Reactivity of mineral oil in the Green Frix process
This study is used to show the reactivity or not of a mineral oil and its modification during the plasma thickening process. The oil studied in this study is mineral oil 600 neutral solvent composed mainly of alkane (C20-C40) with less than one % polyaromatics (w/w). Based on measurements by GPC and signal comparison, it was shown that an additional population appeared on the GPC graph showing the reactivity of this oil in the plasma process (see graph 1). Based on this observation, a first bibliographic study is carried out to highlight the "chemical" phenomenon at the origin of this population (surrounded in green in Figure 1) in the GPC curve.
Figure 1: GPC overlay of a treated sunflower oil only sample (GXT 117-2000) and a sample diluted with 600 NS mineral oil (GXB 118-2000)
Here is a representation of the components present in 600 neutral solvent oil. All alkane chains are not represented for convenience reasons. An example is highlighted for illustration:
Figure 3: Illustration of the plasma polymerization process adapted from
The literature provides information on the reactivity of many species that can react by plasma. Preferably, aromatics, alkenes and cyclic species react faster than alkanes, but their reactivity should not be overlooked. Indeed, the population at about 3.34 (Figure 1) on the Logarithm axis of the molar masses is due to the reactivity of the mineral oil towards the process, thus creating a new population that is not present when the mineral oil is absent from the process.
Estimation of the quantity of mineral oil reacted in the process
Figure 4: Deconvolution of a 1/1 mixture of sunflower vegetable oil and 600 NS
The area obtained for a 1/1 mixture of sunflower vegetable oil and 600 ns is shown in the table below. Since the detector response factor for sunflower vegetable oil was not identical, a correction factor of 1.21 (ratio of area 600 NS/area HVT) was applied to the areas corresponding to 600 NS oil (peak 2,3,4). Indeed, the response is stronger for this substance, so the corrective factor is applied to accurately calculate the percentages of species present. This value is determined by knowing that we have a 1/1 mixture and that therefore the areas corresponding to sunflower oil and 600 NS oil must be equivalent. The corrected and uncorrected values are shown in the table 1.
Table 1: Deconvoluted area of the 1/1 mixture of sunflower vegetable oil and 600 NS
Figure 5: Deconvolution of the 1/1 mixture of sunflower vegetable oil and 600 NS plasma treated (GXB 118-2000)
Table 2: deconvoluted area of the GXB 118-2000 sample (in brown the peaks corresponding to mineral oil, in blue the peak reflecting the reactivity of mineral oil with HVT as discussed in Figure 1)
The sum of the corrected areas corresponding to 600 NS mineral oil only reaches a value of 0.34365363, whereas if the mineral oil did not react, this value should reach 0.54489917. As a result, 36.94% of the mineral oil in the 1/1 mixture reacted with sunflower oil.
Due to the reactivity of mineral oils during the process, it is no longer possible to consider mineral oil as a simple diluent and therefore any GREEN FRIX product reacting with mineral oil during the process must be subject to a different REACH registration than the one under investigation.
 Hudis et al., Journal of applied polymer science, 1972, 6, 2397-2415
 Kobayashi et al., Macromolecules, 1974, 7 (3), pp 277–283
 Skurat and Dorofeev, Die Angewandte Makromolekulare Chemie 216 (1994) 205-224
 Hong et al., plasma and polymers, 2002, 3, 245-260
 Khelifa et al., Chemical reviews, 2016, 116, 3975-4005
 Yasuda et al., Nucl. Instrum. Methods Phys. Res., Sect. A 2003, 515, 15−30