Tip:
Highlight text to annotate it
X
[ Music ]
>> Our next speaker's title is "Potential of Mineral Varieties
for Forensic Soil Analysis".
Andy Bowen works for a private forensic laboratory,
Stoney Forensics in Chantilly, Virginia,
and where he's employed as a forensic scientist.
[ Background noise ]
>> Andrew Bowen: Thank you very much, Chris, and thank you
to NIJ and FBI for the opportunity to speak here,
especially an opportunity to talk about soil,
forensic soil analysis, it's become a strong,
personal interest of mine.
I'm going to go through these first slides relatively quickly.
You can read these when you get back home, to try to get us --
keep us on schedule if possible.
I was basically going to, you know, talk about the fact
that soil evidence has been around for a long time,
and despite that, only about a quarter of labs,
of trace evidence labs in the U.S. currently do soil analysis,
according to results of a survey reported here
at this symposium three years ago.
I think that has something to do with the variety of components
and complexity of soil, along with the variety
of published methods as we heard at the talk earlier today
by Gwyneth, looking into, you know,
what methods are used is a daunting task
because of the great variety that's out there.
And when I say why we should try to change this, it's --
what I'm suggesting we might want to change is the fact
that only about a quarter
of trace labs currently do soil in this country.
There, there are some theoretical reasons why soil has
a lot of potential of evidence, its ubiquity,
how the great variety due to all the possible components
in the various proportions in with which they can combined.
Anyone with children, I've got a 3-year-old and an 18-month-old
that might be watching right now for all I know, you know,
the ease with which they pick up soil and transfer it
to everything in our house is evidence enough for me
that it's got potential.
And if you go to the literature, Raymond Murray's book,
Evidence in the Earth has dozens and dozens of case examples,
some of which pretty astonishing in terms of the value
of the geological evidence.
John McPhee goes into greater detail
in a smaller number of cases.
And I think that -- and I'm really talking
about the United States here, I think the way to change this is
to sort of get soil's foot in the door to more labs
and in order to do that, we need a method that uses,
largely uses the existing personnel and skills,
can't be -- you know with give the current fiscal situation,
there's not a lot of opportunities to hire loads
and loads of geologists just to start doing this.
We also need to use existing equipment
because most lab directors will tell you it's pretty hard
to get funding for that kind of thing right now.
And most importantly, the method needs to work,
it has to take a reasonable amount of time.
And, I think if it gets in --
so the method I'm proposing is not what I think would be the
best method.
I think the best method's probably already out there,
and we'll probably hear it two talks from now.
I think Skip's going to go into some detail, but I think this,
the gap between what currently exists and the kinds of things
that Skip does, or the FBI does, it's a large gap to bridge,
and hopefully if we can find something in between
that allows people to start doing soil
with their existing personnel and equipment,
then sort of organically I would imagine the methods would
increase and people would get additional training and skills
in it and hopefully start doing a more comprehensive
soil analysis.
And what I'm interested in talking about is the potential
for mineral varieties, the role they might play.
The concept is basically that at every --
any given mineral has a range with respect to a number
of its properties, morphological inclusions, texture,
we'll see examples of specific minerals and some
of these properties, and they're analogous
to class characteristics of fiber,
so just like a fiber analyst doesn't stop
when they find nylon in two samples,
they want to know what are the properties o the specific nylons
on this case, and do they compare.
The concept of mineral varieties has been used by geologists for,
probably for over 100 years.
The reference I run into when I read
about this is typically a 1928 article by Alan Brammell,
where he sort of articulated this, I think,
for the first time, and the early studies
in the early 1900s, mid 1900s,
these mineral varieties were generally minerals
that had the same properties grouped
into a single varietal type, most of it was done
by optical microscopy,
and features you can observe with a PLM.
The work is still done currently by sedimentary geologists,
including some in this room, and now they tend
to call it single grain studies as opposed
to mineral variety studies, and they do more dating,
radiometric dating, isotopic analysis, CL imaging,
things that are more quantitative
for comparative purposes.
The reason they do this, primarily for --
so the image is not showing up quite as well
up there unfortunately, this is from that Brammell article
that -- and I've got copies of that, and can tell you
where to find it if you like.
The idea is that if you're trying to figure out the source
of detrital grains in a sediment or sedimentary rock,
you want to know if this sediment in this basin came
from the granite that's on the, you know, up to the north,
or the granite that outcrops to the south,
they might both have similar minerals, they both have zircon,
so the presence of these minerals
in your sediment doesn't allow you to determine its source,
but they notice that the variety,
the zircon from one source might be colorless,
zoned with particular inclusions,
another might be rounded, no zoning, clear, no inclusion,
and so the varieties of these minerals could be used
to associate the detrital grains with their sources.
And the idea to use this in forensic science,
comparing the varieties from a question soil to a known soil,
it's not a new one, it's not my idea.
I'd argue it's been around as long
as forensic soil's been used in criminal cases.
The first case that I know of, George Pop in 1908
in the Margarethe Filbert case, if you read descriptions of that
in various soils, he's comparing to soil from the suspect's shoe,
he describes the quartz in one sample as being splintery,
in another sample as being milky,
so clearly this appreciation for the qualities or properties
of these minerals in addition to their,
strictly their identities was appreciated long ago.
In 2007, in Heavy Mineral and Use,
the chapter that Skip wrote, this quote jumped out at me,
they've come to place more than ordinary importance
on this concept when they do their comparisons
and Skip is the one who brought this to my attention.
So, despite the fact that this has been around I --
looking through the literature,
I would argue there's really only been one systematic study
that's been published that really sort
of uses this concept, now that might change,
there's some cathodal luminescence of feldspars
that was talked about earlier,
and I think that would probably fall into this category.
The other exception is a study that focused on quartz grains,
classifying them into different types based
on their morphology and surface texture.
And although that wasn't specifically called varietal
types here, they were clearly using these properties
to group these grains into different groups.
Unfortunately I don't know of many people in this,
in the U.S. doing this currently, and certainly
within the crime lab setting, there are not too many people
who have the type of expertise
to start applying the SEM in this way.
It might well be a better method than the one I'm proposing,
I think that's a worthy debate to have.
But what I think would be interesting would be
to do some research looking at classifying minerals
into varietal types using PLM, and study perhaps 11
or 12 common minerals.
Look at several hundred soils from different locations,
and go through and get a sense
for how many varieties do we even find,
how different are they from one location to another,
how similar are they within a single location?
The general types of things we're interested in --
whenever we look at a method and applying it
to forensic soil analysis.
So I think that would be really interesting, and that's sort
of the research I'd like to see happen,
hopefully NIJ will fund me to do it.
If not, if someone else does it,
I think that would be great as well.
The reason I think it has the possibility to sort
of meet those admittedly arbitrary criteria I said
at the beginning, it uses basically the same principles
fiber analysis uses.
You're identifying a small subset of materials.
I imagine if we looked at 12 minerals, we'd find that 5
or so really, really show a lot of variety
and would be most useful to focus on,
and there might only be a small, relatively small subset
of possible properties that helped us to group them in,
into varieties, I don't know, but, but it's a hypothesis.
And so I think the ultimate procedure might involve 5
or 6 minerals, all that -- being all that someone needs
to identify in a sample.
And then characterizing them microscopically
with skills they already have.
It would use polarized light microscopes available
in about 90% of trace evidence laboratories.
The data comes from that same survey I mentioned earlier
presented two years ago at this symposium.
And would it work?
You know, research needs to answer that question.
I think if it does work, the fact that we're focusing
on a relatively small subset of the soil, just a single fine,
sand sized mineral fraction would probably all
that would be looked at.
Because you're not considering all the minerals
in there required to identify all of them,
that might speed up the analysis.
We'll have to see what -- you know, see exactly how much time
that takes, but hopefully it would be efficient.
And I think it would be effective in forensic scenarios
for the same reason sedimentary geologists use these.
The sediment in a sedimentary basin is not going to be a one
to one image of its source rock, there are a lot of things
that happen from source to sediment that are going
to change that assemblage.
We've got differential weathering from the rock,
differential transport based on shape, size, and density,
differences of different minerals,
we've got different minerals, different chemical stability,
so some minerals are going to dissolve and, or,
in during digenesis of the rock,
sedimentary rock is formed, or during transport.
So there's all these, these possible things that can happen.
Additional sediments can be mixed with it later,
but the mineral varieties are still the same if they came
from that source rock.
I feel like forensic soil analysis runs
into a lot of these same issues.
We've got the potential
for differential transfer persistence.
I imagine that mica flakes, although I haven't seen evidence
of this in literature, my hypothesis would be
that mica flakes would adhere more readily to evidence,
and so as someone's walking around with mica on their shoes,
than a round mineral like quartz, or equant mineral
like quartz, and I would, I would imagine that a garnet
and a quartz grain, that both having same, same shapes
and same size, the garnet would be more likely
to fall off being much more dense than the quartz would.
So, again, a hypotheses, but I think there's some reason
to suggest in the geologic literature, that these kinds
of fractionation and differential transfer
and persistence are taking place,
and certainly the possibility
to contaminate a forensic soil sample with soil
from another location is a very real one that we need
to consider whenever -- as we think about methods to use
for this particular application.
So, I'm just going to go through a few examples
of commonly occurring minerals and the kinds
of features that could be used.
A lot of this comes from sedimentary geology literature,
geologists, sedimentary geologist, they're interested
in looking at these properties of these minerals to help figure
out what kinds of rocks they came from.
Quartz from volcanic rock is going to look different
than quartz from a plutonic,
or certain types of metamorphic rocks.
So, you'll look at the grain morphology,
something you get a lot more detail about, admittedly by SEM,
but you get some information by PLM.
We've got polycrystallinity.
Quartz can occur as single crystals, it can also occur
as polycrystalline grains.
The number of crystals in that polycrystalline grain,
whether they're all the same size
or different sizes are all things that have been studied
by sedimentary geologists to attribute source rocks
to the, to the sediments.
The undulosity is the measure of the number of degrees
over which extinction occurs in quartz and that's something
that can be used to distinguish quartz from one source
from quartz from another source, inclusions
and then surface features.
So we've got a rounded quartz grain at the top,
a more angular one in the middle,
and then a nice well formed hexagonal bipyramid from a --
it's a volcanic veno cryst at the bottom,
so the morphology clearly varies quartz from source to source.
We've got a nice single crystal at the top
that showed complete extinction,
you'll have to take my word for that.
Polycrystalline grain in the middle,
and then a grain showing undulose extinction
at the bottom, these are some of the properties that vary
that we can observe in quartz.
Varieties of inclusions, both mineral inclusions,
fluid inclusions, sometimes randomly oriented,
sometimes aligned due to either processes
that occurred during crystal growth, or metamorphic episodes.
Surface features, well, these aren't really surface features
so much as we've got a rim of volcanic glass
around the top grain, it's a little hard to see,
but in the picture in the far right, it's a,
the full wave plate you might be able to make it
out along the bottom of the grain.
There's volcanic glass rim
and the other grain there has some adhering iron oxide
coatings, so that would be another potential feature
to use.
Different minerals are going to have different features,
or properties that are more useful
for distinguishing among different varieties,
or classifying them into varietal types.
Morphology's always going to be on the list, but with,
with alkalyd feldspar and other extremely common minerals
in sediments, we've got a variety
of different twinning types, untwinned grains, simple twins,
polysynthetic twinning or tartan twinning and microcline
when you've got polysynthetic twinning on two separate laws.
We've got exsolutions when at high temperatures, sodium and,
and potassium are happy to be mixed in with the high,
the high temperature crystal structure is happy
to have sodium and potassium substitute
for each other readily, and as you get
to lower temperature structure --
in the lower temperature structure, the different sizes
of these cations make it such that they diffuse
within the grain to different regions,
or domains within a crystal, and that's called,
that process is called exsolution,
it results in sodium rich and potassium rich regions
within a single grain.
You can only -- that can only happen
if the cooling is slow enough
that that diffusion can take place.
The textures you get --
these exsolution textures vary significantly depending
on the thermal history of those grains.
Another difference between quartz is the alteration,
feldspar is much less chemically stable than quartz,
so we have a lot more potential for weathering
and here we've got examples of different varieties rounded,
angular, a nice well formed euhedral crystal,
tartan twinning, microcline at the top, simple twin,
although that might be polysynthetic,
but it's certainly coarser twinning, and the bottom,
I believe, we've got a combination of exsolution,
as well as twining in our alkali feldspar example.
Fluid inclusion from a volcanic feldspar on the upper left,
we've got a mineral inclusion at the upper right,
and then this lower one is almost certainly
from a volcanic rock,
where we've got a foliation routile that's been enclosed,
included in the alkali feldspar grain.
Weathering from very fresh crystals at the top,
highly weathered in the middle, the bottom one is also coated
in a thin rim of volcanic glass,
so just some more of the variety.
Two more minerals and then we'll wrap things up.
Calcite, the biggest difference here, I think, is the,
just the tremendous range to of morphological features you get,
particularly with biogenic calcite sources,
so we've got a nice cleavage [inaudible] at the top,
an anhedral crystal, microcrystalline calcite,
foraminifera, a shell fragment, I believe,
although I'm not a micropaleontologist,
and then we've got oid,
where the calcites precipitate in a tiny grain.
You know, by XRD, these are all calcite, so --
but obviously microscopically, the fact that we had calcite
in two samples, if they look
that different should be significant to us.
And then the last example I'm going to throw in here,
tourmaline, simply because pleochroic scheme is something
that can be particularly useful for describing varieties
of particular minerals that exhibit pleochroism,
and here we have got tourmaline three crystals that show,
I think, a lot of different features, highly included,
or a high number of inclusions
in our nice euhedral tourmaline prism at the top that's sort
of a brown to colorless.
And we've got a very, very different pleochroic scheme,
sort of a brownish green blue to a pink in a very,
an anhedral poorly formed crystal,
or no recognizable crystal faces in that middle crystal,
and then we've got a nice prism that's a different pleochroic
scheme from green to colorless in the bottom.
There are going to be some, some interpretation questions
that need to be answered if I find one variety of quartz
and a variety of feldspar uncommon between two samples,
is that like finding a blue nylon and red polyester
from independent sources?
And we've got to be careful about things like that
because certain features or varieties are going
to be more likely from particular types of rocks,
so we might have these varieties similar because they're both
from volcanic rocks, as opposed to entirely different sources.
I would imagine that the different varieties
of the same mineral coming from different rock sources
and mixing would be pretty significant.
Whether we can look at the proportions
of the different mineral varieties is,
is something I'm interested in.
There was a caution in that article that I mentioned by Bull
and Morgan that to make
that claim requires certain assumptions
that may not always be valid, so I'd be interested
to see what the research indicates, but again,
as it's been a theme throughout this meeting,
we need the research before we can make these statements.
So, in summary, I'm proposing what -- or not a new concept,
but a new approach to this concept, I believe,
and I think that if we do this research, I think at best,
it might be a method that could use existing personnel
and skills with minimal training, existing equipment,
and hopefully be robust to some of these differential transfer
and persistent issues.
I think at worse, it will help the people who, like Skip,
who are already using mineral varieties in sort of a less
than quantitative way to assess the, the strength of evidence
when they do have mineral varieties in common,
or different between samples, so this type
of research I think would be potentially
of use to the community.