United States
Environmental Protection
^1 Agency
EPA/600/R-14/057 | April 2014 | www.epa.gov/research
Four Fish Kills Spanning
2011 - 2013 in the Red River
Watershed Beaver Creek to
Lake Texoma, Oklahoma
RESEARCH AND DEVELOPMENT
-------
-------
Four Fish Kills Spanning
2011 -2013 in the Red River
Watershed Beaver Creek to
Lake Texoma, Oklahoma
Prepared by
Tammy Jones-Lepp
Research Chemist
U.S. Environmental Protection Agency
National Exposure Research Laboratory
Environmental Sciences Division
Las Vegas, NV89119
Although this work was reviewed by EPA and approved for publication, it may not necessarily reflect
official Agency policy. Mention of trade names and commercial products does not constitute
endorsement or recommendation for use.
U.S. Environmental Protection Agency
Office of Research and Development
Washington, DC 20460
-------
-------
The information in this document has been funded by the United States Environmental
Protection Agency. It has been subjected to the Agency's peer and administrative review and
has been approved for publication as an external EPA document.
-------
FOREWORD
The U.S. Environmental Protection Agency (EPA) is charged by Congress to protect the nation's
natural resources. Under the mandate of national environmental laws, the EPA strives to
formulate and implement actions leading to a compatible balance between human activities and
the ability of natural systems to support and nurture life. To meet this mandate, the EPA's Office
of Research and Development (ORD) provides data and scientific support that can be used to
solve environmental problems, build the scientific knowledge base needed to manage ecological
resources wisely, understand how pollutants affect public health, and prevent or reduce
environmental risks.
The National Exposure Research Laboratory (NERL) is the Agency's center for investigation of
technical and management approaches for identifying and quantifying exposures to human health
and the environment. Goals of the laboratory's research program are to: (1) develop and
evaluate methods and technologies for characterizing and monitoring air, soil, and water; (2)
support regulatory and policy decisions; and (3) provide the scientific support needed to ensure
effective implementation of environmental regulations and strategies.
The USEPA/ORD-National Exposure Research Laboratory-Environmental Sciences Division
(USEPA/ORD-NERL-ESD) assisted USEPA Region 6 and the State of Oklahoma Department
of Environmental Quality (OKDEQ) in identifying unknown contaminant(s) that were present
during four fish kills in the Red River watershed. These environmental samples were unique in
that they were collected during the active phase of the fish kills along the Red River (Oklahoma,
United States) in 2011, 2012 and 2013. Using liquid chromatography-time-of-flight high-
resolution mass spectrometry (LC-TOFMS), LC-Fourier transform mass spectrometry (LC-
FTMS) and/or liquid chromatography-ion trap mass spectrometry (LC-ITMS), the conditional
assignments of the molecular weights and chemical formulas of the significant unknown
contaminants were determined.
3
-------
EXECUTIVE SUMMARY
This research was conducted under the auspices of the US Environmental Protection Agency's
Office of Research and Development (USEPA ORD) Safe and Sustainable Water Resources
Research (SSWR) program: Theme 1, Q7 (Highly Targeted Programmatic Support). Since
December 2011, the USEPA ORD, National Exposure Research Laboratory-Environmental
Sciences Division (USEPA ORD-NERL-ESD) has assisted EPA Region 6 and the State of
Oklahoma Department of Environmental Quality (OKDEQ) in identifying unknown
contaminants that were present during four fish kills in the Red River watershed. These
environmental samples were unique in that they were collected during the active phase of the
fish kills along the Red River (Oklahoma, United States). There were a total of four fish kills:
two occurred in July and September 2011, one in June 2012, and one in January 2013; they will
hereafter be referred to as fish kill I, II, III and IV.
Using liquid chromatography-time-of-flight high-resolution mass spectrometry (LC-TOFMS),
LC-Fourier transform mass spectrometry (LC-FTMS) and/or LC-ion trap mass spectrometry
(LC-ITMS), the conditional assignments of the molecular weights and chemical formulas of the
significant unknown contaminants were determined. Environmental water samples were
extracted using a solid phase extraction (SPE) method. Sediment samples were extracted using a
modified sonication liquid extraction method. All extracts were screened and analyzed by LC-
ITMS, LC-TOFMS, and/or LC-FTMS. Subsequently, the extracts were then re-analyzed using
collision induced dissociation (CID) (either in the ion trap, or in-source CID for TOFMS and
FTMS) for product ion formation to elucidate chemical structural components, and re-analyzed
by LC-TOFMS and LC-FTMS for accurate mass assignments. Many chromatographic peaks
were present, but most could be attributable to ambient background contamination (e.g.,
surfactants and phthalates). From the screening analyses of the samples, two major unknowns
were discovered in three of the four fish kills, detected at masses m/z 624.3 Da and m/z 639.3 Da.
The unknown at mass m/z 639.3 Da has been unequivocally identified as a porphyrin,
specifically, chlorin e6 trimethyl ester. In fish kill III samples there was no evidence of chlorin
e6 trimethyl ester. Instead, in fish kill III samples, there were two large chromatographic peaks
detected at different masses. The peaks were identified at masses, m/z 562.3760 Da (M+H)+,
C33H48N5O3, and m/z 564.3898 Da (M+H)+ C33H50N5O3. At this time, it would be speculation to
suggest which chemical class these two compounds belong to, whether a porphyrin, a mycotoxin
(as suspected from earlier identification efforts), or some other unknown chemical class.
Another significant unknown was detected in only one sample from the fish kill IV. This
unknown eluted earlier than the porphyrin series and was assigned the chemical formula:
C46H94N6O6, with an accurate mass of m/z 826.72275 (M+ ) [doubly charged ion detected at: m/z
+2
413.36039 (M )]. This chemical has been tentatively identified as belonging to the chemical
class of diquaternary ammonium compounds.
There is evidence that the presence of chlorin e6 trimethyl ester is relational to the dying fish, but
this is just a hypothesis. While the unequivocal identification of one unknown emerging
contaminant was made in fish kill I, II, and IV samples, there are many other unidentified
chromatographic peaks present in both the water and sediment extracts. Only those
chromatographic peaks and masses that were substantially above the chromatographic baseline,
and not detected in the blank samples, were scrutinized.
4
-------
TABLE OF CONTENTS
Notice 2
Foreword 3
Executive Summary 4
Acronyms and Abbreviations 6
Acknowledgments 7
1.0 Introduction and Background 8
2.0 Experimental 9
2.1 Sampling 9
2.2 Water extraction and analysis 9
2.3 Sediment extraction and analysis 9
3.0 Results and Discussion 10
3.1 Major chromatographic unknowns observed in fish kill IV 10
3.2 Other unknowns tentatively identified in fish kill IV sample(s) 12
3.3 Relevance of new data observations to earlier fish kills I and II 12
3.4 Sediment samples 12
4.0 Conclusions 13
Figures
Tables
References
Appendix A - SOP on "Extraction and Analysis of Emerging Contaminants in Aqueous samples
using SPE and LC-ITMS"
Appendix B - GCMS report on the semi-volatile compounds detected in fish kill IV water
samples
5
-------
LIST OF ACRONYMS AND ABBREVIATIONS
ACN
acetonitrile
C
centigrade
CID
collision induced dissociation
CSS
Chemical Safety for Sustainability
CWA
Clean Water Act
ECs
emerging contaminants
EPA
Environmental Protection Agency
ESD
Environmental Sciences Division
ITMS
ion trap mass spectrometer
FTMS
Fourier transform mass spectrometer
GC
gas chromatograph
g
gram
HPLC
high performance liquid chromatography
kg
kilogram
L
liter
LC
liquid chromatograph
(M+-)
intact molecule minus an electron
(M+H)+
intact molecule plus a proton (H+)
MeOH
methanol
mL
milliliter
MOE-Ontario
Ministry of the Environment-Ontario (Canada)
MS
mass spectrometer
MTBE
methyl tertbutyl ether
NERL
National Exposure Research Laboratory
NaCl
sodium chloride
NH4OH
Ammonium hydroxide
OKDEQ
State of Oklahoma Department of Environmental Quality
ORD
Office of Research and Development
ppb
part-per-billion
PPE
personal protective equipment
ppt
part-per-trillion
rpm
revolutions per minute
SIM
single ion monitoring
SPE
solid phase extraction
SSWR
Safe and Sustainable Water Research
TOFMS
time-of-flight mass spectrometer
|j,L
microliter
UPLC
ultra performance liquid chromatography
WWTP
wastewater treatment plant
-------
A cknowledsements
I'd like to thank my colleagues, Dr Wayne Sovocool, Dr Don Betowski, Dr Patrick
DeArmond, Mrs Charlita Rosal, and Dr Vince Taguchi (Canadian Ministry of the Environment-
Ontario) for all of their hard work in analyzing and deciphering the Red River samples. None of
this research would have been possible without their dedication and assistance. I'd also like to
thank my students, Mr Trevor Nance Jr. and Mr Matt Ward for their assistance in extracting
numerous water and sediment samples, sometimes repeatedly for precision and accuracy.
7
-------
1.0 Introduction and Background
Since December 2011, the USEPA ORD-NERL-ESD has assisted EPA Region 6 and the
OKDEQ in identifying unknown contaminants that were present during four fish kill events in
the Red River watershed.
The Red River is a tributary of the Mississippi River, with headwaters in the Texas panhandle,
flowing for 917 kilometers between the borders of Oklahoma (OK) and Texas (TX), before
eventually emptying into the Mississippi River. The fish kills were located in the Red River, or
its tributaries, from Ryan, OK to Lake Texhoma, OK. Three of the fish kills (I, II and III) were
located near the confluence of Red Creek (at Ketchum's Bluff) and the Red River, and one (IV)
was localized to Beaver Creek, which runs alongside Ryan, OK, and flows into the Red River,
north of Ketchum's Bluff. In the first three fish kills, only large bottom feeder fish (i.e., catfish
and buffalo), were observed dead or dying. The last fish kill IV was unique in that not only fish,
but also other animals (i.e., hardshell and softshell turtles) were affected.
In July 2011, the first fish kill (fish kill I) was observed to occur in the Red River near
Ketchum's Bluff, OK. Nearly two months later, in September 2011, another fish kill (fish kill II)
was observed happening further south along the Red River, approximately 130 km downstream
from Ketchum's Bluff near Lake Texhoma. In December of 2011, Region 6 asked the ORD-
NERL-ESD laboratory in Las Vegas, Nevada, for assistance in possibly identifying the unknown
toxicant(s) potentially causing the fish kills. During the active phases of fish kills I and II,
multiple sites were sampled. Originally, the water samples from fish kills I and II had been sent
to various laboratories (i.e., USEPA Region 6, USEPA National Enforcement Investigations
Center-Denver, Oklahoma State Environmental Laboratory Services, and the Texas Parks and
Wildlife Inland Fisheries Environmental Contaminants Laboratories) for routine traditional
analyses (e.g., volatiles, semi-volatiles, metals), and fish necropsies and tissue analyses. With no
reasonable cause identified with these analyses, the archived water and sediment samples were
sent to ORD-NERL-ESD laboratory to perform analyses for emerging contaminants. A
preliminary report of the findings was submitted to USEPA Region 6 and OKDEQ in March
2012.
On June 12, 2012, USEPA Region 6 and OKDEQ notified ORD-NERL-ESD that another fish
kill (fish kill III) was in progress on the Red River, and again assistance was requested. This fish
kill started at almost exactly the same location (Ketchum Bluff area) as fish kill I. On June 20,
2012, ORD-NERL-ESD received 8 water samples (6 water samples and two trip blanks). The
water samples were collected during the observed fish kill (fish dead or actively dying) in the
Red River on June 12 and 13, 2012. ORD-NERL-ESD performed the analyses for emerging
contaminants and a preliminary report on the initial findings was delivered to Region 6 and
OKDEQ in August 2012.
In early February 2013, ORD-NERL-ESD was notified by USEPA Region 6 and OKDEQ that a
fourth fish kill was in progress on Beaver Creek, a small tributary of the Red River, near Ryan,
OK. It was reported to ORD that this fish kill (fish kill IV) was dissimilar in nature to the
previous observed fish kills, reported in the Red River in 2011 and 2012 (Red River fish kills I,
II and III), in that a diversity of animals were adversely affected, not only fish, but also dead and
8
-------
dying hard- and soft-shell turtles. On February 4, 2013, four water samples, including one travel
blank and three environmental water samples, were sent for analyses to ORD-NERL-ESD.
The primary objective of this report is to present what was detected in water samples from fish
kill IV, and relate how these findings impact the conclusions from the earlier fish kills I-III. The
second objective is to discuss what was detected in the sediment samples, and their possible
relationship to the water samples. Also attached to this report, is a standard operating procedure
(SOP): "Extraction and Detection of Emerging Contaminants using solid-phase extraction and
liquid chromatography-ion trap mass spectrometry" (Appendix A). The SOP is provided for
USEPA Region 6 and OKDEQ, such that they can repeat the experiments that were performed
and continue the surveillance of the Red River for possible toxicants.
2.0 Experimental
2.1 Sampling. OKDEQ collected and shipped all water and sediment samples according to their
sampling protocols. All water and sediment samples were collected as grab samples. The initial
water samples sent to ORD-NERL-ESD from fish kills I and II were archived samples,
refrigerated and stored by OKDEQ at < 4°C. The other samples collected during fish kills III
and IV, and during non-fish kill events (background) were stored on ice, or refrigerated, and sent
to ORD-NERL-ESD within one to four days from sampling event.
The sediment samples were collected at various times throughout the last two years. Some of the
sediment samples were collected during the fish kill events, and other sediment samples were
collected at sites upstream from the fish kills during non-events, to be used as background
samples.
2.2 Water extraction and analysis. Briefly, four water samples (a travel blank and three
environmental waters) were extracted using a solid phase extraction (SPE) method. All
Oklahoma fish kill water samples were extracted at a pH < 3. This lower pH was necessary as
OKDEQ reported that the water samples formed a cloudy colloidal suspension when a base was
added to the initial samples from fish kills I and II.
Detection analyses were performed using mass spectrometry, LC-ITMS (in-house) and LC-
TOFMS (in-house), or LC-FTMS [analyses performed by Canadian Ministry of the
Environment-Ontario (MOE-Ontario)]. Splits of the extracts were sent to MOE-Ontario for
analysis by LC-FTMS in order to obtain greater mass accuracy than the in-house LC-TOFMS
could assign.
For in-depth aqueous extraction method details, see Appendix A: SOP on "Extraction and
detection of emerging contaminants using solid-phase extraction and liquid chromatography-ion
trap mass spectrometry."
2.3 Sediment extraction and analysis. Over the last three years, several sediment samples were
received and archived in ORD-NERL-ESD walk-in refrigerator, < 4°C. The sediment samples
were extracted using a crude extraction method. One gram of sediment was weighed into a small
9
-------
(250 mL) beaker, and labeled internal standards were added. The samples were placed under a
hood and allowed to dry (approx. 2hrs). After drying, 5 mL of solvent (5% NH40H/95%
MeOH) was added to each sample. The samples were then placed in a sonicator and sonicated
for 5 minutes. Samples were removed from the sonicator and the solvent layer was transferred to
15 mL capped glass centrifuge vials. These three steps [addition, sonication, and removal of 5
mL of solvent (5% NH40H/95% MeOH)] were repeated two more times until approximately 15
mL of solvent had been collected. Vials (containing solvent layers) were placed in the centrifuge
and spun for 5 minutes at 670 revolutions per minute (rpm). After the 5 minutes the centrifuge
was increased to 1675 rpm for an additional 5 minutes. Each sample was then rinsed with 4 mL
of hexane, a hexane layer was allowed to form, which was then removed and discarded. The
supernatant was poured into 50-mL concentrator tubes, and the solid remaining in the centrifuge
vials was discarded. The concentrator tubes were placed in a semi-automated evaporator
(TurboVap™), the nitrogen stream was set to approx. 7 psi, and the supernatants were
concentrated to 0.5 mL. The concentrated supernatant was subsequently transferred to
autosampler vials for analysis by LC-TOFMS or LC-ITMS.
Screening analyses of sediment extracts were performed using mass spectrometry, LC-ITMS and
LC-TOFMS. See Appendix A for further mass spectrometric analytical details.
3.0 Results and Discussion
3.1 Major chromatographic unknowns observed in fish kill IV. From the screening analyses of
fish kill IV samples two major polar non-volatile unknowns, that were discovered in fish kills I
and II (detected at masses m/z 624.3 Da and m/z 639.3 Da), were again present in significant
amounts in the three environmental water samples. Fortunately enough water sample (2-L) had
been collected with fish kill IV to allow for two sets of extractions. The second set of extracts
was sent to Dr. Vince Taguchi, at Canada's Ministry of the Environment-Ontario (MOE-Ontario)
for further mass spectrometric analyses. MOE-Ontario was able to obtain more detailed accurate
mass and structural information using MOE-Ontario's LC-FTMS than was possible on ORD-
NERL-ESD mass spectrometers.
The information obtained from LC-FTMS gave the following accurate masses: m/z 639.31735
(M+H)+, generating the molecular formula, C37H43N4O6, and m/z 624.31794 (M+H)+, generating
the molecular formula, C36H42N5O5. By searching web resources, it was discerned that the
unknown, at mass m/z 639.31735 (M+H)+, was not a mycotoxin, as had been previously
hypothesized from fish kills I and II. Instead the unknown at mass m/z 639.3 Da was identified
as a geoporphyrin, specifically chlorin e6 trimethyl ester (Figure 1), mw 638.310425 Da,
C37H42N4O6. In order to be indisputably certain that this was the correct identification a standard
of chlorin e6 trimethyl ester was obtained from Frontier Scientific (Logan, Utah). Using the
collision induced dissociation (CID) function of the ORD-NERL-ESD LC-ITMS, a CID mass
spectra of the standard was obtained and compared to the unknown spectra detected at mass m/z
639.4 Da (M+H)+ in fish kill IV extracts, and a positive confirmation was made (Figures 2a and
2b). To ensure accuracy of the identification, verification was completed on a second mass
spectrometer. In-source CID experiments were carried out using the LC-TOFMS, and a
comparison of the spectra in Figure 2d to the spectra in Figure 2c, confirmed the identification.
10
-------
Figure 3 shows the identification of the major fragmentation pathways of the three major product
ions that were formed during CID.
The other major unknown present in fish kill IV extracts, at m/z 624.3 Da (M+H)+, was also
previously detected in fish kill samples I and II. This unknown is chemically related to chlorin
e6 trimethyl ester and is an artifact that was accidentally created during the SPE elution process.
The core chemical structure is very similar to chlorin e6 trimethyl ester. A tentative
identification was assigned as an amide-containing porphyrin by comparing the CID spectra
from the LC-ITMS data, the LC-TOFMS data, and the LC-FTMS data. The molecular formula,
as calculated by LC-FTMS, is C36H42N5O5, m/z 624.31794 (M+H)+. There are three methyl ester
groups that are potential sites for amide formation, and the detection of two major products
[Figure 5(b)], suggests that two of the three possible sites are more accessible to ammonolysis-
type reactions. A series of chemical synthesis experiments were performed to test the hypothesis
that this compound, C36H42N5O5, was an artifact of extracting the samples containing the
porphyrin, chlorin e6 trimethyl ester, with the 95% MeOH/5% NH4OH solution. Figure 5 shows
a (a) chromatogram of unreacted chlorin e6 trimethyl ester, and (b) chromatogram of reacted
chlorin e6 trimethyl ester with 95% MeOH/5% NH4OH solution. The m/z 624.3 Da (M+H)+
ions are nonexistent in the ion chromatogram (a) of the unreacted standard of chlorin e6
trimethyl ester, while the ion chromatogram (b) clearly shows the presence of two ions at two
different retention times with the mass m/z 624.3 Da (M+H)+. Figure 4 is just one possible
structure hypothesized of one of the isomeric amides that was formed by ammonolysis of the
chlorin e6 trimethyl ester. There is no commercial chemical standard available at this time to
compare and confirm with the unknown.
Originally, the masses m/z 624.3 Da (M+H)+ and m/z 639.3 Da (M+H)+, detected in fish kill
samples I and II, were misidentified as mycotoxins. This misinterpretation came about because
the accurate mass that was measured in-house using high-resolution TOFMS, was m/z 624.3175
Da, (M+H)+, and a molecular formula of C36H42N5O5 was generated. At that time a search of the
scientific literature generated a newly discovered mycotoxin, ergosedmine, by Uhlig (Uhlig et al.
2011), assigned molecular formula of C36H42N5O5, but a measured exact mass of 624.3202,
(M+H)+. The difference of 2.7 mmu between the two mass spectral measurements for that
chemical formula equated to a < 4 ppm difference; this is well within acceptable limits for those
types of measurements made by the LC-TOFMS. However, as discussed earlier, enough water
sample during fish kill IV was collected such that a second set of extracts were generated and
sent to MOE-Ontario laboratory for further confirmation. MOE-Ontario has a LC-FTMS, which
is capable of measuring even greater mass accuracy than ORD-NERL-ESD's LC-TOFMS. The
measured mass from the FTMS was m/z 624.31805 Da, generating the molecular formula of
C36H42N5O5. The measured mass from FTMS of the other compound was m/z 639.31736 Da,
generating the molecular formula, C37H43N4O6. Although the molecular formula didn't change
(with regards to the ones originally generated from the TOFMS data), the more accurate mass
allowed for the generation of a more accurate composition of the compound(s), as well as
specific generation of rings plus bonds calculations. Also, FTMS in-source CID of the accurate
mass at m/z 639.31736 Da gave very stable fragment ions, like the porphyrin nucleus with their
ester groups. Using the accurate mass fragments and neutral losses to form fragments, allowed
for re-constructing an accurate chemical structure. Both pieces of information, accurate mass
and accurate mass fragments, from the FTMS data, allowed for the re-computed identification of
11
-------
the unknown emerging contaminant at mass m/z 639.31736 Da (M+H)+ as chlorin e6 trimethyl
ester. It was also fortuitous that a standard of the hypothesized contaminant was commercially
available, leading to an unequivocal identification of the unknown at mass m/z 639.31736 Da
(M+H)+ as chlorin e6 trimethyl ester.
3.2 Other unknowns tentatively identified in fish kill IV sample(s). Another significant polar
non-volatile unknown was detected in only one fish kill IV sample (Beaver Creek Main Street
site) by both ORD-NERL-ESD and MOE-Ontario. This unknown eluted early (before the
porphyrin series) in the total ion chromatogram (TIC), and was assigned the chemical formula:
C46H94N6O6, with an accurate mass of m/z 826.72275 (M+ ) [the doubly charged ion was also
detected at: m/z 413.36039 (M+2)]. From the mass spectra obtained, this chemical has been
identified as belonging to the chemical class of diquaternary ammonium compounds. Using the
accurate mass provided by the LC-FTMS [m/z 826.72275 (M+)] and a search of relevant
chemical databases, it has been tentatively identified as N,N,N,N',N',N'-Hexamethyl-4,20,27,43-
tetraoxo-3,44-dioxa-6,19,28,41-tetraazahexatetracontane-l,46-diaminium; with a theoretical
monoisotopic mass of, 826.722412 Da (ChemSpider, a free on-line chemical data base,
www.chemspider.com). There is no commercial chemical standard available for purchase at this
time for confirmation.
Fish kill IV extracts were also split in-house for gas chromatography-mass spectrometry (GC-
MS) analysis of the semi-volatile fraction. The results of those analyses are reported in
Appendix B - "GCMS Analysis on water extracts dated 2.22.2013." The main finding from the
GC-MS analyses was the discovery of the pharmaceutical gabapentin, and low-levels of several
alkyl organophosphate flame retardants. None of these compounds were totally unexpected due
to reports in the literature of these types of compounds in global surface waters (Kasprzyk-
Horden et al. 2009; Regnery et al. 2010).
3.3 Relevance of new data observations to earlier fish kills I and II. Two water samples from
fish kill I (OK ID 506352 - LV12wat004, and OK ID 506353 - LV12wat008) were re-extracted
and re-analyzed by LC-ITMS, using the optimized CID conditions for detecting chlorin e6
trimethyl ester. These water samples had been archived by ORD-NERL-ESD since January
2012. They were stored in their original V2 gallon plastic jugs in a walk-in refrigerator at ~ 4° C.
The unknown mass, m/z 639.3 Da (M+H)+, that had previously been detected during the first
analyses back in February 2012, was again re-detected in these samples and now positively
confirmed as chlorin e6 trimethyl ester using the CID mass spectrum of the unknowns at m/z
639.3 Da (M+H)+ and comparing them to the chlorin e6 trimethyl ester CID spectra. For the
other archived water samples, from fish kill I and II, the archived CID spectra (if available, LC-
ITMS or LCTOFMS), were compared to the current CID spectra of chlorin e6 trimethyl ester for
affirming confirmation (Table 1).
3.4 Sediment samples. Over the course of the last three years, fifteen sediment samples were
collected and received from OKDEQ, and archived in the ORD-NERL-ESD walk-in refrigerator,
< 4°C. The sediment samples were recently extracted and analyzed by LC-ITMS, see methods
section 2.3. Of the fifteen sediment samples, chlorin e6 trimethyl ester was positively confirmed
in seven sediments (Table 2).
12
-------
4.0 Conclusions
The major unknown identified from Fish Kills I, II, and IV, was chlorin e6 trimethyl ester.
Chlorine6 trimethyl ester belongs to the porphyrin chemical class; for example, chlorophyll and
hemoglobin are considered porphyrins. Some porphyrins are termed geoporphyrins, and many
are chemically fingerprinted to global oil and oil shale deposits. There is one specific group of
geoporphyrins that are unique to the Ordovician Viola and Arbuckle formations found
underneath south central Oklahoma (Michael et al. 1989). It is possible that the geoporphyrin
that was detected in the water samples may belong to these Oklahoma formations. The particular
geoporphyrin that was detected is thought to possibly emanate from an organism unique to this
formation, Gloeocapsamorpha priscas, which was possibly a blue-green alga or large bacterium
present millions of years ago in the primitive oceans (Michael et al. 1989). The reasoning behind
this is the lack of the phytyl group (the chemical side chain for chlorophyll) on the geoporphyrin.
Pickering (Pickering 2009) gives a very good explanation on the possible formation of these
compound in his dissertation "Low temperature sequestration of photosynthetic pigments: Model
studies and natural aquatic environments."
There was no evidence of chlorine6 trimethyl ester in fish kill III samples. Instead two other
large chromatographic peaks were detected, and the masses were identified at m/z 562.3760 Da
(M+H)+, C33H48N5O3 , and m/z 564.3898 Da (M+H)+ C33H50N5O3. At this time it would be
speculation to suggest which chemical class these two compounds belong to, whether a
geoporphyrin, a mycotoxin, or some other unknown chemical class.
It can only be hypothesized as to whether the chlorin e6 trimethyl ester was responsible, or just
relational, to fish kills I, II, and IV. There is some evidence, Figure 6, that the presence of
chlorin e6 trimethyl ester is relational to the dying fish, but it is just a hypothesis at this point in
time.
While the unequivocal identification of one emerging contaminant unknown has been made in
fish kill I, II, and IV samples, there are many other unidentified chromatographic peaks present
in both the water and sediment extracts. We have focused only on those chromatographic peaks
and masses that were substantially above the chromatographic baseline and not detected in the
blank samples.
13
-------
Figure 1. Chlorine6 tri methyl ester, C37H42N4O6
CH
CH
CH
\
CH3
-------
Figure 2a. CID MSMS LC-ITMS: Chlorine6 trimethylester standard, m/z 639.3 (M+H)
Spectrum 1A, 10 415min
BP: 579.4 (7.964e-+-6=1 OQ%), rednver294 xms
4794
506907
¦ 1 I .. L ¦
'2
-------
Figure 2b. CID MSMS LC-ITMS: Unknown m/z 639.3 (M ill) in sample Ivl3wat008
SpKltim -A, 12.115 nin
1COSH
75fe-
Iv13wat008
33%-
25
Ott-
5964
2S13&4
«
565.4
"!26024
ti
i
S&3
573/
41&278
t:
5B1-5
£64
*
5814
186
"TT
607.4
2648$
8»
IE"
653
ma
16
-------
Figure 2. In-source CID TOFMS of: (2c) unknown m/z 639 (M-i-H) in Sample Ivl3wat0082;
and (2d) m/z 639 (M+H) in Chlorine6 trimethylester standard
MaCsl
Mass Error ~ 2.5ppm
18-Apr-2013
10:36:33
1: TOFMS ES+
479.2426
(c)
it
566.2
333.2422 480.2462
866
640.3198
305.2114
765.4543
184.0734 334'2462
231.1742
|tJ [ ,336.2589
ll
760.4990
661.2998 767.4579 ,
976.6091 1052.4065
20131121chlorine6tmeCID 30 (0.494) Cm (11:33)
639.3058
T0FMSES+
1.75e4
(d)
479.2406
566.2805
480,2438
465.2264
105.0422
l.iil.U
677.2653
17
-------
Figure 3. Pathways of production formation from chlorine6 trimethylester ion (M+H) .
H,C
H,C
r
X
m/z 579
(M+H-C2H402)+
H,C
m/z 566
(M+H-C3Hr,02)<
W m/z 607
(M+H-CH40)+
18
-------
Figure 4. Likely ammonolysis transformation product of chlroine6 trimetylester,
yielding m z 624 Da, (M+H)
H+
CH
HN
CH
/CH
H N
2
-------
Figure 5. Ion chromatograms of (a) unreacted chlorine6 trimethylester and (b) reacted chlorin
e6 trimethylester with ammonium hydroxide solution.
chlorin e6 TWIE 1 ng/uL 020314
Resolution Positive W Mode
testfeb14_007
100
0
/ft
0.47 079 1.59
2-56 2.84 3.13 3 84 4.68 , „ ^6.54,
3 o4 5 11 6 20 o 11 i'I'tJi) ' i||kt 1 1 '|'i f| ill I 1 14.82
,',ir;'feTV !;J/. fi'i!1 li 8'^lH/J Alii. il;'
: testfeb 14_007
; 100
0
testfeb 14_007
100
0
0.87
testfeb14_009
100
0
testfeb 14_009
100
0'
testfeb14_009
100
0.07
2.00
2.00
2.00
2.00
2.00
2.00
4.00
4.00
4.00
4.00
4.00
4.00
6.00
6.00^2
. i.hUl I :ll . I I' ' ill I I, ill
11.85 12.6612.97
03-Feb-2014
12:01:23
1: TOF MS ES+
624.3 0.20Da
15.2
14.12
"'14.95
3 8.00
x\
10.00
CO
14.00
1:TOF MS ES+
639.3 0.20Da
9.29e4
8.00
10.00
12.00
J
12.06
7.65
10.30 11:31
12.82 1317
14.00
1: TOF MSES+
TIC
2.04e5
6.00
8.00
4 O)
10.00
9 71 10.40
12.00
6.00 lii
£
14.00
1: TOF MSES+
624.3 0.20Da
1.81e3
11.12
A
J 4
u
8.00
10.00
14.00
1: TOF MS ES+!
639.3 0.20Dai
7.92e4!
6.00
8.00
10.00
12.00
12.09
10.32 11-14
12.82
13.18
14.00
1: TOF MS ES+
TIC
1,74e5
6.00
8.00
10.00
12.00
14.00
Time
20
-------
Figure 6. Fish kill graph and mass spectra of unknown m/z 639.3
Looking back
A water sample from Fish Kill II
25000000
«D«> 06 Uv 201* M 15
MS Prt - W2012 21? PM
Red River Fish Kill Peak Areas
¦ M624.3
¦ M639.3
precursor m/z 639
None Dying Dying None
7/9/11 9/14/11 9/14/11 9/14/11
Site #5 Site #7 Site #6 Site #1 Site #2 Site #3 Site #4 Site #20 Site #19 Site #18
Upstream ¦
->Downstream
21
-------
Table 1. ITMS or TOFMS screen and confirmation of Chlorine6 TME by CID ITMS or TOFMS
Confirmed by CID
ITMS and/or TOFMS
sample ID
Date collected
639.3
Site 1 Ivl2wat002
07/09/11
X
No sample left
Site 2 Ivl2wat007
07/09/11
X
Yes (02/13/12)
Site 2 Ivl2wat004
07/09/11
X
Yes (09/18/13)
Site 3 Ivl2wat011
07/09/11
X
No sample left
Site 3 Ivl2wat008
07/09/11
X
Yes (09/18/13)
Site 4 Ivl2wat013
07/09/11
X
No sample left
Site 5 Ivl2wat014
07/13/11
X
No sample left
Site 6 Ivl2wat017
07/13/11
X
No sample left
Site 7 Ivl2wat020
07/13/11
X
No sample left
Site 18 Ivl2wat023
09/14/11
X
Not confirmed
Site 19 Ivl2wat025
09/14/11
X
Yes
Site 20 Ivl2wat027
09/14/11
X
Yes
sample ID
Date collected
639.3
lvl2watl 10A
06/13/12
nd
lvl2watl 10B
06/13/12
nd
lvl2watlllA*
06/13/12
nd
masses m/z 562.3 and m/z 564.3 present
lvl2watll2A
06/13/12
nd
lvl2watl 12B
06/13/12
nd
lvl2watll3A*
06/13/12
nd
masses m/z 562.3 and m/z 564.3 present
lvl2watl 14 (TB)
06/13/12
nd
lvl2watll5A*
06/12/12
nd
masses m/z 562.3 and m/z 564.3 present
lvl2watl 16A
06/12/12
nd
lvl2watl 16B
06/12/12
nd
Ivl2watll7 (TB)
06/12/12
nd
Ivl2watll8a
06/14/12
nd
lvl2watll9A
06/15/12
nd
22
-------
lvl2watl20A
06/21/12
nd
Ivl2watl21
06/21/12
nd
Ivl2watl22
06/20/12
nd
lvl2watl23A
06/20/12
nd
lvl2watl24A
06/21/12
nd
lvl2watl25A
06/21/12
nd
lvl2watl26A
06/21/12
nd
sample ID
Date collected
639.3
Ivl2watl35
02/14/12
nd
Ivl2watl36
02/14/12
nd
Ivl2watl37
02/14/12
nd
Ivl2watl39
02/14/12
nd
Ivl2watl40
02/14/12
nd
Ivl2watl45
02/14/12
nd
Ivl2watl46
02/14/12
nd
sample ID
Date collected
639.3
Ivl3wat006
01/31/13
nd
Ivl3wat007
01/31/13
X
Yes
lvlSwatOOS*
01/31/13
X
Yes
Ivl3wat009
01/31/13
X
Yes
sample ID
Date collected
639.3
Ivl3wat018
09/04/13
nd
Ivl3wat019
09/04/13
X
Yes
Ivl3wat020
09/03/13
X
Yes
Ivl3wat021
09/04/13
X
Yes
Ivl3wat022
09/03/13
X
Yes
Ivl3wat023
09/03/13
X
Yes
sample ID
Date collected
639.3
These three samples have possibly different geoporphyrins present at masses m/z 562.3 (M+H)+ and m/z 564.3 (M+H)+
Large mass detected at m/z 826.7 Da (M+); nd = not detected; x = detected during screening analysis
-------
Table 2. Sediment data - CID ITMS screening for m/z 639.3, chlorine6 trimethylester.
Site identification
Site #2/Approx. 5.72 mi. US of 1-35
Site #3/Approx. 3,56 mi. US of 1-35
Site #5/Ketcfauiii Bluff
Site #6/Approx. 2.24 mi. PS of BR
Site #7/Priniitive BR (a) Oscar
Hwy 89 near Courtney
Hwy 81/Ryan, Ok (Barn stockpile)
Ketchum Bluff
Co. Rd. 2940
Union Valley Rd. near Oscar
Bub Wilcoxin (lower)
Bub Wilcoxin (upper)
Hwy 32
S11RC East Tribute
S12RC N2900
MS/MS
OK ID
EPA ID
confirmation
506356B
Ivl2sed002
nd
506357 A
Ivl2sed003
nd
506648 A
Ivl2sed004
nd
506649A
Ivl2sed005
nd
506650 A
Ivl2sed006
nd
521445A
Ivl2sed007
Yes
521452A
Ivl2sed009
nd
521446A
Ivl2sed011
Yes
521447A
Ivl2sed013
Yes
521448A
Ivl2sed015
< trace
521449A
Ivl2sed017
nd
521454A
Ivl2sed019
Yes
521450A
Ivl2sed021
nd
Ivl3sed001
Yes
Ivl3sed002
Yes
24
-------
References
Kasprzyk-Horden, B., R. Dinsdale, et al. (2009). "The removal of pharmaceuticals, personal care
products, endocrine disruptors and illicit drugs during wastewater treatment and its
impact on the quality of receiving waters." Water Research 43: 363-380.
Michael, G. E., L. H. Lin, et al. (1989). "Biodegradation of tar-sand bitumens from the
Ardmore/Anadarko Basins, Oklahoma—II. Correlation of oils, tar sands and source
rocks." Organic Geochemistry 14(6): 619-633.
Pickering, M. D. (2009). Low temperature sequestration of photosynthetic pigments: Model
studies and natural aquatic environments. Department of Chemistry, University of York.
Doctor of Philosophy: 239.
Regnery, J. and W. Piittmann (2010). "Seasonal fluctuations of organophosphate concentrations
in precipitation and storm water runoff." Chemosphere 78(8): 958-964.
Uhlig, S., D. Petersen, et al. (2011). "Ergosedmine, a new peptide ergot alkaloid (ergopeptine)
from the ergot fungus, Claviceps purpurea parasitizing Calamagrostis arundinacea."
Phytochemistry Letters 4(2): 79-85.
25
-------
APPENDIX A
STANDARD OPERATING PROCEDURE FOR
EXTRACTION AND DETECTION
OF
EMERGING CONTAMINANTS
USING
SOLID-PHASE EXTRACTION
AND
LIQUID CHROMATOGRAPHY-ION TRAP MASS SPECTROMETRY
January 2014
United States Environmental Protection Agency
Environmental Sciences Division
Environmental Chemistry Branch
Brian Schumacher, Branch Chief Date
Tammy Jones-Lepp, Principal Investigator Date
Ed Heithmar, Branch Quality Assurance Representative Date
ESD Quality Assurance Manager Date
26
-------
STANDARD OPERATING PROCEDURE FOR:
EXTRACTION and ANALYSIS OF EMERGING CONTAMINANTS in
AQUEOUS SAMPLES USING SPE and LC-IONTRAP MS
1.0 Disclaimer
This standard operating procedure (SOP) has been prepared for use of the
Environmental Sciences Division, Environmental Chemistry Branch, National Exposure
Research Laboratory, Office of Research and Development, of the U.S. Environmental
Protection Agency and may not be specifically applicable to the activities of other
organizations THIS IS NOT AN OFFICIAL EPA APPROVED METHOD. This
document has not been through the Agency's peer review process or ORD clearance
process. Additionally, this SOP is equipment and/or instrument-specific.
2.0 Purpose (Scope and Application)
This document describes the procedure for the determination of emerging
contaminants (ECs), in aqueous samples using a Thermo Fisher (formerly Dionex)
Autotrace for an automated solid-phase extraction (SPE) procedure, and an Agilent
(formerly Varian) liquid chromatography-ion trap mass spectrometer (LC-ITMS) for
detection.
3.0 Method Summary
3.1 The method employs high-performance liquid chromatography (HPLC)
coupled with positive (or negative) electrospray ionization (ESI-) ion trap
collision induced (CID) mass spectrometry (MS/MS) for the determination
of emerging contaminants in aqueous matrices.
3.2 Aqueous samples are extracted through Oasis MCX SPE cartridges to
extract the ECs from solution before concentrating the eluants to 0.5 mL.
3.2 Unknown ECs are tentatively identified by using LC-ITMS and searching
known mass spectral databases and operator knowledge of mass
spectrometry. Known ECs can be quantified using select internal
standards.
4.0 Interferences
4.1 All glassware must be washed with detergents free from alkylphenol
ethoxylates. Powdered Alconox does not contain ethoxylated alcohols, but
any comparable detergent free from these interferences may also be used.
This is then followed by acid washing, rewashing in DI water, rinsing with
methanol and heated in an oven. See section 11.5.
27
-------
4.2 Method interferences can be caused by contaminants in glassware,
solvents, and other apparatus producing discrete artifacts or elevated
baselines. These materials are routinely demonstrated to be free from
interferences by analyzing laboratory reagent blanks and method blanks
under the same conditions as the samples.
4.3 Matrix interferences may be caused by contaminants that are co-extracted
from the sample.
4.4 It must also be demonstrated that the AutoTrace is free of contamination.
It has been shown that certain recalcitrant ECs can remain in the Teflon
lines of the Autotrace. After suspected EC contamination the whole
system of the AutoTrace must be cleaned with a mixture of water and
methanol (50:50) until all traces of the contaminants are purged.
4.5 Instrumentation blanks must be analyzed before, during and subsequent to
mass spectral analyses to ensure contamination free analyses. Certain ECs
are recalcitrant and can remain in various parts of the LC-ITMS, causing
"ghost" peaks, and interfering with subsequent analyses. It is incumbent
upon the mass spectrometer operator to ensure contamination free
analyses, and demonstrate this.
5.0 Safety
5.1 All of the samples and chemicals used in this procedure should be handled
only while using proper personal protective equipment such as gloves, lab
coats, safety glasses and fume hoods. The analyst should review the
Material Safety Data Sheet for each chemical in this procedure so that safe
working conditions can be achieved.
5.2 The toxicity of each sample received, and the reagents used in this method
may not be fully established. Each sample and chemical should be
regarded as a potential health hazard, and exposure should be kept as low
as reasonably achievable.
5.3 Waste must be disposed of in appropriate waste containers. Contact the
onsite SHEM Program Manager to dispose of full waste containers.
5.4 Exhaust fumes from the LC-MS must be properly vented.
5.5 All applicable safety and compliance guidelines set forth by the EPA and
by federal, state, and local regulations must be followed during the
performance of this SOP. Stop all work in the event of a known or
potential compromise to the health and safety of any person and
28
-------
immediately notify the SHEM Program Manager and other appropriate
personnel.
5.6 Analysts must be cognizant of all instrumental hazards (i.e., dangers from
electrical shock, heat, or explosion).
6.0 Reagents/Chemicals/Gases
6.1 HPLC-grade methanol
6.2 HPLC-grade water
6.3 HPLC-grade acetonitrile
6.4 Deionized (DI) water: in-house 18 MQ-cm DI water
6.5 ACS research grade methyl t-butyl ether (MTBE)
6.6 ACS reagent grade sodium chloride (NaCl)
6.7 ACS reagent grade ammonium hydroxide (NH4OH), 28% - 30%
6.8 ACS reagent grade hydrochloric acid (HC1), 12 N
6.6 Deuterated internal standards, to be chosen by the mass spectral analyst to
be consistent with possible ECs present. For example, if the target ECs
are pharmaceuticals, then use a 10 ng/|iL mixture of d3-azithromycin, d3-
clarithromycin, and 20 ng/|iL ds-MDMA, in methanol. If target ECs are
aromatase inhibitors, then use 10 ng/|iL of di2-anastrozole, d3-exemestane,
and d5-tamoxifen, and 20 ng/|iL of d4-letrozole, in acetonitrile. The
labeled pharmaceuticals cover a wide mass range and are suitable for use
with unknown ECs.
6.7 LC/MS tuning solutions available from a variety of instrument
manufacturers. Must contain compounds that are in the instrument
manufacturers tuning and mass calibration procedures. Analyst needs to
follow instrument manufacturer's protocol for mass spectrometric tuning
and mass calibration procedures.
7.0 Equipment and Supplies
7.1 HPLC-Ion Trap MS system: (Agilent (formerly Varian) 500MS coupled
with Varian/ASI HPLC and autosampler).
29
-------
7.2 HPLC column (Phenomenex Fusion RP 150 cm x 2.1 mm column, or a
Sigma-Aldrich Ascentis Ci8 100 cm x 2.1 mm column, coupled with a
Varian guard column, MetaGuard 2.0 mm Pursuit XRs 3|im Cig). Other
columns may be used if they provide sufficient retention and separation of
analytes.
7.3 Variable volume standard pipettors (0.5 -10 |iL, 20-200 |iL, 100-1000 |iL)
7.4 Disposable pipet tips
7.5 Glass beakers, volumetric flasks, sized as appropriate
7.6 Disposable borosilicate Pasteur pipets
7.7 Ultra-high-purity grade compressed nitrogen
7.8 1.8 mL autosampler vials with PTFE/silicone septa
7.11 TurboVap concentrator, and 50 mL nipple tubes, 0.5 mL endpoints, for
concentrating samples
7.12 AutoTrace 6-station SPE Workstation
7.13 Oasis MCX SPE cartridges (200 mg, 6 cc size)
8.0 Sample Collection, Preservation, and Storage
8.1 This SOP does not describe sample collection procedures; however, the
following guidelines are followed once samples are received in the
laboratory.
8.2 Samples must be stored at 4°C in a designated sample refrigerator.
8.3 Holding time studies have not been performed on these analytes; however,
samples should be analyzed as soon as possible, and within 28 days.
9.0 Quality Control
9.1 The following are relevant QC criteria for this method.
Table 1. Data Quality Indicators of Measurement Data.
QC Check
Frequency
Completeness
Precision
Accuracy
Corrective Action
Initial known
standard 3-pt
calibration
Prior to sample
analysis
100%
RSD<30%
R2 > 0.98
Review data, re-analyze.
30
-------
Laboratory
blank
One per batch of
samples3
100%
N/A
-------
11.1 Sample preparation
11.1.1 Transfer 500 mL of the environmental water sample (DI water if it's a
blank) into numerically labeled volumetric flasks.
11.1.2 Use pH paper to test pH of each water sample and record the initial pH
of sample in notebook.
11.1.3 Spike internal standard (50 |iL internal standard mix, see section 6.6)
directly into water sample, and indicate spiked samples in notebook.
11.1.3.1 Spike known QC spiking compound(s) directly into one water
sample per extraction batch. This data will be used to evaluate
extraction efficiencies.
11.1.4 Cap and shake volumetric flask.
11.1.5 Add 100 |iL of NH4OH (or 700 |iL of HC1 for acidic pH) to each 500
mL water sample.
11.1.6 Cap and shake volumetric flask.
11.1.7 Test pH of water samples.
11.1.7.1 If basic pH is desired, ensure that each water sample has a pH
>9. If the pH is not > 9, then add more NH4OH (50 |iL at a time)
until a pH of > 9 is attained. Record the final pH and the amount of
NH4OH added in notebook.
11.1.7.2 If acidic pH is desired, ensure that each water sample has a
pH of < 3. If the pH is not < 3, then add more HC1 (100 |iL at a time)
until a pH of < 3 is attained. Record the final pH and the amount of
HC1 added in notebook.
11.1.8 Add 3g of NaCl to each sample.
11.1.9 Cap and shake volumetric flask.
11.2 AutoTrace solid-phase extraction
This method was developed as an automated SPE method. If need be it could
be converted to a manual method using the appropriate SPE cartridges and a
vacuum SPE manifold, and following the steps as outlined below.
32
-------
11.2.1 Place water samples in sample holder. Rinse the outside of the sample
lines with methanol, then DI water, and place into respective prepared water
samples.
11.2.2 Program the computer to input the following conditions as outlined in
steps 11.2.3 through 11.2.6. Including a "pause" step, as noted in step 11.2.4.
Be sure to take note of the time that the machine will "pause" and be ready to
add 50 mL of DI water to each sample.
11.2.3 Load cartridges into AutoTrace SPE Workstation and precondition the
Oasis MCX cartridges with 5 mL methanol, 5 mL de-ionized (DI) water, and
95% water/5%methanol at a flow rate of 1 mL min"1. Divert eluant to waste
stream.
11.2.4 Load 500 mL aqueous sample through the SPE cartridges at a flow
rate of 7 mL min"1. After the 500 mL have loaded through the cartridges,
pause system, rinse the sample volumetric flasks with 50 mL DI water (leave
the rinsate in the flask), and continue to load the rinsate (left in the flask)
through the SPE cartridges.
11.2.5 Dry the cartridges with N2 for 40 min.
11.2.6 Elute cartridges with 5 mL 90% methyl tert butyl ether/10%) methanol,
followed by 10 mL 95%methanol/5%NH40H, at a flow rate of 1 mL min"1.
11.2.6.1 After all samples have loaded, and elution has begun,
remove lines from empty volumetric flasks. Rinse the outside of the
lines with methanol, then DI water, then place lines in container of DI
water.
11.2.6.2 Once method is complete, machine will make a beeping
noise. Press "Cont" to purge lines with a 50:50 mixture of
methanol:DI water. If emerging contaminant levels were high it may
be necessary to repeat his step 2 or 3 times.
11.2.7 Qualitatively transfer the eluate from the AutoTrace collection tube to
a TurboVap 0.5 mL (or 1 mL) endpoint nipple tubes. This involves
gently pouring and subsequent rinses of the AutoTrace elution
collection vials with final solvent to be used for mass spectrometric
analysis, i.e., methanol/1%) acetic acid, or acetonitrile/l%> acetic acid.
11.2.8 Initially set the TurboVap to a gentle nitrogen stream, approximately 3
or 4 psi. As the solvent in the nipple tubes decrease the flow can be
increased to 13 psi. During the evaporation process rinse the sides of
33
-------
the TurboVap tubes at least 4 times, with the final LC-MS compatible
solvent. Concentrate eluants to 0.5 mL.
11.2.9 Transfer the concentrated sample with Pasteur pipette to an
appropriate sized LC-MS autosampler vial, capped with a
PTFE/silicone septa.
11.2.10 Filter the samples, if necessary, with a syringe filter prior to MS
analysis.
11.3 LC-MS analysis
11.3.1 Compositions of the mobile phases were as follows: (A) DI
water/0.5% formic acid, and (B): 82% methanol/18%)
acetonitrile/0.5%) formic acid.
11.3.2 The following LC gradient is used to analyze ECs (column
temperature approx room temp 23°C)
Table 1. LC gradient conditions.
Time (min)
Flow rate
(mL/min)
%A
%B
Initial
0.30
100
0
2
0.30
100
0
5
0.30
30
70
10
0.30
30
70
13
0.30
100
0
15
0.30
100
0
11.3.3 Starting MS analysis conditions: Source conditions: Electrospray
needle voltage: 5000 to 5800kV, Ion Source temperature: 350° C,
Housing chamber 50° C; drying gas, 20 psi; nebulizer gas, 40 psi;
spray shield, 600 V. Capillary voltage and percent radio frequency
(%>RF, on the hexapoles) are set dependent upon the optimized
response of the precursor and product ions of interest. See table 2
for suggested precursor and product ions as produced by the
Varian LC-ITMS. Other LC-MSMS instruments may produce
different product ions, at varying intensities.
11.3.3.1 Operator should turn instrument gases, source and LC on
1/2 hr to 1 hr before operation. Ensure LC flow starts and is in the
inject mode into the MS. This will allow the trap to warmup and
have ions flowing into the trap and towards the detector.
34
-------
11.3.4 Load samples into the autosampler. Program autosampler method
file with correct sample id's, contents, volume injected, vial
position. Refer to manufacturer manual.
11.3.5 Select browse function in sample table and ensure proper method
is loaded.
11.3.5.1 For unknown screening, initially a full-scan method
should be utilized such that the %RF is set to 50% and the
capillary voltage to 40 eV to 60 eV. There will be ions missed at
these voltages and RF, so the operator may want to run a second
pass at different %RF and capillary voltages, if enough sample
extract permits.
11.3.5.2 Once an unknown of sufficient intensity is discovered
(the operator will need to ensure that the unknown is NOT a
background ion by analyzing a sufficient number of instrument and
method blanks) the operator will set up a MS/MS method file, such
that the CID energies are sufficient to produce product ions from
the selected precursor ion(s). A single MS/MS method can be set
up such that multiple precursor ions detected during the screening
phase can be analyzed during a single analytical run. See table 2
for several known ECs and their precursor and product ions as
produced by the Varian LC-ITMS under optimized conditions.
This types of data can be produced by other mass spectrometers
that are capable of isolating and producing product ions. As each
instrument is unique, the optimized settings will need to be set by a
skilled analyst trained in the art of mass spectrometry.
Table 2. MS/MS ions for several known ECs
Compound
Precursor ion
Major Product ion(s)
Urobilin hydrochloride
591.3 (M + H - HC1)+
343.3 [M+H- HC1 - 2(C7Hi0NO)]+
Azithromycin
749.5 (M+H)+
591.4 (M+H-C8Hi602N)+
d3-Azithromycin (ISTD)
752.5 (M+H)+
594.4 (M+H-C8Hi602N)+
Clarithromycin
748.4 (M+H)+
590.1 (M+H-C8Hi602N)+
d3-Clarithromycin (ISTD)
751.4 (M+H)+
593.4 (M+H-C8Hi602N)+
Clindamycin
425.2 (M+H)+
377.2 (M+H-SH-CH3)+
Methamphetamine
150 (M+H)+
119 (M+H-CH3NH2)+
MDMA(Ecstasy)
194 (M+H)+
163 .0 (M-CH3NH2+H)+
ds-MDMA (ISTD)
199 (M+H)+
165.0 (M-CD3NH2+H)+
Pseudoephedrine
166 (M+H)+
148.2 (M+H-H20)+
Hydrocodone
300 (M+H)+
199 (M+H-C5H1 iNO)+
Chlorin e6 trimethyl ester
639 (M+H)+
579 (M+H-C2H402)+, 566 (M+H-C3H502)+
35
-------
11.3.6 Ensure that LC solvent levels are adequate and that there is enough
N2 gas to complete the analyses. Once the instrument is ready,
begin the sample acquisition process.
11.3.7 At the end of each analytical day the operator should open the
source door, gently spray methanol onto the spray shield and wipe
the source surfaces with a clean kimwipe (or similar material).
Hazard: Spray shield is very hot, and there may be toxic
contaminants on the shield, use appropriate personal protective
gear.
11.4 Data Analysis
11.4.1 Inspect each prominent chromatographic peak for a Gaussian
appearance. The peaks may not be Gaussian in appearance due to
the presence of multiple isomers, or interferences. The operator
can try changing the LC conditions to try for better separations on
subsequent analyses.
11.4.3 Identify and confirm the presence of unknown ECs in the samples
by reviewing the total ion chromatograms (TICs) for large, > 20%
intensity above background signal, for masses that are not common
background contaminants, i.e., surfactants (unless looking
specifically for surfactants).
11.4.4 Once an intense unknown EC precursor ion has been selected the
operator will set up a MS/MS method file, such that the CID
energies are sufficient to produce product ions from the selected
precursor ion(s).
11.4.4.1 The operator will then review the CID fragment ions
produced and try to determine a structural assignment to the
fragment ions. Also, using the precursor ion the analyst should be
able to assign a molecular weight to the unknown.
11.4.4.2 If the analyst has available enough sample extract and
access to an accurate mass, mass spectrometer, for example, a
Time-of-flight mass spectrometer (TOFMS), then the extract
should be analyzed by TOFMS for accurate mass of the
unknown(s). Accurate mass can help eliminate many chemical
formulas that are generated by less accurate mass measurements.
Ideally, the ability to generate accurate mass product ions would
36
-------
help greatly in the identification of unknown emerging
contaminants.
11.4.5 To ensure good quality of the sample process on occasion samples
will be spiked with a known compound. These can be quantitated
by using isotope dilution or internal standard techniques. By
adding a known amount of a labeled compound to every sample
prior to extraction, correction for recovery of the native analog of
that compound can be made because the native compound and its
labeled analog exhibit similar chemical properties upon extraction,
concentration, and chromatography.
See Section 10 and Section 17 of the US EPA Method 1694. Note:
During calculations, take into account the concentration factor from
the 500 mL sample down to 0.5 mL following
extraction/concentration.
11.5 Glassware cleaning
11.5.1 Prepare soapy bath with hot water and approximately 1 tsp
Alconox detergent. Scrub glassware with bottle brushes and/or
pipe cleaners until visibly clean (do not scratch glassware with
metal from brushes).
11.5.2 Rinse glassware first with non-DI water, and then with DI water.
11.5.3 Soak glassware in acid bath (3 mL HC1, 3 mL HNO3, 4 L water,
pH 1-2) overnight.
11.5.4 Remove glassware and rinse with DI water; rinse glassware with
methanol and air dry.
111.5.5 Place glassware in oven at 100°C for 6 hours. Let cool in oven.
Remove cooled glassware and put away in appropriate areas.
12.0 Method Performance
12.1 Method performance can be evaluated based on the criteria in Table 1.
12.2 MDLs have not been determined yet because this is a method for
screening and identification of unknown emerging contaminants.
13.0 References
EPA Method 1694. "Pharmaceuticals and Personal Care Products in Water, Soil,
Sediment, and Biosolids by HPLC/MS/MS", 2007.
37
-------
APPENDIX B
GCMS ANALYSIS ON WATER EXTRACTS DATED 2.22.2013
Analysis done by Charlita Rosal and Wayne Sovocool
Extraction was performed by Trevor Nance Jr. and Matt Ward
The following Agilent data file numbers with corresponding sample IDs are provided
below:
#1
13022806
13022808
LV13WAT006 (blank)
DUPLICATE RUN OF
#2
13022810
13022812
LV13WAT007
DUPLICATE RUN OF
#3
13022814
13022816
LV13WAT007 DUP
DUPLICATE RUN OF
#4
13022818
13022820
LV13WAT008
DUPLICATE RUN OF
#5
13022822
13022824
LV13WAT008 spike
DUPLICATE RUN OF
#6
13022826
13022828
LV13WAT009
DUPLICATE RUN OF
Water extracts were provided in 0.5-mL volume of methanol/1% acetic acid. These
extracts were solvent exchanged and brought up to 1 mL with ethyl acetate prior to gas
chromatography-mass spectrometry (GC-MS) analysis. All samples above were
analyzed using an Agilent GC-MS in pulsed-splitless injection, electron impact (EI) scan
mode.
Results and discussion
Except for sample #1 (blank sample), all three environmental water samples contained
gabapentin at unknown concentrations.
Gabapentin (common name Neurontin) is used primarily to treat seizures,
neuropathic pain, including concussions, and hot flashes (http://en.wikipedia.org/wiki/
Gabapentin). Gabapentin (l-(aminomethyl) cyclohexaneacetic acid), has a molecular
formula of C9H17NO2, and mwl71.24 Da. Gabapentin is a white to off-white crystalline
solid with a pKai of 3.7 and a pKa2 of 10.7. It is freely soluble in water and both basic and
acidic aqueous solutions.
The structural formula of gabapentin is:
38
-------
In order to confirm the presence of gabapentin in the water samples, we acquired
prescription gabapentin in 100-mg capsules, as well as a neat standard from Sigma-
Aldrich. We dissolved these materials individually in Dl/methanol and then analyzed by
GC-MS using the same method as the water samples. Retention time and spectra of both
the prescription drug and the neat standard confirmed that what was detected in the
environmental water samples was gabapentin. Using the same approach as was used with
the environmental water samples an extraction of gabapentin was repeated in DI water.
Two 100-mg portions of prescription gabapentin were each dissolved in 500-mL DI
water (200 mg/L), extracted, and analyzed by GC-MS. Preliminary studies suggest that
high levels of gabapentin could be present in the water samples. However, the extraction
procedure used was not optimized for this compound, therefore further method
optimization would be necessary. Also, this compound is not ideal for GC-MS analysis
and confirmation by liquid chromatography-mass spectrometry (LC-MS) would be a
better analytical technique.
Another interesting peak was detected in all of the water extracts (except the blank). This
included those extracts from DI water spiked with neat standard and the prescription
gabapentin. The spectra resemble that of gabapentin's, but with a base peak of m/z 195
Da. This unknown did not show in the neat standard and prescription gabapentin
dissolved in solvent and directly injected into the GC-MS. No further identification was
done on this unknown.
Additionally, low levels of several alkylorganophosphate fire retardants (CAS#'s 13674-
84-5, 115-96-8, and 137909-40-1) were found in the samples. However, a literature
search did not find much evidence for fish toxicity for these compounds. Low levels of
the pesticide terbutylazine (CAS# 5915-41-3) were also found in the samples. This
compound does have some fish toxicity, but at much higher levels than what was
extrapolated as found. The largest peaks detected in the ion chromatograms, that were
not present in the blanks or that were not common contaminants, e.g. plasticizers; were
unsaturated lipids.
39
-------
-------
SEfW
United States
Environmental Protection
Agency
Office of Research
and Development (8101R)
Washington, DC 20460
Official Business
Penalty for Private Use
$300
EPA/600/R-14/057
April 2014
www.epa.gov
Please make all necessary changes on the below label, detach
or copy and return to the address in the upper left hand corner.
If you do not wish to receive these reports CHECK HERE
detach, or copy this cover, and return to the address in the
upper left hand corner.
PRESORTED STANDARD
POSTAGE & FEES PAID
EPA PERMIT No, G-35
Recycled/Recyclable
Printed with vegetable-based ink on
paper that contains a minimum of
50% post-consumer fiber content
processed chlorine free
------- |