If the fertilizer is a straight potassium fertilizer such as potassium chloride (muriate of potash), sulphate or nitrate, it may be extracted by shaking with water. For any fertilizer, including mixed or compound fertilizers, the AOAC official method (No. 983.02B, Preparation of sample) recommends extraction with boiling 4% ammonium oxalate solution. The ammonium oxalate was introduced to precipitate calcium, which could suppress the potassium reading by up to 2.5% in an erratic fashion (Schwer and Conan, 1960). The precipitated calcium oxalate is allowed to settle and an aliquot for analysis taken from the supernatant, or else it is filtered. The final determination uses an automatic analyser and flame photometer (Johnson, 1990b). The SI 1996 No. 1342, however, recommends extraction of straight, compound or fluid fertilizers with boiling water followed by removal of interferences and gravimetric determination of precipitated potassium tetraphenylborate. We will give both extraction methods, and assume the final analysis is by flame photometry. Results using the AOAC method should be reported as oxalate-potassium, thus differentiating them from water-soluble-potassium, which might not give the same values. The one drawback with using flame photometry in fertilizer analysis is the 'phosphate effect'. This is the depression of the intensity of the potassium flame emission signal arising from the presence of higher amounts of phosphate in the fertilizer (Gehrke et al., 1964). These authors found it was the only commonly encountered ion that could cause a serious interfering effect. However, at the normal phosphate levels found in superphosphate, mono- and diammonium phosphate fertilizers, no interference was found by Schwer and Conan (1960). Gehrke et al. (1964) also found that calcium enhanced the signal if present at more than 50 pg Ca ml-1 (but see Schwer and Conan, 1960, who found a depressive effect).
If the Ca:K ratio exceeds 10:1, the interference is significant and the Ca should be removed, which can be achieved by extraction and precipitation using ammonium oxalate. The phosphate effect on a 20 pg K ml-1 standard was such that there was no effect on the K:Li emission intensity (Li is the internal standard for the Technicon flame photometer used by the authors) for concentrations up to 50 pg P2O5 ml-1, an enhancement occurred from 50 to 200 pg P2O5 ml-1, no further effect from 250 to 350 pg P2O5 ml-1, but a suppression occurred from 350 to 750 pg P2O5 ml-1, when a further plateau occurred. This can be largely compensated for by making up the potassium standards with potassium dihydrogen orthophosphate (potassium phosphate, monobasic). The AOAC method, in addition, adds a lanthanum solution (La2O3 in HNO3) to the lithium nitrate internal standard reagent in order to remove the phosphate effect.
In the manufacture of fertilizers, the large numbers of samples prohibit the routine use of the official methods for on-line analysis of potassium. Instead, Hydro Agri (UK) Ltd use a modification of the AOAC method with air-segmented continuous flow analysis and detection by flame photometry using an internal standard. In the UK, official methods in The Fertilisers (Sampling and Analysis) Regulations 1996 - SI 1996 No. 1342 are used to check that the marketed fertilizer conforms to the Fertiliser Regulations 1991 and subsequent amendments (J. Vessey, Hydro Agri (UK) Ltd, 2001, personal communication).
If the acid digestion for total nitrogen has already been carried out according to Methods 6.1a or 6.1b, either of the diluted digest solutions may be used, providing standards are made up in the same concentration of sulphuric acid as the extract. Both methods give the same concentration of potassium in the final diluted solution: a pure potassium chloride sample will yield 524 pg ml-1 K, and a pure potassium nitrate sample gives a solution of 387 pg ml-1 K. With any of the methods, further dilution may be necessary to suit the sensitivity of the flame photometer, and the calculation should be amended accordingly.
Was this article helpful?