**1. Introduction**

206 Mass Transfer - Advanced Aspects

Yakhot, V. & Orszag, A.S. (1986). *Renormalization group analysis of turbulence–Basic theory,* J.

Wilcox, D.C. (1998). *Turbulence Modeling for CFD*, DCW industries. La Canada, California.

Scientific Computing. Vol. 1(1), pp. 3–51.

Electroplating and electroforming are the two electrochemical processes extensively used in metal fabrication. Electroplating provides a thin metal film to bestow the surface with desired property such as abrasion and wear resistance, corrosion protection, lubricity and aesthetic qualities; electroforming leads to a deposition of metal skin onto a mandrel which is then removed and then the metal deposit was thickened to obtain precise fabrication of molds. Both the electrochemical processes are carried out in the bath where sufficient concentration of metal salt is supplied in presence of an electric field. The electrochemical kinetics is determined not only by the strength of electric field but also by the mass transport phenomenon of the electrochemical active ions. The electric field employed in the electroplating is relatively lower and the field distribution is homogeneous. In contrast, the electrical field exerted in electroforming seems to be much stronger and the field distribution becomes less homogeneous.

In 1995, a novel localized electrochemical deposition (LECD) process was pioneered by Hunter [1] to fabricate three-dimensional (3D) metal microstructures. The LECD brings the electrochemical process to a new era. However, in the LECD process, the electrical field exerted at the electroplating site is super high and the distribution of field strength is ultra heterogeneous. The phenomenon of mass transport in such a strong field distributed in extremely heterogeneous is the case which we have never encountered in doing usual electrodeposition. In the process of micro electroplating, the site where LECD taking place was experimentally controlled to along the track guided with a microanode. Accordingly, the micro metallic features could be fabricated electrochemically along the motional track guided by the microanode [2]. Due to this fact, LECD was also named as microanode guided electroplating (MAGE) process. The schematic diagram of MAGE is shown in Fig. 1.

A platinum wire (diameter in the range from 25 to 125 μm) was fixed coaxially, and cold mounted with epoxy resin in polymethylmethacrylate (PMMA) tube (inner and outer diameters are 3 and 5 mm, respectively) to expose a disk (25 ~ 125 μm in diameter) acting as the microanode. The micranode was driven to move by a stepping motor in an electroplating bath thus guiding the micro electroplating way according to the program built in the micro-CPU. The micoanode assembly and microanode was driven to move by a

Mass Transfer Within the Location Where Micro Electroplating Takes Place 209

S3500, Hitachi Co.) were employed to observe their surface morphology. Figure 2 depicts the OM of the micro columns fabricated under certain conditions with different motion modes of the micro anode. As the micro anode was driven to ascend continuously at a constant rate of 1.8 μm s-1 to perform MAGE at 5.0 V, a column appearing in dendrite was formed (Fig. 2(a)). In contrast, if the micro anode was driven to ascend intermittently (with an initial gap of 10 μm within each intermittent cycle) at the same voltage (i.e., 5.0 V) until reaching certain heights, a column revealing periodical nodes (Fig. 2(b)) was established. Comparing with both the columns, we found that they showed a similar diameter (roughly 50 μm), the dendrite tended to decrease the diameter with increasing its height, as shown in Fig. 2(a); however, the nodal one, depicted in Fig. 2(b), tended to vary the diameter periodically with the height. If we conducted the intermittent MAGE under a lower bias (i.e., at 3.5 V), we obtain a micrometer nickel column in uniform diameter (50 μm) with

(a)

Fig. 2. Optical micrographs (OM) of the nickel micrometer columns fabricated by the MAGE process in a sulfate bath where the initial gap between the electrodes was the same (at 10 μm) but their dc-voltage bias and the motion mode of the microanode changed as follows: (a) bias at 5.0 V and the microanode moved continuously at a rate of 1.8 μm s-1, (b) bias at 5.0 V and the microanode moved intermittently and (c) bias at 3.5 V and the microanode moved

Figure 3(a) exhibits SEM morphologies of a nickel column consisting of two segments due to change of voltages in the intermittent MAGE. With respect to fabricating the lower segment (i.e., 3(b) in Fig. 3(a)), we conducted MAGE at 3.2 V until reaching a height of 500 μm. Then we switched the voltage to 4.0 V to continue the MAGE process to grow the upper segment

(b) (c)

smooth morphology (as shown in Fig. 2(c)).

intermittently

stepping motor under precise control, as shown in Fig. 1. In Fig. 1, the cathode was placed horizontally in the electrolytic cell (F) and connected with the negative pole of the dc power supply (V). The microanode assembly (H), connected with the positive pole of the power supply (V), was vertically fixed on a one-dimensional moving table. A servo microstepping motor (M) was used to drive the table through a micro-CPU (C) via the D/A converter (D) and driver (E). A relay (G) was connected with the anode assembly. Through control with dedicated software, the microanode was moved vertically with a resolution of 20 nm per step. Prior to electroplating, intimate contact between the microanode and the cathode was assured through a measurement of null electrical resistance. The microanode was then lifted from the cathode to a variety of gaps (in the range from 1 to 100 μm) to start the MAGE. In this study, a variety of dc-voltage biases (in the range from 3 to 6 V) were employed to conduct MAGE and their corresponding current was monitored with the current sensor (A).

So far we have publish a few papers [2-11] to discuss the electrochemical kinetics with respect to MAGE process. The heterogeneous distribution of very intensive electric field in local sites was determined significantly by experimental parameters such as motion modes of the microanode, applied electric voltage, initial gap between the cathode and microanode, and etc. In the present work, we concentrate ourselves on mass balance of electrochemical active ions those which supplied via mass transport from the bulk solution and to be consumed to turn into metallic micro feature. In terms of various models, we applied the commercial software ANSYS 8.0 to simulate the systems so as to understand the electrochemical mechanism of the MAGE process.

Fig. 1. Schematic diagram of the microanode-guide electroplating system in which the capital letters denote the following. (A) Current sensor, (B) A/D converter, (C) micro-CPU, (D) D/A converter, (E) driver, (F) cell, (G) relay, (H) anode,(M) micro-stepping motor and (V) voltage source
