SEARCH · Engineering Papers
Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.
Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.
Explore the source record for details and available documents.
Hypothesis: The precipitation and dissolution of aluminum-bearing mineral phases in aqueous systems often proceed via changes in both aluminum coordination number and connectivity, complicating molecular-scale interpretation of the transformation mechanism. Here, the thermally induced transformation of crystalline sodium aluminum salt hydrate, a phase comprised of monomeric octahedrally coordinated aluminate which is of relevance to industrial aluminum processing, has been studied. Because intermediate aluminum coordination states during melting have not previously been detected, it is hypothesized that the transition to lower coordinated aluminum ions occurs within a highly disordered quasi-two-dimensional phase at the solid-solution interface. Experiments and simulations: In this work, in situ X-ray diffraction (XRD), Raman and 27 Al nuclear magnetic resonance (NMR) spectroscopy were used to monitor the melting transition of nonasodium aluminate hydrate (NSA, Na 9 [Al(OH) 6 ] 2 ·3(OH)·6H 2 O). A mechanistic interpretation was developed based on complementary classical molecular dynamics (CMD) simulations including enhanced sampling. A reactive forcefield was developed to bridge speciation in the solution and in the solid phase. Findings: In contrast to classical dissolution, aluminum coordination change proceeds through a dynamically stabilized ensemble of intermediate states in a disordered layer at the solid-solution interface. In both melting and dissolution of NSA, octahedral, monomeric aluminum transition through an intermediate of pentahedral coordination. The intermediate dehydroxylates to form tetrahedral aluminate (Al(OH)$_4^–$) in the liquid phase. This coordination change is concomitant with a breaking of the ionic aluminate-sodium ion linkages. The solution phase Al(OH)$_4^–$ ions subsequently polymerize into polynuclear aluminate ions. However, there are some differences between bulk melting and interfacial dissolution, with the onset of the surface-controlled process occurring at a lower temperature (~30 °C) and the coordination change taking place more gradually as a function of temperature. This work to determine the local structure and dynamics of aluminum in the disordered layer provides a new basis to understand mechanisms controlling aluminum phase transformations in highly alkaline solutions.
Metal–organic framework (MOF) heterostructures exhibit unique properties beyond those of individual components, but their design requires an understanding of energetic and kinetic controls at MOF–substrate interfaces. Although the structural relationship has been widely used in heterostructure design, it overlooks the interplay between the organic ligand and the substrate which controls the kinetics and energetics of growth of the final structure. In this study, we used zeolitic imidazolate frameworks (ZIF-8) on ZnO as a model system to evaluate this interplay via in situ monitoring and simulations. Our results demonstrate multiple roles of the 2-methyl-imidazole (2-MIM) ligand as “dissolution-promoter”, “step-pinner”, and “terrace-binder” on the ZnO (001) face and “dissolution-promoter” and “terrace-binder” on the ZnO (100) face. Through these multiple face-specific roles, 2-MIM modulates ZnO dissolution kinetics and, hence, the Zn 2+ release rate, tuning local supersaturations that dictating the characteristic ZIF-8 crystallization kinetics on different substrate faces. The critical thickness for transition from 2D to 3D growth is dictated by competition between interfacial and strain energies. The atomic-scale mechanism of the coupled substrate dissolution and MOF growth, mediated by selective linker–substrate binding, furnishes a new synthesis pathway for other complex heterostructures.
Low-salinity waterflooding (LSWF) has proven to improve oil recovery in carbonate formations through rock wettability alteration, although the underlying mechanism remains elusive. Multivalent ionic exchange and calcite dissolution have usually been investigated using geochemical analysis in secondary coreflooding. In this work, coreflooding, in tertiary mode, coupled with a surface reactivity analysis approach was employed to investigate the interplay of wettability alteration mechanisms such as mineral dissolution, electrostatic bond attraction, and the effect of pH at in situ conditions. Improved oil recovery (IOR) in tertiary mode observed by coreflooding in Indiana limestone rocks showed an ionic strength dependence, that is, reducing brine ionic strength resulted in an increase in oil recovery. Coreflooding results showed that the seawater and low-salinity brines deprived of Mg 2+ ions resulted in the lowest IOR in tertiary mode, indicating the significance of Mg 2+ on IOR in limestone rocks. Similar results were observed through the contact angle measurement showing the limestone rock wettability state dependence on ionic strength and the effect of Mg2+ ions. Surface reactivity analysis showed an increase in solution pH, Ca 2+ and Mg 2+ ions concentration in the effluent solution from the coreflooding in tertiary mode using low salinity brines (about 40 and 20% increase in the effluent composition for Ca 2+ and Mg 2+ , respectively). These changes in solution composition were used to calculate the in situ oil–brine and rock–brine zeta potential using a validated surface complexation model, showing the changes of zeta potential as brine is injected into limestone rocks. The results show that using seawater-like brine in tertiary mode resulted in no mineral dissolution or ionic exchange. However, improved oil recovery (IOR) using such seawater-like brine was due to wettability alteration caused by reduced electrostatic bond attraction associated with Mg 2+ ions [from 2.6 × 10 –13 (mol/m 2 ) 2 for formation water salinity to 1.5 × 10 –13 (mol/m 2 ) 2 for seawater salinity]. Using low-salinity brines in tertiary mode improved oil recovery by mineral dissolution, resulting in oil desorption and an increase in solution pH. Finally, the increase in solution pH also resulted in reduced electrostatic bond attraction which lead to rock wettability alteration using low-salinity brines.
Hybrid halide perovskites display a great tunability of optoelectronic properties and environmental stability by controlling the halogen anions (e.g., I, Cl, and Br). Yet, their water interaction and degradation mechanisms are not completely elucidated. In this work, the interaction of Cs 2 SnCl 6 and Cl-enriched solid solution Cs 2 SnI 0.9 Cl 5.1 with water was systematically studied by in situ synchrotron X-ray diffraction and micro-Raman spectra and compared with the isostructural Cs 2 SnI 6 . Unlike Cs 2 SnI 6 , which experiences a direct dissolution in water, Cs 2 SnCl 6 displays an enhanced stability and the dissolution of the Cs 2 SnCl 6 accompanies with the formation of an amorphous alteration phase. Under controlled dehydration conditions, two-dimensional Cs 2 SnCl 6 flakes can be precipitated out from water solution. Moreover, the mixed halide perovskite (Cs 2 SnI 0.9 Cl 5.1 ) experiences fast iodide dissolution in water solution and transforms to a more chloride-enriched perovskite which shows a behavior similar to Cs 2 SnCl 6 . The mechanistic understanding of the dissolution–precipitation process of Cs 2 SnI x Cl 6–x perovskites is useful for developing new perovskites with varied halogen and controlled environmental stability.
Mineral and glass dissolution is a scientific topic deeply investigated but incompletely understood and of a great interest for the geochemical and materials science communities. If the interfacial dissolution/reprecipitation mechanism seems to be applicable to most of silicate minerals, the debate remains open concerning glass. Here we studied two model glasses, a ternary borosilicate (CJ1) and the same glass doped with 4.1 mol% of Al2O3 (CJ2). The two glasses were altered at 90°C, pH 9, and in conditions far and close to saturation with respect to amorphous silica, to determine the initial and residual rates. Moreover, a specific experiment was conducted for a short duration with a solution highly enriched with 18O and 29Si isotopes to understand how passivating gels form. SEM, TEM and ToF-SIMS characterization, along with Monte Carlo simulations were used to understand the rate limiting reactions at play and infer the role of Al. We show that Al yields a slower matrix dissolution in dilute conditions. However, it slows down he formation and the maturation of the passivating gel and favors alteration by partial hydrolysis of Si and Al entities followed by in-situ reorganization/relaxation into a porous network. Unexpectedly, CJ1 experienced both interfacial dissolution/reprecipitation and partial hydrolysis followed by in situ reorganization of the silicate network during the course of a single experiment. This study offers a unified concept that can pave the way for the future development of a predictive kinetic model based on a detailed description of bond breaking and bond forming as a function of glass composition and alteration conditions.
An accelerated development of durable and affordable sustainable energy technologies is often hindered by a limited understanding of how non-precious materials within these systems degrade. In acidic proton exchange membrane fuel cells and water electrolyzers, metallic cobalt (Co) is considered an unstable component that is often combined with precious metals or other stabilizers. To understand the mechanisms behind Co instability, we employ an experimental platform that quantifies dissolution with on-line inductively coupled plasma mass spectrometry and product formation with electrochemical mass spectrometry during electrochemical testing, along with ex- situ characterization. Under varied conditions (electrocatalysis, time, gas-type saturation, and ion concentration), windows of Co stability are observed that are different than predicted with classical chemical thermodynamics, suggesting new stabilization and degradation mechanisms than previously understood. Notably, Co is active for the hydrogen evolution reaction (HER), with prolonged stability that is ~300 mV different than thermodynamically projected. Additionally, in an oxygenated environment, Co concurrently performs the HER and oxygen reduction reaction (ORR) yet undergoes different morphology changes and dissolution mechanisms. Interestingly, in the absence of electrocatalysis, there is a 22x decrease in dissolution in an oxygen-free environment, proposing a route to decrease Co losses during device shutdown protocols. Lastly, under more extreme operating conditions, Co becomes stable after a substantial amount of dissolution, suggesting that high concentrations of Co 2+ ions in the microenvironment induce the formation of a stable CoHO 2 surface. Altogether, these results can be leveraged to improve the design and development of more robust and cost-effective sustainable energy technologies, as well as promote strategic strategies for prolonged material utilization.
Carbon mineralization to solid carbonates is one of the reaction pathways that can not only utilize captured CO 2 but also potentially store it in the long term. In this study, the dissolution and carbonation behaviors of alkaline solid wastes (i.e., waste concrete) was investigated. Concrete is one of the main contributors to a large carbon emission in the built environment. Thus, the upcycling of waste concrete via CO 2 utilization has multifaceted environmental benefits including CO 2 emission reduction, waste management and reduced mining. Unlike natural silicate minerals such as olivine and serpentine, alkaline solid wastes including waste concrete are highly reactive, and thus, their dissolution and carbonation behaviors vary significantly. Here, both conventional acid (e.g., hydrochloric acid) and less studied carbonic acid (i.e., CO 2 saturated water) solvent systems were explored to extract Ca from concrete. Non-stoichiometric dissolution behaviors between Ca and Si were confirmed under far-from-equilibrium conditions (0.1 wt% slurry density), and the re-precipitation of the extracted Si was observed at near-equilibrium conditions (5 wt% slurry density), when the Ca extraction was performed at a controlled pH of 3. These experiments, with a wide range of slurry densities, provided valuable insight into Si re-precipitation phenomena and its effect on the mass transfer limitation during concrete dissolution. Next, the use of the partial pressure of CO 2 for the pH swing carbon mineralization process was investigated for concrete, and the results were compared to those of Mg-bearing silicate minerals. In the P CO 2 swing process, the extraction of Ca was significantly limited by the precipitation of the carbonate phase (i.e., calcite), since CO 2 bubbling could not provide a low enough pH condition for concrete–water–CO 2 systems. Furthermore, this study showed that the two-step carbon mineralization via P CO 2 swing, that has been developed for Mg-bearing silicate minerals, may not be viable for highly reactive Ca-bearing silicate materials (e.g., concrete). The precipitated calcium carbonate (PCC) derived from waste concrete via a pH swing process showed very promising results with a high CO 2 utilization potential as an upcycled construction material.
Borosilicate glasses are widely used for radioactive waste disposal due to their ability to incorporate a variety of contaminants and radionuclides while exhibiting high durability in various disposal scenarios. This research evaluated the dissolution of borosilicate glass using both single-pass-flow-through (ASTM C1662-18) and product consistency test (ASTM C1285-21) methods with different solutions, including a cementitious-contacted water (called grout-contacted, GC, from this point) and solutions with varying levels of dissolved cementitious species such as Si, Ca, Al. Here, the results indicated that the presence of Ca plays a crucial role in suppressing glass corrosion, as evidenced by the slower normalized dissolution rates, which were one order of magnitude lower for boron and two orders of magnitude lower for rhenium, observed in both Ca-amended and GC solutions compared to the pH 12 buffer solution. This effect is attributed to the formation of a dense, low-porosity, and strongly bonded calcium silicate hydrate (CSH) layer on the glass surface, which implies that a glass corrosion process is influenced by ion exchange involving alkali ions Na + , K + , Ca 2+ , and hydrogen-containing species. A small number of glass particles treated in the GC solution showed minor corrosion pits in the form of shallow craters with an average diameter of approximately 500 μm. This observation is correlated with a significant reduction, 2000 to 3000 times lower, in the cumulative volume of glass pores, indicating that smaller pore voids were “sealed” in the presence of Ca 2+ ions, likely attributed to the formation of CSH precipitation or other corrosion products such as calcium carbonate saturated from the grout solution. These findings suggest that the presence of dissolved Ca in the GC solution can slow down the dissolution of borosilicate glass, contrary to the expected trend of higher dissolution rates resulting from exposure to high alkaline and thus higher pH solutions.
Abstract Direct extraction of used nuclear fuel (UNF) in an organic solution could be more efficient than the previous practice with aqueous solutions. However, the UNF would need to be treated via voloxidiation before being processed using solvent extraction. The voloxidation process can form oxide and/or nitrate compounds. This work investigated the dissolution of uranium/lanthanide oxides in 30 vol % TBP diluted in dodecane (pre-equilibrated with 4 M nitric acid) in a glass reactor with air sparging to ascertain the uranium/lanthanide oxide dissolution behavior prior to scaling the process. The uranium/lanthanide oxides were prepared by co-precipitating uranium/lanthanide nitrates with hydroxide and then calcined to form a mixed oxide. While the relative concentrations of the lanthanides are not representative of used nuclear fuel, the neodymium and erbium allowed ease of tracking dissolution with visible spectroscopy. Cerium was used as a surrogate for plutonium. The dissolution rate of the oxides was similar but incomplete. A miniscule amount of cerium, as cerium oxide, took several months to slowly dissolve; however, when co-precipitated with uranium and other lanthanides, a significant amount of cerium dissolved readily.
The Idaho National Laboratory (INL) is actively engaged in the development of electrochemical processing technology for the treatment of fast reactor fuels using irradiated fuel from the Experimental Breeder Reactor-II (EBR-II) as the primary test material. The research and development (R&D) activities generate a low enriched uranium (LEU) metal product from the electrorefining of the EBR-II fuel and the subsequent consolidation and removal of chloride salts by the cathode processor. The LEU metal ingots from past R&D activities are currently stored at INL awaiting disposition. One potential disposition pathway is the shipment of the ingots to the Savannah River Site (SRS) for dissolution in H-Canyon. Carbon steel cans containing the LEU metal would be loaded into reusable charging bundles in the H-Canyon Crane Maintenance Area and charged to the 6.4D or 6.1D dissolver. The LEU dissolution would be accomplished as the final charge in a dissolver batch (following the dissolution of multiple charges of spent nuclear fuel (SNF)). The solution would then be purified and the 235 U enrichment downblended to allow use of the U in commercial reactor fuel. To support this potential disposition path, the Savannah River National Laboratory (SRNL) developed a dissolution flowsheet for the LEU using samples of the material received from INL.
Byproduct residues generated from Aqueous Chloride Pu and Am recovery operations at Los Alamos National Laboratory have been investigated and determined to contain high weight % chloride salts and minimal hydrogen content. Hydrogen content is the result of limited surface area adsorption of ambient moisture. The assumption used in technical analysis NCS-TECH-19-029 to model potentially hydrogenous items as mixtures of polyethylene is shown to be overly conservative. Our results and literature precedent indicate that water adsorption of 1-3 % by weight is expected for hydroxide cakes and Pu oxides, and a conservative bounding assumption for these materials is 5% water by weight. LOH results for a dissolution heel stored long term showed 6.25% weight loss, which was conservatively assumed to be water. A conservative bounding assumption for dissolution heels is 8 wt.% water content. This equates to 0.55 % hydrogen by weight for hydroxide cakes, and 0.88% hydrogen by weight for dissolution heels. The bulk composition of dissolution heels reflects the composition of the original byproduct feed material, chloride salts of potassium, sodium, magnesium and calcium. The bulk composition of calcined hydroxide precipitation residues is composed of potassium chloride as a result of potassium hydroxide neutralization of hydrochloric acid solutions. Metal hydroxides generated during the hydroxide precipitation process are decomposed under the conditions of calcination to metal oxide forms.
At Savannah River Site (SRS), High-Level Waste is stored in below-grade tanks constructed of carbon steel. This waste is composed of sludge, salt cake, and/or supernate. In part, preparation of this waste for future processing involves dissolution of the salt cake layer. The salt dissolution process can create conditions that leave the carbon steel tanks susceptible to a number of corrosion processes. The salt to be dissolved contains high concentrations of nitrate, that once released, create an environment that may be conducive to pitting corrosion or stress corrosion cracking (SCC) of carbon steel. Additionally, during the salt dissolution process, in the absence of mixing, stratification of the supernatant liquid may occur. This can result in less dense, more dilute waste layers occurring higher in the tank. In these more dilute waste layers, the susceptibility to localized corrosion could potentially differ from that of the more concentrated salt solutions evaluated in previous testing, as the amount of inhibiting and aggressive species, not just the ratios, can affect susceptibility. Evaluation of the susceptibility to localized corrosion in these more dilute waste chemistries could provide insight into the amount of inhibitors required to effectively mitigate pitting corrosion in stagnant salt dissolution environments, as well as other tank farm operations involving dilute waste streams.
We studied the effects of supercritical carbon dioxide (scCO2) on the matrix permeability of reservoir rocks from the Eagle Ford, Utica, and Wolfcamp formations. We measured permeability using argon before exposure of the samples to scCO2 over time periods ranging from days to weeks. We measured permeability (and the change of permeability with confining pressure) when both argon and scCO2 were the pore fluids. In all three formations, we generally observe a negative correlation between initial permeability and carbonate content—the higher the carbonate content, the lower the initial permeability. In clay- and organic-rich samples, swelling of the matrix resulting from adsorption decreased the permeability by about 50% when the pore fluid was scCO2 although this permeability change is largely reversible. In carbonate-rich samples, dissolution of carbonate minerals by carbonic acid irreversibly increased matrix permeability, in some cases by more than one order of magnitude. This dissolution also increases the pressure dependence of permeability apparently due to enhanced mechanical compaction. Despite these trends, we observed no general correlation between mineralogy and the magnitude of the change in permeability with argon before and after exposure to scCO2. Flow of scCO2 through μm-scale cracks appears to play an important role in determining matrix permeability and the pressure dependence of permeability. Extended permeability measurements show that while adsorption is nearly instantaneous and reversible, dissolution is time-dependent, probably owing to reaction kinetics. Our results indicate that the composition and microstructure of matrix flow pathways control both the initial permeability and how permeability changes after interaction with scCO2. Electron microscopy images with Back-Scattered Electron (BSE) and Energy Dispersive Spectroscopy (EDS) revealed dissolution and etching of calcite minerals and precipitation of calcium sulfide resulting from exposure to scCO2.
This study has tested about the deposition and dissolution of Chalk River Unidentified Deposit (CRUD), and investigated the remediation of CRUD. CRUD is an accumulated corrosion product on external fuel rod cladding surfaces and composed of either dissolved ions or solid particles such as Ni and Co. As nuclear power plants are getting older, decontaminating process has been concerned due to increased contamination. Corrosion product can be activated through nuclear reaction to form radionuclide such as Co-60. Because the CRUD is formed at cladding surface and radionuclides can be contained in CRUD, it can affect to human health and environment. Therefore CRUD needs to be removed safely during NPP's decontaminating process by efficient. In order to decontaminate the CRUD, first thing it has to be dissolved and removed. Deposition behavior of corrosion products has been well studied, but the study about behavior of CRUD dissolution is required in various environments. In this study, the CRUD synthesis, the dissolution method of CRUD, and the CRUD decontamination method was investigated. There are three major approaches the remediation of CRUD. First one is well synthesized the CRUD, second one is finding the optimum condition for CRUD dissolution, and last one is applying the Electro-Kinetic (EK) process to CRUD remediation. Synthesis of simulated CRUD for the decontamination experiments was successful. The combination of EK process with chemical decontamination of CRUD proved to be a more economical method for removing CRUD; it can be directly applied to the NPP decommissioning and decontamination process, and it can decrease the volume of secondary radioactive waste as well. (authors)
With the global community committed to significantly expanding nuclear energy capacity, the development of efficient used nuclear fuel (UNF) management strategies has become more critical than ever. These strategies are vital to fostering the widespread adoption of closed fuel cycles, which are essential for sustainable nuclear energy production and security. Achieving this ambitious goal necessitates a comprehensive understanding of radiation effects on next-generation technologies, as radiolysis can often limit the longevity and performance of these systems. This seminar will provide an overview of next-generation UNF reprocessing strategies, highlighting the latest advancements and innovative approaches in the field. Particular attention will be given to two key areas of recent research: 1. Radiation robustness and performance of advanced sulfur chloride-based chlorination technologies. We will explore the efficacy of sulfur chloride-based chlorination processes in the presence of surrogate cladding materials, specifically aluminum. These processes have shown promise in the dissolution, decontamination, and recovery of cladding materials for reuse. Detailed findings on how the composition and performance of these sulfur chloride solvents respond to radiation exposure will be discussed. 2. Impacts of metal ion complexation and direct dissolution conditions on monoamide-based reprocessing strategies. We will delve into the time-resolved and dose accumulation effects of irradiation on the direct dissolution of voloxidized uranium and rhenium using N,N-di-(2-ethylhexyl) butyramide (DEHBA) or N,N-di-(2-ethylhexyl)isobutyramide (DEHiBA) in pre-equilibrated n-dodecane solvent. The implications of these interactions on dissolution efficiency, radiolytic stability, and overall process performance will be examined. These studies aim to underscore the importance of understanding radiation effects in the development of next-generation UNF reprocessing technologies and the global transition towards more sustainable and efficient nuclear energy systems.
Abstract We performed numerical calculations of compaction in aggregates of spherical grains, using Lehner and Leroy's (2004, hereinafter LL) constitutive model of pressure solution at grain contacts. That model is founded on a local definition of the thermodynamic driving force and leads to a fully coupled formulation of elastic deformation, dissolution, and diffusive transport along the grain boundaries. The initial geometry of the aggregate was generated by random packing of spheres with a small standard deviation of the diameters. During the simulations, isostatic loading was applied. The elastic displacements at the contacts were calculated according to Digby's (1981) nonlinear contact force model, and deformation by dissolution was evaluated using the LL formulation. The aggregate strain and porosity were tracked as a function of time for fixed temperature, applied effective pressure, and grain size. We also monitored values of the average and standard deviation of total load at each contact, the coordination number for packing, and the statistics of the contact dimensions. Because the simulations explicitly exclude processes such as fracturing, plastic flow, and transport owing to surface curvature, they can be used to test the influence of relative changes in the kinetics of dissolution and diffusion processes caused by contact growth and packing rearrangements. We found that the simulated strain data could be empirically fitted by two successive power laws of the form, ε x ∝ t ξ , where ξ was equal to 1 at very early times, but dropped to as low as 0.3 at longer times. The apparent sensitivity of strain rate to stress found in the simulations was much lower than predicted from constitutive laws that assume a single dominant process driven by average macroscopic loads. Likewise, the apparent activation enthalpy obtained from the simulated data was intermediate between that assumed for dissolution and diffusion, and, further, tended to decrease with time. These results are similar to the experimental observations of Visser et al.'s (2012), who used an aggregate geometry and physical conditions closely resembling the present numerical simulations.
The use of Mg/Ca in marine carbonates as a paleothermometer has been challenged by observations that implicate salinity as a contributing influence on Mg incorporation into biotic calcite and that dissolution at the sea-floor alters the original Mg/Ca. Yet, these factors have not yet been incorporated into a single calibration model. In this paper, we introduce a new Bayesian calibration for Globigerinoides ruber Mg/Ca based on 186 globally distributed core top samples, which explicitly takes into account the effect of temperature, salinity, and dissolution on this proxy. Our reported temperature, salinity, and dissolution (here expressed as deep-water ΔCO 2- 3 ) sensitivities are (±2σ) 8.7±0.9%/°C, 3.9±1.2%/psu, and 3.3±1.3%/μmol.kg -1 below a critical threshold of 21 μmol/kg in good agreement with previous culturing and core-top studies. We then perform a sensitivity experiment on a published record from the western tropical Pacific to investigate the bias introduced by these secondary influences on the interpretation of past temperature variability. This experiment highlights the potential for misinterpretations of past oceanographic changes when the secondary influences of salinity and dissolution are not accounted for. Finally, multiproxy approaches could potentially help deconvolve the contributing influences but this awaits better characterization of the spatio-temporal relationship between salinity and δ 18 O sw over millennial and orbital timescales.