SAM Dynamics and Chain Length
 |
| Figure 1 |
Self-assembled monolayers (SAMs) formed from organic thiols on Au(111) have been studied extensively for over a decade. The strong interest in SAMs comes from their robust potential for many chemical, physical, and biological applications. The surface chemistry and structure of a SAM can be tuned by changing functional groups. Tying these powerful properties of self-organization and adaptability together, SAMs have been used for lithography, tribology, nano-electronic and biological sensor applications. Alkanethiol SAMs in particular have been studied as a model system for self-assembly. Fundamental studies have focused on the structural phase diagram of alkanethiol SAMs. Two basic phases have been found to be prevalent, a high-density c(4×2) phase formed from large exposures to the alkanethiol, in solution or via vapor deposition, and a low-density (p´Ö3) phase created with smaller doses. There are a variety of other intermediate and low-density phases, which have varying degrees of overlap between alkanethiol chains. The low-density striped phases studied here are all the lowest density phase for that particular chain length. Here we investigate the effects of chain length on the surface vibrations of alkanethiol SAMs.
Data for the C6/Au(111) SAM system are shown in Figure 1. The diffraction in the <110> direction matches the 0.275 Å-1 spacing in parallel momentum space established as the low-density striped phase conformation. We believe this to be an (8.3´Ö3) structure based on the reconstructed (23´Ö3) Au nearest neighbor spacing as discussed in our C10 structure study. The representative inelastic spectrum in Figure 1 also shows two dispersionless modes at DE = 7.3 ± 0.4 and 12.3 ± 0.4 meV. The lower energy mode we assign to the FTz mode found in the longer chain C10 alkanethiol systems. We attribute the higher energy mode, which is also dispersionless, to the 1st overtone of the 7.3 meV FTz phonon.
 |
| Figure 2 |
VENUS, a molecular dynamics code used for normal modes calculations, revealed three distinct low energy modes, with polarization components normal to the surface, which were common for all chain lengths; an up-down motion, a scissor, and a bowing motion. Figure 2 graphically depicts the equilibrium position and an extrema of each mode for a C6 molecule, which is representative for all modeled chain lengths. The calculated energies of the three low-energy modes as a function of the chain length are compared to the experimental energies in Figure 2. In addition, a low-energy mode parallel to the surface, which is a wag in the plane parallel to the surface, is also included in Figure 2. The normal modes analysis was focused on modes with large z-polarizations because of the experimentally observed FTz mode. The planar theoretical modes do, however, demonstrate trends due to the relationship between chain length and vibrational frequency.
The FTz modes for C6, C8, and C10 on Au(111) can be modeled by van der Waals forces between the surface and the hydrocarbon chain. The relationship between the FTz vibrational frequency and chain length in this class of systems agrees in part with prior work on alkanes and physisorbed alkanethiols. Witte et al. showed that alkanes of varying chain length on Cu(111) had essentially the same vibrational energies and further explored the Einstein mode by deuterating n-octane. The scaling of the vibrational energy of n-octane to deuterated n-octane showed an excellent agreement with a simple force constant approximation, E ~Ö(k/m). The work was then expanded to alkanethiols, 1-heptanethiol and ethanethiol, physisorbed on Cu(111) where a 6 meV FTz mode was found, again showing no dependence on chain length. In our study of chemisorbed alkanethiols, the C6 and C8 modes appear, not surprisingly, to be equivalent. For each additional methyl group added to the chain the force constant, k, and mass, m, will increase in proportion keeping the excitation energy, E, constant, with a small perturbation due to the effective mass of the sulfur head group. The equivalence of the C6 and C8 modes implies that the FTz mode is similar to those in the alkanes on Cu(111) and can be modeled by the simplest of attractions between that of the methyl groups and Au(111).
 |
| Figure 3 |
The increased experimental energy of the C10/Au mode indicates that this striped phase is more tightly bound to the surface. One possible explanation for the increased binding energy is a long-range commensurate structure that adds to the van der Waals forces between the hydrocarbon chains and the Au(111). We have previously demonstrated that 1-decanethiol has a unique unit cell that matches the underlying reconstructed Au(111). The (23´Ö3) unit cell of the reconstructed Au(111) coincides with that of twice the C10 repeat spacing, but no commensurate relation exists for the C6 or C8 unit cells. (The theoretical calculations assumed a deconstructed Au(111) surface and, hence, would not account for the extra binding). The relatively small shift in the energy of the C10 vibration suggests that the increase in the van der Waals intermolecular forces is due to a commensurate structure.
The higher energy C6 phonon, exhibited in Figures 1, was assigned by comparing its intensity to that of the fundamental excitation as a function of temperature. A simple relationship between the probability of the mth phonon transition for oscillators and the surface temperature has been developed:
 |
| Figure 4 |
The parameter m is the order of excitation and W(Ts) is the Debye-Waller factor determined for C6/Au(111). The Debye-Waller factor we derived for specular scattering does not account for the parallel mean square displacement and a fitting parameter was used to account for the parallel displacement in the probability calculations. Figure 3 depicts the experimental dependence of the intensity of the C6/Au phonon modes with respect to surface temperature. The background (linear, multi-phonon, and diffuse elastic) has been subtracted to allow for easier comparison. Both the experimental and theoretical ratios of the overtone to the first excitation grow linearly with a slope of W as the surface temperature increases. The agreement between the experiment and predicted probability, shown in Figure 4, along with prior observations of FTz overtones in alkanes and similar physisorbed systems lead to the assignment of the higher energy C6 phonon as an overtone. The discrepancy between the expected energy, i.e. twice that of the first excitation, and the observed energy likely is due to anharmonicity. The lack of an overtone for C8 and C10 is likely be due to their larger Debye-Waller Factors, which would decrease the overall probability, Pm.
We have performed a series of experimental and simulation studies in order to dissect the forces driving the self-assembly of alkanthiolate monolayers. The striped phases of 1-decanethiol, 1-octanethiol, and 1-hexanethiol exhibit FTz phonons at 8.0, 7.3, and 7.3 meV, respectively. The 12.3 meV overtone for the C6/Au(111) striped phase is assigned as the overtone of the FTz mode. Alkanethiols on Au(111) exhibit a distinct single phonon inelastic scattering event with a largely invariant energy. The qualitative molecular dynamics simulations also suggest that the frequencies of the low energy vibrations are relatively constant with respect to chain length. The interactions responsible for these phonons are driven by the van der Waals forces between the methylene/methyl groups and the Au substrate atoms. We also suggest that the modestly increased energy of the C10 mode with respect to C8 and C6 comes from increased binding due to the commensurate structure that C10 has with respect to the Au(111) reconstruction. This work represents a step forward in understanding the forces that govern interfacial self-assembly.
105. "Surface Vibrations in Alkanethiol Self-Assembled Monolayers of Varying Chain Length
