Surface Roughness and Specific Contact Resistance of AuGeNi/InP Ohmic Contacts
- PDF / 1,124,743 Bytes
- 6 Pages / 414.72 x 648 pts Page_size
- 82 Downloads / 213 Views
[1] measured the surface roughness of alloyed AuGe/Ni ohmic contacts to n-type InP annealed using rapid thermal annealing for 30 seconds. They found that the surface roughness increased above 4600C annealing due to melting processes of Au-In binary alloys [1]. Ball [2] found for alloyed AuGeNi/GaAs contacts that the surface roughness was large in the optimum low speci-
fic contact resistance window when using single-step annealing. However, the surface roughness was reduced when employing a two-step annealing sequence, where in the first annealing step the dominant phases for low specfic contact resistance was grown and in the second annealing step it was melted [2]. The two-step annealing technique improved contacts both electrically (lower contact resistance) and with respect to the morphology [2]. In this paper we report on the surface roughness of AuGeNi metallizations to n-type InP using single-step rapid thermal annealing. Also, we report on ways to minimize the surface roughness, while still maintaining the good electrical properties of the AuGeNi/InP contacts. EXPERIMENTAL Ohmic contacts were fabricated by sequential evaporation of Au-12wt%Ge (AuGe), Ni and Au onto highly doped n-type InP wafers. Single-step rapid thermal annealing (RTA) in an AG 389
Mat. Res. Soc. Symp. Proc. Vol. 355 0 1995 Materials Research Society
Heatpulse 410 owen were used to activate the contacts. The n'-wafers were sulphur doped to levels around 6x10 3 cm-3 and the thickness of the layers was 100 nm (AuGe), 25 nm (Ni) and 50 nm (Au). Circular contact dots of different diameters for electrical measurements of the specific contact resistance (r.) were defined in a sputter deposited SiO2 layer prior to the evaporation. Likewise, large-area samples were prepared for surface roughness measurements and Auger electron spectroscopy (AES) analysis. The specific contact resistance was calculated on the basis of the Cox and Strack technique [3]. Scanning surface profiling roughness measurements were done using a Tencor P1 Long Scan Profiler. The surface roughness (p) of the large-area samples was defined in close accordance with the definition put forward by Del Alamo and Mizutani [1] to be able to compare directly with their results. Thus, we measured the maximum peak-to-valley difference from a 100 jtm length scan 10 different places on the samples, and defined p as the average of the peak-to-valley differences over the 10 scans. RESULTS
In Fig. 1 the results from the single-step RTA sequence are shown. The roughness of the asdeposited metallization was found to be 8.9 +0.3nm. From Fig. l it can be seen that the surface roughness remains low up to an annealing temperature of 3700C, but then have a large increase up to an annealing temperature of 420&C (120 nm), where the specific contact has its lowest value (7x10 ncm2). At higher annealing temperatures the surface roughness decreases rapidly with a less rapid increase in the specific contact resistance. The hump in the specific contact resistance at an annealing temperature of 4100C w
Data Loading...
