<p>The function of titanium- and nickel-silicides is to lower the series resistance and contact resistivity in gate, source and drain contacts of an integrated circuit transistor. </p><p>With decreasing dimensions, the low resistivity C54 TiSi₂ is not formed and stays in its high resistivity phase C49. It was found that a layer of niobium interposed between titanium and silicon, which is supposed to promote the C54 phase, led to the formation of the high resistivity C40 (Ti,Nb)Si₂ in both small and large contacts. </p><p>Increased interest in Si_1-xGe_x layers led to the inclusion of the Ni-Si-Ge system in this project. The interaction between nickel and poly-Si_0.42Ge_0.58 was found to be different from nickel and poly-silicon in the meaning of the phases formed during high temperature annealing. High-resistivity NiSi₂ was formed at 750°C, but nickel and Si_0.42Ge_0.58 formed no disilicide. A massive out-diffusion of germanium from the NiSi_1-uGe_u resulted in agglomeration at lower temperatures than for NiSi. This was ascribed to the larger enthalpy of formation for nickel reacting with silicon than with germanium. Ternary phase diagrams, with and without the disilicide phase, were calculated. According to the tie lines, NiSi_1-uGe_u will be in thermodynamic equilibrium with Si_1-xGe_x when u is smaller than x. This was confirmed experimentally, where a balanced germanium concentration in NiSi_1-uGe_u and Si_1-xGe_x, stabilized the germanosilicide. When nickel interacted with strained and relaxed silicon-germanium it was established that a strained substrate led to a morphologically unstable NiSi_1-uGe_u. The germanosilicide was highly textured on both (001) and (111) substrates. The texturing was explained by the absence of Ni(SiGe)₂ which forced NiSiGe to reorient so as not to resemble a digermanosilicide at the film/substrate interface. NiSi_0.82Ge_0.18 formed on p⁺-Si_0.82Ge_0.18 in CBKs grew laterally under the SiO₂, defining the contact hole. The contact resistivity extracted by 3D modelling was 5×10^-8 Ωcm².</p>

The function of titanium- and nickel-silicides is to lower the series resistance and contact resistivity in gate, source and drain contacts of an integrated circuit transistor. With decreasing dimensions, the low resistivity C54 TiSi2 is not formed and stays in its high resistivity phase C49. It was found that a layer of niobium interposed between titanium and silicon, which is supposed to promote the C54 phase, led to the formation of the high resistivity C40 (Ti,Nb)Si2 in both small and large contacts. Increased interest in Si1-xGex layers led to the inclusion of the Ni-Si-Ge system in this project. The interaction between nickel and poly-Si0.42Ge0.58 was found to be different from nickel and poly-silicon in the meaning of the phases formed during high temperature annealing. High-resistivity NiSi2 was formed at 750°C, but nickel and Si0.42Ge0.58 formed no disilicide. A massive out-diffusion of germanium from the NiSi1-uGeu resulted in agglomeration at lower temperatures than for NiSi. This was ascribed to the larger enthalpy of formation for nickel reacting with silicon than with germanium. Ternary phase diagrams, with and without the disilicide phase, were calculated. According to the tie lines, NiSi1-uGeu will be in thermodynamic equilibrium with Si1-xGex when u is smaller than x. This was confirmed experimentally, where a balanced germanium concentration in NiSi1-uGeu and Si1-xGex, stabilized the germanosilicide. When nickel interacted with strained and relaxed silicon-germanium it was established that a strained substrate led to a morphologically unstable NiSi1-uGeu. The germanosilicide was highly textured on both (001) and (111) substrates. The texturing was explained by the absence of Ni(SiGe)2 which forced NiSiGe to reorient so as not to resemble a digermanosilicide at the film/substrate interface. NiSi0.82Ge0.18 formed on p+-Si0.82Ge0.18 in CBKs grew laterally under the SiO2, defining the contact hole. The contact resistivity extracted by 3D modelling was 5×10-8 Ωcm2.