To be of use for a broader community and audience, EPR spectroscopy needs to be routinely applicable to many different samples and questions. This requires standard protocols, high availability, short overall measurement and analysis times, to name but a few aspects.
EPR spectroscopy is intrinsically complex and hence complicated. However, many of the problems preventing the method from being much more routinely applied are not related to the underlying physics (indeed, requiring to be paramagnetic imposes severe limitations to the samples that can be investigated), but to the lack of automation and routine protocols for data acquisition and analysis.
There need to be clear protocols in place to basically characterise a new and hence unknown sample. Experimental parameters such as microwave power and modulation amplitude need to be adjusted for each sample individually. However, this can be fully automated and in a first step carried out following established standard protocols. Own experience shows that such standard protocols seem not to be generally established in the community.
Furthermore, routine operation requires using the simplest setup and type of measurement possible. Of course, pulsed EPR methods have some clear advantages over conventional cw-EPR spectroscopy. However, due to spin relaxation times pulsed operation usually requires performing experiments at cryogenic temperatures. Additionally, cw-EPR measurements are usually much faster than pulsed experiments for the same samples, let alone the difficulty of relating low-temperature measurements to processes usually operating at room temperature.
Besides using always the simplest available method and setup possible to address a given question, EPR spectra usually cannot easily be interpreted without processing and analysing the results, including fitting spectral simulations to the data obtained. While for simulations, the EasySpin package has been widely adopted as standard, there is nothing comparable for fitting of the resulting simulations to the actual data. Due to the usually large number of parameters, robust fitting strategies including sampling of starting conditions and globally fitting different spectra need to be applied and the methods implemented in an easy-to-use and robust software library.