Charging into Battery Analysis with Benchtop NMR

28th February 2023 | Author: James Sagar

Li-Ion Batteries, what is all the fuss about?

You can barely turn on the news or open your web browser these days without seeing something related to batteries, in fact just last week we had national battery day! This isn’t surprising, the drive for electrification to combat climate change has accelerated rapidly in the last 3 years. Looking out of the window of my home office, I see that 2 out of the 10 cars in view are fully electric without even considering the hybrids. This is a huge change from a few years ago when most people’s belief was “electric cars will never work”. At the heart of all this electrification, whether it is cars, personal electronics, grid storage or larger transport vehicles, there are batteries. Within all those batteries is a series of chemistry and materials science challenges that need to be overcome if we are to hit those electrification targets. Those targets are quite something to just keep up with. We need to double how much energy can be stored per kg and quadruple the power density - essentially how easy it is to get that charge in and out of the battery.

The Challenges of Charging

I am very fortunate that in my position I get to observe those analysis challenges and more importantly meet the people trying to push the boundaries, on an almost daily basis. I started my journey into batteries when I used to work with electron microscopy related solutions, this meant that I got to look at novel solid state electrolytes, anodes and cathodes to try to understand failure modes our customers were seeing.

When I moved into benchtop NMR instrumentation 4 years ago, the first thing that I naively said was “do we use these things for looking at batteries?” They tend to be chemistry problems and I think lithium has NMR active nuclei”. It turns out that people do and lithium NMR provides significant insights as to how batteries perform today, as well as giving us guidance how to develop tomorrow’s batteries.

I always like to start by looking at the whole battery when talking about this. The diagram shows just this, the anode, typically graphite at the moment, separated from the Cathode, usually a transition metal oxide, by a polymer separator that is soaked in liquid electrolyte.

That liquid electrolyte is the sweet spot for benchtop NMR spectroscopy. We are now finding more solutions for time domain NMR in electrodes and solid electrolytes, but I’ll discuss those another time.

What does benchtop NMR tell us?

Having wondered about batteries for a while, it wasn’t until we launched X-Pulse, with it’s ability to characterise any NMR active nucleus in September 2019 that we really started to explore this space. Electrolytes, it turns out, just happen to be perfect chemical systems to be comprehensively characterised with a broadband benchtop NMR spectrometer. If we take a typical commercial electrolyte LP30 (1M LiPF₆ in Ethylene Carbonate (EC)/Dimethyl Carbonate (DMC)) we can analyse 5 NMR active nuclei on a single sample. ¹H and ¹³C give you information about the EC/DMC solvent combination, ¹⁹F and ³¹P enable you to accurately characterise your [PF₆]⁻ anion and obviously ⁷Li gives you information about your lithium cations.

That is before we even get into the more esoteric or developmental electrolytes that may also contain boron or sodium, again both elements have NMR active nuclei that are easy to analyse. What was incredible to me, was just how much information the analysis of these spectra provides about your electrolyte. Measuring the ratios of the different peaks in the ¹H will tell you whether you have the appropriate solvent mixture, essentially, do you have LP30, LP40 or LP71 or are your bespoke additives (typically vinylene carbonate) there in the appropriate quantities? The ¹⁹F NMR spectrum turns out to be the most useful NMR spectrum for the analysis of current generation lithium battery electrolytes. I was amazed the first time a 30 second spectrum told us the exact mechanism of failure of an electrolyte as well as informing us of the decomposition path. Previously, this sort of analysis would have taken hours of time analysing with LC-MS. Finally, you get to the ⁷Li spectrum and, as much as it is great for telling us how much lithium is in our solution, the real power comes from measuring the diffusion of these ions. Having sat in my kitchen for the first two months of COVID induced lock-down, a chance videocall with researchers from the University of Oxford highlighted the potential of NMR diffusion analysis for electrolytes. Another month or so later and we had the first results showing that we are quite easily able, within 15 minutes, to quantify the diffusion coefficients of lithium cations as well as [PF₆]⁻ anions. In turn, this enabled ionic conductivity and cation transference values to be calculated. This is a great step forward - not only is benchtop NMR a great tool for QC and failure analysis of electrolytes, but now it can also be used as an active development tool in the lab.

What is the outlook for benchtop NMR in the battery space?

Since those early days, we have seen many different types of electrolytes and we continue to see new “recipes” on a very regular basis. To date we have not found an electrolyte which we have been able to analyse to some extent whether that is liquid polymer cocktail, single ion conduction electrolytes, or ionic liquids.

It seems to me that we are still just scraping the surface of what is possible, and the technique will continue to develop alongside the whole battery industry which itself is in its relative youth.