Estimating stability of vanadinite problem

[mschock]

Hi;

 

I'm trying to develop an upper-bounds estimate for the solubility constant and/or Gibbs free energy of formation of vanadinite (as found in lead pipe corrosion scales), from some drinking water monitoring data and some [hopefully wise] estimations of dissolved lead and vanadium levels in places that can't be directly measured. My strategy is to use SpecE8 with known background water chemistry (which varies somewhat for a few constituents) to derive the activities of Pb+2, VO4-3 and Cl- for some expected ends of chemistry ranges, to get an estimated log Ksp. From that, I can use Act2 to construct an Eh-pH diagram for some different water treatment scenarios.

 

I thought I was OK up until I tried to add vanadinite into the thermodynamic database. I am using a modified version of thermo.com.v8.r6 which I have used for a couple of years, having extensively added to and modified the Pb and Cu sections. It is too big to attach, so I exerpted just the section where I added the vanadinite, and I incremented the number of solids by 1. I wrote the reaction in terms of Cl-, Pb++ and VO4---, but when trying to read the new database into SpecE8, it apparently tries to do something with the VO++/VO4--- couple and crashes with a "reaction mass imbalance for vanadinite" error (attached Word file). I am also attaching one of the representative SpecE8 script files which ran with the thermo database without the added vanadinite.

 

Is there an error in the way I entered the solid, or other changes I need to the aqueous species for V(V) as well? I think I noted a redundancy in the database, too, though I don't think it's part of this problem. Both HVO4-- and VO3OH-- are in the database.

 

As an aside, I originally tried to set Eh and include oxygen, but at the high Eh I tried (0.95 V), the calculations got way off, I think because it might have been oxidizing water. I couldn't find a way to make the calculations work by specifying Eh, which would be useful for looking at Pb solubility vs Eh at different pHs or carbonate concentrations, as the phase changes from cerussite or hydrocerussite to plattnerite. One problem that has some bearing on how I can approach this is that realistically, is many drinking water systems are in redox disequilibrium with the stability domain of water, thanks to the use of highly oxidizing disinfectants like hypochlorous acid. However, modelling based on short-term metastable equilibrium represents the systems well. Observed and inferred (from the presence of PbO2 and other highly oxidized mineral phases) Eh's are often in the 800-1000 mV range. Other systems use chloramines, a weaker oxidant, which are tough to characterize and enter into any thermodynamic database. Therefore, it would be best if I could use basis species that I know co-exist in the systems I want to model, such as VO4-3 and Pb+2, and try to ignore actual redox reactions in the key speciation calculations.

 

Any help with this addition of vanadinite and setup of the V speciation problem would be appreciated.

 

--Mike Schock

[Tom Meuzelaar]

Hi Mike:

 

The "reaction mass imbalance..." error that results when loading a modified database usually occurs because the mole mass for the new species, mineral or gas does not reflect the sum of its component mole masses. When I add the mole masses for Pb, Cl and VO4 in their respective stoichiometric concentrations, I get mole wt. = 1416.2700 g, whereas your entry has 1186.3921 g.

 

I'm not sure if there's a second question. If I understand correctly, you initially tried to model the system using Eh/O2 values and assuming a system in complete redox equilibrium, which did not work. Looking at your water, dissolved components appear to be in redox disequilibrium with the highly oxidizing conditions, so you've decoupled all species and opted to model a metastable system. Is this correct?

 

Regards,

 

Tom Meuzelaar

RockWare, Inc.

[mschock]

Hi Mike:

 

The "reaction mass imbalance..." error that results when loading a modified database usually occurs because the mole mass for the new species, mineral or gas does not reflect the sum of its component mole masses. When I add the mole masses for Pb, Cl and VO4 in their respective stoichiometric concentrations, I get mole wt. = 1416.2700 g, whereas your entry has 1186.3921 g.

 

I'm not sure if there's a second question. If I understand correctly, you initially tried to model the system using Eh/O2 values and assuming a system in complete redox equilibrium, which did not work. Looking at your water, dissolved components appear to be in redox disequilibrium with the highly oxidizing conditions, so you've decoupled all species and opted to model a metastable system. Is this correct?

 

Regards,

 

Tom Meuzelaar

RockWare, Inc.

 

Hi, Tom;

 

Thank you. I figured out it might have something to do with the typo with the mass for vanadinite shortly after I posted the question, and your answer confirmed it. So that problem is solved.

 

I have two remaining questions.

 

First, for this current modeling exercise, I am very confused by the different ways the different thermodynamic databases handle the vanadium aqueous speciation, and how I should combine that with the lead. In this case, I have a system where the Eh is lower, and the likely coexisting valence states are Pb(II), V(V) and P(V)--orthophosphate. The LLNL thermo.com.r8.v6+.dat database that I've been working from and modifying (because of the completeness of some of the metals), seems to be missing some V(V) aqueous species, and it seems also to have some duplicates but with a different set of reactants to create that species. For example, it has both HVO4-- and VO3OH--, though their formation reactions are different. The thermo.dat database seems to have a more comprehensive set of species for V and uses three different basis species representing V(III), V(IV) and V(V), but dependent on redox reactions of V(IV). How can I reconcile these two approaches, to give me the ability to either do Eh-pH diagrams for both Pb and V, but also be able to use SpecE8 with specifying a single Eh? I have tried to incorporate the V(V) reactions into my own database, but the more acidic V(V) species are not being displayed, and I believe it may be becasue their formation reaction is not based on VO4---. I have attached a Word file showing the V(V) species that are in each of these 3 databases, for your reference.

 

Second, is the general question, and that is basically how to properly set up speciation calculations in GWB for high Eh values (above water stability bounds), to prevent water from being removed from the system and producing misleading concentrations. Would I need to make sure all metastable elements (such as would be the case for Pb(IV) species and any other concurrent ligands or metals) are decoupled in some way, and then specify Eh input? I temporarily gave up on this when I kept having the problem with convergence or unrealistically high. Or, would I have to actually make sure the drinking water oxidants are added in to the database as basis and aqueous species (such as hypochlorous acid/hypochlorite, chlorine dioxide, chloramines) so they can participate in the reactions? Or is there some other way.

 

I hope this explanation is sufficiently clear.

 

--Mike

[Tom Meuzelaar]

How can I reconcile these two approaches, to give me the ability to either do Eh-pH diagrams for both Pb and V, but also be able to use SpecE8 with specifying a single Eh?

 

I have little experience with Vanadium thermodynamic data, and cannot speak directly to the consistency and accuracy of the data in the various datasets. For this, I would refer to the original source, or search the forum archives for additional discussion on this topic. The reference for thermo.dat:

 

Delany, J.M. and S.R. Lundeen, 1990, The LLNL thermochemical database. Lawrence Livermore National Laboratory Report UCRL-21658, 150 p.

 

This might help as well - thermo.dat uses V+++ (V3) to represent the vanadium Basis species, whereas thermo.com.v8.r6+ uses VO++ (V4). In each database, the other vanadium oxidation states are related to the vanadium Basis species by redox reactions involving a master redox species, oxygen. So, regardless of whether you constrain your React or SpecE8 Basis concentration for vanadium as V+++ (V3) or as VO4--- (V5), GWB will automatically calculate the equilibrium distribution of aqueous vanadium species for all vanadium oxidation states via the master Eh species, oxygen. This is the single Eh approach you refer to. Theoretically, the choice of Basis species shouldn't matter. If the equilibrium constant data between datasets are consistent, and both datasets have the same selection of aqueous species, minerals and gases for a given component, the results should be the same for different databases at equilibrium. Unfortunately, this is often not the case.

 

The single Eh approach assumes the system is in redox equilibrium. You can also model a system in redox disequilibrium in GWB by "decoupling" any redox reaction. When this is done, the two redox species are no longer tied to one another through the master species, oxygen (or through a single system Eh) and must be constrained separately in the SpecE8/React Basis. As an example, consider the speciation of your two duplicate species from thermo.com.v8.r6+, HVO4-- and VO3OH--. The formation reaction for HVO4-- is written in terms of the vanadium Basis species, VO++ (V4), whereas the formation reaction for VO3OH-- is written in terms of a redox couple species, VO4--- (V5). If you decouple the VO++/VO4--- couple, VO4--- is no longer determined via equilibrium with VO++ and oxygen. Therefore, the only way the model can predict the presence of the VO3OH-- species, is if you constrain VO4--- (V5) in React/SpecE8 as an independent component. On the other hand, if the VO++/VO4--- couple remains coupled, GWB can predict the presence of VO3OH-- based on input of a single vanadium concentration through the equilibrium reaction among VO4---, VO++ and oxygen (reflecting the overall system Eh).

 

Second, is the general question, and that is basically how to properly set up speciation calculations in GWB for high Eh values (above water stability bounds), to prevent water from being removed from the system and producing misleading concentrations. Would I need to make sure all metastable elements (such as would be the case for Pb(IV) species and any other concurrent ligands or metals) are decoupled in some way, and then specify Eh input? I temporarily gave up on this when I kept having the problem with convergence or unrealistically high. Or, would I have to actually make sure the drinking water oxidants are added in to the database as basis and aqueous species (such as hypochlorous acid/hypochlorite, chlorine dioxide, chloramines) so they can participate in the reactions? Or is there some other way.

 

I assume that you mean that the individual Eh values among redox couples are out of equilibrium with the Eh value for oxygen? In that case, yes, you have to decouple each metastable redox couple, and constrain the values for the various redox components independently. GWB will calculate the Eh value for each redox couple independently.

 

I hope that helps,

 

Tom

[mschock]

Thank you. That helps clarify what's going on. I guess I need to work through this in some detail and try to figure out exactly how to decouple and work with a high Eh, and also to learn better how to add appropriate species to the thermodynamic database. What is really confusing to me is the co-existence of identical species but having different different names and using different basis species. From your explanation, I still am not quite sure why this is necessary, and whether or not there is a potential conflict in the calculation. For the time being, I seem to get logical and acceptable results (as best I can determine) using a realistic Eh of 0.7 volts for one of these waters, but that won't always be the case.

 

I need to add some Pb(IV) aqueous species, so the coupling issue is also relevant there.

 

 

 

I have little experience with Vanadium thermodynamic data, and cannot speak directly to the consistency and accuracy of the data in the various datasets. For this, I would refer to the original source, or search the forum archives for additional discussion on this topic. The reference for thermo.dat:

 

Delany, J.M. and S.R. Lundeen, 1990, The LLNL thermochemical database. Lawrence Livermore National Laboratory Report UCRL-21658, 150 p.

 

This might help as well - thermo.dat uses V+++ (V3) to represent the vanadium Basis species, whereas thermo.com.v8.r6+ uses VO++ (V4). In each database, the other vanadium oxidation states are related to the vanadium Basis species by redox reactions involving a master redox species, oxygen. So, regardless of whether you constrain your React or SpecE8 Basis concentration for vanadium as V+++ (V3) or as VO4--- (V5), GWB will automatically calculate the equilibrium distribution of aqueous vanadium species for all vanadium oxidation states via the master Eh species, oxygen. This is the single Eh approach you refer to. Theoretically, the choice of Basis species shouldn't matter. If the equilibrium constant data between datasets are consistent, and both datasets have the same selection of aqueous species, minerals and gases for a given component, the results should be the same for different databases at equilibrium. Unfortunately, this is often not the case.

 

The single Eh approach assumes the system is in redox equilibrium. You can also model a system in redox disequilibrium in GWB by "decoupling" any redox reaction. When this is done, the two redox species are no longer tied to one another through the master species, oxygen (or through a single system Eh) and must be constrained separately in the SpecE8/React Basis. As an example, consider the speciation of your two duplicate species from thermo.com.v8.r6+, HVO4-- and VO3OH--. The formation reaction for HVO4-- is written in terms of the vanadium Basis species, VO++ (V4), whereas the formation reaction for VO3OH-- is written in terms of a redox couple species, VO4--- (V5). If you decouple the VO++/VO4--- couple, VO4--- is no longer determined via equilibrium with VO++ and oxygen. Therefore, the only way the model can predict the presence of the VO3OH-- species, is if you constrain VO4--- (V5) in React/SpecE8 as an independent component. On the other hand, if the VO++/VO4--- couple remains coupled, GWB can predict the presence of VO3OH-- based on input of a single vanadium concentration through the equilibrium reaction among VO4---, VO++ and oxygen (reflecting the overall system Eh).

 

 

 

I assume that you mean that the individual Eh values among redox couples are out of equilibrium with the Eh value for oxygen? In that case, yes, you have to decouple each metastable redox couple, and constrain the values for the various redox components independently. GWB will calculate the Eh value for each redox couple independently.

 

I hope that helps,

 

Tom