Scavenging and fractionation of thorium vs. protactinium in the ocean, as determined from particle–water partitioning experiments with sediment trap material from the Gulf of Mexico and Sargasso Sea

https://doi.org/10.1016/j.epsl.2009.06.029Get rights and content

Abstract

Particle–water partition coefficients (Kd) for thorium (Th) and protactinium (Pa) were calculated for sediment trap material collected from the Gulf of Mexico (GOM) in the upper 140 m and archived sediment trap material collected in the mesopelagic and bathypelagic Sargasso Sea (500, 1500 and 3200 m sediment trap samples from the Oceanic Flux Program (OFP) site off Bermuda). Results showed that Kd values for Th were generally greater than for Pa when values were smaller, but converging to a 1:1 line at higher values. Furthermore, results showed that, though the contents of polysaccharides and calcium carbonate are significantly correlated for all samples, log Kd of Th (and Pa) values are correlated with the polysaccharide content while no correlation is apparent with CaCO3, Mn and Fe. Since polysaccharides are not generally regarded as strongly chelating agents for actinides, we hypothesize that other co-occurring organic phases originating from the matrix of carbohydrate-rich extracellular polymeric substances (EPS), could be responsible for binding the actinides. These data provide, for the first time, direct evidence for the role of carbohydrate-rich EPS being responsible for the differential scavenging of these two radioisotopes, i.e., their fractionation.

Introduction

Thorium and protactinium isotopes have long been used as tracers of oceanic processes. In particular, the ratio of the two longer-lived 231Pa and 230Th, have been used to assess boundary scavenging, ocean circulation, and paleoproductivity (e.g., Anderson et al., 1983a, Anderson et al., 1983b, Kumar et al., 1995, Walter et al., 1997). However, thorium to protactinium ratios in suspended or sinking particles can vary as a function of location, depth and size (Walter et al. 2001). Based on field or laboratory experiments, there are several mutually exclusive theories as to what causes the fractionation between these two isotopes. Most maintain that the cause of Th/Pa fractionation lies in the type of major inorganic sorbents exhibiting different affinities to Pa vs. Th. Anderson et al., 1983a, Anderson et al., 1983b attributed Th/Pa fractionation effects to the role of MnO2. Some, like Taguchi et al., 1989, Rutgers van der Loeff and Berger, 1993, Walter et al., 1997 attribute fractionation effects to preferential removal of Pa by opal, while Luo and Ku (1999) proposed that they are due to changing biogenic opal to clay ratios, or to the clay content (Luo and Ku, 2004). The work by Chase et al., 2002, Chase et al., 2003, on the other hand, suggested that Th/Pa fractionation is due to changing opal to carbonate ratios. Finally, Li (2005) pointed out that one cannot ignore the role of organic matter in controlling Th/Pa fractionation. As a consequence of Li's suggestion, one might expect that Pa(V) is first reduced to the more particle reactive Pa(IV) by redox-active moieties in natural organic matter, e.g., hydroquinones (e.g., Lovely and Blunt-Harris, 1999), allowing it then to be effectively scavenged from the water column.

Work by our group has investigated the importance of different types of macromolecular organic matter on the sorption of actinides onto organic or inorganic particles and have come to the conclusion that it is not only the presence of organic moieties that control the extent of adsorption onto particles, but especially the type of organic matter. For example, laboratory experiments with varying types of organic matter showed that the polysaccharide-enriched fraction of marine colloidal organic matter (COM) had a higher partition coefficient (Kdc) for Th than that of the bulk COM or other known mineral sorbent. Interestingly, the log Kc value for Th of this natural polysaccharide-rich organic material is very similar to the highest value obtained from model acid polysaccharides (Quigley et al., 2002). Further evidence for the importance of organic moieties in controlling Th removal from the ocean is reviewed in Santschi et al. (2006).

Like thorium, protactinium is also a highly particle reactive element. However, the chemistry of protactinium is more complex in large part because it has three possible oxidation states (IV, V, VI) in natural waters (Alhassanieh et al, 1999). The predominant form is likely Pa(V) in oceanic systems (Choppin, 1983, Choppin, 2003, Choppin, 2007), which, as a PaO2+ ion, is quite soluble, can be complexed by F in seawater, and sorbs to particle surfaces to a lesser extent than the four-valent oxidation state ion, Pa(IV). It is likely that Pa(V) must first be reduced to Pa(IV) before stronger adsorption and removal from natural waters can take place, as has been seen for Np(V), which is reduced in the presence of natural organic matter such as humics (Zeh et al., 1999). Organic ligands such as extracellular polymeric substances (EPS) produced by bacteria and phytoplankton are another possible source of reductants. Very little is known not just about the complexation of Pa with EPS, but also about their redox behavior in aquatic systems. What is known is the redox potential for Pa(V) to Pa(IV) reduction is −0.05 V, as shown below.Pa(V)O2+ + 2H2O + e = Pa(IV)(OH)4; Eho =  0.05 V (Greenwood, N.N., and Earnshaw, A. 1998.)Biologically mediated redox reactions, as well as abiotic reactions by simple organic molecules, such as aldehydes, ferridoxin, and semiquinones, have Eho(w) values below −0.05 V at pH of 7 (Stumm and Morgan, 1996), and thus, could be capable of reducing Pa(V) to Pa(IV).

In order to determine the possible carrier phases that cause differential scavenging of thorium and protactinium, i.e., their fractionation, this study compared log Kd values of these radionuclides that were determined from laboratory tracer experiments on board ship a using freshly collected sediment trap material from the upper 150 m in the Gulf of Mexico or in the laboratory using archived Ocean Flux Program (OFP) sediment trap material collected at mesopelagic and bathypelagic depths (500, 1500, and 3200 m) in the Sargasso Sea offshore Bermuda. We then compare our data with those from the literature.

Section snippets

Gulf of Mexico (GOM)

In May 2006, surface sediment traps (65, 90, 120, and 140 m water depth) were deployed for a 48 hour period at three stations in the Gulf of Mexico (Table 1). Station 1 was in a Cold Core Eddy (25° 52.31′N, 92° 31.10′W), Station 2 was in a Warm Core Eddy (26° 54.59′N, 89° 59.45′W and Station 3 in a Cold Core Eddy with coastal upwelling features (27° 52.11′N, 88° 38.448′W). After recovery, material in the plastic tubes was allowed to settle before draining off the surface water. The remaining

Gulf of Mexico (GOM)

Log Kd values in the GOM ranged from 5.22 to 6.62 for 234Th and 3.96 to 4.66 for 231Pa. Log Kd values of Th and Pa increased with depth, except at Station 3 (Fig. 1), which exhibited coastal upwelling and lateral transport features. The log Kd values of the radioisotopes are significantly correlated with each other (p < 0.05 for 233Pa vs. 234Th), with log Kd(Th) > log Kd(Pa).

When the colloidal fraction is included with the particulate (log Kdc), the log Kdc values for 233Pa become closer, but not

Discussion

As can be seen from Fig. 2, log Kd values for both isotopes and from both GOM and OFP agree well with previously determined data from sediment trap material from the Atlantic Ocean (Nyffeler et al., 1984). When all log Kd(Th) values are plotted against log Kd(Pa) values, the F(Th/Pa) values cluster around 1 at higher Kd values, indicating little fractionation between the two radionuclides, while at lower Kd values, F(Th/Pa) values increase to a maximum of 42 (Fig. 2). The fact that the GOM

Conclusions

Potential carrier phases for Th and Pa nuclides in the ocean that had previously been proposed include calcium carbonate, Fe and Mn oxides, lithogenic and organic carbon phases. Using correlations with data from radiotracer partitioning studies to sediment trap material from the Gulf of Mexico and the Ocean Flux Program off Bermuda, only an organic carbon phase, i.e., the polysaccharide fraction of organic carbon, correlated with Log Kd values of Th and Pa radioisotopes. Both log Kd(Th) and log

Acknowledgements

This work was funded, in parts, by NSF project OCE-0351559 to PHS, KAR and CCH, the Welch Foundation (Grant number BD-0046) to XC, and NSF projects OCE-0325627, OCE-0509602 and OCE-0623505 to MHC. We wish to thank the crew of the R/V Seward Johnson for all of their help in the Gulf of Mexico.

References (48)

  • C.-C. Hung et al.

    Distributions of carbohydrate species in the Gulf of Mexico

    Mar. Chem.

    (2003)
  • G.G. Leppard

    The characterization of algal and microbial mucilages and their aggregates in aquatic ecosystems

    Sci. Total Environ.

    (1995)
  • Y.-H. Li

    Controversy over the relationship between major components of sediment-trap materials and the bulk distribution coefficients of 230Th, 231Pa, and 10Be

    Earth Planet. Sci. Lett.

    (2005)
  • S. Luo et al.

    Oceanic 231Pa/230Th ratio influenced by particle composition and remineralization

    Earth Planet. Sci. Lett.

    (1999)
  • S. Luo et al.

    On the importance of opal, carbonate and lithogenic clays in scavenging and fractionating Th-230, Pa-231, and Be-10 in the ocean

    Earth Planet. Sci. Lett.

    (2004)
  • S.M. Myklestad et al.

    A sensitive and rapid method for analysis of dissolved mono- and polysaccharides in seawater

    Mar. Chem.

    (1997)
  • U.P. Nyffeler et al.

    A kinetic approach to describe trace element distribution between particles and solution in natural aquatic systems

    Geochim. Cosmochim. Acta

    (1984)
  • U. Passow et al.

    Organic carbon to 234Th ratios of marine organic matter

    Mar. Chem.

    (2006)
  • M.M. Rutgers van der Loeff et al.

    Scavenging of 230Th and 231Pa near the Antarctic Polar Front in the South Atlantic

    Deep-Sea Res. I

    (1993)
  • P.H. Santschi et al.

    Thorium speciation in seawater

    Mar. Chem.

    (2006)
  • P.K. Smith et al.

    Measurement of protein using bicinchoninic acid

    Anal. Biochem.

    (1985)
  • K. Taguchi et al.

    Particulate removal of 230Th and 231Pa in the biologically productive northern North Pacific

    Earth Planet. Sci. Lett.

    (1989)
  • H.J. Walter et al.

    Enhanced scavenging of 231Pa relative to 230Th in the South Atlantic south of the Polar Front: implications for the use of the 231Pa/230Th ratio as a paleoproductivity proxy

    Earth Planet. Sci. Lett.

    (1997)
  • H.J. Walter et al.

    Shallow vs. deep-water scavenging of Pa-231 and Th-230 in radionuclide enriched waters of the Atlantic sector of the Southern Ocean

    Deep-Sea Res. I

    (2001)
  • Cited by (35)

    • On the cycling of <sup>231</sup>Pa and <sup>230</sup>Th in benthic nepheloid layers

      2021, Deep-Sea Research Part I: Oceanographic Research Papers
    • Kinetics of thorium and particle cycling along the U.S. GEOTRACES North Atlantic Transect

      2017, Deep-Sea Research Part I: Oceanographic Research Papers
    View all citing articles on Scopus
    1

    Present address: Savannah River National Laboratory, Aiken, SC 29808, United States.

    View full text