Chromatographia

https://doi.org/10.1007/s10337-018-3681-3

ORIGINAL

Direct Analysis of Synthetic Phenolic Antioxidants, and Fatty Acid

Methyl Ester Stability in Biodiesel by Liquid Chromatography
and High-Resolution Mass Spectrometry
Marcella Casagrande1 · Chadin Kulsing2,4 · Jalal T. Althakafy2,3 · Clarisse M. S. Piatnicki1 · Philip J. Marriott2

Received: 28 June 2018 / Revised: 7 December 2018 / Accepted: 11 December 2018

© Springer-Verlag GmbH Germany, part of Springer Nature 2018

Abstract
Methods to identify and quantify synthetic phenolic antioxidants, 3-tert-butyl-4-hydroxyanisole (BHA), butylated hydroxy-
toluene (BHT), tert-butyl-hydroquinone (TBHQ) and propyl gallate (PG), in biodiesel samples by using reversed-phase
liquid chromatography (LC) were developed. Using a C18 phase with LC and UV detection showed co-elution between BHT
and fatty acid methyl esters (FAME) in the biodiesel sample, whereas an alkyl phenyl modified stationary phase resulted in
good separation of all antioxidants from the fatty acid matrix, and allowed more accurate quantification of antioxidants in
biodiesel samples. The latter column was applied for further study. Calibration curves for the four antioxidants were con-
structed, and the limit of detection estimated. Good calibration linearity was observed over the investigated concentration
range of 10–80 ppm, with correlation coefficients (R2) ranging from 0.9986 to 0.9995 for all antioxidants. LOD values of
0.010, 0.015, 0.0125 and 0.030 ppm, and recoveries of 70 ± 2, 85 ± 2, 103 ± 2 and 92 ± 4% for PG, TBHQ, BHA and BHT at
injected concentrations of 35 ppm were established, respectively. The method was applied for quantification of antioxidants
in biodiesel without addition of spiked antioxidants, then for biodiesel spiked with the four antioxidants, and a commercial
source of biodiesel with BHT addition. Identification of FAME in the biodiesel samples was performed by using an instru-
ment capable of ultra-high performance LC hyphenated with an electrospray Orbitrap mass spectrometer (UHPLC–ESI-
OrbitrapMS). The stability of antioxidants and FAME in different samples was then investigated. Total FAME C18 content
decreased to 52 ± 4% w/w after 1 week, and 29 ± 6% w/w after 8 weeks in the test sample without antioxidants; FAME content
and antioxidant composition were stable in the samples with antioxidants added, even after 8 weeks exposure to sunlight.

Keywords Antioxidants · Biodiesel · Liquid chromatography · Phenyl stationary phase · Q-Exactive · Quadrupole Orbitrap
mass spectrometer

Abbreviations
BHA 3-tert-Butyl-4-hydroxyanisole
BHT Butylated hydroxytoluene
Published in Chromatographia’s 50th Anniversary Commemorative ESI Electrospray ionisation
Issue.
LC Liquid chromatography
* Philip J. Marriott MS Mass spectrometer
[email protected] PG Propyl gallate
1
SPA Synthetic phenolic antioxidants
Instituto de Química, Universidade Federal do Rio Grande TBHQ  tert-Butyl-hydroquinone
do Sul, Av. Bento Gonçalves, 9500, CP 15003, Porto Alegre,
RS 91501‑970, Brazil HPLC High-performance liquid chromatography
2 UV Ultra-violet
Australian Centre for Research on Separation Science,
School of Chemistry, Monash University, Wellington Road,
Clayton, Melbourne, VIC 3800, Australia
3
Department of Chemistry, Faculty of Applied Sciences,
Umm Al-Qura University, Mecca 21955, Saudi Arabia
4
Department of Chemistry, Faculty of Science, Chulalongkorn
University, Bangkok 10330, Thailand

13
Vol.:(0123456789)
 M. Casagrande et al.

Introduction have been conducted in order to generate faster and more

efficient methods for monitoring oxidation processes regard-
Biodiesel is an alternative renewable fuel comprising long- ing storage time and sample aging [9–13].
chain mono-alkyl esters (fatty acid esters); it is commonly A variety of analytical methods for biodiesel analysis
obtained by reaction between short-chain alcohols (such as have been reported, including electrochemical techniques [7,
methanol or ethanol) and vegetable oil/animal fat in presence 14–16] and capillary electrophoresis [17–19], however gas
of an acid or basic catalyst [1, 2]. This fuel presents some chromatography (GC) [20–24] and high-performance liquid
advantages compared to fossil fuels, e.g. lower toxicity, chromatography (HPLC) are the most widely used methods
higher cetane number and flash point, higher lubricity and [6, 25–30]. Analysis protocols for antioxidant composition
absence of sulfur and aromatic compounds. Furthermore, of such samples often commence with sample extraction,
it can be used in diesel engines without the need of major often with large quantities of solvents, extended time, and an
modification [3]. Despite these advantages, biodiesel can be increased possibility of sample loss during extraction and/
more easily oxidised or autoxidised under long-term storage or derivatisation procedures. Fewer procedures report direct
than conventional diesel, with unsaturated more prone to analysis of the unprocessed sample.
oxidation than saturated esters [4], forming oxidation prod- Methods based on GC offer good repeatability, high
ucts, deposits or polymers that can damage vehicle engines resolution and high peak capacity, and on-line coupling to
and lead to their malfunction. Exposure to light, heat, MS permits reliable library searching, which can confirm
humidity, oxygen and metals can accelerate this oxidation detection of antioxidants in both biodiesel and edible oils
process; therefore, it is necessary to improve biodiesel sta- without complex sample preparation methods [20, 21, 31,
bility as well as to monitor degradation processes [1, 2, 5]. 32]. Application of comprehensive two-dimensional gas
In order to increase shelf life and delay biodiesel oxidation, chromatography (GC × GC) further increases peak capacity,
synthetic phenolic antioxidants (SPAs) are often added to the leading to enhanced resolution of antioxidants from the bio-
biofuel, e.g. 3-tert-butyl-4-hydroxyanisole (BHA), butylated diesel matrix [33–35], as well as improved compound iden-
hydroxytoluene (BHT), tert-butyl-hydroquinone (TBHQ) tification by increased quality of mass spectrometry match-
and 3,4,5-trihydroxybenzoic acid propyl ester (propyl gal- ing, and potential for second dimension retention indices
late, PG) [6, 7]. SPAs interrupt chain oxidation reactions by [36]. Derivatisation is often performed in order to improve
donating a proton, yielding stable free radicals which are peak quality (shape) and/or sensitivity in GC analysis [37,
unable to initiate or continue lipid oxidation. They must be 38]. Alternatively, HPLC can be performed for antioxidant
added to biodiesel prior to use or storage, since they are not analysis, often without requiring derivatisation [6]. Fatty
able to reverse the effects of oxidation processes which have acid methyl ester (FAME) components in biodiesels can be
already occurred [5, 8]. For conventional and alternatively adequately separated, e.g. under gradient elution in reversed-
derived middle distillate fuels, antioxidant ‘packages’ com- phase HPLC and FAME components may be identified by
prising single or multiple SPA are often added to improve MS detection [39].
the stability of biofuels such as those for aviation applica- In this study, analysis of antioxidants in biodiesel was
tions [8]. conducted by using HPLC–UV and HPLC–MS on the whole
Biodiesel stability depends on storage conditions, which biodiesel sample. Limits of detection of antioxidants were
determines the extent of oxidation. The susceptibility of bio- calculated for the UV method. Antioxidants were identified
diesel to oxidation, when compared to fossil fuels, depends by injecting authentic standards, and FAME identification
mainly on the content of unsaturated fatty acid esters that was performed by using ultra-high performance liquid chro-
comes from the feedstock which is used to produce the bio- matography (UHPLC) but operated in conventional HPLC
fuel. It also depends on the presence of natural antioxidants mode, hyphenated with a Q-Exactive Plus Orbitrap MS.
and storage conditions; if not properly stored, e.g. exposure Antioxidant and FAME composition in biodiesel samples,
to light, oxidation products such as aldehydes and alcohols with and without antioxidant addition, were quantified after
may affect properties of the biodiesel. Thermal stability sample exposure to sunlight for different periods. Photo-
towards oxidative degradation relates the effect of temper- oxidation (autoxidation is less rapid) and elevated tempera-
ature to formation of oxidation products, or reduction of ture should be the primary oxidative processes arising from
biodiesel components as a result of oxidation; higher tem- sunlight exposure [40].
peratures may enhance degradation rates. Storage stability
may be described as the ability to resist change under long-
term storage, maintaining the original physical and chemical
characteristics. The standard method for oxidative stability
analysis is the Rancimat method; nevertheless, many studies

13
Direct Analysis of Synthetic Phenolic Antioxidants, and Fatty Acid Methyl Ester Stability…

Materials and Methods (StableBond Analytical 4.6 × 250 mm, 5 µm; Agilent).
Although these columns have relatively large diameter par-
Chemicals and Biodiesel Samples ticle size, they were found to be suited to the sample analysis
reported, and so were used throughout this study. Samples
A 37-component FAME mixture (47885-U) was obtained were separated at 25 °C. 0.1% v/v acetic acid in water and
from Supelco (Sigma-Aldrich, St. Louis, MO). FAME with methanol were used as mobile phases A and B, respec-
different chain length were abbreviated as FAME CX:Y, tively. Gradient elution started at 50% v/v mobile phase B
with the total carbon number and number of double bonds for 17 min and then linearly increased to 100% v/v B. The
of X + 1 and Y, respectively. For example, FAME C18:0 mobile phase content was held at 100% v/v B for 13 min
and FAME C18:1 are methyl octadecanoate and methyl then decreased to 50% v/v B. The column was equilibrated
octadecenoate, respectively. GC-grade dichloromethane at 10% v/v B for 5 min resulting in the total time of 35 min
(DCM) was obtained from Merck Co. (Merck KGaA, prior to the next injection. Samples were diluted 20 times in
Darmstadt, Germany). Water was distilled and deionised methanol and injected (5 µL) at a flow rate of 0.6 mL/min for
using a Milli-Q system (Millipore, Bedford, MA). Metha- 35 min. Chosen wavelengths were 254, 260, 270, 280, 285,
nol ­(SupraSolv® grade) was obtained from Merck KGaA. 290, 295 and 300 nm for selected chromatographic displays
Acetic acid (100% v/v) and phenol (used as internal stand- (detection bandwidth of 2 nm; reference off).
ard) were purchased from Sigma-Aldrich and AnalaR (BDH
Laboratory Supplies), respectively. The original FAME HPLC–Q‑Exactive Orbitrap MS
mixture (10,000 mg L−1 total concentration) was diluted to
1000 mg L−1 with DCM. tert-Butyl-hydroquinone (TBHQ) Standards and samples were analysed by using a Dionex
and butylated hydroxytoluene (BHT) were purchased from analytical ultra-high performance liquid chromatography
Fluka; 3-tert-butyl-4-hydroxyanisole (BHA) and propyl gal- (UHPLC) instrument, hyphenated with a Q-Exactive Plus
late (PG) were purchased from Sigma-Aldrich. Biodiesel Orbitrap mass spectrometer (ThermoFisher Scientific,
samples (both pure and with BHT addition) were provided Scoresby, Australia) which was equipped with a heated elec-
by ARFuels Co. Ltd. (Barnawartha, Australia). According trospray ionisation source (HESI), binary pump, autosam-
to the company, the biodiesel sample has an added BHT pler and Quadrupole-Orbitrap. A Zorbax SB-Phenyl column
concentration ranging from 600 to 800 ppm. (Agilent; as above) was employed as the stationary phase.
To maintain the same retention and resolution obtained in
the above HPLC–UV study, the same column was used, and
UV Degradation of Biodiesel the high pressures characteristic of UHPLC operation were
not employed. The same temperature and mobile phase sys-
A total of 24 biodiesel samples (containing an initial con- tem as that used for HPLC–UV above were applied. The
centration of 600–800 ppm BHT, quantified in this study flow rate was 0.3 mL/min, and the injection volume was
to be 647 ± 10 ppm) were exposed to sunlight for different 15 µL. The HESI source was operated in both positive and
periods, ranging from 1 to 8 weeks. Three replicate sam- negative modes. The source conditions were sheath gas 35,
ples were taken for analysis each week. The same series of auxiliary gas 10, sweep gas 0, spray voltage 3.0 kV, capillary
experiments were also conducted on a corresponding set of T 320 °C and auxiliary gas heater T 300 °C. The MS was
24 pure biodiesel samples (no added SPA), and 24 spiked operated in both positive and negative full scan modes (from
with all 4 SPA; BHA, BHT, TBHQ and PG (700 ppm each). 50 to 400 m/z), applied with MS parameters: resolution of
All samples were characterised by HPLC–UV. Samples were 70,000 FWHM, AGC target of 1 × 106 and maximum injec-
diluted 20 times in methanol, giving a nominal concentra- tion time (IT) of 200 ms. The system was calibrated daily
tion of about 30–40 ppm in each antioxidant, prior to HPLC before analysis for both positive and negative modes.
analysis.
Data Processing
Chromatographic Analysis and Instrumentation
Agilent OpenLAB CDS (ChemStation Edition) for LC and
HPLC−UV LC/MS systems and Microsoft Excel 2007 was used for data
visualisation. Xcalibur 3.0.63 (ThermoFisher) software was
Samples were analysed by using an Agilent 1220 Infinity used to control the Orbitrap instrument and for data analy-
LC system (Agilent Technologies, Mulgrave, Australia) sis. TraceFinder 3.0 software (ThermoFisher) was used for
with a reversed-phase C18 (Alltima 4.6 × 250 mm, 5 µm identification, confirmation and quantification analysis with
particle diameter; Agilent) or Zorbax SB-Phenyl column Microsoft Excel 2007.

13
 M. Casagrande et al.

Fig. 1  a HPLC–UV analysis of four antioxidants (35 ppm each): (below) columns and b results obtained by using different wave-
PG (1), TBHQ (2), BHA (3) and BHT (4), Phenol (IS) in biodiesel lengths according to the same separation conditions used in a for the
matrix obtained at 280 nm by using alkylphenyl (above) and C18 alkylphenyl column

Results and Discussion

Selection of Column and Wavelength with HPLC–UV

In order to investigate effects of column selection and

wavelength, separation of biodiesel samples spiked with
PG, TBHQ, BHA and BHT was performed. The detec-
tion wavelength of 280 nm was selected here to present
an indication of the overall separation, since it resulted in
an acceptable compromise signal for all the studied anti-
oxidants and FAME. Although the C18 column resulted
in good separation of SPA, co-elution between BHT and
matrix interference components was observed; see Fig. 1a
lower trace. Co-elution of SPA and matrix was still observed
when aqueous/methanol mobile phase conditions were var-
ied. Such co-elution leads to unreliable identification and
quantification of BHT in practical samples. The alkyl-phe- Fig. 2  Calibration curves generated in HPLC–UV analysis at 280 nm
nyl modified column was then tested, and resulted in good for PG, TBHQ, BHA and BHT (a–d, respectively) response areas over
the concentration range 10–80 ppm, vs the internal standard peak area.
separation of all the antioxidants from the biodiesel matrix The corresponding calibration relationships are y = − 0.0992 + 0.2544x
(refer to peaks 1–4 in Fig. 1a, upper trace). This column (R2 = 0.9995), y = 0.0432 + 0.0668 × (0.9986), y = − 0.0913 + 0.0674 ×
was selected for further studies. The presence of the phenyl (0.9992) and y = − 0.0127 + 0.0373 × (0.9995), for a, b, c, d, respec-
group on the phase resulted in relatively stronger interaction tively
between the phenyl stationary phase and the antioxidants,
due to enhanced π–π interaction, compared to that of FAME.
Hence the SPA moved to later retention, and toward the Calibration Curves and Detection Limit
FAME matrix, but no overlap occurred. Variation of detec-
tion wavelength revealed that 270, 290, 290 and 280 nm Calibration curves (concentration vs peak area ratio rela-
gave the highest intensities for PG, TBHQ, BHA and BHT, tive to the internal standard peak area at 280 nm) using
respectively; see Fig. 1b for the corresponding peaks 1–4. HPLC–UV analysis were constructed for the four antiox-
Determination of limits of detection was performed accord- idants by injecting the antioxidant mixture over the con-
ing to these selected wavelengths for different antioxidants, centration range 10–80 ppm. The curves were linear over
using UV detection. this range, with correlation coefficients between 0.9985
and 0.9994 for all four antioxidants (Fig. 2). This calibra-
tion range was selected to match the expected concentra-
tion range of antioxidants in real biodiesel samples after
20-fold dilution in methanol. The sensitivity of this method
was sufficient, with limits of detection values estimated at

13
Direct Analysis of Synthetic Phenolic Antioxidants, and Fatty Acid Methyl Ester Stability…

0.010, 0.015, 0.013 and 0.030 ppm for PG, TBHQ, BHA Table 1  Analyte analytical retention and MS data in biodiesel using
and BHT (observed at wavelengths of 270, 290, 290 and HPLC–Q-Exactive Orbitrap MS
280 nm, respectively), which have been identified from the No. Retention Exact mass Mass Possible compound
lowest tested concentrations which resulted in a signal-to- time (min) [M + H]+ or accuracy
noise ratio of the antioxidant peaks being ≥ 3. SPA concen- [M − H]− (ppm)
trations considerably less than 10 ppm were tested for this 1 15.5 211.0607a,c − 0.3 PGb
calculation. 2 16.7 165.0916a,c − 2.3 TBHQb
3 20.8 179.1072a,c − 1.9 BHAb
Identification of Antioxidants in Biodiesel Samples 4 25.4 219.1749a,c 0.4 BHTb
5 25.2, 27.6d 241.2169 − 5.0 FAME C14:1
According to the calibration curves (Fig. 2), each antioxidant 6 25.8, 27.8d 255.2325 − 4.7 FAME C15:1
in spiked biodiesel samples was quantified. The SPA com- 7 26.4, 28.6d 269.2481 − 4.1 FAME C16:1
pounds were spiked into the biodiesel at a concentration of 8 26.6, 28.2d 291.2324 − 3.8 FAME C18:4
700 ppm (after 20-fold dilution, 35 ppm was injected into 9 26.6, 28.5d 293.2480 − 3.4 FAME C18:3
the HPLC). The calculated concentrations for PG, TBHQ, 10 26.9, 29.0d 295.2637 − 3.7 FAME C18:2
BHA and BHT in the spiked biodiesel sample were 458 ± 11, 11 27.2, 29.5d 297.2794 − 4.4 FAME C18:1
12 30.2 299.2949 − 2.9 FAME C18:0
13 29.6 321.2795 − 4.7 FAME C20:3
14 27.7, 30.0d 323.2951 − 5.3 FAME C20:2
15 28.0, 30.6d 325.3107 − 3.4 FAME C20:1
16 30.2, 31.5d 327.3263 − 3.4 FAME C20:0
17 32.1 341.3419 − 3.2 FAME C21:0
a
Compound exact mass values calculated in negative ion mode
b
Compound identity also confirmed by injection of authentic SPA
compounds
c
Pseudo-molecular ion for the [M − H]− ion
d
Compound retention with >1 peak for an exact mass value indicates
the presence of isomers

524 ± 14, 721 ± 17 and 556 ± 18 ppm, respectively. The error

values were calculated standard deviations in the spiked bio-
diesel samples, calculated for three repeat sample prepara-
tions. Based on the injected spiked concentration of 35 ppm,
recoveries of the four antioxidants in biodiesel sample matrix
were 70 ± 2, 85 ± 2, 103 ± 2 and 92 ± 4%, respectively. The
calculated concentration for BHT in the provided biodiesel
sample to which BHT was added, was 647 ± 10 ppm, being
within the range reported (600–800 ppm).

FAME and SPA Profiling by HPLC–Orbitrap MS

The four antioxidant peaks were detected using

HPLC–Orbitrap MS with reduced background interference
in the negative mode, as illustrated by the ion map plot of
retention time vs m/z values (Fig. 3a). The four antioxidant
peaks corresponded to the m/z values (212.20 g mol−1,
166.22 g mol−1, 180.25 g mol−1 and 220.36 g mol−1 for PG,
TBHQ, BHA and BHT, respectively) of pseudo-molecular
ions with loss of a proton (Table 1; see compounds 1–4).
Fig. 3  HPLC–Q-Exactive Orbitrap MS ion map analysis of biodiesel The mobile phase condition was the same as that applied
spiked with PG, TBHQ, BHA and BHT (2 ppm each), labelled as
in HPLC–UV analysis. However, the flow was 0.3 mL/min,
peaks 1–4, respectively, in a negative and b positive modes with cor-
responding extracted ion chromatograms for the four antioxidants in c compared with 0.6 mL/min used in HPLC–UV resulting

13
 M. Casagrande et al.

in longer retention time (tR) observed in MS analysis. The

slow flow facilitates spray formation in ESI–MS analysis. In
addition to antioxidants, the FAME composition in biodiesel
may be detected and identified. The negative ion mode gives
many fewer FAME peaks displayed in the chromatogram
(Fig. 3a), with many more components observed in positive
ion mode (Fig. 3b). In the positive mode, according to the
accurate mass values within ± 5 ppm mass accuracy, the pro-
tonated molecular ions were identified as shown in Table 1,
with the ion map illustrated in Fig. 3b. Since increasingly
hydrophobic compounds are more strongly retained on the
hydrophobic phase, FAME with higher m/z values eluted
later (located towards the top right corner of the ion map),
and saturated FAME (more hydrophobic) eluted later than
unsaturated FAME of similar molar mass (Table 1). These

Fig. 5  Chemical profile in different biodiesel samples after exposure

to sunlight up to 8 weeks: a total FAME C18 contents (compounds
8–12) in blank (times symbol), real biodiesel (open circle) and
spiked biodiesel (open triangle) samples with and without BHT; b
BHT (open square) content in real biodiesel sample and c PG (open
diamond), TBHQ (open triangle), BHA (asterisk) and BHT (open
square) contents in spiked biodiesel sample. The error bars were
obtained from three replicates of the exposure experiments

observed trends are indicative of a typical character of

reversed-phase separation [30].

Biodiesel Stability Test

Stability of biodiesel components under exposure to sunlight

was investigated, as a relatively simple exercise in forced
oxidative degradation. The overlay chromatograms at dif-
ferent exposure periods of 0, 1, 4 and 8 weeks for biodiesel
without antioxidant, biodiesel spiked with all four antioxi-
dants, and the supplied biodiesel sample with BHT additive
Fig. 4  a HPLC–UV analysis of biodiesel without antioxidant, b bio- are shown in Fig. 4a–c. Without added antioxidant, the C18
diesel spiked with PG, TBHQ, BHA and BHT (700 ppm of each FAME isomer content in biodiesel is significantly reduced
SPA), and c the supplied biodiesel sample with BHT additive (600– to about 52 ± 4% w/w in the first week of exposure, decreas-
800 ppm) after 0, 1, 4 and 8 weeks (black, green, purple and pink, ing to 29 ± 6% w/w after 8 weeks exposure (Fig. 5a). For
respectively) under sunlight exposure. The labelled compounds are
BHT (4) and FAME components (8–12) shown in Table 1 with the samples containing antioxidants, the composition from 4
unknown oxidised species indicated in a to 8 weeks did not vary significantly compared with the

13
Direct Analysis of Synthetic Phenolic Antioxidants, and Fatty Acid Methyl Ester Stability…

original assay at 0 weeks e.g. as illustrated for BHT (peak 2. Goulart LA, Teixeira ARL, Ramalho DA, Terezo AJ, Castilho M
4 in Fig. 4b, c with the corresponding content plotted in (2014) Development of an analytical method for the determination
of tert-butylhydroquinone in soybean biodiesel. Fuel 115:126–131
Fig. 5b) and other antioxidants (Fig. 5c). Some FAME with 3. Ng J-H, Ng HK, Gan S (2010) Recent trends in policies, socio-
longer alkyl chains, such as FAME C18, slightly reduced economy and future directions of the biodiesel industry. Clean
to about 90% w/w of the original after 8 weeks exposure Technol Environ Policy 12:213–238
(Fig. 5a). Here, since the unsaturated C18 components are 4. Dunn RO (2008) Antioxidants for improving storage stability of
biodiesel. Biofuels Bioprod Biorefin 2:304–318
not resolved, and the saturated C18 is a minor component 5. Domingos AK, Saad EB, Vechiatto WWD, Wilhelm HM, Ramos
and elutes later, quantification is based on the unsaturated LP (2007) The influence of BHA, BHT and TBHQ on the oxida-
peak, which is more susceptible to oxidation. Although no tion stability of soybean oil ethyl esters (biodiesel). J Braz Chem
particular temperature control was applied, the SPA were Soc 18:416–423
6. Tagliabue S, Gasparoli A, della Bella L, Bondioli P (2004) Quali-
effective in reducing degradation; the effectiveness of SPA quantitative determination of synthetic antioxidants in biodiesel.
at elevated temperature is noted [4]. Riv Ital Sostanze Gr LXXXI:37–40
7. da Silva YP, Dalmoro V, Ruiz YPM, Capeletti LB, Mendonça
CRB, dos Santos JHZ, Piatnicki CMS (2014) Biodiesel water in
oil microemulsions: ferrocene as a hydrophobic probe for direct
Conclusion analysis by differential pulse voltammetry at a Pt ultramicroelec-
trode. Anal Methods 6:9212–9219
This study develops techniques for direct analysis of antioxi- 8. Webster RL, Rawson PM, Evans DJ, Marriott PJ (2014) Synthetic
dants and FAME in biodiesel samples, to explore changes to phenolic antioxidants in conventional and alternatively-derived
middle distillate fuels analysed by gas chromatography with tri-
FAME composition under exposure to sunlight as an accel- ple quadrupole and quadrupole time of flight mass spectrometry.
erated oxidation condition. HPLC–UV was able to separate, Energy Fuel 28:1097–1102
detect and quantify four antioxidants without derivatisation 9. Kivevele T, Huan Z (2015) Influence of metal contaminants and
with low LOD and good linearity of calibration curves. antioxidant additives on storage stability of biodiesel produced
from non-edible oils of Eastern Africa origin (Croton megalo-
HPLC–Q-Exactive Orbitrap MS was applied for identifica- carpus and Moringa oleifera oils). Fuel 158:530–537
tion of antioxidants and FAME according to accurate mass 10. Pereira GG, Alberici RM, Ferreira LL, Santos JM, Nascimento
analysis. Under exposure to sunlight as an accelerated oxi- HL, Eberlin MN, Barrera-Arellano D (2015) A screening
dation process, the presence of antioxidant clearly reduced method to evaluate soybean oil-based biodiesel oxidative qual-
ity during its shelf life. J Am Oil Chem Soc 92:967–974
biodiesel degradation, where FAME content remained con- 11. Rashed MM, Kalam MA, Masjuki HH, Rashedul HK, Ashraful
stant up to the maximum of the study period (8 weeks). The AM, Shancita I, Ruhul AM (2015) Stability of biodiesel, its
developed methods are expected to be applicable for reliable improvement and the effect of antioxidant treated blends on
routine quantification of antioxidant in practical biodiesel engine performance and emission. RSC Adv 5:36240–36261
12. Santos AGD, Souza LD, Caldeira VPS, Farias MF, Fernandes
samples, and investigation of biodiesel stability under differ- VJ Jr, Araujo AS (2014) Kinetic study and thermoxidative deg-
ent treatments. Compared to other methods the main advan- radation of palm oil and biodiesel. Thermochim Acta 592:18–22
tage of the proposed methodology is the reduced time of 13. Yang Z, Hollebone BP, Wang Z, Yang C, Brown C, Landriault
analysis, because of the ease of sample preparation. M (2014) Storage stability of commercially available biodies-
els and their blends under different storage conditions. Fuel
115:366–377
Acknowledgements PJM acknowledges the Australian Research Coun- 14. Almeida JMS, Dornellas RM, Yotsumoto-Neto S, Ghisi M,
cil for a Discovery Outstanding Researcher Award; DP130100217. MC Furtado JGC, Marques EP, Aucélio RQ, Marques ALB (2014)
acknowledges CAPES Foundation for financial support. The authors A simple electroanalytical procedure for the determination of
acknowledge Agilent Technologies and ThermoFisher Scientific for calcium in biodiesel. Fuel 115:658–665
provision of support for some of the facilities used in this study. 15. Chýlková J, Tomášková M, Mikysek T, Šelešovská R, Jehlička J
(2012) Voltammetric determination of BHT antioxidant at gold
Compliance with Ethical Standards electrode in biodiesel. Electroanalysis 24:1374–1379
16. Freitas HC, Almeida ES, Tormin TF, Richter EM, Munoz RAA
Conflict of Interest The authors declare that they have no conflict of (2015) Ultrasound-assisted digestion of biodiesel samples for
interest. determination of metals by stripping voltammetry. Anal Meth-
ods 7:7170–7176
Ethical Approval This article does not contain any studies with human 17. Spudeit DA, Piovezan M, Dolzan MD, Vistuba JP, Azevedo MS,
participants or animals performed by any of the authors. Vitali L, Oliveira MAL, Costa ACO, Micke GA (2013) Simul-
taneous determination of free and total glycerol in biodiesel
by capillary electrophoresis using multiple short-end injection.
Electrophoresis 34:3333–3340
18. Nogueira T, do Lago CL (2011) Determination of Ca, K, Mg,
Na, sulfate, phosphate, formate, acetate, propionate, and glyc-
References erol in biodiesel by capillary electrophoresis with capacitively
coupled contactless conductivity detection. Microchem J
1. Pullen J, Saeed K (2014) Factors affecting biodiesel engine per- 99:267–272
formance and exhaust emissions—part I. Rev Energy 72:1–16

13
 M. Casagrande et al.

19. Piovezan M, Costa ACO, Jager AV, de Oliveira MAL, Micke GA 31. Guo L, Xie M-Y, Yan A-P, Wan Y-Q, Wu Y-M (2006) Simultane-
(2010) Development of a fast capillary electrophoresis method to ous determination of five synthetic antioxidants in edible vegeta-
determine inorganic cations in biodiesel samples. Anal Chim Acta ble oil by GC–MS. Anal Bioanal Chem 386:1881–1887
673:200–205 32. Ding M, Zou J (2012) Rapid micropreparation procedure for the
20. Okullo A, Ogwok P, Temu AK, Ntalikwa JW (2013) Gas chro- gas chromatographic-mass spectrometric determination of BHT,
matographic determination of glycerol and triglycerides in bio- BHA and TBHQ in edible oils. Food Chem 131:1051–1055
diesel from jatropha and castor vegetable oils. Adv Mater Res 33. Marriott PJ, Chin S-T, Maikhunthod B, Schmarr H-G, Bieri S
824:436–443 (2012) Multidimensional gas chromatography. TrAC Trends Anal
21. Hirschegger L, Schober S, Mittelbach M (2014) Efficient and Chem 34:1–21
sensitive method for the quantification of saturated monoacylg- 34. Mogollon NGS, Ribeiro FADL, Lopez MM, Hantao LW, Poppi
lycerols in biodiesel by gas chromatography–mass spectrometry. RJ, Augusto F (2013) Quantitative analysis of biodiesel in blends
Eur J Lipid Sci Technol 116:89–96 of biodiesel and conventional diesel by comprehensive two-
22. Montpetit A, Tremblay AY (2016) A quantitative method of analy- dimensional gas chromatography and multivariate curve resolu-
sis for sterol glycosides in biodiesel and FAME using GC-FID. J tion. Anal Chim Acta 796:130–136
Am Oil Chem Soc 93:479–487 35. O’Neil GW, Culler AR, Williams JR, Burlow NP, Gilbert GJ,
23. Bezerra KS, Antoniosi Filho NR (2014) Gas chromatographic Carmichael CA, Nelson RK, Swarthout RF, Reddy CM (2015)
analysis of free steroids in biodiesel. Fuel 130:149–153 Production of jet fuel range hydrocarbons as a coproduct of algal
24. Farias AFF, Conceição MM, Cavalcanti EHS, Melo MAR, dos biodiesel by butenolysis of long-chain alkenones. Energy Fuel
Santos IMG, de Souza AG (2016) Analysis of soybean biodiesel 29:922–930
additive with different formulations of oils and fats. J Therm Anal 36. Jiang M, Kulsing C, Nolvachai Y, Marriott PJ (2015) Two-dimen-
Calorim 123:2121–2127 sional retention indices improve component identification in com-
25. Ahmed MA, Khan I, Hashima J, Musharraf SG (2015) Sensitive prehensive two-dimensional gas chromatography of saffron. Anal
determination of glycerol by derivatization using a HPLC-DAD Chem 87:5753–5761
method in biodiesel samples. Anal Methods 7:7805–7810 37. Nolvachai Y, Marriott PJ (2013) GC for flavonoids analysis: past,
26. Fedosov SN, Fernandes NA, Firdaus MY (2014) Analysis of oil– current and prospective trends. J Sep Sci 36:20–36
biodiesel samples by high performance liquid chromatography 38. Gao X, Williams SJ, Woodman OL, Marriott PJ (2010) Compre-
using the normal phase column of new generation and the evapo- hensive two-dimensional gas chromatography, retention indices
rative light scattering detector. J Chromatogr A 1326:56–62 and time-of-flight mass spectra of flavonoids and chalcones. J
27. Haagenson DM, Perleberg JR, Wiesenborn DP (2014) Fractiona- Chromatogr A 1217:8317–8326
tion of canola biodiesel sediment for quantification of steryl glu- 39. Holčapek M, Jandera P, Fischer J, Prokeš B (1999) Analytical
cosides with HPLC/ELSD. J Am Oil Chem Soc 91:497–502 monitoring of the production of biodiesel by high-performance
28. Allen SJ, Ott LS (2012) HPLC method for rapidly following bio- liquid chromatography with various detection methods. J Chro-
diesel fuel transesterification reaction progress using a core-shell matogr A 858:13–31
column. Anal Bioanal Chem 404:267–272 40. Knothe G (2007) Some aspects of biodiesel oxidative stability.
29. Carvalho MS, Mendonça MA, Pinho DMM, Resck IS, Suarez Fuel Process Technol 88:669–677
PAZ (2012) Chromatographic analyses of fatty acid methyl esters
by HPLC-UV and GC-FID. J Braz Chem Soc 23:763–769 Publisher’s Note Springer Nature remains neutral with regard to
30. Brandão LFP, Braga JWB, Suarez PAZ (2012) Determination of jurisdictional claims in published maps and institutional affiliations.
vegetable oils and fats adulterants in diesel oil by high perfor-
mance liquid chromatography and multivariate methods. J Chro-
matogr A 1225: 150–157

13