April 2025, a Journal of Ethnopharmacology reviewer asks for Trolox lots on 24 measurements. For three of them, all I have is a phone snapshot of a post-it. The trigger.
PhytoNote had already existed since November 2023 — the founding night is told in The Midnight PhD Student. But the review from the Journal of Ethnopharmacology I am about to tell here triggered a complete pivot toward reagent traceability, and the provenance layer that now structures the app.
In April 2025, I submitted a manuscript to the Journal of Ethnopharmacology. The study looked at the antioxidant activity of hydroalcoholic extracts of Pistacia lentiscus, harvested at four altitudes in the Moroccan Rif, between 320 and 1480 metres. DPPH and ABTS assays, six months of bench work, twenty-four experimental points. I was proud of the dataset. The journal has an impact factor of 5.4, which for a phytochemistry researcher based in Morocco still counts as a serious target.
Six weeks later the verdict landed in my inbox: Major Revision. Seven requests from Reviewer 2. Anonymous, but most likely European — the phrasing on good laboratory practice was too precise to come from anywhere else. The question that took the wind out of me was point 4: "Provide the lot numbers, opening dates and storage temperature of all reference standards used (trolox, ascorbic acid) for each batch of measurements."
Back to the lab notebooks
I pulled my three bench notebooks from the shelf. Spiral binding, gridded paper, that methanol smell that soaks into the pages after months. I went through every single measurement: dilution, blank, reading at 517 nm for DPPH, at 734 nm for ABTS. Spectrophotometer UV-1900 Shimadzu, Greiner 96-well transparent microplate. For twenty-one out of twenty-four measurements, everything was there. Trolox lot number, vial opening date, storage temperature (-20 °C for stock solutions, +4 °C for working dilutions).
For the other three, blank spots. Once I had written "new vial" without giving the date. Once I had jotted the lot number on a post-it that I photographed in the lab in the evening, and the post-it had vanished since. Once I had simply forgotten. Three measurements out of twenty-four. Statistically minor, but enough for a reviewer to ask that the series be redone or removed.
I spent one evening digging through my phone gallery. The post-it was there, in the background of a picture taken at 19:12 on a Tuesday in October. Lot number and date partially legible. I zoomed in, adjusted brightness in Snapseed, recropped. Three characters were still ambiguous. That was the first time I had to lean on a phone snapshot of a post-it to defend a paper.
Six weeks of stress
I went to see my PhD supervisor. His answer was simple: redo the three measurements. Non-negotiable. Except redoing them meant either sourcing the exact same plant lots, or explaining in the rebuttal letter why we were now working on a fresh batch. The lentisk collected at 1480 m had been harvested in late September. Repeating it in June would not carry the same phytochemical meaning. Seasonal variation in polyphenol composition is not trivial in Anacardiaceae.
What I ended up doing: redoing the three measurements with a freshly opened Trolox vial (documented lot, opened on camera), recalibrating the entire series, and writing in the supplementary that the original measurements had been redone "to meet the reviewer's traceability requirements". I included the full table of standard lots. Fourteen pages of supplementary material. I sent the revision back on 11 August 2025.
During those six weeks I slept badly. It was not only the risk of rejection. It was realising that serious work can be undermined by three missing lines in a notebook. My bench hygiene was probably above my lab's average. And still.
The provenance pivot
The conviction that built up during those weeks was not "I need to be more rigorous". I was already rigorous by my own standards. It was something structural: as long as traceability depends on memory and pen, it will eventually fail. Not every day, not for everyone, but one day, for something that matters. And on that day, six weeks of your life evaporate.
That is when I redrew the entire provenance layer of PhytoNote. The app had been handling bench-side capture since 2023, but standards and their history lived in a free-text field. From that review onwards, the central gesture became: opening a standard vial. Not filling a form at the end of the week. The moment you crack the seal, you scan the barcode of the vial (most suppliers print one, or you can scan the QR on the certificate of analysis), or you photograph the COA and an OCR routine extracts lot and date. From that moment on, the vial has an identity. Every measurement I run afterwards inherits its provenance automatically. When I export a table for a manuscript, the metadata is already attached.
The idea was simple. The implementation, less so. I reworked PhytoNote through the autumn, in parallel with my thesis measurements, often until 23:00 in my small office in Rabat. The current version of the app covers the whole cycle: vials, measurements, batches, CSV export with provenance columns enabled by default. The same gap caught me later on a forgotten quercetin vial — that is the story of recit 7.
January 2026, acceptance
The paper went back into review in September 2025, round 2. This time, Minor Revision. Three comments on form, one request for clarification on the multivariate statistics. And one comment from Reviewer 2 that I read three times to be sure: "the provenance documentation is exemplary".
Final acceptance came on 14 January 2026. Online publication ten days later. The DOI landed in my inbox on a Saturday morning while I was making tea. I smiled at the screen on my own. It was not the first publication of my career, but it was the first where I knew traceability had not been a stroke of luck.
Reviewer 2, wherever they are, will probably never know that three lines of feedback reshaped the architecture of my bench tool. Maybe that is for the best.
What stayed with me at the bench
Since that story, I no longer trust my own memory on reagent provenance. Not out of distrust, out of experience. When I reopen a DPPH• vial, I scan. When I prepare a Trolox dilution in HPLC-grade MeOH, I scan the solvent bottle too. It takes three seconds. The benefit is not immediate. It shows up eighteen months later, when a European reviewer asks you to account for everything.
If you want to see how the app handles that flow, the PhytoNote page walks through the standard scenario. I also wrote a short methodological paper that explains the data model and the scientific rationale: it is available here, in open access.