Molecular Dynamics Study of the Aggregation Behavior of N,N,N',N'-Tetraoctyl Diglycolamide.

Abstract

Liquid–liquid extraction is a commonly used technique to separate metals and is a process that has particular relevance to the nuclear industry. There has been a drive to use environmentally friendly ligands composed only of carbon, hydrogen, nitrogen, and oxygen. One example is the i-SANEX process that has been developed to separate minor actinides from spent nuclear fuel. The underlying science of such processes, is, however, both complex and intriguing. Recent research indicates that the liquid phases involved are frequently structured fluids with a hierarchical organization of aggregates. Effective flow-sheet modeling of such processes is likely to benefit from the knowledge of the fundamental properties of these phases. As a stepping stone toward this, we have performed molecular dynamics simulations on a metal free i-SANEX system composed of the ligand N,N,N′,N′-tetraoctyl diglycolamide (TODGA), diluent hydrogenated tetrapropylene (TPH), and polar species water and nitric acid. We have also studied the effects of adding n-octanol and swapping TPH for n-dodecane. It would seem sensible to understand this simpler system before introducing metal complexes. Such an understanding would ideally arise from studying the system’s properties over a wide range of compositions. The large number of components, however, precludes a comprehensive scan of compositions, so we have chosen to study a fixed concentration of TODGA while varying the concentrations of water and nitric acid over a substantial range. Reverse aggregates are observed, with polar species in the interior in contact with the polar portions of the TODGA molecules and the organic diluent on the exterior in contact with the TODGA alkyl chains. These aggregates are irregular in shape and grow in size as the amount of water and nitric acid increases. At a sufficiently high polar content, a single extended cluster forms corresponding to the third phase formation. No well-defined bonding motifs were observed between the polar species and TODGA. The cluster size distribution fits an isodesmic model, where the Gibbs energy change of adding a TODGA molecule to a cluster ranges between 4.5 and 7.0 kJ mol–1, depending on the system composition. The addition of n-octanol was found to reduce the degree of aggregation, with n-octanol acting as a co-surfactant. Exchanging the diluent TPH for n-dodecane also decreased the aggregation. We present evidence that this is due to the greater penetration of n-dodecane into the reverse aggregates. It is known, however, that the propensity for the third phase formation is greater with n-dodecane as the diluent than is the case with TPH, but we argue that these two results are not contradictory. This research casts light on the driving forces for aggregation, informs process engineers as to what species are present, and indicates that flow-sheet liquid–liquid extraction modeling might benefit by incorporating an isodesmic aggregation approach.