Half-flat vs. atomically flat: Alkyl monolayers on morphologically controlled Si(100) and Si(111) have very similar structure, density, and chemical stability

Abstract

Chemists have long preferred the Si(111) surface for chemical functionalization, as a simple aqueous etch can be used to produce ideal, atomically flat H/Si(111) surfaces for subsequent reactions. In contrast, industry-standard etches produce rough H/Si(100) surfaces terminated by nanohillocks. The recent discovery of an aqueous etch that produces morphologically controlled H/Si(100) surfaces with a near atomically flat or "half-flat" morphology challenges the assumption that Si(111) is an inherently preferable starting point for chemical functionalization. This study shows that alkyl functionalization of morphologically controlled, "half-flat" Si(100) surfaces by terminal alkenes produces dense, tightly packed monolayers that are essentially identical to those on atomically flat Si(111). The striking similarity between the infrared spectra on these two surfaces - in terms of absolute absorbance, line shape and position, and polarization dependence - strongly suggests that alkyl monolayers on morphologically controlled Si(111) and Si(100) have essentially identical structures. The principle difference between the two surfaces is the amount of residual H at the Si/organic interface, a difference that is dictated by the structure of the Si(100) surface. Alkyl monolayers on morphologically controlled Si(111) and Si(100) surfaces were shown to be equally resistant to harsh oxidants. As a result, there appears to be no chemical reason to prefer one surface over the other, at least for functionalization with terminal alkenes.

Si/C/H ReaxFF Reactive Potential for Silicon Surfaces Grafted with Organic Molecules

“…Full-coverage methylated surfaces were readily obtained for Si(111) and Si(100)–2 × 1, whereas a coverage of 0.9 was obtained on the Si(100)–hf, although the surface density of methyl groups is higher on this surface. In the time frame of our simulations, the C2 monolayer reaches a coverage of 0.7 on the Si(111) surface and 0.8 on the Si(100)–2 × 1 surface, whereas for C3 and C5, similar coverages are observed on both surfaces (0.5 and 0.6) in agreement with coverages reported in the literature. ,,− In the case of C10, the details of the crystalline surface structure begin to blur as the same coverage was obtained on all surfaces in agreement with experimental observations . Both the DFT calculations and the ReaxFF MD simulations predict that for C10, stable monolayers with coverages of up to 0.8 can be formed, although experimentally coverages around 0.5 have only been reported.…”

“…It has 32 Si atoms in the first layer (dihydrogenated) and 32 Si atoms in the second layer (forming 16 Si dimers, one dimer is colored green for clarity). This structure is somewhat different from the experimentally determined structure, 28,29 in which the elevated rows are terminated by unstrained dihydrides (as in Figure 1c), whereas the trenches are terminated by strained dihydrides, instead of the monohydride-terminated dimers of Figure 1c. As the strained dihydrides in the trenches are likely kinetically trapped structures, which would convert to dimers at higher temperatures to relieve the strain, we decided for simplicity to use the hydrogenated dimers in our high-temperature calculations.…”

“…11,13,15−17 In the case of C10, the details of the crystalline surface structure begin to blur as the same coverage was obtained on all surfaces in agreement with experimental observations. 29 Both the DFT calculations and the ReaxFF MD simulations predict that for C10, stable monolayers with coverages of up to 0.8 can be formed, although experimentally coverages around 0.5 have only been reported. Although these high coverages are feasibly from an energetic point of view, they may be difficult to attain in experiments due to steric effects.…”

“…We also investigated the functionalization of the so-called half-flat Si(100) surface, Si(100)–hf. It is experimentally produced when the silicon oxide is removed by successively immersing the Si(100) wafer in NH 4 F solutions. , This treatment produces a nearly atomically flat surface characterized by missing-row structures along [011] directions. Figure c shows that the top Si atoms of this surface are dihydrogenated, whereas the Si atoms in the second layer form dimers, which are monohydrogenated.…”

“…This is in agreement with recent observations by the Hines's group, where they concluded that the structure, density, and chemical stability of dodecyl (C 12 H 23 −) monolayers on Si(111) and Si(100)−hf surfaces are essentially identical. 29 This implies that the details of the crystalline surface structure become less important as the chain length increases.…”