This laboratory has evaluated 2 automated nitrogen methods, the Kjel-Foss and the Missouri-Technicon with block digestor. A modified Kjeldahl method using a low level (40 mg) of CuS04 as a catalyst was also studied. These 3 methods were compared with the official AOAC Kjeldahl method. The experimental set numbered 22 samples, consisting of amino acids, ammonium salts, meals, grains, forages, and American Association of Feed Control Officials check feed samples. This set was analyzed in duplicate 3 independent times by each method. The accuracy and precision values for the 4 methods were excellent, within a very close range, and were equal or superior to the AOAC method. All 4 methods gave average relative standard deviations of <1 %. The 4 methods were compared for rapidity and timeliness of analysis, applicability, cost per sample, and labor and physical requirements to provide those who are responsible for protein nitrogen analysis with a basis for choosing a nitrogen method that would be most suitable for their needs.

A rapid, accurate, and precise spectrophotometry method, based on the nitroprussidemethionine color reaction, has been developed for the determination of methionine in soybeans, mungbeans, and corn. Amino acid interference was eliminated by partial enzymatic hydrolysis of samples for 4 hr at 50 °C with papain; the samples were analyzed by an automated system. Precision and accuracy were determined by repeated independent analyses of 13 samples of corn, mungbeans, and soybeans; the data were compared to independent classical ion exchange amino acid results. The range for each set of samples varied from 0.02 to 0.04 w/w% methionine. The per cent recovery for 12 of the 13 samples compared to those from ion exchange ranged from 95.2 to 107.1% with an average of 100.2%. The standard deviation varied from 0.01 to 0.02, and the per cent relative standard deviation from 11.8% for a 0.16 w/w% sample to 2.5% for a 0.42 w/w% sample. Recovery of added methionine from corn samples ranged from 92.7 to 102.0% (96.8% average) ; from soybeans 95.0 to 112.0% (101.9% average); and from mungbeans 91.3 to 102.0% (95.8% average). A reliable automated method has been developed which is applicable to different types of biological substances; 20 samples/hr can be analyzed.

The well known chemical interference of phosphorus on the emission of potassium in cool flames was verified and found to cause small but significant error in using the official final action AOAC method of analysis. This effect can be greatly reduced by incorporating the use of lanthanum along with the internal standard, lithium. Lanthanum has been shown to reduce the chemical interference of phosphorus by chemical bonding without affecting potassium or lithium emission characteristics. The resulting manifold can be used to analyze samples at the rate of 6 0 / h r with no significant interference.

Acid-soluble boron in fertilizers is complexed in acid solution with 2-ethyl-l,3-hexanediol and extracted into methyl isobutyl ketone. The extract is nebulized into an air-hydrogen flame and the intensity of emission of the BO2 molecule is measured at 518 inn. The method is convenient and inexpensive and provides detection limits of 0.1 ppm. The method was tested on 6 fertilizer samples with the relative standard deviation varying from 5.6% on a 0.053% (w/w) sample to 2.6% on a 7.81% (w/w) sample. The recovery of added boron ranged from 94.5 to 104.2% with an average recovery of 99.7%.