In the LAWRENCE SMITH method for the determination of the alkali metals in silicates sodium and potassium are isolated finally as the mixed chlorides, Potassium usually being determined as the chloroplatinate or perchlorate and sodium by difference. More recently, modifications have been introduced by MILLER AND TRAVES~, and later by HASLAM AND BEELY~, which avoid the tedious separation of the mixed chlorides. In general, these modifications consist of opening out the silicate by the conventional LAWRENCE Snfrm procedure, making up the leachings to a known volume, and determining sodium and potassium separately on suitable aliquots taken from this solution. Sodium is determined fairly rapidly by precipitation as sodium zinc uranyl acetate, but potassium is first separated by precipitation as the cobaltinitrite and is then converted to the perchlorate, as the final weighing form. These modifications effect some considerable saving of time, but, if it were possible to determine potassium as conveniently as sodium, the total time for the analysis could be reduced still further, and it would be possible to complete it in little over half a day.
Since the present authors had been invited by Panel I of the British Coke Research Association to investigate the possibilities of developing a more convenient method for the determination of sodium and potassium in coal ash than that described in British Standard No. 1016 (which is based on classical procedures), it seemed probable that the modified methods for the determination of these elements in silicates could be applied with little modification. It seemed desirable, however, to investigate the possibility of developing a more convenient procedure for the determination of potassium.
Potassium may be determined accurately by the perchlorate or chloroplatinatc methods, after separation of the alkali metals as the mixed chlorides, but both methods are time-consuming. The cobaltinitrite method will separate potassium fairly rapidly from most other metals, but under the usual conditions of precipitation, the precipitate is of variable composition and cannot be used as such for the weighing form. Potassium is often separated by this process and the precipitate then converted to the perchlorate as the final weighing form. Many claims References p. 481.