Symmetrical Fatty Dialkyl Carbonates as Potential Green Phase Change Materials: Synthesis and Characterisation

Article abstract of DOI:10.1134/S1070363219070223

Symmetrical fatty dialkyl carbonates make a group of chemicals that has great potential for applications as phase change materials (PCMs). In this study, various parameters affecting synthesis of these carbonates are studied on laboratory scale, which includes: reactants ratio, reaction temperature and time, and the amount of catalyst used. Structures of the desired products are identified using FTIR and NMR spectroscopy. Differential scanning calorimetry (DSC) analysis is used for determining potential of the synthesized symmetrical fatty dialkyl carbonates as PCMs for energy-saving applications.

Full text of DOI:10.1134/S1070363219070223

ISSN 1070-3632, Russian Journal of General Chemistry, 2019, Vol. 89, No. 7, pp. 1513–1518. © Pleiades Publishing, Ltd., 2019.
Symmetrical Fatty Dialkyl Carbonates as Potential Green Phase
Change Materials: Synthesis and Characterisation
N. N. Tunga*, T. T. Hunga, N. Q. Trunga, and L. T. Giangb
a Center for Research and Technology Transfer, Vietnam Academy of Science and Technology, Hanoi, Vietnam
*e-mail: tungnguyen.vast@gmail.com
b Vietnam Academy of Science and Technology, Hanoi, Vietnam
Received April 3, 2019; revised July 3, 2019; accepted July 4, 2019
Abstract—Symmetrical fatty dialkyl carbonates make a group of chemicals that has great potential for applica-
tions as phase change materials (PCMs). In this study, various parameters affecting synthesis of these carbonates
are studied on laboratory scale, which includes: reactants ratio, reaction temperature and time, and the amount of
catalyst used. Structures of the desired products are identified using FTIR and NMR spectroscopy. Differential
scanning calorimetry (DSC) analysis is used for determining potential of the synthesized symmetrical fatty dialkyl
carbonates as PCMs for energy-saving applications.
Keywords: symmetrical dialkyl carbonate, dialkyl carbonate, transesterification, phase change material
DOI: 10.1134/S1070363219070223
INTRODUCTION
it can have negative impact on the latent heat values of
the materials [14].
Phase change materials (PCMs) make a unique group
of materials that can store and subsequently release large
amount of energy over a defined narrow temperature
range as they undergo phase transition. Currently, solid-
liquid PCM systems are the most practical and economi-
cal choice for energy storage, since volume change caused
by such phase transition process is negligible [1–4].
In contrast, carbonic esters of fatty alcohols are quite
stable under normal conditions and non-corrosive. Even
upon degradation, such compounds release only CO2 and
relatively non-corrosive respective fatty alcohols. For
these reasons, carbonic esters of fatty alcohols demon-
strate high potential as PCMs [15, 16].
Shaikh and Sivaram [17] studied the synthesis of
several short-chain dialkyl carbonates by carbonate in-
terchange reaction of respective alcohols with dimethyl
carbonate. Azeotropic mixture of methanol and other
volatile compounds was easily removed. Synthesis of
several long-chained dialkyl carbonates by carbonate
interchange reaction of respective alcohols with diethyl
carbonate [18] required at least 20 h for reaching equi-
librium, with maximum symmetrical fatty carbonate
yield 77–80%.
PCMs are used for energy saving [5], heat insulation
[6], solar-energy harvesting, and much more [1, 7–10].
Despite being readily-available and cheap, hydrated
inorganic salts are utilized as PCMs mostly in high tem-
perature range, owning to (original edit is highly mislead-
ing) their significant corrosive effect and degradation over
a number of usage cycles [4]. Currently, paraffin waxes
that are widely utilized do not exhibit corrosive property,
though their energy storage capability per volume unit is
somewhat lower than that of hydrated inorganic salts [1].
In this study, symmetrical fatty dialkyl carbonates
were synthesized by transesterification of dimethyl
carbonate and the respective fatty alcohols. Dimethyl
carbonate is considered as a “green reagent” [19], while
fatty alcohols can be manufactured from readily-available
vegetable oils. We have studied the factors that influence
upon the synthesis of symmetrical 1-octadecyl carbon-
ate, including catalyst, reactants ratio, reaction time and
temperature, and determined its potential as a PCM.
Among promising materials, fatty compounds ex-
tracted from animal fat and vegetable oils proved to
hold great potentials, since they are readily-available
from renewable sources, especially as by-products of
agricultural industry [11–13]. However, most of these
compounds exhibit corrosive effect to some degree, either
directly like fatty carboxylic acids, or through degrada-
tion like fatty carboxylic esters. Such disadvantages can
be minimized by micro-encapsulating technique, though
1513

1514
TUNG et al.
EXPERIMENTAL
held isothermally for 1 min at 0°C and then heated at a
heating rate of 3°C/min to the appropriate temperature
for obtaining a complete melting profile for the samples.
The samples were then cooled down to 0°C at the same
rate for obtaining a complete solidification profile. Heat
flow (mW/g) vs. temperature (°C) curve accumulated by
each scan was analysed using TAUniversalAnalysis 2000
software. Latent heat was obtained by numerical integra-
tion of the area over which a thermal event occurred.
DSC curves indicated melting (Tm) and solidification
(Tc) points. Exothermic signals of DSC are reported in
the upward direction.
Most of chemicals used were of analytical grade
and used without further purification. Octadecan-1-ol,
hexadecan-1-ol, tetradecan-1-ol and dimethyl carbonate
were purchased from Shanghai Macklin Biochemical Co.
Ltd. (China). Dibutyltin oxide was purchased fromAcros
Organics (Belgium). Methylene chloride and acetone
were purchased from Merck Chemicals Ltd. (Germany).
Some experiments were carried out with identical
chemicals as stated above, but of industrial grade. Such
chemicals were acquired from various small vendors in
Vietnam, and their purity was in the range 85 – 95% as
determined by GC/MS analysis. These chemicals were
used without further purification.
GC/MS analysis was carried out on anAgilent 7890B
GC gas chromatograph system equipped with an FID
and an auto sampler/injector, coupled with a 5977A
MSD mass spectrometer. GC was conducted under the
conditions: column flow 1.0 mL/min with helium head
pressure of 15 psi (776 torr), split ratio of 75 : 1, injector
temperature set at 250°C, transfer line temperature set
at 280°C. The oven temperature ramp programs were
utilized: 140°C for 2 min, ramp from 140 to 280°C at
the rate of 20°C per min, then held for 30 min at 280°C.
Electron ionization mass spectrometry was performed at
70 eV, mass range of 50–550 amu. Injector and detector
temperatures were set at 280°C.
Synthesis of symmetrical fatty dialkyl carbonates.
A mixture of octadecan-1-ol (27.0 g, 0.1 mol) with dibu-
tyltin oxide (0.25 g, 0.001 mol) was heated to 80–85°C
under a slight positive nitrogen pressure. Upon melting
the alcohol, dimethyl carbonate (10.0 mL, 0.119 mol) was
added to it, and the mixture was stirred well for ca 30
min. The temperature was raised to 130°C and refluxing
lasted for 5 h. Upon completion, the reaction mixture
was extracted by methylene chloride, then concentrated
in vacuo on a rotary evaporator. The crude product was
stored under vacuum overnight at room temperature be-
fore being weighted.Asmall amount of the crude product
was dissolved in warm acetone. Upon cooling to room
temperature, white crystals were isolated by vacuum
filtration. The purified product was stored under vacuum
overnight at room temperature.
RESULTS AND DISCUSSION
GC/MS analysis was not efficient in determination of
symmetrical fatty dialkyl carbonates because of fatty dial-
kyl carbonates decomposition under GC/MS conditions.
Identification of purified products was successfully
conducted using combined information accumulated
by FTIR and 1H NMR spectroscopy. The representative
FTIR spectrum of the purified product synthesized by
the reaction of octadecan-1-ol with dimethyl carbonate
is presented in Fig. 1.
FTIR spectra were recorded on a Thermo Scientific
Nicolet iS10 spectrometer using KBr discs.
1H NMR spectra were measured on a BrukerAdvance
500 spectrometer using CDCl3 as a solvent.
Thermal analysis, including melting, crystallization
and latent heat, was carried out on a NETZSCH DSC 204
F1 Phoenix calorimeter. A Mettler AE100 balance was
used for weighing all DSC samples. Freshly recrystallized
samples (5.0–10.0 mg) were weighed in an aluminium
DSC pan and sealed. The sealed DSC samples were
stored under a constant N2 stream flowing at 50 mL/min.
The samples were first cycled through the temperature
range of interest above and below the sample’s melting
and freezing points to distribute the materials evenly
throughout the DSC pans and eliminate temperature his-
tory effects. The samples were then cooled down to 0°C
for complete sample solidification. The samples were
The distinctive “tongue shape” broad band at 3200–
3570 cm–1 and another one at 1000–1300 cm–1 were
attributed to the OH and C–O groups of small amount of
unreacted octadecan-1-ol. The band at 1730–1750 cm–1
was attributed to the carbonyl groups and it indicated the
presence of at least one of compounds: carboxylic, ester,
ketone, or aldehyde of symmetrical 1-octadecyl carbon-
ate and 1-octadecyl methyl carbonate. Two bands in the
range of 2860–2970 cm–1 corresponded to C–H bonds
of the alkyl groups, but no band at ca 3000 cm–1 typical
for the C–H bonds in alkene groups was recorded. This
helped to single out the compounds with C=C bonds in
the spectrum of the representative purified product.
RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 89 No. 7 2019

SYMMETRICAL FATTY DIALKYL CARBONATES AS POTENTIAL GREEN PHASE
1515
65
55
45
35
25
15
5
–5
400
300
200
100
Wavenumber, cm–1
Fig. 1. Representative FTIR spectrum of the product of
Fig. 2. Effect of catalytic (1) DBTO and (2) DBTDL ratio to
octadecan-1-ol reaction with dimethyl carbonate.
octadecan-1-ol used on rate of the reaction.
1H NMR spectrum, δ, ppm: 0.88 t (3H), 1.26–1.36 m
(30H), 1.55–1.69 m (2H), 4.15 t (2H). This spectrum was
in the line with the symmetrical structure of 1-octadecyl
carbonate. Similar results were obtained in another study
[18] in which the compound was synthesized by a differ-
ent method. The triplet at 4.15 ppm corresponded to two
α-methylene hydrogens in –CH2–CH2–O–CO– moiety,
and the multiplet at 1.55–1.69 ppm corresponded to two
β-methylene hydrogens in the same fragment.Absence of
other signals, particularly in the range of 2.00–2.50 ppm
typical for α-methylene hydrogens of the carboxylate
groups, ruled out the probability of any other chemical
structures but symmetrical 1-octadecyl carbonate.
Thus, the main alcohol precursor used in most of
the following experiments was octadecan-1-ol, since its
respective symmetrical dialkyl carbonate exhibited the
highest melting point, offering more precise recrystal-
lization results.
Choice of catalyst. Based on the earlier studies
[17,18], were n-butyltin oxide (DBTO) and n-butyltin di-
laurate (DBTDL) were singled out as catalysts of choice.
Preliminary experiments were conducted for determining
the evolution of reaction composition over time, and the
results were quite similar to those reported earlier [18]:
1-octadecyl methyl carbonate was formed alongside with
symmetrical 1-octadecyl carbonate. 1-Octadecyl methyl
carbonate disproportionated further into the symmetrical
1-octadecyl carbonate and the original octadecan-1-ol.
The optimal ratio of each catalyst was determined for the
experiments carried out under the following conditions:
molar ratio octadecan-1-ol : DMC = 0.8 : 1.0, temperature
130°C, reaction time 4 h, and molar percentage of catalyst
to octadecan-1-ol was varied from 0.1 to 1.0 % (Fig. 2).
Composition of the crude product was determined us-
ing the above 1H NMR data for the pure compound that
proved the former one to be more complex, and besides
the symmetrical 1-octadecyl carbonate the spectrum in-
dicated the presence of another compound, most likely
1-octadecylmethyl carbonate.
According to the spectral data, composition of the
particular crude product was quantified in molar fractions
as following: 18.65% octadecan-1-ol, 25.91% 1-octadecyl
methyl carbonate, and 55.44% symmetrical 1-octadecyl
carbonate. The overall yield of symmetrical 1-octadecyl
carbonate based on the initial amount of octadecan-1-ol
used as a precursor was thus determined to be 71.33%.
According to Fig. 2, for DBTO, when the molar ratio
of a catalyst was increased from 0.4 to 0.6% of octadecan-
1-ol, the yield of symmetrical 1-octadecyl carbonate
also increased from 48.40 to 65.90%. However, when
the amount of DBTO catalyst was beyond that point, no
meaningful changes in the yield of symmetrical 1-octa-
decyl carbonate were observed. The same phenomenon
was also observed for DBTDL. The yield of symmetrical
1-octadecyl carbonate increased from 40.15 to 64.19%
when the molar amount of catalyst increased from 0.1 to
0.2% of octadecan-1-ol.
A small amount of the above representative crude
product was recrystallized from acetone which dissolved
well unreacted octadecan-1-ol and 1-octadecyl methyl
carbonate. The maximum yield of recrystallized sym-
metrical 1-octadecyl carbonate was roughly 66%.
RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 89 No. 7 2019

1516
TUNG et al.
80
60
40
20
0
0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75
Reactants ratio (mol DMC/mol octadecan-1-ol)
Fig. 4. Effect of reagents ratio on the rate of reaction.
demonstrates only the results within the temperature
range 130–180°C.
The yield of symmetrical 1-octadecyl carbonate rose
from 66.61% at 130°C (yield 71.33 % after 5 h) to 70.63%
at 180°C (yield 71.43 % after ca 4 h).
At 160°C and higher, crude products became some-
what less transparent in the liquid form, which indicated
some side reactions occurring. Accordingly, the optimal
reaction temperature was determined to be 130°C.
Fig. 3. Effect of temperature on the reaction rate.
The optimum reaction rate was approached under the
above conditions, and molar ratio DMC : octadecan-1-ol
was varied from 0.25 : 1.00 to 1.50 : 1.00 (Fig. 4).
So, the optimal molar ratio between the catalyst and
octadecan-1-ol was determined to be 0.8% for DBTO
(yield 71.33% after 5 h) and 0.6% for DBTDL (yield
71.74% after 4.5 h).
According to the accumulated data (Fig. 4), the maxi-
mum yield of 71.33% was achieved after 5 h with molar
ratio DMC : octadecan-1-ol = 1.25 : 1.00. These results
could be explained by the volatile nature of DMC and
its bp ca 90°C, which was much lower than the reaction
temperature of 130°C.
However, DBTO was a white solid which completely
dissolved in the reaction mixture only when transester-
ifica-tion between fatty alcohol and dimethyl carbonate
started, while DBTDL was a clear liquid that could be
readily dispersed in dimethyl carbonate at room tempera-
ture. This made DBTO to be an excellent visual indicator
for determining the reaction starting point.
Effect of reacting alcohols. The transesterification re-
action of DMC was also carried out with tetradecan-1-ol.
The results achieved in those experiments were found to
be quite similar to the results of octadecan-1-ol, with the
optimal reaction parameters determined as following:
molar ratio between DMC and fatty alcohol 1.25 : 1.00,
molar percentage of DBTO to alcohol 0.8%, temperature
130°C, and reaction time ca 4 h for hexadecan-1-ol (yield
70.47 %), or ca 3.5 h for tetradecan-1-ol (yield 71.17 %).
Therefore, DBTO was chosen as the catalyst for the
following experiments using molar percentage of the
catalyst to octadecan-1-ol to be 0.8%.
Effect of reaction temperature. The optimal reac-
tion temperature was tested within the range 110–180°C
under the following conditions: molar ratio octadecan-
1-ol : DMC = 0.8 : 1.0, reaction time 4 h, and molar
percentage of DBTO to octadecan-1-ol 0.8 %. In the case
reaction temperature was higher than 130°C, the process
was carried at 130°C for the first 2 h, then temperature
was increased to the pre-determined level for the final
2 h (Fig. 3).
The reaction of industrial grade chemicals reached
equilibrium roughly within same time as analytical grade
chemicals. The final yield of symmetrical fatty alkyl car-
bonates was lower (62–66%) before purification.
Comparing to other publications [17, 18], in the
current study the reaction time needed for reaching an
equilibrium was significantly shorter, ca 5 h. The reac-
tion setups were very simple, and did not require sophis-
ticated laboratory equipment. The reagents used were
In most cases transesterification reaction of fatty alco-
hol with dimethyl carbonate did not occur below 130°C,
or proceeded at an impractical rate. Therefore, Fig. 3
RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 89 No. 7 2019

SYMMETRICAL FATTY DIALKYL CARBONATES AS POTENTIAL GREEN PHASE
1517
Fig. 5. DSC results of (a) a representative crude product and (b) a representative purified product.
recognized as environmentally-friendly and renewable
to some extent.
product, with wider phase change temperature range. This
effect could be attributed to 1-octadecyl methyl carbon-
ate, which theoretically had the lowest phase transition
point comparing to symmetrical 1-octadecyl carbonate
and octadecan-1-ol. Fortunately, this compound could
be successfully extracted and recrystallized. The repre-
sentative purified product possessed melting and freezing
points very close to those presented in [18] (see the table).
Comparison of the current study data with those of
[18], indicated relatively low yield achieved in this work.
Another disadvantage was that this setup only worked in
a rather large vessel, otherwise the reaction rate would
become much lower due to excessive dimethyl carbonate
condensed.
The presence of octadecan-1-ol impurity was clearly
traced in DSC curve for solidification transition of the rep-
resentative purified product. Without much impurity, the
representative purified product exhibited much narrower
phase change temperature range, and lower sub-cooling
effect. Even though still not ideal, this material can be
readily used in various real-live applications that do not
require short storing—releasing energy cycle, especially
most domestic applications.
Differential scanning calorimetry. Both crude and
purified products synthesised in the current study by
transesterification of octadecan-1-ol were subjected to
DSC analysis (Fig. 5).
Both products demonstrated good latent heat storage
capacities (Fig. 5) above 300 mJ/mg, however the sub-
cooling effect of these compounds was still evident. This
problem, combined with a wide phase change temperature
range, may create negative effect on their energy stor-
ing–releasing cycle, limiting the possibility of application
under real-live conditions. In particular, the representa-
tive crude product exhibited lower melting and freezing
temperature comparing to the representative purified
CONCLUSIONS
In summary, the optimal parameters for transesterifica-
tion reaction of dimethyl carbonate and fatty alcohols
leading to their respective symmetrical fatty alkyl carbon-
DSC data for symmetrical fatty dialkyl carbonates
Melting
Solidification
Carbonate
Tm, °C
36.50
43.40
55.90
latent heat, mJ/mg
Tc, °C
29.00
36.50
45.10
latent heat, mJ/mg
Symmetrical 1-tetradecyl carbonate
Symmetrical 1-hexadecyl carbonate
Symmetrical 1-octadecyl carbonate
330
313
326
327
312
324
RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 89 No. 7 2019

1518
TUNG et al.
Therm. Energ. Storage ECOSTOCK 2006 “Research on
Thermal Storage using Rock Wool PCM Ceiling Board”,
New Jersey, 2006, p. 59.
ates are determined as follows: temperature 130°C, molar
ratio DMC : alcohol = 1.25 : 1.00, molar percentage of
DBTO to alcohol 0.8%, and reaction time 3.5–5 h which
is much shorter than in the developed earlier methods.
7. Prieto, C., Cooper, P., Fernández, A.I., and Cabeza, L.F.,
Renew. Sust. Energ. Rev., 2016, vol. 60, p. 909. doi
10.1016/j.rser.2015.12.364
The desired high purity symmetrical fatty alkyl car-
bonates are isolated with good yields, and their structures
are supported by FTIR and 1H NMR spectra. According
to DSC the products are characterised by heat of fusion
values ≥ 300 mJ/mg. The developed method allows to
synthesize phase change materials from “green reagents”
and utilize the PCMs in domestic applications.
8. Wang, Y., Wang, S., Wang, J., and Yang, R., Energ.
Buildings, 2014, vol. 77, p. 11. doi 10.1016/j.en-
build.2014.03.036
9. Zhang, Y. P., Lin, K. P., Yang, R., Di, H. F., and Jiang, Y.,
Energ. Buildings, 2006, vol. 38, p. 1262. doi 10.1016/j.
enbuild.2006.02.009
FUNDING
10. Zhou, D., Zhao, C.Y., and Tian, Y., Appl. Energ., 2012,
vol. 92, p. 593. doi 10.1016/j.apenergy.2011.08.025
11. Yuan, Y., Zhang, N., Tao, W., Cao, X., and He, Y., Renew.
Sust. Energ. Rev., 2014, vol. 29, p. 482. doi 10.1016/j.
rser.2013.08.107
We are grateful to the Vietnam Academy of Sci-
ence and Technology (VAST) for financial support
under scientific research project: CT0000.01/18-19 and
TĐPCCC.05/18-20.
12. Kant, K., Shukla, A., and Sharma, A., Energy Reports,
2016, vol. 2, p. 274. doi 10.1016/j.egyr.2016.10.002
13. Rozanna, D., Chuah, T.G., Salmiah, A., Choong, T.S.Y.,
and Sa’ari, M., Int. J. Green Energ., 2005, vol. 1, p. 495.
doi 10.1081/GE-200038722
CONFLICT OF INTEREST
No conflict of interest was declared by the authors.
REFERENCES
14. Németh, B., Németh, Á.S., Ujhidy, A., Tóth, J., Trif, L.,
Gyenis, J., and Feczkó, T., Sol. Energ., 2018, vol. 170,
p. 314. doi 10.1016/j.solener.2018.05.066
15. Kenar, J. A., Sol. Energ. Mat. Sol. C., 2010, vol. 94,
p. 1697. doi 10.1016/j.solmat.2010.05.031
1. Baetens, R., Jelle, B.P., and Gustavsen, A., Energ.
Buildings, 2010, vol. 42, p. 1361. doi 10.1016/j.en-
build.2010.03.026
2. Dincer, I. and Rosen, M., Thermal Energy Storage (TES)
Methods, in book Thermal Energy Storage: Systems and
Applications, New York: John Wiley & Sons, 2010, 2 ed.,
p. 83. doi 10.1002/9780470970751.ch3
16. Kenar, J.A., Eur. J. Lipid Sci. Tech., 2012, vol. 114, p. 63.
doi 10.1002/ejlt.201100043
17. Shaikh, A.A.G. and Sivaram, S., Ind. Eng. Chem. Res.,
1992, vol. 31, p. 1167. doi 10.1021/ie00004a028
18. Kenar, J.A., Knothe, G., and Copes, A.L., J. Am. Oil
Chem. Soc., 2004, vol. 81, p. 285. doi 10.1007/s11746-
004-0897-4
3. Fleischer, A.S., Thermal Energy Storage Using
Phase Change Materials, Cham: Springer, 2015. doi
10.1007/978-3-319-20922-7
4. Kośny, J., PCM-Enhanced Building Components, Man-
chester: Springer, 2015. doi 10.1007/978-3-319-14286-9
19. Kreutzberger, C.B., Chloroformates and Carbonates, in
Kirk-Othmer Encyclopedia of Chemical Technology, New
York: John Wiley & Sons, 2001, vol. 6, p. 290. doi 10.1002/
0471238961.0301180204011312.a01.pub2
5. Cheng, W., Xie, B., Zhang, R., Xu, Z., and Xia, Y.,
Appl. Energ., 2015, vol. 144, p. 10. doi 10.1016/j.apen-
ergy.2015.01.055
6. Kondo, T. and Iwamoto, S., Proc. of the X Int. Conf.
RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 89 No. 7 2019