BUNSEKI KAGAKU Vol. 59, No. 10, pp. 863-871 2010 2010 The Japan Society for Analytical Chemistry 863 X 1 1 2 3 1 X 6 Mn Fe Ni Rb Sr Ba ppm 75 6 6 6 10 2 30 1 1 2 3 4 5 6 AAS 7 ICP-AES 8 1 : 162-8601 1-3 2 : 101-8457 2-2 3 : 150-0001 6-26-1 X X ppm ICP-AES 9 10 100 kv Gd X ppm 11 12 X X X 13 12 X
864 B U N S E K I K A G A K U Vol. 59 2010 Table 1 Number of coffee bean samples Green beans 2 Roast beans Brazil 17 4 Colombia 17 4 Vietnam 17 4 Indonesia 10 7 Tanzania 7 4 Guatemala 7 4 2 1 X Epsilon 5 PANalytical 14 X X 100 kv X 2 2 X X : NIST Wheat Flour SRM 1567a Apple Leaves SRM 1515 Tomato Leaves SRM 1573a Spinach Leaves SRM 1570a NMIJ White Rice Flour CRM 7502a NIES Rice Flour-Unpolished CRM No.10-a 75 27 102 Table 1 X MM 400 RETSCHE 2 25 ml 20 mmf 4 tonf/ cm 2 5 X 13 15 mmf 20 mmf AAS ICP-AES/MS 1 10 X 0.5 4.5 g 2.5 μm : Mylar 3.0 g 20 mmf 2 2 3 X X 10 Al Ti Fe Ge Rh Ag Zr Mo CsI Al 2 O 3 1 X SN 300 7200 1 Lower limit of detection, LLD ppm 15 C I BG LLD = 3 1 I t net C : ppm I net : cps/ma I BG : cps/ma t : s. 2 4 X 11 SRM 1570a Spinach Leaves X FP
: X 865 Fig. 1 A comparison of the XRF spectra of coffee beans obtained by using various materials of secondary targets Measurement time : 3600 s ; (1) Zr ; (2) Mo ; (3) Ag ; (4) Rh ; (5) CsI 2 2 2 3 6 2 5 Soft Independent Modeling of Class Analogy SIMCA 16 Zr Stat Partner 2.0 NEC 3 3 1 10 X Rb Sr Mo Ag Rh Zr CsI Al 2 O 3 Table 2 Secondary target Measurement conditions for XRF analysis of coffee beans Voltage/ kv Current/ ma Analyzed elements Ge 75 8 Mn, Fe, Ni Mo 100 6 Rb, Sr Al 2 O 3 100 6 Ba Fig. 1 Mo Rb Sr Mo X Al 2 O 3 X SN Ba Ge Mo Al 2 O 3 Table 2 3 2 0.5 4.5 g 20 mmf Rb Sr X mm g Rb Sr Kα X Fig. 2 Mo X Kα Fig. 2
866 B U N S E K I K A G A K U Vol. 59 2010 5 Weight (g) 0 1.5 3 4.5 5 Intensity (cps/ma) 4 3 2 4 3 2 Normalized Intensity 1 1 0 0 5 10 Thickness (mm) 0 Fig. 2 Dependence of the XRF intensity on the sample thickness (sample weight) Measurement time : 3600 s ; secondary target : Al 2 O 3, : intensity of Rb and Sr, respectively., : normalized intensity of Rb and Sr, respectively. The XRF intensity was normalized by that of Compton scattering derived from the secondary target. Fig. 3 Effect of counting times on the Rb Kα-line and Sr Kα-line spectra of coffee measured using Mo secondary target Measurement time : (a) 300, (b) 600, (c) 1800 s Rb Sr Kα 6.0 mm 2.5 g X X X 17 4.0 mm 1.5 g 3 3 SN 300 7200 Fig. 3 Rb Sr 300 Ni Ba 3600 Ge Al 2 O 3 3600 Mo 300
: X 867 Fig. 4 Geographical characteristics of elemental concentration in coffee beans (a) Brazil, (b) Colombia, (c) Vietnam, (d) Indonesia, (e) Tanzania, (f) Guatemala 1 LLD Mn 0.41 ppm Fe 0.36 ppm Ni 0.39 ppm Rb 0.21 ppm Sr 0.18 ppm Ba 0.81 ppm ppm 11 12 4 1 10 3 1 2 30 3 4 75
868 B U N S E K I K A G A K U Vol. 59 2010 Table 3 Geographical characteristics of elemental concentration in coffee beans Production area Brazil Colombia Vietnam Indonesia Tanzania Guatemala Sample number (n) 17 17 17 10 7 7 Concentration (ppm) a) Mn 39.7 4.98 47.02 8.55 29.3 2.34 34.6 20.1 37.5 4.76 39.7 10.1 Fe 35.5 3.79 30.7 1.39 52.9 13.2 32.8 3.50 28.3 2.73 29.7 1.46 Ni n.d. b) 0.527 0.285 c) 3.50 0.885 n.d. b) n.d. b) n.d. b) Rb 22.2 5.75 23.5 7.42 42.7 13.8 82.5 13.4 67.2 9.55 37.7 7.67 Sr 3.38 0.684 11.8 8.25 3.32 0.700 5.73 1.09 6.40 2.89 6.58 1.85 Ba 3.04 1.03 8.59 3.80 2.32 1.42 d) 4.31 2.76 7.67 4.01 6.28 2.34 a) Concentration : mean standard deviation. b) n.d. : not detected. c) n 10. d) n 15 Fig. 5 (a) PCA plot for coffee green beans calculated by concentration of 6 elements (Mn, Fe, Ni, Rb, Sr, Ba); (b) Factor loading plot of the principle component 1 and 2 : Brazil, : Colombia, : Vietnam, : Indonesia, : Tanzania, : Guatemala Fig. 4 Table 3 Table 3 Mn Sr Ba b Fe Ni c Rb d e d Mn Fig. 4 Mn 50 ppm 3 7 20 ppm 6 Mn Fe Ni Rb Sr Ba 75 Fig. 5 a Fig. 5 b 44.5 19.7 Fig. 5 a 6 Fe Ni Rb
: X 869 Fig. 6 Comparison of concentration of each element examined the ratio of the analytical values of green beans and those of roast beans Error bar : 1σ 6 6 3 5
870 B U N S E K I K A G A K U Vol. 59 2010 Fig. 7 (a) PCA plot for coffee green beans and roast beans samples calculated by concentration of 6 elements (Mn, Fe, Ni, Rb, Sr, Ba); (b) Factor loading plot of the principle component 1 and 2 : Brazil, : Colombia, : Vietnam, : Indonesia, : Tanzania, : Guatemala.,,,,, : roasted-bean samples 27 1 Fig. 6 0.5 2.0 3 4 75 27 102 Fig. 7 a Fig. 7 b Fig. 5 a Fig. 7 a Mn Fe Ni Rb Sr Ba 6 4 X ppm 5 Mn Fe Ni Rb Sr Ba 6 6 X X 1) : 2 (2003), (NTS). 2) M. Choi, W. Choi, J. H. Park, J. Lim, S. W. Kwon : Food Chem., 121, 1260, (2010). 3) : (Bunseki Kagaku), 49, 405 (2000). 4) K. Ariyama, H. Horita, A. Yasui : J. Agric. Food Chem., 52, 5803 (2004). 5) A. Marcos, A. Fisher, G. Rea, S. J. Hill : J. Anal. At.
: X 871 Spectrom., 13, 521 (1998). 6) P. P. Coetzee, F. E. steffens, R. J. Eiselen, O. P. Augustyn, L. Balcaen, F. Vanhaecke : J. Agric. Food Chem., 53, 5060 (2005). 7) V. Krivan, P. Barth, A. F. Morales : Mikrochim. Acta, 110, 217 (1993). 8) W. P. C. Santos, V. Hatje, L. N. Lima, S. V. Trignano, F. Barros, J. T. Castro, M. G. A. Korn : Microchemi Journal., 89, 123 (2008). 9) : p. 255 (1995), ( ) ; C. Vandecasteele, C. B. Block : Modern Methods for Trace Element Determination, (1993), (John Wiley & Sons, Chichester). 10) K. M. Bisgard, J. Laursen, B. S. Nielsen : X-ray Spectrometry, 10, 17, (1981). 11) : (Bunseki Kagaku), 56, 1053 (2007). 12) : (Bunseki Kagaku), 58, 1011 (2009). 13) Y. Sahin, S. Nas, H. Y. Gokalp : Food SCI. and Tech., 26, 485 (1991). 14) Epsilon 5 (PANalytical) website http://www. panalytical.jp/prod/xrf/system/ep5.html (accessed 2010-7-25). 15) : X 16 (2009), ( ). 16) : p. 253 (2007), ( ). 17) E. Margui, I. Queralt, M. Hidalgo : Trends Anal. Cem., 28, 362 (2009). Determination of Trace Elements in Coffee Beans by XRF Spectrometer Equipped with Polarization Optics and Its Application to Identification of Their Production Area Takao AKAMINE 1, Akiko OTAKA 1, Akiko HOKURA 2, Yuji ITO 3 and Izumi NAKAI 1 1 Department of Applied Chemistry, Faculty of Science, Tokyo University of Science, 1-3, Kagurazaka, Shinju-ku, Tokyo 162-8601 2 Department of Green & Sustainable Chemistry, School of Engineering, Tokyo Denki University, 2-2, Kanda- Nishikicho, Chiyoda-ku, Tokyo 101-8457 3 Kirin Beverage Co., Ltd., 6-26-1, Jingumae, Shibuya-ku, Tokyo 150-0001 (Received 16 April 2010, Accepted 18 August 2010) The production area of coffee beans becomes a brand name, which gives reputations for the products, which is related to the price. This leads to room for mislabeling the products by unscrupulous market dealers. A rapid and easy method for the analysis of trace-element composition of coffee beans, which could be a good indicator of their production area, was studied in the present work. Coffee beans of 6 different regions (Brazil, Colombia, Vietnam, Indonesia, Tanzania, Guatemala) were analyzed by using a highly sensitive X-ray fluorescence spectrometer with three dimensional polarization optics. The experimental conditions were optimized so as to analyze 6 elements (Mn, Fe, Ni, Rb, Sr, Ba) in coffee beans, and linear calibration curves were obtained for the quantitative analysis of those elements. The analytical results were used in principal-component analysis to classify the coffee beans according to the geographical origin, which results in a successful characterization of the 6 production areas. It is found that roasted beans can be used with the same criterion as their green beans. Consequently, a rapid and easy way for the characterization of the geographic origin of coffee beans has been established in this study. Keywords : coffee beans ; X-ray fluorescence analysis ; trace elements ; geographic origin ; multivariate analysis.