H3PW12O40-functionalized tip for scanning tunneling microscopy
In K. Song, John R. Kitchin, and Mark A. Barteau, “H3PW12O40-functionalized tip for scanning tunneling microscopy”, Proceedings of the National Academy of Sciences 99(suppl. 2), 6471-6475 (2002). https://doi.org/10.1073/pnas.072514399
This summary was written by a language model from the paper and has not been checked line by line. Rely on the published paper for details.
(The formula in the title is H3PW12O40, the Keggin-type phosphotungstic acid.)
Question
The resolution and stability of a scanning tunneling microscope depend on the atomic structure and chemical identity of the tip apex, which conventional tip preparation methods do not control reproducibly. Earlier work had shown that attaching C60 molecules or carbon nanotubes to metal tips could give a chemically defined probe. This paper, from a Sackler colloquium on nanoscience, asks whether a heteropolyacid (HPA), a molecular metal oxide cluster with a well-defined structure and characteristic tunneling spectrum, can be attached to a Pt/Ir tip to make a stable, atomically sharp STM probe (Song et al. 2002).
Approach
The authors chose the Keggin-type acid H3PW12O40 for its structural simplicity. This roughly spherical anion, about 11 to 12 Å across, has a central PO4 tetrahedron surrounded by twelve WO6 octahedra, each with an outward-pointing terminal W=O oxo group (bond length 1.7 Å). A mechanically formed Pt/Ir (90/10) tip was dipped into a drop of roughly 0.01 M aqueous H3PW12O40, kept pointing downward, and dried in air for 1 h at room temperature. Tunneling spectra (sample bias swept from -2 to +2 V) on highly oriented pyrolytic graphite were recorded first to confirm molecule attachment, using two different instruments (Topometrix TMX 2010 and LK Technologies LK-1000). Imaging was done in air in constant current mode at +100 mV sample bias and 1 to 1.5 nA. Test samples were bare graphite and monolayers of the Wells-Dawson HPA H6P2W18O62 deposited on graphite. Lattice constants came from 2D fast Fourier transforms of unfiltered images, commensurability was assessed with the EPICALC program, and a multiple-tip image simulation (up to four tips with arbitrary arrangement, weights and heights) was used to test one explanation for unusual images (Song et al. 2002).
Key findings
Spectra taken on bare graphite with the modified tip showed negative differential resistance (NDR), which graphite alone does not show, indicating that electrons tunneled through attached H3PW12O40. The NDR peak voltage averaged -1.15 ± 0.07 V (62 spectra) with the functionalized tip, nearly the same as the -1.14 ± 0.09 V (24 spectra) measured for an H3PW12O40 monolayer on graphite probed by a bare Pt/Ir tip; the instrument voltage resolution was 0.05 V (Song et al. 2002).
The modified tips frequently gave atomic resolution images of graphite with the expected 2.46 Å hexagonal periodicity, and molecular resolution images of ordered H6P2W18O62 arrays with a primitive cell of 14.6 Å sides and a 35.2° included angle, consistent with the 11 Å by 14.5 Å ellipsoidal anion. The authors argue that tunneling occurs through a single terminal oxo group, an electronegative atom bound to W, so the effective tip is atomic rather than the 1 nm molecular scale, and that the rigid molecule fixes the position of this oxo atom (Song et al. 2002).
About 20% of the time, graphite images with HPA tips showed hexagonal superperiodic structures superimposed on the atomic lattice, with periodicities of 70.4 Å (rotation angle 27° or 33° from the graphite lattice) and 14.97 Å (25.3° or 34.7°). EPICALC gave a dimensionless potential of 0.5 in both cases, indicating commensurate overlayers. These features were insensitive to set point, scan angle and scan rate and were never seen with bare Pt/Ir tips on the same instruments. The multiple-tip simulations produced no long-range superperiodicity, so the authors attribute the patterns to Moiré rotation of the top few graphite layers, and attribute the enhanced sensitivity of HPA tips to the discrete electronic levels of the molecular oxide, by analogy with C60-functionalized tips (Song et al. 2002).
Significance
The work extends molecular tip functionalization from conductive C60 to semiconducting metal oxide clusters. Because the NDR voltage of an HPA can be shifted by choice of structure and composition, the authors suggest HPA-modified tips could provide a tunable handle on the electronic properties of STM probes for studying surface and subsurface electronic structure. They also note that controlling the number and orientation of molecules on the tip remains a challenge (Song et al. 2002).