Articles in 2023

The fundamental discoveries that enabled the development of effective mRNA vaccines against COVID-19 were awarded with this year’s Nobel Prize in Medicine.

Post-infusion poly(ethylene glycol) surface modification of chimeric antigen receptor (CAR)-engineered T cells and a subcutaneous chemokine-adsorbing hydrogel address cytokine release syndrome and the neurotoxicity side effects of CAR T cell therapy against tumours.

Piezoelectrics have longitudinal and transverse piezoelectric coefficients that are opposite in sign. Here, by tuning the interface inversion asymmetry in heterostructures, auxetic systems with positive longitudinal and transverse coefficients are realized, with expansion or contraction along all directions in an electric field.

The authors present magnetotransport measurements to demonstrate multistep magnetization switching in orthogonally twisted CrSBr ferromagnetic monolayers.

Polymers made by click chemistry with spirocyclic building blocks form membranes that separate the components of crude oil based on molecular size and type, potentially using far less energy than distillation. Key enablers of this separation are moderate levels of polymer dynamic motion and frustrated chain packing.

A compact, time- and energy-efficient computing architecture — based on ferroelectric-defined reconfigurable two-dimensional photodiode arrays — is shown to be capable of in-memory sensing and computing.

Hybridized electron or hole states across semiconducting van der Waals monolayers in heterotrilayer systems enable the emergence of quadrupolar excitons. Quadrupolar excitons, unlike their dipolar counterparts, have a tunable static dipole moment that responds nonlinearly under an applied electric field.

The discovery of passivating agents for perovskite photovoltaics can be an arduous and time-consuming process. Now, a machine-learning model is reported that accelerates the selection of bifunctional pseudo-halide passivators. The identified pseudo-halide passivators were experimentally shown to enhance the performance of perovskite solar cells.

Heat treatment can transform some moiré superlattices into fully commensurate bilayers, where atoms in opposite layers align perfectly with each other. This structural transformation gives rise to markedly brighter interlayer excitons.

By inserting an epitaxial in-plane buffer layer of Bi₅FeTi₃O₁₅, an artificial flux closure architecture enables ferroelectric polarization from a single unit cell of BaTiO₃ or BiFeO₃.

Three protein interaction surfaces are computationally designed into one protein subunit to enable their accurate assembly into three-dimensional crystals with user-specified lattice architectures.

The monolithic 3D integration of wafer-free all-2D-materials-based electronics can produce an AI processor.

Monolithic 3D integration of electronics based on fully 2D materials is demonstrated in the performance of artificial intelligence tasks.

Optically stimulated vibrational control for materials has the potential to improve the performance of optoelectronic devices. The vibrational control of FAPbBr₃ perovskite solar cells has been demonstrated, where the fast dynamics of coupling between cations and inorganic sublattice may suppress non-radiative recombinations in perovskites, leading to reduced voltage losses.

Early detection of electrical degradation in dielectric polymers is crucial but remains challenging. A general strategy of blending the polymer with chromogenic molecules is reported, which generates a visually discernible colour change as chemically activated by oxygen radicals generated in situ, indicating the early stage of electrical degradation in polymers.

A traditional physical-reservoir device has limited flexibility and cannot perform well across a range of computing tasks, owing to the fixed reservoir properties of the physical system. However, by exploiting the rich magnetic phase spaces of a single chiral magnet, reservoir properties can be reconfigured. This control enables on-demand optimization of computational performance across diverse machine-learning tasks.

Current physical neuromorphic computing faces critical challenges of how to reconfigure key physical dynamics of a system to adapt computational performance to match a diverse range of tasks. Here the authors present a task-adaptive approach to physical neuromorphic computing based on on-demand control of computing performance using various magnetic phases of chiral magnets.

Chemical adsorption of CO on open metal sites enables separation from other gases but leads to selectivity and stability issues. Quasi-open metal sites in metal–organic frameworks are proposed here, which are accessible only by CO-induced structural transformation, enabling CO separation to 9N purity.