期刊
FOOD ANALYTICAL METHODS
卷 9, 期 6, 页码 1627-1637出版社
SPRINGER
DOI: 10.1007/s12161-015-0340-2
关键词
High performance liquid chromatography; Alkaloids; Principal component analysis; Linear discriminant analysis; Coffee; Ethiopia
资金
- University of Gondar, Ethiopia
The alkaloid compositions of 99 green coffee (Coffea arabica L.) bean samples comprising eight varieties (Harar, Jimma, Kaffa, Wollega, Sidama, Yirgachefe, Benishangul and Finoteselam) from the major production regions of Ethiopia were investigated. High performance liquid chromatography was applied for the simultaneous determination of four coffee alkaloids in the aqueous extracts of the beans. The limits of detection for the method were established as 13 mg kg(-1) for trigonelline, 7 mg kg(-1) for theobromine, 8.5 mg kg(-1) for caffeine and 4 mg kg(-1) for theophylline in the dry coffee beans. Theophylline was not detected in any of the samples. The determined concentrations (% w/w dry coffee beans) ranged from 0.98 to 1.32 % for trigonelline, 0.0048 to 0.0094 % for theobromine and 0.87 to 1.38 % for caffeine. The concentrations of the alkaloids varied significantly, depending on the geographical origin of the beans. Theobromine was not detected in coffee beans from the East (Harar coffees), and its absence in samples can be used to ascertain whether the coffee originates from this region. Coffee beans from the Northwest were characterized by higher concentrations of caffeine. Application of linear discriminant analysis provided 75 % correct classification of samples into the respective production regions, with a 74 % prediction success rate. The moderate classification efficiency obtained when using alkaloid data demonstrates the potential of using this class of compounds in discriminant models for determination of the geographical origin of green coffee beans from Ethiopia.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据