Article

pubs.acs.org/est

Low Molecular Weight Components in an Aquatic Humic Substance

As Characterized by Membrane Dialysis and Orbitrap Mass
Spectrometry
Christina K. Remucal,, Rose M. Cory, Michael Sander, and Kristopher McNeill*,

Institute of Biogeochemistry and Pollutant Dynamics (IBP), ETH Zurich, 8092 Zurich, Switzerland
Environmental Sciences & Engineering, University of North Carolina, Chapel Hill, North Carolina, United States

S Supporting Information
*

ABSTRACT: Suwannee River fulvic acid (SRFA) was

dialyzed through a 100500 molecular weight cuto dialysis
membrane, and the dialysate and retentate were analyzed by
UVvisible absorption and high-resolution Orbitrap mass
spectrometry (MS). A signicant fraction (36% based on
dissolved organic carbon) of SRFA passed through the dialysis
membrane. The fraction of SRFA in the dialysate had a
dierent UVvisible absorption spectrum and was enriched in
low molecular weight molecules with a more aliphatic
composition relative to the initial SRFA solution. Comparison
of the SRFA spectra collected by Orbitrap MS and Fourier
transform ion cyclotron resonance MS (FT-ICR MS)
demonstrated that the mass accuracy of the Orbitrap MS is
sucient for determination of unique molecular formulas of compounds with masses <600 Da in a complex mixture, such as
SRFA. The most intense masses detected by Orbitrap MS were found in the 100200 Da mass range. Many of these low
molecular masses corresponded to molecular formulas of previously identied compounds in organic matter, lignin, and plants,
and the use of the standard addition method provided an upper concentration estimate of selected target compounds in SRFA.
Collectively, these results provide evidence that SRFA contains low molecular weight components that are present individually or
in loosely bound assemblies.

smaller than humic acids14,15 and have reported molecular

weight distributions of 2002000 Da.14,16 For Suwannee River
fulvic acid (SRFA), a reference aquatic humic substance,
average molecular weights of 829, 1150, 1360, 1460, and 1000
1500 have been reported as determined by vapor pressure
osmometry,17 eld-ow fractionation,15 high-pressure sizeexclusion chromatography,18 ultracentrifugation,19 and smallangle X-ray scattering,14 respectively. Thus, the estimates of the
average molecular weight of SRFA span nearly a factor of 2 and
provide little information on the distribution of molecular
weights within aquatic DOM.
The identication of molecular formulas of DOM constituents by high-resolution mass spectrometry has provided
additional evidence for the presence of an abundant low
molecular weight fraction. Numerous types of DOM, including
aquatic reference end-members SRFA,2027 Suwannee River
DOM,28,29 and Pony Lake fulvic acid,27 have been analyzed by
Fourier transform ion cyclotron resonance mass spectrometry

INTRODUCTION
Dissolved organic matter (DOM) is a complex mixture of
naturally occurring organic molecules that plays an important
role in many biogeochemical processes, including the carbon
and nitrogen cycles in aquatic systems, metal complexation and
redox reactions, contaminant fate and transport, and microbial
metabolism.13 A comprehensive understanding of the size,
molecular structure, and composition of DOM has remained
elusive despite the importance of these properties in
determining the dynamics of DOM and the interaction of
DOM with organic and inorganic constituents in water.
DOM was traditionally considered to be composed of
relatively large molecules. Early work using gel ltration and
ultraltration reported fractions of DOM with sizes up to 20
2000 kDa.46 Conversely, more recent studies indicate that
DOM is composed of supramolecular assemblies of relatively
small molecules that are stabilized by the hydrophobic eect
and hydrogen bonds.7,8 For instance, 2024% of riverine DOM
was found to pass through ultralters with a nominal pore size
cuto of 1 kDa,912 suggesting that aquatic DOM contains a
signicant fraction of low molecular weight (LMW) compounds. Fulvic acids, the acid soluble fraction of DOM
comprising the majority of aquatic DOM,13 are considered
2012 American Chemical Society

Received:
Revised:
Accepted:
Published:
9350

June 19, 2012

August 3, 2012
August 6, 2012
August 6, 2012
dx.doi.org/10.1021/es302468q | Environ. Sci. Technol. 2012, 46, 93509359

Environmental Science & Technology

Article

(0.22 m nylon) SRFA solution (2 g/L) in unbuered water

(nal pH 4.6) were added to the inside of the dialysis
membrane tubing, which was clamped on one end with a 70
mm dialysis tubing closure (Sigma Aldrich) and tied in a knot
on the other end. The dialysis tube was submerged in 80 mL of
Nanopure water in a beaker, with the clamp used to hold the
membrane in position. The clamp itself was not in contact with
the solution because control experiments showed that the
clamp released light-absorbing material when it was submerged
in water. The beaker was completely wrapped in aluminum foil
to protect the solutions from light, and the dialysate was gently
stirred at 100 rpm at room temperature for 1 week. The SRFA
dialysis experiment was conducted in duplicate. A control
experiment was performed with riboavin (90 M, MW =
376.36 g/mol) and Rose Bengal (100 M, MW = 973.67 g/
mol) in the dialysis membrane. At selected time points, water
(10 mL) outside the dialysis membrane was removed for
analysis and replaced with an equal volume of fresh Nanopure
water. An aliquot (100 L) was removed from inside the
dialysis membrane and diluted by a factor of 100 for SRFA and
10 for the riboavin/Rose Bengal solution. The UVvisible
absorption spectra were measured on a Varian Cary 100
spectrophotometer and non-purgeable organic carbon was
quantied on a Shimadzu total organic carbon (TOC-L)
analyzer. Selected samples were analyzed by Orbitrap MS as
described below.
Orbitrap MS. Samples from the dialysis experiment (5.6
8.9 mg C/L), as well as a concentrated SRFA stock solution
(26.3 mg C/L in water), were diluted 50:50 (v/v) with
methanol for analysis by a Thermo Exactive mass spectrometer
with an Orbitrap mass analyzer. Samples were infused into an
ESI source with a spray voltage of 3.5 kV at a ow rate of 15
L/min and were analyzed in negative ionization mode. Data
were collected in the mass range of 1001000 m/z and at least
50 scans were averaged together using Xcalibur software. The
Orbitrap MS was externally calibrated daily using a calibration
mixture specied by the manufacturer. Additionally, sodium
dodecyl sulfate ([M H] = 265.1479 Da) and sodium
taurocholate ([M H] = 514.2844 Da) were added to an
aliquot of SRFA as internal standards. Correction of the data
with the internal standards resulted in an average mass error of
0.35 1.8 ppm and a root-mean-square error of 1.5 ppm
calculated for the 1016 compounds presented in Figure 3. A
mass error plot is provided in Figure S1 of the Supporting
Information.
FT-ICR MS. The previously reported SRFA comparison data
set was collected with a custom-built 9.4 T FT-ICR MS at the
National High Magnetic Field Laboratory (Tallahassee, FL).27
The SRFA solution was ionized by negative ESI and analyzed in
the mass range of 290800 m/z.
Data Analysis. The averaged m/z values measured by
Orbitrap MS were extracted from the Xcalibur software,
corrected on the basis of the internal standards, and converted
to nominal masses by adding the mass of H+. Elemental
compositions were assigned by comparing masses measured by
Orbitrap and FT-ICR MS with a list of theoretically possible
masses in Excel (Microsoft 2008), similar to the method
employed by Hertkorn et al.23 Formulas containing only C, H,
and O were considered due to the low amounts of N and S in
SRFA (1.17 and 0.54% by mass, respectively).38 Therefore,
only ions with odd masses (i.e., with an even neutral mass)
were analyzed according to the nitrogen rule in the absence of
nitrogen. Additionally, only m/z values for which the 13C112Cn1

(FT-ICR MS), with <1 ppm error with internal calibration and
a resolving power up to 1 000 000 [full width at half maximum
(FWHM), mass to charge ratio (m/z) 400].30,31 The DOM
constituents are typically ionized using negative mode electrospray ionization (ESI),2022,24,26,27,32 but positive mode
ESI,2023,28 atmospheric-pressure chemical ionization, and
atmospheric-pressure photoionization also have been used.23
The results from the FT-ICR studies revealed that DOM is
comprised of thousands of unique molecular formulas, with the
most intense ions at m/z values <800.22,32,33
The mass resolution capabilities of FT-ICR MS have made it
the tool of choice for the analysis of complex samples, such as
DOM. While not a match for the extreme mass resolution of
FT-ICR MS, other types of mass spectrometers are capable of
determining accurate masses.31 The Orbitrap mass analyzer,
which can be used as a standalone mass spectrometer or in
combination with other mass selection technologies (e.g., ion
trap), features <2 ppm error using internal calibration and a
resolving power of 100 000 (FWHM, 200 m/z).30,31 Although
the mass accuracy provided by Orbitrap MS at 500 Da is better
than the 0.1 mDa accuracy required for determination of all C-,
H-, O-, and S-containing compounds,34 Orbitrap MS can be
used for unique formula determination of LMW C-, H-, and Ocontaining constituents in low S-content DOM. To date,
Orbitrap instruments have been used to determine molecular
composition of ambient organic aerosols,35,36 to identify target
compounds in soil matrices,30 and to compare chemically
fractionated humics isolated from volcanic soil.37
The objective of this work was to characterize the low
molecular weight fraction of SRFA, a well-characterized
reference aquatic fulvic acid. To this end, we combined dialysis
of SRFA through a 100500 Da molecular weight cuto
(MWCO) membrane with Orbitrap high-resolution mass
spectrometry analysis of the nondialyzed SRFA, the dialysate,
and the retentate. Furthermore, we compared the negative
mode ESI mass spectrum of SRFA collected by Orbitrap to that
collected by FT-ICR MS. We conducted additional detailed
analyses on the mass spectrum measured by Orbitrap MS in the
100290 m/z range because of the high accuracy and precision
of the Orbitrap MS in this lower mass region. These results
demonstrate the applicability of Orbitrap MS to analyze the
LMW fraction of DOM and to monitor changes in this fraction
in response to environmental perturbations and provide further
evidence for a signicant fraction of LMW constituents in
SRFA.

METHODS
Materials. Suwannee River fulvic acid (2S101F) standard, a
reference aquatic humic substance, was purchased from the
International Humic Substances Society (St. Paul, MN) and
was used as received. SRFA was chosen because it is low in
heteroatom content (e.g., N, S, P)23 and because its spectrum
as collected by Orbitrap MS can be compared to previous
analysis by FT-ICR MS.2023,27 Benzoic acid, 6,7-dihydroxycoumarin (esculetin), umbelliferone, riboavin, and Rose
Bengal were purchased from Sigma Aldrich and used as
received. All solutions were prepared using 18 M cm water
(Barnstead Nanopure Diamond Water Purication System).
Dialysis. Dialysis of SRFA was performed using a 100500
Da MWCO Spectra/Por cellulose ester membrane (Spectrum
Laboratories). The dialysis membrane was stored in sodium
azide to prevent bacterial growth and was thoroughly washed
with Nanopure water prior to use. Aliquots (4 mL) of a ltered
9351

dx.doi.org/10.1021/es302468q | Environ. Sci. Technol. 2012, 46, 93509359

Environmental Science & Technology

Article

Figure 1. (a) Non-purgeable organic carbon measured in SRFA solution in the retentate (diluted 1:100) and in the dialysate. (b) UVvis
absorbance and (inset) area-normalized absorbance spectra of the same solutions. van Krevelen diagrams for compounds with 100600 m/z for (c)
the initial SRFA solution, (d) the retentate, (e) the dialysate, and (f) compounds with dierent relative magnitudes in the initial SRFA solution and
the retentate after 7 days of dialysis. Compounds that increased or decreased in relative magnitude in the retentate after dialysis, as well as
compounds that were absent after dialysis, are shown. Only compounds with magnitudes >100 in the mass spectra are shown, corresponding to 45
50% of identied peaks in this experiment due to dilute samples.

the masses detected by the FT-ICR MS in the 290600 m/z

range with an average error of 0.08 0.24 ppm.
Concentrations of LMW Compounds in SRFA. The
upper limits on the concentrations of selected compounds in
SRFA were determined using the standard addition method.

isotopologue was detected with <5 ppm accuracy at a

separation of 1.0034 Da, indicating a singly charged ion, were
considered. The elemental composition assignment method
was veried using the previously analyzed SRFA data set
collected by FT-ICR MS, resulting in assignment of 100% of
9352

dx.doi.org/10.1021/es302468q | Environ. Sci. Technol. 2012, 46, 93509359

Environmental Science & Technology

Article

Figure 2. (a) Orbitrap ESI negative mass spectrum of SRFA in the 100600 m/z range and SRFA m/z relative magnitude distributions of identied
ions measured by (b) Orbitrap and (c) FT-ICR MS. The relative magnitude cutos of 10.2 and 8.2 for Orbitrap and FT-ICR, respectively,
correspond to 65% of the identied peaks and are indicated by dashed lines. The relative magnitude cuto corresponds to an absolute magnitude
cuto of 100 for the Orbitrap spectra.

Known concentrations (01 M) of benzoic acid, esculetin,

and umbelliferone were added to SRFA (nal concentration =
50 mg/L) and analyzed by Orbitrap MS as described above.
The magnitudes of the m/z values of the spiked compounds
were normalized to the magnitudes of 812 nearby peaks and
plotted versus the added compound concentration to estimate
the maximum concentration of each compound in SRFA.

carbon added to the dialysis tube, 2.24 0.05 and 1.25 0.02
mg of carbon were recovered in the retentate and dialysate
solutions, respectively.
In a control experiment, a solution containing riboavin
(MW = 376.36 g/mol) and Rose Bengal (MW = 973.67 g/mol;
smallest cross-sectional dimension = 14 )39 was dialyzed
through a separate dialysis membrane against Nanopure water.
The concentration of riboavin in the dialysate increased to 3.0
M (268 = 31 600 cm1 M1)40 after 7 days of dialysis,
corresponding to 91% of the expected concentration at
partition equilibrium (Supporting Information, Figure S2).
Conversely, Rose Bengal was conserved in the retentate and
was not detected in the dialysate, as expected based on its MW,
which is larger than the 500 Da MWCO of the dialysis
membrane. It is possible that molecules in SRFA with
molecular masses >500 Da diused through the membrane
provided that they had suciently small cross-sectional areas

RESULTS AND DISCUSSION

Dialysis Separation of LMW Components from SRFA.
A SRFA solution was dialyzed through a 100500 MWCO
dialysis membrane against Nanopure water for 7 days. Over the
course of the dialysis, the non-purgeable organic carbon
(NPOC) decreased in the retentate and increased in the
dialysate (Figure 1a). Using the measured NPOC values and
the known solution volumes, it was possible to close the carbon
mass balance (99.7%). Of the initial mass of 3.50 0.09 mg of
9353

dx.doi.org/10.1021/es302468q | Environ. Sci. Technol. 2012, 46, 93509359

Environmental Science & Technology

Article

after dialysis, suggesting an increase in aromaticity. The

dialysate had a lower SUVA254 value of 2.36, further indicating
that this fraction was depleted in aromatic components relative
to the initial SRFA solution.
The UVvis absorbance data clearly indicate that the SRFA
components that passed through the dialysis membrane
generally had lower absorbance at all wavelengths and
particularly weakened absorbance at longer wavelengths
compared with the initial SRFA solution. The absorbance
indices all point to the same conclusion that this LMW material
was depleted in aromatic and other unsaturated chromophores.
This conclusion is reinforced by high-resolution MS analysis of
the solutions, as discussed below.
High-Resolution Mass Spectra of SRFA. Prior to the
analysis of the retentate and dialysate of SRFA, the mass
spectrum of nondialyzed SRFA was collected on an Orbitrap
instrument to validate the analysis method. The mass spectrum
of SRFA showed a pattern of intense clusters of ions at odd
m/z values and less intense groups at even m/z values (Figure
2a). This pattern is attributable to ions containing only C, H,
and O, which are the dominant atoms in SRFA.22,23,26 The less
intense clusters of ions observed at even m/z values are due to
the presence of heteroatoms (e.g., N and P) and 13C112Cn1
isotopologues from singly charged ions.21 The SRFA mass
spectrum has a repeating pattern with a period of 14.0156 mass
units, corresponding to the addition of CH2 groups.22,23,25 The
same features have previously been reported for DOM spectra
collected by FT-ICR MS.2123,25,26 The Orbitrap-detected mass
spectrum of SRFA showed the highest intensities at m/z values
<200 and intensities decreased with increasing m/z values
(Figure 2a). The overall shape of the mass spectrum in the
Orbitrap was similar regardless of the instrument settings used
(i.e., changing voltages on the ESI source or in the ion optics;
data not shown). Individually identiable ions are observed at
m/z values up to 700 with resolving powers >50 000 in the
Orbitrap MS (data not shown). However, a conservative
nominal mass cuto of 600 Da was used for molecular formula
identication. The mass accuracy of the Orbitrap MS is
suciently high to derive unique chemical formulas for the C-,
H-, and O-containing peaks detected in the negative ESI SRFA
spectrum in this mass range.
The chemical formulas of ions identied by Orbitrap analysis
of the SRFA prior to dialysis and of the retentate and dialysate
after 7 d of dialysis are presented in the form of van Krevelen
diagrams (i.e., the ratio of H:C plotted versus the ratio of O:C;
Figure 1ce). The mass spectra of all three solutions showed
clusters of intense ions at odd m/z values and less intense ions
at even m/z values (Supporting Information, Figure S4),
consistent with the spectrum presented in Figure 2a. Note that
the mass spectra of the SRFA prior to dialysis and the retentate
were collected from dilute solutions to match the intensity of
ions in the mass spectrum of the dialysate and that all formulas
identied in the dialysate were also identied in the initial
SRFA solution (formulas not presented in Figure 1c were
below the magnitude cuto). The spectrum obtained from
analyzing a concentrated SRFA solution is discussed in detail
below.
The van Krevelen plots clearly show that the chemical
formulas of the compounds in the dialysate diered from those
in the retentate (Figure 1ce), conrming that a distinct
fraction of the SRFA starting material was separated by dialysis.
While the retentate showed intense ions in regions of the van
Krevelen diagram where aromatic and lignin compounds are

(e.g., long aliphatic molecules). However, the results from the

control experiment strongly support that the molecules in the
dialysate of SRFA were low in molecular weight (<500 Da).
The relatively fast mass transfer kinetics of riboavin through
the membrane (Supporting Information, Figure S2) suggests
that similarly sized molecules in SRFA had likely approached or
reached partition equilibrium between the retentate and
dialysate after 1 week of dialysis.
The nding of a considerable fraction of SRFA in the
dialysate is consistent with previous studies. For example, up to
10% of DOC from a wastewater treatment euent passed
through a 100500 Da MWCO dialysis membrane,41 20% of
total organic carbon from Suwannee River humic acid passed
through a nanoltration membrane with a 3001000 Da
cuto,42 up to 80% of organic carbon from Aldrich fulvic acid
passed through a 1000 Da MWCO lter,43 and 4345% of
DOC from freshwater DOM passed through a 500 Da MWCO
dialysis membrane in an electrodialysis system.44 We therefore
expect that the observation of LMW fractions in SRFA reported
here also applies to other aquatic humic substances.
UVVis Characterization of the SRFA Retentate and
Dialysate. The UVvis absorbance spectra of the retentate
and dialysate provide evidence that a distinct subset of SRFA
was separated during dialysis. The UVvis absorbance of the
retentate decreased, while the absorbance of the dialysate
increased during dialysis (Figure 1b). Determining the ratio of
absorbance at 300 and 400 nm (A300/A400) demonstrates that
the UVvis spectrum of the initial SRFA solution is notably
dierent from the spectra of the retentate and dialysate after 1
week of dialysis. A300/A400 of the SRFA retentate decreased
from 5.36 to 4.86 after dialysis due to preferential removal of
low molecular mass material with weaker long wavelength
absorbance. This nding is in agreement with trends in
absorbance of Suwannee River humic substances observed after
dialysis through a 5 kDa membrane45 and observations that
fractions of DOM retained by a 10 kDa ultralter are
responsible for water color.46 A300/A400 of the receiving
dialysate was 6.23 and 6.31 after 3 and 7 days of dialysis,
respectively, due to enrichment of compounds with higher UV
and lower visible absorbance relative to the initial SRFA
solution. This is also evident from the area-normalized
absorbance measurements (inset Figure 1b), which show that
the retentate had higher relative absorbance and the dialysate
displayed lower relative absorbance compared to the initial
SRFA solution at wavelengths >250 nm. The opposite is
observed in the 200250 nm wavelength range, with the
dialysate showing higher relative absorption compared to the
retentate. After correcting for solution volumes, summing the
spectra of the retentate and dialysate produces a spectrum that
is identical to the starting SRFA solution (Supporting
Information, Figure S3), demonstrating that the light-absorbing
components of SRFA were conserved.
The NPOC-normalized spectra of the dialysate has a much
weaker specic absorbance than the retentate (Supporting
Information, Figure S3), suggesting that the dialysate is
depleted in UVvis light absorbing moieties relative to the
retentate. For example, the molar absorptivity at 280 nm [L
(mol OC)1 cm1] increased from 420 to 480 in the retentate
and only reached 194 in the dialysate after 7 days, in agreement
with previous observations that lower molecular weight
fractions of fulvic acids have lower molar absorptivities.18
Additionally, the specic UV absorbance at 254 nm [SUVA254;
L (mg C)1 m1]47 of the retentate increased from 4.68 to 5.20
9354

dx.doi.org/10.1021/es302468q | Environ. Sci. Technol. 2012, 46, 93509359

Environmental Science & Technology

Article

of SRFA separated by size-exclusion chromatography (SEC)

and analyzed by a triple quadrupole mass spectrometer.52 In
contrast to the trends observed by Orbitrap MS, the relative
abundance of ions detected by FT-ICR MS is low at both low
and high masses (Figure 2c), with a maximum ion abundance
detected around a center mass.2023 The apparent center mass
in negative mode ESI FT-ICR mass spectra of SRFA varies
between studies from approximately 35026 to 50022,23 to
700.20,21 The center mass observed by positive mode ESI varies
from approximately 40023 to 1700.20
There are several possible explanations for the dierent ion
abundance patterns of the SRFA mass spectra observed by
Orbitrap and FT-ICR MS using the same ionization method
(i.e., negative ESI). Experimental parameters and dierences in
ionization sources can change observed mass distributions,53,54
and dierences in the way ions are introduced to the mass
spectrometers (i.e., tuning, source design, ion optics voltages)
could aect the observed mass distributions. Although ESI is
considered a gentle ionization source,55 it is possible that
fragmentation of higher molecular weight species produces
LMW peaks21,28 through disaggregation of noncovalently or
weakly covalently bound molecules25,33 or through in-source
collision-induced dissociation (CID) processes due to energetic
collisions between ions and residual gas molecules.5456 The
possibility of fragmentation as the source of some LMW
compounds is supported by the observation that the high
molecular weight (HMW) fraction of SRFA, as separated by
SEC and analyzed by QTOF52 or FT-ICR MS,25 consists of
similar LMW subunits also found in the LMW fraction.
Furthermore, molecules with high masses may not be
transmitted as eectively through the ESI source, resulting in
higher intensity ions in lower mass ranges.25,56 Despite the
possibility of fragmentation or disaggregation of larger ions, the
predicted compositions and structures are likely to be present
in DOM, either on their own or as a part of larger molecules.
There is good agreement between the molecular formulas
identied in SRFA as detected by both mass spectrometers
despite the dierences in the trends of ion abundance with m/z.
The previously published FT-ICR MS data set27 contains
masses in the 290800 range. Therefore, only masses between
290 and 600 Da, the conservative upper limit used for Orbitrap
MS data analysis, were compared. Of the 1446 unique formulas
identied in the data sets from either or both instruments in
this mass range, 87 (6.0%) were only detected by the Orbitrap
MS, 852 (58.9%) were only detected by the FT-ICR MS, and
507 (35.1%) were detected by both instruments (Figure 3a). As
evident by the mass distribution of identied formulas
(Supporting Information, Figure S6), most of the formulas
identied only in the FT-ICR MS data set have masses >400
Da. In fact, many of these masses (57%) were detected by the
Orbitrap MS but were excluded from analysis because the
13
C112Cn1 isotopologue was not detected with sucient
condence. The formulas only identied by FT-ICR MS
overlap with those identied in both data sets, with slightly
higher O:C in some cases (Figure 3a), in agreement with
previous observations that HMW fractions of SRFA have
higher O content.25 Note that general shape of the van
Krevelen diagram from the FT-ICR MS data set for 600800
m/z is the same as the 290600 m/z range (Supporting
Information, Figure S7), suggesting that trends in the class
composition of SRFA in the 290600 m/z range may be
applicable to higher masses as well. The compounds that were
unique to the Orbitrap data set generally have masses <400 Da

found (H:C < 1), the dialysate showed ions with higher H:C
and lower O:C ratios, indicating a more aliphatic composition.
The changes in the van Krevelen diagrams must be carefully
interpreted since the abundance of an ion in a single mass
spectrum is not quantitative. The magnitudes of ions produced
by ESI depend on the ionization eciency of their functional
groups (e.g., carboxylic acids, alcohols, amines)29 and also may
depend on the eciency of the detector to detect the ions; they
do not necessarily correlate with the abundance of those
compounds in a mixture such as SRFA.33 Nevertheless, it is
possible to compare relative magnitudes of the same ion before
and after the dialysis to reveal trends.48 Changes in the relative
magnitude (M) of ions were calculated by comparing their
magnitudes in the retentate (t7), normalized to the sum of
magnitudes of all common ion masses in the retentate, to their
normalized magnitudes in the dilute SRFA solution (t0) before
it was dialyzed:
M t 7 / M t 7
all
m/z

log M t 7 /M t 0 = log t 0
t0

M
/
M
m/z
all

(1)

According to eq 1, ions in the retentate after dialysis with

negative and positive log Mt7/Mt0 values represent compounds
with decreasing and increasing abundance, respectively. Dialysis
resulted in an increasing relative abundance of compounds with
aromatic/lignin signatures in the retentate (Figure 1f).
Numerous ions with more saturated formulas (i.e., high H:C
ratios) that were detected in the dilute SRFA solution prior to
dialysis were not detected in the retentate with sucient
resolution of the 13C112Cn1 isotopologue (Figure 1f), but were
found in the dialysate (Figure 1e). The dialysate therefore
contained compounds that were less aromatic and more
saturated than compounds in the retentate, consistent with the
results of the UVvis absorbance analysis. Our results are also
in agreement with previous observations that the amount of
aromatic moieties in aquatic humic substances increases with
increasing molecular weight.18
The relative mass distribution conrms that the retentate
solution had fewer compounds in the 300500 mass range
after dialysis and was slightly enriched in the higher masses
relative to the starting material (Supporting Information, Figure
S5). While some of the low molecular weight components in
the mass spectra of the dialysate may have resulted from
fragmentation of compounds with larger masses, as discussed
below, the predominance of low molecular weight ions in the
dialysate (<500 Da) strongly suggests that the dialysate was
primarily composed of LMW SRFA constituents. These results
imply that a signicant fraction of SRFA, and likely other
aquatic HS, is comprised of LMW components, in agreement
with the current view that humic substances are supramolecular
associations of relatively small molecules that are stabilized by
the hydrophobic eect or other noncovalent binding
mechanisms.7,8,4951
Elemental Composition of SRFA As Detected by
Orbitrap MS and FT-ICR MS. We further assessed the
capabilities of Orbitrap MS to characterize the LMW fraction of
DOM by comparing the mass spectrum of SRFA obtained by
Orbitrap MS to the previously collected spectrum of SRFA
obtained by FT-ICR MS.27 The magnitude of ions in the
Orbitrap MS spectrum of the concentrated SRFA (50 mg/L)
solution was highest in the 100200 m/z range and decreased
with increasing m/z (Figure 2a,b). The mass spectrum
observed here is similar to the spectrum of the LMW fraction
9355

dx.doi.org/10.1021/es302468q | Environ. Sci. Technol. 2012, 46, 93509359

Environmental Science & Technology

Article

(Supporting Information, Figure S6) and are found in the same

section of the van Krevelen diagram as condensed aromatics
and black carbon.2,23,32,5759
It is important to note that the previous FT-ICR MS studies
of SRFA generally did not consider masses below 200
Da.22,23,26 An advantage of the Orbitrap MS is the assessment
of ions in this low mass range. The molecular formulas
identied in the 100200 m/z range occupied a larger space in
the van Krevelen diagram than the ions in the 200290 m/z
range (Figure 3b,c). A similar wide pattern in the van Krevelen
diagram was previously reported for SRFA ions in the 150344
m/z range obtained by SEC-TOF-MS.48
This study demonstrates that Orbitrap and FT-ICR MS
produce complementary information about the chemical
composition of SRFA. For the 290600 mass range, the
agreement between the two methods shows that Orbitrap MS
can be used to determine the molecular composition of
components in complex mixtures, such as SRFA. Additionally,
the high resolving power of the Orbitrap MS in the 100290
mass range and the high magnitudes of these ions in the
detected mass spectra demonstrate that the Orbitrap MS is
well-suited for analyzing the LMW fraction of SRFA and likely
other aquatic humic substances with low N and S contents.
Compound Identication and Quantication. The
complexity of DOM and the presence of multiple compounds
at the same nominal mass limit the application of MSMS for
structural elucidation,2 although insights can be gained by
techniques such as single ion isolation using on-resonance
CID.26 Eorts were made to identify some of the additional
prominent compounds observed in the 100200 mass range,
such as those containing only hydrogen and carbon (Figure
3b). Even with these low masses and only two possible
elements, H and C could be combined in numerous potential
structures, illustrating the diculty with compound identication from mass alone. For example, the neutral mass
116.0623, corresponding to the formula C9H8, matches 41
possible structures in the ChemSpider database,60 of which
indene may be considered the most likely based on its presence
in environmental systems61 and its ability to ionize to form a
relatively stable negative ion.62 By contrast, the mass 120.0935
(C9H12) matches >250 possible structures, with no obvious
naturally occurring compound presenting itself as the clear
assignment.
Nonetheless, it is possible to tentatively identify some
compounds by comparing the unique molecular formulas
observed in SRFA with lists of compounds previously identied
in lignin, DOM, plants, and black carbon (Supporting
Information, Figure S8 and Tables S1S3). Lignin, for
example, is thought to be a major constituent of SRFA.22,63
Although this identication approach does not provide direct
evidence for the presence of these compounds, the
combination of accurate mass determination with knowledge
of the origin of DOM (i.e., derived from plants) does enable
the identication of likely structures. While only one likely
isomer was found for some masses (Supporting Information,
Table S1), multiple potential isomers could be identied for
others (Supporting Information, Table S2), and it is possible
that the masses assigned to a unique structure here could
correspond to other isomers not included in this screening.
Many masses corresponded to those previously reported to
belong to class types such as lignin or carboxyl-rich alicylic
molecules (CRAM),64,65 and no structures were identied
(Supporting Information, Table S3). The tentatively identied

Figure 3. van Krevelen diagram for (a) compounds identied by

negative-ion ESI showing compounds unique to FT-ICR and Orbitrap
MS for 290600 m/z, as well as compounds identied by both
methods. van Krevelen diagrams for compounds identied by
negative-ion ESI by Orbitrap MS for (b) 100200 and (c) 200
290 m/z with magnitudes >100.
9356

dx.doi.org/10.1021/es302468q | Environ. Sci. Technol. 2012, 46, 93509359

Environmental Science & Technology

Article

this approach also may be used for other complex DOM

mixtures with multiple isobaric ions and low N and S contents.
We demonstrated that the low molecular weight fraction in
SRFA can be separated by dialysis over a 100500 Da MWCO
membrane and that the retentate and dialysate have dierent
physicochemical properties, as evident from both spectrophotometric and Orbitrap MS analyses. We anticipate that the
combination of membrane dialysis and high-resolution MS
analysis will allow selectively assessing the dynamics of LMW
fractions of DOM in natural and engineered systems, such as its
availability to microorganisms, its susceptibility to photochemical bleaching, and its reactivity with chemical disinfectants.

compounds span most of the area covered by the van Krevelen

diagram of SRFA, with many of the uniquely identied
compounds found in the 100200 mass range (Supporting
Information, Figure S8).
Three of the tentatively identied compounds were added to
SRFA in increasing amounts according to the standard addition
method. The resulting increase in the relative magnitude of
each mass peak was then used to estimate the maximum
concentration of the compound that could have been originally
present in SRFA (Supporting Information, Figure S9 and Table
S4), assuming that no isomeric compounds contributed to the
original MS peak intensity. Benzoic acid, esculetin, and
umbelliferone were selected because they have low masses,
are known plant-derived compounds,66,67 and their masses
were detected in SRFA with high intensities (Supporting
Information, Figure S8 and Table S1). The calculated
concentrations represent the maximum possible concentration
of the standard compound and do not account for other
isomers. For example, in the case of umbelliferone, it is possible
that other hydroxycoumarins are present in SRFA and would
ionize with comparable eciency. Benzoic acid had a low
response factor, and as a result, the error on the maximum
benzoic acid concentration is high. Neglecting isomers at the
same exact mass, the selected compounds could comprise a
high percentage (ca. 1%) of SRFA by mass (Supporting
Information, Table S4). The linear response of the Orbitrap
MS with increasing model concentration demonstrates that this
method could be used for further quantication of target
compounds in NOM.
Environmental and Technical Implications. The view of
DOM has evolved from it being composed of high molecular
weight polymeric structures to supramolecular associations of
relatively small organic molecules.7,8,50 Fulvic acids are
considered smaller than humic acid14,15 and the average
reported molecular weight of SRFA ranges from 800 to 1500
Da.14,15,1719 Our results show that 36% of the carbon mass of
SRFA passed through a 100500 MWCO dialysis membrane,
demonstrating that a signicant fraction of SRFA falls within
the low (i.e., <500 Da) mass range. The molecular formulas of
some of these compounds corresponded to the formulas of
compounds previously identied in lignin, DOM, plants, and
black carbon. Overall, our results support the view of humic
substances as supramolecular associations.
The presence of a signicant LMW fraction in fulvic acids
that is more aliphatic than the bulk material has implications for
the bioavailability of DOM in aquatic systems. According to the
size-reactivity continuum model, the HMW fraction of DOM is
more bioavailable and leads to the formation of refractory
LMW compounds.9,46 Analysis of freshwater DOM before and
after biodegradation revealed a shift toward lower masses and
removal of compounds with higher O:C ratios.68 Similarly,
DOM from Chesapeake Bay became more H-rich and O-poor
(i.e., more aliphatic) as it was transported to the ocean.59 These
results suggest that the fraction of SRFA with higher masses
and higher O:C ratios retained by the 100500 MWCO
dialysis membrane in this study would be more preferentially
degraded by bacteria than the LMW fraction passing through
the dialysis membrane.
The focus of this work was on the analysis of SRFA by
Orbitrap mass spectrometry. We showed that the resolving
power of the Orbitrap MS allowed the identication of unique
molecular formulas of C-, H-, and O-containing ions in SRFA
produced by negative ESI with masses <600 Da, suggesting that

ASSOCIATED CONTENT

S Supporting Information
*

Additional absorbance spectra and MS data from dialysis

experiments, magnitude distributions and additional van
Krevelen diagrams of identied compounds in SRFA, standard
addition plots for target compounds, and tables of tentatively
identied compounds in SRFA. This material is available free of
charge via the Internet at http://pubs.acs.org.

AUTHOR INFORMATION

Corresponding Author

*E-mail: kris.mcneill@env.ethz.ch; telephone: 0041-(0)44 632

4755; fax: 0041-(0)44 632 1438.
Present Address

Environmental Chemistry and Technology Program, University of WisconsinMadison, Madison, Wisconsin.

Notes

The authors declare no competing nancial interest.

ACKNOWLEDGMENTS
This research was supported by an ETH Postdoctoral
Fellowship Award. We thank Prof. Alan Marshall (Florida
State University) for helpful comments.

REFERENCES

(1) Thomas, J. The role of dissolved organic matter, particularly free

amino acids and humic substances, in freshwater ecosystems.
Freshwater Biol. 1997, 38 (1), 136.
(2) Mopper, K.; Stubbins, A.; Ritchie, J. D.; Bialk, H. M.; Hatcher, P.
G. Advanced instrumental approaches for characterization of marine
dissolved organic matter: Extraction techniques, mass spectrometry,
and nuclear magnetic resonance spectroscopy. Chem. Rev. 2007, 107
(2), 419442.
(3) Stenson, A. C. Fourier transform ion cyclotron resonance mass
spectral characterization of metalhumic binding. Rapid Commun.
Mass Spectrom. 2009, 23 (4), 465476.
(4) Rashid, M.; King, L. Molecular weight distribution measurements
on humic and fulvic acid fractions from marine clays on the Scotian
Shelf. Geochim. Cosmochim. Ac. 1969, 33 (1), 147151.
(5) Gjessing, E. T. Ultrafiltration of aquatic humus. Environ. Sci.
Technol. 1970, 4 (5), 437438.
(6) Ceccanti, B.; Calcinai, M.; Bonmati-Pont, M.; Ciardi, C.;
Tarsitano, R. Molecular size distribution of soil humic substances
with ionic strength. Sci. Total Environ. 1989, 81/82, 471479.
(7) Piccolo, A. The supramolecular structure of humic substances.
Soil Sci. 2001, 166 (11), 810.
(8) Sutton, R.; Sposito, G. Molecular structure in soil humic
substances: The new view. Environ. Sci. Technol. 2005, 39 (23), 9009
9015.
9357

dx.doi.org/10.1021/es302468q | Environ. Sci. Technol. 2012, 46, 93509359

Environmental Science & Technology

Article

biochemical oxidation of dissolved organic matter. Environ. Sci.

Technol. 2010, 44 (10), 36833689.
(28) Kujawinski, E.; Hatcher, P.; Freitas, M. High-resolution Fourier
transform ion cyclotron resonance mass spectrometry of humic and
fulvic acids: Improvements and comparisons. Anal. Chem. 2002, 74
(2), 413419.
(29) Kujawinski, E.; Freitas, M.; Zang, X.; Hatcher, P.; GreenChurch, K.; Jones, R. The application of electrospray ionization mass
spectrometry (ESI MS) to the structural characterization of natural
organic matter. Org. Geochem. 2002, 33 (3), 171180.
(30) Krauss, M.; Singer, H.; Hollender, J. LChigh resolution MS in
environmental analysis: From target screening to the identification of
unknowns. Anal. Bioanal. Chem. 2010, 397 (3), 943951.
(31) Cortes-Francisco, N.; Flores, C.; Moyano, E.; Caixach, J.
Accurate mass measurements and ultrahigh-resolution: Evaluation of
different mass spectrometers for daily routine analysis of small
molecules in negative electrospray ionization mode. Anal. Bioanal.
Chem. 2011, 112.
(32) Kim, S.; Kramer, R.; Hatcher, P. Graphical method for analysis
of ultrahigh-resolution broadband mass spectra of natural organic
matter, the van Krevelen diagram. Anal. Chem. 2003, 75 (20), 5336
5344.
(33) Reemtsma, T. Determination of molecular formulas of natural
organic matter molecules by (ultra-) high-resolution mass spectrometry. J. Chromatogr. A 2009, 1216 (18), 36873701.
(34) Kim, S.; Rodgers, R. P.; Marshall, A. G. Truly exact mass:
Elemental composition can be determined uniquely from molecular
mass measurement at <0.1 mDa accuracy for molecules up to <500
Da. Int. J. Mass Spectrom. 2006, 251 (23), 260265.
(35) Smith, J. S.; Laskin, A.; Laskin, J. Molecular characterization of
biomass burning aerosols using high-resolution mass spectrometry.
Anal. Chem. 2009, 81 (4), 15121521.
(36) Nguyen, T. B.; Laskin, J.; Laskin, A.; Nikorodov, S. A. Nitrogencontaining organic compounds and oligomers in secondary organic
aerosol formed by photooxidation of isoprene. Environ. Sci. Technol.
2011, 45 (16), 69086918.
(37) Nebbioso, A.; Piccolo, A. Basis of a humeomics science:
Chemical fractionation and molecular characterization of humic
biosuprastructures. Biomacromolecules 2011, 12 (4), 11871199.
(38) International Humic Substances Society; http://www.
humicsubstances.org/elements.html.
(39) Srivastava, A.; Doraiswamy, S. Rotational diffusion of rose
bengal. J. Chem. Phys. 1995, 103 (14), 6197.
(40) Koziol, J. Studies on flavins in organic solventsI. Spectral
characteristics of riboflavin, riboflavin tetrabutyrate and lumichrome.
Photochem. Photobiol. 1966, 5 (1), 4154.
(41) Lee, W.; Westerhoff, P. Dissolved organic nitrogen measurement using dialysis pretreatment. Environ. Sci. Technol. 2005, 39 (3),
879884.
(42) Hong, S.; Elimelech, M. Chemical and physical aspects of
natural organic matter (NOM) fouling of nanofiltration membranes. J.
Membr. Sci. 1997, 132 (2), 159181.
(43) Olson, T. M.; Barbier, P. F. Oxidation kinetics of natural organic
matter by sonolysis and ozone. Water Res. 1994, 28 (6), 13831391.
(44) Chen, H.; Stubbins, A.; Hatcher, P. G. A mini-electrodialysis
system for desalting small volume saline samples for Fourier transform
ion cyclotron resonance mass spectrometry. Limnol. Oceanogr. Methods
2011, 9, 582592.
(45) Trubetskoj, O.; Trubetskaya, O.; Richard, C. Photochemical
activity and fluorescence of electrophoretic fractions of aquatic humic
matter. Water Res. 2009, 36 (5), 518524.
(46) Tranvik, L. J. Bacterioplankton growth on fractions of dissolved
organic carbon of different molecular weights from humic and clear
waters. Appl. Environ. Microb. 1990, 56 (6), 16721677.
(47) Weishaar, J. L.; Aiken, G. R.; Bergamaschi, B. A.; Fram, M. S.;
Fuji, R.; Mopper, K. Evaluation of specific ultraviolet absorbance as an
indicator of the chemical composition and reactivity of dissolved
organic carbon. Environ. Sci. Technol. 2003, 37 (20), 47024708.

(9) Amon, R. M. W.; Benner, R. Bacterial utilization of different size

classes of dissolved organic matter. Limnol. Oceanogr. 1996, 41 (1),
4151.
(10) Meyer, J. L.; Edwards, R. T.; Risley, R. Bacterial growth on
dissolved organic carbon from a blackwater river. Microbial Ecol. 1987,
13 (1), 1329.
(11) Benner, R.; Hedges, J. I. A test of the accuracy of freshwater
DOC measurements by high-temperature catalytic oxidation and UVpromoted persulfate oxidation. Mar. Chem. 1993, 41 (13), 161165.
(12) Hedges, J. I.; Cowie, G. L.; Richey, J. E.; Quay, P. D.; Benner,
R.; Strom, M.; Forsberg, B. R. Origins and processing of organic
matter in the Amazon River as indicated by carbohydrates and amino
acids. Limnol. Oceanogr. 1994, 39 (4), 743761.
(13) Aiken, G. R.; McKnight, D. M.; Thorn, K.; Thurman, E.
Isolation of hydrophilic organic acids from water using nonionic
macroporous resins. Org. Geochem. 1992, 18 (4), 567573.
(14) Thurman, E. M.; Wershaw, R.; Malcolm, R.; Pinckney, D.
Molecular size of aquatic humic substances. Org. Geochem. 1982, 4 (1),
2735.
(15) Beckett, R.; Jue, Z.; Giddings, J. C. Determination of molecular
weight distributions of fulvic and humic acids using flow field-flow
fractionation. Environ. Sci. Technol. 1987, 21 (3), 289295.
(16) Leenheer, J.; Rostad, C.; Gates, P.; Furlong, E.; Ferrer, I.
Molecular resolution and fragmentation of fulvic acid by electrospray
ionization/multistage tandem mass spectrometry. Anal. Chem. 2001,
73 (7), 14611471.
(17) Aiken, G. R.; Malcolm, R. L. Molecular weight of aquatic fulvic
acids by vapor pressure osmometry. Geochim. Cosmochim. Acta 1987,
51 (8), 21772184.
(18) Chin, Y. P.; Aiken, G.; OLoughlin, E. Molecular weight,
polydispersity, and spectroscopic properties of aquatic humic
substances. Environ. Sci. Technol. 1994, 28 (11), 18531858.
(19) Reid, P. M.; Wilkinson, A. E.; Tipping, E.; Jones, M. N.
Determination of molecular weights of humic substances by analytical
(UV scanning) ultracentrifugation. Geochim. Cosmochim. Acta 1990, 54
(1), 131138.
(20) Brown, T.; Rice, J. Effect of experimental parameters on the ESI
FT-ICR mass spectrum of fulvic acid. Anal. Chem. 2000, 72 (2), 384
390.
(21) Stenson, A.; Landing, W.; Marshall, A.; Cooper, W. Ionization
and fragmentation of humic substances in electrospray ionization
Fourier transform-ion cyclotron resonance mass spectrometry. Anal.
Chem. 2002, 74 (17), 43974409.
(22) Stenson, A.; Marshall, A.; Cooper, W. Exact masses and
chemical formulas of individual Suwannee River fulvic acids from
ultrahigh resolution electrospray ionization Fourier transform ion
cyclotron resonance mass spectra. Anal. Chem. 2003, 75 (6), 1275
1284.
(23) Hertkorn, N.; Frommberger, M.; Witt, M.; Koch, B. P.; SchmittKopplin, P.; Perdue, E. M. Natural organic matter and the event
horizon of mass spectrometry. Anal. Chem. 2008, 80 (23), 89088919.
(24) Kujawinski, E. B.; Del Vecchio, R.; Blough, N. V.; Klein, G. C.;
Marshall, A. G. Probing molecular-level transformations of dissolved
organic matter: Insights on photochemical degradation and protozoan
modification of DOM from electrospray ionization Fourier transform
ion cyclotron resonance mass spectrometry. Mar. Chem. 2004, 92 (1
4), 2337.
(25) Reemtsma, T.; These, A.; Springer, A.; Linscheid, M.
Differences in the molecular composition of fulvic acid size fractions
detected by size-exclusion chromatographyon line Fourier transform
ion cyclotron resonance (FT-ICR) mass spectrometry. Water Res.
2008, 42 (12), 6372.
(26) Witt, M.; Fuchser, J.; Koch, B. P. Fragmentation studies of fulvic
acids using collision induced dissociation Fourier transform ion
cyclotron resonance mass spectrometry. Anal. Chem. 2009, 81 (7),
26882694.
(27) Cory, R. M.; McNeill, K.; Cotner, J. P.; Amado, A.; Purcell, J.
M.; Marshall, A. G. Singlet oxygen in the coupled photochemical and
9358

dx.doi.org/10.1021/es302468q | Environ. Sci. Technol. 2012, 46, 93509359

Environmental Science & Technology

Article

ultra-high resolution mass spectrometry. Limnol. Oceanogr. 2006, 51

(2), 10541063.

(48) These, A.; Reemtsma, T. Structure-dependent reactivity of low

molecular weight fulvic acid molecules during ozonation. Environ. Sci.
Technol. 2005, 39 (21), 83828387.
(49) Chin, W. C.; Orellana, M. V.; Verdugo, P. Spontaneous
assembly of marine dissolved organic matter into polymer gels. Nature
1998, 391 (6667), 568572.
(50) Conte, P.; Piccolo, A. Conformational arrangement of dissolved
humic substances. Influence of solution composition on association of
humic molecules. Environ. Sci. Technol. 1999, 33 (10), 16821690.
(51) Kerner, M.; Hohenberg, H.; Ertl, S.; Reckermann, M.; Spitzy, A.
Self-organization of dissolved organic matter to micelle-like microparticles in river water. Nature 2003, 422 (6928), 150154.
(52) Reemtsma, T.; These, A. On-line coupling of size exclusion
chromatography with electrospray ionization-tandem mass spectrometry for the analysis of aquatic fulvic and humic acids. Anal. Chem.
2003, 75 (6), 15001507.
(53) Hunt, S. M.; Sheil, M. M.; Belov, M.; Derrick, P. J. Probing the
effects of cone potential in the electrospray ion source: Consequences
for the determination of molecular weight distributions of synthetic
polymers. Anal. Chem. 1998, 70 (9), 18121822.
(54) Gabelica, V.; Pauw, E. D. Internal energy and fragmentation of
ions produced in electrospray sources. Mass Spectrom. Rev. 2005, 24
(4), 566587.
(55) Cole, R. B. Some tenets pertaining to electrospray ionization
mass spectrometry. J. Mass Spectrom. 2000, 35 (7), 763772.
(56) These, A.; Reemtsma, T. Limitations of electrospray ionization
of fulvic and humic acids as visible from size exclusion chromatography
with organic carbon and mass spectrometric detection. Anal. Chem.
2003, 75 (22), 62756281.
(57) Kramer, R. W.; Kujawinski, E. B.; Hatcher, P. G. Identification
of black carbon derived structures in a volcanic ash soil humic acid by
Fourier transform ion cyclotron resonance mass spectrometry. Environ.
Sci. Technol. 2004, 38 (12), 33873395.
(58) Kujawinski, E. B.; Behn, M. D. Automated analysis of
electrospray ionization Fourier transform ion cyclotron resonance
mass spectra of natural organic matter. Anal. Chem. 2006, 78 (13),
43634373.
(59) Sleighter, R. L.; Hatcher, P. G. Molecular characterization of
dissolved organic matter (DOM) along a river to ocean transect of the
lower Chesapeake Bay by ultrahigh resolution electrospray ionization
Fourier transform ion cyclotron resonance mass spectrometry. Mar.
Chem 2008, 110 (34), 140152.
(60) ChemSpider; http://www.chemspider.com/.
(61) Wammer, K. H.; Peters, C. A. Polycylic aromatic hydrocarbon
biodegradation rates: A structure-based study. Environ. Sci. Technol.
2005, 39, 25712578.
(62) Bordwell, F. G.; Drucker, G. E. Acidities of indene and phenyl-,
diphenyl-, and triphenylindenes. J. Org. Chem. 1980, 45 (16), 3325
3328.
(63) Haiber, S.; Herzog, H.; Burba, P.; Gosciniak, B.; Lambert, J.
Two-dimensional NMR studies of size fractionated Suwannee River
fulvic and humic acid reference. Environ. Sci. Technol. 2001, 35 (21),
42894294.
(64) Hertkorn, N.; Benner, R.; Frommberger, M.; Schmitt-Kopplin,
P.; Witt, M.; Kaiser, K.; Kettrup, A.; Hedges, J. Characterization of a
major refractory component of marine dissolved organic matter.
Geochim. Cosmochim. Acta 2006, 70 (12), 29903010.
(65) Lam, B.; Baer, A.; Alaee, M.; Lefebvre, B.; Moser, A.; Williams,
A.; Simpson, A. J. Major structural components in freshwater dissolved
organic matter. Environ. Sci. Technol. 2007, 41 (24), 82408247.
(66) Egan, D.; OKennedy, R.; Moran, E.; Cox, D.; Prosser, E.;
Thornes, R. D. The pharmacology, metabolism, analysis, and
applications of coumarin and coumarin-related compounds. Drug
Metab. Rev. 1990, 22 (5), 503529.
(67) Hollman, P. C. H. Evidence for health benefits of plant phenols:
Local or systemic effects? J. Sci. Food Ag. 2001, 81 (9), 842852.
(68) Kim, S.; Kaplan, L. A.; Hatcher, P. G. Biodegradable dissolved
organic matter in a temperate and a tropical stream determined from
9359

dx.doi.org/10.1021/es302468q | Environ. Sci. Technol. 2012, 46, 93509359