molecules

Article
Polyalcohols as Hydrogen-Bonding Donors in
Choline Chloride-Based Deep Eutectic Solvents for
Extraction of Xanthones from the Pericarp of
Garcinia mangostana L.
Kamarza Mulia * , Farah Fauzia and Elsa Anisa Krisanti
Department of Chemical Engineering, Universitas Indonesia, Depok 16424, Indonesia;
farfazia@gmail.com (F.F.); elsakm@che.ui.ac.id (E.A.K.)
* Correspondence: kmulia@che.ui.ac.id; Tel.: +62-21-7863516

Academic Editor: Giancarlo Cravotto !"#!$%&'(!

!"#$%&'
Received: 9 December 2018; Accepted: 28 January 2019; Published: 12 February 2019

Abstract: Mangosteen (Garcinia mangostana L.) is a fruit that is rich in xanthones, utilized as health
supplements or additives in food products due to their high antioxidant activities. Choline chloride
(ChCl)-based deep eutectic solvents (DESs) with polyalcohols (ethylene glycol, glycerol, propanediols,
and butanediols) as hydrogen bonding donors (HBDs) were used to extract the xanthones from the
pericarp of mangosteen. DESs with 1,2-propanediol, 1,3-propanediol, and 1,2-butanediol as HBDs
(ChCl to HBD mole ratio of 1:3) afforded the highest extraction yields (2.40-2.63%) of ↵-mangostin, the
most abundant component that represents xanthones. These DESs have intermediate Nile Red polar
parameter values similar to that of ethanol and provide extraction yields with a quadratic dependence
on the polar parameter. Polarity and viscosity, the important physicochemical properties to consider
in the selection of DES as an extraction solvent, could be adjusted based on the consideration of the
molecular structure of the polyalcohols. The following factors qualifies the ChCl-1,2-propanediol
DES as a designer solvent for green extraction: It is selected from a set of DESs having a homologous
class of HBDs to deliver the highest ↵-mangostin extraction yield, its extract composition similar
to that obtained using ethanol, it has low or negligible vapor pressure, both of its components are
generally recognized as safe chemicals so that direct utilization of a DES extract is possible, and this
DES is used for utilization of agricultural waste products as the resource of bioactive compounds.

Keywords: deep eutectic solvent; solvent extraction; Garcinia mangostana L.; mangosteen; ↵-mangostin;
choline chloride; polyalcohol

1. Introduction
Mangosteen is a tropical fruit tree that grows in Southeast Asia. The fruit is rich in water,
protein, carbohydrates, fiber, essential nutrients and many kinds of vitamins [1]. Phytochemical
studies have shown that the mangosteen pericarp, which is a biomass waste, contains secondary
metabolites, e.g., oxygenated and prenylated xanthones [2]. Xanthones are bioactive compounds that
have antioxidant, antitumor, anti-inflammatory, antiallergy, antibacterial, antifungal and antiviral
activities [3–5]. Most of the xanthones are found in the pericarp or rind, in an amount sixteen
times higher than that in the aril segments [6]. The most abundant xanthone in the mangosteen
pericarp is ↵-mangostin, while other xanthones are present in less amounts ( -mangostin, -mangostin,
gartanin, and 8-deoxygartanin) [7]. Polyphenol substances, such as xanthones, extracted from plant or
agriculture waste product, have been utilized as antioxidant food supplements. Recently xanthones in
the form of mangosteen pericarp powder have been added to dark and compound chocolates [8].

Molecules 2019, 24, 636; doi:10.3390/molecules24030636 www.mdpi.com/journal/molecules

Molecules 2019, 24, 636 2 of 11

Extraction of bioactive compounds from natural resources usually involves the use of a large
amount of organic solvents. Most of these organic solvents possess a certain degree of toxicity and
volatility, generating waste that is harmful to both human health and the environment. Due to their
nonvolatility at ambient conditions compared with the volatile organic solvents, ionic liquids (ILs)
have received significant attention as alternative solvents [9,10]. ILs are used in a wide range of
applications [11,12], however, they are not commonly used in pharmaceutical and food industries due
to their toxicity and high synthesis cost of some ILs [13–15].
The search of green solvents as substitutes for hazardous organic solvents in order to minimize
environmental problems, improve safety and health, as well as reduce cost, has led to the use of deep
eutectic solvents (DESs). They are a mixture of a hydrogen bond acceptor (HBA) molecule that forms
intermolecular hydrogen bonds with one or more hydrogen bond donor (HBD) molecules, thereby
decreasing the melting point of the mixture to temperatures much lower than those of its individual
components. Abbott et al. were the first to report that DESs could be formed between choline chloride
(ChCl), using a quaternary ammonium salt as a HBA, and a range of amides, carboxylic acids and
alcohols as HBDs [16]. While DESs have physical properties similar to those of ILs [9,16,17], they are
considered as alternatives to ILs due to their low vapor-pressure, ease of synthesis, biocompatibility,
biodegradability, and low production cost [18,19]. Dai et al. [13] suggested the term Natural Deep
Eutectic Solvents (NADES) for DESs that are mixtures of various cellular constituents or primary
metabolites found in many kinds of organisms. NADES that are more stable as liquids at room
temperature, having nontoxic and environmentally friendly characteristics, are potentially useful as
solvents for extraction of bioactive compounds from plants [14].
In this study, extraction yields were measured in terms of ↵-mangostin, the most abundant
xanthone in the mangosteen pericarp. A high extraction yield of ↵-mangostin extracted from
the pericarp of mangosteen could be expected when the hydrogen bonding interactions in
the ChCl-↵-mangostin complex was similar in nature and strength to the interactions in the
ChCl-polyalcohol complex [20,21]. Quantum mechanical and molecular dynamics simulations of the
ChCl-urea, ChCl-ethylene glycol, and ChCl-glycerol DESs showed that the melting point depression
and viscosity data of these DESs could be correlated directly to the strength and nature of the hydrogen
bond network in the bulk liquid [22].
Previously, we reported that ChCl-based DESs with 1,2-propanediol (a polyalcohol) as the HBD
gave ↵-mangostin extraction yields in the range of 2.0–2.5%, much higher than the yields obtained
using DESs with citric acid, glycerol, and glucose as the HBDs (<0.5%) [23]. This finding deserves
further investigation of a homologous class of polyalcohols used as HBDs in ChCl-based DESs,
regarding their effectiveness as extracting solvents for xanthones. It is anticipated that this approach
would allow a systematic investigation of the effect of the molecular structure of the polyalcohols,
and the physicochemical properties of the DESs, on the extraction yields. It would be of interest
to see if a DES could provide a high extraction yield, and also, exemplify a designer solvent for a
green extraction.
In this study, we tested 17 DESs, each consisting of ChCl and a polyalcohol (ethylene glycol,
glycerol, 1,2-propanediol, 1,3-propanediol, 1,2-butanediol, 1,3-butanediol, 1,4-butanediol) as the HBD.
Our first objective was to investigate how extraction yields of ↵-mangostin was affected by the
composition of the DESs, the molecular structure of the HBDs, and the physicochemical properties of
the DESs. The second objective was to determine if ChCl-polyalcohol DESs was a feasible substitute
for ethanol, a common organic solvent used for the extraction of bioactive compounds from plants.

2. Results and Discussion

2.1. Density, Polarity and Viscosity of ChCl-Polyalcohol DESs

Table 1 gives the density, polarity and viscosity data of the ChCl-polyalcohol DESs. The density
data obtained was used to calculate extraction yields, while the polarity and viscosity data was
Molecules 2019, 24, 636 3 of 11

measured to see the effect of these physicochemical properties on the mangostin extraction yields.
DESs were prepared by mixing choline chloride as the HBA and a polyalcohol as the HBD, in three
mole ratios (1:1, 1:2, and 1:3), until a homogenous solution was obtained. Mixtures that were unstable
or took too long to form were not included in this study.

Table 1. Compositions, physical properties, and extraction yields of choline chloride-based polyalcohol
DESs.

↵-Mangostin
ChCl:HBD Density ENR Viscosity
Polyalcohol (HBD) DES Extraction
Mole Ratio (g/mL) (kcal/mol) * (cP)
Yield (%) **
1:1 EG1 1.08 47.4 31.2 0.10 ± 0.01
Ethylene glycol 1:2 EG2 1.05 47.4 30.4 0.12 ± 0.01
1:3 EG3 1.04 47.5 29.7 0.23 ± 0.01
1:1 2P1 1.05 56.5 37.1 1.97 ± 0.11
1,2-Propanediol 1:2 2P2 1.03 56.7 35.2 2.31 ± 0.16
1:3 2P3 1.01 56.6 31.6 2.63 ± 0.11
1:1 3P1 1.04 56.1 40.1 0.41 ± 0.05
1,3-Propanediol 1:2 3P2 1.03 56.7 39.2 1.60 ± 0.13
1:3 3P3 1.10 56.8 38.1 2.46 ± 0.19
1:2 2B2 1.06 50.6 27.6 1.79 ± 0.12
1,2-Butanediol
1:3 2B3 1.06 50.9 25.8 2.40 ± 0.09
1:2 3B2 1.05 50.6 67.0 0.84 ± 0.17
1,3-Butanediol
1:3 3B3 1.04 50.1 62.2 1.15 ± 0.13
1:1 4B1 1.10 57.8 44.8 0.73 ± 0.17
1,4-Butanediol 1:2 4B2 1.05 58.8 43.9 0.84 ± 0.17
1:3 4B3 1.05 60.6 43.7 1.00 ± 0.13
Glycerol 1:1 G1 1.14 55.6 46.8 0.12 ± 0.01
Ethanol - EtOH - 51.6 - 3.27 ± 0.14
* Calculated using Equation (1); ** calculated using Equation (2) and reported as mean ± SD (n = 3).

The polarities of the DESs were determined in terms of their Nile Red polar parameter (ENR )
values, calculated using Equation (1), where a lower ENR value indicated a more polar solvent [24].
Accordingly, the polarity of the DESs in descending order was as follows: DES-EG (ethylene glycol) >
DES-2B (1,2-butanediol) ~ DES-3B (1,3-butanediol) > DES-G1 (glycerol) > DES-2P (1,2-propanediol) ~
DES-3P (1,3-propanediol) > DES-4B (1,4-butanediol).
The viscosity of all DESs decreased with increasing polyalcohol to ChCl mole ratio, due to the
increased solvation effect as more hydroxyl groups were available around the chloride anion of the
ChCl molecule. Reduced viscosity was also observed for DESs with polyalcohols having shorter
methylene chains between the two hydroxyl groups, in the order of DES-EG (ethylene glycol) <
DES-3P (1,3-propanediol) < DES-4B (1,4-butanediol). The position of the OH groups in a polyalcohol
also affected the viscosity of a DES as DES-2B (1,2-butanediol) and DES-3B (1,3-butanediol) had the
lowest and highest viscosities, respectively.

2.2. Optimum Extraction Time

Plots of the extraction yield data as a function of extraction time, obtained using ethanol and
DES-2P3 as the extracting solvents, are shown in Figure 1. The much lower initial extraction rate
of the DES, compared to that of ethanol, could be explained using the Stokes-Einstein equation.
The diffusivity of DES-2P3 was lower than that of ethanol due to its higher viscosity and the larger
size of the ChCl-1,2-propanediol complex, making it more difficult for this solvent to infiltrate the
pericarp matrix. The two plots levelled off after 4 h, an indication that the equilibrium condition had
been attained. Accordingly, an extraction time of 4 h was used in the subsequent experiments.
 Molecules 2018, 23, x FOR PEER REVIEW 4 of 12
Molecules 23,636
2018,24,
Molecules2019, x FOR PEER REVIEW 4 of
of11
12

Figure1.1.Plots
Figure Plotsofof↵-mangostin
α-mangostin yield
yield as
as aa function
function of
of extraction
extraction time.
time. Data
Data obtained
obtained using
using choline
choline
chloride-1,2-propanediol
chloride-1,2-propanediol deep
deep eutectic
eutectic solvent
solvent (DES)
(DES) (mole
(mole ratio
ratio of
of 1:3)
1:3) and
and ethanol
ethanol as
as extracting
extracting
Figure 1. Plots of α-mangostin yield as a function of extraction time. Data obtained using choline
solvents.
solvents.
chloride-1,2-propanediol deep eutectic solvent (DES) (mole ratio of 1:3) and ethanol as extracting
solvents.
2.3. α-MangostinExtraction
2.3.a-Mangostin ExtractionYields
Yields
2.3. ↵-mangostin
α-Mangostin extraction
α-mangostinExtraction yield
yield data
Yields
extraction data obtained ChCl-based DESs,
obtained using 17 ChCl-based DESs, defined
defined as
as the
the mass
mass
percentage
percentageofof α-mangostin
↵-mangostin extracted
extractedfrom the
from dried
the mangosteen
dried powder,
mangosteen is
powder, given in
is given
α-mangostin extraction yield data obtained using 17 ChCl-based DESs, defined as the massTable 1 and
in Table shown
1 and
in Figure
shown in2.Figure
percentage The effect
2. Theof the extracted
effect
of α-mangostin choline chloride
of the choline to dried
polyalcohol
chloride
from the molemole
to polyalcohol
mangosteen ratio, molecular
ratio,
powder, is givenstructure
in Tableof1 the
molecular structure of
and
polyalcohols,
the
shown in Figureand 2.
polyalcohols, polarity
and andof
polarity
The effect viscosity
and of the
viscosity
the choline
of DESs are
the
chloride
DESs discussed inmole
are discussed
to polyalcohol turn.ratio, molecular structure of
in turn.

the polyalcohols, and polarity and viscosity of the DESs are discussed in turn.

d
Figure 2. Extraction yield obtained using ChCl-polyalcohol DESs. The error bar indicates the maximum
Figure 2. Extraction yield obtained using ChCl-polyalcohol DESs. The error bar indicates the
and minimum
maximum and yields (n = 3).yields
minimum Label:(n
EG= (ethylene
3). Label:glycol), 2P (1,2-propanediol),
EG (ethylene 3P (1,3-propanediol),
glycol), 2P (1,2-propanediol), 3P
2B (1,2-butanediol),
(1,3-propanediol), 3B (1,3-butanediol),
2B (1,2-butanediol), 4B (1,4-butanediol),
3B (1,3-butanediol), G (glycerol), EtOH
4B (1,4-butanediol), (ethanol),
G lastEtOH
(glycerol), digit
Figure 2. Extraction yield obtained using ChCl-polyalcohol DESs. The error bar indicates the
(mole ratio of
(ethanol), lastpolyalcohol
digit to
(mole ratioChCl).
of polyalcohol to ChCl).
maximum and minimum yields (n = 3). Label: EG (ethylene glycol), 2P (1,2-propanediol), 3P
(1,3-propanediol), 2B (1,2-butanediol),
2.3.1. Effect of ChCl to Polyalcohol Mole Ratio 3B (1,3-butanediol), 4B (1,4-butanediol), G (glycerol), EtOH
(ethanol), last digit (mole ratio of polyalcohol to ChCl).
The effect of ChCl to polyalcohol mole ratio on extraction yield was exhibited most clearly by
DES-3P1 and DES-3P3 where sixfold improvement in yield (0.41 to 2.46%) was obtained by increasing
ChCl to 1,3-propanediol mole ratio from 1:1 to 1:3, respectively. This effect was due to the presence of
Molecules 2018, 23, x FOR PEER REVIEW 5 of 12

2.3.1. Effect of ChCl to Polyalcohol Mole Ratio

The effect of ChCl to polyalcohol mole ratio on extraction yield was exhibited most clearly by
Molecules 2019, 24, 636 5 of 11
DES-3P1 and DES-3P3 where sixfold improvement in yield (0.41 to 2.46%) was obtained by
increasing ChCl to 1,3-propanediol mole ratio from 1:1 to 1:3, respectively. This effect was due to the
presence
more of more
hydroxyl hydroxyl
groups of thegroups of the surrounding
polyalcohol, polyalcohol, the
surrounding
negativelythe negatively
charged charged
chloride anion chloride
of ChCl,
decreasing the polarity of the DESs, and inducing more dissolution of the non-polar the
anion of ChCl, decreasing the polarity of the DESs, and inducing more dissolution of non-polar
↵-mangostin
α-mangostin
molecules. Themolecules. Thecholine
slightly basic slightly basic in
chloride choline chloride in
DESs increased theDESs increased
extraction the due
efficiency extraction
to the
efficiency due to the weak acidity of the phenolic OH
weak acidity of the phenolic OH groups of ↵-mangostin [25]. groups of α-mangostin [25].

2.3.2. Effect of Molecular Structure of Polyalcohols

Choline chloride was the only hydrogen-bonding acceptor acceptor used in this study, therefore, the
extraction yield data could be analyzed systematically based on the molecular structure of the
polyalcohols (Figure 3) used as as the
the hydrogen-bonding
hydrogen-bonding donor.
donor. The highest extraction yields were
obtained using DES-2P3 (2.63%) and DES-2B3 (2.40%), having HBDs with adjacent OH groups in a
terminal position attached to a methyl group (1,2-propanediol) and an ethyl group (1,2-butanediol),
(1,2-butanediol),
respectively. The alkyl chain allowed the more polar OH groups of ↵-mangostin to draw
respectively. The alkyl chain allowed the more polar OH groups of α-mangostin to closer
draw to the
closer to
chloride
the chlorideanion of ChCl
anion due todue
of ChCl hydrogen bonding.
to hydrogen This meant
bonding. This that
meantthethat
alkane
the chain would
alkane chainbewould
directed
be
away
directedfrom the ChCl,
away increasing
from the the void volume
ChCl, increasing of the
the void DESs of
volume necessary
the DESs fornecessary
extractionfor
of ↵-mangostin.
extraction of
This explanation
α-mangostin. Thiswas consistentwas
explanation with the very with
consistent low extraction yield
the very low of 0.23%yield
extraction observed for DES-EG3
of 0.23% observed
(ethylene
for DES-EG3 glycol). Without
(ethylene any methylene
glycol). group,
Without any the moregroup,
methylene polar ethylene
the more glycol
polarseemed to form
ethylene a
glycol
more dense complex with choline chloride [21], making DES-EG3 not conducive to the solubilization
seemed to form a more dense complex with choline chloride [21], making DES-EG3 not conducive to
of
thea solubilization
large and slightly
of a non-polar molecule
large and slightly such as ↵-mangostin.
non-polar molecule such as α-mangostin.

Figure 3. Chemical structure of ↵-mangostin, choline chloride, and polyalcohols used in this study.
Figure 3. Chemical structure of α-mangostin, choline chloride, and polyalcohols used in this study.
The length of the straight carbon chain between two hydroxyl end groups in HBDs tested was (in
The length
decreasing of1,4-butanediol
order): the straight carbon chain>between
(DES-4B3) two hydroxyl
1,3-propanediol end >groups
(DES-3P3) in glycol
ethylene HBDs (DES-EG3).
tested was
(in decreasing order): 1,4-butanediol (DES-4B3) > 1,3-propanediol (DES-3P3)
DES-3P3, having 1,3-propanediol with one methylene group between the two hydroxyl end groups,> ethylene glycol
(DES-EG3). DES-3P3, having 1,3-propanediol with one methylene group between the
was able to extract a higher amount of ↵-mangostin (2.46%) than DES-4B3 having 1,4-butanediol withtwo hydroxyl
end methylene
two groups, was able (1.00%)
groups to extract a higher having
or DES-EG3 amountethylene
of α-mangostin (2.46%)
glycol with than DES-4B3
no methylene having
group (0.23%).
1,4-butanediol with two methylene groups (1.00%) or DES-EG3 having ethylene glycol
This observation suggested that the molecular structure of 1,3-propanediol is spatially preferable with no
methylene group (0.23%). This observation suggested that the molecular
to form a hydrogen-bonded complex with ChCl, with interactions similar to those between choline structure of
1,3-propanediol
chloride is spatiallyA similar
and ↵-mangostin. preferable
studyto regarding
form a hydrogen-bonded complex
the molecular structure with ChCl, with
of 1,4-butanediol and
interactions similar to those between choline chloride and α-mangostin. A similar
possibility of forming hydrogen bonding with choline chloride has been reported earlier study regarding
[21].
the molecular structure
The position of theofOH1,4-butanediol
groups alongandthepossibility
alkane of forming
chain hydrogen
of the bonding
butanediol withalso
isomers choline
has
chloride has been reported earlier [21].
a significant effect on the extraction yields. The yields for DES-2B3 (1,2-butanediol), DES-3B3
(1,3-butanediol), and DES-4B3 (1,4-butanediol) were 2.40%, 1.15%, and 1.00%, respectively.
As discussed previously, the alkane chain in 1,2-butanediol was the most flexible, allowing the adjacent
terminal OH groups to be attracted to the charged part of choline chloride more easily.
 The position of the OH groups along the alkane chain of the butanediol isomers also has a
significant effect on the extraction yields. The yields for DES-2B3 (1,2-butanediol), DES-3B3
(1,3-butanediol), and DES-4B3 (1,4-butanediol) were 2.40%, 1.15%, and 1.00%, respectively. As
discussed previously, the alkane chain in 1,2-butanediol was the most flexible, allowing the adjacent
terminal OH groups to be attracted to the charged part of choline chloride more easily.
Molecules 2019, 24, 636 6 of 11

2.3.3. Effect of Polarity of DESs

2.3.3. To
Effect of Polarity
study of of
the effect DESspolarity of a DES on extraction of α-mangostin, extraction yield data of
DESsTowith
study the effect of polarity
a ChCl to polyalcohol of aratio
mole DESofon 1:3extraction
was plottedof ↵-mangostin, extraction
against the carbon yield data
to hydroxyl of
(C/OH)
DESs with
ratio of thea HBDs.
ChCl toThepolyalcohol
extractionmole
yieldratio
dataofof1:3 waswith
DESs plotted
ChClagainst the carbonmole
to polyalcohol to hydroxyl
ratios of(C/OH)
1:1 and
ratio of the
1:2 were notHBDs. Theinextraction
included the plot asyield
theydata
did of
not DESs with
reveal ChCl to polyalcohol
a systematic moleinratios
pattern as was of 1:1
the case of
and
DESs 1:2with
werethenot
1:3included in the
ratio. Figure 4aplot
showsas they didhighest
that the not reveal a systematic
extraction yields pattern as was using
were obtained in the DESs
case
of
of DESs
an HBD with thea1:3
with ratio.
C/OH Figure
ratio 4a while
of 1.5, showssignificantly
that the highest
lowerextraction yields
yields were were using
obtained obtained
DESsusing
that
DESs of an HBD with a C/OH ratio of 1.5, while significantly lower yields were obtained using DESs
were not sufficiently polar or too polar, having HBDs with C/OH ratios of 2 and 1, respectively. An
that were not sufficiently
approximately quadratic polar or too
plot was polar, having
obtained, excluding HBDs
thewith
data C/OH ratios of 2 and 1, respectively.
of DES-2B3.
An approximately quadratic plot was obtained, excluding the data of DES-2B3.

Figure 4. Extraction yield as a function of: (a) the C/OH ratio of the polyalcohols, and, (b) polar
Figure 4. Extraction yield as a function of: (a) the C/OH ratio of the polyalcohols, and, (b) polar
parameter ENR of the DESs. (ChCl:polyalcohol mole ratio of 1:3, legend signifies the HBD of the DES).
parameter ENR of the DESs. (ChCl:polyalcohol mole ratio of 1:3, legend signifies the HBD of the DES).
To obtain a better extraction yield dependence on polarity, the Nile Red polar parameter (ENR ) was
To obtain a better extraction yield dependence on polarity, the Nile Red polar parameter (ENR)
used to represent the polarity of the DESs. The resulting plot (Figure 4b) shows that the ↵-mangostin
was used to represent the polarity of the DESs. The resulting plot (Figure 4b) shows that the
yield had an approximately quadratic dependence on the ENR value of the DESs (R2 of 0.91, excluding
α-mangostin yield had an approximately quadratic dependence on the ENR value of the DESs (R2 of
data of ethanol), and DESs with high extraction yields had intermediate ENR values (50.6–56.8 kcal/mol)
0.91, excluding data of ethanol), and DESs with high extraction yields had intermediate ENR values
that bounded the ENR value of ethanol (51.6 kcal/mol). The plot could be useful in the absence of prior
(50.6–56.8 kcal/mol) that bounded the ENR value of ethanol (51.6 kcal/mol). The plot could be useful
experimental data, for example, DES-2B3 (50.9 kcal/mol) and not DES-4B3 (60.6 kcal/mol), could be
in the absence of prior experimental data, for example, DES-2B3 (50.9 kcal/mol) and not DES-4B3
selected as an extraction solvent based on the similarity to the ENR value of ethanol (51.6 kcal/mol).
(60.6 kcal/mol), could be selected as an extraction solvent based on the similarity to the ENR value of
The highest extraction yield of 3.3% was obtained using ethanol, slightly higher than 2.63% yield
ethanol (51.6 kcal/mol).
obtained using DES-2P3. The difference is almost statistically insignificant (tcalc = 4.34, ttable = 4.30,
The highest extraction yield of 3.3% was obtained using ethanol, slightly higher than 2.63%
↵ = 0.05, n = 3). The data obtained in this study showed that three ChCl-polyalcohol DESs were
yield obtained using DES-2P3. The difference is almost statistically insignificant (tcalc = 4.34, ttable =
effective extraction solvents and they could be used as substitutes for ethanol to extract bioactive
4.30, α = 0.05, n = 3). The data obtained in this study showed that three ChCl-polyalcohol DESs were
compounds from plants. Very low extraction yields were obtained using DESs with relatively more
effective extraction solvents and they could be used as substitutes for ethanol to extract bioactive
polar HBDs, ethylene glycol (DES-EG3) and glycerol (DES-G1), both having a C/OH ratio of 1. These
compounds from plants. Very low extraction yields were obtained using DESs with relatively more
results suggest that DES polarity was an important factor to consider in order to obtain high extraction
polar HBDs, ethylene glycol (DES-EG3) and glycerol (DES-G1), both having a C/OH ratio of 1. These
yields of bioactive compounds such as ↵-mangostin.
results suggest that DES polarity was an important factor to consider in order to obtain high
extraction
2.3.4. Effectyields of bioactive
of Viscosity compounds such as α-mangostin.
of DESs

2.3.4.For a certain
Effect DES, theofeffect
of Viscosity DESsof viscosity predominated over the effect of solvent polarity. As shown
in Figure 4a,b, the extraction yield of DES-2B3 was much higher (2.40%) than that of DES-3B3 (1.15%)
For a certain
and DES-4B3 DES,
(1.00%), inthe effect
spite of fact
of the viscosity predominated
that both overand
1,2-butanediol the 1,3-butanediol
effect of solvent polarity.
had the sameAs
shown in Figures 4a and 4b, the extraction yield of DES-2B3 was much higher (2.40%)
C/OH ratio of 2 and very similar ENR values. In the case of DES-2B3, both the viscosity and the than that of
DES-3B3 (1.15%) and DES-4B3 (1.00%), in spite of the fact that both 1,2-butanediol
polarity exerted a synergistic effect on extraction yield. On the other hand, the much lower yield of and
DES-3B3 seemed to be due to its high viscosity value, in contrast to DES-2B3, which had the lowest
viscosity among all of the DESs tested.
 Molecules 2018, 23, x FOR PEER REVIEW 7 of 12

1,3-butanediol had the same C/OH ratio of 2 and very similar ENR values. In the case of DES-2B3,
both the viscosity and the polarity exerted a synergistic effect on extraction yield. On the other hand,
Molecules 2019, 24,
the much 636 yield of DES-3B3 seemed to be due to its high viscosity value, in contrast7 to
lower of 11
DES-2B3, which had the lowest viscosity among all of the DESs tested.

2.4.2.4.
Composition of of
Composition Xanthones
XanthonesininEthanolic
Ethanolicand
andDES
DES Extracts
Extracts
The xanthones
The xanthones recovered
recoveredfrom
fromthe
theDES
DESthat
that gave
gave the
the highest extractionyield
highest extraction yield(DES-2P3)
(DES-2P3) were
were
identified using an LC-MS analyzer. The relative compositions of xanthones in DES and
identified using an LC-MS analyzer. The relative compositions of xanthones in DES and in ethanol, in ethanol,
were determined
were determinedbased
basedononthethe
area percentage
area percentageofofeach
eachspecific
specificpeak
peakcompared
comparedwith
withthe
theoverall
overall area
area of
theofmajor peakspeaks
the major (Figure 5). 5).
(Figure

Figure 5. Chromatogram
Figure 5. ChromatogramofofthetheDES-2P3
DES-2P3(above)
(above)andand ethanolic
ethanolic extracts (below).Numbers
extracts (below). Numbersindicate
indicate
thethe
xanthone components
xanthone componentsidentified
identifiedusing
usingLC-MS:
LC-MS: (1)
(1) 1,7-dihydroxy-3-methoxy-2-(3-methylbut-2-
1,7-dihydroxy-3-methoxy-2-(3-methylbut-2-
enyl)xanthon,
enyl)xanthon,(2)(2)-mangostin,
γ-mangostin,(3)
(3)gartanin,
gartanin, (4)
(4) ↵-mangostin, (5)garcinone
α-mangostin, (5) garcinoneE,E,(6) garcimangosone
(6)garcimangosone B, B,
(7) -mangostin.
(7) β-mangostin.

Table
Table2 provides
2 providesthe thecomposition
compositionofof the
the DES
DES extract that was
extract that was dominated
dominatedby byα-mangostin
↵-mangostin
(↵-mangostin, -mangostin, and a small percentage of -mangostin) followed by gartanin,
(α-mangostin, γ-mangostin, and a small percentage of β-mangostin) followed by gartanin, garcinone
garcinone
E, and garcimangosone
E, and garcimangosone B, present in less
B, present amounts.
in less amounts.Even
Eventhough oneone
though of the compounds
of the compounds present in DES
present in
DES extract, 1,7-dihydroxy-3-methoxy-2-(3-methylbut-2-enyl)xanthon, was not
extract, 1,7-dihydroxy-3-methoxy-2-(3-methylbut-2-enyl)xanthon, was not detected in the ethanolic detected in the
ethanolic
extract, extract,
it was it was
detected asdetected
a minoras a minor compound
compound in another in similar
another study
similar[6].
study
The[6]. The of
result result
the of the
LC-MS
LC-MS analysis showed that DES and ethanol had similar capabilities in
analysis showed that DES and ethanol had similar capabilities in extracting xanthones.extracting xanthones.

Table 2. Composition of xanthones in DES-2P3 and ethanol based on the LC-MS data.
Table 2. Composition of xanthones in DES-2P3 and ethanol based on the LC-MS data.
[M + H]+ Composition * (%)
Compound Xanthone Formula Composition * (%)
Compound Xanthone Formula [M + H]+ m/z
m/z Ethanol DES
1,7-dihydroxy-3-methoxy-2-(3-me Ethanol DES
1 C19H18O5 327 - 5.8
1,7-dihydroxy-3-methoxy-2-(3-methylbut-2-
thylbut-2-enyl)xanthon
1 C19 H18 O5 327 - 5.8
2 enyl)xanthon
γ-mangostin C23H24O6 397 13.0 12.8
2 3 -mangostin
gartanin C23CH HO246 O6
2324 397 397 16.0 13.0 12.312.8
3 gartanin C23 H24 O6 397 16.0 12.3
4 α-mangostin C24H 26O6 411 53.6 52.4
4 ↵-mangostin C24 H26 O6 411 53.6 52.4
5 5 garcinoneEE
garcinone C28CH32HO6 O
28 32 6
465 465 9.5 9.5 10.410.4
6 6 garcimangosoneB B
garcimangosone C24CH HO246 O6
2424 409 409 4.7 4.7 4.2 4.2
7 7 -mangostin
β-mangostin C25CH HO286 O6
2528 425 425 3.3 3.3 2.1 2.1
* Compositionestimates
* Composition estimates based
based on
onthe
thenormalized
normalized area of of
area HPLC
HPLCpeaks.
peaks.

The similarity of both the ↵ -mangostin extraction yields and the composition of the extracted
xanthones indicated that DES-2P3 could be used to extract bioactive compounds, as a substitute for
ethanol, the commonly used organic solvent. Around 39.9% of the ↵ -mangostin in DES-2P3 was
recovered in a single step using ethyl acetate and diethyl ether as the back-extracting solvents, therefore,
multiple back-extraction was required to improve the recovery [18].
Molecules 2019, 24, 636 8 of 11

2.5. ChCl-1,2-Propanediol DES as a Designer Solvent for Green Extraction

DESs are known as designer solvents that allow for a systematic search for the optimum HBA-HBD
combination and mole ratio, suitable for the extraction of the targeted bioactive compounds. In this
study, a set of DESs having a homologous class of polyalcohols was screened for the best solvent for the
extraction of ↵-mangostin. It was found that ChCl-based DESs having 1,2-propanediol, 1,3-propanediol
and 1,2-butanediol as HBDs (in a mole ratio of 1:3) afforded high extraction yields of xanthones, in
terms of ↵-mangostin. The ChCl-1,2-propanediol was the DES with the highest extraction yield of
↵-mangostin among the seventeen polyalcohol-DESs tested, presumably, by forming choline chloride
and a 1,2-propanediol complex with sufficient space, polarity, and viscosity suitable for effective
interaction with ↵-mangostin. This DES exemplified the advantages of DESs as designer solvents
by combining high extraction yield, low or negligible vapor pressures, and good biodegradability
aspects [26,27].
There are two additional advantages of using the ChCl-1,2-propanediol DES. Firstly, choline
chloride and 1,2-propanediol are categorized as generally recognized as safe (GRAS) chemicals [28].
The use of biocompatible HBAs and HBDs opens up the possibility of direct utilization of DES extracts
as part of the bioactive compound formulation, eliminating the back-extraction and solvent recovery
steps. Secondly, the xanthones were extracted from the pericarp of mangosteen, an agricultural
waste product. These advantages are in line with the principles of green extraction proposed by
Chemat et al. [29]. When taken together, the results of this study qualifies the ChCl-1,2-propanediol
DES as a designer solvent for green extraction.

3. Materials and Methods

3.1. Chemicals and Plant Material

Choline chloride (>98%), 1,2-propanediol (99%), 1,3-propanediol (99%), 1,2-butanediol (99%),
1,3-butanediol (99%), 1,4-butanediol (99%), ethylene glycol (99%), glycerol (99%), and analytical grade
Nile Red were purchased from Sigma Aldrich (Singapore). HPLC grade acetonitrile and ethanol were
purchased from Smart Lab (Jakarta, Indonesia). The ↵-mangostin standard (>99.8%) was purchased
from Aktin Chemical Inc., (Chengdu, China). Mangosteen fruit was purchased in Depok (West Java,
Indonesia) and identified as G. mangostana by the Bogoriense Herbarium, Research Center for Biology,
Indonesian Institute of Sciences, specimen voucher number 248/IPH.1.02/If.8/II/2014.

3.2. Preparation of Mangosteen Powder and ChCl-Polyalcohol DESs

Mangosteen pericarp was cleaned from its edible parts, the tough outer skin was peeled, inner
part cut and dried at room temperature, and grind using an electric grinder (IKA, Staufen, Germany).
The powder obtained was further dried in an oven at 65 C until it reached constant weight, screened
using a 20 mesh sieve, and stored in a sealed container to avoid contact with air and exposure to direct
sunlight. DESs were prepared by mixing ChCl and polyalcohols in three mole ratios (1:1, 1:2, 1:3)
under a constant stirring speed at a temperature of 50 and 80 C, for liquid and solid polyalcohols,
respectively. Stirring was continued for 30–90 min until a clear solution was formed [13]. Addition of
water to DESs has been shown to be beneficial in reducing viscosity and increasing extraction yields [9],
however, no water was added into the DESs used in this study to simplify the subsequent solvent
recovery process as part of the overall extraction process.

3.3. Physicochemical Properties of DESs

The physicochemical properties determined in this study were density, polarity, and viscosity.
Density measurement was carried out by weighing a known volume of DES dispensed from a
micropipette previously calibrated using purified water. The Nile Red dye was added to each DES
as a solvatochromic probe to determine max , the wavelength at which the maximum visible light
absorption occurred [30]. A UV-vis spectrophotometer (Model UV-1900, Shimadzu, Kyoto, Japan) was
Molecules 2019, 24, 636 9 of 11

used to scan the DES-dye mixtures in the 400–700 nm range and Equation (1) was used to calculate
the Nile Red polar parameter ENR (in kcal/mol). Viscosity test was performed using a Brookfield
viscometer at room temperature with LV spindle no. 1 at 6, 12, and 30 rpm.

ENR (kcal/mol) = 28,591/ abs,max (nm) (1)

3.4. Extraction of a-Mangostin Using Polyalcohol DESs and Ethanol

The extraction of ↵-mangostin was carried out by mixing 0.2 g of mangosteen powder and 2 g
of DES or ethanol (solid to liquid mass ratio of 1:10) in a sealed reaction tube. The tube was shaken
using a thermomixer apparatus (Model SCIENTZ-100, SCIENTZ, Ningbo, China) at a frequency of 500
oscillations/min at room temperature for 4 h. The suspension was then centrifuged for 15 min at 2000
rpm and the residue was separated using a 0.45 µm membrane filter to obtain the DES and ethanolic
extracts. Each set of extraction was done in triplicate. The extraction time required for the process to
reach equilibrium was determined based on the extraction yield data of a selected DES plotted against
extraction time up to 6 h.

3.5. HPLC Analysis of a-Mangostin Extracted into DESs

The amounts of ↵-mangostin extracted into each DES formulation were determined using a high
performance liquid chromatography (HPLC) analyzer (Model Prominence SPD-20A, Shimadzu, Kyoto,
Japan), operated under conditions similar to those reported by [31]. The HPLC was equipped with an
LC-20AD pump, a UV-Vis detector, a Rheodyne injector with 20 µL loops, and a C18 column with a
length of 250 mm and a diameter of 4.6 mm. An isocratic elution mode was employed with a mobile
phase flow rate of 1 mL/min, consisting of 0.1% (v/v) ortho-phosphoric acid solution and acetonitrile
in equal volumetric flow rate. Sample injection volume was 0.1 mL and UV-Vis detector wavelength
was set at 244 and 320 nm. Prior to injection, each sample was diluted with ethanol and filtered through
a 0.45 µm membrane (Sartorius NY). The extraction yields were calculated as the percent ratio of the
mass of the extracted ↵-mangostin to the mass of the dried mangosteen pericarp powder:

Mangostin yield (%) = 100 ⇥ mass of extracted ↵-mangostin/mass of dried mangosteen powder (2)

3.6. Recovery of Xanthones from DESs

now Niklas
DES with the highest ↵-mangostin extraction yield was chosen for the determination of
↵-mangostin recovery using the following procedure. First, 0.5 mL ethyl acetate was added to 1 mL
mangostin-DES extract, stirred for 1 h, and decanted. An aliquot (0.35 mL) of the upper ethyl acetate
fraction was separated for HPLC analysis. Then, 0.5 mL diethyl ether was added into the remaining
DES mixture, stirred for another 1 h, and decanted. An aliquot (0.4 mL) of the upper fraction was
separated for HPLC analysis. The total amount of ↵-mangostin recovered from the DES was the sum
of ↵-mangostin extracted into the ethyl acetate and diethyl ether fractions.

3.7. LC-MS Analysis of a-Mangostin Extracted Using DES and Ethanol

The compositions of xanthones in the ethanol and the representative DES extract were determined
using an LC-MS analyzer (Model ACQUITY UPLC® H-Class System, Waters, Milford, MA, USA)
equipped with a 1.7 µm pore size, 2.1 mm dia. ⇥ 50 mm BEH C18 column (Waters, Milford, MA, USA)
and a photodiode array detector. The DES was mixed with a mixture of ethyl acetate and diethyl
ether (volume ratio of 1:1), the mixture stirred and decanted, and the organic phase dried using a
vacuum dryer. The dried mangostin extract was diluted using ethanol, filtered through a 0.2 µm
membrane filter, and subjected to qualitative and quantitative analysis using the LC-MS analyzer. The
sample was eluted employing a linear gradient mode over a 15 min period followed by an isocratic
elution over a 2 min period, with a flow rate of 0.4 mL/min and the eluent composition of 95:5 to 5:95
(%-v/v) aqueous solution of formic acid (0.005%-v/v) and acetonitrile. The MS analysis was performed
Molecules 2019, 24, 636 10 of 11

on a Xevo® G2-XS QTof (Waters, Milford, MA, USA), employed in the positive ion mode and the
electrospray ionization. The source temperature and the desolvation temperature were maintained at
120 and 500 C, respectively. The desolvation gas flow was 844 L/h at a collision energy of 10 eV, and
the ramp collision energy was 15–50 V. For comparative purposes, the ethanolic extract was analyzed
without the pretreatment step.

4. Conclusions
The ↵-mangostin extraction yields were highly dependent on the polyalcohol and choline chloride
to polyalcohol mole ratio of the DES used as the extracting solvent. The ChCl-based DESs that had
1,2-propanediol, 1,3-propanediol, and 1,2-butanediol as HBDs afforded high ↵-mangostin extraction
yields, presumably by providing sufficient space and suitable polarity for an effective interaction
between ↵-mangostin and the ChCl-polyalcohol complex. These DESs had intermediate Nile Red
polar parameter values similar to that of ethanol, suggesting that polarity was an important factor
to consider in the selection of a DES in order to obtain high extraction yields. In turn, polarity and
viscosity, the most important physicochemical properties to consider in the selection of DES as an
extraction solvent, could be adjusted based on the consideration of the molecular structure of the
polyalcohols. The results of this study qualifies the ChCl-1,2-propanediol DES as a designer solvent
for green extraction.

Author Contributions: K.M., F.F. and E.A.K.; Conceptualization, K.M. and E.A.K.; Formal analysis, K.M., F.F.
and E.A.K.; Data curation, F.F.; Funding acquisition, K.M.; Investigation, F.F.; Methodology, K.M. and E.A.K.;
Project administration, E.A.K.; Supervision, K.M. and E.A.K.; Writing—Original Draft, E.A.K.; Writing—Review &
Editing, K.M.
Funding: This work was funded by the Indonesian Ministry of Research, Technology and Higher Education
through the Hibah Kompetensi Research Scheme contract number 527/UN2.R3.1/hkp05.00/2018.
Acknowledgments: The authors are grateful to FelitaL
LC-MS analyzer
Terahadi and Sylvania Putri for their support in the
experimental part of the investigation.
C-
Conflicts of Interest: The authors declare no conflict of interest.
M
References S
an
1. Kanchanapom, K.; Kanchanapom, M. Tropical and Subtropical Fruits; Ag Science Inc.: Auburndale, FL, USA, 1998.
2.
al
Matsumoto, K.; Akao, Y.; Ohguchi, K.; Ito, T.; Tanaka, T.; Iinuma, M.; Nozawa, Y. Xanthones induce cell-cycle
arrest and apoptosis in human colon cancer DLD-1yz cells. Bioorg. Med. Chem. 2005, 13, 6064–6069. [CrossRef]
[PubMed] er
3. Akao, Y.; Nakagawa, Y.; Nozawa, Y. Anti-Cancer Effects of Xanthones from Pericarps of Mangosteen. Int. J.
Mol. Sci. 2008, 9, 355–370. [CrossRef] [PubMed]
4. Yu, L.; Zhao, M.; Yang, B.; Zhao, Q.; Jiang, Y. Phenolics from hull of Garcinia mangostana fruit and their
antioxidant activities. Food Chem. 2007, 104, 176–181. [CrossRef]
5. Kondo, M.; Zhang, L.; Ji, H.; Kou, Y.; Ou, B. Bioavailability and antioxidant effects of a xanthone-rich Mangosteen
(Garcinia mangostana) product in humans. J. Agric. Food Chem. 2009, 57, 8788–8792. [CrossRef] [PubMed]
6. Wittenauer, J.; Falk, S.; Schweiggert-Weisz, U.; Carle, R. Characterisation and quantification of xanthones
from the aril and pericarp of mangosteens (Garcinia mangostana L.) and a mangosteen containing functional
beverage by HPLC–DAD–MSn. Food Chem. 2012, 134, 445–452. [CrossRef]
7. Pedraza-Chaverri, J.; Cárdenas-Rodríguez, N.; Orozco-Ibarra, M.; Pérez-Rojas, J.M. Medicinal properties of
mangosteen (Garcinia mangostana). Food Chem. Toxicol. 2008, 46, 3227–3239. [CrossRef] [PubMed]
8. Sim, S.Y.; Ng, J.W.; Ng, W.K.; Forde, C.G.; Henry, C.J. Plant polyphenols to enhance the nutritional and
sensory properties of chocolates. Food Chem. 2016, 200, 46–54. [CrossRef] [PubMed]
9. Dai, Y.; Witkamp, G.-J.; Verpoorte, R.; Choi, Y.H. Natural deep eutectic solvents as a new extraction media
for phenolic metabolites in Carthamus tinctorius L. Anal. Chem. 2013, 85, 6272–6278. [CrossRef]
10. Azmir, J.; Zaidul, I.; Rahman, M.; Sharif, K.; Mohamed, A.; Sahena, F.; Jahurul, M.; Ghafoor, K.; Norulaini, N.;
Omar, A. Techniques for extraction of bioactive compounds from plant materials: A review. J. Food Eng.
2013, 117, 426–436. [CrossRef]
Molecules 2019, 24, 636 11 of 11

11. Earle, M.J.; Esperança, J.M.; Gilea, M.A.; Lopes, J.N.C.; Rebelo, L.P.; Magee, J.W.; Seddon, K.R.; Widegren, J.A.
The distillation and volatility of ionic liquids. Nature 2006, 439, 831. [CrossRef]
12. Earle, M.J.; Seddon, K.R. Ionic liquids. Green solvents for the future. Pure Appl. Chem. 2000, 72, 1391–1398.
[CrossRef]
13. Dai, Y.; van Spronsen, J.; Witkamp, G.-J.; Verpoorte, R.; Choi, Y.H. Natural deep eutectic solvents as new
potential media for green technology. Anal. Chim. Acta 2013, 766, 61–68. [CrossRef] [PubMed]
14. Choi, Y.H.; van Spronsen, J.; Dai, Y.; Verberne, M.; Hollmann, F.; Arends, I.W.; Witkamp, G.-J.; Verpoorte, R.
Are natural deep eutectic solvents the missing link in understanding cellular metabolism and physiology?
Plant Physiol. 2011, 156, 1701–1705. [CrossRef] [PubMed]
15. Angell, C.A.; Ansari, Y.; Zhao, Z. Ionic liquids: Past, present and future. Faraday Discuss. 2012, 154, 9–27.
[CrossRef] [PubMed]
16. Abbott, A.P.; Boothby, D.; Capper, G.; Davies, D.L.; Rasheed, R.K. Deep eutectic solvents formed between
choline chloride and carboxylic acids: Versatile alternatives to ionic liquids. J. Am. Chem. Soc. 2004, 126,
9142–9147. [CrossRef] [PubMed]
17. Pena-Pereira, F.; Namieśnik, J. Ionic liquids and deep eutectic mixtures: Sustainable solvents for extraction
processes. ChemSusChem 2014, 7, 1784–1800. [CrossRef] [PubMed]
18. Maugeri, Z.; Leitner, W.; de María, P.D. Practical separation of alcohol–ester mixtures using Deep-Eutectic-
Solvents. Tetrahedron Lett. 2012, 53, 6968–6971. [CrossRef]
19. Hayyan, M.; Hashim, M.A.; Hayyan, A.; Al-Saadi, M.A.; AlNashef, I.M.; Mirghani, M.E.; Saheed, O.K.
Are deep eutectic solvents benign or toxic? Chemosphere 2013, 90, 2193–2195. [CrossRef]
20. Abbott, A.P.; Harris, R.C.; Ryder, K.S.; D’Agostino, C.; Gladden, L.F.; Mantle, M.D. Glycerol eutectics as
sustainable solvent systems. Green Chem. 2011, 13, 82–90. [CrossRef]
21. Harris, R.C. Physical Properties of Alcohol Based Deep Eutectic Solvents; University of Leicester: Leicester, UK, 2009.
22. Stefanovic, R.; Ludwig, M.; Webber, G.B.; Atkin, R.; Page, A.J. Nanostructure, hydrogen bonding and
rheology in choline chloride deep eutectic solvents as a function of the hydrogen bond donor. Phys. Chem.
Chem. Phys. 2017, 19, 3297–3306. [CrossRef]
23. Mulia, K.; Krisanti, E.; Terahadi, F.; Putri, S. Selected natural deep eutectic solvents for the extraction of
↵-mangostin from mangosteen (Garcinia mangostana L.) pericarp. Int. J. Technol. 2015, 7, 1211–1220. [CrossRef]
24. Deye, J.F.; Berger, T.A.; Anderson, A.G. Nile Red as a solvatochromic dye for measuring solvent strength in
normal liquids and mixtures of normal liquids with supercritical and near critical fluids. Anal. Chem. 1990,
62, 615–622. [CrossRef]
25. Bi, W.; Tian, M.; Row, K.H. Evaluation of alcohol-based deep eutectic solvent in extraction and determination of
flavonoids with response surface methodology optimization. J. Chromatogr. A 2013, 1285, 22–30. [CrossRef] [PubMed]
26. EFSA Panel on Additives and Products or Substances used in Animal Feed (FEEDAP). Scientific Opinion on
safety and efficacy of choline chloride as a feed additive for all animal species. EFSA J. 2011, 9, 2353.
27. Fowles, J.R.; Banton, M.I.; Pottenger, L.H. A toxicological review of the propylene glycols. Crit. Rev. Toxicol.
2013, 43, 363–390. [CrossRef] [PubMed]
28. US Food and Drug Administration. CFR—Code of Federal Regulations Title 21. Available online: https:
//www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfcfr/cfrsearch.cfm?fr=184.1666 (1,2-propanediol) and
https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfcfr/CFRSearch.cfm?fr=182.8252 (choline chloride);
(accessed on 30 January 2019).
29. Chemat, F.; Vian, M.A.; Cravotto, G. Green extraction of natural products: Concept and principles. Int. J.
Mol. Sc. 2012, 13, 8615–8627. [CrossRef] [PubMed]
30. Jessop, P.G.; Jessop, D.A.; Fu, D.; Phan, L. Solvatochromic parameters for solvents of interest in green
chemistry. Green Chem. 2012, 14, 1245–1259. [CrossRef]
31. Pothitirat, W.; Gritsanapan, W. Thin-layer chromatography-densitometric analysis of alpha-mangostin
content in Garcinia mangostana fruit rind extracts. J. AOAC Int. 2008, 91, 1145–1148. [PubMed]

Sample Availability: No samples are available from the authors.

© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).