C.H. Ray, B.R. Ilic, R. Sharma, G. Holland, V. Aksyuk, S.M. Stavis, J.A. Liddle
National Institute of Standards and Technology,
Keywords: nanofluidics, TEM, in situ imaging
Summary:Breakthroughs in material growth and dissolution, electrochemistry, biomineralization, and the study of soft materials are being enabled by the availability of enclosed liquid cells for the transmission electron microscope (TEM). There are a variety of designs for TEM liquid cells, which determine their functionality and fluidic interface. Cells with liquid thicknesses of order 1 µm permit liquid flow, while cells with liquid thicknesses of order 100 nm allow for high-resolution imaging, but have demonstrated little to no control of flow. We recently developed a monolithic liquid cell that maintains a constant thickness of liquid of approximately 100 nm across a viewing area of 200 µm by 200 µm, and enables high-resolution imaging and spectroscopy. Integrating a nanofluidic system to control flow through our cell would dramatically improve its utility by allowing the initiation of chemical reactions at predetermined times during observation and the removal of confounding radiolysis products. However, nanofluidic liquid cells using the typical, pressure-driven approaches to pump fluids require prohibitively high pressures, and, because of the concomitant low flow rates, would suffer from very slow exchange of fluids through macroscopic capillaries. In addition, the use of fluid lines external to the TEM can introduce unwanted vibrations. In order to solve these problems and enable future integration of lab-on-a-chip analysis with the TEM, we have developed an integrated electrokinetic pump. This enables flows to be driven through even nanoscale channels. The introduction of electrodes that must be in contact with the fluid presents process integration challenges. The most significant of these are the requirement that the metallization must be able to withstand a high-temperature (850 ºC) SiNx deposition step, and that the sacrificial layer used to define the fluidic channels be removable without adversely affecting any of the other materials used in the device. We resolve both of these challenges by using Cr2O3 as an encapsulant and sacrificial material. This allows us to construct a modular process flow and device design that can be adapted to enable a variety of in situ measurements.