Ab initio crystal structure determination from powder diffraction data is not yet a straightforward process: it is strongly disrupted by the low quality of the estimated Bragg reflection intensities. In a two-stage method the integrated intensities calculated from a powder pattern are immediately submitted to direct methods to obtain phases. The larger the accuracy of the integrated intensities, the more efficient the phasing process. A systematic decomposition procedure was introduced in the EXPO2004 program to improve the efficiency of the phasing process. The disadvantage of this approach is that a large number of feasible trial structures are generated, among which the correct solution must be recognized. A new procedure is described aiming at introducing strategies to reduce the total number of trials to explore by defining an appropriate figure of merit able to regroup trial structures into different batches, each element of a batch sharing a high percentage of atoms with the other elements of the same batch. The new figure of merit, implemented in an updated version of EXPO2009, is able to discriminate non-solutions from promising trials, corresponding to incomplete or rough models which evolve, after refinement, into the correct solution.

The EXPO package [1] is a computing program able to successfullyprovide the structure solution by starting from minimal information:the experimental powder diffraction pattern and the chemical formulaof the compound to investigate. EXPO can be used for solving organic,inorganic, metalorganic structures for a large variety of applications.Automation and efficiency are suitably combined in EXPO forperforming all the steps of the solution process: indexation, spacegroup determination, intensity extraction, structure solution, modeloptimization, Rietveld refinement.Automation. Default strategies, identified as the most effective,are selected for carrying out the solution process automatically andquickly. They may fail when the experimental data resolution is badand/or the structure complexity is remarkable. In these cases, EXPOcan promptly switch to appropriate strategies by profiting by a userfriendly graphical interface.Efficiency. Innovative theories and computing procedures aimingat making straightforward all the steps of the solution process areintegrated in EXPO and widely tested. EXPO is continuously updatedand optimized in terms of both computing efficiency and graphicalperformances.In EXPO the two stages based ab initio approach (in the first stagethe intensities are extracted; in the second, the reflections are phasedby Direct Methods) [2] is the automatic choice. The model providedby Direct Methods is usually partial and approximate because of wellknown unavoidable problems in powder diffraction (overlapping,background, preferred orientation). It can be optimized by default [3],[4] and/or non-default strategies [5], [6] which are able to reduce theerrors depending on the limited experimental resolution.A very recent study has regarded a new and more effective figureof merit [7], alternative to the classical one. It is able to pick up thephasing trial corresponding to the best solution among several feasibleones.Non ab initio method, requesting the knowledge of the expectedmolecular geometry and based on Simulated Annealing technique[8, 9], can optionally be attempted particularly in case of solution oforganic compounds.Examples of successful automatic runs by EXPO will be discussedin addition to available special strategies.

A new procedure (COVMAP) has been developed with the aim of recovering the full structure from very poor models, such as those provided by direct methods in unfavorable conditions. The procedure is based on the concept of covariance between points of an electron density map, mathematically set out by the authors in a recent paper: i.e. the density at one point depends on the density at another point of the map if their covariance is not vanishing. This concept suggested a procedure of electron density modification that uses pairs of model peaks to restrict the region where the density modification should be applied. Such modified densities lead to additional peaks, which in turn are submitted to two other important phasing tools present in EXPO2011, the resolution bias minimization and weighted least-squares procedures, which relocate, refine or reject these peaks. The procedure is cyclic and often leads to the correct structure even if the starting model is very poor.

Quite recently two papers have been published [Giacovazzo & Mazzone (2011). Acta Cryst. A67, 210218; Giacovazzo et al. (2011). Acta Cryst. A67, 368382] which calculate the variance in any point of an electron-density map at any stage of the phasing process. The main aim of the papers was to associate a standard deviation to each pixel of the map, in order to obtain a better estimate of the map reliability. This paper deals with the covariance estimate between points of an electron-density map in any space group, centrosymmetric or non-centrosymmetric, no matter the correlation between the model and target structures. The aim is as follows: to verify if the electron density in one point of the map is amplified or depressed as an effect of the electron density in one or more other points of the map. High values of the covariances are usually connected with undesired features of the map. The phases are the primitive random variables of our probabilistic model; the covariance changes with the quality of the model and therefore with the quality of the phases. The conclusive formulas show that the covariance is also influenced by the Patterson map. Uncertainty on measurements may influence the covariance, particularly in the final stages of the structure refinement; a general formula is obtained taking into account both phase and measurement uncertainty, valid at any stage of the crystal structure solution.

SIR2014 is the latest program of the SIR suite for crystal structure solution of small, medium and large structures. A variety of phasing algorithms have been implemented, both ab initio (standard or modern direct methods, Patterson techniques, Vive la Différence) and non-ab initio (simulated annealing, molecular replacement). The program contains tools for crystal structure refinement and for the study of three-dimensional electron-density maps via suitable viewers.

Innovative methodologies, introduced in the software EXPO and working both in the reciprocal and in the direct space, can be successfully adopted for solving crystal structure by X-ray powder diffraction data. The principles underlying these methodologies are summarized. Three representative examples of crystal structure solution of the peptides Z-(Aib)(2)-OH, Z-(Aib)(3)-O-t-Bu and Z-(Aib)(4)-OH are discussed in relation to their different degree of structure complexity.

EXPO2011 is a new package for phasing crystal structures from powder diffraction diagrams. It is able to carry out all the steps necessary for crystal structure solution, from pattern indexation up to Rietveld method for structure refinement: for each step, the basic algorithm is described. Phasing is performed via ab initio (e.g., Direct Methods, integrated by real space refinement) and non ab initio techniques (e.g., simulated annealing algorithm, when molecular geometry is a priori known). Some emphasis is given to running procedures: the main commands and directives are described, to allow the user to run default and non-default phasing attempts.

EXPO2013, the heir of EXPO2009, has been enriched by a variety of new algorithms and graphical tools aiming at strengthening the individual steps of the powder structure solution pathway. Particular attention has been addressed to the procedures devoted to improving structural models provided by direct methods in ab initio approaches. In addition, a new procedure has been implemented, working in direct space, which may be chosen by the user as an alternative to the traditional simulated annealing algorithm. © 2013 International Union of Crystallography Printed in Singapore - all rights reserved.

Rietveld refinement succeeds if the model is sufficiently close to the target structure. Severe distortions in the structural parameters or relatively high percentage of missed atoms do not allow the refinement to converge. Ab initio techniques like Direct or Patterson methods may rely only on the reflection intensities provided by full pattern decomposition algorithms. Owing to the relevant errors in the intensity estimates, the structural models provided by such phasing techniques are poor and require supplementary improvements before being submitted to Rietveld refinement. The situation is more favorable for non-ab initio techniques, when they can exploit the prior information on the full molecular geometry: their models can be easily submitted to Rietveld refinement, but often the accuracy of the refined model relies on the geometrical information rather than on the experiment. In this paper ab natio and non-ab initio phasing tools of EXPO2011, the updated and more powerful version of EXPO2009 [1], are described together with ancillary techniques for improving and completing the structural models.

The first structure of an aromatic bis(trifluoroborate) dipotassium salt, elucidated by the combination of crystallography, DFT calculations, topological and non-covalent interaction analysis, discloses a 3D network undergoing spontaneous self-assembly thanks to the massive participation of weak intra- and intermolecular interactions for which fluorine atoms proved to play a leading role.

QUALX2.0 is the new version of QUALX, a computer program for qualitative phase analysis by powder diffraction data. The previous version of QUALX was able to carry out phase identification by querying the PDF-2 commercial database. The main novelty of QUALX2.0 is the possibility of querying also a freely available database, POW_COD. POW_COD has been built up by starting from the structure information contained in the Crystallography Open Database (COD). The latter is a growing collection of diffraction data, freely downloadable from the web, corresponding to inorganic, metal-organic, organic and mineral structures. QUALX2.0 retains the main capabilities of the previous version: (a) automatically estimating and subtracting the background; (b) locating the experimental diffraction peaks; (c) searching the database for single-phase pattern(s) best matching to the experimental powder diffraction data; (d) taking into account suitable restraints in the search; (e) performing a semi-quantitative analysis; (f) enabling the change of default choices and strategies via a user-friendly graphic interface. The advances of QUALX2.0 with respect to QUALX include (i) a wider variety of types of importable ASCII file containing the experimental diffraction pattern and (ii) new search-match options. The program, written in Fortran and C++, runs on PCs under the Windows operating system. The POW_COD database is exported in SQLite3 format.

The new method RAMM (random-model-based method) has been developed and implemented in the EXPO computing program for improving the ab initio crystal structure solution process. When the available information consists of only the experimental powder diffraction pattern and the chemical formula of the compound under study, the classical structure solution approach follows two main steps: (1) phasing by direct methods (or by Patterson methods) in order to obtain a structure model (this last is usually incomplete and/or approximate); (2) improving the model by structure optimization techniques. This article proposes the alternative procedure RAMM, which skips step (1) and supplies a fully random model to step (2). This model is then submitted to effective structure optimization tools present in EXPO - wLSQ (weighted least squares), RBM (resolution bias minimization) and COVMAP (a procedure of electron density modification based on the concept of covariance between points of the map) - which are able to lead to the correct structure. RAMM is based on a cyclic process, generating several random models which are then optimized. The process stops automatically when it recognizes the correct structure.

A strategy for the stereoselective functionalization of thietane 1-oxide has been developed. Mono (C2 substituted) and doubly (C2, C4 disubstituted) functionalized thietanes have been obtained from the readily available thietane 1-oxide by using the corresponding organometallic intermediates that reacted with electrophiles leaving intact the 4-membered ring. This journal is © 2014 the Partner Organisations.

Aziridines are widely used as versatile building blocks for the synthesis of a variety of biologically and pharmaceutically importantmolecules [1]. Among the available methodologies for the preparation of functionalised aziridines, the lithiation/trapping sequence of simpleparent aziridines is growing in importance [2]. Present work reports the results concerning the structural study of aziridine-2-methanolderivatives by dynamic nuclear magnetic resonance (DMNR) [3] and single crystal X-ray diffraction. A careful examination of the aziridine2 (see figure 1) by NMR data, shows that at room temperature the meta protons and the protons of the two ortho methyl groups of the mesitylring gives featureless lumps, likely as consequence of a restricted rotation around the Csp3 -Csp2 bond between the carbinolic carbon andthe ipso carbon of the mesityl ring. Also in the solid state, an almost identical arrangement of the mesityl ring, with respect to the other twophenyl rings, has been observed. In this case, the aziridine nitrogen substituent was found in a syn relationship with respect to the carbinolicgroup so preventing the possibility of forming a hydrogen bond with the hydroxyl group. This evidence underlines a slow rotation of themesityl group instead of a nitrogen inversion as usually is expected. By X-ray analysis of 3 and diast-3 (figure 1), it is estimated that thenitrogen substituents set on the opposite side with respect the carbinolic carbon. Via study of crystal structures, the presence of hydrogenbonds between the hydroxyl group and the aziridine nitrogen lone pair was ascertained for aziridine 3 but not for diast-3 (figure 1). Such a hydrogen bond, which is persistent also in solution, could prevent the formation of conformational diastereoisomers by rotation around the bond between the carbinolic carbon and the aziridine quaternary carbon. This hypothesis has been demonstrated by NMR experiments. This investigation allowed us to calcolate some Csp2-Csp3 and Csp3-Csp3 rotational barriers and highlight a sort of "geared" rotation between aryl and alkyl substituents; this is possible if the structure is not in a locked conformation by an hydrogen bond as demonstrated by NMR and X-ray diffraction.

A new approach is presented here for the structural solution of anhydrous and hydrated metal-organic coordination compounds of alkali and alkaline earth metals by using rigid structure fragments in combination with simulated annealing algorithms. We empirically show how this approach minimizes computation time, while allowing us to obtain the correct result. The structures of two novel ionic co-crystals have been solved from powder data with this approach.

A reciprocal-space resolution bias correction algorithm has been recently suggested, providing suitable corrections for the classical atomic scattering factors. The Fourier maps calculated by using as coefficients the structure factors obtained by the modified scattering factors proved to be less resolution biased. In this paper the correction has been generalized in order to apply it to the experimental structure factor moduli; in this way more useful electron density maps may be calculated. In a less recent paper a direct-space resolution bias correction algorithm was devised and tested on a large set of powder patterns; the algorithm implies the modification of the electron density maps to reduce the truncation errors in the Fourier syntheses. In the present paper direct and reciprocal resolution bias correction algorithms are combined into the dual-space resolution bias correction algorithm. The usefulness of the new algorithm is checked on a set of powder patterns.

The limited resolution of experimental diffraction data distorts the Fourier synthesis so that the electron density map obtained by the structure factors is an imperfect representation of the true density: the worse the experimental resolution, the less accurate the Fourier representation. We have recently developed new methods aiming at reducing the resolution effects by correcting it both in the Reciprocal Space (RS) and in the Direct Space (DS): e.g., by modifying the atomic scattering factors in RS and the electron density map in DS. The Dual-Space (DUS) method combines the RS and DS procedures. In addition, new computing strategies have been developed for improving the Fourier map calculation. The DUS algorithm has been introduced in the EXP02010 package in order to obtain more reliable structure models. It has been successfully applied to several test structures. The main features of the new procedure and the results of our applications will be described.

The big bang-big crunch method is a global optimization approach developed upon the analogy of one of the cosmological theories of the evolution of the universe. It has been suitably combined with a simulated annealing algorithm and used for solving crystal structure from powder diffraction data in direct space. When compared with the traditional simulated annealing method, it provides a significant advance: good solutions are attained in a shorter time. The new method has been implemented in the EXPO package. Its successful application is demonstrated with examples of already known structures.