Michal Leskes and Shira Haber identify issues with the assessment of solid-state interfaces using NMR. For more comprehensive information and tips on how to use ssNMR effectively, check out their paper ‘What Can We Learn from Solid State NMR on the Electrode–Electrolyte Interface?’ in Advanced Materials.
How can ssNMR be used to study the electrode–electrolyte interface?
Solid-state (ss) NMR spectroscopy can be used to detect the majority of the elements in the periodic table, since most elements have at least one isotope with non-zero nuclear spin. As such, NMR can be used to detect most organic and inorganic phases formed at the electrode–electrolyte interface. With ssNMR, we can distinguish different phases by effectively detecting slight changes in the chemical environment of the nuclei. Acquiring a spectrum of the electrode interface therefore provides a compositional map of the interphases formed upon electrochemical cycling.
Moreover, this compositional map is quantitative: it can be used to quantify the formation and evolution of these interphases, depending on the state of charge of the electrode. Since the nuclei detected are essentially small magnets, they also interact with each other, and NMR can be used to measure these interactions, which are an extremely sensitive probes to the distance between atoms. Such measurements provide structural information on the interface in addition to its composition. In principle, this information can be used to build a three-dimensional model for the interface: to determine which phase is formed first and deposited on the electrode, and what phases are deposited upon extended cycling.
Finally, since the interactions between the nuclei and their environment depend on the nuclei mobility, ssNMR can give us insight into the degree of order and dynamics within these phases. Potentially, ssNMR can even be used to probe the process of ions exchanging between the electrode and the electrolyte through the interface! This is extremely useful, since ionic transport is a crucial property for a stable interface, enabling the battery to cycle continuously without increasing interfacial resistance.
What challenges are associated with this method?
The biggest limitation of ssNMR is its low sensitivity. In order to gain this wealth of information, we need to have electrodes with large surface areas and significant formation of interphases.
In addition, NMR-active isotopes are often in low abundance, so we often have to label the electrolyte or the electrode with NMR-active isotopes to enable detection. For example, 13C has only 1% natural abundance, so to get insight into organic interphases we have to enrich them with 13C by using a 13C-enriched solvent for the electrolyte.
Another challenge of this method is that, to date, the study of interfaces and interphases has been limited to ex-situ studies where the battery cell has to be stopped and carefully taken apart. The reason for this is that, to distinguish the different chemical groups, we have to mechanically spin our samples in a process called magic angle spinning. This is essential to average the effect of the samples being anisotropic. To date, this need to spin the sample in order to get decent resolution (and therefore detailed chemical information) has prevented its application in-operando.
How can we overcome these challenges?
To overcome the sensitivity limitation, our group and others have started utilizing a new approach called dynamic nuclear polarization. In the past few years, it has been shown to provide a two- to four-fold increase in sensitivity by using electron spins, which have a much larger magnetic moment than nuclear spins.
Such sensitivity gains can have a huge impact on the amount of information we can get from ssNMR on very challenging interfaces.
Finally, if we were able to spin our batteries to acquire in operando ssNMR data from them, we could get real-time information on the formation and evolution of the electrode–electrolyte interface. If achieved, in operando ssNMR studies of the electrode–electrolyte interface would completely revolutionize our ability to understand these complex and important systems.
Interested in ssNMR? There are currently open positions in the Leskes group. You can also check out these recent battery articles:
- Predicting Rechargeable Battery Performance with Machine Learning
- Magnesium Batteries: An alternative to lithium?
- A New Solution to an Old Challenge: Recharging Cu-Zn Batteries
- Paving the Way for Calcium Batteries