• PDF / 360,211 Bytes
• 6 Pages / 414.72 x 648 pts Page_size
• 56 Downloads / 297 Views

DOWNLOAD

REPORT

---

Although AIN is a wide bandgap semiconductor (6.2 eV), if left undoped its properties are more like that of an insulator. The band offset is believed to be evenly split between the conductance and the valence bands. The relative dielectric constant of about 8 to 9 alleviates the problem of high fields in the dielectric when a high voltage is applied to the blocking junction even though the breakdown electric field of AIN is not well determined. One reason for believing AIN to be better at elevated temperatures is that it could be an effective diffusion barrier, not allowing charged impurities to diffuse at applied bias and thereby shift the flatband voltage of devices, although this needs to be shown in practice. This paper is a preliminary study in which MIS structures using AIN are fabricated and characterised mainly by capacitance-voltage method to determine possible deposition parameters for further work. EXPERIMENTAL

The starting material for the deposition experiments were single crystalline silicon carbide wafers of the 4H or the 6H polytype with the silicon face polished slightly off-axis. All substrates but one had an epitaxial layer of 3 - 4 gim low doped SiC (< 1016 cm-3 ), whereas the bulk was doped in the 1018 cm- 3 range. The epitaxial layers were of the same doping polarity as the substrates (i. e. n on n, or p on p) and were supplied by CREE Research Inc. One substrate had a p-type epitaxial layer doped at about 1018 cm- 3 , which was grown at IMC in Sweden. Prior to deposition the wafers were cut into smaller pieces (1/6 sector of a 30 mm wafer) and cleaned in solvents and acids in the following sequence: 2 minutes of ultrasonic clean in trichloroethylene (TCE), followed by acetone and then methanol. The samples were then cleaned in a 1:1:5 mixture of NH 4 OH:H 2 0 2 :H2 0 heated to about 70 'C for 15 minutes followed by a 1:1:5 mixture of HCI:H 2 0 2 :H2 0, also 70 0 C, for 15 minutes. The samples were rinsed in deionized water between each step. The final step before loading into the chamber was a 30 second dip in 1:50 HF:H 2 0 and a nitrogen blow dry. The AIN growth was performed in a low pressure, vertically configured, metal organic chemical vapor deposition system. The system has two chambers separated by a load lock, which allows the growth to be performed in one chamber and reflection high energy electron diffraction (RHEED) in the other without breaking the vacuum. Precursors for the growth were trimethylaluminum (TMA) and ammonia (NH 3 ), and hydrogen was used as a carrier. The growth temperature was 1200 'C, at a pressure of 10 Torr. Typical growth times for 200 nm thick films were 8 minutes. For some of the samples, silane was introduced during ramp up of the temperature, but shut off before growth was initiated. This was done in an attempt to improve the crystallinity of the AIN films and the interface between AIN and SiC, by avoiding silicon depletion of the surface. After the film thickness was determined, the samples were coated with about 400 nm aluminum in an e-gun evaporator. Som