Noncontact atomic force microscopy imaging of water dissociation products on ${\mathrm{TiO}}_2\left(110\right)$

We have imaged the products of water dissociation on ${\mathrm{TiO}}_2\left(110\right)$ with noncontact atomic force microscopy (NC-AFM). Scanning tunneling microscopy (STM) experiments have revealed that water splits into a hydroxyl group that sits in an O vacancy and an H adatom which lies on a bridging O atom. With this knowledge, STM images acquired in the same area allow an unambiguous identification of these H adatoms in the corresponding NC-AFM images.

Figure 1

Ball model of ${\mathrm{TiO}}_2\left(110\right)$ including an O vac and an H adatom. Large grey spheres depict lattice oxygen with atoms nearer the surface shaded lighter. Small black spheres indicate Ti atoms. The O vac is circled and the H adatom is represented with a small grey sphere.

Figure 2

(Color) $180\mathrm{\AA }\times 250\mathrm{\AA }$ images and line profiles of ${\mathrm{TiO}}_2\left(110\right)$. (a) STM image with bias of $1.4\mathrm{V}$ and tunneling current, $1.0\mathrm{nA}$. Blue and red arrowheads indicate O vacs and H adatoms, respectively. Black crosses indicate unknown contaminants. (b) NC-AFM image recorded with $\mathrm{\Delta }f=\sim -63\mathrm{Hz}$ and applied voltage, $1.4\mathrm{V}$. The NC-AFM image was recorded $\sim 3\mathrm{mins}$ before the STM image in (a). Blue and red arrowheads are marked where depressions correspond to the positions of O vacs and H adatoms in (a), respectively. Green and yellow arrowheads indicate positions where defects are not apparent in (b) but are visible in (a). The green arrowheads clearly lie on areas where the tip is unstable. The defects marked with yellow arrowheads presumably correspond either to defects which cannot be imaged due to a more transient instability or they correspond to defects which have migrated to different positions. Together, the defects which are not imaged clearly in (b) account for $\sim 23\%$ of all the defects in (a). Pink arrowheads indicate defects visible in (b) but not in (a) thereby giving evidence for the migration of defects. For clarity, the images in (a) and (b) are duplicated in (c) and (d), respectively. (e) Representative line profiles taken over an O vac (blue) and an H adatom (red) in the positions indicated by the coloured lines in (a). (f) 16 line profiles taken over depressions in (b) assigned to H adatoms (red) and O vacs (blue).

Figure 3

(Color) $115\mathrm{\AA }\times 185\mathrm{\AA }$ NC-AFM images of ${\mathrm{TiO}}_2\left(110\right)$ recorded with $\mathrm{\Delta }f=\sim -45\mathrm{Hz}$ and applied voltage, $1.4\mathrm{V}$. (a) Recorded in Regime I. (b) Image recorded in Regime II following a tip change between the images. Purple arrowheads indicate coincident type-$A$ defects. The blue circle indicates the position of a defect in (a) which appears to have migrated to the right of the frame and is marked with a blue arrow in (b). The purple circle indicates a defect which is observable in (a) but not in (b). We presume this is also due to migration either away from the scan area or onto the tip (Ref. 7). These two defects which are not in the same position as in (a) amount to $\sim 7\%$ of the total number of defects visible.

Figure 4

(Color) (a) $250\mathrm{\AA }\times 90\mathrm{\AA }$ NC-AFM image of ${\mathrm{TiO}}_2\left(110\right)$ recorded with $\mathrm{\Delta }f=\sim -63\mathrm{Hz}$ and applied voltage, $1.8\mathrm{V}$. The contrast has been adjusted to highlight the upper terrace. (b) As (a) with the contrast optimized for the lower terrace. (c) Simultaneously recorded ${\overline{I}}_{\mathrm{t}}$ image. In (a) and (b), type-$A$ defects are indicated with purple arrowheads and blue lines are drawn over three bright rows. An identical array of arrowheads and lines are superimposed in the same positions on (c). The blue lines in (c) lie on dark rows and the arrowheads point at type-A defects which appear bright and centred on the bright rows in the ${\overline{I}}_{\mathrm{t}}$ image. A circle indicates an unidentified impurity in all three images. In all three images, the ellipse indicates three defects which appear dark in the ${\overline{I}}_{\mathrm{t}}$ image due to the error-contribution. (d) Line profile taken over the red line in (b) showing the position of the defects (between the bright rows) in the topographic image. A sine wave has been fitted to the rows on the right hand side of the graph and is superimposed as a grey curve. (e) Line profile taken over the yellow line in (c) showing the position of the defects (atop the bright rows) in the ${\overline{I}}_{\mathrm{t}}$-image. A sine wave has been fitted to the rows on the right hand side of the graph and is superimposed as a grey curve. (f) A graph showing calculated ${\overline{I}}_{\mathrm{t}}$ vs distance. The grey curve is an exponential fit to the data.