www.elsevier.com/locate/marchem

Marine Chemistry 92

Examining marine particulate organic matter at sub-micron scales

using scanning transmission X-ray microscopy and carbon X-ray

absorption near edge structure spectroscopy

Jay A. Brandesa,*, Cindy Leeb, Stuart Wakehamc, Michael Petersond, Chris Jacobsene,

Sue Wiricke, George Codyf

aMarine Science Institute, The University of Texas at Austin, 750 Channel View Drive, Port Aransas, TX 78373, United StatesbMarine Sciences Research Center, Stony Brook University, Stony Brook, NY 11794-5000, United States

cSkidaway Institute of Oceanography, 10 Science Circle, Savannah, GA 31411, United StatesdSchool of Oceanography, University of Washington, Seattle, WA 98195, United StateseDepartment of Physics, SUNY-Stony Brook, Stony Brook, NY 11794, United States

fGeophysical Laboratory, Carnegie Institution of Washington, 5251 Broad Branch Road NW, Washington, DC 20015, United States

Received 27 October 2003; received in revised form 4 May 2004; accepted 30 June 2004

Available online 3 October 2004

Abstract

Marine sinking particulate organic matter (POM) represents the link between surface primary production and burial of

organic matter in marine sediments. As such, the nature of this material has been the subject of numerous studies attempting

to characterize its composition. The results of these studies have shown that a significant proportion of POM is not

recognizable as known compounds, and that the proportion of uncharacterized material increases with age/depth/diagenesis.

However, few studies have examined the spatial heterogeneity of this material. This study uses a new tool, scanning

transmission X-ray microscopy (STXM), together with carbon X-ray absorption near edge structure (C-XANES), to examine

POM collected from sediment traps deployed at one location in the Arabian Sea as part of the JGOFS program. The results

indicate that POM is composed primarily of four distinct phases: protein, an aliphatic rich phase, a carboxylic-acid-rich

phase, and a phase with complex unsaturated and quinone character. This last phase may be a condensation product between

carbohydrates and proteins, or from degraded plant pigments. Many particles consisted of a single chemical phase; however,

in particles with mixed compositions, individual domains retained distinctive chemical signatures at the instrument’s

resolution limit (50 nm). All major chemical phases were observed in sediment trap particles from 531 to 3369 m depth,

supporting the hypothesis that non-selective degradation dominates particle remineralization, and that overall particle

compositions are determined by near surface processes. Only one particle, out of more than 60 examined, exhibited soot-like

0304-4203/$ - s

doi:10.1016/j.m

* Correspon

E-mail addr

(2004) 107–121

ee front matter D 2004 Elsevier B.V. All rights reserved.

archem.2004.06.020

ding author. Tel.: +1 361 7496756.

ess: brandes@utmsi.utexas.edu (J.A. Brandes).

J.A. Brandes et al. / Marine Chemistry 92 (2004) 107–121108

composition. The lack of a significant black carbon/soot component may be attributable to sampling during the winter

monsoon period.

D 2004 Elsevier B.V. All rights reserved.

Keywords: Organic Geochemistry; Sediment trap; X-ray microscopy; Arabian Sea; Black carbon; Algaenan

1. Introduction

The nature of organic matter leaving the euphotic

zone and its fate have been the subject of much

debate (Lee et al., in press). Wakeham et al. (1997)

were able to identify many compounds directly by

HPLC, GC, and GC/MS, yet, their work pointed to

an inability to identify the majority (N50%) of carbon

in sinking POM. The presence of banalyticallyuncharacterizedQ material in sinking POM (Hedges

et al., 2000) has led to a number of hypotheses about

the nature of this material. Sources may include:

algaenan or waxy insoluble material (Hwang and

Druffel, 2003), soot or ash-derived black carbon

(Deuser et al., 1983), transparent exopolymers

formed from polysaccharides (Chin et al., 1998;

Passow et al., 2001; Lee et al., in press), or organic

matter that has been rendered unidentifiable by

internal polymerization/crosslinking (Hedges, 1978;

Qian et al., 1992). Attempting to reconcile between

these competing hypotheses is difficult as different

methods examine only selective fractions of POM.

There is also the unstated assumption that com-

pounds are homogenously distributed within sam-

ples, but there is very little information to support or

refute this.

One reason that spatial information content of

samples has been largely ignored in biogeochemistry

is that the analytical techniques most commonly used

in this field do not lend themselves to acquiring this

information. Studies of cells using various microscopy

techniques (Mitchell, 2001; Sommer and Franke,

2002; Jamin et al., 2003) serve as reminders that

organic matter is produced with chemical heteroge-

neity on nanoscales. Although the composition of

individual structures may be tremendously altered by

diagenesis, it is not a given that these structures will

be homogenized. In special cases, plant cell structures

have been shown to survive burial for hundreds of

millions of years, with some features still identifiable

by their distinct chemical signatures (Boyce et al.,

2002).

Electromagnetic-radiation-based micro-spectro-

scopy methods such as infrared and X-ray micro-

scopy are increasingly used as tools to investigate

problems in environmental chemistry and biochem-

istry (Myneni, 2002; Dumas and Miller, 2003; Miller

et al., 2003). These spectromicroscopy methods can

generate information on functional group distribu-

tions within samples with enough spatial and

chemical resolution to minimize bsignal averagingQproblems, and with minor sample handling (e.g., no

derivatization or extraction). Infrared microscopy is

diffraction limited to spatial resolutions of roughly

3–10 Am, while X-ray microscopy has a theoretical

resolution limit of less than 10 nm, with current

instruments limited to about 30 nm (Jacobsen and

Kirz, 1998).

Using marine particulate organic matter (POM)

from the Arabian Sea, we have applied scanning

transmission X-ray microscopy (STXM) together

with concurrent measurements of carbon X-ray

absorption near edge structure (C-XANES) spectro-

scopy to examine the potential for identifying

particle compositions and mapping particle hetero-

geneity. This approach allows us to examine the

composition of POM on a particle-by-particle basis.

Although there are limitations in examining large

particles (N200 Am diameter particles were not

analyzed), and particles need to be dried and briefly

heated to embed them in S8 prior to sectioning, our

approach examines all particles without extraction or

derivatization and thus provides an overall view of

particle compositions.

2. Background on X-ray microscopy

X-ray absorption edges arise when an incident X-

ray photon exceeds the threshold energy needed to

Table 1

Approximate energy ranges for primary absorption peaks at the

carbon 1s edge

eV Functionality

284 quinone

284.9–285.5 unsaturated/aromatic

285.8–286.4 aromatic CUOH

286.6 ketone CMO

287.1–287.4 aliphatic

287.7–288.6 carbonyl

289.3–289.5 alcohol

290.3–290.6 carbonate

Sources: Francis and Hitchcock (1992), Hitchcock et al. (1986

1992), Hitchcock and Ishii (1987), Hitchcock and Stohr (1987), and

Urquhart and Ade (2002).

Fig. 1. Schematic representation of X-ray absorption near-edge

structure or XANES spectroscopy. The first peak represents the

primary XANES resonance (e.g., a 1s–k* transition), followed by

the absorption edge. Transitions can be from a low level occupied

orbital (n=1, 2, or 3) to either an unoccupied molecular orbital

(represented by the first arrow) or to the continuum (second arrow),

where electrons are fully ionized by absorption of X-ray energy.

J.A. Brandes et al. / Marine Chemistry 92 (2004) 107–121 109

completely remove (ionize) an electron from an inner-

shell orbital. The result is a step-like rise (the bedgeQin X-ray absorption near edge structure spectroscopy)

in the absorption cross section. Photons with energies

just below the ionization edge (280–300 eV in the

case of carbon 1s electrons) can promote core

electrons into a variety of bound states that correspond

to unoccupied or partially occupied molecular orbitals

(Fig. 1). These near edge absorption bands are

sensitive indicators of the local chemical bonding

environment surrounding the atom in question (Stohr,

1992). In the case of C (1s) XANES, the absorption

spectrum can be used to estimate the abundances of

different functional groups, subject to the caveat that

carbon in different electronic environments may have

vastly different absorption coefficients, and in many

cases absorption bands from different organic func-

tional groups overlap.

Extensive C-XANES studies of polymers and other

well-characterized materials provide generally robust

peak assignment criteria. In the case of carbon K-shell

electrons, the near edge absorption bands span from

280 to 295 eV and correspond to transitions to both

unoccupied k* (antibonding) and low lying j*orbitals. As a general rule, the peak width is given

by the lifetime of the photoexcited transition. Tran-

sitions at lower energies relative to ionization edge

have longer lifetimes, and since the 1s–k* transitions

occur at lower energy than do 1s–j* transitions, the

1s–k* transitions tend to have a narrow band width.

Conjugated k orbitals present multiple unoccupied k*orbitals to which the photoexcited electron can be

promoted. In general, however, these higher energy

1s–k* transitions are considerably weaker than the

lowest energy 1s–k* transition. Table 1 lists the peak

energy assignments for common carbon functional

groups. The lowest absorption bands, occurring at

around 284–285 eV, correspond to molecules with

unusually low energy k* states, e.g., quinones.

Aromatic carbon that is bonded to hydrogen or

reduced carbon exhibits a strong 1s–k* transition at

~285 eV. With the addition of electron withdrawing

elements in the ring (e.g., pyridinic nitrogen) or

substituted on the ring (e.g., carbonyl or oxygen) the

binding energy of the 1s electron is increased shifting

the 1s–k* transitions of aromatic carbon to higher

energies, e.g., ~286 eV for carbonyl and nitrogen, and

~287 eV for oxygen (e.g., phenols). In the case of

carboxyl functional groups, the effect of two oxygens

results in a shift of the 1s–k* transition up to 288.5

eV; carboxyamide’s 1s–k* transition lies slightly

lower at 288.2 eV (see discussions by Urquhart and

Ade, 2002). Saturated carbon exhibits a relatively

strong absorption band near 287.5–288 eV, when

substituted with oxygen (e.g., ethers and alcohols),

this transition shifts up to ~289.5 eV. Although an

overlap exists between different functional groups

(most notably in the 286–289 region), in marine

,

J.A. Brandes et al. / Marine Chemistry 92 (2004) 107–121110

organic matter one can use other information, e.g.,

previous NMR studies, to rule out the presence of

significant aldehyde or ketone functionalities (Hedges

et al., 2001).

Spectra at the full 30–50 nm resolution limit of

present-day X-ray microscopes can be obtained

either by using cross-correlation methods for post-

facto alignment (Jacobsen et al., 2000), or laser-

interferometer scanning stage control (Kilcoyne et

al., 2003), to acquire a large series of images over a

range of energies spanning the XANES spectral

region. This stack of images yields a data set with

spatial dimensions X and Y, and energy dimension E.

All of the particle data presented in this paper were

collected from such three-dimensional data sets. By

taking the ratio of data from images collected below

and at a particular absorption edge, one can map the

concentration of specific functional groups across a

sample. For example, calculation of the ratios of

image maps from 285 and 280 eV yields a map of

the distribution of unsaturated carbon within a

sample, while a ratio of 298 eV (above the XANES

regions) to 280 eV yields a carbon density or optical

density map.

Because we obtain data with 100–200 spectral

points, and several tens of thousands of spatial

locations, multivariate statistical analysis methods are

used to conduct detailed sample analysis. Principal

component analysis (PCA) provides a means of

highlighting successive themes of correlation of

spectra among the (X, Y, E) data (King et al.,

1989; Bonnet et al., 1999; Osanna and Jacobsen,

2000), and to reject poor correlations which are

dominated by photon statistics noise only. We have

used PCA as a method of generating a noise-filtered,

orthogonal space for cluster analysis of these data

(Jacobsen et al., 2003; Lerotic et al., 2004). By

grouping pixels together based on their spectroscopic

signature and with PCA noise-filtering, one can

obtain spectra from representative regions in the

sample with a vastly improved signal to noise ratio.

At the stage of development employed here, cluster

analysis has not always succeeded in finding few-

pixel regions with different spectroscopic signatures,

although efforts to improve this in future work are

underway. This paper therefore makes use of both

cluster analysis and user-selected spectromicroscopy

analysis.

3. Methods

Archived sediment trap samples from the Arabian

Sea JGOFS program were analyzed from station M4

(531, 814, 2222, and 3369 m). The collection

methodologies are described in Lee et al. (1998) and

Wakeham et al. (2002). Small subsamples (~1 mg)

were embedded in reagent grade elemental sulfur by

briefly heating (~120 8C) a mixture of fine S8 grainswith the samples on a clean glass slide until the

mixture melted into a drop. The drop was cooled,

removed from the slide with a razor, and glued to an

epoxy mount using cyanoacrylate glue. The drop was

then subsampled into 70–100 nm thick sections using

a Leica Ultra-Cutk ultramicrotome and a diamond

knife. Sections were transferred to silica monoxide

supported TEM grids for storage and analysis. Images

and stacks were collected at the scanning transmission

X-ray microscope located on beamline X1-A (out-

board branch) at the National Synchrotron Light

Source in Brookhaven National Laboratory (Winn et

al., 2000). Full details on the microscope design,

operation and capabilities are given in Feser et al.

(2000) and Feser et al. (2001).

4. Results

Rather than present spectral libraries of possible

model organic compounds present in sinking POM,

we present below representative spectra found in an

investigation of more than 60 POM particles ana-

lyzed. Although material from four traps was ana-

lyzed, the number of particles observed ranged from 7

(531 m trap) to 20 (2222 m trap). The limited number

of observations, necessitated by an average analysis

time of 2 h per observation (one to five particles per

observation), makes combining all observations the

best choice for the overview analysis presented here.

Particle chemical identifications are based upon peak

positions (Table 1) and relative intensities. The spectra

shown in this paper are in units of optical density,

which is the negative of the logarithm of the ratio of

transmitted to incident flux, �ln(I/Io). As a reference,

a 110-nm-thick, 2 g/cm3 carbon film would have an

optical density of 1.0 at 300 eV. The x-axis is plotted

as X-ray energy in electronvolts. The following

organic matter types were observed.

J.A. Brandes et al. / Marine Chemistry 92 (2004) 107–121 111

4.1. Protein

Proteins exhibit both a 285 eV peak, due to the

presence of aromatic amino acids phenylalanine,

tyrosine and tryptophan, and a strong 288.2 eV peak

(Fig. 2A). The 288.2 eV peak is due to the presence of

amide-linked carbonyl carbon, and is shifted down in

energy by 0.2–0.3 eV from the 1s–k* transition of

carboxylic acids. As noted below, protein may also be

present in several of the other particle types identified.

However, for the purposes of identifying the majority

chemical composition, the presence of a 288–288.2

eV peak was considered diagnostic for the dominance

of protein within a particle.

4.2. Material with a strongly acidic nature

In many particles with a dominant absorption peak

in the 288–289 eV region, the peak center was

observed at 288.5 eV, characteristic of carboxylic

acids (Table 1; Fig. 2B). This material exhibited

consistent spectra, namely a small peak in the 285–

286 eV region followed by a primary peak centered at

288.5 eV. The small 285 eV peak indicates the

presence of either aromatic or olefinic carbon. The

relatively broad nature of the 288.5 eV peak, with

significant energy in the 288.0–288.3 eV range likely

records some protein content. A protein spectrum is

overlaid in Fig. 2b to illustrate the spectral differences

between these particle chemistries. The molar absorp-

tion coefficient of aliphatic carbon is much weaker

than that of carbonyl carbon; nonetheless, there is a

significant absorption component in the 287–288 eV

region, as well as some spread of the 288.5 eV energy

into the 289.5 eV region. Material containing a very

prominent and narrow 288.5 eV peak and consisting

of long, thin bstringsQ was found at 531 and 814 m

depths. Spectrally and structurally, this material

strongly resembled lipid-rich cell walls, although an

absolute confirmation of its source could not be made.

All acid-rich particles typically had a 288.5 eV peak

that was 20–30% greater in intensity than the 290–300

eV range.

4.3. Aliphatic particles

Some particles were notable for lack of a strong

peak in the 287–288 eV region, instead exhibiting a

rise/plateau spectrum in this energy range. This

suggests a chemical composition dominated by

aliphatic carbon (Table 1). This material also con-

tained significant energy absorption in the 286 and

287 eV ranges, but little or no absorption in the

aromatic region. The high C/O ratio and presence of

peaks in the ketone/aldehyde region suggest a waxy or

paraffinic nature of this material.

4.4. Carbohydrates

Sinking particulate organic matter contains a

significant proportion of carbohydrates (Wakeham et

al., 1997). Carbohydrates are dominated by a 289.5

eV absorption peak (Cody et al., 1996; Cody and

Saghi-Szabo, 1999) (Fig. 2D). There is also a minor

286.5 eV peak that may be an artifact of radiation

damage or diagenesis. Although we used techniques

designed to minimize this problem (scanning from

low to high energy range, using minimal dwell times

and imaging the particle as few times as possible prior

to spectral analysis) carbohydrates (as well as other

compounds) are susceptible to radiation damage

(Foster et al., 1992; Cazaux, 1997; Cherezov et al.,

2002; Ito et al., 2002; Ravelli et al., 2002), and this

damage appears as both a decrease in the 289.5 eV

absorption (mass loss) and an increase in the 286.5 eV

absorption. The increase at 286.5 eV is presumably

due to the formation of aromatic heterocycles like

furans (Newbury et al., 1986). The formation of

phenolic carbon is unlikely as there is an insignificant

component present at 285 eV.

4.5. Calcium carbonate

Carbonates exhibit a very characteristic strong

absorption peak at 290.5 eV (Fig. 2E) (Urquhart and

Ade, 2002). In pure CaCO3, only a few other minor

absorption peaks are observed (Fig. 2E), but in our

sediment trap samples, the presence of organic matter

was often detected intermixed with the mineral.

4.6. Particles with quinone/unsaturated character

There were a significant number of particles

exhibiting strong absorbance in the 284–286 eV range

associated with unsaturated or aromatic carbon (Fig.

2F). This organic matter is distinguished by several

Fig. 2. Carbon XANES spectra of different classes of marine particulate matter. (A) Protein, (B) acid-rich (with protein spectrum as a

comparison), (C) aliphatic-rich, (D) carbohydrate, (E) CaCO3, (F) quinone/unsaturated, (G) soot/black carbon, and (H) non-carbon mineral.

J.A. Brandes et al. / Marine Chemistry 92 (2004) 107–121112

Fig. 3. Particle chemistry histogram. Dark bards represent particles

observed without any sign of a CaCO3 peak at 290.5 eV, while

speckled bars represent particles with a CaCO3 peak.

J.A. Brandes et al. / Marine Chemistry 92 (2004) 107–121 113

spectral features: (1) presence of a significant 285 eV

peak with components at 284 and 286 eV associated

with unsaturated carbon, phenols, quinone, and

polynuclear aromatic carbon (Horsley et al., 1985;

Francis and Hitchcock, 1992; Hitchcock et al., 1992;

Hitchcock and Mancini, 1994); (2) significant absorp-

tion in the 287–288 eV range associated with aliphatic

carbon; (3) a lack of a strong absorption peak in the

288 or 289 eV range. Some particles do exhibit small

peaks in the 288.5 and 289.5 region, suggesting that

they also may contain organic acids and alcohol

moieties.

4.7. Black carbon/soot

The presence of soot or combustion products in

sinking POM has been observed (Schmidt and Noack,

2000). Of the ~60 particles that we have examined

from station M4 in the Arabian Sea, only one has

unambiguously fit the profile of a combustion source

particle (Fig. 2G). This particle contained a large and

well-defined 285 eV peak, followed by peaks in the

288.5 and 287.5 eV range. This spectrum is character-

istic of combustion-derived soot (see spectra of NIST

standard soot at http://xray1.physics.sunysb.edu/data/

spectra.php), with the exception of a minor 288.5 peak

that may be the result of adsorbed marine carbon or

UV induced oxidation.

4.8. Non-carbon minerals and elements

Mineral grains and SiO2 frustules have bflatQspectra with no absorption peaks present (Fig. 2H).

The presence of mineral in organic matter is

expressed by a uniform increase in optical density,

which is especially noticeable in the region below the

C edge (b283 eV). Thus, if a particle contains a

significant amount of mineral, the optical density will

be shifted upwards by N0.5 (compare the baseline in

Fig. 2D with that in Fig. 2A for an example). The

only common element with an absorption peak in the

280–300 eV range is potassium, this exhibits 2sYk*shell doublet peaks at 295 and 297 eV. This energy

range lies above the region of interest for organic

carbon; however, it can complicate efforts to deter-

mine particle densities based upon techniques that

integrate the region immediately above the absorption

edge at ~290 eV.

5. Discussion

5.1. Overall distributions

The roughly 60 particles from Arabian Sea sedi-

ment traps examined by C-XANES were dominated

by acid-rich (~25%), aliphatic (15%), protein-rich

(15%), and quinone/aromatic (10%) compositions

(Fig. 3). In many cases, particles consisted of

adjoining regions with different chemical composi-

tions, such as carbohydrate/acid, protein/acid, and

algaenan/acid; in several cases, particles consisted of

three or more dominant chemical types distributed

non-homogenously (denoted as Mix, Fig. 3). Our

estimate of particle chemical distributions is prelimi-

nary; without detailed spectral deconvolution within

each region, we cannot quantify exact chemical

proportions. The presence of a noticeable (N0.1

optical density) peak at 290.5 eV on organic particles

was considered diagnostic for the presence of CaCO3.

Most of the CaCO3-associated material fell into the

acidic particle category (Fig. 3). This is consistent

with the known preferential association of acidic

amino acids and other compounds with CaCO3. The

presence of protein in some of the acid-rich particles

Fig. 4. Carbon optical density maps (ratio of absorbance from 295 to

300 eV to absorbance from 280 to 283 eV) of cell wall remnants

High carbon concentration is shown in bright areas of this optica

density image. Samples from M4, Arabian Sea, 813 m depth.

J.A. Brandes et al. / Marine Chemistry 92 (2004) 107–121114

is also likely, as many of these particles contained

non-trivial (N0.05 optical density) 285 eV peaks

associated with aromatic amino acids (although

unsaturated fatty acids and other olefins are also a

possibility). Thus, our estimate of the number of

protein-containing particles is an underestimate. The

same is true for the aliphatic class; without spectral

deconvolution, it is impossible to assign a proportion

of this composition to particles containing a signifi-

cant amount of protein or acidic materials.

In three cases, we observed carboxylic acid-rich

linear or branched structures (two each in the 814 m

samples, one in the 531 m samples). Examples of

two of these are given in Fig. 4. The strong peak at

288.5 eV in this material is consistent with the

presence of fatty acids. This material may include

the remnants of cell walls, and indeed in Fig. 4A, a

series of linked 4–5 Am length cells is presented.

This material appears devoid of debris or other

cellular remains.

Most particles examined appear to be composed

of only a single chemical phase. About a quarter of

particles consisted of two or more closely inter-

mingled phases. This suggests that the processes of

digestion and aggregation that lead to POM for-

mation do not normally homogenize or mix this

material on tens of nanometer scales, although

smaller scale mixing is possible. Principal compo-

nent analysis on one mixed particle indicated that,

although some of the components appear primarily

to be variations in thickness of the same material

(e.g., Fig. 5D and E), there are at least four

chemically distinct phases present. The material

mapped in Fig. 5C, D, and E are dominated by

aliphatic components, although the outer edges

appear to have some added acidic or proteinaceous

character. There is also evidence of carbohydrate

(Fig. 5G,H), CaCO3 intermixed at a sub-micron scale

with acidic or proteinaceous organic material (Fig.

5I), and inorganic, presumably siliceous, particles

(Fig. 5J). The carbohydrate-rich material appears

more closely associated with the non-CaCO3 con-

taining particles, while the other fractions appear to

be randomly distributed.

In the case of the three-component system shown

in Fig. 6, an inorganic core is surrounded by an

aliphatic-rich layer, followed by a transition to a

proteinaceous/acidic outer layer. The presence of

.

l

defined peaks in the background spectrum shown in

Fig. 6C indicates that the PCA software has integrated

some regions with organic matter in them. However,

the intensity of such peaks is very small (compare the

maximum peak intensity in Fig. 6C at ~0.08 optical

density, with that in Fig. 6D at ~1.2 O.D.). Thus,

although some organic matter has been integrated into

the average spectrum, it is not a major component and

represents very thin regions around the periphery of

the particles in Fig. 6. There is also a small CaCO3

dominated particle in the upper central portion of the

image (Fig. 6G). Even in cases where compounds are

Fig. 5. Optical density and cluster analysis maps of sediment trap sample. Sample was collected from station M4, 2222 m depth. (A) Carbon

optical density map (ratio of absorption integrated from 295 to 298 eV divided by intensity integrated from 280 to 283 eV). (B) Principal

component analysis (PCA) map. Spectra in graphs C–J correspond to integrated signal intensity from regions denoted in color from PCA maps.

J.A. Brandes et al. / Marine Chemistry 92 (2004) 107–121 115

mixed, we observed distinct chemical bzonesQ withinparticles (Figs. 5 and 6). It is interesting to speculate

whether the presence of heterogeneous chemical

bdomainsQ within particles might lead to entanglement

or encapsulation of reactive compounds that could

make it difficult for enzymes to recognize organic

matter. For example, in Fig. 6, there exists a gradient

between acid-rich and aliphatic-rich compositions

within a single particle.

5.2. Particle character

Our results suggest that marine sinking POM

from the Arabian Sea is a mixture of phases,

Fig. 6. Sample from station M4, Arabian Sea, 814 m depth. (A) Cluster analysis map. (B) Carbon optical density map (ratio of absorbance from

295 to 300 eV to absorbance from 280 to 283 eV). (C) Map and spectra of background. (D) Map and spectra of non-carbonaceous material. (E)

Map of amino acid/carboxylic acid rich organic matter. (F) Map of aliphatic-rich material. (G) Map of calcium carbonate.

J.A. Brandes et al. / Marine Chemistry 92 (2004) 107–121116

including several that could be difficult to analyze by

liquid or gas chromatography. One such phase could

be algaenans. The suggestion that algaenans and

other waxy or aliphatic cross-linked phases are

abundant in sinking POM (Hwang and Druffel,

2003) is supported by the presence of a significant

component of aliphatic-rich particles (15%, Fig. 3).

If one includes all particles containing some fraction

of this material, the proportion rises to N20%. While

recognizing that particle abundances do not neces-

sarily correlate to particle mass, this still suggests

that this material is a major component of the

sinking POM. The lack of a strong 288 or 289 eV

peak in this material indicates that it is extremely

oxygen poor, and thus hydrophobic. As the oscillator

strength of carboxyl groups is quite high, even a

small amount of COOH functionality would result in

a noticeable 288.5 eV peak, which was not observed

in most cases (Fig. 2C).

The most intriguing particle composition category

is the quinone/aromatic type (Fig. 2F). This is a

significant particle category (10%, 15% if mix

particles are included). The major cellular biochem-

icals: protein, carbohydrate, lipids, and nucleic acids,

do not exhibit an absorption peak in the 284–285 eV

(quinone) region. However, certain photosynthetic

reaction centers use quinonic moieties for electron

transport, and certain condensation reactions can also

produce quinones. The dissimilarity in spectra

between chlorophyll a (Fig. 7) and the samples

Fig. 7. Carbon XANES spectra of (A) chlorophyll a and (B)

melanoidin produced by the reaction of lysine with glucose.

J.A. Brandes et al. / Marine Chemistry 92 (2004) 107–121 117

indicates that this POM category is not simply

undegraded chlorophyll or even a simple degradation

product of this pigment (e.g., phaeophorbide). Only

minor amounts of pigments and their intact degrada-

tion products are found in sinking POM (Welsch-

meyer and Lorenzen, 1985; Lee et al., 2000;). Another

possible reaction pathway that could produce quinone

moieties is the reaction between proteins and carbo-

hydrates, the Maillard reaction (Hedges, 1978).

Several studies have shown that this reaction is

capable of providing a wide range of polynuclear

aromatic, quinone and unsaturated compounds (e.g.,

Olsson et al., 1977, 1978; Miller et al., 1984). Indeed,

C-XANES of the reaction product of lysine and

glucose (e.g., melanoidin) contains significant spectral

content in the 284–285 eV range (Fig. 7A), but is

again, not spectrally identical to this POM class.

There is also a spectral feature at around 286.3–287.0

in Fig. 2F that may correspond to a peak at 286.4 eV

in the melanoidin spectrum. Without further work, it

is not possible to distinguish between these two

sources, or to rule out another pathway that can

produce this material.

Surprisingly, a particle class that does not appear

to form a major component of sinking POM in

Arabian Sea samples is soot/black carbon (BC). In

this study, we observed only one such particle (out

of N60); if combustion products made up a signifi-

cant percentage of sinking POM, then one might

expect to observe many more. The quinone/aromatic

fraction described above may be ruled out as soot or

combustion derived BC based upon two observa-

tions: (1) Most of that particle class had a

morphology inconsistent with soot, being composed

of strings and membrane-like structures; and (2) soot

and graphitic material does not, in our experience,

exhibit a significant sub-285 eV component. It must

be noted that the Arabian Sea is a marine produc-

tivity dominated regime, and that samples from sites

closer to urban population centers may contain larger

amounts of soot and ash derived black carbon. These

samples were also taken during the winter monsoon

period, when prevailing winds come from Somalia.

The presence of significant amounts of soot and

combustion products in aerosols originating from the

Indian subcontinent (Dickerson et al., 2002; Menon

et al., 2002) would presumably generate a greater

abundance of black carbon in particles collected

when winds from the NE are prevalent. It also may

be possible that very small aerosol particles of BC

would be overlooked by our method (although stack

analysis targeting aromatic rich particles was done

for each data set). However, very small particles

would not sink unless adsorbed or otherwise attached

to larger particles.

Given the high proportion of protein identified in

sinking POM by Wakeham et al. (1997) and Lee et

al. (2000), it is not surprising to see protein as a

major particle composition class (15%, or 27%

including mixtures, Fig. 3). Some particles within

the acid-rich particle class (25%, Fig. 3) also

undoubtedly contain significant amounts of protein,

based upon broadening of the primary peak in the

J.A. Brandes et al. / Marine Chemistry 92 (2004) 107–121118

288.0–288.3 region. It must be noted that other,

non-protein, compounds can give spectra quite

similar to the acid-rich type particles shown in

Fig. 2B. For example, peptidoglycan and chitin both

have primary absorption peaks centered at 288.5 eV.

The chief spectral differences between pure pepti-

doglycan and the particle spectra in Fig. 2B are that

peptidoglycan exhibits no absorption in the aromatic

region (285 eV) and has a significant 289.5 eV

peak associated with alcohol (peptidoglycan, chitin

and other model compound spectra can be found at

the X1A online spectral library http://xray1.physics.

sunysb.edu/data/spectra.php). The lack of strong

289.5 eV peaks in the acidic particles indicates that

either alcohol functionalities are lost early in

diagenesis or that the material began with a lower

alcohol/carbon ratio than typical for carbohydrates.

5.3. Implications for the alteration of sinking POM

Given the results presented above, one can begin

to put different results in the context of the large

body of work examining POM decomposition.

Hedges et al. (2001) suggested that the overall

chemical composition of sinking POM was consis-

tent with passage through the water column.

Although we did not measure enough particles to

make definitive statements about the effect of early

diagenesis upon particle composition (less than 15

particles were measured at each depth), we did

observe all of the major particle chemical categories

at each depth. This would suggest that each particle

type is labile to some extent, as there will be a

significant loss of material between the 531 and

3369 m traps (Lee et al., 1998). In addition, even

material that would presumably be extremely labile,

such as proteins, is present at depths where over 2/3

of the total POM flux has been remineralized

(Wakeham et al., 1997; Lee et al., 1998). Thus, the

overall pattern of particle chemistry, including the

presence of presumably highly altered material,

appears to be set early on in the lifespan of POM.

The large proportion of analytically uncharacteriz-

able material in deep-water POM is consistent with

the presence of waxy, balgaenanQ-like material

(Hwang and Druffel, 2003) as well as heavily

degraded, presumably crosslinked quinone/aromatic

containing organic matter possibly derived from

reactions between proteins and carbohydrates. The

apparent absence of soot or black carbon in this

marine organic matter, as opposed to previous

studies (Masiello and Druffel, 1998; Middelburg et

al., 1999; Tsapakis et al., 2003), may be due to a

lack of input of these materials during the sample

period, or it may be an artifact of the chemical and

physical isolation procedures used.

Many questions still remain and new ones are

raised by this work. Chief among these is why

different compound classes appear to be remineralized

with equal alacrity by heterotrophs. When one

considers the spectrum of particle character, from cell

walls and protein-rich particles to waxy/algaenan-like

and highly condensed quinone/aromatic containing

molecules, the expectation would be that some

material would be removed/remineralized preferen-

tially, consistent with the results obtained by molec-

ular-level studies (Wakeham et al., 1997; Lee et al., in

press). Again, we cannot make definitive statements

as to particle class abundance with depth, but clearly,

given the limited subset of data presented here, we can

state that no one class is completely removed by

diagenesis at depth. It is not possible, given the

overall mass losses experienced by sinking POM, to

state that any of the chemical types presented in Fig. 2

is immune from degradation. A plausible explanation

is provided by Armstrong et al. (2002), where material

that reaches the deep sea is dominated by mineral-

ballasted organic matter. It may be hypothesized that

all of the organic matter particle types presented here

are susceptible to degradation, but at a rate slow

enough so that particles with mineral ballast can reach

the deeper ocean with their initial and overall

chemistries intact.

6. Conclusions

The use of synchrotron-radiation-based soft X-ray

spectromicroscopy has enabled the examination of

individual particle compositions from sediment trap

samples with minimal sample alteration. Particles

collected in traps from 531, 814, 2222, and 3369 m

depths from station M4 in the Arabian Sea JGOFS

program (Lee et al., 1998) showed a variety of

individual particle phases, but were dominated by

four types. Particles with strong 288.5 eV peaks

J.A. Brandes et al. / Marine Chemistry 92 (2004) 107–121 119

characteristic of carboxylic acids formed the highest

percentage of material, but protein, waxy, and

quinone/aromatic particle chemistries were also

observed. Intact cell walls were observed in trap

samples down to a depth of 814 m, but not seen

below that depth. All other particle classes were

observed down to the lowermost trap samples at

3369 m depth. The results suggest that much of the

bmolecularly uncharacterizedQ fraction described by

Wakeham et al. (1997) and Hedges et al. (2000) is

composed of aliphatic-rich and quinone/unsaturated

material. We hypothesize that this latter particle

composition is the result of Maillard-type reactions

between proteins and carbohydrates and or pigments

that take place early in the particle’s history. Only 1

out of more than 60 particles fit the spectral profile

of soot or other bblack carbonQ combustion derived

materials, suggesting that this material is not a major

component of sinking organic matter in the Arabian

Sea during the winter monsoon period. The presence

of all four major particle classes in even the deepest

trap samples also suggests that no one particle

composition is particularly susceptible to degrada-

tion, rather, the survival of individual particles may

be a balance between slow degradation of all particle

classes and sinking rates influenced by the presence

or absence of inorganic minerals. Future work

involving detailed spectral deconvolution or blineshapingQ, as well as examination of more particles

collected from different regions and seasons, will

provide a more complete assessment of particle

chemical compositions.

Acknowledgements

This work was supported by NSF grants OCE-

0221295 and OCE-0118036 (JAB), OCE-9310364

(SGW), and OCE-9312694 (CL). The National

Synchrotron Light Source is a Department of Energy

supported facility. The authors wish to thank Patty

Garlough and Paul Haberstroh (UT) for assistance in

sample handling and sample analysis, and Mirna

Lerotic (SUNY) for assistance in PCA and cluster

analysis. This work is dedicated to the memory of

John Hedges, who provided encouragement, stim-

ulating discussions and samples for this project.

References

Armstrong, R.A., Lee, C., Hedges, J.I., Honjo, S., Wakeham, S.G.,

2002. A new, mechanistic model for organic carbon fluxes in the

ocean based on the quantitative association of POC with ballast

minerals. Deep-Sea Research. Part 2. Topical Studies in

Oceanography 49 (1–3), 219–236.

Bonnet, N., Brun, N., Colliex, C., 1999. Extracting information

from sequences of spatially resolved EELS spectra using

multivariate statistical analysis. Ultramicroscopy 77, 97–112.

Boyce, C.K., Cody, G.D., Feser, M., Jacobsen, C., Knoll, A.H.,

Wirick, S., 2002. Organic chemical differentiation within fossil

plant cell walls detected with X-ray spectromicroscopy. Geology

30 (11), 1039–1042.

Cazaux, J., 1997. A physical approach to the radiation damage

mechanisms induced by X-rays in X-ray microscopy and related

techniques. Journal of Microscopy-Oxford 188, 106–124.

Cherezov, V., Riedl, K.M., Caffrey, M., 2002. Too hot to handle?

Synchrotron X-ray damage of lipid membranes and mesophases.

Journal of Synchrotron Radiation 9, 333–341.

Chin, W.C., Orellana, M.V., Verdugo, P., 1998. Spontaneous

assembly of marine dissolved organic matter into polymer gels.

Nature 391 (6667), 568–572.

Cody, G.D., Saghi-Szabo, G., 1999. Calculation of the 13C NMR

chemical shift of ether linkages in lignin derived geo-

polymers: constraints on the preservation of lignin primary

structure with diagenesis. Geochimica et Cosmochimica Acta

63 (2), 193–205.

Cody, G.D., Botto, R.E., Ade, H., Wirick, S., 1996. The application

of soft X-ray microscopy to the in-situ analysis of sporinite in

coal. International Journal of Coal Geology 32, 69–86.

Deuser, W.G., Emeis, K., Ittekkot, V., Degens, E.T., 1983. Fly-ash

particles intercepted in the deep Sargasso Sea. Nature 305

(5931), 216–218.

Dickerson, R.R., Andreae, M.O., Campos, T., Mayol-Bracero, O.L.,

Neusuess, C., Streets, D.G., 2002. Analysis of black carbon and

carbon monoxide observed over the Indian Ocean: implications

for emissions and photochemistry. Journal of Geophysical

Research, D: Atmospheres 10.1029/2001JD000501.

Dumas, P., Miller, L., 2003. Biological and biomedical applications

of synchrotron infrared microspectroscopy. Journal of Bio-

logical Physics 29 (2–3), 201–218.

Feser, M., Beetz, T., Carlucci-Dayton, M., Jacobsen, C., 2000.

Instrumentation advances and detector development with the

Stony Brook scanning transmission X-ray microscope. In:

Meyer-Ilse, W., Warwick, A., Attwood, D.T. (Eds.), X-ray

Microscopy: Proceedings of the Sixth International Conference.

American Institute of Physics, Melville, NY, pp. 367–372.

Feser, M., Beetz, T., Carlucci-Dayton, M., Jacobsen, C., 2001.

Scanning transmission soft X-ray microscopy at beamline X-1A

at the NSLS—advances in instrumentation and selected

applications. In: Tichenor, D.A., Folta, J.A. (Eds.), Soft X-ray

and EUV Imaging Systems II. Society of Photo-Optical

Instrumentation Engineers, Bellingham, WA, pp. 146–153.

Foster, G.F., Buckley, C.J., Bennett, P.M., Burge, R.E., 1992.

Investigation of radiation-damage to biological specimens at

J.A. Brandes et al. / Marine Chemistry 92 (2004) 107–121120

water window wavelengths. Review of Scientific Instruments 63

(1), 599–600.

Francis, J.T., Hitchcock, A.P., 1992. Inner-shell spectroscopy of

para-benzoquinone, hydroquinone, and phenol—distinguishing

quinoid and benzenoid structures. Journal of Physical Chemistry

96 (16), 6598–6610.

Hedges, J.I., 1978. Formation and clay mineral reactions of

melanoidins. Geochimica et Cosmochimica Acta 42 (1), 69–76.

Hedges, J.I., Eglinton, G., Hatcher, P.G., Kirchman, D.L., Arnosti,

C., Derenne, S., Evershed, R.P., Kogel-Knabner, I., de Leeuw,

J.W., Littke, R., Michaelis, W., Rullkotter, J., 2000. The

molecularly-uncharacterized component of nonliving organic

matter in natural environments. Organic Geochemistry 31 (10),

945–958.

Hedges, J.I., Baldock, J.A., Gelinas, Y., Lee, C., Peterson, M.,

Wakeham, S.G., 2001. Evidence for non-selective preservation

of organic matter in sinking marine particles. Nature 409 (6822),

801–804.

Hitchcock, A.P., Ishii, I., 1987. Carbon K-shell excitation-spectra of

linear and branched alkanes. Journal of Electron Spectroscopy

and Related Phenomena 42 (1), 11–26.

Hitchcock, A.P., Stfhr, J., 1987. K-shell shape resonances and

intramolecular bond lengths—comments on the relationship

between shape resonances and bond lengths. Journal of

Chemical Physics 87 (5), 3253–3255.

Hitchcock, A.P., Mancini, D.C., 1994. Bibliography and database of

inner-shell excitation-spectra of gas-phase atoms and molecules.

Journal of Electron Spectroscopy and Related Phenomena 67

(1), 1–132.

Hitchcock, A.P., Horsley, J.A., Stfhr, J., 1986. Inner shell excitationof thiophene and thiolane—gas, solid, and monolayer states.

Journal of Chemical Physics 85 (9), 4835–4848.

Hitchcock, A.P., Urquhart, S.G., Rightor, E.G., 1992. Inner-shell

spectroscopy of benzaldehyde, terephthalaldehyde, ethyl ben-

zoate, terephthaloyl chloride, and phosgene—models for core

excitation of poly(ethylene-terephthalate). Journal of Physical

Chemistry 96 (22), 8736–8750.

Horsley, J.A., Stfhr, J., Hitchcock, A.P., Newbury, D.C., Johnson,A.L., Sette, F., 1985. Resonances in the K-shell excitation-spectra

of benzene and pyridine—gas-phase, solid, and chemisorbed

states. Journal of Chemical Physics 83 (12), 6099–6107.

Hwang, J.S., Druffel, E.R.M., 2003. Lipid-like material as the

source of the uncharacterized organic carbon in the ocean?

Science 299 (5608), 881–884.

Ito, Y., Yasuda, K., Ishigami, R., Sasase, M., Hatori, S., Ohashi, K.,

Tanaka, S., Yamamoto, A., 2002. Radiation damage of materials

due to high-energy ion irradiation. Nuclear Instruments &

Methods in Physics Research. Section B, Beam Interactions

with Materials and Atoms 191, 530–535.

Jacobsen, C., Kirz, J., 1998. X-ray microscopy with synchrotron

radiation. Nature Structural Biology 5, 650–653. (Suppl.).

Jacobsen, C., Abend, S., Beetz, T., Carlucci-Dayton, M., Feser,

M., Kaznacheyev, K., Kirz, J., Maser, J., Neuh7usler, U.,

Osanna, A., Stein, A., Vaa, C., Wang, Y., Winn, B., Wirick,

S., 2000. New developments in scanning microscopy at Stony

Brook. In: Meyer-Ilse, W., Warwick, A., Attwood, D.T. (Eds.),

X-ray Microscopy: Proceedings of the Sixth International

Conference. American Institute of Physics, Melville, NY, pp.

12–18.

Jacobsen, C., Feser, M., Lerotic, M., Vogt, S., Maser, J., Schafer, T.,

2003. Cluster analysis of soft X-ray spectromicroscopy data.

Journal de Physique. IV 104, 623–626.

Jamin, N., Miller, L., Moncuit, J., Fridman, W.H., Dumas, P.,

Teillaud, J.L., 2003. Chemical heterogeneity in cell death:

combined synchrotron IR and fluorescence microscopy studies

of single apoptotic and necrotic cells. Biopolymers 72 (5),

366–373.

Kilcoyne, A.L.D., Tyliszczak, T., Steele, W.F., Fakra, S., Hitchcock,

P., Franck, K., Anderson, E., Harteneck, B., Rightor, E.G.,

Mitchell, G.E., Hitchcock, A.P., Yang, L., Warwick, T., Ade, H.,

2003. Interferometer-controlled scanning transmission X-ray

microscopes at the advanced light source. Journal of Synchro-

tron Radiation 10, 125–136.

King, P.L., Browning, R., Pianetta, P., Lindau, I., Keenlyside, M.,

Knapp, G., 1989. Image-processing of multispectral X-ray

photoelectron-spectroscopy images. Journal of Vacuum Sci-

ence & Technology. A. Vacuum, Surfaces, and Films 7 (6),

3301–3304.

Lee, C., Murray, D.W., Barber, R.T., Buesseler, K.O., Dymond, J.,

Hedges, J.I., Honjo, S., Manganini, S.J., Marra, J., Moser, C.,

Peterson, M.L., Prell, W.L., Wakeham, S.G., 1998. Particulate

organic carbon fluxes: compilation of results from the 1995 US

JGOFS Arabian Sea Process Study. Deep-Sea Research. Part 2.

Topical Studies in Oceanography 45 (10–11), 2489–2501.

Lee, C., Wakeham, S.G. and Arnosti, C., in press. Particulate

organic matter in the sea: the composition conundrum. Ambio.

Lee, C., Wakeham, S.G., Hedges, J.I., 2000. Composition and flux

of particulate amino acids and chloropigments in equatorial

Pacific seawater and sediments. Deep-Sea Research. Part 1.

Oceanographic Research Papers 47 (8), 1535–1568.

Lerotic, M., Jacobsen, C., Schafer, T., Vogt, S., 2004. Cluster

analysis of soft x-ray spectromicroscopy data. Ultramicroscopy

100 (1–2), 35–57.

Masiello, C.A., Druffel, E.R.M., 1998. Black carbon in deep-sea

sediments. Science 280 (5371), 1911–1913.

Menon, S., Hansen, J., Nazarenko, L., Luo, Y.F., 2002. Climate

effects of black carbon aerosols in China and India. Science 297

(5590), 2250–2253.

Middelburg, J.J., Nieuwenhuize, J., van Breugel, P., 1999. Black

carbon in marine sediments. Marine Chemistry 65 (3–4),

245–252.

Miller, R., Olsson, K., Pernemalm, P.A., 1984. Formation of

aromatic-compounds from carbohydrates: 9. Reaction of d-

glucose and l-lysine in slightly acidic, aqueous-solution. Acta

Chemica Scandinavica. Series B. Organic Chemistry and

Biochemistry 38 (8), 689–694.

Miller, L.M., Smith, G.D., Carr, G.L., 2003. Synchrotron-based

biological microspectroscopy: from the mid-infrared through the

far-infrared regimes. Journal of Biological Physics 29 (2–3),

219–230.

Mitchell, P., 2001. Turning the spotlight on cellular imaging—

advances in imaging are enabling researchers to track more

accurately the localization of macromolecules in cells. Nature

Biotechnology 19 (11), 1013–1017.

J.A. Brandes et al. / Marine Chemistry 92 (2004) 107–121 121

Myneni, S.C.B., 2002. Soft X-ray spectroscopy and spectromicro-

scopy studies of organic molecules in the environment,

applications of synchrotron radiation in low-temperature geo-

chemistry and environmental sciences. Reviews in Mineralogy

and Geochemistry, 485–579.

Newbury, D.C., Ishii, I., Hitchcock, A.P., 1986. Inner shell electron-

energy loss spectroscopy of some heterocyclic molecules.

Canadian Journal of Chemistry 64 (6), 1145–1155.

Olsson, K., Pernemalm, P.A., Popoff, T., Theander, O., 1977.

Formation of aromatic-compounds from carbohydrates: 5.

Reaction of d-glucose and methylamine in slightly acidic,

aqueous-solution. Acta Chemica Scandinavica. Series B.

Organic Chemistry and Biochemistry 31 (6), 469–474.

Olsson, K., Pernemalm, P.A., Theander, O., 1978. Formation of

aromatic-compounds from carbohydrates: 7. Reaction of d-

glucose and glycine in slightly acidic, aqueous-solution. Acta

Chemica Scandinavica. Series B. Organic Chemistry and

Biochemistry 32 (4), 249–256.

Osanna, A., Jacobsen, C., 2000. Principal component analysis for

soft X-ray spectromicroscopy. In: Meyer-Ilse, W., Warwick, T.,

Attwood, D.T. (Eds.), X-ray Microscopy: Proceedings of the

Sixth International Conference. American Institute of Physics,

Melville, NY, pp. 350–357.

Passow, U., Shipe, R.F., Murray, A., Pak, D.K., Brzezinski,

M.A., Alldredge, A.L., 2001. The origin of transparent

exopolymer particles (TEP) and their role in the sedimenta-

tion of particulate matter. Continental Shelf Research 21 (4),

327–346.

Qian, Y., Engel, M.H., Macko, S.A., 1992. Stable isotope

fractionation of biomonomers during protokerogen formation.

Chemical Geology 101 (3–4), 201–210.

Ravelli, R.B.G., Theveneau, P., McSweeney, S., Caffrey, M., 2002.

Unit-cell volume change as a metric of radiation damage in

crystals of macromolecules. Journal of Synchrotron Radiation 9,

355–360.

Schmidt, M.W.I., Noack, A.G., 2000. Black carbon in soils and

sediments: analysis, distribution, implications, and current

challenges. Global Biogeochemical Cycles 14 (3), 777–793.

Sommer, A.P., Franke, R.P., 2002. Near-field optical analysis of

living cells in vitro. Journal of Proteome Research 1 (2),

111–114.

Stfhr, J., 1992. NEXAFS Spectroscopy. Springer-Verlag, Berlin.

Tsapakis, M., Stephanou, E.G., Karakassis, L., 2003. Evaluation of

atmospheric transport as a nonpoint source of polycyclic

aromatic hydrocarbons in marine sediments of the Eastern

Mediterranean. Marine Chemistry 80 (4), 283–298.

Urquhart, S.G., Ade, H., 2002. Trends in the carbonyl core (C 1s, O

1s)Ypi*c=o transition in the near-edge X-ray absorption fine

structure spectra of organic molecules. Journal of Physical

Chemistry. B 106 (34), 8531–8538.

Wakeham, S.G., Lee, C., Hedges, J.I., Hernes, P.J., Peterson, M.L.,

1997. Molecular indicators of diagenetic status in marine

organic matter. Geochimica et Cosmochimica Acta 61 (24),

5363–5369.

Wakeham, S.G., Peterson, M.L., Hedges, J.I., Lee, C., 2002. Lipid

biomarker fluxes in the Arabian Sea, with a comparison to the

equatorial Pacific Ocean. Deep-Sea Research. Part 2. Topical

Studies in Oceanography 49 (12), 2265–2301.

Welschmeyer, N.A., Lorenzen, C.J., 1985. Role of herbivory in

controlling phytoplankton abundance—annual pigment budget

for a temperate marine fjord. Marine Biology 90 (1), 75–86.

Winn, B., Ade, H., Buckley, C., Feser, M., Howells, M., Hulbert, S.,

Jacobsen, C., Kaznacheyev, K., Kirz, J., Osanna, A., Maser, J.,

McNulty, I., Miao, J., Oversluizen, T., Spector, S., Sullivan, B.,

Wang, Y., Wirick, S., Zhang, H., 2000. Illumination for coherent

soft X-ray applications: the new X1A beamline at the NSLS.

Journal of Synchrotron Radiation 7, 395–404.