Comparison of two methods for the direct determination of fat in butter, blended spreads, and margarine

Autor: Ursula Coors, Angelika Miebs, Joachim Molkentin, Jaap M Evers
Rok vydání: 2001
Předmět:
Zdroj: Scopus-Elsevier
ISSN: 1438-9312
1438-7697
Popis: A German (GE) method and a New Zealand (NZ) method for the direct determination of total fat in spreadable fats were compared in a pilot collaborative study to obtain preliminary information regarding the precision and the level of fat content determined. The study comprised six products, including butter, blended fat products, and margarine with a fat content in the range of 39-82%, and four laboratories. The GE method, based on a Soxhlet extraction using petroleum ether, yielded a repeatability estimate of 0.41 g/100 g and a reproducibility estimate of 0.79 g/100g. The NZ method, based on an extraction of the fat by thorough shaking of the sample with petroleum ether followed by phase separation, yielded a repeatability estimate of 0.32 g/100 g and a reproducibility estimate of 0.62 g/100 g. The difference in precision estimates between the two methods was significant for repeatability, but not for reproducibility. The better repeatability of the NZ method was probably partly due to the use of a superior sample preparation procedure which ensured that any inhomogeneity developed inside the sample container during transport or storage was removed. The average fat content as determined by the NZ method was 0.20 g/100 g higher than that determined by the GE method, but for individual samples no constant bias was observed. However, an inverse correlation between the bias and the fat content of the product suggested that the anhydrous sodium sulphate used in the GE method interfered with the fat recovery when analysing samples containing a high percentage of water. Thus, according to this study, the NZ method seems to have certain advantages over the GE method with respect to the results obtained. Une etude pilote compare deux methodes allemande et neo-zelandaise de dosage direct des matieres grasses totales dans les matieres grasses tartinables. Dans quatre laboratoires sont analyses six produits, parmi lesquels le beurre, les melanges tartinables et la margarine, avec un taux de matieres grasses de 39 a 82%. La methode allemande repose sur une extraction Soxhlet avec de l'ether de petrole tandis que la methode neo-zelandaise recourt a une extraction par forte agitation de l'echantillon avec de l'ether de petrole suivie d'une separation de phases. Les deux methodes sont equivalentes vis-a-vis de la reproductibilite mais la methode neo-zelandaise est meilleure quant a la repetabilite.
Databáze: OpenAIRE
Externí odkaz: