Novel Full Hydrogenation Reaction of Methyl Esters of Palm Kernel and Sunflower Oils Into Methyl Stearate Catalyzed by Rhodium, Ruthenium and Nickel Complexes of Bidentate Hexasulfonated o-Phenylendiphosphite Ligands

Article abstract of DOI:10.1007/s10562-018-2642-7

High catalytic activities (TOF = 8680?h^?1) have been achieved by novel rhodium catalysts modified with the chelating sulfonated phosphite ligand hexasulfonated o-phenylendiphosphite in the hydrogenation reaction of renewable methyl esters of sunflower oil under mild reaction conditions and a low rhodium concentration of 50?ppm in methanol. Graphical Abstract: [Figure not available: see fulltext.].

Full text of DOI:10.1007/s10562-018-2642-7

Catalysis Letters
https://doi.org/10.1007/s10562-018-2642-7
Novel Full Hydrogenation Reaction of Methyl Esters of Palm Kernel
and Sunflower Oils Into Methyl Stearate Catalyzed by Rhodium,
Ruthenium and Nickel Complexes of Bidentate Hexasulfonated o-
Phenylendiphosphite Ligands
Niki Kaika1 · Christina Panopoulou1 · Eleni Anagnostopoulou1 · Christos Fakas1 · Panagiotis Lilas1 ·
Dimitra Stavroulaki1 · Georgios Papadogianakis1
Received: 18 September 2018 / Accepted: 18 December 2018
© Springer Science+Business Media, LLC, part of Springer Nature 2019
Abstract
High catalytic activities (TOF=8680 h−1) have been achieved by novel rhodium catalysts modified with the chelating sul-
fonated phosphite ligand hexasulfonated o-phenylendiphosphite in the hydrogenation reaction of renewable methyl esters of
sunflower oil under mild reaction conditions and a low rhodium concentration of 50 ppm in methanol.
Graphical Abstract
O
Isomerization
C18:2 isomer mixture
OCH3
Methyl linoleate (ML)
C18:2 (9c, 12c)
SO3- +NHR3
R3HN+ -O3
S
O
Rh
O
O
P
P
O
O
O
SO3- +NHR3
+ H
SO3- +NHR3
2
SO3- +NHR3
R3HN+ -O3
S
O
Isomerization
.
C18:1 isomer mixture
OCH3
Methyl oleate (MO)
C18:1 (9c)
SO3- +NHR3
R3HN+ -O3
S
O
Rh
O
P
P
O
O
SO3- +NHR3
O
O
+ H
2
SO3- +NHR3
SO3- +NHR3
R3HN+ -O3
S
O
OCH3
Methyl stearate (MS), C18:0
Extended author information available on the last page of the article
Vol.:(0123456789)
1 3

N. Kaika et al.
Keywords Hydrogenation · Methyl esters of palm kernel and sunflower oil · Rhodium, ruthenium and nickel · Bidentate
sulfonated phosphite · Zero trans-fats
In 1989, one of us in cooperation with Fell and Bahr-
mann [56–59] has developed a novel class of ligands i.e.
the sulfonated phosphites which are stable systems in
hydrolysis side-reactions. Rhodium complexes modified
with sulfonated phosphites were used as catalysts in hydro-
formylation reactions of olefins under mild reaction condi-
tions in organic solvents and exhibited higher selectivities
towards the desired linear aldehydes compared to rhodium
complexes modified with conventional triphenylphosphite
or triphenylphosphine ligands under the same reaction con-
ditions [56–59]. Such monodentate sulfonated phosphite
ligands have been also used by Favre et al. [60] to modify
rhodium catalysts for the hydroformylation of olefins in ionic
liquids/organic two-phase systems and the Rh/monodentate
sulfonated phosphite catalyst could be easily separated in the
ionic liquid phase containing the catalyst from the organic
phase containing the products by a simple phase separation.
Recycling experiments proved that the high n/iso-ratios of
the aldehydes obtained with Rh/monodentate sulfonated
phosphite catalysts remained high in two consecutive runs
and that the n/iso-ratios of aldehydes with Rh/monodentate
sulfonated phosphite catalysts were also higher compared to
their corresponding rhodium catalysts modified with conven-
tional phosphine ligands [60].
1 Introduction
Nowadays, renewable vegetable and tropical oils with their
derivatives play an important role in the development of
Green/Sustainable Chemistry because they are feedstocks
for biorefineries with a broad spectrum of applications such
as in foodstuff chemistry, detergents, cosmetics, pharmacy,
plastics, biolubricants and in the energy production with the
manufacture of biofuels with excellent fuel properties thus
contributing to an effective management of greenhouse gas
emissions and because of the depleting trend of conventional
fossil feedstocks [1–10]. Vegetable oils are used for the pro-
duction of 1st generation biodiesel fuel by transesterification
reactions with methanol to fatty acid methyl esters (FAMEs)
and in hydrotreating reactions in conventional oilrefineries
to produce bio gas oil which is a mixture of mainly C11–C18
linear and branched paraffins with a typical gas oil boil-
ing range and belongs to the 2nd generation biodiesel fuel
[1–10]. Full hydrogenation reactions of renewable polyun-
saturated methyl esters of tropical and vegetable oils into
their saturated counterpart product namely methyl stearate
(MS) are industrially interesting conversions for the pro-
duction of surfactants, emulsifiers, cosmetics, emollients,
solid soaps, lubricants and plasticizers [11–14]. Further-
more, hydrogenation reactions of FAMEs to MS are useful
model reactions for studying the full hydrogenation reac-
tion of edible oils into saturated triglycerides which are
further subjected to interesterification reactions with liquid
vegetable oils to yield shortenings, margarines and other
foodstuffs containing zero amounts of trans-fats [13–33]. In
the early 1990s, it was reported that there is a direct associa-
tion between consumption of trans-fats and cardiovascular
disease because of the higher concentrations of LDL-choles-
terol in the plasma [34–52]. Due to these reports new regu-
lations have been introduced worldwide in order to restrict
trans-fats consumption which resulted to a demand for food
products containing zero amounts of trans-fats [34–52].
Since 1911, the industrial partial hydrogenation process of
edible oils makes use of heterogeneous nickel catalysts pro-
ducing high amounts (up to 50%) of undesired trans-fats due
to cis/trans geometric isomerization reactions [15–20, 53,
54]. Nowadays, the edible oils hydrogenation industry pre-
fers the route of full hydrogenation of oils to hard fats with
subsequent interesterification with liquid vegetable oils to
avoid the formation of trans-fats [21–33, 40, 41, 43, 47, 55].
Thus, there is a need for the development of a novel highly
selective industrial partial hydrogenation process of edible
oils to foodstuffs with a zero content of trans-fats.
Recently, we reported [14] the hydrogenation of methyl
esters of palm kernel and sunflower oils to the saturated
methyl stearate catalyzed by rhodium and ruthenium com-
plexes of monodentate sulfonated triphenylphosphite ligands
in the absence or presence of organic solvents. We now
report an expansion of our work on the use of sulfonated
phosphites with the development of the novel full hydrogen-
ation reaction of C=C units of renewable polyunsaturated
methyl esters of palm kernel and sunflower oils into their sat-
urated counterpart methyl stearate (MS) catalyzed by highly
active rhodium, ruthenium and nickel complexes with biden-
tate hexasulfonated o-phenylendiphosphite ligands under
mild reaction conditions in homogeneous systems [13]. The
full hydrogenation of polyunsaturated methyl esters of palm
kernel and sunflower oils into MS are useful model reac-
tions for studying the full hydrogenation reaction of edible
oils into saturated triglycerides which are further subjected
to interesterification reactions with liquid vegetable oils to
yield foodstuffs containing zero amounts of trans-fats. MS
could be also used as starting material for industrial catalytic
hydrogenolysis processes into stearyl alcohol which is an
important industrial fatty alcohol and is further proceeded
for the manufacture of surfactants, emulsifiers, cosmetics,
emollients, solid soaps, lubricants and plasticizers. The
bulkiness of the transition metal bidentate hexasulfonated
1 3

Novel Full Hydrogenation Reaction of Methyl Esters of Palm Kernel and Sunflower Oils Into Methyl…
o-phenylendiphosphite anionic catalytic system bearing
large triisooctylammonium counter cations could offer the
possibility of the easy separation of the catalysts from the
reaction mixtures by means of a membrane [13]. To our
knowledge, this is the first example of a hydrogenation reac-
tion catalyzed by transition metals modified with a chelating
sulfonated phosphite.
the aqueous solvent was saturated with argon. To remove
oxygen the procedure was repeated three times. The renew-
able starting materials methyl esters of palm kernel oil
(MEPKO, Edenor® ME PK 12–18 F) and methyl esters of
sunflower oil (MESO, Sunflower Fatty Acid ME®) were
supplied by Cognis GmbH (today BASF) and used without
any further purification. Aluminium oxide 90 active neutral
(70–230 mesh) was purchased from Merck. Sunflower oil
from Helianthus annuus was purchased from Fluka. Methyl
heptadecanoate was purchased from Fluka.
2 Experimental
2.1 Materials
2.2 Synthesis of the Hydrolysis Stable Bidentate
Hexasulfonated o‑Phenylendiphosphite
Triisooctylammonium Salt (HSPDP) Ligand
Hydrogen (quality 5.0) was purchased from Air Liquide Hel-
las A.E.B.A. (Athens) and was used without further purifi-
cation. RhCl3·3H2O was purchased from Acros Organics,
RuCl3·3H2O and NiCl2·6H2O were obtained from Alfa Aesar
and all three were used as received. 3,5-pyrocatecholdisul-
fonic acid disodium salt monohydrate (1,2-Dihydroxyben-
zene-3,5-disulfonic acid disodium salt monohydrate, Tiron
monohydrate) was purchased from Alfa Aesar. Triisoocty-
lamine and triphenylphosphite were purchased from Sigma-
Aldrich. Sulfuric acid was purchased from Fluka. Methanol
and toluene were purchased from SDS, diethyl ether from
Sigma-Aldrich and methyl acetate from Merck-Schuchardt.
All organic solvents before use were dried through activated
molecular sieves 4 Å which were purchased from SDS and
used after filtration over 0.2 µm filter unit (Millex-FG of
Millipore). Demineralized water was deoxygenated in an
ultrasound bath under vacuum for 2 h. During the deoxy-
genation, the flask was disconnected from the vacuum, and
The pathway for the synthesis of hexasulfonated o-phe-
nylendiphosphite triisooctylammonium salt (HSPDP)
ligand is shown in Fig. 1. Addition of 2 M aqueous H2SO4 to
3,5-pyrocatecholdisulfonic acid disodium salt monohydrate
(1,2-Dihydroxybenzene-3,5-disulfonic acid disodium salt
monohydrate, Tiron monohydrate) aqueous solution yields
3,5-pyrocatecholdisulfonic acid. Reaction of 3,5-pyrocat-
echoldisulfonic acid in water with triisooctylamine (TiOA)
in toluene in an aqueous/organic two-phase system forms
3,5-pyrocatecholdisulfonic acid triisooctylammonium salt
in the organic toluene phase. Subsequent separation of
the upper organic phase, drying and transesterification of
3,5-pyrocatecholdisulfonic acid triisooctylammonium salt
with triphenylphosphite gives (after removal of phenol by
distillation) the HSPDP ligand with a yield of 66.0 mol%.
NaO3S
HO3S
OH
R3HN+ -O3S
OH
OH
OH
H2SO4
+
NR
3
+
2
Na2SO4
OH
OH
SO3H
SO3- +NHR3
SO3Na
SO3- +NHR3
R3HN+ -O3S
R3HN+ -O3S
3
O
O
O
O
OH
OH
[NR3]
PhOH
P
P
P(OPh)
2
+
3
6
SO3- +NHR3
O
O
SO3- +NHR3
SO3- +NHR3
R3HN+ -O3S
HSPDP
Fig. 1 Synthesis of the hexasulfonated o-phenylendiphosphite triisooctylammonium salt (HSPDP) ligand
SO3- +NHR3
( R = C8H17
)
1 3

N. Kaika et al.
2.2.1 Preparation Procedure for the HSPDP Ligand
101.0 g methanol containing 1.61 g of NaOH. The mixture
was stirred at 55–58 °C for 2 h. The course of the reaction
was followed by thin layer chromatography. The reaction
mixture was then cooled and further stirred for 1 h at room
temperature. After phase separation resulted in the isola-
tion of the methyl esters of sunflower oil (MESO1) and the
glycerol. The glycerol phase (bottom layer) was removed and
kept in a separate container. The MESO1 phase (top layer)
was washed with distilled water several times (pH 5.7), dried
over Na2SO4 to obtain 110.4 g of MESO1 mixture which
was used as starting material in the Ni/HSPDP-catalyzed
hydrogenation reaction without further purification. The
composition of MESO1 was determined by GC and is given
in Table 2.
A 2000-ml five-necked, round-bottom flask, equipped with
a mechanical stirrer, a thermometer, a dropping funnel and
a condenser, which was previously evacuated, heated by a
heat gun and filled at room temperature with argon, was
charged with 332.23 g (1.0 mol) 3,5-pyrocatecholdisulfonic
acid disodium salt monohydrate (1,2-Dihydroxybenzene-
3,5-disulfonic acid disodium salt monohydrate) which was
dissolved under stirring in 600 ml of deoxygenated dem-
ineralized water. Through the dropping funnel were added
400 ml of a 2M aqueous H2SO4 solution to the cooled
1,2-Dihydroxybenzene-3,5-disulfonic acid disodium salt
aqueous solution using an ice/water bath in a period of 2 h
under intensive stirring. A mixture of 707.36 g (2.0 mol) tri-
isooctylamine (TiOA) in 800 ml of dried toluene was added
dropwise through the dropping funnel and stirred intensively
overnight at room temperature. After phase separation the
aqueous phase was discarded and the upper organic toluene
phase containing the 3,5-pyrocatecholdisulfonic triisoocty-
lammonium salt was dried over activated Na2SO4 for several
hours. After separation of Na2SO4 by filtration, the 3,5-pyro-
catecholdisulfonic triisooctylammonium salt/toluene organic
phase, was heated under reflux in a Dean–Stark apparatus
overnight to remove last amounts of water from the organic
mixture. A solution of 206.8 g (0.66 mol) triphenylphophite
and 10 g (0.028 mol) TiOA in 600 ml dried toluene was
added dropwise to the 3,5-pyrocatecholdisulfonic triisoocty-
lammonium salt/toluene mixture under stirring at 150 °C for
3 h. The condenser was than replaced by a distillation unit.
Toluene was removed by distillation at 110 °C at first. The
distillation of phenol, which is the byproduct of the transes-
terification reaction, was followed at 30 °C under 1 torr vac-
uum within a period of 70 h to give an amount of 124 g of
removed phenol. The yield of the transesterification reaction,
based on the total amount of removed phenol by distillation,
was 66% and any amounts of unreacted triphenylphosphite
could not be removed by distillation at higher temperatures
up to 70 °C under 0.01 torr vacuum from the mixture and
a yellow, highly viscous liquid remained which was the
bidentate hexasulfonated o-phenylendiphosphite triisoocty-
lammonium salt (HSPDP) ligand. The HSPDP ligand was
analysed by 31P{1H}NMR (121 MHz, referenced to external
85% H3PO4) on a Varian Unity Plus 300/54 spectrometer in
CDCl3 at 25 °C. δ: HSPDP=+130.6 ppm; impurities from
triphenylphosphite=+128.9 ppm.
2.4 Typical Hydrogenation Procedure
The autoclave was thoroughly cleaned and followed by
series of treatment of the autoclave at elevated tempera-
ture (100 °C) and pressure (40 bar of H2) within 1 h each
time in the presence of HSPDP/methanol and the absence
of any transition metals in order to be sure that no memory
effects of the autoclave regarding previous transition metal
catalytic systems are still operative. The hydrogenation
reactions of C=C units of methyl esters of palm kernel and
sunflower oils were than performed in the presence of Rh-
and Ru-HSPDP complexes in organic solvents and typical
hydrogenation reaction conditions were given below. The
Rh- or Ru-HSPD catalyst precursor was first synthesized
by dissolving the amount of HSPDP ligand in 10 ml of
dried organic solvent and complexation with RhCl3·3H2O,
RuCl3·3H2O or NiCl2·6H2O under argon. The metal/HSPDP
catalyst precursor solution was transferred into an Autoclave
Engineers autoclave of a nominal volume of 100 ml which
was previously evacuated and filled with argon together with
the amount of methyl ester of palm kernel or sunflower oils.
After a number of pressurising-depressurising cycles with
hydrogen to remove last traces of air oxygen, the autoclave
was pressured by H2 and contents were heated from 70 up to
170 °C with stirring using a stir bar driven by an IKA mag-
netic stirrer (poorer mixing) with a stirring rate of 750 rpm.
At the reaction temperatures the hydrogen pressures were
from 10 up to 50 bar. After the reaction the autoclave was
cooled to room temperature, depressurized through a vent
and the hydrogenation reaction mixture was removed. The
samples of the reaction mixtures after filtration over neutral
alumina and addition of methyl heptadecanoate as stand-
ard were analyzed by gas chromatography (GC) and the
obtained results are given in Tables 1 and 2. Emphasis has
been placed in all reaction described in Tables 1 and 2 in
order to obtain reproducible results.
2.3 Transesterification Reaction of Sunflower Oil
with Methanol to Obtain MESO1
A 500-ml, three-necked, round-bottom flask, equipped with
a mechanical stirrer, a thermometer and a condenser was
charged with 161.0 g of sunflower oil and a solution of
1 3

Novel Full Hydrogenation Reaction of Methyl Esters of Palm Kernel and Sunflower Oils Into Methyl…
palmitate (C16:0) which were ignored in the calculations to
2.5 Analysis of the Starting Materials Methyl
Esters of Palm Kernel and Sunflower Oils
and the Products of the Hydrogenation
Reaction
adjust the C=C units/Metal molar ratios in the catalytic full
hydrogenation reactions and probably act as a second solvent
for the homogeneous catalytic system together with the
added organic solvent which was usually methanol. The full
hydrogenation of the unsaturated C18 FAMEs part of
MEPKO to yield the desired saturated product methyl stea-
rate (MS, C18:0) was studied in the presence of Rh/HSPDP
and Ru/HSPDP catalytic complexes and the obtained results
are summarized in Table 1. This study deals with the influ-
ence of operating reaction parameters such as temperature,
molar ratios of C=C units/Metal and HSPDP/Rh, hydrogen
pressure, reaction time, addition of different amounts of
methanol as a solvent and of various organic solvents at a
stirring rate of 750 rpm (Table 1). First, we investigated the
influence of reaction temperature on the catalytic activity
and selectivity towards MS (Table 1, entries1/1–1/6). The
catalytic activities and selectivities in the Rh/HSPDP-cata-
lyzed full hydrogenation of the polyunsaturated C18 esters
part of MEPKO towards the desired saturated product MS
increase with increasing temperature from 70 °C up to 90 °C
to give TOF values from 49 up to 110 per hour and selectivi-
ties from 28.0 mol% up to 48.6 mol% of MS at molar ratios
of C=C units/Rh=300 and HSPDP/Rh=1 and a hydrogen
pressure of 40 bar within 60 min of reaction time at a rho-
dium concentration of 50 ppm using 10 ml of methanol as a
solvent (Table 1, entries 1/1–1/3). Raising the reaction tem-
perature higher has a negative effect on both the catalytic
activity and selectivity to MS in the hydrogenation of the
unsaturated esters part of MEPKO to give at 120 °C only 93
TOFs per hour and 42.9 mol% of MS (Table 1, entry 1/6).
This effect that the catalytic activity increases with increas-
ing temperature and above a specific temperature the activity
decreases has been observed in several homogeneously cata-
lyzed hydrogenation reactions of FAMEs [62]. These results
were rationalized by assuming that above 90 °C the catalytic
active Rh-HSPDP key intermediate species are probably
destabilized in the MEPKO hydrogenation reaction mixture.
This may be caused due to the less polar nature of MEPKO
which consists of large amounts (>80 mol%) of low polarity
compounds such as the saturated C10:0-C16:0 FAMEs. The
effect of C=C units/Rh molar ratio in the Rh/HSPDP-cata-
lyzed full hydrogenation of MEPKO at 90 °C, 40 bar hydro-
gen, 60 min of reaction time, a molar ratio of HSPDP/Rh=1
and with the same amount of 10 ml of added methanol as a
solvent is presented in entries 1/7–1/10. A quantitative selec-
tivity towards MS of 100.0 mol% was obtained at a molar
ratio of C=C units/Rh=50 (entry 1/7) whereas at increasing
molar ratios of C=C units/Rh from 100 up to 400 the selec-
tivity to MS decreased from 69.7 mol% down to 34.2 mol%,
respectively (entries 1/8–1/10). The activity and selectivity
to MS in the Rh/HSPDP-catalyzed full hydrogenation of
MEPKO is influenced by the amount of HSPDP ligand
During the hydrogenation reaction an amount of the starting
material containing cis-olefinic bonds undergoes positional
and geometric isomerizations reactions to form several dif-
ferent positional and trans-isomers especially under the con-
ditions were the selectivity to methyl stearate was below
50 mol%. These various cis/trans- and positional FAME iso-
mers obtained as products in the mixture after the reaction
and also the various FAME compounds in the starting mate-
rials of methyl esters of palm kernel and sunflower oils were
identified by comparison of GC and GC/MS analytic data
with data for authentic samples. GC/MS was measured on a
Varian Star 3400CX GC coupled with a Varian Saturn 2000
ion trap MS and equipped with a flame ionization detector
(FID) and a SP-2560 capillary column (100 m× 0.25 mm
i.d. × 0.2 µm film thickness) which was purchased from
Supelco (Athens, Greece). The SP-2560 capillary column
is one of the two columns applied in the approved American
Oil Chemists’ Society (AOCS) official method Ce 1h-05 for
the determination of cis-, trans-, saturated, monounsaturated
and polyunsaturated fatty acids in vegetable or non-ruminant
animal oils and fats by capillary GLC method [61]. Carrier
gas was N2 at 230 kPa. The oven temperature was initially
set at 170 °C for 0 min and then increased to 220 °C with
a rate of 1 °C/min. The injector and detector temperatures
were set both at 220 °C. GC analyses were performed on a
Shimadzu GC-14B equipped with a FID detector and with
the SP-2560 capillary column and the conditions were the
same as described above in GC/MS analyses.
3 Results and Discussion
3.1 Full Hydrogenation Reactions of Unsaturated
Methyl Esters of Palm Kernel Oil (MEPKO)
Catalyzed by Rh‑ and Ru–HSPDP Complexes
The starting material MEPKO applied in the catalytic full
hydrogenation reactions contained only 15.2 mol% of C18
FAMEs with the proportion of 1.78 mol% methyl stearate
(MS, C18:0), 11.6 mol% methyl oleate [MO, C18:1 (9c)]
and 1.82 mol% methyl linoleate [ML, C18:2 (9c,12c)]. The
content of C18 FAMEs was than calculated to 100 mol% to
give 11.7 mol% of C18:0 ester, 76.3 mol% of C18:1 com-
pound and 12.0 mol% of the C18:2 ester (Table 1). The mix-
ture of MEPKO except the C18 FAMEs further contained
the following saturated fatty esters (Table 1): 0.2 mol%
methyl caprate (C10:0), 56.4 mol% methyl laurate (C12:0),
19.1 mol% methyl myristate (C14:0) and 9.1 mol% methyl
1 3

N. Kaika et al.
1 3

Novel Full Hydrogenation Reaction of Methyl Esters of Palm Kernel and Sunflower Oils Into Methyl…
1 3

N. Kaika et al.
added to rhodium precursor (entries 1/8, 1/11, 1/12). As
expected, the highest catalytic activity (TOF=58 h−1) and
therefore higher selectivity to the final product MS
(69.7 mol%) were obtained at the low HSPDP/Rh molar
ratio of 1 (entry 1/8) whereas at a higher ligand/metal molar
ratio the catalytic activity decreases to give at a ratio of
HSPDP/Rh=3 a catalytic activity of 12 TOFs per hour and
a selectivity towards MS of 23.5 mol% (entry 1/12). This
lower catalytic activity at higher HSPDP/Rh molar ratios
could probably be rationalized by assuming that a competi-
tion between the free HSPDP ligand and the C=C units of
unsaturated FAME for a coordination site on rhodium takes
place which may lead to a retardation in the activation of the
MEPKO full hydrogenation reaction. The effect of hydrogen
pressure on the Rh/HSPDP-catalyzed full hydrogenation of
MEPKO at a reaction temperature of 90 °C and a molar ratio
of C=C/Rh=100 within 60 min of reaction time is shown
in Table 1, entries 1/8,1/13,1/14. As expected the catalytic
activity and therefore the selectivity towards the final prod-
uct MS increase with increasing pressure of dihydrogen from
40 to 50 and up to 60 bar to yield MS with 69.7 mol%,
90.6 mol% and 100.0 mol%, respectively (entries 1/8, 1/13
and 1/14). At a shorter reaction time of 30 min (entry 1/15)
the selectivity to MS was lower (38.8 mol%) compared with
the selectivity of 44.5 mol% to MS within 60 min reaction
duration (entry 1/9) under the same reaction conditions. The
selectivity towards the saturated product MS increased fur-
ther at longer reaction times to achieve 64.3 mol% of MS
within a reaction time of 90 min (entry 1/16). The amount
of added methanol as a solvent has a pronounced effect on
the catalytic activity and selectivity towards MS in the Rh/
HSPDP-catalyzed hydrogenation reaction of MEPKO and is
shown in Table 1, entries 1/9 and 1/17–1/19. The catalytic
activity increased from TOF=65 h−1 up to TOF=137 h−1
and the selectivity to MS from 44.5 mol% up to 80.2 mol%
with increasing amount of added methanol as a solvent from
10 up to 40 ml at a low rhodium concentration of 50 and
15 ppm, respectively, a reaction temperature of 90 °C, 40 bar
hydrogen pressure, molar ratios of MEPKO/Rh=200 and of
HSPDP/Rh=1 within 60 min reaction time in the organic
monophasic system (entries 1/9, 1/17, 1/18). The presence
of a larger amount of added methanol has a negative effect
on both the catalytic activity and selectivity to MS in the
hydrogenation of the unsaturated esters part of MEPKO to
give with 50 ml of added methanol a catalytic activity of
TOF=130 h−1 and a selectivity to MS of 76.6 mol% (entry
1/19). Using as solvents diethyl ether and methyl acetate the
catalytic activities were much lower with TOF=3 h−1 and
TOF=4 h−1, respectively (entries 1/20, 1/21) compared with
methanol with TOF=137 h−1 (entry 1/18) using 40 ml of
added organic solvent under the same reaction conditions.
This remarkable rate enhancement achieved with methanol
compared to diethyl ether and methyl acetate in the
hydrogenation of MEPKO may be explained by invoking
interactions between the polar solvent methanol and catalytic
active rhodium key intermediates species modified with the
hexasulfonated o-phenylendiphosphite bidentate ligand in
the organic reaction mixture. We assume that methanol prob-
ably shifts the equilibrium of chelated catalytic intermedi-
ates where both C=C units and esters groups of MEPKO are
coordinated towards non-chelated catalytic complexes,
occupy the vacant coordination and also hydrogen bonding
stabilizes the non-chelated species which obviously facili-
tates subsequent steps in the catalytic cycle. The Ru/HSPDP-
catalyzed full hydrogenation of MEPKO in methanol
(entries 1/22–1/26) proceeds with much lower activities
compared with the catalytic activities exhibited Rh/HSPDP
complexes in methanol (entries 1/1–1/21). The catalytic
activities and selectivities in the Ru/HSPDP-catalyzed full
hydrogenation of MEPKO towards MS slightly increase with
increasing temperature from 70 °C up to 90 °C to give values
from 20 up to 30 TOFs per hour and selectivities from
29.0 mol% up to 40.4 mol% of MS at molar ratios of C=C
units/Ru=100 and HSPDP/Ru=1 and a hydrogen pressure
of 50 bar within 60 min of reaction time at a ruthenium
concentration of 100 ppm using 10 ml of methanol as a sol-
vent (Table 1, entries 1/22–1/24). Raising the reaction tem-
perature higher has a negative effect on both the catalytic
activity and selectivity to MS in the hydrogenation of the
unsaturated esters part of MEPKO to give at 100 °C only 20
TOFs per hour and 32.9 mol% of MS (Table 1, entry 1/25).
At a higher C=C units/Ru molar ratio of 200 in the Ru/
HSPDP-catalyzed full hydrogenation of MEPKO at 90 °C
the catalytic activity and selectivity to MS slightly decreased
to give 20 TOFs per hour and 21.8 mol% of MS (entry 1/26).
3.2 Full Hydrogenation Reactions of Unsaturated
Methyl Esters of Sunflower Oil (MESO, MESO1)
Catalyzed by Rh‑, Ru‑ and Ni–HSPDP Complexes
Table 2 shows the catalytic activity and selectivity of Ru/
HSPDP, Rh/HSPDP and Ni/HSPDP complexes in the full
hydrogenation of unsaturated methyl esters of sunflower oil
to yield the desired product MS in organic monophasic sys-
tems. The influence of various parameters was investigated
on this hydrogenation reaction such as temperature, C=C
units/metal molar ratio, reaction time, dihydrogen pressure,
addition of different amounts of methanol as a solvent and
of various organic solvents (Table 2). We used as start-
ing materials for the hydrogenation reaction two mixtures
of unsaturated methyl esters of sunflower oil the first one
MESO which is a product of Cognis GmbH (today BASF)
and the second one MESO1 which was obtained by trans-
esterification of sunflower oil with methanol. The catalytic
activity and selectivity towards MS in the Ru/HSPDP-cat-
alyzed hydrogenation of MESO increases with increasing
1 3

Novel Full Hydrogenation Reaction of Methyl Esters of Palm Kernel and Sunflower Oils Into Methyl…
temperature from 80 °C up to 120 °C to give TOFs from
200 up to 310 h−1 and selectivities to MS from 20.8 up
to 38.6 mol%, respectively, at molar ratios of C=C units/
Ru = 500 and HSPDP/Ru = 1 and a hydrogen pressure of
50 bar within 60 min of reaction time at a ruthenium con-
centration of 60 ppm using 10 ml of methanol as a solvent
(Table 2, entries 2/1–2/5). The apparent Arrhenius param-
eter of the activation energy of the Ru/HSPDP-catalyzed
full hydrogenation reaction of MESO in methanol was cal-
culated from results obtained from hydrogenation reactions
carried out at temperatures 80–120 °C and are summarized
in Table 2, entries 2/1–2/5. The apparent activation energy
which was calculated with these data (Fig. 2) amounts to
12.75 kJ/mol. The rather low apparent activation energy of
12.75 kJ/mol indicates for the presence of an active ruthe-
nium catalyst modified with the bidentate HSPDP ligand
in the polar medium of MESO in methanol and might sug-
gest that the low reaction rates observed in the Ru/HSPDP-
catalyzed full hydrogenation of MEPKO (Table 1, entries
1/22–1/26) are probably caused from the less polar nature
of MEPKO starting material mixture in methanol compared
to MESO in methanol because MEPKO consists of large
amounts of 84.8 mol% of less polar components which are
the saturated C10:0-C16:0 FAMEs and these compounds
act as a less polar solvent in the MEPKO full hydrogena-
tion reaction mixture. We choose Rh/HSPDP catalysts to
study further the full hydrogenation of MESO because Rh/
HSPDP complexes exhibited much higher catalytic activities
and selectivities to MS compared with those obtained with
Ru/HSPDP catalysts in the full hydrogenation of MEPKO
(Table 1). At the low C=C units/Rh molar ratio of 200 the
selectivity to MS was 97.2 mol% in the Rh/HSPDP-cata-
lyzed full hydrogenation of MESO at a reaction temperature
of 90 °C under 40 bar dihydrogen pressure within 60 min of
reaction time (entry 2/6) and with increasing C=C units/Rh
molar ratios the selectivity to MS drops dramatically to give
at a molar ratio of C=C units/Rh=2000 only 53.0 mol% of
MS (entry 2/9). However, the catalytic activity at a molar
ratio of C=C units/Rh=200 was only 185 TOFs per hour
and with increasing C=C units/Rh molar ratios the catalytic
activity considerably increased to reach at a molar ratio of
C=C units/Rh=2000 a value of 1400 TOFs per hour under
the same reaction conditions (entries 2/6–2/9). The effect of
reaction time in the Rh/HSPDP-catalyzed full hydrogena-
tion of MESO at molar ratios of C=C units/Rh=2000 and
HSPDP/Rh = 1, a reaction temperature of 90 °C under a
dihydrogen pressure of 40 bar and a rhodium concentration
of 50 ppm using 10 ml of methanol as a solvent is presented
in entries 2/9–2/13. At a shorter reaction time of 5 min
the catalytic activity of Rh/HSPDP catalysts was as high
as 8680 TOFs per hour and the selectivity towards MS of
15.0 mol% under the same reaction conditions (entry 2/10).
At longer reaction times the selectivity to MS increased to
achieve within 90 min a selectivity of 56.2 mol% of MS
and at a lower catalytic activity of TOF = 930 h−1 (entry
2/13). The hydrogen pressure has a pronounced effect on
both the activity and selectivity to MS in the Rh/HSPDP-
catalyzed full hydrogenation of MESO in methanol (Table 2,
entries 2/14–2/17). The catalytic activity increased from
TOF = 1380 h−1 up to TOF= 2800 h−1 and the selectivity
towards MS from 10.3 mol% up to 46.4 mol% with increas-
ing hydrogen partial pressure from 10 to 40 bar at the low
reaction temperature of 80 °C, molar ratios of C=C units/
Rh=2000 and of HSPDP/Rh=1 and a rhodium concentra-
tion of 50 ppm within 30 min reaction time using 10 ml
methanol as solvent. However, a relative high formation of
undesired trans-C18:1 esters of 31.5 mol% was observed
with a selectivity to MS of 46.4 mol% at the increased
catalytic activity of 2800 TOFs per hour even at the low
reaction temperature of 80 °C under 40 bar hydrogen pres-
sure in methanol (entry 2/17). The effect of the amount of
added methanol as a solvent in the Rh/HSPDP-catalyzed
full hydrogenation reaction of MESO is shown in Table 2,
entries 2/17–2/19. Both, the catalytic activity and selectivity
to MS increased from TOF=2800 h−1 up to TOF=3800 h−1
and the selectivity to MS from 46.4 mol% up to 100.0 mol%
with increasing amount of added methanol as a solvent
from 10 up to 50 ml at a low rhodium concentration of 50
and 12 ppm, respectively, a reaction temperature of 80 °C,
40 bar hydrogen pressure, molar ratios of MESO/Rh=2000
and of HSPDP/Rh=1 within 30 min reaction time (entries
2/17–2/19). Addition of less polar organic solvents than
methanol such as diethyl ether and methyl acetate has a det-
rimental effect on both the catalytic activity and selectivity
towards MS in the Rh/HSPDP catalysed full hydrogena-
tion of MESO. Thus, in the presence of diethyl ether only
80 TOFs per hour with 6.7 mol% of MS (entry 2/20) and
with addition of methyl acetate 120 TOFs per hour with
5.4 mol% of MS (entry 2/21) were obtained in the full hydro-
genation of MESO compared with 2900 TOFs per hour with
-2.35
0.00252
0.00262
0.00272
0.00282
-2.45
-2.55
-2.65
-2.75
-2.85
-2.95
y = -1533.4x + 1.4843
R² = 0.9656
1/T
Fig. 2 Calculation of the apparent Arrhenius parameter of the activa-
tion energy of the full hydrogenation reaction of MESO catalyzed by
Ru/HSPDP complexes in methanol
1 3

N. Kaika et al.
77.1 mol% of MS in the presence of methanol (entry 2/18)
under the same reaction conditions. We further applied the
less expensive nickel catalyst modified with HSPDP ligands
in the hydrogenation reaction of MESO1 using methanol
as a solvent (Table 2, entries 2/22–2/24). The Ni/HSPDP-
catalyzed full hydrogenation of MESO1 proceeds in gen-
eral with much lower catalytic activities and selectivities
towards MS (entries 2/22–2/24) even under more forcing
conditions due to the 3d transition metal nature of nickel
compared with the activities exhibited both Ru/HSPDP and
Rh/HSPDP catalysts in methanol (entries 2/1–2/21). In the
Ni/HSPDP-catalyzed full hydrogenation of MESO1 at molar
ratios of HSPDP/Ni=1 and C=C units/Ni=500, a reaction
temperature of 120 °C under 50 bar dihydrogen pressure and
a nickel concentration of 35 ppm using 10 ml of methanol
as a solvent within 60 min of reaction time the catalytic
activity was low namely 54 TOFs per hour with a selectivity
to MS of 8.1 mol% (entry 2/22) whereas Ru/HSPDP cata-
lysts exhibited a higher catalytic activity of 310 TOFs per
hour and a higher selectivity to MS of 38.6 mol% under the
same reaction conditions (entry 2/5). At a higher tempera-
ture of 160 °C and a lower molar ratio of C=C/Ni=200 the
Ni/HSPDP catalytic system exhibited a higher activity of
TOF=80 h−1 with a higher selectivity to MS of 15.3 mol %
(entry 2/23). Using Ni/HSPDP catalysts under forcing condi-
tions for a homogeneously catalyzed hydrogenation reaction
namely a temperature of 170 °C and 70 bar of hydrogen
pressure the catalytic activity slightly increases to obtain
90 TOFs per hour with a selectivity to MS of 19.4 mol%
(entry 2/24).
low catalytic activities (TOF = 90 h−1) and selectivities
towards MS (19.4 mol%) were obtained in methanol. The
apparent activation energy of Ru/HSPDP catalyst in the
MESO full hydrogenation reaction was calculated and
amounts to a low value of 12.75 kJ/mol indicating for the
presence of an active ruthenium catalyst modified with the
bidentate HSPDP ligand in the polar medium of MESO in
methanol. A quantitative selectivity to MS was obtained
with, however, much lower catalytic activities of 88 TOFs
per hour in the full hydrogenation reaction of methyl esters
of palm kernel oil (MEPKO) using Rh/HSPDP catalysts
at 90 °C under 60 bar of dihydrogen pressure and molar
ratios of C=C units/Rh=100 and HSPDP/Rh=1 at a low
rhodium concentration of 50 ppm within 60 min of reaction
time in 10 ml of methanol. In the hydrogenation reaction
of MEPKO at a higher amount of 40 ml of added metha-
nol Rh/HSPDP complexes exhibited a catalytic activity of
TOF = 137 h−1 and the selectivity to MS was 80.2 mol%.
The catalytic activities exhibited Ru/HSPDP complexes in
the hydrogenation reaction of MEPKO in methanol were, in
general, lower than those achieved by Rh/HSPDP catalysts
and have not exceeded the range of 30 TOFs per hour with
a selectivity of 40.4 mol% of MS.
Acknowledgements Financial support of this research by the Post-
graduate Studies Programme on “Catalysis and its Applications in the
Industry” of Greek Ministry of Education and by the Special Account
for Research Grants of the Research Committee of the National and
Kapodistrian University of Athens are gratefully acknowledged. We
thank Cognis GmbH (today BASF) for the supply of the starting mate-
rials methyl esters of palm kernel oil (Edenor® ME PK 12-18 F) and
of methyl esters of sunflower oil (Sunflower Fatty Acid ME®).
Compliance with Ethical Standards
4 Conclusions
Conflict of interest The authors declare that they have no conflict of
We have disclosed here a novel study on the full hydrogena-
tion reaction of polyunsaturated methyl esters of palm kernel
and sunflower oils into the saturated methyl stearate (MS)
employing rhodium, ruthenium and nickel catalysts modified
with hexasulfonated o-phenylendiphosphite triisooctylam-
monium salt (HSPDP) ligand using methanol as a solvent.
The catalytic hydrogenation of methyl esters of sunflower
oil (MESO) proceeds smoothly with Rh/HSPDP complexes
to achieve a high catalytic activity of TOF=8680 h−1 at a
low reaction temperature of 90 °C under 40 bar of dihydro-
gen pressure and molar ratios of C=C units/Rh=2000 and
HSPDP/Rh=1 at a low rhodium concentration of 50 ppm
within 5 min of reaction time in 10 ml of methanol. At a
higher amount of 50 ml of added methanol the Rh/HSPDP-
catalyzed full hydrogenation of MESO yields quantitatively
MS at a relative high catalytic activity of 3800 TOFs per
hour even at a lower reaction temperature of 80 °C within
30 min of reaction time. Using less expensive Ni/HSPDP
catalysts under forcing conditions (T=170 °C, PH2 =70 bar)
interest.
References
1. Corma A, Iborra S, Velty A (2007) Chem Rev 107:2411
2. Huber GW, Iborra S, Corma A (2006) Chem Rev 106:4044
3. Moser BR (2014) Biofuels 5:5
4. Issariyakul T, Dalai AK (2014) Renew Sust Energy Rev 31:446
5. Turek T, Trimm DL, Cant NW (1994) Catal Rev Sci Eng 36:645
6. Rupilius W, Ahmad S (2007) Eur J Lipid Sci Technol 109:433
7. Behr A, Döring N, Durowicz-Heil S, Ellenberg B, Kozik C, Lohr
C, Schmidke H (1993) Fat Sci Technol 95:2
8. Huber GW, O’Connor P, Corma A (2007) Appl Catal A 329:120
9. Knothe G (2010) Prog Energy Combust Sci 36:364
10. No S-Y (2014) Fuel 115:88
11. Pozdeev VA, Safronov SP, Levanova SV, Krasnykh EL (2012)
Russ J Appl Chem 85:261
12. Schaff T, Greven H (2010) Lipid Technol 22:31
13. Papadogianakis G, Bouriazos A, Tsichla A, Vasiliou Ch, US
8 334 396 B2 (18.12.2012), EP 2 014 752 A1 (23.06.2007),
WO 2009/000435 A1 (14.06.2008), US 2010/0234625 A1
1 3

Novel Full Hydrogenation Reaction of Methyl Esters of Palm Kernel and Sunflower Oils Into Methyl…
42. Eckel RH, Borra S, Lichtenstein AH, Yin-Piazza SY (2007) Cir-
(16.09.2010) to Cognis IP Management GmbH and National and
Kapodistrian University of Athens
culation 115:2231
43. Mozaffarian D, Katan MB, Ascherio A, Stampfer MJ, Willett WC
(2006) N Engl J Med 354:1601
14. Vasiliou C, Bouriazos A, Tsichla A, Papadogianakis G (2014)
Appl Catal B 158–159:373
44. Schleifer D (2012) Soc Stud Sci 43:54
15. Normann W, DE (DRP) 141 029 (14.08.1902), GB 190 301 515
(21.01.1903) to Leprince & Siveke Herforder Maschinenfett &
Oelfabrik
45. Stender S, Astrup A, Dyerberg J (2012) BMJ Open 2:e000859
46. Stender S, Astrup A, Dyerberg J (2014) BMJ Open 4:e005218
47. Tarrago-Trani MT, Phillips KM, Lemar LE, Holden JM (2006) J
Am Diet Assoc 106:867
16. Leuteritz G (1969) Fette Seifen Anstrichmittel 71:441
17. List GR, Jackson MA (2007) Inform 18:403
18. List GR, Jackson MA (2009) Inform 20:395
19. Philippaerts A, Jacobs PA, Sels BF (2013) Angew Chem Int Ed
52:5220
48. Remig V, Franklin B, Margolis S, Kostas G, Nece Th, Street JC
(2010) J Am Diet Assoc 110:585
49. Borra S, Kris-Etherton PM, Dausch JG, Yin-Piazza SY (2007) J
Am Diet Assoc 107:2048
20. List GR (2012) Inform 23:451
50. Menaa F, Menaa A, Menaa B, Tréton J (2013) Eur J Nutr 52:1289
51. Galvín S, Guillén-Sans R, Galbis J, Guzmán-Chozas M (2016)
LWT Food Sci Technol 65:1066
21. Ribeiro APB, Grimaldi R, Gioielli LA, Gonçalves (2009) LAG
Food Res Int 42:401
22. Adhikari P, Shin J-A, Lee J-H, Kim H-R, Kim I-H, Hong S-T, Lee
K-T (2012) Food Bioprocess Technol 5:2474
52. Patel AR, Dewettinck K (2016) Food Funct 7:20
53. Kuhnt K, Baehr M, Rohrer C, Jahreis G (2011) Eur J Lipid Sci
Technol 113:1281
23. Zhang L, Muramoto H, Ueno S, Sato K (2011) J Oleo Sci 60:287
24. Li D, Adhikari P, Shin J-A, Lee J-H, Kim Y-J, Zhu X-M, Hu J-N,
Jin J, Akoh CC, Lee K-T (2010) LWT Food Sci Technol 43:458
25. Ribeiro APB, Basso RC, Grimaldi R, Gioielli LA, dos Santos AO,
Gardoso LP, Gonçalves (2009) LAG Food Res Int 42:1153
26. Ribeiro APB, Basso RC, Grimaldi R, Gioielli LA, Gonçalves
(2009) LAG J Food Lipids 16:362
54. Menaa F, Menaa A, Tréton J, Menaa B (2013) J Food Sci
78:R377–R386
55. Fitch-Haumann B (1994) Inform 5:668
56. Bahrmann H, Fell B, Papadogianakis G, DE 3 942 787 B1
(23.12.1989), EP 0 435 071 B1 (12.12.1990), US 632 465 (1990),
CA 2 032 371 (1990), JP Hei/2/402 868 (1990), AU 68 368/90
(1990), BR 90 06 501 (1991), TW 79/109 500 (1990), KO 90/20
650 (1990) to Hoechst AG
27. Kim I-H, Lee S-M, Lee B-M, Park H-K, Kim J-Y, Kwon K-I, Kim
J-W, Lee J-S, Kim Y-H (2008) J Agric Food Chem 56:5942
28. Criado M, Hernández-Martín E, López-Hernández A, Otero C
(2007) J Am Oil Chem Soc 84:717
57. Bahrmann H, Fell B, Papadogianakis G, DE 3 942 954 B1
(23.12.1989), EP 0 435 084 B1 (13.12.1990), US 632 464 (1990),
CA 2 032 372 (1990), JP Hei/2/402 869 (1990), AU 68 367/90
(1990), BR 90 06 444 (1991), TW 79/109 499 (1990), KO 90/20
528 (1990) to Hoechst AG
29. Farmani J, Hamedi M, Safari M, Madadlou A (2007) Food Chem
102:827
30. Petrauskaite V, de Greyt W, Kellens M, Huyghebaert A (1998) J
Am Oil Chem Soc 75:489
58. Fell B, Papadogianakis G, Konkol W, Weber J, Bahrmann H
(1993) J Prakt Chem/Chem-Ztg 335:752
31. Schmidt S, Hurtová S, Zemanovič J, Sekretár S, Šimon P, Ains-
worth P (1996) Food Chem 55:343
59. Papadogianakis G, Thesis PhD, “Beiträge zur Rhodium-katalysi-
erten Hydroformylierung mittel- und höhermolekularer α-olefine
im Ein- und Zweiphasen-System”, Rheinisch-Westfälische Tech-
nische Hochschule (RWTH) Aachen, 1990
32. Hurtová S, Schmidt S, Zemanovič J, Sekretár S, Šimon P (1996)
Fett/Lipid 98:60
33. List GR, Mounts TL, Orthoefer F, Neff WE (1995) J Am Oil
Chem Soc 72:379
60. Favre F, Olivier-Bourbigou H, Commereuc D, Saussine L (2001)
Chem Commun 2001:1360
34. Mensink RP, Katan MB (1990) New Eng J Med 323:439
35. Zock P, Katan MB (1992) J Lipid Res 33:399
36. German JB, in: Destaillats F, Sébédio J-L, Dionisi F, Chardigny
J-M (eds), Trans fatty acids in human nutrition, 2nd edn., Oily
Press Lipid Library, Woodhead Publishing, Oxford, 2009, vol 23,
pp xix–xxiv
61. Ratnayake WMN, Hansen SL, Kennedy MP (2006) J Am Oil
Chem Soc 83:475
62. Bouriazos A, Sotiriou S, Vangelis C, Papadogianakis G (2010) J
Organomet Chem 695:327
37. Bysted A, Mikkelsen A, Leth T (2009) Eur J Lipid Sci Technol
111:574
Publisher’s Note Springer Nature remains neutral with regard to
38. Hillyer ChD (2007) Inform 18:356
jurisdictional claims in published maps and institutional affiliations.
39. Clifton PM, Keogh JB, Noakes M (2004) J Nutr 134:874
40. L’Abbé MR, Stender S, Skeaff M, Ghafoorunissa M, Tavella
(2009) Eur J Clin Nutr 63:S50
41. Hunter JE (2005) Nutr Res 25:499
Affiliations
Niki Kaika1 · Christina Panopoulou1 · Eleni Anagnostopoulou1 · Christos Fakas1 · Panagiotis Lilas1 ·
Dimitra Stavroulaki1 · Georgios Papadogianakis1
1
* Georgios Papadogianakis
Industrial Chemistry Laboratory, Department of Chemistry,
papadogianakis@chem.uoa.gr
National and Kapodistrian University of Athens,
Panepistimiopolis-Zografou, 157 71 Athens, Greece
1 3