Enzymatic synthesis of 1-o-galloylglycerol: Characterization and determination of its antioxidant properties

Article abstract of DOI:10.1016/j.foodchem.2019.125479

1-o-Galloylglycerol (GG) was synthesized by the enzymatic glycerolysis of propyl gallate (PG) using a food-grade lipase (Lipozyme 435). The reaction conditions affecting the yield of GG were optimized and a yield of 76.9% ± 1.2% was obtained. GG was chara

Full text of DOI:10.1016/j.foodchem.2019.125479

FoodChemistry305(2020)125479
Contents lists available at ScienceDirect
Food Chemistry
journal homepage: www.elsevier.com/locate/foodchem
Enzymatic synthesis of 1-o-galloylglycerol: Characterization and
determination of its antioxidant properties
Siyu Zhang, Casimir C. Akoh⁎
Department of Food Science & Technology, University of Georgia, Athens, GA, 30602, USA
A R T I C L E I N F O
A B S T R A C T
Keywords:
1-o-Galloylglycerol (GG) was synthesized by the enzymatic glycerolysis of propyl gallate (PG) using a food-grade
lipase (Lipozyme® 435). The reaction conditions affecting the yield of GG were optimized and a yield of
1-o-Galloylglycerol
Lipase-catalyzed glycerolysis
Solubility
Partition coefficient
Antioxidant property
76.9%
1.2% was obtained. GG was characterized by various techniques after being separated from the re-
action mixture using liquid-liquid extraction. The water solubility and hydrophilicity of GG were significantly
higher than those of gallic acid (GA) and PG. The antioxidant properties, measured by the ferric reducing an-
tioxidant power (FRAP) and hydrogen peroxide (H2O2) scavenging assays, showed that GG exhibited the highest
scavenging capacity (GG > GA > PG). From the results of the 1,1-diphenyl-2-picrylhydrazyl (DPPH•) and 2,2′-
azinobis (3-ethylbenzthiazoline-6-sulfonic acid) (ABTS•+) assays, GG and GA exhibited greater scavenging ca-
pacity than PG (GG = GA > PG). These results suggest that GG may be used as a water-soluble antioxidant
alternative to GA for food and cosmetic applications.
1. Introduction
hydroxyl groups of the glycerol moiety. GG should not only readily
dissolve in aqueous media, but could also be transformed to a lipophilic
1-o-Galloylglycerol (GG) is a water-soluble derivative of gallic acid
(GA), and a phenolic secondary metabolite of many plants, such as
Pelargonium reniforme, Phyllanthus emblica, and Rheum rhabarbarum
(Amir, 2016; Latté, Kaloga, Schäfer, & Kolodziej, 2008; Nonaka &
Nishioka, 1983). The structure of GG consists of one gallate moiety
esterified at the sn-1 position of the glycerol backbone with the other
two hydroxyl groups remaining unesterified. GG was first proposed as
an antioxidant and synthesized de novo using arduous chemical
methods (Takasago, Horikawa, & Masuyama, 1976). In terms of pre-
venting lipid oxidation, GG has been reported as being more effective
than propyl gallate (PG) and tocopherols (Song & Xiao, 1988; Takasago
et al., 1976). GG was isolated from a natural source, and also synthe-
sized using direct chemical esterification of gallic acid (GA) and gly-
cerol catalyzed by p-toluene-sulfonic acid, to acquire a sufficient
quantity to confirm the proposed structure of the natural compound
(Nonaka & Nishioka, 1983). GG has been further characterized, re-
vealing its strong ultraviolet (UV) absorbance in the UVB (280–315 nm)
and UVC (200–280 nm) regions (Artamonov, Nigmatullina,
Aldabergenova, & Dzhiembaev, 1999).
form through well-established chemical or enzymatic esterification to
esterify one or two of the remaining hydroxyl groups with fatty acids.
However, the applications of GG as an antioxidant or as a UV filtering
agent are limited because of its low level of occurrence from natural
sources.
GG is conventionally synthesized by chemical esterifications and
transesterifications using strong acids or bases as catalysts at a high
temperature (> 100 °C). These reactions often destroy a significant
amount of the materials, because the GA moiety is heat sensitive and
susceptible to oxidation. As results, the direct chemical esterification
and transesterification methods can only achieve yields of 41% and
12%, respectively (Artamonov et al., 1999; Nonaka & Nishioka, 1983).
To increase the yield of GG, the GA moiety can be protected using ethyl
chloroformate or thionyl chloride to replace the hydrogen atoms of the
aromatic hydroxyl groups (Song & Xiao, 1988; Takasago et al., 1976).
These protective steps significantly increased the yield of GG, but made
the synthesis process more complicated. Another strategy, the Mitsu-
nobu reaction (Mitsunobu & Yamada, 1967), has also been used to
synthesize the esters of gallic acid (Appendino, Minassi, Daddario,
Bianchi, & Tron, 2002). Even though the reaction conditions were mild,
purifying the product was difficult, and the yield was usually low be-
cause of the complicated mechanism and delicate nature of the
As a GA derivative, GG shares many of its characteristics, such as
antioxidant, UV absorbing, and metal chelating properties.
Theoretically, the solubility of GG should be greatly enhanced by the
⁎ Corresponding author.
E-mail addresses: sz25438@uga.edu (S. Zhang), cakoh@uga.edu (C.C. Akoh).
https://doi.org/10.1016/j.foodchem.2019.125479
Received 29 July 2019; Received in revised form 28 August 2019; Accepted 3 September 2019
Availableonline04September2019
0308-8146/©2019ElsevierLtd.Allrightsreserved.

S. Zhang and C.C. Akoh
FoodChemistry305(2020)125479
Fig. 1. Reaction scheme for GG synthesis and side reactions.
Mitsunobu reaction. The Mitsunobu reaction also requires an oxidizing
azo reagent (diethyl azodicarboxylate), and a reducing phosphine re-
agent (triphenylphosphine), which are both toxic, thus preventing the
use of this reaction to produce GG for use in the food, cosmetics, and
pharmaceutical industries.
Unlike chemical synthesis, enzymatic synthesis does not require
harsh reaction conditions, and therefore is more suitable for producing
bioactive compounds. In a previous study, GG was successful synthe-
validated at the milligram-scale, and the purification of GG could only
be achieved using high-performance liquid chromatography (HPLC)
and thin layer chromatography, which are unsuitable for industrial-
scale production. Although GG has long been proposed as a novel an-
tioxidant, its antioxidant properties have only been tested using the
Schaal oven storage stability test (Song & Xiao, 1988; Takasago et al.,
1976). In addition, its water solubility and partition coefficient have not
yet been determined, because the original scientific motive for syn-
thesizing GG was to increase the hydrophilicity of GA.
sized at a high yield (67.1%
1.9%) by the enzymatic glycerolysis of
PG catalyzed by a commercially available food-grade lipase using gly-
cerol as both reactant and solvent (Zhang & Akoh, 2019). Moreover,
due to steric hindrance between gallate moiety and the secondary hy-
droxyl group of glycerol, the ester bond was only formed at the sn-1
position of the glycerol backbone, and the formation of sn-2 ester was
not detected (Zhang & Akoh, 2019). Similar results were also observed
in the reaction of ethyl ferulate and glycerol catalyzed by the im-
mobilized Candida antarctica lipase B (Novozym 435) under solvent-
free condition and in ionic liquids as reaction medium (Sun et al., 2013;
Sun, Hu, Song, & Bi, 2015; Sun, Shan, Jin, Liu, & Wang, 2007). This
steric hindrance effect was also shown during the reaction of free ferulic
acid and glycerol using ferulic acid esterase as catalyst (Matsuo et al.,
2008). However, the synthesis was only performed, optimized, and
Herein, we report the antioxidant properties, water solubility, n-
octanol/water partition coefficient, and molar extinction coefficients of
GG produced by an optimized large-scale enzymatic synthesis method.
The produced GG was purified from the resultant mixture using liquid-
liquid extraction and its purity was tested using HPLC with a diode
array detector (DAD). Fourier-transform infrared spectroscopy (FT-IR)
was used to determine the functional groups of the synthesized com-
pound. One-dimensional and two-dimensional nuclear magnetic re-
sonance spectroscopy (NMR) were used to determine the exact che-
mical structure of the compound. The antioxidant properties of GG was
evaluated by several in vitro assays, including 1,1-diphenyl-2-picrylhy-
drazyl (DPPH) radical scavenging activity assay, 2,2′-azino-bis (3-
ethylbenzothiazoline-6-sulphonic acid) (ABTS) radical scavenging
2

S. Zhang and C.C. Akoh
FoodChemistry305(2020)125479
assay, ferric reducing antioxidant power (FRAP) assay, and hydrogen
peroxide (H2O2) scavenging assay.
residues. A vacuum-rotary evaporator (40 °C, 50 kPa, 100 rpm) was
used to evaporate the solvent. A beige-colored crude GG solid was
obtained after removing the solvent. The GG product obtained was then
further purified by recrystallization in water to form colorless prisms.
After removing the water, a white GG anhydrous powder was obtained
(m.p. 179–180 °C). The purity of the product was examined with HPLC-
DAD. The recovery of GG was calculated as follows:
2. Materials and methods
2.1. Chemicals and reagents
n-Propyl gallate (99.99% purity) was purchased from HiMedia
Laboratories (Nashik, India). Glycerol (99.9% purity) was purchased
from Hoefer Inc. (San Francisco, CA, USA). Lipozyme® 435 (re-
combinant lipase B from Candida antarctica, expressed in Aspergillus
niger, and immobilized on a macroporous hydrophobic resin, with a
specific activity of 8000 propyl laurate unit g−1, and a moisture content
of 1.0%, w/w) was purchased from Novozymes North America, Inc.
(Franklinton, NC, USA). DPPH was purchased from Alfa Aesar (Ward
Hill, MA, USA). Gallic acid, ABTS diammonium salt, horseradish per-
oxidase (≥250 units/mg), phosphate buffer solution (PBS, 1.0 M,
pH 7.4), and ferrous sulfate heptahydrate (FeSO4·7H2O) were pur-
chased from Sigma-Aldrich (St. Louis, MO, US). 6-hydroxy-2,5,7,8-tet-
ramethylchroman-2-carboxylic acid (≥97%, Trolox™), H2O2 solution
(30% w/w solution), and 2,4,6-tri(2-pyridyl)-s-triazine (TPTZ) were
purchased from ACROS Organic (Morris, NJ, USA). Ferric chloride
anhydrous (FeCl3) was purchased from Fisher Chemical (Fair Lawn, NJ,
USA). All chemicals and reagents were used as received without any
further purification.
Moles of GG obtained
Moles of GG in the reaction mixture
Recovery (%) =
(1)
The results were expressed as mean
SD.
2.4. Structural determination and characterization
The structure of GG was further characterized as described below. A
Nicolet Nexus FT-IR 1100 spectrometer (Thermo Fisher Scientific Co.
Ltd., Waltham, MA, USA) equipped with a ZnSe attenuated total re-
flection attachment was used to collect the FT-IR spectra of samples
(νmax was reported in cm−1). Before each experiment, the instrument
was purged with nitrogen for at least 10 min. Then, 50 mg of the
samples were directly placed onto the ZnSe crystal and pressed using
the attached accessory. The spectra were collected from 650 to
4000 cm−1 with a resolution of 4 cm−1 and 32 scans. The data were
processed using Omnilab software (Omnilab Group, Bremen, Germany)
and the KnowItAll® informatic system (Bio-Rad Laboratories, Hercules,
CA, USA).
2.2. Preparation of 1-o-galloylglycerol
NMR spectroscopy analysis was done with 10 mg of the sample
dissolved in D2O with 10 mM acetic-2-13C acid sodium salt as the in-
ternal standard. 1H, 13C, 1He1H gradient correlation spectroscopy
(gCOSY), 1He13C gradient heteronuclear single quantum coherence
(gHSQC), 1He13C gradient heteronuclear multiple quantum coherence
(gHMQC), and 1He13C gradient heteronuclear multiple bond correla-
tion (gHMBC) spectra were recorded at 25 °C using a Varian Unity
Inova 500 MHz NMR Spectrometer (Varian Inc., Palo Alto, CA, USA)
equipped with a 8-mm hydrogen‑carbon‑nitrogen room temperature
probe. The chemical shifts of GG were reported in parts per million (δ/
ppm). Acetic-2-13C acid sodium salt (δH/δC 1.90/25.63, 164.15 ppm)
was used as the internal standard. The chemical shifts were assigned
based on the 1D and 2D NMR spectra (supplementary material Figs.
S1–S5) as follows: 1H NMR (500 MHz, D2O, 13CH313COONa) δ: 7.04 (s,
2H, H2’ 6′.), 4.37–4.13 (m, 2H, H1), 3.97 (q, J = 5.6 Hz, 1H, H2), 3.64
(dd, J = 11.7, 4.6 Hz, 2H, H3), 13C NMR (125 MHz, D2O,
13CH313COONa) δ: 66.01 (C1), 69.68 (C2), 62.82 (C3), 120.68 (arom.
C1’), 109.78 (arom. C2′, 6′), 144.58 (arom. C3’, 5′), 139.14 (arom. C4’),
168.12 (-COO-).
GG was synthesized by the enzymatic glycerolysis of PG (Fig. 1). A
100 mL double-layer jacketed glass reactor equipped with a circulating
water bath was used to carry out the reaction. A SL 2400 StedFast
stirrer (Fischer Scientific Co., Fair Lawn, NJ, USA) fitted with a PTFE
anchor paddle stirring rod was used to mix the substrates and the en-
zyme at 200 rpm. 6.4 g (30 mmol) PG were first dissolved in 69.1 g
(750 mmol) glycerol, then 18.0 g (23.8%, w/w) Lipozyme® 435 were
added into the reactor after the internal temperature of the substrates
stabilized. The reaction parameters, such as reaction time (120h), re-
action temperature (50 °C), substrate ratio (glycerol/PG = 25/1 mol/
mol), and enzyme load (23.8% w/w) were chosen based on previous
report (Zhang & Akoh, 2019). As heat transfer would likely be less ef-
ficient in a large-scale reaction than in a milligram-scale reaction, the
reaction temperature and reaction time were further optimized. The
reaction mixture was sampled (2 μL) periodically (every 24 h) and the
sample was diluted to 1 mL with methanol for quantitative analysis. All
experiments were carried out in triplicate.
The reaction mixture was analyzed as described previously (Zhang
& Akoh, 2019). An Agilent1260 Infinity HPLC system (Santa Clara, CA,
USA) equipped with a diode-array detector (DAD) scanned at 280 nm
were used for quantification purposes. A reverse phase C18 column
(Ultrasphere ODS, 5 μm, 250 × 4.6 mm, Hichrom Ltd., Theale, UK) was
used at a controlled temperature of 35 °C. The yield of GG, conversion
of PG, and hydrolysis during the reaction were calculated as in a pre-
vious study (Zhang & Akoh, 2019). The results were expressed as mean
The UV–Vis spectra of the samples were measured using a UV-1601
UV–Vis spectrophotometer (Shimadzu, Kyoto, Japan). The samples
were dissolved in methanol at a concentration of 50 μM, then scanned
over the range from 190 to 700 nm.
The solubility of GG, PG, and GA were determined as described in
previous studies (Daneshfar, Ghaziaskar,
&
Homayoun, 2008;
Tsuchiyama, Sakamoto, Tanimori, Murata, & Kawasaki, 2007). Briefly,
an excess of the chemicals was suspended in ultrapure water and in-
cubated in a shaking water bath (C76 Water Bath Shaker, New Bruns-
wick Scientific, Edison, NJ, USA) at 100 rpm and 25 °C for 24 h. The
mixture was then centrifuged, and the concentrations of the samples in
the supernatant were determined with HPLC-DAD as described pre-
viously, using the corresponding standard curves for accurate quanti-
fication. The experiments were conducted in triplicate and reported as
values
standard deviation (SD).
2.3. Separation and purification of reaction product
The reaction product was separated from the reaction mixture using
liquid-liquid extraction, according to a previous study with modifica-
tions (Holser et al., 2008). Briefly, a volume of sodium chloride solution
(200 g/L) twice that of the reaction mixture was added to the reaction
mixture to decrease its viscosity. After the enzyme was removed in a
Buchner funnel with a filter paper, ethyl acetate was used to extract GG
from the mixture using a separation funnel. The ethyl acetate phases
were then pooled together and washed with saturated sodium chloride
solution (with 1 mM sodium carbonate) to remove glycerol and GA
mean
SD.
The n-octanol/water partition coefficient was determined using
shaking flask method (Sangster, 1989; Short et al., 2010). Equal amount
(5 mL) of n-octanol and water were added to a flask, then sealed and
stirred in a shaking water bath at 100 rpm and 25 °C for 24 h to reach
mutual saturation of the phases. 10 mg of GG, PG, and GA were added
to the flasks, respectively. The flasks were sealed again and stirred at
3

S. Zhang and C.C. Akoh
FoodChemistry305(2020)125479
the same condition for 72 h. After the layers were separated using a
separating funnel, aliquots of both layers were taken and diluted in
methanol. The concentration of compounds in each layer was de-
termined by the method described previously. The distribution coeffi-
cient (log D) and partition coefficient (log P) were calculated according
to the following equations:
2.7. FRAP assay
The FRAP assay was conducted in accordance with previous studies
(Benzie & Strain, 1996; Ozgen, Reese, Tulio, Scheerens, & Miller, 2006)
with modifications. The FRAP reagent was prepared by mixing solu-
tions of 10 mM TPTZ (in 40 mM HCl), 20 mM FeCl3, and 300 mM
acetate buffer (pH 3.6) at a volumetric ratio of 1:1:10. Aqueous solu-
tions of FeSO4·7H2O were prepared at concentrations of 100, 200, 500,
and 1000 μM, to be used for calibration. Antioxidant samples of GG, PG,
and GA were also dissolved in water at various concentrations (20, 40,
100, 200, and 400 μM). Water was used as the blank. Freshly prepared
FRAP reagent (900 μL) was mixed with 30 μL of the sample solution,
then with 90 μL water. Thus, the final dilution of the sample in the
reaction mixture was 34 times. The reaction mixture was then in-
cubated in the dark at 37 °C for 4 min. The absorbance of the reaction
mixture was monitored at 593 nm. The increase in absorbance was
checked with the calibration curve of FeSO4•7H2O solution to give the
results expressed as Fe2+ equivalents.
log Doctanol/water
Total concentration of the compounds in octanol phase
Total concentration of the compounds in water phase
⎛
⎜
⎞
⎟
= log
(2)
⎝
⎠
log P
octanol/water
Concentration of the unionized compound in octanol phase
Concentration of the unionized compound in water phase
⎛
⎞
⎟
= log
⎜
⎝
⎠
(3)
For unionizable compounds, such as GG and PG, the partition
coefficient and distribution coefficient are the same. For ionizable
compounds, such as GA, the difference between log P and log D is
negligible when the ionization is suppressed by a buffer of suitable pH.
This could also be calculated according to the following equation:
2.8. H2O2 scavenging assay
The hydroxyl radical scavenging ability of the compounds was
measured using the H2O2 scavenging assay (Pick & Keisari, 1980; Sroka
& Cisowski, 2003) with modifications. Solutions of GG, PG, and GA in
methanol were prepared at various concentrations (20, 40, 200, and
400 μM). Equal amounts of the sample solution and 0.002% (w/w)
H2O2 solution were mixed with 0.8 mL PBS and incubated in the dark at
37 °C for 10 min. One milliliter assay reagent, containing 0.2 mg/mL
phenol red and 0.1 mg/mL horseradish peroxidase in PBS, was added
and incubated under the same conditions for 15 min. After incubation,
50 μL of 1 M sodium hydroxide solution were added and the absorbance
of the mixture was measured immediately at 610 nm using spectro-
photometer. H2O2 solutions at different concentrations (0.0002,
0.0005, 0.001, and 0.002%, w/w) were used as the calibration curve.
The decrease in absorbance (compared with 0.002% w/w H2O2 solu-
tion) was expressed against the calibration curve of the H2O2 solutions.
The results were expressed as the percentages of scavenged H2O2.
The results of all four antioxidant assays were calculated using re-
gression analysis and ANOVA by JMP® software (version 13.2.0, SAS
Institute, Inc., Cary, NC, USA), and presented as means followed by
standard errors.
= log Doctanol/water + log (1 + 10(pH−pK )
)
a
log P
(4)
octanol/water
where pH is the measured pH of the octanol/water system, and pKa is
the dissociation constant of the solute. As shown in the Eq. (4), when
the pH is equal or smaller than pKa, the difference between log P and
log D is negligible. The log P value for GA was also calculated according
to Eq. (3) and measured in the octanol/water system using 0.1% tri-
fluoroacetic acid to suppress the ionization of the GA molecules. The
results were expressed as mean
SD of triplicate determinations.
2.5. DPPH• scavenging assay
The assay was performed according to a previous study (Compton,
Laszlo, & Evans, 2012) with some modifications. Samples of GG, PG,
and GA were dissolved in methanol at different concentrations (2, 5, 10,
and 20 μM). DPPH was dissolved in methanol to make a solution at a
concentration of 200 μM. Equal amounts of DPPH• and sample solutions
were mixed and then monitored spectrophotometrically at 517 nm at a
1 s interval for 30 min using the UV-1601UV-Vis spectrophotometer
mentioned previously. Instead of antioxidant solution, ultrapure water
was used in the control groups. The results were expressed as the re-
maining percentage of DPPH• after being reduced by the samples. All
experiments were performed in triplicate. All the samples and reagents
were freshly prepared daily.
3. Results and discussion
3.1. Synthesis of 1-o-galloylglycerol and structure confirmation
2.6. ABTS•+ scavenging assay
Initially, the reaction was carried out in a 50-mL double-layer
jacketed glass reactor stirred with a magnetic stir bar. In this system,
the product yield was only 29.8%, much lower than the previous mil-
ligram-scale synthesis (Zhang & Akoh, 2019). Such a low yield could be
attributed to the limited mass transfer using the magnetic stir bar in the
large reactor. Therefore, the enzymatic glycerolysis of PG was per-
formed in a 100-mL double-layer jacketed glass reactor equipped with a
circulating water bath, stirred with a PTFE anchor paddle stirring rod,
and using glycerol as both the reactant and solvent. Compared with our
previous study (Zhang & Akoh, 2019), the present synthesis was scaled-
up by 30 times from our previous report. Thus, the reaction was further
optimized by a full factorial design via three time-course reactions.
Fig. 2 shows that the yield of GG increased with reaction time, but
the rate of increase declined significantly after 120 h. Quadratic
(second-order) polynomial models could be used to explain the effect of
reaction time on yield at all three temperatures with R2 values of these
models > 0.99, agreeing with our previous study (Zhang & Akoh,
2019). Hydrolysis also increased with reaction time, but the rate of
increase did not significantly change with time (low set of lines). Hy-
drolysis at all three temperatures could be fitted to linear models, with
An improved ABTS radical cation decolorization assay (Re et al.,
1999) with modifications (Phonsatta et al., 2017) was used. Briefly,
ABTS•+ was produced by reacting 7 mM ABTS water solution with
2.45 mM potassium persulfate in the dark at room temperature for 16 h.
The ABTS•+ solution was then diluted with ethanol to obtain an ab-
sorbance of 0.70 ( 0.01) at a wavelength of 734 nm. Samples of GG,
PG, and GG were dissolved in ethanol to obtain solutions at con-
centrations of 10, 20, 40, 50, and 100 μM. Trolox™ was used as the
standard for measuring the antioxidant activity of the samples. Ethanol
solutions of Trolox™ were prepared at concentrations of 10, 20, 50, 100,
and 200 μM. Ethanol was used for the control groups. A 100-μL sample
was mixed with 900 μL of the ABTS•+ solution, then the mixture was
incubated in the dark at 30 °C for 6 min. The absorbance of the mixture
was measured with the spectrophotometer mentioned previously at
734 nm. The results were expressed as the decrease in absorbance after
mixing the samples, compared with the control groups. All experiments
were performed in triplicate. All the samples and reagents were freshly
prepared daily.
4

S. Zhang and C.C. Akoh
FoodChemistry305(2020)125479
Yield in small-scale reaction
100
90
80
70
60
50
40
30
20
10
0
(Zhang & Akoh, 2019).
Yield at 50 °C
The UV–Vis spectrophotometry results (Table 1 and supplementary
material Fig. S6) showed that GG, GA, and PG shared a similar UV–Vis
absorption pattern with slightly different wavelengths of maximum
absorptions (λmax). For all three compounds, strong UV absorbances
were observed around 220 and around 275 nm, with no absorbance in
the visible light region. GG has a higher molar attenuation coefficient
(ε) at around 275 nm than both PG and GA, which may suggest that GG
is a more efficient UV filtering agent than GA and PG. However, as
shown in supplementary material Fig. S6, the UV absorption peaks of
all three compounds were located in UVC region (200–280 nm), only
weak absorbances were observed in UVB region (280–315 nm) while no
absorbance was observed in UVA region (315–400 nm). Thus, the po-
tential of using GG as a broad-spectrum sunscreen ingredient maybe
limited.
The detailed report of the FT-IR spectra and their interpretation are
shown in supplementary material (Fig. S7). The peak at 1686 cm−1
indicated the existence of the ester linkage in GG, which shifted from
the typical signal at 1740 cm−1 owing to the influence of the aromatic
ring. The presence of a pair of sharp bands at 1278 and 1193 cm−1
indicated the asymmetrical CeO bond of the ester linkage. A sharp
signal at 1613 cm−1 represented the C]C stretching of the aromatic
ring, which shifted from 1609 cm−1 in GA, and from 1606 cm−1 in PG.
The alkyl alcohol of the glycerol moiety (R-CH2-OH) in the GG molecule
was detected in the region from 3400 to 3200 cm−1 for the OH hy-
drogen bond stretching, the medium-weak peak at 1446 cm−1 for OH
deformation, and the strong peak at 1024 cm−1 for CeO stretching.
Combining the broad peak in the region from 3400 to 3200 cm−1 (OH
stretching), the strong peak at 1312 cm−1 (OH deformation), and the
strong peak at 1024 cm−1 (CeO stretching), the secondary alcohol of
the glycerol moiety (Ph-CHR-OH) was identified.
Yield at 55 °C
Yield at 60 °C
Hydrolysis in small-scale reaction
Hydrolysis at 50 °C
Hydrolysis at 55 °C
Hydrolysis at 60 °C
0
24
48
72
96 120 144 168 192 216 240
Time (h)
Fig. 2. Effect of the reaction temperature and reaction time on yield and hy-
drolysis (fixed-reaction conditions: glycerol/PG = 25:1, and enzyme load
23.8%).
R2 values equal or > 0.92, meaning that over 92% of the variance could
be explained by the linear models. Considering the effect of the reaction
time on both the yield and hydrolysis, 120 h was selected as the optimal
reaction time. A reaction temperature of 55 °C provided a significantly
higher yield than that at 50 °C or 60 °C. A reaction temperature of 50 °C
produced a temperature at the center of the reaction mixture between
45 and 47 °C after reaching thermal equilibrium. Our previous study
showed that the yield of GG was significantly higher when the reaction
temperature was 50 °C compared to 45 °C (Zhang & Akoh, 2019). Thus,
the lower yield at 50 °C in the present study could be explained by
insufficient heat transfer in the larger system. When the reaction tem-
perature was set to 60 °C, the temperature at the center of reaction
mixture was 55 to 56 °C. Therefore, this lower yield could be explained,
in part, by the partial thermal deactivation of the enzyme at higher
temperatures. At 55 °C, the temperature at the center of the reaction
mixture was 49 to 52 °C, and these temperatures were closest to the
optimum established in our previous report (Zhang & Akoh, 2019).
Thus, the optimal reaction temperature was chosen as 55 °C. Fig. 2
shows that the yield from this reaction under optimal conditions (55 °C,
25:1 glycerol:PG substrate molar ratio, 120 h, and 23.8% enzyme load
relative to the total weight of the substrates) was higher than the op-
3.2. Characterization of GG
The water solubility, log D, and log P of GG, GA and PG are shown
in Table 1. The water solubility of GG was > 4 times and 16 times
higher than that of GA and PG, respectively. This increase could be
attributed to the hydroxyl groups of the glycerol moiety. The solubility
of GG is also less likely to be reduced under acidic pH conditions. The
most acidic pKa of GG was 8.11, calculated using Chemicalize (https://
chemicalize.com/ developed by ChemAxon http://www.chemaxon.
com), indicated that its solubility was stable at pH < 7.0. In alkaline
solutions, the solubility of GG would increase further because of the
ionization of the hydroxyl groups. GA is sparingly soluble in water, with
its solubility being highly dependent on pH. Using the same calculation
for GG, the most acidic pKa of GA was 3.94, while significant ionization
of GA molecule starts at pH 2.5 and 99% being ionized at pH 6.0. Thus,
the solubility of GA decreased significantly at a pH of < 6.0. Con-
sidering its higher solubility and stability in acidic conditions, GG
would be a better antioxidant and UV filtering agent in aqueous ap-
plications, such as fruit juice and water-based sunscreens.
timal yield (69.1%
provement could be attributed to the more efficient mass transfer using
the anchor paddle stirring rod compared with the magnetic stirring bar.
1.9%) obtained in our previous study. This im-
After the extraction, about 92.2%
3.6% of the GG was recovered
from the reaction mixture. The purified product consisted of > 98% of
GG, about 1% GA, and a trace amount of PG. During the extraction, the
amount of GA, PG, and GG were monitored using HPLC with the con-
dition mentioned previously. No hydrolysis of GG or PG was observed.
The purified GG was structurally characterized by UV–Vis, FT-IR and
NMR spectroscopies. The results of the unidimensional NMR and 2D
NMR analyses (Supplementary material Figs. S1–S5) indicated that the
primary hydroxyl group of glycerol had been esterified with the GA
moiety. A detailed interpretation can be found in our previous study
Log P is the standard logarithmic scale for evaluating the hydro-
phobicity of compounds (Andersson & Schräder, 1999) and is measured
using the concentration of the unionized solute (eq. (3)). For ionizable
compounds, log P can be measured either directly by adjusting the pH
Table 1
UV–Vis absorbance, water solubility, distribution coefficient, and partition coefficient of 1-o-galloylglycerol (GG), gallic acid (GA), and propyl gallate (PG).
Log ε (log M−1 cm−1
)
Water solubility (g/L)
Log Da
Log Pa
MeOH
Compounds
λmax
(nm)
GG
GA
PG
221, 276
220, 274
221, 276
4.47, 4.10
4.47, 4.05
4.39, 4.02
55.92
13.19
3.39
0.30
0.09
0.04
−0.63
0.28
1.65
0.01
0.01b
0.03
0.35
0.03
a
Measured in n-octanol/water (5 mL/5 mL) system, log D equals log P when compounds are not ionizable (GG and PG).
The pH of aqueous phase was 3.2 after 72 h.
b
5

S. Zhang and C.C. Akoh
FoodChemistry305(2020)125479
of the aqueous phase or by calculating it from a known log D value
using eq. (4). Unlike log P, log D is the function for all forms of the
compound, both ionized and unionized, as shown in Eq. (2). For io-
nizable compounds, like GA, the concentration of compounds in aqu-
eous phase is contributed by two parts, ionized molecules and union-
ized molecules. When the pKa of a compound and the pH of the aqueous
phase are significantly different, the concentration of the ionized mo-
lecules is not negligible (Sangster, 1989). Table 1 shows that the log P
value of GA was higher than its log D value. This could be explained by
the ionization of GA in the aqueous phase. In particular, the pH of the
aqueous phase changed according to the amount of GA used during the
shaking flask test. The pH of the aqueous phase, in turn, affected the
measured log D values of GA. In many studies (Asnaashari, Farhoosh, &
Sharif, 2014; Farhoosh, Johnny, Asnaashari, Molaahmadibahraseman,
& Sharif, 2016; Lu, Nie, Belton, Tang, & Zhao, 2006), the difference in
the definition between log P and log D was not noticed, so the reported
log P values of GA were actually their log D values. Because the amount
of GA and/or the buffer used as the aqueous phase in these studies were
different, the reported log P values were actually their log D values
under different pH conditions, thereby causing a significant incon-
sistency in the reported values. For example, when a smaller amount of
GA was used, the aqueous phase would have a higher pH, so that more
ionized molecules would have appeared, causing lower log D reading,
so that the reading was mistakenly recorded as log P. In the present
study, the log P value of GA was measured at the aqueous phase with
pH adjusted and also calculated using Eq. (4). While two methods gave
the same log P value, it was significantly larger than the log D value
measured in n-octanol/water system with no buffer to adjust the pH.
Similar to the pattern for water solubility, the log P value of GG was
also significantly decreased, compared with those of its acid and propyl-
ester forms. A high log P value of PG implies a high penetration of the
cellular membrane and hence a high potential for its collapse, which
would result in a higher cytotoxicity than GA (Galati, Lin, Sultan, &
O'Brien, 2006). With the lowest log P value and a much lower acidity
than GA, GG could be a safer alternative to GA for use in food, cosmetic,
and pharmaceutical products. Noticeably, no hydrolysis of PG or GG
was observed throughout the solubility test or during the measurement
of log D and log P.
The DPPH• assay measures the reducing ability of antioxidants to-
wards DPPH•, which is a stable organic nitrogen radical. Figs. 3 and 4a
illustrate that GG and GA showed a similar scavenging ability towards
DPPH• with PG being slightly lower. The scavenging ability towards
DPPH• depends on structural features, such as the dissociation energy of
hydrogen atoms from hydroxyl groups, resonance delocalization of the
phenol radical (PheO•), and steric hindrance arising from bulky groups
substituting hydrogen in the aromatic ring (Shahidi & Naczk, 1995).
After H• is taken by DPPH•, PheO• can react either with DPPH• to form
DPPH-PheO or undergo a termination reaction to form PheO-PheO
(Sánchez-Moreno, Larrauri, & Saura-Calixto, 1998). All three tested
compounds had the same number of phenolic hydroxyl groups, so the
differential ability to scavenge DPPH• could only be affected by the
moiety bonded with the carbonyl group. The lower DPPH• scavenging
ability of PG compared with GA has been observed before (Lu et al.,
2006). This phenomenon was explained by steric hinderance caused by
the n-propyl moiety and the higher hydrophobicity of PG making an-
tioxidants less available for the DPPH radicals in polar systems
(Asnaashari et al., 2014; Lu et al., 2006). The secondary DPPH•
scavenging reaction of PG could also be terminated via the reaction
mentioned above, causing lower scavenging ability (Okuda, 1993). The
glycerol moiety may have caused steric hindrance, but also increased
the hydrophilicity of the compound, so that the DPPH• scavenging
ability of GG was not significantly different to that of GA. Fig. 3 shows
that GG and GA exhibited a faster reduction towards DPPH• than PG,
indicating that the accessibility of the molecule affected its scavenging
activity in the DPPH• assay. Compared with α-tocopherol, other phe-
nolic acids and their esters, such as ferulic acid and caffeic acid, GA and
its esters usually exhibited a much higher scavenging ability towards
DPPH• (Kikuzaki, Hisamoto, Hirose, Akiyama, & Taniguchi, 2002),
suggesting a great potential for GG to be used as an antioxidant.
The ABTS•+ and FRAP assays measure the reducing ability of anti-
oxidants towards two ions (Fe3+-TPTZ and ABTS•+, respectively) with
a similar redox potential under different pH conditions. The FRAP assay
was conducted in acetate buffer (pH 3.6) and the ABTS•+ assay in
ethanol. In both assays, PG exhibited the lowest antioxidant activity,
which agreed with other studies (Phonsatta et al., 2017). GG and GA
exhibited similar antioxidant activities in the ABTS•+ assay while GG
exhibited a higher antioxidant activity than GA in the FRAP assay
(Table 2 and Figs. 4b, c). The lower antioxidant activity of GA observed
in the FRAP assay could be attributed to the acidic pH of the testing
environment as observed previously (Ozgen et al., 2006). As mentioned
earlier, the ionization of GA is suppressed in acidic solutions, so the
mechanism of the FRAP assay is only electron transfer rather than a
combination of electron transfer and hydrogen atom transfer, as in the
DPPH+ and FRAP assays. These results may imply that it is easier for
ionized GA to transfer electrons than for unionized GA. Similar to
DPPH• assay, GA showed higher reducing ability than α-tocopherol,
ferulic acid, caffeic acid, and ascorbic acid in ABTS•+ assay (Miller &
Rice-Evans, 1997).
3.3. Antioxidant activity
The antioxidant properties of GG were determined by four com-
monly used in vitro antioxidant assays. The DPPH•, ABTS•+, FRAP, and
H2O2 scavenging assays are homogeneous antioxidant assays which
evaluate the hydrogen atom and electron donating ability of anti-
oxidants. For the DPPH•, ABTS•+, and H2O2 scavenging assays, the
antioxidant activity of compounds can be expressed as the effective
concentration for obtaining a 50% response (EC50) or the inhibitory
concentration at 50% response (IC50), depending on how the response
is defined. In the present study, IC50 (the concentration of a tested
compound needed to reduce the free radical to 50% of its initial con-
centration) was used to quantify the antioxidant activity of the tested
compounds, while EC50 (the concentration of a tested compound
needed to scavenge 50% of the free radicals) was used in the ABTS•+
and H2O2 scavenging assays. The results of the FRAP assay were ex-
pressed as the effective concentration for obtaining 1 mM Fe2+ (EC1).
The results of all four assays could also be expressed as the standard
equivalent antioxidant activity (usually using Trolox equivalent anti-
oxidant activity, TEAC). Trolox is a water-soluble vitamin E analogue.
The Trolox equivalent was also used for the ABTS•+ and H2O2 assays to
obtain the relative antioxidant activities of GG, GA, and PG, compared
with Trolox.
Unlike the DPPH•, ABTS•+, and FRAP assays, the H2O2 scavenging
assay is a competitive system, where antioxidants have to compete with
phenol red to react with HO• radicals generated from H2O2.
Consequently, the H2O2 scavenging assay measure a combination of
reducing ability and reactive speed towards the HO• radical. GG ex-
hibited the highest H2O2 scavenging ability while GA and PG exhibited
similar activities towards HO• radicals (Table 2 and Fig. 4d). The in-
fluence of substituents bonded to the benzene ring of GA has been
observed before. Pyrogallol (1,2,3-trihydroxybenzene) has been re-
ported to have a lower scavenging ability towards H2O2 than GA (Sroka
& Cisowski, 2003). The different scavenging abilities of the three
compounds tested in the present study may have been caused by the
influence of the different moieties bonded with the carbonyl group of
the GA (n-propyl alcohol, and glycerol) moiety on the hydroxyl groups
of the phenol ring. The glycerol moiety in GG may have increased the
electrophilicity of the carbonyl electron, which would increase the
Table 2 and Fig. 4 show that GG exhibited a higher antioxidant
activity than both GA and PG in the FRAP and H2O2 assays. From the
DPPH• and ABTS•+ assays, GG and GA exhibited a similar antioxidant
activity while PG had a significantly lower antioxidant activity.
6

S. Zhang and C.C. Akoh
FoodChemistry305(2020)125479
Table 2
Antioxidant activities of 1-o-galloylglycerol (GG), gallic acid (GA), and propyl gallate (PG) in DPPH•, ABTS•+, FRAP, and H2O2 assays.
Compounds
DPPH•
ABTS•+
FRAP
H2O2
IC50
EC50
TEAC⁎
EC1
EC50
TEAC⁎
(μM)
(μM)
(μM)
(μM)
GG
GA
PG
10.51
10.50
11.90
0.59a
0.58a
0.61b
28.50
28.50
32.39
0.76a
0.65a
0.59b
5.0
4.95
4.33
0.61a
0.53a
0.41b
287.91
306.64
387.55
5.56a
5.24b
197.27
237.75
235.79
10.75a
7.89b
9.63b
2.05
1.74
1.73
0.36a
0.40b
0.41b
15.32c
a,b,cDifferent letters indicate significant statistical difference at p < 0.05.
All the results are expressed as means (n = 3)
standard errors.
⁎
Trolox equivalent antioxidant activity, molar/molar.
Fig. 3. DPPH• scavenging kinetics in the presence of GG, GA, and PG at 20 (a), 10 (b), 5 (c), and 2 (d) μM in methanol.
hydrogen atom donating ability of the phenolic hydroxyl groups via an
inductive effect. Comparing the chemical shift of two hydrogen atoms
in the benzene ring, GG has been reported to have a higher chemical
shift (7.04 ppm), than GA (6.92 or 6.91 ppm) and PG (6.97 ppm)
(Garrido, Garrido, & Borges, 2012; López-Martínez, Santacruz-Ortega,
Navarro, Sotelo-Mundo, & González-Aguilar, 2015). A higher chemical
shift of H atoms in the benzene ring indicates a decreased electron
density for these two atoms (electrons shifting towards carbon atoms),
which further indicates an increased electrophilicity of the benzene ring
caused by the glycerol moiety. This increased electrophilicity of the
benzene ring makes electron transfer and hydrogen atom transfer of Ph-
OH easier to occur. The increased H2O2 scavenging ability of GG could
also be explained by the formation of an intra-molecular hydrogen
bond, which could remain stable in an aqueous solution. Comparing the
FT-IR spectra of GG and GA (Supplementary material Fig. S7), red shifts
hydrogen atoms, thus increasing the H2O2 scavenging activity of GG.
Notably, the red shifts, observed in the spectrum of GA in the region
from 3000 to 4000 cm−1 compared with the spectrum of PG, could also
have been caused by the inter-molecular hydrogen bond between GA
molecules.
4. Conclusions
A readily water-soluble natural plant component, GG, has been
synthesized and characterized. Unlike commonly used alkyl gallates,
GG is more hydrophilic than GA, and has the potential to be used in
aqueous-based foods, cosmetics, and pharmaceutical products. GG was
be readily prepared enzymatically and was found to be more effective
as an antioxidant in various in vitro assays than PG. Its antioxidant
activity was similar or higher when compared with GA. As GG is less
acidic, less likely to penetrate cellular membranes, and less likely to be
cytotoxic, it may be a safer alternative antioxidant than GA and PG.
were observed in the GG spectrum in the region 3000–4000 cm−1
,
which indicated the possibility of hydrogen bonds being formed be-
tween Ph-OH and the hydroxyl groups of the glycerol moiety. These
intra-molecular hydrogen bonds would promote the transfer of
7

S. Zhang and C.C. Akoh
FoodChemistry305(2020)125479
Fig. 4. Antioxidant activity of GG, GA, and PG, evaluated by DPPH• scavenging assay (a), ABTS•+ scavenging assay (b), FRAP assay (c), and H2O2 scavenging assay
(d).
Declaration of competing interest
The authors declare that they have no conflicts of interest.
Acknowledgement
glycerol derivatives. Industrial Crops and Products, 36(1), 217–221.
Daneshfar, A., Ghaziaskar, H. S., & Homayoun, N. (2008). Solubility of gallic acid in
methanol, ethanol, water, and ethyl acetate. Journal of Chemical and Engineering Data,
53(3), 776–778.
Farhoosh, R., Johnny, S., Asnaashari, M., Molaahmadibahraseman, N., & Sharif, A.
(2016). Structure–antioxidant activity relationships of o-hydroxyl, o-methoxy, and
alkyl ester derivatives of p-hydroxybenzoic acid. Food Chemistry, 194, 128–134.
Galati, G., Lin, A., Sultan, A. M., & O'Brien, P. J. (2006). Cellular and in vivo hepato-
toxicity caused by green tea phenolic acids and catechins. Free Radical Biology and
Medicine, 40(4), 570–580.
Garrido, J., Garrido, E. M., & Borges, F. (2012). Studies on the food additive propyl
gallate: synthesis, structural characterization, and evaluation of the antioxidant ac-
tivity. Journal of Chemical Education, 89(1), 130–133.
Holser, R., Mitchell, T., Harry-O'Kuru, R., Vaughn, S., Walter, E., & Himmelsbach, D.
(2008). Preparation and characterization of 4-methoxy cinnamoyl glycerol. Journal of
the American Oil Chemists' Society, 85(4), 347–351.
The authors would like to thank Food Science Research, University
of Georgia for financially supporting this work. Special thanks go to our
lab technician, Joseph Robert Hyatt, for his technical assistance, and to
Dr. John Glushka from the CCRC NMR spectroscopy facility of UGA for
assisting with the NMR analysis.
Appendix A. Supplementary data
Kikuzaki, H., Hisamoto, M., Hirose, K., Akiyama, K., & Taniguchi, H. (2002). Antioxidant
properties of ferulic acid and its related compounds. Journal of Agricultural and Food
Chemistry, 50(7), 2161–2168.
Latté, K. P., Kaloga, M., Schäfer, A., & Kolodziej, H. (2008). An ellagitannin, n-butyl
gallate, two aryltetralin lignans, and an unprecedented diterpene ester from
Pelargonium reniforme. Phytochemistry, 69(3), 820–826.
Supplementary data to this article can be found online at https://
doi.org/10.1016/j.foodchem.2019.125479.
López-Martínez, L. M., Santacruz-Ortega, H., Navarro, R.-E., Sotelo-Mundo, R. R., &
González-Aguilar, G. A. (2015). A 1H NMR investigation of the interaction between
phenolic acids found in mango (Manguifera indica cv Ataulfo) and papaya (Carica
papaya cv Maradol) and 1, 1-diphenyl-2-picrylhydrazyl (DPPH) free radicals. PLoS
One, 10(11), e0140242.
Lu, Z., Nie, G., Belton, P. S., Tang, H., & Zhao, B. (2006). Structure-activity relationship
analysis of antioxidant ability and neuroprotective effect of gallic acid derivatives.
Neurochemistry International, 48(4), 263–274.
Matsuo, T., Kobayashi, T., Kimura, Y., Tsuchiyama, M., Oh, T., Sakamoto, T., & Adachi, S.
(2008). Synthesis of glyceryl ferulate by immobilized ferulic acid esterase.
Biotechnology Letters, 30(12), 2151–2156.
Miller, N. J., & Rice-Evans, C. A. (1997). Factors influencing the antioxidant activity
determined by the ABTS•+ radical cation assay. Free Radical Research, 26(3),
195–199.
Mitsunobu, O., & Yamada, M. (1967). Preparation of esters of carboxylic and phosphoric
acid via quaternary phosphonium salts. Bulletin of the Chemical Society of Japan,
40(10), 2380–2382.
Nonaka, G., & Nishioka, I. (1983). Tannins and related compounds. X. Rhubarb (2):
Isolation and structures of a glycerol gallate, gallic acid glucoside gallates, galloyl-
glucoses and isolindleyin. Chemical and Pharmaceutical Bulletin, 31(5), 1652–1658.
References
Amir, D. E. (2016). HPLC profiling of Phyllanthus emblica L. extracts. Planta Medica.,
82(05), 10.
Andersson, J. T., & Schräder, W. (1999). A method for measuring 1-octanol− water
partition coefficients. Analytical Chemistry, 71(16), 3610–3614.
Appendino, G., Minassi, A., Daddario, N., Bianchi, F., & Tron, G. C. (2002).
Chemoselective esterification of phenolic acids and alcohols. Organic Letters, 4(22),
3839–3841.
Artamonov, A., Nigmatullina, F., Aldabergenova, M., & Dzhiembaev, B. Z. (1999).
Synthesis of α-monoglycerides of aromatic acids. Chemistry of Natural Compounds,
35(4), 404–408.
Asnaashari, M., Farhoosh, R., & Sharif, A. (2014). Antioxidant activity of gallic acid and
methyl gallate in triacylglycerols of Kilka fish oil and its oil-in-water emulsion. Food
Chemistry, 159, 439–444.
Benzie, I. F. F., & Strain, J. J. (1996). The ferric reducing ability of plasma (FRAP) as a
measure of “antioxidant power”: The FRAP assay. Analytical Biochemistry, 239(1),
70–76.
Compton, D. L., Laszlo, J. A., & Evans, K. O. (2012). Antioxidant properties of feruloyl
8

S. Zhang and C.C. Akoh
FoodChemistry305(2020)125479
Okuda, T. (1993). Natural polyphenols as antioxidants and their potential use in cancer
prevention. Polyphenolic Phenomena, 221–235.
Safety, 73(6), 1484–1489.
Song, Z., & Xiao, Z. (1988). The preparation of 1-glycerol gallate and study on its anti-
oxidation properties. Chemistry and Industry of Forest Products, 3, 001.
Sroka, Z., & Cisowski, W. (2003). Hydrogen peroxide scavenging, antioxidant and anti-
radical activity of some phenolic acids. Food and Chemical Toxicology, 41(6),
753–758.
Sun, S., Hu, B., Song, F., & Bi, Y. (2015). Ionic liquids improve the selective synthesis of
hydrophilic glyceryl ferulates by the enzymatic transesterification of ethyl ferulate
with monostearin: Comparison with organic solvents. European Journal of Lipid
Science and Technology, 117(8), 1171–1178.
Sun, S., Qin, F., Bi, Y., Chen, J., Yang, G., & Liu, W. (2013). Enhanced transesterification
of ethyl ferulate with glycerol for preparing glyceryl diferulate using a lipase in ionic
liquids as reaction medium. Biotechnology Letters, 35(9), 1449–1454.
Sun, S., Shan, L., Jin, Q., Liu, Y., & Wang, X. (2007). Solvent-free synthesis of glyceryl
ferulate using a commercial microbial lipase. Biotechnology Letters, 29(6), 945–949.
Takasago, M., Horikawa, K., & Masuyama, S. (1976). Studies on derivatives of aromatic
hydroxycarboxylic acids as antioxidant. I. Journal of Japan Oil Chemists' Society, 25(1),
16–20.
Tsuchiyama, M., Sakamoto, T., Tanimori, S., Murata, S., & Kawasaki, H. (2007).
Enzymatic synthesis of hydroxycinnamic acid glycerol esters using type a feruloyl
esterase from Aspergillus niger. Bioscience, Biotechnology, and Biochemistry, 71(10),
2606–2609.
Ozgen, M., Reese, R. N., Tulio, A. Z., Scheerens, J. C., & Miller, A. R. (2006). Modified 2,2-
azino-bis-3-ethylbenzothiazoline-6-sulfonic acid (ABTS) method to measure anti-
oxidant capacity of selected small fruits and comparison to ferric reducing anti-
oxidant power(FRAP)and2,2′-diphenyl-1-picrylhydrazyl(DPPH)methods. Journal of
Agricultural and Food Chemistry, 54(4), 1151–1157.
Phonsatta, N., Deetae, P., Luangpituksa, P., Grajeda-Iglesias, C., Figueroa-Espinoza, M. C.,
Le Comte, J., ... Panya, A. (2017). Comparison of antioxidant evaluation assays for
investigating antioxidative activity of gallic acid and its alkyl esters in different food
matrices. Journal of Agricultural and Food Chemistry, 65(34), 7509–7518.
Pick, E., & Keisari, Y. (1980). A simple colorimetric method for the measurement of
hydrogen peroxide produced by cells in culture. Journal of Immunological Methods,
38(1), 161–170.
Re, R., Pellegrini, N., Proteggente, A., Pannala, A., Yang, M., & Rice-Evans, C. (1999).
Antioxidant activity applying an improved ABTS radical cation decolorization assay.
Free Radical Biology and Medicine, 26(9), 1231–1237.
Sánchez-Moreno, C., Larrauri, J. A., & Saura-Calixto, F. (1998). A procedure to measure
the antiradical efficiency of polyphenols. Journal of the Science of Food and Agriculture,
76(2), 270–276.
Sangster, J. (1989). Octanol-water partition coefficients of simple organic compounds.
Journal of Physical and Chemical Reference Data, 18(3), 1111–1229.
Shahidi, F., & Naczk, M. (1995). Food phenolics: Sources, chemistry, effects, applications.
Lancaster, Pa: Technomic Pub. Co.
Short, J., Roberts, J., Roberts, D., Hodges, G., Gutsell, S., & Ward, R. (2010). Practical
methods for the measurement of log P for surfactants. Ecotoxicology Environmental
Zhang, S., & Akoh, C. C. (2019). Solvent-free enzymatic synthesis of 1-o-galloylglycerol
optimized by the Taguchi method. Journal of the American Oil Chemists' Society, 96(8),
877–889.
9