_The Problem_
You are a scientist working in the criminology laboratory of a large
metropolitan city. A detective brings you a minute sample of paint taken
from the clothing of a hit-and-run victim. He has a suspect whose
automobile paint seems to match that sample. The suspect was found in
his parked automobile, not far from the scene of the accident. He seems
to fit the description given by two witnesses, and he is extremely
nervous. You scrape a small sample of paint from a recently damaged area
of the suspect’s car, and, (with the aid of a microscope) find that the
pigment content seems to be the same as that taken from the victim’s
clothing. But, are they really from the same paint?
_The Solution_
You know that paint, like almost everything else, contains very small
quantities of impurities that are present only by accident and do not
affect its properties as a useful material. The trace impurities, as
they are called, will vary from batch to batch of the same paint. Very
rarely will a match be obtained in both type and concentration of trace
impurities in two samples if they are not from the same batch.
By measuring a sufficient number of different elements, the probability
of accidentally matching two samples can be as rare as the duplication
of fingerprints in two individuals. Matching of trace impurities is
often called a “fingerprint” method.
With neutron activation analysis, you can obtain the “fingerprints” of
the two samples to see if they match. Although this kind of evidence may
be difficult to use as proof in court, a positive match will let the
detective know that he is on the right track. Also, the suspect might
confess if he is confronted with the evidence and realizes that he is
“caught”. On the other hand, a mismatch will clear the suspect
completely and the detective will know to look elsewhere for the
criminal.
You seal each sample in a tiny polyethylene bag about ½ inch square. One
sample is taken from the victim’s clothing and the second, about the
same size as the first, taken from the damaged area of the automobile.
In preparing these samples, you handle all the materials with clean
forceps because you realize that the most minute dirt from your fingers
will be detected in the analysis.
The two bags are irradiated together for 1 hour in a nearby reactor and
2 hours later you begin counting the samples with a high-resolution,
lithium-drifted-germanium, gamma-ray spectrometer. This will give you a
match (or mismatch) for elements that yield radioisotopes of fairly
short half-life such as manganese (2.56 hours), copper (12.8 hours),
sodium (15 hours), arsenic (27.7 hours), etc. You plan on “counting” the
samples again later on, if the first counts match, so that you can check
on radioisotopes with longer half-lives such as iron (45 days), chromium
(27 days), silver (270 days), cobalt (5 years), etc.
The two gamma-ray spectra you obtain look like those in the figure on
the opposite page. The gamma rays from the irradiated paint taken from
the victim’s clothing indicate the presence of the common elements
sodium, potassium, and copper, but gold, lanthanum, and europium are
also conspicuously present. The gamma rays from the other sample also
reveal sodium, potassium, and gold but in rather different proportions.
More striking is the absence of copper and the two rare earths, and the
presence of manganese and arsenic, which were not indicated in the first
sample.
The paint samples definitely do not match. Therefore, you inform the
detective that his suspect is innocent after all. You’ve solved your
problem, but he still has his. Perhaps the same technique will provide
positive proof when he finds the real culprit.
[Illustration: Graph: “Counts/10 min per 3.3 KeV channel (arbitrary
scale)” _vs_ “Channel number”]
Element Channel number
.122 MeV Eu (Europium) 35
.328 MeV La (Lanthanum) 95
.344 MeV Eu 96
.412 MeV Au (Gold) 110
.486 MeV La 120
.511 MeV Cu (Copper) 145
.815 MeV La 240
.837 MeV Eu 245
.961 MeV Eu 290
1.37 MeV Na (Sodium) 410
1.53 MeV K (Potassium) 455
1.60 MeV La 475
(channel numbers estimated)
[Illustration: Graph: “Counts/10 min per 3.3 KeV channel (arbitrary
scale)” _vs_ “Channel number”]
Element Channel number
0.412 MeV Au (Gold) 110
0.511 MeV Na 145
0.559 MeV As (Arsenic) 160
0.657 MeV As 195
0.847 MeV Mn (Manganese) 250
1.21 MeV As 370
1.37 MeV Na (Sodium) 410
1.53 MeV K (Potassium) 455
(channel numbers estimated)
Gamma-ray spectra of two samples of paint. These two spectra are
obviously different and, therefore, could not have come from the same
source.