Session V1: Focus Session: Advances in Scanned Probe Microscopy III: SPM of Single Atoms & Molecules

Non-contact friction and the relaxational dynamics of surface defects

When a cantilever oscillates near a sample surface, it experiences a dissipative force. Such non-contact friction is of great practical importance to the ultrasensitive force detection measurements. Previous theories predict the friction coefficient to be six orders of magnitude smaller than the experimental value. Here, inspired by the recent experiments reported in Phys. Rev. Lett. 105, 236103 (2010), we propose two new mechanisms to explain the magnitude, as well as the distance, temperature and frequency dependence of the friction. We assume that the surface defects couple to the cantilever tip either in the spin or charge channel, and their relaxational dynamics gives rise to the observed behavior of both the friction coefficient and the induced spring constant. Nice agreement is found with the experiments.   

Dissecting an organic-metal bond by molecular manipulation

\\ Scanning probe microscopy with a dynamic AFM has been able to answer some fundamental questions of surface science, like the force necessary to move an atom[1]. Recently, we demonstrated the gradual removal of a single 3,4,9,10-perylenetetracarboxylic-dianhydride (PTCDA) molecule from Ag(111) using a dynamic AFM [2]. The continuous force gradient measurement allowed the structural control over the junction. Here, we show how to extract details of the molecule-surface bonding (physisorption, chemisorption, bonding via functional groups) from such an experiment. The importance of a full-fledged simulation of the lifting -including tip oscillation- is emphasized. We point out the necessity of, and fundamental problems related to, a curved tip trajectory. We study PTCDA on Au(111) and Ag(111) and find qualitatively and quantitatively different binding potentials and adsorption energies. The data represents an ideal benchmark for existing and future ab-initio calculations on these systems. Our method should be applicable to various substrate-adsorbate systems and hence has the potential to answer many open questions in the field.\\{[}1] M. Ternes et al., \textit{Science} \textbf{319}, 1066 (2008)\\{[}2] N. Fournier et al., \textit{Phys. Rev. B} \textbf{84}, 035435 (2011)   

Multi-frequency amplitude modulated non-contact atomic force microscopy for nanoscale dielectric measurements

Multi-frequency non-contact atomic force microscopy with amplitude feedback in air was used to obtain the dielectric constant of ultra-thin films on metallic substrates. The cantilever was excited at its second bending mode by applying an AC electric field between the substrate and cantilever. The capacitance gradient between the cantilever tip and sample substrate was obtained by measuring the capacitive force driving the cantilever at its second bending mode. An analytic expression relating capacitance and dielectric constant of thin film was then used to fit the experimental data and the dielectric constant was obtained from the fit parameters. The method was validated by obtaining the dielectric constants of self-assembled monolayers of thiol molecules ($2.0 \pm 0.1$) on gold substrate, and sputtered SiO$_2$ ($3.6 \pm 0.07$) thin film. The high Q-factor of the second bending mode of the cantilever increases the accuracy of capacitive measurements while the low applied potentials minimize the likelihood of variation of dielectric constants at high field strength and of damage from dielectric breakdown of air.   

Anomalously large $g$-factor of single atoms adsorbed on a metal substrate

We performed magnetic field dependent inelastic scanning tunneling spectroscopy (ISTS) on individual Fe atoms adsorbed on different metal surfaces. ISTS reveals a magnetization excitation which is shifting linearly to higher energies in the magnetic field. The data is used to extract the magnetic anisotropies and the $g$-factors of the Fe atoms, as well as the lifetimes of the excitations. We find lifetimes of hundreds of femtoseconds limited by coupling to electron-hole pairs in the substrate and decreasing linearly upon application of the magnetic field. As expected, the magnetic anisotropy strongly depends on the substrate. Astoundingly, we find that the $g$-factor is $g \approx 3.1$ for Ag(111) instead of the regular value of 2 which is observed for the Cu(111) substrate [1]. This very large $g$-shift can be understood when considering the complete electronic structure of both the Ag(111) surface state and the Fe atom, as shown by \textit{ab initio} calculations of the magnetic susceptibility. \\[4pt] [1] A. A. Khajetoorians et al., Phys. Rev. Lett. \textbf{106}, 037205 (2011).   

Spin Parity effects in STM single magnetic atom manipulation

Recent experimental work shows that a spin polarized scanning tunneling microscopy tip can be used both to read and write the spin orientation of a single magnetic spin [1]. Inelastic electron tunneling spectroscopy (IETS) shows that spin of the magnetic atom is quantized [2], like the spin of a molecular magnet. Here we discuss two fundamental problems that arise when a bit of classical information is stored on a quantized spin: quantum spin tunneling and back-action of the readout process. Quantum tunneling is responsible of the loss of information due to the relaxation of the spin coupled to the environment, while the detection induced back-action leads to an unwanted modification of the spin state. We find that fundamental differences exist between integer and semi-integer spins when it comes to both, read and write classical information in a quantized spin.\\[4pt] [1] S. Loth et al, Nature Physics 6, 340 (2010).\\[0pt] [2] C. Hirjibehedin et al, Science 317, 1199 (2007).   

Magnetic Particle Imagining with a Cantilever Detector

We present a novel method for the measurement of the magnetic moment of single micrometer scale particles as a function of the applied field. Our technique is based on magnetic force microscopy (MFM) with a hard magnetic tip in a uniform opposing field. By moving the tip position and using a simple model we can extract the magnetic properties of isolated particles and precisely measure subsurface particle locations in 3D.   

Scanning Probe Microscopy and Spectroscopy of Molecules on Thin Insulating Films

Ultrathin insulating films on metal substrates facilitate the use of the scanning tunneling microscope (STM) to study the electronic properties of single atoms and molecules, which are electronically decoupled from the metallic substrate. The ionic relaxations in a polar insulator lead to a charge bi-stability in some adsorbed atoms and molecules. It is shown that control over the charge-state of individual molecules in such systems can be obtained by choosing a substrate system with an appropriate work function. The distribution of the additional charge is studied using difference images. These images show marked intra-molecular contrast [1]. We discuss how atomic-force-microscopy (AFM) in a combined STM/AFM based on the qPlus-sensor [2] reveals additional information that is truly complementary to the STM data set. In the case of a non-planar molecule that shows two different adsorption geometries, only the AFM channel provides reliable information on the conformation of the molecule. In another example of an artificially formed molecule-metal-molecule complex, the AFM channel provides information on the bonding that is important to understand the STM results. \\[4pt] [1] I. Swart, T. Sonnleitner, and J. Repp, Nano Letters 11, 1580 (2011).\\[0pt] [2] F. J. Giessibl, Appl. Phys. Lett. 76, 1470 (2000).   

Tailoring Dirac Fermions in Molecular Graphene

The dynamics of electrons in solids is tied to the band structure created by a periodic atomic potential. The design of artificial lattices, assembled through atomic manipulation, opens the door to engineer electronic band structure and to create novel quantum states. We present scanning tunneling spectroscopic measurements of a nanoassembled honeycomb lattice displaying a Dirac fermion band structure. The artificial lattice is created by atomic manipulation of single CO molecules with the scanning tunneling microscope on the surface of Cu(111). The periodic potential generated by the assembled CO molecules reshapes the band structure of the two-dimensional electron gas, present as a surface state of Cu(111), into a ``molecular graphene'' system. We create local defects in the lattice to observe the quasiparticle interference patterns that unveil the underlying band structure. We present direct comparison between the tunneling data, first-principles calculations of the band structure, and tight-binding models.   

Synthesizing and Observing Electric and Magnetic Gauge Fields in Molecular Graphene

Molecular graphene is an artificial analogue of graphene that can be built using a scanning tunneling microscope (STM). It is realized by constraining surface-state electrons to a honeycomb lattice, which we have shown reproduces the Hamiltonian of natural graphene. We experimentally demonstrate that creating strains within the honeycomb lattice modifies the Hamiltonian in the same way that true laboratory-frame electric and magnetic fields do. In our experiments we have created artificial magnetic fields that can reach values as high as 60 T, and which can change their magnitude significantly over a distance as short as 2 nm. Some of the new physical phenomena we have been able to observe include: gauge invariance for lattices with very different real-space strains but the same artificial fields, emergence of Landau levels due to constant artificial magnetic fields, occurrence of a confined state at an interface straddling a sign-inversion of the effective magnetic field, and effective chemical doping and potential changes associated with applied electric gauge fields.   

Topologically protected chiral edge state realized on molecular graphene

Graphene has many interesting topological properties arising from the hexagonal lattice shape and sublattice symmetry. Although its edge state is known to be extremely sensitive to the shape of the edge, and is non-topological, it has been shown that graphene with an energy gap produced by broken sublattice symmetry can possess different topologies due to different sign of the carrier mass. By making a junction of two gapped graphene regions with opposite mass, we can realize a topologically protected edge state at the mass domain wall junction, with chirality emerging from the valley degeneracy. Molecular graphene is an artificial honeycomb lattice built by atom manipulation, and due to its extreme tunability, we have realized a topological edge state between gapped molecular graphene regions with opposite signs of Dirac fermion mass. Enhanced density of states restricted only to the junction clearly shows the existence of the edge state. Also, its robustness to the geometrical detail is observed from its persistence over various edge structures, and chirality is revealed by selective scattering at the intersection of edges.   

Measuring Electronic Properties of Topological Defects in Molecular Graphene

With the development of artificial ``molecular'' graphene, it is possible to create a two-dimensional electron system very similar to graphene by assembling molecules in the appropriate geometry on surface states using a scanning tunneling microscope (STM) tip. Using this same system, we recreate many lattice defects that occur naturally in graphene. Such defects have a significant effect on the electronic and transport properties of natural graphene, and are thus of notable interest in the development of nanoelectronics. In particular, we study rotational grain boundaries, which are formed by the rotation of a region of graphene with respect to the rest of the lattice. These include the Stone-Wales defect, the simplest with two adjacent carbon sites rotated by 90 degrees, as well as a common larger topological defect recently identified as the flower defect. Using STM, we examine the electronic properties of these defects in molecular graphene, paying particular attention to the emergence of new states close to the Dirac point and the quasiparticle scattering. These geometries are also studied in hole- and electron-doped variants.   

Scanning Vibronic Spectroscopy of Single Molecules

The dynamics of a single molecule within a quantum point contact or tunnel junction can be quite complex, exhibiting motion over frequency scales ranging from kHz to THz. In this work, we examine in detail the dynamic motion of single molecules by using low-temperature scanning inelastic tunneling, isotopic sorting, and isotopic labeling. Carbon-based molecules such as carbon monoxide, C$_{60}$, and diamondoids function as prototype homonuclear and heteronuclear molecules. Target structures are assembled on surfaces via atomic manipulation or self-assembly in order to tailor local THz-scale vibrational modes and to control kHz-scale molecular motion. In addition we functionalize a scanning tunneling microscope tip and use its own THz-scale vibrational modes as a probe and a tunable perturbation. These techniques reveal structure not visible in traditional STM data. We compare this data to expected local quantum forces.   

Pitfalls in normalizing scanning tunneling spectroscopy and how to do it ``right'' (from an experimentalist perspective)

Scanning tunneling spectroscopy is widely used to probe the local density of states (LDOS) by measuring $\mathrm{d}I/\mathrm{d}V$ typically using a lock in amplifier. According to theory, the data has to be normalized by $(I/V)$ to reveal the true LDOS - of cause still convoluted with the tip density of states. This process can induce its own artifacts and has to be done carefully. In our case the algorithm includes finding the true bias offset (if any), correct any offset in the measured current, rescale the measured $\mathrm{d}I/\mathrm{d}V$ curves using the calculated version $(\Delta I/\Delta V)$, replacing $I(V)$ around $V_{\mathrm{bias}}=0$ by a polynomial and finally performing the normalization. The motivation and consequences of each step will be discussed using example data.