Sythesis and characterisation of ferrocenyl monolayers on silicon surfaces

Abstract

A series of ferrocenyl monolayers on hydrogen-terminated Si(111) single crystal and porous silicon have been prepared and characterised by atomic force microscopy (AFM), X-ray photoelectron spectroscopy (XPS), and cyclic voltammetry (CV). Hydrosilylation modification of hydrogen-terminated Si(111) single crystal and porous silicon with vinylferrocene and ethynylferrocene was shown to produce three different silicon–carbon tethered ferrocenyl surfaces, Si-CH₂-CH₂-Fc, Si-CHCH-Fc, and Si-CC-Fc. Three different grafting procedures were used, a thermally activated method at 110 C in toluene using vinylferrocene produced the Si-CH₂-CH₂-Fc monolayer, EFc. Hydrosilylation at room temperature using ethynylferrocene in the minimum amount of dichloromethane produced the Si-CHCH-Fc surface, VFc. Finally, the reaction of n-butyllithium with ethynylferrocene via a nucleophilic substitution reaction (acetylide anion) at the silicon hydride electrode produced the Si-CC-Fc surface, EnFc. Transmission FTIR analysis confirmed that three different linkages were formed on porous silicon. Data recorded using cyclic voltammetry showed that ferrocenyl silicon monolayers were more stable in acetonitrile than aqueous electrolyte. The maximum surface coverage measured in acetonitrile was 1.3×10-¹º mol cm-₂ for the EFc surface, 2.7×10-11 mol cm-₂ for the VFc surface, and 5.4×10-¹¹ mol cm-₂ for the EnFc surface. However, reproducibility of the EFc, VFc and EnFc surfaces was not easy to control. In order to find a way to improve reproducibility, a series of 50:50 mixtures of vinylferrocene and simple alkenes were investigated. The alkenes used for this purpose were hexene (EFc/C₆), octene (EFc/C₈), decene (EFc/C₁₀), and undecene (EFc/C₁₁). AFM, XPS and CV were used to analyse these surfaces. Different surface coverages for 50:50 mol:mol in bulk solution of EFc/C₆, EFc/C₈, EFc/C₁₀, and EFc/C₁₁ were obtained from CV in aqueous and acetonitrile electrolytes. Higher coverage was observed in acetonitrile than water for all 50:50 mixed monolayers, and the EFc/C₈ surface was the more reproducible surface with a surface coverage of 4.0×10-¹¹mol cm-². Therefore, a range of more dilute EFc/C₈ surfaces, 20:80, 10:90 and 1:99 were prepared via reaction with a vinylferrocene/octene mixture and characterised using AFM, XPS and CV. The cyclic voltammograms of dilute mixed monolayers EFc/C₈ gave different surface coverages in acetonitrile electrolyte. The average observed were 3.03×10-¹¹ mol cm-², 1.84×10-¹² mol cm-² and 7.95×10-¹³ mol cm-² for 20:80, 10:90 and 1:99 EFc/C₈ surfaces respectively. The coverage of mixed EFc/C₈ monolayers determined via CV was found to depend on the mole fraction of vinylferrocene in the initial alkylation solution, and it is possible to control the separation of ferrocene units on the silicon surface by controlling the mole fraction. In addition, longer alkyl chain linkages between the ferrocene and silicon surface were investigated by two different approaches. In the first approach a ferrocenyl group with an extended tether was first synthesised, 1-(but-1-en-3-yne)ferrocene, and was then attached as a complete unit to the silicon surface using a one-step reaction to give the BuFc surface. In the second approach, the ferrocene monolayer was built up from the silicon surface in a two-step reaction procedure. In the first step, the silicon hydride surface was modified with 5-iodo-3-ethyl-2-ethynylthiophene to give the Tp surface. In the second step, the Tp surface was coupled to ethynylferrocene via the Sonogashira reaction to give the TpFc surface. In XPS a new Fe 2p peak and the concurrent disappearance of the I 3d peak was observed for the TpFc surface, which confirmed the replacement of the iodine by ethynylferrocene. The surface coverage obtained from CV for BuFc, (6.25×10-¹¹ mol cm-²), was higher than for TpFc surfaces, (7.2×10-1¹² mol cm-²). This is could be due to a lower overall reaction yield being associated with each reaction in the two-step process. In order to investigate the utility of ferrocene modified silicon surfaces in applications such as charge storage and memory devices, the EFc surface was studied using electrochemical scanning tunnelling spectroscopy which showed enhanced tunnelling current response that is directly related to the presence of the redox active ferrocene unit. The rate of electron transfer from ferrocenyl-monolayers to n-Si(111) was measured for both pure and mixed monolayers using light step photoelectrochemical chronoamperometry (PC). The rate constant associated with ferrocene oxidation is weakly dependent on applied potential, and is independent of the type of linkage, and the length of the n-alkyl chains in the mixed monolayers. The photoelectrochemical experiment provides a simple means to detect the extent to which the monolayer discharges in the dark and therefore can be used to assess the feasibility of charge storage in these monolayers.