Vanadium isotope composition of seawater

The speciation, burial, and isotopic composition of vanadium (V) in seawater is predicted to be closely coupled to the redox state of the oceans. While the speciation and burial terms are reasonably constrained, the V isotopic composition of seawater has remained elusive owing to significant analytical challenges. To this end, for the first time we have developed and validated a new method to purify V from large volume (>= 500 mL) seawater samples that we used to determine the V isotopic composition of seawater. Our method comprises four discrete V-purification steps that exploit ion-exchange chromatography including Nobias chelate and anion exchange resin(s) and measurement by multi-collector inductively-coupled plasma mass spectrometry. Results from several samples with addition of standard V solution with known isotope composition show no isotopic deviation in the chemical and/or analytical procedures and our reproducibility is within typical analytical error for vanadium isotopes measurements. Though V yields were non-quantitative (averaging approximate to 70%) for natural seawater samples, our approach was nonetheless validated with additional experiments. Therein, synthetic seawater solutions of known V isotopic composition with concentration similar to natural seawater were used to confirm that there is limited isotope fractionation during analytical procedures with similar yields. We further tested seawater samples using UV radiation, HNO3/HCl oxidation, and High-Pressure Asher treatments to ensure there was limited effects from potential non-dissolved phases or variable V speciation such as organic ligand binding of V. All tests except the High-Pressure Asher samples had similar recoveries (i.e. >70%) and all recorded similar isotopic values within error which suggest our method is robust and reliable for V isotopic measurement of seawater. Using our most optimal method, we report V isotope data for several seawater samples from surface and subsurface Atlantic Ocean and deep Pacific Ocean for the first time. Inter-laboratory sample comparison shows that the data was within analysis error (similar to 0.15 parts per thousand, 2SD). Initial results imply that the deep ocean is isotopically homogeneous with respect to V, and the uptake of V in surface waters appear to cause very limited, if any, isotope fractionation as it is within analytical uncertainties. Thus, our results suggest a reference seawater V isotope composition of + 0.20 +/- 0.15 parts per thousand relative to V isotope standard AA solution. This work analyzes the first V isotope value of seawater which provides a key foundation for future work to constrain the modern marine V isotope cycle and budget and application for paleoceanographic research. (C) 2018 Elsevier Ltd. All rights reserved.