The tin is (and probably it will rest) one of the most important constituents of the solders, while titanium additions to these materials are prospective for improving their wetting properties, for example. From the other side Si chips are largely used in the electronic devises and solar cells. That is why the interactions of Si with both metals are of fundamental and practical interest. However the system Ti-Sn-Si has been studied partially only [1] though there is some data concerning the corresponding binary diagrams (Ti-Sn, Ti-Si and Sn-Si) [2].
This work is also part of a series of studies on the titanium interactions with some low melting elements (Sn, Zn, Bi) where the laboratory syntheses of Ti containing alloys are done using quartz (SiO2) tubes. Nevertheless the standard Gibbs free energy of the TiO2 has, at the applied temperatures, larger negative values than that of the SiO2 [5]. Consequently the pure titanium might act as reducing agent on the silicon dioxide, thus leading to changes of the specimens’ compositions. In the tubes this reaction usually is slow due to kinetic reasons, but when SiO2 is introduced in the bulk of an alloy the pertinent chemical interactions would be easier to occur. Thus the purpose of the present work is to study the reactions between the Ti and/or titanium-tin alloys with SiO2 as well as to obtain data on the phases formed in the Ti-Sn-Si system.
Pure (5N) bulk titanium (obtained by chemical vapor transport deposition) and tin pellets (p.a.) have been used for the alloys syntheses. Pieces of titanium have been cut and cleaned just before closing under vacuum (10-5 -10-6 Tore) the quartz tubes containing weighted amounts. After thermal treating, in many cases the titanium pieces stuck to quartz walls, making difficult to separate them even after breaking the tubes. Sometimes the tubes have been broken during the cooling cycle either. We observed also that the reaction between the titanium and the tin at relatively low temperatures (e.g. up to about 600 OC) is sluggish (probably due to oxide layers forming on their surfaces).
After the first heating-cooling cycle, small amounts of quartz stuck to titanium, have been left and two specimens sealed again under vacuum.
Optical and scanning electron microscopy (SEM) studies have been performed on polished specimens etched with alcoholic solutions of HNO3. The summarized experimental and literature data is represented in Table 2.
It is interesting to notice that the equilibrium phases Ti6Sn5 and Ti2Sn do not appear in this sample. This fact is obviously due to kinetic reasons [3]. We would like to emphasize that no titanium has been found in the liquid phase (Sn), thus the liquidus line of the binary Ti-Sn phase diagram [3] should be corrected in the Sn-rich region.
In the sample No 2 two equilibrium phases Ti2Sn3 and Liq (Sn) have been observed only. The phase amounts calculated using the lever rule and accepting the lack of Ti solubility in liquid phase (denoted further as Liq), are in agreement with these measured using quantitative metallographic analysis (i.e. 75 % Ti2Sn3 and 25 % Liq).
In Fig. 1 (specimen No 3) six phases can be observed, i.e. more than the number allowed at thermodynamic equilibrium. An intermediate layer of Ti5Si4 has grown between SiO2 and a(Ti,Sn) particles. Regions of Ti3Sn disposed between the a(Ti,Sn) and the liquid phase (Sn) while crystals of the formerly unknown ternary compound with approx. formula Ti7SiSn4 are surrounded by the molten Sn. We suggest that the system has not reached a full thermodynamic equilibrium but a local one only. Nevertheless the existence of a formerly unknown compound Ti7SiSn4 can be supposed because its well-shaped crystals have been largely observed.
References
[1] Bulanova M, Soroka A, Zheltov P, Vereshchaka V, Meleshevich K, Phase equilibria in the Ti-rich corner of the Ti-Si-Sn system, Z. Metallkd. 1999, 90(7), 505-507.
[2] Massalski T, CD ROM: Binary Alloy Phase Diagrams, ASM International, OH, USA, 1996.
[3] Kuper C, Peng W, Pisch A, Goesmann F, Schmid-Fetzer R, Phase Formation and Reaction Kinetics in the System Ti-Sn, Z. Metallkd., 1998, 89 (12) 855-862.
[4] Pondarevskaya O.V., Petrenko O, Sudavtzova V, Lisnyak V, Stus N, Crystallization and thermodynamic properties of titanium stannides, Journal of Thermal Analysis and Calorimetry, 2002, 67 (3) 649-657.
[5] R. Swalin, Thermodynamics of Solids, John Wiley &Sons, N.Y.-London, 1961.