Hydrogen-terminated, chlorine-terminated, and alkyl-terminated crystalline Si(111) surfaces have been characterized using high-resolution, soft X-ray photoelectron spectroscopy from a synchrotron radiation source. The H-terminated Si(111) surface displayed a Si 2p(3/2) peak at a binding energy 0.15 eV higher than the bulk Si 2p(3/2) peak. The integrated area of this shifted peak corresponded to one equivalent monolayer, consistent with the assignment of this peak to surficial Si-H moieties. Chlorinated Si surfaces prepared by exposure of H-terminated Si to PCl5 in chlorobenzene exhibited a Si 2p(3/2) peak at a binding energy of 0.83 eV above the bulk Si peak. This higher-binding-energy peak was assigned to Si-Cl species and had an integrated area corresponding to 0.99 of an equivalent monolayer on the Si(111) surface. Little dichloride and no trichloride Si 2p signals were detected on these surfaces. Silicon(111) surfaces alkylated with CnH(2n+1)- (n = 1 or 2) or C6H5CH2- groups were prepared by exposing the Cl-terminated Si surface to an alkylmagnesium halide reagent. Methyl-terminated Si(111) surfaces prepared in this fashion exhibited a Si 2p(3/2) signal at a binding energy of 0.34 eV above the bulk Si 2p(3/2) peak, with an area corresponding to 0.85 of a Si(111) monolayer. Ethyl- and C6H5CH2-terminated Si(111) surfaces showed no evidence of either residual Cl or oxidized Si and exhibited a Si 2p(3/2) peak approximately 0.20 eV higher in energy than the bulk Si 2p(3/2) peak. This feature had an integrated area of approximately 1 monolayer. This positively shifted Si 2p(3/2) peak is consistent with the presence of Si-C and Si-H surface functionalities on such surfaces. The SXPS data indicate that functionalization by the two-step chlorination/alkylation process proceeds cleanly to produce oxide-free Si surfaces terminated with the chosen alkyl group.

Metal-semiconductor junctions between Hg and chemically modified n-and p-Si(111) surfaces have been prepared and analyzed using current-voltage and differential capacitance-voltage methods. To understand the role of the interfacial dipole on interfacial charge transfer, silicon surfaces were modified with either nonstoichoimetric oxide (SiO x ), terminal monohydride, short (C n H 2n+1 -, n ) 1, 2, 3) saturated alkyl chains, or propynyl (CH 3 -CtC-) groups. X-ray photoelectron spectra of the modified Si electrode surfaces taken before and after exposure to Hg contacts showed no evidence of irreversible chemical interactions between the Si and the Hg. Hg/Si contacts made using H-terminated Si(111) surfaces exhibited Schottky junctions having barrier heights (Φ b ) that were consistent with the known surface electron affinity of Si and the work function of Hg. In contrast, Si coated with a thin, chemically grown oxide formed Hg/Si junctions having barrier heights suggestive of Fermi level pinning. Si(111) surfaces modified with methyl groups yielded Hg junctions having barrier heights in accord with expectations based on the electron affinity (3.67 eV) and surface dipole (0.38 eV) measured on such surfaces by photoemission spectroscopy, attesting to the degree of chemical control that can be exerted over the barrier heights of such systems by surface functionalization methods. Incomplete coverages of functional groups produced by alkylation with ethyl or iso-propyl groups did not greatly impact the observed values of Φ b relative to Φ b values observed for CH 3 -terminated Si(111) surfaces. However, the observed variation in Φ b between nominally identical samples increased as the number of carbons in the functionalizing alkyl group increased. Junctions between Hg and Si(111) surfaces modified with propynyl groups showed nearly identical behavior to that of CH 3 -Si(111)/Hg contacts, both in average Φ b values and standard deviation between samples. The behavior of Si/Hg interfaces modified with short organic functional groups is consistent with the efficacy and utility of passivated surfaces in modifying the properties of surface-based Si devices.

Single-crystal Si(100) surfaces have been functionalized by using a two-step radical chlorination-Grignard (R) MgCl, R ) CH 3 , C 2 H 5 , C 4 H 9 , C 6 H 5 , or CH 2 C 6 H 5 ) alkylation method. After alkylation, no chlorine was detectable on the surface by X-ray photoelectron spectroscopy (XPS), and the C 1s region showed a siliconinduced peak shift indicative of a Si-C bond. The relative intensity of this peak decreased, as expected, as the steric bulk of the alkyl increased. Despite the lack of full alkyl termination of the atop sites of the Si(100) surface, functionalization significantly reduced the rate of surface oxidation in air compared to that of the H-terminated Si(100) surface, with alkylated surfaces forming less than half a monolayer of oxide after over one month of exposure to air. Studies of the charge-carrier lifetime with rf photoconductivity decay methods indicated a surface recombination velocity of <30 cm s -1 for methylated surfaces, and <60 cm s -1 for Si surfaces functionalized with the other alkyl groups evaluated. Soft X-ray photoelectron spectroscopic data indicated that the H-Si(100) surfaces were terminated by SiH, SiH 2 , and SiH 3 species, whereas Cl-Si(100) surfaces were predominantly terminated by monochloro (SiCl and SiHCl) and dichloro (SiCl 2 and SiHCl 2 ) Si species. Methylation produced signals consistent with termination by Si-alkyl bonding arising from SiH-(CH 3 )-, SiH 2 (CH 3 )-, and Si(CH 3 ) 2 -type species.

High-resolution soft X-ray photoelectron spectroscopy was used to investigate the oxidation of alkylated silicon(111) surfaces under ambient conditions. Silicon(111) surfaces were functionalized through a two-step route involving radical chlorination of the H-terminated surface followed by alkylation with alkylmagnesium halide reagents. After 24 h in air, surface species representing Si + , Si 2+ , Si 3+ , and Si 4+ were detected on the Cl-terminated surface, with the highest oxidation state (Si 4+ ) oxide signal appearing at +3.79 eV higher in energy than the bulk Si 2p 3/2 peak. The growth of silicon oxide was accompanied by a reduction in the surface-bound Cl signal. After 48 h of exposure to air, the Cl-terminated Si(111) surface exhibited 3.63 equivalent monolyers (ML) of silicon oxides. In contrast, after exposure to air for 48 h, CH 3 -, C 2 H 5 -, or C 6 H 5 CH 2 -terminated Si surfaces displayed <0.4 ML of surface oxide, and in most cases only displayed ≈0.20 ML of oxide. This oxide was principally composed of Si + and Si 3+ species with peaks centered at +0.8 and +3.2 eV above the bulk Si 2p 3/2 peak, respectively. The silicon 2p SXPS peaks that have previously been assigned to surface Si-C bonds did not change significantly, either in binding energy or in relative intensity, during such air exposure. Use of a high miscut-angle surface (7°vs e0.5°off of the (111) surface orientation) yielded no increase in the rate of oxidation nor change in binding energy of the resultant oxide that formed on the alkylated Si surfaces. Scanning Auger microscopy indicated that the alkylated surfaces formed oxide in isolated, inhomogeneous patches on the surface.