Reports: ND1050630-ND10: Carbon Nanoelectrodes: Generating Hydrogen from Water

B. C. Regan, PhD, University of California (Los Angeles)

Our project is focused on using transmission electron microscopy (TEM) to achieve a better fundamental understanding of photo- and electro-chemistry inside the Debye layer. Progress in this direction will lead to more efficient solar cells, higher capacity batteries, cleaner cars and a healthier economy. This investigation is timely because of a technological innovation, pioneered by Frances Ross at IBM and coworkers, which allows the TEM of wet samples. The basic idea involves building a chamber that sandwiches a thin environmental region between two electron-transparent windows. If the entire system is constructed so as to be sufficiently thin – say, less than a few hundred nanometers thick –it is possible to acquire atomic-resolution images of materials inside the environmental chamber, as various groups have now demonstrated. This technological advance opens up whole new areas for applications of the high-resolution imaging offered by electron microscopy. Whereas previously investigators interested in aqueous electrochemical or biological systems were limited to post-mortem analysis of desiccated or frozen samples, now they can study active electrochemical devices or biological cells in vivo.

Because the field of fluid cell TEM is so new, basic questions about best practices and fundamental limitations to the technique are still being answered. When we initiated work on this project it was still unknown, for instance, what kind of contrast would be generated by a bubble of vapor in water, and over what size ranges vapor bubbles would be stable. Since our goal was to image with the TEM the electrolytic production of hydrogen gas in water, we decided to begin with a study of nanobubble formation.

The stability of nanobubbles is not well understood. According to the Young-Laplace equation, the pressure differential supported by the bubble wall scales inversely with its radius of curvature, which implies a divergent pressure as the bubble radius approaches zero. In our first experiments we patterned a platinum nanowire spanning the TEM field of view through the fluid cell, and applied voltage pulses to the immersed nanowire while observing it in real time. Above a certain threshold, the induced Joule heating nucleated the formation of nanobubbles whose subsequent dynamics could be observed. We found that the vapor-liquid contrast was easily visualized, and that, depending upon the imaging conditions, nanobubbles could be observed over time scales ranging from sub-second to tens of seconds, with radii ranging from tens of nanometers up to several micrometers. The dynamics associated with the Joule-heating induced nanobubbles were particularly interesting; we found it impossible to nucleate and immediately image bubbles with sub-micrometer radii, and likewise we were unable to observe any shrinking nanobubbles achieve radii that were less than ten nanometers. Upon nucleation the bubbles would grow to micrometer size within a single frame (a couple of hundred milliseconds), and their final collapse to zero radius from a few tens of nanometers was equally instantaneous. These observations represent a new record in the high-resolution direct imaging of vapor bubbles in water, but new, faster techniques for visualizing the smallest bubbles are clearly required. During these experiments we also learned that the electron beam itself can influence bubble dynamics, and even nucleate new bubbles if condensed to sufficient intensity. This discovery sounds a cautionary note for investigators intending to examine aqueous phenomena at high spatial and temporal resolution, as high resolution imaging often requires beam intensities of such magnitude as to necessarily alter the system under observation.

The results of the nanobubble investigation, published in Applied Physics Express, encouraged us to more carefully explore the effects of electron dose on aqueous systems. To this end we designed a simpler system: one which had no active electrodes, but rather included ~4 nm platinum nanoparticles deposited on one of the two electron-transparent membranes constituting the windows of the fluid cell. Observing this system at a variety of dose rates, we found that even quite moderate imaging conditions were sufficient to produce dramatic charging effects. As a result of beam-induced secondary electron emission, initially immobile Pt nanoparticles would charge up and be driven out of the field of view. This study, published in Langmuir, provides a useful benchmark (roughly 104 e/nm2) for estimating the threshold dose beyond which one must beware the onset of beam-induced imaging effects.

After our studies of nanobubble formation and beam-induced charge interactions, we turned our attention to demonstrating direct observations of electrochemistry at the nanoscale. We elected to first work with a solution of lead (II) nitrate, reasoning that this salt's large solubility and the cation's large atomic number would both be advantageous for a TEM study. By applying a time-varying electric potential to gold electrodes in the fluid cell, we succeeded in plating and stripping lead in situ. Depending on the rate of the potential change, lead could be induced to deposit on the cathode in a structurally compact layer or in dendrites. The key variables that determine which of these two morphologies is prevalent are of critical interest for several important next-generation rechargeable battery chemistries. We also demonstrated direct imaging of the cation concentration, and direct correlation between the measured electrochemical transport (current-voltage) relationship and the time variation seen in the TEM images. These two firsts, reported in ACS Nano, are key milestones for establishing the power and scope of fluid-cell TEM for addressing fundamental issues in wet electrochemistry.

Funding provided by the American Chemical Society's Petroleum Research Fund over the last year has supported graduate student Edward White, who has been responsible for the body of work summarized above. More detailed descriptions of the research have been provided in the three journal publications mentioned, as well as at the American Physical Society's March Meeting, the 2012 Microscopy and Microanalysis Meeting, and in invited seminars at various academic institutions. This fluid cell work represents a completely new research direction for the PI (B.C. Regan), and efforts on the project are ongoing.