Constant-height Images Research Articles

Molecular identification with atomic force microscopy and conditional generative adversarial networks

Frequency modulation (FM) atomic force microscopy (AFM) with metal tips functionalized with a CO molecule at the tip apex (referred as High-Resolution AFM, HR-AFM) has provided access to the internal structure of molecules with totally unprecedented resolution. We propose a model to extract the chemical information from those AFM images in order to achieve a complete identification of the imaged molecule. Our Conditional Generative Adversarial Network (CGAN) converts a stack of constant-height HR-AFM images at various tip-sample distances into a ball-and-stick depiction, where balls of different color and size represent the chemical species and sticks represent the bonds, providing complete information on the structure and chemical composition. The CGAN has been trained and tested with the QUAM-AFM data set, that contains simulated AFM images for a collection of 686000 organic molecules that include all the chemical species relevant in organic chemistry. Tests with a large set of theoretical images and few experimental examples demonstrate the accuracy and potential of our approach for molecular identification.

Calculations of Constant-Height STM Images of Fullerene C60 Adsorbed onto a Surface

Constant-height scanning tunneling microscopy (STM) images of a C60 molecule adsorbed onto a surface were calculated using symmetry-adapted Hückel molecular orbitals (HMOs). Three adsorption orientations of C60 are considered. The interaction between the C60 molecule and the surface was treated using symmetry arguments only. Projection operators were used to generate symmetry-adapted HMOs of the molecule. These orbitals were then used to construct idealized constant-height STM images using the simple tunneling theory of Tersoff and Hamann. A comparison is made with published experimental STM maps. The results show that, for each orientation of C60, split orbitals of the same symmetry have similar appearances in the constant-height maps. They also show that the map of a molecular orbital of a complete degeneracy is dominated by only one or two of its components.

Molecular Identification from AFM Images Using the IUPAC Nomenclature and Attribute Multimodal Recurrent Neural Networks.

Spectroscopic methods─like nuclear magnetic resonance, mass spectrometry, X-ray diffraction, and UV/visible spectroscopies─applied to molecular ensembles have so far been the workhorse for molecular identification. Here, we propose a radically different chemical characterization approach, based on the ability of noncontact atomic force microscopy with metal tips functionalized with a CO molecule at the tip apex (referred as HR-AFM) to resolve the internal structure of individual molecules. Our work demonstrates that a stack of constant-height HR-AFM images carries enough chemical information for a complete identification (structure and composition) of quasiplanar organic molecules, and that this information can be retrieved using machine learning techniques that are able to disentangle the contribution of chemical composition, bond topology, and internal torsion of the molecule to the HR-AFM contrast. In particular, we exploit multimodal recurrent neural networks (M-RNN) that combine convolutional neural networks for image analysis and recurrent neural networks to deal with language processing, to formulate the molecular identification as an imaging captioning problem. The algorithm is trained using a data set─which contains almost 700,000 molecules and 165 million theoretical AFM images─to produce as final output the IUPAC name of the imaged molecule. Our extensive test with theoretical images and a few experimental ones shows the potential of deep learning algorithms in the automatic identification of molecular compounds by AFM. This achievement supports the development of on-surface synthesis and overcomes some limitations of spectroscopic methods in traditional solution-based synthesis.

Topological Defects Induced High-Spin Quartet State in Truxene-Based Molecular Graphenoids

Topological Defects Induced High-Spin Quartet State in Truxene-Based Molecular Graphenoids

QUAM-AFM: A Free Database for Molecular Identification by Atomic Force Microscopy.

This paper introduces Quasar Science Resources-Autonomous University of Madrid atomic force microscopy image data set (QUAM-AFM), the largest data set of simulated atomic force microscopy (AFM) images generated from a selection of 685,513 molecules that span the most relevant bonding structures and chemical species in organic chemistry. QUAM-AFM contains, for each molecule, 24 3D image stacks, each consisting of constant-height images simulated for 10 tip-sample distances with a different combination of AFM operational parameters, resulting in a total of 165 million images with a resolution of 256 × 256 pixels. The 3D stacks are especially appropriate to tackle the goal of the chemical identification within AFM experiments by using deep learning techniques. The data provided for each molecule include, besides a set of AFM images, ball-and-stick depictions, IUPAC names, chemical formulas, atomic coordinates, and map of atom heights. In order to simplify the use of the collection as a source of information, we have developed a graphical user interface that allows the search for structures by CID number, IUPAC name, or chemical formula.

Analysis of STM images with pure and CO-functionalized tips: A first-principles and experimental study

We describe a first principles method to calculate scanning tunneling microscopy (STM) images, and compare the results to well-characterized experiments combining STM with atomic force microscopy (AFM). The theory is based on density functional theory (DFT) with a localized basis set, where the wave functions in the vacuum gap are computed by propagating the localized-basis wave functions into the gap using a real-space grid. Constant-height STM images are computed using Bardeen's approximation method, including averaging over the reciprocal space. We consider copper adatoms and single CO molecules adsorbed on Cu(111), scanned with a single-atom copper tip with and without CO functionalization. The calculated images agree with state-of-the-art experiments, where the atomic structure of the tip apex is determined by AFM. The comparison further allows for detailed interpretation of the STM images.

Direct Identification and Determination of Conformational Response in Adsorbed Individual Nonplanar Molecular Species Using Noncontact Atomic Force Microscopy.

In recent years atomic force microscopy (AFM) at highest resolution was widely applied to mostly planar molecules, while its application toward exploring species with structural flexibility and a distinct 3D character remains a challenge. Herein, the scope of noncontact AFM is widened by investigating subtle conformational differences occurring in the well-studied reference systems 2H-TPP and Cu-TPP on Cu(111). Different saddle-shape conformations of both species can be recognized in conventional constant-height AFM images. To unambiguously identify the behavior of specific molecular moieties, we extend data acquisition to distances that are inaccessible with constant-height measurements by introducing vertical imaging, that is, AFM mapping in a plane perpendicular to the sample surface. Making use of this novel technique the vertical displacement of the central Cu atom upon tip-induced conformational switching of Cu-TPP is quantified. Further, for 2H-TPP two drastically different geometries are observed, which are systematically characterized. Our results underscore the importance of structural flexibility in adsorbed molecules with large conformational variability and, consequently, the objective to characterize their geometry at the single-molecule level in real space.

Scanning electrochemical microscopy imaging with laser-pulled probes

The fabrication of Pt nanoelectrodes (NEs) and ultramicroelectrodes (UMEs) were explored through the use of a precision laser electrode puller. As a result of the heated pull, the Pt tips of the NEs were exposed immediately during the fabrication process. Newly fabricated NEs were immediately used to observe the topographical and electrochemical characteristics of a TEM grid by scanning electrochemical microscopy (SECM) horizontal line scans. While higher spatial resolution would be offered by NEs, their use in imaging can be restricted due to sample size and instrumental limitations. To accurately characterize an independently addressable microband electrode (IAME) by SECM, a Pt UME of a known size and geometry was used. SECM depth scan and constant height images were obtained of an unbiased IAME. By utilizing a SECM probe that is similar in size to the Pt and glass microbands of the IAME (4.4 vs. 5μm, respectively), redox cycling phenomenon was detected. From these electrochemical images, the Pt microbands appeared to be larger than their physical dimensions, while the glass microbands appeared much smaller. Since redox cycling can be affected by sample tilt, the tilt angle of the IAME was quantified through 3 different methods. The first approach was to use the traditional PAC comparison method over different locations on the sample, while the second and third approaches relied heavily on 3D finite elemental analysis simulations representative of the SECM experiments. The first of these two methods required the fitting of a tilted experimental horizontal sweep against a fixed IAME, while the second approach relied on the direct fitting of the experimental data onto tilted depth scan simulations. Each methodology delivered approximately the same tilt angle, confirming the viability of all the methodologies.

Intelligent Scanning Electrochemical Microscopy Tip and Substrate Control Utilizing Fuzzy Logic

In this paper, we present a technique to intelligently control the motion of Scanning Electrochemical Microscopy (SECM) tip electrodes and automatically align SECM tip electrodes to the substrate with no user interaction. This intelligent method utilizes a custom fuzzy logic algorithm to control the motion of the tip electrode and automate tip/substrate alignment. The fuzzy logic control algorithm rapidly obtains SECM approach curves via a variable speed approach and automatically stops the tip electrode at a preprogrammed tip/substrate distance without “crashing” the tip electrode into the substrate. The fuzzy logic algorithm automatically adjusts the speed of the electrode based on the tip size, the enhancement of the tip current, and the distance that the tip electrode is from a user defined setpoint. We validated that both positive feedback and negative feedback approach curves obtained using the fuzzy logic algorithm matched well with simulated approach curves for both large microelectrodes (sometimes used for Surface-Interrogation SECM, a=175μm) and for conventional SECM ultramicroelectrodes (a=4.2μm). The ability to automatically stop the tip electrode is crucial for the automated sample tilt adjustment protocol. This ability allows us to automate the interactive process of (1) approaching the tip to the substrate, (2) scanning the tip across the substrate to measure the tip/substrate tilt, (3) correcting for the tip/substrate tilt. We demonstrated that this tip/substrate alignment protocol allows us to obtain highly resolved constant-height SECM images over large areas with ease.

Simultaneous measurement of tunneling current and atomic dipole moment on Si(111)-(7 × 7) surface by noncontact scanning nonlinear dielectric microscopy

Non-contact scanning nonlinear dielectric microscopy (NC-SNDM) can resolve the topography and dipole moment distribution of a Si(111)-(7 × 7) surface on an atomic level. We discuss the origin of the atomic contrast in dipole moment images based on simultaneously acquired time-averaged tunneling current images. The dipole moment images are found to have the same characteristics as the simultaneously acquired current images. This similarity between these two images does not originate from circuit crosstalk since it is suppressed by our circuit design that decouples the tip-sample capacitance and the tunneling current. Constant-height images also indicate that the atomic contrast is not artificially caused by feedback crosstalk. These results suggest that the atomic contrast is due to the variation in the tip-sample capacitance caused by modulation of atomic dipole moments by the local density of states of the surface.

Topographical fingerprints of many-body interference in STM junctions on thin insulating films

Negative differential conductance is a nonlinear transport phenomenon ubiquitous in molecular nanojunctions. Its physical origin can be the most diverse. In rotationally symmetric molecules with orbitally degenerate many-body states it can be ascribed to interference effects. We establish in this paper a criterion to identify the interference blocking scenario by correlating the spectral and the topographical information achievable in a scanning tunneling microscopy (STM) single-molecule measurement. Simulations of current-voltage characteristics as well as constant-height and constant-current STM images for a Cu-phthalocyanine on a thin insulating film are presented as experimentally relevant examples.

Simultaneous AFM and STM measurements on theSi(111)−(7×7)surface

We perform simultaneous atomic force microscopy (AFM) and scanning tunneling microscopy (STM) measurements on the $\text{Si}(111)\text{\ensuremath{-}}(7\ifmmode\times\else\texttimes\fi{}7)$ surface. AFM/STM constant height images are obtained at various tip-surface distances. Force/current distance spectroscopy using the same tip apex allows us to estimate the relative tip-surface distance for each image as well as the short/long-range force and the tunneling current. We demonstrate that the tunneling current at tip-surface distances where AFM clearly resolves atoms is much larger than the typical values in conventional STM. In addition, at the tip-surface distances for conventional STM, the short-range force is too small to provide atomic contrast in AFM. We show that the differences in the signal-to-noise ratio of the constant height images between AFM and STM produce different optimal imaging distances. In addition, we note that the different imaging distance also influences images obtained using double tips.

Nanoscale Imaging of Surface Topography and Reactivity with the Scanning Electrochemical Microscope

Over the last 2 decades, scanning electrochemical microscopy (SECM) has been extensively employed for topographic imaging and mapping surface reactivity on the micrometer scale. We used flat, polished nanoelectrodes as SECM tips to carry out feedback mode imaging of various substrates with nanoscale resolution. Constant-height and constant-current images of plastic and Au compact disc surfaces and more complicated objects (computer chips and wafers) were obtained. The possibility of simultaneous imaging of surface topography and electrochemical reactivity was demonstrated. Very fast mass transfer at nanoelectrodes allowed us to obtain high-quality electrochemical images in viscous media under steady-state conditions, e.g., in 1-methyl-3-octylimidazolium-bis(tetrafluoromethylsulfonyl)imide (C(8)mimC(1)C(1)N) ionic liquid. Ion-transfer-based imaging was also performed using nanopipets as SECM tips.

Non-contact Atomic Force Microscopy Simulations of Hydrogen-terminated Si(100) Surfaces with a Methyl

Toward the development of a simulator of scanning probe microscope for systems with adsorbed organic molecules, we simulate non-contact atomic force microscope images of hydrogen-terminated Si(001)2×1 surfaces with an isolated methyl using a Si tip with and without a hydrogen atom at the apex, using a density-functional based tight-binding method. We examine the constant height, constant frequency and energy dissipation images. In addition to normal images of constant height and constant frequency modes, we obtain the exotic energy dissipation images, where a low dissipation spot is surrounded by a ring of high dissipation region. These images can be understood from the site-dependence of force hysteresis.

Imaging nanoclusters in the constant height mode of the dynamic SFM

For the first time, high quality images of metal nanoclusters which were recordedin the constant height mode of a dynamic scanning force microscope (dynamicSFM) are shown. Surfaces of highly ordered pyrolytic graphite (HOPG) were usedas a test substrate since metal nanoclusters with well defined and symmetricshapes can be created by epitaxial growth. We performed imaging of goldclusters with sizes between 5 and 15 nm in both scanning modes, constantΔf mode and constant height mode, and compared the image contrast. We noticethat clusters in constant height images appear much sharper, and exhibit morereasonable lateral shapes and sizes in comparison to images recorded in the constantΔf mode. With the help of numerical simulations we show that only a microscopically smallpart of the tip apex (nanotip) is probably the main contributor for the imagecontrast formation. In principle, the constant height mode can be used for imagingsurfaces of any material, e.g. ionic crystals, as shown for the system Au/NaCl(001).

Simulation of topographic images and artifacts in illumination-mode scanning-near-field optical microscopy

Scanning images in illumination-mode, scanning-near-field optical microscopy (SNOM) are numerically studied by the boundary element method based on rigorous vector electromagnetic theory. Computation results of constant-height and constant-distance images for samples with different topographic features are presented. Effects of the polarization of the input light and the optical parameters of samples on the resolution of SNOM are discussed. The artifacts in constant-distance images are also investigated. Numerical results indicate that the constant-height images for TM input light and constant-distance images for both TE and TM input light give only the local changes of the sample topography because of the loss of the low-frequency component of the topography.

Local detection of the optical magnetic field in the near zone of dielectric samples

We present a study of the influence of the probe composition on the formation of constant-height photon scanning tunneling microscope images when observing a dielectric sample. Dramatic effects due to the metallization of the tip are presented and discussed in detail. We show how the recorded images can look quite different when the probe is dielectric or coated with gold. Comparison with numerical calculations indicate that the experimental signals are of electric or magnetic nature depending on the composition of the tip. For well-defined conditions, gold-coated tips provide images of the distribution of the magnetic field intensity associated with the optical near field. To get more insight into the origin of this effect, we have performed a careful analysis of the plasmon modes associated with a gold-coated dielectric cylinder. We conclude that the detection of the optical magnetic field is related to the conditions of excitation of circular symmetry plasmon, or ring plasmon, in the gold-coated tip.

Shear-force, constant-height, and constant-intensity imaging in scanning near-field optical microscopy with <italic>s</italic>- and <italic>p</italic>-polarized incident light

We present a numerical comparison between shear-force, constant-height, and constant-intensity images in scanning near-field optical microscopy. We demonstrate the general difference between the three imaging modes. Two types of incident light are tested, with polarization perpendicular and parallel to the mean plane of the surface. Merits and demerits of the three images are discussed.

Numerical comparison between constant-height and constant-intensity images in scanning near field optical microscopy

Within a microscopic theory, we present numerical simulations of scanning near field optical microscopy for constant-height, constant-intensity and near-field (under shear-force distance regulation) imaging methods. Our primary interest is to compare the constant-height and constant-intensity images. We clearly demonstrate that the two images are different particularly for near field. Both images may resolve subwavelength features, but constant-height images are more or less associated with far field diffractions. Finally, we use two simple examples to show the following. For the far field components whose distance dependence from the surface is about 1/ z, the two images are somewhat similar, while, for near field components whose distance dependence is as fast as 1/ z 3, the two images are quite different. In other words, the more near field the bigger the difference.

Removing optical artifacts in near-field scanning optical microscopy by using a three-dimensional scanning mode

We demonstrate a method of acquiring near-field scanning optical microscopy data that allow for the construction of three different types of images from one data set: topographic, constantgap, and constant-height. This data set includes the topographic features of the surface and the optical response at various heights above the sample surface. Comparisons are made between the images recorded in this format and both conventional, constant-gap mode images, and pseudoconstant-height mode images constructed using a single retraction curve. Zmotion artifacts are identified by analyzing the optical intensity for a given image as a function of the sample topography. Using this procedure it is shown that significant z-motion artifacts exist in the constant-gap images of gold particles immobilized on a glass slide. These artifacts are avoided by constructing constant-height images.