Professional Documents
Culture Documents
Mitigation Strategies For The Reduction of 2 - and 3-MCPD Esters and Glycidyl Esters in The Vegetable Oil Processing Industry
Mitigation Strategies For The Reduction of 2 - and 3-MCPD Esters and Glycidyl Esters in The Vegetable Oil Processing Industry
Abstract: The refining of vegetable oils leads to the formation of 2- and 3-monochloropropane-1,2-diol esters (2- and
3-MCPD-E), and glycidyl esters (Gly-E). A literature review was performed aiming to provide up-to-date knowledge
on mitigation strategies during oil refining that can reduce the formation of these three processing contaminants. The
review used the database Scopus and covered the period from 2009 to 2017. Most of the 18 papers dealt with palm oil
and two papers with vegetable oil. Most studies focused on 3-MCPD-E, some on Gly-E, and none on 2-MCPD-E.
Water degumming was able to reduce the concentrations of 3-MCPD-E by 84% and Gly-E by 26%. Neutralization
of the oil reduced concentrations of 3-MCPD-E by 81% and Gly-E by 84%. Bleaching with synthetic magnesium
silicate reduced the 3-MCPD-E concentration by 67%. For the deodorization step, several mitigation strategies, such as
double-deodorization, the addition of various antioxidants, or a longer deodorization time, can reduce the formations of
3-MCPD-E by 82% and Gly-E by 78%. Postrefining mitigation, including the use of absorbents, enzymes, or rebleaching
of the oil, has also been reported to produce desirable contaminant reduction. Postrefining treatment with calcinated
zeolite was able to reduce the 3-MCPD-E concentration by 19% and the Gly-E concentration by 77%. Applying
combined mitigation strategies to multiple steps of oil refining is likely crucial in order to adequately reduce levels of
3-MCPD-E and Gly-E.
Keywords: 3-monochloropropane-1,2-diol esters, glycidyl esters, oil refining, processing contaminants, vegetable oil
C 2019 Institute of Food Technologists®
doi: 10.1111/1541-4337.12415 Vol. 18, 2019 r Comprehensive Reviews in Food Science and Food Safety 349
3-MCPD & glycidol mitigation strategies . . .
monoacylglycerols (Cheng, Liu, Wang, & Liu, 2017). However, search was aimed at publications about mitigation strategies during
Gly-E can also be formed from both 2- and 3-MCPD-E under oil refining, including those for individual refining stages. Addi-
certain temperature, time, and pH conditions, and this pathway tional subject area selection was performed, as the search terms
of Gly-E formation is reversible. Cheng et al. (2017) and Shimizu included articles from irrelevant subject areas such as dentistry,
et al. (2012) have explained this reversible conversion of 2- and nursing, immunology, or microbiology. The following subject
3-MCPD-E into Gly-E in more detail. areas were excluded from the selection: environmental science,
Various food products containing refined oils were shown to pharmacology, toxicology and pharmaceutics, energy, dentistry,
have 2-, 3-MCPD-E, and/or Gly-E concentrations of 1 mg/kg immunology and microbiology, physics and astronomy, nursing,
oil or more (European Food Safety Authority, 2013; Hamlet et al., social sciences, business, management and accounting, computer
2011). Several studies have shown that 3-MCPD-E and Gly-E science, mathematics, and psychology. This resulted in 116 hits.
concentrations in infant formula can reach ࣙ1 mg/kg fat, and that Applying more filters or making the search terms more specific
2-MCPD-E concentrations can reach half of that amount (Leigh & could result in unwanted elimination of relevant publications. For
MacMahon, 2017; Wöhrlin, Fry, Lahrssen-Wiederholt, & Preiß- this reason, the 116 hits were manually selected for their relevance
Weigert, 2015). The application of different mitigation strategies to this review by screening their titles and abstracts. For example,
at the various steps of vegetable oil and fat production to limit the publications about analytical techniques to determine the con-
formation of these process contaminants is therefore a high priority centration of only Gly-E or 3-MCPD-E, publications about risk
of many research groups and the food industry. The strategies assessment, or publications about a completely different subject
explored so far have focused on different (sub)processes of oil were excluded from the final selection.
production, from the palm plantation, through novel oil refining
methods, up to methods designed specifically to break down 2-, Results and Discussion
3-MCPD-E, and Gly-E after the refining process (Matthäus & General review results
Pudel, 2013). Although the mitigation strategies developed to In total, 18 relevant papers were obtained. Seventeen papers
date show promising results, current knowledge remains limited focused on mitigation strategies for one or more of the three con-
and scattered, as discussed below. taminants considered during one or multiple stages of vegetable
In the last five years, several reviews of Gly-E and 3-MCPD oil refining. From these, 16 papers dealt with palm oil, and the
have been published, of which two either focused on Gly-E esters remaining two did not refer to palm oil but more generally to
or (unbound) 3-MCPD, and two on both of these compounds. All vegetable or edible oils. This preference towards palm oils is be-
four of these reviews covered the subject from a broad perspective, cause they contain the highest concentration of 3-MCPD-E and
and did not provide in-depth knowledge about the effectiveness related compounds, thus there is an acute need for reliable miti-
of mitigation strategies to limit the formation of 2-, 3-MCPD-E, gation strategies for this type of oil. Mitigation strategies for the
and/or Gly-E. Additionally, the discussed mitigation strategies are degumming step were covered in six papers; for the neutraliza-
mostly limited to either 3-MCPD-E or Gly-E. However, mitiga- tion step in four papers; for the bleaching step in five papers; and
tion strategies should consider all three contaminants at the same for the deodorization step in eight papers. Five additional papers
time, because formation of the three compounds are related to provided alternative insights into the ways of reducing contami-
each other, and the three compounds often are found together in nants by means of postrefining removal. Finally, one of the five
refined oils. Data on mitigation strategies of 2-MCPD are very postrefining papers described a model that tested the combination
limited. This lack of research data on 2-MCPD and its esters of multiple mitigation strategies. The total sum of the articles dis-
is reflected in the aforementioned reviews and in the European cussed in the specific section regarding refining stages is greater
Commission’s regulations, in which no MLs for 2-MCPD and/or than the 18 cited because multiple articles covered more than one
2-MCPD-E have been established to date. mitigation strategy. Therefore, such articles are discussed twice in
This review aims to provide up-to-date insights into the miti- the corresponding and relevant sections. In the current review,
gation strategies for the formation of 2-, 3-MCPD-E, and Gly-E mitigation strategies designed for any other vegetable oils, except
during vegetable oil processing, based on the scientific literature. palm oils, are specifically addressed.
The current state-of-the-art knowledge on potential prevention The level of detail presented in the methods and results sec-
and control strategies during oil production is provided, as well tions varied among the papers selected. To present all results in
as gaps in the data. The results presented will facilitate further re- a comparable and uniform way, it was occasionally necessary to
search on the development of an optimal mitigation strategy. Due perform additional calculations based on the original data in some
to the narrow and precise scope of this paper, the mechanisms un- papers. If contaminant reduction was presented only as concen-
derlying formation, toxicity, and occurrence of the three process tration, we expressed the associated relative reduction as a per-
contaminants are not covered. centage. In another case, data were presented in bar graphs, and
we estimated the percentage of contaminant reduction from these
Materials and Methods graphs.
A literature review was conducted using the online database
Scopus. Articles from 2009 to 2017 were included in the selection. Degumming
The search terms were are as follows: Gums (hydratable and nonhydratable phospholipids) and other
TITLE-ABS-KEY(monochloro∗ OR MCPD∗ OR glycid∗ l phosphorus-containing compounds are removed during degum-
AND oil OR ∗ olein AND mitigat∗ OR reduc∗ OR remov∗ OR ming. Oils high in gum concentration, including sunflower seed
degumm∗ OR neutrali∗ OR wash∗ OR bleach∗ OR deodori∗ oil and rapeseed oil, are usually treated with a water degumming
OR refin∗ ) method to remove the water-soluble gums. Gibon, De Greyt, and
These search terms were assembled in order to cover all three Kellens (2007) reported that palm oil is usually low in phospho-
process contaminants (2-MCPD, 3-MCPD, and glycidol in both lipid and phosphorus contents (10 to 20 ppm) (Gibon et al., 2007).
the esterified and unbound forms) at once. The second part of the Therefore, palm oils are usually degummed with either phosphoric
350 Comprehensive Reviews in Food Science and Food Safety r Vol. 18, 2019
C 2019 Institute of Food Technologists®
3-MCPD & glycidol mitigation strategies . . .
Table 1–Effectivity of various degumming conditions for the mitigation of 2-, 3-MCPD esters, and glycidyl esters.
acid or citric acid. This so-called “dry degumming” results in lower Zulkurnain, Lai, Tan, Abdul Latip, and Tan (2013) published a
oil losses as compared to water washing. However, water degum- physical refining process for the reduction of 3-MCPD-E. They
ming may be more beneficial than dry degumming in terms of reported a reduction of 80% when using 5% water to degum the
lowering the 2- and 3-MCPD-E concentrations. As mentioned by oil without the addition of any acids (from 1.0 mg/kg for 0%
Destaillats et al. (2012), Šmidrkal et al. (2016), and Hamlet et al. water to 0.2 mg/kg for 5% water) (Zulkurnain et al., 2013). This
(2011), the formation of 2- and 3-MCPD-E requires a chlorine can be explained by the removal of 3-MCPD-E precursors by
source that easily could dissociate to create the required chloride water.
ions. Water degumming can be seen as a washing step with water. Based on the presented data, water degumming was shown to be
During this step, chlorine containing polar precursors of 2- and effective in reducing the concentrations of 3-MCPD-E and Gly-
3-MCPD-E can be removed from the oil. Therefore, the elimi- E during the degumming process. However, Zulkurnain et al.
nation of chlorine sources can be an effective method to reduce (2012) and Ramli et al. (2011) noted that the quality of the oil, as
the formation of 2- and 3-MCPD-E. The results of mitigation defined by the terms refined, bleached, and deodorized (RBD),
strategies involving various degumming methods are presented in might be also affected by water degumming. The color of the
Table 1. oil1 is one of the quality parameters that are monitored by re-
Out of the five studies, only one analyzed Gly-E and pro- fineries and suppliers. Ramli et al. (2011) observed strong red
posed a mitigation strategy for this compound during degumming. coloring (3.5 Red on the RYBN color scale) of their final prod-
Matthäus, Pudel, Fehling, Vosmann, and Freudenstein (2011) re- uct, together with a less efficient removal of phosphatides and iron
ported a strategy to reduce 3-MCPD-E and Gly-E in palm oil. In in comparison with their standard procedure. Zulkurnain et al.
this case, the crude palm oil (CPO) was washed with water prior to (2012) reported a stronger red color in the range of 4.0 to 5.9
a simulated deodorization step. They achieved a reduction of 25% in RBD palm oil (Zulkurnain et al., 2012). Gibon et al. (2007)
(from 2.8 to 2.1 mg/kg) for 3-MCPD-E, and 16% (from 3.5 to mentioned that incomplete removal of phosphatides during the
3.0 mg/kg) for Gly-E in comparison with unwashed CPO. Im- degumming step may result in oil darkening during deodoriza-
proved results were obtained by washing the CPO with 75% tion, which negatively impacts the oil flavor (Gibon et al., 2007).
ethanol; in this case, 3-MCPD-E was reduced by 36%, from 2.8 to Interestingly, a reduction of red coloring (approximately 2.0 Red)
1.8 mg/kg, and Gly-E was reduced by 26%, from 3.5 to 2.6 mg/kg can be achieved by using a larger amount of water (from 1% to 5%
(Matthäus et al., 2011). However, 2 years later, Matthäus and Pudel water) to degum the oil (Zulkurnain et al., 2013). In conclusion,
(2013) reported a reduction of 38% of 3-MCPD-E by a water wash water degumming can effectively reduce Gly-E and 3-MCPD-
step. In another study by Pudel et al. (2011), the water degum- E concentrations, with 3-MCPD-E concentrations reduced to
ming step was compared to various water degumming methods <1 mg/kg. Unfortunately, not all the papers mentioned sufficient
using different water percentages, such as water degumming fol- details of the applied washing conditions (that is, amount of water
lowed by a 0.2% phosphoric acid or 0.3% citric acid treatment. or time and temperature of the treatment) for a full comparison
The best reduction was achieved when the oil was washed with of the investigated methods.
5% deionized water only, reducing the combined concentration
of 3-MCPD-E and related compounds (mainly Gly-E) by 64% Neutralization
(from 5.5 to 2.0 mg/kg) compared with unwashed raw palm oil In physical refining, FFAs are removed via strip distillation dur-
(Pudel et al., 2011). Ramli et al. (2011) noted a marked reduction ing deodorization. This is possible because FFAs are more volatile
in 3-MCPD concentration between degumming with (plain) wa- than triacylglycerols. During chemical refining, the FFAs are re-
ter compared with 0.02% phosphoric acid solution. When 0.02% moved earlier in the process during alkali neutralization. This is
phosphoric acid was used, 2.1 mg/kg 3-MCPD-E was found in a necessary step and is unique to chemical oil refining, and it is
the oil, while using water with no added acids reduced the 3- an effective way to neutralize most of the FFAs present in the oil.
MCPD-E concentration to 0.75 mg/kg (64% reduction) (Ramli During chemical refining, FFAs are removed by converting them
et al., 2011). Zulkurnain et al. (2012) reported an even larger re-
duction of 3-MCPD-E using water degumming with 1% water at 1A standardized color chart or a series of color filter combinations can be used
70 °C for 20 min. The 3-MCPD-E concentrations were reduced to help identify and compare the color components. One of the most popular
from 9.8 to 1.6 mg/kg (reduction of 84%) in comparison with acid color scales used to compare oils and fats is the Lovibond R
RYBN color scale
(The Tintometer Ltd, 2016). RYBN stands for red, yellow, blue, and neutral,
degumming with 0.05% phosphoric acid. This might explain the with a corresponding range of 0 to 70 for red, 0 to 70 for yellow, 0 to 40 for
large difference in 3-MCPD ester reduction, since bleaching can blue, and 0 to 3.9 for neutral. Increasing numbers on the scale correlate with
also lower the 3-MCPD-E concentration (as discussed below). darker and more intense tints of that particular color.
C 2019 Institute of Food Technologists® Vol. 18, 2019 r Comprehensive Reviews in Food Science and Food Safety 351
3-MCPD & glycidol mitigation strategies . . .
Table 2–Effectivity of various neutralization conditions for the mitigation of 2-, 3-MCPD esters, and glycidyl esters.
into soap under alkaline conditions. The produced soap stock Bleaching
can then be separated from the oil, removing not only the FFAs Treatment of oil with bleaching clay is used, in addition to ther-
but also the residual phospholipids, oxidized products, and metal mal degradation, to remove colored pigments such as carotenes.
ions that were not completely removed during degumming. Even Bleaching clay is porous, and color pigments are physically en-
though most of the FFAs are removed during the neutralization trapped into the clay. Pigments can also be adsorbed onto bleach-
step in chemical refining, a final deodorization step is still neces- ing clay by van der Waals forces and can be chemically bound via
sary to remove any remaining FFAs, fragrant molecules, and other ionic or covalent bonds (Gibon et al., 2007; Silva et al., 2014). In
volatile compounds. addition to the removal or reduction of pigments from oil, bleach-
Table 2 represents the results of mitigation strategies during neu- ing also removes other impurities such as lipid oxidation products
tralization. Ramli et al. (2011) used 0.2% calcium oxide to neu- and metals (Gibon et al., 2007). Heat treatment and bleaching are
tralize the 0.02% phosphoric acid. This resulted in a 3-MCPD- used in combination because the bleaching step alone is not able to
E concentration of 1.4 mg/kg in the neutralized oil compared remove color pigments completely. Most pigments are not heat-
with a concentration of 2.2 mg/kg in the non-neutralized oil, stable and further removal can be achieved by degradation during
which is a reduction of 36% (Ramli et al., 2011). Matthäus and the subsequent deodorization step. The results of the mitigation
Pudel (2013) reported similar findings. They observed a reduc- strategies during bleaching are presented in Table 3.
tion of 45% and 35%, respectively, when potassium hydroxide Franke et al. (2009) focused on 3-MCPD-E concentrations in
(KOH) or sodium hydroxide (NaOH) was used for neutralization crude and prerefined palm oil and in crude rapeseed oil. Before and
(Pudel et al., 2011). Lastly, Freudenstein, Weking, and Matthäus after every refining process (including degumming, neutralization,
(2013) reported the effect of neutralization with either sodium washing, drying, and bleaching), they analyzed the 3-MCPD-E
carbonate (Na2 CO3 ) or sodium hydrogen carbonate (NaHCO3 ) concentration in each of the three oil samples. Most notable was
in model oil that was spiked with 16 mg NaCl/g to promote the the decrease in 3-MCPD-E concentration after the bleaching pro-
formation of 3-MCPD-E and related compounds (for example, cess in the prerefined oil. Right before the bleaching step, the pre-
Gly-E). Results showed that NaHCO3 was more effective than refined palm oil had a 3-MCPD-E concentration of 6.06 mg/kg,
Na2 CO3 in reducing concentrations of 3-MCPD-E and related while it was 2.48 mg/kg after bleaching, which is a decrease of
compounds. At a concentration of 5 mmol of Na2 CO3 /kg sam- 59% (Franke et al., 2009). In the CPO, the decrease was negligible.
ple, a maximum reduction of 53% (from 5.9 to 2.8 mg/kg) for Right before the bleaching step, the oil contained a 3-MCPD-E
3-MCPD-E and 69% (from 7.0 to 2.2 mg/kg) for 3-MCPD-E and concentration of 1.04 mg/kg, which was reduced to 1.00 mg/kg
related compounds (combined) was achieved. NaHCO3 needed a after the bleaching process (Franke et al., 2009). The 3-MCPD-E
concentration of only 1 mmol/kg sample in order to achieve an concentration in the rapeseed oil was below the detection limit
81% reduction in the concentration of 3-MCPD-E (from 5.9 to (0.4 mg/kg) throughout the entire refining process, except for
1.1 mg/kg) and an 84% reduction for 3-MCPD-E combined with the final product (after deodorization), which had a concentra-
related compounds (7.0 to 1.1 mg/kg) (Freudenstein et al., 2013). tion of 1.04 mg/kg (Franke et al., 2009). As a bleaching agent,
The highest reduction in 3-MCPD-E and Gly-E concentrations Franke et al. (2009) used 1% Tonsil optimum 214 FF bleaching
was obtained when NaHCO3 was used to neutralize the oil af- earth, which is a commercially available as acid-activated calcium
ter degumming. However, the observed reduction was obtained bentonite. It can not only adsorb polar compounds, including
in a model oil spiked with chlorine to facilitate the formation of peroxides and phospholipids, but also carotenoids, which makes it
2- and 3-MCPD-E. The effect of KOH or NaOH addition has suitable for refining vegetable oils. The suffix “FF” stands for fast
been tested in palm oils and, even though the observed reduc- filtration, indicating a uniform grain-size distribution. Similar to
tion was lower compared to using NaHCO3 , the reduction was Franke et al. (2009), Pudel et al. observed a marked reduction of
still considerable. Such levels of reduction can be relevant when 3-MCPD-E concentration by 45% (5.5 to 3.0 mg/kg) when the
combinations of mitigation strategies are needed to achieve the oil was bleached with 1% Tonsil 4191 FF bleaching earth (Pudel
desired reduction of esters. Although an alkali neutralization step et al., 2011).
is not required when physical refining is applied, it is still useful Ramli et al. (2011) compared acid-activated bleaching clay and
to have it in the refining process. Any acids present in the oil, ei- natural bleaching clay in combination with water degumming. In
ther naturally or introduced during the degumming step, must be their control, natural bleaching clay resulted in an average con-
neutralized prior to deodorization, as neutral oils tend to develop centration of 2.2 mg/kg 3-MCPD-E, while acid-activated clay
less 3-MCPD-E. resulted in a concentration of 2.8 mg/kg. In combination with
352 Comprehensive Reviews in Food Science and Food Safety r Vol. 18, 2019
C 2019 Institute of Food Technologists®
3-MCPD & glycidol mitigation strategies . . .
Table 3–Effectivity of various bleaching conditions for the mitigation of 2-, 3-MCPD esters, and glycidyl esters.
water degumming, the acid-activated bleaching clay was able to (added before and after the magnesium slurry). The setup of the
further reduce the 3-MCPD-E concentration to 0.91 mg/kg (68% experiments for wet bleaching was similar to the setup for the dry-
less than the control). The use of neutral bleaching clay, on the bleaching experiments, but they also used a mixture of magnesium
other hand, resulted in a greater reduction of the 3-MCPD-E silicate and activated clay. Wet bleaching initiated with activated
concentration to 0.49 mg/kg (78% less than the control) (Ramli clay, followed by use of magnesium silicate, resulted in the lowest
et al., 2011). The positive effect of the neutral clay, in compar- 3-MCPD-E concentration (0.23 mg/kg) (Zulkurnain et al.,
ison with the acid-activated bleaching clay, was also greater in 2013). However, for all four dry-bleaching conditions the FFA
the water-degummed oil (22% reduction in the control compared concentration observed was the highest (0.10% FFA). However,
with 46% reduction in the water-degummed oils). The observed this elevated FFA value did not exceed the maximum FFA limit
differences between the acid-activated and neutral clay origin- (also 0.10%) set by the Palm Oil Refiners Assn. of Malaysia (PO-
ated from the clay being activated by the addition of hydrochloric RAM) to be qualified as “good” RBD/neutralized, bleached,
acid or phosphoric acid in order to increase its surface area (Ramli and deodorized palm oil (Malaysian Palm Oil Board, 2000). Fi-
et al., 2011; Taylor, 2009). Residual acids, especially hydrochloric nally, although the 3-MCPD-E concentrations were higher in the
acid, in the bleaching clay may act as an external chloride source, dry-bleached palm oils, Zulkurnain et al. (2013) suggested dry
promoting the formation of 3-MCPD-E (and possibly 2-MCPD- bleaching with a mixture of magnesium silicate and activated clay
E) (Destaillats et al., 2012; Matthäus et al., 2011; Šmidrkal et al., as the preferred method because it resulted in a relatively low 3-
2016). Therefore, activation of bleaching clay by hydrochloric acid MCPD-E concentration in combination with satisfactory color
should be avoided. removal.
Zulkurnain et al. (2012) also tested the usage of neutral clay,
but with different results. They observed that neutral clay resulted Deodorization
in a slightly higher 3-MCPD-E concentration than acid-activated Temperature and duration effects. The deodorization step is
clay. When neutral clay was used, the 3-MCPD-E concentration the most critical for the formation of Gly-E, which is tem-
peaked at 1.75 mg/kg, as compared to 1.55 mg/kg for use of perature dependent; Gly-E concentration increases almost ex-
acid-activated clay (Zulkurnain et al., 2012). This result is the ponentially when oils are exposed to temperatures >230 °C
opposite of those reported in a study by Ramli et al. (2011), to 250 °C for a prolonged period (>1 hour) (Craft, Nagy,
in which neutral clay performed better than acid-activated clay. Seefelder, Dubois, & Destaillats, 2012; Hrncirik & van Duijn,
However, Ramli et al. (2011) used a different type of clay. Thus, 2011; Pudel et al., 2011). Note that 3-MCPD-E is formed
no conclusions regarding the use of neutral compared with acid- rapidly at lower temperatures, as low as 175 °C, while it
activated bleaching clays can be made. is maintained at the same level with increasing temperatures.
Zulkurnain et al. (2012) used magnesium silicate (Magnesol Deodorization conditions vary among refining plants and ac-
R60) as a bleaching agent because it has a large surface area and cording to the quality of the crude oil. The most important
the most active basic sites compared to alumina, silica, and ac- deodorization parameters are usually reduced pressure (1 to 5 torr),
tivated carbon (Zulkurnain et al., 2012). Using 1% Magnesol duration (0.5 to 3 hr), and temperature (200 °C to 240 °C). The
R60 as bleaching agent resulted in a reduction of 3-MCPD-E application of mild deodorization conditions is a valuable mitiga-
concentration to 1.55 mg/kg, while acid-activated clay reduced tion strategy. The results of mitigation strategies during deodor-
3-MCPD-E concentration to 0.51 mg/kg. This is a further im- ization are shown in Table 4. An example of a typical temperature-
provement of 67%. Zulkurnain et al. (2013) also explored ef- time profile of a deodorization process is shown in Figure 1(A).
fects of several combinations of bleaching methods using magne- Hrncirik and van Duijn (2011) reported a 3-MCPD-E con-
sium silicate, which acts as an auxiliary adsorbent for chloro-ester centration of approximately 4.8 mg/kg in chemically refined
precursors. During the wet bleaching experiments, they combined palm oil, which had been deodorized at 180 °C for 1 hr
1% of a 20% w/w magnesium silicate (Magnesol R60) slurry alone, (Hrncirik & van Duijn, 2011). Extending the deodorization time
and in combination with 0.5% Taiko Supreme 1B-activated clay from 1 to 5 hr decreased the 3-MCPD-E concentration to 4.1
C 2019 Institute of Food Technologists® Vol. 18, 2019 r Comprehensive Reviews in Food Science and Food Safety 353
Table 4–Effectivity of various deodorization conditions for the mitigation of 2-, 3-MCPD esters, and glycidyl esters.
354 Comprehensive Reviews in Food Science and Food Safety r Vol. 18, 2019
DEO, 1 to 5 mmol/kg
Addition of diacetin during DEO, before ND 50% [ND] ND ND Matthäus and Pudel (2013)
evaporation, amount ND
6% w/w rosemary extracts; 240 °C (1 No additives; 240 °C (1 hour); 82% [2.44 → 0.43] ND ND Zhang et al. (2016)
hour); 0.5 kPa; nitrogen strip-gas 0.5 kPa; nitrogen strip-gas
6% w/w lipophilic tea polyphenols; 240 No additives; 240 °C (1 hour); 75% [2.44 → 0.61] ND ND Zhang et al. (2016)
°C (1 hour); 0.5 kPa; nitrogen strip-gas 0.5 kPa; nitrogen strip-gas
1.8 mg/g TBHQ in palm oil; 4% 4% sn-1,2-dipalmitin; 200 °C (1 hour) ND ND 53% [1.7 → 0.8] Cheng et al. (2017)
sn-1,2-dipalmitin; 200 °C (1 hour)
Nitrogen atmosphere; 200 °C (1 hour) Regular air; 200 °C (1 hour) ND ND 41% [1.7 → 1] Cheng et al. (2017)
Short-path distillation; 0.001 mbar; Regular deodorization; 260 °C 90% [3.0 → 0.3] ND 98% [6.4 → LOD (<0.1)] Pudel et al. (2016)
condenser temp 60 °C; evaporator (1.5 hours); 2 to 3 mbar
temp 170 °C; stirrer speed 100 rpm;
pump frequency 20 Hz; additional
postdistillation deodorization at
180 °C (2 hours)
ND, not defined.
Figure 1–Schematic temperature-time profile of the typical oil refining process. (A) shows the normal process, (B) shows a high-low temperature
profile during deodorization, (C) shows a low-high temperature profile during deodorization.
mg/kg in chemically refined palm oil. The Gly-E concentration E during deodorization. Pudel et al. (2011) performed a labo-
remained unchanged at 0.4 mg/kg (Hrncirik & van Duijn, 2011). ratory experiment in which palm oil was deodorized in a con-
Craft, Nagy, Sandoz, and Destaillats (2012) reported a consid- trolled environment (Pudel et al., 2011). Samples were collected at
erable increase in Gly-E concentration when they deodorized a different combinations of temperatures and deodorization times,
commercially available bleached palm oil at 240 °C in compari- varying from 200 °C to 290 °C and from 90 to 360 min, respec-
son with both 200 °C and 220 °C. At both 200 °C and 220 °C, tively. They found approximately 2.5 mg/kg Gly-E (expressed
the concentration was 0.4 mg/kg Gly-E, while it was 1.7 mg/kg as 3-MCPD-E and related compounds) and no 3-MCPD-E in
Gly-E at 240 °C (an increase of 76%) (Craft et al., 2012). Similar the samples that were deodorized at 240 °C for 240 and 360
results were reported by Hrncirik and van Duijn (2011). Gly- min (Pudel et al., 2011). The highest Gly-E concentration was
E concentration was increased by 76% (from 0.4 to 1.7 mg/kg) found in a sample that had been deodorized at 290 °C for 120
in a physically refined palm oil, which has been deodorized for min (53 mg/kg 3-MCPD-E and related compounds, equal-
3 hr at 230 °C (Hrncirik & van Duijn, 2011). These results ing 48 mg/kg Gly-E). The highest 3-MCPD-E concentration
showed the temperature dependency of the formation of Gly- (6 mg/kg 3-MCPD-E) was found in a sample that was
C 2019 Institute of Food Technologists® Vol. 18, 2019 r Comprehensive Reviews in Food Science and Food Safety 355
3-MCPD & glycidol mitigation strategies . . .
Figure 2–Schematic representation of physical and chemical oil refining. (R.S. Zeldenrust, AOCS Lipid Library, reproduced from lipidlibrary.aocs.org
accessed on February 15, 2018).
deodorized at 250 °C for 90 min. Notably, in the oil sample ment with a very specific diolein oil and an artificial chloride
deodorized at 290 °C, but for a prolonged period of 240 and source.
360 min, the Gly-E concentration decreased to 8 mg/kg and 3 Matthäus and Pudel (2013) studied two double-deodorization
mg/kg, respectively (a reduction of >83%) (Pudel et al., 2011). It strategies and compared those to two standard single-
was suggested that possibly some precursors had been degraded or deodorization conditions (Table 4). In the first double-
distilled at such high temperatures. deodorization strategy, oil was first deodorized at 200 °C for
Several authors tested double-deodorization strategies in which 120 min, followed by a short temperature boost at 250 °C for
oil was deodorized twice, but with different temperature-time 5 min. A decrease of 65% was observed for 3-MCPD-E (from
combinations, to minimize the formations of MCPD-E and Gly- 2 mg/kg to 0.7 mg/kg) and 35% for Gly-E (from 2 mg/kg to 1.3
E. Figures 1(B) and (C) illustrate the possible implementation of mg/kg) (Matthäus & Pudel, 2013). In the second strategy, oil was
this approach in a typical refining process. Shimizu, Weitkamp, first deodorized at 200 °C for 120 min followed by a second de-
Vosmann, and Matthäus (2013) performed double-heating exper- odorization at 270 °C for 5 min. A reduction of 78% was observed
iments with diolein oil prepared from oleic acid and glycerol, for Gly-E (from 16 mg/kg to 3.5 mg/kg) and 3-MCPD-E, which
in which 10 mg/kg chloride per kg oil (as t-butylammonium was reduced by approximately 75% (from 2 mg/kg to 0.5 mg/kg)
chloride) was added (Shimizu et al., 2013). As a control, the oil (Matthäus & Pudel, 2013). These results were the opposite with
was heated at 240 °C for 4 hr, which resulted in a 3-MCPD- respect to those by Shimizu et al. (2013). A direct comparison
E concentration of 1.6 mg/kg and a Gly-E concentration of between the two studies cannot be made because of too many
120 mg/kg. Applying one of the two double-heating strategies differences between the experimental designs, including the tem-
(first at 180 °C for 4 hr, followed by 240 °C for 1 hour) resulted peratures, starting concentration in the oils, heating durations, and
in a decrease of 3-MCPD-E (1.6 to 1.5 mg/kg) and a 17% de- oil compositions. Nonetheless, both studies showed that double-
crease (120 to 100 mg/kg) of Gly-E. Reversing the temperature deodorization might be an effective mitigation strategy. Matthäus
profile from low-high to high-low with similar durations resulted and Pudel (2013) performed their double-deodorization experi-
in a small decrease of 3-MCPD-E. However, using the high-low ments under more realistic conditions than Shimizu et al. (2013)
temperature profile (240 °C for 1 hr, followed by 180 °C for 4 did, but additional confirmatory studies need to be performed
hours), the Gly-E concentration was reduced from 120 mg/kg to with true crude vegetable oil in an industrial refining setting.
50 mg/kg, which corresponds to a reduction of 58% compared Lower deodorization temperature can be beneficial for the final
with the low-high profile. These results showed that 3-MCPD-E concentrations of 3-MCPD-E and Gly-E.
formation was not affected by the deodorization temperature and Effects of additives during deodorization. Another mitigat-
that Gly-E was not stable under heating conditions. There is a fine ing approach is the use of additives in the oil. The ef-
balance between the rate of formation and the rate of degradation fect of using additives has been tested mostly for 3-MCPD-
of Gly-E under high temperatures. 3-MCPD-E is very heat stable E and to a limited extent for Gly-E. No studies have dealt
as shown by the unchanged concentration throughout the heating with 2-MCPD-E. Matthäus and Pudel (2013) mentioned three
experiments and a stability study conducted by these authors. The additives (ethanol, glycerol, and diacetin) that led to a reduction of
authors suggested that the reduction of Gly-E could be caused 3-MCPD-E and/or Gly-E, and their data were reported as per-
by distillation of Gly-E and transformation at low temperatures. centages of reduction (Matthäus & Pudel, 2013). The addition of
An important side note is that Shimizu et al. (2013) performed 0.5% to 2.5% of either ethanol or glycerol into bleached oil prior
this double-deodorization experiment in a laboratory environ- to deodorization reduced the 3-MCPD-E concentration by 30%.
356 Comprehensive Reviews in Food Science and Food Safety r Vol. 18, 2019
C 2019 Institute of Food Technologists®
3-MCPD & glycidol mitigation strategies . . .
Similar results were obtained by Craft et al. (2012). The supposed is its distillation chamber, which is a double-walled glass cylinder.
mechanism consists of the alcohol acting as a chloride scavenger By letting the oil flow down the inner wall, the temperature can
and forming volatile chlorinated adducts that can be stripped away be regulated more carefully. The thin layer of flowing oil formed
during deodorization. Adding 1 to 5 mmol/kg carbonates (either increases the surface area of the oil remarkably for an improved
potassium hydrogen carbonate or sodium bicarbonate) helps to distillation of volatile compounds. Under optimal short-path dis-
prevent the formation of 3-MCPD-E and Gly-E, resulting in a tillation conditions, a reduction of 90% and 98% in 3-MCPD-E
66% decrease of the concentrations (Matthäus & Pudel, 2013). and Gly-E concentrations, respectively, was observed. Under stan-
The third additive that Matthäus and Pudel mentioned is diacetin. dard deodorization conditions, 3.0 mg/kg of 3-MCPD-E and 6.4
Diacetin is a short-chained diacylglycerol (DAG) which can com- mg/kg of Gly-E were formed, while 3-MCPD-E and Gly-E were
pete with other DAGs for the available chlorine. Since diacetin is reported to be <0.1 mg/kg when short-path distillation was ap-
more volatile than other DAGs, it can be completely removed dur- plied (Pudel et al., 2016). However, Pudel et al. (2016) mentioned
ing deodorization. A 50% reduction of 3-MCPD-E was observed that the taste and odor were negatively influenced by short-path
after the addition of diacetin (Matthäus & Pudel, 2013). distillation. Furthermore, the color remained orange-red, which
Zhang et al. (2016) used four different anti-oxidants (α- may not be desirable for applications in some food products, al-
tocopherol, rosemary extract, lipophilic tea polyphenols, and L- though red color indicates the oil is still rich in carotenes, toco-
ascorbyl palmitate) as additives to bleached palm oil. With this pherols, and phytosterols. An extra postdistillation deodorization
mitigation strategy, they also proposed the involvement of free step under mild conditions (180 °C, 120 min) was able to resolve
radicals in the formation of 3-MCPD-E and Gly-E. The antioxi- the taste and odor issue. 3-MCPD-E concentration after deodor-
dants can act as free radical scavengers and prevent the formation ization increased to 0.3 mg/kg, while no increase in Gly-E was
of contaminants. Rosemary extract at 6% w/w (carnosic acid detected.
or rosmarinic acid) significantly reduced 3-MCPD-E concentra- In conclusion, several mitigation strategies during deodoriza-
tion (82.4%, from 2.44 mg/kg in the control to 0.43 mg/kg) tion, such as double deodorization, addition of various antioxi-
(Zhang et al., 2016). However, rosemary extract can have a strong dants, or use of lower deodorization temperatures for an extended
odor, which may add unwanted fragrance to the oil. Lipophilic period, have been shown to reduce the formation of 3-MCPD-E
tea polyphenols (6%, w/w of (-)-epigallocatechin gallate) were and Gly-E.
able to reduce the concentration of 3-MCPD-E by 75% (from
2.44 mg/kg to 0.61 mg/kg). Postrefining removal strategies
More recently, Cheng, Liu, and Liu (2017) investigated the ef- The strategies described in this section are designed to remove
fect of the artificial antioxidant tert-butyl hydroquinone (TBHQ) the process contaminants from already refined oil. Details of the
on the formation of Gly-E in palm oil (prepared in the labora- experimental conditions and concentrations reported in the liter-
tory), camellia oil, soybean oil, and linseed oil. Under laboratory ature are presented in Table 5.
conditions, the addition of an increasing amount of TBHQ re- Adsorbing contaminants onto adsorbents, such as bleaching clays
sulted in an increasing reduction of Gly-E. With the addition of or activated carbon, is a common strategy. This strategy for remov-
1.8 mg/g TBHQ to palm oil, a reduction of Gly-E concentra- ing contaminants is used with bleaching during oil refining but
tion of approximately 53% was achieved, from 1.7 mg/kg Gly-E can also be applied to refined oil. Since most of the undesirable
in the control compared with 0.8 mg/kg in palm oil with 1.8 components in crude oil are already removed during the refining
mg/g TBHQ (Cheng et al., 2017). The results observed for the process, the addition of adsorbents to refined oils might improve
other three oils were comparable to the results with palm oil. The the removal of process contaminants. In this way, the newly added
combination of the results of Zhang et al. (2016) and Cheng et al. adsorbents are more readily available to adhere to 2-, 3-MCPD-E,
(2017) suggested the involvement of a free radical-mediated for- and Gly-E.
mation of 3-MCPD-E and Gly-E. This can be viewed as a possible Strijowski, Heinz, and Franke (2011) tested several adsorbents
additional pathway to those already known. Cheng et al. (2017) such as amorphous magnesium silicate, zeolite, and synthetic mag-
also compared the difference between deodorization under nitro- nesium silicate (Strijowski et al., 2011). The adsorbent was added
gen or regular air. Since nitrogen is nonreactive, it should prevent to the refined palm oil at 80 °C and left to react for 30 min.
any oxygen-based lipid oxidation from occurring. Gly-E concen- The adsorbent was then separated from the oil by means of cen-
tration was markedly reduced by 41% when all four types of oils trifugation. Interestingly, five out of the nine tested adsorbents
were heated under nitrogen (from 1.7 mg/kg under air to 1.0 increased the concentrations of Gly-E to >1 mg/kg, while 3-
mg/kg under nitrogen). No significant differences were observed MCPD-E seems to be less negatively affected (data not shown).
between the four oil types. Two adsorbents, calcinated zeolite and synthetic magnesium sili-
Effects of alternative equipment geometry. Temperature and cate (65% silicon oxide and 15% magnesium oxide), gave promis-
additives as mitigation parameters could be easily implemented ing results. Use of 10% synthetic magnesium silicate resulted in
in existing refining protocols. Adjusting the design of the an insignificant reduction (5%) of the 3-MCPD-E concentra-
deodorizer is a more challenging mitigation strategy. Pudel, tion (from 4.3 to 4.1 mg/kg), while the Gly-E concentration
Benecke, Vosmann, and Matthäus (2016) suggested the use was markedly reduced by 41% (from 2.2 to 0.9 mg/kg). Both
of the short-path distillation process as a replacement for 3-MCPD-E and Gly-E concentrations were markedly reduced
the conventional deodorization process (Pudel et al., 2016). when 10% calcinated zeolite was used. A 19% reduction of 3-
Short-path distillation allows for a gentler removal of volatile MCPD-E (from 4.3 to 3.5 mg/kg) was observed in comparison
compounds without the need to heat the oil at high tempera- to an untreated palm oil control. Gly-E concentration was re-
tures. The vacuum applied during short-path distillation is ap- duced by 77% (from 2.2 mg/kg to 0.5 mg/kg). Strijowski et al.
proximately 10−3 mbar (compared with 2 to 4 mbar in the con- (2011) also conducted sensory and quality tests (oxidation stabil-
ventional process), which reduces the boiling point of the volatile ity) on the treated oils. The use of calcinated zeolite improved the
compounds. The main difference with a conventional deodorizer sensory quality of the oil without leading to a worsening of the
C 2019 Institute of Food Technologists® Vol. 18, 2019 r Comprehensive Reviews in Food Science and Food Safety 357
3-MCPD & glycidol mitigation strategies . . .
Table 5–Effectivity of postrefining processes for the mitigation of 2-, 3-MCPD esters, and glycidyl esters.
oxidation stability. On the other hand, the synthetic magnesium Shimizu et al. (2012) for converting Gly-E to acylglycerols
silicate resulted in oil with poor sensory quality, but improved and glycerols. The main principle is the conversion of un-
oxidation stability. bound 3-MCPD into glycidol by halohydrin dehalogenase, fol-
Another adsorbent that is capable of efficiently removing Gly- lowed by hydrolyzing glycidol into glycerol by epoxide hydrolase
E is acid-washed oil palm wood-based activated carbon (OPAC). (Bornscheuer & Hesseler, 2010). 3-MCPD-E can also be con-
OPAC is an activated carbon made from oil palm wood. The verted into glycerol by the same enzymes, but it requires unbound
activated carbon is washed with acid to achieve high mesoporosity 3-MCPD to initiate the reaction. A lipase was used to hydrolyze
and an acidic environment on its surface. Cheng et al. (2017) 3-MCPD-E into unbound 3-MCPD. In a biphasic system at 5%
showed Gly-E removal efficacy of acid-washed OPAC as part of a v/v water-in-oil ratio, the lipase was able to convert 100% of 3-
kinetic and mechanistic study of the adsorptive capacity of OPAC MCPD oleate esters into 3-MCPD. In a separate experiment using
(Cheng, Liu, Wang, & Han, 2017). At an acid-washed OPAC 5% v/v water-in-oil ratio, the halohydrin dehalogenase was able
concentration of 30 mg/100 mL, a 95% reduction of the level of to convert 100% of the 3-MCPD into glycidol, followed by the
Gly-E was observed (3.75 to 0.2 mg/kg). This adsorbent showed complete conversion (100%) of glycidol into glycerol by epox-
the best reduction of Gly-E in comparison with other adsorbents ide hydrolase. However, in an experiment by Bornscheuer and
including alkaline cellulase, activated clay, and non-acid-washed Hesseler (2010) the concentration of the formed glycidol (7.7
OPAC. Equilibrium was attained after 40 min and the maximum mM) did not correspond with the starting concentration of 3-
adsorption capacity was found to be 36.2 mg Gly-E/g acid-washed MCPD (10 mM). The mechanism of the enzymatic conversion
OPAC. Cheng et al. (2017) mentioned that the mode of Gly-E could not completely be explained by the observed data. Never-
removal was due to both adsorption on the acid-washed OPAC theless, no 3-MCPD could be detected after incubation with halo-
and degradation of Gly-E at activated sites with acidic character. hydrin dehalogenase, nor after incubation with epoxide hydrolase.
No difference in oil quality was observed between before and after It is important to note that the experiments of Bornscheuer and
treatments with acid-washed OPAC, making it a suitable adsorbent Hesseler (2010) were performed under laboratory conditions. The
to adopt in mitigation strategies. However, since the effect of acid- use of enzymes to remove 3-MCPD from vegetable oil is promis-
washed OPAC on the 2- and 3-MCPD-E concentrations was not ing, but it requires more research before the processed oil can be
tested, more experiments need to be performed before it can be regarded as safe for consumption. The safety of degradation prod-
used as an adsorbent for 2-, 3-MCPD-E, and Gly-E. ucts due to enzymatic treatment and the remnants of the enzyme
Shimizu et al. (2012) used acid-activated bleaching clay to re- itself must be investigated. In conclusion, several postrefining re-
move Gly-E from different types of vegetable oils including palm moval strategies are available, including the use of absorbents and
oils and rice bran oils (Shimizu et al., 2012). In all tested oils, enzymes, and have been shown to be effective in removing pro-
the concentration of Gly-E was reduced by 99% or more when cess contaminants from fully refined vegetable oils. 3-MCPD-E
using 1% acid-activated bleaching clay. As an example, the sum concentrations were reduced up to 19% when refined oil was
of all measured Gly-E concentrations (which consisted of palmi- treated with an adsorbent (calcinated zeolite). This method was
tate, stearate, oleate, linoleate, and linolenate esters) was reduced more effective for Gly-E. Finally, postrefining treatment with 1%
from an average of 10.3 mg/kg to below the LOQ of 0.1 mg/kg activated bleaching clay (V2R) was most efficient for Gly-E, with
(Shimizu et al., 2012). Those authors noted that glycidyl palmitate a reduction of 99%.
was modified instead of being absorbed by the activated bleach-
ing clay. An epoxide ring-opening reaction of glycidyl palmitate Combining strategies
with water can occur prior to the interesterification reaction be- Oil refining is a complex process with multiple possibilities
tween monopalmitate and glycerol dioleate. The acidic condition for mitigation strategies to be applied during the process, as dis-
of acid-activated bleaching clay can play a key role in initiating the cussed above. Therefore, a combined mitigation strategy may be
reaction. The end products of the reactions were reported to be promising for limiting the final concentrations of 2-, 3-MCPD-E,
simple acylglycerols and glycerol. and Gly-E in refined oil. For example, Zulkurnain et al. (2013)
A comparable approach to reduce contaminants is the break- applied a modified refining process that was optimized using
down of targeted contaminants with enzymes, as shown by response surface methodology (RSM) in addition to their
358 Comprehensive Reviews in Food Science and Food Safety r Vol. 18, 2019
C 2019 Institute of Food Technologists®
3-MCPD & glycidol mitigation strategies . . .
previously discussed optimization steps (Zulkurnain et al., 2013). ing reported in the reviewed publications. It is known that high
For their RSM optimization, five refining parameters were com- temperature is the main cause for the formation of Gly-E (Destail-
bined. These five parameters are water dosage, phosphoric acid lats et al., 2012). Gly-E formation occurs at a much higher tem-
dosage, degumming temperature, activated clay dosage, and de- perature (>220 °C) than the temperatures at which 2- and 3-
odorization temperature. Their optimized degumming conditions MCPD-E can start to form (Matthäus & Pudel, 2013; Shimizu
were 3.5% water dosage, 0.08% phosphoric acid, and 60 °C et al., 2013) Notably, the temperature applied during degumming,
degumming temperature. Combined with 0.3% of bleaching clay neutralization, and bleaching processes are usually well below 180
and 260 °C as the deodorization temperature, they were able to °C. Therefore, we assume that the missing Gly-E concentrations
reduce the 3-MCPD-E concentration by 87.2%, from 2.95 to for the degumming, neutralization, and bleaching processes are
0.37 mg/kg (Zulkurnain et al., 2013). Some parameters showed not crucial, because it is unlikely that a large amount of Gly-E
contradictory effects, something that can also be seen from several would have already been formed during the mentioned processes
previously discussed strategies. It is important to find the balance prior to the deodorization process. However, this assumption can
between reducing the process contaminants as much as possible only be confirmed if Gly-E concentrations become available from
(3-MCPD-E in this particular case) while maintaining the oil these stages of the refining process.
quality. This review focused on an in-depth analysis of mitigation strate-
gies reported in peer reviewed publications. Most of these studies
General aspects are reported by academic and applied research organizations. From
Designing an efficient mitigation strategy requires more than these studies, it is not clear to what degree these strategies were
one modification in the entire refining process. Many papers have implemented by industry and if they are routinely applied.
focused on a single parameter or a single refining stage, which
may be sufficient for high-quality starting oils, but could still be
Conclusion
inadequate for initial low-quality oils. Degumming, neutralization,
The current review shows that degumming, neutralization,
bleaching, and deodorization are potential targets for the devel-
bleaching, and deodorization are the main processing steps in oil
opment of a multifaceted mitigation strategy. Postrefining removal
refining with the potential to mitigate the formation of 3-MCPD-
could be utilized when dealing with low-grade oils in which resid-
E and Gly-E. Water degumming was able to reduce 3-MCPD-E
ual process contaminants are most likely to still be present. With
concentrations by 84% and by 26% for Gly-E. Neutralization of
the recent MLs set by the European Commission for Gly-E, and
oil with a base such as NaHCO3 or KOH was able to reduce the
possible future regulations for 3-MCPD-E, it is worthwhile for
3-MCPD-E concentration up to 81% and up to 84% for Gly-E.
oil refineries to invest in the development of sound mitigation
Bleaching of oil with neutral clay after a water degumming treat-
strategies over the entire refining process.
ment was able to reduce the 3-MCPD-E concentration by 46%.
Providing a clear, ready-made solution in terms of combining
Magnesol R60 (magnesium silicate) reduced 3-MCPD-E by 67%.
mitigation strategies is subject to several limitations. Crude oil
Deodorizing oil at high temperatures for prolonged periods (for
quality plays a significant role, but also the design of the indi-
example, >230 °C to 250 °C for >1 hr) increased Gly-E con-
vidual refining plant could limit the possibilities. Minor hardware
centration, while 2- and 3-MCPD-E concentrations were more
changes or small changes in standard operating procedures are
dependent on the amount of available chlorine. Other less com-
usually not problematic. However, addition of a water wash step,
mon deodorization methods, including double-deodorization or
varying the amount of acid or bleaching clay, performing dou-
deodorization with certain additives, are very promising mitiga-
ble deodorization, or even double refining is more difficult to
tion strategies that can reduce 3-MCPD-E concentration up to
achieve; nonetheless, it is still manageable. When these changes
82% and Gly-E concentration up to 78%. However, there is no
are optimized, they could significantly decrease contaminants.
effective “one-size-fits-all” single point solution for oil refining.
The availability of resources may also be a limitation for imple-
It is advisable to develop a mitigation strategy combining multiple
mentation. Although time and financial considerations have not
mitigation steps due to the nature of the different reaction mech-
been discussed in this review, they are often the major limita-
anisms of formation. Combination of water washing in order to
tions. Optimizing a single parameter and validating a new method
remove chlorine precursors, pH neutralization prior to any exten-
can be temporally and financially cumbersome for oil refining
sive heat treatments, and double-deodorization with or without
companies. Most mitigation tests are performed in the laboratory
additives to adsorb remaining contaminants can be very effective
and would require a scale-up of research and operation costs for
in reducing 2-, 3-MCPD-E, and/or Gly-E in vegetable oils. The
implementation in a production plant.
combined strengths and specificities of each mitigation strategy can
The mitigation strategies presented show reductions ranging
result in good-quality oil with very low concentration of process
from 4% up to 94% for 3-MCPD-E, and 16% up to 100% for Gly-
contaminants. Noticeably, none of the discussed papers provided
E. Unfortunately, none of the 20 papers reported data regarding
data on 2-MCPD-E. This lack of data cannot be ignored when
2-MCPD-E. This clearly shows that there is a significant gap in
new regulations and/or mitigation strategies are being developed.
the data and knowledge concerning the mitigation of 2-MCPD-
E. It is important to also include 2-MCPD in future studies in
order to prevent reduction of one compound while increasing the Acknowledgments
formation of another. S.B. Oey, S.P.J. van Leeuwen, and H.J. van der Fels-Klerx re-
When, for example, a certain mitigation strategy is able to suc- ceived funding from the Dutch Ministry of Agriculture, Nature
cessfully reduce Gly-E (or one of the other process contaminants) and Food Quality through the Topsector project TKI-AF-16002/
with a certain processing condition, it might be difficult to de- BO-46-002-021. The authors acknowledge the financial contri-
termine whether Gly-E is being physically removed from the bution of three private partners (Special Refining Comp. B.V.,
oil or if it is being converted or degraded into something else. Care Naturkost GmbH and Co. KG, and Spack Trading B.V.)
Finally, we observed that Gly-E concentrations are not always be- involved in this project.
C 2019 Institute of Food Technologists® Vol. 18, 2019 r Comprehensive Reviews in Food Science and Food Safety 359
3-MCPD & glycidol mitigation strategies . . .
References
Author Disclosures
Bakhiya, N., Abraham, K., Gürtler, R., Appel, K. E., & Lampen, A. (2011).
All authors have read and approved the manuscript. Authors Toxicological assessment of 3-chloropropane-1,2-diol and glycidol fatty acid
declare to not have any conflict of interest. esters in food. Molecular Nutrition and Food Research, 55(4), 509–521.
https://doi.org/10.1002/mnfr.201000550
Bornscheuer, U. T., & Hesseler, M. (2010). Enzymatic removal of
Appendix 3-monochloro-1,2-propanediol(3-MCPD) and its esters from oils. European
Chemical and Physical Refining Journal of Lipid Science and Technology, 112(5), 552–556.
Oil refining can be performed via a chemical refining or physical https://doi.org/10.1002/ejlt.200900245
refining process. There are several principal differences between Cheng, W., Liu, G., Wang, L., & Liu, Z. (2017). Glycidyl fatty acid esters in
refined edible oils: A review on formation, occurrence, analysis, and
chemical and physical refining with related advantages and disad- elimination methods. Comprehensive Reviews in Food Science and Food Safety,
vantages. The major difference is that during the chemical refining 16(2), 263–281. https://doi.org/10.1111/1541-4337.12251
process, free fatty acids (FFAs) are removed chemically by saponi- Cheng, W., Liu, G., & Liu, X. (2017). Effects of Fe3+ and antioxidants on
fication with an alkali solution, which results in the formation of glycidyl ester formation in plant oil at high temperature and their
soap. The separated soap stock can be sold for the making of soap influencing mechanisms. Journal of Agricultural and Food Chemistry, 65(20),
4167–4176. https://doi.org/10.1021/acs.jafc.7b00858
or biodiesel (Piloto-Rodriguez, Melo, Goyos-Perez, & Verhelst, Cheng, W., Liu, G., Wang, X., & Han, L. (2017). Adsorption removal of
2014). Physical refining removes the FFAs in a later stage during glycidyl esters from palm oil and oil model solution by using acid-washed
deodorization by means of vacuum steam distillation. oil palm wood-based activated carbon: Kinetic and mechanism study.
Figure 2 shows a schematic illustration of both processes. The Journal of Agricultural and Food Chemistry, 65(44), 9753–9762.
https://doi.org/10.1021/acs.jafc.7b03121
choice between the two refining processes depends on the type
Craft, B. D., Nagy, K., Sandoz, L., & Destaillats, F. (2012). Factors impacting
and quality of the crude oil that needs to be refined. Unwanted the formation of monochloropropanediol (MCPD) fatty acid diesters during
products are easier to remove during the alkali neutralization step palm (Elaeis guineensis) oil production. Food Additives and Contaminants -
in chemical refining. However, physical refining can be applied Part A Chemistry, Analysis, Control, Exposure and Risk Assessment, 29(3),
354–361. https://doi.org/10.1080/19440049.2011.639034
more widely to various types of oils and fats than chemical refin-
Craft, B. D., Nagy, K., Seefelder, W., Dubois, M., & Destaillats, F. (2012).
ing (Gibon et al., 2007). In the case of physical refining, the crude Glycidyl esters in refined palm (Elaeis guineensis) oil and related fractions.
oil is not washed or neutralized before it is deodorized. It is there- Part II: Practical recommendations for effective mitigation. Food Chemistry,
fore preferable to start with a higher quality of oil with limited 132(1), 73–79. https://doi.org/10.1016/j.foodchem.2011.10.034
levels of unwanted compounds such as FFAs, phospholipids, met- Destaillats, F., Craft, B. D., Sandoz, L., & Nagy, K. (2012). Formation
als, and soaps. Optimized degumming and bleaching steps during mechanisms of monochloropropanediol (MCPD) fatty acid diesters in
refined palm (Elaeis guineensis) oil and related fractions. Food Additives &
physical refining helps to increase the general oil quality before Contaminants: Part A, 29(1), 29–37.
deodorization. https://doi.org/10.1080/19440049.2011.633493
It is advantageous to utilize physical refining because the final oil Destaillats, F., Craft, B. D., Dubois, M., & Nagy, K. (2012). Glycidyl esters in
yield is often higher than that of chemical refining. Furthermore, refined palm (Elaeis guineensis) oil and related fractions. Part I: Formation
mechanism. Food Chemistry, 131(4), 1391–1398.
less chemical treatment is needed and there is no soap stock formed https://doi.org/10.1016/j.foodchem.2011.10.006
that needs to be disposed or processed properly, providing a better European Commission. (2018). Amending Regulation (EC) 1881/2006 as
environmental profile. The downside is that more bleaching clay regards maximum levels of glycidyl fatty acid esters in vegetable oils and fats, infant
is needed to purify the oil prior to the deodorization step. Also, formula, follow-on formula and foods for special medical purposes intended for infants
with physical refining the complete deodorization process, often and young children. Retrieved from
https://eur-lex.europa.eu/eli/reg/2018/290/oj
performed at temperatures up to 270 °C, can destroy natural anti- European Food Safety Authority, E. (2013). Analysis of occurrence of
oxidants such as tocopherols and tocotrienols. Careful monitoring 3-monochloropropane-1, 2-diol (3-MCPD) in food in Europe in the years
of temperature and time is therefore important if such minor 2009–2011 and preliminary exposure assessment. EFSA Journal, 11(9),
components are needed to be preserved. 3381. https://doi.org/10.2903/j.efsa.2013.3381
Although Gibon et al. (2007) reported oil losses ranging be- Franke, K., Strijowski, U., Fleck, G., & Pudel, F. (2009). Influence of
chemical refining process and oil type on bound 3-chloro-1,2-propanediol
tween 20% and 30% during chemical refining, it remains the bet- contents in palm oil and rapeseed oil. LWT - Food Science and Technology,
ter solution in terms of removal of unwanted products. The alkali 42(10), 1751–1754. https://doi.org/10.1016/j.lwt.2009.05.021
neutralization applied during chemical refining is better at remov- Freudenstein, A., Weking, J., & Matthäus, B. (2013). Influence of precursors
ing unwanted products than the acid-degumming step performed on the formation of 3-MCPD and glycidyl esters in a model oil under
simulated deodorization conditions. European Journal of Lipid Science and
during physical refining. Oils high in phospholipids, such as sun- Technology, 115(3), 286–294. https://doi.org/10.1002/ejlt.201200226
flower oil and rapeseed oil, also benefit more from water degum- Gibon, V., De Greyt, W., & Kellens, M. (2007). Palm oil refining. European
ming applied during chemical refining, as water degumming is Journal of Lipid Science and Technology, 109(4), 315–335.
able to efficiently remove these phospholipids. As the majority of https://doi.org/10.1002/ejlt.200600307
the FFAs are already removed during the alkali-neutralization step, Habermeyer, M., Guth, S., & Eisenbrand, G. (2011). Identification of gaps in
a less vigorous deodorization step can be applied, and therefore knowledge concerning toxicology of 3-MCPD and glycidol esters. European
Journal of Lipid Science and Technology, 113(3), 314–318.
preserve the previously mentioned minor components. In terms of https://doi.org/10.1002/ejlt.201000314
strategies to mitigate 2-, 3-MCPD esters, and glycidyl esters, both Hamlet, C. G., Asuncion, L., Velisek, J., Dolezal, M., Zelinková, Z., &
the physical and chemical refining should be considered. The focus Crews, C. (2011). Formation and occurrence of esters of
360 Comprehensive Reviews in Food Science and Food Safety r Vol. 18, 2019
C 2019 Institute of Food Technologists®
3-MCPD & glycidol mitigation strategies . . .
3-chloropropane-1,2-diol (3-CPD) in foods: What we know and what we activated bleaching earth. Journal of Oleo Science, 61(1), 23–28.
assume. European Journal of Lipid Science and Technology, 113(3), 279–303. https://doi.org/10.5650/jos.61.23
https://doi.org/10.1002/ejlt.201000480 Shimizu, M., Weitkamp, P., Vosmann, K., & Matthäus, B. (2013).
Hrncirik, K., & van Duijn, G. (2011). An initial study on the formation of Temperature dependency when generating glycidyl and 3-MCPD esters
3-MCPD esters during oil refining. European Journal of Lipid Science and from diolein. Journal of the American Oil Chemists’ Society, 90(10),
Technology, 113(3), 374–379. https://doi.org/10.1002/ejlt.2010 1449–1454. https://doi.org/10.1007/s11746-013-2298-9
00317 Silva, S. M., Sampaio, K. A., Ceriani, R., Verhé, R., Stevens, C., De Greyt,
Jedrkiewicz, R., Głowacz, A., Gromadzka, J., & Namie, J. (2016). W., & Meirelles, A. J. A. (2014). Effect of type of bleaching earth on the
Determination of 3-MCPD and 2-MCPD esters in edible oils, fish oils and final color of refined palm oil. LWT - Food Science and Technology, 59(2P2),
lipid fractions of margarines available on Polish market. Food Control, 59, 1258–1264. https://doi.org/10.1016/j.lwt.2014.05.028
487–492. https://doi.org/10.1016/j.foodcont.2015.05.039 Šmidrkal, J., Tesařová, M., Hrádková, I., Berčı́ková, M., Adamčı́ková, A., &
Leigh, J., & MacMahon, S. (2017). Occurrence of 3-monochloropropanediol Filip, V. (2016). Mechanism of formation of 3-chloropropan-1,2-diol
esters and glycidyl esters in commercial infant formulas in the United States. (3-MCPD) esters under conditions of the vegetable oil refining. Food
Food Additives and Contaminants - Part A Chemistry, Analysis, Control, Chemistry, 211, 124–129. https://doi.org/10.1016/j.foodchem.2016.05.039
Exposure and Risk Assessment, 34(3), 356–370. Strijowski, U., Heinz, V., & Franke, K. (2011). Removal of 3-MCPD esters
https://doi.org/10.1080/19440049.2016.1276304 and related substances after refining by adsorbent material. European Journal
Malaysian Palm Oil Board. (2000). Poram standard specifications for processed of Lipid Science and Technology, 113(3), 387–392.
palm oil. Retrieved April 2, 2017, Retrieved from https://doi.org/10.1002/ejlt.201000323
http://econ.mpob.gov.my/economy/exporters/PORAMSTANDARD.htm Taylor, D. (2009). Chapter 3—Adsorbents. In G. R. List (Ed.), Bleaching and
Matthäus, B., & Pudel, F. (2013). Mitigation of 3-MCPD and glycidyl esters purifying fats and oils (pp. 69–95). Urbana, Illinois: AOCS Press.
within the production chain of vegetable oils especially palm oil. Lipid https://doi.org/10.1016/B978-1-893997-91-2.50009-2
Technology, 25(7), 151–155. https://doi.org/10.1002/lite.201300288 The Tintometer Ltd. (2016). Lovibond RYBN Colour. Retrieved February 13,
Matthäus, B., Pudel, F., Fehling, P., Vosmann, K., & Freudenstein, A. (2011). 2018, Retrieved from http://www.lovibondcolour.com/colour-scale/
Strategies for the reduction of 3-MCPD esters and related compounds in lovibondsupsup-rybn-colour
vegetable oils. European Journal of Lipid Science and Technology, 113(3), Weißhaar, R., & Perz, R. (2010). Fatty acid esters of glycidol in refined fats
380–386. https://doi.org/10.1002/ejlt.201000300 and oils. European Journal of Lipid Science and Technology, 112(2), 158–165.
Organization World Health. (2000). Some industrial chemicals. IARC https://doi.org/10.1002/ejlt.200900137
Monographs on the Evaluation of Carcinogenic Risks to Humans, 77, 469–486. World Health Organization. (2013). Some chemicals present in industrial and
Retrieved from https://monographs.iarc.fr/wp-content/uploads/2018/06/ consumer products, food and drinking-water. IARC Monographs on the
mono77-19.pdf Evalutaion of Carcinogenic Risks to Humans, 101, 349–374. Retrieved from
Piloto-Rodriguez, R., Melo, E. A., Goyos-Perez, L., & Verhelst, S. (2014). https://monographs.iarc.fr/wp-content/uploads/2018/06/mono101-010.pdf
Conversion of by-products from the vegetable oil industry into biodiesel Wöhrlin, F., Fry, H., Lahrssen-Wiederholt, M., & Preiß-Weigert, A. (2015).
and its use in internal combustion engines: A review. Brazilian Journal of Occurrence of fatty acid esters of 3-MCPD, 2-MCPD and glycidol in infant
Chemical Engineering, 31(2), 287–301. Retrieved from formula. Food Additives and Contaminants - Part A Chemistry, Analysis,
http://doi.org/10.1590/0104-6632.20140312s00002763 Control, Exposure and Risk Assessment, 32(11), 1810–1822.
Pudel, F., Benecke, P., Fehling, P., Freudenstein, A., Matthäus, B., & Schwaf, https://doi.org/10.1080/19440049.2015.1071497
A. (2011). On the necessity of edible oil refining and possible sources of Zhang, H., Jin, P., Zhang, M., Cheong, L., Hu, P., Zhao, Y., . . . Xu, X.
3-MCPD and glycidyl esters. European Journal of Lipid Science and Technology, (2016). Mitigation of 3-monochloro-1,2-propanediol ester formation by
113(3), 368–373. https://doi.org/10.1002/ejlt.201000460 radical scavengers. Journal of Agricultural and Food Chemistry, 64(29),
Pudel, F., Benecke, P., Vosmann, K., & Matthäus, B. (2016). 3-MCPD- and 5887–5892. https://doi.org/10.1021/acs.jafc.6b02016
glycidyl esters can be mitigated in vegetable oils by use of short path Zulkurnain, M., Lai, O. M., Latip, R. A., Nehdi, I. A., Ling, T. C., & Tan,
distillation. European Journal of Lipid Science and Technology, 118(3), 396–405. C. P. (2012). The effects of physical refining on the formation of
https://doi.org/10.1002/ejlt.201500095 3-monochloropropane-1, 2-diol esters in relation to palm oil minor
Ramli, M. R., Siew, W. L., Ibrahim, N. A., Hussein, R., Kuntom, A., Abd. components. Food Chemistry, 135(2), 799–805.
Razak, R. A., & Nesaretnam, K. (2011). Effects of degumming and https://doi.org/10.1016/j.foodchem.2012.04.144
bleaching on 3-MCPD esters formation during physical refining. Journal of Zulkurnain, M., Lai, O. M., Tan, S. C., Abdul Latip, R., & Tan, C. P. (2013).
the American Oil Chemists’ Society, 88(11), 1839–1844. Optimization of palm oil physical refining process for reduction of
https://doi.org/10.1007/s11746-011-1858-0 3-monochloropropane-1,2-diol (3-MCPD) ester formation. Journal of
Shimizu, M., Moriwaki, J., Shiiba, D., Nohara, H., Kudo, N., & Katsuragi, Agricultural and Food Chemistry, 61(13), 3341–3349.
Y. (2012). Elimination of glycidyl palmitate in diolein by treatment with https://doi.org/10.1021/jf4009185
C 2019 Institute of Food Technologists® Vol. 18, 2019 r Comprehensive Reviews in Food Science and Food Safety 361