Multiscale simulation of crystal growth from solution

• Applicant: Professor Dr.-Ing. Heiko Briesen
• Subject Area: Chemical and Thermal Process Engineering
• Term: from 2009 to 2021
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 155981807

• Final Report Year: 2021

Final Report Abstract

An in silico prediction of crystal growth and dissolution rates would be of a great benefit to science and industry but is greatly hindered by the molecular nature of the phenomenon. This challenge is raised in the present project. A new multiscale approach to calculate absolute crystal growth and dissolution rates for small drug-like organic molecules is presented. A suitable combination of molecular dynamics (MD), kinetic Monte Carlo (kMC) and continuum simulations allows to cover length and time scales from angstroms to millimeters and from picoseconds to hours, thus, making the results industrially relevant and directly comparable to experiment. The predictions are based only on the molecular structure and crystal packing information. The growth and dissolution processes are first investigated within an MD framework. To see dissolution on MD time scales and to obtain the rates for kMC simulations, we propose the use of a three-dimensional crystal representation in MD simulations, where sticky dummy atoms are introduced into the simulation box to ensure constant undersaturation during the simulation. In case of growth simulations, Constant Chemical Potential Molecular Dynamics method is applied to maintain a region containing the growing crystal and its immediate surrounding at constant supersaturation. MD simulations are used to parametrize kMC simulations. Several MD analysis algorithms are elaborated for providing a simple and universal way to define Markovian states and calculate rate constants for kMC simulations. They can be easily applied to the characterization of dissolution as well as growth processes for arbitrary molecular crystals without the need for intricate and substance specific definition of all possible surface step and kink sites, thus, making the approach promising for high-throughput use in the prediction of crystal growth and dissolution rates for small drug-like molecules. Given the set of states and rate constants, a kMC technique provides a stochastic procedure to produce a state-to-state trajectory, representing a valid realization of the state-to-state dynamics. At the same time, the range of length and time scales accessible to simulation are significantly extended, making it possible to calculate crystal face displacement velocities. In continuum simulations the diffusion equation is solved numerically with the initial concentration corresponding to such from MD and kMC simulations. Thereby, the velocities obtained through kMC simulations are used to describe boundary conditions on the crystal faces. The success of the elaborated approach is demonstrated on the example of aspirin crystal dissolution in water, where the obtained simulation results are in very good agreement with experimental results. For testing and validation of MD analysis algorithms, also paracetamol and glycine crystal were simulated. Computationally demanding MD simulations were performed on the Supercomputer SUPERMUC-NG at Leibniz Supercomputing Centre using CPU time provided by the Gauss Centre for Supercomputing.

Publications

• Multiscale modeling of aspirin dissolution: From molecular resolution to experimental scales of time and size. CrystEngComm, 2016; 18: 5302-5312
  Greiner, M., Choscz C., Eder C., Elts, E. and Briesen, H.
  
• Multiskalenmodellierung von Kristallauflosung: von der atomistischen zur experimentellen Skala. ProcessNet Jahrestagung, Frankfurt am Main, 2016
  Greiner, M., Elts, E. and Briesen, H.
  
• Predicting dissolution kinetics for active pharmaceutical ingredients based on their molecular structure. Cryst. Growth Des., 2016; 16: 4154-4164
  Elts, E., Greiner, M. and Briesen, H.
  
• Bridging from nano- to macroscale for crystal growth and dissolution prediction. ISIC20, Dublin, 2017
  Elts, E., Greiner, M. and Briesen, H.
  
• In Silico Prediction of Growth and Dissolution Rates for Organic Molecular Crystals: A Multiscale Approach. Crystals 2017; 7(10): 288
  Elts, E., Greiner, M. and Briesen, H.
  
• Capturing Crystal Shape Evolution from Molecular Simulations. J. Chem. Inf. Model., 2020, 60(12): 6109-6119
  Elts, E. and Briesen, H.
  
• Crystal face identification using 3D imaging Techniques. ISIC21, online, 2021
  Luxenburger, F., Elts, E., Briesen, H.
  
• Influence of Monovalent Salts on α-Glycine Crystal Growth from Aqueous Solution: Molecular Dynamics Simulations at Constant Supersaturation Conditions, J. Phys. Chem. B., 2021, 125(42): 11732-11741
  Elts, E., Luxenburger, F. and Briesen, H.