Table Of Contents

Previous topic

James Kermode

Next topic


This Page

Research Interests

Hybrid Modelling of Materials Chemomechanics


My research applies multiscale quantum mechanical and classical modelling (in the QM/MM style) to material systems. In particular I am interested in the concerted action of local chemistry and long-range elasticity (‘chemomechanics’). For example, when a brittle material is loaded to the limit of its strength, it fails by nucleation and propagation of a crack. The conditions for crack propagation are created by the concentration of a long-range stress field (illustrated in image, left) at an atomically sharp crack tip, creating a complex and strongly coupled multiscale system.

Macroscopic scattering of cracks by isolated impurities

In an article recently published in Nature Communications, we showed that a single atomic defect can be enough to deflect a crack as it travels through a crystal, overturning the conventional wisdom about how materials fracture. Our research also reveals that, despite these defects, a swift fracture can produce atomically smooth surfaces, potentially useful for improving the efficiency of gem processing and making sharper diamond cutting tools.


The image above how a crack propagating through a silicon crystal ca be scattered by a single isolated boron impurity (left, coloured orange), leading to submicron surface ridges which grow out of the cleavage plane (right, STM image, size 100 x 100 nm). See also this less technical research highlight on the Thomas Young Centre (TYC) webpage.

This movie shows a crack propagating on the (111) surface of an silicon crystal containing a substitutional boron impurity at a speed of 1400 m/s. Propagation is stable, resulting in mirror smooth cleavage surfaces. Silicon atoms coloured dark blue, plus the orange boron atom, are treated at the quantum mechanical level in the multiscale simulations. (Only a small part of the overall system close to the crack tip is shown). When the crack encounters the impurity, it forms an extended crack tip reconstruction promoted by the preference of the boron to form sp2 bonds with three neighbours at angles of 120 degrees, eventually leading to a downwards step in the crack path.

Stress Corrosion Cracking

Stress corrosion cracking (SCC) is of great significant in crystalline and amorphous oxides. In SCC, crack advance is driven by individual chemical reactions with corrosive agents present in the enviornment.


The image to the right shows snapshots from simulations showing for the first time that stress corrosion cracking of silicon crystals can be mediated by O2 dissociative chemisorption, with each oxygen molecule cleaving a single Si-Si bond and leading to a crack advance step. For more details see our recently published joint experimental/theoretical paper in Phys. Rev. Lett., where simulations are used to make predictions in excellent quantitative agreement with experiment. The article has been featured as a TYC research highlight and in an article written for a general audience on the Argonne Leadership Computing Facility website.

Low speed fracture instabilities in silicon


In an earlier work, we discovered a new low speed fracture instability in silicon in which cracks propagating on the (111) cleavage plane below about 1000 m/s become unstable, via a positive feedback process initiated by a novel crack tip reconstruction, leading to experimentally observable surface features. In the image to the left, red atoms are treated at the QM level, while yellow atoms are modelled classically. This work was the main outcome of my PhD and was published in Nature.

‘Learn on the Fly’ hybrid scheme

Our work has shown that accurate fracture simulations can be made practicable by combining a quantum mechanical description of the processes taking place near the crack tip with a classical atomistic model that captures the long-range elastic relaxation. I am one of the developers of the Learn on The Fly (LOTF) scheme, which uses a hybrid QM/MM approach to couple quantum and classical models.

The video above shows a LOTF simulation of brittle fracture on the (111) plane in silicon. Blue atoms are treated at the QM level, and grey atoms with a classical interatomic potential. Atomically smooth fracture surfaces are obtained; note also the \pi-bonded Pandey reconstruction consistenting of alternating pentagons and heptagons.

For a more in depth description of this approach, see a recent tutorial written for a winter school at the International Centre for Theoretical Physics (ICTP) in Trieste.