Forensic Discrimination of Concrete Pieces by Elemental Analysis of Acidsoluble Component with Inductively Coupled Plasma Mass Spectrometry

Transcription

1 ANALYTICAL SCIENCES JUNE 2018, VOL The Japan Society for Analytical Chemistry Forensic Discrimination of Concrete Pieces by Elemental Analysis of Acidsoluble Component with Inductively Coupled Plasma Mass Spectrometry Notes Masaaki KASAMATSU, Takao IGAWA, Shinichi SUZUKI, and Yasuhiro SUZUKI National Research Institute of Police Science, Kashiwanoha, Kashiwa, Chiba , Japan Since fragments of concrete can be evidence of crime, a determination of whether or not they come from the same origin is required. The authors focused on nitric acid-soluble components in the fragments of concrete. As a result of qualitative analysis with ICP-MS, it was confirmed that elements such as Cu, Zn, Rb, Sr, Zr, Ba, La, Ce, Nd, and Pb were contained in the fragments. After the nitric acid-soluble components in the fragments of concrete were separated by dissolving them in nitric acid, the concentrations of these elements in the dissolved solution were quantitatively determined by ICP-MS. The concentration ratios of nine elements compared to La were used as indicators. By comparing these indicators, it was possible to discriminate between the fragments of concrete. Keywords Concrete, elemental analysis, forensic discrimination, ICP-MS (Received January 30, 2018; Accepted April 2, 2018; Published June 10, 2018) Introduction Since concrete is a familiar material used for building structures and block walls, fragments that have occurred in relation to a crime might be physical evidence. Criminal investigations require the evaluation of the similarity between fragmented samples; that is, to determine whether or not they are derived from the same origin. Generally, fragmented samples are investigated for whether there is a region where the physical structure matches. In addition, a morphological observation, such as color and texture, is conducted. These results are used to consider whether or not these fragments are derived from the same origin. However, subjective observation should be eliminated as much as possible because the results may differ depending on the inspector. For these reasons, forensic discrimination is performed based on objective indicators utilizing instrumental analyses, and a method focusing on contained trace elements is effective. In the field of forensic science, discrimination of ceramics, 1 glass, 2 6 and other 7,8 materials are carried out by comparing trace elements contained in the examined physical evidence. Estimation of the chemical composition of concrete is performed by an electron probe microanalyzer, 9,10 but it does not have enough sensitivity for discrimination. Inductively coupled plasma mass spectrometry (ICP-MS) is suitable for trace element analysis based on its high sensitivity, and sequential multi-elemental analysis is widely used for detection of minor and trace elements from petroleum products, food, 14,15 drugs, 16 biological samples, and glass. 2 4,23 However, there are several problems in applying ICP-MS to the analysis of concrete. Glass and metals produced by melting and mixing are relatively homogeneous in elemental component, whereas To whom correspondence should be addressed. kasamatsu@nrips.go.jp concrete is a mixture of cemented heterogeneous aggregates. For this reason, the chemical composition might vary greatly depending on the site from which the sample is collected and is not suitable for analysis by decomposing the whole sample, including the aggregate. In this study, the nitric acid soluble components of concrete were examined, and a solubilized fraction was introduced to ICP-MS. Indices, expressed as the ratios of characteristic elements, were calculated from the obtained results and discrimination was performed based on the difference between intra-sample variation and inter-sample variation of those indices. Experimental Apparatus An ICP-MS instrument (Model 7500c, Agilent Technologies, Tokyo, Japan) was employed for this experiment. The operating conditions for the ICP-MS instrument are given in Table 1. As an internal standard, yttrium was added on-line. Using the ratio of stable isotopes, it was assumed that there was no interference for each measured element. Therefore, cell gas mode was not used. Reagents and samples Purified water was prepared with a Milli-Q preparation system (Elix Advantage 3 and Milli-Q Advantage, Merck Millipore, Tokyo, Japan) throughout this experiment. The nitric acid used for this experiment was ultrapure grade (Kanto Chemical Co. Inc., Tokyo, Japan). Multi-element standard solutions containing or 5 40 ng ml 1 of each element in 2% nitric acid by volume were prepared by mixing single-element standard solutions for atomic absorption spectrometry (Kanto Chemical Co. Inc., Tokyo, Japan) and used as working standard solutions for calibration.

2 730 ANALYTICAL SCIENCES JUNE 2018, VOL. 34 Table 1 Operating conditions for ICP-MS Plasma conditions RF frequency MHz RF power 1.5 kw Plasma gas flow rate 14.8 L min 1 Ar Auxiliary gas flow rate 0.9 L min 1 Ar Carrier gas flow rate 1.1 L min 1 Ar Cell gas None Data acquisition Elements (m/z) Cu(63), Zn(66), Rb(85), Sr(88), Y(89), Zr(90), Ba(137), La(139), Ce(140), Nd(146), Pb(208) Data point 3 points/peak Dwell time 0.3 s Repetition 5 times Table 2 Concrete samples used for the experiments Sample number Condition Source 1 Concrete waste Collect 2 Concrete waste Collect 3 Concrete curbstone Purchase (shop A) 4 Concrete curbstone Purchase (shop A) 5 Concrete block Purchase (shop A) 6 Concrete block Purchase (shop A) 7 Concrete curbstone Purchase (shop B) 8 Concrete curbstone Purchase (shop B) Eight kinds of concrete pieces were collected or purchased randomly from available commercial samples, as shown in Table 2. One dry mortar premixed with cement and aggregate was also purchased and used for this research. Sample preparation A small piece of concrete was chiseled from five different points. Samples collected from each site were individually decomposed. Approximately 100 mg of a concrete piece was placed in a plastic tube; 1 ml of pure water and 1 ml of concentrated nitric acid were sequentially added and left overnight for obtaining an acid soluble fraction. The insoluble aggregate was removed by filtration (qualitative filter paper, ADVANTEC, Tokyo, Japan), and the filtrate that is, the nitric acid soluble component of the concrete was separated. The resulting solution was made up to 50 ml, further diluted if necessary, and qualitative and quantitative analyses were performed by ICP-MS. The insoluble aggregate was approximately 60 to 70% by weight for each concrete piece. Results and Discussion Effect of curing the cement Concrete is prepared by mixing aggregate, cement, and water. The object to be analyzed is the concrete after curing. Here, since the curing of cement is a chemical reaction, there is a possibility that differences may result in the elements detected due to the influence of curing. Therefore, in order to confirm this point, pretreatment and analysis by this method were carried out on the dry mortar sample before and after curing. The obtained spectra are shown in Fig. 1. It should be noted that there was little difference in the shape of the spectra obtained. Fig. 1 Mass spectra of dry mortar (A) before and (B) after curing by ICP-MS. Fig. 2 Analytical results for dry mortar before and after curing. The error bar for each column corresponds to SD. The detected element was quantified and the profile normalized with La concentration, as shown in Fig. 2. Thus, curing cement does not affect the obtained results. Qualitative analysis by ICP-MS In a homogeneous sample such as glass or metal, 24 it is possible to discriminate by sampling a part of the material and completely dissolving and examining the trace elements. However, in heterogeneous samples, such as in the case of

3 ANALYTICAL SCIENCES JUNE 2018, VOL concrete, variation of elemental composition due to the difference of sampling point is too large to get a representative value for discrimination between samples. For this reason, the acid-soluble components in the concrete were selected for analysis. As cement is mixed and agitated with aggregate and water when preparing concrete, the fluid part that includes the cement can be considered close to homogenous in the sample batch. Therefore, it was presumed that discrimination among these samples could be performed by utilizing elemental information of only the cement part of the concrete. The qualitative analysis of the cement acid-soluble fraction confirmed that the fraction contained Cu, Zn, Rb, Sr, Zr, Ba, La, Ce, Nd, and Pb. An example of the result of qualitative analysis Fig. 3 Mass spectrum of sample No. 3 by ICP-MS. by ICP-MS is shown in Fig. 3. Many peaks were observed in each sample in addition to the strong peaks of Sr and Ba. The same kinds of elements were found among different samples in this experiment. Thus, elements that could be used as indicators were selected and quantified. Evaluation and discrimination method In order to calculate the quantitative value from the measured value of the solution, it was premised that the weight of the sample dissolved was known. In this case, part of the sample was dissolved, and, considering the influence of moisture contained in the sample and foaming during dissolution, it was impossible to calculate the weight of only the dissolved part of the sample. However, even if the dilution factor was unknown, the proportion of contained elements should be constant. We surmised that indices could be obtained by standardization with specific elements. La was selected as the normalizing element, which had a high signal intensity and a small deviation in the sample. The result of normalizing each element with La is shown in Fig. 4. Each sample is shown as the mean ± SD of the normalized values. In addition to the measurement error, the error ranges were affected by heterogeneity due to the difference in sampling position; therefore, a large deviation might be observed. However, differences were found in the normalized values between the samples, and discrimination was performed using this value as an indicator. Discrimination between the samples was performed by the following procedure with reference to the previous studies First, a range of the average value ± 2SD was calculated for each sample, and the ranges between two samples were compared. If they did not overlap, a sufficient difference existed, and so it was determined that it could be discriminated. Fig. 4 Analytical results for different concrete samples by ICP-MS. The error bar for each column corresponds to SD.

4 732 ANALYTICAL SCIENCES JUNE 2018, VOL. 34 Fig. 5 Results of discrimination for concrete samples by each discriminating indicator. Discriminated (D), not discriminated (N). If the ranges overlapped, a sufficient difference was not recognized between the examined samples and it was determined that it was difficult to distinguish. Discrimination between the two samples of interest was carried out for all combinations. The results are shown in Fig. 5. The symbol D indicates a discriminable pair determined by the method described above, and the symbol N indicates a non-discriminable pair. For example, the range of the indicator of Cu for samples 1 and 2 did not overlap, so they could be distinguished. Also, as the range of Ba for sample 8 was very large, it could not be discriminated in any sample. In the case of using a single indicator, there were many combinations that could not be distinguished, and the discrimination power between samples was insufficient. Through considering the discrimination between samples, the discrimination power could be improved by combining the nine indicators. If there was more than one indicator that did not overlap in the range of 2 SD, they were judged as discriminable. The results are shown in Fig. 6. Out of 28 combinations, 26 pairs (about 93%) could be discriminated. In this way, by combining indices, the discrimination power was significantly improved. There were two indistinguishable combinations, but these might have been made from the same lot of raw materials. It was difficult to find differences in these products purchased from the same store because many commercially available concrete blocks might be produced at the same time. In fact, when analyzing three concrete blocks obtained from the same shelf, the samples were not distinguishable from each other (data not shown). This result not only shows that this method could distinguish different samples, but also that the samples from the same origin could be the same. Fig. 6 Results of discrimination for concrete samples using nine discriminating indicators. Discriminated (D), not discriminated (N). Conclusions We focused on acid-soluble components that were part of concrete pieces and found that forensic discrimination could be performed by comparing their combination of characteristic ratios. In order to minimize the influence of the aggregate, we did not completely dissolve the sample and targeted only the acid-soluble component cement. Since cement could be considered close to homogeneous when making concrete, focusing on this part was appropriate for forensic discrimination. Because this is a comparison of ratio combinations, it is an excellent method that can be easily pretreated without concern for the dilution ratio of the sample.

5 ANALYTICAL SCIENCES JUNE 2018, VOL References 1. M. Kasamatsu and Y. Suzuki, Jpn. J. Forensic Sci. Tech., 2016, 21, Y. Suzuki, M. Kasamatsu, R. Sugita, H. Ohta, S. Suzuki, and Y. Marumo, Bunseki Kagaku, 2003, 52, T. Trejos, S. Montero, and J. R. Almirall, Anal. Bioanal. Chem., 2003, 376, S. Berends-Montero, W. Wiarda, P. Joode, and G. Peijl, J. Anal. At. Spectrom., 2006, 21, C. M. Bridge, J. Powell, K. L. Steele, and M. E. Sigman, Spectrochim. Acta, Part B, 2007, 62, M. M. El-Deftar, N. Speers, S. Eggins, S. Foster, J. Robertson, and C. Lennard, Forensic Sci. Int., 2014, 241, H. Ohsaki, K. Murakami, M. Kasamatsu, Y. Suzuki, and S. Suzuki, Jpn. J. Forensic Sci. Tech., 2009, 14, Y. Suzuki and S. Suzuki, Bunseki Kagaku, 2012, 61, D. Sawaki, K. Kobayashi, and E. Sakai, Bunseki Kagaku, 2010, 59, D. Sawaki, K. Kobayashi, and E. Sakai, Bunseki Kagaku, 2010, 59, K. Takeda and Y. Arikawa, Bunseki Kagaku, 2005, 54, R. J. Cassella, D. M. Brum, N. F. Robaina, A. A. Rocha, and C. F. Lima, J. Anal. At. Spectrom., 2012, 27, R. Sánchez, C. Sánchez, J. L. Todolí, C. P. Lienemann, and J. M. Mermet, J. Anal. At. Spectrom., 2014, 29, L. R. Drennan-Harris, S. Wongwilawan, and J. F. Tyson, J. Anal. At. Spectrom., 2013, 28, S. H. Choi, J. S. Kim, J. Y. Lee, J. S. Jeon, J. W. Kim, R. E. Russo, J. Gonzalez, J. H. Yoo, K. S. Kim, J. S. Yang, and K. S. Park, J. Anal. At. Spectrom., 2014, 29, C. Liu, Z. Hua, Y. Bai, and Y. Liu, Forensic Sci. Int., 2014, 239, I. Rodushkin and M. D. Axelsson, Sci. Total Environ., 2000, 250, J. P. Goulle, L. Mahieu, J. Castermant, N. Neveu, L. Bonneau, G. Laine, D. Bouige, and C. Lacroix, Forensic Sci. Int., 2005, 153, M. W. F. Bonnie, S. S. Tak, S. K. L. Joseph, and S. Tam, J. Anal. Toxicol., 2007, 31, V. Hollriegl, M. Gonzalez-Estecha, E. M. Trasobares, A. Giussani, U. Oeh, M. A. Herraiz, and B. Michalke, J. Trace Elem. Med. Biol., 2010, 24, M. Kasamatsu, Y. Suzuki, S. Suzuki, W. B. Siong, Y. Oura, and M. Ebihara, Jpn. J. Forensic Sci. Tech., 2014, 19, M. Aramendia, L. Rello, S. Berail, A. Donnard, C. Pecheyran, and M. Resano, J. Anal. At. Spectrom., 2015, 30, T. Trejos and J. R. Almirall, Talanta, 2005, 67, H. Ohsaki, M. Kasamatsu, Y. Suzuki, and S. Suzuki, Bunseki Kagaku, 2007, 56, D. A. Hickman, Forensic Sci. Int., 1983, 23, M. Kasamatsu and Y. Suzuki, Bunseki Kagaku, 2012, 61, C. Munger, K. M. Gates, and C. Hamburg, J. Forensic Sci., 2014, 59, 1351.