Accepted Manuscript

Critical properties, heat capacities, and thermal diffusivities of four saturated

triglycerides

Nataliya S. Bogatishcheva, Mars Z. Faizullin, Alexander P. Popov, Eugene D.

Nikitin

PII: S0021-9614(17)30229-X
DOI: http://dx.doi.org/10.1016/j.jct.2017.07.006
Reference: YJCHT 5124

To appear in: J. Chem. Thermodynamics

Received Date: 21 March 2017

Revised Date: 14 June 2017
Accepted Date: 1 July 2017

Please cite this article as: N.S. Bogatishcheva, M.Z. Faizullin, A.P. Popov, E.D. Nikitin, Critical properties, heat
capacities, and thermal diffusivities of four saturated triglycerides, J. Chem. Thermodynamics (2017), doi: http://
dx.doi.org/10.1016/j.jct.2017.07.006

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Critical properties, heat capacities, and thermal diffusivities of four saturated
triglycerides

Nataliya S. Bogatishcheva, Mars Z. Faizullin, Alexander P. Popov and Eugene D. Nikitin*

Institute of Thermal Physics, Ural Branch of the Russian Academy of Sciences, Amundsen Street, 107a,
620016 Ekaterinburg, Russia

Abstract
The critical temperatures, pressures, heat capacities, and thermal diffusivities of saturated triglycerides
C3H5[CnH2n-1O2]3 with n = 8, 10, 12, 14 have been measured. The experimental critical constants of
triglycerides have been compared with those calculated using the group contribution methods of
Constantinou and Gani, Marrero and Gani, Hukkerikar and co-workers, and Cunico and co-workers.
The heat capacities and thermal diffusivities have been measured in the temperature range from about
the melting point of a triglyceride to 410-425 K. The temperature dependencies of the heat capacities
and thermal diffusivities have been approximated by a third-order and a first-order polynomial,
respectively. The coefficients of these polynomials have been given. Using the experimental data, the
thermal conductivities of triglycerides have been calculated.

Keywords: Critical properties; Heat Capacity; Thermal Diffusivity; Measurements; Triglycerides.

1. Introduction

Biodiesel is well known to be produced through the esterification reaction from vegetable oils and
animal fats which mainly consist of triglycerides. Information about thermophysical properties of
triglycerides is required for optimized biodiesel fuel production. The list of such properties includes but
is not limited to the critical properties, vapor pressures, thermal conductivities, and heat capacities [1,2].
In this paper, we present the results of the measurement of the critical temperatures, critical pressures,
heat capacities and thermal diffusivities of four saturated triglycerides: glyceryl trioctanoate (tricaprylin,
CASRN 538-23-8), glyceryl tridecanoate (tricaprin, CASRN 621-71-6), glyceryl tridodecanoate
____________________________________________________
* Corresponding author.
E-mail addresses: e-nikitin@mail.ru (E.D. Nikitin), bogatishcheva@mail.ru (N.S. Bogatishcheva),
faizullin@itp.uran.ru (M.Z. Faizullin), a-popov@itp.uran.ru (A.P. Popov).

1
(trilaurin, CASRN 538-24-9), and glyceryl tritetradecanoate (trimiristin, CASRN 555-45-3).
Usually vegetable oils contain saturated triglycerides C3H5[CnH2n-1O2]3 with even n from 6 to 24 [3].
However, for completeness, we will discuss all the saturated triglycerides. The critical properties of
triglycerides have probably never been measured. Triglycerides are not very stable compounds.
Measuring the heat capacity of trilaurin, Morad et al. [4] found out that trilaurin began to degrade at
423 K in the air atmosphere. Using nitrogen as a purge gas, they could measure the heat capacity up to
523 K; however, there was a definite change in the slope of the dependence of the heat capacity on
temperature at about 443 K, which indicated some changes in the sample above this temperature. The
critical temperatures of triglycerides measured by us lie in the range 836 to 925 K; it may be assumed
that these triglycerides decompose at their critical points, so we used in this work the pulse-heating
method applicable to thermally unstable compounds [5-9].
We failed to find in the literature any experimental data on thermal conductivity or thermal diffusivity
of pure liquid triglycerides; these thermal properties have been measured only for vegetable oils [10-
15]. Experimental data on the heat capacity of a great number of organic and inorganic liquids including
saturated triglycerides were collected and critically evaluated in the monograph by Zábranský et al. [16]
and the subsequent updates [17,18]. An analysis of the data published shows that heat capacities are
investigated well enough for triglycerides with n = 12, 14, 16, and 18 and there are only scattered data
for the rest of saturated triglycerides.

2. Experimental

2.1 Materials

The samples of triglycerides were purchased from Alfa Aesar and used without any additional
purification. The Chemical Abstract Service Registry Numbers (CASRN’s) of the compounds studied
and the purities of the samples according to the supplier’s certificates of analysis are given in Table 1.

2.2 Measurements of the Critical Properties by the Pulse-Heating Method

The pulse-heating method used for the measurement of the critical temperatures and pressures of
saturated triglycerides has been discussed in detail in previous papers [5-9]. It was developed by us in
early nineties and since then the critical properties of about 250 compounds have been measured. The
method is based on the phenomenon of liquid superheat and consists in measuring the pressure
dependence of the temperature of the attainable superheat (spontaneous boiling-up) [19,20]. When the
2
pressure in the liquid tends to the critical pressure, the temperature of the attainable superheat
approaches the critical temperature. The temperature of the attainable superheat at a given pressure is

measured with the help of a wire probe, 2 × 10−3 cm in diameter and 1-3 cm in length. The probe serves
simultaneously as a heater and a resistance thermometer. The probe is heated by electric current pulses
in such a way that the probe temperature reaches the temperature of spontaneous boiling-up by the end
of a pulse. The time from the beginning of a pulse to the moment of boiling-up in various experiments is
from 0.03 to 1.0 ms. At the moment of boiling-up, a probe temperature perturbation arises because the
conditions of heat transfer from the probe to the liquid change sharply. The probe temperature is
determined from its resistance at this moment. The probe temperature perturbation may be positive or
negative. The temperature perturbation is negative at near-critical pressures for comparatively low-
molar-mass compounds (nonpolymers). When the pressure approaches the critical pressure, the
properties of the liquid and vapour phases come closer and the amplitude of the temperature
perturbation drops. The pressure in the chamber with a substance under study increases step by step
until the negative temperature perturbation decreases to the level of the apparatus sensitivity

( 1× 10−3 K). This pressure and the temperature of boiling-up at this pressure are taken as the measured

values of the critical pressure pcm and the critical temperature Tcm .

A distinguishing characteristic of the pulse-heating technique is that the values of pcm and Tcm are

slightly lesser than the true critical parameters. For stable compounds, the true critical pressure pc and

temperature Tc are calculated using the following equations:

pc = pcm / πo , Tc = Tcm / τ0 , (1)

where 1 / π0 and 1 / τ0 are correction factors [5,9,21]. To calculate the correction factors the acentric

factor or another similarity parameter of a compound under study is needed. We prefer to use the
parameter suggested by Filippov [22]:

pvp ( Tr = 0.625 )
A = 100 ,
(2)
pc

where pvp is the vapour pressure, Tr = T / Tc is the reduced temperature. The sense of the Filippov

parameter is close to that of the well-known Pitzer acentric factor [23], but to determine the Filippov

3
parameter the vapour pressure at Tr = 0.625 is required, while in the acentric factor, this temperature is

equal to Tr = 0.7 . The difference between these reduced temperatures seems insignificant; however,
there are more experimental data on vapour pressure at lower temperatures.
The Filippov parameters and the critical properties were calculated by an iteration method. The

measured values of the critical parameters pcm and Tcm were used as the critical constants for the first

iteration. The vapor pressure at a reduced temperature Tr = 0.625 was estimated using an equation

suggested by Filippov [22]:

P* T *  T *  T *
log10 = 3.9726 log10 + − 1 0.3252 + 0.40529  (3)
pvp T  T  T 

Eq.(3) allows calculating the vapor pressure pvp at any temperature T between the triple and critical

points for a given compound. Here P * and T * are the pressure and temperature characteristic of the
compound. Filippov calls P * and T * the «pseudocritical» parameters of a compound because, if these
parameters are used instead of the critical constants to determine the reduced quantities, the reduced
vapour pressure pvp / P* is a universal function of the reduced temperature T / T * . Filippov obtained

this universal function in the form of Eq.(3). The values of P * and T * can be estimated using two
values of the vapor pressure of the compound at two temperatures. We calculated P * and T * from

pcm , Tcm , and the normal boiling temperatures revised by Wallek et al. [24]. Then the Filippov
parameters, the correction factors and the critical constants were calculated; these values were used in
the second iteration. Two iterations were enough because the correction factors depend only slightly on
the variations of Filippov’s parameter.
The values of the correction factors are given in Tables 4 and 5. These tables show that using the
correction factors increases the critical pressures and temperatures of triglycerides by about 10% and
0.7%, respectively. The correction factors mainly depend on a similarity parameter, say the acentric
factor. For low-molar-mass compounds with comparatively small acentric factors the correction factors
are slightly larger than 1. For instance, the acentric factor of n-hexane is ω = 0.229 and the correction
factors are 1 / π0 = 1.02 and 1 / τ0 = 1.003 . For high-molar-mass compounds like triglycerides the

correction factors are greater. For example, according to our estimation, the acentric factor of glyceride
tritetradecanoate is about 1.3 and 1 / π0 = 1.10 , 1 / τ0 = 1.007 .

4
 The decomposition of an unstable compound in the course of pulse-heating experiments may cause the
dependence of the apparent critical pressure and temperature on the duration of a heating pulse. The
examples of the dependence are given in our previous papers [25-28]. In the experiments with
triglycerides, the durations of heating pulses were 0.035, 0.06, 0.11, 0.22, 0.46, and 0.85 ms; the probe
lengths were 1, 2, and 3 cm. Two or three samples of each compound were employed. As noted above,
the triglycerides degrade at the critical point but we found no obvious dependence of the apparent
critical parameters on the durations of the heating pulses. Probably, the amount of decomposition
products formed during a heating pulse was so small that it did not change the critical properties.

Therefore, the measured values of the critical temperature Tcm and pressure pcm were determined by
averaging over all the heating times, lengths, and samples.
The uncertainties of the critical properties measured by the pulse-heating method are discussed in
detail elsewhere [8,9,29]. It has been shown that for low-molar-mass compounds with low acentric
factors ( ω < 1 ) and without strong hydrogen bonding the uncertainties are δpc = 0.03 pc , δTc = 0.01Tc ,

where Tc is the absolute temperature. For high-molar mass compounds with ω ≥ 1 , including

triglycerides studied, the uncertainties are equal to δ pc = 0.04 pc and δ Tc = 0.015Tc . Here the

uncertainties are the combined expanded uncertainty at the 95% level of confidence.

2.3. Measurements of Heat Capacity

The isobaric heat capacity was measured by differential scanning calorimeter DSC 204 F1 Phoenix,
manufactured by Netzsch. The samples of triglycerides with mass of 15 to 20 mg were welded, gas
tight, into aluminum crucibles. The masses of the samples were measured with the uncertainty of
0.1 mg. The measurements were carried out in the temperature range from the melting point of a

triglyceride to 423 K at the heating rate of 2 K ⋅ min −1 ; argon was used as a purge gas. A sample of
sapphire with mass of 25 mg was employed as reference. The DSC was calibrated on temperature with
standard of cyclohexane, mercury, bismuth, tin, and zinc. The uncertainty of the temperature
measurements was 0.1 K. The performance of the calorimeter was checked with ethylene glycol; in the
temperature range of 300-400 K no deviations from the recommended values [30] larger than 2% were
found. The procedure of the measurements of heat capacity using DSC was discussed in detail by
Hohne et al. [31]. In fact, the heat capacity measured in the way described above is the saturation heat
capacity. However, the vapor pressures of the triglycerides studied are very low at the temperatures of
our experiments. For instance, the vapor pressure of glyceride trioctanoate at 425 K is about 1Pa [32] so
that the heat capacity obtained in our experiments may be considered as isobaric heat capacity at
5
atmospheric pressure. Atmospheric pressure was in the range (730 to 750) mm Hg during our
experiments. Although in Table 3 we give the heat capacities at pressure p = 0.1 MPa, the changes of
the atmospheric pressure give rise to a standard uncertainty of 4 kPa for the pressure. Based on the
results of the calibration of the DSC and the measurement of the heat capacity of ethylene glycol, we
estimate the relative combined expanded uncertainty at 95% of confidence to be 2%. Two or three
samples of each triglyceride were used. Three passes with respect to temperature were made for each
sample. One pass gave about 140 experimental data points which were about 0.5 K apart. The
experimental points for a definite compound were approximated by a third-order polynomial:

C p,m = a0 + a1T + a2T 2 + a3T 3 (4)

The values of the constants of these polynomials, together with the temperature ranges in which they are
valid, are given in Table 2. The heat capacities of triglycerides calculated with the help of these
polynomials with an increment of 5 K are given in Table 3.

2.4. Measurements of Thermal Diffusivity

The thermal diffusivity was measured by the laser flash method using a LFA 457 MicroFlash system
manufactured by Netzsch. The description of the system is presented on the Netzsch site [33]. Some
details of the measurement procedure were given by Blumm et al. [34]. In our experiments, a liquid
under study filled a container made of stainless steel; the thickness of a liquid layer was 0.5 mm. To
check our apparatus, we measured the thermal diffusivities of some liquids: n-hexane, n-tetradecane,
cyclohexane, toluene, and water. For all the liquids studied, the difference between the data obtained by
us and the values calculated using literature data [35] were less than 5%. This approximately
corresponds to the accuracy of the apparatus LFA 457 MicroFlash which is declared by Netzsch. So for
the thermal diffusivity, we estimate the combined expanded uncertainty at 95% of confidence to be 5%.

3. Results and Discussion

3.1. Critical Properties of Saturated Triglycerides

The critical temperatures and pressures of the triglycerides measured in this work are given in Fig.1
and Tables 4 and 5. The internal consistency of the experimental critical pressures was checked by

6
 1/ 2
considering the plot ( M / pc ) versus n, where M is the molar mass. According to the Lydersen

1/ 2
method of the estimation of the critical properties [36,37], the quantity ( M / pc ) should be a linear

function of the number of repeating units in a molecule for a given homologous series. Figure 2 shows
that the experimental critical pressures of the triglycerides studied follow this rule well enough. Usually
to test the consistency of experimental critical temperatures, the plot Tc / Tnb versus n is considered,

where Tnb is the normal boiling temperature. However, the normal boiling points of triglycerides are
high enough and it is difficult to measure them with satisfactory accuracy. [24].
The critical properties of triglycerides were also estimated using some group contribution methods. We
considered only the methods that do not employ the normal boiling temperature as an input parameter
because, as mentioned above, the normal boiling points of triglycerides are unreliable. We used the
method suggested by Constantinou and Gani (CG) as presented by Poling et al. [38] because the original
paper by Constantinou and Gani [39] has some misprints. We also made use of the methods of Marrero
and Gani (MG) [40], Hukkerikar et al. [41], and Cunico et al. [42]. Hukkerikar et al. presented a revised
version of the Marrero/Gani method. In comparison with the classical Marrero/Gani method, they used a
larger experimental data set. They saved the property prediction model but recalculated the parameters
of the model and group contributions by applying two ways. In these ways, the model parameters and
the group contributions were obtained using either a step-wise regression method or a simultaneous
regression one. We shall denote these methods by H-sw and H-s, respectively. Cunico et al. [42] revised
the group contributions and the universal constants of the Marrero/Gani method for lipids. In the paper
by Cunico et al., the contributions for the group CH2CHCH2 needed to calculate the critical
temperatures and pressures of triglycerides are absent. The contributions of this missing group were
estimated by us using the group contribution+ approach suggested by Gani et al. [43].
Percent deviations of the experimental critical temperatures and pressure from the values obtained by
the group contribution methods are given in Tables 4 and 5. All the group contribution methods used
overestimate the critical temperatures of triglycerides. The CG method gives the best results; for it, the
average absolute percentage error (AAPE) is about 1.5% and the maximum absolute percentage error
(MAPE) is only a little more than 2%. The H-sw, H-s, and C methods provide approximately equal
accuracy with the values of AAPE and MAPE about 2% and 3%, respectively. In general, the CG
method and the modern modifications of the MG method (H-sw, H-s, and C ones) give good enough
results in predicting the critical temperatures of triglycerides. Then the MG method follows.
The critical pressures of triglycerides are predicted by the group-contribution methods used in this
study much worse than the critical temperatures. The MG and C methods overestimate the critical
pressures while the CG, H-s, and H-sw ones underestimate them. Table 5 shows that the C method gives
7
the best estimation of the critical pressures of triglycerides and the difference between the experimental
critical pressures and the values predicted by the C method decreases with increasing the number of
carbons in a molecule of the parent acid.

3.2. Heat Capacities of Saturated Triglycerides

The molar heat capacities of triglycerides measured in this work were approximated by Eq.(4). The
coefficients of this equation are given in Table 2 and the heat capacities of triglycerides calculated with
the help of Eq.(4) with an increment of 5 K are given in Table 3. As noted above, heat capacities are
studied well enough for triglycerides with the number of carbons in the parent acid n = 12, 14, 16, 18
while there are only scattered data for the rest of saturated triglycerides. Phillips and Mattamal [44]
measured the heat capacities of triglycerides C3H5[CnH2n-1O2]3 with n = 8 , 10, 12, and 14. The authors
write that the results were reproducible to about 5% but they do not give any estimation of the
uncertainty and the purities of the samples investigated. This considerably reduces the value of these
data. Charbonnet and Singleton [45] measured the heat capacities of glyceride tridodecanoate and
glyceride tritetradecanoate. The purities of the samples they used were 98%, but they do not give the
uncertainty of the data obtained. Measurements of the heat capacity of glyceride tridodecanoate and
glyceride tritetradecanoate were also carried out by Morad et al. [46,47]. The purities of the samples
used by Morad and co-workers were higher than 99%; the authors estimate their experimental heat
capacities to be accurate within ±1 %. Unfortunately, Morad and co-workers in their paper [46] give the
heat capacities of glyceride tridodecanoate and glyceride tritetradecanoate only in the form of equations.
The heat capacities of triglycerides measured in this work are compared with those determined by
various authors in Fig. 3. For glyceride trioctanoate and glyceride tridecanoate, there are only three
experimental data points for each compound determined by Phillips and Mattamal [44]. The drawbacks
of the Phillips and Mattamal experiments were discussed above. The differences between our data and
those of Phillips and Mattamal fall in the range 2.8 to 6.5% for glyceride trioctanoate and 0.9 to 3.3%
for glyceride tridecanoate. For glyceride tridodecanoate and glyceride tritetradecanoate, our data lie
between those of Charbonnet and Singleton and Morrad and co-workers. Our molar heat capacities of
glyceride tridodecanoate are lower than those of Charbonnet and Singleton by 1.6 to 3.05% but higher
than those of Morrad et al. [46] by 0.1 to 3.3%, and higher than those of Morrad et al. [47] by 1.35 to
2.9%. So for glyceride tridodecanoate, the discrepancies between our data and those by Morrad and co-
workers are approximately equal to the combined uncertainty of these methods (3%). For glyceride
tritetradecanoate, our data differ from those of Charbonnet and Singleton by 1.8 to 2.6%, from those of

8
Morrad et al. [46] by 0.01 to 2.0%, and from those of Morrad et al. [47] by 0.5 to 4.3%. The results of
measuring the heat capacity of glyceride tritetradecanoate by Phillips and Mattamal [44] are suspicious.
The molar heat capacities of triglycerides as a function of the number of carbons in the parent acid n at
a temperature of 350 K are shown in Fig.4. This temperature was chosen because at it the maximum
experimental data on the heat capacities of triglycerides in the liquid phase are available in the literature.
Here the heat capacities of triglycerides with n = 4, 6, 16, 18 were calculated using equations suggested
by Zábranský et al. [16,17]; the heat capacities of triglycerides with n = 8, 10, 12, 14 are our data.
Figure 4 demonstrates that the experimental data points fall with good accuracy on a straight line. The
linear dependence of the molar heat capacity on the number of carbons in a molecule at a given
temperature was previously observed for methyl esters of n-alkanoic acids [48].

3.3. Thermal Diffusivities and Thermal Conductivities of Saturated Triglycerides

The thermal diffusivities of saturated triglycerides measured in this work are presented in Table 3. We
failed to find in the literature any information about thermal diffusivities of triglycerides for
comparison. The dependence of the thermal diffusivities of triglycerides on temperature in the
temperature range from about the melting point of a triglyceride to 410-425 K can be written in the form
of a linear function:

a = b0 − b1T (5)

The coefficients of Eq.(5) are given in Table 6. The thermal conductivity of a compound is frequently a
more required quantity than the thermal diffusivity. Using our experimental data on thermal diffusivities
and the heat capacities, and densities measured by Phillips and Mattamal [49], we calculated thermal
conductivities of triglycerides by the formula:

λ = aC pρ , (6)

where ρ is density. The results of the calculation are given in Table 3. The data obtained were
approximated by a linear function:

λ = c0 − c1T . (7)

9
The parameters of Eq.(7) are listed in Table 6.

4. Conclusion

The critical temperatures, pressures, heat capacities, and thermal diffusivities of saturated triglycerides
C3H5[CnH2n-1O2]3 with n = 8, 10, 12, 14 have been measured. The critical constants of triglycerides have
also been estimated by the group contribution methods of Constantinou and Gani, Marrero and Gani,
Hukkerikar and co-workers in two variants, and Cunico and co-workers. The Constantiou and Gani
method provides the best estimation of the critical temperatures of triglycerides while the Cunico and
co-workers method gives the best estimation of the critical pressures. The heat capacities and thermal
diffusivities have been measured in the temperature range from about the melting point of a triglyceride
to 410-425 K. The temperature dependencies of the heat capacities and thermal diffusivities have been
approximated by a third-order and a first-order polynomial, respectively. The coefficients of these
polynomials have been given. It has been shown that the dependence of the molar heat capacity on n (n
= 4-18) at a given temperature is close to linear. The thermal conductivities of triglycerides have been
calculated using the experimental data on heat capacities and thermal diffusivities obtained in this work
and the data on density measured by Phillips and Mattamal [49].

Acknowledgements

The study was supported by the Russian Foundation for Basic Research under Grant No. 14-08-00242 A
and the Complex Program for Basic Research of the Ural Branch of the Russian Academy of Sciences
N 15-1-2-6.

10
Literature Cited

[1] G. Anitescu, Th.J. Bruno, J. Supercrit. Fluids 63 (2012) 133-149.

[2] K.T. Lee, S. Lim, Y.L. Pang, H.Ch. Ong, W.T. Chong, Prog. Energy Combust. Sci. 45 (2014) 54-78.
[3] S.P. Singh, D. Singh, Renew. Sust. Energy. Rev. 14 (2010) 200-216.
[4] N.A. Morad, M. Idress, A.A. Hasan, J. Thermal. Anal. 44 (1995) 823-835.
[5] E.D. Nikitin, P.A. Pavlov, P.V. Skripov, J. Chem. Thermodyn. 25 (1993) 869-880.
[6] E.D. Nikitin, P.A. Pavlov, A.P. Popov, Fluid Phase Equilib. 149 (1998) 223-232.
[7] E.D. Nikitin, P.A. Pavlov, M.G. Skutin, Fluid Phase Equilib. 161 (1999) 119-134.
[8] E.D. Nikitin, A.P. Popov, Yu.G. Yatluk, V.A. Simakina, J. Chem. Eng. Data 55 (2010) 178-183.
[9] E.D. Nikitin, P.A. Pavlov. Method GSSSD ME 163-2010. Method of experimental determination of
the critical temperature and critical pressure of pure substances. The Russian Scientific and Technical
Centre for Information on Standardization, Metrology and Conformity Assessment. FSUO
"STANDARTINFORM", Deposit document No 865a-2010 kq. Moscow, 2010. (in Russian).
[10] J.-F. Hoffmann, J.-F. Henry, G. Vaitilingom, R. Olives, M. Chirtoc, D. Caron, X. Py, Int. J.
Thermal Sci. 107 (2016) 105-110.
[11] E.E.G. Rojas, J.S.R. Coimbra, J. Telis-Romero, Int. J. Food Prop. 16 (2013) 1620-1629.
[12] G. Lara-Hernández, E. Suaste-Gómez, A. Cruz-Orea,·J. G. Mendoza-Alvarez, F. Sánchez-
Sinéncio,·J. P. Valcárcel, A. García-Quiroz, Int. J. Thrmophys. 34 (2013) 962-971.
[13] A. Turgut, I. Tavman, S. Tavman, Int. J. Food Prop. 12 (2009) 741-747.
[14] J. Brock, M.R. Nogueira, C. Zakrzevski, F. de C. Corazza, M.L. Corazza, J.V. de Oliveira, Ciênc.
Tecnol. Aliment., Campinas, 28 (2008) 564-570.
[15] D. Dadarlat, J. Gibkes, D. Bicanic,A. Pasca, J. Food. Eng. 30 (1996) 155-162.
[16] M. Zábranský,V. Růžička, Jr., V. Majer, E.S. Domalski., Heat Capacity of Liquids, Critical Review
and Recommended Values, Monograph No. 6, Vol. I and Vol. II, American Chemical Society,
Washington, D.C., 1996.
[17] M. Zábranský, V. Růžička, Jr., E.S. Domalski, J. Phys. Chem. Ref. Data 30 (2002) 1199-1689.
[18] M. Zábranský, Z. Kolská, V. Růžička, Jr., E.S. Domalski, J. Phys. Chem. Ref. Data 39 (2010)
013103-1-013103-404.
[19] V.P. Skripov, Metastable Liquids, Wiley, New York, 1974.
[20] V.P. Skripov, E.N. Sinitsyn, P.A. Pavlov, G.V. Ermakov, G.N. Muratov, N.V. Bulanov, V.G.
Baidakov, Thermophysical Properties of Liquids in the Metastable (Superheated) State, Gordon and
Breach, London, New York, 1988.
[21] E.D. Nikitin, A.P. Popov, Fluid Phase Equilib. 379 (2014) 191–195.

11
[22] L.P. Filippov, Podobie Svoistv Veshchestv (Similarity of Substance Properties), Moscow
University Press, Moscow, 1978.
[23] K.S. Pitzer, D.Z. Lippmann, R.F. Curl, Jr, Ch.M. Huggins, D.E. Petersen, J Amer. Chem. Soc. 77
(1955) 3433-3440.
[24] T. Wallek, J. Rarey, J.O. Metzger J. Gmehling, Ind. Eng. Chem. Res. 52 (2013) 16966-16978.
[25] E.D. Nikitin, A.P. Popov, Fluid Phase Equilib. 237 (2005) 16–20.
[26] E.D. Nikitin, A.P. Popov, Yu.G. Yatluk, J. Chem. Eng. Data 51 (2006) 1335-1338.
[27] E.D. Nikitin, A.P. Popov, N.S. Bogatishcheva, V.A. Kuznetsov, Fluid Phase Equilib. 355 (2013)
40-45.
[28] E.D. Nikitin, A.P. Popov, Fluid Phase Equilib. 380 (2014) 11–17.
[29] E.D. Nikitin, A.P. Popov, Yu.G. Yatluk, J. Chem. Eng. Data. 51 (2006) 1326–1330.
[30] M.A. Stephens, W.S. Tamplin, J. Chem. Eng. Data. 24 (1979) 81-82.
[31] G.W.H Hohne, W.F. Hemminger, H.F. Flammersheim, Differential Scanning Calorimetry,
Springer–Verlag, Berlin, Heidelberg, 2003.
[32] J.W. Goodrum, M.A. Eiteman, Boiresour. Technol. 56 (1996) 55-60.
[33] www.netzsch-thermal-analysis.com.
[34] J. Blumm, A. Lindemann, S. Min, Thermochim. Acta 455 (2007) 26-29.
[35] N.B. Vargaftik, Handbook of Physical Properties of Liquids and Gases, Springer–Verlag, Berlin,
Heidelberg, 1975.
[36] R.C. Reid, J.M. Prausnitz, T.K. Sherwood, The Properties of Gases and Liquids, 3rd ed., McGraw-
Hill, New York, 1977.
[37] A.L. Lydersen, Estimation of critical properties of organic compounds by the method of group
contributions. University of Wisconsin, College of Engineering, Engineering Experiment Station,
Madison (WI), 1955 April. Report No. 3.
[38] B.E. Poling, J.M. Prausnitz, J.P. O’Connell, The Properties of Gases and Liquids, 5th ed.,
McGraw-Hill, New York, 2000.
[39] L. Constantinou, R. Gani, AIChE J. 40 (1994) 1697-1710.
[40] J. Marrero, R. Gani, Fluid Phase Equilib. 183-184 (2001) 183-208.
[41] A.S. Hukkerikar, B. Sarup, A.T. Kate, J. Abildskov, G. Sin, R. Gani, Fluid Phase Equilib. 321
(2012) 25-43.
[42] L.P. Cunico, A.S. Hukkerikar, R. Ceriani, B. Sarup, R. Gani, Fluid Phase Equilib. 357 (2013) 2-18.
[43] R. Gani, P.M. Harper, M. Hostrup, Ind. Eng. Chem. Res. 44 (2005) 7262-7269.
[44] J.C. Phillips, M.M. Mattamal, J. Chem. Eng. Data 21 (1976) 228-232.
[45] G.H. Charbonnet, W.S. Singleton, J. Am. Oil Chem. Soc., 24 (1947) 140-142.

12
[46] N.A. Morad, M. Idress, A.A. Hasan, J. Therm. Anal. 45 (1995) 1449-1461.
[47] N.A. Morad, A.A.M. Kamal, F. Panau, T.W. Yew, J. Am. Oil. Chem. Soc. 77 (2000) 1001-1005.
[48] M.J. van Bommel, H.A.J. Oonk, J.C. van Miltenburg, J. Chem. Eng. Data 49 (2004) 1036-1042.
[49] J.C. Phillips, G.J. Mattamal, J. Chem. Eng. Data 23 (1978) 1-6.

13
Figure Captions

Figure 1. The critical temperatures (a) and the critical pressures (b) of triglycerides C3H5[CnH2n-1O2]3 vs

the number of carbons in the parent acid n.

Figure 2. Correlation of the critical pressure of triglycerides C3H5[CnH2n-1O2]3 as a function of the

number of carbons in the parent acid n and molar mass M.

Figure 3. Molar heat capacities of triglycerides C3H5[CnH2n-1O2]3: this work ■, Phillips and

Mattamal [44] ▲, Charbonnet and Singleton [45] ○, Morad et al., 1995 [46] ----, Morad et al., 2000 [47]

□.

Figure 4. Molar heat capacities of triglycerides C3H5[CnH2n-1O2]3 vs the number of carbons in the parent

acid n at a temperature of 350 K: this work ■, Zábranský et al. [16,17] ○.

14
Figure

940

920

900

880
T /K
c

860

840 (a)

820
8 9 10 11 12 13 14
n
Figure

0.85

0.80

0.75
p /MPa

0.70
c

0.65

(b)
0.60

8 9 10 11 12 13 14

n
Figure

1.1
1/2
/MPa)]

1.0
c
p

0.9
/kg mol )/(
-1

0.8
M[(

0.7
8 9 10 11 12 13 14

n
Figure

1700

1600 n = 14

1500

1400 n = 12
/(J/(mol K))

1300
p,m
C

1200
n = 10

1100

1000

n=8
900
350 400 450
T/K
Figure

2000

1750

1500
/(J/(mol K))

1250

1000
p,m
C

750

500
5 10 15
n
Table 1

Purities of saturated triglycerides used in measurements (Alfa Aesar, GC and TLC, supplier’s
certificates of analysis).

Compound CASRNa Purity (mol. fraction)

GC TLC
Glyceride trioctanoate 538-23-8 0.99 1.00
Glyceride tridecanoate 621-71-6 1.00
Glyceride tridodecanoate 538-24-9 1.00
Glyceride tritetradecanoate 555-45-3 1.00 1.00
a
Chemical abstracts service registry number.

15
Table 2
Constants of Eq. (4) for the calculation of the molar heat capacity of triglycerides

Triglyceride Temperature a0 ⋅10−3 / a1 / a2 / a3 ⋅104 /

range/K (J/(mol K)) (J/(mol K2)) (J/(mol K3)) (J/(mol K4))

Glyceride trioctanoate 313.4-428.9 -0.61637 10.215 -0.02356 0.20783

Glyceride tridecanoate 317.1-418.1 -3.7642 35.029 -0.08468 0.70781
Glyceride tridodecanoate 326.6-410.6 15.541 -120.99 0.33778 -3.0840
Glyceride tritetradecanoate 347.5-410.0 23.025 -175.51 0.47196 -4.1666

16
Table 3. Experimental values of molar heat capacities C p,m , thermal diffusivities a , and calculated values of thermal conductivities λ of liquid

saturated triglycerides at pressure p = 0.1 MPaa.

T/K C p,m /(J/(mol ⋅ K)) a ⋅107 /(m2/s) λ /(W/(m ⋅ K)) C p,m /(J/(mol ⋅ K)) a ⋅107 /(m2/s) λ /(W/(m ⋅ K))

Glyceride trioctanoate Glyceride tridecanoate

315 913.3 0.838 0.152
320 921.0 0.833 0.152 1093 0.891 0.153
325 928.5 0.828 0.151 1106 0.885 0.153
330 935.9 0.823 0.151 1117 0.879 0.153
335 943.1 0.818 0.151 1128 0.873 0.153
340 950.2 0.814 0.150 1139 0.867 0.153
345 957.1 0.809 0.150 1148 0.861 0.153
350 964.0 0.804 0.150 1157 0.855 0.152
355 970.7 0.799 0.149 1166 0.850 0.152
360 977.4 0.794 0.149 1174 0.844 0.151
365 984.1 0.789 0.148 1182 0.838 0.150
370 990.7 0.784 0.148 1189 0.832 0.150
375 997.2 0.779 0.147 1196 0.826 0.150
380 1004 0.775 0.146 1203 0.820 0.149
385 1010 0.770 0.146 1209 0.814 0.148
390 1017 0.765 0.145 1216 0.809 0.147
395 1024 0.760 0.145 1222 0.803 0.147
400 1030 0.755 0.144 1228 0.797 0.146
405 1037 0.750 0.144 1235 0.791 0.145

17
410 1044 0.745 0.143 1241 0.785 0.144
415 1051 0.740 0.142 1248 0.779 0.143
420 1058 0.736 0.142
425 1065 0.731 0.141

Glyceride tridodecanoate Glyceride tritetradecanoate

330 1316 0.904 0.164

335 1323 0.899 0.163
340 1331 0.893 0.162
345 1340 0.888 0.161
350 1350 0.882 0.161 1547 0.911 0.171
355 1361 0.877 0.160 1557 0.905 0.170
360 1372 0.871 0.160 1568 0.900 0.170
365 1384 0.866 0.159 1580 0.894 0.169
370 1396 0.861 0.159 1592 0.889 0.169
375 1407 0.855 0.158 1606 0.883 0.168
380 1418 0.849 0.158 1619 0.877 0.168
385 1428 0.844 0.157 1632 0.872 0.168
390 1438 0.838 0.156 1645 0.866 0.167
395 1446 0.833 0.155 1657 0.861 0.167
400 1453 0.827 0.154 1668 0.855 0.166
405 1458 0.822 0.153 1678 0.850 0.165
410 1461 0.816 0.151 1686 0.844 0.164

18
a
Standard uncertainties u are u ( T ) = 0.1 K for the measurements of both the heat capacities and the thermal diffusivities, u ( p ) = 4 kPa, the

( )
relative combined expanded uncertainties are U r C p,m = 0.02 , U r ( a ) = 0.05 , and U r ( λ ) = 0.07 with 0.95 level of confidence.

19
Table 4. Critical temperatures Tc /K of saturated triglycerides: experimental values and comparison

with predictive methods; 1 / τ0 is correction factor for critical temperature

Triglyceride 1 / τ0 Exptla CGb MGc H-swd H-se Cf

Glyceride trioctanoate 1.006 836 851.2 887.3 859.8 859.8 850.0

Glyceride tridecanoate 1.007 864 882.8 924.1 890.3 890.3 888.9
Glyceride tridodecanoate 1.007 899 909.7 955.8 916.4 916.4 922.0
Glyceride tritetradecanoate 1.007 925 933.1 983.7 939.3 939.3 950.6
AAPE/%g 1.52 6.44 2.34 2.34 2.47
MAPE/%h 2.18 6.96 3.04 3.04 2.88

a
The relative combined expanded uncertainty U r ( Tc ) = 0.015 with 0.95 level of confidence.
b
method of Constantinou and Gani as presented by Poling et al. [38].
c
method of Marrero and Gani [40].
d
method of Hukkerikar et al. [41], step-wise regression procedure.
e
method of Hukkerikar et al. [41], simultaneous regression procedure.
f
method of Cunico et al. [42].
g
AAPE = (1/ N ) (∑ Yc
exp
)
− Yccalc / Ycexp , where N is the number of experimental data points, Ycexp is

the experimental value of the critical property, and Yccalc is the calculated value of the critical property.
h
MAPE = Ycexp − Yccalc / Ycexp .
max

20
Table 5. Critical pressures pc /MPa of saturated triglycerides: experimental values and comparison with

predictive methods; 1 / π0 is correction factor for critical pressure

Triglyceride 1 / π0 Exptla CGb MGc H-swd H-se Cf

Glyceride trioctanoate 1.08 0.83 0.741 1.026 0.764 0.696 1.005

Glyceride tridecanoate 1.10 0.75 0.589 0.912 0.585 0.519 0.836
Glyceride tridodecanoate 1.10 0.65 0.488 0.838 0.463 0.403 0.729
Glyceride tritetradecanoate 1.10 0.61 0.417 0.787 0.376 0.322 0.657
AAPE/% 22.19 25.79 24.27 33.04 13.10
MAPE/% 31.64 29.01 38.36 47.21 21.08

a
The relative combined expanded uncertainty U r ( pc ) = 0.04 with 0.95 level of confidence.
b
method of Constantinou and Gani as presented by Poling et al. [38].
c
method of Marrero and Gani [40].
d
method of Hukkerikar et al. [41], step-wise regression procedure.
e
method of Hukkerikar et al. [41], simultaneous regression procedure.
f
method of Cunico et al. [42].

21
Table 6. Coefficients of Eq.(5) and Eq.(7) for the calculation of the thermal diffusivities and thermal conductivities of triglycerides

Triglyceride Temperature b0 ⋅106 /(m2/s) b1 ⋅1010 /(m2/(s K)) c0 /(W/(m K)) c1 ⋅104 /(W/(m K2))
range/K
Glyceride trioctanoate 315-425 0.1145 0.9747 0.1847 1.0119
Glyceride tridecanoate 320-415 0.1265 1.170 0.1895 1.0842
Glyceride tridodecanoate 330-410 0.1266 1.097 0.2104 1.4069
Glyceride tritetradecanoate 350-410 0.1300 1.112 0.2067 1.0220

22
Highlights

► Critical properties, heat capacities, thermal diffusivities of four triglycerides were measured. ► The triglycerides studied have 8, 10, 12, 14
carbon atoms in a parent acid.
► The triglycerides decompose at near-critical temperatures.
► Pulse-heating method with short residence times was used.

23