Soil samples are often used in an effort to place a suspect at a particular location, i. e., material from the suspect's shoes,
clothes, car, etc. is compared with soil from the crime area. The trace element concentration in both samples has been used
to establish identity; this has been obtained usually by use of neutron activation analysis. In rural areas and farms more
than half a dozen different kinds of soil may be present, and each soil may consist of several horizons. Horizons in turn
may be broken down into sub-horizons or transitional layers, depending upon the humus present and the amount of weathering.
Thus it would appear that comparison of soil samples would require rather exhaustive knowledge of the terrain, depth, and
history of the samples area since variations of the elements in the soil are affected by eluviation (washing out) or illuviation
(washing in). Such variations are illustrated by showing the change of concentration of elements with soil depth. Even though
most forensic applications require comparisons of only the top inch or two of soil, the samples must be classified as to depth
and type of soil since man often change the soil profile (ploughing, construction, etc.), or nature does (rain, decay, etc.).
We have collected a number of soil samples from a several square mile area in both the horizontal and vertical profile. These
soils were dried, thoroughly mixed, and portions, irradiated for one minute (for short half-life induced activities), and
one week (for long half-life activities) at a flux of ∼10 n·cm−2·sec−1. Standards were irradiated in the same manner as samples. A number of elements were determined non-destructively; chemical
separation was performed to determine others. Our results indicate great variation (factors of two or more occur in the concentration
of many trace elements in going from the surface to several inches below. Similar variations occur in the horizontal plane.
These data ind ate that comparison of soil samples by use of trace element concentrations is subject to considerable uncertainty.
If large variations can be found within a small sampling area, the statistical significance of apparent agreement between
two samples appears highly questionable.
A method was developed to determine thorium and uranium in semiconductor potting plastics. The method is based on neutron activation and subsequent radiochemical separation to isolate and permit measurement of the daughter products233Pa and239Np of the induced233Th and239U. These plastics typically contain macro amounts of silicon, bromine and antimony and nanogram per gram amounts of thorium and uranium. The radiochemical method provides the necessary sensitivity and makes it possible to easily attain adequate decontamination of the tiny amounts of233Pa and239Np from the high levels of radioactive bromine and antimony.
Instrumental neutron activation analysis was used to determine the elemental concentrations in water and air solid samples
collected on Nuclepore and Whatman filters from the Walker Branch Watershed. The results from this study show that the trace
element concentrations removed by water from the watershed vary seasonally, as well as geographically. The data point up the
usefulness and versatility of absolute neutron activation analysis. The NBS Standard Orchard Leaves was assayed in a similar
manner, and good agreement was obtained between results here and standard values.
The forensic activation analyst must often evaluate his own results as they relate to certain legal or moral situations, since
investigative officers, and courts of law are not usually competent to make judgements of the validity or meaning of scientific
data. In providing scientific evidence in court, two criteria for criminal identification must be met: (1) suspect's sample
should be similar to sample found at the scene of the crime, (2) samples relared to other people in the same statistical population
should not generally match that found at the crime site. When two or more specimens are submitted for comparison by NAA they
will usually fall into one of three classes: (a) materials about which we have partial or inconclusive data, (c) materials
with an excellent analytical data background. Ideally all cases would fall in category c; in practice, very few. Some examples
of cases and/or situations that fall into these three categories in both individual and corporate investigations are given.