Cu-N Films grown by reactive MSIP: Constitution, structure and morphology

Abstract

Abstract. Cu-N layers were deposited on Si-(100) wafers and amorphous carbon coated platinum nets at low temperatures (< 100 ~ by means of Reactive Magnetron Sputtering Ion Plating (MSIP). For this, a Cu-target was sputtered in rf-mode in a nitrogen plasma, and the influence of the parameters "sputtering power" and "nitrogen partial pressure" on composition, structure, texture and morphology of the Cu-N layers was investigated. The analysis with EPMA, XPS, HEED, XRD, TEM and SEM yielded the following results: With appropriate settings of the process parameters the nitrogen content of the films could be controlled-in this investigation between 0 to 28 at-%. The HEED analysis shows, that at the near surface region films with a nitrogen content in the range of 16-28 at-% were grown with DO 9 structure of Cu3N. The films were textured: [100] in the case of 16-26.3 at-% nitrogen, [210] + [110] at 28 at-%. XRD analysis of the crystal structure of the bulk of a nitrogen-rich film (26.3 at-% nitrogen) confirm the result of HEED analysis. In the bulk of the nitrogen-poor film (16.4 at-~ two phases were detected, metallic copper with A1 structure and Cu3N with DO 9 structure. XPS analysis revealed that all films were oxidized at the surface to CuzO.Key words: copper nitride, Cu3N, magnetron sputtering ion plating. In this work an account is given on the influence of the process parameters on the composition, structure, texture and morphology of Cu-N films deposited by reactive rf-magnetron sputtering ion plating. Compounds of the system Experimental Coating ProcessCu-N coatings of different composition were deposited in an HV stainless-steel triple cathode reactive rf-magnetron sputtering ion plating system (type L560, Leybold AG). The experimental conditions were chosen as follows.MSIP system. Pumping speed: 1500 l/s; residual gas pressure: < 10 .4 Pa; magnetron type: PK90; target material: Cu (99.99% pure); flow controlled gas inlet (0-100 sccm) for nitrogen (99.999% pure) and argon (99.96% pure); partial presssure measurement by Electron Impact Emission Spectroscopy (EIES); angle between the normal of target and substrate: 30~ distance of centre of target to centre of substrate: 70 mm; substrate heating by radiation. Table 1.Characterization of films by EPMA, XPS, HEED, XRD, TEM and SEM. The composition of the Cu-N coatings was determined by electron probe X-ray microanalysis (EPMA). The measurements were carried out with a Camebax SX50 microprobe analyser equipped with four wavelength dispersive spectrometers. The X-ray signals were calibrated on the following standards: pure freshly polished copper for Cu L~, Fe20 3 for O Kc~ and Fe4N for N K~. To take into consideration possible peak-shape changes in the Cu-L spectra, the intensity was measured by peak integration over the energy range 920-940 eV. For correction of matrix effects the quantitation programm PAP was applied 1-14]. All analyses were carried out at

Design and characterization of a metal-ion beam and nitrogen-radical beam source PVD system with in-situ AES constitution and HEED structure analysis

“…The copper crystals grew with the same [100] texture as the crystals of the phase Cu 3 N. With the assumption that nitrogen is bound exclusively as Cu 3 N, i.e. that no nitrogen is dissolved in the metallic appearing copper and the Cu 3 N, it follows from the AES analysis, that the major component in the near-surface region of the growing layer at a Cu-ion energy of 1-8 keV is Cu 3 N. This is surprising, since it was established in previous investigations [13] that Cu 3 N in the near-surface re- gion was virtually completely reduced by Ar-ion (3 keV) bombardment, see Fig. 7.…”

“…A possible reason for this could be that the nitridic compounds of copper have a great affinity for oxygen. Sputter cleaning is not possible without reduction of the near-surface region of the layer, due to the lack of stability [13].…”

“…Compounds of the system Cu-N are mainly produced either by copper nitriding (reaction of NH 3 with copper to Cu 3 N, CuN 3 and Cu(N 3 ) 2 [3][4][5]), or by implanting nitrogen ions in copper, forming Cu(N 3 ) 2 [6][7][8][9]. Cu 3 N was also prepared by reactive sputtering of a Cu-target in an Ar/N 2 or pure N 2 atmosphere [10][11][12][13]. As shown by [14], a reaction between copper and molecular nitrogen does not take place below 900°C.…”

The Effect of Nitrogen on the Formation of Nanocrystalline Copper Thin Films

“…8). Other authors indicate decomposition for Cu 3 N from 373-673 K. [22][23][24] Some authors 15 advocate the high energy density of the ion beam used in electron microscopy as the reason for the failure to detect nitrogen as an element or as a compound. Nevertheless, this is not a straightforward explanation, because before the thin films were analyzed in the TEM they were thinned by ion milling, and this process could contribute to the heating of the thin foil and allow the nitrogen to escape from the copper thin film.…”

Electrical properties and memory applications of Al-rich Al-based dielectric thin films

“…The film properties can be tailored between those of pure aluminum oxide (AI 203 ) and aluminum nitride (AIN), depending on different applications. The synthesis of AIOxN y films has been commonly reported using physical vapor deposition, such as ion-beamassisted evaporation and sputtering method [166,167,169,171]. As shown in some studies [170,173]…”