See Wee Chee

Real-Time Imaging of the Formation of Au–Ag Core–Shell Nanoparticles

We study the overgrowth process of silver-on-gold nanocubes in dilute, aqueous silver nitrate solution in the presence of a reducing agent, ascorbic acid, using in situ liquid-cell electron microscopy. Au-Ag core-shell nanostructures were formed via two mechanistic pathways: (1) nuclei coalescence, where the Ag nanoparticles absorbed onto the Au nanocubes, and (2) monomer attachment, where the Ag atoms epitaxially deposited onto the Au nanocubes. Both pathways lead to the same Au-Ag core-shell nanostructures. Analysis of the Ag deposition rate reveals the growth modes of this process and shows that this reaction is chemically mediated by the reducing agent.

Linker-Mediated Self-Assembly Dynamics of Charged Nanoparticles

Using in situ liquid cell transmission electron microscopy (TEM), we visualized a stepwise self-assembly of surfactant-coated and hydrated gold nanoparticles (NPs) into linear chains or branched networks. The NP binding is facilitated by linker molecules, ethylenediammonium, which form hydrogen bonds with surfactant molecules of neighboring NPs. The observed spacing between bound neighboring NPs, ∼15 Å, matches the combined length of two surfactants and one linker molecule. Molecular dynamics simulations reveal that for lower concentrations of linkers, NPs with charged surfactants cannot be fully neutralized by strongly binding divalent linkers, so that NPs carry higher effective charges and tend to form chains, due to poor screening. The highly polar NP surfaces polarize and partly immobilize nearby water molecules, which promotes NPs binding. The presented experimental and theoretical approach allows for detail observation and explanation of self-assembly processes in colloidal nanosystems.

Direct observation of the nanoscale Kirkendall effect during galvanic replacement reactions

Galvanic replacement (GR) is a simple and widely used approach to synthesize hollow nanostructures for applications in catalysis, plasmonics, and biomedical research. The reaction is driven by the difference in electrochemical potential between two metals in a solution. However, transient stages of this reaction are not fully understood. Here, we show using liquid cell transmission electron microscopy that silver (Ag) nanocubes become hollow via the nucleation, growth, and coalescence of voids inside the nanocubes, as they undergo GR with gold (Au) ions at different temperatures. These direct in situ observations indicate that void formation due to the nanoscale Kirkendall effect occurs in conjunction with GR. Although this mechanism has been suggested before, it has not been verified experimentally until now. These experiments can inform future strategies for deriving such nanostructures by providing insights into the structural transformations as a function of Au ion concentration, oxidation state of Au, and temperature.Hollow nanoparticles can be synthesized by galvanic replacement or the Kirkendall effect, which are generally regarded as two separate processes. Here, the authors use liquid TEM to follow the entire galvanic replacement of Ag nanocubes, finding experimental evidence that the Kirkendall effect is a key intermediate stage during hollowing.

Hopping Diffusion of Gold Nanoparticles Observed with Liquid Cell TEM

The diffusion of nanoparticles in the microfluidic cells used for liquid cell transmission electron microscopy (TEM) have always been found to be much slower [1-6], often by several orders of magnitude, when compared with bulk diffusion. While this highly suppressed motion is serendipitous for the atomic resolution imaging of nanoparticle nucleation and coalescence events, we still lack a compelling explanation for this anomalous phenomena. Here, we report results from our experiments tracking the motion of Au nanoparticles in water, using a combination of energy filtered imaging and image acquisition at frame rates of 100 Hz.

Probing Nanoparticle Dynamics in 200 nm Thick Liquid Layers at Millisecond Time Resolution

When imaging specimens in liquids using liquid cell TEM, there is a trade-off between spatial resolution and liquid layer thickness [1]. In TEM mode, the imaging resolution in a liquid cell with 1 micron thick liquid degrades to about 12 nm for objects near the top SiN window, due to scattering of electrons by the liquid [1]. While higher resolution can be obtained with thin liquid films (less than 100 nm thick), the desire for sharper images has to be balanced against keeping the liquid layer thick enough such that processes observed are representative of bulk systems. For example, nanoparticle diffusion observed in thin liquid layers with liquid cell TEM had been found to deviate from the Stokes-Einstein relation [2]. Maintaining a thick liquid layer is also crucial for imaging biological specimens, where, so far, in situ imaging is mainly accomplished using STEM [3]. However, STEM imaging cannot match the temporal resolution of TEM imaging for capturing the motion of dynamic objects where state-of-the-art direct detection cameras are capable of recording whole frames in milliseconds.

What Can We Learn from the Shapes of Secondary Electron Puddles on Direct Electron Detectors

While single event detection based imaging, referred to as electron counting, is starting to be widely adopted, several challenges still remain. One of these is, accurately identifying the entry point of the primary electron given the secondary electron puddle produced by it. Succeeding here reduces imaging noise, improves detector efficiency and allows for super-pixel accurate reconstruction [1]. It has been shown that these secondary electron puddles can vary greatly in shape and size [2]. While several techniques have been proposed [2], there is no consensus on which of these is optimal and why they would be better. Here, we investigate the shape and size of these secondary electron puddles to better understand how electron counting can be improved.

Capturing Dynamics in Liquids with High-Speed CMOS Cameras - Opportunities and Challenges

The development of liquid cell transmission electron microscopy (TEM) has allowed us to capture a range of phenomena in liquid samples, such as nanoparticle dynamics, electrochemical reactions and the biological structures within the microscope at high spatial resolution [1]. In principle, these studies should benefit from the capabilities of direct electron detection CMOS cameras, with their higher frame rates and higher sensitivity to electrons. State-of-the art cameras are capable of frame rates up to more than 1000 frames per second (sub-millisecond time resolution). For example, they enabled the capture of Pt nanoparticle nucleation events with atomic resolution at 400 frames per second [2].