Determination of trace cyanide in 3-guanidinopropionic acid by stripping preconcentration/isolation followed by flow-injection analysis with amperometric detection at silver

A method to determine trace levels of cyanide in 3-guanidinopropionic acid (3-GPA) was developed and validated. Instrumentally, a detection limit of less than 1 ppb (1 ng/g) was achieved. However, because of variability in the blank, a routinely achievable detection limit of 5 ppb is quoted. In this method, 2.0 g of 3-GPA is dissolved in 5.0 ml of 2.0 N H3PO4, the HCN stripped with helium at 75 °C at 40 cm3 min−1 for 15 min, trapped as CN in 1.0 ml of 0.05 M NaOH, then measured by flow-injection analysis with amperometric detection at silver. The procedure proved linear over the tested range, 5–100 ppb, yielding an average recovery of 81.7%. To achieve complete recovery would require stripping for a longer time and/or at a higher temperature, both impractical. Precision obtained on eight replicate preparations of 3-GPA fortified with 25 ppb CN was 7.5% R.S.D. With two of the eight preparations removed, the R.S.D. was 2.7%. This difference in precision reflects the variability of the blank, which we found difficult to control. Four lots of 3-GPA were analyzed for cyanide, yielding concentrations of 3.2, 3.0, 2.1, and 1.2 ppb. However, because the signal-to-noise ratio was less than 10 for three of the four lots, we report the results as cyanide present, but not quantifiable.