The influence of silicide formation on the barrier height of Ti/Si MIS Schottky barriers

Abstract

Ti/Si MIS diodes were made by evaporating Ti on a hot substrate at a temperature T,, which could be varied between room temperature and 760 OC. It was found that for T, < 300 OC the Schottky barrier height (OB changed with T, and also depended on the pre-metallisation treatment of the silicon surface. Between 300 OC and 500 'C, ( O B remained independent of T,. These results could be explained by the reaction wherein titanium converts the silicon oxide into titanium oxide and titanium silicide. Further support for this model was obtained from Auger measurements. Between 500 O C and 600 O C a complete conversion of both the interface layer and the Ti top layer into silicide occurred and oxide was no longer present a t the interface. For T, > 600 OC, a (OB typical for a titanium disilicide/Si contact was measured.

Micromachined fabrication of Si cantilevers with Schottky diodes integrated in the tip

“…For the probes investigated, quality factors of 1.7-2.5 were found. The barrier height varied between 0.7 eV and 0.8 eV, which is higher than most values [15,16]. Possibly the surface pretreatment before the Ti deposition plays a decisive role.…”

On the influence of the surface pretreatment of a Si substrate on cobalt silicide formation

“…In our previous work on titanium silicides [4], we developed a method of evaporation on a hot substrate to favour silicidation. This method was also applied in the present work on cobalt silicides up to a substrate temperature T, of 650 "C. However, no silicidation occurred for the S1 to S4 treatments, as checked by XRD and SAM.…”

“…First, we want to discuss the role of the interfacial oxide. The method of evaporation on a hot substrate yielded good silicidation in the case of Ti [4] but not in the case of the present work on Co. This can be explained by considering that Ti is able to convert the native silicon dioxide into titanium oxide and titanium silicide, as reported in our previous work [4], while according to the thermo- To ('C) ment, leading to thinner native oxide (figure 3), showed silicidation while the S1 and S3 samples, having a thicker interfacial oxide (figure 2) [12], did not.…”

“…The method of evaporation on a hot substrate yielded good silicidation in the case of Ti [4] but not in the case of the present work on Co. This can be explained by considering that Ti is able to convert the native silicon dioxide into titanium oxide and titanium silicide, as reported in our previous work [4], while according to the thermo- To ('C) ment, leading to thinner native oxide (figure 3), showed silicidation while the S1 and S3 samples, having a thicker interfacial oxide (figure 2) [12], did not. This shows that if the oxide layer thickness is lower than some critical value silicidation occurs even for Co. Another argument for the decisive role of the interfacial oxide can be found in the results on the Shiraki-treated samples, where the complete treatment led to silicide formation while the partial treatment, without the thermal desorption of the native oxide, did not.…”

“…It is known [4] that Ti converts the native SiO, layer but that CO does not [S]. Consequently, inhomogeneous silicide formation can occur for Co.…”