On the nucleation of PdSi and NiSi2 during the ternary Ni(Pd)/Si(100) reaction

Abstract

During the solid phase reaction of a Ni(Pd) alloy with Si(100), phase separation of binary Ni-and Pd-silicides occurs. The PdSi monosilicide nucleates at temperatures significantly below the widely accepted nucleation temperature of the binary system. The decrease in nucleation temperature originates from the presence of the isomorphous NiSi, lowering the interface energy for PdSi nucleation. Despite the mutual solubility of NiSi and PdSi, the two binaries coexist in a temperature window of 100 C. Only above 700 C a Ni 1-x Pd x Si solid solution is formed, which in turn postpones the NiSi 2 formation to a higher temperature due to entropy of mixing. Our findings highlight the overall importance of the interface energy for nucleation in ternary systems. V C 2013 AIP Publishing LLC.

Effect of Co, Pd and Pt ultra-thin films on the Ni-silicide formation: investigating the sandwich configuration

“…Although the Pd distribution in the sandwich structure (Ni/Pd/Ni/Si) presented in this work (Fig. 12) looks quite similar to the ones reported in in the case of an alloy system (Ni(Pd)/Si) [26,41], still some aspects of the Pd redistribution are different. Our APT results show also that the Pd atoms diffuse through the δ-Ni2Si silicide mainly at grain boundaries but do not penetrate group: Pnma) with a small lattice parameters mismatch [45], and are thus fully miscible [42].…”

“…On the contrary, Schauwen et al [41] have observed by APT the presence of Pd at the grain boundaries of NiSi. This apparent controversy may be linked to the temperatures where the observations were made.…”

“…Concerning the location of the barrier at the Ni/δ-Ni2Si interface (Fig. 5.b), previous work using in-situ RBS have shown that during the reaction of Ni(Pd) alloy film with Si, Pd atoms are pushed back in the metal alloy during the growth of the Ni 2 Si[26,41]. As a result, a Pd enriched layer is observed between the metal and the silicide and ends at the sample surface when the δ-Ni2Si formation is finished.…”

Phase formation and texture of thin nickel germanides on Ge(001) and Ge(111)

“…As the Ni surface energy is only slightly smaller than that of Ge, the Ni peak is superimposed on the plateau resulting from the "infinitely" thick Ge substrate. Therefore, phase transformations are harder to distinguish from these raw data compared to a case where the signals are separated (e.g., for Ni on Si 16,18,35 ) and the individual spectra are much more difficult to analyse. Second, the ramp rate used in these experiments (2 C/min) is much slower than the one used for the in situ XRD measurements (1 C/s).…”

“…Applying this technique in situ during an anneal allows one to accurately probe the atomic diffusion process and the thickness evolution of the growing and shrinking phases during the phase formation sequence that results from the heat treatment. [16][17][18] The in situ RBS measurements were performed in a vacuum chamber with a pressure below 10 À7 mbar using a two-stage ramp anneal. First, a fast ramp rate of 20 C/min to about 110 C was used to reduce the measuring time prior to the onset of the solid-phase reaction.…”