BUNSEKI KAGAKU Abstracts

The Japan Society for Analytical Chemistry (JSAC) has developed various certified reference materials (CRMs) since 1993. CRMs include a wide range of materials : incinerator ash, soil, sediment, river water, plastics, industrial materials, foodstuff and so on. Certified components also include a wide variety, such as dioxins, inorganic elements, or nutrients. The characteristics of JSAC’s reference materials are composites, not pure substances nor their solution with a pure solvent. This means that the sample form is a natural material found in usual laboratories. The certifying method consists of preparing analytical samples whose homogeneity is assured at first, interlaboratory comparison testing, statistical evaluation of reports sent from laboratories applying a robust method, and then certification by experts. The harvests of certification work are not only reference materials, but also some new knowledge for analysis, such as gas chromatographic separation of dioxin isomers, and variations of chlorine substitution in PCB during extraction.

The Japan Society for Analytical Chemistry (JSAC) has been providing proficiency testing schemes based on the ISO/IEC Guide 43-1 : 1997 (JIS Q 0043-1 : 1998) in collaboration with Japan Environmental Measurement & Chemical Analysis Association. The purpose of the proficiency testing is to contribute to the development of a laboratory accreditation system in Japan, and to improve the reliability of chemical analysis. The policy, technical fields, evaluation method, etc. were discussed by the JSAC proficiency testing committee ; planning, execution, the publication of reports, etc. are performed by executive committees organized in accordance with the technical fields. This report summarizes the purpose, scheme and results of the proficiency testing executed from 1999 through 2007, and, as typical examples, discusses mainly technical issues found in the several schemes of proficiency testing ; “Analysis of heavy metal elements in soil”, “Analysis of dioxins”, “Analysis of harmful elements in plastics”, “Food analysis”, and “Testing for understanding of traceability and uncertainty”.

A preparation of new powdery reference materials (RM) was developed for X-ray fluorescence (XRF) analysis of trace amount of Cr, Br, Cd and Pb in polyethylene. Polyethylene powder coated with silane coupling agent were disparsed in the methanol containing appropriate amount of CrCl₃, Cd(NO₃)₂, Pb(NO₃)₂ and tetrabromobisphenol-A (TBBPA). The emalsion was dried and homogenized by rotaly evapolater. The resultant polyethylene powder had a homogeneous elemental distribution and excellent durability for X-ray irradiation. XRF intensity of Hg was remarkably decreased with increase of irradiation time. The calibration curves of Cr, Br, Cd and Pb made by developped RM showed a good linearity. The lower limits of detection of the present method were Cr : 3.9, Br : 3.2, Cd : 5.7, Pb : 1.7 mg kg⁻¹. Certified reference materials of plastic (BCR680, 681) were analyzed to verify the proposed polyethylene RM. Analytical results for Cr, Br, Cd and Pb using calibration with proposed RM were in good agreement with the certified and reference values.

ABS resin pellet certified reference materials (CRMs, NMIJ CRM 8102-a, 8103-a, 8112-a and 8113-a) were developed in this study. Both the sample pretreatment procedure and the measurement one for the determinations of Cd, Cr, Hg and Pb in candidate ABS resin pellet RMs were examined, and several analytical methods, including isotope dilution mass spectrometry (a primary method of measurement), were developed. The analytical results obtained by several analytical methods showed good agreement within the uncertainty estimated for each analytical method. Consequently, it is concluded that these analytical methods established for the determinations of Cd, Cr, Hg and Pb in this study are reliable with high accuracy. From these results, the ABS resin CRMs developed in this study are SI traceable and have international consistency.

A candidate reference material of milled rice flour including 0.2 mg/kg dw (dry weight basis) cadmium in the matrix, which is about half as much as the Codex guideline level of 0.4 mg/kg fw (fresh weight basis), was used for proficiency testing. Codex Alimentarius Commission develops international food standards. Cadmium (Cd), sodium, potassium, magnesium, calcium, phosphorus, iron, zinc, copper and manganese (Mn) were analyzed. While cadmium analysis was required for all participants, other element analyses were discretionary. Fifty-seven participants submitted 59 results. Data analysis was conducted according to the harmonized protocol for proficiency testing, revised cooperatively by the international standardizing organizations IUPAC, ISO, and AOAC International (2006). The assigned value of Cd, derived from 57 data after removing outliers, was 0.189 mg/kg dw, and its expanded uncertainty (k＝2) was 0.004 mg/kg dw. This assigned value was not significantly different (P＞0.05) from the cadmium content (0.195 mg/kg dw) with an expanded uncertainty of 0.006 mg/kg dw of the candidate reference material. Relative uncertainties of 0.8 to 2.3% of the assigned values of the elements derived from this proficiency testing data were almost in the same range of relative uncertainties (1.5 to 2.0%) as the contents of the candidate reference material. The ratios of the number of z scores between −2 and ＋2 to that of all scores were 95% for Cd and 97% for Mn. However, those ratios for other elements were less than 90%. Therefore, participants with z scores outside the range from −2 to ＋2, whose methods should be self-investigated and remedied, were found for all 10 elements, especially many participants for other elements, except Cd and Mn.

Specified standard solutions of six kinds of volatile organic compounds (tribromomethane, dibromochloromethane, bromodichloromethane, trans-1,2-dichloroethylene, 1,2-dichloropropane and 1,4-dichlorobenzene) have been developed based on Japan Calibration Service System (JCSS). Two concentration levels of standard solutions for each organic compound, 100 mg L⁻¹ and 1000 mg L⁻¹, were prepared. The concentration of the prepared standard solution was calculated based on weighing the pure organic compound and the solvent, methanol or hexane (gravimetric method). A storage stability test of six months was carried out, where the standard solutions were subdivided to ampoules and stored at 5°C. The concentration change of all the solutions tested for six months was within −0.2〜0.2%. In JCSS, the specified standard solution calibrate the concentration of the secondary standard solution ; thus, the uncertainty of the concentration of a secondary standard solution calibrated by the specified standard solution was evaluated. Concretely, the values of the uncertainty coming from the following five components were counted : 1) the purity of raw materials and solvents, 2) the preparation process of the specified standard solution, 3) subdivision of the specified standard solution to ampoules, 4) the storage stability of the specified standard solution, and 5) GC measurements for calibration of the secondary standard solution. The uncertainties from the above components except for 1), were calculated by an analysis of the variance for all data of the storage stability test ; 0.2〜0.5% of the expanded uncertainties (k＝2) for the calibrated values of all the secondary standard solutions were obtained.

An interlaboratory study was performed to evaluate a method for determination of ash in macaroni products, based on the Japanese Agricultural Standard (JAS), by heating for 6 hours at 550°C in a muffle furnace. Nine laboratories participated, and analyzed five test materials as blind duplicates. A statistical treatment revealed that the repeatability (RSD_r) of ash measurements was less than 1.4%, and the reproducibility (RSD_R) of those of ash was less than 2.3%. The HorRat values (RSD_R/predicted RSD_R) were 0.12 to 0.56. This method was shown to have acceptable precision by the present study.

A chronoamperometric method for obtaining experimental values of diffusion coefficients with high trueness and precision was described. In this method, the chronoamperograms were analyzed by using the most precise theoretical equation available. All of the variables that appear in the equation had been absolutely determined with instruments calibrated traceably. Experiments were carried out so conscientiously that the demands of the theoretical equation would be realized as well as possible. The obtained chronoamperogram was fitted to the theoretical equation by two adjustable parameters : the diffusion coefficient and the concentration of the diffusing spices. Each chronoamperogram was inspected concerning how it fitted the theoretical equation over the observed time range and how the obtained concentration agreed with the actually known one. By this inspection the trueness of the value of the diffusion coefficient was assessed. For an example of showing the importance of the trueness of the experimental values, the diffusion coefficients of fullerene in references were shown along with the values obtained by this method.

Distillation-coulometric iodometric titration was developed for arsenic determination in white metals based on Japanese Industrial Standard (JIS), which is composed of a two-times distillation of arsenic. Since the recovery of arsenic by the first distillation was 86%, analytical procedures were improved, as described. The addition of water to decomposed the sample solution was reduced, and the first distillation was done for 30 min at 105°C of vapor temperature, followed by the addition of HCl and FeSO₄・7H₂O and distilling similarly for 20 min. By using the improved procedures, the recovery was 99.9% and the time required for sample preparation was reduced to half. Separated arsenic(III) was then titrated with electrogenerated iodine. The end points of the titration were detected amperometrically. The coulometric assey of a white metal reference material (certified value 0.088%) was 0.086% with a relative standard deviation of 0.37%.

The Japan Society for Analytical Chemistry has developed plastic certified reference materials (CRMs) for the X-ray fluorescent analysis of bromine in plastics. Disk samples were prepared by filling Al-cylindrical molds with raw material liquid of polyester resin mixed with a hardener and bromine compounds. Thus, 300 candidate reference materials of JSAC 0651-0655 were obtained. Both surfaces on prepared disks obtained after 12 h of hardening at room temperature were ground and polished by a milling machine at 4.00±0.05 mm thick. The homogeneity of prepared disks was shown to be excellent. The certified values were evaluated by a statistical analysis of the results of an interlaboratory comparison study. An interlaboratory comparison study was performed by participants from 19 laboratories. The medians of the results were applied for the certified values in stead of the averages as before, because there was some lower bias in reports from some laboratories.

Soil standards containing hazardous metals (Cr, As, Se, Cd, Hg, and Pb) were developed for X-ray fluorescence analysis (XRF). Standard soil was prepared by adding appropriate amount of aqueous standard solutions to the base soil (brown forest soil ; Fukuoka, Japan) and then drying and homogenizing. The analytical lines were CrKα, AsKα, SeKα, CdKα, HgLα and PbLβ accompanied by the corrections for the overlapping of SeKβ to PbLβ and PbL&alπha; to AsKα. Specimens for XRF analysis were prepared by a powder briquette molded in the 23 mmφ of an Al ring with 300 kgf cm⁻². The homogeneity and durability of the standards were checked by the fluorescent X-ray intensities and the statistical errors for briquette samples. Each of the intensities varied slightly, and the relative standard deviations of the XRF intensities of Cr, As, Se, Cd, and Pb in 10 briquettes were less than 0.3%. The X-ray resistant performance of the soil standards was checked by the repetitive irradiation of 4 kW X-ray. The fluorescence intensities of Cr, As, Se, Cd, and Pb showed constant values for 240 min of irradiation. The HgLα intensity was almost constant for 60 min of irradiation in air. The calibration curves constructed proposed standards showing good linearity under 3000 mg kg⁻¹ for Cr, As, Se, Cd, Pb, and 300 mg kg⁻¹ for Hg. The lower limits of detection were 0.84〜2.7 mg kg⁻¹.

A measurement uncertainty estimation in the determination of ethyl aminobenzoate by diazotization titration and HPLC with UV detection (HPLC-UV) was carried out. Fishbone diagrams for determining the ethyl aminobenzoate purity by diazotization titration and HPLC-UV were prepared for the identification of factors that affect measurement uncertainties. In the case of diazotization titration, combined standard uncertainties for titers of a sodium nitrite solution used in the standardization and determination were larger than that for other factors. In the case of the HPLC, combined standard uncertainties for the chromatographic peak areas of ethyl aminobenzoate in standard and sample solutions were larger than that for other factors. A determination of the ethyl aminobenzoate purity in the bulk powder was also carried out using diazotization titration and HPLC-UV. The results concerning the ethyl aminobenzoate purity by diazotization titration and HPLC-UV were 100.5% (RSD＝0.26%, n＝6) and 100.6% (RSD＝0.36%, n＝6), respectively, indicating that both analytical results were essentially identical. The estimated combined standard uncertainties were within the 95% confidence interval of SD, which were obtained by the above-mentioned repetitive measurements.

Purity analysis using the freezing-point depression method is qualified as a measurement method traceable to the International System of Units. In particular, it is well-known that it is one of the analytical methods for which absolute quantities can also be obtained for high-purity organic compounds. However, the evaluation of commercially available raw materials for standard solutions is normally performed using chromatography, and few examples exist of the application of the freezing-point depression method. In this study, we applied the freezing-point depression method to perform a purity analysis of polychlorinated biphenyls and other organochlorine compounds, using a differential scanning calorimeter. We then evaluated the uncertainties to examine its practical applicability. The measurement results obtained in this study for the compounds used indicated purity levels of between 99.7% and 99.9%. When we assessed uncertainty, the uncertainty for the unit of purity obtained using the freezing-point depression method, after taking into account the uncertainty in conversion from the amount-of-substance fraction to the mass fraction, could be evaluated at a practically viable level of 0.2% to 0.4%. We expect that the uncertainty evaluation method proposed here can be used to perform a quantitative evaluation of the practicality of applying a differential scanning calorimeter to purity analyses of raw materials for standard solutions.

This report is a validation for an iodometry potentiometric titration analysis for determining the 1,2-glycol content in epoxy resins. An analytical method, JIS K 7146, exists for the analysis of 1,2-glycol in epoxy resins. This method is based on manual titration with a color-change end point. However, the manual titration method sometimes has problems when it is used for the analysis of some solid-type epoxy resins (SERs), which are initially colored when a sample is dissolved during the procedure. This is because the resin color makes it difficult to detect the solutions’ color change at the end point. This problem contributes to poor test precision and increased variability in results determined by different analysts. In this study, improvements in the method were achieved by implementing potentiometric titration. This uses an automatic titrator, which is equipped with a data processor, and disposable mayonnaise bottles as the titration vessel. This method is applicable to the analysis of 1,2-glycol in epoxy resins over the range of 0.05% to 7% (wt/wt). The precision was determined by pooling results from thirteen (13) analyses of two different solid epoxy resins. This corresponds to 11 (pooled) degrees of freedom ; [t_(n−1)＝2.201]. The pooled relative standard deviations (RSD_pooled) was calculated as the square root of the pooled relative variance for the two samples : RSD_pooled＝√Relative variance_pooled ; RSD_pooled＝0.45%. The relative prediction interval at the 95% confidence level [±t_(n−1)×RSD_pooled] is ±0.99%. This relates to future final results determined on similar solid epoxy resin samples. The distribution of the results is assumed to be normal. The validity of this assumption was verified using the Shapiro-Wilk test for normality. The test confirmed that the results could originate from a normal distribution. A recovery study, using 1,2-butanediol (reagent grade) and two epoxy resins (E and F), was carried out, and satisfactory results were obtained : 98.7, 97.4 and 97.3%, respectively. A LOQ (limit of quantitation) study was also carried out. The result indicated that it is possible to measure approximately 0.05% (wt/wt) or less 1,2-glycol in an epoxy resin sample.

The Japan Society for Analytical Chemistry carried out a seventh proficiency testing for the determination of dioxins (polychlorodibenzo-p-dioxins, polychlorodibenzofurans and dioxin-like polychlorobiphenyls) in marine sediment. This proficiency testing of analytical laboratories was based on ISO/IEC Guide 43-1, “Proficiency testing by interlaboratory comparisons” The testing was carried out from September ’05 to January ’06, with the participation of 96 laboratories. As a result, 83 laboratories (86.4%) were estimated as “satisfactory”, 9 laboratories (9.4%) as “questionable” and 4 laboratories (4.2%) as “unsatisfactory”.