a
=
"
—
‘
=
_
~
"
_
'
—
~
‘—
—
=

 

THESIS

If.“

This is to certify that the

thesis entitled

FORENSIC ANALYSIS OF SOIL BASED ON
ITS ORGANIC CONTENT

presented by

John V. Goodpaster

has been accepted towards fulfillment
of the requirements for

M.S. _degree in Wstice

Major pro 55

Date 4-28-2000

 

c-7539 MS U is an Affirmative Action/Equal Opportunity Institution

 

 

LIBRARY

MIchIgan State
University

 

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

I"V 7
DEF l . 2006

“L l 7 71m)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 C'JCIRClDateDuepfifi-pJS

FORENSIC ANALYSIS OF SOIL BASED ON ITS ORGANIC CONTENT
By

John V. Goodpaster

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE

School of Criminal Justice

2000

ABSTRACT

FORENSIC ANALYSIS OF SOIL BASED ON ITS ORGANIC CONTENT
By

John V. Goodpaster

Trace evidence can be a powerful component in a criminal investigation
that may help identify a perpetrator, assist in their prosecution, or influence their
sentencing. However, the probative value of trace evidence is highly dependent
on the discriminatory power and reliability of the analysis to which the evidence is
subjected. In the case of soils, traditional tests have been able to compare and
contrast bulk properties such as color, density, mineralogy and so on. The
organic portion of soil, which is generally formed from the decay of plant and
animal matter, has been largely overlooked. Only recently have researchers
begun to focus on the analysis of this portion of soil, and methods have
remained largely unexamined and unoptimized. Areas which require
investigation and improvement include the methodology for extracting organic
matter from soils, the analysis and identification of the extracted components,
and the optimum conditions for their Chemical separation.

In this work, a systematic study of the types and conditions of analyses for
soil organic matter was completed. First, sequential Soxhlet extractions Of
various soil samples were carried out with five solvents of decreasing polarity:

water, methanol, acetone, hexane, and toluene. In general, the majority of

extractable matter was recovered with the more polar solvents. However, some
pattems were evident. For example, pristine soils which had little to no human
intervention showed a systematic decrease in the amount of extractables as
solvent polarity decreased. Perturbed soils, however, had more random
distributions of organic matter. Such patterns may be useful in distinguishing
natural versus disturbed soil layers.

Spectroscopic analysis of the extracts demonstrated that the polar
solvents contained humic substances, which are widely known but poorly
understood components of soil organic matter. Both UV-visible absorbance and
fluorescence spectroscopy gave results similar to literature spectra for humic
species. Also, it appears that different subpopulations of humic substances were
isolated in the water and methanol fractions respectively. However, despite the
information gained about the nature of the material in the extracts, the
spectroscopic results were very similar for all soils and therefore showed little
promise as a discriminatory tool. One exception was a cultivated corn field which
Showed distinctive features in both its absorbance and fluorescence spectra,
most likely due to agricultural chemicals.

Finally, the extracts were separated using reversed-phase liquid
chromatography. The water extracts did not show any interaction with the non-
polar stationary phase and so only a size-exclusion mechanism was operative
during their separation. Despite this, clear differences were seen between soils
that seemed to reflect the basic soil type. Even more success was obtained with

the reversed-phase separation of the methanol extracts which yielded complex

Chromatograms that were more easily discriminated. Analysis of the acetone
extracts yielded no separation, and these samples (along with the hexane and
toluene extracts) would be best suited for a normal-phase separation.

Overall, it appears that analysis of soil organic matter can be a successful
means to discriminate soil, including samples of the same basic soil type which
differ only in vegetative ground cover. In turn, the ability of these instrumental
methods to more uniquely characterize soil assigns more probative value than
traditional methods. Given the ubiquitous nature of soil, the further development
of powerful and reliable techniques for soil comparisons would be of great

benefit to criminal investigations.

To God and my Family, with Love

ACKNOWLEDGMENTS
“The most tangible way in which science, especially chemistry, can be
concerned with the well-being of society is its use in the maintenance of the
fabric of society as expressed in the constant vigil against crime. ”

Editor, Chemistry in Britain

Only now is it Clear to me how many people have positively influenced my
path in the forensic sciences. To begin at the beginning, my thanks go to Mr.
John Geroux of St. Thomas Academy in Mendota Heights, Minnesota for his
kind and open minded supervision during my senior year independent study into
“forensics” (as I called it then). His encouragement to pursue science with vigor
was a wonderful gift. As an undergraduate, my experience as an Honors lntem
at the FBI Academy in Quantico, Virginia under the supervision of Dr. Edward
Bartick was invaluable. My time there quickly solidified my desire to pursue a
career in criminal justice and law enforcement. My undergraduate advisor, Dr.
Larry Potts of Gustavus Adolphus College in St. Peter, Minnesota, was an
extraordinary mentor who provided me with many and varied opportunities to
develop my interests in forensic science. My four years spent under Larry’s
tutelage helped to form me as an analytical scientist and researcher. His
concern for his students and teaching has earned my utmost admiration and
respect.

The project upon which this thesis is based would not have been
successful nor have even occurred without the dedicated and generous effort of

a number Of people at Michigan State University. My advisor Dr. Jay Siegel,

vi

whose previous research and ongoing enthusiasm about soil analysis has
spurred this effort, has been a continual source of advice, encouragement, and
vision. I also deeply appreciate his full support of my status as the first dual-
degree candidate in forensic science and analytical chemistry at Michigan State.
That status has given me the extreme fortune of working with Dr. Victoria
McGuffin, who provided invaluable assistance and guidance for this project. Her
insights into the chemical problems of soil extractions, spectroscopy and
separations are directly reflected in the content and structure of this study. I
would also like to acknowledge the thoughtful and constructive feedback of the
other members of my thesis committee: Dr. Merry Morash and Dr. Frank Horvath
of the School of Criminal Justice, and Chris Bommarito of the Michigan State
Police Laboratory.

Furthermore, I am very grateful for the assistance of Dr. Steven Boyd and
Dr. Delbert Mokma of the Department of Crop and Soil Sciences. Dr. Boyd
provided valuable expertise at the beginning of the project on soil chemistry. Dr.
Mokma was an energetic and selfless guide during our sampling scheme,
including accompanying us as we tromped through corn fields, climbed fences,
and ducked barbed wire where necessary. Dr. Mokma also provided essential
equipment and information for Obtaining, profiling and processing our samples.
Dr. Kathy Severin Of the Department of Chemistry generously allowed us the use
of departmental instrumentation and supplies, as well as patiently working with

us as needed.

vii

Last, and certainly not least, are two undergraduates whose dedication,
hard work, and long hours truly made this project feasible. Elizabeth Croal
worked as a first year student on the extraction and spectroscopic analysis of the
standard soil, demonstrating the utmost competence and efficiency. Amy
McAdam, an upper level student, distinguished herself through two years of work
on this project. Amy amazed me with her industriousness, commitment to
excellence, and positive attitude. Without her extradorinary efforts, the extent
and scope of this research would have been dramatically curtailed.

Finally, I would like to thank those many people who provided crucial
support to me during my time in graduate school. Thanks to Mom, Dad, Beth
and Katie for their unconditional love, concern, and support. Never before have I
realized how much you mean to me. Thanks to Uncle Bob, Aunt Chris and the
rest of the “Michigan Goodpasters” for their boundless generosity and hospitality
and for always keeping me humble at Scrabble. Thanks to the many friends I
have made at MSU, especially Tom Cullen, Dan Hopkins, Sam Howerton, Peter
Krouskop, Mike Sanregret, and Dan Severs. You have all made this experience
much more bearable and I appreciate that immensely. A special thanks to E.D.,
E.W., K.H., and SC. for the gift of their love. Finally, and most importantly, I
thank God for the many times when there were only one set of footprints in the

sand.

viii

TABLE OF CONTENTS

 

 

 

LIST OF FIGURES ............................................................................................... x
1. INTRODUCTION... .... _ ..... _ ...................................................................... 1
1.1 SOIL ................................................................................................................. 1
1.2 THE VALUE OF SOIL AS PHYSICAL EVIDENCE ....................................................... 2
1.3 ESTABLISHED METHODOLOGIES .......................................................................... 3
1.4 CHROMATOGRAPHIC ANALYSIS OF SOIL ............................................................... 4
1.5 HUMIC SUBSTANCES .......................................................................................... 7
2. EXPERIMENTAL METHODS ............................. - ______ .11
2.1 CHEMICALS ..................................................................................................... 11
2.2 SAMPLING ....................................................................................................... 11
2.3 SEQUENTIAL SOXHLET EXTRACTIONS ................................................................ 1 1
2.4 UV-VISIBLE ABSORBANCE SPECTROSCOPY ....................................................... 12
2.5 FLUORESCENCE SPECTROSCOPY ...................................................................... 12
2.6 HIGH PERFORMANCE LIQUID CHROMATOGRAPHY ............................................... 13
3. RESULTS AND DISCUSSION _ _ -- _ -- - ............... 15
3.1 SEQUENTIAL SOXHLET EXTRACTIONS ................................................................ 15
3.2 UV-VISIBLE ABSORBANCE SPECTROSCOPY ....................................................... 17
3.3 FLUORESCENCE SPECTROSCOPY ...................................................................... 22
3.4 HIGH PERFORMANCE LIQUID CHROMATOGRAPHY ............................................... 45
4. CONCLUSIONS AND FUTURE DIRECTIONS ............................................... 55
5. REFERENCES .................... ....... -- _______ . .................... 59

 

LIST OF FIGURES

Figure 1: Flow chart for the soil analysis procedure ............................................ 14
Figure 2: Results of sequential Soxhlet extractions ............................................ 16
Figure 3: UV-visible absorbance spectra of standard soil extracts ...................... 18
Figure 4: UV-visible absorbance spectra of cornfield soil extracts ...................... 19
Figure 5: UV-visible absorbance spectra of grass lawn soil extracts .................. 20
Figure 6: UV-visible absorbance spectra of forest soil extracts .......................... 21
Figure 7A: Fluorescence excitation-emission matrix of the standard soil water
extract (91 ug/mL) ......................................................................... 25
Figure 7B: Fluorescence excitation-emission matrix of the cornfield soil water
extract (216 ug/mL) ....................................................................... 26
Figure 7C: Fluorescence excitation-emission matrix of the grass lawn soil water
extract (135 ug/mL) ....................................................................... 27
Figure 7D: Fluorescence excitation-emission matrix of the forest soil water
extract (337 ug/mL) ....................................................................... 28
Figure 8A: Fluorescence excitation-emission matrix of the standard soil methanol
extract (98 ug/mL) ......................................................................... 29
Figure 88: Fluorescence excitation-emission matrix of the cornfield soil methanol
extract (195 ug/mL) ....................................................................... 30
Figure 80: Fluorescence excitation-emission matrix of the grass lawn soil
methanol extract (114 ug/mL) ........................................................ 31
Figure 8D: Fluorescence excitation-emission matrix of the forest soil methanol
extract (164 ug/mL) ....................................................................... 32
Figure 9A: Fluorescence excitation-emission matrix of the standard lawn soil
acetone extract (170 ug/mL) .......................................................... 33
Figure 98: Fluorescence excitation-emission matrix of the cornfield soil acetone
extract (1341 ug/mL) ..................................................................... 34
Figure 90: Fluorescence excitation-emission matrix of the grass lawn soil
acetone extract (636 ug/mL) .......................................................... 35
Figure 9D: Fluorescence excitation-emission matrix of the forest soil acetone
extract (325 ug/mL) ....................................................................... 36

Figure 10A: Fluorescence excitation-emission matrix of the standard soil hexane

extract (37 ug/mL) ......................................................................... 37

Figure 103: Fluorescence excitation-emission matrix of the cornfield soil hexane

extract (112 ug/mL) ....................................................................... 38

Figure 100: Fluorescence excitation-emission matrix of the grass lawn soil

hexane extract (52 ug/mL) ............................................................. 39

Figure 10D: Fluorescence excitation-emission matrix of the forest soil hexane

extract (124 ug/mL) ....................................................................... 40

Figure 11A: Fluorescence excitation-emission matrix of the standard soil toluene

extract (181 ug/mL) ....................................................................... 41

Figure 118: Fluorescence excitation-emission matrix of the cornfield soil toluene

extract (1996 ug/mL) ..................................................................... 42

Figure 110: Fluorescence excitation-emission matrix of the grass lawn soil

toluene extract (1894 ug/mL) ......................................................... 43
Figure 11D: Fluorescence excitation-emission matrix of the forest soil toluene
extract (1660 ug/mL) ..................................................................... 44
Figure 12: Comparison of water extract Chromatograms .................................... 48
Figure 13: UV-visible absorbance spectra of soil water extracts at retention time
16.5 min ............................................................................................. 49
Figure 14: Comparison of methanol extract Chromatograms .............................. 50
Figure 15: UV-visible absorbance spectra of soil methanol extracts at retention
time 2.0 min ....................................................................................... 51
Figure 16: Comparison of methanol extract Chromatograms at 210 nm ............. 52
Figure 17: Comparison of methanol extract Chromatograms at 220 nm ............. 53
Figure 18: Comparison of methanol extract Chromatograms at 230 nm ............. 54

xi

1. INTRODUCTION

Despite previous research, instrumental methods are typically not used to
analyze and compare soils in forensic laboratories. This is due in large part to a
lack of knowledge and understanding of the best sample preparation conditions
for soils, an effective means of separating soil components, or detailed
characterization of those components. Hence, this project has sought to
address these deficiencies in a number of ways. First, a sampling scheme was
devised and carried out to acquire soils of differing base soil type as well as
vegetative cover. A systematic study of extraction solvent was then conducted to
determine which solvent yielded the greatest quantity and/or most discriminating
soil components. UV-visible absorbance and fluorescence spectroscopy were
then used to characterize the soil extracts and compare the results to humic
substances. Finally, chromatographic analysis of the water, methanol, and

acetone soil extracts was completed and separation conditions were Optimized.

1.1 Soil

Soil, in its broadest definition, is a loose, heterogeneous, living mixture of
inorganic and organic material formed on the earth’s surface.1 This mixture may
contain solid, liquid and gaseous phases and is constantly exchanging both
energy and matter with its environment. Many materials of natural and
anthropogenic origin may be found in a given sample of soil, including but not

limited to minerals, microorganisms, ash, decomposed plant and animal tissues,

glass, paint, building materials, plastic, fertilizer, pesticides, and hydrocarbons in
varying concentrations.2

The naturally occurring portion of soil is formed in one of two ways. First,
residual soil is formed in place via the weathering of exposed rock and the decay
of organic matter. Second, transported soil is material moved via wind, rain, or
glacial action into place. Overall, the natural soil layer can vary in thickness from
10'3 to 102 m.‘

There are many different types of soil, as well as local variations in both
the horizontal and vertical dimensions. Recent surveys of the United States
have revealed some 20,000 soil types. These types are generally differentiated
based on their particle size distribution, i.e., sand, silt and clay in decreasing
order. Natural variations in these types can be caused by changes in lithology,
which describes geologic induced changes such as streams or glacial action.
Alternatively, large variations can be caused by changes in pedology, which

describes the thickness and moistness of the soil layer.1

1 .2 The Value of Soil as Physical Evidence

To the forensic scientist, soil is more specifically defined as collected
earth material that is relevant to the matter under investigation.‘ As evidence,
soil possesses a number of characteristics which make it valuable. First, soil is
widespread. Hence, it is likely to be found at the scenes of serious crimes,
provided its value is recognized and it is properly collected. Soil is also diverse.

Many different types of soils are known, and they vary considerably from location

to location. In addition, soils have been extensively surveyed and Classified for
agencies such as the US. Department of Agriculture and the US. Geological
Survey. This can aid the forensic scientist in assessing the frequency of a
particular soil type or in locating a region where that type may be found. Finally,
soil evidence is differentiable, in that known natural variations can be
distinguished through microscopic or chemical analysis even to the point of
providing individual Characteristics to some samples.1

In criminal investigations, soil comparisons can form circumstantial links
between a suspect and a crime scene. In addition, the composition of soil tends
to reflect its surroundings. Hence, soil identification can reveal information about
a perpetrator’s environment. Lastly, soil analysis may reveal temporal
information such as a sequence of events if there are discernible seasonal

changes in a soil sample, or if stratigraphy (layering) has been preserved.1

1.3 Established Methodologies

Usually, forensic soil comparisons begin with a microscopic examination
Of the sample after it has been suitably dried. Note that a drying procedure can
change some characteristics of a soil such as salt content, color, mineral
oxidation state, nitrate content, or microbial content and activity. Nevertheless,
such microscopic examinations can add value to a particular comparison by
identifying the more unusual particles in a sample. These comparisons almost
always include examination of the sample under normal and polarized light to

determine the distribution of minerals in the soil.3'4 More recently, the highly

discriminatory technique of scanning electron microscopy/energy dispersive x-ray
spectroscopy has been used to distinguish and Chemically analyze individual
grains of soil samples.5 In this case, it becomes crucial to obtain representative
samples from unknown and control soils. If samples are not representative,
there could be discernible differences between samples originating from the
same source.

Measurement of bulk properties of a soil can also be used for comparison.
Towards that end, a host of techniques have been brought to bear in forensic
soil analysis.1 These include examination and quantitation of soil color,6'7’8
moisture, density gradient, particle size distribution,9'1o acidity,11 conductivity, and

refractive index. Instrumental techniques such as x-ray diffraction,“13

x-ray
fluorescence,14 differential thermal analysis,13 atomic spectroscopy, and infrared
Spectroscopy13 are also popular analysis methods. Overall, for a soil comparison
to fully achieve its potential, the methodologies used should provide an
appropriate amount of discriminating power. Furthermore, it must be realized

that no one technique is sufficient in and of itself to provide a complete

comparison.1

1.4 Chromatographic Analysis of Soil

In all of the above instrumental techniques, the focus lies with the
inorganic/mineral content of soils and their subsequent classification by soil type.
Any organic comparisons conducted are simple measurements of the total

organic content of a soil or its weight loss upon ignition. However, the potential

of using organic compounds found in soils for their comparison has been
increasingly recognized. There are many organic molecules present in soil,
including enzymes, carbohydrates, ketones, lipids, amino acids, organic
phosphorous, organic sulfur, and aromatic hydrocarbons.‘ Of these potential
analytes, biochemical assays of soil enzymes15 and chromatographic analysis of
extractable aromatic hydrocarbons from soil have received the most attention
from the forensic community.

The first example of the latter was reported by Andrasko,16 where
polycyclic aromatic hydrocarbons (PAHS) were extracted from urban and rural
soil samples with methanol and then analyzed using high performance liquid
chromatography (HPLC). PAHS consist of two or more fused aromatic rings, and
are widespread and frequently carcinogenic environmental contaminants. PAHS
are formed whenever organic matter is exposed to heat, such as in combustion
or geological processes. Andrasko identified numerous PAHS in the soil
samples based on their retention times and UV-visible absorbance spectra.
Examples included fluoranthene, pyrene, chrysene, benzo(e)pyrene and
benzo(e)pyrene, with all of the PAHS found at higher concentrations in the urban
samples.

Overall, the method showed good reproducibility and sensitivity but lacked
Chromatographic resolution. Andrasko’s approach was able to discriminate
between samples taken 100-150 m apart, whereas samples taken 5-10 m apart
were largely indistinguishable. No major changes in PAH composition with

depth, time, exposure to UV light, or heating were found. In general, changes in

PAH composition were gradual and tended to reflect the geographic position of a
sample rather than the type of soil.

Reuland and Trinler17 further developed this technique via the analysis of
acetonitrile extracts of soils. Samples from seven different locations were
compared and were found to be distinguishable either qualitatively (in their
number and retention time of peaks) or quantitatively (in their relative peak
heights). In addition, the authors noticed some discernible differences over short
distances within a cultivated field (1 -3 m). This illustrates that this technique can
discriminate local variations within the same soil type. However, while all of the
samples could be differentiated quantitatively, not all could be differentiated
qualitatively and chromatographic resolution was not significantly improved from
Andrasko’s results.

Later, Siegel and Precord18 used a similar extraction methodology on
samples from varying locations. They modified the method of detection to
include the absorbance ratio of the soil components at 254 and 280 nm. All the
samples could be differentiated quantitatively, but only half could be
differentiated qualitatively. Unfortunately, the authors found that the use of
absorbance ratios did not significantly contribute to the discriminating power of
the technique.

I.19 revisited the possibility of

Finally, subsequent work by Reuland eta
using absorbance ratios in the Chromatographic analysis of methanol extracts
from soil samples. In this study, three different absorbance ratios were used

together with single wavelength detection. Out of 91 possible comparisons, 82

were classified as both qualitatively and quantitatively different. Five
comparisons yielded quantitative differences only and four comparisons
indicated chromatographically equivalent samples. The use of ratiograms in this
work tended to support the qualitative or quantitative similarity or dissimilarity of
the samples. Of the four comparisons classified as chromatographically
equivalent in the above work, three were based on samples taken from specific
locations within the same sampling area. In the last case both samples were
taken from heavily traveled roadways, which would further support the
hypothesis that this method strongly reflects the geographic location of a sample
rather than its soil type. Overall, only one out of 91 comparisons showed
chromatographic equivalence between two samples known to be different. This
gives some measure of the rate with which this technique would yield false

positive results.

1.5 Humic Substances

Despite the above developments, there remains a lack of understanding
of the chemical nature of the extracted components from soils. In particular, only
certain PAHS have been identified in soil extracts, whereas a large amount of
other material that may be of value is also present. For example, components of
the naturally occurring humus layer have shown promise for discriminating soil
samples. Humus is the fraction of the organic content of soil remaining after
most animal and plant residues have decomposed. In forested areas, humus

generally lies beneath a 1-2 inch layer of undecomposed leaves and twigs and

above the major soil layer of the area. Humic acid is one of the substances that
is derived from this soil layer and is defined as a mixture of dark-colored organic
matter that can be precipitated by acidifying a dilute-alkali extract Of soil. In
contrast, fulvic acid is also derived from the humus layer but tends to have a
smaller molecular weight and is soluble in water at all pH values.1

The explicit chemical structures of humic and fulvic acid are unknown.
However, elemental characterization of a humic acid from highly weathered coal
found the elements C, O, H, N, and S in order of decreasing concentration. In
addition, 1H NMR results revealed a high concentration of aromatic protons as
well as polycyclic aromatic and quinone functionalities, which reflect the high
degree of humification, or degradation of the substance. However, only trace
amounts of long aliphatic chains, saccharides, methoxyl groups, amines or
olefins were found.20 Subsequent NMR studies also showed that in general,
humic acids tend to have more aromatic carbons, fewer carbohydrate carbons,
and fewer aliphatic protons than fulvic acids.21

Not surprisingly given the aromatic character of the material, humic
substances are fluorescent. The humic acid discussed above absorbed in the
region 300-450 nm and had a featureless emission band centered at 535 nm. It
was noted that this emission maximum was shifted to longer wavelength versus
other humic acids, presumably due to the high degree of aromaticity in the
sample.20 Both the distribution of fluorescence lifetimes22 and phase-resolved
excitation-emission matrices23 of humic acids have also been used to study their

properties. These studies indicated systematic changes in the fluorescence of

humic substances with pH and the ability to compare humic acids from various
sources based on their fluorescence lifetimes.

Vibrational spectroscopy has long been used to determine functional
group information about humic substances. For example, IR spectra of the
weathered coal humic acid discussed above gave evidence for amide, aromatic,
ketone, ester, and carboxylic acid functional groups.20 Early attempts to acquire
Raman spectra of humic substances suffered from a large fluorescence
background but did give results similar to disordered graphite-like materials.24
More recent use of Fourier transform and surface enhanced methods has
allowed information about the carbon backbone of humic substances to be
acquired and has shown that various humic substances from different sources
had very similar backbones.25

Pyrolysis gas chromatography-mass spectrometry has also been used to
characterize humic substances.”27 While the majority of pyrolysis products
formed were light gases such as CO, 002, or CH4, other major pyrolysis
products were aromatic in nature. These included alkylbenzene derivatives and
heterocyclic compounds. However, it is difficult to discriminate between aromatic
building blocks of the original humic substance and those that were formed
during pyrolysis. Independent measures of the aromaticity of the substances
using 13C NMR seem to indicate that aromatic rings with varying lengths of alkyl
chains were important constituents in the humic substance. A comparison of
pyrolysis to other chemical means of digestion showed that it is a valuable tool

with greater ease of use and fewer experimental variables.28

Finally, traditional separation techniques have been used to resolve the
complex mixtures of humic substances found in soils. The most common
method is sizeoexclusion chromatography, which fractionates samples on the
basis of molecular weight. Studies of this kind are generally designed to
determine the Characteristics of different size fractions of naturally occurring
humic acids. However, and of relevance to this work, more recent studies have
indicated that the molecular weight distributions of humic acids may vary
systematically between different soils.29'34 Reversed-phase chromatography
using gradient elution has also been successfully applied to humic and fulvic
acids by separating components based on their relative polarity.”37 As soil
samples contain many unidentified components of uncertain structure,
systematically varying the strength of the mobile phase using a gradient

becomes important for achieving high resolution.

10

2. EXPERIMENTAL METHODS

2.1 Chemicals

A commercially available standard reference soil was obtained for initial
studies (Resource Technology Corporation, CLN SOIL-3). Water used for
Soxhlet extractions was distilled and deionized (Corning Glass Works, Model
MP-3A). Toluene (Aldrich), methanol, acetone, and hexane (Baxter Healthcare,
Burdick and Jackson Division) were high purity, spectroscopic-grade solvents.
These solvents were also used as mobile phases, in addition to HPLC grade

water and 1,4-dioxane (Aldrich).

2.2 Sampling

Three soil samples were acquired from a cultivated cornfield, a nearby
grass lawn adjacent to a rural road, and a wood lot, all on the lands of Michigan
State University in East Lansing, MI. The cornfield and grass lawn sample were
classified as Mariette loam soils while the wood lot sample was classified as a
Colwood-Brookston loam soil.38 Samples were acquired using a solid core
sampler. The upper 10 cm of the core was isolated, air-dried, crushed using a
mortar and pestle, passed through a 2 mm sieve, and stored under refrigeration

(see Figure 1 for a flow chart of the entire experimental procedure).

2.3 Sequential Soxhlet Extractions

Samples of the standard, grass lawn, com field, and forested area soils

weighing ~20 g each were extracted using a Soxhlet apparatus. The extractions

11

lasted for 24 hours with a sequential series of the solvents (water, methanol,
acetone, hexane, and toluene). The samples were placed in double thickness
cellulose thimbles (Whatman, 80 x 20 mm) to prevent silt particulates from
leeching into the extracts. Despite this precaution, the water extracts required
centrifugation (lntemational Equipment Company, Model CL) and the methanol

extracts were filtered using 0.45 um disk filters (Alltech) in order to remove

insoluble material. The solvent was then removed from the extracts by rotary
evaporation in pre-weighed round bottom flasks. Finally, the residue was
reconstituted into the original extraction solvent and transferred to volumetric

flasks.

2.4 UV-visible Absorbance Spectroscopy

UV-visible absorbance spectra of the soil extracts were acquired with a
commercially available spectrometer (ATI, Unicam UV2) in a 1-cm cuvette cell.
The spectrometer contained a deuterium and a tungsten lamp, a photodiode

detector, and was operated with a 2 nm bandpass.

2.5 Fluorescence Spectroscopy

Fluorescence excitation and emission spectra, as well as excitation-
emission matrices (EEM) of the soil extracts were acquired with a commercially
available spectrometer (Hitachi, Model F-4500) in a 1-Cm cuvette cell. The
fluorimeter contained a 150 W xenon lamp, a photomultiplier detector, and was

operated with a 2.5 nm bandpass.

12

2.6 High Performance Liquid Chromatography

Samples were filtered before analysis using 0.45 urn disk filters (Alltech).

Chromatograms were obtained with a commercially available Chromatograph
(Hitachi). The system consisted of a gradient pump (L-6200A), autosampler
(AS-4000), and diode-array absorbance detector (L-4500A). The system used a
reversed-phase octadecylsilica column (Supelco, Hypersil ODS, 250 x 4.6 mm, 5

urn particles, 120 A pore size) with attached guard column (Alltech, Opti-guard

C18).

13

 

 

 

 

 

 

 

 

 

 

 

STANDARD CORNFIELD GRASS LAWN FOREST

 

 

 

 

 

 

isolate top 10 cm of core sample

I

 

 

dry, crush, Sieve
l

 

 

sequential Soxhlet extractions

/

water fraction *-—* UV, EEM spectroscopy HPLC

methanol fraction '— UV, EEM spectroscopy '— HPLC l
I acetone fractiorq— UV, EEM spectroscopy HPLC

hexane fraction ~—- UV, EEM spectroscopy

l toluene fraction UV, EEM spectroscopy

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 1: Flow Chart for the soil analysis procedure

14

3. RESULTS AND DISCUSSION

3.1 Sequential Soxhlet Extractions

The results of the sequential extractions are summarized in Figure 2 in the
form of weight fraction extracted as a function of solvent and soil sample. Based
on these data a number of conclusions can be made. In general, the amount of
extractable material decreased as the polarity of the solvent decreased. This
reinforces what is known about humic substances, which are widely thought of
as large molecular weight polyelectrolytes which would preferentially dissolve in
aqueous and/or polar solvents. Furthermore, any inorganic salts would also be
extracted in these solvents. The latter, less polar solvents would tend to extract
non-humic substances such as carbohydrates, lipids or large fatty acids that are
also found in the organic portion of soil. The profile or pattern formed by these
data seems to indicate that more pristine soils that are largely free from human
activity (the standard and forest soils) show similar profiles, while the grass lawn
and cornfield do not. In particular, the soil isolated from the grass lawn showed a
large amount of material in its acetone extract, which may reflect a larger
proportion of lipid-like material relative to humic substances in the soil. Finally,
the cornfield was largely depleted of organic matter relative to the nearby grass
lawn, which is perhaps indicative of the time of sample collection (after the fall
harvest). Overall, polar solvents were the most successful at extracting more
organic material, and each soil gave a fairly distinctive pattern of extractables in

the various solvents; although less perturbed soils tended to show similar

15

AMOUNT EXTRACTED

(mg/g soil)

   
   
 

 

 

 

 

 

 

 

05- _ Forest
0- Grass Lawn
Cornfield
Standard

Water
Methanol

o
c o
2 C
m a
0 a)
< I

Toluene

Figure 2: Results of sequential Soxhlet extractions

16

Shapes. However, this type of pattern could be used as a marker for soils that

are free of human activity.

3.2 UV-visible Absorbance Spectroscopy

As the concentration of the extracts varied with both solvent and sample,
UV-visible spectra are shown in Figures 3-6 in the form of absorptivity as a
function of wavelength. In general, the material in the water, methanol and
hexane fractions showed the greatest absorptivity. Unfortunately, the strong
absorption of acetone and toluene below 328 and 286 nm, respectively, makes
comparisons of the material in these solvents difficult. The largely featureless
shape of the absorptivity curves in water and methanol are typical for humic
substances as reported in the literature.39 Interestingly, the spectra of the
hexane extracts also have this similar shape, with one important exception. The
hexane extract from the cultivated cornfield absorbed only weakly, but showed
distinctive structure that may be due to aromatic compounds present in the soil
(see Figure 3, the data has been multiplied by a factor of 10 for comparison).
These compounds could be present as the result of fertilization, or the use of
herbicides or pesticides. Such a result may indicate a potential means to

discriminate cultivated soils from other, non-agricultural samples.

17

0.06

 

0.05

 

 

 

0.04 ~ L..- ———-

 

 

0.03

 

0.02 _. ~

 

ABSORPTIVITY (mU(ug*cm))

 

 

 

 

 

200 250 300 350 400
WAVELENGTH (nm)

 

— Water (91 ug/mL)
-— Methanol (98 1.19/le
-— Acetone (170 ILQ/mI-I
—- Hexane (37 ug/mI-I
Toluene (181Il9/mI-I

 

 

 

Figure 3: UV-visible absorbance spectra of standard soil extracts

18

0.06

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0.05
E
«a, 0.04 ~—
3-
:] \
E
E 0.03
2
I'-
o.
8
a) 0.02
m
<
0.01 J
0 . I . 4r ~ I I I
200 250 300 350 400
WAVELENGTH (nm)
— Water (216 ug/mL)
— Methanol (39 ug/mL)
._ Acetone (1341 ug/mL)
-- Hexane (112 ug/mL, x 10)
--~ Toluene (1996 ug/mL)
Figure 4: UV-visible absorbance spectra of cornfield soil extracts

19

0.06

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0.05 ~~ - _-___
’E‘
.9 0.04 +~——- ---___. -- ~
O)
E}
:I
E
t 0.03 -- —- - ~-
E
I—
o.
6
a) 0.02 —
on
<
0.01 -
0 I I -.-- . . T .
200 250 300 350 400
WAVELENGTH (nm)
— Water (135 ug/mL)
— Methanol (114 ug/mL)
—- Acetone (3182 ug/mL)
-- Hexane (52 ug/mL)
----~ Toluene (1894 pg/mL)
Figure 5: UV-visible absorbance spectra of grass lawn soil extracts

20

0.06

 

0.05

 

 

0.04 eeeee

 

 

 

0.03 I

0.02

 

ABSORPTIVITY (mng*cm))

0.01

 

 

 

 

 

 

200 250 300 350 400
WAVELENGTH (nm)

 

— Water (67 ug/mL)
— Methanol (33 ug/mL)
- Acetone (325 ug/mL)
- Hexane (124 ug/mL)
...-.. Toluene (1660 ug/mL)

 

 

 

Figure 6: UV-visible absorbance spectra of forest soil extracts

21

3.3 Fluorescence Spectroscopy

Fluorescence excitation-emission matrix (EEM) spectra for each extract
are shown in Figures 7-11. These figures show contour plots of the fluorescence
of the sample with excitation wavelength shown along the vertical axis and
emission wavelength shown on the horizontal axis. The contour lines represent
steps of 1.0 units in fluorescence intensity. In addition, two cross sections taken
at the excitation and emission maxima of the contour plot are shown on the lower
right and upper right, respectively.

In general, the excitation and emission maxima for all samples fell within
the ranges of 290 - 340 nm and 350 - 440 nm, respectively. Also, with few
exceptions, emission maxima tended to decrease with the polarity of the
extraction solvent. It should be noted that EEM data are particularly sensitive to
the concentration of the extract. This wavelength dependence is due to inner
and outer filtering effects, whereby strong absorption of the excitation and/or
emission beam tends to shift the measured excitation and emission maxima to
longer wavelengths where such absorption is less prevalent. This makes
comparing and/or discriminating samples more difficult, but some chemical
insight can be gained from the results obtained in this study.

For the water and methanol extracts, a broad and weak emission band is
seen which is similar to published fluorescence spectra of humic substances.39
Furthermore, the emission maxima Of the methanol extracts were at shorter

wavelengths than the corresponding water extracts. This may indicate that the

22

material in the water fraction has a higher degree of humification. Chemically,
this implies that the humic components possess a greater degree of aromatic
conjugation (therefore emitting fluorescence at a longer wavelength) in addition
to a greater number of polar groups (therefore increasing their water solubility).

All the acetone extracts showed a large, intense emission band around
420-440 nm which did not vary significantly between samples. Subsequent EEM
spectra of pure acetone showed weak, intrinsic fluorescence in this region,
although not of an intensity to explain the large fluorescence signal seen in the
soil extracts. It is more likely that this signal is due to fluorescence from
condensation products of acetone formed during the 24 hour extraction.

No discernible fluorescence was seen in the hexane extracts, with the
important exception of the cornfield soil. In that case, a structured fluorescence
emission was seen around 380 nm. This further supports the UV-visible
absorbance data by indicating the presence of aromatic compounds in this
cultivated soil.

The toluene extracts, as in acetone, showed a relatively intense emission
band that was largely invariable between samples. The single exception was the
forest soil extract, which had excitation and emission maxima at longer
wavelengths than other soils. This may indicate an increased amount of
conjugated aromatic functionalities in this soil. However, it must be noted that
pure toluene shows weak fluorescence in approximately the same region.
Furthermore, it is possible that a large increase in this signal could occur upon

extraction due to reactions of the solvent to form fluorescent products.

23

Therefore, in the case of both acetone and toluene subtracting a solvent
background which has been refluxed over 24 hours may be useful in
compensating for any inferences.

Overall, the ability of fluorescence spectroscopy to differentiate these
samples is rather limited as the spectra obtained were quite broad and
featureless. The notable exception is the cornfield soil, where trace amounts of

aromatic compounds yielded a clear spectrum with distinct structure.

24

SEE: 5v Swim 22m; :8 Emu—LS» 9: __o xEmE co_mw_Eo-co_§_oxo 3:88.02“. “<5 239....

06mm

Sm

c.09m

 

 

ooo.o

0.0.2”
xm

 

oo.om

 

o.omv

NM

o.oom

 

 

ooo.o

E

0.33

E

 

oo.om

 

0.0mm

 

 

 

 

 

 

 

 

 

 

 

 

 

.om

o.oom

o.om¢

25

 

 

 

 

 

 

 

 

o

0

Ln

<1!

><

to

o

o
o o 02 o o o
o - o o - o
- ZOE o - was 0
o mm - o [xiv-I -
In v o In m o

 

 

 

 

 

 

 

 

26

 

 

650.0

300.0

Figure 7B: Fluorescence excitation-emission matrix of the cornfield soil water extract (216 ug/mL)

 

 

450.0
650.0

 

 

 

 

 

 

o o

o o

o o
o o o o o o

o - I::oN o - om
- o o . o o
0 Eng - o EH5 -
Ln <1- 0 Ln «1 o

o

( 1(a)

 

 

    
 
  

     
     
   

 

(if, é

 

 

 

   

 

 

 

 

'.; INIA”
. .1.- E: 3K6} ;
*STS‘J\\\ ~—.
5
>< o S,
Iii .

O
N

450.0

27

Figure 7C: Fluorescence excitation-emission matrix of the grass lawn soil water extract (135 ug/mL)

SEE: 58 59on 583 :8 $99 9: E xEmE :o_mm_Eo-co_§_oxo 858902”. ”an 9:9“.

0.0mm

2m

o.o

 

 

 

 

o.om¢

o.o

 

 

 

 

on
909.0

EC
o.ovm
xm

oo.om

om
ooo.o

EC
o.omv
2m

oo.om

0.0mm

 

 

 

fl

 

 

 

 

 

 

o.om¢

28

SEE: we 89:8 _ocm£oE :8 2356 on. .o xEmE 5328952386 858202". ”(m 239“.

0.0mm

Sm

o.oom

 

 

Z

ooo.o

E

0.0mm

ifs

oo.om

 

 

o.om.v

xm

o.oom

 

 

25.0

E

0.0:”

5

 

oo.om

 

o. omm<

®o. com

 

 

 

 

o. com

 

 

 

 

 

 

 

 

o.omv

 

29

 

 

 

 

 

 

 

 

o o

o o

In In

«I to

o o

s s
o o o o o o

o - o” o - <:>M
- 505 o - o o
o v-I - o v -
In an o In M o

o

“‘___,_ :3

[,W

 

 ,\\\

 

; A V

 

 

‘ t
n,

 

450.0

30

    
 

ill I“ {IA

0%

Figure BB: Fluorescence excitation-emission matrix of the cornfield soil methanol extract (195 ug/mL)

SEE: .3 5 69:8 _ocm£oE __om Esm. mem 9: E xEmE co_wm_Eo-co_§_oxo 8:83.02". ”0w 2:9”.

  

 

 

 

 

 

 

 

 

 

90$ 2m 0.8m Roms 0.8m
89° Room
ES
93%
xm
8.8 \m
Roma xm Room 3 -. Hm
ooo.o \
S/ . no

EC
o.oaw
Sm

 

 

 

 

 

 

 

 

oo.om o.omv

 

 

 

31

SEE: em: 89:8 _o:m£oE :8 892 9: E waE co_mm_Eo-co_§_oxo 858202: How 2:9”.

o.omw

Em

o.oom

 

 

000.0

E
0.03.”
xm

oo.om

 

 

o.omv

o.oom

 

 

ooo.o

EC
o.ovv
Em

 

oo.om

 

0.0mm 2m

 

 

 

”it.“

 

 

 

o.oom
o.oom

 

o.omv

 

32

SEE: DE 6ng 3058 :8 Emucfim 65 E xEmE co_wm_Eo-co_§_oxo 858903.“. 35 939....

0.0mm

Sm

o.oon

 

 

000.0

E:
0.0mm
xm

 

oo.om

 

o.omv

o.oom

 

 

ooo.o

En
0.02»
Sm

 

oo.om

 

        

OK”.
HI. .. «HIM» H“ .
. I t. MINA/I‘ll It.)
I ‘

T\I\(I

(a S a,
(Jugs P

7
a.

III‘

”WIN.” ..
I’Ubliltu ....
ya I
“sum/II fill
tit"

Em o.oom

 

             
     

:5... a i

?

  

«(I
\3;
3 TS.

if
I?
gig -..
g

 

 

 

 

 

//,

\.

 

 

 

 

Odom

xm

o.omv

33

SEE: Sum: .0858 8:288 :8 92:50 85 E xEmE co_mm_E8-co_§_oxo 8:88.852”. ”mm 859”.

o.omm

o.oom

 

 

ooo.o

EC
o.o¢m
xm

 

oo.om

 

o.omv

o.oom

 

 

ooo.o

o.omv
2m

 

oo.om

 

0.0mm 2m 0.00m
o.oom

 

 

 

 

 

 

 

o.omv

 

 

34

 

 

 

 

 

 

 

 

 

O

0

Ln

<9

0

8
8 °. SN 8
. 05 o .
m M ' In
\0 V O \D

<_/\I/

 

 

 

 

 

\

 

 

450.0

35

650.0

 

Figure QC: Fluorescence excitation-emission matrix of the grass lawn soil acetone extract (636 ug/mL)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

  
      
      
    

  

 

         

 

it E
sis Ea sis is"
,2“.

/.._/-\ v}
/F‘//L~: {V II
L ivy/w
IIII’II'I
\ fill). .7
‘ EWII I'flx .2;
Vi"
IIIIII’I: II fig;
°. 8 98

36

Figure 9D: Fluorescence excitation-emission matrix of the forest soil acetone extract (325 ug/mL)

SEE: 58 69:8 9.38: :8 2856 m5 _o 5.38 co_wm_Em-co_§_oxm 8:83.02“. ”<2 9:9...

o.omm

0.00m

 

 

ooo.o

E:
o.o~.m

oo.om

 

 

o.om¢

o.oo~

 

 

000.0

5

0.0mm

5

 

oo.om

 

0.0mm

x 2m o.oom

 

Odom

fl/ -

 

Ev

Nm

 

 

 

 

 

 

o.om¢

 

 

37

SEE: NF 5 aombxm 388: :8 292.8 9.: 6 x59: co_mm_Eo-co_§_oxo 858202“. ”mop 9:9”.

0.0mm 2m o.oom 0.9mm
ooo.o

a
o
c;
o
m

 

 

E

o.or~

2.5

 

 

 

 

oo.om

 

o.omv xm o.oom
ooo.o

 

E

o.omm

 

5

oo.om

 

 

 

 

 

 

 

 

 

38

 

SEE: va fiomzxm mcmxm; =8 :32 mmmE m5 6 xEmE co_mm_Em-co_§_oxm 8:88.02“. ”Dow 9:9”.

o.omm

2m o.oom

 

 

ooo.o

E

c.0hm

E5

 

oo.om

 

o.omv

xm o.oom

 

 

ooo.o

E

0.0mm

5

oo.om

 

 

90$

Em o.oom

 

     

H'fiIIHIII

H ‘

1':
l
|
‘l

I

A

    

/

  
                      

(NIL: E“;
‘7. I.
’31:; HI;
. .

‘l
I.

 

_ ail/l

 

1”}.

I; :j.

1

I!

m.
I!"

- fig?
h;

§

[
I 7 '1-1 ‘\ '
\‘ I
\ ‘. ..‘ \S r
I: 5 ". r \'
I \: "I

 

   

  

.5"

 

- '-
I

 

  

2f

 

 

 

 

 

NJ

 

 

 

 

o.oom

Xm

o.omv

39

 

SEE: :N: 89:8 8588: :08 88.2 85 :0 xEmE :o_wm_E8-:o_§_ox8 88808882“. ”cop 859“.

o.omw

2m o.oom

 

(A

 

III;

 

 

 

 

06m:

ooo.o

5

c.0hm

i3

oo.om

xm o.oom

 

 

 

 

 

 

 

009.0

E

0.0mm

5

oo.om

c.0mm

,8 98m

 

 

 

Odom

 

 

l/x/

 

 

 

m
m

 

 

 

0.0m:

 

 

40

SEE: 5: 89:8 8:829 .88 99:88 85 8 xEmE co_mm_E8-co_8=ox8 8288902". ”<3 9:9“.

 

 

 

 

 

 

 

osmw zm odor. 93.8 2m 0.8m
89° , /.. o.oo~
I ,« .w' i. . ~ .\\0.
1 M ,r ftp/us \r.»
ca
0.03
xm
8.0m QM
came :8 :62 xm
89°
5
98;
5
8.3. came

 

 

 

 

 

 

 

 

 

41

SEE: 8885 89:8 8:825 .88 28:98 85 8 x58E :o_mw_E8-c059_ox8 898888.82: ”m E 859".

o.omw

o.oom

 

 

ooo.o

EC
o.oom
xm

 

oo.om

 

o.om¢

o.oo~

 

 

000.0
E:

o.omm
2m

oo.om

 

 

9.0mm 2m o.oom

 

MIMI.“ II,
5'" LIIInII».fl.flWfl«J.A|tflHH%u¢Nlllmwwvxl» ,
ttUIled)mflflKfivmmMHl#\I

 

 

 

 

 

 

 

Nm

o.omv

42

SEE: :88: 89:8 8:829 .88 :52 889E 85 8 5sz :o_8m_E8-c8598x8 88888882". ”0: 859...

 

 

         

o.om8 zm o.oom o.om8 =8 o.oom
ooo.o I- ,» I ., ”a y c.oo~

 

 

”MINI , . - I.
$1! \I‘! I‘lpnv.m,u\l\ltlurvfltnIIII .L «I n! I
E amt? nflu:k~f¥hll‘nl.frllwtrll
. Iuaflfil‘l’j «ll. VIII-LI (til-1‘
NW

E
Im..1

 

 

 

oo.om /rIJI\

 

 

 

o.ome x8 o.oom a- /\//I>e\2 x:, x8
ooo.o I.‘
o.omm

Sm \

.<*}

/AS\
I
I IIIHHH M

 

 

 

 

 

 

 

 

 

oo.om A»; , o.om¢

 

 

43

SEE: com: 8858 8:822 .88 892 85 8 x58E co_8m_E8-co=9_ox8 82888882". no I 859”.

0.0mm Em o.oom o.omw 2m o.oom
ooo.o .

 

 

Ed
o.owm
Nm

 

 

oo.om

 

 

came :8 odom x8
:85

 

1?:-

Ed
0.0m:
Em

’12»: ;« z:

‘figtftigim‘

o.omv

 

 

 

 

 

oo.om

 

 

3.4 High Performance Liquid Chromatography

The results of chromatographic analysis for the water and methanol
extracts are shown in Figures 12 - 18. The separation of the water extracts was
particularly difficult and a number of mobile phases were tried without success,
including isocratic methanolzwater mixtures, methanolzwater gradients, and
methanol:0.5% (v/v) acetic acid in water gradients. Eventually, it became clear
that the only detectable components were eluting prior to the void volume of the
system. This indicates that the components of the water extracts were not
interacting significantly with the stationary phase of the column and were being
excluded from the 120 A pores of the silica particles, thereby creating a size
exclusion separation mechanism. Such a mechanism was successfully exploited
by using a 98% water/2% methanol mobile phase at a flow rate of 0.1 mL/min

and a 20 uL injection volume (see Figure 12). These results showed that the

water-soluble humic substances present in the soils could be differentiated
based on their size and, therefore, molecular weight distributions. Furthermore,
the two soils that were located on the same base soil type (the grass lawn and
corn field) both show similar Chromatograms, as would be expected.

Since the chromatographic system was equipped with a diode array
detector, full UV-visible absorbance spectra were acquired of the column
effluent. The spectra for the chromatographic peaks discussed above are shown
in Figure 13. These largely featureless spectra are similar for all extracts.

Furthermore, they are characteristic for humic species and help confirm that the

45

water-soluble organic matter is primarily made up of large molecular weight
humic substances that possess many polar and/or ionizable functional groups.
in turn, this results in no significant interaction with a C18 stationary phase and
high water solubility.

The Chromatograms for the methanol extracts are shown in Figure 14.
Optimization of the separation conditions for these samples indicated that a
mobile phase gradient gave the best results. The data shown uses a linear
methanolzwater gradient from 60 to 100% methanol over 30 minutes with a 15

minute hold at 100% methanol. A 0.85 mL/min flow rate was used with a 20 uL

injection volume. As with the water extracts, the major component in these
extracts is a non-retained peak that is being excluded from pores in the
stationary phase particles. The UV absorbance spectra of these peaks are
shown in Figure 15 and are also very similar between extracts. The presence of
a more defined peak at 230 nm differentiates these samples from the water
extracts. Together with the fluorescence data discussed above, it is apparent
that two separate fractions of humic substances are being isolated in the water
and methanol fractions respectively.

Of greatest importance to this work, however, is the presence of a number
of small, well-defined peaks eluting after 10 minutes in these Chromatograms.
These peaks, shown at 210, 220 and 230 nm in Figures 16-18 show high
potential for discriminating the soil samples. The wavelength that produced the
most complex, and hence the most easily differentiable, Chromatograms was

230 nm. These peaks were found to be reproducible as well as not being

46

present in a blank injection of solvent. It also should be noted that the samples
had to be concentrated above 1 mg/mL in order to achieve sufficient signal-to-
noise ratios for these components.

As a final experimental note, the acetone extracts for the soil samples
were also analyzed. Unfortunately, while a number of mobile phases were used
(consisting of methanol:1,4-dioxane), none produced any discernible retained

peaks, nor was any size-exclusion mechanism noted.

47

mm

on

8E9Eo8EoEo 89:8 8.95 8 comcmano

“NF 859....

 

 

3585 3.9 .68 2858'
3585 F. c 2958'
SEEE 8.3 :33 8890.1:
3585 s. 5 .68 sec“. 1..

 

 

mm

@8555: 8.2:
om m.

o_.

 

 

 

 

 

 

 

JI
Um

 

 

(mm 023) aoweuosev An

48

UV ABSORBANCE (A.U.)

0.5

 

. w’WW

. W'
J- Ivft'»~wx'v

 

 

 

 

 

 

 

 

 

 

250 300 350
WAVELENGTH (nm)

 

.......... Forest Soil (1.7 mg/mL)
”Grass Lawn (0.68 mg/mL)
--Comfield (1.1 mg/mL)

—Standard Soil (0.45 mg/mL)

 

 

 

Figure 13: UV-visible absorbance spectra of soil water
extracts at retention time 16.5 min

49

 

400

mm

8E9E28E950 89.xw 8:858: 8 comtmano ”3 85mm

 

seas em: .68 2888'
3:85 1: 2958......
SEEE v.9 :33 8890...?
SEEE v.3 :8 88883....

3855.5 8.2:
on mm cm or or

n n p _ P

 

 

 

 

 

 

 

(wu ozz) 30Nveuosev An

50

UV ABSORBANCE (A.U.)

0.5

0.4

.0
o.)

.0
N

 

 

 

P

 

 

 

 

 

 

 

 

 

 

 

WAVELENGTH (nm)

 

WForest Soil (4.1 mg/mL)
mGrass Lawn (3.4 mg/mL)
"Cornfield (1.4 mg/mL)

 

 

—Standard Soil (2.9 mg/mL)

 

Figure 15: UV-visible absorbance spectra of soil methanol
extracts at retention time 2.0 min

51

400

on

E: o E 8 mEmESmEoEo 89:8 8:858... 8 88:88:80 ”or 9:9”.

mm

 

 

3282 8.8: .68 2888'
3:82 2: 2828'
SEEE v.9 :33 8890'
SEEE 3; =98. 888m'

 

 

8235.5 8.2:.
om

r _ r

mp

or

 

(mu orz) ao~veuosev An

52

E... 0mm 8 8E9888E98 89:8 8:858E 8 8.28an0 K. 9:9“.

3:82 as .68 2888'
3:82 8: 2828......
SEEE 8.8 $58.. 890%
3:82 :3 .68 828......

82:85 8.2:
8 om mm om m:

P p _ p b _ . . p p b p _ P . P p _ . p P

 

 

 

 

 

 

Essafiflmw 3.41%..

48%; . u

 

 

 

o_.

 

(mu ozz) ao~veuosev An

53

0:

mm

 

E: omm 8 8E9E88EoEo 89:8 8:858E 8 com_88Eoo

 

 

SEEE 8.8V .88 9888'
SEEE v.5 285E801...
SEEE 8.8 $58.. 88903.3...
SEEE 73 .88 8888......

 

 

om

 

 

82:58 8.2:
mm

_

o

N

”mp 9:9“.

9.

 

or

 

(mu oez) aoweaosev An

54

4. CONCLUSIONS AND FUTURE DIRECTIONS

The use of soil as evidence in criminal investigations will surely increase if
techniques are developed which more fully characterize them. Currently, the
analytical focus has remained on the inorganic portion of soil while the use of the
organic component of soil as a source of information has been limited. In this
work, a systematic study of the extraction and analysis of soil organic matter has
been undertaken.

Sequential extraction of soils using various solvents has shown that the
fractionation patterns of soils may be a useful characteristic of comparison.
Furthermore, Soxhlet extractions have proven to be a convenient, exhaustive,
and quantitative method for isolating fractions of a soil sample. However, it
appears that all pristine soils may show similar patterns. This method also has
the disadvantages of being highly time consuming, amenable only to neutral
solvents and requiring lengthy filtering and concentration of extracts for further
analysis. Some alternative approaches to extraction may be more applicable to
soil samples as well as the context of a forensic laboratory. For example, the
use of ultrasound could greatly accelerate extraction times, or supercritical fluid
extraction could also efficiently isolate fractions of soil samples.

The analysis of soil extracts using UV-visible absorbance and
fluorescence spectroscopy gave informative results but unfortunately they were
not highly discriminating. For example, the results showed that the nature of the

humic substances in the water and methanol extracts differed, but little

55

difference was seen between soil types. Only in cases where a soil contains a
unique component (e.g. potential agricultural chemicals in the cultivated corn
field) would absorbance and fluorescence analysis discriminate one sample from
others. Future researchers would be well advised to examine other
spectroscopic techniques such as FTIR spectroscopy, surface-enhanced Raman
spectroscopy, or NMR spectroscopy. These techniques have been successfully
applied to the analysis of humic substances and may be useful for distinguishing
soil extracts after samples have been suitably prepared.40

Lastly, the separation of soil extracts using HPLC showed the greatest
potential for discriminating soils. The analysis of water extracts was based on a
size exclusion mechanism and was able to differentiate soils based on their
underlying soil type. A more efficient size-exclusion separation with a column
whose pore size is optimized for the molecular weight range of humic
substances would likely improve these results greatly. In addition, the use of
FT lR as a detection method could yield more informative results than UV
absorbance. However it must be remembered that such an analysis separates
only on the basis of molecular weight distribution rather than chemical
composition.

The most promising results were generated using a reversed-phase
separation of the methanol extracts. It was clear that there are a number of
methanol soluble components in soils that can be successfully separated using a
gradient elution approach. As long as care is taken to sufficiently concentrate

the samples, this method should be a powerful innovation to soil comparisons.

56

Furthermore, there may be even more discriminatory compounds in the non-
polar extracts (acetone, hexane and toluene). In these cases, the use of a
normal-phase separation may generate the necessary chromatographic
resolution. Finally, another separation technique that has demonstrated
potential for humic substances is capillary electrophoresis, which would be an
efficient and rapid method well suited for aqueous soil extracts.‘I1

Despite these accomplishments and reccomendations, it is important to
note that a thorough and quantitative assessment must be made of the
discriminatory power of this technique. Such an assessment is needed in order
to determine the ability of this approach to successfully compare soil evidence in
criminal investigations. Two main issues to be addressed are discerning
differences between soil samples which originate from slightly different locations
and discerning changes that may occur at a single location over time.
Furthermore, some quantitative criteria must be established for how much
similarity must exist between two samples for them to be deemed as likely
originating from the same source. By conducting a larger and more detailed
sampling study over a suitable period of time, these questions could be resolved
and the practical application of this technique encouraged.

Overall, this study examined the types and conditions of analysis that are
required for forensic soil analysis. Not all methods demonstrated the ability to
differentiate soils, but it has been shown that some do show that potential.
Specifically, both the relative amounts of extractables as a function of solvent

polarity and the separation of these extracts according to molecular weight or

57

solubility yield distinctive patterns. These developments, as well as future work
introducing other spectroscopic or separation methods and developing
quantitative means of comparing data, can only continue to improve the analysis

of soil.

58

5. REFERENCES

 

—L

599°.‘19’9‘PP’N

29.
30.

31.

32.
33.

RC. Murray and J.C.F. Tedrow, Forensic Geology, Prentice Hall:
Englewood Cliffs, 1992.

H. Demmelmeyer and J. Adam, Forensic Sci. Rev. 7, 119 (1995).

WC. McCrone, Microscope 40, 109 (1992).

N. Petraco, Am. Lab. 26, 35 (1994).

M.J. McVicar and W.J. Graves, J. Can. Soc. Forensic Sci. 30, 241 (1997).
R.J. Dudley, J. For. Sci. Soc. 15, 209 (1975).

RR. Antoci and N. Petraco, J. For. Sci. 38, 437 (1993).

R. Sugita and Y. Marumo, For. Sci. Int. 83, 201 (1996).

S. Wanogho, G. Gettinby, B. Caddy, and J. Robertson, J. For. Sci. 34,
823(1989)

S. Wanogho, G. Gettinby, and B. Caddy, For. Sci. Int. 33, 117 (1987).
R.J. Dudley and K.W. Smalldon, For. Sci. Int. 12, 49 (1978).

Y. Marumo, S. Nagatsuka, and Y. Oba, J. For. Sci. 33, 1360 (1988).

Y. Marumo, S. Nagatsuka, and Y. Oba, J. For. Sci. 31, 92 (1986).

Y. Hiraoka, J. For. Sci. 39, 1381 (1994).

J.l. Thornton and AD. McLaren, J. For. Sci. 20, 673 (1975).

J. Andrasko, International Microform Joumal of Legal Medicine 13, 209
(1978)

DJ. Reuland and WA. Trinler, For. Sci. Int. 18, 201 (1981).

J.A. Siegel and C. Precord, J. For. Sci. 30, 511 (1985).

DJ. Reuland, W.A. Trinler, and MD. Farmer, For. Sci. Int. 52, 131 (1992).
Z. Tao, Y. Yang, and F. Sheng, Tox. Env. Chem. 49, 45 (1995).

Z. Tao, J. Zhang, J. Zhai, Anal. Chim. Acta 395, 199 (1999).

LE. McGown, S.L. Hemmingsen, J.M. Shaver, and L. Geng, Appl.
Spectrosc. 49, 60 (1995).

S.L. Hemmingson and LB. McGown, Appl. Spectrosc. 51, 921 (1997).
Y. Yang, 8. Li, and Z. Tao, Spectrosc. Left. 27, 649 (1994).

Y. Yang and HA Chase, Spectrosc. Left. 31, 821 (1998).

J. Poerschmann, F.—D. Kopinke, G. Balcke, and S. Mothes, J.
Microcolumn Sep. 10, 401 (1998).

J. Zhang, J. Zhai, F. Zhao, and Z. Tao, Anal. Chim. Acta 378, 177 (1999).
G. Chiavari, G. Torsi, D. Fabbri, and 6.0. Galletti, Analyst 119, 1141
(1994)

L. Rottmann and K.G. Heumann, Anal. Chem. 66, 3709 (1994).

R. Artinger, G. Buckau, J.l. Kim, and S. Geyer, Fresenius J. Anal. Chem.
364,737(1999)

J.E. Kilduff, T. Karanfil, Y. Chin, and W.J. Weber, Jr., Environ. Sci.
Technol. 30, 1336 (1996).

A. Piccolo, S. Nardi, and G. Concheri, Chemosphere 33, 595 (1996).
J. Peuravuori and K. Pihlaja, Anal. Chim. Acta 337, 133 (1997).

59

 

34.

35.
36.
37.
38.

39.

40.
41.

A.V. Kudryavtsev, l.V. Perrninova, and VS. Petrosyan, Anal. Chim. Acta
407,193(2000)

F.Y. Saleh, W.A. Ong, and D.Y. Chang, Anal. Chem. 61, 2792 (1989).
X. Liu and OK. Ryan, Environ. Technol. 18, 417 (1997).

J.M. Lobbes, H.P. Fitznar, and G. Kattner, Anal. Chem. 71, 3008 (1999).
Soil Survey of lngham County, Michigan, United States Department of
Agriculture Soil Conservation Service in cooperation with the Michigan
Agricultural Experimental Station.

P. MacCarthy and J.A. Rice, Humic Substances in Soil Sediment and

Water: Geochemistry, Isolation, and Characterization, G.R. Aiken, D.
McKnight, R.L. Wershaw, and P. MacCarthy, Eds., John Wiley and Sons:

New York (1985).

Y. Yang and HA Chase, Spectrosc. Left. 31, 821 (1998).

P. Schmitt-Kopplin, D. Freitag, A. Kettrup, N. Hertkom, U. Schoen, R.
Klockig, B. Helbig, F. Andreux, and A.W. Garrison, Analusis 27, 390
(1999) ‘

60

IIIIIIIIIIIIIIIIIIIIIIIIIIIIIII

llljllljjjljjiljjll(Illjljlll‘l