After looking at a very popular but somewhat dubious technique, let’s look at a very obscure but rock solid one. In fact, it is probably the most accurate, replicable, and precise technique I will discuss. This technique is neutron activation analysis. It was famously used in studying JFK’s assassination to claim that the fragments in Governor Connally’s wrist came from the same bullet found in JFK. More on that later! As before, we shall begin by first looking at the necessary background information, then the technique itself, and finally what reasons there are to be skeptical.
Background Reading on Atoms and Atomic Structure
All matter on this gay earth is made up of atoms. Every atom has protons and maybe some neutrons hanging around in the nucleus. Most atoms are surrounded by a constant field of electrons. The number of protons in the nucleus tells us what element it is. The number of neutrons is typically incidental as they just sit there chilling out. Electrons are practically everywhere all the time just buzzing around like assholes you can’t ever look at so you’re pretty much not going to have to worry about technical stuff like shells, ground states, and the like. What we are going to be concerned with primarily, if the title wasn’t obvious enough, is neutrons
On its surface nothing could be simpler than NAA. You take your sample and shove it in a nuclear reactor for a bit. Then you pull it out and use your gamma ray detector to see what kinds of gamma rays come out of it. Unfortunately for people who want to use the technique (but not so unfortunate for nuclear chemists who need work) there’s a lot of science and hardware involved in all of those steps.
First you need something that emits low energy neutrons. Usually this is the core of a nuclear reactor. That is pretty hard to come by for most folks, even prosecutors and forensic scientists. Alternately you can use an element that rapidly emits low energy neutrons. The only element I can remember right now that would be a potential source is Americinium. Of course, that element is manmade and costs thousands of dollars per gram. There are other more viable and common radioactive elements but I can’t remember what they are right now.
Once you have your source of neutrons you expose your sample to the source. You have to control for neutron flux (functionally the density of emission/level of exposure), time, sample size, and the samples potential ability to capture neutrons. If an atom captures a neutron it emits a gamma ray and becomes a little bit heavier. It is then in an excited state for a bit, which ordinary folks call radioactive. After a while, the atom will emit a proton, neutron, electron, gamma ray, or a combination of these things.
The goal of NAA is, classically, to evaluate the timing and energy level of the gamma rays that are emitted. Every element has a different rate of emission and a different energy level gamma ray. If you have an accurate enough stop-watch and an accurate enough gamma ray detector you can determine what elements are present in the sample.
More experimentally, and I believe to such a degree it hasn’t been used in any court I’m aware of, you can measure the gamma ray emitted by an element upon neutron capture. This should be equally scientifically valid but far more difficult to arrange.
However, NAA is not just a qualitative technique. No, sir! It is probably the most accurate quantitative technique on the planet. With the proper procedure and equipment, you can determine the presence and quantity of almost every element in a given sample down to the level of parts per billion. With some elements you can get down to parts per trillion which is, frankly, a stunning level of measurement. Some elements don’t capture neutrons very well (e.g. lead) and some elements emit gamma rays of extremely close energy levels to other similar elements (e.g. barium and antimony). Because of this, it is not uncommon for chemical separations or additional analytical techniques are performed prior or subsequent to NAA.
Under ideal circumstances a sample can be irradiated, observed, then allowed to become non-radioactive and be totally intact and safe to admit at trial! WHICH IS ALSO CRAZY FOR A CHEMICAL ANALYSIS!
So you’ve determined a hair sample has 2ppb of something rare in it, like….Niobium. Then you get a sample of your suspects hair and, sure enough, it too has 2 ppb of niobium! It’s a MATCH! Or is it? Well how rare is niobium in human hair? What industries use niobium? Is there widespread niobium contamination where the suspect works, lives, commutes, or gets his food? I don’t know. Chances are if you asked a still practicing chemist they wouldn’t know either. At best they’d have a hunch that it’s a pretty rare transition metal and pretty expensive so it shouldn’t be prevalent in any substantial quantity. But they wouldn’t know well enough to form an expert opinion on it that would be admissible at trial. Chances are there aren’t any studies in existence on the prevalence of niobium in the average person’s hair, much less in this geographic area.
This becomes a much huger problem when we aren’t talking about rare elements but instead, elements present in some amount in at least one object you interact with on a daily basis. It gets worse still when we aren’t talking about hair samples but bullet, tube, stone, steel or other product samples.
Remember in my previous post I said some questions would be constantly coming back to haunt us?
(1) Is the thing unique?
(2) Can it transfer that uniqueness to something else?
(3) Can that uniqueness be measured accurately enough to distinguish the thing from all other similar things?
Well we have the same problem with NAA. While the technique is amazing, interpreting the results from that analysis is tricky. In some cases it will be impossible. How common are trace elements in a run of bullets? Can you distinguish down to a year of production? A month? A week? A single box of bullets? Or can you distinguish between each individual bullet made? In some cases maybe you can be that precise. In most, you are stuck wondering how significant your results really are. If you are trying to determine whether there was gunpowder residue on someone’s hands, what are the ordinary levels of antimony on a person’s hands? What about if they work in a sporting goods store? Or sell fireworks?
Did the fragments in the governor’s wrist come from the same bullet as JFK? Yeah, but NAA isn’t going to prove that because its unlikely bullets of similar manufacture are chemically distinct enough to make an absolute claim based merely on chemical composition. You will still encounter prosecution experts and prosecutors claiming that because an elemental analysis showed strong simliarities then the killing bullet came from the defendant’s box of bullets. No one, not even defense lawyers oftentimes, asks if any other box of bullets from the same maker would have the same results. Instead lawyers fuss and moan about whether the testimony as a whole is admissible. Judges look at the incredible scientific background and psha at the defense lawyers. Everyone misses the critical question of statistical probability.
Finally, because the technique is so precise, any contamination, even the slightest teeniest tiniest amount will become part of the results and ruin the entire analysis.
In the end, techniques are just a part of the forensics puzzle. The meaning of that result is the biggest piece. Without strenuously questioning how we interpret the results, lawyers, judges, and jurors are overawed by the precision of the technique and ignore the commonality of the result.
NAA is totally admissible in any jurisdiction. Claims of “matches” or the meaning behind the results are up for debate and all good lawyers should be asking for statistical validation of any claim of significance.
Another major issue would be the qualifications of the expert. This should almost never come up. This technique can only be performed at a serious academic laboratory and at least supervised by a full PhD with a post doc in the field. It is tough as shit to do right.