Research on the epitaxial direction control and mechanism of polycrystalline diamond grains

Abstract

The control of diamond film growth is crucial for the application of diamond films. In this article, different diamond film samples were prepared by adding nitrogen and oxygen in different proportions during the growth process. The key conditions for controlling the growth of diamond film with preferred orientation were obtained using scanning electron microscopy (SEM), X-ray diffraction (XRD), and Raman spectroscopy characterizations. During the growth process, the growth of the diamond film with (100) preferred orientation could be achieved by adjusting nitrogen content. It was necessary to introduce nitrogen and oxygen in the growth process, and precisely regulate the ratio of the two to obtain the (111) preferred orientation of the diamond film. When only oxygen was introduced during growth, the preferred orientation of the diamond could not be achieved, leading to a smaller grain size and poorer crystal quality. From the perspective of grain morphology to film morphology evolution, this study obtained the crystal orientation control conditions during the diamond film growth, which has important guiding significance for the application of the diamond film. This study seeks to drive progress in the large-scale growth of readily machinable diamond and the evolution of diamond-grain optical fiber sensing through comprehensive investigations into the correlation between the preferred orientation in polycrystalline diamonds and reaction conditions.

Keywords:

• Diamond
• preferred orientation
• oxygen and nitrogen addition
• morphology

1. Introduction

As one of the representative materials of ultra-wide bandgap semiconductors, diamond has many excellent properties, such as high hardness, thermal conductivity (2200 W/(m·K)) [1], breakdown field strength [2], carrier mobility [3], etc. These excellent properties give diamond great potential applications in microwave power devices and thermal management. In addition, the diamond-related color center structure has high quantum efficiency, perfect photo-stability, and other properties [4–6], which has a very wide application prospect in the field of optoelectronics, biomarkers, etc.

The diamond application needs to be based on controlling material growth. For the growth of polycrystalline diamond films, it is vital to control the preferred grain orientation [7]. On the one hand, the large-size polycrystalline diamond film with good grain orientation consistency has less difficulty in subsequent grinding and polishing processes and has a higher yield. On the other hand, the preferred orientation of the diamond film also has an important influence on the impurity incorporation and the performance of electronic devices [8–12]. Our previous studies investigated the effect of crystal orientation of diamond materials on the performance of Si-V color centers [13] and the hydrogen terminal field-effect transistor (FET) [14].

In the chemical vapor deposition (CVD) growth process of the diamond film, nitrogen and oxygen, as the most common auxiliary precursor gases in growth, are often introduced into the growth atmosphere. Nitrogen can increase the growth rate of diamond [15] and form high-quality nitrogen-vacancy (NV) color centers [9]. Oxygen can effectively inhibit the incorporation of impurity atoms in the diamond growth process [16] to improve the crystal quality of the diamond [17]. However, it also leads to a decrease in the diamond growth rate [18, 19]. At the same time, it has been found that the introduction of nitrogen and oxygen can also change the preferred grain orientation of the diamond film. Yu et al. used hot-filament CVD (HFCVD) to study the effect of nitrogen and oxygen on diamond growth [20]. They proposed that nitrogen promoted the morphology of diamond (100) and oxygen promoted the morphology of diamond (111). Neto et al. grew diamond on the pure cobalt substrate [21]. They found that nitrogen and oxygen affected the orientation morphology of the diamond. When oxygen was introduced, the diamond film showed (111) morphology, and conversely, the film showed (100) morphology. The C.J. Tang et al. found a {111} facet-dominated large-grained polycrystalline diamond film of mixed <211> and <110> texture, and a small amount of nitrogen was beneficial to the formation of (110) nanocrystalline diamond film [22]. Though it has been reported that oxygen and nitrogen can affect the crystal orientation of diamond thin film grains during the growth process, the morphology formation mechanism of the diamond film has not been explained clearly.

In this study, the formation mechanism of diamond film morphology was revealed by changing the ratio of nitrogen and oxygen during the diamond growth process. The morphology of diamond discrete grains was observed under different conditions to realize the controlled growth of CVD polycrystalline diamond film with preferred orientations. Promoting the development and application of the diamond film is of great significance.

2. Experimental sections

In this study, the growth of diamond films was carried out on the intrinsic silicon substrate, with a size of 15 × 15 mm². Before starting the experiment, the silicon substrate was immersed in the diamond powder solution for ultrasonic treatment for 15 min. After the ultrasound was completed, the sample was dried using nitrogen gas. In order to observe the discrete grain morphology, we used a dust-free cloth dipped in deionized water to wipe part of the edge area of the sample. This artificially reduced the density of crystal seeds in this area, avoiding the formation of continuous diamond films to achieve independent diamond grains. The dried silicon substrate was placed in the microwave plasma CVD (MPCVD) chamber, and hydrogen, methane, nitrogen, and oxygen were introduced to grow the diamond. The equipment used for the growth was the WORLDIRAY W-150A-6K MPCVD system. The equipment limit vacuum could reach 6.5 × 10⁻⁷ mbar. During the growth process, the total gas flow rate was 200 sccm, and the ratio of N₂ and O₂ were 0–0.02% and 0–0.5%, respectively. The growth temperature was maintained at 960–980 °C. The chamber pressure, microwave power, and growth time were 185 mbar, 4.2 kW, and 2 h, respectively. A total of nine samples were prepared, numbered S1-S9. Among them, S1, S2, and S3 are the first group of experiments, S4, S5, and S6 are the second group, and S7, S8, and S9 are the third group. Table 1 summarizes the specific growth condition changes of the three experimental groups.

Table 1. The summary of experimental condition variables for preferred orientation growth of diamond.

First group	Samples	S1	S2	S3
	N2 (%)	0
	O2 (%)	0	0.25	0.5
Second group	Samples	S4	S5	S6
	N2 (%)	0.01
	O2 (%)	0	0.25	0.5
Third group	Samples	S7	S8	S9
	N2 (%)	0.02
	O2 (%)	0	0.25	0.5

The surface morphology of the samples was observed using ZEISS field-emission scanning electron microscopy (SEM), and the test equipment model was Sigma300. The Raman and Photoluminescence (PL) spectrum was tested using the German WITec scanning nearfield optical microscope (SNOM) with the testing equipment model Alpha300Rs. The crystal orientation of diamond films was tested by X-ray diffraction (XRD) equipment X ‘pert³ MRD in Malvern Panalytical, the Netherlands.

3. Results and discussion

The substrate after seeding was characterized to ensure that the original seed morphology did not unduly influence the experimental results. The SEM test results of the morphology of seeds planted on the substrate are shown in Figure 1. There is no obvious orientation of diamond crystal seed observed at 40,000 times, which indicates that the crystal seed used in this experiment has little influence on subsequent experiments.

Figure 1. Surface morphology of silicon substrate after seeding (a) 5000 times and (b) 40000 times.

The SEM test results of the first group of three samples (S1–S3) are shown in Figure 2(a1), (b1), and (c1), showing the results of the non-film-forming grain. Figure 2(a2), (b2), and (c2) show the results of the film-forming region of the sample. Figure 2(a3), (b3), and (c3) show the transition region of discrete grains to film formation. Only methane and hydrogen were added to S1 during the growth process. It was observed that the surface of the diamond grain was obviously square-shaped (100) and triangular-shaped (111) [23, 24]. In the growth process of S2, 0.25% O₂ was introduced. Compared with S1, although the characteristic morphologies of (100) and (111) crystal planes also appeared, the diamond grain size was reduced. However, the surface of sample S2 was smoother than S1. This is because oxygen was added during the growth process, and the etching effect of oxygen reduced the diamond grain size. By observing the grain morphology of S3, it can be found that with the increase of the O₂ ratio, the grain size was further reduced due to the further strengthening of etching, and it was difficult to distinguish the typical preferred crystal orientation. These results indicate that the obvious preferred orientation growth of diamond grains could not be induced when only oxygen was introduced.

Figure 2. (a1) Grain morphology and (a2) film morphology and (a3) the transition region of discrete grains to film formation morphology of S1 sample; (b1) Grain morphology and (b2) film morphology and (b3) the transition region of discrete grains to film formation morphology of S2 sample; (c1) grain morphology and (c2) film morphology and (c3) the transition region of discrete grains to film formation morphology of S3 sample.

Further SEM tests were conducted on the film-forming region of the first group of samples (Figure 2(a2)–(c2)). It was found that the grain boundaries of the surface grains of S1 were steep, and the grain size was large. However, the grain shapes of the diamond films on S2 and S3 gradually became smoother, and the grain size became smaller. This is consistent with the above results for independent grains. With the increase of the O₂ ratio, the diamond grain size decreased due to the enhancement of etching and the inhibition of the merger of grains; so, the formation of preferred crystal growth of diamond was not induced. In addition, through the observation of the morphology of the sample, it was found that the effect of oxygen on the growth rate of different diamond crystal orientations was isotropic.

The typical morphology of (100) and (111) orientations was observed for individual diamond grains. However, when the grains merged into films, the edges between (100) and (111) surfaces appeared on the surface, and the preferred diamond film orientation was not observed. To further illustrate the process of grain growth to form thin films, the transition region of discrete grains to film formation was characterized by SEM (Figure 2(a3)–(c3)). It can be observed that the grain boundaries are randomly connected, and after the grain is connected into a film, a certain characteristic plane does not appear in the film region. This also shows that diamond film growth would not show a single preferred orientation in the case of only additional oxygen.

The second group of samples (S4–S6) was kept at the same O₂ flow as the first group to further investigate the conditions that can induce the preferred crystal orientation growth of diamond polycrystals. Meanwhile, the 0.01% N₂ was further introduced during the growth process. The SEM test results of the samples are shown in Figure 3. The S4 and S1 samples were similar in grain morphology (Figure 3(a1)). However, compared with S1, the diamond grain size of S4 was larger. The samples had the same growth time, indicating that the S4 grew faster than the S1. At the same time, it can be observed that in the diamond film region of S4, the quadrilateral platform was connected into a film, showing a square characteristic morphology typical of (100) crystal faces (Figure 3(a2)). However, with the increase of the O₂ ratio, the surface of grain (111) gradually shrank (Figure 3(b1)) and gradually evolved into a cuboid structure with only (100) morphology (Figure 3(c1)). With the addition of O₂, the surface morphology of the typical (100) orientation that was dominant after film formation also changed into the (111) orientation. Finally, it was changed into the typical morphology of the vertex outcrop of the cube with the disappearance of the (111) plane (Figure 3(c2)). It can be observed that the grain boundary preferentially combines and annihilates the non-preferred orientation plane, and after the grain is connected to form a film, a certain preferred orientation characteristic morphology appears in the film region (Figure 3(a3)–(c3)).

Figure 3. (a1) Grain morphology and (a2) film morphology and (a3) the transition region of discrete grains to film formation morphology of S4 sample; (b1) Grain morphology and (b2) film morphology and (b3) the transition region of discrete grains to film formation morphology of S5 sample; (c1) grain morphology and (c2) film morphology and (c3) the transition region of discrete grains to film formation morphology of S6 sample.

The analysis of the second group of samples shows that, on the one hand, the growth rate of diamond could be increased by adding only a trace amount of nitrogen. On the other hand, during the grain consolidation process, the (100) crystal face could be induced upward, and eventually, a film with the (100) face of the grain was formed [25]. Subsequently, with the increasing addition of O₂, it was found that the growth of the (111) face was completely inhibited, forming a cube with only the (100) face. However, in the grain merger process, there was no (100) crystal face upward (Figure 3(c2)), but the top angle of the cube had upward morphology. Finally, the diamond film morphology with (111) surface dominance was formed.

We further introduced 0.02% N₂ in the growth process of the third group of samples based on the second group of experiments. The SEM test results of the sample surface prepared by the third group of experiments are shown in Figure 4. It can be seen from Figure 4 that the morphology variation of diamond samples was similar to that of the second group. With the increase of the O₂ ratio, the preferred orientation of the diamond changed from (100) to (111). Different from the second group, quadratic nucleation occurred on the grain surface due to the increase in the N₂ ratio. Compared with S4, due to the increase in the N₂ ratio, the upward growth rate of the (100) crystal orientation increased, and the area of the (100) crystal surface decreased in S7. After the merger of the grains, the (100) crystal faced upward, and the typical morphology of the (111) crystal face could be significantly observed between the grains. The secondary nucleation can be observed more obviously by observing the transition region between the grains and the film forming region of the third group of samples (Figure 4(a3)–(c3)). A clear preferred orientation can be observed on the surface of the newly formed crystal nuclei. Due to the large area of the non-preferred orientation plane, it is easier to form crystal nuclei on the non-preferred orientation plane, so that more preferred orientation planes appear.

Figure 4. (a1) Grain morphology and (a2) film morphology and (a3) the transition region of discrete grains to film formation morphology of S7 sample; (b1) Grain morphology and (b2) film morphology and (b3) the transition region of discrete grains to film formation morphology of S8 sample; (c1) grain morphology and (c2) film morphology and (c3) the transition region of discrete grains to film formation morphology of S9 sample.

Based on the above morphological characterization results, it can be concluded that the process of grain coalescence and film formation is related to the growth conditions. Under the growth conditions that can induce the preferred orientation of diamond, the non-preferred orientation planes of the grain preferentially contact and merge, and the preferred orientation plane is retained. Conversely, under growth conditions in which there is no obvious diamond preference orientation, the grains will be randomly merged. The position of diamond grain connection is easier for secondary nucleation. In the process of secondary nucleation, the non-preferred orientation plane of the grains is merged and the preferred orientation plane of the diamond is further retained. Therefore, in the film region of the sample with obvious preferred orientation, the preferred orientation plane upward and the non-preferred orientation plane are connected.

XRD tests were conducted on the samples to further characterize the grain orientation of polycrystalline diamond films. The test results are shown in Figure 5. Because the sample growth time was short and the diamond film thickness was thin, the characteristic peak of the silicon substrate could be observed in the test results. The XRD test results of all samples contained the following diffraction peaks: diamond (111) crystal face (D (111)), diamond (220) crystal face (D (220)), and diamond (311) crystal face (D (311)). The peak strength of D (220) and D (311) of all samples did not change basically, but the strengths of the two peaks decreased significantly in the test curve of S9 samples. The peak intensity of D (111) changed obviously with the change in gas proportion.

Figure 5. XRD test results (2θ mode) of (a) samples S1–S3, (b) samples S4–S6, and (c) samples S7–S9.

Figure 5d summarizes the peak intensity of D (111) for nine samples. Observing the first group of samples S1–S3, it was found that the D (111) diffraction peak intensity of the diamond decreased continuously when the O₂ ratio was increased without adding N₂, which is consistent with the morphology observed by SEM. During the growth of samples S1, S4, and S7, no additional O₂ was added, and the content of N₂ was increased. This led to a gradual decrease in the diffraction intensity of the (111) surface while promoting the growth of the (100) surface. The comparison results of S3, S6, and S9 are available. The diffraction intensity of the (111) plane increased gradually when the oxygen content was kept constant and the nitrogen content was gradually increased.

The XRD and SEM results show that the growth rate of the sample could be increased when only N₂ was added during the growth process. However, the surface crystal direction of the grain did not change much. Interestingly, with the increase of N₂, the (100) crystal face was gradually shown to be upward during the process of grain consolidation. When O₂ was added during the growth process, the growth rate and grain size were reduced due to the etching effect of oxygen. However, it did not change the overall morphology of the grain and the preferred orientation of the combined film. During the growth process, when N₂ and O₂ were introduced simultaneously, it was possible to achieve a (111) crystal-free, and the cubic grain comprised six (100) planes (samples S6, S9). However, in the grain consolidation process, because the surface was exposed as the top angle of the cube, the preferred orientation of the surface was (111). Therefore, it can be preliminarily concluded that N₂ should be introduced during the growth process to obtain a diamond film with (100) preferred orientation, and no other gases should be added. Diamond polycrystalline films with the same (100) crystal face can be obtained by adjusting the N₂ flow rate. In addition, N₂ and O₂ can be added simultaneously during the growth process, and the ratio of the two can be adjusted to obtain (111) preferentially oriented diamond films.

Finally, Raman spectrum tests were conducted on the samples to study the crystal quality of diamond films grown under different conditions, and the test results are shown in Figure 6. Figure 6a shows the Raman spectra of the first group of samples. The full width at half maximum (FWHM) of 5.714 cm⁻¹ for S1 reached 58.362 cm⁻¹ for S3 with the continuous increase of the O₂ ratio. This is consistent with the observed decrease in the grain size and the intensity of the (111) diffraction peak. The Raman spectra of the three samples in the second group are shown in Figure 6b. It can be observed that when 0.01% N₂ was introduced, the strength of the G peak of the sample increased with the increase of the O₂ ratio, but the FWHM of the diamond Raman peak decreased. This shows that the crystal quality of the diamond increased with the increase of the O₂ proportion while adding N₂. Figure 6c shows the Raman test results of the third group of samples. Due to the further increase of the N₂ ratio, the Raman spectra of S7 showed a peak of neutral NV (NV⁰). With the increase of the O₂ ratio, the NV⁰ peak disappeared, and the intensity of the G peak increased. The above results show that the material quality and growth rate could be improved effectively by introducing the proper proportion of O₂ and N₂ during the growth process, and the formation of nitrogen vacancies could be inhibited.

Figure 6. Raman spectrum test results of (a) samples S1–S3, (b) samples S4–S6, and (c) samples S7–S9.

4. Conclusion

In this article, the growth of diamond samples with different preferred orientations was realized by controlling the proportion of N₂ (0–0.02%) and O₂ (0–0.5%). In addition, the influence of nitrogen and oxygen on the growth of diamond with preferred orientations was investigated. When only N₂ was added during the growth process, the growth rate of the sample could be improved. With the increase of N₂, the (100) crystal face was shown to be upward during the grain consolidation process. When O₂ was added during the growth process, the growth rate and grain size were reduced due to the etching effect of oxygen. However, the overall morphology of the grain and the preferred orientation of the combined film were not changed with the addition of oxygen. When nitrogen and oxygen were introduced simultaneously during the growth process, it was possible to achieve a cubic grain composed of six (100) planes without (111) crystal orientation. However, in the grain consolidation process, because the surface was exposed as the top angle of the cube, the preferred orientation of the surface was (111). In this study, experiments proved that polycrystalline diamond films with square (100) features can be obtained by introducing only N₂ during the growth process. In addition, by adjusting the N₂ flow rate, polycrystalline diamond films with the same orientation (100) can be obtained. In addition, by adding N₂ and O₂ simultaneously during the growth process and adjusting their ratio, cube-shaped diamond grains and (111) preferentially oriented diamond films can be grown.

Authors’ Contributions

Li Junpeng: Conceptualization, Methodology, Investigation, Formal Analysis, Sample preparation, Writing - Original Draft. Ren Zeyang: Data curation, Sample preparation, Writing- Original draft preparation. Zhang Jinfeng: Resources, Supervision. Zhang Jincheng: Resources, Supervision. Su Kai: Resources, Supervision. Meng Jintao: Investigation, Sample Test. Zhu Liaoliang: Investigation, Sample Test. Li Yijiang: Investigation, Sample Test. Chen Junfei: Resources, Supervision. Wang Hanxue: Resources, Data analysis. Hao Yue: Resources, Supervision.

Disclosure 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 article.

Additional information

Funding

This work was supported in part by National Key Research and Development Program of China (Grant No. 2022YFB3608600), National MCFEnergy R&D Program (Grant No. 2019YFE03100200), National Natural Science Foundation of China (Grant 62127812, 62134006, 62374122, 62204193), Key R&D Plan of Shandong Province (Grant No. 2022CXGC020306), China Postdoctoral Science Foundation (Grant No. 2021TQ0256), Key research and development program of Anhui Province (Grant No. 2023a05020006).