The film growth efficency was found to have an exponential dependence to the normal component of incident translational energy. The growth efficiency, defined as the ratio of the number of carbon atoms deposited to form c-SiC to the total number of carbon atoms in the beam striking the silicon substrate, increases by a factor of five over the kinetic energies studied (Figure 1). X-ray diffraction was used to verfiy the purity of c-SiC grown (Figure 2).

Figure 1. SiC growth efficiency as a function of the normal component of incident HMDS kinetic energy at Ts=1000 K	Figure 2. X-ray diffraction pattern of epitaxial 3c-SiC on Si(100) grown at Ts=1000 K for 4 hours at normal incidence with 2.80 eV HMDS

However, using hexamethydisilane as a beam precursor limited the total c-SiC film thickness to 1500 Å. In comparison, when using methylsilane as a beam precursor, c-SiC growth continued for deposition times beyond two hours (Figure 3).

Although absent from HMDS depositions, changes in methylsilane’s incident kinetic energy caused large changes in c-SiC film morphologies. With kinetic energies around 0.1 eV, the silicon carbide film morphology mimicked c-SiC grown with HMDS: a silicon carbide film broken up by square shaped etch pits and maximum film thicknesses of 1500 Å. More energetic methylsilane beams (0.5 eV) resulted in the growth of a very smooth c-SiC epitaxial film intermixed with 3D Si islands relatively uniform in size and aligned with the Si[001] substrate lattice (Figures 4,5).



Figure 4. Wide range atomic force microscopy (AFM) image of c-SiC film grown with methylsilane	Figure 5.AFM image of a Si island surrounded by 2D c-SiC film on Si[001]

The material growth facilty used in the c-SiC work consists of an ultra-high vacuum chamber equipped with two molecular beams, an 193 nm excimer laser, a differentially pumped UTI mass spectrometer, and a load-lock transfer system. A schematic of the setup is shown below along with an actual photograph.

Future Interests: Diamond Project

The next semiconductor material to study would be diamond films. Its optical transparency, thermal conductivity and large band-gap makes diamond an ideal material for several applications. Since the early 1980s, diamond thin-films have been grown on a variety of materials (Si, Ni, graphite, etc.) through several different gas-phase techniques (chemical vapor deposition, laser desorption, flame-torch, etc.). Nevertheless, an easy means to grow a diamond single crystal still needs to be found before these thin-films can be mass produced for circuit boards, laser-optic coatings, etc.

With our molecular beam facility, we plan to study the morphology of diamond thin-films by using simultaneously an energetic methyl radical source and an atomic hydrogen to deposit homoepitaxial diamond and heteroepitaxial diamond on silicon. Much like the c-SiC work, by utilizing energetic molecular beams, it should be possible to enter a new regime on the CH₃/H diamond phase space. Analogous gas-phase reactions [tertiary-butyl radicals and/or tertiary-butane reacting with methyl radicals or atomic hydrogen] have been found to have activation barriers accessible to molecular beam epitaxial experiments. In the next few weeks, diamond depositions will begin and will be characterized by Raman, SEM and AFM.

	Back To The Top
	Return to Sibener Group Home Page	Contact us with questions or comments