To perform the Si nanopillar formation, the key issue is the proper magnitude of the etching power. There are two kinds of powers included in the ICP-RIE equipment that are the rf and bias power. The rf power decides the isolating level of the etching gases
and the dc bias power controls the vertical level of the side wall. In this experimental process, the experimental tests for the two powers are the ratio and magnitude. From the 45° view angle of the SEM images in Fig. 3.5, we can compare the different formation of Si nanopillars at the instinct power ratio. In this comparison, we chose the optimum results which are the largest length. So, the etching time is also different for the different power ratio. The etching time is 3 min、5min and 50 s when the rf/bias power is 300(W)/100(W)、200(W)/100(W) and 200 (W)/200(W).
(a) 300(W)/100(W) (b) 200(W)/100(W) (c) 200(W)/200(W)
Fig. 3.5 (a) shows the apparently triangle shape of Si nano-pillar. (b) The sharper sidewall of Si nanopillar. (c) A sharp side-wall of Si pillar with insufficient height.
Figure 3.5(a) shows that the by-product appears and the shape of Si nano-pillar resembles a pyramid when the ratio is 3. Then the density is the smallest and the diameter is the largest. The cause is isotropic etching erodes the smaller pillar, and the reason for a pyramid shape is that the vertical etching-rate is not bigger than the side etching-rate so the diameter of the pillar becomes bigger after longer etching time.
Finally, the etching rate is dominated by the power. So we are sure the ratio of 3 is too big to get a straight pillar. Figure 3.5(b) shows the success of the formation of Si nanopillars with a perpendicular sidewall at the ratio of 2. The density and diameter of Si nanopillars are 2.8×1010 cm-2 and 29 nm. Figure 3.5(c) shows the straight but short Si nanopillars at the ratio of 1. This is because the vertical etching is too big to balance the etching selection between Ni and Si, we can see the Ni mask disappeared during the etching.
0 1 2 3 4
Power Ratio (ICP power / Bias power) Density (*109 cm-2 )
30 40 50 60Average diameter (nm)
Fig. 3.6 Different ratio of ICP power over bias power has different influence on the vertical level of Si pillars’ side-wall.
In another word, the appropriate ratio of bias and ICP power is the key point to get sharp sidewall and large height of Si nanopillars. After analyzing the SEM data in Fig.
3.6, we can know the ratio of bias power over ICP power had great influence on the character of Si nano-pillars such as the diameter and the density. In summary, we know that the ratio of 2 is the most suitable for the formation of Si nanopillars. From Fig. 3.7 we can know if the magnitude is bigger, the Si nano-pillars will be longer and the size of them will be wider. The average size of bias/ICP power at 100(W)/200(W) is 29 nm and the average one at 50(W)/100(W) is 44nm. The average height of 100(W)/200(W) is 150nm and the one of 50(W)/100(W) is 250nm. From the cross-session image in Fig.
3.7, Si nanopillars with high aspect-ratio were formed at the bias/ICP power of 2 whatever the magnitude of the powers is.
(a) 100(W)/200(W) (b) 50(W)/100(W) Fig. 3.7 The different magnitude of power with the same ratio of bias and ICP power.
After testing the generally proper rf / bias power ratio and magnitude, the individual effect of the rf and bias powers were also studied. In this part, the etching mask is the Ni nanodots forming after the RTA at 850℃ for120 s and the density of Ni nanodots is about 5x1011 cm-2. Analyzing from Fig 3.8, it is clearly that the magnitude of rf power affects the etching rate and selectivity a lot. Nevertheless, the etching rate is tradeoff with the etching selectivity. In Fig. 3.8 (a), the etching selectivity is high but the etching rate is very low when the rf power is 50 W. The average depth is about 100nm. In Fig.3.8 (b), the etching selectivity and rate are good and the average depth is 170 nm when the rf power is 100 W. In Fig. 3.8 (c), the etching selectivity and rate are the optimum tradeoff for the largest depth of 400nm when the rf power is 150 W. In Fig.
3.8(d), the Si nanopillars disappeared for the low etching selectivity and high etching rate. So, the rf power of 150 W is the most appropriate for balancing the etching selectivity and rate. The average diameter and length of the Si nanopillars are about 50 nm and 400 nm, so the aspect ratio is about 8.
(a) 50 W (b) 100 W
150 (W) 200 (W)
Fig. 3.8 The cross section SEM images of the Si nanopillars were formed at the different RF power from 50 W to 200 W
For the dry etching by ICP-RIE, the vertical etching is controlled by the bias power.
And it affects the etching rate and selectivity more sensitively than rf power. After observing the SEM images in Fig. 3.9, the Si nanaopillars were almost formed at the bias power of 50 W. In Fig. 3.9 (a), the etching depth is nearly 0 nm at the bias power 0 W. From Fig. 3.9(a)-(c), the etching depth is proportion to the magnitude of the rf power. But, the Si nanopillars were etched over on the account of too huge power in Fig.
3.9 (d). Finally, we get the most suitable magnitude of bias power is 50 W. The average diameter and length of the Si nanopillars are 35 nm and 170nm and the aspect ratio is nearly 5.
(a) 0 W (b) 25 W
(c) 50 (W) (d) 75 W
Fig. 3.9 The cross section SEM images of the Si nanopillars were formed at the different bias power from 0 W to 75 W.
3.3-4 Etching Time
By the intuition, the etching depth will be longer if the process takes more time to
etch. From Fig. 3.10 (a)-(c) the etching result matches the intuition. However, the etching time extending over the optimum time would over etch the Si nanopillars.Because the Ni mask also etched by plasma, the long-timed etch would remove the Ni completely. The Si nanopillars without Ni cap would shorten during the etching process. In our etching process, the optimum etching time is 7 min when the pressure 、CF4/Ar flowing and rf/bias power are 0.66 pa、40(sccm)/40(sccm) and 100(W)/50(W). The average diameter and length of the Si nanopillars are 35 nm and 170nm and the aspect ratio is nearly 5.
(a) 3mins (b) 5mins
(c) 7mins (d) 9mins
Fig. 3.10 The cross section SEM images of Si nanopillars were etched for different time from 3 min to 9 min
3.4 Conclusion
Si nanopillars with high aspect ratio were formed on Si substrate with the self-assembled Ni nanodots by ICP-RIE. The etching depth was proportion to the magnitude of CF4 flowing、Ar flowing、the chamber pressure、the etching time、
the bias power and the rf power, but the etching selectivity between the Si and Ni dots is completely opposite. Because of the tradeoff of the etching depth and the etching selectivity, there is an optimum value to every etching parameter. In our experiment, the optimum aspect ratio of Si nanopillars happened with the average diameter and depth of 50 nm and 400 nm in our experiment when the gas flowing of CF4 and Ar、the bias/rf power、the chamber pressure、the etching time were set at
40sccm、40sccm、50(W)/150(W)、0.66pa and 7 min. Comparing to the results of the Korean [3.7], the density of the Si nanopillars in our experiment is over ten times larger than the Korean because of the Ni nandot mask of the high density. Limited to the size of the Ni nanodots, the average diameter of the Si nanopillars is in the range of 30 nm to 50 nm.
3.5 Reference
[3.1] D. Papadimitrioum and A. G. Nassiopoulou, J. Appl. Phys. 84, 1059 (1998)
[3.2] T. V. Torchynska, J. A. Hernandez, A. I. D. Cano, F. G. B. Espinoza, Y.
Goldstein, A. Many, J. Jedrzejewskii, L. Y. Khomenkova, B. M. Bulakh and L.V. Scherbina, Phys. Stat. Sol. 2, 382 (2003)
[3.3] D. D. D. Ma, S. T. Lee and J. Shinar, Appl. Phys. Lett. 87, 033107-1 (2005)
[3.4] Y. Homma, P. Finnie, T. Ogino, H. Noda and T. Nrisu, J. Appl. Phys. 86, 3083 (1999)
[3.5] Y. L. Chueh, L. J. Chou, S. L. Cheng, J. H. He, W. W. Wu and L. J. Chen, Appl. Phys. Lett. 86, 133112 (2005)
[3.6] C. W. Kuo, J. Y. Shiu and P. Chen, Chem. Mater. 15, 2917 (2003) [3.7] M. J. Kim, J. S. Lee, S. K. Kim, G. Y. Yeom, J. B. Yoo and C. Y. Park,
Thin Solid Films 475, 41 (2005)
[3.8] T. Tada and T. Kanayama, J. Vac. Sci. Technol. B 16, 3934 (1998)
[3.9] V. Poborchii, T. Tada, and T. Kanayama, Opt. Commun. 210, 285 (2002).
[3.10] K. Sunouchi, H. Tkato, N. Okabe, T. Yamada,T. Ozaki, S. Inoue, K.
Hashimoto, K. Hieda, A. Nitayama, F. Horiguchi, and F. Masuoka, in IEDM
Tech. Dig. 23 (1989).
[3.11] A. V. Karabutov, V. D. Frolov, A. V. Smakin, and G. A. Shafeev, IEEE Proceedings of the 14th International Vacuum Microelectronics Conference (IVMC2001) 115 (2001).
[3.12] M. Terauchi, N. Shigyo, A. Nitayama, and F. Horiguchi, IEEE Trans.
Electron. Dev. 44, 2303, (1997).
[3.13] F. D. Bauer, Solid-State Electron. 48, 705 (2004).
[3.14] A. G. Nassiopoulos, S. Grigoropoulos, and D. Papadimitriou , Appl. Phys.
Lett. 69, 2267 (1996).
[3.15] W. Beyer, M. Lejeune, J. Muller, U. Zastrow, M. Albert, T. Robler, Proceedings of 3rd World Conference on Photovoltaic Energy Conversion (IEEE Cat. No.03CH37497) 2, 1584 (2003)