Interview with Daniel Kowalczyk, winner of the FAIR4Chem Award 2025
The biggest hurdle is often switching to a digital lab notebook in the first place. However, even incomplete documentation is better than none at all.
What was your submitted data set about?
The data set I submitted was created as part of my doctorate. I work in the Catalight Collaborative Research Centre, which deals with systems for artificial photosynthesis.
We realised that there was a fundamental problem with the reproducibility and comparability of experimental results, even within the CRC. The reason is that in photochemistry the parameter light has to be taken into account as a reaction partner, and the quantification of photons is crucial.
To solve the problem, we developed the modular photoreactor concept, which we can now use to characterise photons. The dataset we submitted documents this modular photoreactor.
Specifically, we have carried out a photonic characterisation. In the photoreactor, we measured how many photons are introduced into the systems in various configurations. We combined two methods to ensure that the calculated photon fluxes are valid: chemical actinometry and radiometry. We spent a lot of time harmonising the two methods in a complementary way, which was ultimately very successful.
Were there any difficulties with the FAIR storage of data?
There were no major problems. I wanted to prevent the photoreactor concept from being forgotten once I had completed my work. That was the motivation to deal with the FAIR principles. We had to clarify how the generated data could be documented FAIR in repositories, because the data sets can be very large. Radiometric two-dimensional scans, for example, in which an area of 50 × 50 centimetres is scanned, comprise over 20,000 individual measurements that cannot be documented in GitHub.
For this reason, we have chosen a two-pronged approach: We publish large data sets on Zenodo and the platform’s technical documentation on GitHub. Beforehand, we developed a specific workflow that takes into account our internal institutional and group requirements for FAIR documentation.
STL files are stored in Git so that we can reproduce all the reactor modules developed to date ourselves. It also contains the complete documentation for the entire photoreactor – from screws and electronics to assembly instructions. This aspect is particularly critical, as publications often lack important information that would be necessary to reproduce a specific setup or set comparable reaction conditions.
Data management plans: Are there DMPs in the CRC?
Data management and data documentation are now mandatory due to DFG funding. However, we do not have any strict guidelines on how this must be implemented, apart from the mandatory use of digital lab journals.
Various systems exist. Chemotion, for example, is well known and suitable for synthesisers. In our group, we use eLabFTW. This platform can be customised very flexibly to the individual needs of individual groups. Workflows can be customised using your own templates.
A key advantage of ELNs is that it is not just a data storage system, but that the data can also be further processed in a meaningful way. For example, measurements from online analytics can be directly linked and automatically documented. The data can then be extracted, converted and further processed.
This opens up opportunities for automation and reaction optimisation. For example, a data set can be analysed, interpreted by an algorithm and used to automate the next experiment. Parameters that are relevant for the optimisation are suggested.
Together with the structured documentation of the metadata, ELNs therefore offer considerable added value compared to pen & paper documentation.
Why did you apply for the FAIR4Chem Award?
There were two main factors:
Firstly, I wanted to check whether my own thoughts on FAIR documentation of the photoreactor concept were in line with expert opinion – a kind of feedback loop to learn from feedback, avoid mistakes in the future and develop better strategies.
Secondly, I know the cost of commercial reactor platforms. For many groups, it is not financially possible to purchase such systems. However, if a reactor can be developed in-house that is similarly or even better characterised, this also opens up the possibility for financially limited groups to achieve research results and thus advance scientific work.
Do you have a recommendation for other researchers who are not yet saving FAIR on how to get started?
My recommendation would be to just get started. That may sound clumsy, but it actually helps a lot. As a rule, getting started with FAIR data documentation shows what works and what can be improved. If a group implements a digital lab notebook and has all members create templates first, the strengths and weaknesses of the system quickly crystallise.
The biggest hurdle is often switching to a digital lab notebook in the first place and thinking about FAIR data documentation during the workflow rather than after the fact. Even initially incomplete documentation is more valuable than no documentation at all. Ultimately, practical application shows which procedures work best.
And why should researchers store FAIR data?
Documenting data FAIR from the outset prevents you from losing track of your own data. It also makes it easy to share data with others. The advantage therefore initially lies with your own data. However, if others also document their data FAIR, this creates an additional advantage for all scientists in this field. So: FAIR documentation is really valuable.
Thank you very much.
