Copper promoter effect on acid-base and redox sites of Fe/Al2O3 catalysts and their role in ethanol-acetone mixture conversion

Catalysis

Science &

Technology

View Article Online

View Journal

PAPER

Copper promoter effect on acid–base and redox

sites of Fe/Al₂O₃catalysts and their role in

ethanol–acetone mixture conversion

ab

Cite this: DOI: 10.1039/c7cy01640a

Antonio Narcisio Pinheiro, ^aRegina Claudia Rodrigues dos Santos,

*

a

Sarah Brenda Ferreira dos Santos, ^aMoacir José da Silva Júnior,

c

a

Valder Nogueira Freire ^band Antoninho Valentini

*

Tiago Pinheiro Braga,

Active species of copper and iron oxide (Cu–Fe) catalysts supported on alumina were prepared by combin-

ing Pechini and wet impregnation methods. The effect of combined acid–base and redox sites of Cu and

Fe species on gas-phase ethanol–acetone mixture conversion was investigated. The catalysts were charac-

terized by chemical analyses, XRD, H₂-TPR, Mössbauer spectroscopy, N₂physisorption, CO₂-TPD, SEM-

EDS, TG/DTA and pyridine adsorption isotherms. N₂adsorption/desorption isotherms and SEM-EDS analy-

sis showed that the addition of copper caused an increase of BET surface area and Cu and Fe oxide disper-

sion. H₂-TPR characterization showed that interactions between Cu and Fe oxides shift the reducibility of

Fe³⁺species to lower temperature improving the redox properties of the catalyst. The partial reduction of

the Cu and Fe oxide species was found to be efficient in inhibiting the side decomposition reactions, im-

proving the catalytic efficiency towards dehydrogenation and hydrogen transfer processes. It was found

that acid–base pairs play an important role in the formation of dehydrogenation, dehydration and conden-

sation products from ethanol, while redox sites are decisive for hydrogen transfer reactions with reduction

of acetone to isopropanol. H₂-TPR and Mössbauer spectroscopy results for the spent catalysts revealed

that the highest catalytic performance of the Cu–FeAl catalysts may be attributed to the good dispersion of

the Cu oxide and the site generated by the partial reduction which produces Cu⁺/Cu⁰and Fe²⁺active spe-

cies. A reaction pathway with the participation of the acid–base and redox sites in the formation of prod-

ucts by consecutive dehydrogenation–condensation or dehydrogenation–hydrogenation reactions has

been proposed.

Received 11th August 2017,

Accepted 27th November 2017

DOI: 10.1039/c7cy01640a

rsc.li/catalysis

may become available as a sustainable and renewable plat-

form molecule for the production of value-added chemicals

1. Introduction

Nowadays, major research effort has been given to the devel-

opment of green processes and use of alternative feedstock,

mainly from renewable biomass resources, to produce fuels

and fine chemical products.¹Biofuels and chemicals pro-

duced from biomass could help to reduce both the world's de-

pendence on petroleum and CO₂emission.^2,3Currently, etha-

nol is an important energy source in Brazil, which is directly

used as fuel or blended with fossil fuels, where it is produced

from sugar cane.⁴However, bio-ethanol, which is ethanol pro-

duced from renewable biomass resources, represents almost

90% of the global biofuel production.³Therefore, ethanol

such as 1-butanol, ethylene, ethyl acetate, acetaldehyde, ethyl

ether, acetone, butadiene and hydrogen.^3,5,6

The conversion of ethanol to useful chemicals can com-

prise both parallel and consecutive reaction pathways, includ-

ing dehydrogenation, condensation, dehydration and decom-

position.³Dehydrogenation and dehydration are typically

endothermic reactions, while condensation presents exother-

mic thermodynamic character; thus, the reaction pathway

should depend on the reaction temperature.^3,7In addition to

reaction temperature, the acid–base properties also have a

meaningful effect on the catalytic performance, such as prod-

uct selectivity and catalyst stability. The ethanol dehydration

process has been extensively studied over a variety of acidic

or amphoteric catalysts;^8,9γ-Al₂O₃and zeolites (e.g., ZSM-5)

have received the most attention due to their high activity

and selectivity.^10,11However, transition metal oxides, mainly

copper-based catalysts, also exhibited a high selectivity to de-

hydrogenation products.^7,12Another interesting catalytic

^aDepartment of Analytical Chemistry and Physical Chemistry, Federal University

of Ceará, Campus of Pici, Fortaleza - CE, 60440-554, Brazil.

E-mail: reginaclaudiasantos@yahoo.com.br, valent@ufc.br

^bDepartment of Physics, Federal University of Ceará, Campus of Pici, Fortaleza -

CE, 60440-554, Brazil

^cInstitute of Chemistry, Federal University of Rio Grande do Norte, Campus

Universitário, Natal - RN, 59078-970, Brazil

This journal is © The Royal Society of Chemistry 2017

Catal. Sci. Technol.

View Article Online

Paper

Catalysis Science & Technology

process, wherewith ethanol may be used as a platform mole-

cule, is the condensation reaction, which forms a carbon–car-

bon bond between two aldehyde or ketone molecules. This

process is generally carried out at mild reaction tempera-

tures, on acidic or basic catalysts.^3,13–15The reaction of etha-

nol condensation can occur through the initial formation of

aldehydes, which sequentially are subjected to an aldol con-

densation reaction. Therefore, it is reasonable to propose that

Therefore, it is interesting to present the discussion of the

effect of combined structural, textural, acid–base and redox

properties on the catalytic performance of the Fe/Al₂O₃cata-

lyst, with and without a copper promoter. The effects of the

reaction temperature, ethanol-to-acetone (Et/Ac) molar ratio

and copper content of the catalyst on the catalytic perfor-

mance were studied. The explanation for the catalytic perfor-

mance is based on characterization of several physical and

chemical properties of the fresh and spent catalysts. There-

fore, in the light of the results obtained, a reaction pathway

for the formation of products from an ethanol–acetone mix-

ture with participation of both acid–base and redox sites is

proposed.

sequential reaction pathways require

a catalyst system

possessing bifunctionality of active sites to achieve greater re-

action rates and product selectivity.^15,16It is especially rele-

vant if a ketone is added to the reaction mixture (ethanol–ke-

tone), where the condensation reaction can occur between

the alcohol and the ketone.¹⁷Therefore, factors such as reac-

tion temperature, the reactant molar ratio, the structure of re- 2. Experimental

actant molecules, and the nature of the catalyst can deter-

mine the selectivity of the process toward important

2.1. Catalyst synthesis

oxygenate compounds.^3,15A recent publication by Onyestyák

et al. with an acetone–ethanol mixture has stated that under

a hydrogen atmosphere, acetone could be reduced to isopro-

pyl alcohol over Pd catalysts.¹⁸However, it is likely that this

process should involve a gas-phase hydrogen transfer step

from the dehydrogenation reaction of ethanol to aldehydes,

which is similar to the Meerwein–Ponndorf–Verley

reduction.^18,19

The reference monometallic catalyst with 10 wt% iron oxide

supported on Al₂O₃(10FeAl) was prepared according to the

polymeric precursor method previously described.²⁸This syn-

thesis route allows preparation of a solid sample with a high

dispersion of the active metal oxide.²⁹Aluminium nitrate

nonahydrate {AlĲNO₃)₃·9H₂O}, iron nitrate nonahydrate

{FeĲNO₃)₃·9H₂O}, citric acid monohydrate (CA) {C₆H₈O₇·H₂O}

and ethylene glycol (EG) {C₂H₆O₂} were used as starting

chemicals. An EG/CA/metal molar ratio of 2 : 3 : 1 was used in

the synthesis of the samples. In brief, an aqueous solution of

metal precursors and CA were mixed and stirred for 1 h at

room temperature for the formation of a metal citrate com-

plex. Subsequently, EG was added, and the mixture was kept

under vigorous stirring at 90 °C until it became a viscous

resin. The resin was heat treated at 250 °C for 1 h in an air

atmosphere. The resulting precursor composite was ground

and heat treated at 500 °C under air flow for 1 h.

The bimetallic CuFeAl catalysts with two different amounts

of copper oxide (0.5 and 1 wt%) were prepared using a wet

impregnation method from the 10FeAl sample, as previously

reported.^12,25The 10FeAl solid was placed in contact with an

aqueous solution containing appropriate amounts of copper

nitrate. This system remained under magnetic stirring at

70 °C until the total evaporation of the solvent. The calcina-

tion process for the dried solids was performed at 500 °C in a

muffle furnace under air flow for 1 h. The bimetallic catalyst

composed of iron and copper oxides supported on Al₂O₃was

denominated as 0.5CuFeAl or 1CuFeAl.

Catalysts containing iron oxide are active in various reac-

tions. Fe-based oxides supported on alumina were found to

be promising in the dehydrogenation of ethylbenzene with

carbon dioxide.^12,20Catalysts such as Ni/Fe₂O₃, Cu/Fe₂O₃and

Pd/Fe₂O₃have been investigated for production of bioderived

compounds²¹in a catalytic hydrogen transfer system with al-

cohols as hydrogen donors. Additionally, bimetallic M_xO_y–

FeO_xcatalysts can catalyse the conversion of biomass re-

sources through the dehydration process and consecutive hy-

drogen transfer reaction.²²However, mixed oxides with inad-

equately balanced surface acid–base properties and redox

sites would catalyse the secondary reactions (cracking and

coking) leading to rapid deactivation of the catalyst by losing

active sites as well as affecting the product distribution.^23,24

Therefore, considering the bi-functional mechanism of the ti-

tle reaction, there is a need to develop catalysts, which pres-

ent a uniform dispersion of the active acid–base and redox

sites. Interestingly, the addition of Cu oxide into Fe oxide-

based catalysts can induce structural changes on the catalyst

and improve the iron dispersion.^25,26Furthermore, the inter-

actions between the Cu and Fe species have an effect on the

catalyst reducibility, enhancing the reactivity of the surface

sites.^26,27

2.2. Catalyst characterization

The relative amount of iron and copper (wt%) was deter-

mined by using atomic absorption spectroscopy (GBC933 AA

model). Previously, the solids were digested in concentrated

nitric acid under heating at 200 °C. The obtained solutions

were diluted to 50 mL to perform the measurements. The

crystalline structure of the solids was characterized by X-ray

diffraction (XRD) using an X'Pert PANalytical diffractometer

operated at 40 kV and 40 mA employing a Cu-Kα radiation

source (λ = 1.5418 Å). The experiments were performed at

Despite the wide application range of Cu- or Fe-based

catalysts, no studies of Cu–Fe catalysts dispersed on Al₂O₃

for the conversion of an ethanol–acetone mixture have yet

been carried out. Additionally, to the best of our knowledge,

the role of both acid–base and redox sites on bimetallic

Cu–Fe catalysts in ethanol and/or acetone conversion to

chemicals has not yet been fully demonstrated based on ex-

perimental results.

Catal. Sci. Technol.

This journal is © The Royal Society of Chemistry 2017

View Article Online

Catalysis Science & Technology

Paper

high diffraction angles (2θ from 10 to 90°). The crystalline

phases were identified using the software X'Pert HighScore

Plus® (PANalytical), and the crystallographic data for all

phases were determined using the Inorganic Crystal Structure

Database (ICSD).

spectrometer working in constant acceleration mode. The

measurements were carried out using standard transmission

geometry with ⁵⁷Co/Rh as the radioactive source. The spectra

were fitted to discrete Lorentzian functions evaluated by

means of a non-linear routine that uses the least squares

method. All the values of isomeric displacement are relative

to those obtained for α-Fe. Thermogravimetric analysis (TGA)

and differential thermal analysis (DTA) were carried out on

spent catalysts to examine the amounts of carbon deposits

using Shimadzu DTA-60H equipment. The TG measurements

were made under air flow (40 ml min⁻¹) from room tempera-

ture to 1000 °C, with a heating rate of 10 °C min⁻¹.

The textural properties of the catalysts were determined by

adsorption/desorption at −196 °C on a Quantachrome instru-

ment, Autosorb-1B model. The samples were degassed at 250

°C prior to the measurement. The specific surface area of the

samples was determined using the BET method. The pore

size distribution was derived from the adsorption branches

of the isotherms using the Barrett–Joyner–Halenda (BJH)

method.³⁰The t-plot method (Harkins and Jura equation)

was applied to determine the volume and area of the micro-

pores.³¹Scanning electron micrographs (SEM) were recorded

using a TESCAN VEGA XMU electron microscope equipped

with a Bruker QUANTAX EDS system coupled to the SEM

microscope. Previously, the samples were deposited on an al-

uminium sample holder and sputtered with gold, and then

SEM measurements were conducted, using an accelerating

voltage of 20 kV.

The fresh and spent catalysts were characterized using the

temperature-programmed reduction technique (TPR) in the

range of 50–950 °C in a quartz reactor using an 8% H₂/N₂mix-

ture flow (30 mL min⁻¹) at a heating rate of 10 °C min⁻¹. Be-

fore the TPR tests, the samples (50 mg) were subjected to heat

treatment at 300 °C under N₂flow for 0.5 h. A thermal conduc-

tivity detector (TCD) monitored the hydrogen consumption.

The acid properties of the samples were evaluated using

pyridine adsorption isotherms, and experiments were carried

out in the liquid phase with a non-polar solvent (cyclohex-

ane). Previously, the samples were heat treated at 250 °C un-

der N₂flow for 60 min and cooled down to room tempera-

ture. Pyridine adsorption experiments were carried out at 25

°C using a closed glass flask containing 50 mg of the solid

sample and 4 mL of cyclohexane. Solutions of known pyri-

dine and n-heptane (used as the internal standard) concen-

trations in cyclohexane were added by injection with a micro-

syringe. After each injection, the pyridine concentration in

the solution was determined by gas chromatography, using a

flame ionization detector (FID) and a capillary DB-5 column

(20 m × 0.18 mm × 0.4 μm). The amount of pyridine

adsorbed by the solid surface area was calculated according

to the equation described in a previous work.²⁴

2.3. Catalytic tests

The catalytic reaction between ethanol and acetone was

performed using a conventional gas-phase system with a

fixed bed reactor at atmospheric pressure. A reaction mixture

containing ethanol and acetone vapours was generated by

passing N₂(total flow of 30 mL min⁻¹) through two flasks

with controlled temperature, one at 25 °C (ethanol) and the

other at 5 °C (acetone). The catalysts were previously acti-

vated in situ under N₂flow (30 mL min⁻¹) at 450 °C for 60

min. The catalytic tests were carried out at 350 °C using an

ethanol (Et) and acetone (Ac) mixture at molar ratios of 0.5

and 2. The amount of reactant fed per unit time to the inlet

of the reactor was 9.87 mmol h⁻¹and 5.40 mmol h⁻¹for etha-

nol and acetone, respectively, for the Et/Ac ratio of 2 and 4.96

mmol h⁻¹and 10.75 mmol h⁻¹for ethanol and acetone, re-

spectively, for the Et/Ac ratio of 0.5. In order to evaluate the

effect of temperature on the catalytic behaviour, catalytic

tests at temperatures of 300 and 350 °C were carried out for

the bimetallic 0.5CuFeAl catalyst. The products generated

were identified using a gas chromatograph coupled with a

mass spectrometer (GC-MS, QP 2010 Plus model from

Shimadzu). The product quantification and the catalytic frac-

tional conversion of reagents were monitored by means of a

gas chromatograph equipped with a FID and a DB-5 capillary

column (20 m × 0.18 mm × 0.4 μm). Taking into account the

difference between the amount of reactant introduced and

the amount that left the reactor and the division by the

amount of reactant introduced, the fractional conversion was

determined. The reactant conversion was taken in a specific

rate in μmol g_cat⁻¹h⁻¹.⁷

The presence of basic sites was determined by temperature-

programmed desorption of CO₂(CO₂-TPD). Prior to TPD, the

samples were heat treated under He flow (20 mL min⁻¹) at 500

°C for 30 min and cooled under the same gas flow to 25 °C.

CO₂adsorption was carried out by using a flow of pure CO₂

through the reactor for 30 min at 25 °C. After baseline stabili-

zation under He flow (20 mL min⁻¹) at room temperature, the

CO₂-TPD run was carried out with a heating rate of 10 °C

min⁻¹from room temperature up to 500 °C. CO₂desorption

was monitored using a thermal conductivity detector (TCD).

⁵⁷Fe Mössbauer spectroscopy measurements were taken at

room temperature for fresh and spent catalysts with a

3. Results and discussion

3.1. Chemical composition and structural features of the

catalysts

Table 1 shows the results of chemical analysis by atomic ab-

sorption spectroscopy. The monometallic 10FeAl and bime-

tallic CuFeAl catalysts show iron and copper content very

close to the nominal values, which are 10 wt% Fe and 0.5 or

1 wt% Cu. This suggests that no Fe loss occurred during the

Cu impregnation process to obtain bimetallic solids. The

higher Fe wt% for the samples 0.5CuFeAl and 1CuFeAl relative

to the sample 10FeAl (Table 1), point out to the Al³⁺leaching

This journal is © The Royal Society of Chemistry 2017

Catal. Sci. Technol.

View Article Online

Paper

Catalysis Science & Technology

Table 1 Iron (Fe) and copper (Cu) contents in catalysts determined by

atomic absorption spectroscopy analysis

properties of the solids, as well as to observe changes in the

oxidation state of the iron oxide due to the reaction condi-

tions employed. Fig. 2 shows the Mössbauer spectra of the

fresh samples while the hyperfine parameters of the doublets

observed are summarised in Table 2.

Samples

Metal

content (wt%)

10FeAl

0.5CuFeAl

1CuFeAl

Fe

Cu

9.36

—

9.48

0.75

10.66

1.17

The Mössbauer spectra of the 10FeAl sample displayed in

Fig. 2 show the presence of two doublets, which is typical of

the Fe³⁺present in the structure of the iron oxide phase (he-

matite).^32,33However, the presence of a second doublet may

be due to the effect of the aluminum oxide on the iron oxide

structure, which results in a different chemical environment.

On the other hand, the Mössbauer spectra of the sample

1CuFeAl show three doublets, which may be due to the pres-

ence of the hematite phase (Fe₂O₃) and copper ferrite

(CuFe₂O₄),³⁴in addition to the aluminium oxide effect.

Therefore, in the 1CuFeAl sample, the iron oxide is in a dif-

ferent chemical environment, resulting in the presence of three

doublets. The presence of doublets in the Mossbauer spectra

suggests that the iron oxide is well dispersed with particles hav-

ing nanometer diameter as indicated by the XRD results

(Fig. 1), which probably has superparamagnetic properties.^32,34

during the Cu²⁺impregnation process. However, take into ac-

count the content of iron oxide value used (10 wt%), this

change in the iron oxide percentage is not significant.

The powder XRD patterns of the samples after the calcina-

tion process at 500 °C are shown in Fig. 1. The profiles

clearly point out a structural change due to copper impregna-

tion. The diffraction pattern of the monometallic 10FeAl cata-

lyst points to a poorly crystalline material, typical of an amor-

phous solid. Considering the calcination temperature of 500

°C, the result suggests a high dispersion of Fe oxide species

on the Al₂O₃matrix.

This feature is commonly found in materials prepared by

the polymeric precursor method.^12,20

The addition of the copper precursor by a wet impregnation

method revealed the formation of broader peaks correspond-

ing to the cubic crystalline phase of γ-Al₂O₃(ICSD – 66559)

with main peaks at 2θ = 36.4, 43.9, 53.5, 71.2 and 79.1° for

both bimetallic samples. Moreover, the diffractograms do not

point to any reflections belonging to crystalline Fe or Cu oxide

phases, indicating that both metal oxides are present in an

amorphous state or with a high dispersion on the alumina ma-

trix. On the other hand, if the crystalline phase of Fe₂O₃is

present, the crystallite diameter is very small.

3.2. Evaluation of reduction behaviour of the catalysts by

H₂-TPR

Temperature-programmed reduction (H₂-TPR) has been a

powerful tool for examining the reduction behaviour and also

the interaction effect between two metal oxide species of cata-

lysts.^34,35Thus, the H₂-TPR technique was used to evaluate

the effects of CuO on the reduction properties of bimetallic

CuFeAl catalysts, and the results are shown in Fig. 3.

The monometallic 10FeAl sample exhibits two broad peaks

of H₂consumption. The first is in the range of 250 to 650 °C

and the second, which overlaps with the first, extends to 970

°C. Indeed, this is a common feature of Fe-based catalysts,

Mössbauer spectroscopy measurement was performed in

order to obtain additional information about the structural

Fig. 2 Mössbauer spectra of the fresh samples at room temperature

Fig. 1 XRD patterns of catalysts calcined at 500 °C under air flow.

(300 K).

Catal. Sci. Technol.

This journal is © The Royal Society of Chemistry 2017

View Article Online

Catalysis Science & Technology

Paper

Table 2 Hyperfine parameters of the Mössbauer spectra for the fresh

mono- and bimetallic catalysts recorded at room temperature (300 K)

peak with a maximum at the lowest temperature (251 °C).

This indicates a characteristic temperature reduction for

CuO. Following the process, it is observed that H₂was con-

sumed up to 600 °C. The shift of the temperature of H₂con-

sumption in the profiles of both samples (0.5CuFeAl and

1CuFeAl) may be interpreted as a strong Cu–Fe interaction,

with the possible formation of a CuFe₂O₄phase.²⁵Further-

more, the sample 1CuFeAl should present a higher amount

of Cu²⁺exposed on the surface, which facilitates the reduc-

tion of oxides. It is known that the reducibility of bimetallic

catalysts is strongly affected by the interaction between two

metal species, where the copper (for instance) can catalyse

the reduction of the other transition metal.^24,35Therefore,

the addition of CuO to FeAl significantly changes the redox

properties of the catalyst with the main peaks shifting to

lower reduction temperatures. It is reasonable to propose

that the displacement of the entire iron oxide reduction band

is only possible if the copper oxide is well dispersed, besides

the strong interaction between Cu²⁺and Fe³⁺species on the

catalysts. The 1CuFeAl sample shows a weak H₂consumption

peak at high temperature (800–950 °C), which suggests the

reduction of the iron spinel structure formed by a stronger

interaction with the support alumina.³⁸

Samples

10FeAl

IS, mm s⁻¹

QS, mm s⁻¹

R.A.%

Fe species

0.218

0.193

0.197

0.198

0.207

0.198

0.204

1.02

1.60

0.82

1.32

0.78

1.19

1.62

61.2

38.8

49.7

50.3

56.8

27.4

15.7

Fe³⁺

0.5CuFeAl

1CuFeAl

IS = isomer shift, QS = quadrupole shift and R.A. = relative spectral

area.

where the phases of iron oxide present support effects.¹²The

first peak with a maximum near 566 °C is assigned to the

partial reduction of Fe³⁺to Fe²⁺species on structures of iron

oxide present as hematite and/or magnetite.^12,36However,

the non-symmetrical profile, which starts near 250 °C, sug-

gests that a fraction of the iron oxide is reduced to metallic

iron in the first H₂consumption band. This is consistent

with the relative amount of H₂consumed in the first band

(250–650 °C), which is similar to that in the second band.

The second large H₂consumption band, centred at about 800

°C, can be ascribed to the reduction of the residual Fe²⁺to

produce metallic iron.³⁶

3.3. Textural and morphological properties of the solids

Interestingly, it is evident from Fig. 3 that the addition of

copper affected the reducibility of the iron species, and the

effect on the reduction is more significant in the sample with

higher Cu content. The TPR profile shows an expressive shift

for the reduction peaks of iron species to lower temperature

in both bimetallic Cu–Fe catalysts. This indicated that the re-

ducibility of the iron oxide species is modified by Cu addi-

tion. The 0.5CuFeAl sample displays only one broad peak of

H₂consumption with a maximum at 380 °C, which is due to

the simultaneous reduction of copper and iron oxide spe-

cies.³⁷However, the H₂consumption signal of the catalyst

with higher Cu content (1CuFeAl) exhibits a strong and sharp

Fig. 4 presents the N₂adsorption/desorption isotherms and

pore size distribution profiles of the samples calcined at 500

°C. The sample 10FeAl presents a typical isotherm profile of

a macroporous material containing a low fraction of meso-

pores, which is responsible for the type H3 hysteresis loop

observed.³⁹

Interestingly, after the wet impregnation with copper, the

isotherm profiles changed significantly. The isotherm of the

10FeAl sample practically does not present a point of identifi-

able monolayer formation, while for the 0.5CuFeAl and

Fig. 3 H₂-TPR profiles of the fresh samples calcined at 500 °C. The

Fig. 4 N₂adsorption/desorption isotherms (a) and BJH pore size

intensity of the signal for the 10FeAl sample was multiplied by 2.5.

distribution from the desorption branch (b) for the fresh samples.

This journal is © The Royal Society of Chemistry 2017

Catal. Sci. Technol.

View Article Online

Paper

Catalysis Science & Technology

1CuFeAl samples, it is easily identified. Therefore, the sam-

ples 0.5CuFeAl and 1CuFeAl should present a higher surface

area, as presented in Table 3. Based on the IUPAC classifica-

tion, the samples 0.5CuFeAl and 1CuFeAl show isotherm pro-

files that seem to be a combination of types I and II, with an

per precursors, where small pores with a size of 3–4 nm are

generated.⁴²These modifications in textural properties

(Table 3) are consistent with the XRD changes (Fig. 1) ob-

served for the CuFeAl catalysts. Therefore, during the wet im-

pregnation with Cu²⁺, aluminium oxide rehydration occurs,

and with the slow water evaporation, the crystallization pro-

cess takes place. The Al₂O₃rehydration process is consistent

with the increase in Fe content observed in Table 1, since a

fraction of the Al may be lost during the process.

H4 hysteresis loop, which is typical of

a mesoporous

structure.^39–41However, the profiles also resemble the H3

hysteresis loop,³⁹suggesting samples with a complex mixture

of pore shapes. For the 10FeAl sample, significant hysteresis

loops at relative pressure (P/P₀) above 0.50 and no distinct

uptake at low relative pressure are observed, which strongly

suggests the predominance of macropores with the presence

of mesopores and absence of microporosity.

On the other hand, it is noticed that the sample with the

highest copper content (1CuFeAl) shows a smaller surface

area relative to the 0.5CuFeAl sample, and such a decrease is

observed in the micropore range. Thus, the above results sug-

gest that wet impregnation with a low amount of copper

should improve the morphological properties of the alumin-

ium oxide. However, a higher copper content could lead to

deposition of a certain amount of oxide species into the po-

rous structure of the support resulting in partial blocking of

the pores, mainly micropores, and formation of larger meso-

pores, as can be seen on the 1CuFeAl catalyst (Table 3).

The morphological aspects and elemental distribution of

the solids are shown by SEM-EDS analyses (Fig. 5 and 6). The

pure 10FeAl sample is composed of irregular and inhomoge-

neous pores with a wide range of shapes and sizes (Fig. 5a). It

has a typical shape due to the synthesis route applied.²⁷The

majority of the pore width detected by N₂adsorption belongs

to a range lower than that detectable by the magnification ap-

plied in Fig. 5. However, the pore width observed by SEM for

the 10FeAl sample is consistent with the pore distribution

(Fig. 4), which pointed to the predominance of macropores.

Fig. 5 shows the evident change in the morphology of the

samples after the incorporation of copper (Fig. 5b and c). Af-

ter wet impregnation, the sponge-like morphology (Fig. 5a) is

completely modified to platelet-shaped particles, whose aver-

age size is evidently larger for the sample 1CuFeAl (Fig. 5c).

The morphology change shown by SEM suggests that, in ad-

dition to the recrystallization process, the increasing surface

area and pore volume (Table 3) are also due to the change of

the particle size and shape. The micropore volume decrease

observed for the sample 1CuFeAl relative to the sample

0.5CuFeAl points to a significant contribution of the space

between the particles, since a large particle size results in a

lower total pore volume between the particles.

On the other hand, the incorporation of copper by wet im-

pregnation into the 10FeAl catalyst led to a notable change in

the form of the isotherm. An increase in the volume of N₂

adsorbed at a low relative pressure (P/P₀< 0.1) and a contin-

uous increase of N₂adsorption along the entire adsorption

branch are observed. This indicates the formation of a new

micro- and mesoporous structure on both bimetallic solids.⁴¹

From the pore size distributions (according to the BJH

method) in Fig. 4(b), it can be seen that the 10FeAl sample is

essentially a macroporous sample with a fraction of meso-

pores. After impregnation of the 10FeAl catalyst with copper,

it still shows the presence of macropores. On the other hand,

the profiles point to the presence of micropores, in addition

to the amount of mesopores improving for both bimetallic

solids, 0.5CuFeAl and 1CuFeAl, if compared to that in the

10FeAl catalyst.

Table 3 shows the detailed textural characteristics of the

catalysts. It is shown that the 10FeAl catalyst has a rather low

surface area and pore volume, pointing to a predominance of

macropores.

The copper impregnation has a large effect on the textural

properties of the catalysts, such as the increase in BET sur-

face area and pore volume. However, as supported by the

presence of the point of identifiable monolayer formation

and by the t-plot results (Table 3), after the addition of cop-

per, the sample presented the presence of micropores. This

textural change can favour the formation of highly disperse

copper species over the 10FeAl catalyst, the effect of which

could be attributed to the rehydration/recrystallization pro-

cess over surface alumina during the impregnation with cop-

In order to evaluate the distribution of copper and iron

species, SEM-EDS analyses were conducted (Fig. 6). Despite

the low magnification of the EDS mapping image, the

10FeAl sample (Fig. 6a and b) demonstrates the presence of

areas with predominance of aluminium (blue spots), which

suggests that the sample has an inhomogeneous surface.

For the bimetallic 0.5CuFeAl catalyst, the elemental EDS

mapping also shows blue spots (Fig. 6c and d); however,

they are small and lower in number. Similar properties are

also observed for the 1CuFeAl sample (Fig. 6e). Therefore,

the wet impregnation with Cu²⁺should promote the

redispersion of the aluminium oxide by means of dissolu-

tion of a fraction of the sample followed by recrystallization.

Table 3 Textural characteristics of the solids studied, from the N₂ad-

sorption/desorption isotherms

Surface area (m²g⁻¹

)

Pore volume (cm³g⁻¹

)

Samples

S_BET

S_micro

S_ext

V_micro

V_meso

V_total

10FeAl

0.5CuFeAl

1CuFeAl

35

431

351

—

129

39

—

302

312

—

0.017

0.003

—

0.592

0.624

0.272

0.609

0.627

S_BET= BET surface area. S_microand V_micro= micropore area and

volume calculated by the t-plot method. V_meso= volume of mesopores

estimated from the equation: V_meso= V_total− V_micro. V_total= pore vol-

ume at a P/P₀of 0.99.

Catal. Sci. Technol.

This journal is © The Royal Society of Chemistry 2017

View Article Online

Catalysis Science & Technology

Paper

Fig. 5 SEM microphotographs of the catalysts: 10FeAl (a), 0.5CuFeAl (b) and 1CuFeAl (c).

In addition, the amount of copper added has an effect on

the morphological properties of the sample, as suggested by

the SEM-EDS images (Fig. 6). The sample 1CuFeAl presented

a larger size of platelets, in addition to the higher ratio of

aluminium/iron, pointing to the iron ratio decrease (spectral

analysis). This is consistent with the dissolution followed by

recrystallization, since the Al₂O₃recrystallization process

may cover the iron oxide with the aluminium oxide. More-

over, this surface property change can somehow affect the

EDS signal. Indeed, it is important to emphasize that the

atomic absorption spectroscopy results (Table 1) pointed to

a higher iron amount for the sample 1CuFeAl.

summarizes the total acid site densities calculated from the

fit curves in Fig. 7, which correspond to the maximum

amount of pyridine adsorption or complete monolayer forma-

tion. The adsorption isotherm data of pyridine on the sam-

ples can be adequately interpreted by nonlinear fitting to a

Langmuir isotherm model (Fig. 7).

The profiles presented in Fig. 7 show a higher amount of

pyridine adsorption on the 10FeAl sample. In addition, it

shows a more pronounced slope at the lowest equilibrium

concentration, suggesting the presence of stronger acid sites

in the sample 10FeAl (Fig. 7). It is known that pyridine mole-

cules interact with surface hydroxyl groups by hydrogen

bonding forming pyridinium ions as well as with Lewis acid

sites involving metal ions,^43–45and both sites can exist on the

surface of the samples presented here. The addition of cop-

per affects the acid site densities, as pointed out in Fig. 7. It

is known that the Cu²⁺interaction in the wet impregnation

process occurs on surface oxygen.⁴⁶This process may result

in Brønsted acid site elimination with new Lewis acid site for-

mation. However, the electronic properties of copper may

also affect the strength of the nearest sites.⁴⁷In fact, with the

addition of copper, the inclination of the curve and pyridine

sorption capacity decrease, indicating weaker acid sites on

the bimetallic copper–iron catalyst.

3.4. The acidic and basic properties

Equilibrium adsorption isotherms of pyridine at 25 °C were

used to investigate the acidic properties of the catalysts.

Fig. 7 shows the results of the amount of pyridine adsorbed

per surface area with the corresponding curves, and Table 4

Fig. 6 EDS mapping images with respective spectra of 10FeAl (a and

b), 0.5CuFeAl (c and d) and 1CuFeAl (e and f) catalysts.

Fig. 7 Adsorption isotherms of pyridine measured on the catalysts at

25 °C. The lines represent the Langmuir model isotherm.

This journal is © The Royal Society of Chemistry 2017

Catal. Sci. Technol.

View Article Online

Paper

Catalysis Science & Technology

Table 4 Surface acid site density determined from isotherms of pyridine

adsorption, considering the Langmuir model for complete monolayer

adsorption

or the bidentate carbonate species. The CO₂-TPD profiles

point to the effect of wet impregnation with copper on the

CO₂adsorption ability. Considering the weak strength, a

higher amount of CO₂adsorption on the 10FeAl sample is

observed, in the proportion of 1.0, 0.74 and 0.86 for 10FeAl,

0.5CuFeAl and 1CuFeAl, respectively. These data suggest a

higher density of –OH groups (hydroxyls) for these samples.

This result is consistent with the pyridine adsorption results

(Fig. 7), since the addition of copper promotes a decrease of

the –OH groups on the surface of the sample. However, the

modification of the 10FeAl catalyst with different amounts of

copper oxide leads to a shift of the first peak of CO₂desorp-

tion to a higher temperature region (from 80 °C for 10FeAl to

134 °C after copper loading). This suggests an increase of the

strength of the basic sites on the surface for the Cu–FeAl cata-

lysts, which is consistent with the superior homogeneity of the

surface suggested by the EDS mapping (Fig. 6). Therefore, the

bimetallic Cu–FeAl composites with higher pore volume pos-

sess a more homogeneous surface basic site distribution, in ad-

dition to their high surface area (see Table 3). However, despite

the small amount of the moderate strength sites, the samples

containing copper show a shift to higher temperature, which is

more pronounced for the sample 1CuFeAl. Thus, the addition

of copper plays an important role in both acidic and basic prop-

erties of the Cu–Fe catalysts supported on alumina.⁵²

Samples

10FeAl

1.193

0.5CuFeAl

0.441

1CuFeAl

1.021

Acid site density^a(μmol m⁻²

)

a

Amount of pyridine estimated by nonlinear fitting to the Langmuir

model.

A more significant change in the amount of pyridine

adsorbed is observed for the sample with lower copper con-

tent. The superior effect of copper for the sample with a

lower amount is possible if the copper dispersion in the

0.5CuFeAl sample is higher than that in the 1CuFeAl sample,

which is consistent with the TPR profile (Fig. 3).

Therefore, according to the literature,^48,49these results in-

dicated that the addition of CuO should promote mainly a

decrease in Brønsted acidity (acidic hydroxyl group) and lead

to an increase of Lewis acid sites. Thus, it is reasonable to as-

sume that this effect is dependent on the dispersion and

interaction between the copper and iron oxide species on the

alumina surface.

The surface basicity of the as-prepared 10FeAl and Cu–

FeAl composites was evaluated by CO₂-TPD experiments, and

the corresponding curves are displayed in Fig. 8.

The CO₂-TPD curves of all samples showed two types of

CO₂desorption peaks, one with the maximum temperature

range of 25 to 230 °C and another broader peak between 230

and 430 °C, which have been attributed to weak and moder-

ate strength basic sites, respectively.²⁴As proposed previously

by various authors,^10,25,50,51the peak at lower temperature

may be assigned mainly to CO₂desorption from weakly basic

monodentate bicarbonate species formed due to the reaction

of CO₂with the OH groups. The moderate strength basic

sites, however, are ascribed to the interaction of CO₂with the

Lewis acid–base pairs (M⁺–O⁻), which form the monodentate

3.5. Catalytic evaluation in ethanol–acetone conversion

Here, the relationship of the surface acid–base and redox

properties with reaction activities of the mono- and bimetal-

lic copper–iron-based catalysts for gas-phase ethanol–acetone

mixture conversion is discussed.

Effect of the ethanol/acetone molar ratio and nature of

catalyst. The catalytic properties of the samples 10FeAl and

0.5CuFeAl were firstly examined by changing the composition

of the reaction mixture. The ethanol (Et)-to-acetone (Ac) molar

ratio (Et/Ac) was changed from 0.5 to 2.0, and the average cat-

alytic conversion and product selectivity are shown in Table 5.

As seen in Table 5, the 0.5CuFeAl sample, which presented

the highest surface area (Table 3), as well as a lower surface

density of acid sites (Fig. 7), shows a superior catalytic activ-

ity for the conversion of both reactants (ethanol and/or ace-

tone) in comparison to the single 10FeAl oxide. In addition,

it is also clearly shown that the ethanol conversion is higher

than the acetone conversion for both Et/Ac molar ratios stud-

ied. This suggests that the bimetallic 0.5CuFeAl catalyst has a

greater ability to adsorb ethanol, resulting in a solid with su-

perior activity for gas-phase ethanol conversion.

The change of the surface properties due to the addition

of copper is evident in the TPR experiments carried out on

solids (Fig. 3). It was observed that copper makes the reduc-

tion of iron oxide species (Fe³⁺to Fe²⁺) easier. Thus, the com-

bined redox and acid–base properties of the 0.5CuFeAl cata-

lyst can contribute to the highest activity observed.

Considering the catalysts separately, it is observed that in-

creasing the Et/Ac molar ratio from 0.5 to 2 resulted in an

Fig. 8 CO₂-TPD curves of the 10FeAl and bimetallic Cu–FeAl catalysts.

The TCD signal from CO₂-TPD was normalized by the surface area

(BET model, Table 3).

Catal. Sci. Technol.

This journal is © The Royal Society of Chemistry 2017

View Article Online

Catalysis Science & Technology

Paper

Table 5 Average catalytic activity and selectivity to light (S_L), acetaldehyde (S_Ac), isopropanol (S_Iso), butanal (S_Bu) and 2-ethylbutanal (S_Et-bu) products

obtained with the 10FeAl and 0.5CuFeAl catalysts for different ethanol–acetone mixtures

Conversion (μmol g⁻¹h⁻¹

)

Selectivity (%)

c

Samples

Ethanol

Acetone

S_L

S_Ac

S_Iso

S_Bu

4.4

2.0

13.8

2.4

S_Et-bu

10FeAl^a

482.4

825.6

582.7

576.8

615.2

243.4

496.5

41.8

37.6

35.6

32.9

18.4

13.8

18.9

14.8

7.7

9.7

11.1

26.0

14.3

19.9

9.6

0.5CuFeAl^a

10FeAl^b

0.5CuFeAl^b

1246.6

10.6

a

b

Reaction conditions: T = 350 °C; w = 0.1 g; time on stream: 5 h; total flow: 30 mL min⁻¹

Light products: methane, ethane, ethene, CO₂.

.

Et/Ac molar ratio = 0.5. Et/Ac molar ratio = 2.

c

increase of the catalytic conversion of ethanol and a decrease

of the acetone conversion. This may be due to the highest re-

activity of ethanol as a polar protic reactant.

via aldol condensation followed by a sequence of consecutive

dehydration/hydrogenation reactions (Scheme 1). In this reac-

tion pathway, the combined pair of Lewis acid–base sites

plays an important role in the dehydration/dehydrogenation

process, while redox species are known to be active for hydro-

gen transfer reactions.⁵⁵The second possibility is the

2-butenal undergoing a new condensation reaction with acet-

aldehyde, leading to 2-ethylbutanal, after the dehydration/de-

hydrogenation reaction step (Scheme 1). This last reaction

pathway is more likely since 2-butenal is more reactive than

butanal. Additionally, the increase of the Et/Ac molar ratio fa-

vours the hydrogen transfer process, which increases the

butanal selectivity, similarly to isopropanol.

Another possibility is the competitive adsorption of etha-

nol and acetone, which is also consistent, since an increase

of the ethanol concentration should favour ethanol adsorp-

tion to the detriment of acetone adsorption. The increase of

the ethanol conversion, due to the Et/Ac molar ratio increase,

is more pronounced for the sample 0.5CuFeAl, and it is con-

sistent with the ability of copper to catalyse dehydration/de-

hydrogenation reactions of alcohols.^4,7,24

Table 5 shows that both catalysts (10FeAl and 0.5CuFeAl)

present higher selectivity to light products (methane, ethane,

CO₂). It is known that at high temperature (>300 °C), side re-

actions such as decomposition or cracking with cleavage of

the carbon–carbon bond of reactants (ethanol and/or ace-

tone) are predominant.²²However, with the combination of

the acid–base and redox properties of the solids investigated,

the formation of useful chemicals, including acetaldehyde,

isopropanol, butanal and 2-ethylbutanal, is observed, which

are products of dehydrogenation with subsequent aldol con-

densation and/or hydrogenation through a hydrogen transfer

reaction using ethanol as the hydrogen donor.

Therefore, the reactant conversion may be dependent on

an adequate balance of active pairs of acid–base and redox

sites. In such a system, copper may play a significant role in

the redox sites, leading to better catalytic performance of the

Moreover, the results presented in Table 5 show that

changing the composition of the reaction mixture has an ef-

fect on the distribution of the products. For both Et/Ac molar

ratios studied, the Lewis acid–base site pairs of the iron oxide

catalyst should dehydrogenate ethanol to acetaldehyde lead-

ing to condensation products (butanal and 2-ethylbutanal).

These condensation compounds are generated via a sequence

of complex mechanisms, combining consecutive dehydroge-

nation/aldol condensation/dehydration and hydrogenation re-

actions, as proposed elsewhere.^15,53Firstly, ethanol is de-

hydrogenated producing acetaldehyde, then it goes through

cross-condensation to form the acetaldol (3-hydroxybutanal),

which dehydrates to 2-butenal, followed by a hydrogenation

reaction of the CC double bond to give butanal.^16,54

Considering the 10FeAl catalyst, the increase in the Et/Ac

molar ratio induced the increase in the selectivity to butanal,

to the detriment of the selectivity to 2-ethylbutanal. There-

fore, we propose that the synthesis of 2-ethylbutanal may go

through a series of reactions. One possibility is the abstrac-

tion of an acidic hydrogen atom at the α-position to the car-

bonyl group of butanal, which may react with acetaldehyde

Scheme

1 Proposed consecutive reactions for the conversion of

ethanol–acetone (350 °C) over bimetallic 0.5CuFeAl catalysts

containing acid–base and redox sites.

This journal is © The Royal Society of Chemistry 2017

Catal. Sci. Technol.

View Article Online

Paper

Catalysis Science & Technology

bimetallic catalyst. This result can be confirmed by H₂-TPR

experiments and acid–base data of the solids (Fig. 3 and

8 and Table 4).

It should also be noted that the increase in the Et/Ac mo-

lar ratio leads to changes in the product distribution over the

0.5CuFeAl catalyst, with a meaningful increase in the forma-

tion of isopropanol. Furthermore, the selectivity to iso-

propanol increases with the addition of copper and with the

Et/Ac molar ratio increase. This result seems to have a direct

correlation with the ethanol conversion; thus, the iso-

propanol selectivity depends strongly on the molar ratio be-

tween the reactants. According to the literature,^17,19iso-

propanol is generated from acetone via a hydrogen transfer

mechanism, which comes from dehydrogenation of ethanol

to acetaldehyde with effective participation of Lewis acid–

base site pairs (Scheme 2a). However, the occurrence of a hy-

drogen transfer process with the participation of surface hy-

droxyl groups is also possible (Scheme 2b).

On the other hand, it is well known that copper-based cat-

alysts are very active for dehydrogenation reactions of alco-

hols to aldehydes.⁷Additionally, we have reported recently

that Cu species play an important role in the redox and acid–

base properties of transition metal oxides, which could catal-

yse hydrogen transfer reactions with reduction of intermedi-

ate chemicals.²³Therefore, the higher conversion of acetone

observed for the 0.5CuFeAl sample may be due to decomposi-

tion reactions, but mainly to the increase in isopropanol pro-

duction. Based on the experimental results and the literature

information, we propose a reaction sequence for ethanol–ace-

tone mixture conversion with the participation of acid–base

and redox sites present on the catalysts studied (see

Schemes 2a and b). The portion highlighted with a rectangle

represents a partially reduced site.

In Scheme 2a, it is proposed that a hydrogen (hydride)

transfer from the ethanol adsorbed on the Lewis acid–base

site pairs to the acetone adsorbed on a Lewis acid site occurs.

Concomitantly, the acetone's oxygen abstracts the hydrogen

(H⁺) from the second ethanol molecule adsorbed, which may

form an ethoxy on the catalyst surface, with the formation of

isopropanol. The adsorption of a new acetone molecule, near

the ethoxy, and a hydride transfer may occur and, with the

participation of a new ethanol molecule, another isopropanol

is formed with partial reduction of the catalyst (Scheme 2b).

On the other hand, although hydrogen transfer reactions

can indeed take place in both mono- and bimetallic catalysts,

with low or high Et/Ac molar ratios, the selectivity to iso-

propanol is considerably lower for the unmodified 10FeAl cat-

alyst. Based on this result, ethanol can be dehydrogenated to

acetaldehyde over a catalyst containing copper or iron oxide,

but the improved redox properties of dual sites play a key

role in the hydrogen transfer reaction mechanism, where ace-

tone can be easily reduced to isopropanol with the metal re-

dox site (Cu⁺/Cu⁰and Fe²⁺) supports.

Scheme 2 Proposed reaction mechanism for hydrogen transfer from

ethanol to acetone over catalysts containing acid-basic and redox

sites. (a) Lewis acid-base site pairs, (b) hydroxyl groups.

ated for the bimetallic (0.5CuFeAl) catalyst and the results

are depicted in Fig. 9.

The results in Fig. 9a show that the conversion of both re-

actants (ethanol and acetone) decreased with time-on-stream

at both temperatures of the reaction (300 °C or 350 °C). Gen-

erally, this result is attributed to loss of the active sites

caused by the deposition of carbon species on the catalyst

surface, which is mainly observed at high reaction tempera-

tures.⁵⁶In addition, acetone is a polar aprotic chemical and

consequently is less reactive than ethanol in any temperature

studied. Reactant conversion increases with increasing tem-

perature due to thermodynamic reasons, since higher tem-

peratures favour the endothermic reactions such as decompo-

sition, dehydrogenation and dehydration.^7,53

Fig. 9b shows the effect of temperature on the selectivity

to main reaction products. As can be seen, at higher reaction

temperature (350 °C), the selectivity to light products

obtained from decomposition of the reactants is superior

Effect of reaction temperature and copper content. After

choosing the better feed composition (Et/Ac ratio), the effect

of the temperature on the catalytic performance was evalu-

Catal. Sci. Technol.

This journal is © The Royal Society of Chemistry 2017

View Article Online

Catalysis Science & Technology

Paper

Scheme 3 Proposed reaction mechanism for ethanol dehydrogenation

over catalysts containing acid-basic and redox sites. (a) Acetaldehyde

or (b) ethoxy formation.

such as acetaldehyde, isopropanol and 2-ethylbutanal. How-

ever, by decreasing the reaction temperature, the selectivity

to butanol increases from 3% to 6%, which is prejudicial to

the posterior aldol condensation for 2-ethylbutanal forma-

tion. Consequently, it may increase the acetaldehyde

selectivity.

Fig. 9 Effect of reaction temperature on the catalytic performance of

the 0.5CuFeAl sample: (a) conversion, (b) product selectivity. Et/Ac

ratio = 2.

compared to the dehydrogenation (acetaldehyde), hydrogena-

tion (isopropanol) and aldol condensation (2-ethylbutanal

and butanal) products. In contrast, decreasing the reaction

temperature to 300 °C results in an increase in the selectivity

to acetaldehyde and butanol. However, the isopropanol selec-

tivity achieved is nearly the same, while the selectivity to light

products decreases by nearly 1/3 (Fig. 9b). At 300 °C, a higher

amount of acetaldehyde is formed, despite the dehydrogena-

tion reactions being typically endothermic.^7,57,58This change

in the product selectivity points to ethanol as the main

source of the light products (methane, ethane, and ethene).

The dehydrogenation of ethanol may generate acetaldehyde

and adsorbed hydrogen⁵⁹or ethoxy (Schemes 3a and b).

Thus, ethanol acts as a hydrogen source, resulting in the

rapid reduction of acetone to isopropanol over active Cu⁺/Cu⁰

and/or Fe²⁺redox sites of the Cu–FeAl catalyst. This is consis-

tent with the TPR results, which point to an easier reducibil-

ity of the CuFeAl catalyst. The direct hydrogen transfer to ace-

tone (Schemes 2a and b) should not be ruled out. Therefore,

our experiments show that acetone acts as the reactant for

isopropanol formation, the reaction of which is limited by

the hydrogen obtained from dehydrogenation of ethanol to

acetaldehyde. Additionally, experiments carried out at 300 °C

also show a noticeable amount of butanol (6.7%). In accor-

dance with the findings,^53,54butanol formation involves path-

ways which proceed through acetaldehyde self-aldolization

with consecutive dehydration and hydrogenation of interme-

diates (Scheme 1).

In heterogeneous catalysis, the combined multi-functional

pathway requires an active site with different functionality

and chemical nature in order to complete the complex net-

work of reaction mechanisms very efficiently.⁶⁰The combina-

tion of redox and acid–base active sites on the CuFeAl cata-

lyst should provide the required bi-functionality for the

ethanol–acetone conversion to highly valuable chemicals.

Thus, we have investigated the effect of the copper content

increase from 0.5 to 1% over bimetallic copper–iron catalysts

on the reactant conversion and product distribution, and the

results are shown in Fig. 10.

The catalytic performance of the solids presented in

Fig. 10a and b clearly indicates better catalytic activity (con-

version of both reactants), for the sample with higher copper

content (1CuFeAl). In addition, the data also suggests that

the increase in copper amount also contributes to an im-

provement in the catalytic stability. According to the N₂ad-

sorption experiments (Fig. 4 and Table 3), both samples are

mesoporous; however, the increase of copper content in the

catalyst composition promotes a decrease in the total (BET)

and microporous surface area with a consequent increase of

the pore size. Therefore, the catalytic conversion results are

consistent with the highest porosity of the 1CuFeAl catalyst,

which contributes to easier reactant and product diffusion in

the catalyst pores leading to better catalytic performance. Ad-

ditionally, its higher activity for reactant conversion should

also be due to the higher density of acid–base active sites on

the surface of the 1CuFeAl catalyst than that of the 0.5CuFeAl

sample (Fig. 7 and 8). Consequently, the catalytic perfor-

mance highlights the contribution of both catalyst acid–base

Fig. 9b suggests that the selectivity to butanol and butanal

becomes negligible compared to the other chemicals formed,

Catal. Sci. Technol.

View Article Online

Paper

Catalysis Science & Technology

ward dehydrogenation/dehydration as well as hydrogen trans-

fer reactions. The maintenance of selectivity to acetaldehyde

and isopropanol with suppression of that to 2-ethylbutanal,

due to the copper content increase, points to the requirement

of a catalyst with a higher copper dispersion in order to pro-

duce a superior amount of condensation products.

The product selectivity profile for the sample with a higher

amount of copper (Fig. 10d) clearly points out the effect of

the formation of reduced species on the hydrogen transfer.

The partial reduction process, which occurs mainly on the

first hour of time-on-stream, improves the isopropanol pro-

duction. In the same range of reaction time, a little decrease

in acetaldehyde and 2-ethylbutanal and an increase in the

light products are observed. This behaviour may suggest that

the partially reduced catalyst presents a lower dehydrogena-

tion activity, which decreases the acetaldehyde selectivity.

Consequently, the amount of products from acetaldehyde

condensation also decreases. However, the partially reduced

catalyst presents superior isopropanol production by the hy-

drogen transfer process.

Therefore, the little increase of the selectivity for the light

products, followed by the decrease of the ethanol conversion

(Fig. 10b), points to a surface property change of the catalyst.

3.6. Characterization of spent catalysts

In order to find a correlation between the catalyst properties

and catalytic performance, characterization of the spent cata-

lysts by thermal analysis (TGA/DTA), H₂-TPR and Mössbauer

spectroscopy was carried out.

Fig. 11 shows the thermal analysis profiles. All the cata-

lysts present a first mass loss event (endothermic) up to near

220 °C, followed by a second event (exothermic), which ex-

tends up to 600 °C. The first, at lower temperature, is due to

desorption of the physically adsorbed water, ethanol or ace-

tone. The second range (220–600 °C) is due to carbonaceous

material burning (alkoxides on the surface), whereas the

Fig. 10 Effect of Cu content on the catalytic performance of CuFeAl

catalysts: (a and c) ethanol and acetone conversion; (b and d) product

selectivity. T = 350 °C, time = 5 h, Et/Ac molar ratio = 2.

characteristics and porosity to the conversion of reactants

(ethanol and acetone).

As pointed out previously (Fig. 10b), the higher Cu content

improved the catalytic conversion and stability for both reac-

tants. Unfortunately, there is an increase in the selectivity to

decomposition products (Fig. 10d) and a decrease in the

2-ethylbutanal ratio. This result confirms again that at the

high reaction temperature employed, the decomposition reac-

tions of reactant molecules are favoured, mainly for the

1CuFeAl catalyst. However, this higher selectivity for decom-

position products is consistent with the relative acidic site

density determined using pyridine adsorption isotherms

(Fig. 7), which showed that the 1CuFeAl sample has much

more acid sites than the 0.5CuFeAl sample.

Interestingly, both samples (0.5CuFeAl and 1CuFeAl) pres-

ent the same selectivity to acetaldehyde, but a little increase

of isopropanol selectivity is observed for the sample 1CuFeAl

(Fig. 10d). It is known that the reduced copper species (Cu⁺/

Cu⁰) generated during the process present a good dehydroge-

nation ability.⁷Hence, the interaction between copper and

iron species in the bimetallic Cu–Fe catalyst may result in the

formation of a higher concentration of reduced metal species

(Cu⁺, Cu⁰and Fe²⁺) highly active in the catalytic process to-

Fig. 11 TGA/DTA (a) and DTG (b) curves of the spent catalysts

obtained under synthetic air flow. Reaction temperature of 350 °C and

Et/Ac molar ratio of 2.

Catal. Sci. Technol.

View Article Online

Catalysis Science & Technology

Paper

inflexion point at 300 °C or 370 °C is relatively low to be con-

sidered coke burning.^25,61All the catalysts show nearly the

same amount of mass loss in the temperature range of 220 to

600 °C: 15 wt% for the 10FeAl sample and 11 wt% for the

samples containing copper. Interestingly, the samples

0.5CuFeAl and 1CuFeAl present two inflexion points (DTA

and DTG) in the range of 220 to 600 °C, while the sample

10FeAl shows only one in the same temperature range. The

new inflexion point near 310 °C suggests burning of the alk-

oxide, which is formed at the copper surface. Therefore, the

burning of the alkoxide at the copper surface occurs at lower

temperature comparatively to the alkoxide on the surface of

the iron site. Additionally, since the event near 310 °C is at-

tributed to the alkoxide burning from the copper surface, a

careful analysis of the profile change due to the copper

amount increase shows that the burning event near 310 °C is

more pronounced for the sample 0.5CuFeAl. This is consis-

tent with the higher copper dispersion suggested by the TPR

profile for the sample 0.5CuFeAl.

The sintering promotes the copper dispersion decrease,

which is consequently ascribed to the typical reduction of

iron oxide in the TPR profile. As pointed out for the 10FeAl

sample, the Et/Ac molar ratio has an effect on the sintering

process, since a higher ethanol amount results in a shoulder

at 226 °C, which is characteristic of CuO reduction

(Cu²⁺),^62,63pointing out that this species (CuO) is formed un-

der reaction conditions of higher ethanol/acetone molar ratio

(Fig. 12b).

That is, during the catalytic process, the surface properties

of the 0.5CuFeAl sample are modified to a certain extent.

However, the superior H₂consumption with a maximum near

400 °C points to the maintenance of a high fraction of sites

with the same properties as the fresh sample (Fig. 3).

On the other hand, the presence of Cu⁺should not be

ruled out since the reduction of Cu⁺may occur in the temper-

ature range of 300 to 430 °C.⁶³The new sites, with H₂con-

sumption in

a temperature range higher than 500 °C

(Fig. 12b), are similar to those observed in the 10FeAl sample,

with lower relative intensity, however.

This behaviour, together with the lower surface area, can

explain the lower catalytic conversion for the sample 10FeAl,

since a stronger interaction means a lower desorption rate

and consequently fewer available sites.

Therefore, these surface changes observed by TPR experi-

ments are consistent with the TG results, which showed two

inflexion points in the range of 220 to 600 °C for the samples

containing copper oxide (Fig. 11). The inflexion point at 375

°C is due to the alkoxide burning from the iron oxide surface,

while that at 310 °C is due to the alkoxide from the region

rich in copper oxide.

The partial reduction of the iron oxide during the catalytic

process is also evidenced in the Mössbauer spectra obtained

at room temperature for the spent samples (Fig. 13).

The corresponding hyperfine parameters are summarized

in Table 6. The profile spectra and the hyperfine parameters

are similar to those of the fresh samples. However, all the

samples show the formation of at last one new doublet,

which is due to the presence of Fe²⁺, as pointed out in

Table 6.

The TPR profiles of the spent catalysts, 10FeAl and

0.5CuFeAl, are presented in Fig. 12. The 10FeAl sample shows

two maxima of H₂consumption, the first near 560 °C and

the second near 800 °C, which are similar to those of the

fresh sample (Fig. 3). However, the spent catalyst (10FeAl)

presents a shoulder at 350 °C, which is not evident in the

fresh sample. The temperature of 350 °C for iron oxide reduc-

tion is characteristic of Fe³⁺partial reduction,^12,27without

the support effect. Therefore, the H₂consumption profile ob-

served for the 10FeAl sample points to the occurrence of a

sintering process, at lower extension. The sintering process

causes an increase of the iron oxide particle diameter, which

consequently presents a lower support effect on the redox

properties. Increasing the Et/Ac molar ratio results in a lower

amount of H₂consumption in the range of 200 to 600 °C,

and the shoulder at 350 °C is less significant. This profile

change, to lower H₂consumption, points to superior partial

reduction of the iron oxide during the catalytic process with

Et/Ac = 2.

On the other hand, the relative spectral area of Fe²⁺and

Fe³⁺shows that the addition of copper produces a lower

amount of Fe²⁺during the reaction process. This suggests a

rather reduction of the copper oxide under the experimental

conditions applied. The Fe²⁺relative spectral area is 33.8, 8.2

and 19.9% for 10FeAl, 0.5CuFeAl and 1CuFeAl, respectively

The TPR profile of the sample 0.5CuFeAl presents a signif-

icant change after the catalytic process. While the fresh sam-

ple presents only one H₂consumption band (Fig. 3), which

was assigned to the high copper oxide dispersion, the TPR

profile of the spent sample shows three H₂consumption

bands. The first, near 400 °C which is the temperature corre-

sponding to maximum H₂consumption, is similar to that in

the fresh sample (Fig. 3). The H₂consumption peaks near

600 °C and higher than 800 °C are ascribed to iron oxide re-

duction. The last one is due to Fe²⁺reduction to Fe⁰, while

the band at 600 °C is due to Fe³⁺reduction to Fe²⁺and reduc-

tion of a fraction of Fe²⁺to Fe⁰.

Fig. 12 H₂-TPR analysis of spent catalysts in the reaction carried out

at 350 °C and an Et/Ac molar ratio of 0.5 or 2: (a) 10FeAl and (b)

0.5CuFeAl samples.

Therefore, similarly to the 10FeAl sample, the TPR profile

points to the sintering process during the catalytic reaction.

Catal. Sci. Technol.

View Article Online

Paper

Catalysis Science & Technology

fall in stability, which is consistent with the reaction pathway

proposed in Schemes 2b and 3.

The partial reduction of the metal oxide (CuO or Fe₂O₃)

in the dehydrogenation of ethanol may occur by various

ways, with participation of the surface hydroxyl groups

(Schemes 2b and 3), or in the Lewis acid–base site pair

(Scheme 2a). All of the proposed reaction pathways imply a

redox process with the oxidation of ethanol and the reduction

of acetone and/or the metal oxide.

4. Conclusion

The role of combined acid–base and redox sites on copper–

iron catalysts supported on γ-alumina in ethanol–acetone

mixture conversion was investigated by consecutive dehydro-

genation–condensation or by dehydrogenation–hydrogena-

tion through hydrogen transfer reactions, leading to valuable

chemicals. The results point out that reaction parameters

such as the temperature and reactant molar ratio affect the

catalytic performance; however, the modification of textural,

acidic/basic and redox properties on the surface of the cata-

lysts has a strong effect. The characterisation of the samples

shows that the high surface area and dispersion of copper

and iron species produce an adequate balance between the

acid–base and redox active sites towards an efficient catalyst.

Furthermore, the addition of an adequate CuO amount posi-

tively modifies the surface properties such acid–base pairs

on the FeAl catalyst, which results in lower decomposition

of the reactants. TPR and Mössbauer measurements after

the catalytic test for the 0.5CuFeAl catalyst pointed out the

role of the reduced species (Cu¹⁺/Cu⁰and Fe²⁺) in the hydro-

gen transfer reaction, which resulted in an increase of the

isopropanol selectivity, to the detriment of the condensation

products.

Fig. 13 Room-temperature Mössbauer spectra (300 K) of spent

catalysts in the reaction carried out at 350 °C and an Et/Ac molar ratio

of 2.

Table 6 Hyperfine parameters of the Mössbauer spectra for the spent

mono- and bimetallic catalysts recorded at room temperature (300 K).

Reaction conditions: T = 350 °C, Et/Ac = 2

Samples

10FeAl

IS, mm s⁻¹

QS, mm s⁻¹

R.A.%

Fe species

0.298

0.722

1.053

0.210

0.212

0.789

0.226

0.230

0.727

0.96

2.32

2.35

0.82

1.31

2.12

0.80

1.28

2.26

66.2

24.2

9.6

58.7

33.0

8.2

47.8

32.2

19.9

Fe³⁺

Fe²⁺

Fe³⁺

Fe²⁺

Fe³⁺

Fe²⁺

0.5CuFeAl

1CuFeAl

Conflicts of interest

IS = isomer shift, QS = quadrupole shift, R.A. = relative spectral area.

There are no conflicts to declare.

(Table 6). The non-linear correlation may be due to the lower

iron oxide dispersion of the sample 1CuFeAl suggested by the

Mössbauer spectra, which present a doublet with a lower QS

(0.80) and an RA of 47.8%. This suggests a higher iron oxide

fraction with a superior symmetric environment,^64,65which is

consistent with a higher particle diameter, despite the fact

that the XRD patterns (Fig. 1) do not present any diffraction

peak from the Fe₂O₃phase. It is important to note that this

finding is consistent with the TPR results (Fig. 3 and 12).

Therefore, the decrease in catalytic activity observed in

Fig. 9 and 10 may be due to several factors. The deposition of

carbonaceous materials may cause catalyst deactivation

(Fig. 12). Similarly, the sintering or segregation phenomenon

during the catalytic reaction has an effect on the active sur-

face area. Additionally, the partial reduction of the active

phase during the catalytic process, as pointed out by

Mössbauer spectroscopy (Fig. 13), should contribute to the

Acknowledgements

The authors acknowledge the “Federal University of Ceará”,

the Brazilian research-funding support agency CNPq (427684/

2016-1), UFC/CT-INFRA/MCTI-SISNANO/CAPES, Dr. J. M.

Sasaki (X-ray laboratory) and Central Analytical of UFC for

SEM-EDX analyses.

References

1 A. Galadima and O. Muraza, J. Taiwan Inst. Chem. Eng.,

2016, 67, 29–44.

2 P. S. Nigam and A. Singh, Prog. Energy Combust. Sci.,

2011, 37, 52–68.

3 J. Sun and Y. Wang, ACS Catal., 2014, 4, 1078–1090.

4 J. M. R. Gallo, J. M. C. Bueno and U. Schuchardt, J. Braz.

Chem. Soc., 2014, 25, 2229–2243.

Catal. Sci. Technol.

View Article Online

Catalysis Science & Technology

Paper

5 A. J. Vizcaíno, A. Carrero and J. A. Calles, Int. J. Hydrogen

Energy, 2007, 32, 1450–1461.

30 E. P. Barrett, L. G. Joyner and P. P. Halenda, J. Am. Chem.

Soc., 1951, 73, 373–380.

6 D. Z. Mezalira, L. D. Probst, S. Pronier, Y. Batonneau and C.

Batiot-Dupeyrat, J. Mol. Catal. A: Chem., 2011, 340, 15–23.

7 R. M. Rioux and M. A. Vannice, J. Catal., 2003, 216,

362–376.

31 W. D. Harkins and G. Jura, J. Am. Chem. Soc., 1944, 66, 1362–1366.

32 I. V. Murin, V. M. Smirnov, G. P. Voronkov, V. G. Semenov,

V. G. Povarov and B. M. Sinel'nikov, Solid State Ionics,

2000, 133, 203–210.

8 T. K. Phung, L. P. Hernández, A. Lagazzo and G. Busca, Appl.

Catal., A, 2015, 493, 77–89.

33 M. Liu, H. Li, L. Xiao, W. Yu, Y. Lu and Z. Zhao, J. Magn.

Magn. Mater., 2005, 294, 294–297.

9 W. Alharbi, E. Brown, E. F. Kozhevnikova and I. V.

Kozhevnikov, J. Catal., 2014, 319, 174–181.

34 I. Nedkov, T. Merodiiska, L. Milenova and T. Koutzarova,

J. Magn. Magn. Mater., 2000, 211, 296–300.

10 R. C. R. Santos, A. N. Pinheiro, E. R. Leite, V. N. Freire, E.

Longhinotti and A. Valentini, Mater. Chem. Phys., 2015, 160,

119–130.

11 B. Chen, J. Lu, L. Wu and Z. Chao, Chin. J. Catal., 2016, 37,

1941–1948.

12 T. P. Braga, A. N. Pinheiro, C. V. Teixeira and A. Valentini,

Appl. Catal., A, 2009, 366, 193–200.

13 L. Faba, E. Díaz and S. Ordóñez, Appl. Catal., B, 2013, 142–

143, 387–395.

35 K. Al-Dalama and A. Stanislaus, Thermochim. Acta, 2011, 520,

67–74.

36 M. Jabłońska, Catal. Commun., 2015, 70, 66–71.

37 J. L. Pinilla, R. Utrilla, R. K. Karn, I. Suelves, M. J. Lázaro, R.

Moliner, A. B. García and J. N. Rouzaud, Int. J. Hydrogen

Energy, 2011, 36, 7832–7843.

38 Z. Z. Qin, X. H. Zhou, T. M. Su, Y. X. Jiang and H. B. Ji,

Catal. Commun., 2016, 75, 78–82.

39 A. H. Fakeeha, A. A. Ibrahim, W. U. Khan, K. Seshan, R. L. Al

and A. S. Al-fatesh, Arabian J. Chem., 2016, DOI: 10.1016/j.

arabjc.2016.06.012.

14 O. Kikhtyanin, D. Kubička and J. Čejka, Catal. Today,

2015, 243, 158–162.

15 L. Wu, T. Moteki, A. A. Gokhale, D. W. Flaherty and F. D.

Toste, Chem, 2016, 1, 32–58.

16 Z. D. Young, S. Hanspal and R. J. Davis, ACS Catal., 2016, 6,

3193–3202.

17 H. V. Fajardo, L. F. D. Probst, A. Valentini, N. L. V. Carreño,

A. P. Maciel, E. R. Leite and E. Longo, J. Braz. Chem. Soc.,

2005, 16, 607–613.

40 M. Thommes, K. Kaneko, A. V. Neimark, J. P. Olivier, F.

Rodriguez-Reinoso, J. Rouquerol and K. S. W. Sing, Pure

Appl. Chem., 2015, 87, 1051–1069.

41 T. K. Phung, A. Lagazzo, M. Á. Rivero Crespo, V. Sánchez

Escribano and G. Busca, J. Catal., 2014, 311, 102–113.

42 S. Storck, H. Bretinger and W. F. Maier, Appl. Catal., A,

1998, 174, 137–146.

18 G. Onyestyák, G. Novodárszki, Á. Farkas Wellisch and A.

Pilbáth, Catal. Sci. Technol., 2016, 6, 4516–4524.

19 V. A. Ivanov, J. Bachelier, F. Audry and J. C. Lavalley, J. Mol.

Catal., 1994, 91, 45–59.

43 G. Li, Y. Liu, Z. Tang and C. Liu, Appl. Catal., A, 2012, 437–

438, 79–89.

44 S. Zhu, P. R. F. Bell and P. F. Greenfield, Fuel, 1988, 67,

1316–1320.

20 T. P. Braga, A. N. Pinheiro, E. R. Leite, R. C. R. dos Santos

and A. Valentini, Chin. J. Catal., 2015, 36, 712–720.

21 D. Scholz, C. Aellig and I. Hermans, ChemSusChem, 2014, 7,

268–275.

22 A. Konaka, T. Tago, T. Yoshikawa, A. Nakamura and T.

Masuda, Appl. Catal., B, 2014, 146, 267–273.

23 J. Sun, R. A. L. Baylon, C. Liu, D. Mei, K. J. Martin, P.

Venkitasubramanian and Y. Wang, J. Am. Chem. Soc.,

2016, 138, 507–517.

45 H. Bouyarmane, S. El Asri, A. Rami, C. Roux, M. A. Mahly, A.

Saoiabi, T. Coradin and A. Laghzizil, J. Hazard. Mater.,

2010, 181, 736–741.

46 L. del R. Silva-Calpa, P. C. Zonetti, D. C. de Oliveira, R. R. de

Avillez and L. G. Appel, Catal. Today, 2017, 289, 264–272.

47 H. Zhan, S. Huang, Y. Li, J. Lv, S. Wang and X. Ma, Catal.

Sci. Technol., 2015, 5, 4378–4389.

48 Z. Fu, J. Wang, N. Zhang, Y. An and Z. Yang, Int. J. Hydrogen

Energy, 2015, 40, 2193–2198.

24 R. C. R. Santos, D. M. V. Braga, A. N. Pinheiro, E. R. Leite,

V. N. Freire, E. Longhinotti and A. Valentini, Catal. Sci.

Technol., 2016, 6, 4986–5002.

25 T. P. Braga, N. Essayem and A. Valentini, RSC Adv., 2015, 5,

93394–93402.

26 X. Lin, R. Li, Y. Zhang, Y. Zhan, C. Chen, Q. Zheng and J.

Ma, Int. J. Hydrogen Energy, 2015, 40, 1735–1741.

27 Z. Bao, K. Xiao, X. Qi, X. Wang, L. Zhong, K. Fang, M. Lin

and Y. Sun, J. Energy Chem., 2013, 22, 107–113.

28 A. Valentini, N. L. V. Carreño, L. F. D. Probst, A. Barison,

A. G. Ferreira, E. R. Leite and E. Longo, Appl. Catal., A,

2006, 310, 174–182.

49 W. Yu, J. Zhu, L. Qi, C. Sun, F. Gao, L. Dong and Y. Chen,

J. Colloid Interface Sci., 2011, 364, 435–442.

50 W. Peng, J. Li, B. Chen, N. Wang, G. Luo and F. Wei, Catal.

Commun., 2016, 74, 39–42.

51 J. I. Di Cosimo, V. K. Díez, M. Xu, E. Iglesia and C. R.

Apesteguía, J. Catal., 1998, 178, 499–510.

52 T. Wang, L. Qi, H. Lu and M. Ji, Biochem. Pharmacol.,

2017, 17, 162–169.

53 F. C. Meunier, J. Scalbert and F. Thibault-Starzyk, C. R.

Chim., 2015, 18, 345–350.

54 K. K. Ramasamy, M. Gray, H. Job, C. Smith and Y. Wang,

Catal. Today, 2016, 269, 82–87.

29 T. P. Braga, B. M. C. Sales, A. N. Pinheiro, W. T. Herrera, E.

Baggio-Saitovitch and A. Valentini, Catal. Sci. Technol.,

2011, 1, 1383–1392.

55 Y. Liu, H. Tüysüz, C.-J. Jia, M. Schwickardi, R. Rinaldi, A.-H.

Lu, W. Schmidt and F. Schüth, Chem. Commun., 2010, 46,

1238–1240.

Catal. Sci. Technol.

View Article Online

Paper

Catalysis Science & Technology

56 P. Lauriol-garbay, J. M. M. Millet, S. Loridant, V. Bellière-

baca and P. Rey, J. Catal., 2011, 280, 68–76.

57 S. E. S. E. A. El-Sharkawy and S. A. El-HaKam, Mater. Lett.,

2000, 42, 331–338.

58 B. A. Raich and H. C. Foley, Ind. Eng. Chem. Res., 1998, 37,

3888–3895.

59 D. Gabriels, W. Y. Hernandez, B. Sels, P. Van Der Voort and

A. Verberckmoes, Catal. Sci. Technol., 2015, 5, 3876–3902.

60 E. Iglesia, D. G. Barton, J. A. Biscardi, M. J. L. Gines and

S. L. Soled, Catal. Today, 1997, 38, 339–360.

61 T. P. Braga, R. C. R. Santos, B. M. C. Sales, B. R. da Silva,

A. N. Pinheiro, E. R. Leite and A. Valentini, Chin. J. Catal.,

2014, 35, 514–523.

62 G. Fierro, M. Lojacono, M. Inversi, P. Porta, R. Lavecchia

and F. Cioci, J. Catal., 1994, 148, 709–721.

63 J. Y. Kim, J. A. Rodriguez, J. C. Hanson, A. I. Frenkel and

P. L. Lee, J. Am. Chem. Soc., 2003, 125, 10684–10692.

64 B. Ganguly, F. E. Huggins, K. R. P. M. Rao and G. P.

Huffman, J. Catal., 1993, 142, 552–560.

65 M. Siddique and M. Riaz, Phys. B, 2010, 405, 2238–2240.

Catal. Sci. Technol.

Article Doi

Copper promoter effect on acid-base and redox sites of Fe/Al2O3 catalysts and their role in ethanol-acetone mixture conversion

DOI: 10.1039/c7cy01640a

Source and publish data:

Authors:

Article abstract of DOI:10.1039/c7cy01640a

Full text of DOI:10.1039/c7cy01640a

Products guided by the article

R&D Labs maybe for 123-72-8

Relevant to this article

Hot Product