Improved analysis of picomole quantities of lithium, sodium, and potassium in biological fluids

The analysis of picomolar lithium, sodium, and potassium by electrothermal atomic absorption spectrophotometry was studied using a Perkin-Elmer Zeeman 3030 spectrophotometer. With ordinary pyrolytically coated graphite tubes, a number of interference effects associated with the sample matrix were observed. In particular, the lithium and potassium absorbance signal was depressed by chloride, an effect shown to be dependent on the preatomization heating. When an in situ tantalum-coated atomization surface was used, matrix interferences observed in lithium and potassium analyses were abolished, and the linear range for the potassium assay was extended. Technical difficulties encountered during sodium analysis at the primary wavelength were effectively circumvented by analysis at a less-sensitive wavelength (303.3 nm), at which tantalum coating also prevented significant chloride interference. The improved microanalyses were employed to reevaluate the handling of lithium, sodium, and potassium along the proximal convoluted tubule (PCT) of the anesthetized rat. The average tubular fluid-to-plasma concentration ratios for lithium [(TF/P)Li] and sodium [(TF/P)Na] were 1.13 +/- 0.08, n = 26, and 0.99 +/- 0.07 (n = 26), respectively. The tubular fluid-to-plasma ultrafiltrate concentration ratio for potassium [(TF/UF)K] was 1.09 +/- 0.05 (n = 13). Ratios did not change significantly with puncture site along the PCT for any of the ions. (TF/P)Li and (TF/UF)K were significantly greater than (TF/P)Na, indicating that lithium and potassium reabsorption do not directly parallel sodium reabsorption in the PCT.