In-Situ Observation of Electrodeposition in a Microfluidic Device with through Silicon Via Model

Abstract

Nowadays, the high performance of integrated circuits is due to packaging technologies, and 3d-packaging using TSVs (Through Silicon Vias) have entered use in commercial products. Copper electroplating is used to form TSVs, since bottom-up filling is available using a combination of several additives, but the filling mechanism is not clear yet. Recently, we have observed the adsorption and deactivation behavior of typical suppressing additives by using a disk electrode with a diameter of 100 μm in a microchannel that allows precise control of mass transport [1]. In this study, in order to understand the details of the TSV filling behavior, we propose a microfluidic device that has TSV feature in the channel, and in-situ observation of the electrodeposition behavior is enabled. The microfluidic device was fabricated using the silicon substrate. First, Au is sputtered on the silicon substrate and the Au film is patterned by photolithography and wet etching. Then, the micro channel is defined by a 19 μm thick photoresist pattern on the silicon substrate. In the micro channel, a rectangular area on Au film is exposed. The exposed Au film and the surrounding resist pattern model the TSV. The width, height, and depth of the via are 20 μm, 19 μm and 100 μm, respectively. Finally, a glass cover plate is put on the substrate, and the microchannel is mounted on a base. A mercury sulfate electrode (MSE) is connected to the resin base and works as a reference electrode. A copper wire is inserted into the outlet tube and works as a counter electrode. An acid copper sulfate plating solution was prepared as a VMS (Virgin Makeup Solution). Initially, additive-free VMS was supplied, and the working electrode was pre-plated with copper. Then, the solution was switched to the VMS + Leveler and open circuit potential was maintained for 120 s. Finally, -600 mV vs. MSE was applied on the working electrode, and plating behavior was observed. Figure 1 shows the appearance of the working electrode along time. At 5 s, clear interface was observed at the depth around 20 μm from the opening of the via. Below the interface, the working electrode became dark because the deposition proceeded. While no obvious change was observed above the interface and it was found that the deposition was strongly suppressed by the leveler adsorption. The interface line inclined and it was assumed that the leveler penetrated deeper in downstream side because of the solution flow. Reference [1] T. Akita, M. Tomie, R. Ikuta, H. Egoshi, and M. Hayase, J. Electrochem. Soc., 166(1), D3058 (2019). Figure 1