Solvent-free “green” amidation of stearic acid for synthesis of biologically active alkylamides over iron supported heterogeneous catalysts

Article abstract of DOI:10.1016/j.apcata.2017.06.006

Stearoyl ethanolamine was synthesized by amidation of stearic acid with ethanolamine in solventless conditions. Iron containing heterogeneous catalysts supported on SiO₂, Al₂O₃, Beta (BEA), ZSM-12 (MTW) and Ferrierite (FER) were used in this work. Sn-modified Ferrierite and H-Ferrierite were also studied for comparison. Fe-modified catalysts synthesized using solid state ion-exchange and evaporation impregnation methods, were thoroughly characterized with X-ray powder diffraction, scanning electron microscope, FTIR with pyridine, nitrogen adsorption, energy dispersive X-ray microanalysis and M?ssbauer spectroscopy. The highest conversion was obtained with Fe-H-FER-20 at 140?°C in 1?h giving 61% conversion and 98% selectivity towards the desired amide. The catalytic performance in terms of turnover frequency per mole of iron was achieved with the catalyst exhibiting the largest amount of Fe³⁺ species, optimum acidity and a relatively low Br?nsted to Lewis acid site ratio.

Full text of DOI:10.1016/j.apcata.2017.06.006

Applied Catalysis A, General 542 (2017) 350–358
Contents lists available at ScienceDirect
Applied Catalysis A, General
journal homepage: www.elsevier.com/locate/apcata
Feature Article
Solvent-free “green” amidation of stearic acid for synthesis of biologically
active alkylamides over iron supported heterogeneous catalysts
a
a
a
a
a
a
Päivi Mäki-Arvela , Jeanne Zhu , Narendra Kumar , Kari Eränen , Atte Aho , Johan Linden ,
b
b
c
c
a,⁎
Jarno Salonen , Markus Peurla , Anton Mazur , Vladimir Matveev , Dmitry Yu. Murzin
Åbo Akademi University, Turku/Åbo, Finland
University of Turku, Turku, Finland
St. Petersburg State University, Russia
a
b
c
A R T I C L E I N F O
A B S T R A C T
Keywords:
Zeolite
Amidation
Fatty acid
Stearoyl ethanolamine was synthesized by amidation of stearic acid with ethanolamine in solventless conditions.
Iron containing heterogeneous catalysts supported on SiO , Al O , Beta (BEA), ZSM-12 (MTW) and Ferrierite
2
2 3
(FER) were used in this work. Sn-modified Ferrierite and H-Ferrierite were also studied for comparison. Fe-
modified catalysts synthesized using solid state ion-exchange and evaporation impregnation methods, were
thoroughly characterized with X-ray powder diffraction, scanning electron microscope, FTIR with pyridine,
nitrogen adsorption, energy dispersive X-ray microanalysis and Mössbauer spectroscopy. The highest conversion
was obtained with Fe-H-FER-20 at 140 °C in 1 h giving 61% conversion and 98% selectivity towards the desired
amide. The catalytic performance in terms of turnover frequency per mole of iron was achieved with the catalyst
3
+
exhibiting the largest amount of Fe species, optimum acidity and a relatively low Brønsted to Lewis acid site
ratio.
1. Introduction
reaction operating at the temperature close to the boiling point of acetic
acid yielding 96% of amides within 1- 5 h depending on the reactant
Amidation of fatty acids is an important reaction, in which fatty
[9]. In addition acetic acid was also amidated with aniline using Fe-
Beta zeolite with Si/Al ratio of 15 at 117 °C with an excess of acetic acid
under microwave irradiation giving 52% conversion of stearic acid in
15 min [10]. Furthermore, cheap alumina balls were used for produc-
tion of n-isopropylheptanamide from heptanoic acid and iso-
propylamine resulting in 95% yield at 140 °C in 3 h [11]. In this work
[11] the alumina balls were calcined prior to the experiment facilitating
their function as adsorbents during amidation reaction. The reactant
was a relatively short carboxylic acid and the drawback with alumina
alkanolamines are formed as products. Applications of these products
are related not only to surfactants, but also to high value-added phar-
maceuticals exhibiting several biological effects, such anti-carcinogenic
ones [1] and activity against the Alzheimer disease [2]. The use of
heterogeneous catalysts makes the process more industrially benign and
cheaper, avoiding cumbersome catalyst separation.
Fatty acid amides are typically synthesized with homogeneous
catalysts, for example by reacting a fatty acid ester with an alkanola-
mine in the presence offatty acid chlorides [3] or using potassium hy-
droxideas reagents [4], which are not environmentally benign proce-
dures. Utilization of fatty acids is beneficial from the viewpoint of
expanding the feedstock base and typically acidic and metal modified
zeolites and mesoporous materials were used as catalysts. Direct ami-
dation of fatty acids with alkanolamines has been scarcely investigated
over heterogeneous catalysts [5–8]. These reactions were mainly con-
ducted in the presence of solvents. In some cases, supported iron cat-
alysts have been applied in acetic acid amidation [9,10]. This acid
underwent amidation with several phenylamine derivatives over iron
on activated carbon catalyst. A high excess of acid was used in this
balls is that they cannot be used in autoclave. When Cu-SiO
2 2 3
, Cu-Al O
and Cu-TiO catalysts were studied in α-amidation of cyclic ethers [12],
2
it was observed that the reaction was influenced by the support type
and copper dispersion. Application of non-conventional, sustainable,
eco-friendly and highly efficient process technology for production of
various types of amides is of immense importance. Oxidative amidation
of alcohols with amines was carried out using heteropolyanion-based
ionic liquids under microwave irradiation and solvent free reaction
media [13]. Formation of primary, secondary and tertiary alkyl amides
was reported with bifunctional catalysts. Direct amidation of carboxylic
acids with amines using ultrasonic irradiation was investigated using
⁎
Corresponding author.
E-mail address: dmurzin@abo.fi (D.Y. Murzin).
http://dx.doi.org/10.1016/j.apcata.2017.06.006
Received 10 January 2017; Received in revised form 5 April 2017; Accepted 2 June 2017
Available online 02 June 2017
0926-860X/ © 2017 Elsevier B.V. All rights reserved.

P. Mäki-Arvela et al.
Applied Catalysis A, General 542 (2017) 350–358
solid acid sulfonated graphene oxide [14]. Very high yields of amides
under shorter reaction time was reported by the authors when sono-
chemical amidation of carboxylic acid was applied with a reusable solid
acid catalyst.
Besides zeolites per se, transition and noble metal modified ones
have found applications in several industrial processes in the fields of
petrochemistry, oil refining, exhaust gas emission abatement, produc-
rotavapor for 24 h. The aqueous solution was finally evaporated, the
catalyst was dried for 7 h at 100 °C and calcined in a muffle oven.
For the tin containing catalyst, the same synthesis procedure was
applied with 0.905 g of tin sulfate (Fluka, ≥96%) or 1.6976 g of tin
chloride (Sigma, 99%). The step calcination procedure was different
from iron catalysts: initial heating rate was 2.3 °C/min up to 200 °C,
maintaining this temperature for 65 min, heating to 400 °C with 2.9 °C/
min, holding at 400 °C for 3 h, thereafter cooling the catalyst to 25 °C.
To prepare iron catalysts of 5 wt% by SSIE, 3.6 g of ferrite nitrate
nonahydrate (Fluka) was mixed with 10 g of the catalyst support during
ball milling for 6 h. Every 2 h, the equipment was stopped to manually
mix and crush the solid mixture. After 6 h of ball milling, the catalyst
was dried in an oven at 100 °C overnight. Then, the catalyst was cal-
cined applying the same step calcination procedure as for iron catalysts
prepared by EIM.
tion of specialty and fine chemicals. In the current study besides SiO
Al and TiO supports also, Beta (BEA), Ferrierite (FER), and ZSM-12
MTW) zeolites were modified with Fe using solid state ion-exchange
2
,
2
O
3
2
(
and evaporation impregnation synthesis methods. Sn modified FER was
synthesized using evaporation impregnation method. ZSM-12 (MTW)
belongs to high silica zeolite group with unidimensional channel sys-
tems, 12 membered rings with pore dimensions 0.57 × 0.61 nm. The
pore dimensions are slightly larger than for ZSM-5 (MFI) zeolite. ZSM-
1
[
2 zeolite exhibits shape selectivity and is resistant to coke formation
15,16].
Application of zeolites is not only limited to amidation of carboxylic
For the tin catalyst, the same synthesis procedure was applied with
0.905 g of tin sulfate (Fluka, ≥96%).
acids, as they have been used as catalysts for amidation of alcohols [17]
and ketones [18]. In the former case for example sec-butanol was re-
acting with acrylonitrile at 160 °C giving 76% selectivity to the corre-
sponding amide with 72% conversion after 8 h [17]. Furthermore, there
was an optimum Si/Al ratio observed giving the highest amide se-
lectivity. In the latter case, in the amidation of benzophonene with
hydroxylamine hydrochloride the isolated yield of amide using micro-
wave irradiation and HY-zeolite as a catalyst of 94% was achieved in
2.2. Catalyst characterization methods
The specific surface area was determined using nitrogen adsorption
with Sorptomatic 1900. The samples were outgassed prior to mea-
surements at 150 °C for 3 h. The specific surface areas were calculated
using Dubinin’s method [19,20].
XRD measurements were performed to identify the structure of
zeolites using Philips X’Pert Pro MPD instrument and monochromated
CuKα radiation at 40 kV/50 mA using beam collimation of 0.25° di-
vergence slit and a fixed mask of 20 mm. Philips X’Pert HighScore and
MAUD programs were used for analysis.
SEM analyses were performed with a LEO Gemini 1530 scanning
electron microscope. Thermo Scientific UltraDry Silicon Detecto (SDD)
was used for morphological analysis of samples. The equipment con-
tained both secondary and backscattered electron detectors and an In-
Lens detector.
2
min [18].
The aim in this work was to investigate the possibilities to synthe-
size stearoyl ethanolamide from stearic acid and ethanolamine in the
absence of any volatile solvent and for the first time to use cheap
supported iron catalysts in this reaction. The purpose was also to study
feasibility of overall technology comprising environmentally benign
methods of catalyst preparation by solid state ion exchange and reac-
tions with inexpensive heterogeneous catalysts. The studied catalysts
were characterized by several physical-chemical methods, including
Mössbauer spectroscopy, gas phase pyridine adsorption desorption, li-
quid phase adsorption of 2-phenylethylamine, solid state NMR, ni-
trogen adsorption and SEM. The main parameters were the type of the
support and acid sites, concentration of the latter and the oxidation
state of iron.
−1
The amounts of Brønsted and Lewis acid sites at bands 1545 cm
−1
and 1455 cm , respectively were quantified with pyridine (Sigma-
Aldrich, > 99.5%) adsorption/desorption by FTIR using ATI Mattson
instrument and using molar extinction coefficients from Emeis [21].
Adsorption of 2-phenylethylamine on zeolites and metal modified
zeolites was investigated using 0.03 M 2-phenylethylamine as an ad-
sorbate in distilled water [22]. Typically 50 mg of the dried catalyst was
combined with 3 ml of 2-phenylethylamine (Acros Organics, 99%) so-
lution at 24 °C. The solution was stirred with a magnet for 2 h, which
was sufficient to achieve the equilibrium. The supernatant liquid after
filtration of the catalyst was analyzed with UV–vis (Shimadzu UV-2550)
at 252 nm. The adsorbed amounts of 2-phenylethylamine were calcu-
lated by subtracting the amount of 2-phenylethylamine present in the
liquid phase from the initial adsorbent concentration. The experimental
error in absorbance was 2%.
2. Experimental
2.1. Catalyst synthesis
Several commercial zeolites and oxides were used as catalysts and
supports. Silica gel 60 (Merck) was sieved with a 90 μm sieve. Titanium
IV) hydroxide granules (Alfa Aesar) were crushed and sieved with ball
milling to the size below 90 μm. NH -Beta-300, NH -Beta-150, NH
2 3
Beta-25 and NH -FER-20, in which the number represents SiO to Al O
(
4
4
4
-
4
2
ratio, were purchased from Zeolyst International and calcined in order
to get the proton forms. The step calcination procedure was: heating
rate of 3.8 °C/min up to 250 °C, holding at this temperature for 40 min,
subsequent heating with 2.1 °C/min up to 400 °C, maintaining this
temperature for 4 h, thereafter cooling zeolite to 25 °C.
2.3. Catalyst evaluation for amidation reaction and analytical procedure
Typically the experiments were performed in an autoclave using
equimolar amounts of the fatty acid (Sigma Aldrich, 95%) and etha-
nolamine (0.14 mol, Sigma Aldrich, > 95%) under 20 bar Ar (AGA)
under high stirring speed, 1100 rpm to minimize mass transfer limita-
tions using 0.5 g catalyst. External mass transfer was suppressed by
applying small catalyst particles, below 90 μm. The liquid volume was
44 ml and the initial concentration of fatty acid was 3.0 M. The samples
of reaction mixtures were silylated as follows: the solid sample was
dissolved in toluene with the concentration of 1 mg/ml. Thereafter
about 1 ml of the sample was silylated with 120 μl of bis(trimethylsilyl)
trifluoroacetamide (BSTFA) and 0.4 μl of trimethylchlorosilane (TMCS)
at 70 °C for 1 h. The samples were cooled and analyzed with a GC
equipped with HP-1 column using the following temperature program:
100 °C (1.5 min) – 12 °C/min – 340 °C (20 min). The injector and
To prepare iron catalysts with 5 wt% by evaporation impregnation
method (EIM), 3.6 g ferric nitrate nonahydrate (Fluka) was dissolved in
250 ml of water. The pH was recorded prior and after adding 10 g of the
catalyst support. The round bottom flask was put in the rotator eva-
porator for rotation at 60 °C during 24 h. Thereafter, water was eva-
porated under vacuum and the catalyst was dried at 100 °C overnight.
The catalyst was then calcined in a muffle oven at a rate of 3 °C/min up
to 250 °C, maintaining this temperature for 50 min, followed by heating
rate of 3.3 °C/min up to 450 °C and holding this temperature for 3 h,
thereafter cooling the catalyst to 25 °C. For Fe-SiO
2
-50% US-EIM the
slurry was first subjected to ultrasound for 2 h, followed by rotation in
351

P. Mäki-Arvela et al.
Applied Catalysis A, General 542 (2017) 350–358
Fig. 1. Diffractogram of H-FER-20, Fe-FER-20-SSIE and
2 3
Hematite (Fe O ).
4
2
000
000
0
H-FER-20
Fe-FER-20-SSIE
Hematite
5
10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85
2
Theta °
detector temperatures were 330 °C and 340 °C, respectively. The peaks
were confirmed with GC–MS and NMR analogously to our previous
publication [5].
often quite long and narrow. The Fe-Beta-25-EIM crystals are circular,
whereas Fe-Al -EIM crystals have a shape of large balls with small
crystal growth inside. Fe-TiO crystals are very small and circular in
shape, while Fe-SiO crystals have sharp and thick edges.
2 3
O
2
2
3. Results and discussion
3
.1.3. Surface area and pore volume measurements using nitrogen
3
3
.1. Catalyst characterization results
physisorption
The specific surface areas of the catalysts are given in Table 1. The
.1.1. X-ray powder diffraction
The structure and phase purity of Fe modified Beta-25, ZSM-12-35
2
largest specific surface area (953 m /g) was measured for Fe-Beta-25-
EIM zeolite catalyst, whereas Fe-TiO -SSIE exhibited the lowest specific
2
surface area (90 m /g). The reason for a very large surface area for Fe-
Beta-25-EIM zeolite catalyst is attributed to the presence of a large
amount of micropores. In addition Fe-FER-20-SSIE zeolite catalyst ex-
hibited even a lower specific surface area (384 m /g), which was 17%
lower than the specific surface area (464 m /g) of pristine H-FER. Fe-H-
ZSM-12-35-EIM catalyst displayed also a low specific surface area.
and Ferrierite-20 zeolite catalysts were determined using X-ray powder
diffraction. The diffractogram for Fe-FER-20-SSIE is shown in Fig. 1.
The corresponding diffraction patterns were similar to those of the
pristine unmodified Beta (Fig. 2) [23], ZSM-12 and Ferrierite zeolites
2
2
[
24], indicating that Fe modification did not influence the basic
2
structures. In order to investigate if Fe particles, which presence was
2 3
O
confirmed by Mössbauer spectroscopy and discussed, are visible in XRD
of Fe-FER-20-SSIE, XRD for pristine ferrierite was also made and
compared with XRD of hematite [25]. The main Braggs reflection peak
of hematite, 104 is found at 2Θ 33° followed by the second largest peak,
3
.1.4. Determination of acidity
Acidity of the catalysts was elucidated by pyridine adsorption-des-
orption from the gas phase and via adsorption of phenyl ethylamine
from distilled water. The results from pyridine adsorption-desorption
showed that Fe-SiO
1
10 at 36° were also visible in Fe-FER-20-SSIE. It was, however, chal-
lenging to clearly differentiate hematite particles from ferrierite in TEM
images due to low loading of iron. XRD data indicated that hematite
particles in Fe-FER-20-SSIE would be larger than 3 nm.
2
exhibited a very low acidity as expected. Fe-
TiO –SSIE showed only Lewis acidity. Fe-ZSM-12-35-EIM displayed
2
very low acidity according to pyridine FTIR. Due to its small pore sizes,
it can be, however, concluded that pyridine is not capable to enter the
pores of ZSM-12 giving thus very low acidity values reflecting only the
external surface. The Brønsted acidity of H-FER was the highest one
featuring also large amounts of strong acid sites (Table 1). When Fe was
loaded on H-FER, the concentration of Brønsted acid sites of the re-
sulting Fe-H-FER-20-SSIE was only 54% of H-FER initial acidity. The
amount of Brønsted acid sites in Fe-Beta-25-EIM catalyst was 13%
higher than that of H-Beta-150-SSIE as expected. Noteworthy is, how-
ever, that the amount of strong acid sites increased in the following
order: Fe-Beta-25-EIM > Fe-Beta-150-SSIE > Fe-FER-20-SSIE. When
comparing the acidity values measured with pyridine adsorption des-
orption to that of amine adsorption (Fig. 5), it can be seen that Fe-Beta-
3
.1.2. Scanning electron microscopy
The morphology of the catalysts was studied using scanning elec-
tron microscopy. The SEM images of some selected catalysts are shown
in Figs. 3 and 4. The shape of Fe-FER-20-SSIE crystals is rectangular,
4000
3500
3000
2500
2000
1500
1000
25-EIM exhibited high acidity according to both methods. Un-
expectedly, the amine adsorption capacity of Fe-Beta-150-EIM was
higher than that of Fe-Beta-25-EIM (Fig. 5), although they have almost
the same total capacity for adsorption of pyridine. The latter catalyst
contains also more Brønsted acid sites than the former one. On the other
hand, acidity of especially H-FER measured by amine adsorption was
lower than determined by pyridine adsorption-desorption method. Iron
5
00
0
0
20
40
60
80
2
Theta
Fig. 2. X-ray powder diffraction patterns of H-Beta-25 zeolite.
352

P. Mäki-Arvela et al.
Applied Catalysis A, General 542 (2017) 350–358
Fig. 3. Scanning electron micrographs of (a) 1.8 wt%
Fe-FER-20-SSIE and (b) H-FER-20.
loaded Fe-FER-20-SSIE exhibited a rather low acid site concentration
based on amine adsorption. These results can be correlated with the fact
that 2-phenylethylamine is a rather large molecule having difficulties in
penetrating into the pores of FER. Noteworthy is the acidity measured
line with the pyridine adsorption-desorption results. As mentioned above
acidity of ZSM-12 is related to its pore size.
3
.1.5. Measurement of the oxidation states of Fe using Mössbauer
spectroscopy
The Mössbauer spectra of Fe-H-FER-SSIE, Fe-H-Beta-25-EIM and Fe-
-EIM catalysts are given in Fig. 6a–c. The oxidation states of Fe i.e.
2
for Fe-TiO -SSIE, which was relatively large based on 2-phenylaethy-
lamine adsorption in comparison to pyridine adsorption-desorption.
Both methods indicated very low acidity of Fe-ZSM-12-35-EIM re-
flecting difficulties of adsorbate penetration into narrow pores of this
Al
FeO, α-Fe
presence of magnetic, paramagnetic and super magnetic oxides. 1.8 wt
Fe-Ferrierite-20-SSIE, the most active and selective catalyst in ami-
2 3
O
0
2 3 3 4
O , Fe O and Fe studied in the above catalysts showed the
material.
27Al NMR showed clearly that tetrahedrally and octahedrally co-
%
ordinated Al species at 55 ppm and close to zero ppm were clearly visible
in the spectra [27]. The ratio between tetrahedrally to octahedrally co-
ordinated Al-species decreased in the following order: Fe-H-Beta-25-EIM∼
Fe-ZSM-12-35-EIM > > > Fe-H-Beta-150-SSIE (Table 2), which is in
dation of stearic acid, exhibited two paramagnetic compounds with
isomer shifts of 0.20 mm/s and 0.21 mm/s attributed to Fe (III). Fe-
FER-20-SSIE sample contained also large amounts of magnetic iron
2 3
oxides which was probably hematite, Fe O , about 60% and the two
353

P. Mäki-Arvela et al.
Applied Catalysis A, General 542 (2017) 350–358
2
Fig. 4. Scanning electron micrograph of Fe-SiO US-
5
0%-EIM catalyst.
other iron species with oxidation state of III which were paramagnetic.
The influence of Fe content on the oxidation states of Fe in Ferrierite
zeolite was previously studied using Mössbauer spectroscopy [28]
showing that the amount of Fe in Ferrierite influenced the oxidation
states. Isomer shifts for Fe (II) and Fe (III) were reported to be depen-
dent on the amount of Fe present in Ferrierite, in the study where the
iron loading was very low, below 0.3the molar ratio of Fe/Al [29].
4.8 wt% Fe-H-Beta-25-EIM zeolite prepared by the evaporation
impregnation method showed two paramagnetic compounds with
isomer shifts < 0.5 mm/s both tentatively attributed to Fe (III) at
0.4 mm/s, with one component being slightly shifted towards divalency
(
isomer shift 0.40 mm/s). It is also noteworthy to mention, that isomer
shifts of Fe were observed to be influenced by the type of structures for
Beta (BEA), Ferrierite (FER) and ZSM-5 (MFI) zeolites [29].
2 3
For 3.8 wt% Fe-Al O –EIM catalyst prepared using evaporation-
impregnation method, three paramagnetic compounds with isomer
shifts obtained were at 0.44 mm/s, 3.0 mm/s and 0.13 mm/s. The
compound with a negative isomer shift could be due to the presence of
Fe(IV). Analogously to 4.8 wt% Fe-H-Beta-25-EIM, all iron was also
Fig. 5. Acid site concentration from pyridine adsorption-desorption with gas phase versus
2-phenylethylamine adsorption from aqueous phase.
paramagnetic in 3.8 wt% Fe-Al
Mössbauer spectra results of the studied catalysts, Fe isomer shifts were
influenced by the amount of Fe, structure and acidity of the catalysts.
2
O
3
–EIM. Hence, based on the
Furthermore, catalytic activity and selectivity in stearic acid amidation
were also observed to be correlated with the isomer shifts of the studied
catalysts, and thus with the presence of different iron species.
Table 1
Brønsted and Lewis acidity of selected used catalysts.
Catalyst
Specific surface
Pore volume
(cm /gcat)
Brønsted acid sites (μmol/gcat
)
Lewis acid sites (μmol/gcat
)
Refs.
2
3
area (m /gcat
)
2
50 °C
350 °C
450 °C
250 °C
350 °C
450 °C
4
3
4
6
5
4
1
4
.6 wt% Fe-SiO
.8 wt% Fe-Al
.5 wt% Fe-TiO
.4 wt% Fe-ZSM-12-35-EIM
.3 wt% Fe-Beta-150-SSIE
.8 wt% Fe-Beta-25-EIM
.8 wt% Fe-Fer-20-SSIE
.3 wt% Sn-FER-20-EIM
2
-50% US-EIM
-EIM
-SSIE
389
261
90
109
467
953
384
949
464
0.68
1.13
0.27
0.04
0.59
0.65
4
12
0
4
12
0
2
6
0
0
45
17
74
5
3
62
0
1
12
0
0
4
[26]
[26]
2
O
3
140
117
12
119
95
2
This work
This work
[26]
This work
[24]
0
0
0
193
216
176
137
168
169
41
36
15
5
4
19
H-FER-20
326
284
170
8
7
4
[24]
354

P. Mäki-Arvela et al.
Applied Catalysis A, General 542 (2017) 350–358
3.2. Catalytic evaluation of stearic acid amidation
Table 2
Results from 27Al MAS NMR for different Fe-modified zeolites.
3
.2.1. Thermal amidation of stearic acid
Thermal amidation of stearic acid with an equimolar amount of
Catalyst
Al(tetrahedral Al(octahedral)
Al(tetrahedral)
/Al(octahedral)
(
ppm)
(ppm)
ethanolamine proceeded at 180 °C in an autoclave very rapidly in the
absence of any solvent giving 90% conversion within 1 h (Table 3). This
is much higher conversion than achieved at the same temperature in
hexane as a solvent, when conversion of stearic acid of only 61% after
5.3 wt% Fe-Beta-150-SSIE
4.8 wt% Fe-H-Beta-25-EIM
6.4 wt% Fe-ZSM-12-35-EIM
53.74
54.46
51 ±
−0.64
−0.65
−0 ±
4.0 ± 0.1
9.0
25
1
1
5
h was reported [6]. Thus it can be concluded that a higher reaction
rate is related to a higher initial stearic acid concentration. High ac-
tivity of undiluted stearic acid was supported by experimental data at
160 °C giving conversion 73% after 1 h. An important question is re-
lated to a potential impact of reactor walls and impeller material made
of stainless steel as they could also act as a catalyst in addition to the
2 3
Fig. 6. Mössbauer spectra (a) 1.8 wt% Fe-H-FER-20-SSIE zeolite catalyst, notation: 1. line corresponds to magnetic hematite, 2. and 3. lines correspond to paramagnetic trivalent Fe O ,
(
Fe
b) 4.8 wt% Fe-Beta-25-EIM zeolite catalyst, notation: the sample is fully paramagnetic, 1. line corresponds to trivalent iron, 2. line is most probably trivalent, but slightly shifted towards
2+
2 3
and (c) 3.8 wt% Fe-Al O -EIM catalyst, notation: all iron paramagnetic, 1. line trivalent iron, 2. line, trivalent, 20.3%, 3. line, corresponds possibly to tetravalent, 23.4%.
355

P. Mäki-Arvela et al.
Applied Catalysis A, General 542 (2017) 350–358
Table 3
Results from the amidation of stearic acid with an equimolar amount of ethanolamine in a batch reactor.
Entry Catalyst
Temperature
°C)
Mass of catalyst Time Conversion
Mole converted SA per Selectivity to amide
Selectivity to esteramide
(mol-%)
(
(g)
(h)
(%)
mol Fe
(mol-%)
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
No
No
No
No
No
No
180
160
140
140
120
120
120
120
120
120
120
120
120
120
–
–
–
–
–
–
2
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
1
1
1
5
1
5
1
5
5
1
5
5
1
5
5
90
73
7
78
4
15
17
26
17
22
32
11
32
32
29
–
–
92
95
0
0
–
–
–
2
104
–
72
106
37
75
76
95
97
100
95
95
90
–
96
97
93
94
97
100
0
0
0
3
0
4
0
0
0
3
0
Fe
3.8 wt% Fe-Al
Al
2
O
3
2
O
3
-EIM
2 3
O
0
1
2
3
4
5
4.6 wt% Fe-SiO
4.6 wt% Fe-SiO
4.5 wt% Fe-TiO -SSIE
2
6.4 wt% Fe-ZSM-12-35-EIM
6.4 wt% Fe-ZSM-12-35-EIM
4.6 wt% Fe-SiO
2
2
-50% US-EIM
-50% US-EIM
2
and H-6.4 wt% 120
ZSM-12-35-EIM
16
17
18
19
20
21
22
23
24
1.8 wt% Fe-FER-20-SSIE
1.8 wt% Fe-FER-20-SSIE
1.8 wt% Fe-FER-20-SSIE
Sn-FER-20-EIM
120
140
120
120
120
120
120
120
120
0.5
0.5
0.5
0.5
0.5
0.5
0.5
–
1
1
5
5
5
5
5
5
5
26
61
33
10
23
25
22
12
4
219
538
277
30
–
81
109
–
97
98
100
93
89
97
90
93
71
0
2
0
0
0
0
4
0
0
a
H-FER-20
4.8 wt% Fe-Beta-25-EIM
5.3 wt% Fe-Beta-150-SSIE
b
no
b
5.3 wt% Fe-Beta−150-SSIE
0.5
42
a
b
Calculated per mole of Sn.
Reactive distillation with stepwise addition of ethanolamine, 0.8 g per batch, corresponding to the molar ratio of SA:EA (stearic acid = SA, ethanol amine = EA) of 1:1 in each
3
0 min.
4
0
thermal reaction. A decrease of the reaction temperature resulted in
non-negligible conversions of stearic acid equal to 7% and 4%, re-
spectively at 140 °C and 120 °C. The subsequent step was to investigate
the catalytic effect of iron.
30
3.2.2. Iron catalyzed amidation of stearic acid
2 3
To investigate the effect of bulk iron, 2 g of Fe O was used as a
catalyst for 0.31 mol of stearic acid with an equimolar amount of
ethanolamine at 120 °C. The result showed that even bulk iron oxide
has a catalytic effect, since the conversion after 1 h was 17% (Table 3),
being in the absence of any catalyst only 4%.
A systematic catalyst screening study for amidation of stearic acid
was performed at 120 °C, since under these conditions noncatalytic
stearic acid conversion was only 4% and 15% after 1 and 5 h, respec-
20
10
2 3
tively. When the supported iron catalyst (Fe-Al O -EIM) was used, the
0
conversion after 5 h was 26% showing clearly that it acts more effi-
ciently as a catalyst than plain hematite. In addition, catalyst pro-
ductivity, calculated as the moles of converted stearic acid per mole of
Thermal
FeTiO2
FeAl2O3
FeSiO2
FeZSM12 FeBeta25 FeBeta150 Fe FER
Sn FER
H FER
Fig. 7. Conversion of stearic acid in its amidation with an equimolar amount of etha-
nolamine at 120 °C after 1 h (red bar) or 5 h (blue bar) using 0.5 g of different catalyst. In
comparison results from thermal amidation are given. (For interpretation of the refer-
ences to colour in this figure legend, the reader is referred to the web version of this
article.)
iron was much higher for Fe-Al
2 3
O -EIM compared to hematite
(
Table 3). When only Al was used, conversion of 17% was achieved
2 3
O
in 5 h also reflecting the catalytic effect of iron.
To avoid interference of sampling the solvent-free amidation of
stearic acid with an equimolar amount of ethanolamine was conducted
by analyzing only the final mixtures for experiments lasting either 1 or
using the Gibbs free energy at 120 °C, calculated with HyperChem
programme [7]. ΔG393K was equal to −0.1 kJ/(K mol) giving the
equilibrium conversion of 50% showing that the experimental conver-
sion was lower than corresponding to the equilibrium one.
5
4
h. The results revealed that conversion of stearic acid increased with
.6 wt% Fe-SiO -EIM and 1.8 wt% Fe-FER-20-SSIE when the reaction
2
time was prolonged from 1 h to 5 h (Fig. 7). On the other hand, with the
medium pore size zeolite, ZSM-12-35-EIM the same conversion levels
were obtained after 1 and 5 h indicating strong deactivation which can
be explained by its small pore size compared to the size of the reactants
and products leading to pore blockage. The conversion levels, however,
remained in the range of 32 − 33% after 5 h experiment for 4.6 wt%
Fe-SiO -50% US-EIM, 6.4 wt% Fe-ZSM-12-35-EIM and 1.8 wt% Fe-FER-
2
SSIE-20. In order to check if these conversion levels correspond to the
equilibrium, the latter were determined for amidation of stearic acid
In order to compare catalytic activity in terms of turnover fre-
quencies (TOF) of different catalysts, the particle sizes of different iron
species, for example hematite, should be determined. While XRD results
indicate presence of hematite particles larger than 3 nm, it is, however,
difficult to approximate the sizes due to partially overlapping peaks in
pristine Ferrierite. Iron particle sizes could be determined by CO che-
misorption at −78 °C [30], which is not feasible with the setup used in
this work. Moreover, in the current case the catalyst was not reduced
prior to the reaction, as done for example in Fischer-Tropsch synthesis,
356

P. Mäki-Arvela et al.
Applied Catalysis A, General 542 (2017) 350–358
Scheme 1. Reaction scheme for stearic acid (1) amidation with ethanolamine (2). Notation: stearylethanolamide (3) and esteramide (4).
Fe-ZSM-12-35
allowing correlation of hydrogen uptake with catalytic activity [31].
Due to such specific features of iron containing catalyst, the catalytic
activity is typically correlated only with the oxidation state of iron as
for example in benzene oxidation to phenol [32].
Fe-FER-20-SSIE
Fe-SiO -50% US
30
2
Fe-Al O -EIM
2 3
When the batch operation mode was used, the formed water was
accumulating in the reactor, shifting the reaction towards the reactants
Fe-Beta-25-EIM
H-FER-20
Fe-Beta-150-SSIE
(
Scheme 1). Thus some experiments were also performed under atmo-
20
spheric pressure using a reactive distillation option, when water can be
evaporated continuously from the reactor while ethanolamine was
added in several batches. Acceptable levels of ethanolamine vapor
pressure at 120 °C permitted injection of ethanolamine into the stearic
acid melt containing the catalyst. The conversion remained, however,
at a quite low level, most probably due to high viscosity of the reaction
mixture retarding water evaporation.
Fe-TiO -SSIE
2
10
The effect of specific surface area on iron dispersion and subse-
quently on stearic acid reactivity in amidation was clearly visible, for
example with oxide supported Fe-catalysts. The lowest conversion
0
0
5
10
15
20
25
30
35
40
45
Brønsted acid sites concentration/Lewis acid site concentration
among Fe-oxide catalysts was achieved with Fe-TiO
2
–SSIE exhibiting
the lowest specific surface area of 90 m /gcat followed by Fe-Al -EIM
and Fe-SiO -EIM. Fe-FER-20-SSIE, which also gave a relatively high
2
2 3
O
Fig. 8. Conversion of stearic acid in its amidation with an equimolar amount of etha-
nolamine at 120 °C after 5 h using 0.5 g of different catalysts as a function of the ratio of
Brønsted to Lewis acid site concentrations.
2
conversion of stearic acid, possesses a high specific surface area. It
should, however, be stated that due to strong acidity of some zeolites,
with high specific surface area and high dispersion of iron, e.g. Fe-Beta-
150-SSIE, catalyst deactivation is prominent. When comparing the
performance of different catalysts in terms of productivity (moles of
converted stearic acid per moles of iron), Fe-FER-20-SSIE is much better
than Fe-SiO -50% US-EIM and Fe-ZSM-12-35-EIM, with the corre-
2
sponding ratios being 277, 106 and 76, respectively. In addition,
bearing in mind that for both Fe-FER-20-SSIE and Fe-ZSM-12-35-EIM
all iron and acidic sites are not accessible, the result is even more ap-
parent indicating that Fe-FER-20-SSIE is the best catalyst among tested
due to its optimum acidity.
The effect of the type, strength and concentrations of acid sites in
the studied catalysts on stearic acid amidation can be seen in Figs. 8 and
9
. The results showed that the highest conversion was obtained with Fe-
ZSM-12-35, Fe-SiO –EIM and Fe-FER-20-SSIE, which exhibit different
2
acidities. The acid sites in Fe-ZSM-12-35-EIM are most probably not
accessible for the reactants due to too small pores of the zeolite. Fe-
SiO –EIM exhibits only very low acidity, being still active in stearic acid
2
amidation. Fe-FER-20-SSIE with a relatively high acidity and especially
a high amount of strong acid sites seems to be beneficial for the reac-
tion. Noteworthy is that Fe-Beta-25-EIM with relatively large pores, but
only a small amount of strong acid sites, gave conversion lower than Fe-
FER-20-SSIE. This result indicates that also presence of strong Brønsted
acid sites is beneficial for amidation. On the other hand, when only H-
FER was used as a catalyst, conversion of stearic acid was low, H-FER
exhibited nearly two fold strong acid sites compared to Fe-FER-SSIE-20.
It should also be noted that Sn-FER-20-EIM did not have any catalytic
effect, moreover, conversion with this material was even lower than
thermal amidation per se. The effect of the ratio between Brønsted to
Lewis acid site concentration on stearic acid conversion is also visible in
Fig. 9. Conversion of stearic acid in its amidation with an equimolar amount of etha-
nolamine at 120 °C after 5 h using 0.5 g of different catalyst as a function of the sum of
Brønsted and Lewis acidity of the catalyst (measured by pyridine adsorption).
Fig. 8. It can be clearly seen that conversion was the lowest with the
catalysts having only Lewis acidity, such as Fe-TiO
low specific surface area. A relatively high conversion was achieved
with mildly acidic Fe-SiO -50% US-EIM with a large specific surface
2
-SSIE, exhibiting
2
area. When the ratio between Brønsted to Lewis sites increased, the
conversion of stearic acid increased. A high conversion was also
achieved with Fe-FER-20-SSIE. H-FER exhibited too high concentration
of Brønsted acid sites leading to catalyst deactivation. These results
357

P. Mäki-Arvela et al.
Applied Catalysis A, General 542 (2017) 350–358
indicate that too high amount of strong Brønsted acid sites is not fa-
vorable for the amidation of stearic acid or alternatively that Fe is
highly dispersed on a non-acidic catalyst. Selectivity to amide was al-
ways very high and only a limited amount of esteramide was formed.
From the above mentioned results, it can be concluded that sup-
ported iron catalysts have a clear catalytic effect in stearic acid ami-
dation with an equimolar amount of ethanolamine in the absence of any
solvent. High iron dispersion is beneficial for stearic acid amidation. An
optimum amount of strong Brønsted acid sites in addition to high Fe
dispersion is also required for catalyst efficiency. For the catalyst most
active at 120 °C, Fe-FER-20-SSIE, amidation of stearic acid was also
performed at 140 °C. A blank experiment in the absence of any catalyst
was also conducted. The results revealed that the non-catalyzed ami-
dation at 140 °C gave only 7% conversion after 1 h of reaction time,
whereas with Fe-FER-20-SSIE the conversion was 61%. In addition to a
relatively low concentration of Brønsted acid sites, this catalyst con-
are needed to fully optimize the catalyst properties, application of
supported iron catalysts in solvent-free conditions is a promising and an
environmentally benign route to produce fatty alkanolamides.
References
[1] L. Hamtiaux, J. Masquelier, G.G. Muccioli, C. Bouzin, O. Feron, B. Gallez,
D.M. Lambert, BMC Cancer 12 92 (2012) 1–14.
[
2] R. Romani, R. Galezzi, G. Rosi, R. Fiorini, I. Pirisinu, A. Ambrosini, G. Zolese,
Biochimie 93 (2011) 1584–1591.
[3] S. Chaudhari, J. MacDougall, J. Peters, J. Ramsbottom, Preparation of amino acid
fatty acid amides, US patent 2008/0200704 A1.
[
4] Y.F. Suen, S.L. Jensen, Method for preparing mono or dialkanol amides,
WO2013154689, 2013, Chevron Oronite Company LLC
[5] A. Tkacheva, I. Dosmagambetova, Y. Chapellier, P. Mäki-Arvela, I. Hachemi,
R. Savela, R. Leino, C. Viegas, N. Kumar, K. Eränen, J. Hemming, A. Smeds, D. Yu
Murzin, ChemSusChem 8 (2015) 2670–2680.
[
6] P. Mäki-Arvela, A. Tkacheva, I. Dosmagambetova, Y. Chapelliere, I. Hachemi,
N. Kumar, A. Aho, D. Yu Murzin, Top. Catal. 59 (2016) 1151–1164.
[7] P. Mäki-Arvela, N. Kumar, Y. Chapelliere, I.L. Simakova, D. Yu. Murzin, React.
3
+
tained mostly iron as Fe
species which are apparently promoting
Kinet. Mech. Catal. 120 (2017) 15–29.
amidation of stearic acid. This result is very promising showing clearly
that a bifunctional Fe-modified zeolite is active and selective catalyst
for production of stearoyl ethanolamide in the absence of any solvent.
[
8] M. Musteata, V. Musteata, A. Dinu, M. Florea, V.T. Hoang, D. Trong-On,
S. Kaliaguine, V.I. Parvulescu, Pure Appl. Chem. 79 (2007) 2059–2068.
[9] B. Sheedhar, V. Bhaskar, Ch. Sridkar, T. Srinivas, L. Kotai, K. Szenthályi, J. Mol.
Catal. A: Chem. 191 (2003) 141–147.
[
[
10] K.V.V. Krishna Mohan, N. Narender, S.J. Kulkarni, Green Chem. 6 (2006) 368–372.
11] S. Ghosh, A. Bhaumik, J. Mondal, A. Mallik, S. Sengupta (Bandyopadhyay),
C. Mukhopahahya, Green Chem. 14 (2012) 3220–3229.
4. Conclusions
[
[
[
[
12] R. Gava, A. Biffis, C. Turbaro, F. Zaccheria, N. Ravasio, Catal. Commun. 40 (2013)
Solvent-free amidation of stearic acid with ethanolamine for pro-
duction of surfactants and pharmaceuticals was investigated for the first
time using inexpensive iron-supported catalysts.
6
3–65.
13] R. Fu, Y. Yang, W. Feng, Q. Ge, Y. Feng, X. Zeng, W. Chai, J. Yi, R. Yuan,
Tetrahedron 72 (2016) 8319–8326.
14] M. Mirza-Aghayan, M.M. Tavana, R. Boukherroub, Ultrason. Sonochem. 29 (2016)
2 2 3 2
Iron-modified SiO , Al O , TiO as well as Fe containing Ferrierite,
3
71–379.
ZSM-12 and Beta zeolite catalysts were synthesized using solid state
ion-exchange and evaporation impregnation methods. Structural in-
tegrity of the Fe modified catalysts studied with X-ray powder diffrac-
tion and scanning electron microscopy showed that the parent struc-
tures of the support were kept intact. Mössbauer spectroscopy of the
most active and selective catalyst in the amidation of stearic acid,
15] R.B. La Pierre, A.C. Rohrman Jr., J.L. Schlenker, J.D. Wood, M.K. Rubin,
W.J. Rohrbaugh, Zeolite 5 (1985) 346–348.
[16] X. Wei, Ps G. Smirniotis, Microporous Mesoporous Mater. 97 (2006) 97–106.
[
[
[
17] K.V.N.S. Srinivas, E. Bolla Reddy, B. Das, SynLett 4 (2002) 625–627.
18] S.-C. Wu, J.S.B. Wang, T.-C. Tsai, Top. Catal. 53 (2010) 1419–1429.
19] M.M. Dubinin, Carbon 27 (1989) 457–467.
[20] S.G. Chen, R.T. Yang, Langmuir 10 (1994) 4244–4249.
[
21] C.A. Emeis, J. Catal. 141 (1993) 347–354.
1
4
.8 wt% Fe-Ferrierite-20-SSIE exhibited the presence of Fe (III) species.
.8 wt% Fe-Beta-25-EIM contained Fe (II) and Fe (III), whereas iron in
[
22] V. Rac, V. Rakic, D. Stosic, O. Otman, A. Auroux, Microporous Mesoporous Mater.
1
94 (2014) 126–134.
2 3
mildly acidic Fe-Al O -EIM was present as magnetite with Fe (IV) and
hematite containing Fe (III), respectively.
[23] W. Guo, C. Xiong, L. Huang, Q. Li, J. Mater. Chem. 11 (2001) 1886–1890.
[
24] N. Kumar, P. Mäki-Arvela, S. Faten Diaz, A. Aho, Y. Demidova, J. Linden,
A. Shepidchenko, M. Tenhu, J. Salonen, P. Laukkanen, A. Lashkul, J. Dahl, I. Sinev,
A.-R. Leino, K. Kordas, T. Salmi, D. Yu Murzin, Top. Catal. 56 (2013) 696–713.
The most active and selective catalyst for amidation of stearic acid
with ethanolamine was 1.8 wt% Fe-Ferrierite-20-SSIE containing
[25] D. Peng, S. Beysen, Q. Li, Y. Sun, L. Yang, Particuology 8 (2010) 386–389.
[26] M. Stekrova, N. Kumar, A. Aho, I. Sinev, W. Grunert, J. Dahl, J. Roine,
S. Arzumanov, P. Mäki-Arvela, D. Yu Murzin, Appl. Catal. A. Gen. 470 (2014)
3
+
mainly Fe
especially a low amount of Brønsted acid sites. The highest conversion
61%) of stearic acid and selectivity (98%) to amide at 140 °C in 1 h
species and an optimum concentration of acid sites,
1
62–176.
(
[27] P.P. Pescarmona, K.P.F. Janssen, C. Strobants, B. Molle, J.S. Paul, P.A. Jacobs,
B.F. Sels, Top. Catal. 53 (2010) 77–85.
was obtained over this catalyst, which was superior to other catalysts in
terms of catalyst productivity per mole of iron. In addition, a nonacidic
Fe-SiO with large pores and slightly higher iron loading gave also high
2
conversion. Noteworthy was also that Fe-ZSM-12-35-EIM with a smaller
pore size facilitated high conversion of stearic acid despite accessibility
of only external surface.
[
28] E. Tabor, K. Zavet, N.K. Sathu, Z. Tvaruzkova, Z. Sobalik, Catal. Today 169 (2011)
6–23.
[29] K. Jisa, J. Novakova, M. Schwarze, A. Vondrova, S. Sklenak, Z. Sobalik, J. Catal. 262
2009) 27–34.
1
(
[
[
30] P.H. Emmett, S. Brunauer, J. Am. Chem. Soc. 59 (1937) 310–315.
31] C.-H. Zhang, Y. Yang, B.-T. Teng, T.-Z. Li, H.-Y. Zheng, H.-W. Xiang, Y.-W. Li, J.
Catal. 237 (2006) 405–415.
These results indicated that although there are several parameters
influencing selectivity in amidation of stearic acid and further studies
[32] E. Hensen, Q. Zhu, P.-H. Liu, K.J. Chao, R. van Santen, J. Catal. 226 (2004)
66–470.
4
358