• Korean Journal of Metals and Materials
• /
• v.50 no.2
• /
• pp.152-158
• /
• 2012

The effect of current waveform on Cu filling into TSV (through-silicon via) and the bottom-up ratio of Cu were investigated for three dimensional (3D) Si chip stacking. The TSV was prepared on an Si wafer by DRIE (deep reactive ion etching); and its diameter and depth were 30 and $60{\mu}m$, respectively. $SiO_2$, Ti and Au layers were coated as functional layers on the via wall. The current waveform was varied like a pulse, PPR (periodic pulse reverse) and 3-step PPR. As experimental results, the bottom-up ratio by the pulsed current decreased with increasing current density, and showed a value of 0.38 on average. The bottom-up ratio by the PPR current showed a value of 1.4 at a current density of $-5.85mA/cm^2$, and a value of 0.91 on average. The bottom-up ratio by the 3-step PPR current increased from 1.73 to 5.88 with time. The Cu filling by the 3-step PPR demonstrated a typical bottom-up filling, and gave a sound filling in a short time.

• Korean Journal of Metals and Materials
• /
• v.49 no.5
• /
• pp.388-394
• /
• 2011

High speed copper filling into TSV (through-silicon-via) for three dimensional stacking of Si chips was investigated. For this study, a tapered via was prepared on a Si wafer by the DRIE (deep reactive ion etching) process. The via had a diameter of 37${\mu}m$ at the via opening, and 32${\mu}m$ at the via bottom, respectively and a depth of 70${\mu}m$. $SiO_2$, Ti, and Au layers were coated as functional layers on the via wall. In order to increase the filling ratio of Cu into the via, a PPR (periodic pulse reverse) wave current was applied to the Si chip during electroplating, and a PR (pulse reverse) wave current was applied for comparison. After Cu filling, the cross sections of the vias was observed by FE-SEM (field emission scanning electron microscopy). The experimental results show that the tapered via was filled to 100% at -5.85 mA/$cm^2$ for 60 min of plating by PPR wave current. The filling ratio into the tapered via by the PPR current was 2.5 times higher than that of a straight via by PR current. The tapered via by the PPR electroplating process was confirmed to be effective to fill the TSV in a short time.

• Corrosion Science and Technology
• /
• v.9 no.6
• /
• pp.276-280
• /
• 2010

Copper(Cu) as an interconnecting metal layer can replace aluminum (Al) in IC fabrication since Cu has low electrical resistivity, showing high immunity to electromigration compared to Al. However, it is very difficult for copper to be patterned by the dry etching processes. The chemical mechanical polishing (CMP) process has been introduced and widely used as the mainstream patterning technique for Cu in the fabrication of deep submicron integrated circuits in light of its capability to reduce surface roughness. But this process leaves a large amount of residues on the wafer surface, which must be removed by the post-CMP cleaning processes. Copper corrosion is one of the critical issues for the copper metallization process. Thus, in order to understand the copper corrosion problems in post-CMP cleaning solutions and study the effects of DC biases and post-CMP cleaning solution concentrations on the Cu film, a constant voltage was supplied at various concentrations, and then the output currents were measured and recorded with time. Most of the cases, the current was steadily decreased (i.e. resistance was increased by the oxidation). In the lowest concentration case only, the current was steadily increased with the scarce fluctuations. The higher the constant supplied DC voltage values, the higher the initial output current and the saturated current values. However the time to be taken for it to be saturated was almost the same for all the DC supplied voltage values. It was indicated that the oxide formation was not dependent on the supplied voltage values and 1 V was more than enough to form the oxide. With applied voltages lower than 3 V combined with any concentration, the perforation through the oxide film rarely took place due to the insufficient driving force (voltage) and the copper oxidation ceased. However, with the voltage higher than 3 V, the copper ions were started to diffuse out through the oxide film and thus made pores to be formed on the oxide surface, causing the current to increase and a part of the exposed copper film inside the pores gets back to be oxidized and the rest of it was remained without any further oxidation, causing the current back to decrease a little bit. With increasing the applied DC bias value, the shorter time to be taken for copper ions to be diffused out through the copper oxide film. From the discussions above, it could be concluded that the oxide film was formed and grown by the copper ion diffusion first and then the reaction with any oxidant in the post-CMP cleaning solution.

• Journal of the korean academy of Pediatric Dentistry
• /
• v.36 no.1
• /
• pp.30-37
• /
• 2009

Early enamel caries is commonly remineralized by the patient‘s improved oral hygiene or fluoridation, however the result is clinically unreliable. As an alternative, we tried to seal the lesions with low-viscosity light-curing resin. The aim of the present study was to search the proper methods of the adequate pretreatment prior to applying adhesive resin on natural proximal caries lesions. Thirty nine extracted deciduous molar teeth showing proximal early caries lesion were used for this study. They were divided into 5 groups : Group 1; only carefully cleaned with water, group 2; etched with 15% HCl for 15s, group 3; etched with 35% phosphoric acid for 15s, group 4; etched with 35% phosphoric acid for 30s, and group 5; cleaned with 0.5% NaOCl. Following results were obtained by evaluating with SEM and CLSM after applied with adhesive resin. 1. As a result of SEM evaluation, group 2 showed clearly removed surface layer, group 3,4 showed partially removed surface layer irregularly, group 5 showed slightly removed surface layer. 2. Group 2 showed the deepest infiltration depth, followed by group 4, group 3, group 5, group 1 and besides group 5, other groups showed significantly deep infiltration depth. (p < 0.01) In conclusion, the best methods of the adequate pretreatment on natural proximal caries lesion for deep infiltration of adhesive resin was to etch with 15% HCl for 15s.