Chemistry at the air-sea interface

December 17, 2011


The oceans cover roughly 70% of the Earth’s surface, therefore, discounting average ice cover, the air-sea interface accounts for just under half of the geographic area of the planet (Cunliffe and Murrell 2009). Through this thin layer occurs all atmospheric gas exchange and particulate deposition. Particularly, the sequestration of CO2 in the marine system is modulated by physical and chemical sea surface properties.

Because it is nearly two-dimensional, the surface micro-layer (SML) is inherently differentiated from bulk water. Surface chemistry explains some of the distinction, and will be discussed in the context of chemical partitioning (the ratio of SML concentration to bulk water concentration). Ideally, the SML might be considered a mono-molecular layer, on the order of 100 angstrom thick (Lion and Leckie 1981).

MacIntyre hypothesized in 1974 that this sheet of single molecules was composed of successive layers of decreasingly hydrophobic compounds (cited in Wurl and Holmes 2008). Later studies revisited this idealized model, demonstrating surface dominance of intertwined organic polymers (Cunliffe and Murrell 2009), as predicted by John Sieburth in 1983. This gelatinous mat is nearly homogenous for 50 μm, with a steep physical and chemical gradient between it and bulk water (Zhang et al 2003). Apparent partitioning of chemical species and biological activity has increased with finer sampling resolution, and this paper will discuss how the depth and even the concept of the SML has changed as sampling techniques have evolved.

Influx in the form of Aeolian dust is responsible for excess surface trace metals, which complex with organics to form ligands (Lion and Leckie 1981). Additionally, floating biogenic material (FBM) accumulates through buoyant flux (Azetsu-Scott and Passow 2004) and is photolysed over time by the intense UV irradiation at the surface (Zhou and Mopper 1996). At high concentrations, surface organics form visible sea slicks and stable foams. Physical processes in the upper ocean influence the distribution of the buoyant and adsorbed material, which can support microbial and macrofaunal communities in convergence zones typically considered oligotrophic (Dandonneau et al 2008).

This paper reviews pertinent literature on the physico-chemical processes controlling SML partitioning, and the feedback connections with microbial respiration. Gases are also partitioned in the SML, but only the effects of excess surface organic matter on atmospheric gas exchange are included in the following discussion.


Methods for sampling generally utilize surface adhesion of the SML to some substrate. Water is collected over an often arbitrary depth (Zhang etal 2003), so that the end-member properties are diluted by the inclusion of subsurface water. Glass and Teflon plates, dipped vertically into water, collect surface material over a depth proportional to the withdrawal rate. A metal or polyester screen collects the surface layer by wetting of the 1 mm mesh in a similar manner. These methods achieve SML resolutions of 60-100 μm and 200-400 μm respectively (Lion and Leckie 1981; Kuznetsova etal 2004). Rotating drums similar to oil slick remediation devices are also used: the drum wall continually adheres surface material, which is scraped off and collected during rotation. Collections of this type can be run for several hours (Kuznetsova etal 2004) and are able to accumulate trace elements above necessary nanomolar detection thresholds. The sampling depth for the rotating drum is approximately 60 μm (Lion and Leckie 1981). Sampling sea foam also provides a good approximation of SML concentrations (Lion and Leckie 1981) since this material is formed during the percolation of air bubbles through the SML. Using several of these methods to vary the depth of measurement, Zhang et al (2003) demonstrated that concentrations of a number of elements and nutrients changed rapidly over a 10 µm thick layer centered at a depth of 50 µm, which establishes the functional depth of the SML in accordance with estimates from tracer experiments.

By adhering the surfactant material to a germanium prism Baier (1970, referenced in Lion and Leckie 1981) was able to optically analyze the upper 30 nm of lake water using multiple attenuated internal reflection spectroscopy (MAIR). In MAIR, infrared light passes through the prism and loses energy during each reflection relative to material in contact with the outside surface. He established that proteinaceous organic molecules dominated the interface.

Other indirect and proxy measurements are used. The neuston provides a unique opportunity for the study of heavy metal distribution (Schulz-Baldes 1992), as these organisms live and feed in an environment enriched in toxic species. The assimilated concentrations of Cu, Cd, and Pb in surface feeding zooplankton serve as a proxy for in situ surface concentrations, as will be discussed later in conjunction with a closer look at biology in the neuston (7). Satellite remote sensing of surface optical properties is a valuable tool in the study of primary production, and colored dissolved organic matter (CDOM). It may also be of use in observing SML chemistry, if a systematic relationship can be established between surface abundance transparent organic matter (Dandonneau et al 2008) and overall seawater optical properties. Ocean modeling demonstrates the divergent transport of surface and neutrally buoyant particles, and will be further addressed in the discussion of physical processes (6).


Nearly all trace metals are partitioned in the SML, and can be four orders of magnitude greater than the average in the water column (Lion and Leckie 1981; Zhang et al 2003). Surface excess is maintained by surface adhesion, wind blown dust deposition (Fe), and scavenging by rising bubbles and buoyant particles. Pb is enriched by atmospheric deposition of gasoline combustion products. Trace metals compete with more common ionic species (Mg2+ and Ca2+) for limited binding sites on organic molecules. Elements with the greatest binding affinity will be preferentially transported by particle adsorption, including Fe3+ and Cr3+ (Lion and Leckie 1981). These complexes may exist either in equilibrium with free ions, or in stable bound forms. Metal-organic ligands also serve to reduce toxicity and increase trace metal bio-availability (Schulz-Baldes 1992) .

Surface active molecules (surfactants) preferentially occupy the air-sea interface, a solution behavior described by the Gibbs adsorption isotherm (Lion and Leckie 1981) ,

-dgamma = SUM{1,n}(RHO_i(dmu_i))

describes the relationship between surface tension and surface enrichment of chemical species, where γ is surface tension, μi is chemical potential of species i, and Γi is the surface excess (moles cm-2). Those which decrease surface tension will be preferentially adsorbed, while those which increase surface tension will be excluded. Because of the limited adsorption sites, as material is concentrated by physical processes high-molecular-weight molecules will preferentially occupy the interface, and inorganic molecules are excluded. When pollutants are introduced to sea water, they can become a major surface constituent, as in the Mediterranean Sea, where 85% of SML molecules can be pollutant hydrocarbons (Lion and Leckie 1981). The surface exclusion of inorganic metal species means that upward particle transport must be substantial to counteract diffusion and scavenging by sinking particles.

In a systematic survey of coastal and open ocean sites in the South China Sea, Zhang et al (2003) observed partitioning of NO3, NO2, NH4, PO4, and SiO3. This is due to dust deposition and net surface respiration. Films and foams trap nutrients, and are crucial in supporting surface microbial communities.

Biogenic inputs

The current model of the SML is as a gelatinous matrix of polysaccharides, proteinaceous material and lipids (Cunliffe and Murrell 2009; Wurl and Holmes 2008). Relative to each other these exist in a ratio of 10:4:1. The lipids are not abundant enough to occupy surfaces sites, and instead are mostly adhered to particles (Wurl and Holmes 2008). When the concentration of surface-confined biogenic material is great enough, surface films form, apparent as slicks of relative smoothness on the water. Partitioning of organic matter occurs through in situ production or buoyant transport. This section will address the formation and aggregation of buoyant organic matter. The excretion of DOM and uptake of aldehydes is more usefully mentioned in reference to the microbial community (5).

Since there is negative net primary production in the SML, some upward flux of organic material is required to fuel high surface respiration rates (Dandonneau etal 2008). Inverted sediment traps show upward particulate flux is typically 5-10% of downward flux at 600 m, and could be much greater closer to the surface where more organic matter is produced (Dandonneau etal 2008). Ascending transparent exopolymer particles (TEP) balance this deficit. TEP is part of the POM pool. It adheres and aggregates particles and microbes, building marine snow from suspended material. Fresh TEP, produced during the spontaneous aggregation of polysaccharide DOM (Lion and Leckie 1981), is initially positively buoyant and collects at the surface, transporting bacteria and diatoms as it ascends (Azetsu-Scott and Passow 2004; Dandonneau et al 2008; Wurl and Holmes 2008). The ascension rate can be calculated from the change in surface concentrations of pure TEP (Ct-C0), over time t,

WTEP = (Ct-C0) * V over pi * rc^2 * C_0 * t

where V is the surface volume, and rc is the radius of the experimental settling column. On average, TEP ascends at 14 m day-1 (Azetsu-Scott and Passow 2004). Mean TEP concentrations around 1000μg Xeq L-1 (expressed in xanthan gum equivalency) are typical during bloom periods. Blooms in growth phase produce the more precursor DOM than standing stocks (Kuznetsova et al 2004).

Subsurface TEP concentration is logarithmically related to chlorophyll in the water column, but this relationship breaks down in the SML due to the accumulation of degradation-resistant refractory organic matter (Wurl and Holmes 2008). The relative stickiness of TEP (α, probability of adhesion given a collision) can be modeled using the ratio of TEP and chlorophyll,

α = 6.38·10-4  (TEP:Chl-a) - 0.0033

Stickiness is greater in open ocean (α = 0.34 ± 0.29) than in estuaries (α = 0.022 ± 0.015), due to the high concentration of suspended particulate matter in coastal ecosystems (Wurl and Holmes 2008).

Sulfate half-ester groups increase carbohydrate adhesion by forming metal ion bridges and hydrogen bonds, and therefore increase scavenging of reactive species (Zhang et al 2003). Surface partitioning of sulfate half-esters enhances aggregation of free TEP in the SML, and will promote the abiotic formation of TEP/POM around rising bubbles (Lion and Leckie 1981; Wurl and Holmes 2008). TEP partitioning is higher in the open ocean than in estuaries (Wurl and Holmes 2008). This can be explained by high subsurface production rates in coastal systems, but the increased wave and wind action in the open ocean could also enhance near surface TEP formation.

Bubble walls can be stabilized by the glycoprotein-proteoglycan matrix and silica diatom fragments, forming persisting sea foams (Wurl and Holmes 2008). Polar groups in the adsorbed layer repel each other and slow the loss of fluid from bubble wall (Lion and Leckie 1981). As previously discussed, the amount of any foam constituent is indicative of the amount in the SML. Stable foams are degraded over time by wave action and UV radiation (Zhou and Mopper 1997) and collapse and return their material to the OM pool. Amino acids concentration (a proteinaceous fraction of OM pool) in the SML can be 50 times that of the underlying water column (Kuznetsova etal 2004), and like DOM is controlled in part by the surface microbial community.

Degradation of DOM by UV radiation produces low-molecular-weight (LMW) carbonyl compounds including formaldehyde, acetaldehyde, propanol, glyoxal, and methylglyoxal (Zhou and Mopper 1997). Formaldehyde and acetaldehyde are not surface active, and respective surface partitioning factors of 8.9 and 11.4 are attributable only to irradiance. The conversion process is greatly enhanced in sea foams because of the high DOM concentration. The loss of aldehydes to the atmosphere is limited by organic surfactants, making the residence time of LMW carbonyl species on the order of hours to days instead of minutes (Zhou and Mopper 1997). This allow nano- and pico-molar concentrations to be efficiently utilized by SML microbes.


Surface concentrations of bacteria can be several orders of magnitude greater than in bulk water (DiSalvo 1973). The partial pressure of CO2 is greater in the upper few centimeters, indicating enhanced respiration (Dandonneau etal 2008). Open ocean and estuarine surface communities contain species adapted to metal toxicity (Schulz-Baldes 1992), varying temperature and salinity (Katsaros and Buettner 1969), and irradiance (Cheng 1981). Vibrio and Pseudomonas bacteria both thrive in the nutrient-rich TEP matrix. The dynamic environment mean decreased diversity of SML species compared to the general marine assemblage (Lion and Leckie 1981). Viruses are also abundant, with a partitioning ratio around 15 (Kuznetsova etal 2004). Bacterial membranes damaged or lysed by viruses leak amino acids into the surrounding water. DOM partitioning is correlated with bacterial surface abundance, suggesting that in situ production is a major source. Surface concentrations of particulate amino acids (POM) are not correlated with bacterial abundance (Kuznetsova etal 2004), and therefore must be buoyantly transported as previously mentioned. Damaged bacteria are more particle reactive and will be vertically transported, although it is unknown whether this contributes to or reduces surface partitioning. In fact, lower morbidity rates are observed in the bacterioneuston than in the subsurface community or at depth.

Physical processes

The transport of surface-adhered and buoyant material necessarily deviates from that of density driven circulation, with distribution dictated by wind-driven and geostrophic surface currents (Dandonneau et al 2008). FBM is formed in productive upwelling areas and accumulates in convergence zones, where its positive buoyancy resists downward surface water velocities (Dandonneau etal 2008). In the Atlantic and Pacific, FBM is produced at the equator, and is transported poleward on the leading edge of tropical instability vortices to collect in high concentration bands along 5°N and 5°S, where vertical water velocities are about -0.25 m day-1 (Dandonneau etal 2008). In the Indian Ocean FBM accumulated in a thin band north of the equator, and in the western Arabian Sea.

The model does not have a fine enough resolution to account for entrainment of FBM in mesoscale eddies, Langmuir cells and turbulent flow which are likely responsible for small scale variability (Schulz-Baldes 1992; Wurl and Holmes 2008). It does, however, successfully demonstrate a decoupling of surface material from patterns of primary productivity. Chlorophyll concentrations were three times greater in upwelling areas than in oligotrophic convergence zones, while FBM concentrations were ten times greater in convergence zones than in upwelling areas. This transported OM pool fuels open ocean ecosystems, from bacteria to pelagic apex predators like sharks (Dandonneau et al 2008).

Convergent physical processes which concentrate surface material also inhibit atmospheric exchange of CO2, N2O, O2, N2 and CH4 (among other gases) by establishing a molecular barrier (Lion and Leckie 1981; Wurl and Holmes 2008; Cunliffe and Murrell 2009). This areal compression increase molecular encounter rates, and leads to increased POM formation and aggregation (Wurl and Holmes 2008). Surface biofilms also contribute to the damping of capillary waves (observable as a decrease in sea surface roughness) and reduce the surface area available for gas exchange.

High winds can mix the SML and subsurface waters, but surface films quickly reform since much of the material is buoyant. Increased air bubbles in subsurface water may even promote abiotic POM formation, although how this compares to physical breakdown by wave action is unknown. At high wind speed, the surface area of water spray can exceed the surface area of the underlying ocean. Spray and fog contain water from the SML and have a similar level of chemical partitioning (Lion and Leckie 1981) which facilitates the exchange of surface enriched elements.

The neuston

Organisms which comprise the neuston include the microbial community, primary producers, and the zooplankton which graze on them. The pleuston includes organisms which actually live in contact with air-water membrane (Schulz-Baldes 1992). For instance, the water striders of the genera Halobates are marine insects with live spent skating on the interface, supported by surface tension (Cheng 1981). Choanoflagellates hang suspended from the interface, or colonize TEP, and feed on microbes (Lion and Leckie 1981). These organisms are sometimes considered oddities, and not necessarily major players in the larger ecological picture. They can, however, be used as bioassays of heavy metal surface enrichment.

Cadmium, for instance, has a dissolved concentration of about 1.0 ng L-1 in seawater. Neustonic organisms contribute a sizable pool of 0.2 ng L-1 while the subsurface community contributes only 0.125 ng L-1 (Schulz-Baldes 1992). Schulz-Baldes (1992) showed higher Cd, Cu, and Pb concentrations in Atlantic Halobates micans than in copepods and hyperiid amphipods, which have vertical migration patterns. Uptake in Halobates is by drinking surface water and feeding directly at the enriched interface (Cheng 1981).


Partitioning in the SML is a ubiquitous but highly variable phenomena. Trace elements and microbes may have concentrations ten thousand times greater than in subsurface water (Lion and Leckie 1981; DiSalvo 1973), while DOM partitioning can be only by a factor of two (Zhang et al 2003). Chemical, physical and biological processes control the level of enrichment. Of particular importance is the buoyant flux of organic matter in form of TEP, which scavenges microbes and trace elements and brings them to the surface. High insolation reduces photosynthesis in the SML, and net respiration is responsible for breaking down TEP.

While this layer has been studied for decades, it is only recently that buoyant flux has been considered a major contributing process. This led to a re-evaluation of the SML definition, establishing that it is a gelatinous polysaccharide matrix that supports an extensive microbial community. This model of surface layer gradients will continue evolving (Zhang et al 2003).

Surface currents transport material to unproductive convergence zones, where TEP can represent a substantial portion of the carbon pool. HMW pollutants will be transported in the same manner, and may interfere with regenerated production. The SML also has implications for atmospheric exchange of green house gases, which can be enhanced or retarded by sea state and biogenic partitioning. Generalized partitioning models are hard to develop, because of small scale variability due to physical processes. I expect that as remote sensing algorithms continue to improve, detection of specific sea surface components will be possible over large geographic area. With this data for ground-truthing fine-scale ocean models, overall SML contribution to ocean processes will be more readily estimated.


  1. Azetsu-Scott K, U Passow. 2004. Ascending marine particles: Significance of transparent exopolymer particles (TEP) in the upper ocean. Limnology and Oceanography 49(3), 741-748.
  2. Cheng L. 1981. Biology of Halobates (Heteroptera: Gerridae). Annual Review of Entomology 30, 111-135.
  3. Cunliffe M, JC Murrell. 2009. The sea-surface microlayer is a gelatinous biofilm. The ISME Journal 3, 1001-1003.
  4. Dandonneau Y, C Menke, O Duteil, T Gorgues. 2008. Concentration of floating biogenic material in convergence zones. Journal of Marine Systems 69, 226-232.
  5. DiSalvo LH. 1973. Contamination of surfaces by bacterial neuston. Limnology and Ocenaography 18(1), 165-168.
  6. Katsaros K, KJK Buettner. 1969. Influence of rainfall on temperature and salinity of the ocean surface. Journal of Applied Meteorology 8, 15-18.
  7. Kuznetsova M, C Lee, J Aller, N Frew. 2004. Enrichment of amino acids in the sea surface microlayer at coastal and open ocean sites in the North Atlantic Ocean. Limnology and Oceanography 49(5), 1605-1619.
  8. Lion LW, JO Leckie. 1981. The biogeochemistry of the air-sea interface. Annual Review of Earth and Planetary Science 9, 449-486.
  9. Schulz-Baldes M. 1992. Baseline study on Cd, Cu and Pb concentration in Atlantic neuston organisms. Marine Biology 112, 211-222.
  10. Wurl O, M Holmes. 2008. The gelatinous nature of the sea-surface microlayer. Marine Chemistry 110, 89-97.
  11. Zhang Z, L Liu, C Liu, W Cai. 2003. Studies on the sea surface microlayer II. The layer of sudden change of physical and chemical properties. Journal of Colloid and Interface Science 264, 148-159.
  12. Zhou X, K Mopper. 1996. Photochemical production of low-molecular-weight carbonyl compounds in seawater and surface microlayer and their air-sea exchange. Marine Chemistry 56, 201-213.