In-Situ Observation of Stress in Cu/Pd Multilayers
- PDF / 441,412 Bytes
- 6 Pages / 414.72 x 648 pts Page_size
- 39 Downloads / 243 Views
EXPERIMENT The samples were grown and the stress behavior observed in a sputter deposition chamber equiped with three independently-shuttered magnetron sources and a laser-based substrate curvature system. The base pressure of the system was 5 x 10-9 Torr, and deposition rates 279 Mat. Res. Soc. Symp. Proc. Vol. 382 01995 Materials Research Society
were monitored with an oscillating quartz crystal rate monitor. Deposition of Pd and Cu were performed at the rate of IA/second in 3 mTorr of Ar. A 300A Pd seed layer was
deposited on the glass substrates to induce (111) texture in the multilayer films. For this study a series of Cu/Pd films with varying bilayer periods and relative thickness ratios was prepared. The structure was examined using x-ray diffraction, and the surface morphology was examined using scanning tunneling microscopy. The substrate curvature measurement system consisted of externally mounted HeNe laser, mirrors, lenses and position sensitive detector as shown schematically in Fig. 1. By rotating the scanning mirror, the laser is swept over the surface of the substrate, and the curvature can be calculated from the deflection of the reflected light measured at the position sensitive detector. By scanning before and after the deposition of a film, the film stress a can be determined from the change in curvature using the Stoney equation: Et2 orI_8
G6tf( -v
-
) (XR _t_
)I
where E8 , t, and v, are the Young's modulus, thickness and Poisson's ratio for the substrate, tf is the film thickness, and R0 and R are the radii of curvature before and after the film deposition. With this scanning method, we can accurately determine the absolute values of film stress. However, on our apparatus, each scan takes about 20 seconds, which limits the time resolution of this technique in real time growth studies. To achieve relatively rapid measuring rates appropriate for typical sputter deposition growth rates, we position the incident laser at a fixed position on the edge of the substrate and measure the deflection of the reflected light as a function of time. This deflection method allows us to monitor the stress behavior of films during growth with a time constant corresponding to sub-monolayer coverage. Class Window
Optical Table Stationary Flat mirror
Plano-Convex | I Lens
ustrate
.1
Sensitive Detector
•••'-Position u
•
Sputtering Chamber \ UHVTarget
",,,1"-Scanning Mirror Foal Plane of Ln
~~Laser' Figure 1: Schematic in-situ of laser curvature apparatus. A growing film with a constant level and sign of stress would cause a linear increase in deflection as a function of time. We plot the deflection versus time using the convention that a negatively sloped line would result from a growing film with compressive stress, and a positively sloped line would result from a film with tensile stress. It is important to realize that the slope of this line is just the apparent stress in the film, and that extracting the stress of a given increment of film by differentiating the displacement versus thickness cu
Data Loading...
