Vol.2 No.1 2009

19/88

Research paper : Expansion of organic reference materials for the analysis of hazardous substances in food and the environment (T. Ihara et al.)−16−Synthesiology - English edition Vol.2 No.1 (2009) number of molecules in the analyte can be obtained. The mass (weight) and molecular weight of the target substance can then be used to determine the purity of the analyte[3]. Therefore, quantitative NMR is, in principle, a primary ratio method which can be used to obtain traceable measurement values for the number of protons — that is, amounts of substance in a sample.In the example in Fig. 3, both the analyte and the PS are pure substances. After weighing the two substances individually, they are dissolved in a deuterated solvent, and quantitative NMR is used to measure the purity of the analyte using the mass ratio of the two substances. Working RMs, in contrast, are often supplied in the form of solution. If supplied at a certain concentration (about 0.1 %), quantitative NMR can be applied by dissolving the working RM in an appropriate deuterated solvent. The concentration of working RM can be found from the number of molecules obtained for the analyte, the mass of sample solution added, and the number of molecules in the analyte.4.3 Feasibility of quantitative NMRNational metrology institutes in several countries (including AIST), which are members of the Consultative Committee for Amount of Substance (CCQM)Term 12, have shown interest in the possibility of applying quantitative NMR as a primary ratio method, which was first suggested by Germany’s Federal Institute for Material Research and Testing (BAM). In 2001, the Laboratory of the Government Chemist (LGC) in the United Kingdom and BAM served as pilot laboratories to conduct an international comparisonTerm 13 for the quantitative analysis of ethanol in aqueous solution, with the participation by 10 institutes in key countries. On this occasion, measurements were conducted on the same sample using conventional analytical approaches such as gas chromatography (GC) as well as quantitative NMR[4]. The sample was precisely produced by LGC, one of the pilot laboratories. The ethanol concentration was 1.072 mg/g ± 0.006 mg/g, but this value was not disclosed to the participants. Also, BAM separately supplied a deuterated water solution of PS (3-trimethylsilyl sodium propionate-d4) of known concentration to the participating institutions that declared to conduct the quantitative NMR measurement.The measurement results were reported individually to the pilot laboratory. Figure 4 is a summary of the results. Each data point represents a reported result. The adjacent error bar is the measurement uncertainty estimated by each participating institution (95 % confidence interval). The uncertainty of the quantitative NMR results from most institutions was in the range that could be described as percentage, and some of the results deviated significantly from the preparation values. In short, it was found that the quantitative NMR lacked accuracy compared to the conventional analytical methods such as GC. From the result of this international comparison, it was determined that the quantitative NMR did not offer sufficient technical accuracy. This view remains essentially unchallenged in the international scientific community today.At the same time, Fig. 4 shows that the value reported by AIST closely matched the preparation value and its uncertainty was considerably smaller than the quantitative NMR findings of other participating institutions. This is why AIST takes a different stance on quantitative NMR. The uncertainty AIST reported to the pilot laboratory for quantitative NMR in the international comparison is illustrated in Fig. 5. Upon evaluating the relative standard uncertainties of each component, we found that the greatest factor was the uncertainty of the concentration of 1H PS supplied by the pilot laboratory. Because the uncertainty of AIST’s quantitative NMR measurement was much smaller, it became clear that a much smaller measurement uncertainty would have resulted if AIST had supplied its own more accurate PS.It should be emphasized that the quantitative NMR offers a major advance in versatility. Whereas GC and other 1.031.051.071.091.191.211.111.131.151.17mg g-12(NMR)6(GC)9(GC)6(NMR)4(NMR)8(GC)7(GC)1(GC)2(GC)5(GC)3(NMR)8(NMR)10(NMR)7(NMR)5(NMR)1(Titration)0.00070.150.00040.100.0500.00140.0003【Source of uncertainty】Molecular weight (PS)Molecular weight (ethanol)Number of 1H nuclear (PS)Number of 1H nuclear (ethanol)Concentration of PSPreparation and analysisRelative standard uncertainty (%)Fig. 4 Results of international comparison on quantitative analysis of ethanol in aqueous solution.The solid line indicates the preparation value; the dotted line indicates uncertainty for the preparation value. No. 6 is the result for NMIJ/AIST.Participants: BAM (Germany), KRISS (Korea), LGC (UK), LNE (France), NIST (USA), NMi (Netherlands), NMIJ (Japan), NRC (Canada), NRCCRM (China), and VNIM(Russia). Fig. 5 Uncertainty for 1H NMR in the international comparison on quantitative analysis of ethanol in aqueous solution.

※このページを正しく表示するにはFlashPlayer9以上が必要です