TAGS: Science-based Formulation
With the help of Hansen Solubility Parameters (HSP), the modern formulator is able to find compatible ingredients faster. The standard method for measuring the HSP of any material such as a polymer, a solute or a nanoparticle is to put the sample into, say, 20 different solvents and judge whether the sample is “happy” (Score =1) or “unhappy” (Score=0) in the solvent. A standard fitting routine then creates a sphere that (ideally) includes all the 1s and excludes all the 0s. The center of the sphere is the HSP and the radius defines, under those specific measurement conditions, what is a good or bad solvent.
With modern techniques, it is now possible to get objective measurements of how samples interact with the solvent. Examples include:
-
solubility for solutes,
-
% swelling for partially crosslinked polymers and
-
settling time for particles in the solvents.
“Why not use these objective data rather than subjective 1/0 scores?”, says Prof. Steven Abbott, a leading independent scientist, who is passionate about science-based formulation.
He continues, “
One possibility, used in the HSPiP software for many years, is to fit the data to a formula that gives an exponential decrease in the measured parameter with increasing HSP Distance. A more recent technique is the Optimal Binary Fit (OB-Fit). It has been developed by Prof Dietmar Lerche of LUM GmbH and Prof Doris Segets at U Duisberg-Essen.”
The starting point was a technique developed by LUM to get an objective measure of the sedimentation time of nanoparticles. This method uses a centrifuge with a sensor array that can measure the light transmitted through each tube. And as it spins past, it can see when a homogeneously absorbing dispersion changes to high absorption at the bottom of the tube and a clear upper layer, depending on:
-
the speed of the centrifuge,
-
the quality of the particle/solvent interaction and
-
the densities and viscosities of the solvents.
The diagram above shows that the initial optical density (OD) in orange changes quite slowly after a given time for a good solvent. And for a bad solvent, that same initial OD changes rapidly as the particles settle out.
After correcting the sedimentation time, ST, to account for viscosity and density effects, the relative settling time, RST, captures the difference between good and bad solvents. The table on the left of the figure is sorted by RST in the Score column.
“
It is clear that the sample (in this case a carbon black) is very happy in THF and ethyl acetate and deeply unhappy in xylene and hexane. But where should we draw the line between good and bad solvents? Maybe we say that everything down to #4 NMP is good, or maybe we should go to #7 ethylene glycol monobutyl ether.”, says Abbott.
Lerche, Segets and their colleagues have come up with an ingenious scheme
1. This involves:
Calculating the HSP using just the top 2 solvents as good.
Next, calculating it with the top 3, the top 4, right down to near the end of the table, leaving just the worst 1 or 2 solvents out of the calculations.
As the number of solvents defined as good changes, the calculated HSP value changes. But at the optimal point (hence, the name Optimal Binary), the fit is least sensitive to including/excluding a solvent.
Further, the OB Fit technique within HSPiP provides the required data.
As the number of good solvents increases, the distance between consecutive calculated values oscillates then reaches a minimum with 6 solvents. So, the HSP [17.2, 8.5, 11.5] calculated with these 6 good solvents is taken as the optimal value.
Because no algorithm is perfect, the user can choose a manual Split High/Low if they want to over-ride the automatic value.
The technique works for any numerical data, not just RST data from centrifuges.
“
Since its introduction into v5.1.07 of HSPiP, it has proved to be popular with the increasing number of users who have numerical data thanks to modern analytical techniques. I would like to thank Profs Lerche and Segets for their kind permission to use their algorithm and to the team at LUM for testing an early version within HSPiP”, concludes Prof. Abbott.
1 Sebastian Süß, Titus Sobisch, Wolfgang Peukert, Dietmar Lerche, Doris Segets, Determination of Hansen parameters for particles: A standardized routine based on analytical centrifugation, Advanced Powder Technology 29 (2018) 1550–1561
About Science-based Formuation Group (SBFG)
SpecialChem S.A. with Van Loon Chemical Innovations (VLCI) and Professor Steven Abbott joined forces to establish the Science-Based Formulation Group (SBFG). The three parties have one mutual intention: endorsing the digitalization of formulation data to enable ingredient suppliers and formulators to accelerate the trend towards Science-Based Formulation (SBF). By using scientific models and relevant data at a large scale across a broad range of formulation issues, formulators can now move away from trial-and-error formulation.
Formulators, chemical and material suppliers, as well as industry experts are invited to join the
Science-based Formulation Community (SBFC) LinkedIn group to contribute their data and ideas and help promote the use of scientific models to make formulations more data-driven and more predictive.
Source: Science-based Formuation Group (SBFG)