It is important – for example in the formulation of printing inks – to combine resins with the right process oil. In the right combination, printing inks have the best rheological properties, which in turn influence the inks' tack and misting as well as other printing properties.
Tests are required to assure best performance, but it is always an advantage if the test series can be limited from the start through knowledge of which oils have the necessary prerequisites.
Solubility parameters
Specifications for process oils state aniline point and Viscosity Gravity Constant (VGC) as measurements of solubility. Other ways of describing solubility are Seal Compatibility Index (SCI) and Critical Solution Temperature (CST). CST is above all a good indication of the soap-oil interactions in greases (see Naphthenics Magazine 3/99). Unfortunately these parameters are not the ones normally stated for resins. Resins are often characterised by a solubility parameter instead. If the solubility parameters of oils were known, it would be possible to judge which oil might be interesting to use together with a particular resin, and vice versa.
Hildebrand's solubility parameter –d– is a measurement of the dispersive forces between the molecules in a chemical substance. It can be calculated for example from a substance's boiling point, heat of vaporisation, refractive index and molar volume. When it comes to oil, it is not possible to obtain the specific values of these properties, since oil consists of a very large number of chemical compounds.
We must therefore investigate whether it is possible to determine the solubility parameter for an oil indirectly. This should be possible by determining the solubility of the oil with one of the methods that are proven to work and comparing this with substances of known solubility parameter values.
It is, however, not sure that we can actually determine the solubility parameter for an oil. Nonetheless, if the value of the solubility parameter obtained with the aid of indirect methods is the same regardless of whether it is determined by comparing with the aniline point, VGC or SCI, then we can be fairly sure that the value gives a relevant measurement of the oil's solubility. This theory must therefore be tested.
Different values with different methods
Nynas R&D department embarked on a series of tests to determine the value for aniline point, VGC, SCI and CST for a number of solvents with known solubility parameters. It proved to be difficult to find suitable solvents with documented solubility parameters. One approach was to use mixtures of various solvents and calculate the parameter of the mixture. However, testing has shown that mixtures can be a source of error, not least in the determination of aniline point.
The values for aniline point, VGC, SCI and CST were plotted against the known or estimated values for Hildebrand’s solubility parameter, d. Based on these curves, the solubility parameters of three oils were determined. The oils tested were: Nytex 801, an unlabelled naphthenic oil with moderate aromatic content; Nyflex 222 B, which is a naphthenic technical white oil; and Nytex 820, an oil with a slightly different crude base and with a slightly more paraffinic character than the other two.
It was hoped that each oil's d value would be the same regardless of which curve was referred to. If that had been the case, it would have confirmed that d was a usable parameter for oils. However, it was not the case. Each oil's d-value varied rather strongly depending on which analysis formed the basis of the curve used to read off the value (see table).
| |
Aniline Point |
VGC |
SCI |
CST |
| Nytex 801 |
69.1 |
0.863 |
9.3 |
113.0 |
| Nyflex 222 B |
103.3 |
0.826 |
1.5 |
115.7 |
| Nytex 820 |
84.5 |
0.864 |
3.6 |
114.6 |
| |
δ(Mpa)
aniline point curve |
δ(Mpa)
VGC curve |
δ(Mpa)
SCI curve |
δ(Mpa)
CST curve |
| Nytex 801 |
16.7 |
16.5 |
15.6 |
16.5 |
| Nyflex 222 B |
16.5 |
15.6 |
13.8 |
14.5 |
| Nytex 820 |
16.0 |
16.6 |
14.5 |
16.2 |
It seems that the d-value varied quite considerably, depending on the measuring method. The variation was too great for the figures to give any usable information. Moreover, the d-values for the various oils did not correlate with the results of measurements using the other methods.
Three forces determine solubility
Is it then inappropriate to try to work with solubility parameters for oils? If we look at what Hildebrand's solubility parameter actually describes, we see it is just the dispersive forces between the molecules, i.e. forces that are dependent on the size of the molecules. In the case of oil, it is not only these forces that affect solubility.
An oil consists of three main groups of molecules: paraffinic, naphthenic and aromatic. The paraffinic molecules consist mainly of straight hydrocarbon chains, the naphthenic molecules of hydrocarbon rings and the aromatic molecules of benzene rings. There are also molecules present that contain other atoms than carbon and hydrogen, primarily nitrogen and oxygen.
Above all the aromatic groups in the molecules have polar characteristics. Naphthenic groups are, if not directly polar, at least less non-polar than the paraffinic groups. Between molecules that contain oxygen or nitrogen atoms and other hydrocarbons, hydrogen-binding forces also act. Both polar forces and hydrogen-binding forces affect solubility properties. In actual fact it is the somewhat greater polar forces that give naphthenic oil considerably better solubility properties than paraffinic oils, and oils with a high aromatic content even better solubility properties.
The problem with Hildebrand's solubility parameter is that it only pays regard to one of the three forces that determine solubility, i.e. the dispersive forces. Consequently, the Danish chemist Charles Hansen has developed a solubility parameter that, in addition to Hildebrand's parameter, also takes account of the two other forces. Hansen's solubility parameter can be described by a simple numerical value (although it rarely is); since it is affected by three forces, the value is described as a sphere in a co-ordinate system with three dimensions. Dependent on the size of the three components that contribute to the Hansen parameter, a compound has different solubility properties. Even if the total value for two different compounds is the same, they can have totally different properties depending on which component of the solubility parameter is the dominant in each of them.
It will therefore be a logical continuation of our work to study whether Hansen's solubility parameter correlates with oil's solubility properties.
Even if this study did not lead us all the way to the results we had hoped for, it did at least provide certain information that can be worth reflecting on. Compare the values that Nytex 801 and Nytex 820 got for VGC and aniline point. While they have almost the same value for VGC, on the one hand, the value Nytex 801 obtains for aniline point is a little over 20 percent lower than Nytex 820:s value. Both the oils have approximately the same degree of refining. What differentiates them apart from viscosity is that Nytex 820 is an oil with a different crude basis and a somewhat more paraffinic character than Nytex 801, a naphthenic oil made from Venezuelan crude. From experience we know that Nytex 801 has better solubility than Nytex 820. This gives a clear effect on the aniline point and SCI but not on VGC.
Our study has shown that it is not possible to simplify "the solubility concept" to any great degree. That could result in what is known as a "nonsense correlation", where a model is constructed that applies to recently conducted tests. If you change a parameter, which in this case could be the degree of hydrotreatment or the crude base, then the connections and the model no longer apply. To plot the aniline point or the VGC against d and then make use of this connection, without paying regard to the influence of factors such as the degree of hydrotreatment or crude base, would probably lead to wrong conclusions being drawn.
There is doubtless much room for discussion as to which parameter in the end is actually right. The only means that can currently be applied to characterise solubility properties with safety are tests with well-defined formulations.
Our efforts continue to find the method that best describes solubility as the important property it is, and which is possible to use.
