Simultaneous Determination of Sodium and Potassium in Small Volumes of Fluid by Flame Photometry

INTRODUCTION In two earlier papers (Ramsay, 1950; Ramsay, Falloon & Machin, 1951) a method was described by which it was possible to determine the concentration of sodium in samples of fluid of the order of io~ cu.mm. in volume. This method has now been extended to include the determination of potassium simultaneously upon the same sample. It is still in the experimental stage of development; we are fully conscious of its shortcomings and have in mind various possibilities for its improvement. Nevertheless, it is being used in its present form for biological work, some of which is already published (Ramsay, 1952), and it is therefore necessary that some description of it should be given. The general principles and technique of flame photometry are becoming well known; various instruments have been described and some are on the market In the majority of these instruments the solution to be investigated is atomized into the flame and the steady level of emission is measured. The method now under discussion differs mainly in that the whole of the sample is passed into the flame and the total emission is integrated. A small volume—of the order of io" cu.mm. —of the fluid to be investigated is drawn up into a silica capillary pipette to a mark. This sample is then transferred to a platinum wire and dried. The platinum wire is thrust into the flame by a mechanical device. The emission from the flame enters a spectrometer; the spectrum falls upon a mask having two slits, one passing the sodium doublet (5890-96A.) and the other passing the potassium doublet (7665-99 A.), and these emissions are allowed to reach two photomultiplier tubes. The photomultipliers are connected to integrating circuits with meter presentation. The apparatus is calibrated with known solutions drawn up to the same mark in the capillary pipette. In essence, this is the same as the method previously described, differing only in the provision of a second channel for potassium. But for operational reasons further complications become necessary. In the earlier method the meter was first brought to rest near zero and the sample was passed into the flame; the meter made a rapid excursion and then fell back slowly. The operator noted the maximum deflexion, subtracted the zero reading and recorded the net deflexion. This was