## Answer

The structured records contain no explicit "program start" date — there is no field marking when the 9076 program began. The start can only be inferred from the earliest dated experiment entries.

Earliest dated experiments on record:

• Batch 1 — the first numbered batch — has no explicit date field. Its sample alias S 9076 #L01301224 encodes an L0130-12-24 date (Dec 24, 2024), and its application tests are dated 02.01.2025. These are the earliest FOIL *measured* data points in the drop.
• Batch 5 is dated 27.12.22 (27 Dec 2022) — an earlier calendar date, but it appears on a later batch number, so the numbering does not map cleanly to time.

So the program's *documented measured work* spans from at least 27.12.22 (Batch 5) through 04.05.2026 / 02.06.2026 (Batches 63–69), with Batch 1's lab sample carrying a 24 Dec 2024 date code and trials on 02.01.2025.

Caveats / gaps:

• There is no field that literally says "program started on …". I am inferring a start from per-experiment dates, not stating one.
• A few early batches (1, 2, 3, 4, 6…) have date=- (no date captured), so the true first experiment may be earlier than the dated entries but is not dated in this data drop.
• The 2021-01-26 date (unknown, notebook p19, Phosphate E3+2) is a STUB / unlinked sample, not a confirmed 9076 program experiment.

If you need a definitive "kickoff" date, it is not in the structured records — only the earliest dated experiments are.