5 Feb 2015 Rational chemical replacement with HSP
The Massachusetts Toxics Use Reduction Institute, TURI, asked me to give a webinar showing how Hansen Solubility Parameters, HSP, can make it much more straightforward to replace a current chemical (which has fallen into regulatory disfavour) with something both safer and with good performance.
The key is that many of the things we like to use such as solvents, polymers, plasticisers, nanofillers can be characterised by their HSP values and if the "distance" calculated between two sets of HSP values is small then they are likely to be compatible. So if one solvent is taken from you, finding a replacement solvent from a list of alternatives can be done straightforwardly if you have a way of knowing the distances of each solvent in a list from your target. This is most conveniently done in the HSPiP software that I co-authored with Dr Charles Hansen and Dr Hiroshi Yamamoto, though you can download an Excel workbook from my Practical Solubilty site.
The problem is that so many solvents have been declared "bad" that it is becoming impossible to find a single solvent to replace another. That is when HSP (and HSPiP) become especially useful. It turns out that a mix of bad or not-very-good solvents can, under the right circumstances, give an excellent solvent. So now the challenge becomes how to select 2 or 3 solvents which, when combined, have a small distance from your bad solvent. Again HSPiP does this very well and is one of the key reasons it is used by formulators around the world.
You can do more than optimise solvents. Suppose your favourite plasticisers have been declared bad (e.g. phthalates). It is no good just using a green plasticiser such as a Citroflex. If the HSP distance is too large it is guaranteed to bloom to the surface over time. Because these plasticiser tests can take months it is important to choose molecules with a low distance in order to increase the probability of success.
The distance approach also lets you think about diffusion through plastic films - such as bad things diffusing into a food container or good things (like flavour molecules) diffusing out. Although the diffusion coefficient will depend on the polymer and the molecular size, if the HSP distance is small then there is likely to be a high concentration of the molecule in the first 1nm of the polymer and therefore a large concentration gradient across the polymer which drives diffusion. So if you have a set of fragrance molecules, those with a small distance are likely to be lost much faster than those with a large distance.
One reason that food packaging is so good at keeping good things in and bad things out is that the thin films are multi-layer, with very different HSP polymers in each layer. If the first layer is PE then things like alcohols will not get through but aromatics and alkanes will get through rather easily. If they next meet an EVOH layer they are so insoluble in EVOH that despite it being only a few μm thick it is a very good barrier. The EVOH would be useless for the alcohols, but they never get there because the PE is too good a barrier.
So with some simple principles of knowing HSP, calculating the distance and basically invoking the principle that "like is compatible with like" you can do a large amount of intelligent replacement of toxic chemicals - along with many other formulation tasks.
It was a real pleasure to work with the professionals at TURI to prepare the webinar and as it was a "sold out" attendance I'm repeating it on 20 Feb. Clearly there is a lot more that can be done with TURI to provide, for example, HSPiP datasets of "good" and "bad" chemicals so our joint work will continue in the months ahead.