Mobile NMR for the plant sciences: Engineering for sensor-like, outdoor use

Some of the most important indicators of plant performance and yield are water content, dry matter content and sap flow velocity. None of these parameters are easy to measure without harvesting or destroying the plant, yet have shown to be fairly straightforward to measure by means of NMR relaxometry, MR imaging and MR velocimetry. It would therefore be attractive to make such NMR methods available for agronomic outdoor use. However, despite the exciting developments that have been made in the realm of mobile NMR in the last decade, the application of NMR methods on plants has mostly remained restricted to the laboratory. The most important limitations for more widespread use are cost, complexity and poor suitability for outdoor and glasshouse use.

The largest challenges in bringing NMR to the field are in the hardware. The magnet needs to be light and small enough to be mobile, yet for integrative (non-unilateral) measurements should have a homogeneous region that is large enough to contain the object of interest, and should offer access to the samples of interest - stems, branches or fruit. The RF coil, the gradient set (if applicable) and the shielding should be open or openable as well. It further should be possible to mount the system in an arbitrary position in, under or around the plant part of interest. Finally, the setup should be robust and be able to cope with the presence of moisture.

Possibly the most challenging factor when working under field conditions is temperature. Not only the magnet should be temperature controlled or temperature stable, so should the spectrometer. The temperature of the sample, on the other hand, should (if possible) be allowed to change with the environment, but should be measured and recorded in order to compensate for temperature induced changes in the Boltzmann equilibrium.

In this contribution we present a number of ways to address these challenges and bring mobile NMR to the field. We tested a number of magnet designs and explored their suitability for use in a number of simple NMR devices. For sensor-like applications we explored the use of simplified relaxometric methods to estimate water- and solid content. Such measurements could be done in a highly automated fashion, either leaving the spectrometer to run autonomously for days or controlling it remotely, and were found to produce data of surprisingly high information content on subjects ranging from growing seeds to droughted trees. We further demonstrate the application of a mobile low-end NMR device as a fully battery-driven imager.