Supercritical fluid chromatography as a rapid single-step method for the determination of mineral oil saturated and aromatic hydrocarbons in purified mineral oils for food and

Journal of Chromatography A 1614 (2020) 460713

Contents lists available at ScienceDirect

Journal of Chromatography A
journal homepage: www.elsevier.com/locate/chroma

Supercritical fluid chromatography as a rapid single-step method for
the determination of mineral oil saturated and aromatic hydrocarbons
in purified mineral oils for food and cosmetics applications
Alan Rodrigo García-Cicourel a,∗, Bas van de Velde a, Gerry Roskam a, Hans-Gerd Janssen a,b,c
a
b
c

University of Amsterdam, Van‘t Hoff Institute for Molecular Sciences, Analytical Chemistry Group, P.O. Box 94157, 1090 GD Amsterdam, the Netherlands.
Wageningen University, Agrotechnology and Food Sciences Group, Laboratory of Organic Chemistry, Stippeneng 4, 6708 WE Wageningen, the Netherlands.
Unilever Research and Development, P.O. Box 114, 3130 AC Vlaardingen, the Netherlands.

a r t i c l e

i n f o

Article history:
Received 6 September 2019
Revised 11 November 2019
Accepted 14 November 2019
Available online 15 November 2019
Keywords:
Mineral oil analysis
Supercritical fluid chromatography

Argentation chromatography
Parallel detection
MOSH/MOAH deconvolution

a b s t r a c t
Mineral oil hydrocarbons are used in the consumer goods sector for the elaboration of a wide range of
foods and cosmetics. Traditional methods for determining their levels and composition are time consuming and laborious, besides requiring complex instrumentation. Here a simple and fast method was developed that uses columns packed with silver-modified silica in supercritical fluid chromatography with
flame ionization and UV detection (SFC-FID/UV) for the determination of mineral oil saturated hydrocarbons (MOSH) and mineral oil aromatic hydrocarbons (MOAH) in purified mineral oil samples. The method
requires no sample preparation apart from dilution. Direct quantification of MOSH and MOAH was possible for samples with MOSH/MOAH ratios around one. For other samples deconvolution of the MOSH
and MOAH humps in the FID chromatogram using the UV signal was needed since baseline separation
of the two fractions could not be obtained. Validation of the method performance showed an excellent
linearity (R2 > 0.9995) in the range of concentrations tested (2.5–100 mgmL−1 ) and a better repeatability
than the standard methods (<5%). MOAH detection limits were better than 0.36% MOAH, which makes
the method sufficiently sensitive for analysis of all but the highest purity mineral oils. The proposed
SFC-FID/UV method was suitable for the analysis of mineral oils with viscosities and molecular weights
below approximately 56 mm2 s−1 and 450 gmol−1 . The quantitative results of the new method were not
statistically significantly different from those obtained with the standard SPE-GC-FID method where the
new method has the advantages of a better repeatability, simpler operation and a significantly shortened
analysis time. This new method could potentially also be used for the analysis of mineral oil contaminations in consumer products such as foods. However, in this case additional sample clean-up and preconcentration steps are needed for reducing matrix interferences from e.g. triglycerides and olefins and for
improving the detection limits.
© 2019 The Authors. Published by Elsevier B.V.
This is an open access article under the CC BY-NC-ND license.
( />
1. Introduction
The presence of mineral oil hydrocarbons in foods and cosmetics is currently receiving a great deal of attention in the consumer
goods sector. Mineral oil hydrocarbons can enter these products
non-intentionally as a contaminant from e.g. packaging or transport, but purified mineral oil fractions are also widely used as ingredients for the elaboration of a range of foods and cosmetics.
Detailed analyses are needed to establish the level of residual aro-

∗

Corresponding author.
E-mail address: (A.R. García-Cicourel).

matic species in the purified oils and to determine the extent and
type of a mineral oil contamination. In both application areas separation of the mineral oil saturated hydrocarbons (MOSH) from the
mineral oil aromatic hydrocarbons (MOAH) is essential. Several liquid chromatographic methods to do so have been described [1–
5]. Combined with a universal detector, generally a flame ionization detector (FID), the separated MOSH and MOAH fractions can
be quantified. In general, however, these methods are time consuming, laborious and require difficult to operate two-dimensional
heart-cut LC-GC instruments [6].
Separation of complex hydrocarbon mixtures based on aromaticity (polarity) is difficult because it requires a very high se-

/>0021-9673/© 2019 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license. ( />

2

A.R. García-Cicourel, B. van de Velde and G. Roskam et al. / Journal of Chromatography A 1614 (2020) 460713

lectivity towards the aromatic ring system of the molecules [7].
Whereas gas chromatography (GC) can provide this selectivity for
very small molecules, GC selectivity for the very high molecular
weight compounds encountered in typical MOSH/MOAH analyses
is insufficient. Still, GC is widely used as a second dimension in
MOSH/MOAH analyses because of its easy coupling to the FID,
which is often preferred due to its equal response for all hydrocarbons irrespective of their structure [8,9]. LC stationary phases
on the other hand can provide a good selectivity for MOSH/MOAH
fractionation [1,7,8], but unfortunately LC systems cannot be coupled to the FID. MOSH/MOAH analyses hence require coupled LC
and GC methods: LC using normal phase columns (NPLC) for fractionation of the mineral oil based on aromaticity, followed by GC
analysis using either FID for quantification, or mass spectrometry
(MS) or vacuum ultraviolet detection (VUV) for structural characterization [6,8,10]. This combined LC-GC approach was first described by Biedermann et al. in 2009 and has been further optimized and improved in a series of articles by the same authors

and other users of the method [11–14]. A significant improvement
was the use of silver-loaded silica columns which made the determination of the cut-point for the MOSH/MOAH separation less
critical [10,13]. Barp et al. optimized the original Biedermann and
Grob method by speeding up the LC and GC analyses. An analysis
time of 23 min was achieved for the analysis of each fraction, resulting in a total analysis time of 46 min when both MOSH and
MOAH levels have to be determined [12]. Despite these improvements, the long analysis time and the complicated coupled LC-GC
set-up required still make the application of the two-step method
in routine analysis expensive and complex.
In many applications supercritical fluid chromatography (SFC)
has been shown to combine the advantages of LC and GC [15].
The technique, for example, is used in the mineral oil industry
in a number of methods for the analysis of lighter fuels. Method
D5186–19 from the American Society for Testing and Materials
(ASTM), as an example, describes SFC with packed columns as an
official method for the determination of the aromatic and polynuclear aromatic content of light diesel- and aviation fuels [16]. In
SFC, all LC stationary phases can be used. Silica columns have been
extensively applied [17], and also silver-loaded silica columns can
be used [18]. If these columns would be capable of separating the
heavier mineral oils into MOSH and MOAH using pure carbon dioxide as the mobile phase, MOSH and MOAH levels could be determined directly using just one single chromatographic method,
SFC-FID. Additionally, SFC could also solve other issues related to
the online LC-GC method, such as incomplete elution of the high
molecular weight compounds in the GC-FID quantification step.
In this contribution, the applicability of packed-column SFC as
a faster and simple method for the analysis of MOSH and MOAH
in crude and purified mineral oil samples is assessed. Different stationary phases are evaluated for their ability to separate MOSH and
MOAH and provide information on the aromaticity of the MOAH
species. Parallel detection using UV and FID is studied to maximize sensitivity and selectivity of the method. Pure CO2 is used
as the mobile phase in all experiments to allow coupling to the
FID. The applicability and merits of the new SFC-FID/UV method
are discussed, and quantitative data obtained are compared with

those from the standard off-line solid phase extraction – SPE-GCFID method for MOSH/MOAH analysis.
2. Materials and methods
Standards to study the selectivity in SFC were purchased
from Sigma-Aldrich (Zwijndrecht, the Netherlands). These
comprise the aliphatic hydrocarbons n-undecane (C11 ), bicyclohexyl (Cycy), n-tridecane (C13 ) and 5-α -cholestane (Cho),
the mono-aromatics 1,3,5-tri-tert-butylbenzene (TBB) and 1,4-

di-dodecylbenzene (D12B), and finally the poly-aromatics 1methylnaphthalene (1MN), biphenyl (BP) and anthracene (Ant).
Hexane and dichloromethane (DCM), both HPLC grade, were
purchased from Biosolve BV (Valkenswaard, the Netherlands).
SFC grade carbon dioxide (grade 4.8) was obtained from Praxair
(Vlaardingen, the Netherlands). Silver-nitrate impregnated silica
gel (230 mesh – 65 μm) loaded at approximately 10 wt.% for the
SPE experiments was purchased from Sigma-Aldrich. Eight mineral
oil samples from different stages in the purification process were
obtained from a local supplier of white oils. The physicochemical
properties of the mineral oils are summarized in Table 1. Stock
solutions of the standard compounds (at 10 mgmL−1 ) and of the
mineral oils (at 500 mgmL−1 ) were prepared in hexane and stored
in the refrigerator when not in use.
2.1. SFC-FID/UV selectivity
For the experiments to map the SFC selectivity, mixtures of
aliphatic marker compounds (C11 , C13 , Cycy and Cho) and selected
aromatic standards (1MN, BP, TBB, D12B and Ant) were prepared in
hexane at levels around 300 μgmL−1 per compound. Separation of
the compounds was performed on an Acquity UPC2 SFC instrument
with PDA detection (Waters, Etten-Leur, the Netherlands) coupled
to an external FID. For FID detection, part of the CO2 mobile phase
flow was diverted to the FID. To do so, a low dead-volume Tpiece (Valco Instrument Company Inc., Schenkon, Switzerland) was
connected to the exit of the column just prior to the UV detector. A 90 cm long, 50 μm internal diameter fused-silica capillary

was used to transfer the mobile phase flow to the FID of an adjacent Agilent G1530 GC system (Agilent, Amstelveen, the Netherlands). At the end of the capillary an integral tapered restrictor
was prepared using the method described by Guthrie and Schwartz
[19]. In short, the capillary was first sealed by heating it in a butane/nitrous oxide flame and then re-opening it by gently abrading it by hand using a wet abrasive sheet. The re-opening is done
with the capillary connected to the SFC pump. Gas bubbles escaping from the capillary indicate the restrictor is open. Abrading was
continued until a gaseous CO2 flow of approximately 20 mLmin−1
was obtained. The outlet of the transfer capillary was positioned
just below the flame jet of the FID. The temperatures of the GC
and FID were set at 60 and 300 °C, respectively. Hydrogen and air
flows for the FID were 36 and 400 mLmin−1 , respectively.
The injection volume in the SFC experiments was 1 μL. For the
separation of the compounds a column set consisting of either a
bare silica column, 150 × 4.6 mm Polaris 5 Si-A 5 μm, or two serially connected 100 × 4.6 mm silver-loaded silica Chromsep SS
5 μm columns (Agilent) was used. Pure CO2 was used as the mobile phase at a flow rate of 1 mLmin−1 . Temperature and pressure
were set at 60 °C and 110 bar, respectively. Data acquisition was
performed at 20 Hz for the FID and 25 Hz for the UV detector.
UV chromatograms were acquired and processed using Empower
3 (Waters) either in the MaxPlot mode (190–400 nm) for deconvolution, or at 254 nm for maximum selectivity towards aromatic
species.
The same mixtures of aliphatic and aromatic compounds were
analyzed by LC on a Shimadzu 10Avp HPLC system equipped with
UV and refractive index (RID) detection (Shimadzu Benelux, ‘sHertogenbosch, the Netherlands). To do so, 20 μL of each mixture were injected onto the two serially connected 100 × 4.6 mm
silver-loaded silica columns previously described. Compounds were
eluted using hexane as the mobile phase at 0.5 mLmin−1 .
2.2. Optimization of the SFC-FID/UV method for mineral oil analysis
Optimization of the SFC parameters, i.e. temperature, pressure
and flow rate, was done using a two-level factorial experimental


A.R. García-Cicourel, B. van de Velde and G. Roskam et al. / Journal of Chromatography A 1614 (2020) 460713

3

Table. 1
Physicochemical properties of seven mineral oil samples.

Mineral
Mineral
Mineral
Mineral
Mineral
Mineral
Mineral
Mineral

oil
oil
oil
oil
oil
oil
oil
oil

1
2
3
4
5
6
7

8

Density (gcm−3 )

Viscosity at 40 °C (mm2 s−1 )

Molecular weight (gmol−1 )

Number of carbon atoms

0.8570
0.8631
0.8694
0.8984
0.8742
0.8859
0.9158
0.8284

16.53
10.32
30.29
22.56
56.41
91.21
153.40
18.62

352.0
284.0

379.0
312.0
452.0
519.0
423.0
388.0

C15 –C40 ,
C14 –C31 ,
C14 –C40 ,
C13 –C40 ,
C16 –C44 ,
C17 –C52 ,
C18 –C48 ,
C14 –C31 ,

Table. 2
High and low levels of the selected parameters
for the optimization of the SFC-FID/UV method
by a two level factorial experimental design.
Parameter

Low level

High level

Temperature (°C)
Pressure (bar)
Flow (mLmin−1 )

50
110
1.0

60
138
2.0

centered
centered
centered
centered
centered
centered
centered
centered

around
around
around
around
around
around
around
around

C24
C23
C25
C23

C28
C33
C30
C23

(3 min) at 15 °C/min. The FID temperature was set at 350 °C, hydrogen and air flows were 36 and 400 mLmin−1 , respectively and
the data collection rate was 200 Hz. The MOSH/MOAH composition of the mineral oils was calculated using the total area and the
areas of the respective fractions.

3. Results and discussion
3.1. SFC-FID/UV selectivity

design. The high and low levels for the three parameters are shown
in Table 2. The response to optimize was the MOSH/MOAH resolution. In these experiments 1 μL of mineral oil 1 was injected at a
concentration of 10 mgmL−1 . For the calculation of the resolution,
retention times and peak widths of the MOSH and MOAH bands
were calculated both from the FID and UV chromatograms. These
data were also used to assess extra-column band broadening in the
capillary to the FID.
2.3. Validation of the SFC-FID/UV method
To evaluate the performance of the proposed SFC-FID/UV
method, linearity, repeatability, limit of detection (LOD) and limit
of quantification (LOQ) were assessed following the Eurachem
guideline for method validation [20]. The experiments were done
using solutions of mineral oil 8 in hexane. This oil was aromatics free. Linearity was evaluated in the concentration range of 2.5–
100 mgmL−1 . Repeatability was evaluated at three concentration
levels, 7.5, 15 and 30 mgmL−1 . LOD and LOQ were calculated by
injecting three different solutions of mineral oil 8 at 0.25 mgmL−1
(five times each). For all validation experiments, 1 μL of the different samples were injected. These experiments were performed
at 60 °C and 138 bar at a supercritical CO2 mobile phase flow of

1 mLmin−1 . The other instrumental parameters were maintained
as mentioned in the Section 2.1.
2.4. Analysis of real samples
Solutions of mineral oils 1–7 at concentrations of 10 mgmL−1
in hexane were analyzed by SFC-FID/UV. The injection volume
was 1 μL. CO2 temperature, pressure and flow were set at 60 °C,
138 bar and 1 mLmin−1 , respectively. All other settings were as
indicated in Section 2.1. To allow comparison of the quantitative
data obtained using SFC-FID/UV with those of the standard methods, also off-line SPE-GC-FID analyses were performed according
to the procedure described previously [6]. The GC-FID analysis of
the isolated MOSH and MOAH fractions was carried out using an
Agilent 6890 N instrument equipped with a Focus-PAL autosampler (GL Sciences, Eindhoven, the Netherlands). The sample (1 μL)
was injected in the splitless mode (2 min splitless time) at 350 °C
using a liner packed with glass wool. The capillary column was a
15 m × 0.32 mm × 0.1 μm DB5-HT column (Agilent) and was operated at a constant flow of 2 mLmin−1 using helium as carrier
gas. The temperature program ran from 60 °C (3 min) to 350 °C

To understand the interaction of mineral oil constituents with
the stationary phase under SFC conditions, mixtures of standard
compounds were injected first. Fig. 1 shows the SFC-FID chromatograms obtained for the separation of the aliphatic and aromatic standards on the bare silica and the silver-loaded silica
columns using pure carbon dioxide as the mobile phase. All compounds elute in less than 20 min from both columns. The alkanes C11 and C13 coelute with the solvent peak (hexane) for the
bare silica column (Fig. 1a), indicating the absence of retention
for these compounds. Unfortunately, neither the bare silica column nor the silver-loaded silica give a full separation of the compounds based on aromaticity. Unexpectedly, 5-α -cholestane (Cho),
an aliphatic compound that was expected to show little interaction
with the stationary phase, coelutes with the aromatics. Moreover,
for both stationary phases, 1,4-di-dodecylbenzene (D12B), a monoaromatic compound, elutes after 1-methylnaphthalene (1MN) and
biphenyl (BP), two di-aromatic species. Aliphatic/aromatic selectivity was better for the silver-loaded silica column, where Cho eluted
slightly before the di-aromatic compounds. The stronger retention
of the aromatics on the silver loaded phase is most likely due to
a strong interaction of the aromatic π electrons with the vacant

orbitals of the silver ions [21,22]. Nevertheless, based on the results of this experiment it is clear that retention and selectivity in
SFC cannot be explained by the interaction of the analytes with the
stationary phase only.
To better understand the influence of the mobile phase on
MOSH/MOAH retention and selectivity in SFC, the same standard
compounds were also analyzed in NPLC using hexane as the
mobile phase. Fig. 2 shows the chromatograms obtained for the
NPLC-UV/RID and the SFC-FID separation of the aliphatic and
aromatic standards on the silver-loaded silica column. In contrast
to the situation in SFC, there is no separation of the individual
compounds in NPLC. In the latter mode the analytes are clearly
separated according to their aromaticity, with little effect of the
size of the molecule or the presence of side chains [10]. All
aliphatic compounds elute together at the beginning of the chromatogram followed by the mono-, di- and tri-aromatic compounds
in the order of increasing aromaticity. This confirms that retention
and selectivity in SFC not only depend on the interaction with the
stationary phase but are also influenced by the solubility of the
analytes in the CO2 .
Retention and selectivity in packed column SFC have been
extensively studied in the past [23]. In these studies three main


4

A.R. García-Cicourel, B. van de Velde and G. Roskam et al. / Journal of Chromatography A 1614 (2020) 460713

Fig. 1. SFC-FID separation of aliphatic (black) and aromatic (red) standards on (a) bare silica and (b) silver-loaded silica columns. Standards: C11 (1), C13 (2), Cycy (3), Cho
(4), TBB (5), 1MN (6), BP (7), D12B (8) and Ant (9). CO2 at 60 °C, 110 bar and 1 mLmin−1 was used as mobile phase. (For interpretation of the references to color in this
figure legend, the reader is referred to the web version of this article.)

effects are identified: (i) solubility in the mobile phase, (ii) interaction with the stationary phase-mobile phase complex (the
mobile phase, CO2 , is absorbed on the stationary phase forming
a mono- or multilayer [24]) and (iii) displacement of the analytes
from the stationary phase by the mobile phase molecules. Since
each of these mechanisms depends on temperature and pressure
in its own way, prediction of retention in SFC is difficult at best.
Although the above mentioned three mechanisms-model is a
simplification that partly violates the principle that in chromatography mobile and stationary phase effects are inseparable, it is
helpful in understanding the current observations. Apparently, for
the system studied here, at the pressure and temperature applied,
supercritical CO2 is a weaker solvent than hexane. Moreover, the
interaction of the analytes is not only with the stationary phase
but with the stationary phase covered with CO2 . As a result of
these effects, aliphatic compounds can be retained even though
they are expected to have limited interaction with the stationary
phase. Thus, retention of the compounds in SFC is related to their
aromaticity as this determines the interaction with the stationary

phase, but also by their molecular weight which affects their solubility in the mobile phase. As a result, small aliphatic compounds
will elute rapidly, whereas larger aliphatic species will elute
slower due to their low solubility. They can then coelute with the
smaller mono-aromatic compounds. In the same way, large monoaromatic compounds can elute after smaller di- or poly-aromatic
compounds as is observed for 1,4-di-dodecylbenzene (D12B).
However, if two molecules with a similar solubility, or similar
molecular weight, have different aromaticity, they will be separated since their interaction with the stationary phase is different.
Whether a good MOSH/MOAH resolution is obtained depends on
the difference in retention between the ‘average MOSH’ and the
‘average MOAH’ molecule on the one hand, and the width of the
MOSH and MOAH bands on the other. The width of these bands is
determined not just by chromatographic band broadening mechanisms, but more importantly by retention differences between

the various molecules within a group, i.e. the polydispersity of the
MOSH and MOAH fractions in terms of size and structure of the
molecules.


A.R. García-Cicourel, B. van de Velde and G. Roskam et al. / Journal of Chromatography A 1614 (2020) 460713

5

Fig. 2. Chromatograms of the separation of the aliphatic (black) and aromatic (red) standard compounds on a silver-loaded silica column by (a) NPLC-UV/RID and (b) SFCFID. Standards: C11 (1), C13 (2), Cycy (3), Cho (4), TBB (5), D12B (6), 1MN (7), BP (8) and Ant (9). Hexane was used as mobile phase in the NPLC-UV/RID separation. CO2 at
60 °C, 110 bar and 1 mLmin−1 was used as mobile phase in the SFC-FID separation. (For interpretation of the references to color in this figure legend, the reader is referred
to the web version of this article.)

Fig. 3 shows the SFC-FID/UV MOSH/MOAH separation of mineral oil 1 on the silver-loaded silica column. Three peaks are observed in the chromatogram. The first peak is the solvent, hexane,
followed by the MOSH and the MOAH bands. Separation of the
MOSH and MOAH fractions is confirmed by the UV chromatogram,
which evidently only shows the aromatic compounds. The resolution obtained is insufficient for accurate MOSH and MOAH
quantification. Further optimization of temperature and pressure is
needed to improve MOSH/MOAH resolution.
3.2. Optimization of the SFC-FID/UV method for mineral oil analysis
As previously described, retention in SFC is affected by three
mechanisms, the two most important ones being the interaction of
the analyte with the CO2 -covered stationary phase and the solubility of the analytes in the CO2 mobile phase. For this latter mechanism the density of the CO2 , i.e. whether it is more liquid-like or
more gas-like, is very important. High pressures and low temper-

atures give the CO2 a more liquid-like nature improving the solubility of the analytes and reducing retention and band broadening
due to the polydispersity of the MOSH and MOAH fractions. On the
contrary, low pressures and high temperatures decrease the analyte solubility increasing their retention and peak width [7]. Moreover, like in all forms of chromatography the mobile phase flow
rate will also influence analyte retention, peak width and resolution of the analytes. Thus, optimization of the method should consider these three parameters and identify conditions where maximum resolution is obtained.
To identify the optimal chromatographic conditions for the

MOSH/MOAH separation, a two-level factorial experimental design
optimization was performed with temperature, pressure and flow
as factors. The high and low levels chosen for the three factors
are shown in Table 2. The response selected for optimization was
MOSH/MOAH resolution. Resolution values obtained for the different experiments are given in Table 3. For the MOSH fraction,
peak width was obtained from the FID chromatogram. MOAH peak


6

A.R. García-Cicourel, B. van de Velde and G. Roskam et al. / Journal of Chromatography A 1614 (2020) 460713

Fig. 3. SFC-FID/UV separation of mineral oil 1 on a silver-loaded silica column. First peak corresponds to the solvent, hexane, followed by MOSH and MOAH. CO2 at 60 °C,
110 bar and 1 mLmin−1 was used as mobile phase. MOSH/MOAH resolution is 0.9057. The UV chromatogram was plotted using the MaxPlot mode.
Table. 3
Resolution values obtained for every experiment for the optimization of the SFC-FID/UV
method. Every combination was evaluated in triplicate.
Experiment

Temperature (°C)

Pressure (bar)

Flow (mLmin−1 )

Resolution

1

50

110

1.0

2

50

110

2.0

3

50

138

1.0

4

50

138

2.0

5

60

110

1.0

6

60

110

2.0

7

60

138

1.0

8

60

138

2.0

0.9973
1.0117
0.9883
1.0303
1.0293
1.0151
1.0310
1.0756
1.0324
0.9315
0.9040
0.9447
0.9232
0.9258
0.9464
1.0110
1.0133
1.0245
1.0529
1.0783
1.0595
1.0488
1.0590
1.0603

width was obtained from the UV chromatogram where it could
be more accurately determined due to the absence of the MOSH
signal. Prior to starting the evaluation, the FID and UV signals for
the MOAH band were compared to see if possibly extra-column

band-broadening occurred in the T-piece or the transfer capillary
to the FID. This was done using a pure MOAH fraction isolated
from an oil using SPE. No extra band broadening was observed,
although a slight shift in the retention times was seen, with the
FID signal being slightly later than that of the UV detector. Statistical analysis using the Analysis of Variance (ANOVA) method
showed that the flow rate applied did not have a significant influence on MOSH/MOAH resolution, whereas the effects of temperature and pressure were significant. Details of the statistical analysis, the optimization of the response and the equation describing
the influence of each parameter on MOSH/MOAH resolution are

shown in the supplementary material (SM-1). Fig. 4 shows resolution as a function of pressure and temperature at a flow rate
of 1 mLmin−1 . Resolution is seen to increase at higher pressures
and temperatures. Under these conditions the aliphatic compounds
will have less interaction with the stationary phase and elute in
a narrower band improving the resolution between the two mineral oil fractions. A maximum resolution of 1.08 is obtained when
working at 138 bar and 60 °C. This resolution value is sufficient
if the MOSH and MOAH bands are of similar height, but is too
low for an accurate MOSH/MOAH determination of samples with
high MOSH/MOAH ratios, such as encountered in highly purified
white oils. Given the shape of the resolution function observed in
Fig. 4, it was logical to test even higher pressures and temperatures. Unfortunately, at those conditions the MOSH band started to
overlap with the solvent peak leading to an incorrect quantifica-


A.R. García-Cicourel, B. van de Velde and G. Roskam et al. / Journal of Chromatography A 1614 (2020) 460713

7

Fig. 4. Response surface of the resolution as a function of temperature and pressure. Flow is set at 1 mLmin−1 . The maximum resolution, 1.08, is obtained at 60 °C and
138 bar.

Fig. 5. Deconvolution process of the MOSH and MOAH peaks in the FID chromatogram using the UV signal, (a) normalized SFC-FID and SFC-UV chromatograms, (b) SFC-FID

chromatogram after subtraction of the normalized UV signal, MOAH peak is removed, and (c) deconvoluted SFC-FID chromatogram of mineral oil 1. MOSH/MOAH separation
was carried out on a silver-loaded silica column and CO2 at 60 °C, 138 bar and 1 mLmin−1 was used as mobile phase. The UV chromatogram was plotted using the MaxPlot
mode.

tion of the MOSH fraction. Hence, the conditions chosen for the
SFC MOSH/MOAH separation were 138 bar, 60 °C and 1 mLmin−1 .
In case of determinations where only the MOAH levels are needed,
working at higher temperatures and pressures is possible. Pressure
gradients were briefly studied as a means to improve MOSH/MOAH
separation while avoiding coelution of the MOSH band with the
solvent peak. However, no large improvement was observed and
the option of pressure programming was not further studied. This
also because quantification would be complicated as the split ratio
between UV and FID will vary during pressure programming. Additionally, the FID response might be affected by the increasing CO2
flow.
To solve the partial overlap issue, the possibility to deconvolute the MOSH and MOAH peaks using the combined FID and UV
chromatograms was studied. The principle of this approach is described in Fig. 5. Fig. 5a shows the overlay of the SFC-FID and SFCUV chromatograms of mineral oil 1 at the optimal temperature

and pressure conditions after normalization of the two signals. For
the UV signal, the MaxPlot (190–400 nm) trace was used as this
showed the best similarity to the FID signal. Using the normalized
MOAH peak from the SFC-UV chromatogram, the MOAH contribution in the SFC-FID chromatogram can be removed by subtraction
(see Fig. 5b) leaving just the signal of the MOSH. Normalization of
the response was done using the following equation:

FIDnew (t ) =

UV(t )( FID )
UVmax

+ FIDmin

where FIDnew (t) is the MOAH FID signal at time t, UV(t) represents the UV signal at time t,
FID is the difference between the
FID response at the time of the maximum UV response and the
baseline FID response along the chromatogram, UVmax is the maximum UV response along the UV chromatogram and FIDmin is the
minimum, i.e. baseline, FID response. Fig. 5c shows the deconvoluted SFC-FID chromatogram of mineral oil 1 and its comparison


8

A.R. García-Cicourel, B. van de Velde and G. Roskam et al. / Journal of Chromatography A 1614 (2020) 460713

Fig. 6. Comparison of the deconvoluted SFC-FID chromatogram of mineral oil 1 and the SFC-FID chromatograms of the individual MOSH and MOAH fractions of the same
mineral oil after separation by SPE.

with the original chromatogram. Both the individual MOSH and
MOAH signals match perfectly with the original chromatogram, but
the two signals are now separated. With the deconvoluted chromatograms better integration and more accurate quantification of
the two fractions is possible. The comparison of quantitative data
obtained using this deconvolution strategy with those obtained by
the standard SPE-GC-FID method will be discussed in the next sections. The FID MOSH/MOAH deconvolution was further corroborated by injecting isolated MOSH and MOAH fractions, obtained
using SPE, individually into the SFC system. Fig. 6 shows the comparison of the deconvoluted FID chromatogram and the SFC-FID
chromatograms of the individual MOSH and MOAH fractions. Although the deconvoluted bands and the SPE fractions do not match
perfectly due to small differences in concentrations and retention
time shifts, the figure clearly shows that the MOSH/MOAH deconvolution gives very similar peaks to those obtained for the individual fractions. This confirms that the UV signal of the SFC analysis
can be used for deconvolution of the FID signal in order to allow a
more accurate integration of the MOSH and MOAH peaks and obtain more accurate quantification of the two fractions.
3.3. Validation of the SFC-FID/UV method
In order to evaluate the performance of the proposed SFCFID/UV method, the main performance characteristics, i.e. linearity,

repeatability, limit of detection (LOD) and limit of quantification
(LOQ) were assessed following the Eurachem guide for method validation [20]. This evaluation was done using mineral oil 8, an oil
free of aromatics, hence only the MOSH fraction was considered
to assess the performance characteristics. Trueness of the method
was assessed by comparing the results of the new method with
that of off-line SPE fractionation and GC-FID quantification (see
Section 3.4). Since the FID responds unselectively to hydrocarbons
and MOSH and MOAH have very similar response factors, the performance characteristics determined using MOSH are also valid for
MOAH [1]. The results obtained in the validation experiments and
the comparison with the performance characteristics of the SPEGC-FID method are presented in Table 4 and the calculations and
graphs are shown in the supplementary material (SM-2). An ex-

Table. 4
Performance characteristics for the validation of the
SFC-FID/UV and SPE-GC-FID methods.

Linearity
LOD(mgmL−1 )
%MOAH
LOQ(mgmL−1 )
%MOAH
Repeatability (CV%)

SFC-FID/UV

SPE-GC-FID

0.9995
0.0360
0.36

0.1200
1.20
<4.84

0.9641
0.0191
0.19
0.0637
0.63
<6.77

cellent correlation (R2 > 0.9995) between the FID response and
the mineral oil concentration was found in the concentration range
2.5–100 mgmL−1 . Repeatability of the method, evaluated at three
concentration levels, resulted in RSD values below 5%. Unlike in
classical two-step MOSH/MOAH methods, variation of the results is
only related to the peak integration since no other sample handling
steps are present. LOD and LOQ, evaluated statistically as the RSD
of the injection of three different samples at low concentration,
were 0.0360 and 0.1200 mgmL−1 , respectively. Thus, when working
with a mineral oil concentration of 10 mgmL−1 and considering a
mineral oil with similar physicochemical characteristics as mineral
oil 8, the LOD and LOQ values for the MOAH fraction calculated
as the percentage of MOAH in the sample are 0.36 and 1.2%, respectively. This is sufficiently sensitive for all but the highest purity white oils. It should be noted here that the LOQ cannot be
improved by injecting a larger sample volume or a more concentrated sample. This because LOQ/LOD values are not determined by
the detector sensitivity, but rather by the MOSH/MOAH separation,
with this separation being compromised when the MOSH/MOAH
ratio of the sample gets too high. To confirm the trueness of the
calculated LOQ value for MOAH, a solution at 10 mgmL−1 of a mixture of mineral oil 1 (24% aromatics) and mineral oil 8 (virtually
aromatic free), with a final MOAH concentration of 1.2%, was injected in the SFC-FID/UV system. After deconvolution of the FID

chromatogram and calculation of the aromatics content based on
the MOSH/MOAH area ratios, an aromatics level of 2% was obtained, so well above the calculated value of 1.2%. Fig. 7 shows


A.R. García-Cicourel, B. van de Velde and G. Roskam et al. / Journal of Chromatography A 1614 (2020) 460713

9

Fig. 7. SFC-UV chromatograms of mineral oil 8 and a mixture of mineral oil 1 and 8 with a final MOAH concentration of 1.2% with respect to the total mineral oil. The
double MOAH peak in the mixture could be the result of a high naphthene concentration in mineral oil 8. CO2 at 60 °C, 138 bar and 1 mLmin−1 was used as mobile phase.
UV chromatograms were plotted using the MaxPlot mode.

Fig. 8. SFC-UV chromatogram of Cycy (1) and Cho (2) at a concentration of 500 μgmL−1 . CO2 at 60 °C and 138 bar was used as mobile phase. The UV chromatogram was
plotted using the MaxPlot mode.

the SFC-UV chromatograms of mineral oil 8 and of the mixture of
mineral oil 1 and 8 previously described. Unexpectedly, the SFC-UV
MaxPlot chromatogram of the aromatics-free mineral oil 8 shows
a peak around 3 min which matches with the first peak observed
in the chromatogram of the mixture. This elution position suggests the signal is not caused by aromatic species but results from
aliphatics. Knowing that in the preparation of aromatics-free mineral oils hydrogenation is applied and that this will result in the
formation of naphthenes, i.e. cyclic aliphatics, the SFC-UV response
of the two naphthenes in the set of standard compounds, Cycy
and Cho, was studied. Fig. 8 shows the resulting SFC-UV MaxPlot
chromatogram. Even though no response of the two naphthenes

was expected, two clear peaks are observed in the chromatogram.
Comparison of the absorbance spectrum of the standards and the
peak seen for mineral oil 8 showed very similar spectra that are
clearly different from those obtained for aromatic species. As a

consequence, the UV MaxPlot signal of the mixture of mineral oil
1 and 8 is formed by two humps, one for the naphthenes of mineral oil 8 eluting in the MOSH window, and one for the MOAH.
Plotting the UV chromatogram using the aromatics wavelength of
254 nm rather than the MaxPlot mode confirms that the peak
around 3 min are not aromatic species. Apparently the naphthenes
show a strong absorbance in the MaxPlot region of 190–400 nm.
Deconvolution of the FID signal considering only the MOAH peak


10

A.R. García-Cicourel, B. van de Velde and G. Roskam et al. / Journal of Chromatography A 1614 (2020) 460713

Fig. 9. SFC-FID chromatograms of mineral oil 1 (black), mineral oil 5 (red) and mineral oil 6 (blue). The three mineral oils have different viscosities of 16.53, 56.41 and
91.21 mm2 s−1 , respectively. MOAH peak is visible in mineral oil 1 and 5 but not in mineral oil 6. (For interpretation of the references to color in this figure legend, the
reader is referred to the web version of this article.)

Fig. 10. Comparison of the aromatic percentage of five different mineral oil samples using the SFC-FID/UV and the off-line SPE-GC-FID standard method.

resulted in an aromatic content of 1.4% in the mixture of mineral
oils, a value close to the expected level, corroborating the trueness
of the method.
3.4. Analysis of real samples
Following the optimization and validation of the SFC-FID/UV
method, seven different mineral oils from different stages in the
white oil purification process were analyzed. The mineral oils were
chosen with the intention to cover a wide range of samples in
terms of viscosities and molecular weights. Fig. 9 shows the SFC-

FID chromatograms of three different mineral oils (mineral oil 1,

5 and 6). Mineral oil 6 has the broadest MOSH hump followed
by mineral oil 5 and 1. The higher the viscosity or the molecular
weight of the sample, the broader the MOSH and MOAH humps. A
broader MOSH hump results in more severe coelution of the MOSH
and MOAH humps complicating the deconvolution process. The
MOAH hump is visible for mineral oils 1 and 5, but not for mineral
oil 6. Consequently, the SFC-FID/UV method for MOSH/MOAH analysis is limited to lighter mineral oils with a viscosity and molecular
weight below approximately 56 mm2 s−1 and 450 gmol−1 , respectively. Fig. 10 compares the MOAH composition of five different


A.R. García-Cicourel, B. van de Velde and G. Roskam et al. / Journal of Chromatography A 1614 (2020) 460713

mineral oils that meet the viscosity/molecular weight requirements
obtained using the new SFC-FID/UV method with data obtained by
the classical off-line SPE-GC-FID method. Statistical analysis using
ANOVA showed that the results are not statistically significantly
different between the two methods. Moreover, the results obtained
with the newly proposed method show lower RSD values, probably
as a result of the fact that the method only consists of one simple
dilution step. In terms of ease of use and options for automation
the new method also compares favorably to the existing methods.
4. Conclusions
A simple and fast SFC-FID/UV method for the determination of
the MOSH and MOAH content of purified mineral oil samples was
developed. The optimal CO2 temperature, pressure and flow rate
were determined using a two-level factorial experimental design
resulting in optimum settings of 60 °C, 138 bar and 1 mLmin−1 ,
respectively. Unlike the two-step conventional methods for mineral oil analysis, MOSH and MOAH content could be determined in
a single run. For samples for which partial MOSH/MOAH coelution
occurred, a deconvolution method that allowed accurate determination of MOSH and MOAH peak areas was developed. This was

possible thanks to the parallel detection capabilities of the new
method. Analysis of the individual MOSH/MOAH fractions using the
proposed method confirmed that the results obtained after peak
deconvolution were accurate. Unfortunately, viscosity, or molecular
weight, of the mineral oil affects the MOSH/MOAH separation resulting in coelution of both peaks. The SFC-FID/UV method is suitable for the analysis of mineral oils with viscosities and molecular weights below approximately 56 mm2 s−1 and 450 gmol−1 .
The proposed method showed to be linear (R2 > 0.9995) in the
range of concentrations selected (2.5–100 mgmL−1 ) and has a better repeatability than the standard methods (<5%). MOAH detection limits were better than 0.36%, which makes the method sufficiently sensitive for analysis of all but the highly purified mineral oils. Analysis of different samples gave identical results as obtained with the standard methods with the advantage of a better
repeatability, simple operation and a shorter analysis time of less
than 10 min. This new method could potentially also be used for
the analysis of mineral oil contaminations in consumer products
such as foods. However, in this case additional sample clean-up
and preconcentration steps are needed for reducing matrix interferences from e.g. triglycerides and olefins and for improving the
detection limits.
Acknowledgments
The authors thank the Consejo Nacional de Ciencia y Tecnología
de México (CONACyT) for the scholarship awarded to A.R. GarcíaCicourel (No. 410740). The authors acknowledge Joshka Verduin for
her assistance with the SPE-GC-FID experiments.
This research did not receive any further specific grant from
funding agencies in the public, commercial, or not-for-profit sectors.
Supplementary materials
Supplementary material associated with this article can be
found, in the online version, at doi:10.1016/j.chroma.2019.460713.
References
[1] M. Biedermann, K. Fiselier, K. Grob, Aromatic hydrocarbons of mineral oil origin in foods: method for determining the total concentration and first results,
J. Agric. Food Chem. 57 (2009) 8711–8721, doi:10.1021/jf901375e.

11

[2] S. Moret, L. Barp, G. Purcaro, L.S. Conte, Rapid and sensitive solid phase
extraction-large volume injection-gas chromatography for the analysis of mineral oil saturated and aromatic hydrocarbons in cardboard and dried foods, J.

Chromatogr. A 1243 (2012) 1–5, doi:10.1016/j.chroma.2012.04.040.
[3] M. Biedermann, C. Munoz, K. Grob, Update of on-line coupled liquid chromatography – gas chromatography for the analysis of mineral oil hydrocarbons
in foods and cosmetics, J. Chromatogr. A 1521 (2017) 140–149, doi:10.1016/j.
chroma.2017.09.028.
[4] S. Weber, K. Schrag, G. Mildau, T. Kuballa, S.G. Walch, D.W. Lachenmeier, Analytical methods for the determination of mineral oil saturated hydrocarbons
(MOSH) and mineral oil aromatic hydrocarbons (MOAH)— a short review, Anal.
Chem. Insigths 13 (2018) 1–16, doi:10.1177/1177390118777757.
[5] B. Chuberre, E. Araviiskaia, T. Bieber, A. Barbaud, Mineral oils and waxes in
cosmetics : an overview mainly based on the current European regulations
and the safety profile of these compounds, JEADV 33 (2019) 5–14, doi:10.1111/
jdv.15946.
[6] A.R. García-Cicourel, H.-.G. Janssen, Direct analysis of aromatic hydrocarbons
in purified mineral oils for foods and cosmetics applications using gas chromatography with vacuum ultraviolet detection, J. Chromatogr. A (2019) 1–8,
doi:10.1016/j.chroma.2019.01.015.
[7] R. M’Hamdi, D. Thiébaut, M. Caude, Packed column SFC of gas oils part I :
hydrocarbon group separation using pure carbon dioxide, J. High Resol. Chromatgr. 20 (1997) 545–554.
[8] M. Biedermann, K. Grob, On-line coupled high performance liquid
chromatography-gas chromatography for the analysis of contamination by
mineral oil. Part 1: method of analysis, J. Chromatogr. A 1255 (2012) 56–75,
doi:10.1016/j.chroma.2012.05.095.
[9] M. Biedermann, K. Grob, On-line coupled high performance liquid
chromatography-gas chromatography for the analysis of contamination
by mineral oil. Part 2: migration from paperboard into dry foods: interpretation of chromatograms, J. Chromatogr. A 1255 (2012) 76–99,
doi:10.1016/j.chroma.2012.05.096.
[10] A.R. García-Cicourel, B. van de Velde, J. Verduin, H.-.G. Janssen, Comprehensive
off-line silver phase liquid chromatography × gas chromatography with flame
ionization and vacuum ultraviolet detection for the detailed characterization of
mineral oil aromatic hydrocarbons, J. Chromatogr. A 1607 (2019), doi:10.1016/
j.chroma.2019.460391.
[11] P.Q. Tranchida, M. Zoccali, G. Purcaro, S. Moret, L. Conte, M. Beccaria, P. Dugo,

L. Mondello, A rapid multidimensional liquid-gas chromatography method for
the analysis of mineral oil saturated hydrocarbons in vegetable oils, J. Chromatogr. A 1218 (2011) 7476–7480, doi:10.1016/j.chroma.2011.06.089.
[12] L. Barp, G. Purcaro, S. Moret, L.S. Conte, A high-sample-throughput LC-GC
method for mineral oil determination, J. Sep. Sci. 36 (2013) 3135–3139, doi:10.
10 02/jssc.20130 0114.
[13] M. Zoccali, L. Barp, M. Beccaria, D. Sciarrone, G. Purcaro, L. Mondello, Improvement of mineral oil saturated and aromatic hydrocarbons determination in edible oil by liquid-liquid-gas chromatography with dual detection, J. Sep. Sci. 39
(2016) 623–631, doi:10.1002/jssc.201501247.
[14] M. Zoccali, P.Q. Tranchida, L. Mondello, On-line liquid chromatographycomprehensive two dimensional gas chromatography with dual detection for
the analysis of mineral oil and synthetic hydrocarbons in cosmetic lip care
products, Anal. Chim. Acta 1048 (2019) 221–226, doi:10.1016/j.aca.2018.10.069.
[15] A. Tarafder, Metamorphosis of supercritical fluid chromatography to SFC: an
overview, Trends Anal. Chem. 81 (2016) 3–10, doi:10.1016/j.trac.2016.01.002.
[16] A.S.T.M. D5186-19, Standard test method for determination of the aromatic
content and polynuclear aromatic content of diesel fuels by supercritical fluid
chromatography, ASTM Int. West Conshohocken, 2019.
[17] E. Lesellier, C. West, The many faces of packed column supercritical fluid chromatography – a critical review, J. Chromatogr. A 1382 (2015) 2–46, doi:10.1016/
j.chroma.2014.12.083.
[18] P. Sandra, A. Medvedovici, Y. Zhao, F. David, Characterization of triglycerides in
vegetable oils by silver-ion packed-column supercritical fluid chromatography
coupled to mass spectroscopy with atmospheric pressure chemical ionization
and coordination ion spray, J. Chromatogr. A 974 (2002) 231–241, doi:10.1016/
S0021- 9673(02)01311- 0.
[19] E.J. Guthrie, H.E. Schwartz, Integral pressure restrictor for capillary SFC, J. Chromatogr. Sci. 24 (1986) 236–241, doi:10.1093/chromsci/24.6.236.
[20] Eurachem, Eurachem Guide: The Fitness for Purpose of Analytical Methods – A Laboratory Guide to Method Validation and Related Topics, 2014
978-91-87461-59-0.
[21] J.C. Aponte, J.T. Dillon, R. Tarozo, Y. Huang, Separation of unsaturated organic
compounds using silver-thiolate chromatographic material, J. Chromatogr. A
1240 (2012) 83–89, doi:10.1016/j.chroma.2012.03.082.
[22] W. Moreda, A. Cert, New method of stationary phase preparation for silver ion
column chromatography. Application to the isolation of hydrocarbon fractions

in vegetable oils, J. Chromatogr. A 823 (1998) 291–297.
[23] H.G. Janssen, P.J. Schoenmakers, C.A. Cramers, Mobile and stationary phases for
SFC: effects of using modifiers, Mikrochim. Acta 104 (1991) 337–351, doi:10.
1007/BF01245520.
[24] J.R. Strubinger, H. Song, J.F. Parcher, High-pressure phase distribution isotherms
for supercritical fluid chromatographic systems. 1. Pure carbon dioxide, Anal.
Chem. 63 (1991) 98–103, doi:10.1021/ac0 0 0 02a0 03.