Solubility of Collinin and Isocollinin in Pressurized Carbon Dioxide: Synthesis, Solubility Parameters, and Equilibrium Measurements

Article abstract of DOI:10.1021/acs.jced.9b00234

Collinin is a derivative of coumarin that has shown remarkable potential against cancer, tuberculosis, periodontitis, and other prevalent diseases, and is usually extracted from plants of the Rutaceae family at a very low yield. In this work, collinin and a position-isomer herein called isocollinin were synthesized at different scales (from 1 to 50 g of precursor) by a route consisting of two parallel and two sequential chemical reactions. The isomers were characterized by ¹H NMR, ¹³C NMR, nuclear Overhauser enhancement spectroscopy NMR, melting temperature, and melting enthalpy. For each isomer, the Hansen solubility parameters and the radius of its solubility sphere were experimentally determined by solubility tests in 15 common solvents and two solvent blends. The solubility of each isomer in pressurized CO₂ was determined at 30 and 50 °C from 72.2 to 112.9 bar, by an in situ high-pressure spectrometry technique, which was validated with the anthracene-CO₂ system. The solubility of both isomers in CO₂ increased with pressure in the range of temperatures and pressures considered, but that of collinin exhibited an asymptotic behavior around 80.8 and 104.8 bar, at 30 and 50 °C, respectively.

Full text of DOI:10.1021/acs.jced.9b00234

Article
pubs.acs.org/jced
Cite This: J. Chem. Eng. Data XXXX, XXX, XXX−XXX
Solubility of Collinin and Isocollinin in Pressurized Carbon Dioxide:
Synthesis, Solubility Parameters, and Equilibrium Measurements
́
̃
̃
Camilo Pardo-Castano, Andres C. García, Paola Benavides, and Gustavo Bolanos*
School of Chemical Engineering, Universidad del Valle, CL 13 100-00 Building E27, Cali, Colombia
S
* Supporting Information
ABSTRACT: Collinin is a derivative of coumarin that has shown remarkable potential against cancer, tuberculosis,
periodontitis, and other prevalent diseases, and is usually extracted from plants of the Rutaceae family at a very low yield. In this
work, collinin and a position-isomer herein called isocollinin were synthesized at different scales (from 1 to 50 g of precursor)
1
by a route consisting of two parallel and two sequential chemical reactions. The isomers were characterized by H NMR, 13C
NMR, nuclear Overhauser enhancement spectroscopy NMR, melting temperature, and melting enthalpy. For each isomer, the
Hansen solubility parameters and the radius of its solubility sphere were experimentally determined by solubility tests in 15
common solvents and two solvent blends. The solubility of each isomer in pressurized CO2 was determined at 30 and 50 °C
from 72.2 to 112.9 bar, by an in situ high-pressure spectrometry technique, which was validated with the anthracene−CO2
system. The solubility of both isomers in CO2 increased with pressure in the range of temperatures and pressures considered,
but that of collinin exhibited an asymptotic behavior around 80.8 and 104.8 bar, at 30 and 50 °C, respectively.
high-yield synthesis procedure reported by Curini and co-
workers, two isomers were obtained and not enough evidence
about their respective identification was reported. In addition,
their proposed synthesis route has only been tested on a
milligram scale. Because it is well known that synthesis yield at
a larger scale is usually much lower than that obtained at
milligram scale, it is important to obtain information on the
synthesis yield of collinin at a gram scale to favor the research
of this product as a potential pharmaceutical component.
In this work, collinin and an isomer, henceforth called
isocollinin, were synthesized and purified at a gram scale, by
using an improved synthesis procedure based on that reported
1. INTRODUCTION
Collinin is a geranyloxycoumarin that was originally isolated
from the bark of the Australian tree Flindersia collina Bail by
Anet et al.1 in 1949, and it is known to be present in plants of
the Rutaceae family. This compound has shown important
activity both in vitro and in vivo against diseases such as
cancer,2−6 tuberculosis,7 and periodontitis,8 among several
others.9−12 Such activity and the prevalence of these
pathologies make this substance an interesting leading
compound for the development of effective pharmaceutical
products, and thus, a knowledge on its properties, such as
solubility in conventional and nonconventional solvents, is of
interest to the pharmaceutical industry.
At present, collinin is extracted from the leaves of
Zanthoxylum schinifolium, a plant commonly used in the
medical traditions of China, Japan, Korea, and Taiwan.9
Unfortunately, solvent extraction of collinin provides very low
yields (around 0.011 wt%)3,7 and purities not higher than 97%.
As a result, several routes for chemical synthesis have been
proposed. For example, Curini et al.13 and Maes et al.14 studied
and reported routes of chemical synthesis with overall molar
yields as high as 24.6 and 6.8%, respectively. However, in the
by Curini et al.13 Both isomers were characterized by H
1
NMR, 13C NMR, nuclear Overhauser enhancement spectros-
copy (NOESY) NMR, melting enthalpy, melting temperature,
and purity as estimated from melting point depression.
Solubility experiments for both isomers were performed at
room temperature (ca. 28 °C) in 15 common organic solvents
Received: March 15, 2019
Accepted: August 2, 2019
© XXXX American Chemical Society
A
DOI: 10.1021/acs.jced.9b00234
J. Chem. Eng. Data XXXX, XXX, XXX−XXX

Journal of Chemical & Engineering Data
Article
a
Table 1. List of Reagents and Solvents
reagents and solvents
pyrogalol
propiolic acid
geraniol
phosphorus tribromide
sodium hydride
DBU
CAS
source
purity (mol %)
analysis
87-66-1
Alfa Aesar, Tewksbury, MA
99.6
99.5
99.2
99.9
GC
GC
GC
471-25-0
106-24-1
7789-60-8
7646-69-7
6674-22-2
74-88-4
60, oil dispersion
99.4
99.9
GC
GC
iodomethane
Merck, Darmstadt, Germany
dichloromethane
sulfuric acid
acetic acid
diethyl ether
ethyl acetate
acetone
acetonitrile
methanol
ethanol
75-09-2
7664-93-9
64-19-7
60-29-7
141-78-6
67-64-1
75-05-08
67-56-1
64-17-5
110-54-3
109-99-9
67-63-0
7757-82-6
7647-14-5
144-55-8
7487-88-9
119-61-9
7440-23-5
77-92-9
≥99.8
96.2
GC
alkalimetric
alkalimetric
GC
GC
GC
GC
GC
GC
GC
GC
GC
alkalimetric
argentometric
acidimetric
100.0
≥99.7
≥99.5
≥99.5
≥99.9
≥99.8
≥99.9
≥96
≥99.9
≥99.9
100.0
99.7
100.0
99.9
99.9
>99.8
100.0
100.0
99.9
hexane
tetrahydrofuran
2-propanol
sodium sulfate anhydrous
sodium chloride
sodium hydrogen carbonate
magnesium sulfate
benzophenone
sodium
citric acid 1-hydrate
dimethylformamide
cyclohexane
anthracene
isooctane
toluene
carbon dioxide
nitrogen (gas)
7,8-dihydroxycoumarin
geranyl bromide
7-hydroxy 8-geranyloxycoumarin
7-geranyloxy 8-hydroxyacoumarin
collinin
Sigma-Aldrich, St. Louis, MO
Carlo-Erba Reagents, Barcelona, Espan
GC
̃
a
́
Panreac Quimica, Barcelona, Spain
acidimetric
GC
68-12-2
110-82-7
120-12-7
540-84-1
108-88-3
124-38-9
7727-37-9
486-35-1
6138-90-5
Fisher Chemicals, Fairlawn, NJ
98.7
99.9
99.9
99.90
99.9990
DSC
HPLC
GC
Mallinckrodt Baker, Phillipsburg, NJ
Cryogas INDURA Group, Cali, Colombia
this work
1H and 13C NMR, DSC
1H NMR
1H and NOESY NMR, DSC
97.4
94.5
98.0
98.5
34465-83-3
1H and 13C NMR, DSC
1H, 13C, and NOESY NMR, DSC
isocollinin
a
GC: gas chromatography, DSC: differential scanning calorimetry, DBU: 1,8-diazabicyclo[5.4.0]undec-7-ene
and two solvents blends, to estimate the dispersion (δD), polar
(δP) and hydrogen-bonding (δH) Hansen solubility parameters
(HSPs), and the radius of the solubility sphere (Ro) of each
isomer. Results were compared to solubility parameters of CO2
at different temperatures and pressures to predict whether each
isomer would dissolve in pressurized CO2 at any conditions.
An experimental apparatus was assembled in our laboratory
to perform high-pressure equilibrium measurements, using an
in situ spectrometric technique. Its performance was validated
by measuring the solubility of anthracene in pressurized CO2 at
40 °C. Finally, the solubilities of both isomers were determined
in pressurized CO2 at 30 and 50 °C.
vacuum and collected in a flask, which contained 3 Å molecular
sieves to keep THF dry. In addition, tap water was purified in
our laboratory by two successive steps of filtration, followed by
deionization and distillation.
Anthracene was characterized by using differential scanning
calorimetry (TA Instruments, DSC25 model, New Castle,
DE), to check its purity, as it was used in a validation
procedure. Purity (mol %) approximation was computed
(software TRIOS v4.4.0), following the ASTM E928
method.15 A melting temperature of 217.0 °C with a melting
depression of 0.3 °C was calculated. The corresponding
thermogram has been included as Supporting Information.
These values are consistent with the 216.7
0.8 °C value
reported by Radomska and Radomski.16
2. EXPERIMENTAL SECTION
2.2. Synthesis of Collinin and Isocollinin. The synthesis
of collinin (10) and isocollinin (11) carried out in this work
was based on the works by Curini et al.13 and Hanessian et
al.,17 and consists of two parallel reactions followed by two
sequential reactions, which are shown in Scheme 1. Important
2.1. Materials. Table 1 shows the list of reagents and
solvents along with its purity and supplier. All compounds were
used without further purification except for tetrahydrofuran
(THF), which was dried by using sodium as desiccant and
benzophenone as indicator. Dried THF was distilled under
B
DOI: 10.1021/acs.jced.9b00234
J. Chem. Eng. Data XXXX, XXX, XXX−XXX

Journal of Chemical & Engineering Data
Article
scale, the change in yield was smaller, which indicates that at a
larger scale, the yield might approximate a constant value.
2.2.2. Synthesis of Geranyl Bromide (8). Phosphorus
tribromide (5) (6 mL, 64.8 mmol) was added to a mixture
of geraniol (4) (28 mL, 162.1 mmol) and diethyl ether (350
mL) in a light-isolated flask, under nitrogen atmosphere. After
3 h at 0 1 °C and constant stirring, the resulting crude was
neutralized using a saturated sodium bicarbonate solution (350
mL) and extracted twice with diethyl ether (1 × 100, 1 × 150
mL). The organic extracts were collected and dried using
sodium sulfate and evaporation under vacuum to obtain
geranyl bromide (6) (33.5 g, 154.3 mmol) as a golden oil
(95.2% yield). This procedure was repeated for different
amounts of geraniol (11 and 23 mL). The results indicate that
molar yield does not significantly decrease (96.5 and 95.0%,
respectively) with increasing synthesis scale.
Scheme 1. (a) 125 °C, 50 min, H2SO4; (b) 0 °C, 2 h, Light-
Free; (c) 28 °C, 5 h, Acetone, DBU; (7) R1 = R2 = Geranyl
Group; (8) R1 = H, R2 = Geranyl Group; (9) R1 = Geranyl
Group, R2 = H; (d) 28 °C, 18 h, THF, NaH, MeI; (10) R1 =
Geranyl Group, R2 = CH3; (11) R1 = CH3, R2 = Geranyl
Group.
2.2.3. Synthesis of 7-Hydroxy 8-Geranyloxi Coumarin (8)
and 7-Geranyloxi 8-Hydroxy Coumarin (9). Geranyl bromide
(6) (12.2 g, 56.1 mmol) was added to a solution of daphnetin
(3) (10.0 g, 56.1 mmol) in acetone (150 mL) and DBU (8.5 g,
56.1 mmol). The mixture was stirred at 28 1 °C for 5 h. The
resulting crude was neutralized using a 2 wt% citric acid
solution (300 mL) and extracted with dichloromethane (2 ×
250 mL). The organic phase was dried by using sodium sulfate
and evaporation under vacuum.
Three compounds 7−9 were isolated from the resulting
solid as follows. The dialkylated compound was extracted as an
oil (7) (4.2 g, 7.8 mmol) using (2 × 200 mL) hexane (13.9
yield %). The rest of the mixture was first separated from
insolubles (3.7 g) in both hexane and dichloromethane; then,
it was subjected to column chromatography using a silica gel
60 (70−230 mesh)/mixture ratio of 20 and to mixtures of
dichloromethane/methanol (97/3 v/v)/hexane (50/50, 60/
40, 70/30, and 80/20 v/v, 500 mL each, sequentially), to
afford isomers 8 (5.5 g, 17.4 mmol) and 9 (1.4 g, 4.3 mmol) as
white solids, once the solvent was completely evaporated under
vacuum (31.0 and 7.7% yield, respectively). The selectivities
were 26, 59, and 15% for compounds 7, 8, and 9, respectively.
Most of the solvent used for the chromatographic separation
was recovered by distillation in a packed column.
variations to the original procedure were included to improve
the synthesis yield and efficiency, especially at larger scales.
Briefly, 7,8-dihydroxycoumarin (3), commonly known as
daphnetin, was synthetized using pyrogallol (1), propiolic
acid (2), and concentrated sulfuric acid as catalyst, at 125 °C
during 50 min. In parallel, geranyl bromide (6) was synthetized
using geraniol (4) and phosphorus tribromide (5), under light-
free conditions, at 0 °C during 2 h. The position isomers 7-
hydroxy 8-geranyloxi coumarin (8) and 7-geranyloxi 8-hydroxy
coumarin (9) were then obtained by geranylation of 3 using
1,8-diazabicyclo[5.4.0]undec-7-ene (DBU) or potassium car-
bonate (K2CO3), at 28 °C, during 5 h. The use of DBU led to
a higher selectivity and yield of the position isomers 8 and 9
than the use of K2CO3. In both cases, when we looked for total
consumption of the raw material, the formation of 7,8-
dialkylated coumarin (7) was unavoidable. Finally, methylation
of 8 and 9 using sodium hydride and methyl iodide in THF, at
28 °C for 12 h, gave collinin (10) and isocollinin (11),
respectively.
2.2.1. Synthesis of 7,8-Dihydroxycoumarin (3). A mixture
of pyrogallol (1) (10 g, 79.3 mmol), propiolic acid (2) (10 g,
142.7 mmol), and sulfuric acid (nine drops) was stirred at 125
1 °C for 50 min. The resulting crude was allowed to cool
and then washed with ethyl acetate (100 mL). The suspended
solid was separated by vacuum filtration and washed with a
saturated solution of sodium bicarbonate, until no evolution of
carbon dioxide was observed. The mixture was rinsed with
water until the yellowish coloration disappeared. The solid was
oven-dried at 80 °C for 40 min to obtain 3 (6.6 g, 37.6 mmol)
as a light brown solid (47.0% yield). Purification of the target
compound present in the ethyl acetate extract was performed
by extraction with a 5 wt% sodium bicarbonate solution, drying
over sodium sulfate, evaporation under vacuum, and
recrystallization in diethyl ether, to achieve a global yield of
51.4%.
It is worth mentioning that Curini et al.13 proposed a
mixture of dichloromethane/methanol (99/1 v/v) to purify a
precursor of a collinin isomer at a retention factor of 0.5. Even
though that mixture allowed us to isolate the three compounds
on a small-scale synthesis (e.g., 0.3 g), it was not the case for
larger-scale synthesis (e.g., 10 g). At larger scales, when silica
gel 60 (230−450 mesh) was used, the retention factor
difference was smaller, leading to an inefficient separation
procedure. Conversely, when silica gel 60 (70−230 mesh) was
used, the mobile phase flowed so fast that the compounds were
not retained enough in the stationary phase, also leading to an
inefficient but faster procedure. As a consequence, a different
solvent mixture must be used to purify the isomers.
The synthesis and purification procedures were repeated at
different scales in terms of the daphnetin mass. The results
indicate that for the 0.3 and 20 g scale syntheses, the three
products 7−9 were obtained at (17.6, 41.7, 10.4) and (11.6,
22.3, 5.6) molar yield %, respectively. As expected, the molar
yield decreases as scale increases, particularly at this alkylation
stage of daphnetin where the large number of purification steps
led to mass losses. However, one might expect an asymptotic
behavior like the one observed for the first reaction stage.
Molar yields of 9.3, 16.1, and 4.0% were estimated for a 30.2 g
The described procedure was repeated at scales of 1, 20, and
50 g of precursor. As a result, molar yields of 53.4, 43.5, and
42.7% were obtained, respectively. As expected, the yield
decreased as the synthesis scale increased, but after the 20 g
C
DOI: 10.1021/acs.jced.9b00234
J. Chem. Eng. Data XXXX, XXX, XXX−XXX

Journal of Chemical & Engineering Data
Article
scale synthesis. Table 2 presents molar yields of compounds
7−9 estimated in terms of the pyrogallol mass.
Table 3. Molar Yield of Compounds Collinin (10) and
Isocollinin (11) Estimated in Terms of Pyrogallol (1) Mass
a
molar yield (%)
Table 2. Molar Yield of Compounds 7,8-
Digeranyloxycoumarin (7), 7-Hydroxy 8-
pyrogallol (1) (g)
isomer (8) (g)
isomer (9) (g)
10
11
1
10
20
50
0.55
4.30
6.20
8.59
0.14
1.07
1.54
2.14
46.6
37.0
35.5
34.2
62.9
55.7
54.5
53.3
Geranyloxycoumarin (8), and 7-Geranyloxy 8-
Hydroxyacoumarin (9), Estimated in Terms of Pyrogallol
a
(1) Mass
molar yield (%)
a
Refer to Scheme 1 for specification of compounds.
pyrogallol (1) (g)
daphnetin (3) (g)
7
8
9
Table 4. Overall Yield of Compounds Collinin (10) and
Isocollinin (11) Estimated in Terms of Pyrogallol (1) Mass
1
10
20
50
0.8
7.3
12.3
30.2
17.3
15.1
13.5
9.3
41.4
33.6
28.6
16.1
10.3
8.4
7.1
a
overall molar
yield (%)
4.0
a
Refer to Scheme 1 for specification of compounds.
pyrogallol (1) (g) isomer (10) (g) isomer (11) (g)
10
11
1
10
20
50
0.14
1.07
1.54
2.14
0.55
4.30
6.20
8.59
2.6
1.6
1.1
0.6
13.9
9.6
6.8
2.2.4. Synthesis of Collinin (10) and Isocollinin (11).
3.7
Mineral oil from a 60% sodium hydride dispersion (1.1 g) was
removed with anhydrous THF (20 mL), and additional 120
mL of anhydrous THF were added to the residual sodium
hydride. Compound 8 (3.0 g, 9.6 mmol) was dissolved in
anhydrous THF (30 mL) and added to the suspension of
sodium hydride. All previous procedures were performed at 0
1 °C under nitrogen atmosphere. The produced hydrogen
was purged from the flask using nitrogen. The resulting
solution was constantly stirred for 2 h at room temperature
(ca. 28 °C), after which iodomethane (5.5 mL, 88.4 mmol)
was added to the solution. Methylation proceeded at room
temperature (ca. 28 °C) during 16 h, time at which all of the
precursor material was consumed.
To the resulting reaction mixture, 200 mL of a saturated
solution of sodium chloride was added. The aqueous layer was
extracted with diethyl ether (4 × 400 mL), and the organic
extracts were dried over sodium sulfate and distilled under
vacuum. The distilled mixture was then purified by column
chromatography, using a silica gel 60 (230−450 mesh)/
mixture ratio of 26 and a solvent mixture of hexane/methanol
(96/4 v/v, 1500 mL) to afford isomer 11 as a golden oil, which
crystallized as a white solid (57.0% yield) after refrigeration at
4 °C. Most of the solvent used in the chromatographic
separation was recovered by distillation. The same synthesis
and purification procedures were also performed for
compound 9 (1.0 g, 3.2 mmol) to afford the isomer 10
(0.39 g, 1.2 mmol) as a white solid (37.9% yield). Compound
10 required recrystallization (hexane/ethanol 96% v/v) due to
the presence of impurities, including raw material that was not
consumed.
The complete procedure was repeated at different scales in
terms of the mass of compounds 8 (4.0 and 5.0 g) and 9 (0.25
and 0.5 g). As a result, molar yields of (56.5 and 55.4%) and
(44.3 and 39.1%) were obtained, respectively. Table 3 presents
the molar yields of compounds 10 and 11 estimated in terms of
the pyrogallol mass.
Table 4 presents the overall molar yields of compounds 10
and 11 estimated for each overall scale in terms of the
pyrogallol mass. Overall molar yield values show a logarithmic
decay, which tends to a constant value as the overall scale
increases. Future work will be focused on making the overall
synthesis and purification process more efficient and environ-
mentally friendly.
a
Refer to Scheme 1 for specification of compounds.
2.3. Characterization. Compounds 3, 6, and 8−11 were
1
characterized by H NMR. Some of them, 8, 9, and 11, were
also characterized by using one-dimensional (1D) NOESY
NMR, and 10 and 11 by 13C NMR. Table 5 shows the H
1
NMR chemical shifts of the compounds 3, 6, and 8−11. Figure
1
1
1 shows the H NMR spectrum of compound 10. H NMR
spectra of precursors 3 and 6 and products 10 and 11 are
shown in the Supporting Information (see Figures S1−S4),
where the spectra are presented as obtained in the NMR
processor.
1
The H NMR chemical shifts of compound 3 are different
from those obtained by Curini et al.13 However, they are
consistent with those obtained by Venditti et al.18 and Gao et
al.19 The largest discrepancies were 0.04 and 0.06 ppm,
respectively. Apparently, there are differences in the proton
assignments regarding those reported by Curini et al.,13 even
so, the deviation is significant in the chemical shift of proton
1
H5 (0.17 ppm). On the other hand, the H NMR chemical
shifts of compound 6 are also consistent with those obtained
by Van Der Klei et al.,20 with 0.04 ppm as the largest
difference.
1
The H NMR chemical shifts of the isomers 8 are not
reported in the literature, and those of isomers 9 were
apparently reported in Curini’s work.13 As expected, the
discrepancies found in product 3 remained in products 9 after
the geranylation of 3. Besides, there are differences in the
chemical shifts of protons H5′, H6′, and H7′ of at least 0.1
ppm, which correspond to the methyl groups in the geranyl
group chain.
As mentioned in the synthesis procedure, Curini et al.13
assured they obtained 7-geranyloxy 8-hydroxy coumarin (9)
concentrating the fraction with a retention factor of 0.5. In
preliminary experiments, we tried to replicate Curini’s work
and observed that fractions a, b, and c were obtained at
retention factors of 0.27, 0.50, and 0.70, respectively. The first
fraction was the dialkylated coumarin (7), while the two
remaining ones corresponded to isomers 8 and 9. Even though
1
the H NMR chemical shifts of these two isomers presented
small differences between them, this information was not
enough to identify each isomer.
D
DOI: 10.1021/acs.jced.9b00234
J. Chem. Eng. Data XXXX, XXX, XXX−XXX

Journal of Chemical & Engineering Data
Article
b
1
Table 5. H NMR Chemical Shifts of Compounds 7,8-Dihydroxycoumarin (3), Geranyl Bromide (6), 7-Hydroxy 8-
a
Geranyloxycoumarin (8), 7-Geranyloxy 8-Hydroxyacoumarin (9), Collinin (10), and Isocollinin (11)
compound
protons
1H(3)
1H(4)
1H(5)
1H(6)
3H(OCH3)
2H(1′)
1H(2′)
4H(3′,4′)
1H(5′)
3H(6′)
3H(7′)
3H(8′)
multiplicity
3
6
8
9
10
11
d
d
d
d
s
d
t
m
m
s
6.20
7.85
7.01
6.84
6.27
7.66
7.14
6.92
6.29
7.65
7.01
6.88
6.26
7.64
7.16
6.88
4.00
4.71
5.51
2.12
5.09
1.61
1.67
1.77
6.24
7.63
7.17
6.87
3.94
4.70
5.59
2.02
5.05
1.58
1.66
1.68
4.06
5.56
2.11
5.11
1.64
1.68
1.76
4.88
5.53
2.07
5.07
1.62
1.70
4.73
5.52
2.13
5.10
1.63
1.69
1.78
s
s
a
b
The presented reference molecule corresponds to compound 10. Refer to Scheme 1 for specification of compounds. Obtained in a Bruker
Avance II UltraShield 400 MHz NMR spectrometer (Billerica, MA).
Scheme 2. NOESY NMR Experiments of Compounds (a) 8
and (b) 9; Coupling of Protons H6 and H1′.
1
Figure 1. H NMR chemical shifts of compound 10.
Selective 1D NOESY NMR experiments were performed to
see the couplings of proton H6 and H1′. The selective 1D
NOESY NMR spectra are shown in the Supporting
Information (see Figures S5 and S6). Scheme 2a,b shows the
coupling of protons H6 and H1′ in compounds 8 and 9,
respectively. This indicates that for compound 8 or fraction b
on the thin-layer chromatography (TLC) plaque, the H6
proton couples to H5, and H1′ couples to H2′ and H8′,
indicating that the geranyloxy chain is linked in position 8. In
both cases, a barely visible coupling of H6 and H1′ with the
hydroxyl group −OH appears. On the other hand, for
compound 9 or fraction c on the TLC plaque, H6 couples
to H5 and H1′, indicating that the geranyloxy chain is linked in
position 7. These results strongly suggest that fractions b and c
correspond to compounds 8 and 9.
The 1H NMR chemical shifts of the isomers 10 are
consistent with those reported by Kim et al.3 and Chen et al.10
for the so-called collinin compound. However, Curini et al.13
did not report 1H NMR for the isomer synthesized and
purified in their work, but instead they quoted Chen et al.’s
work.10 The H NMR chemical shifts of isomer 11 present
1
some small but not conclusive deviations in regard to those
already reported for the so-called collinin compound. Thus,
selective 1D NOESY NMR experiments were also performed
to look at the couplings of H6 proton in isomer 11. The
E
DOI: 10.1021/acs.jced.9b00234
J. Chem. Eng. Data XXXX, XXX, XXX−XXX

Journal of Chemical & Engineering Data
Article
selective 1D NOESY NMR spectra are shown in the
Supporting Information (see Figure S7). As expected, the
results indicate that for compound 11, the H6 proton is
coupled to protons H5 and methyl protons −OCH3. These
results validate that the compound synthesized after the
methylation of 8 corresponds to isocollinin (11). Therefore,
methylation of 9 leads to collinin (10).
Compounds 3 and 8−11 were also characterized by
differential scanning calorimetry (DSC). The DSC experi-
ments were carried out using an instrument (TA Instruments,
DSC25 RCS90 model), whose performance was verified using
an indium standard (series number LGC2601), a multimeter
(Fluke 179, serial number 35840218), and a flowmeter (Cole-
Parmer, serial number 155533). The melting temperature of
Table 6 shows the 13C NMR chemical shifts of the products
10 and 11. 13C NMR spectra are shown in the Supporting
Table 7. Thermal Properties and Purity of Compounds 7-
Hydroxy 8-Geranyloxycoumarin (8), 7-Geranyloxy 8-
b
a
Table 6. 13C NMR Chemical Shifts of Compounds Collinin
Hydroxyacoumarin (9), Collinin (10), and Isocollinin (11)
a
(10) and Isocollinin (11)
compound
b
u
8
9
10
11
melting enthalpy (J/g)
melting temperature (°C)
melting point depression (°C)
purity
8.7
0.1
0.1
0.2
100.3
105.6
1.0
88.5
97.2
2.0
76.2
45.6
0.7
98.0
b
73.3
43.7
0.5
97.4
94.5
98.7
compound
a
Refer to Scheme 1 for specification of compounds Standard
uncertainty P = 900.6 mbar (Patm), u(P) = 2.5 mbar.
carbons
12
13
C(2)
CH(3)
C(4)
160.66
113.69
143.68
113.38
122.66
110.28
155.03
141.89
148.18
61.38
160.55
113.63
143.71
113.29
122.9
108.46
156.13
142.65
148.53
56.35
the indium standard was measured as 157.2 °C (reference
value: 156.6 1.3 °C), and the melting enthalpy was 28.5 J/g
C(4a)
(reference value: 28.7
0.8 °C). All experiments were
CH(5)
CH(6)
C(7)
C(8)
C(8a)
OCH3
CH2(1′)
CH(2′)
C(2′a)
CH2(3′)
CH2(4′)
CH(5′)
C(5′a)
CH3(6′)
CH3(7′)
CH3(8′)
performed using a sample size of ca. 7 mg, a scanning rate of
0.5 °C/min, and processed by onset temperature. Figures
S10−S15 in the Supporting Information show the thermo-
grams for anthracene and all of the synthesized solid-state
compounds.
Table 7 shows the melting enthalpy, melting temperature,
melting point depression, and purity of compounds 8−11.
These values were estimated according to the methodology
described in the ASTM E92815 standard. The melting
properties and purity of compound 3, however, were not
included. This compound presents a complex thermal
behavior, which is shown in the Supporting Information as
Figure S11. Although a melting peak for this compound
appears at 266.6 °C, its thermogram indicates that
decomposition takes place and purity cannot be estimated
from this information.
The melting temperature of compound 3 is comparable to
that obtained by Gao et al.,19 which goes from 265.7 up to
267.2 °C. However, it differs from those obtained by Curini et
al.13 (255−256 °C) and by Liu and Tian21 (253−255 °C).
Melting point temperatures of isomers 8 and 9 are comparable
to those obtained by Curini et al.,13 which go from 97 up to 99
°C, while melting points of isomers 10 and 11 are not
comparable to those obtained by Curini et al.13 (67−69 °C),
Kim et al.3 (66.5−67 °C), and Maes et al.14 (67−68 °C).
Differences among the melting temperatures obtained in this
work and those reported in the literature might be explained as
a result of the compound’s purity and purification method. In
fact, it is well known that molecules in the solid state can
arrange in different configurations, leading to different melting
temperatures for the same compound, a fact that depends on
many factors, such as crystallization method and temperature,
and solvent(s) used in the process, among others. It is
important to note that many of the references cited above do
not report information about their melting point determination
technique or the purity of the compound.
66.34
69.81
119.02
136.69
39.49
119.62
134.76
39.59
26.24
26.43
123.65
131.88
16.76
17.73
25.68
123.89
131.57
16.41
17.64
25.64
a
The presented reference molecule corresponds to compound 10.
b
Refer to Scheme 1 for specification of compounds. Obtained in a
Bruker Avance II UltraShield 100 MHz NMR spectrometer (Billerica,
MA).
Information file (see Figures S8 and S9). The 13C NMR
chemical shifts of compound 10 are consistent with those
reported by Kim et al.,3 but not with those obtained by Curini
et al.13 Instead, 13C NMR chemical shifts obtained for
compound 11 are consistent with those obtained by Curini
et al.13 The main differences between the two isomers are
found in the chemical shifts of carbons OCH3 and CH2(1′),
where the position of the geranyloxy group differs for the two
isomers. For instance, Curini et al.13 reported chemical shifts of
55.9 and 69.8 for the OCH3 and CH2(1′) carbons,
respectively, while chemical shifts of 61.60 and 66.50 were
reported by Kim et al.,3 for the same carbons. As seen in Table
6, the signals of the aforementioned carbons appear at
chemical shifts of 61.38 and 66.34 for compound 10, and
56.35 and 69.81 for compound 11. These facts strongly suggest
once again that compounds 10 and 11 correspond to collinin
and isocollinin, respectively.
2.4. Solubility Parameters. The Hansen solubility
parameters (HSPs) of collinin and isocollinin, as well as the
radius of the solubility sphere (Ro) were estimated from a set
F
DOI: 10.1021/acs.jced.9b00234
J. Chem. Eng. Data XXXX, XXX, XXX−XXX

Journal of Chemical & Engineering Data
Article
Figure 2. Schematic representation of the experimental apparatus for determination of solubility in pressurized CO2.
−1.25
of solubility tests by using the “good/bad” test proposed by
δD,ref
δD
ν
ν
i
y
ref
j
j
j
z
z
z
Hansen.22 Fifteen common organic solvents, namely, diethyl
ether, dichloromethane, acetone, methanol, tetrahydrofuran,
hexane, ethyl acetate, ethanol, cyclohexane, acetonitrile, 2-
propanol, isooctane, toluene, acetic acid, and dimethylforma-
mide, were selected for solubility tests of collinin and
isocollinin. Solvent blends of hexane−ethanol were also tested
for each isomer, from 0 vol % ethanol until dissolution was
observed. For each solvent, 1 mL was added to a vial
containing 23 mg of the isomer, and the mixture was then
constantly stirred by using a vortex (Thermo Scientific,
MaxiMix III model, Waltham, MA) at 1400 rpm for 6 h.
Scores of 1 or 0 were assigned to “good” or “bad” solvents,
respectively. “Good solvents” were those that completely
dissolved the solute, and “bad solvents” those that did not. The
dispersion (δD), polar (δP), and hydrogen-bonding (δH)
solubility parameters for each solvent were gathered from
Hansen’s book22 and were used to calculate the so-called
“distance” parameter (Ra), according to
=
(2)
k
{
−0.50
δP,ref
δP
ν
i
y
ref
j
z
z
= j
j
z
ν
(3)
k
{
Ä
É
Ñ
Ñ
Ñ
Ñ
Ñ
Ñ
Å
Å
Å
Å
Å
0.50
δH,ref
δH
ν
ν
i
j
j
j
y
z
z
z
−
3
ref
= exp −1.32 × 10 (T − T) − ln
Å
ref
Å
Å
Å
Ñ
Ñ
Ñ
k
{
(4)
Å
Ç
Ñ
Ö
where Tref = 298 K is the reference temperature, νref = 39.13
cm3/mol is the reference molar volume, and δD,ref = 15.6
MPa1/2, δP,ref = 5.2 MPa1/2, and δH,ref = 5.8 MPa1/2 are the
reference HSPs of CO2. All of these reference values were
gathered from the literature.23 Briefly, reference HSPs were
obtained using a procedure similar to that mentioned above for
the good/bad test. However, in this case, the good solvents
were those in which the solubility of CO2 was greater than the
ideal solubility at 298 K and 1 atm. The reference molar
volume was estimated using Huang’s equation of state to
obtain a total reference solubility parameter of δt = 17.4
MPa1/2, at 917 bar.
2.5. Experimental Apparatus for High-Pressure
Equilibrium Measurements. Figure 2 shows a schematic
representation of the experimental apparatus for the determi-
nation of the solubility of collinin or isocollinin in pressurized
CO2. The apparatus uses a pumping system to feed CO2 into a
high-pressure view cell, in which an amount of collinin or
isocollinin has been loaded, and a UV−vis spectrometry
assembly that detects changes in the absorbance of the mixture
until it reaches equilibrium.
Ra = 4(δD2 − δD1)2 + (δP2 − δP1)2 + (δH2 − δH1)2
(1)
where the subscripts 1 and 2 refer to the collinin isomer and
the solvent, respectively.
For a given collinin isomer, the Ra parameter gives the
relative distance between the solubility parameters of the
solvent and those of the solute, in a three-dimensional space.
An iterative procedure was performed to determine the
solubility parameters and the radius of a solubility sphere of
the collinin isomer that includes all of the good solvents, but
excludes the bad ones. This target was achieved when the Ra/
Ro ratio was less than 1 for all of the good solvents and greater
than 1 for the bad solvents.
The HSPs obtained for the collinin isomers were compared
to the parameters of CO2 at different temperatures and
pressures, to estimate whether they dissolve in CO2 at any
condition. The HSPs of CO2 were calculated from the
equations23
In more detail, liquid CO2 is pumped from a refrigeration
system (A−C) to the high-pressure view cell (F) using a
pneumatic pump (CP250V225 model, Milton Roy, Ivyland,
PA) (D). The pressure inside the cell is finely adjusted using a
pressure generator (62-6-10 model, HiP, Erie, PA) (E). This
cell is an assembly of a central body and a cap made of stainless
steel 316, designed and built in our laboratory to operate up to
300 bar. The central body has four lateral cavities, each one
holding a UV-fused-silica window (diameter, 10.00
mm, thickness, 12.0 0.3 mm, surface flatness, < λ/10 at 633
0.25
G
DOI: 10.1021/acs.jced.9b00234
J. Chem. Eng. Data XXXX, XXX, XXX−XXX

Journal of Chemical & Engineering Data
nm, Thorlabs, Sao Carlos, SP, Brazil), which in turn is
maintained in place by threaded bushings. The fluid inside the
cell is stirred using a magnetic bar, which is externally activated
by a magnetic stirrer (PC-420 model, Corning Inc., Corning,
NY).
The view cell is located on a thermoelectric heating and
cooling system (G). This system consists of a type J
thermocouple that measures the temperature of a thermally
isolated copper jacket, which covers the visual cell. The signal
is transmitted to a controller (SDC50D controller, Pacific
supercool, Klongtoey, BK, Thailand), which finally acts on a
Peltier plaque (TEC-12715 model). Heat from the Peltier
plaque is dissipated in an aluminum extended surface
refrigerated by an ethylene glycol solution pumped from a
refrigeration unit. This system can control the temperature of
Article
̃
was purged. No solute losses were detected when the solute
was rediluted and analyzed by UV spectrometry. The empty
cell was then taken as a blank, and the system was pressurized
with CO2 and stirred constantly, until it reached equilibrium
(i.e., the absorbance was constant) at the desired values of
pressure and temperature. At this moment, the absorbance was
recorded and the solute was pumped out from the cell at a
pressure 34 barg higher than the intended equilibrium
pressure, until the absorbance signal approached zero. Finally,
the cell was washed twice with acetone to remove any traces of
solute inside the visual cell and was dried at 60 °C in an oven.
The latter procedure was repeated at a higher pressure until
there was no significant increase of absorbance of the solute as
a function of pressure.
To illustrate, Figure 3 shows the absorbance of the
anthracene−CO2 system as a function of pressure, at 40 °C,
the copper jacket between 5 and 100 °C to a precision of
°C.
1
Temperature of the fluid in the visual cell was measured to a
precision of 0.3 °C by using a type-K thermocouple (Omega,
Norwalk, CT) and a digital thermometer (52-II model, Fluke,
Everett, WA) (H), both calibrated against a high-precision
RTD (1512A model, 0.0006 °C, Fluke), whose calibration is
in turn traceable to an NIST standard.
Pressure in the visual cell was measured to a precision of
5
psig (0.3 barg) by using a 12 in. precision Bourdon manometer
(No. A4A133994, 0−7500 psig, 10 psig subdivisions, 0.1%
accuracy, Ashcroft, Stratford, CT) (I), calibrated against a
digital manometer (DPI 104 IS model, 0−68 950 kPa, 1 kPa
subdivision, 0.05% accuracy, GE DRUCK, Billerica, MA).
Finally, a light beam from a source (DH-2000, OceanOptics,
Winter Park, FL) (J) is directed by an optical fiber probe
(XSR-BX model, 600 μm core diameter, 0.25 m length,
OceanOptics) (K) and collimator lens (7 V model, Ocean-
Optics) (L) to pass through the mixture in the visual cell. The
light signal that leaves the visual cell is directed by an identical
optical fiber probe to a minispectrophotometer (USB4000
model, OceanOptics) (M) and is analyzed using a computer
software (Spectrasuite OceanOptics) (N) to determine
absorbance.
2.5.1. Validation of the Experimental Apparatus. The
validation of the experimental apparatus was performed by
determining the solubility of anthracene in pressurized CO2, a
system for which there is plenty of information reported in the
literature.24 The solubility of anthracene was measured
following a methodology similar to that proposed by Ngo et
al.,25 who used a spectrometry technique to determine the
solubility of several solids in supercritical carbon dioxide,
including anthracene. Even though this methodology is limited
by the detection limit of the spectrometry apparatus, it does
not require sampling or calibration of the solute concentration
in the solvent, and it is very accurate.
All experiments were performed at 40 °C and pressures from
85.6 to 108.8 bar. A calibration curve was built to correlate the
amount of a solution of anthracene in cyclohexane (9.9−70.5
mg, 2.0 mg/g) with the volume measured with a micropipette
(20−100 μL). Each amount was measured by triplicate, and a
linear model was fitted to serve as a calibration curve (R2 =
0.9940). In a typical experiment, the amount of solution
required (in μL) was computed from the linear model, verified
experimentally eight times, and loaded into the visual cell. The
solvent was then evaporated at 60 °C for ca. 30 min, during
which there is no trace of solvent according to an evaporation
curve previously built. The cell was sealed and the air inside
Figure 3. Absorbance of the anthracene−CO2 system at 40 °C as a
function of pressure for 0.257 mg of anthracene.
for a 0.257 mg sample of anthracene. The absorbance was read
at the wavelength of the maximum absorbance peak (λmax),
which varied as a function of pressure. In this case, the
maximum absorbance peak was read at wavelengths from
346.58 to 348.90 nm. The solubility data were obtained at the
intersection of two straight lines fitted to the experimental
data, where the absorbance of the system begins to be
constant. At this intersection point, a solubilization pressure of
88.6 bar was obtained.
Table 8 lists the values of solubility of anthracene in
pressurized CO2 at 40 °C, and Figure 4 shows a comparison of
the values of solubility of anthracene measured in this work
and those reported in the literature.25−27 As observed, our
experimental data follow the tendency of data selected from
the literature. In particular, our data are consistent with those
obtained by Ngo et al.25 As mentioned above, these authors
Table 8. Solubility of Anthracene in CO2 at 40 °C, as a
Function of Pressure
a
b
pressure
(bar)
CO2 density
(mol/L)
CO2 mass
(mol)
y (106) (mole fraction) u(y)
85.6
88.6
92.5
97.8
108.8
8.44
10.15
12.04
14.40
15.50
0.17
0.21
0.25
0.30
0.32
4.1
6.9
8.7
10.3
26.4
0.4
0.2
0.1
0.1
0.1
a
b
Standard uncertainty: u(T) = 0.1 °C, u(P) = 1.2 bar. Calculated
from temperature and pressure using miniREFPROP, version 9.1.30
H
DOI: 10.1021/acs.jced.9b00234
J. Chem. Eng. Data XXXX, XXX, XXX−XXX

Journal of Chemical & Engineering Data
Article
temperature, and liquid CO2 was pumped until a desired
pressure was reached. The equilibrium was determined by
following the absorbance of the solutes until no variation was
observed. At this moment, the absorbance was recorded. This
procedure was repeated until no significant variation of the
absorbance (<0.01 units) was observed as pressure was
increased. Each determination of solubility was repeated at
least twice.
3. RESULTS AND DISCUSSION
3.1. Solubility Parameters. Table 9 shows solubility
scores of collinin and isocollinin in 15 common organic
solvents and two solvent blends. As observed, only the
solubility score of the isomers in cyclohexane was different
among all common organic solvents tested, which completely
dissolved isocollinin, but partially dissolved collinin. In
addition, a 99.4 vol % hexane−ethanol mixture partially
dissolved collinin, but completely dissolved isocollinin, and a
96.0 vol % hexane−ethanol mixture completely dissolved both
isomers. These results indicate that collinin is a little more
polar (or less apolar) than isocollinin, which was also observed
by thin-layer chromatography. We speculate that this fact could
be explained by the spatial arrangement of the geranyl chain
linked to positions 7 or 8, but a rigorous explanation is beyond
the scope of this work.
Figure 4. Solubility of anthracene in pressurized CO2 at 40 °C.
○
Comparison with data reported in the literature: this work ( ), Ngo
26
27
28
■
●
▲
et al. ( ), Lou et al. ( ), Zerda et al. ( ).
used an apparatus and methodology similar to the one
described for our work.
2.6. Solubility of Collinin and Isocollinin in Pressur-
ized CO2. Solubility experiments of collinin and isocollinin in
pressurized CO2 were performed at 30 and 50 °C. The
preparation of each run was similar to that described for the
measurement of solubility of anthracene in pressurized CO2. In
detail, solutions of collinin or isocollinin were prepared in
ethanol (2.0 0.1 mg/g, each). Reference volumes of 33 and
72 μL of the collinin solution and 72 μL of the isocollinin
solution were measured eight times in a micropipette and
weighted on a scale (0.1 mg). As a result, reference amounts of
20.8 0.6 and 49.3 0.8 mg of the collinin solution and 49.6
0.9 mg of the isocollinin solution were used to add a desired
amount of each isomer into the cell.
Table 10 shows the solubility parameters and the radii of the
solubility sphere of collinin and isocollinin. As expected, there
Table 10. Solubility Parameters and Radius of the Solubility
Sphere for Collinin and Isocollinin
Hansen solubility parameters
(MPa1/2
)
isomer
δD
δP
δH
R0 (MPa1/2
)
collinin
isocollinin
15.0
15.8
6.8
6.3
11.3
11.2
12.3
12.8
Once a desired amount of solution was added into the cell,
this was heated up to 60 °C during 30 min to evaporate the
solvent. The visual cell was then sealed, the air was purged, and
a blank reading was taken. The system was cooled to a desired
were small but important differences in the solubility
parameters of the isomers. The dispersion or apolar solubility
Table 9. Solubility Scores of Collinin and Isocollinin in Different Common Organic Solvents and Solvent Blends
a
Hansen solubility parameters
score
solvent
diethyl ether
dichloromethane
acetone
methanol
tetrahydrofuran
hexane
ethyl acetate
ethanol
cyclohexane
acetonitrile
2-propanol
isooctane
toluene
acetic acid
dimethylformamide
hexane−ethanol 99.4 vol %
hexane−ethanol 96.0 vol %
δD
δP
δH
collinin
isocollinin
14.5
18.2
15.5
14.7
16.8
14.9
15.8
15.8
16.8
15.3
15.8
14.3
18.0
14.5
17.4
14.9
14.9
2.9
6.3
10.4
12.3
5.7
0.0
5.3
8.8
0.0
18.0
6.1
0.0
1.4
8.0
13.7
0.1
0.4
4.6
6.1
7.0
22.3
8.0
0.0
7.2
19.4
0.2
6.1
16.4
0.0
1
1
1
1
1
0
1
1
0
1
1
0
1
1
1
0
1
1
1
1
1
1
0
1
1
1
1
1
0
1
1
1
1
1
2.0
13.5
11.3
0.1
0.8
a
1: Completely soluble, 0: not soluble or partially soluble.
I
DOI: 10.1021/acs.jced.9b00234
J. Chem. Eng. Data XXXX, XXX, XXX−XXX

Journal of Chemical & Engineering Data
Article
parameter (δD) of collinin is 0.8 MPa1/2 smaller than that of
isocollinin, while the polar (δP) and the hydrogen-bonding
(δH) solubility parameters are 0.5 and 0.1 MPa1/2 greater than
those of isocollinin. These results further support the fact that
collinin is a little more polar than isocollinin. The radius of the
solubility sphere of isocollinin was 0.5 MPa1/2 greater than that
of collinin, as a consequence of the number of solvents and
solvent blends that completely dissolved each isomer.
collinin and isocollinin in pressurized CO2, respectively, as a
function of temperature and pressure. We note that at 30 °C
both solutes are solid and CO2 is a liquid, while at 50 °C, the
solutes are liquid and CO2 is supercritical. Solubilities in mole
fraction from 0.97 × 10−6 to 7.07 × 10−6 were obtained for
collinin, and from 2.12 × 10−6 to 6.11 × 10−6 for isocollinin.
Figure 6 shows the solubility of collinin and isocollinin as a
function of pressure, at 30 °C. The solubility of isocollinin was
higher than that of collinin at 30 °C, not only in the range of
pressure considered, but at higher pressures, as the tendency of
the data suggests. The solubility data for collinin indicate an
asymptotic growth as seen from 74 to 81 bar, while for
isocollinin, they indicate an exponential growth from 72 to 74
bar.
Figure 5 shows the behavior of the Ra/Ro ratio for the
collinin−CO2 and isocollinin−CO2 systems, at 30 and 50 °C
Figure 7 shows the solubility of collinin and isocollinin as a
function of pressure, at 50 °C. At this temperature, the
solubility of collinin is higher than that of isocollinin, and the
data show an asymptotic growth for collinin and an
exponential growth for isocollinin, as seen from 99 to 105
bar and from 105 to 113, respectively.
Although CO2 has been considered as an apolar solvent due
to its low dielectric constant and nule dipole moment, it is also
a charge-separated molecule with a significant quadrupolar
moment, and thus, it can dissolve both dipolar and nondipolar
molecular systems.28 In addition, it is well known that the
apolar character of CO2 increases as its density increases,
which occurs as pressure increases at constant temperature or
as temperature decreases at constant pressure.29 These facts
explain why, in the experimental range considered, isocollinin
is more soluble in CO2 than collinin at 30 °C, but less soluble
at 50 °C, where a decrease in CO2 density is expected.
Nevertheless, as observed in Figure 7, the solubility of
isocollinin at 50 °C also tends to be higher than that of
collinin at pressures higher than 113 bar, where a more
significant contribution of the apolar character of CO2 is
presumed, compared to the polar one.
Figure 5. Ra/Ro ratio for collinin−CO2 and isocollinin−CO2 at 30
and 50 °C.
and pressures from 1 to 1000 bar. As observed, both isomers
could be dissolved in pressurized CO2, at 30 °C from pressures
around 100 bar, and at 50 °C from pressures around 200 bar.
At these conditions, the solubility parameters of CO2 are
within the solubility sphere of both isomers (Ra/Ro ≤ 1.0). The
behavior of the Ra/Ro ratio also suggests that CO2 has a little
more affinity for collinin than for isocollinin, from 100 up to
600 bar. At pressures above 600 bar, however, this tendency is
reversed. These observations are consistent with the fact that
the apolar character of CO2 increases with its density, and as
mentioned, collinin is less apolar than isocollinin. Nonetheless,
as it is well known, the solubility parameter theory is too
simplified to quantitatively predict solubilities.
4. CONCLUSIONS
Collinin and its position-isomer isocollinin were synthesized by
a reaction route conformed by two parallel and two sequential
reactions. The syntheses were successfully performed at several
different scales in terms of the precursor amount (1, 10, 20,
and 50 g of pyrogallol). Both isomers were characterized by 1H
NMR, 13C NMR, NOESY NMR, melting temperature, melting
enthalpy, and purity as estimated from melting depression.
3.2. Solubility of Collinin and Isocollinin in Pressur-
ized Carbon Dioxide. Tables 11 and 12 list the solubilities of
a b
,
Table 11. Solubility of Collinin (y) in Pressurized CO2 as a Function of Temperature and Pressure
c
T (°C)
P (bar)
u(P)
CO2 density (mol/L)
y (106) (mole fraction)
solute−CO2 phases
solid−liquid
30
74.0
75.4
75.8
77.1
80.8
1.2
1.6
1.5
1.2
1.2
1.2
1.2
1.2
1.2
1.2
1.2
14.71
15.13
15.22
15.49
16.05
8.57
8.98
9.03
9.09
9.42
0.97
1.88
2.80
3.14
3.55
0.70
1.59
4.73
6.26
6.68
7.07
50
99.4
liquid−supercritical
100.9
101.1
101.3
102.5
104.8
10.07
a
b
c
Standard uncertainty: u(T) = 0.1 °C Relative standard uncertainties: ur(masses) = 0.05, ur(y) = 0.05. Calculated from temperature and pressure
using miniREFPROP, version 9.1.30
J
DOI: 10.1021/acs.jced.9b00234
J. Chem. Eng. Data XXXX, XXX, XXX−XXX

Journal of Chemical & Engineering Data
Article
a b
,
Table 12. Solubility of Isocollinin (y) in Pressurized CO2 as a Function of Temperature and Pressure
c
T (°C)
P (bar)
u(P)
CO2 density (mol/L)
y (106) (mole fraction)
solute−CO2 phases
30
72.2
73.0
73.1
73.6
73.7
104.6
106.4
110.6
110.8
112.9
1.2
1.3
1.2
1.2
1.2
1.2
1.5
1.2
2.1
1.2
13.59
14.27
14.32
14.56
14.60
10.01
10.51
11.56
11.61
12.06
2.12
3.10
4.02
4.94
5.92
1.47
2.80
3.82
5.08
solid−liquid
50
liquid−supercritical
6.11
a
b
c
Standard uncertainty: u(T) = 0.1 °C Relative standard uncertainties: ur(masses) = 0.05, ur(y) = 0.05 Calculated from temperature and pressure
using miniREFPROP, version 9.1.30
and 12.8 MPa1/2) for isocollinin, respectively, by using the
“good/bad” solubility test with 15 common organic solvents
and 2 solvent blends. Comparison of the isomer solubility
parameters to those of CO2 at different temperatures and
pressures allowed us to estimate whether the isomers would
dissolve in CO2 at about 100 and 200 bar, and 30 and 50 °C,
respectively.
Solubilities of collinin and isocollinin in pressurized CO2
were determined at 30 and 50 °C. In the experimental range
considered in this work, the solubility of collinin presented an
asymptotic behavior and a lower solubility than isocollinin at
30 °C, whereas at 50 °C, the asymptotic behavior was also
present, but its solubility was higher than that of isocollinin. In
both cases, the solubility of isocollinin tends to be higher than
that of collinin, at high pressures.
⧫
●
Figure 6. Solubilities of collinin ( ) and isocollinin ( ) in
pressurized CO2 at 30 °C as a function of pressure.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.jced.9b00234.
■
S
1H NMR spectra of compounds 3, 6, 10, and 11; 1D
NOESY NMR spectra of compounds 8, 9, and 11; 13C
NMR spectra of compounds 10 and 11; and DSC
thermograms of anthracene and the produced solids 3
and 8−11 (PDF).
AUTHOR INFORMATION
Corresponding Author
*E-mail: gustavo.bolanos@correounivalle.edu.co.
■
ORCID
Gustavo Bolanos: 0000-0002-1740-2888
̃
Notes
The authors declare no competing financial interest.
⧫
●
Figure 7. Solubilities of collinin ( ) and isocollinin ( ) in
pressurized CO2 at 50 °C as a function of pressure.
ACKNOWLEDGMENTS
■
The authors express their gratitude to Administrative Depart-
ment of Science, Technology and Innovation of Colombia
Colciencias, for financial support of this research (Grant 1106-
715-51462) and for a fellowship for doctoral studies granted to
Analysis by NMR allowed us to demonstrate that the precursor
of isocollinin was the most selective compound in the synthesis
pathway. Thus, overall yields of (2.6, 1.6, 1.2 and 0.6%) and
(13.9, 9.6, 6.8 and 3.7%) were obtained for collinin and
isocollinin, respectively.
The dispersion (δD), polar (δP), and hydrogen-bonding (δH)
Hansen solubility parameters (HSPs) as well as the radius of
the solubility sphere were experimentally estimated as (15.0,
6.8, 11.3, and 12.3 MPa1/2) for collinin and (15.8, 6.3, 11.2,
Camilo Pardo-Castan
̃
o.
REFERENCES
■
(1) Anet, F. A. L.; Blanks, F. R.; Hughes, G. K. The Chemical
Constituents of Australian Flindersia Species. I. Collinin-7-Geranoxy-
8-Methoxycoumarin. Aust. J. Sci. 1949, 2, 127−131.
K
DOI: 10.1021/acs.jced.9b00234
J. Chem. Eng. Data XXXX, XXX, XXX−XXX

Journal of Chemical & Engineering Data
Article
(2) Li, W.; Sun, Y. N.; Yan, X. T.; Yang, S. Y.; Kim, E. J.; Kang, H.
K.; Kim, Y. H. Coumarins and Lignans from Zanthoxylum schinifolium
and Their Anticancer Activities. J. Agric. Food Chem. 2013, 61,
10730−10740.
(3) Kim, J. S.; Lee, S. T.; Han, C. R.; Jun, D. Y.; Woo, M. H.; Kim, Y.
H. Induction of Apoptosis by Collinin from Zanthoxylum schinifolium
Is Mediated via Mitochondrial Pathway in Human Jurkat T Cells.
Process Biochem. 2013, 48, 945−954.
(4) Tanaka, T.; Yasui, Y.; Ishigamori-Suzuki, R.; Oyama, T. Citrus
Compounds Inhibit Inflammation- and Obesity-Related Colon
Carcinogenesis in Mice. Nutr. Cancer 2008, 60, 70−80.
(5) Kohno, H.; Suzuki, R.; Curini, M.; Epifano, F.; Maltese, F.;
Gonzales, S. P.; Tanaka, T. Dietary Administration with Prenylox-
ycoumarins, Auraptene and Collinin, Inhibits Colitis-Related Colon
Carcinogenesis in Mice. Int. J. Cancer 2006, 118, 2936−2942.
(6) Tanaka, T. Development of an Inflammation-Associated
Colorectal Cancer Model and Its Application for Research on
Carcinogenesis and Chemoprevention. Int. J. Inflammation 2012,
2012, 1−16.
(7) Kim, S.; Seo, H.; Al Mahmud, H.; Islam, M. I.; Lee, B.-E.; Cho,
M.-L.; Song, H.-Y. In Vitro Activity of Collinin Isolated from the
Leaves of Zanthoxylum schinifolium against Multidrug- and Extensively
Drug-Resistant Mycobacterium Tuberculosis. Phytomedicine 2018, 46,
104−110.
(22) Hansen, C. Solubility Parameters  An Introduction. Hansen
Solubility Parameters: A User’s Handbook; CRC Press Taylor & Francis
group: Boca Raton, FL, 2007; Chapter 1.
(23) Williams, L. L. Determination of Hansen Solubility Parameter
Values for Carbon Dioxide. Hansen Solubility Parameters: A User’s
Handbook; CRC Press Taylor & Francis group: Boca Raton, FL, USA,
2007; Chapter 10.
(24) Gupta, R. B.; Shim, J.-J. Solubility in Supercritical Carbon
Dioxide; CRC Press Taylor & Francis Group: Boca Raton, FL, 2007.
(25) Ngo, T. T.; Bush, D.; Eckert, C. A.; Liotta, C. L. Spectroscopic
Measurement of Solid Solubility in Supercritical Fluids. AIChE J.
2001, 47, 2566−2572.
(26) Lou, X.; Janssen, H. G.; Cramers, C. A. Temperature and
Pressure Effects on Solubility in Supercritical Carbon Dioxide and
Retention in Supercritical Fluid Chromatography. J. Chromatogr. A
1997, 785, 57−64.
(27) Zerda, T. W.; Wlegand, B.; Jonas, J. FTIR Measurements of
Solubilities of Anthracene In Supercritical CO2. J. Chem. Eng. Data
1986, 31, 274−277.
(28) Raveendran, P.; Ikushima, Y.; Wallen, S. L. Polar Attributes of
Supercritical Carbon Dioxide. Acc. Chem. Res. 2005, 38, 478−485.
(29) Turk, M. Particle Formation with Supercritical Fluids: Challenges
̈
and Limitations; Elsevier: Amsterdam, Netherlands, 2014.
(30) NIST Applied Chemicals and Materials Division. MiniREF-
PROP, version 9.1; Standard Reference Data, NIST: Gaithersburg,
MD.
(8) Santos, J.; Marquis, A.; Epifano, F.; Genovese, S.; Curini, M.;
Grenier, D. Collinin Reduces Porphyromonas Gingivalis Growth and
Collagenase Activity and Inhibits the Lipopolysaccharide-Induced
Macrophage Inflammatory Response and Osteoclast Differentiation
and Function. J. Periodontol. 2013, 84, 704−711.
(9) Epifano, F.; Genovese, S.; Marcotullio, M. C.; Curini, M.
Chemistry and Pharmacology of Collinin, Active Principle of
Zanthoxylum Spp. Mini-Rev. Med. Chem. 2008, 8, 1203−1208.
(10) Chen, I. S.; Lin, Y. C.; Tsai, I. L.; Teng, C. M.; Ko, F. N.;
Ishikawa, T.; Ishii, H. Coumarins and Anti-Platelet Aggregation
Constituents from Zanthoxylum schinifolium. Phytochemistry 1995, 39,
1091−1097.
(11) Curini, M.; Cravotto, G.; Epifano, F.; Giannone, G. Chemistry
and Biological Activity of Natural and Synthetic Prenyloxycoumarins.
Curr. Med. Chem. 2006, 13, 199−222.
(12) Choi, J. Y.; Kim, H. M.; Lee, J. M.; Choi, K.; Ku, J.; Park, K. W.;
Cho, E. J.; Lee, S. Antibacterial Activity of Zanthoxylum schinifolium.
Nat. Prod. Sci. 2012, 18, 137−140.
(13) Curini, M.; Epifano, F.; Maltese, F.; Marcotullio, M. C.;
Gonzales, S. P.; Rodriguez, J. C. Synthesis of Collinin, an Antiviral
Coumarin. Aust. J. Chem. 2003, 56, 59−60.
(14) Maes, D.; Vervisch, S.; De Kempe, N. New Synthesis of
Hydrangetin and Collinin. Org. Prep. Proced. Int. 2007, 39, 395−398.
(15) ASTM D-E928-08. Standard Test Method for Purity by
Differential Scanning Calorimetry. American Society for Testing and
Materials: West Conshohocken, PA. 2014.
(16) Radomska, M.; Radomski, R. Calorimetric Studies of Binary
Systems of 1,3,5-Trinitrobenzene with Naphthalene, Anthracene and
Carbazole. I. Phase Transitions and Heat Capacities of the Pure
Components and Charge-Transfer Complexes. Thermochim. Acta
1980, 40, 405−414.
(17) Hanessian, S.; Cooke, N. G.; DeHoff, B.; Sakito, Y. The Total
Synthesis of (+)-Ionomycin. J. Am. Chem. Soc. 1990, 112, 5276−5290.
(18) Venditti, A.; Sanna, C.; Lorenzetti, L. M.; Ballero, M.; Bianco,
A. New Coumarinyl Ethers in Daphne Oleoides Schreb Collected
from Sardinia Island. Chem. Biodiversity 2017, 14, No. e1700072.
(19) Gao, W.; Li, Q.; Chen, J.; Wang, Z.; Hua, C. Total Synthesis of
Six 3,4-Unsubstituted Coumarins. Molecules 2013, 18, 15613−15623.
(20) Van Der Klei, A.; Jong, R. L. P.; De Lugtenburg, J.; Tielens, A.
G. M. Synthesis and Spectroscopic Characterization of [1′-14C]-
Ubiquinone-2,[1′-14C]-5-Demethoxy-5-Hydroxyubiquinone-2, and
[1′-14C]-5-Demethoxyubiquinone-2. Eur. J. Org. Chem. 2002,
3015−3023.
(21) Liu, Z.; Tian, X. The Components of Cacalia tangutica. Bull.
Korean Chem. Soc. 2004, 25, 1078−1080.
L
DOI: 10.1021/acs.jced.9b00234
J. Chem. Eng. Data XXXX, XXX, XXX−XXX