carlo simulation methods [41,42]. The surface reactions using TMG and NH3 were investigated based on ab initio simulations [43,44,45]. However, the effectiveness of these methods is limited by the ability to provide accurate kinetic data [46]. Density functional theory (DFT) is a well-known computational method that provides accurate kinetic information of ALD processes. Qi et al. [47] reported first-principle calculations of GaN (0001) growth using TMG and NH3 versus TEG and N2H4 in 2015. DFT calculation in the growth of GaN has been widely reported [48,49,50,51,52,53,54,55,56]. The sequence of chemical bond breakage during the adsorption is the key to growing the ALD film at RT.
This paper identifies the initial surface reaction of semiconductor GaN semiconductor film formation in the ALD process using TMG and NH3. NH3 has the potential to serve as the nitrogen source in the GaN deposition process [57]. The effects of Si–NH2 and Si2–NH substrates on the initial surface adsorption were the focus of this study. The optimization of chemical structures and energy profiles was reported, using Si clusters as the substrates, such as Si15H18–(NH2)2 and Si15H20–(NH)2, respectively.

2. Materials and Methods

The initial surface reaction of GaN film grown using the ALD process was investigated based on DFT. TMG molecules and NH3 were used as gallium and nitrogen sources, respectively, for the adsorption and nitridation steps [57]. The Si15H18–(NH2)2 and Si15H20–(NH)2 clusters were used as representatives of 2–dimers of Si(100) substrate, respectively. This surface model consisted of four silicon layers, which were used to study the metal oxide and nitride films in the ALD process [58,59,60,61]. The Gaussview 5.0 and Gaussian 09 program packages [62] were used for visualization and calculation, respectively, of the characteristics of surface reactivity. The molecular structures along with the ALD steps were optimized by finding stationary points. The transition states (TS) were optimized using the Berny method [63,64]. All molecular structures were optimized without any constraints applied [63,64]. The nature of the optimized molecular structures was confirmed by the frequency calculation. The transition state and minima structures were contained with one and zero orders of imaginary frequency, respectively. The frequencies in the infrared region of the product structures were predicted and scaled by 0.96. Intrinsic reaction coordinate (IRC) calculations were performed to confirm the transition states that connected reactants to products. The molecular optimizations, frequency calculations, and IRC calculations were performed based on the DFT calculation at the DFT with the Becke, 3–parameter, Lee-Yang-Parr hybrid functional (B3LYP) [65,66,67,68] using a basis of 6–31G(d) level [63,64,65,66,67,68]. The reaction mechanisms of GaN grown in the ALD process were evaluated based on their energy profiles. The effects of surface cluster sizes and the number of NH2 adsorption sites were investigated using the 6-31G(d) basis set and systematically compared in Section 2.1 and Section 2.2, respectively. The influence of basis set selection was examined in Section 2.3, where 6-311+G(2d,p) was identified as the most reliable. Accordingly, it was used for computing relative energies of the optimized structures along all reaction pathways presented in Section 3.1, Section 3.2, Section 3.3, Section 3.4. Zero-point energy collection was not included in this study. The energies reported in the energy pathways were referenced to the reactant level (sum of the individual ground-state energies of all the reactants before the reaction took place) and were presented in electron volt (eV). All energy levels were reported to two decimal places for accuracy and for comparison with future studies. The charge and multiplicity of the system were set at zero and a singlet, respectively. Only atoms in the reaction mechanism were labeled to clarify the adsorption atoms on the surface.

2.1. Effects of Surface Cluster Sizes

To evaluate the effect of cluster sizes, the Si9H12–(NH2)2 and Si15H18–(NH2)2 surface clusters were used as one dimer and two dimers, respectively, of NH2 passivated Si(100) substrates. The reaction pathways of TMG adsorbed on one dimer and two dimers are compared in Figure 1. The product 2 structures show that TMG is capable of being adsorbed within a dimer or between two neighboring dimers of the Si(100) surface by consuming two adsorption sites. The reactions proceeded according to the same trend. However, there was a noticeable difference in the first transition state, leading to the differentiation of activation barriers. The barrier at TS 1 of the one-dimer pathway was significantly less than for the two-dimers pathway (0.32 eV). In ALD simulations, the substrate model should yield consistent results with increasing system size. Thus, a one-dimer Si(100) model is inadequate for this study.
As simulating the full Si(100) substrate is impractical, an optimal balance between model reliability and computational cost must be achieved. Accordingly, the most appropriate model is the smallest one that yields stable relative energy. To identify a suitable surface model, we simulated the adsorption of TMG on the two-, three-, and four-dimers of the Si(100) surface. They are compared in Figure 2. It was found that the reactions proceeded with the same trend. A significant differentiation in activation barriers was only observed for the one-dimer of the silicon substrate that might have been caused by steric effects. The smaller surface model allowed TMG molecules to approach closer to the surface than in the larger model during the adsorption stage. This led to an unrealistic small surface model (one-dimer). Therefore, the Si15H18–(NH2)2 surface cluster (two-dimers) was used as the substrate in this work.

2.2. Effects of NH2 Adsorption Site Number

In the adsorption step, the neighboring adsorption sites might affect the reaction. The reaction pathways of TMG adsorbed Si15H18–(NH2)2 and Si15H18–(NH2)4 substrates are compared in Figure 3 to evaluate the effect of neighboring adsorption sites during the ALD step. It was found that the number of adsorption sites did not affect the reaction tendency; however, the relative energy in the adsorption stage was affected, which resulted in process reactivity. In addition, the difference in the relative energy at the adsorption stage caused a slight difference in the activation barrier (0.13 eV). This showed that the interaction with neighboring adsorption sites slowed down the adsorption of TMG. In this work, Si15H18–(NH2)2 substrate was used as a silicon substrate to diminish the interactions of neighboring adsorption sites. Additionally, this surface model had two adsorption sites, which was similar to the cluster Si15H20–(NH)2 used as the Si2–NH cluster substrate. Therefore, it should provide a better comparison.

2.3. Effects of Basis Sets

To evaluate the effects of basis sets on the energy profile, we calculated the reaction pathway of TMG adsorption on the Si15H18–(NH2)2 surface (to diminish the interactions of neighboring adsorption sites) with various basis sets, as shown in Figure 4. The 6-31G(d,p), 6-311G(d,p), 6-311+G(2d,p) and 6-311++G(2df,2p) basis sets were ranked from small to large and were represented by blue, orange, yellow, and gray lines, respectively. It was found that all the reaction pathways showed the same trend. The differentiation in relative energies was reduced with the larger basis sets. The pathway was stable at the 6-311+G(2d,p) level of basis set (yellow line) that was very close to the larger basis of the 6-311++G(2df,2p) level (gray line). The activation barriers calculated from the two basis sets differed only by 0.01 eV, but the 6-311+G(2d,p) basis set took less time to calculate. Therefore, the 6-311+G(2d,p) basis set was used to perform the energy calculations in this work. In the current work, (Section 3.1, Section 3.2, Section 3.3, Section 3.4), the molecular geometries were optimized at the B3LYP 6-31G(d) level, while the relative energies were calculated at the B3LYP 6-311G+(2d, p) level.

3. Results

3.1. Reaction Mechanisms of TMG Adsorption on NH2 Passivated Si(100) Surface

The reaction mechanisms of TMG adsorption on the Si–NH2 and Si2–NH surfaces during the ALD steps are explained in Figure 5. The corresponding molecular structures are labeled (a)–(g) for adsorption on the Si–NH2 surface and (h)–(n) for adsorption on the Si2–NH surface, as shown at the bottom of the figure. The TMG adsorbed the silicon surface by consuming two adsorption sites and releasing two by-product molecules (CH4). The GaN2 species was located between two neighboring silicon dimers. The reaction was initiated by the adsorption step when the TMG molecule approached the Si–NH2 surface. Then, at the first transition state (TS 1–1), the Ga–C1 bond was broken at a distance of 2.73 Å while the Ga–N1 bond formed. The CH3 released in this step attacked one of the H atoms that was released from the NH2 adsorption site to form a CH4 by-product molecule in the next step (product 1). In the PC 1–1 + CH4 stage, a CH4 molecule was removed, while the Ga–N1 bond formed with a length of 1.92 Å. There was an interaction between a Ga atom and an N2 atom of the NH2 adsorption site, leading to the second transition state (TS 2–1). At TS 2–1, the Ga–C2 bond was broken at a distance of 2.25 Å while the Ga–N2 bond formed. In the final step (PC 2–1 + CH4), the Ga–N2 species formed completely while the second CH4 by-product was removed. The N2Ga–CH3 species was the new surface. In addition, we observed that the Ga–C1 bond was broken while the Ga–N1 bond formed during the TS 2–1 stage; however, the N–H bond of the Si–NH2 surface site still remained. This showed that the Ga–C bond was broken together with the formation of the Ga–N1 bond, followed by the rupture of the N–H bond in the first adsorption. Knowing the order of breaking and forming in chemical bonds during the adsorption step is the key factor in ALD processes. We considered that the energy of precursor decomposition defined the process temperature of ALD. In the current work, the H atom of the Si–NH2 surface did not contribute to dissociation of the Ga–C bond (within a TMG precursor) in the transition states (TS 1–1 and Ts 2–1). Therefore, large activation barriers (2.00 eV) are needed to break the Ga–C bond directly, resulting in the relative energy for this stage being 1.65 eV above the reactant level. Therefore, the ALD–GaN thin film TMG precursor must be deposited at a high temperature (above 375 °C) [47]. These computational results were comparable to previous experimental reports. For example, in 2015, Pansila et al. [32] reported the mechanisms of TMG adsorbed on a GaN surface in the ALD process. The reaction mechanisms were characterized using infrared (IR) spectroscopy. In the adsorption step, TMG was adsorbed on the Ga–NH2 surface site by consuming one– or two–surface adsorption site(s). The adsorption on two–adsorption sites is considered more favorable. CH3 and H2 molecules were released as by-products. These results were consistent with Jaclyn’s work reported in 2016 [30]. However, in Jaclyn’s study, the CH4 molecule was released as a by-product, similar to the current work, where TMG was adsorbed on the one- and two-adsorption sites of the Si–NH2 surface species, respectively. Then, the Ga–CH3 species was a new surface species, while CH4 molecules were released as by-products. In the current work, we considered that the CH3 groups were released during transition states and were not always combined with an H atom to form the CH4 molecule because the H atom did not contribute to breaking the Ga–C bond during the TS stages. This could be seen from the remaining NH bonds during the TS 1–1 and TS 2–1 stages. Therefore, the CH3 and CH4 molecules could both possibly be released as reaction by-products. This agreed with the simulation of TMG adsorbed on the GaN surface in the ALD process as reported by Qi et al. in 2015 [47]. Qi proposed that the H atom and the CH3 and CH4 molecules were possibly released as by-product molecules. Regarding the geometry characteristics, the Ga–C and N–H bond lengths of the TMG (gas phase) and NH2 surface sites in the reactant step were 1.99 and 1.01 Å, respectively. Before the adsorption, the H–N–H angles were about 108° and 110°. The chemical geometries changed during the ALD steps. The Ga–N bonds formed at PC 1 and PC 2 were 1.92 and 1.86 Å, respectively, which was in good agreement with Qi et al. [47], who reported N–H bond lengths and an H–N–H angle of the NH2 surface site of 1.02 Å and 109°, respectively. After the adsorption, Qi reported that a Ga–N bond formed with a length of 2.13 Å. The results reported in the current work were comparable to these previous reports; however, we found that the Ga–N bond would be shortened when the TMG is adsorbed by consuming two-surface adsorption sites.

3.2. Effects of Surface Adsorption Site (NH and NH2 Adsorption Sites)

In the adsorption of TMG on a silicon surface, another possible surface is the Si2–NH species (nitrogen bridge). In 2015, Pansila et al. [57] reported the adsorption of NH3 and plasma NH3 on a cleaned Si(100) surface. They found that the adsorption of NH3 molecules and NH3 plasma on the Si surface was able to generate Si–NH2 and Si2–NH surface sites at RT. Si–NH2 and Si2–NH surface species co-existed at 25–300 °C, and the ratio of these species varied with the temperature. To investigate the effects of Si–NH2 and Si2–NH surfaces on the TMG adsorption during an ALD process, we also simulated the adsorption of TMG on the Si2–NH surface for comparison. The reaction mechanism of TMG adsorption on the Si2–NH surface in the ALD step is shown in Figure 5. Similar to the Si–NH2 adsorption sites, TMG was adsorbed on the surface by forming Ga–N bonds and breaking Ga–C bonds and then releasing a CH3 group and an H atom to form a CH4 molecule (by-product), respectively. TMG was consumed at the first and second NH surface sites, respectively. The chemical geometries were changed during the adsorption step and compared to the adsorption on the Si–NH2 surface; there were differences in the adsorption characteristics. The differentiation in the adsorption distances, in the breaking distances of Ga–C and N–H bonds, and in forming the distances of the Ga–N bonds during the TMG adsorption steps resulted in differences in the energy profiles.

3.2.1. TMG Adsorption on Si–NH2 Surface

The reaction pathway of TMG adsorption on NH2 passivated Si(100) substrate is shown in Figure 5 (black line). It is a strong exothermic reaction that released 1.33 eV to form the product structure (product 2). The TMG absorbs the silicon surface by consuming two adsorption sites and releasing two by–product molecules. This must be carried out through two successive activation barriers. In the transition states, TS 1–1 and TS 2–1 required the activation barriers of 2.00 and 1.44 eV, respectively. This showed that reactivity was low at TS 1 and then increased at TS 2. When considered at TS 1–1 and TS 2–1 stages, we found that the bonds of Ga–C1 and Ga–C2 were broken with different activation barriers. This allows us to know that the strength of Ga–C1 and Ga–C2 covalent bonds, which are the same type (Ga–C bond), might be differenced or changed during the adsorption. This information is difficult to obtain from experimental characterization techniques such as IR spectroscopy but is obtainable using simulation. This agreed with the first-principle calculations of TMG adsorption on GaN surface reported by Qi in 2015 [47]. The Ga–C bonds within a TMG molecule were reported to be dissociated with the activation barriers of 1.28 and 1.84 eV, respectively. However, the cause of this phenomenon has never been discussed in the publications but is considered in the current work. The difference between TS 1 and TS 2 may be explained by the CH3 binding capacity of the Ga atom being reduced upon adsorption at the Si–NH2 surface site. This can be seen from the breaking distances of the shortened Ga–C1 and Ga–C2 bonds during the TS 1 and TS 2 states, respectively (Figure 5). The breaking distance was substantially reduced from 2.73 Å (TS1) to 2.25 Å (TS2). The hypothesis for this phenomenon is that the N atom has greater electronegativity than the C atom. When the Ga–C bond is replaced by a Ga–N bond, the Ga electrons transfer to the N region. Therefore, the C atom decreases the Ga electron attraction used to form the Ga–C covalent bond. As a result, the remaining Ga–C bonds in TS 2–1 are weakened. In addition, the structure of PC 2 is more stable than for PC 1, resulting in the reaction tending to continue as PC 2 (consume two-adsorption sites) rather than remaining as PC 1.

3.2.2. TMG Adsorption on Si2–NH Surface

The reaction pathway of TMG adsorption on Si2–NH passivated Si(100) substrate is shown in Figure 5 (red line). In the adsorption stage, the TMG molecule appeared to be relatively inert to the NH surface. Only 0.02 eV were released in this stage. The molecule of TMG was closer to the NH2 surface (2.03 Å) than to the NH surface (2.24 Å) during the adsorption. In the transition stage, TMG was adsorbed on the NH surface with a substantially lower activation barrier (1.11 eV). This may explain why the CH3 binding capacity of the Ga atom was reduced when TMG was adsorbed on the Si2–NH surface. This can be seen from the shortening of the breaking distance of the Ga–C1 bond from the TS stage. The breaking distances of the Ga–C1 bond at TS stages were 2.24 and 2.73 Å when TMG was adsorbed on Si2–NH and Si–NH2 surfaces, respectively. In addition, we found that the Ga–N bond formed in the TS stage of the Si2–NH adsorption site (2.09 Å) was slightly shorter than for the Si–NH2 adsorption site (2.13 Å), indicating that TMG was attracted by the Si2–NH adsorption site with a higher reactive force than for attraction to the Si–NH2 adsorption site. We assumed that a nitrogen (N1) atom at the Si2–NH surface has a greater electron density than at the Si–NH2 surface because the N1 atom at the Si2–NH surface site is connected to the 2–Si atoms and 1–H atom, while the N1 atom at the Si–NH2 surface site is connected to the 1–Si atom and the 2–H atoms. For the Si2–NH surface, the Si atom has less electronegativity than the H atom, so most electrons spend more time in the N region, resulting in the stronger and easier formation of the Ga–N1 bond. When a Ga–N1 bond is formed, electrons are transferred from the Ga atom to the N atom, resulting in decreased ability of the Ga atom to hold CH3 groups. As a result, TMG adsorption on the Si2–NH surface is easier than on the Si–NH2 surface. In the first adsorption, TMG adsorption on the Si2–NH surface site tends to be more facile and can occur at a lower temperature due to its lower relative energy and smaller activation barrier in the TS 1–2 stage. However, in the second adsorption, the TS 2–2 stage has a higher relative energy and activation barrier compared with the TS 2–1 stage (black line). In addition, the relative energy at the TS 2–2 stage is higher than at TS 1–2. Thus, the reaction could not proceed through the TS 2–2 stage at the same temperature as for TS 1–2. Furthermore, the relative energy in the PC 2–2 stage is higher than the relative energy in the PC 1–2. This is an endothermic reaction (compared with PC 1–2 structure formation) that is a reversible reaction at higher temperature. Therefore, TMG adsorption on Si2–NH surface tends to stop at the PC 1–2 structure by consuming a one-surface adsorption site rather than continuing to a PC 2–2 structure. This work was elucidated on the conditions of TMG adsorption, whether it is adsorbed on surfaces by consuming one surface adsorption site or two. In the case that the reaction stops at the PC 1–2 structure, the reaction mechanism of NH3 adsorption on this structure (nitridation step) is discussed in Section 3.3.
In 2015, Pansila reported the use of NH3 and NH3 plasma to generate Si–NH2 and Si2–NH adsorption sites on cleaned Si(100) substrates [32]. After the exposure of NH3 or NH3 plasma, Si–NH2 and Si2–NH adsorption sites coexisted on the surface at RT. The ratio of Si2–NH to Si–NH2 was substantially increased with NH3 plasma exposure compared to non-plasma NH3 exposure. We considered that the use of NH3 plasma to create a Si2–NH surface was highly beneficial for the deposition of GaN thin films in low-temperature ALD processes. In addition, we found that the reaction mechanisms of TMG adsorbed on Si–NH2 surfaces reported in the current work were similar to the mechanisms of TMG adsorbed on the GaN surface. In 2015, Qi et al. [47] reported that the TMG molecule is physically adsorbed on the top of an NH2 molecule with a heat of reaction of −0.24 eV. The distance between the TMG and Ga and the N surface in NH2 is 2.27 Å. Thus, a CH3 group reacts with an H from the NH2 surface site to release CH4 as a by-product. The first and second Ga–C bonds were dissociated with activation barriers of 1.28 and 1.84 eV, respectively. These mechanisms were consistent with the current work. However, there was a differentiation in the energy profile. In Figure 5 (black line), the first and second Ga–C bonds were dissociated with activation barriers of 2.00 and 1.44 eV, respectively, showing that the adsorption of TMG on Si–NH2 and Ga–NH2 substrates requires different process conditions. In the ALD process, the film thickness is controlled by varying ALD cycles. Multilayers of thin films were deposited on substrate by multicycles of ALD reaction. For the initial layer (first layer), TMG is adsorbed on the silicon surface. However, for the second and subsequent layers (second–nth layer), TMG is adsorbed on the GaN thin film layer. The initial layer is very important in determining the desired thin film quality. The current work hypothesized that the initial layer requires different process conditions to the following layers.

3.3. NH3 Adsorption on PC 1–2 Structure

When TMG is adsorbed on the Si2–NH surface by consuming a one-adsorption site, the product of the reaction would be the PC 1–2 + CH4 structure (Figure 5). In this case, we simulated the adsorption of NH3 on this structure. The reaction mechanism of the nitridation step is described in Figure 6, where the PC 1–2 + CH4 structure is used as the substrate. The reaction starts from the adsorption 1–3 stage (a), where the first NH3 molecule approaches the Ga atom at a distance of 2.13 Å. Then, in the transition 1–3 stage (b), the Ga and N3 atom distances are reduced to 1.99 Å, and the Ga–C2 bond is broken at 2.29 Å. The Ga–N3 bond is formed with a length of 1.83 Å in the PC 1–3 stage (c), while a CH4 molecule is released as a reaction by-product in the PC 1–3 + CH4 stage (d). Then in the adsorption 1–4 stage (e), the second NH3 molecule approaches the Ga atom at a distance of 2.11 Å. In the transition 1–4 stage (f), the Ga and N4 atom distances are reduced to 1.98 Å, and the Ga–C3 bond is broken at 2.27 Å. The Ga–N4 bond is formed with a length of 1.86 Å in the PC 1–4 stage (g), while a CH4 molecule is released as a reaction by-product in the PC 1–4 + CH4 stage (h). The N4–Ga–N3 species is formed in the PC 1–4 + CH4 stage as a final product, while the second CH4 by-product is released. The energy profile for this reaction is shown in Figure 6. The two NH3 molecules were adsorbed on the surface according to similar mechanisms. Therefore, they have similar reaction pathways. Energies of 0.71 and 0.74 eV were released when the first and second NH3 molecules approached the surface in the adsorption 1–3 and 1–4 stages, respectively. Activation barriers of 1.88 and 1.92 eV are needed to overcome the transition stages, TS 1–3 and TS 1–4, respectively. NH3 adsorptions were exothermic, where the final product (PC 1–4) structure was stably formed at 0.94 eV below the reactant level. The NH3 molecules were adsorbed on the surface with activation barriers slightly less than the barrier for TMG adsorption on the Si–NH2 surface (2.00 eV), indicating that the nitridation processes were preceded by a more facile reaction. In addition, the relative energies at transition stages (TS 1–3 and TS 1–4) were substantially less than the relative energy at TS 1–2 (1.65 eV), indicating that the nitridation process can occur at a lower process temperature. This work suggested that if TMG were adsorbed on the Si2–NH surface by consuming a one-adsorption site, the residual Ga–C bonds would be eliminated in the nitridation steps, which have lower energy barriers and lower process temperatures compared to the TMG adsorbed on the Si–NH2 surface.
For other cases, Figure 5 shows the reaction pathways leading to possible four-product structures that can be generated by the adsorption of TMG on Si2–NH and Si–NH2 passivated Si(100) surfaces. The relative energies of PC 1–1, PC 2–1, PC 1–2, and PC 2–2 structures were 0.95, 1.33, 0.73, and 0.51 eV, respectively, below the reactant level, indicating that all the product structures could coexist on the initial surface. This was similar to the adsorption of TMG on the GaN surface reported by Pansila et al. in 2015 [32], with possible TMG adsorption models for low-temperature ALD proposed in their work. The Ga2(CH3) and Ga(CH3)2 species became new surface sites when TMG was adsorbed on the surface by consuming one- and two-adsorption site(s), respectively. The current work suggested that the ratio of each product structure involved in surface composition depends on the reaction barrier used to form them. Furthermore, the existence of Si2–NH species would support the adsorption of TMG at a low temperature as well as enhancing process reactivity due to the substantially lower activation barrier. The reaction mechanisms of NH3 adsorption on PC 2–1 and PC 2–2 structures (nitridation) are discussed in Section 3.4.

3.4. Reaction Mechanisms of NH3 Adsorption on TMG Adsorbed Si(100) Surface

After the TMG adsorption step, we assumed that TMG was adsorbed on the Si2–NH and Si–NH2 surfaces by consuming two-adsorption sites. Therefore, for the final products of the first-half ALD reaction, PC 2–1 and PC 2–2 structures were used as substrates in the nitridation step. In this section, we simulated the nitridation step of GaN thin film grown using TMG and NH3 precursors in the ALD process. The reaction mechanisms of NH3 adsorption on the residual TMG adsorbed silicon surfaces (2–1 and PC 2–2 substrates) are shown in Figure 7. The reactions started from the adsorption stages. In the adsorption 3–1 stage, the NH3 molecule approached the N2–GaCH3 surface. The distance between N3 and Ga was 2.15 Å. In the adsorption 3–2 stage, the distance between N3 and Ga was 2.11 Å, which was shorter than for adsorption on the PC 2–1 substrate, suggesting that the NH3 molecule could approach the PC 2–2 substrate closer than the adsorption on the PC 2–1 substrate.
During the adsorption stages, NH3 was found on the top of the GaN surface, resulting in N–H bond lengths (in NH3) of approximately 1.01 Å for both substrates. The H–N–H angles (in NH3) were 108°, 107°, and 107° for the PC 2–1 substrate and 107°, 108°, and 108° for the PC 2–2 substrate. These adsorption characteristics were comparable to other work. In 2015, Qi et al. [47] reported the adsorption of NH3 on a GaN surface. The distance between the N in the NH3 and Ga at the surface site was 2.10 Å. During the adsorption stage, NH3 was found on the top site of the GaN surface, resulting in N–H bond lengths of 1.03 Å and H–N–H angles of 108.6, 108.7, and 111.2 Å. In the current work, the Ga–C1 bonds were broken in the TS 3–1 and TS 3–2 stages at distances of 2.28 and 2.23 Å, respectively. Differences in the adsorption mechanisms of the two substrates were observed in the transition state. The N–H bond was broken in the TS 3–2 stage but remained in the TS 3–1 stage. The H atom released during the TS 3–2 stage attracted the CH3 group and combined to generate CH4 as a by-product. Next, Ga–N bonds were formed with lengths of 1.84 and 1.95 Å in the PC 3–1 and PC 3–2 structures, respectively. The crystal Ga–N bond has been reported to be 1.992 Å [47]. In the final step, the CH4 molecules were eliminated, while the N2Ga–NH2 species were left as new surface adsorption sites. The NH3 molecule not only eliminated the CH3 group but also regenerated the NH2 surface site to accommodate the new TMG adsorption in the next step of ALD.
The energy pathways of NH3 adsorption on the residual TMG adsorbed silicon surfaces (PC 2–1 and PC 2–2) are shown in Figure 7 by black and red lines, respectively. In the adsorption stages (adsorption 3), 0.58 and 0.94 eV were released when NH3 approached the PC 2–1 and PC 2–2 substrates, respectively, indicating that NH3 was more active with the PC 2–2 substrate than the other substrate. Compared to adsorption on the PC 2–1 substrate, the reaction pathway of NH3 adsorbed on the PC 2–2 substrate was more thermodynamically and kinetically favorable. In the transition state, the activation barriers in the TS 3–1 and TS 3–2 stages were approximately 2.00 ± 0.02 eV. The relative energy levels of the TS 3–1 and TS 3–2 stages were 1.38 and 1.06 eV, respectively, above the reactant level. The reaction pathways show that NH3 adsorbs onto both types of surfaces via similar reaction mechanisms, resulting in comparable nitridation kinetics.
In the ALD precursor such as a TMG molecule, the Ga atom is the core of the molecule connected to the methyl groups (the basis of the molecule). In the adsorption step, only the Ga atoms are required, while the methyl groups must be eliminated. In developing a new precursor for low temperature processes, the ease of decomposition of these methyl groups should be considered. Furthermore, the decomposition of the molecule must not contaminate the film. In addition, the residual precursor (the remaining species) should be highly reactive with the NH3 molecules when they adsorb on the surface. Using a co-reactant to help in the dissociation of these basis groups should be investigated.
The activation energies for ALD–GaN thin-film growth under various conditions in this study are summarized in Table 1. This table compiles the activation energies of TMG and NH3 molecules adsorbed on different surfaces during the ALD process, along with comparative values from previous reports. The data show that when TMG molecules react with a surface, they decompose significantly more easily than in the gas phase, as reflected by the high activation energy of up to 3.27 eV for gas-phase decomposition. Additionally, the adsorption of TMG and NH3 on different surfaces leads to variations in activation energy, indicating differing requirements for process temperatures. Therefore, tuning the surface adsorption sites of the substrate directly influences both the reaction kinetics and the process temperature. From the table, it is evident that preparing the surface in the Si2–NH configuration reduces the activation energy from 2.00 eV to 1.11 eV (in the case where TMG is adsorbed by consuming a single surface site). This configuration also improves the reactivity during subsequent NH3 adsorption. Interestingly, this approach to surface preparation could potentially be extended to other nitride thin-film ALD processes as well.

3.5. Prediction of Infrared Frequencies of Ga–N and Ga–CH3 Species of Product Structures

The current work proposed reaction pathways leading to possible four-product structures that formed in the first half of ALD (TMG adsorption) and possible four-product structures that formed in the second half of ALD (NH3 nitridation). The vibrational characteristics of these structures were calculated and compared to other experimental results [11,32]. Their predicted infrared frequencies are listed in Table 2. According to the reaction mechanisms of TMG adsorption on the Si–NH2 and Si2–NH surfaces in the adsorption step, TMG is adsorbed on the surfaces by forming Ga–N bonds. The IR frequencies of Ga–N stretching bonds that formed in PC 1–1, PC 2–1, and PC 1–2 stages were at wave numbers of 531, 530, and 537 cm−1, respectively. This was in good agreement with the work of Banerjee that reported an IR peak of the Ga–N stretching bond at a wave number of 534 cm−1 [11]. The Ga–CH3 stretching peaks formed in the PC 1–1, PC 2–1, and PC 1–2 stages were at wave numbers of 2917 cm−1. This was in good agreement with the work of Pansila that reported an IR peak of the Ga–CH3 stretching bond at a wave number of 2912 cm–1 [32]. Additionally, Ga–N peaks were observed to shift slightly as TMG was adsorbed on different surface locations (SiN–H2 or Si2–NH), and the Ga–N peaks also slightly shifted when the adsorption occupied one or two surface adsorption site(s). In the reaction step (NH3 nitridation), NH3 adsorbs on the surfaces by forming a Ga–N bond. The IR frequencies of the Ga–N stretching bonds that formed in the PC 3–1 and PC 3–2 stages were at wave numbers of 547 and 539 cm−1 , respectively. Similar to the TMG adsorption step, the Ga–N peaks were observed to shift slightly as the Ga–N bonds formed on different substrates (PC 2–1 or PC 2–2).

4. Conclusions

Density functional theory (DFT) calculations were used to investigate the growth mechanism of GaN thin films via the ALD process, using trimethylgallium (TMG) and NH3 as the gallium and nitrogen precursors, respectively. In the TMG adsorption step, TMG binds to both Si–NH2 and Si2–NH surfaces by occupying one or two surface site(s). On the Si2–NH surface, the average activation barrier (1.59 eV) is significantly lower than that on the Si–NH2 surface (1.72 eV) when TMG is adsorbed by occupying two surface sites. Moreover, the activation barrier further decreases to 1.11 eV when TMG is adsorbed by occupying only one surface site. This mechanism also improves the reactivity during subsequent NH3 adsorption. The results clearly indicate that tuning the surface adsorption sites of the substrate directly influences both the reaction kinetics and the process temperature. The present study suggests that increasing the density of Si2–NH surface sites accelerates the initial ALD reactions, enabling lower deposition temperatures, shorter cycle times, and improved GaN film quality.

Author Contributions

Conceptualization, P.P.P., T.P. and S.S. (Seckson Sukhasena); methodology, T.P. and P.P.P.; validation, T.P., S.S. (Saksit Sukprasong), W.S., S.S. (Seckson Sukhasena), W.T. and T.J.; formal analysis, P.P.P., T.P. and S.S. (Seckson Sukhasena); investigation, P.P.P. and T.P.; resources, P.P.P. and T.J.; data curation, T.P. and P.P.P.; writing—original draft preparation, T.P. and P.P.P.; writing—review and editing, P.P.P., S.S. (Saksit Sukprasong), S.S. (Seckson Sukhasena), W.S., T.J. and W.T.; visualization, T.P. and P.P.P.; supervision, P.P.P.; project administration, P.P.P. and T.P.; funding acquisition, P.P.P. and T.P. All authors have read and agreed to the published version of the manuscript.

Funding

Thailand Research Fund (TRF), grant number MRG6180194.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

The original contributions presented in this study are included in the article. Further inquiries can be directed to the corresponding author.

Acknowledgments

This work was supported by the Thailand Research Fund (TRF), currently integrated under the National Research Council of Thailand (NRCT), under project number MRG6180194. We gratefully acknowledge the Faculty of Science at Sriracha, Kasetsart University, for the partial financial support of the unit research project in 2022. The authors would also like to thank the anonymous referees for their helpful comments on this article.

Conflicts of Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this research paper.

References

  1. Alfredo Mameli Andrew, V. Teplyakov, Selection Criteria for Small-Molecule Inhibitors in Area-Selective Atomic Layer Deposition: Fundamental Surface Chemistry Considerations. Acc. Chem. Res. 2023, 56, 2084–2095. [Google Scholar]
  2. Oviroha, P.O.; Akbarzadeha, R.; Panb, D.; Coetzeea, R.A.M.; Jen, T.C. New development of atomic layer deposition: Processes, methods and applications. Sci. Technol. Adv. Mater. 2019, 20, 465–496. [Google Scholar] [CrossRef]
  3. George, S.M. Atomic layer deposition: An overview. Chem. Rev. 2010, 110, 111–131. [Google Scholar] [CrossRef] [PubMed]
  4. Lencer, D.; Salinga, M.; Grabowski, B.; Hickel, T.; Neugebauer, J.; Wuttig, M. A map for phase-change materials. Nat. Mater. 2008, 7, 972–977. [Google Scholar] [CrossRef]
  5. Mayer, T.M.; Elam, J.W.; George, S.M.; Kotula, P.G.; Goeke, R.S. Atomic-layer deposition of wear-resistant coatings for microelectromechanical devices. Appl. Phys. Lett. 2003, 82, 2883. [Google Scholar] [CrossRef]
  6. Ngo, C.Y.; Yoon, S.F.; Loke, W.K.; Cao, Q.; Lim, D.R.; Wong, V.; Sim, Y.K.; Chua, S.J. Investigation of semiconductor quantum dots for waveguide electroabsorption modulator. Nanoscale Res. Lett. 2008, 3, 486. [Google Scholar] [CrossRef]
  7. Notten, P.H.L.; Roozeboom, F.; Niessen, R.; Baggetto, L. 3-D integrated all-solid-state rechargeable batteries. Adv. Mater. 2007, 19, 4564–4567. [Google Scholar] [CrossRef]
  8. Lee, F.; Marcus, S.; Shero, E.; Wilk, G.; Swerts, J.; Maes, J.W.; Blomberg, T. Atomic layer deposition: An enabling technology for microelectronic device manufacturing. In Proceedings of the 2007 IEEE/SEMI Advanced Semiconductor Manufacturing Conference, Stresa, Italy, 11–12 June 2007; pp. 359–365. [Google Scholar]
  9. Wang, J.J.; Deng, X.; Varghese, R.; Nikolov, A.; Sciortino, P.; Liu, F.; Liu, X.J. Filling high aspect-ratio nano-structures by atomic layer deposition and its applications in nano-optic devices and integrations. Vac. Sci. Technol. B 2005, 23, 3209–3213. [Google Scholar] [CrossRef]
  10. Hwang, C.H.; Yoo, C.Y. Introduction. Atomic Layer Deposition for Semiconductors; Springer: New York, NY, USA, 2014; pp. 3–12. [Google Scholar]
  11. Banerjee, S.; Aarnink, A.A.I.; Gravesteijn, D.J.; Kovalgin, A.Y. Thermal atomic layer deposition of polycrystalline gallium nitride. J. Phys. Chem. C 2019, 123, 23214–23225. [Google Scholar] [CrossRef]
  12. Maruska, H.P.; Stevenson, D.A.; Pankove, J.I. Violet luminescence of Mg-doped GaN. Appl. Phys. Lett. 1973, 22, 303–305. [Google Scholar] [CrossRef]
  13. Nakamura, S.; Senoh, M.; Senoh, M.; Mukai Takashi Mukai, T. P-GaN/N-InGaN/N-GaN Double-Heterostructure Blue-Light-Emitting Diodes. J. Appl. Phys. 1993, 32, L8. [Google Scholar] [CrossRef]
  14. Ambacher, O.; Smart, J.; Shealy, J.R.; Weimann, N.G.; Chu, K.; Murphy, M.; Schaff, W.J.; Eastman, L.F.; Dimitrov, R.; Wittmer, L.; et al. Two-dimensional electron gases induced by spontaneous and piezoelectric polarization charges in N- and Ga-face AlGaN/GaN heterostructures. J. Appl. Phys. 1999, 85, 3222–3233. [Google Scholar] [CrossRef]
  15. Trivedi, M.; Shenai, K. Performance evaluation of high-power wide band-gap semiconductor rectifiers. J. Appl. Phys. 1999, 85, 6889–6897. [Google Scholar] [CrossRef]
  16. Zhu, T.G.; Lambert, D.J.H.; Shelton, B.S.; Wong, M.M.; Chowdhury, U.; Kwon, H.K.; Dupuis, R.D. High-voltage GaN pin vertical rectifiers with 2 µm thick i-Layer. Electron. Lett. 2000, 36, 1971–1972. [Google Scholar] [CrossRef]
  17. Rivera, C.; Pau, J.L.; Pereiro, J.; Munoz, E. Properties of Schottky barrier photodiodes based on InGaN/GaN MQW structures. Superlattices Microstruct. 2004, 36, 849–857. [Google Scholar] [CrossRef]
  18. Bardsley, N.; Bland, S.; Chwastyk, D.; Monasterio, C.; Pattison, L.; Pattison, M.; Welsh, F.; Yamada, M. Manufacturing Roadmap Solid State Lighting Research and Development; U.S. Department of Energy: Washington, DC, USA, 2014. [Google Scholar]
  19. Li, G.; Wang, W.; Yang, W.; Lin, Y.; Wang, H.; Lin, Z.; Zhou, S. GaN-based light-emitting diodes on various substrates: A critical review. Rep. Prog. Phys. 2016, 79, 056501. [Google Scholar] [CrossRef]
  20. Zhang, Y.; Piedra, D.; Sun, M.; Hennig, J.; Dadgar, A.; Yu, L.; Palacios, T. High-Performance 500 V Quasi- and Fully-Vertical GaN-on-Si pn Diodes. IEEE Electron. Device Lett. 2017, 38, 248–251. [Google Scholar] [CrossRef]
  21. Zhu, D.; Wallis, D.J.; Humphreys, C.J. Prospects of III-nitride optoelectronics grown on Si. Rep. Prog. Phys. 2013, 76, 106501. [Google Scholar] [CrossRef]
  22. Morkoc, H.; Strite, S.; Gap, G.B.; Lin, M.E.; Sverdlov, B.; Burns, M. Large-band-gap SiC, III-V nitride, and II-VI ZnSe-based semiconductor device technologies. J. Appl. Phys. 1994, 76, 1363–1398. [Google Scholar] [CrossRef]
  23. Yoshida, S.; Misawa, S.; Gonda, S. Improvements on the electrical and luminescent properties of reactive molecular beam epitaxially grown GaN films by using AlN-coated sapphire substrates. Appl. Phys. Lett. 1983, 42, 427–429. [Google Scholar] [CrossRef]
  24. Ross, J.; Rubin, M.; Gustafson, T.K. Single crystal wurtzite GaNon (111) GaAs with AlNbuffer layers grown by reactive magnetron sputter deposition. J. Mater. Res. 1993, 8, 2613–2616. [Google Scholar] [CrossRef]
  25. Lamb, H.H.; Lai, K.K.; Torres, V.; Davis, R.F. Deposition of Gan Films Using Seeded Supersonic Jets. MRS Online Proc. Libr. 1995, 388, 265–270. [Google Scholar] [CrossRef]
  26. Ferguson, B.A.; Sellidj, A.; Doris, B.B.; Mullins, C.B. Supersonic-jet-assisted growth of GaN and GaAs films. J. Vac. Sci. Technol. A 1996, 14, 825–830. [Google Scholar] [CrossRef]
  27. Chtchekine, D.G.; Fu, L.P.; Gilliland, G.D.; Chen, Y.; Ralph, S.E.; Bajaj, K.K.; Bu, Y.; Lin, M.C.; Bacalzo, F.T.; Stock, S.R. Properties of low-pressure chemical vapor epitaxial GaN films grown using hydrazoic acid (HN3). J. Appl. Phys. 1997, 81, 2197–2207. [Google Scholar] [CrossRef]
  28. Neumayer, D.A.; Ekerdt, J.G.; Growth of group III nitrides. A review of precursors and techniques. Chem. Mater. 1996, 8, 9–25. [Google Scholar] [CrossRef]
  29. Molnar, R.J.; Moustakas, T.D. Growth of gallium nitride by electron-cyclotron resonance plasma-assisted molecular-beam epitaxy: The role of charged species. J. Appl. Phys. 1994, 76, 4587–4595. [Google Scholar] [CrossRef]
  30. Sprenger, J.K.; Cavanagh, A.S.; Sun, H.; Wahl, K.J.; Roshko, A.; George, S.M. Electron Enhanced Growth of Crystalline Gallium Nitride Thin Films at Room Temperature and 100 °C Using Sequential Surface Reactions. Chem. Mater. 2016, 28, 5282–5294. [Google Scholar] [CrossRef]
  31. Chun, J.; Hwang, Y.; Choi, Y.; Kim, J.-J.; Jeong, T.; Baek, J.H.; Ko, H.C.; Park, S. Laser lift-off transfer printing of patterned GaN light-emitting diodes from sapphire to flexible substrates using a Cr/Au laser blocking layer. Scr. Mater. 2014, 77, 13. [Google Scholar] [CrossRef]
  32. Pansila, P.; Kanomata, K.; Miura, M.; Ahmmad, B.; Kubota, S.; Hirose, F. Growth kinetics for temperature-controlled atomic layer deposition of GaN using trimethylgallium and remote-plasma-excited NH3. Appl. Surf. Sci. 2015, 357, 1920–1927. [Google Scholar] [CrossRef]
  33. Gupta, G.; Banerjee, S.; Dutta, S.; Aarnink, A.A.I.; Schmitz, J.; Kovalgin, A.Y.; Hueting, R.J.E. Charge carrier transport and electroluminescence in atomic layer deposited poly-GaN/c-Si heterojunction diodes. J. Appl. Phys. 2018, 124, 084503. [Google Scholar] [CrossRef]
  34. Alevli, M.; Haider, A.; Kizir, S.; Leghari, S.A.; Biyikli, N. Comparison of trimethylgallium and triethylgallium as “Ga” source materials for the growth of ultrathin GaN films on Si (100) substrates via hollow-cathode plasma-assisted atomic layer deposition. J. Vac. Sci. Technol. A 2016, 34, 01A137. [Google Scholar] [CrossRef]
  35. Ozgit-Akgun, C.; Goldenberg, E.; Okyay, A.K.; Biyikli, N. Hollow cathode plasma-assisted atomic layer deposition of crystalline AlN, GaN and AlxGa1−xN thin films at low temperatures. J. Mater. Chem. C 2014, 12, 2123–2136. [Google Scholar] [CrossRef]
  36. Motamedi, P.; Dalili, N.; Cadien, K. A route to low temperature growth of single crystal GaN on sapphire. J. Mater. Chem. C 2015, 3, 7428–7436. [Google Scholar] [CrossRef]
  37. Sumakeris, J.; Sitar, A.; Ailey-Trent, K.S.; More, K.L.; Davis, R.F. Layer-by-layer epitaxial growth of GaN at low temperatures. Thin Solid Film. 1993, 225, 244–249. [Google Scholar] [CrossRef]
  38. Shahmohammadi, M.; Mukherjee, R.; Sukotjo, C.; Diwekar, U.M.; Takoudis, C.G. Recent advances in theoretical development of thermal atomic layer deposition: A review. Nanomater 2022, 12, 831. [Google Scholar] [CrossRef]
  39. Sibanda, D.; Oyinbo, S.T.; Jen, T.C. A review of atomic layer deposition modelling and simulation methodologies: Density functional theory and molecular dynamics. Nanotechnol. Rev. 2022, 11, 1332–1363. [Google Scholar] [CrossRef]
  40. Mustard, T.J.L.; Kwak, H.S.; Goldberg, A.; Gavartin, J.; Morisato, T.; Yoshidome, D.; Halls, M.D. Quantum mechanical simulation for the analysis, optimization and accelerated development of precursors and processes for atomic layer deposition (ALD). J. Korean Ceram. Soc. 2016, 53, 317–324. [Google Scholar] [CrossRef]
  41. Elliott, S.D. Atomic-scale simulation of ALD chemistry. Sci. Technol. 2012, 27, 074008. [Google Scholar] [CrossRef]
  42. Sengupta, D.; Mazumder, S.; Kuykendall, W.; Lowry, S.A. Combined ab initio quantum chemistry computational fluid dynamics calculations for prediction of gallium nitride growth. J. Cryst. Growth 2005, 279, 369–382. [Google Scholar] [CrossRef]
  43. Fu, K.; Fu, Y.; Han, P.; Zhang, Y.; Zhang, R. Kinetic Monte Carlo study of metal organic chemical vapor deposition growth dynamics of GaNthin film at microscopic level. J. Appl. Phys. 2008, 103, 103524. [Google Scholar] [CrossRef]
  44. Cardelino, B.H.; Cardelino, C.A. Dissociative Chemisorption of Trimethylgallium, Trimethylindium, and Ammonia on Gallium and Indium Nitride Substrates. A Computational Study. J. Phys. Chem. C 2011, 115, 9090–9104. [Google Scholar] [CrossRef]
  45. Bermudez, V.M. Chemisorption of NH3 on GaN (0001)-(1 × 1). Chem. Phys. Lett. 2000, 317, 290–295. [Google Scholar] [CrossRef]
  46. Timoshkin, A.Y.; Bettinger, H.F.; Schaefer, H.F. DFT Modeling of Chemical Vapor Deposition of GaN from Organogallium Precursors. 1. Thermodynamics of Elimination Reactions. J. Phys. Chem. A 2001, 105, 3240–3248. [Google Scholar] [CrossRef]
  47. An, Q.; Jaramillo-Botero, A.; Liu, W.-G.; Goddard, W.A. Reaction pathways of GaN (0001) growth from trimethylgallium and ammonia versus triethylgallium and hydrazine using first principle calculations. J. Phys. Chem. C 2015, 119, 4095–4103. [Google Scholar] [CrossRef]
  48. Rapcewicz, K.; Nardelli, M.B.; Bernholz, J. Theory of surface morphology of wurtzite GaN (0001) surfaces. Phys. Rev. B 1997, 56, 12725–12728. [Google Scholar] [CrossRef]
  49. Fritsch, J.; Sankey, O.F.; Smith, K.E.; Page, J.B. Ab initio calculation of the stoichiometry and structure of the (0001) surfaces of GaN and AlN. Phys. Rev. B 1998, 57, 15360–15371. [Google Scholar] [CrossRef]
  50. Northrup, J.E.; Neugabauer, J. Indium-induced changes in GaN (0001) surface morphology. Phys. Rev. B 1999, 60, 8473–8476. [Google Scholar] [CrossRef]
  51. Wang, F.H.; Kruger, P.; Pollman, J. Electronic structure of GaN (0001) and surfaces. Phys. Rev. B 2001, 64, 035305. [Google Scholar] [CrossRef]
  52. Timon, V.; Brand, S.; Clark, S.J.; Gibson, M.C.; Abram, R.A. First-principles calculations of 2 × 2 reconstructions of GaN(0001) surfaces involving N, Al, Ga, In, and As atoms. Phys. Rev. B 2005, 72, 035327. [Google Scholar] [CrossRef]
  53. Rosa, A.L.; Neugebauer, J. First-principles calculations of the structural and electronic properties of clean surfaces. Phys. Rev. B 2006, 73, 205346. [Google Scholar] [CrossRef]
  54. Elsner, J.; Haugk, M.; Jungnickel, G.; Frauenheim, T.H. Theory of Ga, N and H terminated GaN (0001)/) surfaces. Solid State Commun. 1998, 106, 739–743. [Google Scholar] [CrossRef]
  55. Van de Walle, C.G.; Neugabauer, J. Structure energetics of nitride surfaces under MOCVD growth conditions. J. Cryst. Growth 2003, 248, 8–13. [Google Scholar] [CrossRef]
  56. Northrup, J.E.; Van de Walle, C.G. Indium versus hydrogen-terminated GaN(0001) surfaces: Surfactant effect of indium in a chemical vapor deposition environment. Appl. Phys. Lett. 2004, 84, 4322–4324. [Google Scholar] [CrossRef]
  57. Pansila, P.; Kanomata, K.; Ahmmad, B.; Kubota, S.; Hirose, F. Nitrogen Adsorption of Si(100) Surface by Plasma Excited Ammonia. IEICE Trans. Electron. 2015, E89-C, 395–401. [Google Scholar] [CrossRef]
  58. Carlos, F.J.; Teplyakov, V. Chemistry of Organometallic Compounds on Silicon: The First Step in Film Growth. Chem. A Eur. J. 2007, 13, 9164–9176. [Google Scholar]
  59. Carlos, F.J.; Teplyakov, V. Chemistry of diffusion barrier film formation: Adsorption and dissociation of tetrakis (dimethylamino) titanium on Si (100)-2 × 1. J. Phys. Chem. C 2007, 111, 4800–4808. [Google Scholar]
  60. Carlos, F.J.; Teplyakov, V. Surface Transamination Reaction for Tetrakis(dimethylamido)titanium with NHX-Terminated Si(100) Surfaces. J. Phys. Chem. C 2007, 111, 16498–16505. [Google Scholar]
  61. Carlos, F.J.; Teplyakov, V. Chemisorption of Tetrakis(dimethylamido)titanium on Si(100)-2 × 1: C–H and C–N Bond Reactivity Leading to Low-Temperature Decomposition Pathways. J. Phys. Chem. C 2008, 112, 9695–9705. [Google Scholar]
  62. Frisch, M.J.; Trucks, G.W.; Schlegel, H.B.; Scuseria, G.E.; Robb, M.A.; Cheeseman, J.R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G.A.; et al. Gaussian 09, Revision B. 01; Gaussian, Inc.: Wallingford, UK, 2009. [Google Scholar]
  63. Promjun, T.; Rattana, T.; Pansila, P.P. Kinetic study on initial surface reaction of titanium dioxide growth using tetrakis(dimethylamino)titanium and water in atomic layer deposition process: Density functional theory calculation. Chem. Phys. 2022, 562, 111653. [Google Scholar] [CrossRef]
  64. Promjun, T.; Phothisonothai, M.; Sriboon, W.; Sukprasong, S.; Pansila, P.P. Comparison of H2O2 and H2O oxidations on TDMAT absorbed on silicon (100) surface during reaction step of ALD–TiO2 process: A DFT study. Mater. Today Commun. 2024, 38, 108125. [Google Scholar] [CrossRef]
  65. Becke, A.D. Density-functional exchange-energy approximation with correct asymptotic behavior. Phys. Rev. A 1988, 38, 3098. [Google Scholar] [CrossRef] [PubMed]
  66. Becke, A.D. Density-functional thermochemistry. III. Role Exact exchange. J. Chem. Phys. 1993, 98, 5648–5652. [Google Scholar] [CrossRef]
  67. Lee, C.; Yang, W.; Parr, R. Development of the Colle-Salvetti correlation-energy formula into a functional of the electron density Phys. Rev. B 1988, 37, 785. [Google Scholar] [CrossRef] [PubMed]
  68. Hariharan, P.C.; Pople, J.A. The influence of polarization functions on molecular orbital hydrogenation energies. Theor. Chim. Acta 1973, 28, 213–222. [Google Scholar] [CrossRef]
  69. Sekiguchi, K.; Shirakawa, H.; Chokawa, K.; Araidai, M.; Kangawa, Y.; Kakimoto, K.; Shiraishi, K. Thermodynamic analysis of trimethylgallium decomposition during GaN metal organic vapor phase epitaxy. Jpn. J. Appl. Phys. 2018, 57, 04FJ03. [Google Scholar] [CrossRef]
  70. Hong, L.S.; Chen, W.L.; Chen, S.W.; Hsu, Y.J.; Jiang, J.C. Atomic Layer Epitaxy of Gallium Nitride using Trimethylgallium and Ammonia. In Asian Pacific Confederation of Chemical Engineers Congress Program and Abstracts; The Society of Chemical Engineers: Tokyo, Japan, 2004. [Google Scholar]
  71. Hinshelwood, C.N.; Burk, R.E. The thermal decomposition of ammonia upon various surfaces. J. Chem. Soc. Trans. 1925, 127, 1105–1117. [Google Scholar] [CrossRef]
  72. Cooper, D.A.; Ljungstroem, E.B. Decomposition of ammonia over quartz sand at 840–960.degree.C. Energy Fuels 1988, 2, 716–719. [Google Scholar] [CrossRef]
  73. Cooper, D.A.; Ghardashkani, S.; Ljungstroem, E.B. Decomposition of ammonia over calcined and sulfated limestone at 725–950.degree.C. Energy Fuels 1989, 3, 278–283. [Google Scholar] [CrossRef]
Applsci 15 07487 g001
Figure 1. Reaction pathways of TMG adsorption on one dimer and two dimers of Si(100) substrates are illustrated by the red and black lines, respectively. The corresponding molecular structures are labeled (a)–(g) for the one-dimer pathway and (h)–(n) for the two-dimer pathway, as shown at the bottom of the figure.
Figure 1. Reaction pathways of TMG adsorption on one dimer and two dimers of Si(100) substrates are illustrated by the red and black lines, respectively. The corresponding molecular structures are labeled (a)–(g) for the one-dimer pathway and (h)–(n) for the two-dimer pathway, as shown at the bottom of the figure.
Applsci 15 07487 g001
Applsci 15 07487 g002
Figure 2. Reaction pathways of TMG adsorption on two-, three-, and four-dimer Si(100) surfaces are illustrated by the black, red, and green lines, respectively.
Figure 2. Reaction pathways of TMG adsorption on two-, three-, and four-dimer Si(100) surfaces are illustrated by the black, red, and green lines, respectively.
Applsci 15 07487 g002
Applsci 15 07487 g003
Figure 3. Reaction pathways of TMG adsorption on Si15H18–(NH2)2 and Si15H18–(NH2)4 substrates are illustrated by the black and red lines, respectively.
Figure 3. Reaction pathways of TMG adsorption on Si15H18–(NH2)2 and Si15H18–(NH2)4 substrates are illustrated by the black and red lines, respectively.
Applsci 15 07487 g003
Applsci 15 07487 g004
Figure 4. Reaction pathway of TMG adsorption on the Si15H18–(NH2)2 substrate using various basis sets. The pathways calculated with 6-31G(d), 6-311G(d,p), 6-311+G(2d,p), and 6-311++G(2df,2p) basis sets are illustrated by the blue, orange, yellow, and gray lines, respectively.
Figure 4. Reaction pathway of TMG adsorption on the Si15H18–(NH2)2 substrate using various basis sets. The pathways calculated with 6-31G(d), 6-311G(d,p), 6-311+G(2d,p), and 6-311++G(2df,2p) basis sets are illustrated by the blue, orange, yellow, and gray lines, respectively.
Applsci 15 07487 g004
Applsci 15 07487 g005
Figure 5. Comparison of reaction pathways of TMG adsorption on Si–NH2 and Si2–NH surfaces, illustrated by the black and red lines, respectively. The corresponding molecular structures are labeled (a)–(g) for adsorption on the Si–NH2 surface and (h)–(n) for adsorption on the Si2–NH surface, as shown at the bottom of the figure.
Figure 5. Comparison of reaction pathways of TMG adsorption on Si–NH2 and Si2–NH surfaces, illustrated by the black and red lines, respectively. The corresponding molecular structures are labeled (a)–(g) for adsorption on the Si–NH2 surface and (h)–(n) for adsorption on the Si2–NH surface, as shown at the bottom of the figure.
Applsci 15 07487 g005
Applsci 15 07487 g006
Figure 6. Energy profile of NH3 adsorption on residual TMG adsorbed on the Si2–NH surface (PC 1–2 substrate). The corresponding molecular structures of NH3 adsorption are labeled (a)–(h), as shown at the bottom of the figure.
Figure 6. Energy profile of NH3 adsorption on residual TMG adsorbed on the Si2–NH surface (PC 1–2 substrate). The corresponding molecular structures of NH3 adsorption are labeled (a)–(h), as shown at the bottom of the figure.
Applsci 15 07487 g006
Applsci 15 07487 g007
Figure 7. Comparison of reaction pathways of NH3 adsorption on TMG-adsorbed Si–NH2 (black line) and Si2–NH (red line) substrates. The corresponding molecular structures are labeled (a)–(d) for adsorption on the Si–NH2 surface and (e)–(h) for adsorption on the Si2–NH surface, as shown at the bottom of the figure.
Figure 7. Comparison of reaction pathways of NH3 adsorption on TMG-adsorbed Si–NH2 (black line) and Si2–NH (red line) substrates. The corresponding molecular structures are labeled (a)–(d) for adsorption on the Si–NH2 surface and (e)–(h) for adsorption on the Si2–NH surface, as shown at the bottom of the figure.
Applsci 15 07487 g007
Table 1. Activation energies of TMG and NH3 on different substrate surfaces for ALD–GaN processes.
Table 1. Activation energies of TMG and NH3 on different substrate surfaces for ALD–GaN processes.
ReactionActivation Energy (eV)
TGM AdsorptionFirst
Activation Energy		Second
Activation Energy		Average
Activation Energy		Reference
On Si–NH2 surface of Si(100)2.001.441.72This work
On Si2–NH surface Si(100)1.112.071.59This work
Gas phase decomposition--3.27[69]
On NH2 terminated Ga(0001)1.281.841.56[47]
On GaN surface of Sapphire--1.20[70]
NH3 adsorption
On PC 1–2 surface Si(100)1.881.921.90This work
On PC 2–1 surface Si(100)--1.96This work
On PC 2–2 surface Si(100)--2.00This work
On GaN surface on Sapphire--1.14[70]
On Platinum Wire surface--6.1[71]
On Tungsten surface--1.7[71]
On quartz sand surface--1.52[72]
On calcined and sulfated
Limestone surface		--1.02, 1.51[73]
On Ga Surface of Sapphire--1.14[70]
On GaN surface of Si(111)--1.34[11]
Table 2. Prediction of infrared frequencies of Ga–N and Ga–CH3 species calculated from product structures compared to other experimental studies, with vibrational frequencies calculated at B3LYP 6–31G(d) level and scaled using 0.96.
Table 2. Prediction of infrared frequencies of Ga–N and Ga–CH3 species calculated from product structures compared to other experimental studies, with vibrational frequencies calculated at B3LYP 6–31G(d) level and scaled using 0.96.
StructureVibration Mode
(Symmetric Stretching)		Wavenumber (cm−1)Other Studies [11,32] (cm−1)
Products (first half)
PC 1–1 + CH4Ga–N531534
Ga–CH329172912
PC 2–1 + CH4Ga–N530534
Ga–CH329172912
PC 1–2 + CH4Ga–N537534
Ga–CH329172912
PC 2–2 + CH4Ga–N574-
Ga–CH32939-
Products (second half)
PC 1–3 + CH4Ga–N557534
PC 1–4 + CH4Ga–N509-
PC 3–1 + CH4Ga–N547534
PC 3–2 + CH4Ga–N539534
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Pansila, P.P.; Sukhasena, S.; Sukprasong, S.; Sriboon, W.; Temnuch, W.; Jamnongkan, T.; Promjun, T. DFT Study of Initial Surface Reactions in Gallium Nitride Atomic Layer Deposition Using Trimethylgallium and Ammonia. Appl. Sci. 2025, 15, 7487. https://doi.org/10.3390/app15137487

AMA Style

Pansila PP, Sukhasena S, Sukprasong S, Sriboon W, Temnuch W, Jamnongkan T, Promjun T. DFT Study of Initial Surface Reactions in Gallium Nitride Atomic Layer Deposition Using Trimethylgallium and Ammonia. Applied Sciences. 2025; 15(13):7487. https://doi.org/10.3390/app15137487

Chicago/Turabian Style

Pansila, P. Pungboon, Seckson Sukhasena, Saksit Sukprasong, Worasitti Sriboon, Wipawee Temnuch, Tongsai Jamnongkan, and Tanabat Promjun. 2025. "DFT Study of Initial Surface Reactions in Gallium Nitride Atomic Layer Deposition Using Trimethylgallium and Ammonia" Applied Sciences 15, no. 13: 7487. https://doi.org/10.3390/app15137487

APA Style

Pansila, P. P., Sukhasena, S., Sukprasong, S., Sriboon, W., Temnuch, W., Jamnongkan, T., & Promjun, T. (2025). DFT Study of Initial Surface Reactions in Gallium Nitride Atomic Layer Deposition Using Trimethylgallium and Ammonia. Applied Sciences, 15(13), 7487. https://doi.org/10.3390/app15137487

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop