Transcript Modelling metal speciation in natural waters

Spatial and Temporal Assessment of high Resolution Depth profiles Using novel
Sampling Technologies
Modelling metal speciation and
behaviour in sediments
Lille, March 18, 2005
VLIZ
Modelling metal speciation and
behaviour in sediments
3 different modelling approaches:
1. Trace metal speciation (multi2.
3.
ligand model)
Kinetics of metal
remobilisation in sediments
(DIFS model)
Diagenetic model
From DGT-data
DGT in sediments
• Separates species according to size:
– only small labile species can diffuse through the
diffusive gel and be bound on the Chelex gel
• DGT separates species according to their
lability:
– DGT device acts as a sink for the metals in
porewaters: the metal concentration in porewater will
decrease unless there is a fast remobilisation from the
solid phase
– Remobilisation rate can be estimated by:
• DGT devices with different deployment times or different
diffusive gel thicknesses
• Independant measurement of labile metal species
I. Modelling metal speciation in
porewaters
Adriano Agnese and Marco Santon
Master Thesis (March 2005)
Promotor W. Baeyens and M. Elskens
Objectives
• To obtain information about the extent to which metals
are organically bound in natural waters
• Two principal aims:
– Firstly to use DET and DGT measurements to determine total
and labile metal fractions with a high resolution profile.
– Secondly to use these measurements in a multi-element multiligand interaction model to provide free metal concentrations
and investigate the extent to which these metals are organically
bound.
Assumptions
• Metotal = Melabile + Menon-labile
where
Metotal = DET
Melabile
= DGT
Menon-labile
= DET-DGT
It is assumed that the labile fraction mainly represents
inorganic metal species and the non-labile fraction the
organically bound metal (strongly bound)
Tools
• Multi-element multi-ligand
interaction model to provide
free metal concentration and
investigate the extent of
inorganic metal complexation
• Single site complexing model
for the complexation of metals
with humic materials:
m

j
DGTMe

Me



Me

L

i
i
j
i
j


j1

n

 TL  L 
i  Mei  L j j

j
j


i1
Hum  L free 
DGT  DET
Mefree
• Two-site complexing model for 1  L free  1  2  L2
the complexation of metals
with amino-acids
free

DGT  DET
Mefree
Procedures
• The multi-element multi-ligand interaction model represents a set of
non-linear equations that are solved with a Weighted Least Squares
technique
• The free ion activity coefficients required to make activity
corrections in the model is performed with the Davies Equation. It
takes into account ionic strength and temperature effects.
• Uncertainty on the final model results are quantified with MonteCarlo simulations
• Stability constants for humic and amino-acids interactions are
assessed using DOC profile and average molecular mass given by
Schwarzenbach et al. [1993]
Results: Humic acid (log K)
Ref
Site
Cu
Cd
Zn
This study
Warneton
9.0–10.9
8.0-10.3
7.0-8.9
Mantoura et
al. 1978
In vitro
8.9.0–11.4
4.6-5.1
4.8-5.9
Vanden berg
et al. 1987
Scheldt
11.8-14.0
Baeyens et
al. 1993
Scheldt
10.6
8.6-10.6
8.9
9.1
Results: Amino acids (log K1 ~ k2)
Ref
Site
Cu
Cd
Zn
This study
Warneton
5.1-6.3
4.7-5.9
4.0-5.1
Valenta et
al. 1986
In vitro
6.7-8.6
4.0-5.0
5.0
Baeyens et
al. 1993
Scheldt
7.8
6.9
7.0
II. DGT induced fluxes in
sediments (DIFS)
(Harper, 1998)
DGT device in sediments
Resinlayer
(b)
(c)
Sediment
Concentration
Ca
a) Diffusion only case
b) Fully sustained case
c) Intermediate case
(a)
diffusiongel
filter
}diffusionlayer
DIFS: Assumptions
• Only two labile pools (dissolved, sorbed)
• First order reversible reaction
C soln
k1
Csolid
K-1
• Only passive mobilisation due to a decrease in
•
•
metal concentration in porewater
1D model
Homogeneous sediment
DIFS: model components
Labile Kdl = C-labile-solid
Resin Diffusive
gel
gel
diffusion
C soln
Response time
diffusion
C soln
removal
k1
K-1
Tc =
1
C lsolid
K1 + k-1
Large Tc- slow response
Small Tc-rapid response
DIFS
• R value (remobilisation rate)
R= C-DGT
C-labile porewater
C-labile porewater: measured or calculated by
speciation model
• Introduce R in the model: Kdl and Tc can
be calculated
• Introduce all parameters in the model:
simulation of DGT behaviour
III. Modeling reactive transport
in aquatic sediments
Diagenetic modeling
(CEMO, Yrseke)
Diagentic modeling
• Pathways of organic matter mineralisation
• Coupling among the biogeochemical cycles
of C, N, O, S, Mn, Fe, ...
• Recent developments: build models in
commercially available software: FEMLAB
(older models Fortran codes)
Diagentic modeling
• Identify for the test site (Warneton) most important
reactions
– Mineralisation
– Precipitation/dissolution
– Equilibria, ...
• Establish mass balance calculations
• Building of the model for each element which reactions
•
are important
Model can then be used to reproduce porewater and solid
phase constituents:
– Under present conditions
– Under varying environmental conditions: higher oxygen
concentrations, less organic matter, ...