The Cavendish Surface Physics Group and Atom Scattering Facility invite undergraduate applications to an online summer school and program of collaborative summer research projects, which will be conducted remotely, given current and likely restrictions on travel. The program will cover areas such as ultrafast surface-dynamics, scattering of atoms, molecules, and neutrons from surfaces, and helium atom microscopy.
The summer school will aim to bring students up to speed with both theoretical and computational concepts associated with the area of the program. Research projects will follow the summer school, and will generally be computational, theoretical, or related to scientific instrument development. There will be two types of projects; supervised and unsupervised.
Unsupervised projects will normally address problems for educational purposes, and will typically last around 3-5 weeks. An example project could be the writing of a Kinetic Monte Carlo simulation with inter-adsorbate interactions, and investigation of the 2D formation of nano-clusters as function of temperature. Students will receive general guidance, and will be invited to exchange ideas and to discuss the projects in dedicated online forums.
Academic Enrichment Supervised projects will be offered by members of the surface physics group and national/international collaborators. Those projects will normally address original scientific problems, and will last 8 weeks (with an option for a short extension). For last years’ projects, see https://www.smf.phy.cam.ac.uk/jobs/surfacesummer2020. A list of projects for the 2021 program is now available below.
The program will take the form:
- Week 1-2: Summer school. Lectures, Educational Package of assignments, seminars on selected topics. Guided reading of relevant literature. Small group discussions. Allocation to projects based on interest and availability.
- Weeks 3-8: Individual work on projects, guided by a project mentor. Weekly group meetings with the cohort of project students and mentors, normally including a short overview of progress from each student. Mid-project review.
- Week 9-10: Completion of the research project and support with writing a high quality report on the activities. Final review of the activities.
In addition, students will be invited to join our usual ‘virtual tea-time’ meetings, which are the online version of the ‘tea-times’ which have served for decades as a platform for exchanging ideas within the Cavendish laboratory.
Requirements & Commitment: As projects will be conducted remotely, individuals will need their own (basic) computer and a reliable internet connection suitable for video conferencing. We will provide remote access to any high performance computers required for particular projects, as well as any software necessary.
Individuals will be able to be very flexible with their work pattern, but will be expected to commit to the full ten week programme (starting on 28th June 2021) and to at least weekly group video meetings with the project cohort and with their project mentor.
Supervisor:
Dr Anton Tamtögl
Nature of project:
computational
Requirements:
Part 1B level physics or above
Understanding the difference in neutron and atom
scattering from large organic molecules
Sattering techniques are routinely used to study the sructure of
thin films and molecules adsorbed at surfaces. Helium atom
scattering provides some unique advantages over other scattering
techniques. However, helium scattering from isolated molecules is
hardely understood in contrast to neutron and X-ray scattering.
While neutrons are directly scattered by the ion cores in a
molecule, He atoms are scattered by the elctron density “around”
the ion cores of the molecule and thus a precise description of
helium scattering from large molecules requires knowledge of the
so-called form factor.
1. Understand how intensity oscillations upon helium scattering
from a large organic molecule are related to the form factor of
the molecule and the difference between neutron and helium
scattering
2. Try to evaluate helium scattering intensities form an isolated organic molecule based on either simple analytical assumptions or scattering calculations form a simple potential. Compare the results with neutron scattering calculations to illustrate the difference in the form factor
The project requires some programing, at least in Matlab. Python
may be useful, e.g. for comparison with neutron scattering the
open source package https://pypi.org/project/jscatter/
can be used.
Supervisor:
Dr Bill Allison
Nature of project:
computational
Requirements:
Part 1B level physics or above
An approach to quantum inelastic scattering of atoms from surfaces
During surface scattering, phonons in the solid create a heat bath that gives energy gains and losses for the scattered particle. Elastic scattering is the zero-phonon case, single-phonon scattering gives discrete energy transitions while multi-phonon scattering generates a broad range of energy transitions. A full quantum description of the process is a formidable task and is beyond existing computational capabilities. In the present project we will approach the problem using perturbation techniques developed recently by our collaborators, [1,2]. Here, both the static surface corrugation, which defines the extent of diffractive scattering, and the bath oscillators, responsible for inelastic scattering, are treated as perturbations to the scattering problem. The unperturbed dynamics will be treated classically with the perturbations handled within a semi-classical picture. We will adopt a numeric approach that eliminates some of the approximations involved in analytic solutions. Complexity will be added in an incremental way; thus, we start with purely elastic scattering, [1], before attempting the more complicated problem of inelasticity, [2]. In this way the student will develop necessary numerical skills alongside the theoretical ideas that underlie the analysis.
[1] E Pollak and S Miret Artes, J. Phys. Chem. C, 119, 14532 (2015)
[2] S Daon and E Pollak, J. Chem. Phys. 142, 174102 (2015)
Supervisor:
Dr David Ward
Nature of project:
computational
Requirements:
Part 1B level physics or above
Efficient pixel acquisition for imaging with helium atoms
Imaging with helium involves moving a sample in a beam of helium atoms while measuring the number of helium atoms scattered to a fixed detector; moving the sample allows a real space scattering map of the sample to be formed. It’s been shown that the data contains information related to the topography of the sample and scattering dependant features such as diffraction, inelastic scattering etc. While the benefits of the technique are significant, it typically takes in excess of 300ms per pixel to collect enough data and therefore images of, typically thousands of pixels square take many hours to acquire. The project involves making a sampling method to reconstruct images, acquiring as much information as possible and can consider a wide range of issues such as low resolution pre-imaging to guide higher resolution imaging, multiple-pass techniques, prior knowledge of the sample, accuracy vs speed, existing methods such as compressed sensing, in-painting etc. used either in compression algorithms or in other acquisition techniques.
The intention for the project is to use simulated and experimental data and for candidates to investigate methods to gather the maximum information from unknown samples with the fewest measured pixels.
Supervisor:
Dr David Ward, Alek Radić
Nature of project:
computational
Requirements:
Part 1B level physics or above
Application of GPU (CUDA) to atom ray tracing
Ray tracing has become the de facto-standard simulation method for evaluating helium atom imaging. The framework we use has been developed in house using Matlab and C++. In the project the student will work on several aspects of the method which will improve its applicability. Possible routes that may be interesting are implementation of a graphical user experience, adding features for sample manipulation and geometries including incorporating the next generation development platform. There is an MPhil student who has an interest in the project who will work with the candidate to develop the project.
There are many practical ways in which the project can be conducted including through access to the high performance computing capability in Cambridge, using a GPU enabled system in our group or on the candidates own system if they possess a recent Nvidia GPU in their own equipment.
Supervisor:
Dr David Ward, Alek Radić
Nature of project:
computational
Requirements:
Part 1B level physics or above
Rotating stages & tilted planes
Scanning helium microscopy relies on rastering a sample in a beam of helium and measuring the scattered helium intensity. Recently the sample manipulation system for helium microscopy has been upgraded to enable azimuthal rotation which allows for a range of new imaging capabilities including diffraction from small patches of the sample, diffraction intensity as a new contrast mechanism akin to light and dark field imaging in optical microscopy. A less obvious capability is that through point tracking of the sample through rotation, information about the height of the sample can be obtained, that can be used to measure the 3D structure of the areas of the sample, and the tilt of surface facets. The project will consider additions and enhancements to our existing ray-tracing framework to enable these new hardware capabilities to be modelled and new imaging data to be analysed.
Supervisor:
Dr David Ward, Alek Radić
Nature of project:
computational
Requirements:
Part 1B level physics or above
Gas flows, pressures and pumping speeds in mote-Carlo simulation
Helium imaging (SHeM) has thus far used pinhole apertures mounted in plates machined from metal or printed in plastic. Put simply the project involved making optimised plates that are suitable for high resolution imaging, assessing the capabilities of high efficiency atom optic elements and for angular and spatially resolved plates for our new development framework. The project is flexible in approach but it’s expected that CAD modelling, ray tracing simulations and pressure modelling would be used. There are existing tools for these areas that can be used (Autodesk Inventor, in house ray tracing and molflow).
Supervisor:
Dr David Ward, Nick A. von Jeinsen
Nature of project:
computational
Requirements:
Part 1B level physics or above
Temperatures and heat flows around samples in vacuum
We have recently been experimenting with the addition of a heating capacity for samples in scanning helium microscopy (SHeM). In the sample environment a Kapton bound heating filament is mounted beneath the sample. There are constraints on the system including the power of the heater, the thermal conductivity of the sample holder and the maximum operating temperature of the sample manipulators. The project will consider thermal modelling of the region, including the manipulator, sample cartridge and aperture plates and determine mitigation measures to improve performance and efficiency such as coatings to optimise the black body functions. The project will go on to consider measurements of potentially interesting samples under heating, considering the measured data and modelling sample proposals of interest.
Supervisor:
Dr David Ward
Nature of project:
computational
Requirements:
Part 1B level physics or above
The internet of scientific things
Our helium scattering instruments consist of a large range of hardware that needs to be networked to read the data associated with imaging. There are many considerations such as the rate of data acquisition, ease and modularity of adopting hardware, requirements for real time data analysis and storage and logging capabilities. There is a legacy system in place iCON which is difficult to maintain and relied on legacy hardware capabilities. Work has progressed on hardware elements based on mini-computers such as raspberry pi’s and microcontrollers implementing protocols such as MQTT. The project aims to bring together the work that has been done so far and design, unit test and implement the next generation of instrument control system. Capabilities in C and or C++ would be advantageous but are not required.
Supervisor:
Dr David Ward
Nature of project:
computational
Requirements:
Part 1B level physics or above
Modelling diffraction with multiscat + ray tracing
We have achieved recent successes in application of the technology for helium imaging to traditional surface science systems. One of the key results is the demonstration that diffraction can be measured from spatially resolved surface patches. The project will use helium beam modelling (multiscat) to find and use helium-sample potentials to calculate scattering intensities in diffraction channels and then use them to derive scattering probabilities which can be used
for ray-tracing. The combination of experimental techniques with these two simulation methods is novel and has the capabilities to measure many otherwise inaccessible quantities. In the project several samples will be considered and simulated to form proposals for experimental measurements in the future.
Supervisor:
Dr Andrew Jardine
Nature of project:
computational
Requirements:
Part 1B level physics or above
Quantum Scattering from Disordered Surfaces
Understanding how atoms scatter from atomically disordered surfaces is of fundamental importance and underpins using atom-scattering for advanced material characterisation. The aim of this project will be to develop computational methods to perform quantum scattering from models of randomly rough surfaces, using different atom-surface interaction potentials. Scattering will be performed using wavepacket propagation [1], enabling full simulation with quantum-mechanical precision and including multiple scattering effects. The results from the project will also inform the emerging technique of scanning helium atom microscopy (SHeM), as the scattering distributions determines how images are formed [2].
[1] A. R. Alderwick, A. P. Jardine, W. Allison, J. Ellis, “An evaluation of the kinematic approximation in helium atom scattering using wavepacket calculations”, Surface Science 678, 65-71 (2018).
[2] S.M. Lambrick, L. Vozdecky, M. Bergin, J.E. Halpin, D.A. MacLaren, P.C. Dastoor, S.A. Przyborski, A.P. Jardine and D.J. Ward, “Multiple scattering in scanning helium microscopy”, Applied Physics Letters 116, 061601 (2020).
Supervisor:
Dr Andrew Jardine
Nature of project:
computational
Requirements:
Part 1B level physics or above
Characterising Surface Adsorption with Nanoscale Resolution
Atom scattering methods are widely used to characterise adsorption of atoms and molecules on surfaces, in order to understand (for example) the growth of materials and devices and chemical reactions. However, such experiments are usually restricted to measuring average properties over 2D model surfaces. The aim of this project will be to explore the feasibility of studying adsorption and desorption using the helium atom beam within a scanning helium microscope [1]. The goal is to unlock a unique capability to characterise adsorption on realistic surfaces with atomic precision. The project will likely involve numerical simulations using a combination of several existing codes.
[1] M. Barr, A. Fahy, J. Martens, A.P. Jardine, D.J. Ward, J. Ellis, W. Allison, and P.C. Dastoor, Nature Comms. 7, 10189 (2016).
Supervisor:
Dr John Ellis
Nature of project:
computational
Requirements:
Part 1B level physics or above
A ‘Permanent Magnet’ ‘solenoid’ high efficiency detector for atom and molecular beam detection
The surface physics group specializes in probing the dynamics and
structure of atoms at surfaces via helium atom scattering by two methods. Firstly we use an NMR like ‘spin echo’ approach to study atom motion on Angstrom length and sub-ps to ns time scales and secondly we are pioneering a ‘helium atom microscope which aims to achieve 30nm resolution in the short to medium term, with an ultimate resolution of perhaps 10nm with particular application to sensitive objects that are degraded by the beam in an electron microscope.
Key to these projects are the detectors used to monitor the scattered helium atoms. We have developed detectors that achieve gains of ~1000 over commercially available instruments, but they rely on large, high power consuming solenoids to hold the electrons that ionize the beam atoms. we have begun to design a new detector that replaces these solenoids with permanent magnetic configurations, and this project aims to model and design a ‘production’ version of such an instrument.
The project will entail using ‘in house’ and commercial finite element analysis software to model the electric and magnetic fields and the charged particle trajectories inside the detector and will be suitable for anyone interested in learning finite element analysis methods and designing new experiments.
Supervisor:
Prof. Gil Alexandrowicz, Dr. Helen Chadwick and Dr. Josh Cantin
Nature of project:
computational
Requirements:
Part 1B level physics or above
Simulations of magnetically manipulated atomic and molecular beams
This summer project is related to a major European research project aimed at controlling the rotational quantum states of molecules during a collision with a solid surface (“Rotational Waves” – [1]). The computational work carried out by the student will use classical and quantum mechanics based simulations to calculate the trajectories and spin states of molecules through the experimental beam line. The student will learn the physical principles of the new experimental technique ( [2] ) and gain experience in programming, executing and analysing numerical simulations.
The results of this project will be used for the interpretation of experimental results from past and future molecular beam experiments.
Due to Covid-19 restrictions the project will be supervised remotely, although once restrictions are lifted, students will be invited to visit the laboratory and the see the experimental setup they simulated.
[1] https://cordis.europa.eu/project/rcn/214696/factsheet/en
[2] https://www.nature.com/articles/ncomms15357, https://www.nature.com/articles/s41467-020-16930-1
Supervisor:
Prof. Gil Alexandrowicz, Dr. Helen Chadwick
Nature of project:
computational
Requirements:
Part 1B level physics or above
Simulating nuclear spin flips in a molecule-surface collision
This summer project is related to a recent UKRI project aimed at exploring the possibility of nuclear spin flips during a collision of an atom / molecule with a solid surface. ([1]). The computational work carried out by the student will use quantum mechanics based simulations to examine the effect of a nuclear spin flip on the signal of a magnetic molecular interferometry experiment. This work expands the idea of controlling and measuring rotational states ( [2] ) to nuclear spin states which are generally believed to not change during a collision with a surface. Due to Covid-19 restrictions the project will be supervised remotely, although once restrictions are lifted, students will be invited to visit the laboratory and the see the experimental setup they simulated.
[1] https://gow.epsrc.ukri.org/NGBOViewGrant.aspx?GrantRef=EP/V048589/1
[2] https://www.nature.com/articles/s41467-020-16930-1
Supervisor:
Dr Nadav Avidor
Nature of project:
computational
Requirements:
Part 1B level physics or above
Stochastic representation of inter-adsorbate interactions in molecular dynamics simulations
In molecular dynamics simulations of surface diffusion, the interactions between adsorbates are normally represented by pair-wise inter-adsorbate interaction potentials. At each time step, the force between each pair of particles is calculated. Such representation, while crucial for accurate representation of the surface-system, is computationally time consuming, and challenging in cases where the interaction is anything beyond a simple dipole-dipole repulsion. It would be of interest to model the interactions using a more computationally economical approach, which provide an intuitive insight to the interactions. For example, using random force drawn out of coloured noise, with a noise spectrum which represents the time and strength characteristics of the interactions.
The project is to demonstrate that computationally such representation is possible, via two steps: (a) Simulate a known system such as the diffusion of sodium on hexagonal surface, using explicit inter-adsorbate interactions (b) simulate the system using Generalised Langevin MD simulator, without explicit interactions, but with a random force as explained above. By that, show that non-interacting Generalised Langevin simulations (with “memory friction”) can be used to represent inter-adsorbate interactions.
Extensions to the project can be (a) look for mathematical relations between the two representations of the inter-adsorbate interactions, and (b) model a more challenging system such as water diffusion on a topological insulator, effectively extending a recent published work ( https://www.nature.com/articles/s41467-019-14064-7 ).
Supervisor:
Dr Nadav Avidor
Nature of project:
computational
Requirements:
Part 1B level physics or above
Position dependent friction
In analysing measurements of surface diffusion, we normally assume that nanoscale friction is not position dependent (can be averaged over the unit-cell). The project would be to explore and characterise when this assumption breaks.
Supervisor:
Dr Nadav Avidor
Nature of project:
computational
Requirements:
Part 1B level physics or above
Quantum mechanical scattering calculations (multiple projects)
Helium scattering unlocks experimental regimes which are important for science of thin films, such as the growth of 2D materials and quantum technology devices.
Understanding the scattering of a helium atom from a surface requires full quantum mechanical calculations, yet we often use approximations such as the kinematic approximation. A few projects are available in this context. Using either close coupled or wavepacket propagation calculations (in 3D):
1. Evaluate the kinematic approximation for scattering from diffusing atoms.
2. Evaluate the kinematic approximation for scattering from a rotating benzene molecule.
In these projects the students will use available codes, however some programing will be required (mostly in Matlab, however some may need Fortran, and some may need CUDA C, depending on the project).
Supervisor:
Dr Nadav Avidor
Nature of project:
computational
Requirements:
Part 1B level physics or above
Surface Diffusion of Molecule Carrier
The surface diffusion of the organic molecule anthraquinone is fascinating. It is suspected to pivot around its oxygen atoms, in a stepping-like motion, maintaining an overall linear diffusion. It was also found that anthraquinone can attach and detach CO2 molecules while it diffuses, making it a molecule carrier [Science, DOI: 10.1126/science.1135302].
So far, information reported in the literature (on anthraquinone diffusion) was deduced from STM cryogenic measurements, where degrees of freedom in non-simple molecules are usually frozen. Using the Helium-3 Spinecho spectrometer (home built by the SMF group at the cavendish lab, first and uniqe in the world), we were able recently to measure the diffusion of anthraquinone at eleveted temperatures, where technologically relevant processes takes place.
This project is to perform analysis of our recent experimental measurements, by applying molecular dynamics simulations combined with global bayesian statistics.
Supervisor:
Dr John Ellis and Dr Nadav Avidor
Nature of project:
computational
Requirements:
Part 1B level physics or above
In-beam elements for correction of precession solenoids inhomogeneity
Supervisor:
TBD
Nature of project:
computational
Requirements:
Part 1A level physics or above
Unsupervised and Loosely supervised Projects (some suitable for 1A students)
Last summer we ran a few unsupervised and loosely supervised projects, which turned to be very successful, and even generated published work. We intend to offer such projects again.