Session A01: Advances in Scanned Probe Microscopy I

Data Mining in Scanning Probe Microscopy

Scanning probe microscopy instrumentation and techniques for data acquisition have made great leaps over the past decade, producing ever more complex and large multidimensional datasets. This necessitates the development of new instrumentation and techniques for data analysis, in order to better understand this new wealth of data. Fortunately, these developments are paralleled by the general rise of big data and the development of a number of data mining techniques which can extract hidden structure from large datasets. In this talk, I will describe the use of unsupervised learning techniques (e.g. clustering, feature extraction, and spectral mixing) that can identify patterns in data by simultaneously analyzing multiple variables. Using these methods we can uncover statistically significant components and trends within large multidimensional datasets, then compress and highlight these components, allowing us to both better visualize the data and extract physically relevant information.   

Cancelling Picometer Vibrations from a Scanning Tunneling Microscope by Post Processing

Scanning tunneling microscopy (STM), a crucial tool for atomic resolution measurements of the electronic structure of condensed matter systems, typically requires an extremely stable environment to operate. Modern ultra-low vibration laboratories are designed to reduce external vibrations from the micron scale to the nanometer scale, by employing large pneumatic isolators to float inertial blocks. Improvements beyond this benchmark require careful STM head design to minimize weight and reduce internal resonances. However, even the best microscopes experience picometer vibrations in the tip-sample junction. Here we introduce a software algorithm that can reduce residual picometer vibrations by more than 50%. Our method focusses on accurately measuring the transfer function for vibrations to reach the tip. By measuring vibrations with a sensitive geophone, the transfer function allows us to predict their induced noise at the tip. We apply this technique to topographic data acquired on a passively isolated STM, and are able to reduce the tip-sample rms displacement from 0.6 pm to 0.35 pm. We anticipate extending this method to reduce noise in spectroscopic measurements of the density of states.   

Studies of Magnetic Substrates for SP-STM

Spin-polarized scanning tunneling microscopy (SP-STM) is a technique that combines atomic scale spatial resolution with magnetic sensitivity. To perform detailed measurements of novel spin systems, it is necessary to accurately characterize the magnetization and typical behavior of different spin-polarized tips using a well-known magnetic substrate. Here, we present our work on Cr(001), where alternating terraces are antiferromagnetically coupled, and Fe/Ir(111) which hosts a variety of magnetic orderings depending on film thickness in the few layer regime. For Cr(001), a 1200° C flash anneal results in the appearance of a complex surface reconstruction, island formations, and self-assembled Cr nanowires. Tunneling spectroscopy of these features typically display Kondo-like resonances that may reflect spin-screening or spin-polarized surface states. In preliminary work on Fe/Ir(111), we find the surface is partially covered by monolayer films and bilayer islands. Resolving magnetic contrast in these films will allow us to precisely characterize the magnetization of our tips in three spatial dimensions, which can then be used to study and map spin in other systems.   

Inter-Molecule Interaction for Magnetic Property of Vanadyl Tetrakis(thiadiazole) Porphyrazine Film on Au(111)

Metal phthalocyanine has been widely investigated and its magnetic behaviour of their film attracts much attention in relation to the spintronic application. Controlling the magnetic properties with changing molecule-assembly is a great challenge. However, it’s not easy to control the assembly with C-H terminated Pc molecules. Here, we investigate a film formation and a magnetic property of vanadyl-tetrakis(thiadiazole)porphyrazine(VOTTDPz) molecules on Au(111) with using STM.
Figure 1 shows the well-ordered film of VOTTDPz molecules on Au(111). Unlike the square-like lattice of VOPc on Au(111), the film shows the three-fold symmetry with nearest-neighbor distance of ~1.1 nm which is shorter than that of 1.4 nm for VOPc/Au(111), suggesting stronger attractive force than that formed by CH. Figure 2 shows a STS near the Fermi level. Sharp Kondo peaks derived from the molecules unpaired spin. The presence of a spin in VOTTDPz is consistent with previous EPR data and there exists S=1/2 at the V atom.
We detected asymmetric and weak Kondo peak for outer magnetic field of 0 T, which becomes strong and symmetric peak at 5 T, which is understood by the shift of the spin center of the Kondo resonance from V 3d to delocalized p state in ring with the magnetic field.   

Nanoscale thermal imaging of dissipation from individual atomic defects in graphene

Energy dissipation is a fundamental process governing the dynamics of classical and quantum systems. Despite its vital importance, direct imaging and microscopy of dissipation in quantum systems is currently impossible because the existing thermal imaging methods lack the necessary sensitivity and are unsuitable for low temperature operation. We developed a scanning nanoSQUID with sub 50 nm diameter that resides at the apex of a sharp pipette [1] acting simultaneously as nanomagnetometer with single spin sensitivity and as nanothermometer providing cryogenic thermal imaging with four orders of magnitude improved thermal sensitivity of below 1 µK/Hz^1/2 [2]. The non-contact non-invasive thermometry allows thermal imaging of minute energy dissipation down to the fundamental Landauer limit of 40 fW for continuous readout of a single qubit at 1 GHz at 4.2 K. By varying potential between the SQUID-on-tip and the sample a nanoscale spectroscopic analysis of the dissipation process can be attained. Using this scanning nano-thermometry we visualize and control phonon emission due to inelastic electron scattering off individual atomic defects in graphene [3]. The inferred electron-phonon “cooling power spectrum” exhibits sharp peaks when the Fermi level comes into resonance with electronic quasi-bound states at such defects, a hitherto uncharted process. The atomic defects are very rare in the bulk but abundant at the edges, acting as switchable atomic-scale phonon emitters that establish the dominant dissipation mechanism in graphene.

[1] Vasyukov et al., Nature Nanotech. 8, 639 (2013).
[2] D. Halbertal, J. Cuppens, M. Ben Shalom, L. Embon, N. Shadmi, Y. Anahory, H. R. Naren, J. Sarkar, A. Uri, Y. Ronen, Y. Myasoedov, L. S. Levitov, E. Joselevich, A. K. Geim, and E. Zeldov, Nature 539, 407 (2016).
[3] D. Halbertal, M. Ben Shalom, A. Uri, K. Bagani, A.Y. Meltzer, I. Marcus, Y. Myasoedov, J. Birkbeck, L.S. Levitov, A.K. Geim, and E. Zeldov, arXiv:1710.01486 (Science in press).   

Visualizing the Coulomb blockade in graphene quantum dots, Part I

The charge of a conductor separated from particle reservoirs by tunnel junctions is quantized in units of the elementary charge, a phenomenon known as Coulomb blockade. Recent experiments have enabled the creation of graphene QDs with fixed build-in potentials inside circular p-n junctions by ionizing impurities in the boron nitride underlying insulator. In these small nanometer sized circular resonators the quasi-bound resonances in zero magnetic field can be confined further by the application of a perpendicular magnetic field forming quantized Landau levels (LL) inside the graphene QD. The LLs at high magnetic fields form a series of metallic rings, separated by highly insulating impressible rings, allowing tunnel barriers to be created between the LL quantum liquids and the sample bias electrode. The isolated LL metallic rings are then accessible by Coulomb blockade spectroscopy between the STM probe and graphene sample. Using scanning tunneling spectroscopy we provide direct spatially and spectroscopically resolved measurements of the formation of the LL rings and their charging characteristics. We investigate the addition energy spectrum of the LL rings and analyze their charging characteristics in terms of capacitances and QD energy level structure.   

Visualizing the Coulomb blockade in graphene quantum dots; Part II

The Coulomb blockade (CB) is one of the most characteristic phenomena of nanoscale artificial atoms. Here, we created circular quantum dots (QD) in exfoliated graphene on hexagonal boron nitride (hBN), by locally ionizing defects in the hBN using the tip of a scanning tunneling microscope (STM). At high magnetic fields, the gaps between Landau levels (LLs) create insulating barriers inside and around the QDs, enabling a capacitive interaction with the STM tip, sample, and back gate electrodes. Inside the QDs, dI/dV spectra reveal a series of CB peaks, alongside local density of states peaks due to LLs. By sweeping the gate voltage, we construct spectroscopic gate maps in which the Coulomb peaks appear as lines, whose slope is governed by the capacitances between dot, tip, and sample electrodes, and whose offsets reveal the addition spectrum of the QDs. Each LL has its own series of charging lines, creating anticrossings whose characteristics reflect the interactions between electrons in different LLs, and depend strongly on both the magnetic field and the gate voltage, especially at weaker fields. By moving the STM tip, we can tune the tip-dot capacitance, and tunnel into different parts of the dot, enabling a full characterization of the anticrossings of these coulomb peaks.   

Efficient Editing of Atomic-Scale Silicon Devices and Memories

A dangling bond (DB) on the surface of hydrogen-terminated silicon forms an atomic silicon quantum dot, localizing up to two electrons. Circuits based upon DBs can enable future device architectures. Among these are atomic scale logic devices, where they are predicted to reduce power consumption by several orders of magnitude. The fabrication of these devices requires the precise control over the placement of DBs to achieve proper functionality. Scanning Tunneling Microscopes (STMs) are employed to fabricate many of these devices. However, at the atomic scale fabrication presents unique challenges due to the extreme sensitivity to uncertainties in the STM tip position and geometry. These uncertainties can result in the creation of misplaced DBs during fabrication. Recently we’ve discovered several methods to cap individual DBs with single atoms of hydrogen to efficiently erase misplaced DBs and edit structures. This newfound ability has greatly improved our fabrication yields, allowing for the creation of truly prefect DB structures. With these techniques, new experiments and applications are now within reach, including ultra-high density, room-temperature stable memory.   

Using Quantum States of Molecules as Probes

Scanning tunneling microscopy (STM) and inelastic electron tunneling spectroscopy (IETS) are powerful, versatile tools for characterizing and manipulating atoms and molecules on surfaces, but it is difficult to distinguish the contributions of different interactions. STM-IETS probes the local density of states of the system between the surface and the tip and can excite quantum states of a molecule in the tunneling junction. By attaching single molecules to the apex of the STM tip, we show that certain excitations of the molecule on the tip are sensitive to distinct aspects of the local environment on the surface and by monitoring the response of these excitations as we vary the position of the tip over the surface, we can identify the nature of the contribution. We show that single molecule probes that have different kinds of quantum states can serve as tools for probing an array of different physical properties beyond the capabilities of conventional STM can achieve.   

Sequence Identification and Structural Mapping of Single DNA and RNA Molecules via Quantum Tunneling Spectroscopy

We present a method that uses scanning tunneling spectroscopy (STS) measurements on DNA and RNA molecules to directly identify the nucleotide sequence as well as to map both secondary structure and epigenetic modifications. The method relies on non-perturbative STS to probe the molecular orbitals of each base,[1] combined with a molecular identification algorithm that uses machine learning to recognize unique electronic fingerprints of the bases and their modifications.[2] Structural mapping of RNA at the single molecule level can be achieved using conformation-dependent chemical labeling followed by mapping of the labels via STS. We further describe the use of a ‘smear parameter’ to quantify the extent of smearing out of the nucleotide orbitals due to entropic factors, and we show how these effects can be mitigated through engineered surface chemistry to reduce conformational entropy, which plays an important role in molecular recognition.[3] This paves the way for probing the genome and transcriptome with a previously unattainable level of detail.

[1] Ribot, JC; Chatterjee, A & Nagpal, P. J. Phys. Chem. B 2015, 119, 4968.
[2] Korshoj, LE; Afsari, S; Khan, S, et al. Small 2017, 13, 1603033.
[3] Korshoj, LE; Afsari, S; Chatterjee, A, et al. J. Am. Chem. Soc. 2017, 139, 15420.   

Applying and detecting tip-induced local strain on monolayer MoS2/graphite with scanning tunneling microscopy and inelastic electron tunneling spectroscopy

Strain engineering of two-dimensional (2D) materials gained great attention because of their superior mechanical properties and applicability to novel devices such as wearable electronics. The extremely small size of modern electronic devices necessitates a method to investigate strain effects at the nanoscale in these materials. Here we utilize scanning tunneling microscopy (STM) and inelastic electron tunneling spectroscopy (IETS) to apply and detect the local strain on monolayer MoS₂ grown on a graphite substrate. Monolayer MoS₂ behaves as a mechanical and tunneling barrier, and by controlling the distance between the tip and the sample, STM can apply local strain and simultaneously detect the change in phonon modes by IETS. IETS revealed that the phonon energy of graphite decreased by the tip-induced strain. Density functional theory calculation of phonon density of the states also showed the phonon-softening by strain, which substantiated our observation.   

Single-atom thermometer

Measuring temperature with nanoscale resolution has proven extraordinarily difficult because scanning probes lack the thermal isolation needed to achieve high spatial resolution. We propose a novel method to measure temperature with sub-nanometer resolution using individual physisorbed atoms as surface thermometers. We study the diffusive dynamics of aluminum adatoms adsorbed on graphene nanoflakes subject to a thermoelectric bias leading to a nontrivial spatial temperature distribution. The adatoms couple to local thermal fluctuations of both the vibrational and electronic degrees of freedom. The electronic coupling is mediated by Coulomb interactions, and is treated using the nonequilibrium Green's function method (NEGF). The adatom motion can be followed by STM in real time, and we find that they seek out the cold spots on the surface, thereby serving as thermometers.