Fig. 2-11 presents the XRD spectra of the nickel germanides formed on the 100 nm
18
19
poly-Ge at different annealing temperatures. The peaks at 34.7o, 36.7o, 42.7o, 44.2o, 45.6o, 53.5o are identified as NiGe(111), (120), (021), (211), (121), and (002), respectively. No peaks correspond to other phases of nickel germanides such as Ni3Ge2, Ni2Ge, etc. This indicates that poly-NiGe is the only phase at the annealing temperature ranging from 250 ℃ to 500 ℃. The average grain sizes of poly-NiGe can be calculated by using the Scherrer relation. From the XRD spectra and the Scherrer relation, it was observed that the grain size of NiGe gradually increases with increasing temperature. Fig. 2-12 shows the sheet resistance of the NiGe as a function of annealing temperature. Clearly, it can be seen that the increase of sheet resistance occurs at the temperature above 400 ℃. We attribute the phenomenon of the increase in sheet resistance to the grain growth of poly-Ge. From the previous XRD spectra, we found that the peak corresponding to the phase of poly-Ge(111) increases and the full-width half-maximum of the peak narrows at the temperature above 400 ℃, and this result provides the evidence for grain growth of poly-Ge grain. Further, the relative intensity of the peaks of the grown poly-Ge(111) were investigated by magnifying further XRD spectra surrounding the peaks and the average crystallite size (L) of the Ge was estimated by using the Scherrer relation, as shown in Fig. 2-13.
Obviously, the average grain size (L) of poly-Ge increases with increasing temperature at above 400 ℃. In fact, similar results have been observed in the studies
20
of silicides formed on poly-Si [34]-[35], and it has been ascribed to the fact that the poly-Si growth beneath the upper silicide film induced the broken silicide film during annealing. Hence, the growing grain of poly-Si accounts for the drastic increase of the resistance of the silicide film. It is believed that the silicide film enhances the grain growth of the poly-Si because of the reduction in grain boundary and interface energy in the poly-Si. From previous results, we infer that NiGe enhances the grain growth of the poly-Ge during annealing, especially at the temperature above 400 ℃. As the grain of poly-Ge is gradually enlarged, the grain penetrates into the upper NiGe film and the NiGe decomposes. Further, the metal atoms from the decomposed NiGe diffuse to the poly-Ge/NiGe interface to form new NiGe and then trigger the formation of the column NiGe. Finally, the NiGe and poly-Ge interlace. The interlaced grains break the continuous NiGe film and account for the increase of the sheet resistance [34]-[36].
The surface images of NiGe formed at 300 ℃, 400 ℃, 450 ℃, and 500 ℃ are shown in Fig. 2-14. From the AFM images, it can be clearly seen that the phenomenon of growing grain is becoming more and more obvious as the annealing temperature increases. The same result can be achieved from the SEM images as shown in Fig. 2-15. Therefore, we believe that the dramatic increase of sheet resistance is induced by the gradually growing grain of poly-Ge when increasing the
21
annealing temperature.
In Fig. 2-16, the XRD analysis of as-deposited sample and nickel germanides formed on the 30 nm poly-Ge at various annealing temperatures from 250 ℃ to 500
℃ is shown. Unlike the nickel germanides formed on the 100 nm poly-Ge, poly-NiGe was not the phase formed on the 30 nm poly-Ge anymore and there were no peaks corresponding to poly-Ge(111) by comparing the XRD spectra of as-deposited sample and germanides samples which indicated that the poly-Ge was fully consumed. At the annealing temperature of 250 ℃, Ni2Ge corresponding to the peaks (210), (202), (013) and (211) at 42.4o, 43.2o, 44.2o, and 44.3o is the only phase formed on the poly-Ge.
However, as the temperature increases, it can be seen that there are two phases of Ni2Ge and Ni3Ge whose peaks are identified as (111), (200), and (220) at 43.9o, 51.2o, and 75.3o are simultaneously found in the Ge film at the temperature ranging from 300 ℃ to 400 ℃. Then, while the annealing temperature rises to 450 ℃, Ni3Ge is the only phase observed in the poly-Ge. Finally, Ni5Ge3 whose peaks are identified as (002), (403), (203), (602), and (420) at 36.3o, 43.7o, 45.6o, 46.7o, and 47.8o is the only phase that appears in the Ge film. For 30 minutes annealing, these results indicate that Ni2Ge is the only phase formed at low temperature and there is a phase competition between Ni3Ge and Ni2Ge as the temperature increases upon 300 ℃. With increasing temperature, Ni2Ge is gradually consumed whereas Ni3Ge slowly grows. Finally,
22
Ni2Ge has been fully consumed and Ni3Ge is the only phase found in the poly-Ge at the temperature of 450 ℃. Ni3Ge would be further consumed to form Ni5Ge3 which is the only phase observed at 500 ℃. Because Ni5Ge3, Ni2Ge and Ni3Ge are the transition phase [37]-[42], we believe that the phase of NiGe would be formed as the annealing time increases to fully consume the transition phase regardless of annealing temperature. Therefore, for applying to Ge thin film, the thickness of Ni film must be decreased to the extent that it is fully consumed by Ge or the annealing time must be long enough to form NiGe which possesses the advantage of low sheet resistance appropriate for high speed device application, especially at low temperature.