Nanoporous Structure Formation on the Surface of InSb by Ion Beam Irradiation

Nanoporous structures have a great potential for application in electronic and photonic materials, including field effect transistors, photonic crystals, and quantum dots. The control of size and shape is important for such applications. In this study, nanoporous structure formation on the indium antimonide (InSb) surface was investigated using controlled focused ion beam irradiation. Upon increasing the ion dose, the structures grew larger, and the shapes changed from voids to pillars. The structures also became larger when the ion flux (high-dose) and accelerating voltage were increased. The structure grew obliquely on the substrate by following the ion beam irradiation of 45°. The shapes of the structures formed by superimposed ion beam irradiation were affected by primary irradiation conditions. The nanostructural features on the InSb surface were easy to control by changing the ion beam conditions.


Introduction
Nanoporous structures on semiconductor surfaces have a great potential for application in electronic and photonic materials, including field effect transistors (FETs), photonic crystals, and quantum dots. In these materials, both the FET gate length and the quantum dot size must be less than 20 nm and the photonic crystals must be approximately 200 nm in size. Controlling the nanoporous structure size and shape is important for such applications. In this study, nanoporous structure formation is investigated on the surface of indium antimonide (InSb) by using controlled focused ion beam (FIB) irradiation. InSb is a narrow-gap (direct-gap) compound semiconductor with a band gap of 0.17 eV and can be applied in electro-photonic devices. Nanoporous structure formation on InSb using ion beam irradiation has been reported [1][2][3][4][5][6][7][8][9], and the structures obtained were similar to those formed on materials such as gallium antimonide (GaSb) [1,[10][11][12][13][14][15][16][17][18], germanium (Ge) [19][20][21][22][23][24][25][26][27], Si 1−x Ge x alloys [28], and GaAs 1−x Sb x alloys [29]. The formation mechanisms on these materials are dominated by the self-assembly of irradiation-induced point defects (Frenkel pairs, an interstitial atom, and a vacancy). Many point defects are generated near the surface by the collision cascade under ion irradiation. Small voids or elevations are formed in the early stage of irradiation. The surface roughness increases through the migration of vacancies and interstitials, thus resulting in the formation of nanoporous structures on the surface. Herein, we report the effects of ion beam conditions (ion beam dose, flux, accelerating voltage, and irradiation angle) on the sizes and shapes of the resulting nanoporous structures. In addition, the effects of superimposed ion beam irradiation were examined using different ion beam doses in the first and second irradiations. The accelerator was used for FIB, which was easy to change ion beam conditions.  Figure 1o. The void shape changed from round to not round. The surface had voids in addition to roughness. Under high-dose irradiation (1 × 10 19 ions/m 2 scan; Figure 1p-t), pillar structures were observed instead of voids. Whereas the voids formed via vacancy aggregation [12], the pillars formed from re-deposition resulting from ion beam sputtering [5,6]. In this experiment, those phenomena depended on different ion doses. Different ion dose irradiation induced changes in the features of the structure features. The pillar had a facet in the structure. It was shown that the pillar was made via recrystallization by sputtered atoms. Figure 2 shows surface SEM images of InSb irradiated with 30 kV Ga + ions at high-flux irradiation. Compared to the images in Figure 1 (low-flux irradiation), voids were formed on the surface irradiated at doses of 1 × 10 17 ions/m 2 scan, as shown in Figure 2d,e (low dose). In Figure 2g-j, irradiated at a dose of 1 × 10 18 ions/m 2 scan, the voids also were formed. The average void diameter was 45 nm in  Figure 2m, and 122 nm in Figure 2o. As the ion dose increased, the small voids observed in Figure 2k disappeared, and the size of the resultant structure became larger; thus, the trends observed with increasing dose were the same as those seen in low-flux irradiation. However, the void and pillar sizes were larger under high-flux irradiation than under low-flux irradiation. High flux was effective at producing large structures. This is because it was thought that the induced vacancies presented with dense distribution under the surface in the short time. The vacancies could easily aggregate, resulting in large voids under high-flux irradiation. In addition, highly efficient sputtering and re-deposition also occurred under high-flux irradiation.
The diameters of the void structures on the InSb samples observed in Figures 1 and 2 are presented as a function of irradiation dose in Figure 3. The void diameter appears to increase roughly linearly with the ion dose. For low doses, the points are all clumped together. The void dimensions under high-flux irradiation were larger than under low-flux in high-dose irradiation.
roughly linearly with the ion dose. For low doses, the points are all clumped together. The void dimensions under high-flux irradiation were larger than under low-flux in high-dose irradiation.                Figure 4k-o)). The pillar size for the accelerating voltage of 30 kV was larger than that obtained at 16 kV, whereas the pillar density was lower. The total sputtering yields (atoms per ion) of In and Sb on InSb irradiated with Ga + ions were calculated by SRIM simulation [30][31][32], and are summarized in Table 1. The total number of atoms per ion is smaller for the accelerating voltage of 16 kV (7.975 atoms/ion) than for 30 kV (8.537 atoms/ion). Thus, larger structures are expected to form on samples irradiated at 30 kV.  Figure 4k-o)). The pillar size for the accelerating voltage of 30 kV was larger than that obtained at 16 kV, whereas the pillar density was lower. The total sputtering yields (atoms per ion) of In and Sb on InSb irradiated with Ga + ions were calculated by SRIM simulation [29,30], and are summarized in Table 1. The total number of atoms per ion is smaller for the accelerating voltage of 16 kV (7.975 atoms/ion) than for 30 kV (8.537 atoms/ion). Thus, larger structures are expected to form on samples irradiated at 30 kV.    (Figure 6f-j,p-t). Tilted voids and pillars grew on the samples' surfaces. The pillar of the highest dose at 45° irradiation was observed to be more sputtered by the ion beam. The total number of sputtering atoms is 8.537 atoms/ion in 30 kV and 14.698 atoms/ion in 30 kV, at 45° tilt as by calculated SRIM simulation ( Table 1). The sputtering rate was higher under tilted irradiation, thus resulting in the formation of smaller structures.     Table 2 shows the atomic percentage of EDX quantification. It also indicated a low intensity of Sb atoms and a richness in In atoms. These results indicate that the top of the pillar was made from re-deposited In atoms by sputtering. The ratio of In atom sputtering was higher than that of Sb atom sputtering, as calculated by the SRIM simulation in Table 1.    Table 2 shows the atomic percentage of EDX quantification. It also indicated a low intensity of Sb atoms and a richness in In atoms. These results indicate that the top of the pillar was made from re-deposited In atoms by sputtering. The ratio of In atom sputtering was higher than that of Sb atom sputtering, as calculated by the SRIM simulation in Table 1.   Table 2 shows the atomic percentage of EDX quantification. It also indicated a low intensity of Sb atoms and a richness in In atoms. These results indicate that the top of the pillar was made from re-deposited In atoms by sputtering. The ratio of In atom sputtering was higher than that of Sb atom sputtering, as calculated by the SRIM simulation in Table 1.   Table 2. EDX quantification (atomic %) of InSb irradiated with a 30 kV Ga + ion beam at a dose of 3 × 10 20 ions/m 2 , and a flux of 5.3 × 10 18 ions/m 2 s, at room temperature.

Ga
3.5 In 52 Sb 45 Figure 8 shows surface SEM images of InSb irradiated with two doses of a 30 kV Ga + ion beam. The first and second irradiation doses were 1 × 10 19 and 1 × 10 20 ions/m 2 scan, respectively. The images shown in Figure 8a-d (reshown as Figure 2k-n) were collected after the first irradiation at a dose of 1 × 10 19 ions/m 2 scan, whereas those shown in Figure 8e-h were obtained after both irradiations (1 ×10 19 and 1 × 10 20 ions/m 2 scan). As shown in Figure 8a-h, the voids grew with increasing the ion dose, and the surface became rough. On the other hand, under superimposed irradiation, the void diameter was small. The small voids were observed in the sample depicted in Figure 8a, they did not grow larger by superimposed irradiation. The initial structure was important for the formation of the InSb structures. The induced interstitial migrated through the initial structure wall of void, resulting in the growth of voids.    Figure 8a-h, the voids grew with increasing the ion dose, and the surface became rough. On the other hand, under superimposed irradiation, the void diameter was small. The small voids were observed in the sample depicted in Figure 8a, they did not grow larger by superimposed irradiation. The initial structure was important for the formation of the InSb structures. The induced interstitial migrated through the initial structure wall of void, resulting in the growth of voids.  Figure 9 shows surface SEM images of InSb with superimposed irradiation with a 30 kV Ga + ion beam at a first irradiation dose of 1 × 10 20 ions/m 2 scan (Figure 2p-s), and a second irradiation dose of 1 × 10 19 ions/m 2 scan. In contrast to the results shown in Figure 8, the first irradiation dose was high (1 × 10 20 ions/m 2 scan), and the second irradiation dose was low (1 × 10 19 ions/m 2 scan). Pillar structures were observed on the sample surface after the first irradiation dose. Upon superimposed irradiation, pillars were not formed. This suggests that the mechanism of pillar formation was dominated by ion beam sputtering. The low-dose irradiation did not induce sputtering and void formation.   (Figure 2p-s), and a second irradiation dose of 1 × 10 19 ions/m 2 scan. In contrast to the results shown in Figure 8, the first irradiation dose was high (1 × 10 20 ions/m 2 scan), and the second irradiation dose was low (1 × 10 19 ions/m 2 scan). Pillar structures were observed on the sample surface after the first irradiation dose. Upon superimposed irradiation, pillars were not formed. This suggests that the mechanism of pillar

Conclusions
Ion beam conditions affected the formation of nanoporous structures on the InSb surface in this study. The structure's size became large as the ion dose, flux (high-dose), and accelerating voltage increased. The structure's shape changed from voids to pillars with increasing the ion dose. The oblique structure was obtained by tilting the sample by 45 degrees with respect to the ion beam radiation. Under the superposed ion beam irradiation, the structure's shape was affected by the primary structure formed during the first irradiation dose in Figures 8 and 9. The nanostructural features were easy to control by changing the conditions of ion beam irradiation.

Conclusions
Ion beam conditions affected the formation of nanoporous structures on the InSb surface in this study. The structure's size became large as the ion dose, flux (high-dose), and accelerating voltage increased. The structure's shape changed from voids to pillars with increasing the ion dose. The oblique structure was obtained by tilting the sample by 45 degrees with respect to the ion beam radiation. Under the superposed ion beam irradiation, the structure's shape was affected by the primary structure formed during the first irradiation dose in Figures 8 and 9. The nanostructural features were easy to control by changing the conditions of ion beam irradiation.