Chemical characterization and charge carrier dynamics of crystalline silicon(111) surfaces modified with surface-bound organic functional groups

The structural morphology of the CH3-terminated Si (111) surface was investigated by low-energy electron diffraction (LEED), which revealed that this surface retained a flat, non-reconstructed (1 x 1) structure. CH3-terminated Si(111) surface collected at an incident beam angle of 74◦ and 30◦off surface normal.

Charge Carrier Dynamics at Semiconductor Surfaces and Interfaces

The net positive charge in the semiconductor is instead distributed over a much larger part of the crystal, called the depletion or space charge region. Electron-hole pair recombination (a) in the bulk and (b) on the surface of a semiconductor crystal.

Figure 1.1: Energy band diagram of a n-type semiconductor immersed in a solution con- con-taining an electron donor-acceptor pair (E 0 (A/A − ))

Chemical Modification of Silicon(111) Surfaces

Given a surface atom density of 7.8 x 1014 cm−2 on Si(111), 1S determination can lead to a direct determination of the number of Si surface defects. The HREELS spectra, however, were unable to estimate the overall surface coverage of the Si-C feature.16.

Figure 1.3: Structure of two common silicon crystal faces. (a) Si(100) surface; (b) Si(111) surface.

Summary

On the CH3-terminated surface, two regions were closely examined to identify Si–C stretching. These protection strategies should prevent extensive oxidation of the Si(111) surface while maintaining a low level of H-terminated surface recombination of the Si(111) surface.

Experimental

Materials
Sample Preparation

Preparation of H-Terminated Si(111) Surfaces
Preparation of Cl-Terminated Si(111) Surfaces
Preparation of C n H 2n+1 -Terminated Si(111) Surfaces

After removal from the etching solution, the sample was rinsed thoroughly with H2O and dried under a stream of N2(g). The sample was then removed from the reaction solution, rinsed with tetrahydrofuran (THF) and CH 3 OH and then blown dry under a flowing stream of N 2 (g).

Table 2.1: Alkylmagnesium halide reagents and reaction times for Si(111) surface alkyla- alkyla-tion with X = Cl or Br

Introduction

Due to the importance of silicon surfaces and interfaces in modern electronic devices, XPS of Si surfaces, usually prepared under ultrahigh vacuum (UHV) conditions, has been investigated and reported in the literature. Both factors prevented previous XPS investigations of the alkylated Si(111) surface from identifying possible Si atoms that could be bonded to the carbon atom of the alkyl group.

Experimental

Materials and Methods
Instrumentation

XPS Measurements
SXPS Measurements

The escape depth of Si 2p photoelectrons was calculated using an empirical relation described by Seah.21 The size of the Si atom, aSi, was determined using Eq. The distance between the first and second Si layers along a vector perpendicular to the Si(111) crystal face is 1.6 ˚A, which means that an electron escape depth of 3.5 ˚A will sample 2.2 monolayers of the Si crystal.

Results

H-Terminated Si(111) Surfaces

XPS Results
SXPS Results

Cl-Terminated Si(111) Surfaces

XPS Results
SXPS Results

Alkyl-Terminated Si(111) Surfaces

XPS Results
SXPS Results

High-Resolution SXPS of CH 3 -Terminated Si(111) Surfaces Col-

The low binding energy tail in the Si 2p3/2 region shown in Figure 3.2(b) showed a small signal that was not explained by the peak fitting procedure. Spectra are shown after background subtraction and before spin–orbit stripping as a function of relative binding energy (eV) above the center of the Si 2p3/2 signal.

Figure 3.1: Survey XPS of functionalized Si(111) surfaces prepared through a two-step chlorination/alkylation method: (a) H-Si(111), (b) Cl-Si(111), (c) CH 3 -Si(111), (d) C 2 H 5 -Si(111)

Discussion

The lack of Cl signals on the methylated Si surface is consistent with the formation of an almost full monolayer of Si-CH3 bonds. The Si–C bond energy shift is no more than 0.2 eV larger than that of Si–H, making reliable resolution of the two peaks difficult, even under these high resolution conditions. However, fitting the single integrated peak at +0.19 eV above the Si 2p3/2 bulk peak to two peaks at known Si–C and Si–H bond energy shifts resulted in a deconvolution of the spectrum consistent with expectations for termination of ' a monolayer of Si on top of atoms on such surfaces with either Si–C or Si–H bonds.

Conclusion

Furthermore, the small full width at half maximum (fwhm) of the Si-Cl stretching and bending modes on the Cl-terminated Si(111) surface indicated that the surface was very homogeneous. This value is an approximation of the degree of band bending at the CH3-terminated Si(111) surface. Subsequent alkylation of the Cl-terminated Si surface was then performed as also described in section 7.2.1.3.

Experimental

Instrumentation

Functionalization of the surface with CH3- and C2H5- was achieved following the procedures in Chapter 2, with several modifications. After the TIRS data was collected, the sample was again quickly air-returned to the antechamber of the N2(g) purged glove box. The single beam spectrum of each surface was dominated in the low wavenumber region by Si lattice phonon vibrations, and so each sample was referenced to the surface of its precursor to subtract common components of the signal.

Figure 4.1: Configuration of TIRS experiments. For maximum signal to the detector, θ = 74 ◦ , while for polarization-type experiments, θ = 30 ◦ .

Results

H-Terminated Si(111) Surfaces
Cl-Terminated Si(111) Surfaces

Experimental Results
Calculations of the Cl-Si(111) Surface

CH 3 -Terminated Si(111) Surfaces
C 2 H 5 -Terminated Si(111) Surfaces

Measurement of Si–Cl bending mode regions when observed at an IR incident angle of 30° gave similar results. New prominent C–H stretching mode peaks appeared around 2900 cm−1 on the CH3 terminated surface shown in Figure 4.7. This was close to the energy of the peak representing the Si–H stretching motion (2083 cm−1) observed in the spectrum of the freshly etched Si(111) surface previously seen in Figure 4.2.

Figure 4.2: TIRS and proposed peak assignments of the H-terminated Si(111) surface col- col-lected at an incident beam angle of 74 ◦ (bottom) and 30 ◦ (top) off surface normal

Discussion

620 cm−1, which was often obscured by a large Si-Si phonon vibration, which was particularly difficult to reduce in the difference spectrum. If the Si-Si phonon background is properly subtracted, any features remaining below that region should be revealed after the subtraction is complete. However, this procedure revealed nothing on the CH3-terminated surface, even after the Si-Si lattice phonon appeared to be fully accounted for (data not shown).

Conclusion

This was then used as a background spectrum that was subtracted from the difference spectrum of the CH3-terminated Si(111) surface. High-resolution soft XPS was used to determine the electronic band structure of the CH3-terminated Si(111) surface. The degree of surface band bending and electron affinity of the CH3-terminated Si(111) surface is summarized in Figure 7.6.

Experimental

Instrumentation

After the wafer was cut, the orientation of the sample relative to the flat was marked and monitored throughout by STM data collection. After functionalization, the sample was mounted on the STM stage and rapidly introduced into the UHV chamber of the STM. Data were collected at the BESSY synchrotron facility in Berlin by Bengt Jaeckel of the Institute for Materials Science at the Technische Universitat¨at Darmstadt, Darmstadt, Germany.8.

Results

The arrangement was quite robust, with relatively few defects observed over large areas of the surface. The orientation of the pits in Figure 5.2(a) therefore provided independent confirmation of the orientation of the low-index lattice planes. This analysis determined that the C-H bonds of the methyl groups were rotated 7◦ towards the underlying Si-Si back bond.

Figure 5.1: LEED pattern of CH 3 -Si(111) taken with electron energies of 40, 45, and 50 eV

Discussion

Using these three independent verifications of the orientation of the Si crystal in the STM it was possible to identify the surface unit cell, which is indicated by the parallelogram drawn in Figure 5.3(b). The minimum energy structure for such a surface was calculated to be an angle6 A–B–D of 30◦, close to the angle of 23◦ measured from the STM data, and confirmed that the interactions governing the packing of the CH3 - terminated Si surfaces are mainly the repulsions between hydrogen atoms on adjacent methyl groups in the functionalized organic top layer. However, the repulsive methyl–methyl interactions are tempered by an attractive preference for the eclipsed conformation of the C–H bonds over the underlying Si–Si bonds.

Figure 5.4: Proposed structural model of the CH 3 -terminated Si(111) surface, viewed nor- nor-mal to the <111>plane

Conclusion

This type of interaction is expected to result in a 60 A–B–D angle of 60◦ for an isolated methyl group attached to the Si(111) surface. This is manifested by a favorable interaction resulting in a 6 A–B–D angle of 23° on the Si surface with the methyl end, in contrast to the repulsion between the methyl C–H bonds and the S–Si bonds leading from the Si bound to the methyl group , which would be expected to produce an angle of 6 A–B–D>30◦. While the CH3-terminated Si(111) surface is the simplest and most ideal functionalized surface available for study, understanding the morphology of surfaces alkylated with longer or larger functional groups is important to obtain a complete picture of the alkylated surface.

Experimental

Instrumentation

SXPS Measurements
XPS Measurements
SAM Measurements

The distance between Si layers along a vector perpendicular to the Si(111) crystal face is 1.6 ˚A, which means that an electron leakage depth of 3.5 ˚A will sample 2.2 monolayers of the Si crystal. Secondary electron images of the sample were used to identify a debris-free region of the sample surface. Because the amount of SiO2 on the surface of alkylated samples was very small, it was not possible to collect spectra of the SiO2 signal (72-76 eV) in a short enough time to prevent radiation damage.

Results

SXPS Analysis of Alkylated Surfaces

Cl-Terminated Si(111)
CH 3 -Terminated Si(111)
C 2 H 5 -Terminated Si(111)
C 6 H 5 CH 2 -Terminated Si(111)

Oxidation Rates of Stepped Surfaces
SAM of Alkyl-Terminated Si(111) Samples

The scale is the binding energy (BeV) relative to the center of the bulk Si 2p3/2 peak of the freshly prepared surface. The scale is the binding energy (BeV) relative to the center of the Si 2p3/2bulk peak of the freshly prepared surface. The scale is the binding energy (BeV) relative to the center of the bulk Si 2p3/2peak of the freshly prepared surface.

Figure 6.1: Chlorine-terminated Si(111) surface freshly prepared then allowed to sit in air for 12, 24, 36, and 48 hr

Discussion

To overcome this problem, two spectral maps of the surface were collected: one of the Si LVV region at 92 eV and one of the O KLL region at 510 eV. In a regular SEM image (Figure 6.8 (a)) bleached parts of the Si surface appeared in patches that ran roughly parallel to each other. O signal; iv) composite SAM image of superimposed maps of Si LVV (red) and O KLL (blue) energy regions.

Figure 6.8: Auger electron spectroscopy of a C 2 H 5 -terminated Si(111) surface exposed to air for 8 days

Conclusion

Detailed investigation of the oxidation of functionalized Si(111) surfaces in air was described in Chapter 6. Surface charge carrier recombination rates are our most sensitive measure of the electronic quality of a surface and should be the standard by which it is judged. surface passivation. The observed dynamics of surface charge carrier recombination were correlated with chemical measurements and band structure in an attempt to explain any apparent electronic passivation of the alkyl-terminated Si(111) surface.

Experimental

Materials
Functionalization by Chlorination/Alkylation

Instrumentation

XPS
Surface Recombination Velocity Measurements
High-Resolution Core Level Photoelectron Spectroscopy 133

XPS Results
Surface Recombination Velocity Results
High-Resolution Core Level Photoelectron Spectroscopy Results

The detailed scan of the Si 2p region was used to determine the amount of Si oxides present. 7.5, the SiOx:Si 2p peak area ratio is independent of the photoionization cross section of all species present in the scan. The decay lifetime of the fundamental charge carrier, τ, is dependent on the lifetime of the charge carriers both in the bulk and on the surfaces of the sample.

Figure 7.1: Survey scan XP spectra of a Si(111) surface (a) cleaned and etched in 40%

Discussion

Upon short-term exposure to air, all surfaces prepared by chlorination/alkylation routes were much more chemically inert to oxidation than H- or Cl-terminated Si(111) surfaces. The rapid oxide growth observed upon air exposure of H- and Cl-terminated surfaces was correlated with a sharp drop in the surface charge carrier lifetime. In fact, all samples functionalized via the chlorination/alkylation route showed increased lifetime with longer exposure to air.

Conclusion

Introduction

Comparison of electrical properties and chemical stability of crystalline silicon (111) surfaces alkylated using Grignard reagents or olefins with Lewis acid catalysts. In order for this promising field to be developed further, some direct comparisons of the chemical and electrical stabilities of different modification techniques need to be made. Surface chemical oxidation due to air exposure determined by XPS was correlated with surface S values to determine whether changes in the surface recombination rate could be related to changes in the chemical composition of the functionalized Si surfaces.

Experimental

Materials
Functionalization by Chlorination/Alkylation
Lewis Acid-Mediated Terminal Alkene Reduction
Electrochemical Reduction of CH 3 MgI

Instrumentation

Occasionally, wafers were inserted directly into the vacuum-sealed pre-chamber of the N2(g)-flushed purge box to prevent surface scratches caused by contact with the XPS sample holder. A Teflon two-chamber cell was fabricated so that both sides of the double-polished wafer could be exposed to the electrolysis solution. The cell was then disassembled and the top and bottom ohmic contacts were scratched off to leave only the portion of the wafer that had been exposed to the reaction solution.

Results

Surfaces Functionalized by Chlorination/Alkylation Methods
Lewis Acid-Mediated Terminal Alkene Reduction

XPS Results

Electrochemical Reduction of CH 3 MgI

XPS Results

After 200 hours in air, the C8H17-functionalized surface exhibited a normalized SiOx:Si 2p peak area ratio of 0.21, compared to a SiOx. Si 2p peak area ratio of 0.12 observed for a C8H17-terminated surface prepared by PCl5 chlorination/C8H17 alkylation. The magnitude of the C 1s:Si 2p peak area ratio was larger than expected for alkyl chains at least 12 carbon atoms in length,5 indicating that the C1s signal was not entirely due to surface-bound methyl groups.

Figure 8.1: Survey scan XP spectrum of C 6 H 13 -terminated Si(111) prepared by reduction of 1-hexene catalyzed by C 2 H 5 AlCl 2 .

Discussion

It is clear that the reaction chemistry of crystalline Si (111) is quite sensitive to the preparation method of the alkyl layer on the functionalized Si surface. Infrared spectroscopic studies have shown that the electrochemical anodization method produces a degree of methylation of the Si(111) surface.33 The XPS data described above however show that this surface also has a large O 1s XPS signal, as well as from other elements used in the functionalization step. Consequently, these surfaces exhibited lower chemical and electrical properties than any alkylated surface that was prepared by either chlorination/alkylation route.

Figure 8.4: Proposed mechanism of Si surface alkylation through the Lewis acid-catalyzed reduction of a terminal olefin.

Conclusion

At all points in the surface functionalization chemistry described here, the degree of contamination of the Si surface by silicon oxides or random hydrocarbon species was low. TIRS of the C2H5-terminated surface identified peaks that could be attributed to Si–H stretching and bending motions, confirming the presence of a Si–H species on the alkylated surface. This is probably due to the high sensitivity of radio frequency (rf) photoconductivity decay methods in v.

Introduction

Data Entry and Manipulation