Multicomponent nanocrystal superlattices represent an interesting class of material that derives emergent properties from mesoscale structure, yet their programmability can be limited by the alkyl-chain-based ligands decorating the surfaces of the constituent nanocrystals. Polymeric ligands offer distinct advantages, as they allow for more precise tuning of the effective size and ‘interaction softness' through changes to the polymer's molecular weight, chemical nature, architecture, persistence length and surrounding solvent. Here we show the formation of 10 different binary nanocrystal superlattices (BNSLs) with both two- and three-dimensional order through independent adjustment of the core size of spherical nanocrystals and the molecular weight of densely grafted polystyrene ligands. These polymer-brush-based ligands introduce new energetic contributions to the interparticle potential that stabilizes various BNSL phases across a range of length scales and interparticle spacings. Our study opens the door for nanocrystals to become modular elements in the design of functional particle brush solids with controlled nanoscale interfaces and mesostructures.

One of the key challenges facing liquid-phase transmission electron microscopy (TEM) of biological specimens has been the damaging effects of electron beam irradiation. The strongly ionizing electron beam is known to induce radiolysis of surrounding water molecules, leading to the formation of reactive radical species. In this study, we employ DNA-assembled Au nanoparticle superlattices (DNA-AuNP superlattices) as a model system to demonstrate that graphene and its derivatives can be used to mitigate electron beam-induced damage. We can image DNA-AuNP superlattices in their native saline environment when the liquid cell window material is graphene, but not when it is silicon nitride. In the latter case, initial dissociation of assembled AuNPs was followed by their random aggregation and etching. Using graphene-coated silicon nitride windows, we were able to replicate the observation of stable DNA-AuNP superlattices achieved with graphene liquid cells. We then carried out a correlative Raman spectroscopy and TEM study to compare the effect of electron beam irradiation on graphene with and without the presence of water and found that graphene reacts with the products of water radiolysis. We attribute the protective effect of graphene to its ability to efficiently scavenge reactive radical species, especially the hydroxyl radicals which are known to cause DNA strand breaks. We confirmed this by showing that stable DNA-AuNP assemblies can be imaged in silicon nitride liquid cells when graphene oxide and graphene quantum dots, which have also recently been reported as efficient radical scavengers, are added directly to the solution. We anticipate that our study will open up more opportunities for studying biological specimens using liquid-phase TEM with the use of graphene and its derivatives as biocompatible radical scavengers to alleviate the effects of radiation damage.

Graphene liquid cell electron microscopy has the necessary temporal and spatial resolution to enable the in situ observation of nanoscale dynamics in solution. However, the chemistry of the solution in the liquid cell during imaging is as yet poorly understood due to the generation of a complex mixture of radiolysis products by the electron beam. In this work, the etching trajectories of nanocrystals were used as a probe to determine the effect of the electron beam dose rate and preloaded etchant, FeCl 3 , on the chemistry of the liquid cell. Initially, illuminating the sample at a low electron beam dose rate generates hydrogen bubbles, providing a reservoir of sacrificial reductant. Increasing the electron beam dose rate leads to a constant etching rate that varies linearly with the electron beam dose rate. Comparing these results with the oxidation potentials of the species in solution, the electron beam likely controls the total concentration of oxidative species in solution and FeCl 3 likely controls the relative ratio of oxidative species, independently determining the etching rate and chemical potential of the reaction, respectively. Correlating these liquid cell etching results with the ex situ oxidative etching of gold nanocrystals using FeCl 3 provides further insight into the liquid cell chemistry while corroborating the liquid cell dynamics with ex situ synthetic behavior. This understanding of the chemistry in the liquid cell will allow researchers to better control the liquid cell electron microscopy environment, allowing new nanoscale materials science experiments to be conducted systematically in a reproducible manner.

Using in situ high-resolution TEM, we study the structure and dynamics of well-defined edge dislocations in imperfectly attached PbTe nanocrystals. We identify that attachment of PbTe nanocrystals on both {100} and {110} facets gives rise to b = a/2⟨110⟩ edge dislocations. Based on the Burgers vector of individual dislocations, we can identify the glide plane of the dislocations. We observe that defects in particles attached on {100} facets have glide planes that quickly intersect the surface, and HRTEM movies show that the defects follow the glide plane to the surface. For {110} attached particles, the glide plane is collinear with the attachment direction, which does not provide an easy path for the dislocation to reach the surface. Indeed, HRTEM movies of dislocations for {110} attached particles show that defect removal is much slower. Further, we observe conversion from pure edge dislocations in imperfectly attached particles to dislocations with mixed edge and screw character, which has important implications for crystal growth. Finally, we observe that dislocations initially closer to the surface have a higher speed of removal, consistent with the strong dislocation free surface attractive force. Our results provide important design rules for defect-free attachment of preformed nanocrystals into epitaxial assemblies.

Mechanisms of kinetically driven nanocrystal shape transformations were elucidated by monitoring single particle etching of gold nanocrystals using in situ graphene liquid cell transmission electron microscopy (TEM). By systematically changing the chemical potential of the oxidative etching and then quantifying the facets of the nanocrystals, nonequilibrium processes of atom removal could be deduced. Etching at sufficiently high oxidation potentials, both cube and rhombic dodecahedra (RDD)-shaped gold nanocrystals transform into kinetically stable tetrahexahedra (THH)-shaped particles. Whereas {100}-faceted cubes adopt an { hk0}-faceted THH intermediate where h/ k depends on chemical potential, {110}-faceted RDD adopt a {210}-faceted THH intermediate regardless of driving force. For cube reactions, Monte Carlo simulations show that removing 6-coordinate edge atoms immediately reveals 7-coordinate interior atoms. The rate at which these 6- and 7-coordinate atoms are etched is sensitive to the chemical potential, resulting in different THH facet structures with varying driving force. Conversely, when RDD are etched to THH, removal of 6-coordinate edge atoms reveals 6-coordinate interior atoms. Thus, changing the driving force for oxidation does not change the probability of edge atom versus interior atom removal, leading to a negligible effect on the kinetically stabilized intermediate shape. These fundamental insights, facilitated by single-particle liquid-phase TEM imaging, provide important atomic-scale mechanistic details regarding the role of kinetics and chemical driving force in dictating shape transformations at the nanometer length scale.

Graphene liquid cell electron microscopy provides the ability to observe nanoscale chemical transformations and dynamics as the reactions are occurring in liquid environments. This manuscript describes the process for making graphene liquid cells through the example of graphene liquid cell transmission electron microscopy (TEM) experiments of gold nanocrystal etching. The protocol for making graphene liquid cells involves coating gold, holey-carbon TEM grids with chemical vapor deposition graphene and then using those graphene-coated grids to encapsulate liquid between two graphene surfaces. These pockets of liquid, with the nanomaterial of interest, are imaged in the electron microscope to see the dynamics of the nanoscale process, in this case the oxidative etching of gold nanorods. By controlling the electron beam dose rate, which modulates the etching species in the liquid cell, the underlying mechanisms of how atoms are removed from nanocrystals to form different facets and shapes can be better understood. Graphene liquid cell TEM has the advantages of high spatial resolution, compatibility with traditional TEM holders, and low start-up costs for research groups. Current limitations include delicate sample preparation, lack of flow capability, and reliance on electron beam-generated radiolysis products to induce reactions. With further development and control, graphene liquid cell may become a ubiquitous technique in nanomaterials and biology, and is already being used to study mechanisms governing growth, etching, and self-assembly processes of nanomaterials in liquid on the single particle level.

Surface ligands impact the properties and chemistry of nanocrystals, but observing ligand binding locations and their effect on nanocrystal shape transformations is challenging. Using graphene liquid cell electron microscopy and the controllable, oxidative etching of gold nanocrystals, the effect of different ligands on nanocrystal etching can be tracked with nanometer spatial resolution. The chemical environment of liquids irradiated with high-energy electrons is complex and potentially harsh, yet it is possible to observe clear evidence for differential binding properties of specific ligands to the nanorods’ surface. Exchanging CTAB ligands for PEG-alkanethiol ligands causes the nanorods to etch at a different, constant rate while still maintaining their aspect ratio. Adding cysteine ligands that bind preferentially to nanorod tips induces etching predominantly on the sides of the rods. This etching at the sides leads to Rayleigh instabilities and eventually breaks apart the nanorod into two separate nanoparticles. The shape transformation is controlled by the interplay between atom removal and diffusion of surface atoms and ligands. These in situ observations are confirmed with ex situ colloidal etching reactions of gold nanorods in solution. The ability to monitor the effect of ligands on nanocrystal shape transformations will enable future in situ studies of nanocrystals surfaces and ligand binding positions.

Formation mechanisms of dendrite structures have been extensively explored theoretically, and many theoretical predictions have been validated for micro-or macroscale dendrites. However, it is challenging to determine whether classical dendrite growth theories are applicable at the nanoscale due to the lack of detailed information on the nanodendrite growth dynamics. Here, we study iron oxide nanodendrite formation using liquid cell transmission electron microscopy (TEM). We observe "seaweed"-like iron oxide nanodendrites growing predominantly in two dimensions on the membrane of a liquid cell. By tracking the trajectories of their morphology development with high spatial and temporal resolution, it is possible to explore the relationship between the tip curvature and growth rate, tip splitting mechanisms, and the effects of precursor diffusion and depletion on the morphology evolution. We show that the growth of iron oxide nanodendrites is remarkably consistent with the existing theoretical predictions on dendritic morphology evolution during growth, despite occurring at the nanoscale.