United States
Environmental Protection
Agency
Environmental Monitoring
Systems Laboratory
P.O. Box 93478
Las Vegas NV 89193-3478
EPA 600/X-92/001
January 1992
Research and Development
&ER&
Evaluation of a Portable
Supercritical Fluid
Extraction Apparatus
-------
PREFACE
.\1
•-4 This is the final report for Sub-task 2 of Task 70.29 for FY 1991, EPA Contract 68-CO-0049,
. . conducted at Lockheed Engineering & Sciences Company. The project was directed by Neal Amick.
^ This report was written by Neal Amick. Technical support for the project was provided by John
t>) Zimmerman and Don Hilke. We would like to acknowledge the assistance of Bob Wright of Battelle
Pacific Northwest Laboratories who provided the prototype instrument used in this study.
111
-------
ABSTRACT
A method for extracting semi-volatile organic compounds from soil samples was evalusted as part
of an ongoing effort by the U.S. Environmental Protection Agency to identify alternative techniques for
characterizing contamination in soils from hazardous waste sites. Current methods include Soxhlet
extraction and sonication. These methods are time-consuming, require large volumes of solvent, and
generate a substantial amount of waste. Supercritical fluid extraction (SFE) was evaluated in this study
using a prototype unit designed for field use. The technique was compared to the established EPA
methods and was evaluated for accuracy and precision. The results of this study indicate SFE is an
appropriate method for use in the field as a screening method
IV
-------
CONTENTS
Notice ii
Preface iii
Abstract iv
Figures vi
Tables vii
1.0 Introduction 1
2.0 Conclusions 2
3.0 Experimental 3
3.1 Apparatus 3
3.1.1 Supercritical Fluid Extraction System 7 . . 3
3.1.2 Gas Chromatograph 4
3.1.3 GC/MS System : 4
3.2 Materials 4
3.3 Sample Description 6
3.4 Procedures 1
3.4.1 SFE Procedure 1
3.4.2 Soxhlet Procedure 1
3.4.3 CLP Procedure 1
3.4.4 QTM Procedure 1
3.4.5 Sample Spiking Procedure 1
3.4.6 GC Calibration Procedure 8
3.4.7 GC/MS Calibration Procedure 8
4.0 Results and Discussion 9
4.1 Preliminary Evaluation 9
4.2 Spiked Soils 10
4.3 Environmental Samples 20
4.3.1 Air Force Base Samples -• 20
4.3.2 Scrubber-Sludge Pond Sediment Samples 20
4.3.3 Creosote-Contaminated Soil Sample 31
4.3.4 Standard Reference Materials 31
4.4 Interlaboratory Comparison Study 46
5.0 References 50
-------
FIGURES
1 GC/FID chromatogram of an extract of a blank soil extracted for 15 min by SFE. . 12
2 GC/FID chromatogram of an extract of clean soil spiked with n-hydrocarbons. ... 14
3 GC/FID chromatogram of an extract of clean soil spiked with PAHs 16
4 Percent recoveries of PAHs from spiked soil matrices 17
5 GC/FID chromatogram of an extract of clean soil spiked with phenols 19
6 GC/FID chromatogram of an SFE extract of a soil sample from an Air Force Base
compared with a chromatogram of an SFE extract of clean soil spiked with JP-4. . . 21
7 GC/FID chromatograms of extracts of the low-level scrubber-sludge pond
sediment obtained by SFE and sonication 26
8 Comparison of extraction techniques for the low-level scrubber-sludge pond
sediment sample 29
9 Comparison of extraction techniques for he high-level scrubber-sludge pond
sediment sample 30
10 Comparison of extraction techniques for the soil sample containing creosote 36
11 GC/FID chromatograms of extracts of the soil sample containing creosote
obtained by SFE and Soxhlet 37
12 CG/FID chromatogram of an SFE extract of SRM HS-3 40
13 GC/FID chromatogram of an SFE extract of SRM 1941 44
VI
-------
TABLES
1 HP-5890 Instrument Settings Used for the Analysis of SFE Extracts 5
2 Percent Recoveries of Aliphatic Hydrocarbons from Spiked Soil 13
3 Percent Recoveries of PAHs from Spiked Soil Using SFE 15
4 Percent Recoveries of Phenols from Spiked Soil Using SFE 18
5 SFE of a Soil Sample from an Air Force Base 22
6 Concentrations of PAHs Extracted by SFE from a Low-Level Scrubber-Sludge
Pond Sediment Sample 23
7 Concentrations of PAHs Extracted by SFE from a Low-Level Scrubber-Sludge .
Pond Sediment Sample 24
8 Concentrations of PAHs Extracted by SFE from a High-Level Scrubber-Sludge
Pond Sediment Sample 25
9 Comparison of SFE of the Low-Level Scrubber-Sludge Pond Sediment Sample
with Sonication Extraction 27
10 Comparison of SFE of the High-Level Scrubber-Sludge Pond Sediment Sample
with Sonication Extraction 28
11 PAH Concentrations of a Sample Containing Creosote Extracted by SFE -
for Comparison with those Obtained by Sonication Extraction (Table 13) 32
12 PAH Concentrations of a Sample Containing Creosote Extracted by SFE -
for Comparison with those Obtained by Soxhlet Extraction (Table 14) 33
13 Comparison of SFE with Sonication Extraction for a Sample Containing
Creosote 34
14 Comparison of SFE with Soxhlet Extraction for a Sample Containing Creosote ... 35
15 Concentrations of PAHs Extracted by SFE from SRM HS-3 - Day 1 38
Vll
-------
TABLES (CONTINUED)
16 Concentrations of PAHs Extracted by SFE from SRM HS-3 - Day 2 39
17 Comparison of Percent Recoveries for PAHs Extracted from HS-3 - SFE vs
Sonication 41
18 Concentrations of PAHs Extracted by SFE from SRM 1941 43
19 Comparison of Results from SFE Extracts of SRM HS-3 and SRM 1941 Using
GC-FID and GC-MS 45
20 Percent Recoveries of Phenols Extracted by the Portable SFE Instrument
Compared with the Mean Percent Recoveries Obtained by Laboratories
Participating in an Interlaboratory Comparison Study 47
21 Percent Recoveries of PAHs from SRM HS-3 Extracted by the Portable SFE
Instrument Compared with the Mean Percent Recoveries Obtained by
Laboratories Participating in an Interlaboratory Comparison Study 48
22 Percent Recoveries of PAHs from SRM 1941 Extracted by the Portable SFE
Instrument Compared with the Mean Percent Recoveries Obtained by
Laboratories Participating in an Interlaboratory Comparison Study 49
via
-------
SECTION 1
INTRODUCTION
The purpose of this study was to determine the performance characteristics of a transportable
supercritical fluid extraction (SFE) instrument designed for extraction of organic compounds from
solid matrices.
The Environmental Protection Agency (EPA) is interested in analytical methods that can be
performed on or near hazardous waste sites. Field screening methods can help in the selection of
locations for subsequent sampling and provide for more efficient selection of samples to send to a
laboratory for full analyses, and problems associated with chain-of-custody, sample transport, and
holding times are reduced. A field analytical method should be easy to perform in a reasonable
amount of time utilizing materials and equipment that can be transported and set up in an
environment less controlled than an analytical chemistry laboratory. Ideally, the data quality of a field
screening method should be as good as, or better than, the laboratory method being replaced.
Extractions of organic species from solid environmental samples are usually performed with liquid
solvents either in a Soxhlet apparatus or via sonication. Soxhlet extraction methods, although
providing high recoveries of organic analytes from environmental samples, generally take several hours
to perform, require large amounts of solvent, and require fragile equipment not suitable for field
transport Although sonication methods are quicker and more easily adapted for field use, analyte
recoveries are generally not as good as with the Soxhlet extraction, and the methods are more labor-
intensive.
A prototype portable SFE instrument, developed by Battelle Pacific Northwest Laboratories for the
Electric Power Research Institute, was evaluated by Lockheed ESC under contract to the EPA. This
prototype instrument was designed to be used in the field for extracting semi-volatile organic
compounds from soil samples. All our experiments were performed under laboratory conditions to
generate the performance data included in this report. Polynuclear aromatic hydrocarbons (PAHs),
straight-chain hydrocarbons, and phenols were the target compounds investigated. Instrument blanks,
- matrix blanks, matrix spikes, and characterized samples were used in the experiments. The SFE
extracts were analyzed by gas chromatography with flame ionization detection.
-------
SECTION 2
CONCLUSIONS
The results of this study indicate that SFE has potential as a field method when properly designed
instrumentation is used. Even though the instrument used in this study contained only the basic
components to decrease the weight and to achieve the ruggedness required for field transport, it
performed as well as SFE instruments designed for use in the laboratory during an interlaboratory
comparison study. Environmental samples extracted by the field SFE method and by accepted EPA
methods gave comparable recoveries.
Some problems were encountered during sample extractions; however, these were easy to detect and
isolate. The restrictor tube can break, resulting in an audible hiss of escaping carbon dioxide. The
restrictor tube or extraction vessel can clog, resulting in reduced bubbling in the collection chamber.
Insufficient extraction pressures or temperatures can be monitored with digital readouts. Undetected
problems, which can produce false-negative data, appear not to be a major problem with this
instrument.
The results of this study indicate that SFE is appropriate for use in the field as a screening method.
However, further study is warranted. The recoveries of analytes appear to be matrix-specific.
-------
SECTION 3
EXPERIMENTAL
3.1 APPARATUS
3.1.1 Supercritical Fluid Extraction System
The supercritical fluid extraction system used for this study was a prototype instrument designed to
evaluate the feasibility of performing SFE on-site at hazardous waste locations. The instrument,
constructed with commercially available components, was to be light and compact to facilitate field
transport. The prototype instrument weighed about 50 pounds and was 18 inches wide, 18 inches deep
and 12 inches tall. Carbon dioxide, SFC grade or cleaner, from a cylinder equipped with eductor tube
and helium headspace, and a cylinder content from 285 g to 50 Ibs (Scott Specialty Gases,
Plumsteadville, PA) is pressurized by a high-pressure liquid chromatograph pump before flowing
through a heated stainless steel extraction vessel. A restrictor (uncoated fused-silica tubing) is placed
between the extraction vessel and the collection vessel containing an organic solvent. A condenser is
placed at the outlet of the collection vessel.
The carbon dioxide was cooled with dry ice placed around the delivery tube immediately before
entering the pump. This was to ensure that the carbon dioxide was in its liquid state so it could be
pumped. A Rainin (Woburn, MA) reciprocating single-piston pump pressurized the CO2 to a pre-set
pressure (0-420 bars). For these experiments, a pressure of 400 bar was used. The pump delivery rate
was set at 10 mL/min. The pump would pump at this rate until 400 bar was reached, then stop until
the pressure fell below 400 bar. This resulted in pressure transients above and below the 400 bar set*
point. Extractions were performed with a 2.0-mL vessel (0.5-cm ID x 10-cm length) obtained from
Keystone Scientific Inc. (Bellefonte, PA). The extraction vessel was heated with a heating jacket
maintained at 100°C The thermocouple was inside the thermal jacket between the jacket and the
extraction vessel, therefore the temperature inside the extraction vessel may have differed from 100°C
- The restrictor was a 60-cm length of 50-Aon ID x 375-^m OD uncoated fused-silica tubing obtained
from Polymicro Technologies Inc. (Phoenix, AR). The restrictor was passed through a 5-cm ceramic
-------
electrical resistance heater to prevent water vapor from condensing and blocking the restrictor. This
heating element was variable, with the highest setting employed for these experiments. The extract
was collected by inserting the outlet restrictor into 10 mL of an organic solvent contained in a 20-mL
evaporator flask obtained from Ace Glass (Vineland, NJ). Methylene chloride or hexane was used for
the collection solvent, as specified in the results section.
v
3.1.2 Gas Chromatograph
A Hewlett-Packard (Avondale, PA) 5890 equipped with a flame-ionization detector (FID) was used to
analyze all SFE extracts. A 15 m x 0.53-mm ID RTX-5 fused-silica column (0.5-txn film thickness)
obtained from Restek Corporation (Bellefonte, PA) was the analytical column. Table 1 lists the
instrument conditions used for each set of analytes.
3.1.3 GC/MS System
A Finnigan MAT (San Jose, CA) ITS™ 40 (Ion Trap GC/MS system) was used for the analysis of
SFE extracts prepared for the interlaboratory comparison study. These extracts were split and analyzed
on both the GC/MS and the GC/FID. Details of the GC/MS analyses have been reported (4).
A Hewlett-Packard 5988 GC/MS was used for the analyses of extracts prepared by the CLP High
Concentration Organics Method. A 30 m x 0.32-mm ID fused-silica column (1.0-^m film thickness)
was used under the conditions prescribed in the CLP Method (2).
3.2 MATERIALS
Analytical reference standards for PAHs and phenols were obtained from the EPA Pesticides and
Industrial Chemicals Repository (Research Triangle Park, NC). These compounds were obtained as
composite mixtures in methylene chloride (2000 /^g/rnL). Reference standards for aliphatic
hydrocarbons were purchased as neat compounds from Chem Service (West Chester, PA) (purity-
Carbon dioxide was SFC-grade liquid obtained from Scott Specialty Gases (Plumsteadville, PA).
Hexane and methylene chloride were pesticide-grade, obtained from Burdick and Jackson (Muskegon, MI).
-------
TABLE 1. HP-5890 INSTRUMENT SETTINGS USED FOR THE ANALYSIS OF SFE
EXTRACTS
INSTRUMENT
SETTING
Injection
Temperature
Detector
Temperature
Initial Column
Temperature
Initial Hold
Programming Rate
Final Temperature
Carrier Gas
Carrier Flow Rate
n-Hydrocarbons
250°C
300°C
50°C
2 min
12°C/min
260°C
He
10 mL/min
PAHs
250°C
300°C
70°C
2 min
12°Qmin
260°C
He
10 mL/min
Phenols
250°C
300°C
65°C
3 min
10°C/min
200°C
He
8 mL/min
-------
3.3 SAMPLE DESCRIPTION
Some of the environmental samples used for this evaluation were obtained during field investigations
at EPA National Priority List (NPL) sites. As the purpose of this study is method evaluation, not site
characterization, the source of these materials is described only in generic terms.
Unadulterated soil - Soil used for spiking with analytes was obtained from a site in the Pacific
Northwest. It is a glacially deposited soil composed of clay, silt, sand, gravel, and glacial till. The soil
was sieved through a 2-mm mesh screen and homogenized. It had a total organic carbon content of
0.23 percent.
AFB samples -- These samples were collected at an Air Force Base on a site contaminated with
hydrocarbon fuels. The samples were sandy with a high, but unknown, moisture content.
Scrubber-sludge pond sediments — These samples were obtained from a drained and graded scrubber-
sludge pond at an aluminum production plant. The sediment had a high, but unknown, carbon
content from spent electrochemical cells. The reduction cells (refractory bricks, carbon blocks, and
carbon paste) became contaminated with PAHs during the production process. The samples were
sieved through a number-10 (2-mm) brass sieve and homogenized.
Creosote-contaminated soil samples — These samples were collected at a wood preserving treatment
plant, where creosote is used in wood processing.
Standard Reference Materials (SRMs) ~ SRM # HS-3 was purchased from the National Research
Council of Canada (Halifax, Nova Scotia, Canada). This material was prepared from sediments
collected from four harbours in Nova Scotia. It was freeze-dried, sieved through a 125-fjm sieve, and
homogenized. The material was certified for 15 PAHs (see Table 15 for values). The combined
results for benzo(b)fluoranthene and benzo(k)fluoranthene, which co-elute on the GC column used in
this study, are reported as benzo(b+k)fluoranthene. SRM #1941 was obtained from the National
Institute of Standards and Technology (Gaithersburg, MD). This material is a marine sediment
collected from the Chesapeake Bay. The sediment was air-dried, pulverized, sieved (150 /zm),
. homogenized, and sterilized before being certified for 9 PAHs (Table 18). The certified values were
based on the results obtained from the analyses of this material using three different sample
-------
preparation procedures and analytical techniques. Non-certified values were supplied with this SRM
for additional PAHs analyzed by less than three analytical procedures.
3.4 PROCEDURES
3.4.1 SFE Procedure
Samples were extracted for 20 min using the instrument settings described in Section 3.1.1., unless
otherwise stated in the result section.
3.4.2 Soxhlet Procedure
EPA Method 3540 was used (3). Two g of the sample was mixed with 8 g clean sand (to fill
extraction thimble) and 10 g anhydrous sodium sulfate. This mixture was extracted with 300 mL of a
1:1 hexane/acetone mixture for 20 hours. The final extract was concentrated to 1 mL using Kuderna-
Danish evaporation.
3.4.3 CLP Procedure
The high-concentration organics procedure of the CLP was used (2). A one-g sample was extracted
with methylene chloride using sonication. This extract was cleaned up using gel-permeation
chromatography.
3.4.4 QTM Procedure
This was a quick-turnaround procedure utilizing sonication extraction followed by solid-phase
extraction (SPE) clean-up (5). A one-g sample was extracted with methanol using sonication. The
methanol extract was diluted with water (5:100) and passed through C-8 solid-phase extraction
cartridges. The cartridges were eluted with 1 mL hexane that was analyzed directly via GC/FID.
3.4.5 Sample Spiking Procedure
A one-gram sample of unadulterated soil was placed inside the extraction vessel. The standards were
-------
then spiked onto the soil, by syringe, from the end of the vessel where the CO2 enters. The extraction
vessel was sealed and attached to the SFE apparatus, and the spiked sample was extracted.
3.4.6 GC Calibration Procedure
Three standards for each target analyte, encompassing the detection range of the gas chromatograph,
were analyzed in triplicate. A calibration curve was prepared from the results, and calibration factors
were determined for each target analyte.
3.4.7 GC/MS Calibration Procedure
Three standards for each target analyte, encompassing the detection range of the gas chromatograph,
were analyzed in triplicate. A calibration curve was prepared from the results, and calibration factors
were determined for each target analyte.
-------
SECTION 4
RESULTS AND DISCUSSION
4.1 PRELIMINARY EVALUATION
Some problems encountered during the start-up phase of the evaluation had to be solved before
extractions could be performed smoothly.
One major difficulty with the instrument was the performance of the cooling system that cools the
carbon dioxide inlet line and the condenser on top of the collection vessel. The coolers supplied with
the instrument were piezo-electric ceramic devices: one side is cold, the other side is hot. Heat was
removed from the hot side with a circulating water radiator. This cooling system was continually
breaking down, with either the electronic connections breaking or the plumbing lines leaking. This
cooling system was replaced by dry ice placed next to the areas to be cooled. Aluminum foil shaped
like a cup held the dry ice in place next to the carbon dioxide inlet line and next to the condenser.
This modified cooling system functioned well and improved ruggedness of the portable instrument.
Another problem encountered was clogging of the restrictors during extraction of samples containing
sulfur or high levels of moisture. Placing copper turnings at the egress end of the extraction vessel to
chemically bind the sulfur resolved the clogging problem for sulfur-containing samples. The
instrument had a heating element for the restrictors to prevent moisture from condensing; however,
samples with high water content (>20%) would still cause clogging. For these samples, a small heat
gun (hair dryer) was aimed, as needed, at the restrictor to prevent moisture buildup. The restrictor
should be replaced when it is clogged and when carry-over from a previous sample is suspected. To
prevent carry-over from the restrictor between high-level samples, a clean empty extraction vessel
should be extracted for 10 min, and when the extract is not blank, the restrictor should be replaced.
The fittings to the extraction vessel were delicate and difficult to manipulate. On the one hand, when
the fittings were too loose, the restrictor tube would blow off, resulting in partial loss of the sample.
On the other, the fittings were easy to strip by tightening too much. The fitting that stripped and
broke during this study was the nut that attached the 1/16" CO2 line to the extraction vessel. When
the nut broke, the nut, ferrule, and end fitting had to be replaced, but not the entire extraction vessel.
The restrictor by its nature was delicate and easily broken. This problem was overcome mainly
-------
through operator experience. Before the instrument is taken to a field site, an operator should have a
good feel for how tight the fittings should be, as well as carry a large supply of restrictors and fittings.
Even though these problems were initially encountered, after the modifications were made and the
personnel became familiar with the instrument, the extractions were performed with relative ease. The
simplicity of the instrument design had advantages; problems were easy to detect and to solve.
Extractions of samples were less time-consuming and required less manipulation than established soil
extraction methods.
4.2 SPIKED SOILS
Unadulterated soil (see sample description section) was spiked with methanolic dilutions of target
analytes. The unspiked soil was extracted and the extract analyzed before each set of spiked soil was
extracted to ensure no interfering compounds were present in the matrix, instrument, or solvent. No
interfering compounds were detected; an example of a chromatogram of a clean-soil extract is
presented in Figure 1. Caution must be exercised in the interpretation of results from spiked samples
because the removal of a spiked compound from a matrix is usually much easier than removal of
"incorporated" or "native" compounds from an environmental sample. Data from spiked samples are
useful, however, as an indication of the potential of the extraction method; when the spiked
compound cannot be recovered, then it is unlikely that it can be recovered from an environmental
sample.
Aliphatic, straight-chain hydrocarbons were spiked onto 2.5 g of clean soil to determine extraction
precision and accuracy for these compounds. Triplicate extractions by SFE were performed, with
results presented in Table 2. An example chromatogram of the extract is shown in Figure 2. Good
recoveries with high precision were observed for these compounds, indicating that the method may be
useful for extracting soil samples contaminated with hydrocarbon fuels.
Table 3 shows the recoveries of 15 PAH compounds spiked onto 2.5 g of clean soil. The
chromatogram of the extract is shown in Figure 3. The results indicate quantitative recoveries, except
for a few of the higher-molecular-weight compounds. These results compared favorably with the
results of other studies performed with laboratory-grade instrumentation (1). A comparison of PAH
10
-------
extraction efficiencies using the field SFE unit and a Suprex (Pittsburgh, PA) Model SE-50 SFE
system is presented graphically in Figure 4. Although different matrices were used for these two
studies (clean sand for the Suprex instrument and clean soil for the portable instrument), similar
extraction efficiencies were noted for individual analytes.
Good recoveries were obtained for 14 phenolic compounds spiked onto 2 g of clean soil, as shown in
Table 4. Figure 5 is an example chromatogram of an SFE extract of these phenols. Only the recovery
of 4-nitrophenol was low. The extractions were quite reproducible, as indicated by the low percent
relative standard deviations of triplicate extracts.
Overall, the recoveries of the organic compounds selected for this study were good when spiked
matrixes were used. Only one matrix type was used for this study, a relatively dry soil that was very
low in organics. Although different recoveries may be obtained using other matrix types, the basic
adequacy of the instrument extraction capability was demonstrated with these experiments.
11
-------
0
a
(V. 05
FIGURE 1. GC/FID chromatogram of an extract of a blank soil extracted for 15 min by SFE.
-------
TABLE 2. PERCENT RECOVERIES OF ALIPHATIC HYDROCARBONS FROM
SPIKED SOIL"
Compound
Decane
Dodecane
Tridecane
Hexadecane
Heptadecane
Octadecane
Nonadecane
1
98
101
99
99
96
95
93
Extract #
2
97
111
115
113
109
107
104
3
95
102
102
108
109
108
106
Mean
97
105
105
106
104
103
101
Percent
RSD
1.4
4.3
6.3
5.5
5.7
5.5
5.7
* The sample size was 2.5 g of dry matrix soil. The spiking level was 100 ug/g for each
compound. The extraction time was 15 min.
13
-------
n
0)
c
flj
o
o>
a
0)
c
(0
o
0)
TJ
O
a
\fl
0)
c
(TJ
u
-------
TABLE 3. PERCENT RECOVERIES OF PAHs FROM SPIKED SOIL
USING SFE"
Compound
no.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Compound
Naphthalene
Acenaphthylene
Acenaphthene
Fluorene
Phenanthrene
Anthracene
Fluoranthene
Pyrene
Benzo(a)anthracene
Chrysene
Benzo(b + k)fluoranthene
Benzo(a)pyrene
Indeno(l,2,3-cd)pyrene
Dibenzo(a,h)anthracene
Benzo(g,h,i)perylene
Concentration
found
C/'g/g)
50.5
47.0
50.1
51.7
51.1
44.5
49.4
49.9
49.0
28.8
89.7
42.1
38.5
27.8
22.5
Percent
recovery
101
94
100
103
' 102
89
99
100
98
58
90
84
77
56
45
a The spiking level was 50 /^g/g. 2.5 g of dry matrix soil was extracted. The extraction time was
15 min. Single determinations.
15
-------
Legend
n
00 *
f>
"I
f,
10
1
2
'3
4
5
•s ,
•» 6
7
- 8
9
10
11
12
13
Z 14
* 15
12
00
13
• ^ *
• •» ir. •»
' .j> ^ -
r* K >J
I J L r .
I?
15
/ 1^
Naphthalene
Acenaphthylene
Accnaphthene
Fluorene
Phenanthrene
Anthracene
Fluoranthene
Pyrene
Benzo(a)anthracene
Chrysene
Benzo(b-t-k)nuoranthene
Ben2o(a)pyrenc
Indeno{l,2,3-al)pyfene
Dibenzo{a,h)anthraccne
Benzo(grh,i)perylene
FIGURE 3. GC/FID chromatogram of an extract of clean soil spiked with PAHs.
-------
o
6789
Compound Number
12 13
M
15
FIELD SFE
SUPREX SFE
FIGURE 4.
Percent recoveries of PAHs from spiked soil matrices - comparison of results from two instruments.
-------
TABLE 4. PERCENT RECOVERIES OF PHENOLS FROM SPIKED SOIL
USING SFE'
Compound
Phenol
2-Methylphenol
4-Mcthylphenol
2-Chlorophenol
2-Nitrophenol
4-Nitrophenol
2,4-Dimcthylphenol
2,4-Dichlorophenol
4-Chloro-3-methylphenol
2,4-Dinitrophenol
2,4,6-Trichlorophenol
2,4,5-Trichlorophenol
2-Methyl-4,6-dinitrophenol
Pentachlorophenol
1
94
82
81
81
86
25
85
91
86
84
87
90
87
88
Extract #
2
92
80
80
80
85
29
85
90
85
82
87
89
86
88
3
90
80
80
78
84
28
83
89
84
79
85
88
83
84
Mean
percent Percent
recovery RSD
92
81
80
80 _
85
27
84
90
85
81
86
89
86
87
1.9
1.0
0.8
1.4
1.1
6.4
1.0
1.2
0.8
2.6
1.1
1.1
2.1
1.8
* Soil spiled with 200 jtg/g for each compound. The amount of dry matrix soil extracted was 2.0 g.
The extraction time was 20 min.
18
-------
.n
Legend
n
oo
8 a>
o
10
•o
•j-
«•>
•»>
12
<3T\
13
1 Phenol
2 2-Chlorophenol
3 2-Mcthylphenol
4 4-Methylphenol
5 2-Nitrophenol
6 2,4-Dimethylphenol
7 2,4-Dichlorophenol
8 4-Chloro-3-methyIphenol
9 2,4,6-Trichlorophenol
10 2,4,5-Trichlorophenol
11 2.4-Dinitrophenol
. 12 4-Nitrophenol
13 2-Methyl-4,6-dinitrophcnol
14 Pentachlorophenol
a.
o
FIGURE 5. GC/FID chromatogram of an extract of clean soil spiked with phenols.
-------
43 ENVIRONMENTAL SAMPLES
4.3.1 Air Force Base Samples
Fourteen soil samples collected from an Air Force base (AFB) for an unrelated EPA field study were
extracted by SFE. These samples, which had a soggy appearance, were suspected of being
contaminated with diesel or JP-4 jet fuel. Although no confirmatory extractions were done on these
samples, the results demonstrate the potential utility of SFE for the extraction of soils contaminated
with hydrocarbon fuels. Figure 6 compares a chromatogram of an AFB soil sample extracted by SFE
with a chromatogram of an extract of blank soil spiked with JP-4 jet fuel standard. The similarity
suggests JP-4 as the contaminant at this site. The results from triplicate sample extracts prepared by
SFE are presented in Table 5. The precision obtained from this real sample is not as good as the
precision obtained from the soil matrix spiked with hydrocarbons (Table 2). A lack of sample
homogeneity may explain the lesser precision, as only minimal homogenization was performed on this
sample to avoid the loss of the more volatile components that may result from too much sample
manipulation.
4.3.2 Scrubber-Sludge Pond Sediment Samples
These samples were obtained and analyzed during a previous EPA field analytical method
demonstration (5). Replicate extractions of one sample (labeled "low-level") were performed by SFE
on two separate occasions. The analytical results of the first set of extractions are presented in Table
6, and the results of the second set of extractions are presented in Table 7. Another sample (labeled
"high-level") was extracted in triplicate on one occasion; the results are presented in Table 8. These
samples had been extracted by two independent procedures: the CLP sonication/GPC method (2) and
the quick-turnaround sonication/SPE method (5). The CLP sonication extracts were analyzed by
GC/MS, and the QTM extracts were analyzed by GC/FID. In Figure 7, a chromatogram of an SFE
extract is compared with a chromatogram of a QTM sonication extract The mean results from the
SFE extracts are compared with the results of extracts obtained by other methods in Tables 9 and 10.
The two extraction techniques are compared graphically in Figures 8 and 9.
20
-------
FIGURE 6. GC/FID chromatogram of an SFE extract of a soil sample from an Air Force base (top)
compared with a chromatogram of an SFE extract of clean soil spiked with JP-4
(bottom).
21
-------
JIGURE 6. GC/FJD chromatogram of an SFE extract of a soil sample from an Air Force base (top)
compared with a chromatogram of an SFE extract of clean soil spiked with JP-4
(bottom).
21
-------
TABLE 5. SFE OF A SOIL SAMPLE FROM AN AIR FORCE BASE'
Compound11
Nonane
Decane
Undecane
Dodecane
Tridecane
Tetradecane
Pentadecane
Hexadecane
Heptadecane
Octadecane
Nonadecane
Eicosane
1
31
203
287
255
258
202
115
39
22
5
5
3
Extract #
2
Mg/g
34
286
407
373
363
261
150
42
25
6
10
3
3
18
184
289
294
312
241
110
35
21
5
9
5
Mean
t
Mg/g
28
224
328
307
311
235
125
39
23
5
8
4
%RSD
25
20
17
16
~ 14
10
14
7
7
11
27
22
*The sample size was 2.5 g. Triplicate samples extracted for 20 min.
b Peaks were identified by GC retention times.
22
-------
TABLE 6. CONCENTRATIONS OF PAHs EXTRACTED BY SFE FROM A LOW-
LEVEL SCRUBBER-SLUDGE POND SEDIMENT SAMPLE *
#
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Compound
Naphthalene
Acenaphthylene
Acenaphthene
Fluorene
Phenanthrene
Anthracene
Fluoranthene
Pyrene
Benzo(a)anthracene
Chrysene
Benzo(b + k)Quoranthene
Benzo(a)pyrene
Indeno( 1 ,2,3-cd)pyrene
Dibenzo(a,h)anthracene
Benzo(g,h,i)perylene
1
NDC
ND
ND
ND
3.6
ND
7.9
6.3
8.6
30.5
37.7
10.2
4.0
2.5
3.9
Concentration
Extract #
2
ND
ND
ND
ND
2.4
ND
7.7
6.0
8.8
31.4
37.6
10.5
3.6
2.0
3.5
(fg/g)
b
3
ND
ND
ND
0.8
1.7
1.5
7.7
6.0
8.9
36.4
49.3
14.4
7.3
4.0
5.0
Mean
ND
ND
ND
0.3
2.6 -
0.5
7.8
6.1
8.8
32.8
41.5
11.7
5.0
2.8
4.1
%RSD
NAd
NA
NA
NA
36
NA
2
3
2
10
16
20
42
4
19
1 First sample set. Extractions were performed at 400 atm and 100°C for 12 min.
bThe sample size was 2.4 g.
c ND - not detected; estimated detection limit 0.2 »ig/g.
" NA - not applicable.
23
-------
TABLE 7. CONCENTRATIONS OF PAHs EXTRACTED BY SFE FROM A LOW-
LEVEL SCRUBBER-SLUDGE POND SEDIMENT SAMPLE '
#
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Compound
Naphthalene
Acenaphthylene
Acenaphthene
Fluorene
Phenanthrene
Anthracene
Fluoranthene
Pyrene
Benzo(a)anthracene
Chrysene
Benzo(b + k)fluoranthene
Beazo{a)pyrene
Indeno(l,2,3
-------
TABLE 8. CONCENTRATIONS OF PAHs EXTRACTED BY SFE FROM A HIGH-
LEVEL SCRUBBER-SLUDGE POND SEDIMENT SAMPLE "
#
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Compound
Naphthalene
Acenaphthylene
Acenaphthene
Fluorene
Phenanthrene
Anthracene
Fluoranthene
Pyrene
Benzo(a)anthracene
Chrysene
Benzo(b + k)fluoranthene
Benzo(a)pyrene
Indeno(l,2,3-cd)pyrene
Dibenzo(a,h)anthracene
Benzo{g,h,i)perylene
1
NDe
ND
ND
1.6
8.8
4.9
73.1
81.5
653
118.9
148.0
30.0
5.9
3.7
15
Concentration
Extract #
2
ND
ND
ND
1.9
11.0
5.9
70.7
76.8
57.1
95.4
104.3
23.9
4.8
Z8
5.8
(Mg/g)
b
3
ND
ND
ND
ND
7.0
4.1
53.4
61.8
49.7
81.6
10Z5
21.5
4.8
2.5
5.2
MEAN
ND
ND
ND
1.7
8.9
5.0
65.7
73.4
57.4
98.7
118.2
25.1
5.2
3.0
6.2
%RSD
NAd
NA
NA
11
23
18
16
14
14
19
22
18
13
21
19
' Extractions were performed at 400 atm and 100°C for 12 min.
bThe sample size was 12 g.
CND - not detected; estimated detection limit 0.2 ng/g.
* NA - not applicable.
25
-------
Ill
_l
CD
FIGURE 7. GC/FID chromatograms of extracts of the low-level scrubber-sludge pond
sediment obtained by SFE (top) and sonication (bottom).
26
-------
TABLE 9. COMPARISON OF SFE OF THE LOW-LEVEL SCRUBBER-SLUDGE
POND SEDIMENT SAMPLE WITH SONICATION EXTRACTION
Compound Compound
#
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Naphthalene
Acenaphthylene
Acenaphthene
Fluorene
Phenanthrene
Anthracene
Fluoranthene
Pyrene
Benzo(a)anthracene
Chrysene
Benzo(b + k)fluoranthene
Benzo(a)pyrene
Indeno( 1 ^,3-cd) pyrene
Dibenzo(a,h)anthracene
Benzo(g,h,i)perylene
SOIL
CONCENTRATION
(ug/g)
SFE • CLP b
NDd
ND
ND
ND
2.6
ND
7.8
6.1
8.8
32.8
41.5
11.7
5.0
2.3
4.1
ND
ND
ND
ND
0.6
0.3
8.6
3.9
9.4
37.0
39.8
12.5
9.2
4.1
13.0
QTMC
ND
ND
ND
ND
ND
ND
2.2
1.3
1.6
17.0
21.0
6.0
22.0
2.1
9.0
*SFE - Supercritical fluid extraction with GC/FID detection.
bCLP - Sonication extraction with GPC cleanup and GC/MS detection.
CQTM - Sonication extraction with SPE and GC/FID detection.
d ND - not detected; estimated detection limit 0.2
27
-------
TABLE 10. COMPARISON OF SFE OF THE HIGH-LEVEL SCRUBBER-SLUDGE
POND SEDIMENT SAMPLE WITH SONICATION EXTRACTION
Compound Compound
#
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Naphthalene
Acenaphthylene
Acenaphthene
Fluorene
Phenanthrene
Anthracene
Fluoranthene
Pyrene
Benzo(a)anthracene
Chiysene
Benzo(b + k)fluoranthene
Benzo(a)pyrene
Indeno( l,2,3-cd)pyrene
Dibenzo(a,h)arithracene
Benzo(g,h,i)perylene
SOIL CONCENTRATION (/^g/g)
SFE1 CLPb QTMC
NDd
ND
ND
1.7
8.9
5.0
65.8
73.4
57.4
98.7
118.2
25.1
5.2
3.0
6.2
ND
ND
ND
0.6
18.1
5.1
170.0
100.0
5.9
6.0
254.0
ND
7.2
ND
ND
ND
ND
ND
- 1.7
6.4
1.2
90.2
76.0
ND
88.0
53.2
15.1
2.8
2.3
3.6
1SFE - Supercritical fluid extraction with GC/FID detection.
bCLP - Sonication extraction with GPC cleanup with GC/MS detection.
CQTM - Sonication extraction with SPE with GC/FID detection.
dND - not detected; estimated detection limit 0.2
28
-------
SFE COMPARED TO CLP EXTRACTS
I
£
o
15
SFE COMPARED TO OTM EXTRACTS
\
o
g 10 n
Compound Number
12
13
V4
IZZJ
OTM
FIGURE 8. Comparison of extraction techniques for the low-level scrubber-sludge
pond sediment sample.
29
-------
COMPARISON OF SFE AND CLP EXTRACTS
6 a io
COMPOUND NUMBER
SFE
CLP
120
no -
no
90 -
80 -
70 -
60 -
50 -
4O -
3O -
20 -
ro -
o
COMPARISON OF SFE AND QTM EXTRACTS
COMPOUND NUMBER
SFE
OTM
FIGURE 9. Comparison of extraction techniques for the high-level scrubber-sludge
pond sediment sample.
30
-------
The comparative results of the extraction procedures for these two samples are difficult to interpret.
The SFE compares well with the CLP extraction for the low-level sample but not so well for the
high-level sample. If fact, the high-level sample extract from SFE compares better with the QTM
extract than with the CLP extract.
4.3.3 Creosote-Contaminated Soil Sample
A sample contaminated with creosote was extracted by SFE in triplicate. This experiment was
performed twice, each time with a comparison analysis performed by another extraction method, once
by the CLP sonication/GPC method and once by the Soxhlet method. The results of the SFE analyses
are presented in Tables 11 and 12. The comparison with the CLP high-organics method is presented
in Table 13, and the comparison with the Soxhlet method is presented in Table 14. These
comparisons are shown graphically in Figure 10. All extracts for this sample were analyzed by
GC/FID. A comparison of chromatograms for this sample is shown in Figure 11.
For this sample, the SFE compared favorably with the Soxhlet extraction and yet required minimal
time and solvents. The Soxhlet extracts showed lower %RSD values than the SFE extracts; however,
the concentration values were similar for both methods. The CLP sonication values were consistently
lower than the SFE values. These results are consistent with the nature of the extraction methods
employed. The CLP sonication method was designed for rapid analysis and does not stress efficient
recoveries, whereas the Soxhlet extraction ~ requiring a long extraction time with a large amount of
solvent -- provides efficient analyte recoveries.
4.3.4 Standard Reference Materials
Two SRMs were extracted. The SRM HS-3 was extracted on two separate occasions, the first time in
quadruplicate; the results of the GC/FID analysis are presented in Table 15. This material was later
extracted in triplicate, and the results of the GC/FID analysis are presented in Table 16. An example
chromatogram of an SFE extract of this SRM is shown in Figure 12. The results of these SFE
!
extractions are compared with those of the same SRM extracted by the CLP sonication method and
analyzed by GC/MS (Table 17).
31
-------
TABLE 11. PAH CONCENTRATIONS OF A SAMPLE CONTAINING CREOSOTE
EXTRACTED BY SFEa - FOR COMPARISON WITH THOSE OBTAINED
BY SONICATION EXTRACTION (TABLE 13)
#
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Compound
Naphthalene
Acenaphthylene
Acenaphthene
Fluorene
Phenanthrene
Anthracene
Fluoranthene
Pyrene
Benzo(a)anthracene
Chrysene
Benzo(b+k)fluoranthene
Benzo(a)pyrene
Indeno{ 1 23-cd)pytCM
Dibenzo(a,h)anthracene
Benzo{g,h)i)perylene
1
409
54
313
221
428
68
171
103
24
24
22
14
12
ND"
5
Concentration (fig/g)
Extract #
234
411
53
287
197
337
52
113
67
16
16
16
11
10
ND
4
429
58
338
239
464
72
188
114
27
28
25
16
13
'ND
6
449
64
364
259
502
83
196
117
25
26
19
13
6
ND
3
Mean
424 .
57
325
229
433
69
167
100
23
24
20
13
10
ND
5
%RSD
4
8
9
10
14
16
19
20
19
20
16
13
25
NAC
28
' The extractions were performed at 400 atm and 100°C for 20 min. The sample size was 2.5 g.
b ND - not detected.
c NA - not applicable.
32
-------
TABLE 12. PAH CONCENTRATIONS OF A SAMPLE CONTAINING CREOSOTE
EXTRACTED BY SFEA - FOR COMPARISON WITH THOSE OBTAINED
BY SOXHLET EXTRACTION (TABLE 14)
#
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Compound
Naphthalene
Acenaphthylene
Acenaphthene
Fluorene
Phenanthrene
Anthracene
Fluoranthene
Pyrene
Benzo(a)anthracene
Chrysene
Benzo(b+ k)fluoranthene
Benzo(a)pyrene
Indeno( l,2,3-cd)pyrene
Dibenzo(a,h)anthracene
Benzo{g,h,i)perylene
1
105
5
233
170
360
46
139
108
24
23
21
7
2
ND"
3
Concentration (pg/g)
Extract #
2 3
126
5
273
198
408
50
156
120
27
24
22
8
3
ND
3
98
12
215
138
309
42
115
68
16
19
12
5
2
ND
2
V
Mean
110
7
240
169
359
46
136
99
22
22
18
7
2
ND
3
%RSD
11
45
10
14
11
7
12
22
21
9
26
17
12
NAC
11
' The extractions were performed at 400 atm and 100°C for. 20 min. The sample size was 1.0 g.
bND - not detected.
c NA - not applicable.
33
-------
TABLE 13. COMPARISON OF SFE WITH SONICATION EXTRACTION FOR A
SAMPLE CONTAINING CREOSOTE
#
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Compound
Naphthalene
Acenaphthylene
Acenaphthene
Fluorene
Phenanthrene
Anthracene
Fluoranthene
Pyrene
Benzo(a)anthracene
Chrysene
Benzo(b+k)fluoranthene
Benzo{a)pyrene
Indeno( l,2^-cd)pyrene
Dibenzo{a,h)anthracene
Benzo(g,h,i)perylene
Concentration
SFE'
Extract
424
57
325
229
433
69
167
100
23
24
20
13
10
ND
5
Wg)
Sonication b
Extract
116
20
117
85
172
27
76
45"
12
13
11
8
6
ND
2
" Average of four determinations (Table 11).
b Single determination. Sample size was 1.0 g.
e ND - not detected
SURROGATE RECOVERIES FOR SONICATION EXTRACTION
Compound
% Recovery'
Phenol-d5
Nitrobenzene-d5
2-Fluorobiphenyl
2,4,6-Tribromophenol
Terphenyl-dl4
26
46
46
70
59
* Single determination.
34
-------
TABLE 14. COMPARISON OF SFE WITH SOXHLET EXTRACTION FOR A
SAMPLE CONTAINING CREOSOTE
#
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Compound
Naphthalene
Acenaphthylene
Acenaphthene
Fluorene
Phenanthrene
Anthracene
Fluoranthene
Pyrene
Benzo(a)anthracene
Chrysene
Benzo(b+k)fluoranthene
Benzo(a)pyrene
Indeno(l,2,3-cd)pyrene
Dibenzo(a,h)anthracene
Benzo(g,h,i)perylene
Concentration (jig/g)
SFE1 SFE
Extract %RSD
110
7
240
169
359
46
136
99
22
22
18
7
2
ND
3
11
45
10
14
11
7
12
22
21
9
26
17
12
NA
11
Soxhlet"
Extract
65
34
208
• 150
341
39
126
76
18
18
14
9
NDe
ND
1
Soxhlet
%RSD
9
9
- 8
6
6
3
5
5
5
5
13
14
NA"
NA
73
1 Average of triplicate determinations (Table 12).
b Average of triplicate determinations. The sample size was 2.0 g.
"ND-not detected.
d NA - not applicable.
35
-------
45O
COMPARISON OF SFE AND CLP SON1CATION
8
O
6769
COMPOUND NUMBER
SFE ES883 SONOATION
15
4OO
350 -
3OO -
250 -
2OO -
COMPARISON OF SFE AND SOXHLET
ISO -
WO -
6760
COMPOUND NUMBER
K>
M 15
\7~7\ SFE
SOXHLET
FIGURE 10. Comparison of extraction techniques for the soil sample containing
creosote.
36
-------
n
r
•- -£
->i^*
FIGURE 1L GC/FID chromatograms of extracts of the soil sample containing creosote
obtained by SFE (top) and Soxhlet (bottom).
37
-------
TABLE 15. CONCENTRATIONS OF PAHs EXTRACTED BY SFE FROM
SRM HS-3 - DAY 1"
ANAJLYTES
Naphthalene
Acenaphthylene
Acenaphthene
Fluorene
Phenanthrene
Anthracene
Fluoranthene
Pyrene
Benzo(a)anthracene
Chrysene
Benzo(b + k)fluoranthene
Benzo{a)pyrene
Indeno( l,2,3-cd)pyrene
Dibenzo(a,h)anthracene
Benzo(g,h,i)perylene
1
1.4
ND"
1.2
5.2
50.8
1.1
39.4
22.9
2.6
•LI
2.7
12.7
ND
ND
4.0
Concentration (
-------
TABLE 16. CONCENTRATIONS OF PAHs EXTRACTED BY SFE FROM
SRM HS-3 - DAY 2a
Compound
Naphthalene
Acenaphthylene
Acenaphthene
Fluorene
Phenanthrene
Anthracene
Fluoranthene
Pyrene
Benzo(a)anthracene
Chrysene
Benzo(b+k)fluoranthene
Benzo(a)pyrene
Indeno( 1 ,2,3-cd)pyrene
Dibenzo(a,h)anthracene
Benzo(g,h,i)perylene
Concentration (Mg/g)
Extract #
1 2 3
3.6
0.3
1.9
7.4
50.4
0.5
35.3
22.4
1.0
23
7.0
1.8
14.2
1.5
12
4.4
0.3
1.9
7.1
48.3
0.7
31.7
20.9
1.8
2.9
7.7
1.8
14.3
1.5
0.5
4.6
0.4
2.3
8.4
59.6
1.1
47.4
30.8
0.1
4.6
10.8
1.9
18.2
0.2
2.4
Percent CertiOed
Mean RSD Value
4.2
0.3
2.0
7.6
52.8
0.8
38.1
24.7
1.0
3.3
8.5
1.8
15.6
1.1
1.4
10
14
10
8
9
34
18
18
69
31
19
2
12
58
59
9.0
0.3
4.5
13.6"
85.0
13.4
60.0
39.0
14.6
14.1
10.5
7.4
5.4
1.3
5.0
Percent
Recovery
47
113
45
56
62
6
64
63
7
23
81
25
288
83
27
" Sample size was 1.5 g. Extraction time was 20 min. Extracts were analyzed by GC/FID.
39
-------
FIGURE 12. GC/FID chromatogram of an SFE extract of SRM HS-3.
-------
TABLE 17. COMPARISON OF PERCENT RECOVERIES FOR PAHs EXTRACTED
FROM SRM HS-3 - SFE vs SONICATION
#
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Compound
Naphthalene
Acenaphthylene
Acenaphthene
Fluorene
Phenanthrene
Anthracene
Fluoranthene
Pyrene
Benzo(a)anthracene
Chrysene
Benzo(b 4- k)fluoranthene
Benzo(a)pyrene
Indeno( l,2,3-cd)pyrene
Dibenzo(a,h)anthracene
Benzo(g,h,i)perylene
SFE1
DAY 1
32
0
51
48
57
6
57
52
9
15
30
131
11
0
86
SFE1
DAY 2
47
113
45
56
62
6
64
63
7
23
81
25
288
83
27
Sonicationb
53
133
73
65.
54 -
48
95
57
82
82
144
61
44
85
54
'Extracts analyzed by GC/FID.
b Extracts analyzed by GC/MS.
41
-------
The results from the extract of SRM 1941 obtained by SFE are presented in Table 18, with an
example chromatogram shown in Figure 13.
The analytical results from these SFE extracts are quite variable with respect to the certified values
provided for these SRMs. Some of this variability is due to the complex nature of components present
in these SRMs, as can be seen from the FID chromatograms of these samples. Detection by FID is
not selective; co-eluting non-target substances give positive interference. The SFE extracts of these
two SRMs were analyzed by both GC/FID and by GC/MS; the results are presented in Table 19. It
can be seen that the GC/MS results of the same extract are generally in closer agreement to the
certified values. The GC/FID results are presented to demonstrate that real-world samples extracted
by SFE and analyzed without clean-up procedures may contain compounds that interfere with the
target analytes listed in this method, if analyzed by GC/FID.
42
-------
TABLE 18. CONCENTRATIONS OF PAHs EXTRACTED BY SFE FROM SRM 1941*
Compound
Naphthalene
Acenaphthylene
Acenaphthene
Ruorene
Phenanthrene
Anthracene
Fluoranthene
Pyrene
Benzo(a)anthracene
Chrysene
Benzo(b + k)fl uoranthene
Benzo(a)pyrene
Indeno{ l,2,3-cd)pyrene
Dibenzo{a,h)anthracene
Benzo(g,h,i)perylene
Concentration (ng/g)
Extraction #
1 2 3
0.97
0.28
0.31
2.36
0.98
4.31
2.09
2.12
0.13
0.64
1.47
0.67
2.94
2.22
0.48
0.57
0.11
0.23
1.83
0.89
3.78
1.52
1.41
0.10
0.36
0.90
0.62
8.13
3.11
0.40
0.70
0.18
0.25
1.81
0.80
3.90
1.84
1.%
0.06
119
1.51
1.77
1.28
2.19
0.26
Mean
0.75
0.19
0.26
2.00
0.89
4.00
1.82
1.83
0.10
1.06
1.29
1.02
4.11
2.50
0.38
%RSD
22
38
14
13
8
6
13
16
30
75
22
52
71
17
23
Certified
Value
1.32"
0.12"
0.05"
0.10"
0.58-
0.20
1.22
1.08
0.55
0.45"
1.22
0.67
0.57
c
0.52
Percent
Recovery
57
161
503
1923
154
1980
149
169
18
237
106
152
723
NA"
74
* The sample size was 1.0 g. The extraction time was 20 min. Analyzed by GC-FID.
b Value not certified value.
c No certified value
d NA - not applicable.
43
-------
FIGURE 13. GC/FID chromatogram of an SFE extract of SRM 1941.
-------
TABLE 19. COMPARISON OF RESULTS FROM SFE EXTRACTS OF SRM HS-3
AND SRM 1941 USING GC-FID AND GC-MS
Compound
Naphthalene
Acenaphthylene
Acenaphthene
Fluorene
Phenanthrene
Anthracene
Fluoranthene
Pyrene
Benzo(a)anthracene
Chrysene
Benzo(b 4- k)fluoranthene
Benzo(a)pyrene
Indeno( l,2,3-cd)pyrene
Dibenzo(a,h)anthracene
Benzo(g,h,i)perylene
SRM
FID
47
113
45
56
62
6
64
63
7
23
81
25
288
83
27
Percent
HS-3
MS
57
67
70
60
94
40
99
83
68
73
82
41
32
67
18
recoveries
SRM
FID
57
161
503
1923
154
1980
149
169
18
237
106
152
723
b
74
1941
MS
a
a
a
a
72
69
71
55
57
a
281
99
43
b
34
* Analyte not determined by GC-MS.
b No certified value available.
45
-------
4.4 INTERLABORATORY COMPARISON STUDY
The portable SFE apparatus was used in an interlaboratory comparison study of SFE. Ten
laboratories participated in this study, each laboratory extracting four samples: a matrix blank, the
matrix blank spiked with phenols, SRM HS-3, and SRM 1941. The extracts from the individual
laboratories were shipped to EMSL-LV and analyzed on the same instrument to eliminate analytical
variance from the study. The results of this study have been published (4); the portable SFE
instrument (subject of this current study) was used by Laboratory 12. Overall, the results obtained
with the portable SFE instrument compare adequately with those with the other instruments (4).
The percent recoveries of phenols extracted from the spiked matrix by the portable instrument are
compared with the mean recoveries obtained by all laboratories in Table 20. The analytical data
presented in Table 20 for the SFE extracts were obtained by a Finnigan TSQ 45 MS using an external
standard quantitation method.
The percent recoveries of PAHs from SRM HS-3 by the portable instrument are compared with the
mean recoveries obtained by all laboratories in Table 21. The analogous information for SRM 1941 is
presented in Table 22. For all but one phenol, and for all but four PAHs (in SRM HS-3), the
recoveries using the portable SFE exceeded the averages of those obtained by the other, laboratory-
grade instruments.
In general, these data suggest that the portable instrument is capable of performing extractions
comparable in efficiency to those of laboratory-grade instruments.
46
-------
TABLE 20. PERCENT RECOVERIES OF PHENOLS EXTRACTED BY THE
PORTABLE SFE INSTRUMENT COMPARED WITH THE MEAN
PERCENT RECOVERIES OBTAINED BY LABORATORIES
PARTICIPATING IN AN INTERLABORATORY COMPARISON STUDY'
Compound
Phenol
2-Methylphenol
4-Methylphenol
2-Chlorophenol
2-Nitrophenol
4-Nitrophenol
2,4-Dimethylphenol
2,4-Dichlorophenol
4-Chloro-3-methylphenol
2,4-Dinitrophenol
2,4,6-Trichlorophenol
2,4,5-Trichlorophenol
2-Methyl-4,6-dinitrophenol
Pentachlorophenol
Field
SFEb
50
62
81
56
59
39
65
66
63
51
66
64
54
58
Percent Recovery
Mean of all
Instruments
34
44
56
38
50
45
53
48
61
50
63
63
51
53
* Extractions from spiked matrix blanks.
bThe extractions were performed at 400 atm and 100° C for 20 min. The sample size was 2.0 g.
47
-------
TABLE 21. PERCENT RECOVERIES OF PAHs FROM SRM HS-3 EXTRACTED BY
THE PORTABLE SFE INSTRUMENT COMPARED WITH THE MEAN
PERCENT RECOVERIES OBTAINED BY LABORATORIES
PARTICIPATING IN AN INTERLABORATORY COMPARISON STUDY
Compound
Naphthalene
Acenaphthylene
Acenaphthene
Fluorene
Phenanthrene
Anthracene
Fluoranthene
Pyrene
Benzo(a)anthracene
Chrysene
Benzo(b+k)fluoranthene
Benzo(a)pyrene
Indeno( 1 ,2,3-cd)pyrene
Dibenzo(a,h)anthracene
Benzo(g,h,i)perylene
Field
SFE1
57
67
70
60
94
40
99
83
68
73
82
41
32
67
18
Percent Recovery
Mean of all
Instruments
41
78
72
54
81
46
80
63
47
54
69
45
19
56
11
* The extractions were performed at 400 atm and 100° C for 20 min. The sample size was 1.0 g.
48
-------
TABLE 22. PERCENT RECOVERIES OF PAHs FROM SRM 1941 EXTRACTED BY
THE PORTABLE SFE INSTRUMENT COMPARED WITH THE MEAN
PERCENT RECOVERIES OBTAINED BY LABORATORIES
PARTICIPATING IN AN INTERLABORATORY COMPARISON STUDY
Compound
Phenanthrene
Anthracene
Fluoranthene
Pyrene
Benzo(a)anthracene
Benzo(b + k)fluoranthene
Benzo(a)pyrene
Perylene
Indeno( l,2,3-cd)pyrene
Benzo(g,h,i)perylene
Percent
Field
SFEa
72
69
71
55
57
281
99
74
43
34
Recovery
All
Labs
44
37
70
38
33
114
35
40
18
15
1 The extractions were performed at 400 atm and 100° C for 20 min. The sample size was 1.0 g.
49
-------
REFERENCES
(1) U.S. Environmental Protection Agency, "Method for the Supercritical Fluid
Extraction of Soils/Sediments", Project Report, EPA/600/4-90/026, September 1990.
(2) U.S. Environmental Protection Agency, "High Concentration Organics Statement of
Work", Solicitation IFB W801397D1, Exhibit D, September 1988.
(3) U.S. Environmental Protection Agency, 'Test Methods for Evaluating Solid Waste",
SW-846, Method 3540, September 1986.
(4) U.S. Environmental Protection Agency, "An Interlaboratory Comparison Study of
Supercritical Fluid Extraction for Environmental Samples", EPA 600/X-91/041, April
1991.
(5) U.S. Environmental Protection Agency, "Field Evaluation of a Quick-Turnaround
Method for the Analysis of Polynuclear Aromatic Hydrocarbons", Letter Report,
EMSL-LV, August 1990.
50
-------
------- |