ICP-MS for Forensic Applications

More than 30 years ago, Sam Houk, professor of chemistry at Iowa State University (Ames, Iowa), and a scientist in the Ames Laboratory of the United States Department of Energy, was the first scientist to use an inductively coupled plasma (ICP) as an ionization source for mass spectrometry (MS). Spectroscopy recently spoke to Houk about the current state of ICP-MS, including its use in forensics.

Spectroscopy: In what new applications areas is ICP-MS being used?

Houk: ICP-MS is still very widely used in its traditional applications — environmental analysis, semiconductor analysis, isotopic and materials analysis in the nuclear industry, and so on —but there are some very interesting new ones, such as forensic matching of evidence based on its trace-element composition. There’s much more work in biological analysis now, including speciation or measurement of the chemical forms of the elements, at trace and ultratrace levels. Usually that requires coupling chromatographic separations with ICP-MS. And there’s even a group that is finding previously undiscovered elements by doing ICP-MS experiments at very high mass. This is done mainly by a group in Israel, and [they have] found evidence for low levels of unexpected elements with reasonably long radioactive half lives. So there are some very interesting new applications.

Spectroscopy: What kinds of analysis are you doing in forensics?

Houk: Well, we are supported by the [United States] Department of Energy to develop and evaluate ICP-MS for measurement of trace actinides, particularly to try and diagnose illicit enrichment of uranium or production of plutonium. So there’s a large area within the Department of Energy that is tasked with supporting possible programs to look for nuclear weapons development. We work on laser ablation (LA)-ICP-MS to this end.

We also use LA-ICP-MS for forensic matching of material. We just recently had a program supported by the [United States] National Institute of Justice for that, in which we tried to match evidence, say collected from a crime scene, with evidence collected from a suspect, based on their trace-element composition. This has been shown to work by other groups as well, and I will mention particularly John Watling in Australia and José Almirall at Florida International University in Miami. A lot depends on whether the trace-element signature is diagnostic and unique. This is being sorted out for a number of kinds of materials.

Spectroscopy: José Almirall is also using laser-induced breakdown spectroscopy (LIBS) in that work. How do you see the balance between LA-ICP-MS and LIBS for this kind of application?

Houk: LIBS is another way, in general, to do the same sort of analysis. Isotopic information, which is of value for some elements — I would say mainly uranium and lead — is a lot easier to extract from ICP-MS measurement. As usually done, the detection limit is probably better in ICP-MS, although LIBS is improving to close to those levels, with a variety of experiments. So LIBS is another way to do most of what ICP-MS would do for that.

Spectroscopy: In the forensic matching of materials, what kinds of samples are you talking about?

Houk: Well, to my knowledge, the original application of this, done by John Watling of Australia, was for matching trace elements in gold. José Almirall, and perhaps others, did comprehensive analysis of glass samples and found that the trace-element composition can be compared to a database of many kinds of glass samples for matching purposes. Other materials of interest are metals. For example, we have been looking at copper speaker wire. We’ve also been looking at duct tape, which is often used to commit crimes.

One thing we find is that the spatial homogeneity of the trace-element levels in these materials is something of an issue. So experiments need to include comprehensive evaluations of the spatial heterogeneity of the trace elements in the materials. I should say that [the analysis of] many such materials can be envisioned, but before they can properly be used, some sort of database or global information is needed about the general trace-element composition of all the sources of the particular kind of material. There has been a lot said about trace elements in bullet lead being used improperly.

Spectroscopy: Can you explain a bit further the issue of the spatial homogeneity of the trace materials?

Houk: Suppose you have a solid sample, and you want to analyze it and still keep it intact; you don’t want to put the entire thing in solution. So one thing we have found in recent times, and perhaps others found this earlier, is that before you can try to match two different kinds of paint samples, for example, you have to be sure that the trace-element composition in the section you’re sampling represents that of the whole material. And we’ve found that we end up having to ablate much larger portions of the material than we thought would be necessary initially.

Spectroscopy: Have you or others had to develop new methods to be able to do those sorts of forensic analyses?

Houk: I would say in recent times we have been applying methods that have already been demonstrated by others, particularly improvements to laser ablation, such as use of a laser with a very short pulse duration of about 200 femtoseconds. Rick Russo and his group at Lawrence Berkeley National Lab pioneered the use of femtosecond-lasers for LA-ICP-MS. We find that the advantages described by that group are very applicable for the more application-oriented work we have been doing.

Spectroscopy: What do you plan to focus on next in the field of ICP-MS?

Houk: Well, that’s a good question. We’ll continue work on the laser ablation experiments, but a couple of things that we have done in recent years that I think would be interesting to look into would be to use ICP-MS for characterization of nanoparticles. There are other groups working on this. One I know of is Niemax’s group in Dortmund, Germany. I believe it should be possible to introduce individual streams of finely collimated nanoparticles and get composition and sizes of the nanoparticles, on a particle-by-particle basis.

We also did experiments here at Ames in the mid 2000s, in collaboration with Dan Armstrong, in which we measured trace-element levels in individual biological cells, which were grown in trace metal–enriched media. And I think there are a number of energy-related applications relevant to that. So those are areas I’d like to get back into sometime.