We have succeeded in the fabrication of Si-based metal-oxide-semiconductor diodes with an ultrathin ( i.e. , ∼1.4 nm thick) silicon dioxide layer whose leakage current density is lower than those for thermally grown layers with the same thickness. The leakage current density of the as-grown layers formed in concentrated nitric acid solutions ( i.e. , ∼3 A/cm 2 at the forward gate bias, of 1 V) is as high as those for thermal layers. The leakage current density is decreased to ∼1.5 and ∼1 A/cm 2 at V by postoxidation annealing (POA) at 400 and 500°C (in hydrogen), respectively, and the decreases are mainly attributed to the elimination of interface states and slow states, respectively. The vibrational frequency of longitudinal optical (LO) phonons of Si-O asymmetric stretching vibration increases with the POA temperature, while that of the transverse optical (TO) phonons remains constant. From the analysis of these vibrational frequencies, the atomic density of the as-grown layer and that after POA at 500°C are estimated to be 2.34 and cm 3 , respectively, i.e. , higher than that of crystalline bulk of cm 3 . Measurements of the valence band spectra for the structure show that the valence band discontinuity energy at the interface is increased from 4.3 to 4.6 eV by POA at 500°C. Therefore, the other reason for the decrease in the leakage current density by POA is a reduction in the tunneling probability through caused by the increase in the band discontinuity energy. Postmetallization annealing treatment at 200°C in hydrogen performed after POA further decreases the leakage current density, showing that atomic hydrogen passivates defect states more effectively than molecular hydrogen.
No takes yet. Share an insight, caveat, or question.
Asuha et al. (2004) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: