Exploring In Situ Liquid Ambient Transmission Electron Microscopy Techniques for the First Time

I. Why study liquid ambient transmission electron microscopy?

Most liquids, including water and other organic solvents, have large saturated vapor pressure and cannot exist in the high vacuum environment of transmission electron microscope, therefore, when studying the behavior of nanomaterials in liquid environment, it is necessary to construct a liquid storage cell to isolate the liquid from the high vacuum environment in the electron microscope, which requires the use of Liquid cell TEM.Liquid cell TEM actually Liquid cell TEM is to make a liquid cell by micro-nanofabrication, and then fix it on the head of a common sample rod or a special liquid sample rod and put it into the electron microscope for observation.

Figure 1: Liquid Cell Structure Schematic

II. History of the development of in situ liquid transmission electron microscopy technology

In-situ Liquid cell TEM can be traced back to 1934, when Morton, from the Free University of Brussels, Belgium, made the first attempt of transmission electron microscopy of living biological samples by wrapping the samples with two pieces of aluminum foils, but due to the thickness of the aluminum foils and the liquid layer, the resolution of the liquid cell was only on the order of micrometers.

In recent years, thanks to the advances in micro and nanofabrication technology and microfluidic technology, the preparation of Liquid cell has made a breakthrough, and the in-situ electrochemical Liquid cell chip designed and fabricated by F. M. Ross in 2003 is a milestone in the preparation of Liquid cell in recent times. Its structure is shown in Fig. 2, the bottom silicon wafer deposits a layer of polycrystalline gold electrodes, which are glued with the top silicon wafer through SiO2 ring spacers to form an electrochemical reactor, and the top silicon wafer has two containers, which lead to two electrodes for applying electric bias voltage respectively. When used, the liquid is injected and flows into the observation window through capillary action, and then the Liquid cell is sealed and placed into the electron microscope for observation. Since the imaging electron beam needs to pass through a 100 nm silicon nitride film window, and the spatial resolution of the liquid layer close to 1 μm is only 5 nm, this kind of chip, which forms a liquid chamber between the two layers of silicon wafers, and adopts a silicon nitride film as the observation window, is a prototype for the development of many subsequent improvements of the Liquid cell.

Figure 2: (A). Schematic diagram of Liquid cell, (B) optical photograph of two-electrode Liquid cell (Rosset al., Nat. Mater., 2003, 532)

Currently there are two main types of Liquid cell fabrication, one is closed cell and the other is flow cell which contains liquid flow pipes (see Figure 3).

Figure 3: A. Schematic of the three-dimensional structure of the closed cell, B. Schematic of the structure of the cross-section along the horizontal line in Figure A (Zhenget al., Science, 2009, 1309) C. Schematic of the structure of the flow cell (de JongeN et al., PNAS, 2009, 106).

In 2009, Zheng Haimei reported an ultra-thin silicon nitride window Liquid cell as in Fig. 3A&B, whose thickness of silicon nitride film is only 25 nm, and the upper and lower layers of the chip are bonded with ultra-thin indium spacers to form a Liquid cell chamber, and the thickness of the liquid layer in the observation window is about 200 nm. based on this, in 2014, Liao et al. improved the ultra-thin silicon nitride window Liquid cell technique, further reducing the thin silicon nitride film degree to 13 nm and the liquid layer thickness to about 100 nm, effectively increasing the spatial resolution to the atomic level.

In 2009 Neils de Jonge et al. designed the open Liquid cell, as in Fig. 3C, to observe individual molecules in intact cells in situ without the need for freezing and drying. The thickness of its liquid layer is about 7 μm and the spatial resolution can reach 4 nm.

In addition to the use of silicon nitride film as an observation window, in 2012, Jong Min Yuk first proposed the use of graphene film to prepare a liquid cell and studied the growth process of palladium nanocrystals in-situ, as shown in Fig. 4. The use of graphene as an observation window material can effectively reduce or even ignore the electron scattering and thus realize the atomic-level resolution. Subsequently, a variety of complex graphene Liquid cell structures have been derived by utilizing graphene as an electron beam transmission window. In particular, in 2014, JongMin Yuk observed the anisotropic lithiation process on the surface of silicon nanoparticles using Liquid cell, which made it possible to use graphene Liquid cell for electrochemical studies. However, the use of graphene Liquid cell to study electrochemical processes is still very limited due to the fact that graphene films are very thin and it is difficult to place conventional electrochemical electrodes.

Figure 4: Graphene Liquid cell schematic (Li et al., Science 2010, 330).

Liquid cell TEM can not only be used to observe the behavior of nanomaterials in liquid environments in situ, but also integrate heating and freezing elements on Liquid cell chips and liquid rods for functional testing of nanomaterials, which greatly broadens the research scope of transmission electron microscopy. For example, Kai-Yang Niu et al. from Haimei Zheng's group used a heatable Liquid cell to study in situ the formation process of bismuth oxide hollow nanoparticles under the action of Kirkendall.K.Tai used a freezing platform to study the phase transition in ice during crystallization and the dynamic interactions between the surface and gold particles before crystallization.

Figure 5: A.Hollownanoparticle growth dynamics via Kirkendall effect （Paul Alivisatoset al., Nano Lett,2013,13）B.The dynamic interactions of Aunanoparticles at the ice crystallization front （Dillon et al.,Microsc. Microanal, 2014, 330）).

To summarize, the current Liquid cell chip is mostly based on silicon substrate processing, the window material is generally used ultra-thin silicon nitride film, Haimei Zheng's group can be silicon nitride film to 13nm or so, other groups and commercial Liquid cell window material is generally to 30nm or so, the size of the window is 50*50 μm. the resolution can be up to Atomic level, close to the intrinsic resolution of the electron microscope. The resolution can reach atomic level, close to the intrinsic resolution of the electron microscope, and can be integrated with heating and freezing functions, but it is not easy to realize the high requirements on liquid cell stability.

III. Application of in situ liquid transmission electron microscopy technology

The process of nanoparticle nucleation and growth can be observed using In-situ Liquid cell TEM to experimentally demonstrate issues that have been controversial, such as whether the dominant mechanism during nanoparticle liquid-phase growth is monomer attachment or particle fusion.

Figure 6: Video images showing simple growth by means of monomer addition （left column） or growth by means of coalescence （right column). （Zheng et al., Science, 2009, 1309）).

Heterogeneous nanocrystal growth processes can be studied

Figure 7: Comparison of Pdgrowth on 5 and 15 nm Au seeds. （a, d）Starting dark-field STEM images of a 5 nm（a） and a 15 nm （c） Au nanoparticles in 10 μM aqueous PdCl2 solution （samescale）; （b,e） The same two particles after Pd deposition （84 s total beamexposure）;（c, f） Schematic illustration of the Pd growth morphology for thetwo sizes of Au seed nanoparticles （E. A. Sutter et al., Nano Lett, 2013, 13）.).

Nanoparticle self-assembly processes can be studied

Figure 8: TEM images of NPassembly formed under electron beam irradiation （a,b） and drop casting （c,d） onSiNx TEM grid. The scale bar is 100 nm （Jungwon Park et al., ACS NANO, 2012, 6）.).

The lithiation process of lithium-ion batteries can be investigated.Huang et al. investigated in situ the swelling, elongation and helical behaviour of ZnO nanowires during the lithiation process of lithium-ion batteries in an open Liquid cell.

Figure 9: Schematic of the experimental setup（Li et al.,Science 2010,330）

It can also be used to observe some biological samples.

Figure 10: Image of the edge of a fixed COS7 cell after 5-min incubation with EGF-Au（de Jonge N et al., PNAS, 2009, 106）

Of course, the research content of Liquid cell TEM is not only limited to these, interested parties can read Hong gang Liao's 2016 review article Liquid Cell Transmission Electron Microscopy published in Annu. Rev. Phys. Chem.

Seeing this, one may ask how to exclude the effect of electron beam on the reaction process during the research. Electron beams are indeed a love-hate situation, as you need to use them for imaging, but you don't want them to interact with the materials under study to affect the results of your experiments. However, don't worry, Liquid cell TEM Ross has provided you with the theoretical basis for quantifying the effect of electron beam! Speaking of which, I can't help but marvel that Ross is a bull with deep academic attainments and a willingness to share. I had the honour to ask Ross for advice at a conference, and she was very nice to encourage my broken English and immature ideas, and patiently explained to me, which is the kind of idol we need when we are just starting out in the research field.