Effects of Leveler Concentration in High Aspect Ratio Via Filling in 3D SiP

Abstract

3D packaging field is actively being studied in order to obtain better characteristics, such as shorter interconnection, reduction in signal delay, etc. Electroplating copper via filling is the most important technology in 3D stacking interconnection of SiP. Copper is inexpensive electrode material that has excellent electrical properties and easily obtained. In this study, the effects of leveler concentration in high aspect ratio via filling was investigated without the addition of other additives such as inhibitor and accelerator. Tetronic 701 was used as leveler. The effects of leveler on copper deposition was investigated using galvanostatic, polarization and cyclic voltammetric techniques. High overpotential of copper deposition in tetronic 701 added solution was confirmed. Finally, the optimum conditions of copper via filling in high aspect ratio via (diameter 10 μm, depth 150 μm, AR 15) was obtained.

1. Introduction

Most portable devices require smaller part size, multifunction capability, and higher performance. Au wire bonding has been widely used for the interconnection between stacked chips. Although wire bonding can reduce footprint of packaged chip area, it cannot reduce RC-delay due to its long wiring length. 3D packaging is a fast growing field that encompasses different types of technologies. One of the most promising technologies in 3D packaging is through silicon via (TSV).^1,2) It can achieve high density of packaging and minimize signal delay. Low electric resistance is achieved by interconnecting through via holes due to its short wire length. It is also compatible with multilayer and back-end processes.^1–3) In this study, high aspect ratio via was filled by electroplating method. The inside of via hole should have no void or seams and it can be achieved by bottom-up deposition in fine trenches called ‘superfilling’.^4,5) For successful bottom-up superfilling, which is a prerequisite condition for defect-free copper via filling, organic additives such as inhibitor, accelerator and leveler are necessary in a copper plating bath.^6–10)

However, the decomposition of these organic additives process have been known to cause contamination of bath.^1,2) This contamination problem can cause reliability problems in the devices and reduce the life time of the solution.^11,12) Therefore, an alternative plating process employing a plating solution which has lower additive concentration has been studied.^11–17) The objective of this study is to find optimum leveler concentration to fill via without using accelerator and suppressor. In this study, accelerator and suppressor were not added to reduce the concentration of organic additives. Specifically, the effect of only leveler (tetronic 701) concentration in high aspect ratio via filling were investigated.

2. Experimental Procedure

Prior to copper via filling, electrochemical deposition behaviour of copper with various concentration of tetronic 701 was studied by electrochemical methods such as galvanostatic plot, polarization plot and cyclic voltammograms. Copper rotating disk electrode (RDE) was used for the polarization plots. Area of the copper RDE was 0.1 cm², the deposition current density was fixed at 10 mA/cm², and the scan rate was fixed at 50 mV/s. The potential was scanned from 0.15 V higher than open circuit voltage (OCV) to −0.4 V (vs Ag/AgCl) cathodically. The rotational speed of RDE was fixed at 300 rpm and the plating bath was agitated by a magnetic stirrer at 300 rpm. The same copper RDE was used in cyclic voltammograms. The potential was scanned from OCV to −0.6 V (vs Ag/AgCl) cathodically and reversed to 0.1 V before returning to the initial OCV.

Two size of Si via specimens were used. One is with 10 μm in diameter, 55 μm in depth, aspect ratio (AR) 5.5 and another is with diameter 10 μm in diameter, 150 μm in depth, AR 15. Ti and Cu were sputtered on Si via for the diffusion barrier and seed layer respectively. Only one additive (tetronic 701) was selected and used in this experiment. The concentration of each components in copper plating bath was shown in Table 1. Pre-treatment of the plating pickling with 5% H₂SO₄ solution was conducted for one minute to remove copper oxide on copper seed layer. A periodic pulse current was applied for electroplating of copper. The current density at on-time was 10 mA/cm². On/off ratio was set to 1:1 during 1～10 hours plating. After electroplating, cross-section morphologies of vias were observed using FESEM.

Table 1 Bath compositions for copper via filling.

	Chemical	Concentration
Stock Solution	CuSO4 H2SO4	0.5 M 1 M
Additive	Tetronic701	0～200 ppm

3. Results and Discussion

The effects of tetronic 701 concentration on electrode potential and current density were analysed using galvanostatic and polarization method and then the results were shown in Fig. 1. In Fig. 1(a), copper deposition potentials at 10 mA/cm² were analysed with time. The addition of tetronic 701 increased the copper deposition overpotential. The deposition potential of copper in stock solution was around 0.02 V (vs Ag/AgCl), on the other hand those in tetronic 701 added solution were lowered to −0.1 V and the overpotential difference with additive concentration was less than 0.02 V. Tetronic 701 showed very strong levelling power even at low concentration. In Fig. (b), the polarization plot also showed the levelling behaviour of tetronic 701. Tetronic 701 completely suppressed the copper deposition between OCV and −0.15 V regardless of tetronic 701 concentration. Cyclic voltammetric technique was used to analyse the levelling power of the additive and the results were shown in Fig. 2. Copper was deposited immediately and the current was increased from just below OCV in stock solution. However copper was not deposited and the current did not increased until −0.18 V (vs Ag/AgCl) in tetronic 701 added solutions regardless of their concentrations. Over10 ppm of tetronic 701, cyclic voltammograms showed almost identical hysteresis loops which proved the levelling effects of the tetronic 701 even at low concentration. Thus, copper deposition was suppressed as much as 200 mV due to high overvoltage in tetronic 701 added solution.

Fig. 1

Effect of tetronic701 concentration on (a) galvanostatic plot (b) polarization plots.

Fig. 2

Cyclic voltammograms at different tetronic 701 concentrations.

Copper filling behaviour in AR 5.5 via was investigated and the cross section of via was shown in Fig. 3. In stock solution, without additive, copper deposition rate at top of via was faster than that at bottom of via. However, the deposition of copper at top of via was completely suppressed in tetronic 701 added solutions. The concentration of tetronic 701 at bottom was lower than that at top and then the copper deposition at top of via was completely suppressed due to high overpotential in copper deposition. As tetronic concentration was increased, copper deposition rate at bottom was decreased.

Fig. 3

Filling morphology in low aspect ratio via (AR5.5).

Copper filling behaviour in AR 15 via was investigated and the cross section of via was shown in Fig. 4. The low concentration of tetronic 701 cannot prevent the early closure of via entrance. The bottom of via cannot be filled in 10 and 20 ppm tetronic 701 added solution. Since the copper deposition at middle wall cannot be completely suppressed, and then via was blocked eventually. Once via was blocked the plating solution cannot reach to the bottom of via. Superfilling behaviours were observed at 30 and 40 ppm, however the growth of copper at the top of via was also observed. When 100 ppm tetronic 701 was added, the growth of copper at the top of via was completely suppressed and then superfilling behaviour was observed. The same conditions were applied to the via filling and successful superfilling without defects was obtained. The cross section of copper filled high AR via was shown in Fig. 5.

Fig. 4

Filling Morphology in high aspect ratio via (AR15).

Fig. 5

Bottom up superfilling in high aspect ratio via (AR15).

4. Conclusions

In this study, the effects of leveler concentration in high aspect ratio via filling was investigated to reduce the use of additives. Galvanostatic, polarization and cyclic voltammetric techniques were used to verify the levelling power of additive. Tetronic701 showed higher levelling effects and increase of overpotential in copper deposition was as much as 200 mV. The optimum concentration of tetronic701 was dependent on the aspect ratio. Filling morphologies changed with different tetronic701 concentrations. 10 or 20 ppm of tetronic 701 was enough to obtain defect free copper filling in AR5.5 via by controlling the plating time. However higher concentration of tetronic 701 was required for high aspect ratio via filling. In AR15 via, 100 ppm tetronic 701 was added to achieve defect free via filling.

ACKNOWLEDGEMENTS

This work is supported by 2014 Hongik University Research Fund and KITECH

REFERENCES

• 1)  M.H. Lee and J.K. Cho: J. Korean Inst. of Surf. Engr. 46 (2013) 153. 10.5695/JKISE.2013.46.4.153
• 2)  J.J. Sun, K. Kondo, T. Okamura, S.J. Oh, M. Tomisaka, H. Yonemura, M. Hoshino and K. Takahashi: J. Electrochem. Soc. 150 (2003) G355. 10.1149/1.1572154
• 3)  W.P. Dow, C.C. Li, Y.C. Su, S.P. Shen, C.C. Huang, C. Lee, B. Hsu and S. Hsu: Electrochim. Acta 54 (2009) 5894. 10.1016/j.electacta.2009.05.053
• 4)  T.P. Moffat and D. Josell: J. Electrochem. Soc. 159 (2012) D208. 10.1149/2.040204jes
• 5)  D. Josell, T.P. Moffat and D. Wheeler: J. Electrochem. Soc. 154 (2007) D208. 10.1149/1.2434684
• 6)  W.P. Dow, H.S. Huang, M.Y. Yen and H.C. Huang: J. Electrochem. Soc. 152 (2005) C425. 10.1149/1.1901670
• 7)  M. Hayase and M. Nagao: J. Electrochem. Soc. 160 (2013) D3216. 10.1149/2.036312jes
• 8)  J. Tang, Q. Zhu, Y. Zhang, X. Zhang, J. Guo and J. Shang: ECS Electrochem. Lett. 4 (2015) D28. 10.1149/2.0101509eel
• 9)  M. Tan, C. Guymon, D.R. Wheeler and J.N. Harb: J. Electrochem. Soc. 154 (2007) D78. 10.1149/1.2401057
• 10)  M.C. Kang and A. Gewirth: J. Electrochem. Soc. 150 (2003) C426. 10.1149/1.1572152
• 11)  A. Frank and A.J. Bard: J. Electrochem. Soc. 150 (2003) C244. 10.1149/1.1557081
• 12)  R. Kimizuka, H. Toda, T. Eda, K. Kishimoto, R. Oh, H. Honma and O. Takai: J. Electrochem. Soc. 162 (2015) D584. 10.1149/2.0781512jes
• 13)  T.P. Moffat, D. Wheeler, C. Witt and D. Josell: Electrochem. Solid-State Lett. 5 (2002) C110. 10.1149/1.1521290
• 14)  M.J. Kim, H.C. Kim, S. Choe, J.Y. Cho, D. Lee, I. Jung, W.S. Cho and J.J. Kim: J. Electrochem. Soc. 160 (2013) D3221. 10.1149/2.037312jes
• 15)  Y. Zhang, G. Ding, H. Wang, P. Cheng and J. Luo: ECS Electrochem. Lett. 4 (2015) D18. 10.1149/2.0111506eel
• 16)  S.H. Jang, K.S. Choi and J.H. Lee: J. Microelectron. Packag. Soc. 22 (2015) 99. 10.6117/kmeps.2015.22.4.099
• 17)  S.J. Lee, Y.J. Jang, J.H. Lee and J.P. Jung: J. Microelectron. Packag. Soc. 21 (2014) 91. 10.6117/kmeps.2014.21.4.091