Nucleation & crystallisation

Crystallisation of compounds is a wondrous process that captures the imagination of children – be it in the preparation of copper sulfate crystals† at school, the growth of crystal gardens‡ at home, or the flowering of a Crystal Christmas tree§ in the Christian yuletide festivities – as well as adult amateur and professional scientists. The latter need crystals for purification, processing, characterisation and control of properties of materials. Therefore, achieving the formation of crystals is a very important enterprise, accentuated this year which is the International Year of Crystallography (see http://www.iycr2014.org/). The formation of crystals is governed by the steps of nucleation and growth, which are the topics of the reviews that are gathered here in this themed issue of Chemical Society Reviews.

But if this process is so well known, carried out by schoolchildren and used on a day to day basis in academic labs and industrial reactors, why the interest and why the need for these reviews now? After all, it is over a hundred years since the thermodynamic principles of nucleation were established. The reason is that although the formation of crystals is done, the mechanisms of their nucleation and growth are not fully understood and while control over crystallisation can be achieved, there are still a number of outstanding fundamental questions that cannot be answered. Equally, there are a growing number of ways to help promote nucleation and control growth of periodically ordered solids. The pursuit of novel challenges in these areas is as valid now as it has ever been.

Interestingly, the current trend of studying out of equilibrium processes in chemical systems in solution finds synergic interests in the growth of ordered structures. The inherently kinetically-governed non-equilibrium processes that give rise to nucleation and subsequent crystal growth can afford a remarkable diversity of structure: the vast possibilities that can be explored when metastable states can be templated or isolated under particular conditions by trapping them are only beginning to be explored.

The idea of this themed issue of Chemical Society Reviews was born from discussions between the guest editors – an organic chemist and solid state physicist by training, working in a Materials Science Institute – who realised the great deal in common that they had (for completely different reasons) in respect to the growth of ordered structures of the materials, and the effects that nucleation, mass transport and growth parameters had on their experiments in disparate chemical systems. It occurred to us that a collection of viewpoints of crystallisation from across chemistry would be a rare and useful volume because of the great advances made over the years in the different disciplines that are at first sight unrelated, from the growth of thin films to the formation of gigantic gypsum crystals in Mexican caves! And here we see that fusion, which we hope will provide the authors and readers with a key reference for understanding and inspiration as well as a springboard to further reading of the seminal works that are cited.

The subject of crystal growth is wide and deep: here we selected authors whom we were sure would give accessible and fresh views of topics that we feel are still very challenging, fundamentally important and yet, perhaps, not well understood generally. New techniques for growth and characterisation also feature predominantly. A background of knowledge in the topics covered here can be gleaned from the reviews and books that are referred to in articles here.

One of the most amazing crystal crops on our planet is that found in the Naica mine in Mexico. “Nucleation and growth of the Naica giant gypsum crystals” is a tutorial review by Fermín Otálora and JuanMa García-Ruíz which leads us from the mystery and awe surrounding the Cave of Giant Crystals to the origins of the growth of these remarkable objects (DOI: 10.1039/c3cs60320b). The very low supersaturation and very low nucleation rate are discussed in a clear and quantitative manner, with one of the most startling observations being the thousand-fold increase in nucleation rate by decreasing temperature only 0.8 °C!

So, where do these “nuclei” come from, what is their nature? This question – despite our incredible analytical tools and relative computational power – taunts scientists still. The review article by Denis Gebauer et al. (DOI: 10.1039/c3cs60451a) takes us through the different theories of nucleation, and focuses on pre-nucleation clusters that are inherent to the current thinking that many scientists have concerning this first step in the formation of crystals. Recent results of experimental and theoretical research are discussed, and the link with phase separation is proposed. Furthermore, the many important (and tempting!) challenges within the field are set out in a set of tantalising questions in the review.

Some consider that crystals can arise from a glass-like state, and therefore the crystallisation process in glasses is particularly relevant for understanding nucleation. The formation of ordered areas in oxide glasses is reviewed in a tutorial by Karpukhina, Hill and Law (DOI: 10.1039/c3cs60305a). It is shown how the mechanism of crystallisation in glasses depends on the processes taking place in the glass when heated in specific ways at the pre-nucleation stage, and the use of multiple analytical techniques to probe these processes is proved to be beneficial for understanding the mechanisms that occur. The systems can be designed such that nanocrystals can be formed. The final materials are composed of different areas with crystalline and amorphous character. The nature and proportion of these areas have a dramatic effect on the final properties of the glasses as materials, so understanding the reactions that take place aids the development of these glass ceramics.

Of course, the understanding of phase behaviour of materials is essential for a true comprehension and control over crystallisation processes. Gérard Coquerel provides a tutorial review that shows beautifully how control over the phase diagram of molecules in solution can lead to an exquisite rationalisation of conditions for crystallisation, even for quite complex mixtures of components (DOI: 10.1039/c3cs60359h). Dominion over the supersaturation is essential for success in a crystallisation, and the author shows how the presence of transient metastable states can be explained by considering the pathways that a process takes.

The use of solution-based chemical synthesis methodologies for growing inorganic nanostructured materials has become a dramatically expanding field in recent years. The flexibility in designing chemical precursors, using different types of solvents and shaping processes has become a great advantage as compared to top-down based preparation techniques. Many different synthetic approaches have been explored to achieve nanomaterials with very diverse functionalities: ferroelectrics, ferromagnetics, catalysts, ionic conductors, superconductors, etc.

In all cases understanding and controlling the nucleation and growth processes are key to achieving novel or advanced functionalities. Preparing materials with low dimensionality, for instance, is a real challenge. Mari-Ann Einarsrud and Tor Grande (DOI: 10.1039/c3cs60219b) show how hydrothermal synthesis and molten salt methods can lead to 1D oxide nanostructures following different strategies to control growth along a single specific crystallographic direction. The thermodynamic and kinetic principles to follow are described clearly for different types of materials. Following a similar path, Carretero-Genevrier, Mestres et al. (DOI: 10.1039/c3cs60288e) describe how ternary ferromagnetic oxide nanowires can be grown using porous polymeric templates deposited on substrates and precursor chemical solutions filling them. We shall see another use of nanoscopic templates later on for organic compounds. For the oxides the complex interplay of the confinement effect, the misfit with the substrate at the interface and the growth temperature to define a given crystallographic structure and specific nucleation events are described and the potential of using such a methodology to prepare novel oxide nanostructures emphasized. In another review, Yuen Wu, Dingsheng Wang and Yadong Li (DOI: 10.1039/c3cs60221d) make an additional contribution to describe the challenges faced to define low dimensional nanostructures from chemical solutions, in particular nanocrystals for heterogeneous catalysis. The challenge here is to achieve specific crystal shapes where selected facets with an enhanced catalytic activity are obtained. The authors emphasize how useful, and how challenging, the use of the presently available techniques for in situ control of nucleation and growth of nanoparticles at real time and in real space still is.

Modelling the formation of crystals is a challenging yet potentially enlightening enterprise. The nature of nucleation in simple liquids via dense amorphous precursors is the subject of a tutorial review by László Gránásy et al. in which they present the use of density functional theory with the phase-field crystal model (DOI: 10.1039/c3cs60225g). It is surmised that even homogeneous nucleation can take place in two steps in this way, and that the crystal phase appears by heterogeneous nucleation on the dense precursor. Time dependent modelling is necessary to probe these kinetically determined processes. The mismatch between lattices (a large mismatch can lead to amorphous layers) determines the wetting at the interfaces and therefore the thermodynamic barrier to the nucleation prior to crystal growth. In a contrasting approach, model systems involving colloids have been used to simulate single-particle dynamic processes in crystallisation (Tian Hui Zhang and Xiang Yang Liu, DOI: 10.1039/c3cs60398a). For example, polymer spheres are shown to order in ways that emulate phenomena one would imagine at the atomic or molecular levels. The assembly of the colloidal spheres show the non-classical steps such as multistep processes and supersaturation-driven mismatch nucleation. Therefore, this kind of system could be useful for studying fundamental aspects of crystallisation, as well as other phase transition processes. Indeed, the self-assembly of various colloidal systems into one-dimensional nanocrystals is reviewed by Ming-Yong Han and co-workers (DOI: 10.1039/c3cs60397k). The assembly is described at interfaces as well as in solutions where chemical bonding, depletion attraction forces and linker-mediated interactions drive the formation of the nanocrystals of inorganic materials that can form rods or wires. The assembly of the rod-like colloids is especially interesting with regard to its possible relation to other systems. It is hoped that these controlled assemblies – formed thanks to the low size distribution in the colloids – will lead to useful materials thanks to their unique properties.

The crystallisation of molecular systems is important across a range of disciplines, from the preparation of the solid forms of therapeutically important drugs to the organisation of molecular materials. One of the major challenges in the area is in the prediction of the solid state structures, which in general terms is done in a qualitative way based on known structures and employing crystal engineering ideas. Sarah Price (DOI: 10.1039/c3cs60279f) details the basics and state of the art in crystal structure prediction of organic compounds using theoretical modelling. The huge recent advances in computer power have aided the correct prediction of increasing numbers of structures. Yet while the present modelling techniques focus on determining the “static” thermodynamic minima, there can be several structures that lie close to this minimum and therefore the technique can help identify possible polymorph occurrence.

Obtaining crystals of molecular systems can be challenging, and three reviews discuss techniques for their preparation. Krishna Kumar and Jonathan Steed (DOI: 10.1039/c3cs60224a) provide a historical perspective and contemporary view of the use of supramolecular gels for the crystallisation of compounds, where exquisite control over crystal form can be achieved. The crystallisation of compounds in nanoscale confined spaces is intriguing, as the length scale corresponds to the critical size of nuclei. Qi Jiang and Michael Ward (DOI: 10.1039/c3cs60234f) present recent advances in the crystallisation of organic compounds in a variety of nanoporous materials that contain millions of nucleation sites in the channels of the hosts. Glimpses of the earliest stages of growth can be achieved and screening for polymorphs is feasible. If the pores are aligned, oriented crystal growth can also be achieved. On the other hand, the crystallisation of materials in microfluidic systems is an area that is gaining momentum. Josep Puigmartí-Luis (DOI: 10.1039/c3cs60372e) shows how the very precise control of fluid interphases can lead to exquisite and sometimes surprising control over the crystallisation of a wide variety of chemical systems and how the systems have been made into microscale parallel systems for screening conditions.

Flow is important in a number of areas in crystallisation, not least in polymers. Indeed, the crystallisation of polymers in general is a fascinating area that is beautifully demonstrated by Günter Reiter (DOI: 10.1039/c3cs60306g). The similarities and contrasts with small molecule crystallisation are shown and the benefits of using thin films for the study of crystal growth of polymers are made clear. The folding and unfolding of polymer chains can change their degree of metastability, and the reorganisation process is somewhat more facile during nucleation than in growth. Intriguing “inheritance of orientation” in re-grown crystals is quite unique to polymer systems. Gaetano Lamberti (DOI: 10.1039/c3cs60308c) shows how solidification can be monitored and reasons why extensional flow is far more effective than shear flow in polymer systems. The use of models to explain the results of flow induced crystallisation are shown to be valuable in a predictive sense, including giving information on the morphology of the material.

The preparation of thin films on perovskite oxide substrates by precise physical vapour deposition techniques can lead to epitaxial layers with atomic precision in thickness. Florencio Sánchez et al. (DOI: 10.1039/c3cs60434a) show that this method gives control over the chemical functionality at the newly formed surface, which is grown in an epitaxial manner ensuring a high degree of order. The control of these factors is crucial for the function of these interesting materials. Atomic force microscopy was shown to be important for the characterisation of the outer surface in that case, and scanning tunnelling microscopy (STM) is also very important for studying other epitaxial nanostructures. Matthew Marshall and Martin Castell, after discussing the workings of STM, show how it can probe at the atomic level and in real time the formation of crystalline areas on conducting surfaces (DOI: 10.1039/c3cs60458f). The resolution of these experiments provides precise information concerning the formation of nanocrystalline materials, and the spectroscopy facility also permits analysis of their physical characteristics.

The field of creating inorganic epitaxial or textured functional films and nanostructures has been traditionally occupied by physical vapour deposition techniques but in recent years progress in chemical solution deposition (CSD) has made this methodology very competitive in extensive applications owing to the much lower capital investment cost required and the ability to coat large areas or long lengths. Here understanding how the substrate facilitates the heterogeneous nucleation process as compared to homogeneous nucleation is essential to control the final microstructure and functionalities of these coatings and nanostructures. The use of a polymer-assisted deposition (PAD) method to prepare very different epitaxial inorganic films is described by McCleskey, Jia et al. (DOI: 10.1039/c3cs60285k). They show that the use of metal ions bonded to a polymer helps achieve a controlled heterogeneous nucleation and thus very high quality epitaxial films are easily obtained. They emphasize that this novel CSD approach has great potential because of the remarkably broad range of materials where it can be applied. The field of ferroelectric oxide films and nanostructures grown by CSD is reviewed by Nazanin Bassiri-Gharb et al. (DOI: 10.1039/c3cs60250h). This is a field where a relatively large tradition already exists and so the versatility of the technique to develop controlled microstructures in specific substrates has been widely investigated. In spite of this, no consensus exists yet about the mechanisms relating nucleation and growth with the final microstructure of the films, as discussed in depth. Concerning the fabrication of ferroelectric oxide nanostructures it is claimed that the field is still a young and developing research area and different bottom-up approaches towards achieving reliable nanostructures for electronic applications are described and their advantages and limitations emphasized. The great potential of using CSD as a low cost tool to grow epitaxial oxide nanostructures is discussed by one of our groups (Obradors, Puig et al., DOI: 10.1039/c3cs60365b). The principles governing self-assembly and self-organization of epitaxial nanostructures has been widely investigated in the past for in situ growth methodologies, mainly in the scope of semiconductor quantum dot investigation. However, these principles have still been less investigated in ex situ growth techniques such as CSD. The review shows that the strain engineering principles can indeed also be fully applied to obtain oxide nanostructures and the involved thermodynamic and kinetic features have been described. The potential of this bottom-up approach as a low cost growth technique to achieve wide area functional nanostructures is stressed.

Three reviews deal with the expanding field of crystallisation techniques involving light. Marek Grzelczak and Luis Liz-Marzán show how metal-based nanoparticles of varying shapes can be prepared through light-assisted synthesis (DOI: 10.1039/c3cs60256g). This effect can happen through both homogeneous and heterogeneous nucleation, for example on nanotubes. The contrast between continuous and pulsed UV light-assisted nucleation and growth of oxide films is presented in a tutorial review by Tomohiko Nakajima et al. (DOI: 10.1039/c3cs60222b). This technique allows the formation of crystalline systems at lower temperatures than the habitual one for them. Local heating under pulsed conditions is particularly beneficial for ordered growth from the film surface. On the other hand, high quality dense films can be prepared upon continuous irradiation. This technique is useful for the patterned formation of functional crystalline materials on substrates. Pulsed lasers can also be used for the nucleation and seeding of high quality protein crystals, as shown by Hiroshi Yoshikawa et al. (DOI: 10.1039/c3cs60226e). In solution, laser ablation at very specific points induces nucleation at low supersaturation by photomechanical and photochemical processes, yet with low damage to the sample. It is foreseen that this method may be used to speed up automated search methods for protein crystallisation and subsequent structure determination.

We thank most deeply our colleagues who have contributed to this themed issue of Chemical Society Reviews; it was an exciting project that we hoped would provide a collection that would stimulate the field and encourage synergies between the participants, and the authors have risen to this challenge with aplomb. For sure the readers of these articles will find as much interest and inspiration as we have.

Footnotes

† See, for example: http://www.rsc.org/chemistryworld/podcast/CIIEcompounds/transcripts/coppersulfate.asp and http://www.youtube.com/watch?v=9ihUJ2CRmYQ

‡ See, for example: http://prospect.rsc.org/blogs/cw/2010/12/16/chemistry-by-any-other-name/, http://www.youtube.com/watch?v=jvmW8OSvuK0 and http://www.youtube.com/watch?v=SRtxWYSL0Pw

§ See, for example: http://prospect.rsc.org/blogs/cw/2010/12/21/crystal-growing-christmas-trees/ and http://www.youtube.com/watch?v=vXtCzJT2CKo

---