Saturday, 8 March 2014

Can surfactant micelles cheat solubility products?

©2014 Federico I. Talens-Alesson

What we are going to discuss here is whether a charged surface changes the solubility of a salt. If you have followed my blogs, you know that I postulate that solution chemistry is (mostly) the chemistry between of surface excesses of chemicals. A normal boundary to a solution is either the air/water surface or a wall (the container's) or a probe. However, seed crystals and colloidal particles would also be surfaces against which a surface excess may develop. This is the underlying reason for all sorts of fouling and cryytal growth/nucleation phenomena.

In the particular case of a micellar solution of an ionic surfactant (which is a type of colloidal system), it also demands saturation of the air/water interface with surfactant (which means that it also demands saturation of any other surface) forming a charged surface layer. At the time I was still publishing in scientific journals, and I went as far as to show that the critical conditions for micellisation is ionic pairing between the surface excess of an ionic surfactant and its counter ion's surface excess (Talens-Alesson, Langmuir 2010, 26(22), 16812–16817)

However, as the surface excess and concentration are related, I also showed that using Bjerrum's correlation it is possible to estimate the relative bindings of different polyvalent cations onto anionic surfactant micelles. Biding of cations onto anionic micelles is a particular case of ionic pairing, and ion pairing is key to precipitation and other reactions between electrolytes. It is through forming of ionic pairs that the building blocks for crystal assembly take place (Talens-Alesson, J. Phys. Chem. B 2009, 113, 9779–9785).

In short, the higher charge counter ion predominates and for as long as the half length of the hedge of a cube of volume (volume of solution/concentration of higher charge counter ion times Avogadro's Number) is less than Bjerrum's distance for ionic pairing, then this counter ion would bind onto the micelle. This works well for the binding of Al3+ and Zn2+ and subsequent flocculation of the surfactant SDS (sodium dodecylsulfate).

This required assigning an apparent charge of 5e- to the “counter ion adsorption patch” of an SDS micelle. This was required even though the binding of counter ions onto an SDS spherical micelle follows a ratio 1 counter ion (regardless of charge) per three SDS molecules in the micelle. This means that the binding attraction is more than it should be expected. And this brings us to the question of whether enhanced interaction of electrolytes near micelles is possible.

In practice, research on ionic surfactants focuses on how they help other chemicals (essentially organic ones) to dissolve through micellization. It also focuses on the solubility of ionic surfactants themselves in the presence of a range of counterions. The latter has to do with the presence of certain cations available (and therefore in soluble form) either in nature (soils, for remediation or oil recovery, for example) or in water in technological applications (detergency). Of course, how the surfactant may affect the solubility of something already precipitated seems a rather bizarre subject of research, particularly if one believes that chemical reactions in solution are bulk phase phenomena. Of course, I don't. And it should not be so difficult to realize that a broader investigation of how micelles affect chemical reactions should take place, considering that there is some evidence in its favor. An example of a micelle-enhanced Fenton reaction is given by Talens-Alesson (Chem, Eng Technol 2003, 26(6) 684-687).


Let us consider an aged data set, dating back to 1994, by a team from the IAST at the University of Oklahoma in Norman (Scamehorn et al., Separation Sci Technology, 1994, 29(7) 809-830). It describes binding of single cations and their mixtures onto micelles of anionic surfactant SDS.

In Figure 1 I plot my own calculation for experimental binding ratios for the cations investigated by the researchers at IAST, following:
total cation concentration minus final cation concentration in solution
one third the difference between total SDS concentration minus final SDS concentration in solution

together with my own predicted 'Bjerrum's” binding ratio as discussed above. There is a constant offset between the actual binding and the predicted one using Bjerrum's correlation.

If Bjerrum's correlation was correct, then binding would be lower than it actually is. But then, Bjerrum's correlation is derived on the assumption that solution chemistry is bulk phase concentration chemistry. Therefore, it is possible that there is a condition for diluted concentrations still matching the true critical ionic pairing condition at the surface excess regions. But this means that ionic pairing can occur well below Bjerrum's condition for ionic pairing: it would occur for the “surface excess” condition for ionic pairing, which would be masked by Bjerrum's use of bulk phase concentration.

This deviation is consistent with the fact that above saturation the surface excess is constant: above the saturation surface excess the value is independent of the concentration, and therefore the condition for ionic pairing appears to be a constant. The additional binding at a lower concentration would be a constant amount, related to the difference between the saturation surface excess and the critical surface excess for ionic pairing. Figure 2 shows this.

This does not imply that the linear relationship can be stretched indefinitely: Figure 3 shows some data by IAST researchers for 4mM counter ion concentrations, below a Bjerrum's concentration (would be 4.5 mM for a divalent cation against an SDS micelle). In these experiments the variation is in the SDS concentration: from 100 mM, to 200 mM and on to 400 mM. The concentration of counter ion is always 4 mM. As the concentration of surfactant increases, the binding drops which is consistent with the fact that there is more surfactant colloidal surface for the same amount of counter ion. As the ratio surface to volume increases, with the ratio surface excess to concentration being basically constant at low concentrations, then the total amount of solute is distributed over a lower surface excess matching a lower concentration. The binding ratio decreases but the total binding slightly increases. Binding of counterions onto ionic micelles may therefore exceed theoretical values and micelles are attractors of counter ions.


Paton and Talens (Langmuir, 2002, 18 (22), pp 8295–8301) describe competitive binding of Al3+ and Zn2+ onto flocculating SDS micelles. The results indicate that the total charge of cations exceeds that of surfactant in the floc in some cases, with apparent charge ratios higher than 1 and sometimes up to 1.8 (Figure 4). The only explanation is ionic pairing between said cations and sulfate anions present in the solution. This leaves them as lower charge species and more of them are required to reach electroneutrality around the micelles.

This means that the flocculate is a mixture AlxZnyDSzSO4t. It is important to notice that the concentrations of all the chemicals are below their solubility limits. Aluminum sulfate and zinc sulfate are soluble at higher concentrations than the ones the flocculate forms. The excess chemicals are “opted in” into the flocs irrespective of their solubilites.

This means that a discontinuity in the solution, with the ability to cause enhanced adsorption of counter ions, may lead to enhanced capture of ionic pairs/salts. In a subsequent paper it was proposed that ionic pairs AlSO4+ and ZnSO40 may co-adsorb onto the micelles (Talens-Alesson, J. Phys. Chem. B 2009, 113, 9779–9785), explaining this apparent charge inversion. However, this explanation does not change the fact that well below their saturation conditions, fragments belonging to the Al2(SO4)3 and ZnSO4 salts are incorporated onto the flocs.

This is a peculiar experimental result, as it is reported as part of a fraudulent effort consisting in two papers in which staff at the University of Nottingham (one of them eventually moving to Oxford) plagiarized work of mine to pretend that they were active in the research of micellar flocculation.

In the second of them, published in Separation Purification Technol (2008) they claim flocculation of AL(DS)3 from solutions in which the residual Al and SDS concentrations they claim to have found (far below normal results in micellar flocculation of Al(DS)3) could only be explained by a strange charge ratio of around 4/1 between Al and DS (assuming the concentrations given where Al3+ and not aluminum sulfate, in which case it would be around 8/1). While in pH adjusted solutions it is possible to have rather high ratios Al3+/SDS (2 to 1), this only happens because in the adequate range of pH Al13 is present. The authors of this document do not state such pH adjustement, and the concentrations of surfactnat and SDS do not justify the charge inversion, as observed in previous work. A number of documents on the case of this story on plagiarism and fraud are available on the net.

However, there is a question: considering that various other statements make absolute no sense (stating that benzoic acid is a reagent for the analysis of anionic surfactants in two phase titration, that the formula of aluminum sulfate is AlSO4, or that phenol is an alkali, or that the Critical Micelle Concentration of a surfactant can ever be tolerable in a water targeted for processing for drinking purposes) even though the result seems unlikely there is no obvious reason why the authors, in their overall ignorance, should realize that the expected Al concentration would be unreasonably high considering they expect to remove a concentration of phenol which would be unrealistic (too low for an attempt to recycle/recovery). In principle this should be an irrelevant question if the experiment was a regular micellar flocculation one.

But there is a peculiarity in this experiment. Unlike previous work, it was carried out while air was bubbled through the solution as flocculation took place. This presents an interesting question. It is known that foam can be used to remove the foaming surfactant but also other chemicals present, adsorbed on the surfactant layer of the foam. There is for example work on the removal of moderate amounts of phenol in SDS foams.

What if the foam actually provides a barrier that may be crossed by ionic pairs or ionic molecules? What if on the air side they can remain for long enough for a shift in the effective equilibrium of solution, becoming an electrolyte sink? Figure 5 illustrates the idea.

On the grounds of previous information it is suggested that foams may assist to cause an “enhanced” insolubilistion of salts, which a variety of potential technological applications.
The way in which this may be brought to happen would be to use a micellar solution of surfactant, and cause air to bubble in the presence of the electrolytes targeted to produce the desired insoluble salt.