First row transition metals on the ethanol Guerbet reaction: Products distribution and structural behavior of mixed metal oxides as catalysts

Article abstract of DOI:10.1016/j.apcata.2021.118272

Described is the transformation of ethanol into 1-butanol, acetaldehyde and other products by the Guerbet reaction (GR) over mixed metal oxides (MMOs) as catalysts. The MMOs, in which Mg²⁺ was partially (20 mol%) substituted by Fe²⁺, Co²⁺, Ni²⁺, Cu²⁺, and Zn²⁺, were obtained from hydrotalcite precursors and the reactions conducted in a fixed bed flow reactor. EPR was used to observe oxygen vacancies after the catalytic reactions, which may be related to the ethanol reactivity due to the basicity enhancement of the catalyst. Carbon deposition, mostly filamentous, was observed in all catalysts and a trend between the metal-carbon bond energy and the percentage of deposited carbon was established. This correlation and the catalyst product distribution can be useful to tailor new catalysts. Using four parameters, ethanol conversion, 1-butanol and side-product selectivities and percentage of carbon deposition, Zn₂₀MMO proved to be the best choice for GR.

Full text of DOI:10.1016/j.apcata.2021.118272

Applied Catalysis A, General 623 (2021) 118272
Contents lists available at ScienceDirect
Applied Catalysis A, General
journal homepage: www.elsevier.com/locate/apcata
First row transition metals on the ethanol Guerbet reaction: Products
distribution and structural behavior of mixed metal oxides as catalysts
a
,
a
a
Gustavo Metzker *, Jorge Andres Mora Vargas , Livia Padilha de Lima ,
a
a
b
Olavo Micali Perrone , Marcos Rechi Siqueira , Laudemir Carlos Varanda ,
a
,
Mauricio Boscolo *
a
Sao Paulo State University (UNESP), Chemistry and Environmental Sciences Department (IBILCE), Sao Jose do Rio Preto, SP, Brazil
University of S a˜ o Paulo (USP), Sao Carlos Institute of Chemistry (IQSC), Sao Carlos, SP, Brazil
b
A R T I C L E I N F O
A B S T R A C T
Keywords:
Described is the transformation of ethanol into 1-butanol, acetaldehyde and other products by the Guerbet re-
Ethanol
2+
action (GR) over mixed metal oxides (MMOs) as catalysts. The MMOs, in which Mg was partially (20 mol%)
Guerbet reaction
First-row transition metals
2+
2+
2+
2+
2+
substituted by Fe , Co , Ni , Cu , and Zn , were obtained from hydrotalcite precursors and the reactions
conducted in a fixed bed flow reactor. EPR was used to observe oxygen vacancies after the catalytic reactions,
which may be related to the ethanol reactivity due to the basicity enhancement of the catalyst. Carbon depo-
sition, mostly filamentous, was observed in all catalysts and a trend between the metal-carbon bond energy and
the percentage of deposited carbon was established. This correlation and the catalyst product distribution can be
useful to tailor new catalysts. Using four parameters, ethanol conversion, 1-butanol and side-product selectivities
and percentage of carbon deposition, Zn20MMO proved to be the best choice for GR.
1
-Butanol
1
. Introduction
One of the challenges of twenty-first-century chemistry is discov-
the desired route.
The ethanol condensation mechanism was proposed by Guerbet at
the beginning of the twenty-century [16,17] and has been studied in
detail since then. The GR is always catalyzed, and many classes of cat-
alysts, homo- and heterogenous, have already been proposed. [18–30].
Long-chain carbon alcohols can be obtained by ethanol dehydrogena-
tion into acetaldehyde, which reacts with another acetaldehyde mole-
cule in an aldol condensation reaction. The condensation product
(crotonaldehyde) is obtained by a dehydration reaction and the final
product, 1-butanol, is formed by the hydrogenation of the olefinic
double bond and the aldehyde. The presence of acetaldehyde in the
reaction medium allows the formation of side products such as ethers
and esters. Focusing on the GR products, the simplified mechanism is
depicted in Fig. 1.
ering new synthetic routes, catalysts and industrial processes to try and
obtain raw chemicals and fuels, which nowadays are almost exclusively
derived from petrol [1–4]. From this perspective, ethanol has been the
choice starting material for the aforementioned purposes since it has an
intrinsically renewable nature (obtained by fermentation/distillation of
sugars from vegetables) and there is already infra-structure installed for
high-volume production. These enable it to be a new renewable carbon
source [5–8].
There are two main approaches to using ethanol as a starting mate-
rial for obtaining raw chemicals. The first one is the ethanol reforming
2
reaction, aiming to produce hydrogen (H ) [9,10]. The second approach
is the ethanol condensation reaction (the Guerbet reaction (GR)), aiming
to produce raw chemicals with long-chain carbon compounds, mostly
alcohols (1-butanol, hexanol, octanol) and aldehydes, which are the
reactive intermediates for GR [11–17]. Although the products from both
approaches, hydrogen and condensed products, are interesting from the
economic and environmental point of view, the routes are competitive.
Thus, the catalyst and the reaction parameters must be tuned to favor
As depicted in Fig. 1, the acetaldehyde reactivity is the key-step for
4
GR, since it can give rise to esters and to CO and CH . These reforming
reaction products present a well-established route to carbon deposition
over the catalyst (nanostructured and amorphous carbon through acet-
aldehyde decomposition (1), methane dehydrogenation (2), and the
Boudouard reaction (3) that can poison the catalyst [31–41]. Thus, the
control of carbon deposition is a key-issue for long-term stability of the
*
Corresponding authors.
E-mail addresses: g.metzker@unesp.br (G. Metzker), mauricio.boscolo@unesp.br (M. Boscolo).
https://doi.org/10.1016/j.apcata.2021.118272
Received 9 March 2021; Received in revised form 17 June 2021; Accepted 20 June 2021
Available online 25 June 2021
0
926-860X/© 2021 Elsevier B.V. All rights reserved.

G. Metzker et al.
Applied Catalysis A, General 623 (2021) 118272
catalysts.
corrected to 10 using 1.0 M solution of NaOH. In 160 mL of water, 0.096
Due to the nature of the coupling reaction, the catalyst must perform
mols of Mg(NO
metal nitrate and 0.04 mols of Al(NO
dropped into Na CO solution, checking the pH and keeping it at 10
3
)
2
⋅6H
2
O, 0.024 mols of the desired first-row transition
both dehydrogenation and hydrogenation reactions. This dual behavior
can be obtained by using mixed metal oxide (MMO) catalysts, which
have already shown to be active in methanol, ethanol and butanol cat-
alytic transformations [20,42–47]. The MMOs are a large class of ma-
terials synthesized by many different routes [48–50]. Our research
group has been using hydrotalcite-like materials (HTC), composed pri-
3
)
3
⋅9H O were dissolved and slowly
2
2
3
using the 1.0 M NaOH solution. After the addition of the nitrate solu-
◦
tions, the resulting slurry was aged for 24 h in oven at 60 C. The
resulting material was filtered, and exhaustively washed with deionized
water, until the washing solution reached pH 7. The resulting solid was
marily by Mg2 and Al , as a precursor of MMOs, due to its simple
+
3+
dried in an oven at 110 C for 12 h, powdered and kept in vacuum
◦
2
+
synthetic route, reproducibility and the possibility of replacing Mg or
desiccator. The same procedure was utilized to synthetize the reference
material, without magnesium substitution (using 0.12 mols of Mg
3
+
Al ions by transition metals or lanthanide ions [18,29,43,51,52]. The
2
+
main research focus is the substitution of Mg ions by divalent ions of
the first-row transition metal series due to their lower cost and higher
abundance, both being essential prerequisites for the development of
new catalysts for upgrading ethanol into added-value products. Also,
there are no reports in the literature comparing the reactivity of a series
of MMOs in which Mg2 was partially substituted by divalent metal ions
of the first-row transition metal series and the catalyst behavior of these
materials in GR.
(NO
3
)
2
⋅6H
2
O, and 0.04 mols of Al(NO
3
)
3
⋅9H
2
O). The HTCs were labeled
as M20HTC, in which M represents the transition metal cation that
2
+
substituted the Mg , 20 means 20 mol% of substitution, and HTC
means hydrotalcite and characterized by X-Ray diffraction (XRD). The
◦
HTCs were calcined at 450 C for 5 h under an atmosphere of air to
+
obtain the MMOs. The MMOs were labeled as M20MMO and MMO
following the same labeling procedure as described before. The MMOs,
before and after the reaction, were characterized by XRD, surface area
by the BET method, scanning electron microscopy (SEM),
energy-dispersive X-ray spectroscopy (EDS), electron paramagnetic
resonance (EPR) and thermogravimetric analysis (TGA). The full
description of the equipment and methods employed for HTC and MMO
characterization, before and after the catalytic reactions, is described in
the Supplementary Material.
Here, the synthesis and characterization of a series of HTCs con-
taining Fe2 , Co , Ni , Cu
+
2+
2+
2+
and Zn as precursors of MMOs
2+
employed as catalysts for GR is described. The reactions were carried out
◦
in a fixed bed flow reactor (FBFR), operating at 450 C, at atmospheric
pressure and the distribution of the products by class (alcohols, alde-
hydes, ethers, and esters) was observed. The temporal behaviors for the
main reaction products were compared for each catalyst. The catalysts
were also characterized after the catalytic runs, aiming to gain insights
into the catalyst stability, reactivity, and carbon deposition.
2.2. Fixed bed flow reactor (FBFR)
The research group developed the FBFR employed in the catalytic
reactions. It was composed of a Jasco pump (model PU-980), an oven
equipped with a heater/thermocouple controlled by a Watlow micro
processed based auto-tuning temperature controller (model series-96).
The reactor column was machined using stainless steel (316 L alloy),
with dimensions of 12.0 × 1.0 × 0.5 cm (L x O.D. x I.D.). The column was
2
. Experimental section
2
.1. Catalysts synthesis and characterization
The hydrotalcites (HTCs), which are the mixed metal oxide (MMO)
◦
precursors, were synthesized using the co-precipitation method as
coupled with a stainless-steel condenser in which cold water (4 C)
described in the literature, using an Mg2 /Al ratio of 3:1 [48,49,53].
The Mg(NO O (99 %), Al(NO ⋅9H O (98 %), Na CO (99.5 %),
⋅6H
NaOH (97 %), FeCl O (98 %), Co(NO ⋅6H O (>98 %), Ni
⋅4H
NO O (> 97 %) Cu(NO ⋅3H O (99 %), Zn(NO ⋅6H O (98 %)
⋅6H
+
3+
continuously circulated to condense the outcoming products. The
condenser was coupled to a rubber septum glass flask to collect the
products. The flask has a gas output, connected to a graduated cylinder
to measure the gas volume generated during the reaction. The reactor
scheme is presented in Fig. S1.
3
)
2
2
3
)
3
2
2
3
2
2
3
)
2
2
(
3
)
2
2
3
)
2
2
3
)
2
2
salts were purchased from Sigma-Aldrich and used without further pu-
rification. Deionized water was utilized in all synthetic procedures and
solutions. For the HTC synthesis, 0.2 mols of Na
2
CO
3
was dissolved into
◦
1
20 mL of water and heated at 60 C under stirring. The solution pH was
Fig. 1. Major ethanol reaction pathways for the Guerbet reaction (GR) and undesired side products.
2

G. Metzker et al.
Applied Catalysis A, General 623 (2021) 118272
2
.3. Catalytic reactions
3.1.1. Ethanol consumption and catalyst stability
Aiming to investigate the catalyst stabilities and the ethanol con-
sumption, all the catalysts and the reference material were submitted to
All the reactions were carried out using ethanol 95 % (5 % of water)
◦
pumped (flow of 0.2 mL/min) through a heated 15 cm stainless steel
tube, which allowed ethanol to vaporize before entering the reactor
column head. No carrier gas was employed. The reactor column was
filled with a mixture of 0.3 g of M20MMOs and 1.4 g of silica to disperse
the catalyst homogeneously. After packing the column and before
heating the reactor, the ethanol flow was turned on and the volume was
reactions in the FBFR at 450 C for 6 h and the ethanol consumption
followed by GC–MS. The results are summarized in Fig. 2.
As can be observed in Fig. 2, after 60 min of reaction, all the catalysts
reached the maximum ethanol consumption. However, values were
distinct depending on the metal substitution, the highest being for
Zn20MMO (85 %) and the lowest for Fe20MMO (47 %).
◦
checked before starting the heating ramp (20 C/min). Once the tem-
Analyzing the temporal behavior of ethanol consumption for each
catalyst, the maximum values for ethanol consumption were not main-
tained during the 6 h period for all MMOs tested. At this point, it is
possible to divide the catalysts into two distinct groups: MMO,
Ni20MMO, and Zn20MMO were the ones that presented a small decrease
in the ethanol consumption and then kept it constant; Fe20MMO,
Co20MMO, and Cu20MMO had the ethanol consumption gradually
decreased, indicating that these catalysts were deactivated during the
reaction period of 6 h. It is important to mention that Cu20MMO reached
95 % of ethanol consumption with 30 min of reaction but lost almost 50
% of its activity after 6 h (47 %).
◦
perature reached 450 C, the catalysts were activated in situ by ethanol,
which corresponds to the first 30 min of reaction until the products
reached a plateau. The collector flasks were changed every 30 min in a
total of 6 h of reaction. The flasks containing the condensed products
◦
were kept in a freezer (ꢀ 20 C) until the analysis.
2
.4. Products analysis and quantification
The reaction product identification and quantification were carried
out using a GC Clarus 500 gas chromatograph coupled with a 600 T mass
spectrometer detector (Perkin Elmer). Chromatographic conditions: He
◦
as carrier gas (2.0 mL/min), injector temperature at 250 C, injection
volume of 1
μ
L, BD-EN14103 column (30 m x0.320 mm x 0.25
μ
m,
3.2. Product distribution
◦
◦
Agilent), temperature program: 40 C for 10 min, 10 C/min until 160
◦
◦
◦
◦
◦
C, 2 C/min until 180 C, 10 C/min until 240 C holding 2 min. Mass
All the catalysts were able to produce four classes of compounds in
the following order of abundance: alcohols, aldehydes, ethers and esters.
The abundance of each class depended on divalent transition metal ions
spectrometer conditions: an m/z ratio ranging from 40 to 300 and an
ionization energy of 70 eV. Before GC–MS analysis, the samples were
prepared adding 20
μ
L of the collected solution, 10
μ
L of ethyl dodec-
S
incorporated in the MMO . Fig. 3A shows the product distribution by
anoate (internal standard) and 970
μL of MS grade methanol. The
class for each catalyst, and Fig. 3B shows the main products for each
class.
products were quantified by the correlation factor determined for a se-
ries of classes of compounds (alcohols, ketones, ethers, esters, five-
member rings, six-member aromatics) as described in a previous paper
As shown in Fig. 3, Ni20MMO is the catalyst that produces more al-
cohols, followed by Cu20MMO, MMO, and Zn20MMO. For all catalysts,
[
43].
3
. Results and discussion
3
.1. Catalyst characterization
The catalyst precursors, HTCs, were characterized by XRD. All the
materials presented the XRD pattern characteristic of HTCs with seven
2
+
2+
peaks (Fig. S2), indicating that the partial substitution of Mg by Fe ,
2
+
2+
2+
2+
Co , Ni , Cu , and Zn was successfully achieved [43,48,51]. Also,
the calculation of the crystallographic parameters for each material
(
Table S1) indicates that incorporating the above-mentioned ions in the
hydrotalcite structure did not drastically change the hydrotalcite
structure.
◦
After the HTC calcination at 450 C for 5 h, the materials were
converted into MMOs. The MMOs were characterized by XRD, indi-
cating the complete change in the material structure, since the charac-
teristic 7-peak pattern of the HTC was replaced by the 3-peak
characteristic patterns of the MMOs (Fig. S3) [43].
The MMOs were also characterized by surface area, and acidity and
basicity measurements. The surface area of the MMOs is coherent with
2
the values reported in the literature [26,28,29,43] of around 100 m /g
(
Table S2), with Zn20MMO being the catalyst with the highest surface
2
area (132.9 ± 1.4 m /g) and Co20MMO that with the lowest (92.6 ± 1.4
2
m /g). Analyzing the values of acidity and basicity and comparing them
2
+
with the reference material (MMO), the partial substitution of Mg by
2
+
2+
2+
2+
2+
Fe , Co , Ni , Cu , and Zn induced a decrease in the basicity and
an increase of the acidity values (Table S2). These changes are important
to the reactivity of the catalysts, since it is already reported in the
literature that the balance between the acidity and basicity in the
catalyst is crucial for its bifunctional behavior, which is important for
the GR.
Fig. 2. Temporal behavior for the ethanol consumption over MMOs. Reaction
conditions: T =450 ◦C, 6 h reaction, sampling every 30 min.
3

G. Metzker et al.
Applied Catalysis A, General 623 (2021) 118272
Fig. 3. (A) Product distribution for catalytic transformation of ethanol by the MMOs. Reaction conditions: T =450 ◦C, 6 h of reaction. The bars represent the sum of
products for each class during the 6 h of reaction. (B) Chemical structures of the main products for each class. The red structures represent the more abundant product
for each class of compound. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
the major alcohol obtained is 1-butanol (>95 %) with small amounts of
-hexanol (mostly for the Cu20MMO catalyst). Considering the alde-
Taking into account the reaction pathways shown in Fig. 1, diethyl
ether and ethyl acetate can be considered unwanted products for GR. For
diethyl ether, the temporal profiles for each catalyst changed regarding
1-butanol and acetaldehyde. Fe20MMO showed a constant increase
reaching 32 % selectivity at 360 min while Co20MMO, Ni20MMO,
Cu20MMO and Zn20MMO reached a plateau at distinct reactions times,
with selectivities of 16.7, 25.9, 12.2 and 6.3 % respectively. For MMO,
the diethyl ether reached a maximum of 40 % selectivity at 30 min of
reaction and decreased until reaching a plateau at 19 %.
1
hydes, the catalysts Co20MMO and Zn20MMO produce almost twice as
much than the other catalysts, acetaldehyde (ca. 75 %) being the major
aldehyde obtained, followed by butyraldehyde (ca. 25 %). The third
class of products, ethers, was higher for the MMO and Fe20MMO and
lower for the Zn20MMO, diethyl ether being the most abundant product
for this class, followed by butyl ethyl ether. Finally, the last class in
terms of abundance were esters, which contributed to less than 15 % of
the product composition for all MMOs, ethyl acetate and ethyl butanoate
being the major ones. Note that no esters were observed in the product
distribution for the reference MMO.
Finally, regarding ethyl acetate, the selectivities for all catalysts were
by far the lowest compared with the other three products. For MMO and
Ni20MMO, almost no ethyl acetate was detected. Fe20MMO presented
oscillatory behavior with selectivity ca. 2.8 %. For Co20MMO, Cu20MMO
and Zn20MMO, a plateau was reached with selectivities of 4.8, 3.9 and
3
.3. Product temporal behavior
1
0.6 %, respectively.
The selectivities for the main products of each class (1-butanol,
Taking into account the ethanol conversion and temporal behavior of
acetaldehyde, diethyl ether and ethyl acetate) were followed by 6 h for
all catalysts as shown in Fig. 4. The other identified and quantified
products for each catalyst are shown in Tables S3–S8.
1-butanol, acetaldehyde, diethyl ether and ethyl acetate, it is possible to
rank the catalysts in order of choice for GR. Ethanol conversion is time-
stable for MMO, Ni20MMO and Zn20MMO, making the other three cat-
alysts less attractive since the continuous loss of ethanol conversion
impacts on the desired products (1-butanol and acetaldehyde) for GR.
Thus, Fe20MMO, Co20MMO and Cu20MMO, even with good selectivities
for the desired products in the temporal behavior analysis, would not be
the best choice. Taking into account the temporal behavior of the
products, MMO, Ni20MMO and Zn20MMO seem to be the best choices for
GR because, apart from ethanol conversion, they presented high selec-
tivities for 1-butanol and acetaldehyde and lower selectivities for the
side products (diethyl ether and ethyl acetate).
In general, the four products mentioned above reached maximum
production in the range from 30 to 60 min. This period is coherent with
the maximum ethanol consumption observed in Fig. 2 and may be
explained by initial ethanol reforming over the fresh catalyst surface,
generating metal hydrides and hydrogen. After 60 min, the main prod-
uct temporal behavior changed according to the product and catalyst.
Taking into account 1-butanol, the target product for ethanol GR,
only the Fe20MMO catalyst diminished the production, while the cata-
lysts MMO, Ni20MMO, and Zn20MMO kept constant with selectivities of
3
8.5, 46.2 and 30.8 %, respectively. The catalyst Co20MMO, after 270
min of reaction, started to increase the 1-butanol production reaching
6.5 %. For Cu20MMO 1-butanol production continuously increased
There was no correlation between the ethanol conversion and
product selectivities with physicochemical catalyst properties such as
acidity, basicity, acidity/basicity ratio.
3
reaching a maximum of 56 % of selectivity at 360 min. Although the
observation of an increase in 1-butanol production could be interpreted
as a positive characteristic, it is important to recall that both Co20MMO
and Cu20MMO showed a decrease in ethanol consumption, as seen in
Fig. 2.
3.4. Catalyst characterization after reaction
After the catalytic assays, the materials were submitted to charac-
terization by means of EPR, XRD, TGA, SEM and EDS, aiming to better
understand the behavior of MMOs during the course of the reaction.
The EPR analysis for the catalysts containing Fe, Co, Ni, Cu, and Zn
presented a paramagnetic signal with g value around 2.0106 and line-
width ca. 6.0 G. For the reference material (MMO), the EPR spectrum
A second important product for GR is acetaldehyde which is a re-
action intermediate for 1-butanol. For MMO and Zn20MMO, a plateau
was reached with maximum selectivities of 15.7 and 23.5 % respec-
tively. For Ni20MMO, the maximum selectivity was reached after 30 min
and then continuously decreased until 330 min with a small increase at
2
+
was silent. For the Cu containing catalyst, the signal of Cu was seen
3
60 min. Co20MMO presented oscillatory behavior, although with a
with g = 2.1532 and for the other metal ions, no signal was observed
◦
tendency to decrease. For the Cu20MMO catalyst, the maximum of 25 %
selectivity was reached after 120 min, then decreasing until 330 min,
stabilizing at 8.1 % selectivity.
since the spectra were obtained at 25 C. The g value and linewidth
found were not coherent with paramagnetic transition metal ions [54].
Fig. 5 shows the EPR spectra for all catalysts after the reaction.
4

G. Metzker et al.
Applied Catalysis A, General 623 (2021) 118272
Fig. 4. Temporal behavior of ethanol catalytic conversion to 1-butanol, acetaldehyde, diethyl ether, and ethyl acetate by MMOs.
◦
Thus, the EPR signals mentioned above can only be correlated with
show the presence of OV. For the other catalysts, a mass gain ca. 690 C,
oxygen vacancies (OV) in the catalyst structure during the ethanol re-
action [54–58]. The OV can enhance the reactivity of the catalyst, since
it is already reported that the catalyst basicity increases with the
removal of oxygen from the surface [54,55]. Also, OV are related with
oxidation reactions and strong interaction with carbonyl groups [54].
Using TGA analysis, the OV presented in the catalysts were quanti-
fied [59–61] (Fig. S4 and Table S9). As observed by EPR, MMO did not
ranging from 0.20 to 0.75 % was observed. These values are in agree-
ment with those for other metal oxides reported in the literature
[60–62].
The OV may have an important role in ethanol dehydrogenation into
acetaldehyde and may enhance the acetaldehyde reactivity in the aldol
condensation, by its stabilization through the carbonyl group on the
catalyst surface. On this matter, Li and co-workers established a linear
5

G. Metzker et al.
Applied Catalysis A, General 623 (2021) 118272
reactions occurred in any extensive way. The MMO acidity and basicity
values followed similar behavior; there was no clear correlation between
the crystallographic parameters and the product distributions found.
To complement the XRD analysis of MMO structures, TGA, SEM, and
EDS analyses were also carried out. The TGA analysis of the catalysts
◦
showed thermal events in the temperature range from 300 to 700 C. It is
well-reported in the literature that carbon species are oxidized to CO and
CO
2
in this temperature range and it is possible to infer the type of
◦
carbon present in the samples (amorphous carbon ca. 250–400 C,
◦
nanostructured materials like nanofibers and nanotubes ca. 400–700 C)
[
64]. Thus, TGA and differential thermal analysis (DTA) of the MMOs
after the reaction showed the deposition of two distinct types of carbon
over the catalysts: amorphous and nanostructured, as seen in the ther-
mograms of Figs. S6–S11. The percentage of carbon in each sample was
determined, ranging from 9.5 to 36.4 % of the initial mass. Deconvo-
luting the DTA spectra and integrating the curves (Figs. S12–S17)
determined the percentage of each type of carbon deposited on the
MMOs. The amorphous carbon corresponds to the lower percentage
found, ranging from 0.7 to 19.2 %, the nanostructured carbon corre-
sponding to 80.8–99.3 % of the deposited carbon. Table S11 summarizes
the data.
SEM images showed the formation of clusters after the reaction,
indicating that the catalytic process induces particle aggregation
(
Figs. S18–S23). Also, SEM images indicated filamentous carbon (C
F
)
formation in the catalysts Fe20MMO, Co20MMO, Ni20MMO, and
Cu20MMO. Surprisingly, MMO and Zn20MMO images did not display the
F
formation of C , corroborating with the lower percentage of carbon
deposition found in TGA. The EDS analysis indicated carbon in all
samples, although with much lower content for MMO and Zn20MMO
catalysts (Table S12). The percentages of carbon found in EDS were very
distinct from the ones found by TGA. This observation is coherent since
EDS quantifies only spots of the sample, while TGA furnishes a complete
panorama of the sample. Fig. 6 depicts the SEM for Ni20MMO before the
reaction (A) and after 6 h of reaction (B) showing the C
The deposition of carbon species over catalysts is associated with
and CO presence in the reaction medium and is usually found in
F
Fig. 5. EPR spectra of the catalysts after reaction.
CH
4
relationship between the concentration of OV (ranging from 30 to 46 %)
and yield of condensation products (butyraldehyde, crotonaldehyde
catalytic approaches to the ethanol steam reforming reaction [37,39,41,
65]. Although amorphous carbon deposition is associated with catalyst
among others) employing CeO
2
with distinct morphologies and redox
F
deactivation due to the blocking of active sites, C is not associated
states [63]. However, it was not possible in our study to directly
correlate the OV percentages in the MMOs and the condensation product
yields.
directly with catalyst poisoning since the catalyst reaction sites are not
blocked. However, over long operational periods, the continuous for-
F
mation of C can inactivate the catalyst because a sintering process at the
The presence of OV are also related to ethanol dehydration into
ethylene [54]. Thus, OV may contribute to ambivalent behavior in the
catalytic process. It may enhance the ethanol reactivity for the Guerbet
reaction pathway and introduce a competitive reaction pathway
catalyst active sites can be induced [37]. Also, there was no correlation
of the MMO activity and the percentage of deposited carbon (amorphous
or nanostructured).
Thus, the TGA and SEM data indicated that, as expected, in our
catalytic approach, a fraction of ethanol in the reactor undergoes a
(
ethanol dehydration), diminishing the GR products.
It is also important to recall that only the MMOs with transition
4
reforming reaction and CO and CH are transformed into carbon species
metals in their structure presented OV. This behavior may be due to the
since it is not possible to direct the ethanol reaction along the GR
pathway only.
enhancement of acidity due to the substitution of Mg2 by the first-row
transition metal ions, facilitating the removal of oxygen from the MMO
surface. This hypothesis proved to be coherent since Fe20MMO and
Cu20MMO, which had the highest and lowest acidity values (Table S2),
also presented the highest and lowest OV concentrations, respectively.
The MMOs were also characterized using XRD, aiming to follow any
structural changes. After 6 h of reaction, all catalysts still presented the
characteristic 3 peak pattern of the fresh catalysts (Fig. S5). However,
calculating the crystallographic parameters (a: cation-cation distance; D:
crystallite size) for all the catalysts, changes could be noticed. Table S10
resumes the parameters calculated for the catalysts.
+
As shown in Fig. 1, acetaldehyde is the key-intermediate for both GR
4
and the ethanol reforming reaction yielding CH and CO. Thus, the
carbon species formation is directly related to the catalyst ability to
promote acetaldehyde decomposition (1), methane dehydrogenation (2)
and the CO Boudouard reaction (3) [65]. Catalysts containing first-row
transition metals, except Zn, were reported to be active in the ethanol
steam reforming reaction and thus are expected to promote the carbon
deposition by the pathways described above.
3.5. Catalyst efficiency for ethanol GR
The crystallite size (D) increased for the Co20MMO, Ni20MMO and
Cu20MMO (25, 12 and 16 % respectively) and this corroborates with the
increase in the cation-cation distance (a). The increase of these two
parameters indicates that during the reaction, the catalyst structure is
modified in some way, promoting a relaxation in its structure. However,
the absence of new peaks suggests that no phase segregation or redox
After the quantification of the GR products and catalyst character-
ization, mostly by the percentage of carbon deposition, these parameters
can be correlated with the best catalyst choice for ethanol conversion
into 1-butanol, which is the main goal of GR. However, not only the 1-
butanol selectivity should be taken into account, but also the main side
6

G. Metzker et al.
Applied Catalysis A, General 623 (2021) 118272
F
Fig. 6. SEM images for Ni20MMO before the reaction (A) and after the reaction (B). C can be observed in (B).
product formation (sum of diethyl ether and ethyl acetate) and the
percentage of carbon deposition over the catalyst surface. Large
amounts of side products diminish the 1-butanol selectivity and the
separation process becomes more complicated, thinking forward to up-
scaling. Also, the percentage of carbon deposition (amorphous or fila-
mentous) will diminish the catalyst activity over time, demanding car-
bon removal from its surface or replacing the catalyst in the reactor.
Aiming to rank the MMOs utilized as GR catalysts in the present
study, four variables, ethanol conversion, 1-butanol selectivity, main
side-product selectivities and the percentage of carbon deposition were
compared, as shown in Fig. 7. Notice that Fe20MMO, Co20MMO and
Cu20MMO were excluded from the comparison, since these catalysts
presented a drop in ethanol consumption over 6 h of reaction.
Thus, taking into account the data of Fig. 7, Zn20MMO catalyst seems
to be the best choice for GR. Although Ni20MMO presents a higher 1-
butanol selectivity, Zn20MMO presented lower side product produc-
tion and a considerably lower carbon deposition over the catalyst in 6 h
of reaction.
Fig. 8. Correlation between percentage of deposited carbon on the catalyst and
metal-carbon bond energy (M C).
–
carbon atoms to the transition metal present in the catalyst composition.
Thus, the tailoring of catalysts aiming to better perform GR should
consider this relationship. For example, the Ni-containing catalyst pre-
sented the best conversion of ethanol into 1-butanol but presented a high
3
.6. Catalyst tailoring for ethanol GR
F
amount of C formation. On the other hand, the Zn-containing catalyst
produced 30 % less 1-butanol than the Ni-containing catalyst. Thus, for
future catalyst development, it seems that a bimetallic catalyst with Ni-
Zn could be the right candidate that associates both high ethanol con-
version to 1-butanol and, also, a lower carbon deposition over the
catalyst surface.
The literature has also reported the role of transition metals in the
catalytic growing of carbon nanotubes, related to the energy of the
metal-carbon bond at the catalyst surface [66]. Thus, it was possible to
establish a correlation between the percentage of carbon found in TGA
analysis of the MMOs and the metal-carbon bond energy calculated for
growing carbon nanotubes. Fig. 8 shows the correlation.
It has already been reported in the literature that Zn addition to Cu
[67–69], Pt-Sn [70], Pt [71], and Pd [72] catalysts enhance their sta-
bility and conversion rate, as well as changing the product distribution
for the water-gas shift [67], hydrogenation [68], methanol steam
reforming [69], and dehydrogenation [70–72] reactions. However, until
now, there was no clear correlation between the Zn-promoted catalytic
enhancing capacity and carbon deposition over the surface, although it
As shown in Fig. 8, there is a clear linear relationship between carbon
deposition in the catalyst and the first-row transition metals employed as
dopant.
The data obtained from TGA, SEM and EDS experiments indicate that
the carbon deposition is an intrinsic process related to the affinity of
Fig. 7. Parameters utilized to determine the best MMO for ethanol GR. Ethanol consumption, 1-butanol and side product selectivities refer to the points in which the
plateau was reached in Fig. 4.
7

G. Metzker et al.
Applied Catalysis A, General 623 (2021) 118272
has already been shown that Zn-containing catalysts diminished the
carbon deposition [71]. Thus, the correlation presented in Fig. 8 helps to
explain the behavior of Zn-containing catalysts.
[11] W.R. da S. Trindade, R.G. dos Santos, Renew. Sustain. Energy Rev. 69 (2017)
6
42–651.
[
[
[
12] P. Benito, A. Vaccari, C. Antonetti, D. Licursi, N. Schiarioli, E. Rodriguez-Castell o´ n,
A.M. Raspolli Galletti, J. Clean. Prod. 209 (2019) 1614–1623.
13] J.A. Barrett, Z.R. Jones, C. Stickelmaier, N. Schopp, P.C. Ford, ACS Sustain. Chem.
Eng. 6 (2018) 15119–15126.
4
. Conclusions
A series of MMOs containing Fe2 , Co , Ni , Cu , and Zn were
14] H. Li, A. Riisager, S. Saravanamurugan, A. Pandey, R.S. Sangwan, S. Yang,
R. Luque, ACS Catal. 8 (2018) 148–187.
+
2+
2+
2+
2+
[15] Z. Sun, A.C. Vasconcelos, G. Bottari, M.C.A. Stuart, G. Bonura, C. Cannilla,
F. Frusteri, K. Barta, ACS Sustain. Chem. Eng. 5 (2017) 1738–1746.
tested as catalysts for GR in a fixed bed flow reactor. Considering the
ethanol conversion, GR products, side products, and carbon deposition
over the catalyst surface in 6 h of reaction, the catalyst containing Zn
presented the best results. Although the Ni-containing catalyst produced
[
[
16] M. Guerbet, CR Acad. Sci. Paris 128 (1899) 1002–1004.
17] D. Gabriels, W.Y. Hernandez, B.F. Sels, P. Van Der Voort, A. Verberckmoes, Catal.
¨
´
Sci. Technol. 5 (2015) 3876–3902.
[
18] O.M. Perrone, F. Lobefaro, M. Aresta, F. Nocito, M. Boscolo, A. Dibenedetto, Fuel
Process. Technol. 177 (2018) 353–357.
F
a higher quantity of 1-butanol, the high deposition of C over the cata-
[
19] S. Cimino, L. Lisi, S. Romanucci, Catal. Today 304 (2018) 58–63.
lyst surface may be a drawback in future applications. A correlation
between the percentage of deposited carbon and the metal-carbon bond
energy was established. This relationship, along with the catalysis re-
sults, can be useful in tailoring new catalysts for GR. Bimetallic catalysts
containing Zn would be a starting point since Zn presented a low-carbon
deposition, and its combination with other transition metals may
enhance its reactivity and selectivity for GR products.
[20] G.R.M. Dowson, M.F. Haddow, J. Lee, R.L. Wingad, D.F. Wass, Angew. Chemie -
Int. Ed. 52 (2013) 9005–9008.
[
[
21] N.V. Kulkarni, W.W. Brennessel, W.D. Jones, ACS Catal. 8 (2018) 997–1002.
22] P.L. Burk, R.L. Pruett, K.S. Campo, J. Mol. Catal. 33 (1985) 1–14.
[23] F. Hayashi, M. Iwamoto, ACS Catal. 3 (2013) 14–17.
[
[
[
[
24] M.D. Jones, C.G. Keir, C. Di Iulio, R.A.M. Robertson, C.V. Williams, D.C. Apperley,
Catal. Sci. Technol. 1 (2011) 267–272.
25] E.V. Makshina, W. Janssens, B.F. Sels, P.A. Jacobs, Catal. Today 198 (2012)
338–344.
26] D.L. Carvalho, R.R. De Avillez, M.T. Rodrigues, L.E.P. Borges, L.G. Appel, Appl.
Catal. A Gen. 415–416 (2012) 96–100.
CRediT authorship contribution statement
27] S. Chakraborty, P.E. Piszel, C.E. Hayes, R.T. Baker, W.D. Jones, J. Am. Chem. Soc.
1
37 (2015) 14264–14267.
[
[
28] J. Quesada, L. Faba, E. Díaz, S. Ord o´ n˜ ez, Appl. Catal. A Gen. 559 (2018) 167–174.
29] M. Rechi Siqueira, O. Micali Perrone, G. Metzker, D.C. de Oliveira Lisboa, J.
C. Thom ´e o, M. Boscolo, Mol. Catal. 476 (2019), 110516.
Gustavo Metzker: Conceptualization, Methodology, Investigation,
Writing - original draft, Writing - review & editing. Jorge Andres Mora
Vargas: Investigation. Livia Padilha de Lima: Investigation. Olavo
Micali Perrone: Investigation. Marcos Rechi Siqueira: Investigation.
Laudemir Carlos Varanda: Conceptualization, Writing - original draft.
Mauricio Boscolo: Conceptualization, Writing - original draft, Funding
acquisition, Supervision.
[30] H. Aitchison, R.L. Wingad, D.F. Wass, ACS Catal. 6 (2016) 7125–7132.
[
31] D. Ayillath Kutteri, I.W. Wang, A. Samanta, L. Li, J. Hu, Catal. Sci. Technol. 8
(
2018) 858–869.
[
32] M. Bowker, R. Holroyd, N. Perkins, J. Bhantoo, J. Counsell, A. Carley, C. Morgan,
Surf. Sci. 601 (2007) 3651–3660.
[33] C.H. Bartholomew, Appl. Catal. A Gen. 212 (2001) 17–60.
[
34] C. Montero, A. Remiro, B. Valle, L. Oar-Arteta, J. Bilbao, A.G. Gayubo, Ind. Eng.
Chem. Res. 58 (2019) 14736–14751.
[
[
[
35] M.A. Vannice, Catal. Sci. Technol. 3 (1982) 139–198.
36] R. Brukh, S. Mitra, Chem. Phys. Lett. 424 (2006) 126–132.
37] J. Rostrup-Nielsen, D.L. Trimm, J. Catal. 48 (1977) 155–165.
Declaration of Competing Interest
The authors report no declarations of interest.
Acknowledgments
[38] A. Ochoa, J. Bilbao, A.G. Gayubo, P. Casta n˜ o, Renew. Sustain. Energy Rev. 119
(
2020).
[
[
[
39] I. Alstrup, J. Catal. 109 (1988) 241–251.
40] V. Palma, C. Ruocco, M. Cortese, M. Martino, Catalysts 10 (2020).
41] M.A. Ermakova, D.Y. Ermakov, G.G. Kuvshinov, L.M. Plyasova, J. Catal. 187
(
1999) 77–84.
The authors acknowledge the Sao Paulo Research Foundation
[
42] A.G. Sato, D.P. Volanti, D.M. Meira, S. Damyanova, E. Longo, J.M.C. Bueno,
(
FAPESP) for their financial support (grant no. 2020/02471-4) and the
fellowship to J.A.M.V (grant no. 2017/09520-8) and O.M.P (grant no.
019/11058-6). The authors would also like to thank the National
J. Catal. 307 (2013) 1–17.
[43] J. Mora Vargas, L.H. Tofaneli Morelato, J. Orduna Ortega, M. Boscolo, G. Metzker,
Green Chem. 22 (2020) 2365–2369.
2
[
[
44] I.C. Marcu, D. Tichit, F. Fajula, N. Tanchoux, Catal. Today 147 (2009) 231–238.
45] H. Idriss, E.G. Seebauer, J. Mol, Catal. A Chem. 152 (2000) 201–212.
Council for the Improvement of Higher Education (CAPES) for the PNPD
post-doctoral fellowship to G.M (grant no. 1760327). The authors are
also indebted to Professors Benedito dos Santos Lima-Neto and Daniel
Rodrigues Cardoso (IQSC-USP) for the use of EPR spectrometer equip-
ment (FAPESP grant no. 92/03515-3).
[46] W.Y. Hern ´a ndez, K. De Vlieger, P. Van Der Voort, A. Verberckmoes, ChemSusChem
9 (2016) 3196–3205.
[
[
[
[
47] I.C. Marcu, N. Tanchoux, F. Fajula, D. Tichit, Catal. Letters 143 (2013) 23–30.
48] F. Cavani, F. Trifir o` , A. Vaccari, Catal. Today 11 (1991) 173–301.
49] W.T. Reichle, Solid State Ion. 22 (1986) 135–141.
50] K.K. Ramasamy, M. Gray, H. Job, D. Santosa, X.S. Li, A. Devaraj, A. Karkamkar,
Y. Wang, Top. Catal. 59 (2016) 46–54.
Appendix A. Supplementary data
[
51] J. de Souza Rossi, O.M. Perrone, M.R. Siqueira, D.P. Volanti, E. Gomes, R. Da-Silva,
M. Boscolo, Renew. Energy 133 (2019) 367–372.
Supplementary material related to this article can be found, in the
online version, at doi:https://doi.org/10.1016/j.apcata.2021.118272.
[52] O.M. Perrone, M.R. Siqueira, G. Metzker, D.C. de Oliveira Lisboa, M. Boscolo,
Environ. Prog. Sustain. Energy 40 (2021).
[
[
[
53] M. Chui, G. Metzker, C.M. Bernt, A.T. Tran, A.C.B. Burtoloso, P.C. Ford, ACS
Sustain. Chem. Eng. 5 (2017) 3158–3169.
54] L.H. Chagas, P.C. Zonetti, C.R.V. Matheus, C.R.K. Rabello, O.C. Alves, L.G. Appel,
ChemCatChem 11 (2019) 4011–4020.
References
55] Y. Xi, Y. Wang, D. Yao, A. Li, J. Zhang, Y. Zhao, J. Lv, X. Ma, ChemCatChem 11
[
1] N.M. Eagan, M.D. Kumbhalkar, J.S. Buchanan, J.A. Dumesic, G.W. Huber, Nat.
Rev. Chem. 3 (2019) 223–249.
(
2019) 2607–2614.
[
[
56] W. Mokwa, D. Kohl, G. Heiland, Surf. Sci. 117 (1982) 659–667.
57] A.A.H. Elbadawi, L. Ge, J. Zhang, L. Zhuang, S. Liu, X. Tan, S. Wang, Z. Zhu, Chem.
Eng. J. 392 (2020) 123739.
[
[
2] J. Sun, Y. Wang, ACS Catal. 4 (2014) 1078–1090.
3] C. Angelici, B.M. Weckhuysen, P.C.A. Bruijnincx, ChemSusChem 6 (2013)
1
595–1614.
[
[
58] T.J. Huang, H.J. Lin, T.C. Yu, Catal. Letters 102 (2005) 175–181.
59] A.M. D’Angelo, A.C.Y. Liu, A.L. Chaffee, J. Phys. Chem. C 120 (2016)
[
[
[
4] G.W. Huber, S. Iborra, A. Corma, Chem. Rev. 106 (2006) 4044–4098.
5] P. Kostetskyy, G. Mpourmpakis, Catal. Sci. Technol. 5 (2015) 4547–4555.
6] E.G. Derouane, J.B. Nagy, P. Dejaifve, J.H.C. van Hooff, B.P. Spekman, J.
C. V ´e drine, C. Naccache, J. Catal. 53 (1978) 40–55.
1
4382–14389.
[
[
60] Z. Wang, W. Wang, L. Zhang, D. Jiang, Catal. Sci. Technol. 6 (2016) 3845–3853.
61] W. Ren, T. Ding, Y. Yang, L. Xing, Q. Cheng, D. Zhao, Z. Zhang, Q. Li, J. Zhang,
L. Zheng, Z. Jiang, X. Li, ACS Catal. 9 (2019) 8772–8784.
[
7] A.D. Patel, S. Telalovi ´c , J.H. Bitter, E. Worrell, M.K. Patel, Catal. Today 239 (2015)
5
6–79.
[
[
62] M.H. Casta n˜ o, R. Molina, S. Moreno, Catalysts 5 (2015) 905–925.
63] Z. Liang, D. Jiang, G. Fang, W. Leng, P. Tu, Y. Tong, L. Liu, J. Ni, X. Li,
ChemistrySelect 4 (2019) 4364–4370.
[
[
8] E. Cs ´e falvay, G.R. Akien, L. Qi, I.T. Horv ´a th, Catal. Today 239 (2015) 50–55.
9] J.M.R. Gallo, J.M.C. Bueno, U. Schuchardt, J. Braz. Chem. Soc. 25 (2014)
2
229–2243.
[
64] A.G. Bannov, M.V. Popov, P.B. Kurmashov, J. Therm. Anal. Calorim. 142 (2020)
[
10] J.L. Contreras, J. Salmones, J.A. Colín-Luna, L. Nu n˜ o, B. Quintana, I. C o´ rdova,
B. Zeifert, C. Tapia, G.A. Fuentes, Int. J. Hydrogen Energy 39 (2014) 18835–18853.
3
49–370.
8

G. Metzker et al.
Applied Catalysis A, General 623 (2021) 118272
[
[
65] L.V. Mattos, G. Jacobs, B.H. Davis, F.B. Noronha, Chem. Rev. 112 (2012)
094–4123.
66] F. Silvearv, P. Larsson, S.L.T. Jones, R. Ahuja, J.A. Larsson, J. Mater. Chem. C 3
[69] F. Tonelli, O. Gorriz, L. Arrúa, M.C. Abello, Quim. Nova 34 (2011) 1334–1338.
[70] Y. Zhang, Y. Zhou, J. Shi, X. Sheng, Y. Duan, S. Zhou, Z. Zhang, Fuel Process.
Technol. 96 (2012) 220–227.
4
(
2015) 3422–3427.
[71] C. Yu, H. Xu, Q. Ge, W. Li, J. Mol. Catal. A Chem. 266 (2007) 80–87.
[72] D.V. Glyzdova, T.N. Afonasenko, E.V. Khramov, N.N. Leont’eva, M.V. Trenikhin, I.
P. Prosvirin, A.V. Bukhtiyarov, D.A. Shlyapin, Top. Catal. 63 (2020) 139–151.
[
[
67] C.Y. Shiau, S. Chen, J.C. Tsai, S.I. Lin, Appl. Catal. A Gen. 198 (2000) 95–102.
68] H.S. Na, S.Y. Ahn, J.O. Shim, K.W. Jeon, H.M. Kim, Y.L. Lee, W.J. Jang, B.H. Jeon,
H.S. Roh, Korean J. Chem. Eng. 36 (2019) 1243–1248.
9