Tracing A Threat

Everything leaves a trace—fingerprints, ballistics, DNA. Law enforcement agents have an extensive forensic toolbox to link perpetrators to crime scenes and murder weapons. But what happens when the murder weapon is a molecule? Scientists are discovering that even a chemical holds clues that can point to its source.

This nascent field is known as chemical forensics. Its goal is to take analytical techniques that have been used for forensic analysis, such as impurity profiling and stable isotope analysis, and use them to attribute weaponized toxic chemicals or related substances to their sources. A chemical forensic analysis could, for example, trace a chemical threat agent back to the specific lot of the precursor that was used to make it.

By developing chemical forensics, scientists “have a greater chance of finding perpetrators of chemical attacks or their sources of materials before they can strike again,” explains Carlos G. Fraga, a chemical forensics expert at Pacific Northwest National Laboratory (PNNL).

There’s been a flurry of published chemical forensic research in the past couple of years, all of it supported by the Chemical Forensics Program, sponsored by the Department of Homeland Security. DHS hopes the program will develop and maintain scientific expertise for collecting, preserving, and forensically analyzing chemical threat agents so that DHS is prepared in case of an attack. It also wants such science to be admissible in criminal prosecutions.

The Chemical Forensics Program’s origins can be traced back, in part, to the 2001 anthrax attacks, says Randolph Long, acting director of DHS’s Chemical & Biological Division, part of the agency’s Science & Technology Directorate. Five people died and 17 were sickened after coming in contact with mail laced with anthrax spores.

Numerous samples were collected after that attack, Long notes. “They were stored for many years, awaiting the development of technology that could definitively provide evidence that could link them to a perpetrator,” he says. The Chemical Forensics Program hopes to put technologies in place now to trace future threats.

Since the 2001 anthrax attacks, the threat of a biological attack has loomed large in the public imagination. But the threat of a chemical attack may be even more likely, according to a recent report from the Center for a New American Security (CNAS), a think tank specializing in national security issues.

In 2008, CNAS began interviewing members of the Aum Shinrikyo cult who were imprisoned for their role in what has become the most notorious act of chemical terrorism: the 1995 sarin attack on the Tokyo subway in which 13 people died and more than 6,000 sought hospital treatment. CNAS’s goal was to gain insights into how terrorists develop biological and chemical weapons.

CNAS released a report in July 2011 noting that even though Aum had tried to develop both kinds of weapons, its biological program failed while its chemical program flourished. The report says “a number of factors suggest that chemical weapons are likely to be more accessible than biological capabilities for terrorist groups intent on killing substantial numbers of people.”

According to the report, “Aum’s history suggests that with access to the relevant literature, a generally skilled chemist can produce a chemical weapon, whereas a generally trained biologist is likely to have more difficulty propagating and conserving an unfamiliar pathogen.” The authors conclude: “We believe that other terrorists may endeavor to use biological and chemical weapons in our future.”

With that in mind, Fraga says, “we’re getting ready now instead of waiting for something bad to happen.” Fraga has been working in an area known as impurity profiling, which involves looking at the various impurities that turn up in a sample of a chemical threat agent. These impurities—which can be impurities in the starting material, unreacted starting material, degradation products, and side products, to name just a few—make up a chemical threat agent’s so-called attribution signature.

That signature can help reveal the route used to make a compound, the conditions a compound was prepared under, and even what specific batch of a precursor was used. “We’re trying to get into the recipe box of the synthetic chemist” who made the compound, explains Joseph Chipuk, director of chemical sciences at Signature Science, a technical consulting firm in Austin, Texas.

Impurity profiling has been used for decades to follow the trail of illegal drugs. Recent advances in instrumentation and new data analysis techniques have made impurity profiling useful for determining attribution signatures of chemical threat agents. The information from those signatures is helpful for law enforcement, Chipuk says. It can indicate, for example, which brand of chemical or what specific equipment officers should look for.

Fraga and colleagues at PNNL and Battelle Memorial Institute in Columbus, Ohio, recently used impurity profiling to link samples of sarin they synthesized to the specific batches of its precursor, methylphosphonic dichloride. They found they could tell what commercial manufacturer supplied the precursor and even what chemical lot it came from (Anal. Chem., DOI: 10.1021/ac202340u).

“We were able to show that most of the impurities that were in the precursor transferred over into the product,” Fraga explains. “I was pretty surprised by that because there’s a distillation step that’s done, two solvent extraction steps, and we still found these impurities.”

Saphon Hok, a chemist at Lawrence Livermore National Laboratory (LLNL), in collaboration with the Swedish Defence Research Agency, has been using impurity profiling to link chemical threat agents to the method used to make them. At the American Chemical Society national meeting in Denver last fall, Hok reported that he could differentiate among four production routes to make the nerve agent known as Russian VX. He has done similar work on sarin, VX, and the blister agent sulfur mustard.

“I go through all the possible ways an agent can be synthesized, and then I go into the laboratory and make these compounds on a small scale as crudely as possible,” Hok explains. “We try to mimic someone who is trying to do this synthesis in their backyard or in their garage.”

He then analyzes the compounds using liquid chromatography/mass spectrometry, gas chromatography/mass spectrometry (GC/MS), and nuclear magnetic resonance (NMR) spectroscopy. The data, he says, reveal different signatures that pinpoint the production method used.

Signature Science recently looked at both a toxic bicyclocarboxylate and a toxic bicyclophosphate that are known to act upon the nervous system. “A lot of the chemicals we look at may not be quite as toxic as some of those conventional war gases and nerve agents, but they’re often easier to manufacture or the materials are more available,” Chipuk points out. “We tend to focus on things that are more likely to be in the terrorist profile. Most terrorists don’t want to go through a six-step synthetic route to something. That’s simply too much trouble.”

To figure out signatures based on various synthetic routes and conditions, Chipuk says that the synthetic chemists on his team will make the same chemical threat agent as many as 2,000 times in an “almost robotic manner,” following a database that tells them exactly what conditions to use. They then hand off the product to the analytical chemists, who look at all the tiny impurities that turn up along with the toxic chemical—“the stuff that’s down in the weeds,” as Chipuk describes it. From there, the hundreds or, in some cases, thousands of spectra that are collected go to statisticians and computer scientists who work their magic to tease out the unique attribution signatures.

Scientists are also using impurity profiling to differentiate routes used to purify the deadly protein ricin, which comes from castor seeds. ­David S. Wunschel and colleagues at PNNL prepared ricin by four methods—three from kitchen recipes that appear in “anarchist literature” and one from a relatively simple laboratory procedure. They made derivatives of the carbohydrates and fatty acids found with the ricin, analyzed them by GC/MS, and were able to differentiate among the preparations used to go from seed to protein (Anal. Chem., DOI: 10.1021/ac1006206).

As components are stripped from the seed, the material is going to change, Wunschel explains. For example, castor oil accounts for roughly half of the contents of the castor seed, and the fatty acid ricinoleic acid accounts for a large portion of that oil. If someone were to try to isolate ricin, they’d first want to strip away that oil, resulting in a loss of ricinoleic acid. Wunschel and his coworkers found that ricinoleic acid levels will vary depending on the purification method and how well someone succeeds in purifying the protein.

“You’re likely to remove seed components that have different carbohydrate markers,” Wunschel adds. Specifically, he and his colleagues found changes in arabinose and mannose concentrations after precipitation of the ricin protein. These carbohydrates originate in the castor seed’s cell wall. Some are lost and some are enriched depending on how the ricin was isolated.

“Just being in possession of castor seeds is not a crime,” Wunschel points out. But applying this sort of analysis to residues found in some clandestine lab could reveal whether someone was trying to extract oil or remove cell wall components in an attempt to purify ricin, he says.

Impurities aren’t the only clue that chemists can use to trace a chemical threat agent back to its source. The atoms that make up the chemical threat agent itself can also point to its origins. Light elements—such as hydrogen, nitrogen, oxygen, carbon, and sulfur—have multiple stable isotopes: carbon-12 and carbon-13 or oxygen-16 and oxygen-18, for example. The ratio of these stable isotopes can vary depending upon the natural source of a material or how it was manufactured, and these isotopes can be measured using MS instruments outfitted with a detector designed to pick up these tiny differences in mass.

“Stable isotope analysis has been used for a long time in the geological and hydrological sciences,” says Michael J. Singleton, a staff chemist at LLNL who is using stable isotope analysis to characterize chemical threat agents. “The strength of using the stable isotopes is that you can tell the difference between substances with the same exact chemical composition, so it works for highly purified compounds where there may not be a signature from impurities,” Singleton explains.

“If a chemical was used in an attack and investigators found a source of that chemical, you would want to be able to link those two so that it could be proven in a court of law that the person actually had the compound that was used in an attack,” Singleton says. “By measuring the stable isotope composition, you can show that that’s the case. But you need to show that there’s enough variation in the compounds that they don’t all look alike.”

To that end, Singleton recently used stable isotope analysis to look at the toxic industrial chemical ammonium metavanadate (NH₄VO₃) and the rodenticide tetramethylenedisulfotetramine, commonly known as TETS. In the case of NH₄VO₃, Singleton and his team purchased 19 samples of the material, all from different suppliers, and analyzed their N, H, and O isotope compositions. They determined that those 19 samples originated from six sources of the compound, which they were able to differentiate with “a very high level of confidence” (Forensic Sci. Int., DOI: 10.1016/j.forsciint.2011.01.005).

For TETS, a substance that’s banned throughout the world, Singleton and his LLNL coworkers had to synthesize the compound. They used different sources of its sulfamide precursor and found they could use stable isotope analysis to link the TETS to its sulfamide source.

Scientists are also using stable isotope analysis to study cyanide. In 1982 seven people died in the Chicago area after taking Tylenol tainted with the substance. That case remains unsolved. Helen W. Kreuzer, an expert in stable isotope analysis at PNNL, along with colleagues there and at Oak Ridge National Laboratory, recently examined a collection of 65 different samples of cyanide and found that stable isotope analysis was sufficient to differentiate 95% of the samples (J. Forensic Sci., DOI: 10.1111/j.1556-4029.2011.01946.x).

PNNL’s Fraga combined Kreuzer’s data with ion chromatography analyses that Fraga had done with potassium cyanide, in which he was able to use anionic impurities to trace the material to its country of origin (Talanta, DOI: 10.1016/j.talanta.2010.08.017). By integrating their data, Kreuzer and Fraga think it may be possible to trace cyanide used in an attack back to its source. These kind of data “would give you some leads about where to start and if you’re trying to figure out whether the cyanide used in this attack was the same stuff as what was used in somebody’s garage,” Kreuzer says. “It’s not a turnkey solution to a chemical crime, but it certainly could provide evidence that would be helpful in an investigation.”

Although chromatography equipment and mass spectrometers are the workhorses of chemical forensics, PNNL senior staff scientists John R. Cort and Herman M. Cho are among a handful of researchers who have been using NMR to gather clues about chemical threat agents.

The PNNL team used the diastereomeric ratios of the rodenticide brodifacoum to track the poison back to its manufacturer and in some cases, its specific batch (Forensic Sci. Int., DOI: 10.1016/j.forsciint.2011.08.003).

Brodifacoum has two stereogenic centers and four stereoisomers. All of brodifacoum’s isomers are lethal, and because makers of rat poison aren’t hung up on preparing a single isomer of the material, its diastereomeric ratio varies depending on the synthesis. For example, Cort tells C&EN, in one brodifacoum sample the diastereomeric ratio was 70:30 while in another it was close to 50:50.

Cort acknowledges that the technique is limited to chemical threats that are produced as diastereomers or as isomers of cis and trans alkenes, as certain pesticides are. Even so, he says it provides useful information that’s intrinsic to the chemical threat. “We’re not looking for tiny impurities that might be hard to find,” he says. “These are actually characteristics of the active ingredient itself, and you can see these signatures pretty clearly.”

DHS, as well as the scientists working in chemical forensics, want the public to know that they are making preparations should there be a chemical attack in the future. But they also hope that by making it known that they have many tools to trace chemical threats back to their source, they will actually deter someone from planning a chemical attack.

“The fact is that technology continues to improve, instrumentation continues to improve, and computers continue to improve. The chances of someone being able to slip by undetected are getting smaller and smaller,” says Signature Science’s Chipuk. “If you were to choose to do something like this, the science is going to catch up to you.”