Molecular characterization of dissolved organic matter

organic matter which acts as carbon sink in the ocean. ... inorganic salts ... able for detection of ionized organic compounds, because of its ultrahi...

0 downloads 145 Views 557KB Size
Anal Bioanal Chem (2013) 405:109–124 DOI 10.1007/s00216-012-6363-2

REVIEW

Molecular characterization of dissolved organic matter (DOM): a critical review Antonio Nebbioso & Alessandro Piccolo

Received: 30 June 2012 / Revised: 12 August 2012 / Accepted: 14 August 2012 / Published online: 11 September 2012 # Springer-Verlag 2012

Abstract Advances in water chemistry in the last decade have improved our knowledge about the genesis, composition, and structure of dissolved organic matter, and its effect on the environment. Improvements in analytical technology, for example Fourier-transform ion cyclotron (FT-ICR) mass spectrometry (MS), homo and hetero-correlated multidimensional nuclear magnetic resonance (NMR) spectroscopy, and excitation emission matrix fluorimetry (EEMF) with parallel factor (PARAFAC) analysis for UV–fluorescence spectroscopy have resulted in these advances. Improved purification methods, for example ultrafiltration and reverse osmosis, have enabled facile desalting and concentration of freshly collected DOM samples, thereby complementing the analytical process. Although its molecular weight (MW) remains undefined, DOM is described as a complex mixture of low-MW substances and larger-MW biomolecules, for example proteins, polysaccharides, and exocellular macromolecules. There is a general consensus that marine DOM originates from terrestrial and marine sources. A combination of diagenetic and microbial processes contributes to its origin, resulting in refractory organic matter which acts as carbon sink in the ocean. Ocean DOM is derived partially from humified products A. Nebbioso (*) : A. Piccolo Centro Interdipartimentale di Risonanza Magnetica Nucleare per l’Ambiente, l’Agroalimentare e i Nuovi Materiali (CERMANU), Università degli Studi di Napoli Federico II, via Università 100, 80055 Portici, Italy e-mail: [email protected] A. Piccolo Dipartimento di Scienze del Suolo, della Pianta, Dell’Ambiente e delle Produzioni Animali (DISSPAPA), Università degli Studi di Napoli Federico II, via Università 100, 80055 Portici, Italy

of plants decay dissolved in fresh water and transported to the ocean, and partially from proteinaceous and polysaccharide material from phytoplankton metabolism, which undergoes in-situ microbial processes, becoming refractory. Some of the DOM interacts with radiation and is, therefore, defined as chromophoric DOM (CDOM). CDOM is classified as terrestrial, marine, anthropogenic, or mixed, depending on its origin. Terrestrial CDOM reaches the oceans via estuaries, whereas autochthonous CDOM is formed in sea water by microbial activity; anthropogenic CDOM is a result of human activity. CDOM also affects the quality of water, by shielding it from solar radiation, and constitutes a carbon sink pool. Evidence in support of the hypothesis that part of marine DOM is of terrestrial origin, being the result of a long-term carbon sedimentation, has been obtained from several studies discussed herein. Keywords Dissolved organic matter . Carbon cycle . Humic substances . Water . Dissolved organic carbon . Natural products

Environmental relevance of DOM Natural organic matter (NOM) is a product of plant and animal tissue decay, and, together with the biota, is of pivotal importance in the global carbon cycle. NOM is found in soil, sediments, and natural water and its biological mineralization contributes, with anthropogenic emissions, to increasing the carbon dioxide content of the atmosphere. Any control of the environmental processes of NOM transformation would become possible only if additional knowledge about its molecular composition were accumulated [1]. Molecular characterization of NOM has therefore become a primary research objective in environmental and

110

ecological chemistry [2]. The recent development of instrumental techniques, for example NMR spectroscopy, hyphenated mass spectrometry, and X-ray spectroscopic methods, has greatly enhanced our ability to detect and characterize single organic compounds and, occasionally, homogeneous mixtures. However, because of the great complexity and heterogeneous composition of NOM, its characterization remains a challenge. Dissolved organic matter (DOM) cannot be correctly regarded as a chemical solution; it is, rather, a very fine colloidal suspension. The distinction between particulate organic matter (POM) and DOM is only operationally defined [3]: DOM is assumed to pass through a 0.45-μm filter pore whereas POM is blocked (Fig. 1). DOM and its composition are important in natural-water ecosystems because of the number of processes in which it becomes involved. DOM acts as a strong chelating agent for metals, thus affecting their solubility, transport, and toxicity [4]. It is fundamentally involved in the transport of organic pollutants [5], formation of colloidal particles (and affects their surface area) [6], aqueous photochemical reactions [7], nutrients cycling and availability [8], pH-buffering [9], and the distribution of ions between aqueous and solid phases [10]. DOM and POM are also important sources of energy in river-water ecosystems [11]. The most abundant pool of DOM is that in the oceans, also known as marine DOM. This is generally believed to be the decay products of phytoplankton and consists of 25–50 % proteins, 5–25 % lipids, and up to 40 % carbohydrates [12–14]. However, DOM from terrestrial sources, for example biomass, plant litter, and soil organic matter, also reaches ocean waters by transport from rivers, lakes, glaciers, and other natural sources. Such transfer of terrestrial carbon is an important link in the global carbon cycle. Estimates of total global transport of organic carbon to the oceans range from 0.4 to 0.9×1015 g−1 year−1 [15–17]. Fig. 1 Size range of particulate (POM) and dissolved organic matter (DOM) and organic compounds in natural waters. AA, amino acids; CHO, carbohydrates; CPOM, coarse particulate organic matter; FA, fatty acids, FPOM, fine particulate organic matter; HA, hydrophilic acids; HC, hydrocarbon; VPOM, very fine particulate organic matter. Adapted from Ref. [3]

A. Nebbioso, A. Piccolo

Genesis of DOM Terrestrial DOM is the result of biological degradation and progressive concentration of organic compounds particularly resistant to degradation. Degradation of vascular plants furnishes DOM containing approximately 10 % proteins, 30–50 % carbohydrates (mainly cellulose), some lipids concentrated in the roots and leaf cuticles [18], 15–25 % lignin [19], and other biomacromolecules. Moreover, it seems there is a correlation between environmental conditions and type of terrestrial DOM derived from soil [20]. Lignin, an important tracer for terrestrial OM [21], consists of repeating phenylpropanoid units which are randomly linked to each other by ether and carbon–carbon bonds. This branched macromolecular network confers on lignin chemical stability that is assumed to resist extensive microbial degradation [22, 23]. Proteins and carbohydrates are, in contrast, biolabile compounds, because of the susceptibility of peptide and glycosidic bonds to hydrolysis by a variety of enzymes [24–26]. However, even labile molecules are preserved in marine sediments under specific conditions, e.g., when they are protected in the frustules of marine organisms [27] and within recalcitrant complex structures formed with DOM [28–30]. Such highly complex OM is also referred to as molecularly uncharacterized (MU-OM), because its composition can be hardly resolved at a molecular level by conventional analytical techniques [31, 32]. MU-OM may account for up to 80 % of marine sediments [31, 33]. Highly unsaturated compounds are also present in DOM of natural waters as strongly resistant organic components, and are also referred to as black carbon (BC) or the product of either incomplete burning of biomass or recycled kerogen. Even bacteria-derived OM, for example recalcitrant peptidoglycan from cell walls, can be found in DOM [34, 35].

Molecular characterization of dissolved organic matter

As a result of this heterogeneous and complex chemical nature, research on molecular characterization of DOM is conducted widely by use of several different approaches and strategies. The purpose of this review is, thus, to evaluate the most recent advances in analytical methods for characterization of DOM in natural water bodies.

Characterization of DOM State of the art of DOM isolation and purification DOM in the environment is found, with rare exceptions (e.g. Suwannee River DOM), at extremely low concentrations (0.5–1.0 mgL−1 in the oceans); inorganic salts exceed this value by several orders of magnitude. Therefore, specifically designed techniques are generally used to increase the concentration of DOM and to remove salts. Retention-based methods involving XAD resins [36] or C18 stationary phases [37] have been extensively investigated, only to reveal that a variable but substantial part of DOM is lost because of incomplete retention [38]. Ultrafiltration (UF) [39–42] has been used to remove large volumes of water through membrane pores which are restrictive for DOM but not for water molecules and small ions. It is a formidable desalting method, that is also used to purify water from excessive DOM content [43]. UF is also affected by incomplete recovery of organic carbon, but to a lesser extent. Reverse osmosis (RO) is an improvement of UF. RO operates similarly to UF by allowing water through membranes with a restrictive cut-off for DOM. However, in RO the solution is forced by a pressure gradient to flow against osmotic flow (hence reverse osmosis). Different and more restrictive membranes are used for RO than for UF, resulting in greater retention of ions in DOM samples. Such retention consists mainly of ions derived from H4SiO4 and H2SO4. Development of methods such as RO coupled with electrodialysis [44] and pulsed electrodialysis [45] was, in fact, intended to minimize these inorganic impurities. These processes are now rapidly becoming conventional for DOM purification [46] and are in constant development, optimization, and standardization [47]. Treatment of sediments to enable solid-state crosspolarization magic-angle spinning (CPMAS) NMR spectroscopy of their labile organic matter is routinely based on acid washing and chemical digestion with HCl and/or HF [48]. However, adoption of sensitive instrumental methods for DOM analysis, for example ultrahigh-resolution mass spectrometry, may prevent such sample pretreatment [49], thus limiting the formation of artefacts resulting from the processes of DOM concentration and pH modification [50]. A promising application for DOM separation seems to

111

be the development of carbon nanotubes as solid-phase extraction (SPE) stationary phases which exploit the affinity of nano-structures for organic compounds in solution [51]. Such stationary phases have, however, so far resulted in limited recovery that ranges between 30 and 80 %, depending on DOM type, and specific selectivity for lowmolecular-weight DOM fractions. There is increasing interest in DOM from other sources, for example atmospheric aerosols [52]. Discovery, characterization and quantitative assessment of such alternative sources is critical for understanding the relevance of DOM in global carbon dynamics. This, however, is a challenging task, because sampling and analysis of fog-water-derived DOM is more difficult than for surface or ground water. Interest in this subject is increasing, and, recently, a thorough review of analytical methods for airborne DOM aerosols was published by Duarte and Duarte [53]. These authors emphasized the environmental significance of this underestimated source of organic carbon and discussed the inherent use of NMR, IR, and MS methods for its analysis.

Main analytical methods for DOM Several problems may affect the analysis of NOM—difficulty of complete dissolution [54], lack of proper molecular separation [54], extreme heterogeneity of samples, mutual interference from different classes of compound, tendency of association in complex superstructures [55]. DOM is no exception. However, substantial analytical improvement has recently resulted from the introduction of Fourier-transform ion cyclotron (FT-ICR) mass spectrometry (MS). This is the most advanced instrumentation available for detection of ionized organic compounds, because of its ultrahigh resolution and because it is usually coupled with non-destructive ion sources, for example electrospray ionization (ESI). The impact of FT-ICR MS on NOM analysis has been outstanding [56], and it has rapidly become one of the first choices in DOM studies [57]. A consequence of this breakthrough, the number of masses characterized in DOM analysis has increased to such an extent that results can be efficiently reported only in simplified diagrams, as plots sorting m/z ratios by homologous series (Kendrick) and by O/C and H/C ratios (Van Krevelen) [58]. An aromaticity (AI) index has also been proposed [59], to highlight aromatic (>0.5) or condensed aromatic (>0.67) empirical formulae. FT-ICR MS has substantially improved the capacity to identify DOM molecules. Two specific applications of FTICR MS are worth mentioning—detection of either hydrogen-deficient aromatic compounds or nitrogenous organic molecules. In fact, FT-ICR MS high resolution scanning is effective for detection of:

112

1. ions with H/C ratios <0.9 and <0.25, which imply a large number of double bond equivalents (DBE) and thus restricts possible structures to condensed aromatic rings; and 2. ions with heteroatoms such as nitrogen. Analysis of condensed aromatic rings in sea water near the Antarctica continent [60] and the Southern Ocean between South Africa and Antarctica [61], revealed 224 different empirical formulae with large DBE. By use of basic organic chemistry, the minimum and maximum number of rings for the MS empirical formulae observed can be calculated; it was found that only structures with five to eight condensed rings were plausible [60]. Moreover, analysis of very hydrophilic (HPI) fractions from sea water of the Atlantic Ocean revealed the presence of nitrogencontaining compounds. The empirical formulae showed this ocean DOM contained compounds with up to three nitrogen atoms [62]. Analysis of sediments from Mangrove Lake, Bermuda, and Mud Lake, Florida, has confirmed the presence of homologous series of CHNO compounds which are likely to contain alkyl amide structures [63]. Unfortunately FT-ICR MS is not easily hyphenated, and coupling with liquid chromatography (LC) is only possible in the off-line mode [64]. Off-line FT-ICR MS coupled with HPLC is widely applied and yields outstanding results [65]. Hyphenated methods based on online chromatographic configurations are only possible with other types of MS. High-performance direct and reversed-phase adsorption [66] and size-exclusion liquid chromatography [67] interface well with negative-mode ESI-MS, which remains the most frequent option for MS analysis of dissolved humic substances. The ESI source may be coupled with ion trap [67], quadrupole TOF [66], or triple-quadrupole [68] MS, depending on the type of information required. In fact, there is a general consensus that ion-trap mass spectrometers have superior sensitivity whereas quadrupole instruments, especially those with the triple configuration, have better mass accuracy. Triple quadrupoles are also generally capable of high-resolution analysis. Other high-resolution MS, for example isotope ratio MS, is often used to measure stable isotopes, e.g. to investigate the 13C signature of DOM, by breaking down analytes and measuring atomic masses [69]. High-performance size-exclusion chromatography (HPLC) coupled with either the traditional UV detection [70] or combined multi-detectors [71] is also used for qualitative and quantitative evaluation of DOM. HPLC has also been coupled with NMR spectrometers in on-line mode, thus enabling investigation of the structure of separated DOM molecules [72]. Finally, a very specific HPLC technique named HPAEC-PAD (high-performance anion-exchange chromatography with pulsed amperometric detection),

A. Nebbioso, A. Piccolo

applied within a rigorous analytical procedure, has enabled investigation the complexity of DOM polysaccharides [73]. These MS methods have become conventional for DOM analysis because dissolved organic molecules are readily ionized, especially in negative-ion mode. However, many problems remain unsolved. First, non-ionizable compounds cannot be characterized by MS. Second, ionization of terrestrial HS or DOM is a complex phenomenon prone to irreproducible results, because of molecular interferences as a result of complex inhomogeneous, supramolecular associations [68, 74, 75]. These limitations thus prevent reliance on MS methods alone to achieve structural identification of DOM molecules. NMR spectroscopy has become fundamental in complementing DOM characterization, because ionization is not required for the NMR excitation and detection of 1H and 13 C nuclides. Both solution and solid-state NMR spectroscopy are well established tools in the environmental sciences [48, 76–78]. Molecular structural information has been obtained from conventional mono, bi, and tri-dimensional NMR spectra [79–81], and information about molecular diffusion properties and stacking arrangements of organic matter is obtained by use of DOSY [82] and relaxation time (T1H and T1ρH) techniques [83, 84]. Moreover, the highresolution magic-angle spinning (HR-MAS) technique, which is applied to semi-solid samples, is increasingly being used to characterize colloidal humic matter and DOM [85].

Ultraviolet and fluorescence spectroscopy of DOM Because of several limitations of even the most advanced NMR and MS techniques in DOM analysis, there is an increasing interest in advanced applications of UV and fluorescence spectroscopy, because of their qualitative and quantitative reliability. WETStar instrumentation is an example of modern fluorimetry applied to DOM[86]. The main advantage of this approach over use of traditional fluorimeters is the better accuracy achieved for unfiltered samples. This advantage enables in-situ analysis of water samples without preliminary purification steps [86]. In-situ DOM measurements with WETStar fluorimeters are comparable with values obtained on filtered samples by use of traditional fluorimeters. Another innovative application of spectrophotometry for DOM analysis is combination of 3D spectrofluorimetry with HPLC and capillary electrophoresis [87]. This combined technique is capable of differentiating marine from fresh-water DOM, thereby extending its potential to profiling of water samples. DOM quantification methods have been also developed for large-scale monitoring by remote sensing. This approach is simpler than methods requiring direct sampling; it is, therefore, attracting much interest. A remarkable application of this approach [88] uses a combination of:

Molecular characterization of dissolved organic matter

1. below-sea-surface spectral downward scalar irradiance to calculate a radiative transfer model (STAR) corrected for clouds by use of TOMS UV reflectivity; 2. surface-ocean spectral diffuse attenuation coefficients and absorption coefficients for chromophoric dissolved organic matter retrieved from SeaWiFS ocean colour by use of the SeaUV/SeaUVc algorithms; and 3. spectral apparent quantum yield for the inherent photochemical reactions. This approach enabled resolution of the contribution of surface chromophoric DOM (CDOM) from total DOM. Unfortunately, one disadvantage is a lack of detailed structural characterization of the DOM, so applicability is confined to large-scale environmental studies, for example CO2 management. Conversely, excitation emission matrix fluorimetry (EEMF), although unsuitable for remote sensing, is a useful tool for DOM molecular characterization, because of the specific spectrum obtained from each known fluorophore. Nevertheless, EEMF may be compatible with large-scale analysis when combined with a network of sampling stations, for example that described by Singh et al. [89]. In fact, by coupling EEMF with parallel factor (PARAFAC) analysis, scientists have also been able to discriminate qualitative differences in CDOM throughout the Barataria Basin [89]. The combination of EEMF with PARAFAC analysis has been attracting attention as a reliable method, and literature descriptions of its application are increasing [90, 91].

Marine DOM Size distribution and molecular structure There is no model for characterization of marine DOM. DOM arising from natural bodies and stored in the ocean is a major carbon sink, and redistributes DOM in regions where it is less available, for example the oceans depths, glaciers, pores, or polar caps. Marine DOM is also formed as a product of the biochemistry of life forms, for example diatoms, bacteria, algae, or microfauna. These may either actively release organic matter or undergo natural decay after cell death; their metabolic diversity is one of the factors resulting in the heterogeneity of DOM. Unlike soil organic matter (SOM), which is deposited as solid, marine DOM is subjected to much faster dynamics because it is dissolved in a water mass. DOM is therefore transferred or transformed in larger amounts than SOM, and its availability for enzyme activity, oxidation, or metal complexation is significantly larger than that of SOM. Moreover, DOM is usually found at lower concentrations and greater ionic strength (salinity) than SOM, and both

113

conditions affect the mutual association capacity of DOM. These factors contribute to differentiation of natural organic matter in its marine and terrestrial form. Unfortunately, both SOM and marine DOM suffer from the same analytical limitations with regard to determination of molecular weight and structure. The molecular weight (MW) of marine DOM cannot be defined simply. The current general consensus is that lowmolecular-weight substances coexist with increasingly larger biomolecules, for example proteins, polysaccharides, and exocellular macromolecules [92], and with inorganic ions, which stabilize intermolecular association by formation of complexes. Such a heterogeneous mixture of substances is involved in several mutual interactions and associations, which result in a wide range of apparent molecular weights. DOM scientists operationally distinguish low molecular weight (LMW <1 kDa) from high molecular weight (HMW >1 kDa) fractions when measuring the MW of DOM associations. The current view of DOM structure required an interdisciplinary approach, varying from marine geochemistry to microbiology and polymer physics. This was because of the three-dimensional DOM network that arises from the well known interactions between small organic compounds, biomacromolecules, and microgels [93]. Microgels are a product of the transformation of bacterial or diatom exopolymeric substances [94]. They acquire progressively different physicochemical properties and their particle size changes from the original size [93]. It is known that part of DOM originates from humic matter, and thus shares its properties, including self-association in supramolecular structures [55, 68]. Thus, one can assume that the selfassociated molecules in DOM may also interact with microgels, and this may explain the observed changes in the chemical properties of the latter. The occurrence of interactions between DOM small molecules and polymers was revealed experimentally by in-vitro experiments with amphiphilic exopolymers of Sagittula stellata [30]. Cations seem to be of pivotal importance in the formation of microgels and to their stability, probably because of development of ion bridges across branches of polymers [30, 95]. Amphiphilicity has been observed in supramolecular associations of small DOM molecules of terrestrial origin, suggesting that interactions between suprastructures and microgels are plausible [29]. Moreover, coastal mixing zones have been indicated as the most likely environment for these aggregation phenomena. Transition to saline water has been suggested as a possible cause of the contraction of such associations, which may facilitate the intertwining with microgels in solution. The hypothesis of coexistence of autochthonous and terrestrial substances in marine DOM was confirmed by studies on the HMW fraction of marine DOM collected

114

from the mid-Atlantic Bight coastal region, in which an acylpolysaccharide (APS) and other marine and terrestrial humic-like compounds were identified [40]. On the basis of analysis of its spatial distribution, the authors remarked that APS is most abundant in surface waters, being synthesized there, but rapidly decreases in deep waters. This suggests that the marine DOM transported as a microgel may be a significant component of the oceanic carbon pool. A more recent experiment compared 1H NMR spectra and analytical chemical profiles of DOM extracted from a diatom after microbial degradation with those of an APScontaining DOM obtained from sea water. It was found that both DOM had relatively high stability against microbial degradation of hydrolysable sugars during incubation for 40 days [96]. More evidence of bacterial polymers in marine DOM arises from analysis of samples from the Pacific, Atlantic, and northern oceans, which revealed the presence of beta hydroxylated, branched, and other aliphatic acids of typical bacterial origin [97]. This further suggests that diverse procaryotic and eucaryotic species contribute with their biomacromolecules to marine DOM. In this case, it is interesting that both free and ester-bound hydroxy acids were found, implying that the extent of degradation of bacterial membranes by hydrolysis is variable, and results in products with a range of molecular weights. However, DOM size fractions are sometimes reported as strongly related to their sugar chemical composition [92], with amino, deoxy, and methylated sugars being more abundant in HMW fractions and hexose in LMW [98]. This may indicate that modified sugars are predominantly present as building blocks of large polymeric structures, whereas hexose sugars are differently associated. In fact, periodate oxidation of HMW DOM yielded 6-deoxy and methylmonosaccharides, and, because the consumption of oxidant was greater than for a linear polysaccharide, a branched structure was suggested for the HMW DOM [99]. It is plausible that additional information about diagenetic processes may contribute to elucidation of chemical implications of marine DOM. In fact, on the basis of results from analysis of neutral and amino sugars, it is known that diagenetic transformation affects HMW more than LMW DOM [100], thus implying that microgels probably undergo the same fate. Further evidence comes from analysis of DOM harvested in the Hawaii sea water [77], which revealed a decrease in the relative contribution of carbohydrates and a concomitant increase of lipids with increasing depth. Other biomacromolecules also found in DOM are proteins, which are of different significance to marine biochemistry [92, 101, 102]. Degradation of proteins occurs naturally in the environment and produces progressively smaller fragments down to single amino acids. The survival of proteins as refractory organic matter is a highly controversial subject and the mechanism underlying the partial

A. Nebbioso, A. Piccolo

protection of proteinaceous materials is not yet clear. It is possible that peptides have a specific function in the formation of DOM associations, for example being part of a supramolecular structure or present simply hosted in unbound forms. An experimental model with bovine serum albumin incubated in sea water suggested that the fragments formed contained fewer than 40 amino acids, thus maintaining an intermediate MW between that of polymers discussed above and that of single molecules [25]. The idea that proteins in marine DOM originate mainly from bacterial membranes [101] was recently challenged [102], because of evidence based on sequences peculiar to typical cytosolic enzymes. After cell death, it is plausible that most of the protein pool is metabolized by microbes, and that only a fraction is protected as recalcitrant peptides by either physical protection or selective preservation [28]. However, studies on the interaction of proteinaceous materials with microgel polymers has produced controversial results. Microscopy of simulated marine aerosols has shown separated clusters of stained proteinaceous material and exopolymers (Fig. 2), suggesting no reciprocal interaction among these classes of compounds [103]. In contrast, characterization of microgels by fluorescence and proteomic methods achieved detection of peptidases, proteases, and ATPases [104]. Moreover, evidence for the presence of whole bacterial enzymes indicates other possible means of stabilization of proteins and peptidoglycans, for example the formation of Pseudomonas-type vesicles [35]. Characterization of proteins in marine DOM has also been attempted by marine proteomics [105]. In these methods peptides are sequenced by tandem mass spectrometry. This “cataloguing” of proteins has improved our understanding of: 1. evolution and cycling of carbon pools within the ocean; and 2. how different proteomes adapt to different conditions, thereby providing insight into which proteins are preserved in the environment. Macromolecules and molecules, though self-associating in large suprastructures, are mostly constituents of HMW DOM fractions, whereas other low-MW species are predominantly found in the LMW fraction of DOM. The classes of compounds which have been found in DOM include lipids [97, 106], BC [60, 61], and carboxyl-rich alicyclic molecules (CRAM) [107, 108], all supposedly of terrestrial thermogenic origin. There may also be smaller unbound compounds in DOM, for example those found in soil organic matter. However, whereas SOM requires extensive fractionation to yield mass spectra from which meaningful structures can be suggested [2], this requirement may be less strict for DOM. At this stage, available data describe the structure of marine DOM as a system containing constituents with

Molecular characterization of dissolved organic matter

115

Fig. 2 A, B, Proteinaceous particles and C, D, transparent exopolymeric particles (TEPs) that contain polysaccharides in simulated aerosol (spray). Adapted from Ref. [103]

polydispersed MW. The HMW DOM fraction contains exopolymeric materials of autochthonous origin which form stable microgels with polysaccharides and lipids, and coexist, or even interact with, proteinaceous matter and with terrestrial-derived humic-like supramolecular structures. The LMW DOM fraction is composed of terrestrial material originating from the decay of biomacromolecules, the degree of mutual binding of which is not yet clarified. Chromophoric DOM (CDOM) in marine environments CDOM is the fraction of organic matter capable of interacting with light and adsorbing its energy. The importance of this fraction is connected with the marine biota that need light for photosynthesis and it is established as a primary factor that affects biological cycles and phytoplankton life [109]. For DOM to have chromophoric properties it must contain unsaturated and conjugated groups, for example aromatic or quinoid structures. The conventional technique for investigating CDOM quality and dynamics is excitation emission matrix fluorescence (EEMF) spectroscopy combined with parallel factor (PARAFAC) analysis. CDOM is classified as terrestrial, marine, anthropogenic, or mixed, depending on its origin. Terrestrial DOM is transferred to sea water via estuaries, as suggested by the correlatio ns found be tween river output and C DOM concentration [110]. Conversely, autochthonous CDOM is directly formed in sea water, as indicated by its more homogeneous marine distribution [111]. Recently evidence that submarine hydrothermal vents release CDOM has been obtained, revealing a new source for autochthonous CDOM

in sea water [112]. Terrestrial CDOM discharged to sea water also includes anthropogenic CDOM; unlike natural CDOM, the anthropogenic material is likely to cause environmental stress because it contains PCBs [113]. It is generally agreed that normal CDOM fluorescence spectra contain peaks from the main chemical constituents, for example humic-like and protein-like materials. Modern fluorescence spectroscopy combined with PARAFAC easily differentiates these two types of CDOM in ocean waters [114], and reveals the dilution which estuarine CDOM undergoes when mixing with oceans. Numerous studies have traced CDOM flows spatially and temporally, to characterize DOM dynamics along mixing zones and across water columns. The approach has been used in a variety of geographical areas and this type of study is growing in number. The results tend to agree, irrespective of geographical area [110, 115–122]. These results further confirm that CDOM is of pivotal importance in global carbon cycles and more efforts should be made to characterize its dynamics. Moreover, CDOM in ocean ecosystems also acts as a carbon sink, because its interaction with light causes partial photochemical mineralization, leading to dissolved inorganic carbon (DIC). This carbon, which is mostly produced off-shore, is not released into the atmosphere and hence does not contribute to the greenhouse effect [123]. Photochemical oxidation products of CDOM are not limited to DIC. In fact, carbon monoxide (CO), which is of fundamental and often underestimated importance in carbon cycling, is also formed at different depths and may be measured by means of the spectroscopic methods discussed herein [88].

116

Molecular characterization of marine DOM Molecular characterization of DOM is important to better understand how DOM molecules are organized and mutually related. Moreover, detailed elucidation of the structure of the components of DOM would enable recognition of biomarkers present in specific geographical areas [124]. Study of biomarkers and their association with well characterized DOM has great potential for understanding the environmental dynamics of major carbon reservoirs by measuring. Advanced analytical instrumentation nowadays enables characterization of single compounds from highly complex molecular mixtures. Such a molecular identification is possible for purified, desalted samples even without pretreatment, thereby preserving the molecular structures as they are found in the environment. Identification of polysaccharides and proteins still requires hydrolytic pretreatment before instrumental determination of the structure of single units. Pretreatment is followed by MS and NMR molecular characterization to define the structures of the sugars and lipids composing marine DOM [40, 125, 126]. However, a recent study on mucilage polysaccharide used a direct analytical approach without preliminary hydrolysis and achieved molecular characterization of the sequence of monosaccharides by use of a tandem MS technique [127]. Fig. 3 Molecular structures and alternative isomers proposed for each (CH2)n homologous series of condensed polyaromatic compounds. It is assumed that a maximum number of oxygen atoms is present in carboxyl functional groups. From Ref. [60]

A. Nebbioso, A. Piccolo

Fourier transform ion cyclotron resonance (FT-ICR) mass spectrometry is also becoming the instrument of choice for molecular characterization of marine DOM, because of the outstanding quality of results obtained by use of this mass spectrometric technique. Some studies using FT-ICR MS have focused on one particular class of molecule whereas others have preferred to attempt comprehensive sample characterization. An example of the first approach is the identification of hydroxy fatty acids, as lipid biomarkers, in ultrafiltered organic matter from ocean waters to evaluate potential inputs from bacterial membrane lipids [97]. The second approach consists in comprehensive characterization of all the empirical formulae found by FT-ICR MS accurate mass measurements [60, 61, 128, 129]. Structures characterized with this method are reported in Fig. 3; substances with multiple condensed rings were attributed to BC, as a product of the partial oxidation of DOM [60]. FT-ICR MS may also be used as a complementary and confirmative technique for characterization of natural organic matter. This approach has been used in studies on partial products of DOM chemical oxidation, in which there is still significant interest. Solid-state 13C NMR spectroscopy has provided experimental evidence of the existence of structural products of chemical oxidation in DOM [32]. Later, it was shown [130] that oxidized products similar to black carbon,

Molecular characterization of dissolved organic matter

including perylene and benzopyrene derivatives with a variety of substituents, were present in DOM. Their detection was also accomplished by FT-ICR MS in DOM samples from very different geographical areas [60, 61, 128, 129], suggesting that the genesis and molecular structure of black carbon products in DOM is consistent throughout the oceans. Hence, FTICR MS has been proved to suitable for molecular characterization of DOM constituents, although its use in combination with other techniques, for example NMR spectroscopy, should be still recommended. High-resolution mass spectrometry of complex samples can provide a large number of empirical formulae from DOM samples, including those of lignins, tannins, amines, amides, carboxyl-rich alicyclic molecules (CRAMs), lipids, amino sugars, and carbohydrates (Fig. 4). The inherent complexity of the analytical response obtained must be simplified by use of Kendrick and Van Krevelen (KvK) plots [65, 107]. However, with only empirical formulae, although systematized in KvK plots, information about the chemical structure of DOM components remains rather poor. There is, thus, growing pressure to develop innovative scanning methods involving tandem MS techniques. In fact, ion fragmentation and daughter ion scans achieved by secondary tandem mass spectrometry may enable the required structural elucidation of the entire molecular complexity of DOM. Application to DOM of the most recent advances in highresolution tandem mass spectrometry have improved our understanding of environmental carbon cycling. In fact, despite variability in oceanic environments, because of different depth, geographic position, proximity of estuaries, and anthropogenic modifications, results obtained for DOM analysis have been quite consistent [110, 115–122]. As a result of such advanced studies, there is a general consensus that marine DOM: 1. is a combination of components of terrestrial and marine origin, which can be easily differentiated according to their genesis;

Fig. 4 Molecular assignment of DOM components from Chesapeake Bay, by use of Van Krevelen diagrams based on FT-ICR MS experimental data. From Ref. [107]

117

2. is formed by multiple metabolic pathways which originate in terrestrial and marine environments via a combination of natural processes (e.g. forest fires, plant decay, etc.), microbial activity, and environmental phenomena (e.g. solar irradiation, oxidation); and 3. may have a long residence time, because of the refractory nature of its components, and is, thus, an ocean carbon sink.

Understanding fresh-water DOM Size distribution and molecular architecture Although fresh-water and oceanic DOM share some characteristics, they also differ substantially, because of the different environments in which they are formed, transformed, accumulated, or transported. A specific difference is that fresh-water DOM normally flows into the oceanic pool whereas the opposite flux is very rare and may only occur during natural disasters. Thus, marine DOM is found to have characteristics similar to those of fresh-water DOM, because of to their common terrestrial origin. Numerous studies have been conducted on estuarine mixing areas, and have shown that both types of DOM are spatially distributed by streams and tides [131]. Because discrimination of these DOM types can be challenging, here we emphasize the differences between terrestrial DOM molecules before and after contact with the marine environment. Fresh-water DOM is derived from terrestrial soil organic matter (SOM) that underwent specific transformations to increase its affinity for an aqueous environment. SOM is traditionally and operationally divided in three pools: fulvic acids, humic acids, and humin, according to their solubility in acids and alkali. Contemporary understanding regards SOM as an aggregate of numerous heterogeneous molecules of relatively small molecular mass held together by weak non-covalent bonds [2, 55, 132, 133]. There is experimental evidence to show that DOM is also arranged in similar supramolecular associations [68]. Because of the similarity with terrestrial SOM, genesis of fresh-water DOM is therefore strongly related to terrestrial vegetation. This is suggested by measurements of δ13C of river sediments and autochthonous plants, which were found to be strongly correlated [134]. In this work, analytes of plankton origin were found in both river and estuary DOM, although in smaller amounts than in marine DOM samples. Transportation of riverine DOM to the marine environment was also related to complex transformations. Combination of mass spectrometric characterization with fluorescence analysis of CDOM distribution indicated that DOM apparent mass decreased as salinity increased [135].

118

The composition of fresh-water DOM is believed to depend on the transformation of plant compounds into humic-like substances. Investigation of river and lake DOM composition with different techniques supports the hypothesis of plant genesis. In particular, NMR spectroscopy using two dimensional long-range correlation techniques has enabled characterization of compounds directly related to decay of terpenes [72], for example CRAM and material derived from linear terpenoids (MDLT), in agreement with previous DOM literature [136]. The same NMR techniques also enabled detection of hexopolysaccharides and aromatic structures, possibly of lignin origin. ESI FT-ICR MS investigation of river DOM revealed evidence of hydrogen-deficient compounds assigned to black carbon (BC), thereby confirming the hypothesis of transport of terrestrial material in marine DOM [128]. Once formed, these black carbon substances have a dynamic distribution in natural bodies. Recent FT-ICR MS studies on lacustrine DOM from Lake Superior emphasized differentiation among fresh-water areas such as swamp, creek, and river, depending on spatial and temporal distribution [137]. Furthermore, larger amounts of lignin-like compounds were reported in this lake than in its tributaries. Lignins are regarded as refractory substances and, in fact, their contribution to total DOM accumulated along the riverine paths, owing to slower mineralization, and was found to reach the greatest concentration in the oceans [137]. In this scenario, the chemical properties of compounds are bound to affect their distribution in water. Although the contribution of lipids is expected to be limited because of limited aqueous solubility, they are, nevertheless, found in lacustrine DOM [138]. This finding inevitable leads to the inference that the supramolecular structure of DOM enhances the solubility of specific hydrophobic molecules by forming complex associations with them. The dynamics of hydrophobic compounds in natural bodies is complex, because they are also important components of particulate organic matter (POM). However, although it is well established that hydrophobic OM is an abundant component of POM and sedimentary matter [139], it is not yet clear whether hydrophobic DOM and POM are related to each other. Because of the further complexity introduced by the tendency of DOM molecules in solution to associate, assessment of the size of DOM particles is not straightforward. Interestingly, whereas mass spectrometry of DOM indicates molecular masses lower than 1000 Da for most compounds [68, 128, 137], size-exclusion chromatography (SEC) profiles of the same sample suggests a much larger hydrodynamic volume. This discrepancy confirms that single molecules are prone to spontaneous association [55, 132], but more evidence should be gathered on how this structure is organized.

A. Nebbioso, A. Piccolo

The findings reported here have led to a plausible structure for fresh-water DOM, which is currently described as an aggregation in spontaneous self-associated superstructures formed by plant-derived products of natural decay, for example lipids, amino sugars, sugars, CRAM and other terpene derivatives, aromatic condensed structures (BC), and lignin-derived compounds. Chromophoric DOM in the fresh-water environment Interaction with solar radiation is a fundamental property of fresh-water DOM and is very relevant in fresh-water environmental interactions. The formation of chromophoric DOM (CDOM) is still debated, but experimental evidence over the last decade suggests that organic matter derived from phytoplankton, initially colourless, is processed by microbial flora into fluorescent DOM. In fact, CDOM isolated after incubation of algae was found to grow concomitantly with microbial mass [140]. Further evidence of the involvement of phytoplankton in the formation of lacustrine DOM came from quantitative assessment of average and daily rates of in-situ production [141]. It has also recently been reported that fluorescence absorption peaks for humic and fulvic acids increased proportionally with the amount of DOM. These acids are probably formed under terrestrial conditions and then transported in natural water bodies, thereby affecting fluorescence response [142]. The different chemical composition of autochthonous and humic DOM in fresh water necessitated more systematic description. Hence, the humification index (HIX) and the index of recent autochthonous contribution (BIX) were usefully introduced [143]. The chemical origin of the colloidal properties of CDOM have also been investigated by flow-field flow fractionation [144], assuming differentiation between humic and/or fulvic-like and protein-like compounds. Whereas the origin of the latter was attributed to fresh-water autochthonous life, the sources of the former materials are believed to be terrestrial. Furthermore, it seems there is a strict correlation between the size fraction and the composition of the colloidal phase, with protein-like materials occurring primarily in the smaller size fraction and humic-type materials in the larger [145]. All these findings suggest that the genesis of freshwater DOM is similar to that of marine DOM, because both contain compounds with well differentiated optical properties. In fact, differentiation of riverine from marine DOM is also possible in water sampled in estuarine mixing zones [131]. In this work, two rivers flowing into the same gulf supplied DOM with different optical behaviour, thereby affecting the spatial distribution of the resulting marine DOM. In agreement with this finding, an interesting experiment has further clarified the effect of salinity and dilution on DOM fluorescence spectra, by revealing that changes in salinity has no significant effect on fluorescence lifetime

Molecular characterization of dissolved organic matter

[146]. However, the transition in other optical characteristics from riverine to marine DOM has been related to increased salinity [147], thereby indicating that ionic strength is of pivotal importance in the association of DOM molecules. It can be concluded that although salt concentration seems to affect DOM and its optical properties, fluorescence lifetime is not affected. As a result of developments in fluorescence methods, recent research has successfully differentiated fresh-water DOM from several geographical sites [89, 90, 148–150]. It is currently believed that CDOM degradation simultaneously involves microbial metabolism, which seems to have a preference for allochthonous CDOM [151], photodegradation [152–154], and, to a lesser extent, adsorption on suspended particles [155]. Recent evidence suggests that protein-like and humic CDOM are predominantly degraded by microbial and UV phenomena, respectively [156]. Particularly enlightening is a report that CDOM shields aquatic life from potentially harmful radiation [157]. Not surprisingly, techniques for large-scale monitoring of CDOM are in constant development [158, 159]. Molecular characterization of fresh-water DOM The molecular composition of fresh- DOM has been studied less than that of marine DOM, probably owing to the greater effect of oceanic DOM on the geochemical carbon balance. Nevertheless, several studies have tried to remedy this and characterize fresh- substances in detail. A noteworthy NMR spectroscopy experiment performed on lacustrine DOM [72] has shown the potential of this technique in recognizing and quantifying functional groups even in such a complex DOM system. In this study, by simple 1H monodimensional spectroscopy, aliphatic, carbohydrate, aromatic, and CRAM molecules were differentiated by chemical shift analysis. Furthermore, 13C investigation by multidimensional heteronuclear techniques, for example HMQC (heteronuclear multiple quantum coherence) and HMBC (heteronuclear multiple bonding coherence), enabled characterization of specific regions assignable to well known organic species, for example anomeric carbons from carbohydrates, conjugated double bonds from aliphatic acids, CRAM, N-acetyls, and others. Interestingly, differentiation between terpenederived CRAM and MDLT has been achieved by HMBC (Fig. 5), thereby providing a potential tool for investigation, in further detail, of terpene metabolism in DOM.

Fig. 5 2D 1H–13C HMBC spectra of LO-DOM. (A) Expansion in-„ cluding the full CRAM (I) and MDLT (II) regions. (B) Expansion of (A) showing the 5–115 ppm (carbon) region. (C) Expansion of (A) showing the carboxyl region. The notation La, Lb, Lc, and Ld is used to identify crosspeaks in the HMBC spectrum. From Ref. [72]

119

120

A different approach based on the use of pyrolysis coupled to MS has been proposed by Guo et al. [160]. This work showed that pyrolysed products of the high molecularweight fraction of DOM contained compounds, for example furfural, methylfurfural, dimethylbenzene, phenol, and methylcyclopentenone, which may be assigned to sugar and aromatic structures. Another study on lacustrine DOM used Fourier-transform infra-red (FT-IR) spectroscopy in combination with direct temperature-resolved MS (DTMS), and achieved excellent characterization of chemical groups by processing the spectroscopic data by use of mathematical methods [161]. The last two are examples of inexpensive but effective approaches which may act as a substitute for more advanced instrumentation. However, the conventional technique for molecular investigation of DOM is, again, FT-ICR MS, because of its resolving power, which is capable of revealing hundreds of empirical formulae and furnishing plausible molecular structures for each unknown compound. Characterization of condensed aromatic compounds (BC) [128] (Fig. 6) and amides [63] has been achieved by use of FT-ICR MS. Transport of terrestrial DOM to oceans The long-debated hypothesis that part of marine DOM is of terrestrial origin seems to have been confirmed by several studies. In particular, analysis of spatial distribution in

A. Nebbioso, A. Piccolo

estuarine systems showed that riverine DOM is highly preserved in the ocean, despite structural rearrangements, which occur mostly because of to changes in salinity [135]. In a review of recent breakthroughs in arctic biogeochemistry, Dittmar and Kattner [162] observed that terrestrial DOM was persistently refractory in DOM pools of the arctic ocean, being the result of a long-term carbon sedimentation and the end-product of a natural carbon sink. A molecular investigation of transported DOM by HPLC and MS revealed it is composed of both hydrophilic and hydrophobic fractions [65]. Low-MW compounds were predominantly found in the former whereas larger-MW constituents were identified in the latter. In the same work, comparison of the molecular composition of a wood extract and DOM revealed several characteristics in common, thereby indicating that lignin-derived compounds are a major component of refractory DOM. Further insight into lignin transformation [163] suggested the involvement of photochemical processes, implying that chemical transformation of DOM is not only a result of microbial activity. Consistent with such findings, an increase of oxidation state was measured in DOM from Chesapeake Bay flowing in-shore to off-shore [107], confirming that environmental conditions in the mixing zone affected the transformation of transported DOM. The same work also suggested the presence of terrestrial black-carbon structures in transported DOM, as a pyrogenic contribution from carbonized OM leaching through soil and reaching fresh water. The complexity of the process of formation and transport of refractory DOM to oceanic carbon pools would greatly benefit from a dedicated study which included samples from different estuaries. From the published literature it can be inferred that ocean DOM is derived partially from: 1. the humified products of plant decay, which subsequently dissolve in fresh water and are transported to the ocean; and 2. phytoplankton proteins and polysaccharide products which are altered and transformed in situ by bacteria to form another pool of refractory organic matter. During transport, chemical transformations may also occur, thereby increasing its refractory properties. This organic matter is subsequently included in the oceanic carbon pool for a long-term storage. DOM in pore water

Fig. 6 Possible chemical structure of one peak (C27H18O8) in the mass spectrum of DOM from McDonalds Branch Basin located in the New Jersey Pine Barrens (USA). From Ref. [128]

Pore water is water in physical spaces isolated by sediments in which geochemical transformations differ from those occurring in outside spaces. MALDI-TOF-MS evidence indicates that protein breakdown in pores is not as extensive as in open marine waters [25]. The accumulation and dynamics of DOM are dramatically dependent on whether the

Molecular characterization of dissolved organic matter

pore water conditions are anoxic, suboxic, or mixed-redox (suboxic, or oscillating between oxidizing and reducing), which result in very different reactive products and in variable rates of degradation [164]. Moreover, DOM accumulation has been found to be generally limited in the mixedredox zone relative to the anoxic zone, and humic-like fluorescence intensity also differed between mixed-redox and anoxic zones of the sediment, such that anoxic pore waters were relatively enriched in fluorescent, humic-like compounds [164]. Pore waters are also found within peat soils, in which they pose a problem to water quality. In fact, because of microbiological degradation of pore DOM, “tea”-coloured water is released, and requires purifying treatment before use [165]. Pore waters contain CDOM material, and, as for other types of DOM, humic-like and protein-like fluorophores [166]. Humic-like DOM apparently increased with depth whereas no particular trends were observed for protein-like DOM. Moreover, humic-like CDOM may be a lowmolecular-weight fraction of refractory pore DOM. As expected, experimental evidence showed that humic-like CDOM was better preserved under anoxic conditions [166]. Molecular characterization of pore-water DOM from sediments of the Iberian peninsula revealed relatively high abundance of highly oxygenated aromatic compounds or nitrogen-bearing compounds with low H/C ratios [167]. These structures were likely to originate from lignin phenols of terrestrial origin. In fact, their amounts decreased as the material was transported along the shelf. Pore DOM is, therefore, a material produced by geomorphological phenomena similar to those which produce freshwater and marine DOM, and its composition is similar in several ways to those of open-water types of DOM. However, the compartment in which it is confined affects its chemical and chromophoric composition, which ultimately depend on conditions such as oxygen availability, depth, and microbial population [167].

Conclusions Characterization of DOM has advanced substantially in the last decade, mainly because of the development of advanced instrumental analytical methods, for example FT-ICR MS, FEEM-PARAFAC, multidimensional NMR, tangential flow ultrafiltration, and reverse osmosis, but also because of the introduction of innovative ideas regarding DOM genesis and transformation. The recognition of components from terrestrial sources, the discovery of microgels, advances in understanding the effect of microorganisms in reworking DOM and the supramolecular nature of humic substances, and the discovery of new mechanisms for carbon storage in the environment, are the current milestones in the science of natural

121

OM. The knowledge acquired encompasses not only DOM itself and its fate, but also that of terrestrial and marine biota, and of larger environmental domains, for example continental shelves, sediments, and natural streams. The content and structures of proteins, lipids, polysaccharides, lipopolysaccharides, lignin, terpene derivatives, and black carbon components must be characterized to enable understanding of the genesis, structure, and chemical reactivity of DOM. Determination of chromophores in DOM has been useful in enabling extensive monitoring of fluxes and streams, and in revealing anthropogenic interference. Future challenges in environmental science are: 1. standardization of procedures for assigning structures on the basis of MS, NMR, and fluorescence experimental data; 2. clarification of the involvement of microorganisms in transforming and degrading DOM, i.e. the extent to which they produce OM and the extent to which they transform existing OM; and 3. understanding the correlation between DOM composition and its properties, for example recalcitrance, action as a radiation shield, and interaction with organic and metal pollutants. It is expected that research based both on the use of advanced instrumentation and on the methodological experience acquired so far will enable successful tackling of these challenges in the future.

References 1. Piccolo A (2012) The nature of soil organic matter and innovative soil managements to fight global changes and maintain agricultural productivity. In: Piccolo A (ed) Carbon Sequestration in Agricultural Soils. Springer Verlag, Heidelberg, pp 1–20 2. Nebbioso A, Piccolo A (2011) Biomacromolecules 12:1187–1199 3. Thurman EN (1985) Organic Geochemistry of Natural Waters. Nijhoff/Junk, Dordrecht 4. Schnitzer M, Khan SU (1972) Humic Substances in Environment. Marcel Dekker, New York 5. Carter CW, Suffet IH (1982) Environ Sci Technol 16:735–740 6. Tipping E (1986) Mar Chem 18:161–169 7. Zafiriou OC, Joussot-Dubien J, Zepp RG, Zika RG (1984) Environ Sci Technol 18:358A–371A 8. Stewart AJ, Wetzel RG (1981) Arch Hydrobiol 92:265–286 9. Oliver BG, Thurman EM, Malcolm RL (1983) Geochim Cosmochim Acta 47:2031–2035 10. Jenne EA (1975) Sorption phenomenon in organics of bottom sediments. In Progress in Water Technology Vol. 7., ed. P. K. Krenkel. pp. 131–136 11. Fisher SG, Likens GE (1973) Ecol Monogr 43:421–439 12. Raymont J (1983) Plankton and productivity in the oceans. Pergamon Press, Oxford 13. Leenheer JA, Croue JP (2003) Environ Sci Technol 37:18A– 26A 14. Schumacher M, Christl I, Vogt RD, Barmettler K, Jacobsen C, Kretzschmar R (2006) Biogeochemistry 80:263–275

122 15. Schlesinger WH, Melack JM (1981) Tellus 33:172–187 16. Jahnke R (1996) Glob Biogeochem Cycles 10:71–88 17. Griffith DR, McNichol AP, Xu L, McLaughlin FA, Macdonald RW, Brown KA, Eglinton TI (2012) Biogeosciences 9:1217– 1224 18. Killops S D, Killops V. J. (2005) Introduction to organic geochemistry. Blackwell Publishing 19. Hatakka A (2001) In: Hofrichter M, Steinbüchel A (eds) Biodegradation of Lignin. Lignin, Humic Substances and Coal, vol 1. Wiley–VCH, Weinheim, pp 129–180 20. Christ MJ, David MB (1996) Soil Biol Biochem 28:1191–1199 21. Hedges JI, Mann DC (1979) Geochim Cosmochim Acta 43:1803–1807 22. Sarkanen KV, Ludwig CH (1971) Lignins: Occurrence, Formation, Structure and Reactions. Wiley, New York 23. Verma P, Junga U, Militz H, Mai C (2009) Holzforschung 63:371–378 24. Hollibaugh JT, Azam F (1983) Limnol Oceanogr 28:1104–1116 25. Nunn BL, Norbeck A, Keil RG (2003) Mar Chem 83:59–73 26. Piontek J, Lunau M, Händel N, Borchard C, Wurst M, Engel A (2010) Biogeosciences 7:1615–1624 27. Ingalls AE, Lee C, Wakeham SG, Hedges JI (2003) Deep-Sea Res Pt II 50:713–738 28. Hedges JI, Eglinton G, Hatcher PG, Kirchman DL, Arnosti C, Derenne S, Evershed RP, Kögel-Knabner I, de Leeuw JW, Littke R, Michaelis W, Rullkötter J (2000) Org Geochem 31:945–958 29. Dittmar T, Kattner G (2003) Mar Chem 82:115–123 30. Ding YX, Chin WC, Rodriguez A, Hung CC, Santschi PH, Verdugo P (2008) Mar Chem 112:11–19 31. Wakeham SG, Lee C, Hedges JI, Hernes PJ, Peterson MJ (1997) Geochim Cosmochim Acta 61:5363–5369 32. Simpson MJ, Hatcher PG (2004) Org Geochem 35:923–935 33. Burdige DJ (2007) Chem Rev 107:467–485 34. Keil RG, Tsamakis E, Hedges JI (2000) Early diagenesis of particulate amino acids in marine systems. In: Goodfriend GA, Collins MJ, Fogel ML, Macko SA, Wehmiller JF (eds) Perspectives in Amino Acid and Protein Geochemistry. Oxford University Press, New York, pp 69–82 35. Yamada N, Tanoue E (2006) Prog Oceanogr 69:1–18 36. Stürmer DH, Harvey GR (1977) Deep-Sea Res 24:303–309 37. Mills GL, Quinn JG (1981) Mar Chem 10:93–102 38. Perdue EM, Ritchie JD (2003) Dissolved organic matter in fresh waters. In Surface and Ground Water, Weathering, Erosion and Soils (ed. J. I. Drever) Vol. 5, Treatise on Geochemistry (eds. H. D. Holland and K. K. Turekian), Elsevier–Pergamon, Oxford, pp. 273–318 39. Benner R, Biddanda B, Black B, McCarthy M (1997) Mar Chem 57:243–263 40. Aluwihare LI, Repeta DJ, Chen RF (2002) Deep-Sea Res Pt II 49:4421–4437 41. Wong H, Mok KM, Fan XJ (2007) Desalination 210:44–51 42. Walker BD, Beauprè SR, Guilderson TP, Druffel ERM, McCarthy MD (2011) Geochim Cosmochim Acta 75:5187–5202 43. Tansakul C, Laborie S, Cabassud C (2011) Water Res 45:6362– 6370 44. Koprivnjak JF, Perdue EM, Pfromm PH (2006) Water Res 40:3385–3392 45. Gurtler BK, Vetter TA, Perdue EM, Ingall E, Koprivnjak JF, Pfromm PH (2008) J Membrane Sci 323:328–336 46. Koprivnjak JF, Pfromm PH, Ingall E, Vetter TA, Schmitt-Kopplin P, Hertkorn N, Frommberger M, Knicker H, Perdue EM (2009) Geochim Cosmochim Acta 73:4215–4231 47. Huguet A, Roux-de Balmannc H, Parlanti E (2009) J Membr Sci 326:186–196 48. Gélinas Y, Baldock JA, Hedges JI (2001) Org Geochem 32:677– 693

A. Nebbioso, A. Piccolo 49. Sleighter R, McKee G, Hatcher P (2009) Org Geochem 40:119– 125 50. Zsolnay A (2003) Geoderma 113:187–209 51. Sanchez-Gonzalez J, Garcia-Otero N, Moreda-Pineiro A, Bermejo-Barrera P (2012) Microchem J 102:75–82 52. Birdwell JE, Valsaraj KT (2010) Atmos Environ 44:3246–3253 53. Duarte RMBO, Duarte AC (2011) Trend Anal Chem 30:1659– 1671 54. Schijf J, Zoll A (2011) J Colloid Interface Sci 361:137–147 55. Piccolo A (2001) Soil Sci 166:810–832 56. Kujawinski EB, Freitas MA, Zang X, Hatcher PG, Green-Church KB, Jones RB (2002) Org Geochem 33:171–180 57. Kim S, Simpson AJ, Kujawinski EB, Freitas MA, Hatcher PG (2003) Org Geochem 34:1325–1335 58. Kim S, Kramer RW, Hatcher PG (2003) Anal Chem 75:5336–5344 59. Koch BP, Dittmar T (2006) Rapid Commun Mass Spectrom 20:926–932 60. Dittmar T, Koch BP (2006) Mar Chem 102:208–217 61. Dittmar T, Paeng J (2009) Nat Geosci 2:175–179 62. Reemtsma T, These A, Linscheid M, Leenheer J, Spitzy A (2008) Environ Sci Technol 42:1430–1437 63. McKee G, Hatcher P (2010) Geochim Cosmochim Acta 74:6436–6450 64. Koch BP, Ludwichowski KU, Kattner G, Dittmar T, Witt M (2008) Mar Chem 111:233–241 65. Liu Z, Sleighter RL, Zhong J, Hatcher PG (2011) Estuar Coast Shelf Sci 92:205–216 66. Mawhinney DB, Rosario-Ortiz FL, Baik S, Vanderford BJ, Snyder SA (2009) J Chromatogr A 1216:1319–1324 67. Dittmar T, Whitehead K, Minor EC, Koch BP (2007) Mar Chem 107:378–387 68. Peuravuori J, Bursáková P, Pihlaja K (2007) Anal Bioanal Chem 389:1559–1568 69. Alberic P (2011) Rapid Commun Mass Spectrom 25:3012–3018 70. Lepane V, Tõnno I, Alliksaar T (2010) Procedia Chem 2:101–108 71. Kawasaki N, Matsushige K, Komatsu K, Kohzu A, Nara FW, Ogishi F, Yahata M, Mikami H, Goto T, Imai A (2011) Water Res 45:6240–6248 72. Lam B, Baer A, Alaee M, Lefebvre B, Moser A, Williams A, Simpson AJ (2007) Environ Sci Technol 41:8240–8247 73. Engel A, Händel N (2011) Mar Chem 127:180–191 74. Nebbioso A, Piccolo A, Spiteller M (2010) Rapid Commun Mass Spectrom 24:3163–3170 75. Wickramasekara S, Hernández-Ruiz S, Abrella L, Arnold R, Chorover J (2012) Anal Chim Acta 717:77–84 76. Paytan A, Cade-Menun BJ, McLaughlin K, Faul KL (2003) Mar Chem 82:55–70 77. Sannigrahi P, Ingall ED, Benner R (2005) Deep-Sea Res I 52:1429–1444 78. Abdulla HAN, Minor EC, Dias RF, Hatcher PG (2010) Geochim Cosmochim Acta 74:3815–3838 79. Kaiser E, Simpson AJ, Dria KJ, Sulzberger B, Hatcher PG (2003) Environ Sci Technol 37:2929–2935 80. Simpson AJ, Kingery WL, Hatcher PG (2003) Environ Sci Technol 37:337–342 81. Simpson AJ (2001) Soil Sci 166:795–890 82. Smejkalova D, Piccolo A (2007) Environ Sci Technol 42:699– 706 83. Piccolo A, Conte P (1998) Advances in Nuclear Magnetic Resonance and Infrared Spectroscopies of Soil Organic Particles. In: Huang PM, Senesi N, Buffle J (eds) Structure and Surface Reactions of Soil Particles, IUPAC Environmental and Physical Series: Environmental Particles. Vol. 4, Analytical and Physical Chemistry of Soil. John Wiley and Sons, New York, pp 184–250

Molecular characterization of dissolved organic matter 84. Conte P, Spaccini R, Smejkalova D, Nebbioso A, Piccolo A (2007) Chemosphere 69:1032–1039 85. Simpson AJ, McNally DJ, Simpson MJ (2010) Prog Nucl Magn Res Spectrosc 58:97–175 86. Belzile C, Roesler CS, Christensen JP, Shakhova N, Semiletov I (2006) Estuar Coast Shelf Sci 67:441–449 87. Parlanti E, Morin B, Vacher L (2002) Org Geochem 33:221–236 88. Fichot CG, Miller WL (2010) Remote Sens Environ 114:1363– 1377 89. Singh S, D’Sa EJ, Swenson EM (2010) Sci Total Environ 408:3211–3222 90. Zhang Y, Yin Y, Feng L, Zhu G, Shi Z, Liu X, Zhang Y (2011) Water Res 45:5110–5122 91. Guéguen C, Granskog MA, McCullough G, Barber DG (2011) J Mar Syst 88:423–433 92. Mecozzi M, Pietroletti M, Conti ME (2008) Int J Environ Pollut 32:527–549 93. Verdugo P, Alldredge AL, Azam F, Kirchman DL, Passow U, Santschi PH (2004) Mar Chem 92:67–85 94. Bhaskar PV, Grossart HP, Bhosle NB, Simon M (2005) FEMS Microbiol Ecol 53:255–264 95. Passow U, Dunne J, Murray JW, Balistrieri L, Alldredge AL (2006) Mar Chem 100:323–336 96. Gogou A, Repeta D (2010) Mar Chem 121:215–223 97. Wakeham SG, Pease TK, Benner R (2003) Org Geochem 34:857–868 98. Minor EC, Simjouw JP, Boon JJ, Kerkhoff AE, van der Horst J (2002) Mar Chem 78:75–102 99. Quan TM, Repeta DJ (2007) Mar Chem 105:183–193 100. Kaiser K, Benner R (2009) Mar Chem 113:63–77 101. Tanoue E, Nishiyama S, Kamo M, Tsugita A (1995) Geochim Cosmochim Acta 59:2643–2648 102. Powell MJ, Sutton JN, Del Castillo CE, Timperman AT (2005) Mar Chem 95:183–198 103. Kuznetsova M, Lee C, Aller J (2005) Mar Chem 96:359–377 104. Orellana MV, Petersen TW, Diercks AH, Donohoe S, Verdugo P, van den Engh G (2007) Mar Chem 105:229–239 105. Nunn BL, Timperman AT (2007) Mar Ecol Prog Ser 332:281–289 106. Goutx M, Guigue C, Striby L (2003) Org Geochem 34:1465–1473 107. Sleighter R, Hatcher P (2008) Mar Chem 110:140–152 108. Hertkorn N, Benner R, Frommberger M, Schmitt-Kopplin P, Witt M, Kaiser K, Kettrup A, Hedges JI (2006) Geochim Cosmochim Acta 70:2990–3010 109. Mei ZP, Saucier FJ, LeFouest V, Zakardjian B, Sennville S, Xie H, Starr M (2010) Cont Shelf Res 30:2027–2042 110. Keith DJ, Yoder JA, Freeman SA (2002) Estuar Coast Shelf Sci 55:705–717 111. Yamashita Y, Tsukasaki A, Nishida T, Tanoue E (2007) Mar Chem 106:498–509 112. Yang L, Hong H, Guo W, Chen CTA, Pan PI, Feng CC (2011) Mar Chem 128–129:64–71 113. Tedetti M, Cuet P, Guigue C, Goutx M (2011) Sci Total Environ 409:2198–2210 114. Murphy KR, Stedmon CA, Waite TD, Ruiz GM (2008) Mar Chem 108:40–58 115. Granskog MA, Macdonald RW, Mundy CJ, Barber DG (2007) Cont Shelf Res 27:2032–2050 116. Luciani X, Mounier S, Paraquetti HHM, Redon R, Lucas Y, Bois A, Lacerda LD, Raynaud M, Ripert M (2008) Mar Environ Res 65:148–157 117. Kowalczuk P, Durako MJ, Young H, Kahn AE, Cooper WJ, Gonsior M (2009) Mar Chem 113:182–196 118. Kowalczuk P, Cooper WJ, Durako MJ, Kahn AE, Gonsior M, Young H (2010) Mar Chem 118:22–36 119. Du C, Shang S, Dong Q, Hub C, Wu J (2010) Estuar Coast Shelf Sci 88:350–356

123 120. Fellman JB, Spencer RGM, Hernes PJ, Edwards RT, D’Amore DV, Hood E (2010) Mar Chem 121:112–122 121. Lønborg C, Álvarez-Salgado XA, Davidson K, Martínez-García S, Teira E (2010) Mar Chem 119:121–129 122. Omori Y, Hama T, Ishii M, Saito S (2011) Mar Chem 124:38–47 123. Johannensen S, Miller W (2001) Mar Chem 76:271–283 124. Burns KA, Greenwood P, Benner R, Brinkman D, Brunskill G, Codi S, Zagorskis I (2004) Cont Shelf Res 24:2373–2394 125. Repeta DJ, Quan TM, Aluwihare LI, Accardi AM (2002) Geochim Cosmochim Acta 66:955–962 126. Amon RMW, Benner R (2003) Deep-Sea Res Pt I 50:151–169 127. Cappiello A, Trufelli H, Famiglini G, Pierini E, Capellacci S, Penna A, Ricci F, Ingarao C, Penna N (2007) Water Res 41:2911– 2920 128. Kim S, Kaplan LA, Benner R, Hatcher PG (2004) Mar Chem 92:225–234 129. Hockaday WC, Grannas AM, Kim S, Hatcher PG (2006) Org Geochem 37:501–510 130. Dittmar T (2008) Org Geochem 39:396–407 131. Chen RF, Gardner GB (2004) Mar Chem 89:103–125 132. Piccolo A (2002) Adv Agron 75:57–134 133. Nebbioso A, Piccolo A (2012) Anal Chim Acta 720:77–90 134. Goni MA, Teixeira MJ, Perkey DW (2003) Estuar Coast Shelf Sci 57:1023–1048 135. Stabenau ER, Zika RG (2004) Mar Chem 89:55–67 136. Leenheer JA, Nanny MA, McIntyre C (2003) Environ Sci Technol 37:2323–2331 137. Minor EC, Steinbring CJ, Longnecker K, Kujawinski EB (2012) Org Geochem 43:1–11 138. Penezic A, Gasparovic B, Buric Z, Frka S (2010) Estuar Coast Shelf Sci 86:625–636 139. Shi W, Sun MY, Molina M, Hodson RE (2001) Org Geochem 32:453–467 140. Rochelle-Newall EJ, Fisher TR (2002) Mar Chem 77:7–21 141. Zhang Y, van Dijk MA, Liu M, Zhu G, Qin B (2009) Water Res 43:4685–4697 142. DePalma SGS, Arnold WR, McGeer JC, Dixon DG, Smith DS (2011) Appl Geochem 26:394–404 143. Huguet A, Vacher L, Relexans S, Saubusse S, Froidefond JM, Parlanti E (2009) Org Geochem 40:706–719 144. Stolpe B, Guo L, Shiller AM, Hassellov M (2010) Mar Chem 118:119–128 145. Boehme K, Wells M (2006) Mar Chem 101:95–103 146. Clark CD, Jimenez-Morais J, Jones G, Zanardi-Lamardo E, Moore CA, Zika RG (2002) Mar Chem 78:121–135 147. Jaffè R, Boyer JN, Lua X, Maie N, Yang C, Scully NM, Mocka S (2004) Mar Chem 84:195–210 148. Spencer RGM, Baker A, Ahad JME, Cowie GL, Ganeshram R, Upstill-Goddard RC, Uher G (2007) Sci Total Environ 373:305–323 149. Huguet A, Vacher L, Saubusse S, Etcheber H, Abril G, Relexans S, Ibalot F, Parlanti E (2010) Org Geochem 41:595–610 150. Guéguen C, McLaughlin FA, Carmack EC, Itoh M, Narita H, Nishino S (2011) The nature of colored dissolved organic matter in the southern Canada Basin and East Siberian Sea. Deep-Sea Res Pt II, in press doi:10.1016/j.dsr2.2011.05.004 151. Boyd TJ, Orsburn CL (2004) Mar Chem 89:189–210 152. Mazzuoli S, Focardi S, Bracchini L, Falcucci M, Loiselle SA, Rossi C (2005) Ecol Model 186:55–61 153. Dalzell BJ, Minor EC, Mopper KM (2009) Org Geochem 40:243–257 154. Shank GC, Zepp RG, Vähätalo A, Lee R, Bartels E (2010) Mar Chem 119:162–171 155. Guo W, Stedmon CA, Han Y, Wu F, Yu X, Hu M (2007) Mar Chem 107:357–366 156. Massicotte P, Frenette JJ (2011) Ecol Appl 21:2600–2617

124 157. Bracchini L, Cozar A, Dattilo AM, Loiselle SA, Tognazzi A, Azza N, Rossi C (2006) Chemosphere 63:1170–1178 158. Carstea EM, Baker A, Bieroza M, Reynolds D (2010) Water Res 44:5356–5366 159. Baghoth SA, Sharma SK, Amy GL (2011) Water Res 45:797–809 160. Guo L, White DM, Xu C, Santschi PH (2009) Mar Chem 114:63–71 161. Minor E, Stephens B (2008) Org Geochem 39:1489–1501 162. Dittmar T, Kattner G (2003) Mar Chem 83:103–120

A. Nebbioso, A. Piccolo 163. Bianchi TS, DiMarco SF, Smith RW, Schreiner KM (2009) Mar Chem 117:32–41 164. Komada T, Reimers CR, Luther GW III, Burdige DJ (2004) Geochim Cosmochim Acta 68:4099–4111 165. Aguilar L, Thibodeaux LJ (2005) Chemosphere 58:1309– 1318 166. Burdige DJ, Kline SW, Chen W (2004) Mar Chem 89:289–311 167. Schmidt F, Elvert M, Koch BP, Witt M, Hinrichs KU (2009) Geochim Cosmochim Acta 73:3337–3358