Predicting surface stress response of crystals through code

Elastically bendable molecular crystals can be used in flexible semiconductor devices without significant loss in electron mobility.

Custom-developed computer code that accurately calculates the mechanical properties of molecular crystals is revolutionising materials discovery and providing deeper scientific insights into the complex behaviour of molecular crystals.

MechaPredict, a Python code developed by Sharmarke Mohamed’s group in the Green Chemistry and Materials Modelling Laboratory at Khalifa University of Science and Technology, is at the heart of this groundbreaking work. The code relies on elastic constants generated using density functional theory, a quantum mechanical method for modelling solid-state materials.

“The code takes as input a crystallographic information file and the elastic constants for the crystal, as well as a user-specified set of Miller indices for crystal surfaces of interest. It then computes the facet-dependent mechanical properties of the crystal within seconds,” explains Mohamed. “The estimates of the mechanical properties from MechaPredict are in excellent agreement with experimental nanoindentation measurements for crystals that comprise strong cohesive intermolecular forces, such as amino acids.”

Unlike crystals derived from metals and ceramics, where the composition is less variable, molecular crystals are composed of covalently bonded chemical fragments that can be reconfigured in an endless number of ways via bond-breaking and bond-formation events. This brings numerous opportunities for customizing the crystal’s behaviour.

“Small changes in the chemical structure of these fragments can have significant effects on the properties of the crystal. This in turn offers more scope for discovering new molecular crystals for various technological applications,” says Mohamed.

Globally, there is now intense activity around investigating molecular crystals, with researchers keen to identify new useful examples and better understand their behaviour. In a recent Chemical Science paper1 authored by Mohamed and colleagues, they showed that elastically bendable molecular crystals can be used in flexible semiconductor devices without significant loss in electron mobility.

“Our code allows high-throughput screening of the mechanical properties of crystals without doing a number of serial and time-consuming nanoindentation experiments,” Mohamed says. “The code also allows us to gain a deeper understanding of how molecular crystals deform under applied stress. By doing careful density functional theory simulations of the facet-dependent changes in the stress response of crystals, we are now in a better position to understand, for example, why some crystals crack under stress while others can withstand significant stresses before they fracture.”

MechaPredict is compatible with both Windows and Linux operating systems. Its user-friendly graphical interface makes it accessible and valuable to both theoreticians and experimentalists. Looking to the future, Mohamed’s team is exploring the idea of integrating MechaPredict within a machine-learning workflow to predict not just the mechanical properties of molecular crystals but also the possible experimental conditions for targeting their discovery in the laboratory.


  1. Samanta, R.; Das, S.; Mondal, S.; Alkhidir, T.; Mohamed, S.; Senanayak, S.P. and Reddy, C.M., Elastic organic semiconducting single crystals for durable all-flexible field-effect transistors: insights into the bending mechanism. Chem. Sci., 14 (6), 1363-1371, 2023. | Article

Related articles