Abstract

A rapid and accurate method has been developed for the determination of boron in serum, plasma and urine by inductively coupled plasma mass spectrometry. The memory effects of B were examined using different diluents/rinse solutions, including water, nitric acid, Triton X-100, ammonia and mannitol in water, in nitric acid and in ammonia. A combination of ammonia with mannitol, as both diluent and flush solution, gave the best precision, the minimum memory effect and the lowest background. A sample dilution of 20-fold was simply made for serum and plasma and 100-fold for urine for determination with a single calibration curve. Beryllium was employed as the internal standard to control matrix effects and to compensate for possible fluctuation and instrument drift. The isotope ¹⁰B⁺ was utilised to avoid spectral overlap by the intense ¹²C⁺ isotope. The final solution of blank, standards and samples contained 0.25% w/v mannitol, 0.1 M ammonia and 20 ng ml⁻¹ of Be. Six samples, including human and horse serum, human and horse plasma, and human urine, were analysed to test the reliability of the method. A limit of detection (3σ) of 0.015 ng ml⁻¹ was obtained and the recoveries of spiked boron (two spiking levels for each matrix) from the selected samples ranged from 98% to 104%. Much higher concentrations of B in urine (≈1 µg ml⁻¹) were found compared to those in serum and plasma samples (32.8–61.1 ng ml⁻¹).