Structure sensitive in situ Raman study of iron passive films using SERS-active $\mathrm{Fe}∕\mathrm{Au}\left(111\right)$ substrates

This work describes the preparation of well-defined $\mathrm{Fe}∕\mathrm{Au}\left(111\right)$/mica substrates for in situ Raman studies of the iron passive film with surface-enhanced Raman scattering (SERS). It is a two-step technique in which a SERS active (111) epitaxial gold substrate is prepared by resistive evaporation on mica. An epitaxial $\mathrm{Fe}\left(110\right)$ film is then electrodeposited to serve as iron electrode. It is shown that the SERS enhancement factor depends primarily on the good matching between the gold film plasmon resonance wavelength ${\lambda }_{\mathrm{P}}$ and the excitation wavelength. The iron thickness is the second main parameter controlling the SERS enhancement factor with a maximum found for a thickness of 20 monolayers. Under optimized conditions an amplification factor of $5\times {10}^4$ is demonstrated with respect to the case of a bulk polycrystalline iron substrate. This technique allows the recording of Raman spectra of nm-thick passive films within a few seconds only, which opens up to nearly real-time bias-dependent investigations of the chemistry at the electrochemical interface. In addition, taking advantage of the well-defined structure of the $\mathrm{Fe}\left(110\right)∕\mathrm{Au}\left(111\right)∕$mica substrates, we present a preliminary structure-sensitive in situ Raman study of the iron passive film formed in a borate solution of $p\mathrm{H}8.4$. It is thought that this simple approach of promoting SERS should find more general interest for the electrochemistry community.

Figure 1

(A) and (B) $2.5\times 2.5{\mu \mathrm{m}}^2$ AFM images and cross sections of a SERS-active evaporated $\mathrm{Au}∕\mathrm{mica}$ film (film $\mathrm{Au}\text{−}2$, see Table ) and SERS-inactive sputtered $\mathrm{Au}∕\mathrm{mica}$ substrate. (C) Absorbance spectra of $\mathrm{Au}∕\mathrm{mica}$ substrates showing that the surface plasmon resonance wavelength ${\lambda }_{\mathrm{P}}$ depends on the substrate: (a)–(d) spectra of $\mathrm{Au}\text{−}1$ to $\mathrm{Au}\text{−}4$, respectively (see Table ). The reference substrate is the one imaged in (B).

Figure 2

Characterization of $\mathrm{Fe}\text{−}\mathrm{Au}$ electrodes: (a) in situ STM image of a $5\text{−}\mathrm{ML}$-thick iron film grown at $-1.5\mathrm{V}$ on a flat top $\mathrm{Au}\left(111\right)$ grain of a polycrystalline gold film evaporated on mica. See text for the deposition conditions. All steps are monatomic. Numbers in the image indicate the local iron thickness in ML. (b) Cyclic voltammogram of a $20\text{−}\mathrm{ML}$–thick electrodeposited $\mathrm{Fe}$ film in a $0.05M{\mathrm{H}}_3{\mathrm{BO}}_3+0\mathrm{.0125}M{\mathrm{Na}}_2{\mathrm{B}}_4{\mathrm{O}}_7·{10\mathrm{H}}_2\mathrm{O}$ solution of $p\mathrm{H}8.4$. The $\mathrm{Fe}\text{−}\mathrm{Au}$ electrode was prepolarized at $-1.5\mathrm{V}$ for $300\mathrm{s}$ prior to starting the positive-going sweep of potential $\left(5\mathrm{mV}∕\mathrm{s}\right)$ from $-1.5\mathrm{V}$ to $+0.2\mathrm{V}$ and back to $-1.5\mathrm{V}$. Inset: same for the poly-$\mathrm{Fe}$ electrode.

Figure 3

(A) Ex situ Raman spectra of native oxide films grown on $18\text{−}\mathrm{ML}$-thick $\mathrm{Fe}\left(110\right)$ films electrodeposited on different $\mathrm{Au}∕\mathrm{mica}$ substrates. Spectra (a)–(c) were recorded in $60\mathrm{s}$ using $\mathrm{Au}\text{−}4$, $\mathrm{Au}\text{−}3$, and $\mathrm{Au}\text{−}2$ as gold substrates, respectively (see Table ). (B) Variations of the Raman intensity as a function of iron thickness in the case of the $\mathrm{Au}\text{−}2$ substrate. Note that no spectrum could be measured on poly-$\mathrm{Fe}$.

Figure 4

In situ Raman results. (a) Spectrum of an anodic oxide film grown at $+0.2\mathrm{V}$ on an $18\text{−}\mathrm{ML}$-thick $\mathrm{Fe}$ film electrodeposited on $\mathrm{Au}\text{−}2$. (b) Same for a poly-$\mathrm{Fe}$ electrode. Inset of (b): spectrum of maghemite powder $\gamma {\text{−}\mathrm{Fe}}_2{\mathrm{O}}_3$. Spectrum (a) was recorded in $60\mathrm{s}$ and spectrum (b) in $1800\mathrm{s}$.

Figure 5

(a) same as Fig. 4a after reduction of the oxide film at $-1.1\mathrm{V}$ for $30\mathrm{s}$. The spectrum was recorded in $3\mathrm{s}$. (b) Orientation dependence of ex situ Raman of a (111) ${\mathrm{Fe}}_3{\mathrm{O}}_4$ single crystal. Spectra (a) and (b) were obtained with the (111) axis, respectively, $\Vert$ and $\perp$ to the incident laser beam (respectively $\perp$ and $\Vert$ to the light polarization).