Modelling the strain build-up in nitrogen implanted tungsten films on silicon substrates

A model of the strain accumulation in nitrogen implanted polycrystalline tungsten films deposited on 110 single-crystalline silicon substrates is presented. The films, with the nominal thickness of 500 nm, were implanted at room temperature, in a direction perpendicular to the sample surface with an energy of 60 keV N2+ ions and concentrations between 0.5 × 1017/cm2 and 5 × 1017/cm2. The proposed model is based on the simulations of the X-ray diffraction 2θ-ω scans in the vicinities of the 110 W most intense Bragg peak. Up to a fluence of 2 × 1017/cm2, the increase of asymmetry towards lower angles of the 110 W peak is interpreted as a linear increase in strain affecting the entire implanted region. Then, the solubility of N in W saturates and phase transition to fcc β-W2N is observed. The phase transition occurs at the surface strain regions of the implanted crystal volume while the lowest strained regions are masked by the β-W2N barrier and even evidence some relaxation. After 3 × 1017/cm2, the strain deduced for the 111 and 200 planes of the fcc β-W2N formed pseudo-layer increases at different rates. While the former comprises a maximum tensile deformation of 3.8 %, the latter is more resilient to nitrogen irradiation developing less than half of the strain magnitude (1.5 %). After 4 × 1017/cm2, Rutherford backscattering spectrometry measurements suggest partial sputtered N while X-ray diffraction indicates no strain-relief at the remaining surface fcc β-W2N. The N concentration derived by the ion beam technique and the combined 110 W/ (111, 200) β-W2N strain profiles as functions of depth agree with the Simulations of Ions and Radiation in Matter theoretical predictions.