Dissolutive wetting occurs when a liquid spreads over a solid surface with simultaneous dissolution of the solid into the liquid. This process is of great interest for both fundamental research and several industrial processes, an important example being soldering in microelectronics fabrication processes [1]. Several studies, performed for various liquid metal/solid metal systems, have shown that for millimetresized droplets the spreading time in dissolutive wetting ranges from a few to several hundred seconds [2–6]. This time is orders of magnitude higher than the spreading time found in liquid metal/solid metal systems with negligible miscibility, which is typically around 10 ms [7–11]. Despite the progress made over the last 10 years in the understanding of dissolutive wetting, several points remain obscure concerning both the driving force and kinetics of this type of wetting. The aim of the work reported in this paper is to contribute to this subject by performing a comparative study of spreading on the same substrate (monocrystalline Si) at 1100 C with pure copper and Si-presaturated copper. In the past, a similar attempt to study spreading in the same system in equilibrium and non-equilibrium conditions was made using Ag/Cu couple [12]. However, with the sessile drop technique used in these experiments, the initial stages of spreading were obscured by metal melting. In the present study this difficulty is overcome by using the dispensed drop technique, which enables the processes of melting and spreading to be separated (see for instance [13]) According to the Cu–Si phase diagram, at 1100 C a liquid CuSi alloy containing 52 at %Si is in equilibrium with solid Si (the solubility of Cu into solid Si is negligible) [14]. As for the surface tension of pure Cu at 1100 C, 1280 mN/m [15], it is much higher than that of molten Si, which, extrapolated to 1100 C from the melting point of Si, is close to 800 mN/m [16, 17]. As a consequence, dissolution of Si into Cu is expected to decrease the surface tension of the liquid. Wetting was studied in a metal furnace under a vacuum of (1–5) 9 10 Pa. The experiment involved heating pure Cu or the CuSi alloy (purity higher than 99.999%) in an alumina crucible placed above the Si substrate. At the experimental temperature, the liquid was extruded from the crucible through a capillary, forming droplets with a diameter ddr lying between 1.3 and 2 mm. In view of the high sensitivity of Si to oxidation and to improve surface cleaning, a prior heat treatment of the substrates was performed at 1250 C before depositing the drop at 1100 C. The wetting process was filmed by a camera (500 frames per second) connected to a computer, enabling automatic image analysis. The characteristic dimensions of the drop (drop base diameter d and visible contact angle h) were extracted with an accuracy of ±2 for h and ±2% for d. The (111) surfaces of electronic purity Si have an average roughness of 1–4 nm after polishing with diamond paste up to 0.1 lm. Figure 1 gives the temporal variation in contact angle h and the normalized drop base diameter d/ddr for the wetting of Si by a Cu droplet. Due to the resolution of 2 ms, most of the non-reactive spreading, which occurs at t \ 2 ms, is missing. However, on the drop base diameter curve, it can be clearly seen that the triple line velocity vanishes at point A, corresponding to time t & 4 ms. This is just a tendency because, after slightly receding to point B, d starts to P. Protsenko Department of Colloid Chemistry, MSU, Moscow, Russia