Relating surface structure and growth mode of gamma ' Fe4N

We have grown thin layers of gamma'Fe4N on Cu(100) substrates by molecular beam epitaxy in a flow of atomic nitrogen, delivered by a radio-frequency (RF) plasma source. This nitride phase is a ferromagnetic metallic conductor and has interesting properties for device applications. In addition it has an intriguing growth mechanism. In earlier work we found that pure crystalline layers can be grown at substrate temperatures higher than 250degreesC, with excess nitrogen and in the presence of hydrogen.(1) To gain insight into the growth mechanism, we studied the structure and composition with scanning tunneling microscopy (STM), Auger electron spectroscopy (AES), low-energy electron diffraction (LEED) and X-ray diffraction (XRD). This was done for a coverage range of Fe4N on Cu(100) between 0.5 and 30 monolayers (ML) equivalent of Fe, deposited at 400degreesC, or at 300degreesC in one case. Here a preliminary account of this study is presented. We found that at sub-ML coverage, first "depressed" (with respect to the Cu surface) islands of Fe-N are formed. Then, on top of these islands a second layer is growing. Subsequently the space between the islands is filled up by a Fe-N layer growing directly on Cu. This gives rise to a smooth surface with patches differing in height by only 0.5 A. The following layers grow by step-flow growth. The smooth terraces still show patches with a 0.5 A height difference. The phase is gamma'Fe4N with a distorted (p4g-like) structure as observed with LEED and STM, where a p(2 x 2) symmetry is seen. The c(2 x 2) symmetry expected for gamma'Fe4N is observed after growing 30 ML or more. A model for the growth mechanism based on our observations is proposed.