Physical Deposition of YBCO

An important deposition technique for a variety of materials involves the use of an electron beam to vaporize metallic atoms from a solid ingot. This study considers the deposition of super-conducting films of YBa2Cu3O7-x that are of great technological interest in numerous applications. The background pressure in an experimental deposition chamber is maintained at 2~5E-5 Torr using vacuum pumps. Atoms are vaporized from yttrium, barium and copper ingots by a high-energy electron beam, and the vapor jets proceed through the low-pressure chamber toward a deposition substrate. A critical component for the process is the manner in which the atoms are transported from the ingot molten surfaces to the substrate. There is a need to understand in detail the gas dynamics of the expansion process. The direct simulation Monte Carlo (DSMC) method has been used to simulate the three dimensional gas dynamic process. Besides the translational energy mode, the atomic electronic energy is taken into account. Some important issues such as the atomic collision cross-sections for metal vapors and hyperfine electronic structure of atomic absorption spectra are addressed.

Figure 1 shows the total number density field given by DSMC calculation. The source fluxes (mol cm-2s-1) are Y: Ba: Cu=0.84: 1.86: 2.52. The atoms that vaporize from the surfaces of the yttrium, barium and copper ingots, at the bottom of the figure, quickly undergo expansion in the deposition chamber that results in a rapid decrease of the number density. The flow is therefore brought into the non-continuum regime that needs to be studied based on the kinetic viewpoint.


Figure 2 Comparison of deposition thickness distributions along the symmetrical line on the substrate for cases where only the yttrium source is evaporating at rates 8.4E-7 mol cm-2s-1(left) and 1.1E-5 mol cm-2s-1 (right), respectively.

Figures 2 compares DSMC deposition thickness profiles along the symmetrical line of the substrate with measured data for two cases where only the yttrium source is evaporating. The source fluxes are 8.4E-7 and 1.1E-5 mol cm-2s-1, respectively, and the corresponding experimental deposition time are 30 and 12 minutes. The DSMC profiles agree very well with the measurement. Because of few collisions between the atoms for the case with the low evaporation rate, the collision-less distribution is close to the DSMC and measured results. For the high evaporation rate case, however, the collision-less distribution is much lower than the DSMC and measured results. The collisions impede to some extend atomic diffusion to the block plates and chamber wall that are assumed to be perfectly sticking, which is well satisfied because the plate and wall temperatures are low in the study. Therefore the possibility of the atoms to transport from the vaporized ingot surface to the substrate is increased by the atomic collisions.


Figure 3 Comparison of DSMC and measured atomic absorption spectra for yttrium at central frequencies of 668nm (left) and 679nm (right), respectively.

Figure 3 compares DSMC and measured atomic absorption spectra at two frequencies along an aperture close to the substrate symmetrical line for the high evaporation rate case of pure yttrium, i.e. only the yttrium source evaporates at rate of 1.1E-5 mol cm-2s-1. Because of the different hyperfine electronic structures at the central frequency of 668nm and 679nm, the peak structures are inclined to the right and left, respectively. The DSMC and measured Doppler widths and peak structure details are in excellent agreement.

Investigators

Jing Fan, Postdoctoral Associate
Iain D. Boyd, Associate Professor

 

Acknowledgments

This work was developed as part of the Office of Naval Research/3M "Models, Sensors, and Controls for E-Beam Deposition" program, Agreement No. N00014-98-3-0015. The content does not necessarily reflect the position or policy of the Government and no official endorsement should be inferred.

 

Recent Publications