Improved performance of crystal structure solution from powder diffraction data through parameter tuning of a simulated annealing algorithm

Abstract

Synopsis ignificant gains in the performance of the global optimisation algorithm in the DASH software package have been realised using the irace automatic configuration tool and are demonstrated using over 100 powder diffraction data sets. Abstract Significant gains in the performance of the simulated annealing algorithm in the DASH software package have been realised by using the irace automatic configuration tool to optimise the values of three key simulated annealing parameters. Specifically, the success rate in finding the global minimum in intensity  2 space is improved by up to an order of magnitude. The general applicability of these revised simulated annealing parameters is demonstrated using the crystal structure determinations of over 100 powder diffraction data sets.

Pushing the Limits of Molecular Crystal Structure Determination From Powder Diffraction Data in High-Throughput Chemical Environments

“…Table 6 illustrates that even with the relatively short data collection times used, structures are solved with a high level of success and accuracy. As expected, as molecular and crystallographic complexity rises, success rates drop off (Kabova et al, 2017). Nevertheless, in all cases, high-quality crystal structures were obtained that showed excellent agreement with the reference single crystal structures, as evidenced by the low RMSD values for the 15 molecule overlays performed using Mercury (Table 7 and Figure 3).…”

“…Since the late 1990s, developments in crystal structure determination from powder diffraction data (SDPD) methods have meant that it is often relatively straightforward to solve complex molecular crystal structures from laboratory-based PXRD data alone (Kabova et al, 2017;Shankland et al, 2013;Fernandes et al, 2007;Bushmarinov et al, 2012;Chernyshev et al, 2010;Chernyshev et al, 2009) without the need for the high instrumental resolution that a synchrotron source allows. In principle, therefore, SDPD could be much more routinely accessible to the laboratory chemist.…”

“…Starting models for SDPD were generated from crystal structures stored on the Cambridge Structural Database (CSD) Version 5.38 (Groom et al, 2016), and the relevant CSD "refcodes" are shown in Table 2. The number of simulated annealing (SA) DASH runs, the number of SA moves utilised in each run and the SA control parameters were based on recommendations made by Kabova (Kabova et al, 2017). For data collected using the reflection flat plate instrument, a March-Dollase correction for preferred orientation was included in DASH as optimisable parameter.…”

Co-Crystal Structures of Furosemide:Urea and Carbamazepine:Indomethacin Determined from Powder X-Ray Diffraction Data

“…Where data for full structure determination were required, a variable-counting time scheme [12] was used. In the case of FUR:URE, the total data collection time was 18 h (see Table 1); in the case of CBZ:IND form II, the total data collection time was scaled to 42 h. Structure indexing and structure solution were performed using the DASH software package [13][14][15]. The starting molecular models used in the DASH global optimization procedure were obtained directly from the Cambridge Structural Database (CSD) [16] and are summarized in Table 2.…”

Hybrid Approach with Improved Genetic Algorithm and Simulated Annealing for Thesis Sampling

“…When iterating to update the feasible solution, a solution worse than the current solution is accepted with a certain probability, so it is possible to get the optimal global solution by jumping out of the local optimal solution [12]. In [13][14][15], the simulated annealing algorithm was improved or combined with other algorithms to achieve a better performance in experiments. Figure 2 is an example.…”