State-of-the-art lithium-ion batteries (LIBs) employ graphite anodes and transition-metal-oxide cathodes. If in lieu of graphite, Li metal is used as an anode, these batteries can be made more energy dense (increasing both weight-specific and volumetric performance). Unfortunately, Li-metal anodes suffer from nonuniform deposition/dissolution electrochemistry. This leads to needle-like deposits that could cause an internal short circuit in the extreme event and pose a safety risk. This phenomenon is generically referred to as dendrite formation, and is a major challenge for cycle life and commercialization of compact LIBs employing Li-metal anodes. A related aspect of futuristic batteries, especially for automotive applications, is the ability to charge fast. For example, a battery being charged at 6C would recharge in 10 min (1C implies a 60-min charging time). These operational requirements put additional constraints on practical application of Li-metal electrodes.

James M. Tour’s research group at Rice University has come up with an innovative approach to resolving the dendrite problem as well as stable, fast-charging electrodes. The study was recently published in ACS Nano. Instead of working with bare Li metal electrodes, the researchers developed a porous carbon host for Li electrochemistry using asphalt (Asp). Graphene nanoribbons (GNRs) were added to ensure sufficient electronic conduction. Such a porous Asp-GNR electrode provides better cyclability (hundreds of cycles) and electrode performance (i.e., smaller overpotential), even at higher rates of operation (as high as 10C). Interestingly, the morphology of deposited Li in this host is uniform, rather than the typically observed dendritic features on Li anodes—noting that the term “dendrite” is used here to signify the whole set of nonuniform Li deposition morphologies.

The superior response of these Asp-GNR electrodes is attributed to a complex interplay among interfacial and bulk phenomena. The extremely high surface area of the Asp-GNR composite (>3000 m²/g) gives rise to more efficient electrochemical reactions at the electrode–electrolyte interface (i.e., lower overpotential), while reasonable conductivity resulting from GNR provides sufficient bulk conduction. Use of asphalt is important as it forms the necessary high-surface-area backbone structure. Additionally, it has very low levels of graphitization, which in turn reduces the propensity for Li intercalation in the host material, and promotes Li deposition at the surface. This intercalation-free approach also favors ultrafast operation, as sluggish solid-state diffusion—which is a characteristic of the intercalation process—is completely bypassed. Thus, the proposed Asp-GNR electrodes exhibit a reasonable balance between different transport processes, which leads to stable and fairly uniform cycling. The use of a porous host structure also provides structural rigidity and the space to accommodate volume changes upon Li deposition and dissolution.

While discussing the implications of this work, Martin Bazant from the Massachusetts Institute of Technology (not related to the published work) applauded the improvement in the efficiency of cycling Li-metal anodes in an asphalt-graphene matrix, and felt that it would be interesting to identify the mechanism, either modifying transport or reaction kinetics, in future work.

Thus, the Asp-GNR porous structure offers an interesting solution to the quest for uniform, stable Li-metal electrochemistry. It simultaneously combines augmentative characteristics such as reduced electrode impedance and uniform Li electroplating. Future work relies on further improvement of such electrodes to make them a commercial reality.

Originally published in the December 2017 issue of MRS Bulletin.