See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/323538202

Unravelling the chemical reactions of fatty acids and triacylglycerides under

hydrodeoxygenation conditions based on a comprehensive thermodynamic
analysis

Preprint  in  Biomass and Bioenergy · March 2018

DOI: 10.1016/j.biombioe.2018.02.014

CITATIONS READS

5 750

6 authors, including:

Diego Valencia Isidoro García-Cruz

Instituto Mexicano del Petroleo Instituto Mexicano del Petroleo
27 PUBLICATIONS   464 CITATIONS    44 PUBLICATIONS   556 CITATIONS   

SEE PROFILE SEE PROFILE

Victor Uc Luis Verduzco

Metropolitan Autonomous University Instituto Mexicano del Petroleo
19 PUBLICATIONS   327 CITATIONS    35 PUBLICATIONS   1,038 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Heavy crude oil Emulsions View project

Measurement and prediction of physical and chemical properties of compounds and substances View project

All content following this page was uploaded by Isidoro García-Cruz on 03 March 2018.

The user has requested enhancement of the downloaded file.

 Biomass and Bioenergy 112 (2018) 37–44

Contents lists available at ScienceDirect

Biomass and Bioenergy

journal homepage: www.elsevier.com/locate/biombioe

Research paper

Unravelling the chemical reactions of fatty acids and triacylglycerides under T

hydrodeoxygenation conditions based on a comprehensive thermodynamic
analysis
Diego Valenciaa,∗, Isidoro García-Cruza, Víctor Hugo Ucb, Luis Felipe Ramírez-Verduzcoa,
Myriam A. Amezcua-Allieria, Jorge Aburtoa
a
Dirección de Investigación en Transformación de Hidrocarburos, Instituto Mexicano del Petróleo, Eje Central Lázaro Cárdenas 152, Col. San Bartolo Atepehuacan, C.P.
07730, Mexico City, Mexico
b
CBI, Universidad Autónoma Metropolitana-Azcapotzalco, Av. San Pablo 180, Col. Reynosa Tamaulipas, Mexico City, C.P. 02200, Mexico

A R T I C L E I N F O A B S T R A C T

Keywords: Triacylglycerides (TAGs) and their related fatty acids (FAs) are abundant in biomass (vegetal oils and animal
Biomass fats). Here, we studied by theoretical calculations the most representative FAs and their chemical reactions
Hydrogenation under hydrogen atmosphere to produce O-free hydrocarbon molecules. In general, the ΔHr0 and ΔGr0 do not vary
Fatty acid due to the size of the molecule. Selected TAGs were also studied by theoretical calculations. Similar as FAs, the
Triacylglyceride
overall reaction energies do not depend on the size of the molecules. Possible reaction networks for the hy-
Density functional theory
drogenation and hydrogenolysis processes are proposed for such biomolecules under hydrogen environments.
We also studied the profiles of the enthalpy and Gibbs energy of reaction for TAGs at typical operating conditions
of temperature or pressure. This self-consistent study is willing to contribute for a better understanding of the
reactions of the biomass that take place under hydrogen atmosphere.

1. Introduction
Interest for renewable energies is growing for their potential alter-
natives to bring similar raw materials or energy than those obtained
from fossil oils [1,2]. The generation of bio- or green fuels since early
this decade has become a major research and industrial topic [3–5].
Vegetable oils, especially palm oil, have been good candidates to de-
velop environmental friendly and high quality fuels [6] if we keep in
mind sustainable production techniques. However, these vegetable oils
cannot be used immediately as fuels due to their high oxygen content,
which leads to low heating value, immiscibility with fossil fuels, a
tendency for polymerization, thermal instability and high viscosity [7].
Catalytic hydrotreating (HDT) is a promising technology for the con-
version of biomass to liquid biofuels. Hydrodeoxygenation (HDO) is the
main catalytic process to eliminate oxygen contained in the organic
molecules by hydrogenation and hydrogenolysis reactions [8–10]
Nowadays, several research efforts are trying to eliminate oxygen from
the vegetable oils by their chemical reactions under hydrogen atmo-
sphere [11,12]. Those have faced several challenges and future op-
portunities about chemistry, catalytic processes, biorefinery concepts,
engineering changes and solutions are emerging [13].
Target molecules in vegetable oils for biofuels production are tria-
cylglycerides (TAGs) that are meant to be converted into fatty acids
(FAs). After their reaction under hydrogen atmosphere, TAGs and FAs
are transformed into hydrocarbon chains that are able to be used as fuel
or incorporated in existing refinery infrastructure [14,15]. Their con-
version is complicated since the typical catalyst used for HDT processes
is poisoned by coke deposition and unstable in the presence of water
[16]. On the other hand, detailed steps of chemical reaction of the TAG
or FAs under hydrogen atmosphere to produce O-free hydrocarbons are
almost scarce in the literature. There are studies of the chemical reac-
tions of TAGs for the production of biodiesel, but in this process TAGs
react with H2O or alcohols to produce FAs or esthers [17]. It is well-
known that FAs under hydrogen atmosphere can react towards three
main routes, namely deoxygenation, decarbonylation or decarboxyla-
tion.
Detailed thermochemical data is necessary for a better under-
standing of the factors that affect the elimination of oxygen from bio-
mass molecules. Thermodynamic properties such as ΔH or ΔG have
been reported for selected FAs and some of them showed discrepancies
[18–20]. Those values were calculated by different approaches, in some
cases based on very rough models or adapting engineering data of

∗
Corresponding author.
E-mail address: dvalencia@imp.mx (D. Valencia).

https://doi.org/10.1016/j.biombioe.2018.02.014
Received 21 August 2017; Received in revised form 31 December 2017; Accepted 16 February 2018
0961-9534/ © 2018 Elsevier Ltd. All rights reserved.
D. Valencia et al. Biomass and Bioenergy 112 (2018) 37–44

estimated energy for the bonds within the FAs. For instance, dec-
arboxylation and hydrodeoxygenation are reported at certain tem-
perature or pressure and sometimes the values do not agree well be-
tween different reports. To the best of our knowledge, full
thermochemical data for the reactions of TAGs under hydrogen atmo-
sphere have not been reported yet and those values are very much
needed to understand their conversion into green fuels. Moreover, re-
action steps and intermediate structures are necessary to draw a clear
reaction scheme for the TAGs under HDO conditions. Those studies are
very difficult to conduct by experimental methods. On the other hand,
accurate theoretical calculations are currently possible with the existing
supercomputer infrastructures, for which these biomolecules are very
attractive. Therefore, we are presenting a detailed and consistent the-
oretical study about the chemical reactions and their energies for FAs
and TAGs under hydrogen atmosphere.

2. Computational details
Fig. 1. Optimized geometries of fatty acids: (a) stearic acid, (b) oleic acid, and (c) linoleic
acid.
FAs were chosen as they are found in the nature and are relevant for
the production of green fuels (C8 to C18) such as caprylic, capric, lauric,
palmitic, stearic acids and others with unsaturation in the aliphatic Table 1
HOMO (εHOMO), LUMO (εLUMO) energies and their gap for fatty acids.
chain. Ground state structures of the molecules were obtained by means
of Density Functional Theory (DFT) calculations. The optimized geo- Name (C:D) εHOMO εLUMO ΔH-L
metry for each compound was obtained using the hybrid meta GGA
density functional M06 [21] with the 6-31+G(d,p) standard basis set. (eV) (eV) (eV)
We obtained the optimized geometry of tricaprylin and related products
Caprylic acid (8:0) −8.27 −0.65 7.62
at the same level of theory. However, TAGs bigger than caprylin were Capric acid (10:0) −8.27 −0.66 7.61
studied at M06/6-31G(d,p) level due to the size of the system. It was not Lauric acid (12:0) −8.26 −0.66 7.60
computationally possible to get geometry optimization at higher level Myristic acid (14:0) −8.26 −0.66 7.60
of theory for TAGs larger than C10, due to the diffuse function makes the Palmitic acid (16:0) −8.25 −0.66 7.59
Stearic acid (18:0) −8.23 −0.66 7.57
calculation complicated for its convergence. Thus, we chose the M06/6- Palmitoleic acid (16:1) −6.83 −0.67 6.16
31G(d,p) level to get the optimized geometry of TAGs and then single Oleic acid (18:1) −6.83 −0.67 6.16
point calculation were done at the same level of theory than FAs for a Linoleic acid (18:2) −6.78 −0.76 6.01
proper comparison of the energy values varying the size of the TAG. All
harmonic vibrational frequencies were real; therefore the structures are
minima on the potential energy surface. electronic properties of those molecules are shown in Table 1. The FAs
We chose the M06 functional because it has shown its excellent from caprylic (C8) to stearic acid (C18) showed slight changes in their
performance to get accurate thermodynamic properties [22]. The average values for HOMO (−8.25 eV) and LUMO (−0.66 eV) energies,
thermal correction to the total energy was obtained with the con- which pointed out that the size of the aliphatic chain does not have a
tribution of the translation, electronic, rotational and vibrational mo- significant influence on the electronic structure of those FAs. The band
tions. Then ΔH and ΔG were computed for the chemical processes that gap of the linear molecules slightly decreases as the number of carbon
have been described experimentally for the FAs and for a proposed atoms increase, caprylic acid 7.62 eV and stearic acid 7.57 eV. Oleic
network of TAGs under H2 atmosphere. We also conducted a thermo- acid (C18:1Δ9) exhibited a change in its HOMO and LUMO energy
chemical study by varying temperature and pressure for the HDO, values respect to its saturated counterpart. The gap was decreased to
decarbonylation and decarboxylation processes of selected FA. On the 6.16 eV. However, a comparison with palmitoleic acid (C16:1Δ9)
other hand, we proposed a reaction network for the partial FA elim- showed the same values than the ones for oleic acid. The addition of
ination from the TAGs and its partial reduction under H2 environments. another C=C bond (linoleic acid) also decreased the values of the
ΔH and ΔG as a function of T and P were also calculated for the TAGs frontier MOs. Naturally, the number of the unsaturation is more im-
chemical reactions under H2 atmosphere. Enthalpy and Gibbs energy portant on their electronic properties than the total number of carbon
were calculated at M06/6-31+G(d,p). atoms.
All the calculations were carried out with Gaussian 09 [23], con- Chemical reactions of FAs under HDO conditions are depicted in
sidering all electron and closed shell systems. Fig. 2. We studied the consecutive reduction reactions of the -COOH
groups from the FAs to form the corresponding fatty alcohol (R1-R3), its
3. Results and discussion final reduction to hydrocarbon (R4) and the overall reaction from the
FA to the O-free hydrocarbon (R5). Other two reactions that take place
3.1. Fatty acids under HDO conditions are their thermal decomposition namely dec-
arbonylation (R6) or decarboxylation (R7) reactions. We studied the
We studied the most abundant FAs in biomass ranging from C8 to thermochemical data of those chemical reactions for the most re-
C18 and including those with unsaturations. Fig. 1 showed the opti- presentative FAs present in biomass. ΔHr0 and ΔGr0 values are shown in
mized geometries for selected FAs, namely stearic (18:0), oleic (18:1) Table 2. Reduction of the fatty acid moiety to produce the first hy-
and linoleic (18:2) acids. Those molecules contain the same number of drogenated product namely its aldehyde species showed positive values
carbon and oxygen atoms, but they have a different number of un- for both thermodynamic quantities. This value agrees with previous
saturated C=C bonds within their structures. Stearic acid is a linear studies where the first reduction reaction is not well promoted at
molecule with sp3 hybridization of all its carbon atoms, while oleic and standard conditions (ΔH0 and ΔG0 > 0) [19]. The ΔS0 of this process is
linoleic acids contain one and two C=C bond, respectively, with a cis positive (39.26 JK−1 mol−1), therefore it is a spontaneous reaction. We
configuration. It produces fatty acids with nonlinear structure. The could state that this reaction limits all the HDO process through the

38
D. Valencia et al. Biomass and Bioenergy 112 (2018) 37–44

(approximately ΔHr0 = −88 kJmol-1 and ΔGr0 = −97 kJmol-1). It is

Fig. 2. Chemical reactions of FA and their derivatives under HDO reaction conditions.
R1 = Reduction of the carboxylic acid to its aldehyde, R2 = Reduction of the aldehyde to
the alcohol species, R3 = overall reaction from R1 and R2, R4 = Reduction of the alcohol
group to the O-free hydrocarbon, R 5 = Overall HDO reaction of FA,
R6 = Decarbonylation reaction and R7 = Decarboxylation reaction.
reduction reactions with H2. Its initial chemical species (aldehyde) in
this pathway must be further transformed. The following reduction
reaction (R2) from the aldehyde to produce the alcohol species is highly
exothermic and exergonic. This reaction might promote the reduction
processes. The overall reaction from the fatty acid to the alcohol species
(R3) is an exothermic and exergonic reaction because of both chemical
reactions involved in this process (R1 and R2). The values are not
changing very much due to the size of the aliphatic chain (approxi-
mately ΔHr0 = −34 kJmol-1 and ΔGr0 = −8 kJmol-1). It means that all
the FAs would almost react to form the alcohol with almost the same
reaction energy. The values of enthalpy and Gibbs energies for the HDO
reaction of the alcohol species (R4) are also negative. Such values are
almost the same for the FAs no matter the size of the aliphatic chain
important to mention that this hydrogenolysis reaction is much more
favorable than the previous reduction reactions (R1 and R2). The
overall HDO reaction of the FAs (R5) is an exothermic and exergonic
process, which contains all the energetic contributions from the con-
secutive reduction and hydrogenolysis reactions. The values of the
thermodynamic quantities is the sum of the R1, R2 and R4. There is
indeed no changes in those values due to the size of the molecules.
Decarbonylation reaction (R6) is an endothermic and exergonic pro-
cess, which means that it would be favoured by increasing the tem-
perature. In contrast with the first step in the HDO process, this thermal
decomposition could occur at standard conditions (ΔG0 < 0). Like the
other reactions, the values for the energy reaction do not vary sig-
nificantly by the size of the molecules. The other thermal decomposi-
tion reaction (decarboxylation) of the FAs is also an exothermic and
exergonic process. Again the energy reaction values of this process are
almost the same no matter the FA analyzed. Finally, we calculated the
hydrogenation reaction of the C=C bond for the unsaturated com-
pounds. Palmitoleic (16:1), oleic (18:1) or linoleic (18:2) acids reacts
with H2 to produce their saturated counterparts with almost the same
energy values ΔH0 = −139 kJmol-1 and ΔG0 = −95 kJmol-1, which
compared to the hydrogenation of the C=O bond is much more fa-
vorable. Based on this thermodynamic values, it is expected that the
hydrogenation of C=C would take place first.
In general, the electronic properties and reaction energy values for
FAs under hydrogen atmosphere do not vary significantly by the size of
the aliphatic chain, similar results have been obtained for non-con-
jugated systems [24]. The electronic properties mainly changed due to
the presence of C=C bonds in the fatty acids. We could explain the
scarce variation on the reaction energies due to the size of the mole-
cules in terms of the very similar chemical nature of the substituents.
The molecules vary in the number of -CH2-CH2- groups, which make the
molecules very similar in their chemical character. A comparison be-
tween the reactions of FAs under hydrogen atmosphere showed that the
reduction of the -COOH group to the respective aldehyde and alcohol
species are less favorable than its final reduction reaction of the -OH
group (hydrogenolysis reaction) to form the hydrocarbon. It would
indicate that the thermodynamic limiting step is the first reduction of
the FAs compounds to its aldehyde and alcohol version. A catalyst (that
is not considering in this study) must overcome this thermodynamic
limitation imposed by the chemical reaction. The following reaction is
much more favorable and it produces the desired product of the HDO
process. On the other hand, the decarbonylation process is en-
dothermic, which can be tuned by temperature reaction and it is less
favorable than the reduction reactions. It might suggest this process
could be properly allowed by tuning reaction conditions. Based on the
thermochemical results in this paper, the decarboxylation process is
more favorable than decarbonylation process, but competitive with
reduction reactions. It is important to note that the decarboxylation is

Table 2
Enthalpy and Gibbs energies for the reactions (R1-R7) depicted in Fig. 2. Values are given at standard conditions (298 K and 1 atm).

Name (C:D) Reaction energy (kJmol−1)

R1 R2 R3 R4 R5 R6 R7

ΔHr0 ΔGr0 ΔHr0 ΔGr0 ΔHr0 ΔGr0 ΔHr0 ΔGr0 ΔHr0 ΔGr0 ΔHr0 ΔGr0 ΔHr0 ΔGr0

Caprylic acid (8:0) 33.1 21.4 −67.6 −29.8 −34.5 −8.3 −88.3 −98.2 −122.8 −106.5 48.9 −5.9 −24.9 −67.0
Capric acid (10:0) 33.2 21.8 −67.8 −30.4 −34.6 −8.6 −88.2 −98.0 −122.8 −106.6 48.8 −6.3 −24.9 −67.4
Lauric acid (12:0) 33.3 22.1 −67.9 −30.9 −34.6 −8.8 −88.1 −97.9 −122.7 −106.7 48.8 −6.4 −25.0 −67.5
Myristic acid (14:0) 33.2 22.8 −68.0 −31.0 −34.6 −8.8 −88.1 −97.9 −122.7 −106.7 48.7 −7.0 −25.0 −68.2
Palmitic acid (16:0) 33.3 22.5 −68.0 −31.2 −34.6 −8.8 −88.1 −97.9 −122.7 −106.7 48.8 −6.8 −25.0 −67.9
Stearic acid (18:0) 33.4 22.5 −68.0 −31.3 −34.7 −8.8 −88.1 −97.9 −122.8 −106.7 48.8 −7.0 −25.0 −68.1
Palmitoleic acid (16:1) 33.1 23.3 −68.1 −32.6 −35.0 −9.4 −88.2 −96.8 −123.2 −106.2 45.2 −7.9 −28.6 −69.0
Oleic acid (18:1) 33.7 24.7 −68.7 −33.5 −35.0 −8.7 −87.5 −95.2 −122.5 −103.9 48.5 −6.1 −25.3 −67.2
Linoleic acid (18:2) 32.8 29.2 −66.6 −34.4 −34.4 −7.1 −88.4 −98.8 −122.8 −105.9 49.5 −8.6 −24.2 −69.7

39
D. Valencia et al.
Fig. 3. Reaction energy values for hydrogenation reactions for fatty acid to produce its aldehyde ( ), aldehyde to the alcohol ( ) and alcohol to hydrocarbon ( ), calculated for caprylic
acid as a function of temperature or pressure: (a) ΔHr (T, 1 atm), (b) ΔGr (T, 1 atm), (c) ΔHr (473 K, P) and (d) ΔGr (473 K, P).
more favorable than the reduction of the -COOH group, but less than its
total reduction at standard conditions.
Effect of temperature and pressure on the thermodynamic quantities
were plotted in Fig. 3 for the reduction reactions of caprylic acid (C8:0).
It is important to mention that the calculated value in the present work
at 573 K and 1 atm (ΔG573 = −87.7 kJmol-1) agrees well with the re-
ported one in the literature (ΔG573 = −86.1 kJmol-1) for the HDO of
stearic acid [25]. As we discussed above, the effect of the aliphatic
chain on the values of the reaction energies is almost negligible. We
then chose the smallest molecule in this study to have an overview of
the profiles of reaction energies for the hydrogenation reactions (R1, R2
and R4). ΔHr showed minor changes due to the effect of the tempera-
ture on the three reaction schemes studied. Those values decreased or
become more exothermic by increasing reaction temperature. Gibbs
energy values changed due to the reaction temperature too. The re-
duction of the fatty acid to produce its aldehyde species becomes more
favorable at higher temperatures, while the following reaction has an
opposite effect at higher temperatures. Both reactions (R1 and R2) are
consecutive in the presence of reductive atmosphere, this is very im-
portant for a better understanding of HDO process. At higher con-
centrations of aldehyde species in the reaction mixture, decreasing
temperature reaction might allow higher yields of HDO due to the fa-
vorable production of alcohol species at lower temperatures. The hy-
drogenolysis reaction of the alcohol species (R4) has a similar profile
than the R1 respect to the reaction temperature, more favorable at
higher temperatures. This reaction is the most energetically preferred
and the last one on the reduction pathway.
Effect of pressure on the reaction energies was analyzed at 473 K
because typically the HDO process of FAs takes place at temperatures
above this value. ΔHr showed constant values along the pressure range
in this study (5–80 atm), there is not important effect of the reaction
pressure on the values of the enthalpy. This could be related to the
40
Biomass and Bioenergy 112 (2018) 37–44
reactions studied or to the limitation of the computational model that
uses the ideal gas equation to calculate volume. The Gibbs energy for
the reduction of the aldehyde to its alcohol species changes with
pressure, it is more exergonic at higher pressures and almost the same
value above 40 atm. This thermodynamic quantity for the other two
reduction reactions is almost constant in the range analyzed.
A comparison between the overall HDO reaction and the thermal
decomposition processes is depicted in Fig. 4. The overall HDO reaction
(R5) decreased their Gibbs energy at higher temperatures, as we dis-
cussed above this is related to the sum of the three reduction reactions
that are needed from the carboxylic acid to the O-free hydrocarbon
molecule. The main component for this reaction is the hydrogenolysis
reaction from the alcohol to the hydrocarbon. The energetic preference
of decarbonylation and decarboxylation processes (R6 or R7) is in-
creased at higher temperatures, but yet decarboxylation is still much
more favorable over decarbonylation reaction. We can expect a higher
competitive zone of FAs between reduction and decarboxylation reac-
tions at about 500 K. One can tune thermal decomposition reaction
without the use of hydrogen at higher temperatures and with an ap-
propriate catalyst. On the other hand, ΔGr values showed an opposite
trend as a function of pressure than the effect of temperature. In-
creasing pressure, would increase the ΔGr values for reduction reactions
and decrease the values for the thermal decomposition reactions. The
main contribution to the overall HDO came from the hydrogenolysis
reaction and the shape is given by the reduction of the aldehyde to the
alcohol as we mentioned above. It reached a maximum value at about
50 atm where the change on the reaction Gibbs energy with pressure is
very little. This result is in good agreement with the possibility of re-
lease a gas under higher pressure, which makes it less exergonic dec-
arboxylation or decarbonylation processes. It has been clearly shown
that ΔG exhibited different profiles by changing reaction conditions.
Thermochemical data for the reaction of FAs under HDO conditions
D. Valencia et al. Biomass and Bioenergy 112 (2018) 37–44

Fig. 4. Reaction energy values for the overall hydrodeoxygenation ( ), decarbonylation ( ) and decarboxylation ( ) reactions for caprylic acid as a function of temperature or pressure:
(a) ΔHr (T, 1 atm), (b) ΔGr (T, 1 atm), (c) ΔHr (473 K, P) and (d) ΔGr (473 K, P).

has been obtained by theoretical calculations. These results described

the thermodynamic preference of the molecules to react towards hy-
drogenation, decarboxylation or decarbonylation at given T or P.
However, the use of heterogeneous catalysts might direct the reaction
to a product over another one due to different transition states or me-
chanism between the FAs and the catalytic surface. In this line, a cat-
alyst will never change the thermodynamic quantities of the molecules
at given reaction conditions. This thermodynamic information is useful
and necessary for a deeper discussion of the energetic limitations due to
the type of reactions. If we compare thermochemical data obtained in
this work with existing literature, the MoS2 supported catalysts are in
good agreement with the preference of HDO over the thermal decom-
position reactions, where the major product is the paraffin [24].
However, Ni-based or Pt-based catalysts preferred, in most of the cases,
react through decarboxylation and decarbonylation processes over Fig. 5. Optimized geometry of selected triacylglyceride: tripalmitin. This compound is
HDO [26]. In general, production of CO2 is larger than CO for the present in higher quantities in several vegetal oils or animal fats. The optimized geo-
sulfided catalysts as well as zeolite-based ones [27,28]. The present metries for the other triacylglycerides showed a similar structure.
study is willing to contribute to a clear understanding of the preference
of the molecules to react at given T or P without the presence of any compounds of TAGs, since it is well-known that the non-symmetric
catalytic entity. A catalyst would accelerate a process to a desired TAGs are most likely present in vegetable oils or animal fats [29,30].
product with the intrinsic limitation due to the thermodynamic pre- The electronic properties showed that they are very similar on the
ference of the molecules. HOMO and LUMO energies. Production of FAs from TAGs in hydrogen
atmosphere was studied through direct fatty acid elimination (DFAE)
3.2. Triacylglycerides route, shown in Fig. 6. The values for the TAGs are shown in Table 3.
The ΔHr0 and ΔG r0 for the TAGs do not vary significantly by changing
TAGs were also studied by theoretical calculations. They are the raw the aliphatic chain, similar as for the chemical reactions studied for the
materials for the fatty acids from biomass. They consist of three fatty FAs. The process for breaking TAGs and for the production of FAs and
acids bonded to a propane moiety. The optimized geometry for tri- propane is an exothermic and exergonic process with ΔHr0 = −285
palmitin is shown in Fig. 5. The optimized geometry exhibited a non- kJmol-1 and ΔGr0 = −345 kJmol-1. Therefore, as model compounds,
linear structure, different than its usual 2D representation. We studied the symmetric or non-symmetric structures would not change the re-
several TAGs, most of them symmetric and one non-symmetric (com- action energies under H2 environments. There is a slight difference for
posed by C14-C16-C18 units of FA). We chose those molecules as model

41
D. Valencia et al. Biomass and Bioenergy 112 (2018) 37–44

Fig. 6. Overall reaction of the direct fatty acid elimination (DFAE)

route of the triacylglycerides under hydrogen atmosphere.

Table 3 conditions, namely carboxylic partial reduction (CPR) route. The ther-
HOMO (εHOMO), LUMO (εLUMO) energies and their gap for selected triacylglycerides. modynamic quantities indicate that this route is less favorable com-
Enthalpy and Gibbs energies for the total FA direct elimination depicted in Fig. 5. Values
pared to the DFAE route. The reaction of the initial TAG to form the
are given at standard temperature and pressure (298 K and 1 atm).
respective alcohol intermediate has a small reaction energy value
Triacylglyceride εHOMO εLUMO ΔH-L Reaction energy (ΔHr0 = −71.2 kJmol-1 and ΔGr0 = −54.2 kJmol-1) compared to the
first step of DFAE. It is important to mention that the value of ΔHr0 is
(eV) (eV) (eV) ΔHr0 (kJmol−1) ΔGr0 (kJmol−1) higher than the respective of ΔGr0. The same profile is observed for the
Tricaprylin −7.99 −0.87 7.12 −283.7 −344.2 following reduction reaction within this reaction network.
Tricaprin −8.01 −0.88 7.13 −283.6 −340.1 Hydrogenation of the C=O bond is thermodynamically possible, but
Trilaurin −7.97 −0.88 7.09 −285.4 −347.3 less than the C-C scission in the DFAE route. The values of the ther-
Trimyristin −7.98 −0.87 7.11 −286.0 −345.5 modynamic quantities have the same trend ΔHr0 > ΔGr0. Totally re-
Tripalmitin −7.97 −0.88 7.09 −284.4 −346.6
duced TAG product still contains oxygen atoms in its structure and yet
Tristearin −7.98 −0.87 7.11 −286.0 −351.6
C14-C16-C18 −7.98 −0.88 7.10 −284.5 −342.9 no hydrocarbon molecules have been formed.
The last reaction network studied was the scission of the C-O bond
to produce a fatty alcohol and propane, we called this process the fatty
tristearin ΔGr0 = −351.6 kJmol-1, but within the range of the studied alcohol elimination (FAE) route. All reactions were exothermic and
TAGs. A quick comparison between the reaction energies for the TAGs exergonic, which means those could take place at certain reaction
and the FAs suggested that the first reaction from vegetable oils or conditions. More importantly, the scission of C-O bond is more favor-
animal fats under HDO conditions are much more exothermic and ex- able than the reduction of the C=O bond. The first alcohol elimination
ergonic. Those higher values on their reaction energies indicate that is the most favorable reaction with ΔHr0 = −78.7 kJmol-1 and
chemical reactions of the TAGs would be much more favorable than ΔGr0 = −102.3 kJmol-1. The second step is also favored with
their further oxygen elimination from their FAs products. It could also ΔHr0 = −62.7 kJmol-1 and ΔGr0 = −79.6 kJmol-1. Final cleavage of C-
have consequences on the overall exothermic processes within a reactor O bond is energetically more preferred than the second step
that can alter the thermal or chemical parameters. ΔHr0 = −68.1 kJmol-1 and ΔGr0 = −83.4 kJmol-1. The overall reac-
The reaction energies for the FAs and TAGs do not show drastic tion forms three molecules of the caprylic alcohol and propane.
changes due to the variation of the size of the molecules, but a clear Caprylic alcohol could be further reduced with energy values as shown
trend depending on the type of chemical reaction was observed. To the in Fig. 2 (R4).
best of our knowledge, reaction networks for the TAGs under hydrogen The overall chemical reaction of the routes described in Fig. 7 are
atmosphere have been suggested based on products distribution, but thermodynamically favorable, but the highest values were found for the
not fully elucidated [31,32]. Isolation and purification of all the inter- DFAE route. The overall enthalpy and Gibbs energy of the reactions
mediate molecules are very complicated and difficult, but a detailed were computed by varying temperature and pressure to have a better
reaction network is very necessary. Those schemes would provide im- overview of the changes within those quantities (Fig. 8). The effect of
portant information in terms of the chemical characteristics and reac- the temperature on the three routes is almost the same on the ΔHr,
tion energy involved in the TAGs transformation. Then, we conducted a higher temperatures would increase the negative value of enthalpy. The
detailed study for tricaprylin reaction networks under H2 atmosphere at case for ΔGr as a function of temperature showed different trend for the
standard conditions (Fig. 7). Three different reaction networks were CPR route, it becomes less exergonic at higher temperature. This
studied from the TAGs to its direct C-C bond scission between FAs and change in the profile of ΔGr as a function of the temperature is desirable
propane and its partial reduction. We focused the attention to get as the formation of an intermediate would be reduced and the pro-
theoretically intermediate steps of TAGs under H2 atmosphere for a duction of free fatty acids or alcohols increases at higher temperatures.
better understanding of the biomass processing. The effect of pressure was also evaluated. ΔHr values do not change
TAGs direct scission of the C-C bond produces FAs and propane with the pressure, similar as for the FAs, this could indicate there is not
through the DFAE route. The first FA elimination from the TAG is a dependence on this property at the studied range or the theoretical
exothermic (ΔHr0 = −113.3 kJmol-1) and exergonic (ΔGr0 = −137.5 model has limitations for such property. Nevertheless, the values of ΔGr
kJmol-1) and it leaves a molecule of FA and diacylpropane (DAP). The showed differences by the type of chemical reaction. DFAE and FAE
second cleavage of the C-C bond between FA and DAP is also an exo- routes that involves a C-O bond scission (hydrogenolysis reactions)
thermic (ΔHr0 = −94.7 kJmol-1) and exergonic reaction showed no variation on the ΔGr values along the range of pressure
(ΔGr = −117.1 kJmol ), which produces FA and monoacylpropane
0 -1
studied. The CPR route changed their values depending on the pressure.
(MAP). The last reaction gives propane and FA with energy values of They indicate that increasing the pressure, more of the reduced pro-
ΔHr0 = −75.7 kJmol-1 and ΔGr0 = −89.6 kJmol-1. DFAE from trica- ducts will be formed. Similar dependence of ΔGr values due to pressure
prylin revealed information and intermediate products from TAG re- was found for the FAs, the values of this property for the CPR route
action under H2 atmosphere. The first reaction is the most thermo- reach a maximum at about 50 atm.
dynamic favorable and the last step the less one. Therefore, TAG Conversion of TAGs to FAs can be achieved by DFAE route mostly at
subjected to H2 environments reacts towards DFAE with high reaction higher temperatures. The purpose of HDO is the production of O-free
energies. hydrocarbons from TAGs, FAs or their alcohols. All the routes of TAGs
Secondly, we studied the reduction of the C=O bond in the TAG, a under H2 environments are feasible at higher temperatures and the
possible scenario, which would take place within certain reaction effect of pressure on the thermodynamic quantities is not drastic. The

42
D. Valencia et al. Biomass and Bioenergy 112 (2018) 37–44

Fig. 7. Reaction network of TAGs under H2 environments. (A) Direct fatty acid elimination (DFAE) route, where TAG reacts with H2 to release FAs and propane, (B) Carboxylic partial
reduction (CPR) route, where the carboxylic groups within TAG molecule is reduced to form an ether group and water and (C) Fatty alcohol elimination (FAE) route, the reduced molecule
of TAG under H2 eliminates a fatty alcohols, by the cleavage of the C-O bond, and propane. All the reaction energies correspond to tricaprylin and related products. Values are given in
kJmol−1.

Fig. 8. Effect of temperature and pressure on the reaction energy of tricaprylin for direct fatty acid elimination (DFAE) ( ), carboxylic partial reduction (CPR) ( ) and fatty alcohol
elimination (FAE) ( ) routes. (a) ΔHr (T, 40 atm), (b) ΔGr (T, 40 atm), (c) ΔHr (473 K, P) and (d) ΔGr (473 K, P).

43
D. Valencia et al.
CPR route is less exergonic at higher temperatures and lower pressures.
More specifically, the reduction of C=O moiety within the TAGs is
higher at lower temperatures and higher pressures, while C-O bond
scission is more exergonic at higher temperatures and no important
effects of pressure were found. Compared to the reduction of the C=O
moiety in the FAs, its behavior seems to be very similar as the one
observed for the TAGs. It is less favored at higher temperatures and
more favorable at higher pressures. The reaction energies for hydro-
genolysis reactions showed similar profiles for either TAGs or FAs as a
function of temperature or pressure. These results might indicate a di-
rect correlation between the chemical reaction and their energy reac-
tions as a function of pressure or temperature. Future studies can take
these thermodynamic properties to selectively direct the chemical re-
actions of TAGs or FAs to specific products by tuning their different
routes under H2 environments. Or as inputs for a chemical engineering
design in specialized softwares.
In the context of clean energy and sustainability, vegetable oils or
animal fats are very promising biomass feedstocks for the production of
new biofuels. The reaction routes of FAs or TAGs with hydrogen are
thermodynamically possible under the studied range of T and P. From a
given value of T and P, the TAGs could react towards DFAE, CPR or FAE
routes, while the respective FAs are subjected to the same T and P
conditions and they can undergo to the HDO or thermal decomposition
routes. The present reaction networks of the biomass-derived products
might be important to understand their chemical reactions at different
operating conditions. Their intermediate products to form free fatty
acids or hydrocarbons involve complex chemical structures that must
be further transformed. The development of new technical tools for
biomass processing might be interested in this chemical and thermo-
dynamic data.
4. Conclusions
A systematic study was conducted to provide a better understanding
of the chemical reactions of fatty acids and triacylglycerides under
hydrogen atmosphere. ΔHr0 and ΔGr0 values of several FAs or TAGs do
not vary due to the size of the molecule. Thermodynamic properties for
decaboxylation, decarbonylation and HDO processes indicate that those
can be properly tuned by temperature or pressure changes. On the other
hand, the chemical reactions of TAGs are much more favorable than the
oxygen elimination from their FAs products. TAGs reaction network
under H2 atmosphere has been suggested for three different possible
scenarios. Direct fatty acid elimination or fatty alcohol elimination
routes are favored at higher temperatures and the effect of pressure on
those processes is almost negligible. The effect of temperature and
pressure on the ΔGr0 values showed an inverse trend between hydro-
genation and hydrogenolysis reactions. The carboxylic partial reduction
is more exergonic at lower temperatures and higher pressures. This
comprehensive study showed the different preferences of FAs or TAGs
to react towards a specific route at a given condition.
Acknowledgements
The authors wish to thank Project Y.61023 CEMIE BIO Clúster
bioturbosina (Instituto Mexicano del Petróleo), and Laboratorio de
Visualización y Cómputo Paralelo at UAM–I for the access to their
computer facilities.
References
[1] S. Chu, A. Majumdar, Opportunities and challenges for a sustainable energy future,
Nature 488 (2012) 294–303.
[2] N.L. Panwar, S.C. Kaushik, S. Kothari, Role of renewable energy sources in en-
vironmental protection: a review, Renew. Sustain. Energy Rev. 15 (2011)
1513–1524.
[3] A. Corma, G.W. Huber, L. Sauvanaud, P. O´Connor, Processing of biomass derived
44
View publication stats
Biomass and Bioenergy 112 (2018) 37–44
oxygenates in the oil refinery: catalytic cracking (FCC) pathways and role of cata-
lysts, J. Catal. 247 (2007) 307–327.
[4] L.R. Lynd, C.E. Wyman, T.U. Gerngross, Biocommodity engineering, Biotechnol.
Prog. 15 (1999) 777–793.
[5] C.E. Wyman, N.D. Hinman, Fundamentals of production from renewable feedstocks
and use as a transportation fuel, Appl. Biochem. Biotechnol. 24 (1990) 735–753.
[6] P. Tamunaidu, S. Bhatia, Catalytic cracking of palm oil for the production of bio-
fuels: optimization studies, Bio Technol. 98 (2007) 3593–3601.
[7] M. Mohammad, T.K. Hari, Z. Yaakob, Y.C. Sharma, K. Sopian, Overview on the
production of paraffin based-biofuels via catalytic hydrodeoxygenation, Renew.
Sustain. Energy Rev. 22 (2013) 121–132.
[8] P.M. Mortensen, J.-D. Grunwaldt, P.A. Jensen, K.G. Knudsen, A.D. Jensen, A review
of catalytic upgrading of bio-oil to engine fuels, Appl. Catal., A 407 (2011) 1–19.
[9] E. Furimsky, Catalytic hydrodeoxygenation, Appl. Catal., A 199 (2000) 147–190.
[10] M.J. Girgis, B.C. Gates, Reactivities, reaction networks, and kinetics in high-pres-
sure catalytic hydroprocessing, Ind. Eng. Chem. Res. 30 (1991) 2021–2058.
[11] T.V. Choudhary, C.B. Phillips, Renewable fuels via catalytic hydrodeoxygenation,
Appl. Catal., A 397 (2011) 1–12.
[12] M. Toba, Y. Abe, H. Kuramochi, M. Osako, T. Mochizuki, Y. Yoshimura,
Hydrodeoxygenation of waste vegetable oil over sulfide catalysts, Catal. Today 164
(2011) 533–537.
[13] G.W. Huber, A. Corma, Synergies between bio- and oil refineries for the production
of fuels from biomass, Angew. Chem. Int. Ed. 46 (2007) 7184–7201.
[14] A.A. Lappas, S. Bezergianni, I.A. Vasalos, Production of biofuels via co-processing in
conventional refining processes, Catal. Today 145 (2009) 55–62.
[15] E. Furimsky, Hydroprocessing challenges in biofuels production, Catal. Today 217
(2013) 13–56.
[16] J. Horácek, D. Kubicka, Bio-oil hydrotreating over conventional CoMo & NiMo
catalysts: the role of reaction conditions and additives, Fuel 198 (2017) 49–57.
[17] K. Wang, X. Zhang, J. Zhang, Z. Zhang, C. Fan, P. Han, Theoretical study on free
fatty acid elimination mechanism for waste cooking oils to biodiesel over acid
catalyst, J. Mol. Graph. Model. 66 (2016) 41–46.
[18] S. Lestari, P. Maki-Arvela, J. Beltramini, G.Q. Max Lu, D.Y. Murzin, Transforming
triglycerides and fatty acids into biofuels, ChemSusChem 2 (2009) 1109–1119.
[19] M.F. Wagenhofer, E. Baráth, O.Y. Gutiérrez, J.A. Lercher, Carbon–carbon bond
scission pathways in the deoxygenation of fatty acids on transition-metal sulfides,
ACS Catal. 7 (2017) 1068–1076.
[20] Q. Smejkal, L. Smejkalová, David Kubicka, Thermodynamic balance in reaction
system of total vegetable oil hydrogenation, Chem. Eng. J. 146 (2009) 155–160.
[21] Y. Zhao, D.G. Truhlar, The M06 suite of density functionals for main group ther-
mochemistry, thermochemical kinetics, noncovalent interactions, excited states,
and transition elements: two new functionals and systematic testing of four M06-
class functionals and 12 other functionals, Theor. Chem. Acc 120 (2008) 215–241.
[22] Y. Zhao, D.G. Truhlar, Density functionals with broad applicability in chemistry,
Acc. Chem. Res. 41 (2008) 157–167.
[23] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb, J.R. Cheeseman,
G. Scalmani, V. Barone, B. Mennucci, G.A. Petersson, H. Nakatsuji, M. Caricato,
X. Li, H.P. Hratchian, A.F. Izmaylov, J. Bloino, G. Zheng, J.L. Sonnenberg, M. Hada,
M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda,
O. Kitao, H. Nakai, T. Vreven, J.A. Montgomery Jr., J.E. Peralta, F. Ogliaro,
M. Bearpark, J.J. Heyd, E. Brothers, K.N. Kudin, V.N. Staroverov, R. Kobayashi,
J. Normand, K. Raghavachari, A. Rendell, J.C. Burant, S.S. Iyengar, J. Tomasi,
M. Cossi, N. Rega, J.M. Millam, M. Klene, J.E. Knox, J.B. Cross, V. Bakken,
C. Adamo, J. Jaramillo, R. Gomperts, R.E. Stratmann, O. Yazyev, A.J. Austin,
R. Cammi, C. Pomelli, J.W. Ochterski, R.L. Martin, K. Morokuma, V.G. Zakrzewski,
G.A. Voth, P. Salvador, J.J. Dannenberg, S. Dapprich, A.D. Daniels, Ö. Farkas,
J.B. Foresman, J.V. Ortiz, J. Cioslowski, D.J. Fox, Gaussian 09, Revision B.01,
Gaussian, Inc., Wallingford CT, 2009.
[24] D. Valencia, J. Aburto, I. García-Cruz, Electronic structure and mesoscopic simu-
lations of nonylphenol ethoxylate surfactants. A combined DFT and DPD study,
Molecules 18 (2013) 9441–9450.
[25] E. Santillán-Jiménez, M. Crocker, Catalytic deoxygenation of fatty acids and their
derivatives to hydrocarbon fuels via decarboxylation/decarbonylation, J. Chem.
Technol. Biotechnol. 87 (2012) 1041–1050.
[26] M.W. Schreiber, D. Rodríguez-Niño, O.Y. Gutiérrez, J.A. Lercher,
Hydrodeoxygenation of fatty acid esters catalyzed by Ni on nano-sized MFI type
zeolites, Catal. Sci. Technol 6 (2016) 7976–7984.
[27] J. Horacek, D. Kubicka, Bio-oil hydrotreating over conventional CoMo & NiMo
catalysts: the role of reaction conditions and additives, Fuel 198 (2017) 49–57.
[28] F. Gong, Z. Yang, C. Hong, W. Huang, S. Ning, Z. Zhang, Y. Xu, Q. Li, Selective
conversion of bio-oil to light olefins: controlling catalytic cracking for maximum
olefins, Bioresour. Technol. 102 (2011) 9247–9254.
[29] J. Gresti, M. Bugaut, C. Maniongui, J. Bezard, Composition of molecular species of
triacylglycerols in bovine milk fat, J. Dairy Sci. 76 (1993) 1850–1869.
[30] M.J. Ramos, C.M. Fernández, A. Casas, L. Rodríguez, A. Pérez, Influence of fatty
acid composition of raw materials on biodiesel properties, Bioresour. Technol. 100
(2009) 261–268.
[31] L.M. Orozco, D.A. Echeverri, L. Sánchez, L.A. Ríos, Second-generation green diesel
from castor oil: development of a new and efficient continuous-production process,
Chem. Eng. J. 322 (2017) 149–156.
[32] A.F.H. Studentschnig, T. Telser, S. Schober, M. Mittelbach, Hydrotreating of non-
food feedstocks over Raney nickel for the production of synthetic diesel fuel,
Biofuels 7 (2016) 279–287.