In-Situ UHV Tem Investigations of the Initial Oxidation Stage of Copper Thin Films

  • PDF / 1,824,548 Bytes
  • 6 Pages / 414.72 x 648 pts Page_size
  • 112 Downloads / 214 Views

DOWNLOAD

REPORT


initial oxidation stage. We have developed a semi-quantitative model to explain our observation of the initial oxidation stage of Cu where oxygen surface diffusion is a dominant process for transport and nucleation [2] and oxygen surface diffusion plus direct impingement describe the initial growth of Cu 20 [3]. We now wish to understand the processes of the nucleation. Venables et al. have extensively modeled the initial nucleation processes, such as nucleation, surface diffusion, and desorption (Figure 1) [1, 4, 5]. This model has been used to describe metal heteroepitaxy[6], where comparison with the rate theory gives quantitative information, such as the activation energy of surface diffusion. As yet, there are no controlled experiments of a chemical reaction and the comparison to nucleation rate theory. To obtain a unique solution to nucleation rate theory, it is necessary to examine all possible regimes of temperature and pressure. In particular, nucleation rate theory makes strong predications concerning the asymptotes of nucleation, i.e. the initial nucleation regime and the final saturation density. We have examined the initial nucleation stages of copper under various temperature and pressure conditions and compared our experimental data of the asymptotes to the predictions from nucleation rate theory. We found qualitative agreement, but not quantitative agreement. Copper was chosen as a model metal system, since it is a simple face-centered cubic metal, with a lattice parameter a = 3.61A and a melting temperature of 1083°C [7]. Copper is also of interest because of its low resistivity and good electromigration properties, as compared to aluminum, making it an attractive interconnect material for ULSI (ultra-large scale integrated circuits) [8]. Two types of oxides, CuO and Cu 2O, have been reported to form on copper (001) due to oxidation [9]. CuO has monocinic structure (space group C2/6, a=4.662A, b=3.416A, c=5.118A, 3=99.490) [10] and Cu 20 is cubic, with space group Pn-3m (a=4.217A) [10]. 557 Mat. Res. Soc. Symp. Proc. Vol. 481 ©1998 Materials Research Society

R = deposition rate

nceoE E

l•IE

desorption

nucleation diffusion

Figure 1. Schematic diagram of the various processes during nucleation as described by nucleation rate theory.

EXPERIMENT Single crystal 99.999% pure 1000A copper films were grown on irradiated NaCl, which was cleaved alony the (001) plane, in an UHV e-beam evaporator system, where the base pressure was 10-' torr. The conditions of the evaporation within the e-beam evaporator system were the following: deposition rate was 5A/sec and substrate temperature was 250'C. The copper film was removed from the substrate by floatation in deionized water, washed and mounted on a specially prepared Si mount. The Si mount and the microscope specimen holder allow for resistive heating of the specimen up to 1000'C.

The microscope used for this experiment is a modified JEOL200CX. The spatial resolution of the JEOL200CX is 2.5A. The modifications permit the introduction of gases directly