Nitric acid extraction into TODGA

Transcription

1 Available online at Procedia Chemistry 7 (2012 ) ATALANTE 2012 International Conference on Nuclear Chemistry for Sustainable Fuel Cycles Nitric acid extraction into TODGA K. Bell a,*, A. Geist b, F. McLachlan a, G. Modolo e, R. Taylor a, A. Wilden e a National Nuclear Laboratory, Central Laboratory, Sellafield, Seascale, Cumbria, CA20 1PG, UK b Karlsruhe Institute of Technology, INE, Karlsruhe, Germany e Forschungszentrum Jülich, IEK-6, Jülich, Germany Abstract Advanced solvent extraction processes, namely DIAMEX, SANEX or GANEX, for the separation of the minor actinides (americium and curium) are under development within Europe. The tridentate diglycolamide ligand, TODGA, shows many interesting properties and is under investigation in conjunction with a variety of other extractants for the DIAMEX and SANEX processes as well as the GANEX process. In order to successfully demonstrate these processes, understanding the acid extraction into the organic phases is critical to process flowsheet design and modelling. Here nitric acid extractions into TODGA have been measured and models produced using an equilibrium based approach accounting for nitric acid activities in the aqueous phase The Authors. Published by by Elsevier B.V. B.V. Selection and /or peer-review under responsibility of the Chairman of the ATALANTA Selection and/or 2012 peer-review Program Committee under responsibility Open access of under the Chairman CC BY-NC-ND of the license. ATALANTE 2012 Program Keywords: nitric acid; diglycolamide; activity; GANEX; SANEX; DIAMEX 1. Introduction The European Union has a vision for the operation of Fast Neutron Reactors with closed fuel cycles by 2050 to provide a sustainable nuclear energy source and reduce the burden of radioactive waste for geological disposal.[1] One of the key requirements for this sustainable nuclear fuel cycle is the ability to fully recover the actinide elements from spent fuel and refabricate them into new fuel, for burning in Generation IV (fast) * Corresponding author: katie.j.bell@nnl.co.uk The Authors. Published by Elsevier B.V. Selection and /or peer-review under responsibility of the Chairman of the ATALANTA 2012 Program Committee Open access under CC BY-NC-ND license. doi: /j.proche

2 K. Bell et al. / Procedia Chemistry 7 ( 2012 ) reactors, or into targets for transmutation in dedicated facilities. Both heterogeneous and homogeneous recycling routes are under investigation.[2] To achieve these goals, processes to recover uranium (U), plutonium (Pu), neptunium (Np) and other minor actinides, americium (Am) and curium (Cm), require development. Whilst separation of U and Pu by the PUREX process is well established,[] there remains a challenge to recover the remaining actinides particularly from the chemically similar lanthanides.[4] Separation of the trivalent actinides by solvent extraction requires new extractant molecules that have a high specificity for Am and Cm. In the heterogeneous route, a number of processes have been proposed that separate the minor actinides (MAs) from the highly active raffinate (HAR) stream after the PUREX process. In Europe these processes have been termed DIAMEX and SANEX.[4-6] An alternative concept under investigation within Europe is the Grouped ActiNide EXtraction process (GANEX) in which all of the actinides are routed together.[7-9] Prospective extractants are diglycolamide (DGA) ligands which have proved very successful in DIAMEX and SANEX process development[10-12] and show many ideal properties for grouped actinide extractions.[1, 14] The tridentate ligand TODGA (N,N,N,N- tetraoctyl diglycolamide) is the leading example of this class of ligands.[15-19] Indeed, a number of DIAMEX-style flowsheet trials have been performed with 0.2 M TODGA / 0.5 M TBP (tri-n-butyl-phosphate) in an alkane diluent, successfully demonstrating the complete co-extraction of trivalent lanthanides and minor actinides.[10-12] An improved formulation (0.2 M TODGA / 5 % octanol) has been proposed and tested for its use in an innovative(i)-sanex process aimed at separating the trivalent actinides from the lanthanides.[20] M TODGA has also been used as a phase modifier with the CyMe 4 BTBP (6,6 -bis(5,5,8,8-tetramethyl-5,6,7,8-tetrahydrobenzo[1,2,4]-triazin--yl)-[2,2 ]-bipyridine) ligand to selectively extract trivalent actinides from a simulated HAR in a 1-cycle SANEX process.[21] Recently, a system containing TODGA and the malonamide, DMDOHEMA (N,N -dimethyl-n,n -dioctylhexylethoxymalonamide) has been shown to have favourable properties for applications in a GANEX process.[1] In conjunction with experimental programs to explore the properties of extractants the development of models to simulate the separation processes is an essential part of process development. The extraction of acid into the solvent phase and the resulting changes in acidity of the aqueous and organic phases has a profound impact on the extraction process. The carrying over of acid through a solvent extraction process can determine the success or failure of the system and as such it is essential to have a good understanding of the acid extraction. Using a simple concentration based model, Mowafy[22] found that the 1:1 HNO :TODGA complex was the extracted species at 1 M HNO with a conditional equilibrium constant K H = 0.27 ±0.04 in benzene diluent. Guoxin[2] reported HNO extraction into a related DGA molecule N,N -dimethyl-n,n -dicotyl-- oxadiglycolamide (DMDODGA). Using the same approach as Mowafy, they also find 1:1 HNO :TODGA complexation from extractions at 2 M HNO with a conditional equilibrium constant K H = 0.8 ±0.04 in chloroform diluent. Ansari[15] confirms 1:1 complexation with TODGA in n-dodecane (aqueous phase of 1 M HNO ) with a conditional equilibrium constant K H = 4.1 ±0.4. Sasaki finds HNO extraction varies with diluent with the apparent solvation number around 1 for dodecane[24, 25]. Modolo[10], however, shows that the apparent solvation number varies with TODGA concentration, from (HNO ) 2 TODGA at 6 M HNO and 0.1 M TODGA to (HNO ) 1.5 TODGA with 0.6 M TODGA in TPH (tetrahydrogenated propylene) and dodecane diluents. In extractions in TPH with 5 % octanol, Geist[20] also proposes (HNO ) 2 TODGA species must be present above M HNO since the extracted nitric acid exceeds the TODGA concentration, although he notes the 1:1 complex is the dominant species.

3 154 K. Bell et al. / Procedia Chemistry 7 ( 2012 ) The use of activity coefficients in the modeling of solvent extraction systems has been reported by a number of authors. HNO extraction into TBP has been studied extensively[26-6] and several models using activity coefficients have been reported.[27, 7] Chaiko and Vandegrift calculated molal activity coefficients using a method by Bromley[8] before converting them into molar units[7] while Davis recalculated in molar units molal activity coefficients produced from freezing point data.[27] Both then looked at the extraction of a number of complexes into the organic solution. Models for extraction of HNO into DMDOHEMA[9] and octanol[40] have been reported, again using activity coefficients but calculated using Specific Interaction Theory.[41]. Experiments have been carried out to measure nitric acid extraction into TODGA and the results used to produce models of this extraction based on an equilibrium approach, accounting for nitric acid activities in the aqueous phase. Future work will extend the modelling to mixed extractant systems of relevance to prospective advanced solvent extraction processes (e.g. TODGA plus DMDOHEMA). 2. Nitric acid extractions Solutions of nitric acid of varying concentration have been contacted with solutions of TODGA, in odourless kerosene (OK) and the resulting acid concentrations in the organic and aqueous phase measured. The concentration of the extractant has also been varied to see the effect this has on the extraction.. Model development and results Equilibrium equations should correctly be expressed in terms of the activities of the species involved but as activities are not always known they are commonly expressed in terms of concentrations. This is most valid when the solute is at a trace concentration compared to the extractant concentration, a situation that does not apply with the extraction of nitric acid. The approach taken here is to try to follow a theoretical activity based approach as far as possible. The nitric acid activities are calculated using correlations to literature data as the range of concentrations of interest are out with those recommended for the use of other methods of activity calculation, such as the SIT (Specific Ion Interaction Theory) method.[41].1. Modelling procedure The complexes formed in solution with the extractant (A) are considered in order to adequately model the experimental values,[42, 4] with the equilibria shown in Equations 1- taken as the starting point.[7] (Note that in this paper the molal activity of a species X is denoted by {X} and the molar concentration by [X].) 11 H NO A HNO. A H NO 12 2A HNO. A H 2NO A( HNO ). A 2 11 where where where HNO. A H NO HNO. A A free 2 H NO 2 ( HNO A free 2 2 H NO ). A 2 A free (1) (2) ()

4 K. Bell et al. / Procedia Chemistry 7 ( 2012 ) Assuming activity coefficients in the organic phase are unity the activities of organic phase species can be replaced by concentrations. Also, as no other species are present in the aqueous phase the assumption is made that the activities of the nitrate and proton are equal. As the distribution value, D HNO, for acid extraction is defined as in (4), substitution of the equilibrium equations gives (5). Accounting for the mass balance of the extractant (A Total ) gives a solution for the free ligand concentration, A free. Values for the equilibrium constants can then be established by comparison of calculated distribution data with experimental data using Microsoft Excel Solver to minimise the differences.[42] D D HNO HNO [ HNO, [ HNO 11 org, aq ] [ HNO. A] [ HNO. A ] [ HNO ] 2[( HNO ) ] 2, aq H [ A ] H [ A ] 2 H free 12 [ HNO free, aq ] 21 4 [ A 2. A] free ] (4) (5).2. HNO extraction with TODGA The above method was used to fit values of the equilibrium constants to NNL data on the acid extraction into solutions of TODGA in odourless kerosene diluent and the calculated values for organic acid compared with experimental values. The equilibrium constants calculated are shown in Table 1 (Case 1). As can be seen from Figure 1 (Case 1) the calculated values for the extracted acid are much lower at the highest acidities than the experimental values. Studying the experimental data reveals that the concentration of extracted acid at the highest acidities is more than twice the concentration of TODGA present in solution. This indicates that complexes containing more than two acid molecules to one TODGA must be formed at high acidity. Table 1. Values for the equilibrium constants for HNO extraction into TODGA/OK Case ß 11 ß 12 ß 21 ß 1 ß x x x10-7 The involvement of the equilibrium shown in Equation (6) was therefore proposed, firstly where x =. The :1 complex was added to the system by modifying Equation (5) to include the contribution from this complex to the organic phase acidity and the equilibrium constants recalculated. The values found are shown in Table 1 (Case 2). Including this complex significantly improved the fit to the experimental values and allowed the concentration of extracted acid at the highest acidities to be more closely modelled (results not shown).

5 156 K. Bell et al. / Procedia Chemistry 7 ( 2012 ) [HNO] org, eq, mol/l M experimental 0.2 M Experimental- NNL 0.2 M Experimental- KIT-INE 0.2 M Experimental-FZJ 0. M experimental calculated-1 calculated [HNO] aq, eq, mol/l Figure 1. HNO extraction into TODGA in OK/TPH as a function of equilibrium aqueous HNO concentration, experimental data compared with two versions of the equilibrium model (lines) corresponding to (1) calculated using equilibrium reactions 1- and () calculated using reactions (1,) and (6) with x= and x=4. To explore further complexes the 4:1 complex (i.e. x = 4 in Equation (6)) was then added to the system and while the resulting value for the equilibrium constant for this complex was low it did improve the fit over Case 2. The value of the equilibrium constant ß 12 produced in this case was also very low (1x10-8 ) and so it was decided to remove this complex from the system to assess its impact. Thus the equilibrium constants were recalculated for the system using equations (1), () and (6) with x = and x = 4 (Table 1 Case ). The results are shown in Figure 1 (Case ). Indeed, simplistic plots of log D HNO against log[todga] give gradients close to 1 (1.09, 1.02 and 1.00 for 1, and 6 M HNO respectively) indicating that only one TODGA is present in the extracted complexes. This is in agreement with the literature where 1:1 and 2:1 complexes have been reported but no evidence has been found for 1:2 complexes. [10, 15, 20, 22-25] Most published modeling approaches and experimental data for systems of this kind restrict the acidity of the solutions to a lower range than those considered here and as can be seen in Figure 1 the extracted acid can be reasonably well modelled at lower acidities without the need for complexes containing more than two HNO molecules. Figure 1 also shows the experimental data collected at KIT-INE and FZJ, both with 0.2 M TODGA in Exxsol D80. The values obtained by KIT-INE are in very good agreement with the NNL data and while the FZJ data show higher values of extracted acid they are in reasonable agreement taking into account experimental errors. xh xno x1 A( HNO ) x A ( HNO ) x. A where x1 (6) x x H NO A free

6 K. Bell et al. / Procedia Chemistry 7 ( 2012 ) TODGA is known to self-aggregate in solution[15, 44-46] and the strength of these aggregates will have an impact on the ability of the extractant to form complexes with other species. The effect of the aggregates may be to reduce the available free extractant concentration. Alternatively, the aggregates may themselves extract species from the aqueous phase forming large complexes which may average out as those complexes which have been modelled here. Further studies of the solvent systems are necessary to ascertain how aggregation effects observed spectroscopically affect HNO extraction equilibria. 4. Conclusions Using activity based calculations (at least for the aqueous phase) thermodynamically based models for HNO extraction into TODGA odourless kerosene solvents can be developed and information obtained on the likely complexes formed. While there are a number of issues that ideally would be addressed to improve the modelling, it is clear that good estimates can be made of HNO extraction in this system simply by accounting for nitric acid activity in the aqueous phase. Acknowledgements Financial support for this research was provided by the European Commission (project ACSEPT Contract No FP7-CP ). NNL also thank Sellafield Ltd. and the UK Nuclear Decommissioning Authority for support. References [1] A contribution to the EU Low Carbon Energy Policy: Demonstration Programme for Fast Neutron Reactors Concept Paper, in: T.E.S.N.I.I. (ESNII) (Ed.), [2] D. Warin, Future nuclear fuel cycles: Prospect and challenges for actinide recycling, IOP Conference Series: Materials Science and Engineering, 9 (2010) [] I.S. Denniss, A.P. Jeapes, Reprocessing irradiated fuel, in: P.D. Wilson (Ed.) The Nuclear Fuel Cycle, Oxford Science Publications, Oxford, [4] C. Madic, M. Lecomte, P. Baron, B. Boullis, Separation of long-lived radionuclides from high active nuclear waste, Comptes Rendus Physique, (2002) [5] C. Madic, Testard, F., Hudson, M. J., Liljenzin, J.-O., Christiansen, B., Ferrando, M., Facchini, A., Geist A., Modolo, G., Gonzalez- Espartero, A., De Mendoza, J., PARTNEW New solvent extraction processes for minor actinides, in: Final Report CEA-R-6066, [6] D. Serrano-Purroy, P. Baron, B. Christiansen, R. Malmbeck, C. Sorel, J.P. Glatz, Recovery of minor actinides from HLLW using the DIAMEX process, Radiochimica Acta, 9 (2005) [7] S. Bourg, C. Hill, C. Caravaca, C. Rhodes, C. Ekberg, R. Taylor, A. Geist, G. Modolo, L. Cassayre, R. Malmbeck, M. Harrison, G. de Angelis, A. Espartero, S. Bouvet, N. Ouvrier, ACSEPT--Partitioning technologies and actinide science: Towards pilot facilities in Europe, Nuclear Engineering and Design, 241 (2011) [8] M. Miguirditchian, L. Chareyre, X. Hérès, C. Hill, P. Baron, M. Masson, GANEX: Adaptation of the DIAMEX-SANEX Process for the Group Actinide Separation, in: Proc. GLOBAL 2007, Advanced Nuclear Fuel Cycles and Systems, American Nuclear Society, Boise, Idaho, 2007, pp [9] M. Miguirditchian, Roussel; H., L. Chareyre, P. Baron, HA demonstration in the Atalante facility of the GANEX 2nd cycle for the grouped TRU extraction, in: GLOBAL 2009, American Nuclear Society, Paris, France, 2009, pp [10] G. Modolo, Asp, H., Schreinemachers, C., Vijgen, H., Development of a TODGA based Process for Partitioning of Actinides from a PUREX Raffinate Part I: Batch Extraction Optimization Studies and Stability Tests, Solvent Extraction and Ion Exchange, 25 (2007)

7 158 K. Bell et al. / Procedia Chemistry 7 ( 2012 ) [11] G. Modolo, Asp, H., Vijgen, H., Malmbeck, R., Magnusson, D., Sorel, C., Demonstration of a TODGA-Based Continuous Counter- Current Extraction Process for the Partitioning of Actinides from a Simulated PUREX Raffinate, Part II: Centrifugal Contactor Runs, Solvent Extraction and Ion Exchange, 26 (2008) [12] D. Magnusson, B. Christiansen, J.-P. Glatz, R. Malmbeck, G. Modolo, D. Serrano-Purroy, C. Sorel, Demonstration of a TODGA based Extraction Process for the Partitioning of Minor Actinides from a PUREX Raffinate -- Part III: Centrifugal Contactor Run using Genuine Fuel Solution, Solvent Extraction and Ion Exchange, 27 (2009) [1] J. Brown, F. McLachlan, M.J. Sarsfield, R.J. Taylor, G. Modolo, A. Wilden, Plutonium loading of prospective grouped actinide extraction (GANEX) solvent systems based on diglycolamide extractants, Solvent Extraction and Ion Exchange, 0 (2012) [14] J. Brown, M.J. Carrott, O.D. Fox, C.J. Maher, C. Mason, F. McLachlan, M.J. Sarsfield, R.J. Taylor, D.A. Woodhead, Screening of TODGA/TBP/OK solvent mixtures for the grouped extraction of actinides, IOP Conference Series: Materials Science and Engineering, 9 (2010) [15] S.A. Ansari, P.N. Pathak, V.K. Manchanda, M. Husain, A.K. Prasad, V.S. Parmar, N,N,N,N-Tetraoctyl Diglycolamide (TODGA): A Promising Extractant for Actinide-Partitioning from High-Level Waste (HLW), Solvent Extraction and Ion Exchange, 2 (2005) [16] Y. Sugo, Sasaki, Y., Tachimori, S., Studies on hydrolysis and radiolysis of N,N,N,N-tetraoctyl--oxapentane-1,5-diamide, Radiochimica Acta, 90 (2002) [17] Z.-X. Zhu, Sasaki, Y., Suzuki, H., Suzuki, S., Kimura, T., Cumulative study on solvent extraction of elements by N,N,N',N'-tetraoctyl-- oxapentanediamide (TODGA) from nitric acid into n-dodecane, Analytica Chimica Acta, 527 (2004) [18] Y. Sasaki, Sugo, Y., Suzuki, S., Tachimori, S., The Novel Extractants, Diglycolamides, For The Extraction Of Lanthanides And Actinides In HNO n-dodecane System, Solvent Extraction and Ion Exchange, 19 (2001) [19] Y. Sasaki, Tachimori, S., Extraction of actinides(iii), (IV), (V), (VI), and lanthanides(iii) by structurally tailored diamides, Solvent Extraction and Ion Exchange, 20 (2002) [20] A. Geist, Modolo, G., TODGA process development: an improved solvent formulation, in: GLOBAL 2009, ANS, Paris, France, 2009, pp [21] A. Wilden, C. Schreinemachers, M. Sypula, G. Modolo, Direct Selective Extraction of Actinides (III) from PUREX Raffinate using a Mixture of CyMe 4 BTBP and TODGA as 1-cycle SANEX Solvent, Solvent Extraction and Ion Exchange, 29 (2011) [22] E.A. Mowafy, Aly, H. F., Synthesis of some N,N,N,N-Tetraalkyl--Oxa-Pentane-1,5-Diamide and their Applications in Solvent Extraction, Solvent Extraction and Ion Exchange, 25 (2007) [2] S. Guoxin, Min, L., Yu, C., Meilong, Y. Shaohong, Y., Synthesis of N,N-Dimethyl-N,N-Dioctyl--Oxadiglycolamide and its Extraction Properties for Lanthanides, Solvent Extraction and Ion Exchange, 28 (2010) [24] Y. Sasaki, Rapold, P., Arisaka, M., Hirata, M., Kimura, T., Hill, C., Cote, G., An Additional Insight into the Correlation between the Distribution Ratios and the Aqueous Acidity of the TODGA System, Solvent Extraction and Ion Exchange, 25 (2007) [25] Y. Sasaki, Morita, Y., Kitatsuji, Y., Kimura, T., Extraction Behavior of Actinides and Metal Ions by the Promising Extractant, N,N,N,N-Tetraoctyl-,6-dioxaoctanediamide (DOODA), Solvent Extraction and Ion Exchange, 28 (2010) [26] J.W. Codding, Haas, W. O. Jr., Heumann, F. K., Tributylphosphate-Hydrocarbon Systems. Organizing Equilibrium Data, Industrial and Engineering Chemistry, 50 (1958) [27] W.J. Davis, Thermodynamics of Extraction of Nitric Acid by Tri-n-butyl Phosphate-Hydrocarbon Diluent Solutions I. Distribution Studies with TBP in AMSCO at Intermediate and Low Acidities, Nuclear Science and Engineering, 14 (1962) [28] P. Leroy, Etude du Solvant Phosphate Tributylique 0 Percent-Dodecane (Study of the Tributyl Phosphate--0 Percent Dodecane Solvent), in: Other Information: UNCL. Orig. Receipt Date: 1-DEC-67, 1967, pp [29] P.E. Burns, Hanson, C., Distribution of Nitric Acid between Tri-n-butyl Phosphate and Water, Journal of Applied Chemistry, 14 (1964) [0] T.V. Healy, Brown, P. E., Extraction of Alkali Salts into Organic Solvents, in, AERE Harwell, [1] C.J. Jones, Scargill, D., Watermand, M. J., Commissioning Trials on an Automated Distribution Data Collection Facility Bases on an AKUFVE Mixer Centrifuge Unit, in, AERE Harwell, [2] P. Mistry, Baker, D. D., Partridge, B. A., Experimental Measurements on the Solubility of Water in TBP/OK Solutions Containing HNO and UO 2 (NO ) 2 and on the Partition of CsNO, in, 1989.

8 K. Bell et al. / Procedia Chemistry 7 ( 2012 ) [] D.R. Olander, Donadieu, L., Benedict, M., The distribution of nitric acid between water and tributyl phosphate - hexane solvents, AIChE Journal, 7 (1961) [4] G. Petrich, Kolarik, Z., The 1981 Purex distribution data index, in, Kernforschungszentrum Karlsruhe, [5] Z. Kolarik, Tachimori, S., Nakashima, T., Extraction of uranium(iv,vi), plutonium(iii) and nitric acid by 0% tributyl phosphate (TBP) in dodecane, Solvent Extraction and Ion Exchange, 2 (1984) [6] R. Prasanna, Suresh, A., Srinivasan, T. G., Vasudeva Rao, P. R., Extraction of nitric acid by some trialkyl phosphates, Journal of Radioanalytical and Nuclear Chemistry, 222 (1997) [7] D.J. Chaiko, G.F. Vandegrift, A thermodynamic model of nitric acid extraction by tri-n-butyl phosphate, Nuclear Technology, 82 (1988) [8] L.A. Bromley, Thermodynamic properties of strong electrolytes in aqueous solutions, AIChE Journal, 19 (197) [9] A. Geist, Equilibrium Model for the Extraction of Am(III), Eu(III), and HNO into DMDOHEMA in TPH, in: Atalante 2008 (Nuclear Fuel Cycles for a Sustainable Future), Montpellier, France, [40] A. Geist, Extraction of Nitric Acid into Alcohol: Kerosene Mixtures, Solvent Extraction and Ion Exchange, 28 (2010) [41] R. Guillaumont, T. Fanghänel, J. Fuger, I. Grenthe, V. Neck, D.A. Palmer, M.H. Rand, Update on the Chemical Thermodynamics of Uranium, Neptunium, Plutonium, Americium and Technetium, Elsevier Science, 200. [42] M.J. Sarsfield, H.E. Sims, R.J. Taylor, Modelling Np(IV) Solvent Extraction between 0% Tri-Butyl Phosphate and Nitric Acid in the Presence of Simple Hydroxamic Acids, Solvent Extraction and Ion Exchange, 29 (2011) [4] M.J. Sarsfield, H.E. Sims, R.J. Taylor, Extraction of Neptunium (IV) Ions into 0% Tri-Butyl Phosphate from Nitric Acid, Solvent Extraction and Ion Exchange, 27 (2009) [44] S. Nave, Modolo, G., Madic, C., Testard, F., Aggregation Properties of N,N,N,N-Tetraoctyl--oxapentanediamide (TODGA) in n- Dodecane, Solvent Extraction and Ion Exchange, 22 (2004) [45] T. Yaita, Herlinger, A. W., Thiyagarajan, P., Jensen, M. P., Influence of Extractant Aggregation on the Extraction of Trivalent f-element Cations by a Tetraalkyldiglycolamide, Solvent Extraction and Ion Exchange, 22 (2004) [46] P.N. Pathak, Ansari, S. A., Kumar, S., Tomar, B. S., Manchanda, V. K., Dynamic light scattering study on the aggregation behaviour of N,N,N',N'-tetraoctyl diglycolamide (TODGA) and its correlation with the extraction behaviour of metal ions, Journal of Colloid and Interface Science, 42 (2010)