When the standard is not the standard

A calibration standard can satisfy the chromatographic system suitability parameters monitored in routine use and still be an order of magnitude from its nominal concentration. This report documents two such events in a single method development exercise on coenzyme Q10 (ubidecarenone), together with the solubility measurements that explain them. A stock solution prepared at 1 mg/mL in methanol:ethanol 65:35 (v/v) — approximately five times the equilibrium solubility of coenzyme Q10 in that diluent — was forced into solution by sonication and warming to approximately 36–37 °C, then held overnight. Calibration levels prepared from it the following morning were approximately ten-fold low, with a response-factor relative standard deviation of 14.9% and a coefficient of determination of 0.985. Throughout that sequence the instrument reported plate counts of 2,489–2,641, tailing factors of 1.272–1.296, injection repeatability of 0.23% across a four-fold concentration range, and carryover of 0.000%. No chromatographic parameter indicated a fault. Apparent equilibrium solubility was subsequently determined by saturation against a gravimetrically prepared calibration curve, both values lying above its highest level and so obtained by extrapolation: approximately 206 µg/mL at 24.5 °C and 119 µg/mL at refrigerator temperature, a decrease of 42%. Dissolution from the solid was found to require approximately 20 hours at 88 µg/mL and approximately 40 hours at 150 µg/mL at 24.5 °C, while crystallisation on cooling occurred overnight. A second and differently arising discrepancy was identified in the same study: a solution prepared below the solubility limit, judged visually clear after 20 hours, assayed at 82.3% of its nominal concentration. If the deficit is attributed entirely to incomplete dissolution, it corresponds to approximately 0.78 mg of the 4.4 mg weighed, a quantity not visible in the flask. The present data establish the deficit but do not independently establish its mechanism. Two minor peaks sharing the analyte's chromophore, at relative retentions of 0.745 and 1.373, are consistent with coenzyme Q9 and coenzyme Q11 on the compendial relative retention and on homologous-series behaviour, and showed relative peak-area percentages threefold and 1.3-fold higher in the saturated solution than in an undersaturated one, the direction solubility predicts. Both failures are invisible to inspection and undetected by system suitability. Practical recommendations are given.