**3. Identification of arsenic oxide powders**

The curry was cooked in two pots in a garage of a festival organizer's house. The curry was cooked from noon till 3 pm there, and then moved to the festival venue. During the noon and 3 pm, the curry pots were kept boiling by housewives of the organizers in turn. One of the housewives there, Mrs. H, was arrested on October 4 and prosecuted as the murder on December 29. She was sentenced to death on December 11, 2001, at the Local Court of Wa‐ kayama. Then again she was sentenced to death at Osaka Court of Appeal. Finally, May 18, 2009, the death penalty has been fixed at the Supreme Court of Japan. She has denied from the

The only evidence was a paper cup found near the cooking site. This paper cup might have been used to poison the curry pot bringing arsenic. Powder of about 35 mg arsenic oxide, As2O3, was left inside of the paper cup. Her husband had arsenic oxide powders as white ant pesticide, as his job was white ant exterminator. Therefore the key forensic analysis was the identification of arsenic oxide powders between her husband's and the powder adsorbed on the inner surface of the paper cup. "High concentration arsenic" was found on one of her hairs, which was one of the several hundreds of hairs cut on December 9, 1998, by the police. These two evidences are the main reasons of her death penalty. The hair was analyzed by synchrotron radiation X-ray fluorescence (SR-XRF) and also by atomic absorption spectrometry (AAS). Several impurity elements in the arsenic oxide powders were analyzed by the SR-XRF and inductively-coupled plasma atomic emission spectrometry (ICP-AES), as was reviewed by Kawai [2]. Infrared (IR), ion chromatography/inductively-coupled plasma mass spectrometry (IC/ICP-MS), X-ray diffraction (XRD), scanning electron microscope-energy dispersive X-ray

analysis (SEM–EDX), and many other chemical analysis techniques were used.

prosecutor, and again this murder case becomes discussed in Japan.

Wakayama case are as follows.

Because the chemical poison was used by the Tokyo subway sarin attack in 1995, the Wakaya‐ ma arsenic case attracted large attention by mass media, such as television, newspapers, and gossip magazines, at that time for about 1 year duration. The forensic analyses were performed mainly by the National Research Institute of Police Science, Tokyo University of Science, St. Marianna University School of Medicine, Osaka Electro-Communication University, and Hiroshima University. It was well known to the public at that time that SPring-8, one of the third generation synchrotron radiation facility, a 1.5 km circumference accelerator ring of 8 GeV, was used for the forensic analysis. The forensic analysis of SPring-8 was just 1 year after it became in use. Since 2012, Kawai, the author of the present paper, found many faults in the forensic analyses in this case, of which documents were submitted to the court from the

**2. Influence on the academic researches of Wakayama arsenic murder case**

The sarin attack at Tokyo subway was just a few years before this arsenic murder case. Therefore, many academic research papers on Wakayama murder case were published, which discussed the relation of the subway sarin attack. Some examples of papers related to the

first until now, but she is now in the death row.

28 Forensic Analysis - From Death to Justice

There were eight kinds of arsenic oxide evidences. Mr. M, who was a brother of Mrs. H, kept arsenic oxide powders, which were originally used as the white ant pesticide by H's husband, long before the incidence. The evidences were as follows: (1) Paper cup, (2) M's green 50-kg can, (3) M's milk can, (4) M's white can marked "Heavy", (5) M's brown Tupperware, (6) A milk can found at H's old house (This house was at that time Mr. T's house, and we call this "T's milk can"), (7) A plastic container found at H's kitchen, but a few particles of arsenic powders were attached on the inner surface of the container, and (8) arsenic oxide crystals found in curry pot. These are tabulated in **Tables 1** and **2**.


**Table 1.** Arsenic oxide powder evidences.


NA: Not available.

NA: Not available.

+: Positive.

0(+): Three tests not detected, one test detected, out of 4-time tests.

+(0): Three tests detected, one test not detected.

**Table 2.** Light element concentrations of arsenic oxide powder evidences (ppm).

The elemental concentrations analyzed using ICP-AES by the National Research Institute of Police Science are shown in **Table 1**. The judge wrote the death sentence by describing that one of the arsenics powders from evidences (2)–(7) was brought by H using the paper cup and put into the curry pot. However it is strange that the arsenic oxide powder concentration of the paper cup was 98.7 wt%, but evidences (4)–(6) were significantly lower than the paper cup. Finger prints were not found on the paper cup. It was known from testimony that the H's husband bought the 50-kg green can (2) more than 10 years before, and distributed into several small cans, (3)–(6). The prosecutor guessed that H brought the arsenic powder from her old house (6) using the container (7) to her new house kitchen at house-moving. Then she brought the arsenic powder using the paper cup (1) to the curry pot.

can, (3) M's milk can, (4) M's white can marked "Heavy", (5) M's brown Tupperware, (6) A milk can found at H's old house (This house was at that time Mr. T's house, and we call this "T's milk can"), (7) A plastic container found at H's kitchen, but a few particles of arsenic powders were attached on the inner surface of the container, and (8) arsenic oxide crystals

**No. Evidence meaning As wt% [34, 35] As2O3 wt% [36]**

(7) H's kitchen container Not available Not available (8) Curry pot crystal Not available Not available

**No. Evidence meaning Na Mg Al P Ca Fe Zn Ba Starch** (1) Paper cup 393 16 138 7 79 146 297 5 0 (2) M's green 50-kg can 35 6 0 5 3 36 203 0 0 (3) M's milk can 32 5 0 5 6 28 201 0 0 (4) M's "Heavy" 59 105 308 85 3965 303 178 2 + (5) M's brown tupper 70 49 170 86 147 861 205 21 0(+) (6) T's milk can 87 203 2266 234 >1% 153 124 7 +(0) (7) H's kitchen container NA NA NA NA NA NA NA 24–36 0 (8) Curry pot crystal NA NA NA NA NA NA NA 23 NA

The elemental concentrations analyzed using ICP-AES by the National Research Institute of Police Science are shown in **Table 1**. The judge wrote the death sentence by describing that one of the arsenics powders from evidences (2)–(7) was brought by H using the paper cup and put into the curry pot. However it is strange that the arsenic oxide powder concentration of the paper cup was 98.7 wt%, but evidences (4)–(6) were significantly lower than the paper cup. Finger prints were not found on the paper cup. It was known from testimony that the H's

(1) Paper cup 74.80 98.7 (2) M's green 50-kg can 77.0±3.4 101.6 (3) M's milk can 77.6±4.0 102.4 (4) M's white can, "Heavy" 68.6±2.2 90.6 (5) M's brown tupper 65.7±1.6 86.7 (6) T's milk can 48.7±0.8 64.3

found in curry pot. These are tabulated in **Tables 1** and **2**.

**Table 1.** Arsenic oxide powder evidences.

30 Forensic Analysis - From Death to Justice

0(+): Three tests not detected, one test detected, out of 4-time tests.

**Table 2.** Light element concentrations of arsenic oxide powder evidences (ppm).

+(0): Three tests detected, one test not detected.

NA: Not available. NA: Not available. +: Positive.

The National Research Institute of Police Science also analyzed five impurity elements, Se, Sn, Sb, Pb, and Bi, in (1)–(6). The chemical properties of these elements were quite similar to that of As, that is to say, Se is neighbor to As in the periodic table, Sb and Bi are in the same column, Sn and Pb are neighbors of Sb and Bi respectively, and thus they co-existed from earth crust. The concentrations of these elements were plotted as "radar chart" as shown in **Figure 1**, by the National Research Institute of Police Science (NRIPS). Similar radar charts of arsenic oxide powders from different industries are shown in Refs. [22] and [23]. However I found that the concentration ratios of Se/As, Sn/As, Sb/As, Pb/As, and Bi/As were multiplied by 1,000,000, then the logarithms were calculated and the radar chart was plotted by NRIPS. The pentagon radar charts of six evidences well overlap each other and looks like the arsenic oxide powders were identical as shown in **Figure 1**. However, both the 1,000,000 times and the logarithm are unfolded, and the radar chart is replotted including As concentration [37], then the hexagon of the paper cup (1) is significantly different from those of H's arsenic powders (2)–(6), as is shown in **Figure 2**.

**Figure 1.** Radar chart of six evidences (1)–(6) taken from a document of the National Research Institute of Police Sci‐ ence, Japan. The document is a public document, not a copyrighted matter.

The hexagon radar chart is interpreted as follows.

**1.** The regular hexagons mean the same root as M's green 50-kg can (2).


The root or origin of the paper cup (1) is significantly different from M's green can (2). The multiplication of 1,000,000 and logarithmic calculation was in order to hide this truth.

**Figure 2.** Unfolded radar chart taken from Ueba and Kawai [37] with permission.

The legal document of the National Research Institute of Police Science concluded that the arsenic oxide in the paper cup (1) and H's arsenic oxide powders were less than 50% identi‐ cal [38]. The small difference between paper cup (1) and H's arsenic shown in **Figure 2** could not be recognized by the XRF spectra of SPring-8. Two representative SPring-8 XRF spectra measured by the Tokyo University of Science are shown in **Figures 3** and **4** [39], from which one cannot recognize the difference of the root. Though it has been revealed that spectra were measured only once for most of the evidences [40], the Tokyo University of Science conclud‐ ed [39] that paper cup (1) was 100% identical to the H's arsenic oxide powders (2)–(7) in **Table 1**. The details of the SR-XRF method at SPring-8 was reported in Refs. [27, 28], and it is found from these papers that the precision of the SR-XRF quantitative analysis was not high enough for the present forensic analysis, and thus the discrimination was not possible. The SR-XRF analysis conclusion was a false conclusion forced by the prosecutor [37].

**2.** Large regular hexagon means that the arsenic powder was more diluted. H's husband

**3.** A slight distortion from the regular hexagon means the error of quantitative analysis as well as inhomogeneity of arsenic oxide in a can. Evidences (2)–(6) were sampled five times and then analyzed five times. The As concentrations of evidences (2)–(6) in **Table 1** are

**4.** Significantly distorted hexagon means different roots, such as the paper cup (1). Once, one of the Bi concentration data of M's "Heavy" (4) had a small analytical error in its concentration out of five data of NRIPS, and the hexagon was distorted. Such one-time error in five measurements can be detected by the distortion: very sensitive to different

The root or origin of the paper cup (1) is significantly different from M's green can (2). The multiplication of 1,000,000 and logarithmic calculation was in order to hide this truth.

The legal document of the National Research Institute of Police Science concluded that the arsenic oxide in the paper cup (1) and H's arsenic oxide powders were less than 50% identi‐ cal [38]. The small difference between paper cup (1) and H's arsenic shown in **Figure 2** could not be recognized by the XRF spectra of SPring-8. Two representative SPring-8 XRF spectra measured by the Tokyo University of Science are shown in **Figures 3** and **4** [39], from which one cannot recognize the difference of the root. Though it has been revealed that spectra were measured only once for most of the evidences [40], the Tokyo University of Science conclud‐ ed [39] that paper cup (1) was 100% identical to the H's arsenic oxide powders (2)–(7) in

used diluted arsenic oxide powders for white ant pesticide.

**Figure 2.** Unfolded radar chart taken from Ueba and Kawai [37] with permission.

root.

32 Forensic Analysis - From Death to Justice

displayed as averages ± standard deviations of five time analyses.

**Figure 3.** SR-XRF spectrum of paper cup (1), taken from the legal document of the Tokyo University of Science [39]. The document is a public document, not a copyrighted matter.

**Figure 4.** SR-XRF spectrum of M's green 50-kg can (2), taken from the legal document of the Tokyo University of Sci‐ ence [39]. The document is a public document, not a copyrighted matter.
