Control and Prediction of the Organic Solid State: Translating the Technology

Lead Research Organisation: University College London
Department Name: Chemistry

Abstract

Many organic molecules can adopt more than one solid crystalline form, including polymorphs which differ only in the arrangement of the molecules. This can be exploited in the development of new speciality materials with optimised physical properties such as non-linear optical coefficients or molecular electronics, but it can prove disastrous for quality control when the new form appears unexpectedly during production or storage, as has occurred for some pharmaceuticals. However, the factors that control the diversity of possible solid forms are poorly understood. It has been said that the number of polymorphs of a molecule depends on the time and money spent looking for them. This is certainly a factor, when about 90% of systems studied for pharmaceutical companies are found to have multiple solid forms, and less than 5% of entries in the Cambridge Structural Database have two or more structures. The development of new organic materials and quality control in their manufacture requires a more scientific approach to understanding the possible range of crystal structures.The CPOSS project has been developing a computational method of predicting which crystal structures of a molecule are sufficiently low in energy that they might be able to form. We have also developed a workflow, including a uniquely powerful automated solvent crystallisation system, which is discovering many new crystalline forms even for the highly studied systems we used to validate the approach. By combining these results with other experimental and computer simulation work, the multi-disciplinary CPOSS team are developing a molecular level picture of what determines which of the computed low energy structures can be found experimentally for several specific molecules, ranging from the influence of solvent, to that of surfaces and even impurities. We even predicted possible new crystal structures of aspirin, paracetamol and piracetam before they were experimentally found! This Translation grant will enable the core of the CPOSS team to use the computational and experimental technologies they have developed in work with a broad range of collaborators on their particular organic solid state problems. These areas include molecular electronics and optical materials, energetic materials, pigments, agrochemicals and pharmaceuticals. Thus, we will develop methods of accurately computing which crystal structures are possible, and establish how to use the computed structures and experimental results to effectively determine which crystal structures might be found and the most efficient route to obtaining them. The CPOSS databases of known and hypothetical crystal structures and their properties, when covering a wide range of organic molecules, will provide a basic technology for polymorph prediction and control. This project will make available these complementary techniques to both academic and industrial organic solid state research scientists.

Publications

10 25 50
 
Description We have developed a computational method of predicting which crystal structures of a molecule are sufficiently low in energy that they might be able to form, and a uniquely powerful experimental workflow for identifying those that do. The combined multi-disciplinary approach can obtain a molecular level picture of what determines the range of possible outcomes of crystallisation of a molecule. During the Translation Grant, we have more than doubled, to over 170, the number of molecular systems whose computed crystal energy landscapes, the range of thermodynamically plausible crystal structures, are stored on our Crystal Navigator database. This includes a significant extension into multi-component systems, such as salts, cocrystals and hydrates of different stoichiometries, assessing the ability to determine which phase will form. There has been a qualitative change in the size and flexibility of molecule that can be tackled through algorithmic development, with a new program DMACRYS able to model crystals of much larger organic molecules. This allowed us to successfully predict the structure of a molecule more representative of modern pharmaceuticals in development, in the International Blind Test of Crystal Structure Prediction. The calculated crystal structure have complemented significant advances in our methodology for experimental investigations of the organic solid state. We have been able to characterise completely many microcrystalline powders, using the calculations to place the protons. We have learnt how to interpret the crystal energy landscape to suggest possible types of disorder, which can significantly affect the reproducibility of crystalline properties and reproducibility, and quantify when disorder is thermodynamically inevitable. The prediction of a propensity for forming a wide range of solvates is also possible, as seen in our studies of olanzapine yielding 56 solvates as well as 3 polymorphs. We are gaining confidence in determining which of the computed low energy structures may be found as metastable polymorphs, and our investigations of molecule/surface interactions show that heterosurfaces (designed, or from impurities) can produce new polymorphs that are not accessible by conventional experiments. We have used crystal energy landscapes to choose a templating experiment that produced the first crystal of the anti-epileptic, carbamazepine with catemeric hydrogen bonding, demonstrating that predicting crystal structures can lead to new polymorphs that would not otherwise be found.
Exploitation Route The crystal energy landscapes can be used as a complementary technique as part of the pharmaceutical development process, not only to confirm that the most stable form is known, but also to suggest which other structures may be found as polymorphs and facilitate characterising these structures. This information on the range of crystal packings of a molecule should allow a qualitative improvement in our ability to design organic solid state products and crystallisation processes suitable for their manufacture. This research has developed organic crystal structure software and experimental crystallisation and characterisation techniques, and algorithms for combining and interpreting the results, that can be applied to other organic molecules. These procedures have and continue to be disseminated into industry and academia.
Sectors Chemicals,Pharmaceuticals and Medical Biotechnology

URL http://www.cposs.org.uk
 
Description The methods developed have been used in industrial collaborations in conjunction with their screening results, resulting in some publications. More companies are becoming interested in using commercial crystal structure prediction (CSP) services, and Prof Price has spoken at companies such as Merck, Pfizer and Lilly central sites in USA to help inform their decisions. Since this award ended, Price's group have performed CSP studies in collaboration with Eli Lilly scientist's experimental work, showing how the application of CSP can help complete and give confidence to a molecular level picture of the crystallisation behaviour of a drug molecule. This has since considerably increased the interest in industry in CSP, as shown by the number of companies now offering this service, and the vision developed in the Faraday Discussion on Crystal Structure Prediction in 2018.
First Year Of Impact 2010
Sector Pharmaceuticals and Medical Biotechnology
Impact Types Economic

 
Description Commonwealth Scholarship
Amount £60,000 (GBP)
Organisation Government of the UK 
Department Commonwealth Scholarship Commission
Sector Public
Country United Kingdom
Start 09/2009 
End 12/2012
 
Description ELI Lilly and Company
Amount £142,733 (GBP)
Funding ID A PSR&D Technology S Reutzel-Edens 
Organisation Eli Lilly & Company Ltd 
Sector Private
Country United Kingdom
Start 08/2011 
End 02/2013
 
Description EPSRC
Amount £5,939,782 (GBP)
Funding ID EP/I033459/1 
Organisation Engineering and Physical Sciences Research Council (EPSRC) 
Sector Academic/University
Country United Kingdom
Start 10/2011 
End 09/2016
 
Description EPSRC
Amount £499,000 (GBP)
Funding ID EP/K004670/1 
Organisation Engineering and Physical Sciences Research Council (EPSRC) 
Sector Academic/University
Country United Kingdom
Start 04/2012 
End 03/2013
 
Description EPSRC
Amount £79,217 (GBP)
Funding ID EP/K504129/1 
Organisation Engineering and Physical Sciences Research Council (EPSRC) 
Sector Academic/University
Country United Kingdom
Start 04/2013 
End 04/2015
 
Description EPSRC
Amount £79,430 (GBP)
Funding ID K504117/1 
Organisation Engineering and Physical Sciences Research Council (EPSRC) 
Sector Academic/University
Country United Kingdom
Start 11/2012 
End 10/2015
 
Description EPSRC
Amount £4,348,959 (GBP)
Funding ID EP/K503289/1 
Organisation Engineering and Physical Sciences Research Council (EPSRC) 
Sector Academic/University
Country United Kingdom
Start 10/2012 
End 09/2018
 
Description Eli Lilly Research Agreement
Amount $167,578 (USD)
Organisation Eli Lilly & Company Ltd 
Sector Private
Country United Kingdom
Start 11/2014 
End 05/2016
 
Description Eli Lilly Research Funding
Amount $1,409,787 (USD)
Organisation Eli Lilly & Company Ltd 
Sector Private
Country United Kingdom
Start 08/2017 
End 07/2021
 
Description Fujifilm
Amount £80,000 (GBP)
Funding ID Strathclyde EPSRC KTA project EP/H50009X/1 
Organisation Fujifilm 
Sector Private
Country Japan
Start 04/2011 
End 03/2012
 
Description GlaxoSmithKline (GSK)
Amount £40,000 (GBP)
Funding ID KT Secondments EP/H500278/1 UCL 
Organisation GlaxoSmithKline (GSK) 
Sector Private
Country Global
Start 08/2010 
End 03/2011
 
Description Horizon2020-FETOPEN
Amount € 2,886,323 (EUR)
Funding ID 736899 MagnaPharm 
Organisation European Research Council (ERC) 
Sector Public
Country European Union (EU)
Start 01/2017 
End 12/2019
 
Description Lilly Research Awards Program
Amount $297,371 (USD)
Organisation Eli Lilly & Company Ltd 
Sector Private
Country United Kingdom
Start 01/2013 
End 01/2015
 
Description Scottish Funding Council
Amount £720,000 (GBP)
Funding ID RKES-GS0280 
Organisation Government of Scotland 
Department Scottish Funding Council
Sector Public
Country United Kingdom
Start 10/2010 
End 09/2014
 
Description Syngenta
Amount £40,000 (GBP)
Funding ID Strathclyde EPSRC KTA project EP/H50009X/1 
Organisation Syngenta International AG 
Sector Public
Country Global
Start  
 
Title DMACRYS 
Description DMACRYS - Energy minimisation package to simulate rigid molecules with multipoles This package models crystals of rigid molecules, allowing lattice energy minimisation and the calculation of second derivative properties such as phonon frequencies and mechanical properties. It is designed to use anisotropic atom-atom model intermolecular potentials, particularly distributed multipole electrostatic models, but also has the capability of using distributed polarizabilities and anisotropic repulsion. 
Type Of Technology Software 
Year Produced 2014 
Impact The program is being licensed by UCL-B, and maintained and updated, with the 2014 release version being 2.0.8. There are licenses to industry and a large number of academic research groups throughout the world. 
URL http://www.cposs.org.uk/