Two operators, same reactor, same chemistry, different results. The divergence almost always traces to three habits: how the reactor was sealed, how temperature was controlled, and how samples were taken. Standardizing these three turns a vessel into an instrument.
Seal the same way every run: same torque sequence on clamped flanges, same cap condition, same septum freshness. Verify the seal before adding value, pressure hold for pressurized runs, inert-purge verification for oxygen-sensitive chemistry. A seal that "usually holds" is a run that usually works, which is not the standard we publish on.
Control the reaction temperature, not the bath temperature, they differ under irradiation by amounts that shift kinetics. Put the probe in the liquid or against the wall at the reaction zone, log it through the run, and treat setpoint-only records as incomplete. When comparing runs, compare temperature curves, not settings.
Sample with a fixed protocol: same volume, same position, same quench method, same time stamps relative to light-on. Heterogeneous reactions change with every withdrawal, suspension fraction leaves with the aliquot, so either account for it or keep sampling fractions small and identical. Filter or centrifuge immediately to stop the reaction in the sample.
Log seal verification, temperature curve, sampling times and any deviation, a run sheet of ten lines is the difference between data and anecdote. The next operator, including future you, inherits the run sheet, not the memory.
See instrument instructions for Perfectlight reactor operating procedures and application solutions.
