Lethal nerve gas attacks in the city of Matsumoto in 1994, and in the Tokyo subway system in 1995, led to the deaths of 19 people, as well as to a large number of injuries. These attacks caused great shock, in that they constituted an illegal use of chemical warfare agents against a defenceless public. These acts of terrorism were carried out by the members of Aum Shinrikyo, the Japanese doomsday cult.The Japanese police had to cope with a new type of organised crime.
The Chemistry Section of the National Research Institute of Police Science (NRIPS) has been engaged in forensic investigations of both incidents. The NRIPS is an institute attached to the National Police Agency, and is engaged in research and development as it relates to police activity, in the identification and analysis of criminal evidence, as well as in the training of scientists in local forensic science laboratories.This article describes the Sarin gas attacks, and outlines the subsequent related forensic efforts which contributed both to the understanding of how to manage the consequences of chemical terrorism, and also to the entire criminal investigation,which culminated in trials in a court of law.There is also a brief discussion of how police efforts involved cooperation on both domestic and international levels.
Sarin Gas Attacks and Aum Shinrikyo
Aum Shinrikyo was established by Shoko Asahara, and it disseminated a unique doctrine. It taught that one could kill another human being who continues to commit evil deeds, and is thus destined to go to hell. This killing benefits both the killers and the one who is killed. The leader Asahara prophesied the coming of Armageddon. After failing to win support in the 1989 general election, the cult transformed itself into a terrorist group that produced arms and toxic gases. While expanding its influence in Japan, Aum also set up branch offices overseas in the United States, Germany and Sri Lanka. In Russia, it is said that over 30,000 people joined the cult. Aum secretly plotted to buy arms and drugs, and sent its members for training in shooting. Aum also opened affiliated business firms in Sri Lanka and Taiwan. In Australia, the cult purchased a farm, manufactured chemicals, and tested the chemicals on sheep. The South Australian Forensic Science Laboratory detected Sarin hydrolysis products from soils and sheep wool, and informed the Japanese police about the results of their detection. The nerve gas attacks occurred within a one-year period, from 1994 to 1995. In addition to the Matsumoto and Tokyo subway Sarin incidents, there were several other antipersonnel incidents of terrorism. The interrogation of the arrested suspects revealed that they had used VX, Sarin, and hydrogen cyanide in their murders and in their attempted murders, and forensic examinations support these conclusions. In 1993, the leader Asahara directed his cult members to begin the mass production of Sarin, and in November 1993, they had succeeded in synthesising it. In a trial run, the cult successfully sprayed Sarin gas in Matsumoto, and so a chemical plant for Sarin production was started in the autumn of 1994. The cult also succeeded in synthesising a VX-agent. However, police investigators suspected that the Matsumoto Sarin incident had been committed by the cult.Aware that the police were conducting a criminal investigation, the cult stopped running its chemical plants in early 1995, and tried to hide the evidence.
The act of nerve gas terrorism occurred in a quiet residential area in the city of Matsumoto in Nagano. At 11:30 p.m., the Matsumoto police station received an urgent report from the ambulance team of the city Fire Defence Bureau, informing them that some patients had been transported to the hospital. The police immediately began to rescue the injured, and they began to conduct investigations. Many injured people were suffering from ocular pain, and from numbness of the hands.The next day, dead fish and crayfish were found in a pond near the scene. The carcasses of dogs, sparrows, a dove, and a large number of caterpillars were found under some trees. Trees and grass on the scene were observed withering, and the colour of the leaves had changed. Nearly all the casualties were discovered in a sector-shaped residential area, within a radius of 150 metres from the centre, near the pond. People near open windows, or in air-conditioned rooms, were severely exposed to toxins. Some victims had seen a fog with a pungent and irritating smell floating by very slowly. But the cause of the deaths and the poisonings remained a mystery for the first two days. Five dead residents were discovered in their apartments, and two died in the hospital immediately after admission. A total of 274 people were treated in the hospital. Typical symptoms included darkened vision, ocular pain, nausea, miosis, and a decrease in serum cholinesterase (ChE) activities. Autopsy findings showed intense post-mortem lividity, miosis, pulmonary edema, increased bronchial secretion, and congestion of the parenchymatous organs. The Nagano police investigation headquarters assumed that the causative toxic gases originated near the pond, and the police staff of the Identification Section performed an on-site inspection to gather evidence samples.The cult members had sprayed the Sarin at 10:40 p.m., and the Matsumoto police station was informed fifty minutes later. A routine analysis of on-site samples by the Nagano Forensic Science Laboratory (FSL) failed to find any toxic substances. Three days later, the
Nagano FSL and the Nagano Public Health Institute detected Sarin in pond water samples, and the NRIPS confirmed this finding. One week later, police investigation headquarters announced the results to the press. We performed a toxicological analysis of the crime scene evidence samples sent from the Nagano FSL.The forensic elucidation of the causative substance ‘Sarin’ crucially helped the Criminal Investigation Section in their search for possible suspects.
The act of terrorism involving Sarin occurred in the center of metropolitan Tokyo two months after the great Hanshin Awaji earthquake. Inside the trains on three subway lines during the peak of the morning rush hour, on March 20th 1995, a large number of passengers and station personnel suffered toxic effects from an unknown gas. Some people escaped from the trains themselves, and went to the hospital, but many victims were in need of immediate medical treatment. The patients complained of ocular pain and vomiting. So many patients rushed to nearby hospitals, that the hospital staff panicked and could not adequately cope with dispensing medical care.Twelve passengers and station personnel were killed, and approximately 5000 people were injured. Typical symptoms included darkened vision, ocular pain, nausea, miosis, hyperaemia, and nosebleeds. The autopsy findings were almost the same as those in the Matsumoto Sarin incident. In cooperation with the Metropolitan Fire Department, the Metropolitan Police Department (MPD) immediately started to carry out rescue operations, to conduct traffic control, and to investigate the incidents. Wearing protective suits and gas masks, police investigators proceeded to the dangerous crime scene. On-site evidence samples, such as containers, newspapers, and other remains were collected, and immediately transported to the FSL of the MPD.The liquid had been released from plastic bags at passengers, who were exposed to the toxic gas. Sarin was released separately in five subway cars at 8 a.m. Early on, the first responses from the Metropolitan Fire Department mistakenly announced that acetonitrile had been detected. At 10 a.m., the Tokyo FSL, using laboratory gas chromatography/mass spectrometry (GC/MS), detected Sarin in a crime scene sample.This speedy result was possible because of the prior experience, which had been acquired during the Matsumoto Sarin incident.At 11 a.m., the Criminal Investigation Department announced the detection of Sarin to the press.The Defence Agency quickly responded to this incident, and helped the police to perform an on-site inspection and to carry out a decontamination of any poisoned sites.
For the previous ten years, the cult had started to construct facilities in a rural area of Yamanashi refecture near Mt. Fuji. In July 1994, just after the Matsumoto Sarin incident, inhabitants in the vicinity of a cult facility reported foul smells.A criminal investigation team from the Matsumoto Sarin incident analysed the chemical process for the synthesis of Sarin, and found that dummy companies operating on behalf of the cult had purchased large quantities of chemical raw materials. In addition, a Sarin hydrolysis product was detected and removed for forensic investigation from the soils taken near the site of the foul smell. The criminal investigators then began to strongly suspect the cult. Police secretly searched cult facilities, but before the search took place, the cult carried out the Tokyo subway Sarin gas attack. Two days after the Sarin incident, a simultaneous raid of Aum facilities was launched by 2500 police in connection with the imprisonment of a notary public manager who had died while being detained.
To help cope with the Sarin release, the Defence Agency provided the police investigators with anti- toxic gas suits and masks.The search uncovered a large amount of chemicals, including phosphorus trichloride, as well as a chemical plant in the No.7 Satian building. Outside the building, there were many pipes connected to the plants; inside the building, various types of equipment needed for manufacturing chemicals were found.The criminal investigation led the police to strongly suspect that Sarin was being produced.The police identification section gathered evidence samples from various reaction bottles.The NRIPS conducted a chemical examination of the on-site samples, and thus was involved in all the various aspects of this Sarin manufacturing case. Almost all the cult perpetrators were arrested within the next two months.
The special nature of nerve gases slowed down police efforts in connection with the Sarin incidents. First, nerve gases are subject to the law on the Ban of Chemical Weapons and the Regulation of Specific Substances (in force since 1995). Second, nerve gases are highly toxic, and are lethal in trace amounts, making them difficult to detect; they are also difficult to detect because they evaporate very rapidly. Third, they are volatile, and the handling of samples requires special caution. Fourth, nerve gases decompose easily in a body and in the environment. It is necessary to identify the degradation products. Sarin, soman, tabun and VX are representatives of organophosphorus types of chemical warfare agents. They invade the body through the skin, and are extremely toxic, given that they strongly inhibit cholinesterases. The physicochemical properties of nerve gases vary. Sarin is volatile and easily hydrolysed in water, whereas VX is not volatile and is rather resistant to hydrolysis. Because of their extremely lethal nature, it is sometimes impossible to detect nerve gases or their hydrolysis products in a victim’s samples.Therefore, a decrease in blood cholinesterase activity is a good index of exposure to nerve gases.
Nerve gases can be analysed by GC/MS. Retention indices are useful indicators in identifying chemical warfare agents. Wiped samples and soils are extracted with both dichloromethane and water at neutral pH. Blood samples are deproteinised with perchloric acid, and the resulting supernatants are adjusted to neutrality, and extracted with dichloromethane. Nerve gases, synthetic intermediates and by-products are extracted into the organic solvent fraction, and after concentration under mild conditions, the residue is analysed by GC/MS using an apolar capillary column DB-5, and a multi-step temperature programme starting at 45°C. The hydrolysis products of nerve gases are extracted into the aqueous fraction, and after derivatisation, are analysed by GC/MS.
In water, nerve gases are easily hydrolysed and produce characteristic methyl phosphorus compounds that are metabolically stable, and water-soluble. These alkylmethylphosphonates are specific for the original nerve gases. These are finally hydrolysed to methylphosphonate (MPA). Therefore, detecting the hydrolysis products gives indirect proof of the presence of a nerve gas.Ter tiarybutyldimethylsilylation is adopted to conver t the hydrolysis products to volatile derivatives, because its derivatisation efficiency is good, and the resulting derivatives are stable.The aqueous fraction is taken to dryness,and N-methyltert-butyldimethylsilyltrifluoroacetamide and acetonitrile are added, heated at 60°C for 1 hour, and injected into GC/MS.Two kinds of GC detection are used. Mass spectrometry using electron impact ionisation (EI) and isobutane chemical ionisation (CI) provide acceptable structural and chemical information.
Atomic emission detection is selective in monitoring particular elements. Molecules in the cavity are degraded to their elements, and the atomic emissions are detected under selected wavelength.Their retention indices are distinctive; EI-mass spectra provide representative fragmentation patterns, and CI-mass spectra give clear quasi-molecular ions.
Using these established forensic toxicological techniques, we performed a forensic investigation on the evidence samples from the Sarin incidents. In the Matsumoto Sarin incident, Sarin gas was sprayed to murder the local court judges, but nearby residents also became exposed to toxins. The cholinesterase (ChE) activities of all of the fatal casualties were greatly decreased compared to the control blood levels. If Sarin were to be sprayed over a garden, the Sarin left on the leaves would undergo hydrolysis, producing hydrofluoric acid, and under these strong acidic conditions, the plant leaves would wither.We examined the levels of fluoride ions in aqueous extracts of the leaves by capillary electrophoresis. Background levels were detected after we analysed control leaves from a similar area. Extraordinarily high levels of fluoride and chloride were detected in the withered leaves. At that time, we were not able to prepare authentic Sarin. However, we could confirm the existence of Sarin by the four criteria of a retention index, an electron impact (EI) mass spectrum, a CI mass spectrum, and the presence of phosphorus. From the organic solvent extract of the pond water, a Sarin peak was detected on an extracted ion chromatograph of m/z 99, with a very similar retention index to values mentioned in reference papers.The EI-mass spectrum was identical to that of Sarin (described in the literature and in reference papers).The CI-mass spectrum yielded quasimolecular ions of m/z of 141.An atomic emission monitoring with a phosphorus emission line, at the position of m/z 99 peak in the EI-MS resulted in the observation of a phosphorus peak. Sarin was also identified in the nasal mucosa of one victim. The Nagano FSL and PHI also identified Sarin from other on-site samples of water, soil, air, and from wiped samples from the scene. From the tertbutyldimethylsilyl (TBDMS) derivatised solution of various evidence samples, two peaks appeared on extracted ion chromatograms of m/z 153 and 267, with identical retention times, and the mass spectra of the reference TBDMS isopropylmethylphosphonate (IMPA) and MPA. We have not presented quantification data for the levels of IMPA and MPA in the blood.However, the lowest detectable blood levels under GC/MS analytical conditions are 0.5 _g/ml for MPA and 1 _g/ml for IMPA. IMPA and MPA were detected from almost all the victims’ blood, as well as from the blood taken from a dead dog. Diisopropylmethylphosphonate was also identified in the blood of nearly all the fatalities. The perpetrators sprayed Sarin from evaporator-type spray containers in the car park, and the released fog-like gas moved and spread, causing many residents to be poisoned. The spraying device was made from refrigerator-car equipment, and had a heating-pot and a fan.The Sarin solution used in the Matsumoto Sarin incident was synthesised by the following procedure. Isopropyl alcohol was added to a mixture of methylphosphonyl difluoride and methylphosphonyl dichloride. But, because the choice of reaction conditions was poor, the synthetic yield was low, and reaction byproducts were produced.After storing the solution for almost 3 months, they sprayed the crude solution by means of evaporation, which was brought about by ten minutes of forced heating. High levels of Sarin hydrolysis products were detected, even from the injured people. This fact suggests that the crude Sarin solution was composed, not only of Sarin, but of other synthesis precursors and byproducts as well.
In the Tokyo subway Sarin incident, the cult decided to use Sarin in trains on three subway lines, all of which stop at Kasumigaseki station near the MPD. Sarin was released in the locations.The blood samples taken at the autopsies were sent to our laboratory. A scattered distribution pattern of blood plasma butyrylcholinesterase (BuChE) and red blood cell acetylcholinesterase (AChE) activities were observed in 11 of the fatalities. In only six victims were both ChE activities significantly decreased.They were admitted to the hospital in a state of cardiopulmonary arrest, and died immediately.
Compared to the ChE activities in the Matsumoto Sarin incident, the degree of the decrease in activity was not significant, and there were some victims in whom blood activity levels were not decreased at all. Some victims were admitted to the hospital alive, but despite the fact that extensive medical procedures were performed, including the administration of the antidote PAM, they died from brain damage at an early stage. The rather high levels of ChE activities might have been due to the early discovery of patients, which then led to rapid medical treatment, and the blood ChE activity may have been restored by the reactivation mechanism. In the Tokyo subway Sarin gas attack, Sarin-containing plastic bags were used for the indiscriminate murder of defenceless people. The perpetrators boarded the subway trains with plastic bags containing Sarin, and released the gas by prodding the bags open with the metal tips of umbrellas.
At that time, we had no authentic Sarin. Therefore, we measured the Sarin content in the evidence sample, by quantifying the hydrolysed Sarin product, IMPA, by means of capillary electrophoresis. From the analytical data of IMPA in the hydrolysed solution, the Sarin content was deduced to be about 30%. n-Hexane and N, N-diethylaniline (DEA) were also identified as major components.IMPA was detected from blood samples taken from only two of the victims. Considering the detection limit in TBDMS GC/MS analysis, blood levels of IMPA were below 1 ppm, except in the case of 2 victims.TNO Prins Maurits Laboratory in the Netherlands, using sophisticated retrospective analytical methods, detected Sarin-related compounds in the Japanese victims’ blood samples.The Sarin solution used in Tokyo subway gas attack was synthesised by the following procedure. Isopropyl alcohol was added to 1.4 kg of methylphosphonyl difluoride, using n-hexane as a solvent, and DEA as an acid neutraliser. The resulting solution, about 7 litres, was divided into 11 bags. Our analytical result, of about a 30% Sarin oncentration, supports the testimony of the suspects who were arrested.
In investigating the suspicions as to how the Sarin was manufactured, our laboratory at the NRIPS and the FSL of the Tokyo MPD performed forensic investigations on hundreds of evidence samples taken from the crime scene of the suspected Sarin manufacturing location. When the police investigated the documents seized in the cult office, they discovered the synthetic route to the mass manufacture of Sarin. The process involved 5 steps. In the first step, phosphorus trichloride was reacted with methanol to produce trimethylphosphite. In the second step, trimethylphosphite was converted to imethylmethylphosphonate (DMMP) through a rearrangement caused by the application of heat. In the third step, DMMP was reacted with phosphorus pentachloride by applying heat, to produce methylphosphonyl dichloride. In the fourth step, methylphosphonyl dichloride was reacted with sodium fluoride to produce methylphosphonyl difluoride. In the final step, methylphosphonyl difluoride and methylphosphonyl dichloride were mixed with isopropyl alcohol to produce Sarin. Under the restricted conditions used by the cult members, a forensic investigation was carried out. From the wiped samples taken from the first-step equipment, trimethylphosphate, n-hexane and DEA were detected. From the second-step equipment, trimethylphosphate,DMMP, iodine and DEA were detected. From the third-step equipment, MPA, DEA, phosphorus oxychloride and sodium chloride were detected. From the fourth-step equipment, MPA, DEA, sodium chloride and sodium fluoride were detected. From the final-step equipment, IMPA, MPA, DEA, DMMP and sodium chloride were detected. From the chemical analysis of evidence samples taken from the manufacturing plant, only stable substances corresponding to the synthetic routes have been identified, but these serve to verify the synthesis of Sarin in the Aum plant facility.
The GC/MS analysis of Sarin hydrolysis products after TBDMS derivatisation was effective for confirming Sarin exposure, usage and production, and was used as definitive evidence in the court trial.The early detection of Sarin also contributed to the speedy criminal investigation. To save the lives of severely affected victims, early and fast aid, such as cardiopulmonary resuscitation, may be important. In the Sarin poisoning, red blood cell AChE was inhibited more strongly than plasma BuChE. Measuring only serum BuChE is somewhat risky, in that it can lead to a misjudgement of poisoning. The accurate identification of causative toxic substances and the verification of poisoning should be anticipated as providing evidence for the court. In addition, we must be involved in managing the consequences when an act of chemical terrorism occurs.The rapid detection of causative toxic substances is necessary, both in forensic laboratories and sometimes, if possible, at the crime scene. As specialists,we must make our contribution to helping the police investigators and other first responders by providing scientific and technical information about chemical warfare.
Published in Synthesis, June 2001