Outline. Wir schaffen Wissen heute für morgen. Validation of uptake processes of radionuclides such as Cs on clay minerals by EXAFS

Transcription

1 Nuclear Energy and Safety Research Department Laboratory for Waste Management Wir schaffen Wissen heute für morgen Paul Scherrer Institut Rainer Dähn Validation of uptake processes of radionuclides such as Cs on clay minerals by EXAFS Outline Introduction Validation of uptake mechanisms as function of aqueous chemistry by: Characterization of solid phases Conventional EXAFS and polarized EXAFS modeling Molecular modeling approach Status of Cs research at PSI Potential for quantitative modeling Potential role to support Fukushima challenges Conclusions

2 Geological waste repository concept Engineered and natural barrier materials to retard RN transport into the environment Backfill material bentonite consists 75% montmorillonite Molecular-level understanding of chemical processes (sorption, redox, dissolution) Modeling mobility of radionuclides Development and validation of sorption models e.g. high level waste repository NAGRA NTB -5 Uptake mechanisms desorption organic complexation surface precipitation outer-sphere inner-sphere After Manceau et al. () incorporation

3 Approach Investigations of predominant mineral phases: Wet chemistry Development of sorption models Input for XAFS measurements Approach Investigations of predominant mineral phases: Wet chemistry Development of sorption models Input for XAFS measurements Bulk XAFS (powder and P-EXAFS) Structural information (oxidation state, coordination numbers, bond distances, system disorder) Determination of uptake processes Characterization of uptake complexes Input for sorption models and molecular modeling

4 Approach Investigations of predominant mineral phases: Wet chemistry Development of sorption models Input for XAFS measurements Bulk XAFS (powder and P-EXAFS) Structural information (oxidation state, coordination numbers, bond distances, system disorder) Determination of uptake processes Characterization of uptake complexes Input for sorption models and molecular modeling Investigations of heterogeneous samples: Micro-XRF, Micro-XAS and Micro-XRD techniques Determination of reactive phases Identification of newly formed phases Speciation along diffusion profiles b c :1 clays e.g. vermiculite + montmorillonite T (Si) O (Al) cis vacant (cv) C O OH Si M1 M M vacant Montmorillonite: dioctahedral clay a trans sites (M1) and cis sites (M) b c predominant cis vacant up to mmol/kg Zn incorporated T (Si) Why Zn: Is an analogue for 3 Ni Allows to measure down to mmol/kg in the presence of 1 wt% Fe Zn contamination from smelters, mining, industrial and municipal waste trans vacant (tv) C/M

5 Polarized EXAFS Problem: Powder-EXAFS signal is averaged Filtration Self-supporting clay film P-EXAFS on highly oriented self-supporting clay films = "in-plane" Beam direction a) b) Zn-Al c* Zn-Si c* = 9 "out-of plane" Increased sensitivity to distinguish between IS/OS complexes, important for Cs uptake by clays T O T Zn Al, Vacancies Si Conventional vs. MM based EXAFS modelling Conventional MM based Bond Distance: Fitting parameter Fixed by MM Coordination: Fitting parameter Fixed by MM Disorder: Fitting parameter Fitted/Fixed by MM Multiple Scattering: Can be considered Full account Flowchart for conventional EXAFS modelling 1. Measured spectra. Shell model fit (Inter-atomic distances and coordination) 3. What are the structures consistent with Inter-atomic distances? Flowchart for MM based EXAFS modelling 1. MM modeling of potential structures. Calculate EXAFS spectra based on structures from MM modeling 3. Linear fit of calculated EXAFS spectra to measured ones Both approaches should give consistent results!

6 Reference plane Case study: molecular simulations of Zn uptake Rc Restraint potentials ab initio molecular dynamics and geometry optimization using CPK -package Meta-dynamics with classical Force Fields CLAYFF - Zn-Al and Zn-Si coordination number System composition K Mg Al 1 Si O (OH) 1 5H O Zn + System dimensions Å k 3 (k) Characterisation of the solid phase P-EXAFS: Zn incorporated in montmorillonite (MILOS) Measured data 1 k (Å -1 ) FT -1 Fourier transformed data FT(k 3 (k)) R + R(Å) k 3 (k) 1 Milos-incorporated 35 Fit of Milos-incorporated k (Å -1 ) Data were fitted using theoretical phase and amplitude function (FEFF) for Zn-O, Zn-Al and Zn-Si backscattering pairs Dähn et al., GCA 11

7 P-EXAFS data modelling P-EXAFS: Zn incorporated in montmorillonite (MILOS) k 3 (k) FT(k 3 (k)) Measured data 1 k (Å -1 ) Fourier transformed data Shell CN R ij Zn-O.().7(1) Zn-Al 3.3() 3.() Zn-Si.() 3.() R + R(Å) P-EXAFS: Zn incorporated in montmorillonite (MILOS) k 3 (k) FT(k 3 (k)) P-EXAFS data modelling Measured data 1 k (Å -1 ) Fourier transformed data CN Zn-Al ~ R + R(Å) Shell CN R ij Zn-O.().7(1) Zn-Al 3.3() 3.() Zn-Si.() 3.() Zn is NOT filling empty octahedral positions In that case: CN Zn-Al ~

8 P-EXAFS data modelling P-EXAFS: Zn incorporated in montmorillonite (MILOS) k 3 (k) FT(k 3 (k)) Measured data 1 k (Å -1 ) Fourier transformed data R + R(Å) Shell CN R ij Zn-O.().7(1) Zn-Al 3.3() 3.() Zn-Si.() 3.() Only one Si distance Zn is in trans positions Zn incorporated in montmorillonite montmorillonite cis vacant (C symmetrie) O OH Si M1 M Zn a b c Zn is incorporated into montmorillonite trans sites

9 Zn incorporated in montmorillonite cis vacant (cv), C Modelled Measured k 3 (k ) trans vacant (tv), C/M 1 k [Å -1 ] Zn incorporated in montmorillonite cis vacant (cv), C Modelled Measured k 3 (k ) trans vacant (tv), C/M 1 k [Å -1 ] Zn is incorporated to cv-montmorillonite into trans sites

10 Validation of uptake mechanisms Zn uptake by montmorillonite (. M NaClO ) log R d (L kg -1 ) 5 3 ph dependent C C C (a) Na-STx-1 init RD eq eq cation exchange ph V m surface complexation log Zn sorbed (mol kg -1 ) isotherm (ph 7.) SWy 1997 Milos 1 STx 1 model log Zn equilibrium concentration (M) SPNE SC/CE sorption model SPNE SC/CE = site protolysis non electrostatic surface complexation and cation exchange (Bradbury M. H. and Baeyens B., J. Contam. Hydrol. 1997)

11 Strong sites SPNE SC/CE = site protolysis non electrostatic surface complexation and cation exchange strong sites strong : < mmol/kg capacity Weak sites SPNE SC/CE = site protolysis non electrostatic surface complexation and cation exchange weak sites weak : < mmol/kg capacity

12 Planar sites SPNE SC/CE = site protolysis non electrostatic surface complexation and cation exchange planar sites planar sites play a role: low ph low ionic strength at high loadings Planar sites SPNE SC/CE = site protolysis non electrostatic surface complexation and cation exchange planar sites planar sites play a role: low ph low ionic strength at high loadings in this study: ph 7. M NaClO - mmol/kg

13 Strong/weak sites? Main question: Do the strong and weak sites have a crystallographic/physical meaning, or are they just fit parameters? Preparation of EXAFS samples weak strong Model is used to design the experiment and find favourable conditions for strong and weak sites

14 Zn uptake at medium absorber concentration 3 mmol/kg Zn sorbed to STX-1 k 3 (k) FT(k 3 (k)) 1 Zn incorporated Zn mmol/kg sorbed Zn 3 mmol/kg sorbed - 1 k (Å -1 ) R + R(Å) Shell CN R ij RSF peaks decrease with Zn-O ~5.5.5 increasing loading: higher disorder Zn-Al ~1. 3. complexes are sorbed Zn-Si ~ close to = 5.7 Zn uptake at medium absorber concentration 3 mmol/kg Zn sorbed to STX-1 +/ P-EXAFS is unable do determine the predominant uptake mode => molecular modeling

15 15 1 EXAFS data fitting at medium loadings (weak sites) Modeled Measured Texas Texas W S S S Zn Zn k 3 (k ) 5 Milos S S Zn Texas W S Zn Best Fit % (1) 1% (11) -5 1 k [Å -1 ] % (xxx) 11% Total Modeled structures can only explain uptake at low loadings! Preparation of EXAFS samples weak strong Model is used to design the experiment and find favourable conditions for strong and weak sites

16 Zn uptake at low absorber concentrations mmol/kg Zn sorbed to Milos and STX-1 k 3 (k) FT(k 3 (k)) Milos ( mmol/kg inc.) k (Å -1 ) k 3 (k) FT(k 3 (k)) k (Å -1 ) STx-1 (. mmol/kg inc.) R + R(Å) R + R(Å) Strong P-EXAFS dependency Zn uptake at low absorber concentrations O OH Si M1 M Zn mmol/kg Zn sorbed to Milos and STX-1 Shell CN R ij Zn Si1 Si1 Si3 Si3 Zn1 Si3 Si3 Zn5 Zn-O ~5.5.5 Zn-Al ~ Zn-Si 1 ~1. 3. Zn-Si ~. 3. b a c Zn Zn3 Si1 Si1 two Si distances

17 Zn uptake at low absorber concentrations O OH Si M1 M Zn mmol/kg Zn sorbed to Milos and STX-1 Shell CN R ij Zn Si1 Si1 Si3 Si3 Zn1 Si3 Si3 Zn5 Zn-O ~5.5.5 Zn-Al ~ Zn-Si 1 ~1. 3. Zn-Si ~. 3. b a c Zn Zn3 Si1 Si1 Reduced Coordination Numbers Zn uptake at low absorber concentrations O OH Si M1 M Zn mmol/kg Zn sorbed to Milos and STX-1 Shell CN R ij Zn Si1 Si1 Si3 Si3 Zn1 Si3 Si3 Zn5 Zn-O ~5.5.5 Zn-Al ~ Zn-Si 1 ~1. 3. Zn-Si ~. 3. a Si1 b c Zn Zn3 Si1 Dähn et al., GCA 11 Mixture of cis- and trans-like edge sites

18 Sorption from solution at (11) surface Zn-O coord. num..... n[o ] Tot n[o ] SiAl n[o ] w n[o ] Al Reaction coordinate, [Å] Structure search for sorption sites Edge surfaces cv- and tv-(1), (11), (13), (11) Substitutions cv-(1)-zn T, cv-(1)-zn C, tv-(1)-zn C, k 3, < (k)> k^3*< (k)> k, [angstrom^-1] 3 k i a ( k) i i k, [Å] exp ( k) a i min

19 EXAFS data fitting at low loadings (strong sites) k 3 (k ) Milos S S Zn Modeled Measured 1 k [Å -1 ] k 3 (k ) Texas S S Zn Modeled Measured 1 k [Å -1 ] cv-(1)-znt 5% cv-(11)-znt 7% cv-bulk -ZnT 5% Total 1% cv-(1)-znt % cv-(11)-znt 5% cv-bulk -ZnT 1% Total 1% Comparison EXAFS vs. MM S Texas S Zn Milos S S inc Milos Zn Zn R CN Al R R ZnSi 1 CNO ZnO ZnAl CN Si1 CN Si R ZnSi 5.1/..5/.1 1.1/.1 3./3. 3./3. 1.3/1. 1.3/. 3./3.9 3./3..1/..1/. 3./3../..5/.1.5/ /.5 3./3. 1./1. 1./. 3.1/ / /. 3./3.7./..7/.1 3.3/3. 3./3.5./. 3./3. model data Churakov S. V. and Dähn R. ES&T 1 Good agreement between molecular modeling and EXAFS

20 Summary low and medium Zn(II) loadings show different uptake behaviours. At low loadings, Zn(II) is structurally incorporated in octahedral edge positions. At medium loadings, the EXAFS spectra indicate that Zn(II) is sorbed as a mixture of inner-sphere complexes. This study confirms the assumption of the strong and weak site concept in the SPNE SC/CE sorption model. Molecular modeling can quantify uptake processes Status of Cs research at PSI Cs uptake by illite and argillaceous rocks Wet chemistry and modeling Bottom-up approach EXAFS in planning Cs diffusion in Opalinus Clay Micro-spectroscopic studies of Cs uptake by Opalinus Clay X-ray micro tomography microxrf and LA-ICPMS

21 Potential for quantitative modelling Generalised Cs model for illite (Bradbury & Baeyens, ) To model the sorption isotherm on illite: - 3 site types are required: Frayed Edge Sites (FES), Type II sites and Planar sites - Selectivity coefficients of Cs with respect to other cations e.g. Na and K are required ph 7,.1 NaClO Cs sorption model summary Site type Site capacity CEC illite = meq kg -1 FES Type II Planar.5 % of CEC % of CEC % of CEC Cation exchange reaction Na-FES + K K-FES + Na Na-FES + Cs Cs-FES + Na Na-II + K K-II + Na Na-II + Cs Cs-II + Na Log K c Boda Claystone Formation, Southern Hungary 5 Ma 1 Ma Deposition in a shallow lacustrine environment Opalinus clay, Northern Switzerland Deposition in shallow marine environment Boda Opalinus Depth ~1 m ~ 5 m

22 Blind predictions of the sorption isotherms Modelling constraints/assumptions: Generalised Cs and SPNE SC/CE sorption models for illite is also valid for muscovite (1 Å) and ilite/smectite mixed layers Site capacities scaled to the illite/muscovite and/or illite/smectite mixed layer contents Sorbing species: only free cations and hydrolysed species, no carbonate or other aqueous complexes Nagra/PSI 1/1 thermodynamic data base to calculate radionuclide speciation in the porewater No further adjustment of the model parameters i.e. no modification of surface complexation constants Illustration of the bottom-up approach U(VI) on Boda Claystone Experimental data

23 Illustration of the bottom-up approach U(VI) on Boda Claystone Experimental data Illite model Illustration of the bottom-up approach U(VI) on Boda Claystone Experimental data Illite model scaled to the clay amount

24 Illustration of the bottom-up approach U(VI) on Boda Claystone Experimental data Illite model scaled to the clay amount Porewater composition Blind predictions for Cs(I) on Boda clay Cs(I) Only Type II sites and planar sites active The frayed edge sites may be blocked with stable Cs (slow desorption kinetics) or absent in the case of muscovite

25 Parameters: 17 wt.-% illite [K] =. x 1-3 M [Na] =.1 M Blind predictions for Cs(I) on OPA clay Cs(I) Active sites:type II and FES, planar sites > 1-3 M [Cs] eq Same approach can be used for soils with a high clay content Water and cation diffusion in the clay interlayer Molecular dynamics simulations: M k d R dt k U ( R) R k Cs-Na-Montmorillonite T=3K, P=1bar R ( t t) R ( t) nd k k t Probability density D interlayer /D bulk arbitrary units Cs Na O w - - z [Å] water content (Interlayer distance) Froideval, et al., J. Nuc. Mat. 11 Kosakowski, et al., Clays Clay Miner.

26 Modeling of Cs diffusion in Opalinus Clay Experimental setup Filter 1 Clay sample Filter Reservoir cell C(,t) = f(t) Measurem. cell C(L,t) x 1 x L x Diffusion direction Competition aspects of ion transport in clays Diffusive Diffusive caesium flux [molm m - s -1 ] Experiment Sorption Isotherm model Isotherm Time [day] [ Cs C t [ Cs Cs D C x i e i [ Cs ] Sorbed Cs Planar K Planar Sorbed T II Cs K K Sorbed Cs [ ] FES K ] K K T II FES [ Cs [ K [ Cs [ K [ Cs [ K R Non-linear isotherm Sorbed ] K d ( C)[ Cs Illite sorption model Solution Solution Solution Solution Solution Solution i Solution ] Q ] ] Q ] ] Q ] ] Planar T II FES Cs sorbed [mol kg -1 ] E-5 1E-5 Mont Terri OPA Measured on crushed and compacted clay Cs in solution [mol m -3 ] 1 Nonlinear sorption isotherm can not adequately explain Cs diffusion in Opalinus Clay Multi-species reactive transport simulations based on mechanistic sorption model describe batch sorption data and diffusion in compacted clays consistently Competition with K is by far the most important factor for Cs transport Jakob. et al., Geochim. Cosmochim. Acta 9

27 Mineralogical heterogeneities in Opalinus Clay X-ray micro tomography measurement In Opalinus Clay θ m to mm scale X-ray tomography measurement In Opalinus Clay. mm Detector Clay Incidental X rays 3.5 mm Moving stage (θ,z) TOMCAT beamline PSI Resolution: 3x3x3 m 3 Calcite z y x 3.5 mm 9x9x11 mesh TOMCAT beamline PSI Resolution: 3x3x3 m 3 Diaz, et al., Project Rep. 11 Cs transport in Opalinus Clay (OPA) at m scale In-situ diffusion experiment Cs-adsorption edge measurements θ Measured 3D Cs tracer distribution Detector Incidental X rays 3.5 mm Reservoir [Cs] =.1M Moving stage TOMCAT beamline PSI Resolution: 3x3x3 m 3 t = 1h 9x9x11 mesh Resolution: 3x3x3 m 3 t = h, 1h, 1h, 3h Diaz, et al., Project Rep. 11

28 Cs transport in Opalinus Clay: m => mm Direct random walk simulation at m resolution 9x9x11 equidistant mesh with 3x3x3 m 3 resolution. mm Clay Calcite z y x 3.5 mm 9x9x11 mesh Churakov 11 Simulated versus measured Cs concentration A 3D Cs distribution in OPA A D Cs distribution Measured Measured max Simulated Simulated min 9x9x11 mesh

29 1D Cs profile: Linear and non-linear isotherms Non-linear isotherm: D e =.1-11 m.s -1 Linear isotherm: D e = x1-11 m s -1 K d = 1-3 m -3 kg -1 Cs uptake by OPA Comparison of D chemical images recorded by SRmicroXRF (3x3 m) and LA-ICPMS (1x1 m) Wang et al., Chimia 1

30 Cs uptake by OPA Wang et al., Anal. Chem. 11 Potential role to support Fukushima challenges Most studies concentrated on the investigation of pure mineral Phases and on potential argillaceous host rocks Opalinus Boda Opalinus Boda Mineralogy wt. % wt. % Porewater Illite & Illite-Smectite mixed layer Illite/Muscovite Chlorite Kaolinite Quartz Calcite Dolomite/Ankerite K-feldspar/Albite Siderite Pyrite Hematite Analcime 17 & ph I (M) 7..3 Dissolved constituents (M) Na K Mg Ca Cl CO 3 /HCO 3 SO Si M Fukushima soils are inherently more complex

31 Example: Zn smelter in France Topsoil: Depth: -1 cm= B horizon Zn: 11 ppm ph: TOC: g/kg CaCO 3 : 1 g/kg Clays: % ( g/kg) Silt: % (3 g/kg) Sand: 13% (1 g/kg) Subsoil : Depth: 3- cm= C horizon Zn: 7 ppm ph:.5 TOC: 1 g/kg CaCO 3 : 1 g/kg Clays: % ( g/kg) Silt: 5% (57 g/kg) Sand: 1% (17 g/kg) C B Vespa et al., ES&T 1 Bottom-up approach should be applicable in clay-rich soils Zn in soils: Bulk+P-EXAFS Vespa et al., EST 1 Bulk soil sample: Aa: 11 ppm Bb: 7 ppm P-EXAFS clay fraction: Aa: 5 ppm Bb: 1 ppm Extraction of the clay fraction and perform P-EXAFS

32 Conclusions Advanced spectroscopic and modeling methods, complemented with wet chemistry, are powerful tools to develop remediation strategies of contaminated soils Combination of bulk-, polarized, micro-xas/xrf/xrd is a key element to investigate complex heterogeneous systems Results can help to predict the fate of contaminants in Fukishima soils and quantify the uptake processes System understanding gives strong public credibility Geochemical implications: deep understanding of uptake processes at the water/clay interface at a molecular level Acknowledgements B. Baeyens, M.H. Bradbury, S.V. Churakov, D. Grolimund, L.v. Loon, M. Marques

33 Cordially invite you to participate in the ANXAS-1 at PSI Mineralogy & porewater composition Opalinus Boda Mineralogy wt. % wt. % Illite & Illite-Smectite mixed layer Illite/Muscovite Chlorite Kaolinite Quartz Calcite Dolomite/Ankerite K-feldspar/Albite Siderite Pyrite Hematite Analcime 17 & Porewater ph I (M) Opalinus 7..3 Dissolved constituents (M) Na K Mg Ca Cl CO 3 /HCO 3 SO Si Boda.1. M Opalinus Boda

34 Boda Claystone Formation, Southern Hungary 5 Ma 1 Ma Deposition in a shallow lacustrine environment monotonous, albititic, partly cross-bedded or laminated poorly bedded siltstone with conglomerate, sandstone, and dolomitic intercalations. Red and reddish brown, reflecting the dominantly oxidizing nature of the depositional environments Depth ~1 m Opalinus clay, Northern Switzerland Deposition in shallow marine environment dark grey claystone, five subunits can be distinguished, on the basis of slight variations in the carbonate/clay ratio ~ 5 m Bulk-XRD Identification of Zn species in the matrix in the Topsoil bulk-xrd AD: air dried EG: ethylene glycol Increases Sensitivity to smectite Presence of both di- and trioctahedral clays Vespa et al., EST 1

35 Micro-XRD Identification of Zn species in the matrix in the Topsoil -XRD reflection Characteristic for clays Presence of Fe-Al smectite Zn incorporated in Fe-containing dioctahedral clay Micro-XRF Zn Fe Pb Fe-hydroxide µm phyllomanganate Zn-phyllosilicate Zn immobilized in clay minerals Subsoil: -1 ppm Zn Zn-Al containing clay -> natural forming dioctahedral clay Topsoil: 1-3 ppm Zn Zn-Al-Fe containing clay -> anthropogenic forming dioctahedral clay => low mobility Zn is associated with Pb in phyllomanganate & Fe-(hydro)xides precipitates => mobile in reducing conditions