Quantum approach to modelling high resolution Kelvin Probe Force Microscopy

Project Information

QMKPFM

Grant agreement ID: 845060

DOI

10.3030/845060

Project closed

EC signature date 10 April 2019

Start date 1 January 2020

End date 4 March 2022

Funded under

• EXCELLENT SCIENCE - Marie Skłodowska-Curie Actions

Total cost

€ 202 680,96

EU contribution

€ 202 680,96

Coordinated by

AALTO KORKEAKOULUSAATIO SR
Finland

Periodic Reporting for period 1 - QMKPFM (Quantum approach to modelling high resolution Kelvin Probe Force Microscopy)

Reporting period: 2020-01-01 to 2021-12-31

Summary of the context and overall objectives of the project

In this project we were focusing on understanding physics of exceptionally precise surface physics measurements proceeded with Scanning Probe Microscopy through computer modelling and quantum physics and chemistry computations. These measurements done at extremely stable conditions such as ultra-high vacuum, ensuring unchanging of the surface on the atomic scale for many hours, and temperature lower than 10 Kelvins. Together with careful preparation of the Scanning Probe Microscope (SPM) tip via appending a flexible and non-reactive apex to it – like a CO molecule or Xe atom- a resolution of individual atoms in organic molecules is possible. Our research focused on the interaction between the tip apex and sample inn the presence of electrostatic field, where the experimental data have not been understood - these measurements are called Kelvin Probe Force Microscopy (KPFM).
Better understanding of these measurements, leads towards getting more information from these measurements and in the long term can make SPM into widely usable tool for analysis of chemical compounds on surfaces with extraordinary spatial and chemical resolution. This goes very well with a contemporary trend of usage of on-surface synthesis for new promising compounds and materials, that cannot be synthetized otherwise. Thus, the results of this projects is also contributing to discovery of new materials and enhancement of characterization methods.

Work performed from the beginning of the project to the end of the period covered by the report and main results achieved so far

In our first [1], second [2] and fourth [4] paper, we used SPM simulations to identify a novel organic molecule and novel 2D materials synthesizable only through combination of classical wet and on-surface chemistry. The combination of experimental measurements and theoretical calculations confirmed atomic geometries. In works [1] and [4] it showed the occurrence of otherwise unstable eight-carbon rings both, that have not been synthetized and observed before. Our work also identified understand the electronic structures of the novel 2D materials [2, 4]. All these work opened a pathway-towards creation of materials with technically interesting properties, like magnetic ordering, or quantum hall effect The fourth work [4] was also publicly disseminated through Aalto webpages and Finnish translation was also sent to newspapers.
In our third paper [3] and our last work [5], we employed our SPM simulations and experiments in collaboration with machine learning. These were used for effective finding of the most stable molecular geometryof a bulky 3D molecule on a metallic surface [3] and electrostatic field around measured molecules [5]. Our simulations were crucial for identification of the experimentally observed geometries [3] and for creating the database to train the machine learning model [5]
Two further works are currently submitted to peer-reviewed-journals.
The work on this project was presented already on 6 conferences and workshops and also two university seminars. All the computational data and metadata for the published work were openly published on scientific databases, allowing for further reusability and next research.
The main project, was facing several difficulties mainly caused by employment of third-party open-source quantum chemistry code, which is one of few, that can be used for such calculations. These difficulties slowed down the project, however we ended up with working code [6], that is published on software repository GitHub, where is described with Wiki pages and is open to other scientists.
The main investigator participates on organizing talks discussing and covering topics of gender diversity and inclusiveness at Aalto University, School of Science.

Literature:
[1]K. Nakamura, et al., On-Surface Synthesis of a π-Extended Diaza[8]circulene, J. Am. Chem. Soc. 142, p. 11363–11369 (2020)
DOI: 10.1021/jacs.0c02534
Green open-access: https://tinyurl.com/57b825z7

[2] L. Yan, et al., Synthesis and Local Probe Gating of a Monolayer Metal‐Organic Framework, Adv. Funct. Mater. 2021, 2100519 (2021)
DOI: 10.1002/adfm.202100519 (Open Access)

[3] J. Järvi, et al., Integrating Bayesian Inference with Scanning Probe Experiments for Robust Identification of Surface Adsorbate Configurations, Adv. Funct. Mater. 2021, 2010853 (2021)
DOI: 10.1002/adfm.202010853 (Open Access)

[4] Q. Fan,et al., Biphenylene network: A nonbenzenoid carbon allotrope, Science 372, 852-856 (2021)
DOI: 10.1126/science.abg4509
Green open access: https://tinyurl.com/bdhskuyy

[5] N. Oinonen, et al., Electrostatic discovery atomic force microscopy, ACS Nano 16, 89–97 (2022)
DOI: 10.1021/acsnano.1c06840 (Open Access)

[6] https://github.com/SINGROUP/KPFM_sim

Progress beyond the state of the art and expected potential impact (including the socio-economic impact and the wider societal implications of the project so far)

During the project we took the state-of-the-art KPFM theory [7] and expand it to two mechanical models that reflect possible contribution to the KPFM signal when, the flexible tip-apices are relaxing. One of the models is a direct extension of the original theory taking into account the relaxing tip-apex, while the second one is coming with a new possible contribution to the KPFM signal. Direct comparison of these models with published experiment [8] suggests that both of these models are contributing to the KPFM signal, but the ratio between these contributions remains unknown (Fig. 1). These models are published as part of Probe Particle Model [9]. These models allows for quick simulations and will be used in future SPM research.

We also took a completely different approach for probing the electrostatics around samples measured with flexible-tip scanning probe microscopy, through using two tip method [10] and fully-automatized this method via machine learning [5]. This removes lots of hand work with image processing. Together with our previous work[11], it allows for quick and precise recognition of atoms and electrostatic field around them from two set of non-contact Atomic Force Microscopy images, acquired with different flexible tip-apices (Fig 2). This work will allow for faster, automatized evaluation of SPM experiments of new materials and molecules. It comes with more quantitative way for the experiment evaluation and thus propel the field towards measurements of larger, and more technologically/chemically interesting systems than just simple molecules.

Finally, we came up with a completely new method for simulating the electric field, that is between the metallic tip with flexible tip-apex and sample with metallic substrate on the ab-initio level. This method allows for simulations of images at various heights [6]. Results of this method are in close agreement with the state-of-art theory for some range heights (Fig. 3). However, to fully confirm this method, further calculations will be needed.

Further Literature:
[7] L. Gross, et al., Investigating atomic contrast in atomic force microscopy and Kelvin probe force microscopy on ionic systems using functionalized tips, Phys. Rev. B 90, 155455 (2014).
DOI: 10.1103/PhysRevB.90.155455

[8] F. Mohn et.a Imaging the charge distribution within a single molecule, Nat. Nanotechnol. 7, 227–231 (2012)
DOI: 10.1038/NNANO.2012.20

[9] https://github.com/ProkopHapala/ProbeParticleModel/wiki/KPFM-approximations

[10] P. Hapala, et al., Mapping the electrostatic force field of single molecules from high-resolution scanning probe images, Nat. Commun. 7, 11560 (2016)
DOI: 10.1038/ncomms11560

[11] B. Alldritt, et al., Automated structure discovery in atomic force microscopy, Sci. Adv. 6, eaay6913 (2020)
DOI: 10.1126/sciadv.aay6913

Fig. 1: Mechanistic model for possible KPFM contribution and their comparison with experiment

Fig. 3: Our model for creating the electric-field on the ab-initio level and results of the model.

Fig. 2: Electrostatic Field from 2 set of AFM images via Convolutional Neural Network.