Nucleation is the first step in the formation of many materials; understanding its microscopic dynamics is crucial for improving synthesis of existing materials and predicting under what conditions novel materials will form. The simple picture of nucleation that prevailed for more than a century does not account for complex nucleation pathways observed in recent years in experiments and simulations. A more general framework is needed to explain reported phenomena; such a framework must account for the peaks and valleys in the free-energy landscape across which nucleation takes place and for the microscopic dynamic factors that dictate how a system explores this landscape. The articles of this issue illustrate and describe the many complex nucleation pathways seen across a range of material systems.

Crystallization is a key process in materials science, and most materials are made by processes that involve crystallization. Crystallization starts with nucleation, a process that is poorly understood for two reasons. First, nucleation occurs in contact with the typically uncharacterized surface of an impurity in the system. Second, we typically have little direct data on the microscopic mechanism of nucleation. We have a theory called classical nucleation, but when a simple application of the theory disagrees with experiment, it is unclear whether the theory is wrong, or if some feature of the surface is missing from the model. This article briefly reviews recent work on nucleation and its mechanisms. We are not alone in working with a stochastic process whose underlying mechanism is poorly understood. Engineers often have this problem and have developed powerful statistical models for stochastic processes. Surprisingly, even though they are sometimes used by materials scientists in different contexts, these are not used to model and predict nucleation behavior. We could advance the field with their use.

Colloidal systems offer ideal conditions to study the nucleation process, both from an experimental viewpoint, due to their relatively large size and long time scales, and from a modeling point of view, due to the tunability of their interactions. In this article, we review recent studies on the process of colloidal crystallization from a microscopic perspective. In particular, we focus on nonclassical pathways to nucleation, where the appearance of solid crystals involves fluctuations of two or more order parameters. Nonclassical behavior is interpreted as a decoupling of positional and orientational symmetry breaking. We then consider how the nucleation pathway determines which polymorph is selected upon nucleation from the melt. The study of nucleation pathways not only sheds new light on the microscopic mechanism of nucleation, but also provides important information regarding its avoidance, suggesting a deep link between crystallization and vitrification.

Conformational changes, and the formation of densely packed ordered aggregates or crystals, are behaviors that profoundly affect the properties of a molecule. Using the example of biological macromolecules, we discuss two types of interactions between these two behaviors. First, we demonstrate that shape change may be driven by crystallization if the gain in crystallization free energy is sufficient to overcome the transition to an unfavorable molecular conformation. Hence, the crystal structures of flexible molecules may be a poor representation of their free-phase atomic arrangements. Second, molecules with conformational variability, such as proteins, may facilitate the nucleation of their crystals by forming dense liquid clusters enriched in domain-swapped or misassembled oligomers. In the clusters, the nucleation barrier is reduced due to the lower surface free energy of the crystal/dense liquid interface, and nucleation is significantly faster.

Nanoparticles (NPs) have emerged as new functional blocks for optical, energy, and biomedical applications, opening a new frontier of rational self-assembly of materials. One of the most controllable assembly strategies relies on programming interparticle interactions using the complementarity of DNA strands, providing selective and reversible interactions between particles of different sizes and shapes. Much progress has been achieved in DNA-guided assembly of particle superlattices. By tuning the interactions, sizes, and shapes of NPs, a wide variety of structures have been assembled. This article discusses the most significant achievements and challenges in assembly of DNA-programmable particle superlattices.

A review of recent literature supports the notion that organisms may achieve nanoscale control over the hierarchical assembly of organic–inorganic materials by compartmentalizing reactions into small volumes containing specifically functionalized macromolecules. Such confinement may introduce a degree of determinism to the stochastic process of nucleation by greatly reducing the number of nucleation events, allowing an organism to control nucleation using “soft” organic substrates. In this way, the polymorph, orientation, shape, and size of a crystalline building block can be selected, and its assembly into a larger structure orchestrated by the organic matrix.

Research aimed at designing and optimizing open framework materials for commercial applications tend to focus on two critical objectives: identifying synthesis conditions that yield crystals with tailored physicochemical properties, and unlocking the untapped design space to achieve theoretical structures that far outnumber the list of synthetically realized materials. Accomplishing these goals requires detailed knowledge of nucleation in order to cultivate efficient, facile, and economical methods of controlling crystallization. The vast number of open framework materials that can be engineered through the judicious selection of inorganic or organic building units hold the promise for future discovery of materials with unique and superior properties compared to available porous materials. Herein, we review what is known about the nucleation of open framework crystals, highlighting the voids in our understanding of nucleation pathways, and we offer guidelines for advancing crystal engineering in this exciting area of research.

Universal access to abundant scientific data, and the software to analyze the data at scale, could fundamentally transform the field of materials science. Today, the materials community faces serious challenges to bringing about this data-accelerated research paradigm, including diversity of research areas within materials, lack of data standards, and missing incentives for sharing, among others. Nonetheless, the landscape is rapidly changing in ways that should benefit the entire materials research enterprise. We provide an overview of the current state of the materials data and informatics landscape, highlighting a few selected efforts that make more data freely available and useful to materials researchers.