United States
Environmental Protection
Agency
Environmental Monitoring and
Support Laboratory
Cincinnati OH 45268
Research and Development
EPA-600/S4-84-006 Mar. 1984
4MEPA Project Summary
Analysis of the NBS
Sediment by the MRI
Sludge Protocol
Viorica Lopez-Avila, Raymond V. Northcutt, Jon Onstot, and Margie
Wickham
An extraction technique, involving
homogenization of a sediment sample
with dichloromethane at dual pH and
phase separation by centrifugation,
was used in the determination of 51
organic priority pollutants as identified
in a standard reference sediment
sample. These compounds were spiked
into the sediment and equilibrated for a
defined period of time. The extraction
was performed first at pH > 11 to
isolate the base/neutral compounds;
acidic compounds were extracted at <
2. The extract containing the base/neu-
tral compounds was fractionated by gel
permeation chromatography (GPC).
Separation and identification of individ-
ual compounds in the silica gel fractions
was accomplished by gas chromatogra-
phy/mass spectrometry (GC/MS).
Method precision and accuracy are
discussed. Tentative identifications of
other organic compounds found in the
sediment are given.
This Project Summary was developed
by EPA's Environmental Monitoring
and Support Laboratory, Cincinnati,
Ohio to announce key findings of the
research project that is fully documented
in a separate report of the same title (see
Project Report ordering information at
back).
This project investigated the application
of the EPA Interim Protocol for the
Analysis of Extractable Organic Priority
Pollutants in Industrial and Municipal
Wastewater Treatment Sludge for isola-
tion of selected test compounds spiked
into a sediment sample. The extraction
procedure involved homogenization of a
sediment/water slurry with dichlorome-
thane at dual pH followed by centrifuga-
tion. A silica gel fractionation of the
base/neutral extract and a GPC fraction-
ation of the acidic extract were performed.
The compounds selected for this study
were 51 priority pollutants representative
of nonpolar and polar, basic and acidic,
stable and reactive compounds. A stand-
ard reference sediment material available
from the National Bureau of Standards
(NBS) was used as matrix.
Experimental
The standard reference sediment
material was purchased from the NBS
and is identified as NBS Standard
Reference Material No. 1645. According
to NBS, the material was dredged
from the bottom of the Indiana Harbor
Canal near Gary, Indiana. The material
was screened to remove foreign objects,
freeze dried, and sieved (particle size <
180 /j). The material was sterilized by
radiation to minimize biological activity.
The individual stock solutions for each
of the 51 compounds (concentration
approximately 4 mg/mL) were made in
methanol, except for hexachlorobenzene,
chrysene and dibenzo(a,h)anthracene
stock solutions, which were made in
tetrahydrofuran, and anthracene and
benzo(a)pyrene, which were made in
acetone. Internal standards naphthalene-
da, anthracene-dto, and chrysene-dia
were purchased from Merck, Sharp and
Dohme, Stable Isotope Division (Quebec,
Canada).
A 20-g aliquot of the NBS Standard
Reference Material was loaded into a
225-mL centrifuge bottle and diluted with
60 ml_ purified water. The sample was
homogenized with Tekmar Tissuemizer
blending probe for 30 sec to enhance the
-------
wetting of the sediment. The spike was
added immediately following homogen-
ization and allowed to equilibrate with the
sediment through continuous mechanical
tumbling for 24 hours at 4°C. Three
sequential extractions with 80-mL aliquots
of dichloromethane were performed first
at pH > 11 and then at pH < 2. Following
each addition of dichloromethane, the
sample was homogenized for 30 sec with
a Tekmar Tissuemizer blending probe,
and centrifuged for 30 min at 2,500 rpm
to promote phase separation. Three
distinct layers consisting of an aqueous
layer on top, a dark solvent layer in the
middle, and a firm solid layer on the
bottom were observed. The base/neutral
extracts were combined and dried by
passage through a short column of
anhydrous sodium sulfate prior to con-
centration to 5 mL in a Kuderna-Danish
evaporator. Following extraction at pH <
11, the sediment/water slurry was
adjusted to pH < 2 with 6 N hydrochloric
acid, and the extraction, extract drying,
and concentration steps were repeated to
isolate the acidic compounds.
Cleanup of the base/neutral extracts
was performed by silica gel chromatogra-
phy using 3% water deactivated silica
(Silica Woelm, 70/150 mesh, ICN). The
elution patterns and the recoveries of the
42 base/neutral compounds from the
silica gel column were determined in
duplicate at two levels (8 fjg, 80 (jg, per
column), corresponding to the spiking
levels in sediment samples. The compounds
were loaded onto the silica gel column by
mixing appropriate aliquots of the stock
solution with 2.5g of 3% water-deactivated
silica gel in a 10-mL beaker. The solvent
was removed under a gentle stream of dry
nitrogen, and the adsorbent containing
the spikes was transferred to the silica gel
column. Four fractions were collected:
Fraction I eluted with 100 mL of hexane;
Fraction II eluted with 50 mL of 10%
dichloromethane in hexane; Fraction III
eluted with 50 mL of 50% dichlorometh-
ane inhexane; and Fraction IV eluted with
150 mL of 5% acetone in dichloromethane.
The sediment extracts were processed
identically.
Extracts containing the acidic compounds
were fractionated by GPC using Biobeads
SX-3 gels (BioRad, Inc.). The GPC column
was calibrated using solutions of corn oil,
n-butylbenzylphthalate, and phenols (e.g.
phenol, 2,4-dichlorophenol, and penta-
chlorophenol) as described in the test
procedure. The eluting solvent was
dichloromethane. GPC recoveries of
phenols were determined at Qfjg and 80
//g per column, corresponding to the
spiking levels of the sediment samples.
A Varian 3700 gas chromatograph
equipped with an effluent splitter and a
fused silica capillary column (30 m x 0.25
mm ID) coated with SE-54 (film thickness
0.25 /;) was used for screening the
various fractions prior to GC/MS analysis.
The oven was held at 50°C for 4 min
during injection and then temperature
programmed at 4°C/min to 300°C (10
min hold) for the base/neutral fractions.
The oven was held at 60 °C for 5 min and
temperature programmed at 8°C/min to
325°C for the acid fractions. A flame
ionization detector (FID) and an electron
capture detector (ECD) were used simul-
taneously (effluent split ratio 1:1).
GC/MS analyses were performed on
a Finnigan 4021 quadrupole mass spec-
trometer interfaced to a Finnigan 9600
gas chromatograph and operated under
the control of Incos 2300 data system. All
separations were performed by splitless
injection using a SE-54 fused silica
capillary column (30 m x 0.25 mm ID)
connected directly to the ion source. The
mass spectrometer was operated in the
electron impact mode. Calibration for the
mass spectrometer was checked daily
using decafluorotriphenylphosphine (DFTPPJ:
resolution; focusing was adjusted to meet
the DFTPP criteria. Emission current was
in the 0.4-0.6 mA range; ionization
energy was eV, and the electron multiplier
was operated at 1.7 kV with the preampli-
fier set at 10"8A/V.
Results and Discussion
The sample preparation technique
described here was originally developed
for POTW sludges. In order to be applied
to the NBS sediment, an aliquot of the dry
sediment is first slurried with water and
then extracted with dichloromethane
using a high speed homogenizer. If the
sediment particles are not within a
relatively narrow particle size range (e.g.
20 to 200 fj), difficulties in isolating the
organic layer from the aqueous and the
solid layers will be experienced.
Although the evaluation experiments
were conducted using a set of compounds
representative of halogenated hydrocar-
bons, polycyclic aromatic hydrocarbons,
phenols, etc., the results indicate that
this extraction technique can be used to
isolate other nonpolar and weakly polar
compounds.
The cleanup of the base/neutral and
the acidic fractions was performed by
silica gel and GPC, respectively. Most
of the halogenated hydrocarbons (e.g.,
dichlorobenzenes, trichlorobenzene,
hexachlorobutadiene, hexachlorocyclo-
pentadiene, hexachlorobenzene) eluted
in Fraction I, whereas the polycyclic
aromatic hydrocarbons (e.g. fluoranthene,
pyrene, benzo(a)pyrene, dibenzo(a,h)an-
thracene) eluted in Fraction III. The more
polar compounds such as bis(2-chloro-
ethyl) ether, N-nitrosodi-jvpropylamine,
bis(2-chloroethoxy)methane, dinitrotolu-
enes, phthalates) were identified in
Fraction IV. Several compounds (e.g.,
naphthalene, chloronaphthalene, x-BHC,
p,p'-DDE) were found to spread over two or
three silica gel fractions; however, for
each compound, at least 75% of the
amount spiked eluted in only one of the
fractions. The data indicated that with a
few exceptions, the recoveries of the
base/neutral compounds from the clean-
up procedures were quantitative and
reproducible.
All phenols were recovered from spiked
blanks at both spike levels. The chromato-
graphic behavior of dinitrophenols may
have precluded their identification at
levels below 8 ng//;L in the GPC
fractions. Due to these problems, these
compounds were not reported for the
spiked sediment.
The results of the precision and
accuracy determinations for the base/
neutral compounds are given in Table 1.
Of all recovery determinations in Table 1,
51% fell within the range 60 to 120%.
An additional 16% were below 60%
recovery, and 8% were above the 120%
recovery. Due to the high concentrations
of several compounds in the unspiked
sediment, their recoveries could not be
determined at the lower spike level. For
example, fluoranthene, pyrene, chrysene,
and benzo(a)pyrene had levels above
10,000 ng/g dry sediment. Furthermore,
because of the complexity of the sediment
sample, coeluting compounds may have
also contributed to the high standard
deviations of the recoveries of many of
the spikes and the very poor recoveries of
several compounds. Polar and reactive
compounds such as hexachlorocyclopen-
tadiene, benzidine, 3,3'-dichlorobenzidine,
heptachlor epoxide (only at 400 ng/g dry
sediment), endrin, and cr-endosulfan
were not recovered.
The results of the precision and
accuracy determinations for the acidic
compounds are given in Table 2. Of all
recovery determinations, 64% were
greater than 70%, 29% were between
40% and 70%, and 7% were zero.
Quantitative recoveries (>70%) were
found at both spike levels for phenol, 2-
chlorophenol, 2,4,-dichlorophenol, 2,4-
dimethylphenol, and pehtachlorophenol.
4-Nitrophenol was not detected at 400
ng/g; however, its recovery was satisfac-
tory at the higher spike level. The
chromatographic behavior of this compound
-------
Table 1. Results of the Precision and Accuracy Evaluations for the Sludge Protocol Applied to the Analysis of Base/Neutral/Pesticide
Compounds in the Spiked NBS Sediment
% Recovery"
Compound
No.
1
2
3
4
5
6
7
8
9
70
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
Compound
Bis(2-chloroethyl) ether
1,3-Dichlorobenzene
1,2-Dichlorobenzene
Hexachloroethane
N-Nitroso • di-n_ -prop y famine
Nitrobenzene
Bis(2-chloroethoxy)methane
1,2,4- Trichlorobenzene
Naphthalene
Hexachlorobutadiene
Hexachlorocyclopentadiene
2-Chloronaphthalene
Acenaphthylene
2, 6-Dinitrotoluene
Acenaphthene
2, 4 -Dinitrotoluen e
Fluorene
4-Chlorophenyl phenyl ether
Diethylphthalate
4-Bromophenyl phenyl ether
Hexachlorobenzene
Phenanthrene
Anthracene
Di-n-butylphthalate
Fluoranthene
Pyrene
Benzidineb
Endosulfan sulfate
n-Butylbenzylphthalate
Chrysene
3,3'-Dichlorobenzidine
Bis(2-ethylhexyl)phthalate
Benzo(a)pyrene
Dibenzo(a,h)anthracene
y-BHC
Aldrin
Heptachlorepoxide
p.p'-DDE
Endrin
p.p'-DDD
jw'-DDT
a-Endosulfan
Unspiked sediment*
concentration
(ng/g dry sediment)
ND
ND
ND
ND
ND
ND
ND
ND
880 ± 850
ND
ND
ND
1,820 ± 350
ND
850 ± 120
ND
1,580 + 130
ND
ND
ND
ND
3.770 ± 300
2,090 ± 72
580 ± 65
14,760 ± 2,810
20.020 ± 4.630
ND
1,160.455
920
27.000 ± 6,500
ND
7,560. 10.790
22.850 ± 6.200
5,960 ± 420
3,050 ± 1.270
ND
ND
340 ± 100
ND
120
840 ± 20
ND
Spike level
400 ng/g
dry sediment
96 ± 23
63 ± 9
66 ± 14
41 ± 15
95 ± 38
43 ± 3
94 ± 39
67 ± 4
64C
63 ± 2
0
72 ± 10
d
110± 51
d
49C
d
86 ± 42
33 ± 14
91 ± 50
46 ± 10
d
d
156C
d
d
0
97. 137
d
d
0
d
d
d
d
44°
0
41C
0
121, 186
d
0
Spike level
4,000 ng/g
dry sediment
112± 19
68 ± 16
75 ± 22
62 ± 5
89 ± 11
102 ± 26
81 ± 11
58 ± 13
70 ± 35
71 ± 17
0
94 ± 20
86 ± 24
102 ± 36
104 ± 24
64 ± 50
82 ± 7
74 ± 7
109 ± 13
89 ± 2
53 ± 4
83 ± 26
25+ 11
100. 137
174 ± 78
148. 252
0
0
103C
d
0
d
128 ± 23
113± 42
50 ± 19
22 ± 13
54 ± 9
76 ± 32
0
104 ± 18
155 ± 43
0
ND = Not detected.
3 Mean ± standard deviation for three determinations.
b Benzidine not detected by GC/MS in any of the standards at concentrations § 80 ng/fjl.
c Single determination reported; levels in the other replicates were below the unspiked sediment level.
" Concentrations determined in the spiked sediment were below the unspiked sediment level.
may have precluded its identification at
low spike levels.
The GC/MS analyses of the silica gel
fractions and the GPC fraction generated
from the NBS sediment extract resulted
in the identification of over 100 base/
neutral compounds and 11 acidic compounds,
without considering the various isomers
of many of these compounds.
Because authentic reference standards
were not analyzed for each compound,
the identifications are tentative (a few
compound identifications were confirmed
with standards). Most of the compounds
give mass spectra that matched the mass
spectra in the NBS library. Identification
of specific positional isomers (e.g.
dimethylnapthalenes, trimethylnapha-
lenes, etc.) could not be obtained from the
mass spectrometric information.
Whenever an identification was made,
either by library search routines or by
manual interpretation, mass chromato-
grams for the most intense ions in the
mass spectrum were obtained. In many
instances, this allowed identification of
other isomers and assignment of coelut-
ing peaks. Although in many cases the
library search did indicate the positional
isomers that best match a particular
unknown mass spectrum, this was not
sufficient for confirming the compound
identity. Consequently, no identification
has been made about the position of the
substituent or the particular alkyl radical.
Fractionation of the base/neutral
extract by silica gel chromatography
reduced sample complexity and, conse-
quently, made possible the identification
of minor components in the sample.
Furthermore, the chromatographic beha-
vior of several classes of compounds (e.g..
-------
Table2.
Results of the Precision and Accuracy Evaluations for the Sludge Protocol Applied to the Analysis of Acidic Compounds in the Spiked NBS
Sediment
Compound
No,
1
2
3
4
5
6
7
8
9
Compound
Phenol
2-Chlorophenol
2,4-Dimethylphenol
2,4-Dichlorophenol
2, 4, 6- Trichlorophenot
2,4-Dinitrophenol
4-Nitrophenol
4. 6-Dinitrocresol
Pentachlorophenol
Unspiked sediment
concentration
(ng/g. dry sediment)
1.810. 2.005
ND
ND
ND
ND
ND
ND
ND
ND
% Recovery
Spike level
400 ng/g dry sediment
112. 128
106 ± 11
99 ± 12
99 ± 16
67 ± 8
0
0
0
95 ± 2
Spike level
4.000 ng/g dry sediment
78 ±21
78+ 18
55 + 3
76 ±5
63 ± 10
0
46 ±11
0
71+3
ND = Not detected.
' - % Recovery determined from triplicate measurements;
Recovery - Spikedsed. cone - unspikedsed. cone
Spike added
sulfur containing heterocyclics, aldehydes
keytones) on silica gel helped in the
assignment of the identifications.
For example, polar compounds such as
benzofluorenone and methylbenzofluo-
renone (m/z 230) was identified by reverse
library search in Fraction IV; however, no
mass spectral match was found for the
mass spectrum with a parent ion at m/z
244. Assigning a mass spectrum with a
parent ion at m/z 244 methylbenzofluo-
renone was based on two considerations:
similar mass spectral fragmentation pat-
tern with benzofluorenone and similar
chromatographic behavior on silica gel of
benzofluorenone and methylbenzofluo-
renone.
The GC/MS chromatograms showed
that most of the compounds of environ-
mental significance elute in Fractions II
and III. Fraction I contains mostly
saturated hydrocarbons (alkanes) that
were not resolved under the conditions
used. Similarly, Fraction IV contains polar
compounds (naturally occurring materials)
that would interfere in the analysis, had
the sample fractionation not been performed.
Summary
Given the complexity of the matrix, the
precision and accuracy results presented
here demonstrate that the extraction
method, involving homogenization of the
wet sediment with dichloromethane
followed by centrifugation, can be applied
with minor modifications to the analysis
of sediments. Although these experiments
were conducted using selected compounds,
the results indicate that other nonpolaror
weakly polar compounds can also be
isolated. In fact, the qualitative information
presented here does indicate that a large
number of polycyclic aromatic hydrocar-
bons and sulfur-containing heterocyclics
were isolated from the sediment matrix.
Polar compounds or reactive compounds
containing nitro or epoxy groups were not
amenable to the analysis procedures
chosen for evaluation.
Fractionation of the sediment extract
containing the base/neutral compounds
by silica gel chromatography and of the
acidic extract by GPC helped reduce the
sample complexity, thus improving the
quantitation of the priority pollutants
without any significant effect on the
analyte recovery for many of the compounds.
Analysis of the NBS sediment extracts
containing the base/neutral compounds
and the acidic compounds was success-
fully performed using a fused silica
capillary column coated with SE-54. The
inertness and excellent resolution of the
capillary column allowed the detection of
nanogram levels of phenols in the
presence of high levels of interfering
coextractants.
Tentative identifications were provided
for the organic compounds extracted for
sediment by the extraction and analysis
procedures described in this paper. Most
of the compounds identified in the silica
gel and GPC fractions were polycyclic
aromatic hydrocarbons, alkyl substituted
polyaromatics, sulfur-containing hetero-
cyclics, alkyl substituted sulfur-contain-
ing heterocyclics, and several polycyclic
ketones.
Viorica Lopez-Avila, Raymond V. Northcutt, Jon Onstot, and Margie Wickham are
with Midwest Research Institute. Kansas City. MO 64110.
Stephen Billets was the EPA Project Officer (see below for present contact).
The complete report, entitled "Analysis of the NBS Sediment by the MRI Sludge
Protocol," (Order No. PB 84-133 750; Cost: $16.00. subject to change) will be
available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA22161
Telephone: 703-487-4650
For information James E. Longbottom can be contacted at:
Environmental Monitoring and Support Laboratory
U.S. Environmental Protection Agency
Cincinnati, OH 45268
------- |