Conceptual Models > Database selection and modification

Potassium permanganate oxidation of Manganese(II)

(1/1)

Nick:
Dear Dr Parkhurst,

I have been working on a tailings dam water treatment project where we intend to remove manganese by chemical oxidation using potassium permanganate. However, I have been having some difficulties finding the log K and delta H values for KMnO4 and was wondering if you might be able to help. I also have a few question regarding my model setup.

The model setup is as follows: Initial solution to define the water quality, equilibrium phase to allow different precipitates and fix pH, portlandite phase defined for fixing pH, potassium permanganate phase defined to be used as a reactant and a reaction data block used to react potassium permanganate with the initial solution.

I have noticed that if I change the log K and delta H values that the potassium permanganate dose requirement to reduce the solution manganese concentration to below 0.5 mg/L does not change. Hence, I am a little unsure if the reaction is set up correctly. The model results in terms of the amount of potassium permanganate required to remove manganese also do not reflect my experimental results, with the model predicting a potassium permanganate dose of half the experimentally derived dose.

Please find my PHREEQC model input file attached.

Any help would be much appreciated.

Thank you,
Nick

dlparkhurst:
A quick check of the solubility of KMnO4 is 63 g/L, which is 0.4 mol/L. This is much greater than the amount of KMnO4 that you are adding, so there should be no solubility limit on the dissolution of KMnO4.

In addition, you are adding the permanganate as a REACTION, which does not consider thermodynamics or solubility, it simply increases the solution concentrations by the specified amounts. The log K is irrelevant here, so you could skip defining the "Permanganate" in PHASES, and simply use the formula KMnO4 in the REACTION definition.

 The general sequence of the simulation is to form hausmannite, then make manganite while dissolving hausmannite, and then form pyrolusite while dissolving manganite. The simulation assumes instantaneous reactions. A couple things could be going on in your experiments relative to the calculation. First, the simulated mineral sequence may not be right. You may form different minerals with different redox states. Perhaps more importantly, the first minerals that form may be less stable than the ones that are simulated. It is expected that less stable phases form first, which then convert to more stable phases over time. So it could be complex as the mineral valence state changes from <3 to 3 to 4, while the stability of the individual solids is also changing. Finally, the rate of oxidation of Mn may be kinetically controlled.  It may not be the nice equilibrium system that is simulated. Slow kinetics may appear to require more permanganate. Maybe you have some information on this if you repeated an experiment with different wait times between doses.

Note that you have not included "Manganite" in your sludge production totals, so you are missing some of the precipitate.

Navigation

[0] Message Index