Antioxidant activity of two edible isothiocyanates: Sulforaphane and erucin is due to their thermal decomposition to sulfenic acids and methylsulfinyl radicals

Article abstract of DOI:10.1016/j.foodchem.2021.129213

Sulforaphane (SFN) and erucin (ERN) are isothiocyanates (ITCs) bearing, respectively, methylsulfinyl and methylsulfanyl groups. Their chemopreventive and anticancer activity is attributed to ability to modulate cellular redox status due to induction of Phase 2 cytoprotective enzymes (indirect antioxidant action) but many attempts to connect the bioactivity of ITCs with their radical trapping activity failed. Both ITCs are evolved from their glucosinolates during food processing of Cruciferous vegetables, therefore, we studied antioxidant behaviour of SFN/ERN at elevated temperature in two lipid systems. Neither ERN nor SFN inhibit the oxidation of bulk linolenic acid (below 100 °C) but both ITCs increase oxidative stability of soy lecithin (above 150 °C). On the basis of GC-MS analysis we verified our preliminary hypothesis (Antioxidants 2020, 9, 1090) about participation of sulfenic acids and methylsulfinyl radicals as radical trapping agents responsible for the antioxidant effect of edible ITCs during thermal oxidation of lipids at elevated temperatures (above 140 °C).

Full text of DOI:10.1016/j.foodchem.2021.129213

Food Chemistry 353 (2021) 129213
Contents lists available at ScienceDirect
Food Chemistry
journal homepage: www.elsevier.com/locate/foodchem
Antioxidant activity of two edible isothiocyanates: Sulforaphane and erucin
is due to their thermal decomposition to sulfenic acids and
methylsulfinyl radicals
Jakub Cedrowskia, Kajetan Dąbrowab, Paweł Przybylskia, Agnieszka Krogul-Sobczaka,
,
*
Grzegorz Litwinienkoa
a University of Warsaw, Faculty of Chemistry, Pasteura 1, 02-093 Warsaw, Poland
b Polish Academy of Sciences, Institute of Organic Chemistry, Kasprzaka 44/52, 01-224 Warsaw, Poland
A R T I C L E I N F O
A B S T R A C T
Keywords:
Isothiocyanates
Sulforaphane
Erucin
Sulforaphane (SFN) and erucin (ERN) are isothiocyanates (ITCs) bearing, respectively, methylsulfinyl and
methylsulfanyl groups. Their chemopreventive and anticancer activity is attributed to ability to modulate
cellular redox status due to induction of Phase 2 cytoprotective enzymes (indirect antioxidant action) but many
attempts to connect the bioactivity of ITCs with their radical trapping activity failed. Both ITCs are evolved from
their glucosinolates during food processing of Cruciferous vegetables, therefore, we studied antioxidant behaviour
of SFN/ERN at elevated temperature in two lipid systems. Neither ERN nor SFN inhibit the oxidation of bulk
linolenic acid (below 100 ◦C) but both ITCs increase oxidative stability of soy lecithin (above 150 ◦C). On the
basis of GC-MS analysis we verified our preliminary hypothesis (Antioxidants 2020, 9, 1090) about participation
of sulfenic acids and methylsulfinyl radicals as radical trapping agents responsible for the antioxidant effect of
edible ITCs during thermal oxidation of lipids at elevated temperatures (above 140 ◦C).
Antioxidant
Radicals
Sulfenic acids
Oxidation
Lipids
1. Introduction
& Kostov, 2012; Fahey et al., 2001; Palliyaguru, Yuan, Kensler, & Fahey,
2018; Posner, Cho, Green, Zhang, & Talalay, 1994; Yuanfeng et al.,
– –
Isothiocyanates (ITCs), the organic compounds with –N
C
S
2020; Zhang, Talalay, Cho, & Posner, 1992), see Scheme 1B. Both ITCs
are the components of our diet when consuming common Cruciferous
– –
group, are sulfur analogues of isocyanates (R-NCO) and isomers of
thiocyanates (R-SCN). Numerous natural ITCs can be found in vegeta-
bles and due to their specific odour and taste they formerly were called
mustard oils. These compounds are present in plants in the form of
glucosinolates, ie., metabolites with S-β-D-glucopyrano unit anomeri-
cally connected to an O-sulfated (Z)-thiohydroximate function, see
ˇ
´
vegetables (Blazevic et al., 2020; Fahey et al., 2001; Fenwick et al.,
1983): SFN is present in broccoli, some cultivars of cabbage and Brussels
sprouts while ERN is predominantly found in rocket (Eruca vesicaria)
(Azarenko, Jordan, & Wilson, 2014; Fahey et al., 2001; Fahey, Zhang, &
Talalay, 1997; Fenwick et al., 1983).
ˇ
´
Scheme 1 (Blazevic et al., 2020; Fahey, Zalcmann, & Talalay, 2001;
Fenwick, Heaney, Mullin, & VanEtten, 1983). During cutting, slicing, or
grinding, they are enzymatically transformed giving a gentle, specific
flavor and taste. Formation of ITCs from glucosinolates occurs in the
presence of myrosinase (EC 3.2.3.1.147), a glycoprotein present in plant
tissues, and proceeds via Lossen rearrangement (Hayes, Kelleher, &
Eggleston, 2008). Recently, the great attention has been paid to
naturally-occurring ITCs with methylsulfinyl and methylsulfanyl moi-
eties, namely sulforaphane (4-methylsulfinylbutyl isothiocyanate, SFN)
and erucin (4-methylthiobutyl isothiocyanate, ERN) (Dinkova-Kostova
SFN and ERN are promising compounds for the use in medicines,
food supplements and cosmetics. Some dietary supplements contain
glucoraphanin with active myrosinase or stabilized “free” SFN (Avma-
col®, TrueBroc®, and Prostaphane®) and a few cosmetics with SFN are
accessible on the market, with the names relevant to broccoli and sul-
foraphane (BroccoFusion® Sulforaphane Lotion, Broccoli serum and
CBD/Sulforaphane serum). SFN is also being tested as a potential drug in
about 70 clinical trials (according to clinicaltrials.gov., date:
24.07.2020). Medical applications (eg., treating autism, psoriasis, and
bladder cancer) and cosmetic applications of ITCs are subject of
* Corresponding author.
E-mail address: litwin@chem.uw.edu.pl (G. Litwinienko).
https://doi.org/10.1016/j.foodchem.2021.129213
Received 17 November 2020; Received in revised form 5 January 2021; Accepted 24 January 2021
Available online 1 February 2021
0308-8146/© 2021 The Author(s).
Published by Elsevier Ltd.
This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).

J. Cedrowski et al.
Food Chemistry 353 (2021) 129213
numerous patents. Moreover, ERN is listed in the flavor library of Flavor
Extract Manufacturers Association (FEMA) as cabbage flavor (no. 4414).
Talalay and co-workers (Fahey et al., 1997; Posner et al., 1994;
Zhang et al., 1992) discovered and documented that among natural ITCs
sulforaphane is the most potent inducer of carcinogen detoxification
enzymes (eg., quinone reductase, glutathione-S-transferase, heme oxy-
genase) and SFN is one of the most potent phytochemical with the
inducing activity of cytoprotective enzymes (Fahey et al., 1997;
Yagishita, Fahey, Dinkova-Kostova, & Kensler, 2019; Zhang et al.,
1992). ITCs are considered as chemopreventive agents (Dinkova-Kos-
tova & Kostov, 2012; Hayes et al., 2008; Zhang et al., 1992) due to their
role in the induction Phase 2 enzymes of the xenobiotic metabolism (this
process involves Nrf2 signaling) and inhibition of Phase 1 enzymes. SFN
or ERN are also recognized as inducing some apoptotic pathways, anti-
inflammatory effects, inhibition of angiogenesis, and their H2S-donating
ability might be important in redox homeostasis (Cho, Lee, & Park,
2013; Citi et al., 2019; Houghton, 2019; Melchini & Traka, 2010).
Health benefits of dietary intake of SFN include prevention of different
types of cancer (eg., prostate, bladder and lung cancer), asthma,
schizophrenia, neurodegenerative diseases (such as Alzheimer and
Parkinson diseases), diabetes, cardiovascular diseases (CVDs), chronic
obstructive pulmonary disease (COPD) or autistic spectrum disorders
(ASDs) (Houghton, 2019; Tarozzi et al., 2013; Yagishita et al., 2019).
ERN, a less extensively studied analogue of SFN, also have a beneficial
impact on human health, being a potential chemopreventive and anti-
cancer agent (Azarenko et al., 2014; Citi et al., 2019; Melchini & Traka,
2010; Posner et al., 1994).
radical-trapping (chain-breaking) antioxidants (Barillari, et al., 2005;
Papi et al., 2008). Other reports indicate that SFN and ERN react with
hydroxyl radicals (•OH) and also can scavenge artificial model radicals,
like 2,2′-diphenyl-1-picrylhydrazyl radical (DPPH•) or 2,2′-azinobis(3-
ethylbenzothiazoline-6-sulfonic acid) radical cation (ABTS+•) (Ligen
et al., 2017; Montaut et al., 2017; Yuan, Yao, You, Xiao, & You, 2010).
However, the results of •OH, DPPH•, and ABTS+• assays give the in-
formation about overall reducing ability of a tested compound toward
particular ROS/RNS (Amorati & Valgimigli, 2015), not about the ability
to inhibit peroxidation of organic materials or biomolecules. The
conclusion from literature search is that SFN and ERN are not radical
trapping antioxidants at ambient temperature, however, there is no in-
formation about their chain-breaking activity at temperatures higher
than 40 ◦C. As ITCs are present in food that is subjected to mechanic
operations or/and thermal operations like boiling or frying, there is a
need to get knowledge about thermal behaviour of this class of com-
pounds at elevated temperature, at conditions similar to those in stan-
dard cooking in the presence of lipids as the constituents most sensitive
to oxidation (Yuanfeng et al., 2020). In our very recent communication
we put the hypothesis that ITCs can act as antioxidant agents during
high temperature processing of food (Cedrowski, Dąbrowa, Krogul-
Sobczak, & Litwinienko, 2020). In order to confirm the hypothesis, we
checked the oxidative stability of two model lipids of different oxidative
stability, we also performed chromatographic analysis of decomposition
products in order to explain the mechanism of the antioxidant action.
In our former studies we elaborated fast and convenient methodol-
ogy of measurement of oxidative stability of fats and oil by using dif-
ferential scanning calorimetry (DSC) in non-isothermal mode, with
5–10 mg samples heated with linear heating rate (β) and recorded heat
Health effects of ITCs have been connected with their ability to
modulate cellular redox status due to induction of Phase 2 cytoprotective
enzymes (Dinkova-Kostova and Talalay, 2008). As reviewed by Valgi-
migli and Iori (Valgimigli & Iori, 2009), there is a general agreement
that the enhancement of antiradical defense system is not due to a direct
scavenging of radicals and such effect has been named the indirect
antioxidant action of ITC. Indeed, the attempts to demonstrate that ITCs
are capable to effectively react with reactive oxygen or nitrogen species
(ROS, RNS) and prevent the start of peroxidation were only partially
successful. Barillari et al. showed that glucoerucin (the parent glucosi-
nolate of ERN, see Scheme 1B) efficiently decomposes hydrogen
peroxide and tert-butylperoxide with the rate constants 3.3 × 10ꢀ 2
Mꢀ 1sꢀ 1 and 2.0 × 10ꢀ 3 Mꢀ 1sꢀ 1 at 25 ◦C, respectively, proving that
glucoerucin is a preventive antioxidant (Barillari et al., 2005), and they
suggested that ERN also should be a preventive antioxidant. Experi-
ments with autoxidation of styrene demonstrated that ERN and SFN
react too slowly with peroxyl radicals and neither ERN nor SFN are
effect of oxidation (Litwinienko, 2001; Litwinienko, Daniluk,
&
Kasprzycka-Guttman, 2000; Litwinienko Kasprzycka-Guttman,
&
1998a,c, 2000). The methodology of measurements, typical for studies
of oxidative stability of polymers and biopolymers (Czochara, Kusio, &
Litwinienko, 2017; Ziaja, Jodko-Piorecka, Kuzmicz, & Litwinienko,
2012) can be also applied for food, as alternative to isothermal accel-
erated tests (isothermal conditions at 100–140 ◦C, detection of polar
volatile products of oxidation and decomposition of the hydroperoxides)
as, for example, the oxidative stability index (OSI) or the Rancimat
method (RM). Recently, a new methodology of oximetric measurement
of the rate of lipid oxidation has been introduced based on the moni-
toring of the oxygen consumption during high temperature oxidation by
means of the fluorescence probe sensing oxygen in head space (Mollica
et al., 2020), and this method has great advantages because, unlike OSI
and RM, it monitors the oxygen consumption instead of detecting
Scheme 1. A) Reaction scheme of ITC formation during hydrolysis of glucosinolate catalysed by myrosinase. B) Structures of sulforaphane, erucin and their cor-
responding glucosinolates.
2

J. Cedrowski et al.
Food Chemistry 353 (2021) 129213
secondary products of oxidation. However, DSC non-isothermal method
brings some additional advantages, the samples are small, thermal effect
of oxidation is interpreted as due to formation of hydroperoxides (pri-
mary products of oxidation) (Litwinienko & Kasprzycka-Guttman,
1998a) and, depending on β, the single run takes 10–30 min
comparing to several hours in OSI or RM.
products were analyzed and the obtained NMR signals were identical
with the literature data for natural and synthetic SFN and ERN (Ganin
et al., 2013), see Supplementary Material.
2.3. Kinetic measurements
In this report we used DSC to observe the effects of SFN and ERN on
the thermal behaviour of linolenic acid (LNA) and soy lecithin (SL)
under non-isothermal oxidation conditions. LNA is an example of
essential fatty acid and this polyunsaturated component of lipids is one
of the most sensitive toward peroxidation. LNA undergoes relatively fast
oxidation at elevated temperature, giving a clear and strong thermal
effect in DSC. On the other hand, soy lecithin is an important component
of a great number of food products and also neutraceutical and food
supplement ingredient with many health benefits (List, 2015). Due to
lipophilic/hydrophilic character, lecithin is frequently employed as a
very good emulsifying agent but there are a number of other applica-
tions, and some of them are connected with high temperature opera-
tions, lecithins are used as release agents and as wetting agents in
industrial bakery and confectionery and also used as pan (also mold and
belt) release agents, product separation aids (nonsticking aids in sliced
food), and other applications based on heat resistance and surface
coating feature of lecithin (List, 2015).
Preparation of lipid samples containing ERN and SFN for DSC anal-
ysis is described in the Supplementary Material. The final concentrations
of ITCs in samples were 0.17–2.0 mg ITC /g of LNA and 0.2–2.3 mg of
ITC/g of SL. The methodology of DSC measurements has been described
elsewhere (Czochara et al., 2017; Czochara, Kusio, Symonowicz, &
Litwinienko, 2016; Ulkowski, Musialik, & Litwinienko, 2005) and is
detailed in the Supplementary Material. The calorimetric measurements
were carried out using a DSC apparatus Du Pont 910 differential scan-
ning calorimeter with Du Pont 9900 thermal analyzer and a normal
pressure cell, under oxygen flow 6 dm3/min. 3–5 mg samples were
heated from 40 to 200 ◦C in open pan with linear heating rate β (2.5–20
K/min). The way of determination of kinetic parameters from series of
DSC curves is discussed in section 3.1.
2.4. GC–MS measurements
The procedure for thermal decomposition of ITCs and GC–MS mea-
surements of the products is described below. Samples of ITC (1–2 mg)
were placed (under air) in ~ 2 ml glass ampoule along with a drop of
water and sealed. The ampoule with the sample was heated at 100 ◦C
and 160 ◦C for 5 h and 30 min respectively. The reaction mixture was
then cooled down to ꢀ 70 ◦C (dry ice/acetone bath), the head of ampoule
The presented work is an extension of the short communication in
which we described that SFN and ERN inhibit oxidation of sunflower oil
but do not inhibit oxidation of linolenic acid (Cedrowski, et al., 2020).
Here, we compare the kinetics of oxidation of LNA and SL as two lipids of
different oxidative stability and dramatically different polarity and both
are used for high temperature operations with vegetables containing
isothiocyanates, therefore, we decided to use both lipid matrices in our
calorimetric studies of the antioxidant effect of SFN and ERN.
was broken and the content was dissolved in 500
μl of methylene
chloride (DCM). The solution was moved to vial containing 1 ml of
water. The DCM layer phase was subjected to GC–MS analysis directly.
GC–MS analysis was performed with HP5 capillary column (0.32 mm
2. Materials and methods
diameter, 30 m length, 0.25 μm thickness, Hewlett-Packard) with GC-
17A Ver.3 Shimadzu and GCMS-QP5050A Shimadzu mass detector.
Mass spectrum was obtained by electron impact ionization mode,
scanning from 33 m/z to 350 m/z. The flow rate of helium gas was 1.5
ml/min, and the split ratio was 5:1. Inlet temperature was 210 ◦C and
oven temperature program was as follows: an initial step starts at 40 ◦C
(isothermal for 5 min); raising with the rate 3 ◦C/min to 220 ◦C.
2.1. Materials
Detailed description of all chemicals, reagents, synthesis of ERN and
SFN, equipment (NMR, GC–MS) and experimental procedures (DSC,
preparation of samples) is included in the Supplementary Material.
Linolenic acid (98%, all-cis-9,12,15-Octadecatrienoic acid, LNA, Sigma-
Aldrich) and soy lecithin (95%, SL, MP Biomedicals LLC) were used as
lipid matrices. The main constituents of SL as declared by the producer
(certified analysis for catalog no. 102147, in %w): phosphatidylcholine
(25%); phosphatidylethanolamine (22%); phosphatidylinositol (16%);
phosphatidic acid (7%). Linoleic acid (cis-cis-9–12-octadecadienoic
acid) stands for 70% of fatty acid residues in SL.
3. Results and discussion
3.1. Kinetic parameters for oxidation of pure linolenic acid (LNA) and
soy lecithin (SL)
Fig. 1 presents DSC curves of oxidation obtained for heating rates β
ranging from 2.5 to 20 K/min for LNA and 5–20 K/min for SL (for β <
5 K/min thermal effect is extended during long time of experiment and
DSC curve is much flatter than the curves obtained for bigger β).
Linolenic acid (C18:3) is polyunsaturated fatty acid with two bis-allyl
2.2. Synthesis of erucin and sulforaphane
The target compounds ERN and racemic SFN were prepared by using
method originally proposed by Schmidt and Karrer (Schmid and Karrer,
1948) and modified by us. Full description of procedures, yields, iden-
tification by 1H and 13C NMR is provided in Supplementary Material.
Here, the symbols S1-S5 denote the compounds in Scheme S1. Potas-
sium phthalimide (S1) and 1,4-dibromobutane (S2) were reacted in the
presence of a catalytic amount of tetrabutylammonium bromide
(TBABr) giving 1-phthalimido-4-bromobutane (S3, yield 77%). This
intermediate was reacted with sodium thiomethoxide (prepared from
dimethyl disulfide and sodium) giving 1-phthalimido-4-methylthiobu-
tane (S4a) in very good yield (93%). Compound S4a was partially
oxidized using proper amount of H2O2 in glacial acetic acid providing
sulfoxide (S4b, yield 80%). Next, phtalimide group was deprotected
using hydrazine hydrate in refluxing ethanol providing amines S5a-b in
moderate to good yields (49–81%). The crude amines were reacted
without purification with thiophosgene (CSCl2) in a biphasic system
(10% NaHCO3/DCM) providing the target products (55–70%). The
–
C
H bonds and is one of the most sensitive toward oxidation. Thermal
effect of LNA oxidation is clear, with start of oxidation at temperature
90–120 ◦C (depending on β). In contrast, soy lecithin is more stable, with
linoleic acid (C18:2) residues as the main fatty acid represented, thus, the
oxidation of SL starts attemperaturesca. 150 ◦C, and, asa consequenceof
higher temperature, a shape of DSC curve is more sheer (compare Fig. 1A
and B). Non inhibited oxidation of LNA and SL was carefully discussed
and interpreted in our previous publication (Ulkowski et al., 2005), so, a
brief information will be presented here. The observed thermal effect is a
consequence of formation of hydroperoxides as primary products of
oxidation. Hydroperoxides decompose at temperatures above 200 ◦C
forming second peak on DSC curve (not shown on the plots in Fig. 1, see
DSC at temp. range 50–350 ◦C in our previous work (Ulkowski et al.,
2005)).
For low degrees of conversion
α (low extent of reaction) and at ox-
ygen pressure bigger than 13 kP the autoxidation can be described as a
3

J. Cedrowski et al.
Food Chemistry 353 (2021) 129213
Fig. 1. DSC curves obtained during thermoxidation of linolenic acid (LNA) shown in panel A, and soy lecithin (SL) shown in panel B. The number above each thermal
curve is the heating rate (β in K/min). Te is extrapolated start of oxidation. The curves have been vertically shifted for better visual clarity. Panel C: plots of logβ
versus 103/Te for LNA and SL constructed on the basis of Te values determined from DSC curves.
Table 1
Temperatures of extrapolated start of oxidation (Te) of linolenic acid and soy lecithin obtained for experiments with various heating rates (β), parameters a and b of
equation (4) with standard deviations for the slope (
σ
and σ90%)a, square regression coefficients (R2), and calculated kinetic parameters: overall activation energy Ea,
pre-exponential factor Z, and overall rate constant of oxidation (k) calculated for 100 ◦C and 150 ◦C.
Linolenic acid (LNA)
Soy lecithin (SL)
β [K/min]
T
e [◦C]
90.8
kinetic and statistical parameters
β [K/min]
Te [◦C]
kinetic and statistical parameters
2.5
a=
ꢀ 4.3468
a=
ꢀ 5.5859
5.0
101.8
107.1
110.0
115.2
115.8
120.0
120.3
b=
12.3207
5.0
7.5
151.2
156.5
159.4
162.7
166.8
168.4
171.6
b=
13.8830
7.5
R2=
0.9922
R2=
0.9891
10.0
12.5
15.0
17.5
20.0
σ
=
0.16
10.0
12.5
15.0
17.5
20.0
σ
=
0.262
σ90%
Ea=
Z=
=
0.29
σ90%
Ea=
Z=
=
0.487
79 ± 5 kJ/mol
4.54 × 1010 minꢀ 1
0.38 minꢀ 1
7.7 minꢀ 1
102 ± 9 kJ/mol
1.29 × 1012 minꢀ 1
7.49 × 10ꢀ 3 minꢀ 1
0.36 minꢀ 1
◦
◦
k100
=
=
k100
=
=
C
C
C
C
◦
◦
k150
k150
a Standard deviations (
σ) and errors calculated for confidence level 90% (σ90%).
first order process with the rate law depending on the amount of lipid,
LH:
Fig. 1A and 1B, the higher β the higher temperature of start of oxidation
(Te, see Table 1). Te values correspond to a constant conversion (α =
const) and for series of experiments with different β, the straight line
dependence of logβ values plotted versus 1/Te (expressed in kelvins) is
obtained (see Fig. 1C):
(
)
Ri
2kt
v = kp
1/2[LH] = k[LH]
(1)
a
Te
where the rate of initiation (Ri) together with the rate constant for
propagation (kp) and termination (kt) can be considered as the overall
rate constant k = kp(Ri/2kt)1/2, thus, the chain process of oxidation can
be described by the Arrhenius equation:
logβ = + b
(4)
with the slope a = ꢀ 0.4567 Ea/R and reciprocal b = -2.315 + log
(ZEa/R), with R = 8.314 J molꢀ 1 Kꢀ 1 (Litwinienko et al., 2000).
(
)
ꢀ Ea
k(T) = Z exp
(2)
RT
3.2. Kinetic parameters for oxidation of linolenic acid and soy lecithin
with ITCs
where Z is a pre-exponential factor, Ea is activation energy, R is the gas
constant. During non-isothermal experiment the starting temperature of
DSC curves presented in Fig. 1A and 1B indicate that LNA is more
sensitive to oxidation (lower Te) and the kinetic parameters calculated
from DSC curves listed in Table 1 confirm this observation. The acti-
vation energy for LNA oxidation is in reasonable agreement with
determined in our previous works, ranging from 65 to 78 kJ/mol
(Czochara et al., 2016; Ulkowski et al., 2005) and Ea for oxidation of soy
lecithin is in perfect agreement with 97 ± 8 kJ/mol (Ulkowski et al.,
2005). The rate constants clearly demonstrate that below 150 ◦C
oxidation of LNA is faster (at 100 ◦C is 50 times faster than the oxidation
of SL). Therefore, we have two materials with different oxidative sta-
bilities which can be used for assessment of the antioxidant behaviour of
ITCs at two different thermal conditions. Choice of more than one lipid
model is an important aspect during screening the antioxidant activity of
a compound because our previous work with fullerene C60 and its de-
rivatives revealed that antioxidant activity depends on the nature of the
the sample (T0) is increased to T = T0 + βτ with a linear heating rate β =
dT/dτ (τ is time). The amount of heat (H) released per time unit (or
temperature unit) corresponds to the reaction rate v, and can be
expressed in term of conversion (dH/d
τ
= v = dα/d
τ
= 1/β d
α
/dT), thus,
equation (1) defines the rate as a change of conversion:
dH/d = d /d = βd /d = k[LH] = k(1 ꢀ α)
τ
α
τ
α
τ
(3)
So called isoconversional methods of calculation of kinetic parame-
ters are based on combination of eq. (2) and (3).
Ozawa and, independently, Flynn and Wall proposed a relatively
simple way for determination of the kinetic parameters from non-
isothermal experiments, nowadays this methodology, named Ozawa-
Flynn-Wall method, is one of the most frequently used method in ther-
mal analysis (Flynn and Wall, 1966; Ozawa, 1970). As can be seen in
4

J. Cedrowski et al.
Food Chemistry 353 (2021) 129213
oxidized material: the same compound can behave as non-inhibitor in
unsaturated hydrocarbons but it turns into an effective antioxidant
during oxidation of saturated hydrocarbons (like stearic acid, polymers
etc.) at higher temperatures (Czochara, Grajda, Kusio, & Litwinienko,
2018; Czochara, Kusio, & Litwinienko, 2017, 2019; Czochara et al.,
2016). There are several reasons for such behaviour, as, for example,
different mechanism of high temperature oxidation (oxidation of poly-
mers might be mediated by alkoxyl instead of peroxyl radicals) or not
sufficient ratio kinh/kp, where kinh is the rate constant for reaction of
inhibitor with the propagating radicals Y• (reaction (5)).
behaviour would be observed). However, methodology used in DSC al-
lows to calculate the kinetic parameters that are more informative and
can be used for prediction of the oxidative stability of the samples in
other temperatures, including temperature before the rapid, sponta-
neous oxidation starts (corresponding to lag phase in isothermal ex-
periments). Table 2 collects Ea, Z and k parameters determined for
thermo-oxidation of LNA and SL containing ITCs. The results for LNA
confirm the first observations from Table S15: isothiocyanates do not
improve the oxidative stability of polyunsaturated lipid, in contrast to
conventional phenolic antioxidants like 2,6-di-tert-butyl-4-methyl-
phenol (BHT) or tocopherol (Litwinienko & Kasprzycka-Guttman,
1998b; Litwinienko, Kasprzycka-Guttman, & Jamanek, 1999, 1997;
Musialik & Litwinienko, 2007).
Inhibitor + Y⋅→non radical or stable radical; kinh
(5)
Usually reaction (5) is reduction of radical Y• to Y-H or addition of Y•
to the inhibitor. Considering the large stoichiometric excess of oxidized
substrate (neat compounds or at molar concentrations) over the amount
of antioxidant (micromolar concentrations), kinh should be 103 times
larger than kp to effectively suppress the autoxidation. Consequently, in
order to keep the proper ratio kinh/kp > 1000, the moderate inhibitor
will effectively break the propagation chain during oxidation of a ma-
terial with relatively high oxidative stability (small kp), otherwise, the
propagation chain will not be sufficiently broken and the inhibiting ef-
fect will not be observed.
However, the experiments for thermal oxidation of soy lecithin
containing ITCs resulted in the kinetic parameters which differ from the
ones obtained for the oxidation of SL without additives. The same
Table 2 presents the Ea, Z and k values for three different concentrations
0.21–2.3 mg ITC/g SL (corresponding to 1, 5, and 10 milimoles of ITC
per mol of lipid, we used here mg/g units instead of molar concentration
because SL is solid). For each compound an optimal concentration can
be found at which the lipid is protected against the oxidation. Taking
into account the activation energies, a small increase is observed for
sulforaphane (present at concentration 1.2 mg SFN/g SL) that,
combining with the increase of pre-exponential factor, gives significant
decrease of k calculated for 100 ◦C. Other two concentrations of SFN are
not effective, the small concentration 0.23 mg/g is not significant, while
ten-fold bigger amount, CSFN = 2.3 mg/g, generates pro-oxidative effect
(at 100 ◦C the oxidation is twice faster than oxidation of pure SL), see
Fig. 2.
Addition of an inhibitor should increase the oxidative stability (lag
phase, inhibition phase) and during non-isothermal oxidation such lag
phase should be manifested as an increasing temperature when the
oxidation starts (Litwinienko & Kasprzycka-Guttman, 1998b; Litwi-
nienko, Kasprzycka-Guttman, & Studzinski, 1997; Musialik & Litwi-
nienko, 2007). Table S15 (Supplementary Material) presents the
temperatures Te (for β = 5 K/min) determined for oxidation of LNA and
SL containing various concentrations of ITCs. Unfortunately, the addi-
tion of the studied compounds at any concentration within the range
0.17–2.0 mg /g of lipid (corresponding to three concentrations: 1 mM, 5
mM, and 10 mM) did not caused any increase of oxidative stability of
LNA and even the opposite effect is observed, oxidation of LNA con-
taining ITCs starts at lower temperatures than Te for neat LNA (Te is
lower in the range 2.8–4.1 ◦C for the lowest concentrations of ITCs, and
for the highest concentrations of SFN and ERN a decrease of Te is ca.
11.5 ◦C, see Table S15). Moreover, addition of ITCs to SL also did not
improve the oxidative stability of this lipid. In other accelerated
(isothermal) tests like Rancimat and OSI, such results would be
concluded as no antioxidant effect (and apparent pro-oxidative
ERN at the smallest concentration (0.21 mg per g of SL) is also not
effective, but for five-fold and ten-fold bigger concentrations of ERN the
increased values of Ea and Z were determined, giving smaller rate con-
stants for temperatures below 150 ◦C, see Table 2. For the middle con-
centration (1.1 mg ERN/g) the rate of oxidation at 100 ◦C is reduced to
half of the value obtained for pure SL. This tendency is also kept for ERN
present at concentration 2.1 mg/g: the activation energy increased to
160 kJ/mol and the rate constants calculated for 100 ◦C is almost 10-
times smaller. The effect of addition of ERN and SFN on the kinetics of
oxidation of soy lecithin confirms our very recent observation for
oxidation of sunflower oil (Cedrowski, et al., 2020): Ea for oxidation of
the pure oil was 103 ± 4 kJ/mol and increased to 119 ± 8 kJ/mol in the
presence of 10 mM ERN or 5 mM SFN, which corresponds to increased
resistance towards oxidation (k calculated at 100 ◦C ie., during the lag
phase for oxidation of pure sunflower oil was 5.8 × 10ꢀ 3 minꢀ 1 but in
the presence of 10 mM ERN was almost two-fold smaller.
Table 2
The values of the apparent activation energy (Ea), pre-exponential factor Z, and
overall rate constant of oxidation (k) calculated for 100 and 150 ◦C for oxidation
of linolenic acid (LNA) and soy lecithin (SL) containing two isothiocyanates
(ITC) at concentrations 0.17–2.0 mg ITC /g LNA and 0.21–2.3 mg ITC /g SL.
3.3. Plausible mechanism(s) of antioxidant action of ITCs at higher
temperature
CITC [mg/g] Ea [kJ/mol] Z [minꢀ 1
]
k (100 ◦C) [minꢀ 1
]
k (150 ◦C) [minꢀ 1
]
pure LNA
sulforaphane
0.20
79
5
4.5 £ 1010 0.38
7.73
Oxidation of LNA was inhibited neither by SFN nor ERN but the same
compounds inhibited oxidation of SL (the lipid undergoing oxidation at
higher temperatures). We exclude the possibility that ITCs could be
activated by the products of SL decomposition because thermogravi-
metric and calorimetric studies of soy lecithin made by Ross and co-
workers (Ross, Lemay, & Takacs, 1985) demonstrated no significant
change of mass during heating up to 200 ◦C, these results were
confirmed by Weete, who reported that soy lecithin did not change the
composition even if heated at 150 ◦C during 1 h in “forced-air oven”
(Weete, 1994). From the other side, we also exclude a decomposition of
isothiocyanate functional group because alkyl isothiocyanates are stable
71 ± 5
79 ± 6
91 ± 6
3.3 × 109
5.0 × 1010
4.5 × 1012
0.42
0.45
0.97
6.2
9.1
30
1.0
2.0
erucin
0.17
90 ± 4
86 ± 7
98 ± 5
1.5 × 1012
6.6 × 1011
6.4 × 1013
0.45
0.62
1.1
14
16
45
0.85
1.7
pure SL
sulforaphane
0.23
102
9
1.3 £ 1012 7.5 £ 10ꢀ 3
0.36
114 ± 12
129 ± 12
92 ± 12
4.9 × 1013
2.4 × 1015
8.0 × 1010
5.3 × 10ꢀ 3
2.2 × 10ꢀ 3
1.2 × 10ꢀ 2
0.41
0.30
0.39
1.2
– –
and –N
C S group is not eliminated during rather harsh conditions of
GC ana–lysi–s with high temperature of injector and increasing tempera-
2.3
erucin
0.21
112 ± 10
130 ± 16
160 ± 11
2.2 × 1013
4.1 × 1015
2.4 × 1019
5.2 × 10ꢀ 3
2.9 × 10ꢀ 3
9.1 × 10ꢀ 4
0.37
0.41
0.41
ture of GC column (up to 220 ◦C) (Jin, Wang, Rosen, & Ho, 1999). We
also exclude a reaction of isothiocyanate group with amine functionality
in lecithin (such reaction with formation of alkylated thiourea would be
possible for primary amines (Jin et al., 1999), thus, in order to check
1.1
2.1
5

J. Cedrowski et al.
Food Chemistry 353 (2021) 129213
Fig. 2. Comparison of rate constants (k at 100 ◦C) for thermoxidation of LNA (left panel) and SL (right panel) containing ITCs: sulforaphane and erucin at various
concentrations (in mg of ITC per g of lipid).
Fig. 3. Stacked 1H NMR plots (400 MHz) in the range of 6–7.5 ppm exemplifying no reaction between SFN (5.6 mg) and lecithin (176 mg) containing phospha-
tidylethanolamine (~2 equivalents to SFN) leading to thiourea formation, even after prolonged heating in sealed vial in a CDCl3:acetone‑d6 (1:1 v/v) solvent mixture;
inset: corresponding thiourea derivative obtained by heating the SFN (20 mg) and ethanolamine (8 mg) in MeCN (2 ml) at 60 ◦C for 2 h, distilling of the solvent and
redissolving the oily residue in CDCl3:acetone‑d6 (1:1 v/v).
whether our assumption is correct, we placed SFN with SL (containing
two-fold molar excess of phosphatidylethanolamine over SFN) in glass
vial with CDCl3: acetone‑d6 (1:1 v/v), the sample was thermostated at
temperature 80 ◦C during 4 days and we monitored the 1H NMR signals.
Fig. 3 demonstrates that no new signals appeared at 6.0–7.5 ppm range,
in contrast to control experiment, with ethanolamine instead of soy
lecithin, where the NH protons for thiourea were easily recorded (Fig. 3,
inset). Additionally, we analysed (1H NMR) the mixture of SFN with soy
lecithin heated without solvents at 100 ◦C during 2 h. The residue was
dissolved in CDCl3 containing 10% mol DMSO‑d6 and no signals from
the alkylthiourea group were detected (data not shown).
– –
Thus, a functionality other than –N
C S must be responsible for
– –
“antioxidant activation” of SFN and ERN at higher temperatures. The
difference between SFN and ERN is the oxidation state of sulfur in sul-
fides and sulfoxides, however, the literature search indicates that at
higher temperatures and in the presence of oxygen sulfides are con-
verted into sulfoxides, and then, sulfoxides undergo decomposition with
evolution of the radical trapping agents. Scheme 2 presents examples of
such possible conversions at elevated temperatures. First three mecha-
nisms (A-C) include the formation of sulfenic acids R-S-OH, from dialkyl
6

J. Cedrowski et al.
Food Chemistry 353 (2021) 129213
Scheme 2. Mechanisms of formation of sulfenic acids: A) decomposition of dialkyl sulfoxides in a Cope elimination (during autoxidation of squalene in chloro-
benzene at 60 ◦C (Bateman et al., 1962) or autoxidation of neat tetralin at 60 ◦C (Koelewijn & Berger, 1972)); B) oxidation of alkylthio-derivatives bearing electron
withdrawing group (EWG) at β-carbon with subsequent thermolysis in non-polar solvent (Gupta & Carroll, 2014; Jones, Cottam, & Davies, 1979); C) thermolysis of S-
methylcysteine sulfoxide (in water, sealed tube, 80–200 ◦C (Kubec et al., 1998)); D) and E) thermal decomposition of sulforaphane with generation of methylsulfinyl
radical (in water, 50–100 ◦C (Jin et al., 1999)).
sulfides and sulfoxides (Bateman, Cain, Colclough, & Cunneen, 1962;
reaction was proposed for diallyl thiosulfinate (allicin - a component of
garlic and onion) (Block, 1992; Lynett et al., 2011; Vaidya et al., 2009).
Another example of formation of sulfenic acid is pictured in panel B
of Scheme 2 when initial oxidation of sulfides to sulfoxides make them
prone to stepwise β-elimination of acidic hydrogen (Gupta & Carroll,
2014). Electron withdrawing group (EWG) at β-position seems to be
helpful but is not necessary (see panel A) and an example of such
decomposition not assisted by β-EWG was observed by Kubec et al.
(Kubec et al., 1998), see panel C, for S-methylcysteine sulfoxide heated
at 120 ◦C during 1 h (complete conversion in the presence of 10% of
water, for dry compound the process was “somewhat slower”). The same
authors reported the presence of methyl methanethiosulfinate and
dimethyl disulfide (Scheme 2, panel E) as thermal oxidation products,
the latter compound responsible for unpleasant smelling during too long
frying, baking or roasting (when the food reaches high temperature and
the amount of water becomes very low). The formation of unstable
sulfenic acid was also observed for cysteine during oxidative stress,
meaning that extensive heating is not necessary to initiate this decom-
position route (Gupta & Carroll, 2014).
´
Gupta & Carroll, 2014; Koelewijn & Berger, 1972; Kubec, Drhova, &
ˇ
Velísek, 1998; Lynett, Butts, Vaidya, Garrett, & Pratt, 2011; Vaidya,
Ingold, & Pratt, 2009).
–
Sulfenic acids are transient species with very weak O H bonds
(about 70 kcal/mol) which can react with alkylperoxyl radicals with a
near diffusion-controlled proton-coupled electron transfer mechanism
(Lynett et al., 2011). Formation of sulfenic acids from dialkyl sulfoxides
at elevated temperatures were proposed as early as in sixties (Bateman
et al., 1962) for sulfoxides with accessible H atoms at β-position, mainly
with tert-butyl groups (Scheme 2 , panel A). Interestingly, the authors
noticed that sulfoxides inhibited autoxidation initiated with hydro-
peroxides but not with other kinds of initiators, like peroxides or azo-
compounds and they interpreted this inhibition effect as a result of
inactivation of hydroperoxides due to intermolecular interactions. Ten
years later Koelewijn and Berger (Koelewijn & Berger, 1972) ruled out
the hypothesis about the role of molecular complexes of hydroperoxides
and they proposed that antioxidant action of sulfoxides in tetralin comes
from their ability to form transient sulfenic acids during thermal
decomposition, and they measured the rate constants for unimolecular
decomposition of sulfoxides (kdecomp for di-tert-butylsulfoxide at 60 ◦C is
1.3 × 10ꢀ 5 sꢀ 1 and for less reactive di-n-dodecylsulfoxide kdecomp = 10ꢀ 5
sꢀ 1 at 130 ◦C). They also measured rate constants for reaction of sulfenic
acids with alkylperoxyl radicals:
Fourth mechanism we found in the literature is presented in panel D
of Scheme 2 (Hanschen, Lamy, Schreiner, & Rohn, 2014) and this is a
thermal degradation of SFN with formation of methylsulfinyl radicals.
Although the presence of sulfenic acid was not postulated by the au-
thors, the mechanism D also can be used as an explanation of the anti-
oxidant action of sulforaphane, because methylsulfinyl radicals are able
(8)
to scavenge peroxyl radicals. Moreover, methylsulfinyl radical can be
also produced from sulfenic acid (reaction (6) modified to the form:
CH3SOH + RO•2 → CH3SO• +ROOH), therefore, mechanism D does not
exclude the formation of sulfenic acid (Scheme 2, mechanism A) as an
intermediate. Pratt and collaborators in their work on antioxidant action
of sulfenic acids (Lynett et al., 2011) paid attention to alkylsulfinyl
radicals R-SO• as the products of H atom abstraction from sulfenic acid,
and studied thermochemistry of possible subsequent reactions starting
from the trapping another alkylperoxyl radical (equation (8)).
R-SOH + RO2⋅→non ꢀ radical products
(6)
to be higher than 107 Mꢀ 1sꢀ 1 at 60 ◦C. If peroxyl radicals are not
accessible, alkylsulfenic acids undergo self-condensation giving S-alkyl
alkanethiosulfinates (Bateman et al., 1962):
2CH3SOH→CH3(S = O)SCH3 + H2O
(7)
and this thiosulfinate easily decomposes via Cope-type elimination
into sulfenic acid and thioformaldehyde. The similar sequence of
7

J. Cedrowski et al.
Food Chemistry 353 (2021) 129213
Scheme 3. Proposed path(s) of thermal oxidative conversion of erucin (ERN) to sulforaphane (SFN) and subsequent decomposition to the products detected by GC-
MS and showed in Table 3. Wavy bond in 3A means Z/E isomers.
–
Although formation of peroxysulfinate ester CH (S O)OOR is 40
of ITCs that can be important for antioxidant action (see Scheme 3 and
Table 3).
–
3
kcal/mol exothermic, it decomposes because of exceptionally weak
–
O
O bond (8–10 kcal/mol). If such decomposition produced additional
All three products 3A, 3B and 3C were detected for both starting
compounds, ERN or SFN, proving that oxidation of ERN to SFN is a first
step of oxidative thermal conversion. SFN undergoes further processes:
dehydratation and formation of 3A (both Z/E isomers were detected),
product of β-elimination 3B (accompanying sulfenic acid was not
radicals, no antioxidant action would be observed, but sulfonyl and
alkoxyl radicals recombine within the cage to form stable sulfonate ester
–
–
which is 60 kcal/mol lower in enthalpy than CH (S O)OOR, and 100
3
kcal/mol lower than the starting sulfinyl and peroxyl radicals (Lynett
et al., 2011). Therefore, the mechanism D with thermal decomposition
of sulforaphane and release of sulfinyl radicals that can scavenge peroxyl
radicals also can be considered as probable explanation of inhibiting
effect of this compound during thermal autoxidation of lipids above 100
◦C.
–
detected because of instability) and cleavage of S C bond with gener-
ation of methylsulfinyl radical which, after recombination with meth-
ylsulfinyl radical gives S-methyl methanethiosulfinate, further oxidized
to 3C (this process is also depicted in Scheme 2, panelE). Compound 3A
was also detected by Jin (Jin et al., 1999), moreover, Papi et al. (Papi,
et al., 2008) described its selective cytotoxic/apoptotic activity and
ability to decompose peroxides, however, 3A was not a good chain-
breaking inhibitor of autoxidation of styrene and methyl linoleate.
Comparison of the kinetic parameters collected in Table 2 indicates
the highest Ea and Z parameters obtained for oxidation of SL containing
1.2 mg SFN and 2.1 mg ERN per gram of SL. The antioxidant action of
both ITCs is dependent on the rate of their decomposition to sulfenic
acid (ie., depends on temperature and concentration of ITC) therefore,
the observed increase of Ea and Z is an effect of the presence of a new
chain-breaking component formed from ITCs. Then, for increasing
concentration of SFN from 1.2 to 2.3 mg/g we observe a decrease of
effectiveness but such decrease is also observed for other conventional
antioxidants when present in too high concentration.
–
Mechanism D proposes a cleavage of S C bond proximal to meth-
ylsulfinyl group but mechanism B predicts the oxidation of methyl sul-
–
fide to methyl sulfoxide as a first step, instead of cleavage of S C bond.
–
S
We can justify such different paths because of different stability of C
bond in sulfides and sulfoxides: bond energy in dialkyl sulfides,
depending on the length of alkyl group, is within the range 72 kcal/mol
–
(for CH3S-C5H11) to 77 kcal/mol (for CH3S-CH3), so, C S bond is at
least 15 kcal/mol stronger than C-(SO) bond in sulfoxides (55 kcal/mol
for dimethylsulfoxide) (Luo, 2007). Thus, terminal CH3S- group is
oxidized to CH3S(=O)- group (ERN is easily oxidized to SFN by perox-
ides even under mild conditions (Barillari et al., 2005)), subsequently
decomposes to CH3SO• or undergoes Cope elimination to sulfenic acid
(and then, to CH3SO•), see Scheme 2. We performed experiments with
thermal degradation of a few milligrams of SFN and ERN placed in glass
sealed vials (aerobic conditions) and kept at temperatures 100–160 ◦
C
3.4. The problem of peculiarity of start of oxidation at 150 ◦
C
during 5 h at 100 ◦C (or 0.5 h at 160 ◦C) and we analysed the volatile
products by GC-MS. The conditions used in our experiments differ from
the experimental conditions described in the caption to Scheme 2A-D,
however, we succeed in detecting three products of thermal degradation
One more problem that needs to be additionally explained is why
temperatures of start of oxidation (Te) of lecithin and lecithin containing
ITCs are almost the same, regardless the different kinetic parameters
Table 3
Names, retention times and mass spectral data for three products of thermal decomposition of neat SFN/ERN at temperatures 100 ◦C and 160 ◦C.
a
Compound, symbol, RT
Experimente
MS spectral data m/z (relative intensity)
4-methylthio-3-butenyl isothiocyanateb 3A, RT = 34.2 min
SFN: 100 ◦C, 160 ◦
C
C
161 [(M + 2)+, 9], 87 (22), 85 (24), 72 (42), 61 (100), 45 (34)
ERN: 160 ◦
C
but-3-enyl isothiocyanatec 3B, RT = 11.6 min
SFN: 100 ◦C, 160 ◦
113 (M+, 50), 85 (10), 72 (100), 55 (30), 39 (71)
ERN: 160 ◦
C
S-methyl methanethiosulfonated 3C, RT = 10.5 min
SFN: 160 ◦
C
128 [(M + 2)+, 8], 126 (M+, 73), 111 (11), 81 (5), 79 (50), 64 (40), 47 (10), 45 (20)
ERN: 160 ◦
C
a RT = retention time. b Identified basing on NIST Mass Spectrometry Data Center, SRD1A, NIST MS numbers 5649 and 414,629 (accessed on 14.08.2020). c Identified
basing on NIST Mass Spectrometry Data Center, SRD 69, NIST MS number 414,530 (accessed on 14.08.2020), d Identified basing on NIST Mass Spectrometry Data
e
Center, SRD 69, NIST MS number 236,079 (accessed on 14.08.2020). This column indicates in which experiment the compounds were detected for thermal
decomposition of SFN /ERN.
8

J. Cedrowski et al.
Food Chemistry 353 (2021) 129213
Fig. 4. Comparison of the plots of log k(T) versus reciprocal of absolute temperature 363 K-513 K (90–240 ◦C) for oxidation of pure soy lecithin (SL) and SL
containing SFN and ERN. Rate constants were calculated from equation (2), using Ea and Z values listed in Table 2. The vertical arrows indicate the temperatures of
intersection of the straight lines (see text).
characterizing the process with and without SFN and ERN, indicating no
inhibition effect at temperatures 150 ± 15 ◦C. Indeed, the rate constants
calculated for 150 ◦C (start of spontaneous oxidation) are very close to
each other, within the range 0.30–0.41 minꢀ 1, see Table 2. However,
such levelling of oxidative stability of inhibited and non-inhibited sys-
tems at particular temperature is not surprising and corresponds to
commonly found inversion of the relative rates of the processes moni-
tored by thermal analysis techniques. The rate constants of two pro-
cesses described with different sets of pairs of Ea and Z might cross each
other at isokinetic temperature Tiso (Litwinienko & Jodko-Piorecka,
2015; Grzegorz Litwinienko, 2005; Ulkowski et al., 2005). Below Tiso
the process with smaller activation barrier is faster but at temperature >
responsible for breaking the autoxidation chain. Additional possibility is
a thermal decomposition of methylsulfinyl end group with subsequent
release of sulfinyl radicals that can react with alkylperoxyl radicals.
Interestingly, reaction of sulfenic acids with peroxyl radicals also pro-
duces sulfinyl radicals. Possible mechanisms of recombination of the
formed sulfinyl radicals with sulfenyl radicals gives also S-methyl
methanethiosulfinate, therefore, several organosulfur species and the
transient products presented in Schemes 2 and 3 can be responsible for
the antioxidant action of SFN and ERN at higher temperatures.
CRediT authorship contribution statement
T
iso the process described by higher Ea proceeds faster. On the basis of
Jakub Cedrowski: Conceptualization, Funding acquisition, Inves-
tigation, GC-MS and DSC measurements, NMR and MS analysis, Visu-
alization, Writing - review & editing. Kajetan Dąbrowa: Investigation,
Synthesis, GC-MS measurements, NMR and MS analysis, Visualization,
Writing - review & editing. Paweł Przybylski: Investigation, DSC
measurements, Visualization, Writing - review & editing. Agnieszka
Krogul-Sobczak: Investigation, Writing - review & editing. Grzegorz
Litwinienko: Conceptualization, Funding acquisition, Writing - review
& editing, Supervision, Project administration.
the Ea and Z from Table 2, we plotted the logk as a function of reciprocal
absolute temperature for oxidation of pure lecithin (red line) and for
oxidation of lecithin containing SFN and ERN, see Fig. 4.
The intersections of the red line with other lines are at temperatures
135◦ and 160 ◦C for SFN and 145 ± 2 ◦C for ERN, indicating that during
DSC experiment the temperatures of the start of oxidation (Te from 150
to 170 ◦C) overlap with Tiso 135–160 ◦C giving completely erroneous
impression that both isothiocyanates are kinetically insignificant during
autoxidation. However, inspection of the rate constants calculated for a
bit lower temperatures, during the induction period (lag phase) for 100
◦C, (see Table 2) clearly shows the inhibiting effect at temperatures
below Tiso. Moreover, some important information can be gained from
the above considerations that isothiocyanates at high temperature (ca.
100 ◦C and higher) are activated via thermal decomposition into reac-
tive transient products which can effectively inhibit oxidation but only
at temperatures below 150 ◦C. Practically, it is possible to decompose
ITCs at temperatures above 150–160 ◦C but after such decomposition
the system should be immediately cooled to temperature below Tiso, to
reach the optimal antioxidant activity of decomposed isothiocyanates.
Declaration of Competing Interest
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
the work reported in this paper.
Acknowledgements
J.C. gratefully acknowledges financial support from the National
Science Center of Poland, NCN grant PRELUDIUM 8 No. 2014/15/N/
ST5/02939. Part of this work was financed from the research grant
OPUS 18 No. 2018/31/B/ST4/02354 received from NCN by G.L.
4. Conclusions
Sulforaphane (SFN) and erucin (ERN) do not inhibit autoxidation of
linoleic acid (LNA) at temperatures lower than 100 ◦C, but both ITCs
manifest significant inhibiting effect during autoxidation of soy lecithin
(SL) at temperatures above 100 ◦C. Our results, taken together with the
results described in preliminary communication (Cedrowski et al.,
2020), suggest that thermal decomposition of ITCs can produce sec-
ondary compounds able to trap peroxyl radicals. Literature search in-
dicates that the isothiocyanate (-N=C=S) group is thermally stable and
does not exhibit antioxidant activity, thus the other functional groups
might contribute to the observed inhibition effect. At higher tempera-
tures methylsulfanyl- (in ERN) and methylsulfinyl- (in SFN) groups can
be converted into sulfenic acids and such transient species are
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.foodchem.2021.129213.
References
Amorati, R., & Valgimigli, L. (2015). Advantages and limitations of common testing
methods for antioxidants. Free Radical Research, 49(5), 633–649. https://doi.org/
10.3109/10715762.2014.996146.
Azarenko, O., Jordan, M. A., & Wilson, L. (2014). Erucin, the Major Isothiocyanate in
Arugula (Eruca sativa), Inhibits Proliferation of MCF7 Tumor Cells by Suppressing
Microtubule Dynamics. PLOS ONE, 9(6), Article e100599. https://doi.org/10.1371/
journal.pone.0100599.
9

J. Cedrowski et al.
Food Chemistry 353 (2021) 129213
´
ˇ
Barillari, J., Canistro, D., Paolini, M., Ferroni, F., Pedulli, G. F., Iori, R., & Valgimigli, L.
(2005). Direct Antioxidant Activity of Purified Glucoerucin, the Dietary Secondary
Metabolite Contained in Rocket (Eruca sativa Mill.) Seeds and Sprouts. Journal of
Agriculture and Food Chemistry, 53(7), 2475–2482. https://doi.org/10.1021/
jf047945a.
Kubec, R., Drhova, V., & Velísek, J. (1998). Thermal Degradation of S -Methylcysteine
and Its SulfoxideImportant Flavor Precursors of Brassica and Allium Vegetables.
Journal of Agriculture and Food Chemistry, 46(10), 4334–4340. https://doi.org/
10.1021/jf980379x.
Ligen, Z., Yuanfeng, W. U., Yuke, S., Lei, Z., Mupunga, J., Jianwei, M., & Shiwang, L.
(2017). Broccoli seed extracts but not sulforaphane have strong free radical
scavenging activities. International Journal of Food Science & Technology, 52(11),
2374–2381. https://doi.org/10.1111/ijfs.13521.
Bateman, L., Cain, M., Colclough, T., & Cunneen, J. I. (1962). 697. Oxidation of organic
sulphides. Part XIII. The antioxidant action of sulphoxides and thiolsulphinates in
autoxidizing squalene. Journal of the Chemical Society, 3570–3578. https://doi.org/
10.1039/JR9620003570 (Resumed).
List, G. R. (2015). Soybean Lecithin: Food, Industrial Uses, and Other Applications. In
M. U. Ahmad, & X. Xu (Eds.), Polar Lipids (pp. 1–33). Elsevier.
ˇ
´
ˇ
Blazevic, I., Montaut, S., Burcul, F., Olsen, C. E., Burow, M., Rollin, P., & Agerbirk, N.
(2020). Glucosinolate structural diversity, identification, chemical synthesis and
metabolism in plants. Phytochemistry, 169, 112100. https://doi.org/10.1016/j.
phytochem.2019.112100.
Litwinienko, G. (2001). Autooxidation of unsaturated fatty acids and their esters. Journal
of Thermal Analysis and Calorimetry, 65(2), 639–646. https://doi.org/10.1023/A:
1017974313294.
Block, E. (1992). The Organosulfur Chemistry of the GenusAllium - Implications for the
Organic Chemistry of Sulfur. Angewandte Chemie (International ed. in English), 31(9),
1135–1178. https://doi.org/10.1002/anie.199211351.
Litwinienko, Grzegorz (2005). Analysis of Lipid Oxidation by differential Scanning
Calorimetry. In A. Kamal-Eldin, & J. Pokorny (Eds.), Analysis of Lipid Oxidation.
AOCS Publishing, Article doi:10.1201/9781439822395.ch7.
Cedrowski, J., Dąbrowa, K., Krogul-Sobczak, A., & Litwinienko, G. (2020). A Lesson
Learnt from Food Chemistry—Elevated Temperature Triggers the Antioxidant Action
of Two Edible Isothiocyanates: Erucin and Sulforaphane. Antioxidants, 9(11), 1090.
https://doi.org/10.3390/antiox9111090.
Litwinienko, G., Daniluk, A., & Kasprzycka-Guttman, T. (2000). Study on autoxidation
kinetics of fats by differential scanning calorimetry. 1. Saturated C12–C18 fatty acids
and their esters. Industrial and Engineering Chemistry Research, 39(1), 7–12. https://
doi.org/10.1021/ie9905512.
Cho, H. J., Lee, K. W., & Park, J. H. (2013). Erucin Exerts Anti-Inflammatory Properties in
Murine Macrophages and Mouse Skin: Possible Mediation through the Inhibition of
NFκB Signaling. International Journal of Molecular Sciences, 14(10), 20564–20577.
https://doi.org/10.3390/ijms141020564.
Litwinienko, G., & Jodko-Piorecka, K. (2015). Recent Developments in DSC Analysis to
Evaluate Thermooxidation and Efficacy of Antioxidants in Vegetable Oils. Chapt 5.
In E. Chiavaro (Ed.), Differential Scanning Calorimetry: Applications in Fat and Oil
Technology. CRC Press, Taylor & Francis.
Citi, V., Piragine, E., Pagnotta, E., Ugolini, L., Di Cesare Mannelli, L.Testai, L., … (2019).
Anticancer properties of erucin, an H2S - releasing isothiocyanate, on human
pancreatic adenocarcinoma cells (AsPC-1) : Erucin, an H2S - donor, exerts anticancer
properties on AsPC-1. Phytotherapy Research, 33(3), 845–855. https://doi.org/
10.1002/ptr.6278.
Litwinienko, G., & Kasprzycka-Guttman, T. (1998a). A DSC study on thermoxidation
kinetics of mustard oil. Thermochimica Acta, 319(1-2), 185–191. https://doi.org/
10.1016/s0040-6031(98)00410-9.
Litwinienko, G., & Kasprzycka-Guttman, T. (1998b). The influence of some chain-
breaking antioxidants on thermal-oxidative decomposition of linolenic acid. nd
Calorimetry, 54(1), 203–210. https://doi.org/10.1023/A:1010189507864.
Litwinienko, G., & Kasprzycka-Guttman, T. (1998c). Oxidation of saturated fatty acids
esters: DSC investigations. Journal of Thermal Analysis and Calorimetry, 54(1),
211–217. https://doi.org/10.1023/A:1010193624702.
Czochara, R., Grajda, M., Kusio, J., & Litwinienko, G. (2018). Expanding the antioxidant
activity into higher temperatures - Fullerene C60 conjugated with α-Tocopherol
analogue as a hybrid antioxidant in saturated lipid systems. Bulgarian Chemical
Communications, 50(C), 268–274.
Czochara, R., Kusio, J., & Litwinienko, G. (2017). Fullerene C60 conjugated with phenols
as new hybrid antioxidants to improve the oxidative stability of polymers at elevated
temperatures. RSC Advances, 7(70), 44021–44025. https://doi.org/10.1039/
C7RA08764K.
Litwinienko, G., & Kasprzycka-Guttman, T. (2000). Study on the Autoxidation Kinetics of
Fat Components by Differential Scanning Calorimetry. 2. Unsaturated Fatty Acids
and Their Esters. Industrial and Engineering Chemistry Research, 39(1), 13–17. https://
doi.org/10.1021/ie990552u.
Czochara, R., Kusio, J., & Litwinienko, G. (2019). Fullerene C60 conjugated with N, N-
dimethylaniline - a hybrid antioxidant acting at high temperature. Bulgarian
Chemical Communications, 51, 224–230. https://doi.org/10.34049/bcc.51.2.4896.
Czochara, R., Kusio, J., Symonowicz, M., & Litwinienko, G. (2016). Fullerene C60
Derivatives as High-Temperature Inhibitors of Oxidative Degradation of Saturated
Hydrocarbons. Industrial and Engineering Chemistry Research, 55(37), 9887–9894.
https://doi.org/10.1021/acs.iecr.6b02564.
Litwinienko, G., Kasprzycka-Guttman, T., & Jamanek, D. (1999). DSC study of
antioxidant properties of dihydroxyphenols. Thermochimica Acta, 331(1), 79–86.
https://doi.org/10.1016/S0040-6031(99)00058-1.
Litwinienko, G., Kasprzycka-Guttman, T., & Studzinski, M. (1997). Effects of selected
phenol derivatives on the autoxidation of linolenic acid investigated by DSC non-
isothermal methods. Thermochimica Acta, 307(1), 97–106. https://doi.org/10.1016/
S0040-6031(97)00366-3.
Dinkova-Kostova, A. T., & Kostov, R. V. (2012). Glucosinolates and isothiocyanates in
health and disease. Trends in Molecular Medicine, 18(6), 337–347. https://doi.org/
10.1016/j.molmed.2012.04.003.
Luo, Y.-R. (2007). Comprehensive Handbook of Chemical Bond Energies. CRC Press,
Taylor&Francis Group. Chapt 8.
Lynett, P. T., Butts, K., Vaidya, V., Garrett, G. E., & Pratt, D. A. (2011). The mechanism of
radical-trapping antioxidant activity of plant-derived thiosulfinates. Organic &
Biomolecular Chemistry, 9(9), 3320–3330. https://doi.org/10.1039/C1OB05192J.
Melchini, A., & Traka, M. H. (2010). Biological Profile of Erucin: A New Promising
Anticancer Agent from Cruciferous Vegetables. Toxins, 2(4), 593–612. https://doi.
org/10.3390/toxins2040593.
Dinkova-Kostova, A. T., & Talalay, P. (2008). Direct and indirect antioxidant properties
of inducers of cytoprotective proteins. Molecular Nutrition & Food Research, 52(S1),
S128–S138. https://doi.org/10.1002/mnfr.200700195.
Fahey, J. W., Zalcmann, A. T., & Talalay, P. (2001). The chemical diversity and
distribution of glucosinolates and isothiocyanates among plants. Phytochemistry, 56
(1), 5–51. https://doi.org/10.1016/S0031-9422(00)00316-2.
Mollica, F., Lucarini, M., Passerini, C., Carati, C., Pavoni, S., Bonoldi, L., & Amorati, R.
(2020). Effect of antioxidants on high-temperature stability of renewable bio-oils
revealed by an innovative method for the determination of kinetic parameters of
oxidative reactions. Antioxidants, 9(5), 399. https://doi.org/10.3390/
antiox9050399.
Fahey, J. W., Zhang, Y., & Talalay, P. (1997). Broccoli sprouts: An exceptionally rich
source of inducers of enzymes that protect against chemical carcinogens. Proceedings
of the National Academy of Sciences, 94(19), 10367–10372. https://doi.org/10.1073/
pnas.94.19.10367.
´
Fenwick, G. R., Heaney, R. K., Mullin, W. J., & VanEtten, C. H. (1983). Glucosinolates
and their breakdown products in food and food plants. C R C Critical Reviews in Food
Science and Nutrition, 18(2), 123–201. https://doi.org/10.1080/
10408398209527361.
Montaut, S., Benson, H. J., Kay, M., Guido, B. S., Mahboob, S. S., Chenier, J., , …
Gasparetto, J.-L, & Joly, H.A.. (2017). Probing the free-radical scavenging activity of
the extract, the major glucosinolate and isothiocyanate of Eruca sativa Mill. and
Lepidium densiflorum Schrad. seeds. Journal of Food Composition and Analysis, 61,
52–58. https://doi.org/10.1016/j.jfca.2017.02.015.
Flynn, J. H., & Wall, L. A. (1966). A quick, direct method for the determination of
activation energy from thermogravimetric data. Polymer Letter 4, 323-328.
Ganin, H., Rayo, J., Amara, N., Levy, N., Krief, P., & Meijler, M. M. (2013). Sulforaphane
and erucin, natural isothiocyanates from broccoli, inhibit bacterial quorum sensing.
Med. Chem. Commun., 4(1), 175–179. https://doi.org/10.1039/C2MD20196H.
Gupta, V., & Carroll, K. S. (2014). Sulfenic acid chemistry, detection and cellular lifetime.
Biochimica et Biophysica Acta (BBA) - General Subjects, 1840(2), 847–875. https://doi.
org/10.1016/j.bbagen.2013.05.040.
Musialik, M., & Litwinienko, G. (2007). DSC study of linolenic acid autoxidation
inhibited by BHT, dehydrozingerone and olivetol. Journal of Thermal Analysis and
Calorimetry, 88(3), 781–785. https://doi.org/10.1007/s10973-006-8507-0.
Jones, D. N., Cottam, P. D., & Davies, J. (1979). Addition of sulphenic acids to
unactivated alkynes to give alkenyl sulphoxides. Tetrahedron Letters, 20(51),
4977–4980. https://doi.org/10.1016/S0040-4039(01)86766-6.
Ozawa, T. (1970). Kinetic analysis of derivative curves in thermal analysis. Journal of
Thermal Analysis, 2(3), 301–324.
Hanschen, F. S., Lamy, E., Schreiner, M., & Rohn, S. (2014). Reactivity and Stability of
Glucosinolates and Their Breakdown Products in Foods. Angewandte Chemie
International Edition, 53(43), 11430–11450. https://doi.org/10.1002/
anie.201402639.
Palliyaguru, D. L., Yuan, J.-M., Kensler, T. W., & Fahey, J. W. (2018). Isothiocyanates:
Translating the Power of Plants to People. Molecular Nutrition & Food Research, 62
(18), 1700965. https://doi.org/10.1002/mnfr.201700965.
Hayes, J. D., Kelleher, M. O., & Eggleston, I. M. (2008). The cancer chemopreventive
actions of phytochemicals derived from glucosinolates. European Journal of Nutrition,
47(S2), 73–88. https://doi.org/10.1007/s00394-008-2009-8.
Papi, A., Orlandi, M., Bartolini, G., Barillari, J., Iori, R., Paolini, M., … Ferroni, F. Grazia,
F. M., Pedulli, G. F., Valgimigli, L.. (2008). Cytotoxic and Antioxidant Activity of 4-
Methylthio-3-butenyl Isothiocyanate from Raphanus sativus L. (Kaiware Daikon)
Sprouts. Journal of Agriculture and Food Chemistry, 56(3), 875–883. https://doi.org/
10.1021/jf073123c.
Houghton, C. A. (2019). Sulforaphane: Its “Coming of Age” as a Clinically Relevant
Nutraceutical in the Prevention and Treatment of Chronic Disease. Oxidative
Medicine and Cellular Longevity, 2019, 1–27. https://doi.org/10.1155/2019/
2716870.
Posner, G. H., Cho, C.-G., Green, J. V., Zhang, Y., & Talalay, P. (1994). Design and
synthesis of bifunctional isothiocyanate analogs of sulforaphane: Correlation
between structure and potency as inducers of anticarcinogenic detoxication
enzymes. Journal of Medicinal Chemistry, 37(1), 170–176. https://doi.org/10.1021/
jm00027a021.
Jin, Y. I., Wang, M., Rosen, R. T., & Ho, C.-T. (1999). Thermal Degradation of
Sulforaphane in Aqueous Solution. Journal of Agriculture and Food Chemistry, 47(8),
3121–3123. https://doi.org/10.1021/jf990082e.
Koelewijn, P., & Berger, H. (1972). Mechanism of the antioxidant action of dialkyl
sulfoxides. Recueil des Travaux Chimiques des Pays-Bas, 91(11), 1275–1286. https://
doi.org/10.1002/recl.19720911102.
Ross, R. A., Lemay, R., & Takacs, A. M. (1985). The thermal decomposition of granular
and fluid soybean lecithins. Canadian Journal of Chemical Engineering, 63(4),
639–644. https://doi.org/10.1002/cjce.5450630417.
10

J. Cedrowski et al.
Food Chemistry 353 (2021) 129213
Schmid, H., & Karrer, P. (1948). Synthese der racemischen und der optisch aktiven
Formen des Sulforaphans. Helvetica Chimica Acta, 31(6), 1497–1505.
Tarozzi, A., Angeloni, C., Malaguti, M., Morroni, F., Hrelia, S., & Hrelia, P. (2013).
Sulforaphane as a Potential Protective Phytochemical against Neurodegenerative
Diseases. Oxidative Medicine and Cellular Longevity, 2013, 1–10. https://doi.org/
10.1155/2013/415078.
Yagishita, Y., Fahey, J. W., Dinkova-Kostova, A. T., & Kensler, T. W. (2019). Broccoli or
Sulforaphane: Is It the Source or Dose That Matters? Molecules, 24(19), 3593.
https://doi.org/10.3390/molecules24193593.
Yuan, H., Yao, S., You, Y., Xiao, G., & You, Q. (2010). Antioxidant Activity of
Isothiocyanate Extracts from Broccoli. Chinese Journal of Chemical Engineering, 18(2),
312–321. https://doi.org/10.1016/S1004-9541(08)60358-4.
Ulkowski, M., Musialik, M., & Litwinienko, G. (2005). Use of Differential Scanning
Calorimetry To Study Lipid Oxidation. 1. Oxidative Stability of Lecithin and
Linolenic Acid. Journal of Agriculture and Food Chemistry, 53(23), 9073–9077.
https://doi.org/10.1021/jf051289c.
Zhang, Y., Talalay, P., Cho, C. G., & Posner, G. H. (1992). A major inducer of
anticarcinogenic protective enzymes from broccoli: Isolation and elucidation of
structure. Proceedings of the National Academy of Sciences, 89(6), 2399–2403. https://
doi.org/10.1073/pnas.89.6.2399.
Vaidya, V., Ingold, K. U., & Pratt, D. A. (2009). Garlic: Source of the Ultimate
Antioxidants-Sulfenic Acids. Angewandte Chemie International Edition, 48(1),
157–160. https://doi.org/10.1002/anie.200804560.
Yuanfeng, W., Chengzhi, L., Ligen, Z., Juan, S., Xinjie, S., Yao, Z., & Jianwei, M. (2020).
Approaches for enhancing the stability and formation of sulforaphane. Food
Chemistry. https://doi.org/10.1016/j.foodchem.2020.128771. In press.
Ziaja, P., Jodko-Piorecka, K., Kuzmicz, R., & Litwinienko, G. (2012). Calix[4]
pyrogallolarenes as novel high temperature inhibitors of oxidative degradation of
polymers. Polymer Chemistry, 3(1), 93–95. https://doi.org/10.1039/c1py00494h.
Valgimigli, L., & Iori, R. (2009). Antioxidant and pro-oxidant capacities of ITCs.
Environmental and Molecular Mutagenesis, 50(3), 222–237. https://doi.org/10.1002/
em.20468.
Weete, J. D. (1994). Preferential degradation of noncholine phosphatides in soybean
lecithin by thermalization. Journal of the American Oil Chemists’ Society, 71(11),
1195–1199. https://doi.org/10.1007/BF02540536.
11