This paper is devoted to the characterization of point defects arising in mercury cadmium telluride, Hg1-xCdxTe, at different stages of the device manufacturing process. A research program was focused on the study of: (i) point defects in an as-grown material; (ii) atomic diffusion via vacancy sites during annealing at elevated temperatures; and (iii) defect transport accompanying formation of n-p junctions produced by boron implantation. We used, as a main tool, high-resolution x-ray diffraction (HRXRD) in triple axis and Bond modes of measurement. High-resolution scanning electron microscopy (HRSEM), energy dispersive spectroscopy (EDS) in SEM, secondary ion mass spectrometry (SIMS), atomic force microscopy (AFM), optical (Fourier transform infrared transmission - FTIR), and electrical measurements served as complementary techniques for advanced sample characterization. Clusters of point defects (down to 10 nm in diameter) were observed in HRSEM images taken under high magnification (x400K) from as-grown Hg1-xCdxTe layers. It was found that these clusters lead to the swelling or contraction (depending on the defect type: Te-antisites or Ilgvacancies, respectively) of the crystalline lattice, which influences the diffraction peak position in the HRXRD spectra. Based on these findings, a novel monitoring method of the Hg1-xCdxTe epilayers was developed, which combines precise x-ray measurement of the absolute values of lattice parameters (in Bond mode) and determination of the Cd-content, x, by FTIR. We used HRXRD to study atomic diffusion in Hg1-xCdxTe at elevated temperatures. A new simulation routine was developed in order to calculate the HRXRD spectra affected by the diffused atoms. Successful fittings of the measured and the simulated HRXRD spectra allowed us to determine (in a non-destructive manner) the temperature-dependent Hg diffusion coefficients and the activation energy of the jumping diffusion via vacancy sites in the Hg1-xCdxTe lattice. HRXRD measurements in boron-implanted Hg1-xCdxTe showed presence of an additional peak in the diffraction spectra, which originated in the interstitials-rich damaged layer. Angular peak positions in samples with various Cd contents were used to study post-implantation dynamics of interstitials. A percolation problem in the migration of the Cd interstitials was discovered in specimens with x < x(c) (x(c) = 0.265 is the percolation threshold). The percolation phenomenon is explained in terms of migration of the dumbbell-like defects through the fee-sub-lattices of Hg1-xCdxTe. Due to the percolation problem, we observed a completely different atomic transport below and above x, which resulted in pronounced concentration-dependent effect on surface recovery in implanted samples, visualized by HRSEM. Percolation problem in samples with x < x, seems to be responsible (at least partially) for the limited mobility of implanted boron and for the difficulties in boron activation in Hg1-xCdxTe-based devices for the 8-12 mu m atmospheric transparency window.
