UNITED STATES STF1EL, LORAIN, OHIO, WORKS, BLACK- RIVER SURVEY:
ANALYSIS FOR HEXANE ORGANIC EXTRACTA3LES AND FOLYNUCLEAR AKCMATIC HYDROCARBONS
Herbert J. Brass, Supervisory Chemist
Walter C. Elbert, Chemist
Mary Ann Feige, Chemist
Edward M. Click, Chemist
Arthur W. Lington, Chemist
Organic Chemistry Laboratories
National Field Investigations Center - Cincinnati
Office of Enforcement and General Counsel
U. S. Environmental Protection Agency
5555 Ridge Avenue
Cincinnati, Ohio 1*5268
OCTOBER 197U
-------
TABLE OF CONTENTS
Page Numbei
SUMMARY 1
INTRODUCTION 3
EXPERIMENTAL 5
A. HEXANE EXTPACTA3LES (OIL AND GREASE) 5
B. LIQUID CHRCMATOGRAFHIC FRACTIONATION 6
C. GAS CHROMATCGPAPHY 10
D. GAS CHROMATCGRAPHY/MASS SPECTROMETRY (GC/MS) 12
RESULTS 15
A. WATER SAMPLES 15
B. HEXANE EXTRACTABLES (OIL AND GREASE) 17
C. POLYNUCLEAR AROMATIC HYDROCARBON ANALYSIS IN
SEDIMENT SAMPLES 20
1. Liquid Chromatographic Separations
of PAH ' 20
2. Gas Chromatographic Analysis 25
3. Gas Chromatographic/Mass Spectrometric
(GC/MS) Analysis " 33
DISCUSSION 6k
A. SURVEY FINDINGS 6k-
B. OTHER COMMENTS 66
REFERENCES 6?
-------
SUMMARY
Three water and twelve sediment samples, collected at the U. S.
Steel, Lorain, Ohio, works, were analyzed as part of an extensive
biological and chemical survey of the Black River in the vicinity of
the U. S. Steel site. Two water samples, taken from blast furnace
effluents, were free of contamination. The coke sewer (002) sample
did show the presence of small amounts of organic compounds, possibly
paraffins, cycloparaffins, and/or olefins. The concentration of any
individual component was less than 20 ppm (wt/v). Small amounts
(<1 ppm) of polynuclear aromatic hydrocarbons (PAH) may have been
present in the coke sewer sample. Water sampleswere analyzed by gas
chromatography/mass spectrometry (GC/MS).
The sediment samples were extracted with ji-hexane to determine
organic solvent extractables - "Oil and Grease". Values ranged from
810 to 26,700 mg/kg based on dried sample weight. Highest values were
at sampling sites in close proximity to the plant.
PAH were determined in sediment samples by gas chromatography (GC)
and GG/MS. Liquid chromatography (LC). was used to separate PAH from other
classes of organic compounds in hexarie extracted residues. Quantitation of
ten PAH in eight samples was then made by GC. These PAH represented an en-
velope of polynuclears of increasing ring complexity ranging from phenan-
threne and anthracene (molecular weight 178) through two dibenzanthracenes
-------
(MW 278). Several of the PAH quantitated are known carcinogens. Values of
benz[a]- plus benz[e]pyrene ranged from 0.91 nig/kg (sample 19^66) to
19 mg/kg (sample 19^72). Values for dibenzanthracenes ranged from 0.37
(19U671 to 12 mg/kg (19472). Sample 19^72, collected just downstream
from plant outfall 002, contained the largest concentrations of PAH;
this sample was examined in detail "by GC/MS. A "broad spectrum of PAH
was seen to be present: these included unsubstituted as well as methyl
substituted polynuclears. Specialized GC/MS data system techniques
were used to determine the presence of PAH. Isotope abundances were
used to confirm molecular formulas.
-------
INTRODUCTION
At the request of EFA Region V, NFIC-Cincinnati's Organic Chemistry
Laboratories vere ad-ted to analyze twelve sediment and three water samples, •
collected by NFIC-Cincinnati personnel, at the U. S. Steel, Lorain, Ohio,
works; this plant discharges into the Black River. This request is part
of an extensive biolo-rical and chemical survey of the Black River in the
vicinity of the U. S. Steel site in conjunction with EPA's pending court
case against U. S. Steel. The sediment samples were to be analyzed for
compounds extractable into an organic solvent; this procedure is better
known as an Oil and Grease Determination. In addition the sediments
were to be scanned for Polynuclear Aromatic Hydrocarbons (FAH). Certain
PAH, for example benz[a]pyrene, are known to be powerful carcinogens. The
three water samples were to be analyzed for the presence of organic com-
pounds and in particular PAH. Chain of custody procedures were followed
in handling the samples.
The request for PAH analysis arose out of the survey, conducted by
Mr. Allen Lucas of these laboratories, in which substantial concentrations
of benz[a]pyrene (BaP) were found to be present in sediment samples collected
at a U. S. Steel - Calumet site. Dr. Paul Clifford, who analyzed these
samples, reported concentrations of BaP ranging from 1 to 380 mg/kg (dry
sample weight) using gas chromatography/mass spectrometry (GC/MS) and
fluorescence spectroscopy as analytical tools.
The approach taken in our laboratory to analyze for PAH in the Lorain
sediment survey was different than in the Calumet case. In the latter
-------
instance FAji v.-ere found as part of a general scan for organic compounds;
phenanthrene, anthracene, and EaP were then selected as measures of PAH
and carcinogenic content. In the Lorain case it was decided to analyze
the samples for a broad spectrum of PAH from phenanthrene and anthracene
(KW=178) to the dibenzanthracenes (MW= 2?8). It was felt that it vould
be of greater value scientifically to examine selected sediments for many
individual FAH than all samples for, say, EaP. In fact a detailed PAH
scan was obtained for eight of the twelve Lorain sediment samples.
-------
EXPERIMENTAL
A. HEXAiYE EXTRACTA^LES (OTL Ai-JD GREASE)
The method used for hexane organic extractables is essentially the
2
same as outlined in tr.e Standard Methods nrocedure.
Samples were weighed into individual wide mouth porcelain crucibles. Before
removal each sample was_agitated in its original container to assure'adequate
mixing. A control was prepared. It contained 25 g of silica gel and 25 ml
of water. Samples and the control, ca 50 g T"ret weight, v/ere air dried at
30°C overnight in the individual crucibles. They were then crushed
to fine particles and allowed to dry an additional two hours. Each residue
was placed in a soxhlet thimble and extracted in individual soxhlet apparat-
uses into "distilled-in-glass" n-hexane for 10 hours. Hexane was evaporated
to 15 ml by allowing a stream of nitrogen to flow over the solvent in the
extraction flasks.which were heated by a steam bath held at 80°C. The re-
maining sample and solvent from each flask were transferred to appropriate
size vials. The extraction flasks were washed with small portions of n-hexane
which were transferred to the vials. No apparent residue remained in the ex-
traction -flasks. A stream of nitrogen was allowed to flow over each sample
vial, at room temperature, until the hexane solvent had evaporated. Checks
were made to be certain that samples had reached a constant weight indi-
cating complete hexane loss. Oil and grease is reported as mg, organic
residue per kg, dry sample. It should be noted, that in all procedures,
attempts were made to keep sample exposure to light in the laboratory to a
minimum to prevent decomposition of compounds to be analyzed for in the residue.
-------
3. LIQUID C?I?.CMA7C'J?--.?r:IC F?ACTICNATTO:i
Polynuclear aro-acic hydrocarbons were separated 'from other classes
of organic compounds in the hexane extractable residues by a liquid
chromatographic ' LC' procedure given by Hoffmann and Wynder. In order
to determine recovery of FAH from the LC column an experiment was
performed.
Florisil, oO/ICO :nesh, was heated for an hour at 110°C and cooled
in a dessicator. Forty grams of the activated Florisil was weighed and
placed in a column chroraatography tube whose dimensions were hO X 2 cm
(id). The volume occupied by the Florisil was 20.5 X 2 cm (id). Aluminum
foil was placed around the column to prevent any possible photochemical
reactions. With the stopcock in an open position, hexane was poured into
the column. (it is imperative that the solid packing be uniform and with-
out air bubbles). The initial hexane elutant was discarded. The volume
of eluting solvent passing through the column was determined to be 37 ml.
Five ml of a 1 mg/nl standard of BaP in cyclohexane was applied and ad-
sorbed onto the column.
The first eluting solvent was n-hexane - 200 ml. The flow rate was
375 ml/hour. Twenty-five ml fractions were collected and viewed under
ultraviolet (UV) light (long wavelength - 365 nm) and observed for fluor-
esence. No fluorescence was observed'for all eight fractions.
The second eluting solvent was an 8:1 hexane/benzene (V/V) solution.
The volume of this solution to be applied was determined experimentally.
A critical variable which affects the total volume of the solvent used
-------
is the flow rate (i.e., slower flow rates require less eluting volume).
A flow rate of 3^0 ml/hour was established. Fifty ml' fractions were
collected in 100 ~1 reagent bottles. Each fraction collected was covered
with aluminum foil and viewed under UV (365 nml light. The total volume
after which no further fluorescence was observed was 700 ml. The frac-
tions were pooled and the total volume (TOO ml) was concentrated in a
1 liter round bottom flask by use of a Euchi ROTAVAPOR-R. After the sol-
vent was evaporated to approximately 30 ml, the concentrate was trans- .
ferred to a 250 ml wide-mouth flask. The original container was washed
with 8:1 hexane/benzene and the washings transferred to the 250 ml flask.
Under a steam bath (at c_a 70°C) and a gentle stream of nitrogen, the volume
was further reduced to 2-5 ml. This solution was transferred to a 15 ml
centrifuge tube and the 250 ml flask washed with 8:1 hexane/benzene;
washings were added to the centrifuge tube and this volume was reduced to
ca 2 ml and stored in a refrigerator until ready for analysis by GC.
Samples were chromatographed in a fashion similar to that in the BaP
recovery experiment. The Scheme (Page 8} outlines the procedures used. The
samples (i.e., hexane extractable organic residues) were dissolved in 6 ml
n-hexane and applied to:Florisil columns. n-Hexane was the first solvent;
it was employed to elute non-polar paraffinic type material. Depending on
/
/
the sample, 250 to 1000 ml of n-hexane were used. The n-hexane fraction
is defined as fraction A. For one sample, number 19U?0, where a duplicate
oil and grease determination was made, ten 100 ml fractions were collected
-------
FLOWCHART OF TYPICAL SAMPLE WORKUP
fR~~l
I LJi I
\\i || II \ i-ifi^mlf t
8=1
HEXANE/
BENZENE
{SAMPLE IN
|e Ml ti*-HEXANE
FLORISIL
ICoL. CHROM
N-HEXANE ML
'25O-1OOO
1-3
LITER
B
CONCEN.
ML
B P
N) T RQMETHANF
EXTRACTION
HEXANE
LAYER
EVAP. DRYNESS
PVEOISSOLVE
N—HEXANE
GC
GC/MS
SCHEME
-------
in 125 ml Erlersneyer flasks. No fluorescence was observed indicating no
"breakthrough of PAH. Fractions 2, U, 6, and 10 were concentrated to ca
5 ml on a steam bath under a stream of nitrogen. The concentrates of
these fractions were examined by gas chromatography.
After ri-hexane, an 8:1 n-hexane/benzene (V/V) solution was used to
elute fraction B which contains the PAH. As fractions were collected
they were checked for fluorescence. Certain of the samples less concen-
trated in PAH, particularly in higher molecular weight PAH, displayed no
fluorescence after c_a one liter of solvent had been eluted from the column.
For these samples elution was terminated when fluorescence was no longer
observed. Other samples, for example 19^72, displayed fluorescence after
elution by 3«5 liters of solvent. For this sample eluate from 0-2500 ml
and 2500-3500 ml were separated. •
Fraction B was concentrated to 10 ml by previously described methods.
It was then extracted four times with 10 ml portions of nitromethane (NM). '
The NM extracts (fraction B_T in Scheme) were pooled and evaporated to
dryness and redissolved in n-hexane—1 to 5 ml depending on the sample.
-------
c. GAS CHROMATC:RAPHY
A gas chrorcatcgraphic (GC) method was employe^ to analyze for specific
PAH, based oh the system of Lao et al. A 10 ft X 1/8 in stainless steel
column was employed which was packed with 6% Dexsil 300 on 80/100 mesh Supel-
coport. A similar column packed with 3$ Dexsil 300 was equally efficient
and eluted compounds in a shorter time. A Varian lUOO GC equipped with a
flame ionization detector (FID) and a Hewlett Packard Model 7100B, 1 MV
recorder, were used. Injector temperature was 225°C and detector tempera-
ture 325°C. After injection the column-was held at l65°C for two minutes
then programmed at U°C per minute to 295°C and held until the
dibenzanthracenes had eluted. Total analysis time for the 6$ arid 3% Dex-
sil columns were ca 120 and 80 minutes, respectively. The carrier gas
was heliuin and its flow rate 30 ml/min. The cylinder pressure regulator
V
was set to deliver 80 psig to"minimize pressure drop accross the column
during the temperature program. Extreme care had to be taken to prevent
helium leaks in the carrier gas line. A Gow-Mac Model 21-110 Leak Detector
was useful in detecting leaks in gas lines,
Hamilton 10 nl syringes were employed for injection of standards and
samples onto the GC column. Standards (1000 ppm) were prepared in Burdick
and Jackson distilled-in-glass ri-hexane using PAH obtained from commercial
sources. Perylene was dissolved in benzene due to its low solubility in
n-hexane. For quantitation, a standard mixture of ten PAH (see Results)
was prepared in ri-hexane. The standard mixture consisted of phenanthrene,
10
-------
anthracene, fluorar.thene, pyrene, benz[a]anthracene, benzfajpyrene,
benz[e]pyrene, perylene, dibenz[a,h]anthracene, and dibenz[a,c]anthracene.
The concentration of each PAH in the standard was 50 ppra (wt/v) except for
dibenz[a,h]anthracene v/hich was 25 ppra. Concentrations of PAH in the
samples were calculated by comparison with the standard mixture. Peak
heights were used to compare the earlier eluting PAH (through benz[a]anthra-
cene); peak areas were used for quantitations of BaP and BeP and later eluting
compounds. In calculating concentrations of PAH components, corrections
were applied for volumes of samples and standards analyzed and for total
sample volumes. For comparison, the standard mixture was analyzed prior
to sample analysis. When more than one sample was to be analyzed, the
standard was run after every 2-3 samples.
In analyzing the paraffinic fraction A, a Perkin-Elmer Model 900 gas
chromatograph equipped with a dual FID system and a Honeywell potentiometric
recorder was employed. The f>C was operated in the dual differential mode.
Two 6 ft X 2 mm (id) glass columns packed with 3% OV-1 on 60/80 mesh
Supelcoport were employed. Injector and detector temperatures were 220°
and 320°C, respectively. The column was held at 80°C for five minutes after
injection and then programmed at 10°C/minute to 220°C. The temperature was
then held for ca 10 minutes.
11
-------
D. GAS CHHOMATOCRA~-Y/:.IASS S?ECTRO>-ETRY (GC/MS1
Analysis of standards and samples by OC/MS employed a Finnigan Model
1015 quadrupole GC/MS equipped with a System Industries Model 250 data
system. Inasmuch as the C-C on the Finnigan unit is a modified Varian lUOO
instrument, columns could be interchanged with the GC possessing the FID.
The mass spectrometer (MS) was scanned by applying mass set voltages
to the quadrupole rods from a 15-bit digital-to-analog converter. At a
given mass set voltage, only ions of a specific mass-to-charge ratio pass
7
through the quadrupcle field to the electron multiplier detector. The high
_y
mass range of the MS was used; the preamplifier setting was 10 amp/volt and
2600-2800 volts were applied to the electron multiplier.
Direct aqueous injection of water samples by GC/MS was accomplished
Q
using the method of Budde et al. A 6 ft X 2 mm id, glass column was packed
with Chroraosorb 101 and conditioned at 270°C for two days. Five ul of
sample was injected and the column oven held at 80°C for two minutes and
then programmed at h° or 8°C/minute to 220°C. The final temperature was
held so that the total analysis time was 35-UO minutes. The MS ionizer .and
data system were turned on after water had eluted from the GC column and been
pumped from the ion source. It was allowed to scan from 33-200 amu.
Water samples were also extracted and the extracts examined by GC/MS.
One liter of sample was extracted twice with distilled-in-glass methylene chlor-
ide. The extracts were combined and dried by passage through two inches of
12
-------
prewashed (with methylene chloride) anhydrous sodium sulfate-in a glass col-
umn. After the combined extracts had been dried, 100 ml of distilled-in-
glass acetone was poured through the column. The extract plus acetone was evap-
orated to 5 mi employing a Kuderna-Danish apparatus. In this way the
methylene chloride (BP 39.8°C) was removed, leaving the sample dissolved in
acetone (BP 56.1°C).
A 6 ft X 2 mm (id) glass column packed with 1.5$ OV-17 + 1.95$ QF-1
on Supelcoport, 100/120 mesh, was used to analyze extracted water samples.
It was conditioned at 250°C for two days prior to analysis. Two to four ul
of sample was injected onto the column and .the MS .was "turned on" after the
solvent, acetone, had been pumped from the ion source, ca 1 to 2 minutes,
and allowed to scan from 33-^50 amu.
For sediment samples the 3$ Dexsil column described in the section on
gas chromatography was used. At the high temperatures needed for these
analyses (ca 3CO°C) GS/MS separator and transfer lines had to be adjusted
to 280°C and 230°C, respectively.
The System 250 data system has certain special features which were
p.
applicable to these analyses.7 These include choosing mass set voltages
so that one can acquire data for selected ions of interest. This technique
has the advantage of allowing the'detection of compounds at 10-100 times
/
lower concentrations than by normal''scanning techniques. Another feature
is being able to search the reconstructed gas chromatogram (RGC) for a
limited number of ions. The RGC is a plot of ion current, normalized to 100$
for the spectrum yielding the highest'ion current against spectrum number.
13
-------
The limited mass search allows the operator to ignore background and
search for contributions to the ion current by selected ions of interest
as a function cf spectrum number.
1U
-------
RESULTS
A. WATER SA..-r?LZS
Three water samples were received. The first was taken at the 1, 2,
5 blast furnace (east clarifier) , the second, at the 3, U blast furnace (west
clarifier), and the third at the 002 discharge coke plant sewer. The samples
were scanned for organic compounds, using the laboratory's gas chromatograph/
mass spectrometer (3C/MSX- unit. Two approaches were taken. First, a direct
aqueous injection of sample was made into the GC/MS. This allows the
determination of lev.- molecular weight, volatile, organic compounds of
varying types. The second procedure involved extraction of the sample into
raethylene chloride, addition of acetone, evaporation, and concentration (see
Experimental Section).
The water samples were found to be essentially free of contamination.
The direct aqueous injection technique showed that the two blast furnace
effluents were free of lower molecular weight compounds though one of them
probably contained a small amount of acetone. The coke plant sewer sample
did show small quantities of organic compounds; the concentration of any
individual component was likely less than 20 ppm (wt/v). While individual
components were not identified, two or three classes of compounds were
found possibly to be present; these are paraffins, cycloparaffins, and olefins.
These classes were searched for by allowing the computer to scan the GC/MS
reconstructed gas chromatograph for the presence of particular ions -
m/e U3 and 57 for paraffins and m/e 55 and 69 for cycloparaffins and
olefins.
15
-------
Few organic compounds were present in samples extracted with methylene
chloride, even though the extraction procedures concentrated the samples
some 200 fold. Blast furnace effluents showed no detectable compounds,
while the coke plant sewer again showed small GC/MS peaks. For this
sample the GC/MS was set to acquire data solely for ions corresponding to
base peaks, which are also the molecular ions, of certain PAH. The ions
were m/e 128 (naphthalene), 166 (fluorene), 1?8 (phenanthrene and anthracene),
202 (fluoranthene and pyrene), 228 (benz[a]anthracene and chrysene), and
252 (benz[a]pyrene and benz[e]pyrene).
GC/MS peaks were obtained, for the coke sewer' extracted sample cor-
responding to m/e 166, 1?8, and 202 with 1?8 being the largest. These data
indicate the presence of low concentrations (
-------
B. HEXANE EXTRACT.-^ LES (OIL AND GREASE)
Table I presents data for the n-hexane extraction of organic compounds-
from the twelve sedir.ent samples collected at the Lorain site. Table II
gives a description of the sampling sites. Values ranged from 810 to 26,700
mg/kg based on the -.-.-eight of samples dried overnight at 30°C. A blank
containing 25 g of silica gel and 25 ml of water yielded no significant
oil and grease. Five duplicate determinations were made and yielded excel-
lent reproducability.
17
-------
TABLE I
"OIL AND GREASE" DATA
Lorain Sediment Samples
Weight Organic
Sample
Number
Blank *
1Q1*65
19'i66
19*167
19li67(A)
19U68 '
191i69
i9i(69(A)
191*70
191*70 (A)
19^71
19^72
191*72 (A)
19^73
191171*
19'*7U(A)
19^75
19W6
Wet Sample
V/t . , Grams
70.1*6
50.83
53-02
52.98
51.96
60. 1*3
57.82
52.67
1*9.70
1*9-57
55.53
55-27
1*9.08
61*. 79
62.3U
50.83
73-05
Dry Sample
Wt., Grams
30.79
25.62
28.8U
28.85
33.35
3**.53
33.30
32.69
30.83
20.1*6
36.71*
36.09
29.08
39-26
36.31
31.26
37.25
Percent
Volatiles
56.1*
'19.6
1*5-6
1*5.6
35.9
1*2.9
1*2 ".5
38.0
38.0
58.8
33.9
3^.8
39-6
39-5
1*1.8
38.7
1*9.1
Residues ,
Grams
0.0031
0.0258
0.0595
0.1170
0.0956
0.051*3
0.3683
0.3319
• 0.3080
0.3007
0.2889
0.9539
0.9919
0.5297
0.9758
0.9560
0.2065
0.2061
Oil and Grease, pprn
mg/kg Dry Semple
837
2,320
if ,050
3,310
1,620
10,700
9,960
9,lf20
9,750
1'* , 100
25,900
27,1)80
18,200
2l|,800
26,1*10
6,580
5,530
* Blank contained 25 g. Silica Gel and 25 ml. water
A - Duplicate determinations
18
-------
TABLE II
DESCP.IPT:::: c? BLACK RIVIR SAKPLIIIG STATIONSa
Sample
Iromber
19U65
19U66-
19U67
19^68
19^69
19^70
Sample
Station
1
2
3B
UB
5C
6B
?o
-
-
0.
1.
2.
2.
ile
ir.t
C
0
~>
75
Site Description
Lake water Treatment Intake
Mouth of Harbor: Lake Erie -Black
River
Center of Black River (River Mouth)
Center of River
Right Hand Side of River, Looking
Middle of Turning Basin Opposite
Upstream
Outfalls
003 and
19U71 7B 2.9 Head of River Turning'Basin - End of
Commercial Navigation
19^72 8B 3.U ca 500 ft. Downstream from Outfall 002
19U73 9B k.O Near Plant Water Intake
19V7U 10 5.0 ca 0.2 mi. Downstream from Outfall 001
19^75 11 0.2 0.2 mi Upstream on French Creek
19^76 12 5.3 ca 0.2 mi. Upstream from Outfall 001
a
'All samples taken from the surface of the sediment layer.
19
-------
C. POLYNUCLEAR AROMATIC HYDROCARBON ANALYSIS IN SEDIMENT SAMPLES
The organic residues obtained in the Oil and Grease determinations
were used to analyze for PAH. These residues are complex mixtures of
organic compounds of varying classes and a method of separating the PAH
(fraction) was necessary. The largest body of information on the separa-
tion and analysis of PAH, other than in the air pollution field, lies in
the tobacco smoke chemistry area. PAH determinations in organic residues
obtained from sediments should be.similar to determinations of PAH in
tobacco tars. Separation of PAH was affected by the method of Wynder and
Hoffmann. Several gas chromatographic (GC) ~. .and gas chromatographic/
mass spectrometric (GC/MS) " methods of measuring PAH are available. Tech-
niques for GC and GC/MS analysis advanced by Lao et al. appeared to be
most applicable and were employed.
1. Liquid Chromatographic Separations of PAH
PAH were separated by the liquid chromatographic (LC) procedures
outlined in the Scheme and detailed in the Experimental Section. Paraffinic
type compounds are eluted from an LC Florisil Column by n-hexane (fraction A)
The next solvent is an 8:1 hexane/benzene solution which elutes the PAH
(fraction B). B is concentrated and PAH extracted into nitromethane
(fraction Bj^) to separate the PAH from-other organic compounds that are
eluted in fraction B.
Prior to analysis of the sediment organic residues an experiment was
devised to measure PAH recovery from the LC column. Five ml of a 1 rag/ml
(1000 ppm) benz[a]pyrene (BaP) solution' was applied to a column and eluted
20
-------
as in the Scherae except that the nitromethane (KM) extraction was omitted.
The recovery of EaF was 97$.
It was deterained that 1 liter of n-hexane was likely a sufficient
amount of solvent to elute fraction A. For sample 19^70 1 liter of
n-hexane was eluted from the LC column with ten 100 ml fractions "being
collected. Fractions 2, h, 6 and 10 were concentrated and examined "by
gas chromatography.using OV-1 as the liquid phase; OV-1 is efficient for
non-polar materials. Fractions 2 and U contained a broad envelope of
paraffinic type material which eluted from the OV-1 column as temperature
was programmed. Fraction 6 was free of GC peaks until a high column
temperature was reached, ca 200°C. Reasonably, this indicates high mol-
ecular weight compounds which were late eluters from the LC column.
Fraction 10 contained no GC peaks, thus signifying a complete separation
of total fraction A.
.Table III lists the volumes of n-hexane used to elute fraction A in
each sample. Prior to GC analysis of the n-hexane cuts of sample 19^70,
it was felt that 200-UOO ml of n-hexane would be sufficient to elute a
total fraction A. Clearly this is not the case. Fortunately, however, any
carryover of paraffinic material into Fraction B did not appear to adversely
affect the final analysis; the nitromethane extraction apparently is almost
/
exclusive for PAH, leaving behind other organic compounds in the hexane
layer. For example GC/MS analysis of sample 19U?2 (see below) revealed
only PAH; few other types of compounds were present. An exception was
21
-------
VOLL7-ES C? ^-HEXAiTE USED TO ELUTE FRACTION A
Sample 'lumber Volume n-Kexane, ml
19^66 1000
19U67 ^00
19^69 Uoo
19^70 Uoo
19U71 1000
19U72 200
19U7U 200
19^76 1000
19^70 (Duplicate)3" 1000
aA second "Oil and Grease" determination was made on this sample
see Table I.
22
-------
sample igUyk where an exceedingly high paraffin content (see explanation
below) did interfere somewhat with PAH analysis. Elution of fraction A
with greater than 200 nl of ji-hexane was obviously necessary for this
sample.
No fluorescence was observed in the 1 liter of n-hexane eluted for
sample 19^7U. However, a low level of fluorescence was detected for
samples 19^66 and 19^76. The possibility does exist that small amounts
of low molecular weight PAH were lost.
. One to three liters of 8:1 hexane/benzene was used to elute Fraction B.
Table IV gives .the volumes of solvent used. Elution of samples 19^66 and
19U76 was terminated at 1600 and 1250 ml respectively since fluorescence
was no longer observed in solvent passing through the column. For sample
19^72 2.5 liters of solvent were eluted; this was followed by elution with
an additional liter which was separated from the initial 2.5 liters. Each
was concentrated and extracted with nitromethane (NM); the MM extracts
(fraction B__) were evaporated to dryness, redissolved in n-hexane
and analyzed by GC. All PAH through the dibenzanthracenes (MW 278) were
found to be in the initial 2.5 liters fraction. Since fluorescence was.
observed in the 2.5 to 3.5 liter fraction, including the last 100 ml of
this fraction, higher molecular weight strongly fluorescing PAH are no
doubt responsible for fluorescence at high elution volumes. Sample 19^72
contained the highest amounts of PAH. Thus, it was acceptable to term-
inate elution after 2 liters of solvent on all other samples (see Table IV).
23
-------
TABLE IV
VOLUME OF 8:1 HEXANE/BENZENE USED TO ELUTE FRACTION B
Sample Number Volume 8;1 Hexane/Benzene, ml
19U66 I600a
19^67 2000
19U69 . 2200
. 19U70 2100
19^71 2000
19^72 3500
19^7U 2050
19^76 1250a
ft
No fluorescence observed when these volumes had eluted.
-------
2. Gas Chronatographic Analysis
Excellent separations of PAH were achieved employing Dexsil 2%
and 6% columns. Figure 1 shows a gas chromatographic trace representing
a typical analysis of a ten component PAH standard mixture. Table V
gives the structures and properties of these compounds. Figure 2 shows
a gas chromatographic trace representing the analysis of sample 19^72;
the sample represents the first 2.5 liters of 8:1 hexane/benzene from
the LC column (see page 9).
In Figure 1 it can be seen that phenanthrene and anthracene (peaks 1
and 2) are partially separated, benz[a]pyrene (BaP.) and benz[e]pyrene
(BeP) (Peaks 6 and 7"1 are not separated and dibenz[a,h]- and dibenz-[a,c]-
anthracene (peaks 9 and 10) are not- separated. These observations repro-
duce the data of Lao et al.
Components in the standard mixture were chosen to represent a PAH
envelope that might be observed in sediment samples. Phenanthrene and
anthracene are three ring PAH with molecular weights (MW) of 178. The
dibenzanthracenes are five ring PAH with MW's of 278. In between there
can be PAH with varying number of rings, ring structures, and substitu-
tions on the rings. Over 100 PAH are known between MW 178 and 278. •
Effective separations are both critical and difficult; with caution they
can be accomplished and compound assignments be made.
Figure 2 shows assignment of PAH to components of the standard mix-
ture. In addition to comparing gas chrbmatograms of standards and samples,
a comparison with the data of Lao was useful.in chromatogram interpretation.
25
-------
26
FIGURE I. Recorder trace of detector response in gas chromatographlc analysis of PAH standard
mixture using 6$ Dexsil 300 column; Attenuation and amplifier Bettings are given in
the figure as IB recorder chart speed; Peak numbers correspond to the following PAJ1:
1. phenanthrene; 2. anthracene; 3. fluoranthene; *4. pyrene; 5. benzfa]anthracene;
6 and 7. benz[a]pyrene -and benz[e]pyrene; 8. perylene; 9 and 10. dibenz[a,c]anthracene
and dibenz[a,h]anthracene.
-------
FIGURE II. GC trace for analysis of sample 19!472, 0-2.5 liter, 8:1 hexatie/benzene fraction.
Attenuation, amplifier, and recorder settings are as indicated in the figure.
* T » 1 * .«.«.• «.« 1 * • * » •
-------
28
TABLE V
STRUCTURES AND PROPERTIES OF CERTAIN POLYNUCLEAR AROMATIC HYDROCARBONS
PAH
Molecular
Formula Weight
Structure
Carcinogeni c
Behavior
Phenanthrene
Anthracene
Fluoranthene
Pyrene
3enz[a]anthracene
Eenz[a]pyrene
Benz[e]pyrene
Perylene
Vw 178
ci4Hio
C16H10
C16H10
C18H12
C20H12
C20H12
C20H12
Dibenz[a,c]anthracene CppHlU
Dibenz[a,h]anthracene
178
202
202
228
252
252
252
278
278
oo
O
o
++++
-------
a Ratings of carcinogenic cehavior obtained from reference 20'; (-)
indicates not carcinogenic; ( + ) uncertain or weakly carcinogenic,
( + ) carcinogenic; (-+, +++, +-H-+) strongly carcinogenic.
29
-------
The first -.r.a.jor pe?-k (labeled 1 and 2) in Figure 2 represents phenan-
threne and anthracene. The assignment of these compounds is clear. However,
since only one 1C pea.-: appeared it is probable that either phenanthrene or •
anthracene predc~ir.2.-es. One cannot distinguish unequivocally between the two
and quantitation IT.UST; be nade as the sum of phenanthrene and anthracene;' thi6
situation obtains for all other samples. For sample 19^-72 GC/MS experiments
suggest that the r.a.jcr peak (l and 2) is predominantly phenanthrene with a
minor contribution from anthracene (see Figure 5, page k2).
The peaks labeled 3 and h can be assigned to fluoranthene and pyrene.
Further support for these (and other) assignments -will be given in the
GC/MS data to be presented below. Peak 5 certainly represents at least
two overlapping peaks. This is consistent with the observation of Lao
et al which notes that benz[a]anthracene is not separated from chrysene
and triphenylene using the Dexsil GC system. Peak 5, then, should be
assigned to these three compounds; no other PAH of MW 228 elutes near these com
pounds and GC/MS data show that the peak is due to PAH having molecular
po
weights of 228. Benz[a]anthracene and chrysene are carcinogens while
triphenylene is not.
Peaks labeled 6 and 7 are assigned to EaP and EeP; peak 8 is assigned
to perylene. GC/MS data show these peaks to have molecular weights of
252. The doublet consisting of a large and a small peak elating before
BaP and BeP also have molecular weights of 252 (see below). Based on the
data of Lao the compounds represented by these peaks may be the benzofluor-
anthene isomers.
30
-------
Peaks 9 ?-r.d -"'' ^"e assigned, to the dibenzatythracenes. Other PAH
g
vhich elute i:i "he ~.iben~an'hracene region and vrhic'n may be present in
certain sairrples are ^he cenzocnrysenes, picene, the benzoperylen.es,
o_-phenylenepyrer.e , ar.d the bensotetraphenes .
Table VI reports the concentrations of'FAH determined in sediment
samples (from frac'icr.s ET_ ; . Matching of individual components in -the
samples was not always as sinple as in 19^72 described above and in Fig-
ure 2. For samples vhere certain components could not be compared with the
standard mixture a value is not reported in Table VI.
Samples 1QU66 and 19^76 display the lowest levels of PAH. This is
consistent with their Low oil and grease values and the fact that fluor-
escence terminated after 1-1.5 liters of solvent in collecting fraction B.
It is also consistent with the locations of these two sampling sites.
Sample 19^-72 being just downstream from outfall 002 shows by far the high-
est concentrations of PAH. No explanation can be offered for the absence
of lower molecular weight PAK in sample 19^71-
Sample 19U71* was different than all other sediments collected. Frac-
tion A showed an apparently high paraffinic content. This was observed by
analysis of fraction A employing the Dexsil column. Fractions A of several
samples were scanned for qualitative information as to paraffinic content.
Sample 19U?1*- contained at least a ten fold higher paraffinic level than any
other sample examined. In addition examination of Fraction ETT for PAH
showed a different shaped GC elution pattern. The GC trace in Figure 2
-------
TABLE VI
CONCENTRATIONS OF PAH DETERMINED IN LORAIN SEDIMENT SAMPLES
Concentration, mg/kg (dry sample weight)a
Sample
Number
19*i66
19*i67
19*l69
19)i70
191*71
19)172
19!i7*l
19)176
Phenanthrene
& Anthracene
0.^3
2.8
2.6
3-2
b
18
~3b
0.68
Fluoranthene
b
2.5
3-9
3.3
e
29
b
0.75
Pyrene
b
2.9
5.2
h.k
e
30
b
0.61
Benz[ a] anthracene
0.83d
i.6d
2.9d
2.8d
3-id
2*1
b
1.2d
Benz[a]pyrene &
Benz[elpyrene
0.91
1.5
It. 6
2-9
li.Y
.19
1 . 1
1.5
Perylene
0.10
o.oH
e
e
3-9
2.0
e
0.35
Dibenzanth race nes
0.67C
0.37
2.6
e
3-9
I?.
c
o.66c
Q
Data calculated by comparison of PAH in the sample fraction B T and a standard mixture of ten PAH.
Overlapping peaks in the gas chromatogram precluded analysis of this PAH.
Other major peaks occurred in the dibenzanthracene area of the GC. Concentration reported represents only
GC peak at retention time of dibenzanthracenes in the standard mixture.
The GC peak clearly represents more than one compound, as reported in reference 6. The values given are for
the total PAH present based on the benz[a]anthracene standard.
g
Less than detectable.
-------
for 19^72 is typical cf samples analyzed except 19^7^. Consistent with
our observations concerning sample 19V/U is the fact that U.S. Steel
apparently discharges waste oils from outfall 001; sample I^kjk was taken
0.2 miles downstream from this outfall.
3. Gas Chromatcgraphic/Mass Spectrometric (GC/MS) Analysis
Sample 19^72 was examined in detail by GC/MS. In addition the ten
component FAH solution was analyzed. Both analyses employed the 3$ Dexsil
column. The "real--cine" and reconstructed gas chromatogram (RGC) for
the standard mixture showed, separations to be identical to these shown
in Figure 1. Mass spectra for the standard compounds were consistent
6 21 22
with spectra for known PAH. ' '
Figure 3 shows the "real-time" gas chromatograra for the analysis
of sample 19^72. (All further mass spectral data reported are for this
sampled. The GC is essentially the same as the trace shown in Figure 2.
In GC/MS analysis the total ion current is plotted, as in Figure 3, after
each spectrum scan during the run. Over sixty peaks can be counted.
Figure U shows the reconstructed gas chromatogram (RGC) of the same
run shown in Figure 3- The data system has normalized the largest peak
in the RGC to an amplitude of 100. Because of the way the data system
acquires and processes information relative peak heights between the
"real-time" and reconstructed chromatograms are not always the same.
Further explanation cannot be given in this section, owing to space limi-
tations.
33
-------
FIGURE 3: "Real-tine" gas chroraatograa of GC/MS analysis of sample 19^72,
Times in minutes and seconds written above each peak represent
retention tines of the eluting components.
FIGURE U: Reconstructed gas chroinatogram for GC/MS analysis of sample
19^72.
FIGURE 5: Reconstructed gas chromatogram for .GC/MS analysis of sample
19^72 employing a limited mass search for ions m/e 178, 202,
228, 252, and 278.
FIGURE 6: Reconstructed gas chromatogram for GC/MS analysis of sample
19^72 employing a limited mass search for 'ions m/e 192, 216,
2U2, 266, and 292.
-------
35
FIGURE 3
J. CJ Id W{3 3
I!
-------
36
FIGURE 3 (COOT)
.40-14*103
-------
37
FIGURE 3 (CONT)
-------
38
FIGURE
C-131TZ 3>CE>OL IfSrOS
8
B_
S.
P.
8_
£
R.
8.
!„..,... ,,.,,,,.MI..,,,.,,, p.,.,....,...,!...,,,...,.., ...,,.., ,
o laaoao-ttBeiwns
100 110 ie8 13J l-KJ ISO ICO 170 183 133 ZD3 CIS tiiJ Z3J 213 ZS3 260 Z70 2K> Z23 ->3
-------
39
FIGURE U (CONT)
310
'
« sa a i
-------
Uo
FIGURE U (CONT)
•T-i'-l l""i""l '"I I '"I |'"i'"|
-------
Figure 3 gives ?.r. RGC limited mass search fcr the parent ions cor-
responding to the cc-pcunds in the PAH standard mixture. The ions
causing the RGC-peaks are indicated above each individual peak. A peak
corresponding to n/e 278 is not observed since it is small compared to
the 202 peak on which the RGC is normalized. The mass spectra of PAH
ajjnost always display intense parent ions. The parent is usually the
base peak. Thus performing a limited mass search for the parent is an
effective way of defining particular PAH. A cluster is often present
about the parent representing the P-l, 'P-2, P-3, P+l, and P+2 masses; - that
is the parent less one, two, and three and plus one and two arr.u.
Figure 6 gives an RGC limited mass search for the methyl-substituted
derivatives of compounds in the PAH standard mixture." Thus the unsubsti-
tuted parents -t-lU amu were searched. Again, the molecular ions causing the
RGC peaks have been indicated.
A limited mass search for dimethyl derivatives, the unsubstituted
parents +2$ amUj was r.ade but the data were not conclusive and are not being
presented.
Searching for unsubstituted ions in the standard and their-methyl
derivatives produced thirty-three separate GC/MS peaks indicating at least
that many compounds. In certain cases two or more PAH were present for
a specific HGC peak indicating more than one PAK was eluting from the
Dexsil column into the ion source at the same time; this observation
is quite reasonable. Searching for the dimethyl derivatives produced an
additional eight distinct GC/MS peaks.
Ul
-------
1*2
FIGURE 5
t-12-rrz 3>CEX1U KB-SG
8.
S_
P
8
«L
]
RH
s.
<.nAftci4 -
i I ' I '"'I I |....|....|..i.|i.i| ! , r.|.'"| [ j | .1 ,.... .
10 zo x> 10 EO eo TO eo ao ica no ieo 123 I-KJ isa iea no ico ix za> z\o vza
ZED zco 270 CDS zsi 200 310
.AH-iu HMO a
-------
FIGURE 5 (CONT)
lea
• llllillll I I i I I I'" I' I' 'I I I "'I I 'I' I ' I
jC3 3Q3 1C3 118 tZD 130 Vt9 1S3 1CO T78 K
S SIO SZ8 ba £« SSO ECO £78 SSO EO 633 «13 6Z3
-------
uu
FIGURE 5 (CONT)
i i | • i
i|i if• ji
era «23 ccs era 'twa csa 7ca TIB •KD '"no 713 7S9 7C9 770 780
I
I""!—-T I « | • | • | * I ' »
73O sea cio KB eaa ow csa KO sno sw eao
310 2O5
-------
8.
9s.
FIGURE 6
B
8.
LIMITED MASS
H/C iqa.au,
'-r-r.
0
so' to
M/QQI
w' aa' in no iza li "Aw ioi ici ni iw isi
210
210 iLJ
I.L.
-------
U6
FIGURE 6(CONT)
It.
««
'«" TO""'-!^'
•*fflM«M»M •_„..»».• tWM-M*nM
| MMMIflM
OT bo s
o
-------
Mass spectral iata shew that the sample extract contained almost exclu-
sively ?AK. The largest peal-: in the RGC (Figure k) was not a PAH but a
phthalate derivative with a strong base peak of m/e lU9. The presence of
this compound cannon ~e explained.
Presentation of all niass spectra is precluded by space limitations.
In Figure 5, however, it can be seen that peaks for phenanthrene +
anthracene (m/e 173), flucranthene (first m/e 202), pyrene (second large
m/e 202), benz[a]anthracene and probably chrysene and tetraphene (large
broad m/e 228), benz[a]- and benz[e]pyrene (large peak of second m/e 252
doublet1) and perylene (last m/e 252 peak) are present. Comparison should
be made with Figure 2.
Presentation of selected spectra is useful. Figures 7 through 11
show the mass spectra of major RGC peaks (Figure 5) having m/e 252 as
the base peak - the benzfluoranthene, benzpyrene, perylene area.
Figures 12 through 15 show the mass spectra of major RGC peaks (Figure 6)
having m/e 2U2 as the base peak. These likely represent methyl-PAH which
are not identified. Other unsubstituted PAH having molecular weights of
2^2 are possible but unlikely. It should be noted the computer was
allowed to subtract background from spectra of interest. Thus in Figure 7
spectrum 526-520 signifies that spectrum 520 has been subtracted from 526.
Another method in identifying and confirming molecular formulas by
mass spectrescopy employs isotope abundances of the common elements. For
PAH one need consider only carbon and hydrogen. For every carbon atom,
-------
FIGURES 7, 8, 9, 10, 11: Mass spectra of PAH having n/e 252 as the base
peak in sample 19^72.
FIGURES 12, 13, 1^, 15: Mass spectra of FAH having m/e 2U2 as the base
peak in sample 19^72.
-------
sc« - sj
FIGURE 7
B
8.
0.
SJ-l
- S2fl
E'ISITZ 31I. 16S-23S
g
r
Ea
- '*>-
,-R
1
U J-
, ,„„ ,.,.,. , ., ,11,..
ZSO ZSO ZTO Z83 2
i V'"| •!' 'l"'i- | I ' I i 'I | |"l"'| |'i«|'i-| i' | '. riM|.i.r..|".| !
10 SO CO 70 88 20 1(53 HO IZ0 130 1TO ISO 163 170 1C3 ISO ZOO 21(1
flX E
2ZO Z30
30O 318 3SO 3jO 310
-------
50
FIGURE 8
StTTRM HfBEH StO - SZO
EM317Z 3-COOV. 1CS-2SS
s.
I....I...1
.ft
•39 0 2Z8 Z3O 210 ZSfl Z*O ZTO OJO Z30 3OO 310 3Zfl
JV C
lira
•• » .
-------
n ti |P
Tltl.f.
.1.UIXII. I(.J-2«S»«/M MOLD
51
FIGURE" 9
8
R.
3_
a - sst
C'WTZ 3XXX1L 1C5-Z3S
50 'to SO 63 13 63 S3 100 110 >Z3 133
nx E
ISO ICO 17O IBS 133 ZOO 210
Z33 S1i3 Z£3 2W Z70 ZC3 ZX) 39IJ 319 2ZJ 333
-------
52
FIGURE 10
TITI H
*..»*» »•»."< I
t,.T»l to" Cl'i-*l>ti -l..l:>tl
SPECTFUI Hfeen see - ssi
S
y
I ' I ""I1 I I "• I '"I .,-..-, ,---
33 to so co 70 ao ao ico ito izo 133
OXE
r«4-
•"""•, > '! -r
^2
iso 10 no im jao 200 «o zo oa z-w zca zso ZTO zsa oa 300 aia sza aai
'•
•"»'« f IMUMIMIKKnCH
-------
53
FIGUBE 11
»Tl I HI .in
IOLII lOC-llA/V PMOOtf
8
. R.
8.
gp.
fcS.
£2
SFEOHJ1 t*rtE« SS2 - S7S
E'13172 aiODOl. 1£C-«3C
30 na S3 CO
MX E
I
8(5
jilll! "I1 'I" I'" 1' I 'I' ' |""1 " | "'I'1"! " I'" |""1' ' \< "('" |"
108 HO >zo 130 IM tso ico 17O ioo i» ze» zio zzo zao zio iso iaw era zsa zso sea aia azo
...II..
-------
FIG'JRE 12
TITLE.
r' .- L . I 'I Nl-l^lMi
Trtol. ION CUHHtHTt M.'Ju.l
i«.s-r«:>»i/M IIIILII ioc-7i»/« PKOOU
SPEOHJ1
s.
,1. I.. r ill
,'l, , I
33 « £0 SO 78 00
nx e
"| I' "I | i"| 'ii'v
103 110 1ZO 133 1W ISO ICO 170 ICO ISO 209 Z10 C20 CO 2'J S3 EC8 ZTfl UO C23 300 318 XO XC 1fl
J
-------
FIGURE 13
r/t"«T .irnrrn. i«,s-t>osr./n HOLD IO<-T>A/«
8
8_
_
F-
£o
t~ *" —
Is-
o
—
' 1 1 1"'"' I1 ' 'l
S» tO S3 60 7C
nx E
•es
ill 1 1
i w M ieo
, hi
no
t. «< ^* WAI •
t
IZO
n*^»
T-
\3a
I
"•r"'"t"
110 1
i M
"i"T"'i'"i r 'i'"1! "<(
U ICO 170 183 130
l,.l J|
ZOO 210 ZZO 239 Z
l! Ml1^ ' M " /I. ,' | i
13 ZS3 ZSO Z70 Z00 ZSO 300 310 3ZB 333 .
:
_2
0
Mfl
-------
56
FIGURE lU
TtTI.Ki t«l«472 3;:UEXII. I4S-«9SI>«/N HOLD IOI-7>/>/V 2IK10V
fMKCT-l'M NUMHtMt «60 - «S3
HOSE PKXl 20577
T(iT»l. inn CIIRKFXTt IROI63
8
S
SPETTTUI roBOT tea - «a
E-15TTZ >CEX1L 1GS-23S
JJH JrJ.itl t
SO CO TO SO
in no >a> «, ',« ',» in no ira i» zw ttB'ea za T« CT
,' i..
cm cw
a.a
aad'
-------
57
FIGURE 15
TITIT1
T.'..VL 10" C''K
r»i*4t> sr.nKxiL I»S-«»S»«/N MOLD in<-i>*/v tnoiw
8
3.
|SJ
hSL
|SJ
STCTPUI oretn icz -
C-1M7Z XXX1L 1CS-235
b
n i"
W
•: . -.
ls8
*" *** M "*
fc8 3CO «10 3M 3»
..l.
-------
13
ca 1.08$ of the nolecules contair.ir.fi carbon possess a C atom and produce a
P + 1 (parent -1) pesJc. Za.ch atom in an ion contributes an amount to the
intensity of the isotope peak that is equal to the relative abundance
2k
of the isotope of that atom. "Thus if a strong molecular ion is present,
as in the case of PAH, the P+l abundance is useful in identifying a par-
?\±
ticuiar molecular formula. The P+2 is also considered, for there exists.
a probability that two JC atoms exist in a single ion. Values of P+l
and P+2 have been calculated for the various -.olecular formulas and are
23 25
presented in tables.
Table VII list ?+l and P+2 isotope abundances for selected PAH of
interest. The values are theoretical except for two molecular formulas
which represent experimentally determined literature data. Table VIII gives the
P+l and P+2 values for the PAH in the ten component .standard mixture.
Excellent agreement is observed.
Table IX gives the values found for certain major PAH present in
sample 19^72. The values were obtained from spectra defined by the
reconstructed gas chromatograms shown in Figure 5 for unsubstituted and
Figure 6 for methyl substituted PAH. Agreement for the unsubstituted PAH
is good. In a sample containing often overlapping GC peaks one would not
expect agreement to be ideal. This is especially true for the P+2 isotope
abundances which is a second-order effect. In addition, spectra were defined
by limited mass reconstructed gas chromatographic searches. Overlapping
compounds present which are not seen in the limited mass search can easily
-------
interfere with isotope abundar.ee calculations. The total amounts of
methyl substituted PAH present are considerably smaller than unsubsti-
tuted polynuclears. Thus, one would not expect isotope abundance calcu-
lations to be nearly as satisfactory for these cases; this is exactly
the situation.
59
-------
TABLE VII
ISOTOPE A2UIvDANCES FOR MOLECULAR FORMULAS
CORKESFO;TDING TO PAH OF
•Formula
C1UH10
C15H12
cl6Hio
C17H12
C18H12
C19H1U
C20H12
C22"lU
Molecular
Weisht
178
192
202
216
228
2U2
252
278
P+l
15.29
16. Uo
17.^5
18.56
19.6*1
20.76
21.9
23.5
P+2
1.09
1.26
1.U3
1.62
1.82
2.0^
2.U
3.2
9.
Values are theoretical (reference 23) except for MW 252 and 278;
these values are experimental findings (reference
60
-------
TABLE VIII
ISOTCFE Ar.lJT-rDA.NCE FOR PAH STANDARD MIXTURE
Compound Assigned
Phenanthrene )
Anthracene )
Fluoranthene
Pyrene
Eenz [a] anthracene
Benz[a]pyrene )
Benz[e]pyrene )
Dibenz [a, c] anthracene )
Dibenz[a,h]anthracene )
Formula
C1UH10
ClUH10
C16H10
C16H10
C18H12
C20H12
C22Hli|
Molecular
Weight
178
178
202
202
228
252
278
P+l
15. 2a
15. 5a
17.6
17.7
19-7
21.1b
23. 5b
P+2
l.3a
l.Ua
1.6
!•?
2;i
2.2b
3.1b
Q
Phenanthrene and anthracene were slightly separated as in Figure U.
Values for each side of the peak were calculated.
Values given are averages for several sections of each peak; each
peak represents two compounds J see Figure 2.
61
-------
TABLE IX
ISOTOPE A?.UT!DA::CIS rOR CERTAIN PAH FOUND IN SAMPLE
PAH
Spectrum Parent Peak P+l P+2
A. Unsubstituteci
211-206
215-206
337-333
3**3-333
353-3^9
1*33-1*30
1*36-1*30
526-520
51*0-520
563-557
566-557
582-576
78U-775
821-805
B. Methyl Substituted
268-262
278-262
372-369
375-369
378-369
386-383
393-390
1*6-1*1*2
1*57-1*53
1*60-1*53
1*62-1*53
178
178
202
202
202
228
228
252
252
252
252
252
278
278
192
b
216
216
216
216.
216 ,'
2U2
2l*2
2l*2
2U2
15. U
15.6
18.3
16.8
17.9
20.1
20.0
20.3
21.9
19.6
20. U
21.5
26.7
19-3
18.2
b
16.0
16.6
16.0
b
b
19-6
20.7
22.1*
23.2
1.1
1.2.
1.1*
0.9
2.0
l.U
1.2
2:r
1.9
1.9
1.9
2.0
e
e
1.2
b
c
U.5
c
b
b
c
c
c
<~c
62
-------
TABLE IX (CONT)
a. Refer to Fig-i^re ? fcr unsubstituted FAH and Figure 6 for methyl
substituted "A:-!.
b. Overlapping of tv;o or :nore ccmpour.ds prevented analysis by this method,
c. P+2 value sns.ll cr net nresent.
63
-------
DISCUSSIC::
A. SURVEY FI:.^::;^
Water sa.-r.ples vere :'cu::d not to be contaminated to any significant
degree. The cc>e rev.-er ''CC2^ did show small quantities of organic com-
pounds and likely ccr.«ai::ed lover molecular weight polynuclear aromatic
hydrocarbons 'PAH;.
The data of Table I shows the presence of substantial amounts of
hexane extractable organic residues in sediment samples. All Lorain
sediments vrere surface samples. ji-Hexar.e was vised as the extracting
solvent for two reasons. First it was felt that use of a halogenated
solvent such as Freon would later interfere with gas chroinatographic/
mass spectrcmetric analysis of the residues for PAH. In addition,
n-hexane was the extracting solvent used in the U. S. Steel Calumet survey
and it was felt that comparisons between the two surveys would need to be
drawn.
26
Calumet data show "oil and grease" values as high as'120,000 ppm
for one sample and over 60,000 ppm for many others; some of these were
core samples. Lorain data show, overall, slightly lower oil and grease
values. This could be consistent with the fact that the Black River has
recently been dredged.
Reasonably, the highest "oil and grease" values in the Lorain survey
were found in the vicinity of the plant discharge points. The lowest
values determined were at the harbor mouth and 2-3 miles downstrean from
the outfalls. The lev; value for sample 19U68 cannot be explained.
-------
It is terr.p^ir1.:; zo try to drav: analogies between amounts of organic
residue and particular classes of organic compounds.present in sediment
samples. In certain cases this can no doubt be done; however, caution
must be exercised. ~or example while samples 19^72 and 19^?U yielded
similar amounts of organic residue, their chemical compositions are
quite different.
Table VI and ether data presented in the Results section shows
that a broad spectrum, of FAK is present in the 'sediment samples. The
highest concentrations are in the vicinity of the plant outfalls with
the levels in 1QV72 being 2-10 fold greater, depending on specific PAH,
than for any other sample analyzed. Lowest values are found for sample
19U66 (harbor mouth), 1QU67 (center of the river mouth), and 19^76
(0.2 miles upstream from outfall 001). It is reasonable that PAH were
found in all samples since the sampling sites were in close proximity
to the plant. A lake effect is known to occur carrying water upstream
27
at least five miles from the river mouth. Because this survey rep-
resented the first attempt in this laboratory to quantitatively analyze
for many PAH, one probably should place wide tolerance limits on the
accuracy of the data in Table VI. The data certainly reflect the levels
of PAH in the analyzed, samples; however, errors of + 15$ would not be
unexpected.
65
-------
Many of tr.e ~AH round in the sediment samples are without question
carcinogenic. T'r.ese include not only compounds reported in Table VI
(see also Table "•/'• . Since ?. broad envelope of PAH was shown' qualita-
tively to be present, rr.ar.y unidentified compounds containing four or
more rings, are r.o doubt carcinogenic. In particular are the methyl sub-
stituted. FAK which have been shown to possess a high degree of carcino-
genicity.5'20
3. OTHER COMMENTS
The general philosophy taken at the outset of the survey was to
analyze for a broad envelope of PAH rather than for one or two compounds.
This approach proved to be successful and we believe that future surveys
should be handled in a similar fashion.
The analysis for complex mixtures of PAH constitutes one of the most •
difficult determinations in organic chemistry.; the data presented in this
report are clear evidence of this fact. Optimum conditions for separating
PAH from other classes of organic compounds had to be determined. GC and
GC/MS analyses techniques had to be perfected. The PAH fraction (B,.,) is
complex and requires careful and detailed analysis and interpretation.
These criteria were met and meaningful data were obtained.
In subsequent'surveys, in other geographical locations, our capa-
bilities will be extended. For example, we intend on obtaining a stock of
/
PAH primary standards available from'the Canadian Air Pollution Control'
r,0
Directorate, Department of the Environment, Ottawa. These will allow
us to quantitate upwards of 50 PAH.
66
-------
REFERENCES
1. P. R. Clifford, ''Organic Analysis of the Grand Calumet Oil and
Grease Samples.'1 I-fr'IC-Cincinnati Report, January, 1973.
2. "Standard Methods for the Examination of Water and Wastevater,"
Thirteenth Edition, I-'.J. Taras , A. E. Greenberg, R. D. Hook, and
M. C. Rand, Eds., American Public Health Association, Washington,
D. C., 1971, pp. U09-U13.
3. D. Hoffmann, ar.d E. L. wynder, Cancer, 27, 8U8 (1971).
U. D. Hoffmann and E. L. Wynder, Anal. Chem. , 32, 295- (1960^ .
5- £. Wynder and D. Hoffmann, "Tobacco and Tobacco Smoke," Academic
Press, New York, 1°67 pp. 330-3UU.
6. R. C. Lao, P.. S. Thomas, H. Oja, and L. Dubois, Anal. Chem., 1+5,
903 (1973).
7. J. W. Eichelberger, L. E. Harris, and W. L. Budde, ibid , U6, 227
(197*0.
8. W. L. Budde et al ibid. , in press
9. "System 150 GC/MS Data Processor: System Specifications," System
Industries, Inc., Sunnyvale, California,
10. J. B. Knight, "Computerized GC/MS For Quick Qualitative Identifi-
cation of Hydrocarbon 'Types Included in Chromatographic Peaks,"
Finnigan Corporation Application Tip No. U3, Finnigan Corporation,
Sunnyvale, California, 1972.
11. K. Bhatia, Anal. Chem., U3, 609 (1971).
12. W. Lijinsky, I.I. Dorasky, and J. Ward, J. Gas Chrom., 3, 152 (1965).
13. V. Cantuti, 0.- P. Cartoni, A. A. Leherti, and A. G. Tori, J.
Chromatography, 17, 60 (1965' .
lU. N. Carugno and S. Rossi, J. Gas/Chroin. , 5_, 103 (1967).
15. H. J. Dawson, Anal. Chem., 9, 1852 (196U) .
16. H. J. Davis, Talanta, 16, 621 (1969).
67
-------
17. K.'j. Davis, Anal. Chem., Uo, 15S3 (1968).
10. D. A. Lane,. H. K. Mre, and M. Katz, it id., U5, 1776 (1973)
19. R. F. Skinner, J. B. Knight, D. M. Taylor, and E. J. Eonelli, "The
Determination of Folycylic Aromatic Hydrocarbons in Air Pollution
Samples," l-'ir.nigan Corporation Application Tip No. 52, Finnigan
Corporation, Sunr.yvale, California., 1973-
20. "Particulare Folycyclic Organic Matter," National Academy of Sci-
ences, Washington, D. C., (1972), Chapter 2.
21. L. Dubois, R. C. Lao, and R. S. Thomas, "Mass, Ultraviolet, and
Fluorescence Spectrometric Data of Polycyclic Hydrocarbons Found
in Urban Air and Cigarette Smoke," l6th Spectroscopy Symposium,
Montreal, Canada, 1969.
22. "Atlas of Mass Spectral Data," E. Stenhagen, S. Abrahamsson, and
F. W. McLafferty Eds., Wiley-Interscience, New York, 1969.
23. R. M. Silverstein and G. C. Bassler, "Spectrometric Identification
of Organic Compounds," Second Edition, John Wiley, New York, 1967,
Chapter 1.
2U. S. R. Shrader, "Introductory Mass Spectrometry," Allyn and Bacon,
Inc., Boston, 1971, Chapter.1.
25. J. H. Beynon and A. E. Williams, "Mass and Abundance Tables for Use
in Mass Spectrometry," Elsevier, Amsterdam, 1963.
26. "Grand Calumet River Sediment Measurements - Collection and Analysis,"
NFIC-Cincinnati Report, 1972.
27. T. Braidech, private communication.
28. J. L. Monkman, "Annual Report, International Bank of Polycyclic
Aromatic Hydrocarbons," World Health Organization, Lyon, France,
December, 1973.
68
------- |