Distribution of PAHs among Sediment Size
Fractions Determined by Thermal Desorption
Gas Chromatography Mass Spectrometry
Jejal Reddy Bathi1, Robert Pitt1, Robert Findlay2,
Shirley E. Clark3
1Department
of Civil, Construction and Environmental Engineering
University of Alabama, Tuscaloosa, AL 35487
2Department of Biological Sciences, University of Alabama
Tuscaloosa, AL 35487
3Penn State - Harrisburg, School of Science, Engineering and
1
Technology, Middleton, PA - 17507
Overview of Presentation
Introduction
 Thermal Desorption (TD) Method
Development
 Sample Collection and Analyses
 Conclusions
 Question and Answers

2
Introduction
Steps for PAH analyses
 Sample preparation
 Extraction and Concentration
 Detection and Quantification
 Liquid-liquid extraction by separatory funnel
and Solid Phase Extraction (SPE) are the
most common PAH extraction methods for
liquid samples.
 SPE method is ineffective for liquid samples
with high concentrations of solids.

3
Introduction


For solid samples, Soxhlet, automated Soxhlet, and
ultrasonic extraction are the most common PAH
extraction methods.
Disadvantages of traditional methods include,
 Time consuming
 Labor intensive
 Use of large amounts of toxic solvents
 Exposure of toxic organic solvents to the
operators
4
Introduction - Fugacity Modeling

Contaminant partition constants (KOW, KOC, Henry’s constant) are
key factors in their fate.
Fugacity Level I equilibrium model (Mackay et al. 1992) was used for
predicting the phase partitioning of PAHs. The following graph
illustrates the effect of KOW and KOC on partitioning onto solids.
Log (Kow)
Log (Koc)
100
% of PAH with Solids

80
60
40
20
0
2
2.5
3
3.5
4
4.5
5
5.5
Log (Kow), Log (Koc)
6
6.5
7
7.5
5
After Hurricane Katrina, and other
devastating hurricanes of the 2005 season,
EPA reviewed several monitoring and
cleanup programs in the days following the
events. The selection of the right analytical
and sampling methods for water and
sediment were all found to be very critical
in reducing the risk to the cleanup crews,
emergency workers, and residents.
Katrina oil refinery damage (EPA photo)
Fuel and other material leakage from
damaged boats. Hurricane Katrina
(NOAA photo)
Waste sorting area with field labs (EPA
photo)
6
Analytical Method Development




The main transport mechanisms and fate of PAHs in the aquatic
environment are closely related to their major associations with
particles. Therefore, it is essential that rapid and sensitive analytical
tools be used that can measure PAHs in samples having large
amounts of suspended sediment.
During the aftermath of natural disasters, homeland security
incidents, and accidental releases of hazardous and toxic materials, the
rapid analyses of samples is needed to identify areas for priority
cleanup and in preventing dangerous exposures to cleanup personnel
and residents.
Analytical tools need to be developed and tested on a variety of
samples to ensure their applicability to a wide range of emergency
situations.
With these objectives in mind, a TD-based GC/MSD method was
developed and tested using urban stream sediments for PAH
contamination.
7
TD Method Development

The TD analytical method is rapid, uses no solvent, and is
less labor intensive than other PAH analytical methods,
especially for samples having high sediment concentrations.

TD uses elevated temperatures as a means to directly
transfer the analytes from solid sample matrices to the
gaseous analytical system of the GC/MSD.
The desorbed analytes are concentrated in a cyrotrap at
the head of the GC column and are then re-vaporized into
the column for separation and final detection.

Final desorption temperature and temperature holding
times are two important factors to be optimized for better
recovery of PAHs from solid matrices.

8
Direct thermal extraction and
analysis technique
SIS Thermal desorption unit
9
TD Method Development


A desorption temperature of 350oC produced the largest
peak areas for the individual PAHs, for the tested range of
temperatures 250oC to 375oC with 25°C step changes.
A final temperature holding time of 15 min produced
maximum peak areas for the individual PAHs in the
resulting chromatograms.
10
TD Method Interferences
Chromatogram obtained by TD/GC/MS
method for the standard NIST PAH sediment
sample which was collected at Chesapeake Bay
at the mouth of Baltimore Harbor

Other major findings included the addition of small amounts of
copper to avoid interfering sulfur compounds and freeze drying the
samples to reduce the moisture content in the samples to prevent ice
blockages in the GC column. Local urban stream sediments were
11
sufficiently dried using a conventional drying oven.
TD Method Detection Limits

Calculated Amount of Analyte (ng)

DL = Y0 + SyZα
(McCormick and Roach 1987)
NIST sediment was tested in
the weight range 3 – 60mg
R2
Detection
Limit (ng)
Naphthalene
98.5
9.63
Fluorene
96.4
2.62
Phenanthrene
97.1
6.37
Anthracene
94.1
3.83
Fluranthrene
98.8
2.67
60
Pyrene
93.2
9.81
50
Benzo(a)anthracene
96.8
2.25
40
Chrysene
96.5
2.63
Benzo(b)flouranthrene
97.6
4.15
Benzo(a)pyrene
97.5
4.2
Indeno(1,2,3-cd)pyrene
93.1
1.34
Dibenz(a,h)anthracene
52.2
1.05
Benzo(ghi)perylene
97.1
0.97
PAH
Naphthalene
80
y = 1.2481x + 6.1919
R2 = 0.9854
70
30
20
10
0
0
10
20
30
40
50
Amount of Analyte in NIST Standard (ng)
60
12
TD Method Recovery Calculations
% Recovery
from Solid
Samples
Acceptable Range of %
Recovery from EPA
Methods (Aqueous
Samples)
Acceptable Range of %
Recovery from Standard
Methods (Aqueous
Samples)
Naphthalene
125
D – 122
21 – 133
Fluorene
142
D – 142
59 – 121
Phenanthrene
110
D – 155
54 -120
Anthracene
104
NG
NG
Fluranthrene
62
14 – 123
26 – 137
Pyrene
33
D – 140
52 – 115
Benzo(a)anthracene
109
33 – 143
33 – 143
Chrysene
140
17 – 168
17 – 168
Benzo(b)flouranthrene
35
24 – 159
24 – 159
Benzo(a)pyrene
41
17 – 163
17 – 163
PAH
Indeno(1,2,3-cd)pyrene
34
NG
NG
Dibenz(a,h)anthracene
46
NG
NG
Benzo(ghi)perylene
43
NG
NG
13
Effect of Recovery by Particle Size

Concentrations of PAHs in three replicates of two coarse samples were
compared with PAH concentrations in the same samples that were
ground to <180 µm to test the effect of particle sizes on PAH recovery.
Only one PAH of one sample was significantly affected by grinding.
PAH
Naphthalene
Fluorene
Phenanthrene
Anthracene
Fluranthrene
Pyrene
Benzo(a)anthracene
Chrysene
Benzo(b)flouranthrene
Benzo(a)pyrene
Indeno(1,2,3-cd)pyrene
Dibenz(a,h)anthracene
Benzo(ghi)perylene
ANOVA P Value (95% C.I) of Comparing Concentrations
Comparing the Coarser 710 - 1400µm Comparing the Coarser 1400 - 2800µm
with Its Ground Sample
with Its Ground Sample
0.122
0.128
0.064
0.118
0.618
0.052
0.776
0.204
0.786
0.135
0.516
0.076
0.052
0.368
0.36
0.249
0.342
0.409
0.048
0.45
0.175
0.67
0.376
0.294
14
0.100
0.660
Sampling Sites

Sediment samples were collected from three
urban creeks in the Tuscaloosa and
Northport, Alabama, areas
Site 1: Cribbs Mill Creek
 Source areas: medium density two story
family home residential area
 No history of sanitary sewage contamination
(Pitt, et al 2005)
Site 2: Hunter Creek
 Source areas: automobile service commercial
areas, heavy traffic along McFarland Blvd.,
and runoff from trailer park residential areas
Site 3: Carroll Creek
 Source areas: A residential area on one side
and forested lands on the other side of the
creek
 Has a recent history (in 2006) of sanitary
sewer overflows (SSOs) into the creek
(ADEM Consent Order No. 07-139-CWP to
City of Northport)
15
Sample Particle Size Distributions

Five sediment samples were collected from each of the three
creeks.
All sediments were fractionated into nine size (µm) ranges <45,
45-90, 90-180, 180-355, 355-710, 710-1400, 1400-2800,
>2800(w/o large organic matter (LOM)) and > 2800 (LOM only).
Hunter Creek
% Less than

Cribbs Mill Creek
Carroll's Creek
100
90
80
70
60
50
40
30
20
10
0
10
100
1000
Size (um)
10000
16
COD Associations by Particle Size
COD analyses were conducted to indicate the amount of organic
material in the samples (a rapid and commonly available analytical
method), an important factor in PAH associations.
Median mass of the sediment COD was associated with 355 µm
particles.


Carroll's
Hunter
Cribb's Mill
Carroll's
Cribbs
2.E+08
100
1.E+08
80
1.E+08
COD (mg/Kg)
% of Total COD Less Than
Hunter
60
40
1.E+08
8.E+07
6.E+07
4.E+07
20
2.E+07
0.E+00
0
10
100
Size (um)
1000
10000
< 45
45 -
90
90 -
180
180
5
0
00
00
- 35 55 - 71
- 14 00 - 28
3
4
710
1
Size Range (µm)
> 28
00
COD in large organic material fraction >2800 µm (LOM), COD much larger than in
other sample sizes:
Cribbs Mill Creek
Hunter Creek
Carroll's Creek
1.60E+09
1.50E+09
1.30E+09
17
Thermal Chromatography used to Identify Major Sample
Components
Material Lost at These
Temperatures
Temperature (oC)
up to 104
Moisture
104 - 240
Paper debris
240 - 365
Leaves and grass
365 - 470
Rubber
470 - 550
Asphalt
Above 550
Remaining material is inert
Carroll's
20
18
16
14
12
10
8
6
4
2
0
< 45
45
-9
0
90
180
180
355
355
710
0
0
M)
140
280
LO
o
0
w/
710
140
0(
0
8
>2
Size Range (µm)
Temperature range of 104 – 550 °C
(total volatile fraction)
Hunter
Carroll's
104 - 240
8.6
20.4
240 - 365
42.5
33.5
365 - 470
28.3
3.9
470 - 550
1.3
0.9
% Inert Material
19.2
41.4
Hunter
Cribb's
% Of Weight Loss
% of Weight Loss
Hunter
LOM % Weight Loss
Temperature
Range (ºC)
Carroll's
Cribbs
18
16
14
12
10
8
6
4
2
0
5
<4
45
-9
0
90
-1
80
018
5
35
535
0
71
071
)
0
00
80
OM
14
-2
oL
/
0
0
w
14
0(
80
2
>
Size Range (µm)
Temperature range of 240˚C – 365 ˚C
18
(mostly leaves and grass fraction)
COD vs. Expected Leaves and Grass Fraction
Cribbs Mill Creek
Hunter Creek
2.E+08
y = 5E+06x + 5E+07
1.E+08
R2 = 0.808
9
1.E+08
Log(COD)
COD (mg/Kg)
1.E+08
P < 0.005
8.E+07
6.E+07
7
Log(y) = 1.0739Log(x) + 6.9425
R2 = 0.805
P < 0.005
5
4.E+07
2.E+07
0.E+00
0
4
8
12
16
20
24
3
-0.5
0
% Total Weight Loss
0.5
1
1.5
Log(% Total Weight Loss)
Log(COD)
9
7
Creek
Log(y) = 1.1694Log(x) + 6.9732
P < 0.005
3
0
0.5
1
Log(% Total Weight Loss)
Carroll’s Creek
1.5
2
2.5
ANOVA P Value of
Regression
Slope
Constant
Cribbs Mill Creek
< 0.005
< 0.005
Hunter Creek
< 0.005
< 0.005
Carroll's Creek
< 0.005
< 0.005
2
R = 0.9393
5
2
19
6000
5000
4000
3000
2000
1000
0
<
45
45
0
-9
90
-
0
18
0
18
-
5
35
5
35
-
0
71
0
71
0
40
-1
00
14
0
80
-2
>
00
28
Le
es
av
Size Range (µm)
Fluorene
3500
Concentration (µg/Kg)
3000
2500
2000
1500
1000
500
0
<
45
45
0
-9
90
80
-1
0
18
55
-3
5
35
10
-7
0
71
0
40
-1
00
14
0
80
-2
>
00
28
a
Le
s
ve
Size Range (µm)
Anthracene
4000
Concentration (µg/Kg)
PAH concentration
trends with sediment
size are similar to the
trends for COD and
the leaves and grass
fractions.
3000
2000
1000
The smaller
(<180μm) and larger
(>710μm) particles
were found to have
higher concentrations
of PAHs than the
intermediate sizes
6000
Concentration (µg/Kg)
Concentration (µg/Kg)
7000
5000
4000
3000
2000
1000
0
<
45
45
0
-9
90
80
-1
0
18
55
-3
5
35
10
-7
0
71
0
40
-1
00
14
0
80
-2
>
0
80
-2
>
0
80
-2
>
00
28
a
Le
s
ve
Size Range (µm)
Benzo(g,h,i)perylene
8000
7000
Concentration (µg/Kg)
Concentrations
of PAHs by
Particle Sizes
8000
Dibenz(a,h)anthracene
7000
6000
5000
4000
3000
2000
1000
0
<
45
45
0
-9
90
80
-1
0
18
55
-3
5
35
10
-7
0
71
0
40
-1
00
14
00
28
a
Le
s
ve
Size Range (µm)
Indeno(1,2,3-cd)pyrene
4000
Concentration (µg/Kg)
Naphthalene
9000
3000
2000
1000
0
0
<
45
45
0
-9
90
80
-1
5
35
10
-7
0
71
0
40
-1
00
14
Size Range (µm)
0
80
-2
>
00
28
Le
e
av
s
<
45
45
0
-9
90
80
-1
0
18
55
-3
5
35
10
-7
0
71
0
40
-1
Size Range (µm)
00
14
00
28
20
a
Le
s
ve
20000
15000
10000
5000
0
<
45
45
0
-9
90
80
-1
0
18
55
-3
5
35
10
-7
0
71
0
40
-1
00
14
0
80
-2
>
00
28
a
Le
s
ve
Size Range (µm)
% Mass Association with Sediment
Fluorene
25000
20000
15000
10000
5000
0
<
45
45
0
-9
90
80
-1
0
18
55
-3
5
35
10
-7
0
71
0
40
-1
00
14
0
80
-2
>
00
28
Almost all of the
sediment PAHs are
associated, by mass, in
the intermediate
particle size fraction
(90 to 710 µm), even
though this size range
had the lowest
concentrations
es
av
Le
Size (µm)
Size Range (µm)
% Mass Association with Sediment
% Mass Association with Sediment
25000
Dibenz(a,h)anthracene
35000
30000
25000
20000
15000
10000
5000
0
<
45
45
0
-9
90
5000
0
<
45
90
0
18
55
-3
5
35
10
-7
0
71
5
35
10
-7
0
71
0
40
-1
00
14
0
80
-2
>
00
28
0
80
-2
>
0
80
-2
>
a
Le
s
ve
50000
40000
30000
20000
10000
0
% Average
dry weight
of Sediment
<
45
45
0
-9
90
80
-1
0
18
55
-3
5
35
10
-7
0
71
0
40
-1
00
14
00
28
a
Le
s
ve
Size Range (µm)
-
00
14
Size Range (µm)
00
14
-
00
28
>
00
28
Le
es
av
<45
45 – 90
90 - 180
180 - 355
355 - 710
710 - 1400
1400 - 2800
Large organic
matter
1.1
2.8
14.4
46.4
27.4
3.5
1.8
% Mass Association with Sediment
% Mass Association with Sediment
10000
80
-1
55
-3
Indeno(1,2,3-cd)pyrene
15000
0
-9
0
18
Benzo(g,h,i)perylene
Anthracene
20000
45
80
-1
Size Range (µm)
% Mass Association with Sediment
Mass of PAHs
by Urban
Sediment
Particle Sizes
Naphthalene
30000
35000
30000
25000
20000
15000
10000
5000
0
<
0.3
45
45
0
-9
90
80
-1
0
18
55
-3
5
35
10
-7
0
71
0
40
-1
Size Range (µm)
00
14
00
28
21
a
Le
s
ve
Conclusions




When discharged into the aquatic environment, PAHs are mostly
associated with particulate matter, rather than in the dissolved
form.
Thermal desorption (TD) is rapid, minimizes the likelihood for
human errors and contamination, and uses no solvents.
The leaf and grass content, COD and PAH concentrations in the
urban creek sediments showed similar trends: smaller and larger
sized particles were associated with higher concentrations
compared to the intermediate sized particles.
However, most of the mass of the PAHs were associated with the
intermediate particle sizes (90 – 710 µm) in the stream sediments
which were in the greatest abundance, even though their PAH
concentrations were the lowest.
22
Acknowledgments



NSF EPSCoR, COSS research center for funding this research
work
Fellow graduate students in the department of civil engineering
at The University of Alabama
Disclaimer:
This material is based upon work supported by the National
Science Foundation under Grant No. EPS-0447675. Any
opinions, findings, and conclusions or recommendations
expressed in this material are those of the author(s) and do not
necessarily reflect the views of the National Science Foundation.
23
Thank you
Questions?
24