Stir Bar Sorptive Extraction modelling for hydrometallurgical extracting molecules, and LC-MS analytical method for environmental applications

Transcription

1 Stir Bar Sorptive Extraction modelling for hydrometallurgical extracting molecules, and LC-MS analytical method for environmental applications Hélène MALANDAIN, Xavier MACHURON-MANDARD nd workshop SBSE Paris, February 1 th, février 013 PAGE 1/9

2 Plan Context DEHPA extraction modelling with PDMS Thermodynamics approach Kinetics aspects Experimental validation of the model A LC-MS method to quantify DEHPA in aqueous solutions Experimental kinetics aspects Extraction modelling versus experimental data Conclusions nd workshop SBSE, February 1 th, 013, Paris PAGE /9

3 Industrial and analytical context Detection of extracting molecules at ppb concentration levels in aqueous samples. Modelling of the sampling conditions for general applications 19 février 013 CEA 10 AVRIL 01 PAGE 3 nd workshop SBSE, February 1 th, 013, Paris PAGE 3/9

4 Hydrometallurgical process for uranium production : From ore mining to spent fuel reprocessing plants Ore mining Stérile 1 Extraction du minerai minerai brut UO + Extraction de l uranium Yellow cake U 3 O 8 3 Raffinage UO Entreposage Usine de retraitement MOX U recyclé 6 35 U appauvri 5 Enrichissement 4 Conversion UF 4 Power plants, and spent fuel reprocessing facilities Combustible UO U métal UO enrichi UF 6 nd workshop SBSE, February 1 th, 013, Paris PAGE 4/9

5 Hydrometallurgical process for uranium production : From ore mining to spent fuel reprocessing plants Production of uranium or plutonium by Purex*-like hydrometallurgical processes Uranium conversion industrial process *Purex : PURification by EXtraction Yellow cake dissolution Liquid-liquid extraction [HNO 3 ] concentré "Yellow cake" + NO 3 + UO NO 3 + UO + [UO (NO 3 ) ], H O [UO (NO 3 ) ](TBP) Uranium nitrate [UO (NO 3 ) ](TBP) [UO (NO 3 ) ], H O Ammonium diuranate Calcination Uranium dioxide Hydrofluorination (UO (NO 3 ) ) + 6 NH 4 OH (NH 4 ) U O 7 + 4NH 4 NO H O (NH 4 ) U O 7 UO 3 + NH 3 + H O N UO 3 + H UO + H O UO 4 HF UF 4 H O + + Uranium hexafluoride UF 4 + F UF 6 CEA 9 JANIVER 013 PAGE 5/9

6 Hydrometallurgical process for uranium production : Two-phase equilibrium and extraction mechanism Two-phase process (Liquid-liquid process) Molecular extraction mechanism (complex formation and partition equilibria) Pulsed columns for counter-flow liquid-liquid extraction nd workshop SBSE, February 1 th, 013, Paris PAGE 6/9

7 Hydrometallurgical process for uranium production : Two-phase equilibrium and extraction mechanism Typical extracting molecules Extracted metal ion complex structure Typical extraction solvents β-dicetones TOPO (phosphine oxides) TBP (phosphoric acid ester) (phosphinic acids) (phosphonic acids) UO (NO 3 ), TBP (Carboxylic acids) DEHPA (phosphoric acids) nd workshop SBSE, February 1 th, 013, Paris PAGE 7/9

8 DEHPA : bis-(-ethylhexyl) phosphoric acid DEHPA is a very good liquid-liquid extraction agent for uranium in phosphoric acid media, it has potential environmental impact which requires analytical methods for environmental monitoring at trace concentration levels (µg/l), Important today : Development of efficient analytical procedures based on green analytical chemistry (solvent-free methods, direct and rapid analysis procedures), Due to its physical and chemical properties (hydrophobic and acidic properties, dimerization in organic solvents, complex equilibria with uranyl ions, ), DEHPA is a good standard for extraction process modelling. nd workshop SBSE, February 1 th, 013, Paris PAGE 8/9

9 Aim of this R&D project Development of green analytical methods for DEHPA detection at trace levels : On site and solvent-free sampling technique (no sample or solvent transportation), Optimization of sampling conditions, from modelling concepts, Development of a versatile model that could be applied for other extracting agents, Analytical method permiting rapid detection and quantification at trace levels (µg/l), Sampling by Stir Bar Sorptive Extraction, Application of basic analytical chemistry and liquid-liquid partition equilibria for extraction modelling, LC-MS analytical method development. nd workshop SBSE, February 1 th, 013, Paris PAGE 9/9

10 DEHPA extraction modelling with PDMS 19 février 013 nd workshop SBSE, February CEA 1 10 th, AVRIL 013, Paris 01 PAGE 10/9

11 DEHPA extraction modelling : thermodynamics approach Starting point of the model : polydimethylsiloxan (PDMS) is considered as a non-polar solvent, where chemical equilibria occur similarly compared with other organic liquid solvents Aqueous sample PDMS Chemical equilibria considered for DEHPA partition modelling (from known chemical properties) with with with with with (Two-phase partition equilibrium) (Acid-base equilibrium in aqueous phase) (Dimerization equilibrium in organic phase) nd workshop SBSE, February 1 th, 013, Paris PAGE 11/9

12 DEHPA extraction modelling : thermodynamics approach Aqueous phase PDMS phase The extraction process is thermodynamically ruled by three physical and chemical equilibria based on thermodynamic constants : K P HA = HA eq eq K p =K ow (hypothesis) = 10 6,07 K a = K H Di + eq HA = H A eq A HA eq eq eq = = ,47 4,50 CEA 9 JANIVER 013 PAGE 1/9

13 PAGE 13/9 0 0 n n n n n eq A H HAeq PDMS + = = P P Di P A pka ph P A pka ph P Di P Di P A pka ph P A pka ph PDMS K K K C K K K K K K C K K C n n β β γ β γ β β β γ β γ Three modelling parameters depending of the molecule extracted K a, K p, K Di Three operating parameters to optimize the extraction γ A- : activity coefficient (aqueous media) β : volume ratio for the two phases (V sample / V PDMS ) ph: ph of the aqueous solution (sample) DEHPA extraction modelling : thermodynamics approach Application of the law of mass action and the law of conservation of matter to calculate the extraction yield : Extraction yield Distribution coefficient

14 DEHPA extraction modelling : thermodynamics approach 14/31 Parameters used : C 0 = 1 nm, β = 83.3, γ A -=1 nd workshop SBSE, February 1 th, 013, Paris PAGE 14/9

15 DEHPA extraction modelling : thermodynamics approach Activity coefficient = 0,7 (ex. I=0. M, or ph=0,7) Activity coefficient = 1 (ex. I=0 M, or ph = 7) 15/31 Parameters used : C 0 = 1 nm, β = 83.3 nd workshop SBSE, February 1 th, 013, Paris PAGE 15/9

16 DEHPA extraction modelling : thermodynamics approach β = 83.8 (V sample = ml) β = 08.3 (V sample = 5 ml) β = (V sample = 10 ml) 16/31 16/4 Parameters used : C 0 = 1 nm, γ A- = 1 nd workshop SBSE, February CEA 91JANIVER th, 013, Paris 013 PAGE 16/9

17 DEHPA extraction modelling : thermodynamics approach Aqueous phase PDMS phase K a : Literature (acidity constant) K Di : Literature (dimerization constant) K p : K ow (hypothesis) Test of the K p value hypothesis to «tune»the model parameters 17/31 nd workshop SBSE, February 1 th, 013, Paris PAGE 17/9

18 DEHPA extraction modelling : kinetics aspects «La thermodynamique impose, la cinétique dispose» In the convection zone, the solution is homogeneous In the diffusion zone (or diffusion layer) (with thickness δ), pure diffusion of matter occurs according to Fick s law : r J r = D gradc Simplified empirical model similar to first order kinetics model : da dt t = k. A A = τ abs( t) A0 1 e A. Prietoa, O. Basauria, R. Rodilb, A. Usobiagaa, L.A. Fernándeza, N. Etxebarriaa, O. Zuloagaa, Stir-bar sorptive extraction: A view on method optimisation, novel applications, limitations and potential solutions, Journal of Chromatography A, 117, 64-66, (010)

19 DEHPA extraction modelling : kinetics aspects Amount of absorbed matter A abs( t) A0 1 e = τ t t 0,95 = -τ x ln(0.05) =.99 x τ t 0.95 is a pseudo-equilibrium time Time Operating parameters with an impact on the equilibrium time - Temperature (Viscosity, Brownian movement of molecules) - Sample volume (Amount of matter to absorb, transfer kinetics of matter toward the sample-twister interface) 19/31 - Rotation speed of the twister (Sample mixing and diffusion layer thickness) - Sample vial (Sample mixing efficiency) nd workshop SBSE, February 1 th, 013, Paris PAGE 19/9

20 Experimental validation of the model 19 février 013 CEA 10 AVRIL 01 PAGE PAGE 0 nd workshop SBSE, February 1 th, 013, Paris 0/9

21 A LC-MS method to quantify DEHPA in aqueous solutions Chromatography (elution) conditions: Mobile phase : Methanol / ammonium acetate 50 mm (70/30 v/v); Flow rate: 1ml/min -1 Column : Poursuit C18 (50x4,6 mm d p :5µm) Retention time for DEHPA : 34 min. MS detection by APCI (atmospheric pressium chemical ionisation) : Ionization mode : negative mode DEHPA detected at : m/z = 31u. Limit of detection : 3,0 µg/l Limit of quantitation : 7,0 µg/l nd workshop SBSE, February 1 th, 013, Paris PAGE 1/9

22 DEHPA extraction modelling : experimental kinetics aspects Tributylphosphate (TBP) Sample volume : ml Bath temperature : 5 C Rotation speed of the twister : 700 rpm DEHPA Magnetic stirrer Absorption n absorbé / n initial (%) yield (%) Temps Time (min) Absorption % absorbé yield (%) DEHPA Concentration 8, M Sample ph,5 V sample Temps Time (min) ml V PDMS 4 µl β 83,3 Temperature 5 C Rotation speed Pseudo-equilibrium time (t 0.95 ) estimate : 60 (±6) min Use of an effective time of 70 min to carry out further model validation experiments 700 rpm

23 DEHPA extraction modelling : experimental kinetics aspects Influence of operating conditions evaluated with TBP due to faster analytical procedures if compared with those for DEHPA Sample volume : ml Magnetic stirrer Bath temperature : 5 C Rotation speed of the twister : 700 rpm Temperature ( C) t 0.95 (min) Absorption yield (%) n absorbé / n initial (%) Sample composition Water Buffered water Sample volume (ml) 10 t 0.95 (min) 4 46 t 0.95 (min) 4 94 Equilibrium time ratio 1 Equilibrium time ratio 0 Rotation speed of (rpm) t 0.95 (min) Equilibrium time ratio Temps Time (min) nd workshop SBSE, February 1 th, 013, Paris PAGE 3/9

24 DEHPA extraction modelling versus experimental data DEHPA Concentration Equilibrium time 8, M 70 min Absorption yield (%) K P observed =10,3 K P =K O/W =10 6,07 K P =K O/W =10 6,07 V sample ml V PDMS 4 µl β 83,3 Temperature 5 C Rotation speed 700 rpm 4/31 nd workshop SBSE, February 1 th, 013, Paris PAGE 4/9

25 DEHPA extraction modelling versus experimental data Why K o/w cannot be used in our model? Publication by Paul Ruelle : P. Ruelle, The n-octanol and n-hexane/water partition coefficient of environmentally relevant chemicals predicted from the mobile order and disorder (MOD) thermodynamics, Chemosphere, 40, , (000) PDMS is not octanol! For polar molecules (ex. R-OH or molecules with proton exchange properties), K O/W and K Hexan/W are different Molecule Log K H/W Log K O/W 4-ethylphenol Heptanol ,4 -Cl - bisphenol F But the K H/W value for DEHPA is not available, at the moment! nd workshop SBSE, February 1 th, 013, Paris PAGE 5/9

26 DEHPA back-extraction modelling DEHPA back-extraction yield : 50 to 80% after 4 to 5 replicate leaching. The method probably suffers a kinetics-limited reverse process. Hypothesis : is it due to slow diffusion of DEHPA in PDMS? nd workshop SBSE, February 1 th, 013, Paris PAGE 6/9

27 Conclusions 19 février 013 nd workshop SBSE, February CEA 1 10 th, AVRIL 013, Paris 01 PAGE 7/9

28 Conclusions Modelling of DEHPA extraction by PDMS (twister) achieved, from solution chemistry and liquid-liquid extraction basic concepts, But the impact of other operating parameters must be evaluated, Variation of β Variation of C 0 Application of such model to other similar molecules should be possible (to be continued), But data for molecules in PDMS (as a solvent) are lacking (K PDMS/W K O/W ), And the back-extraction seems to be limited (kinetics limitation? hypothesis to confirm). nd workshop SBSE, February 1 th, 013, Paris PAGE 8/9

29 Thank you for your attention 19 février 013 nd workshop SBSE, February CEA 1 10 th, AVRIL 013, Paris 01 PAGE 9/9