Guest–host systems offer potential for various applications like gas storage and separation, sensor design, catalysis and (electro)chemical energy conversion and storage. The nanoporous host confines the adsorbed guest phase, thus enforcing interactions between the mobile guest components and the host surfaces at the interfaces. These guest–host interactions crucially influence macroscopic properties like sorption, catalytic reactivity, mass, and charge transport, and they are vital to developing next-generation materials. This chapter highlights essential observations and strategies to analyse and understand how guest–host interactions convert into macroscopic properties. This is challenging, as the underlying processes span nano- to micrometre length and picosecond to second time scales. NMR crystallographic strategies are unique for this task, as they allow for probing of length- and time-scale-dependent processes. While diffraction experiments are sensitive to the long-range structure, NMR spectroscopy provides element-selective information about local structural elements of the guest and host and probe connectivities. Additionally, NMR spectroscopic techniques offer access to the local dynamics and the long-range guest transport. Combined with DFT and MD simulations, the information about structure, dynamics and transport can be linked to the guest–host interactions.
A historical perspective on the development of NMR crystallography is presented. The concept of crystallography and the role of NMR in generating structural and crystallographic information on solids are discussed. A brief survey of topics in modern NMR crystallography covers areas such as experimental and computational methodologies and applications to a range of materials.
Nuclear magnetic shielding and spin–spin coupling constants are the fundamental parameters that can be extracted from nuclear magnetic resonance experiments. These parameters contain structural information and can be used to deduce the structure of unknown molecules. The traditional approach based on empirical correlations between structure and spectral parameters may be of insufficient accuracy for a unique assignment of the molecular structure. An alternative approach is to calculate the NMR parameters from first principles for possible structural candidates. The accuracy of the calculated parameters depends on several components, of which the basis set used for expanding the orbitals is one ingredient. The present chapter reviews how basis sets can be constructed to allow a systematic reduction of basis set incompleteness, with specific focus on basis sets for calculating NMR parameters.
In most NMR crystallography applications experimental techniques are used to build an appropriate structural model, which can be later refined using quantum-chemical calculations. In some cases, this can be viewed as an obstacle, in particular when structural constraints extracted from the experimental data are ambiguous or not abundant enough. One of the most promising solutions to this problem is crystal structure prediction (CSP). On the other hand, for complicated, flexible and/or multicomponent systems the number of degrees of freedom (DOF) which need to be accounted for in CSP starts to be overwhelming, thus limiting the applicability of this computational method. In such instances, structural constraints extracted from solid-state NMR spectra can help to reduce this vast number to a perfectly manageable number of DOFs, making a combination of NMR crystallography and CSP calculations a very powerful approach. This chapter focuses on the applicability of CSP in the context of NMR crystallography, including a brief overview of modern CSP approaches, together with their advantages and limitations.
This chapter will look at situations where diffraction methods do not provide a complete description of all atomic positions in a crystal. For example, because hydrogen atoms diffract poorly, their locations are often better determined via NMR.
In this chapter, we discuss recent applications of NMR crystallography in bulk organic molecular solids enabled by the sensitivity enhancement offered by MAS DNP. We draw upon examples from recent literature to illustrate how MAS DNP has significantly expanded the scope of NMR crystallography of organic molecular solids at natural isotopic abundance by clearing several longstanding bottlenecks caused by the low sensitivity of NMR. Specifically, for samples at natural isotopic abundance, we discuss how MAS DNP enabled: the detection of low-receptivity nuclei within the bulk of organic solids; the observation of spectral correlations between sparsely distributed spins; the investigation of the morphology of multi-domain solid particles on the nm- to µm-scale; the detection and characterization of structural changes in samples that evolve over time. For each of these advancements, we present and discuss a curated selection of applications where the contribution of DNP proved to be essential.
This chapter describes the use of NMR crystallography for the study of disordered inorganic solids. After briefly discussing the different types of disorder that are encountered in these materials and approaches for modelling disorder, recent advances in the application to systems including minerals, biomaterials, ceramics, energy materials and porous solids are presented. The additional challenges posed by systems exhibiting dynamic disorder are also discussed before the future outlook in this area is considered.
This chapter reviews the application of the polarization propagator/response methods to the ab initio calculation of the indirect NMR spin–spin coupling constants. Starting with the first applications in the 1970s, it describes the journey to the present state-of-the-art. It shows that all four contributions to the spin–spin coupling constant are polarization propagators and how these terms may be calculated directly without invoking the sum over states. Various approximations to the polarization propagator such as TDA, RPA, SOPPA, MCRPA, SOPPA(CCSD), CC2, CC3 and CCSD are described. The computational issues for these methods are discussed as are the trends in the numerical results. A discussion of the strengths and future challenges of the field ends the presentation. It is concluded that we have come a long way but that there is still more to be done.
This chapter provides an overview of the progress in solid-state nuclear magnetic resonance (SSNMR) research on investigating non-covalent interactions in molecular crystals. All relevant interactions are examined: from common strong hydrogen bonds (e.g., O–H⋯O, N–H⋯N, O–H⋯N, or O–H⋯N), to the halogen bond, to the recently rediscovered tetrel, pnictogen, chalcogen, and osme bonds, ending with other weak interactions including π⋯π interactions and weaker hydrogen bonds (e.g., C–H⋯N, C–H⋯O, C–H⋯F, and C–H⋯π). The emphasis is mainly, though not exclusively, on molecular organic, pharmaceutical, and inorganic solids. It aims to offer valuable insights to both experienced and novice researchers by addressing challenges, unresolved issues, technique solutions, and limitations. This chapter also reports on significant applications of SSNMR methods, covering various parameters and studied nuclei relevant to characterize weak interactions. Additionally, computational techniques, increasingly applied and fundamental in NMR crystallography studies, are also discussed.
Assessing the accuracy of predictions of NMR parameters and understanding the limitations of the computational methods are crucial aspects of NMR crystallography research. This chapter focuses on computational methods that go beyond the complexity of the most commonly used DFT approximations and static calculations, which typically ignore molecular dynamics in the solid state. Particular attention is given to a number of effects, including fast molecular motion, vibrational motion, and nuclear quantum effects (nuclear delocalization and tunneling), on solid-state NMR parameters.
Methods of computation of nuclear magnetic resonance (NMR) observables from first principles in extended solids are discussed. The emphasis is on an overview of the underlying theories, highlighting their similarities, differences, and limitations, and focusing on the two most commonly studied observables, the chemical shielding and nuclear quadrupole interaction. Methods of how the resulting computations may be interpreted are discussed as well. Finally, extensions of these approaches to other observables are overviewed, as well as extensions to the basic theories. It is hoped that this chapter will form a bridge for those working on solid state NMR, between the basic task of learning to run the different available codes, and the highly technical literature where the underlying theories and their implementation are described in detail.
NMR calculations with relativistic four-component Dirac–Kohn–Sham models are computationally demanding. Therefore, scalar-relativistic one-component and spin–orbit two-component approaches were developed for applications to larger molecules. In particular, the development of the zeroth-order regular approximation (ZORA) in the 1990s represents a cornerstone and is still widely used in computational chemistry. Today, ZORA is superseded by the formally more accurate exact two-component (X2C) theory. In X2C, the positive- and negative-energy states of the four-component one-electron Hamiltonian are decoupled by a straightforward diagonalization and derivatives are formed by solving one-electron response equations. This way, the X2C approach is much more efficient than its parent four-component ansatz. However, further approximations are still desirable to achieve optimal performance. This is of particular interest for highly efficient density functional treatments using density-fitting and seminumerical integration techniques to calculate the NMR spectra of molecules with more than 100 atoms on low-cost computer hardware in a few minutes or hours. Here, local relativistic approximations are pivotal, and this chapter reviews recent endeavours for NMR coupling constants and shifts with local X2C. The accuracy of the introduced approximations is assessed in detail, showing that the respective errors are insignificant, and illustrative applications to extended organometallic compounds are presented.
Machine learning is becoming increasingly important in the prediction of nuclear magnetic resonance (NMR) chemical shifts and other observable properties. This chapter provides an introduction to the construction of machine learning (ML) models for predicting NMR properties, including the discussion of feature engineering, common ML model types, Δ-ML and transfer learning, and the curation of training and testing data. Then it discusses a number of recent examples of ML models for predicting chemical shifts and spin–spin coupling constants in organic and inorganic species. These examples highlight how the decisions made in constructing the ML model impact its performance, discuss strategies for achieving more accurate ML models, and present some representative case studies showing how ML is transforming the way NMR crystallography is performed.
Nuclear Magnetic Resonance (NMR) spectroscopy is a powerful analytical technique widely used in chemical and biological research to elucidate molecular structure and dynamics. In this chapter, we explore the intricate relationship between intra- and inter-molecular hydrogen bonding (HB) and NMR spectroscopy parameters. Through a combination of experimental and modern quantum computational approaches for some case examples, we investigate how HB can be a pathway for the transmission of Spin–Spin Coupling Constants (SSCC) and can influence the chemical shift in NMR spectra. Specifically, we examine the effects of HB on the transmission of a hetero-nuclear (JFH) coupling constant in solution-state NMR. By understanding the role of hydrogen bonds in shaping NMR spectra, researchers can increase the accuracy and sensitivity of NMR-based analyses and gain deeper insights into the behavior of complex molecular systems. This chapter provides an overview of the current understanding of the effects of HB on NMR spectra.
This chapter is concerned with the analysis of calculated NMR parameters aimed at understanding how molecular structure and bonding leads to observed trends. The quantum-theoretical approach to the calculation of nuclear magnetic shielding, indirect nuclear spin–spin coupling, and electric field gradients at quadrupolar nuclei is outlined, followed by a general discussion of different ways by which to analyze the calculated results in terms of orbital or spatial contributions. A number of examples are provided to illustrate the applications of NMR parameter analyses in a broad range of research areas.
Metal halide perovskite and perovskite-inspired materials could lead to the next generation of cost-effective semiconducting materials for various optical and electrical applications. This chapter describes the characterization advances of low- and high-dimensional metal halide perovskite materials made possible using solid-state nuclear magnetic resonance (NMR) spectroscopy, specifically, that of quadrupolar nuclei. Furthermore, advantages are discussed when using the complementary nuclear quadrupole resonance (NQR) spectroscopy technique when faced with large quadrupole moments. The chapter is organized by defining common NMR interactions for quadrupolar nuclei, common structure types of metal halide perovskite compounds of interest, 2H/14N NMR for dynamics, alkali and halogen NMR probe nuclei, and the exploration of more exotic nuclei often attributed to the B-site of these materials.
Paramagnetic NMR measures the chemical shifts of molecules in the presence of a paramagnetic center. In this chapter, we present equations to model the part of the chemical shifts due to the presence of the paramagnetic center. We pay particular attention to cases where the orbital contribution is important, either only partially quenched or due to strong spin–orbit coupling. This is illustrated with modeling using the quantum chemistry of pNMR chemical shifts in transition metal, lanthanide and actinide complexes.
A nearly universal component of NMR crystallography is the ranking of candidate structures based on how well their first-principles predicted NMR parameters align with the results of solid-state NMR experiments. This chapter reviews the statistical concepts underpinning the structure selection process and presents methodologies for assigning relative probabilities to competing models. An emphasis is placed on robust approaches grounded in classical parametric statistics, supplemented by Bayesian probability analysis. To bridge theory and practice, Monte Carlo simulations are used to illustrate important concepts and their practical application to the structure selection problem in NMR crystallography.
Protein structure determination at the atomic scale is of utmost importance and serves as the cornerstone for understanding biological processes at the molecular level. Three-dimensional structures, preferably in conjunction with protein dynamics, report on protein function and can guide the development of inhibitors and possibly of new pharmaceutical drugs and vaccines. The protein data bank (PDB) holds over 200 000 structures determined mainly by X-ray crystallography, nuclear magnetic resonance (NMR), and cryo-electron microscopy (cryo-EM). In this chapter we focus on the use of magic-angle spinning solid-state NMR for protein structure determination at atomic-resolution. We discuss sample preparation techniques, NMR methodologies and structure calculation techniques, and provide typical examples for structures including early studies of crystalline proteins, as well as examples from metalloproteins, amyloids, membrane proteins, and DNA/RNA binding proteins. All the structures discussed are available in the protein data bank.
