Synthesis of C4 and C8 Chemicals from Ethanol on MgO-Incorporated Faujasite Catalysts with Balanced Confinement Effects and Basicity

Article abstract of DOI:10.1002/cssc.201501518

A new type of catalyst has been designed to adjust the basicity and level of molecular confinement of KNaX faujasites by controlled incorporation of Mg through ion exchange and precipitation of extraframework MgO clusters at varying loadings. The catalytic performance of these catalysts was compared in the conversion of C₂ and C₄ aldehydes to value-added products. The product distribution depends on both the level of acetaldehyde conversion and the fraction of magnesium as extraframework species. These species form rather uniform and highly dispersed nanostructures that resemble nanopetals. Specifically, the sample containing Mg only in the form of exchangeable Mg²⁺ ions has much lower activity than those in which a significant fraction of Mg exists as extraframework MgO. Both the (C₆+C₈)/C₄ and C₈/C₆ ratios increase with additional extraframework Mg at high acetaldehyde conversion levels. These differences in product distribution can be attributed to 1) higher basicity density on the samples with extraframework species, and 2) enhanced confinement inside the zeolite cages in the presence of these species. Additionally, the formation of linear or aromatic C₈ aldehyde compounds depends on the position on the crotonaldehyde molecule from which abstraction of a proton occurs. In addition, catalysts with different confinement effects result in different C₈ products.

Full text of DOI:10.1002/cssc.201501518

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
DOI: 10.1002/cssc.201501518
Full Papers
Synthesis of C4 and C8 Chemicals from Ethanol on MgO-
Incorporated Faujasite Catalysts with Balanced
Confinement Effects and Basicity
Lu Zhang,[a] Tu N. Pham,[a] Jimmy Faria,[b] Daniel Santhanaraj,[a] Tawan Sooknoi,[a]
Qiaohua Tan,[a] Zheng Zhao,[a] and Daniel E. Resasco*[a]
A new type of catalyst has been designed to adjust the basici-
ty and level of molecular confinement of KNaX faujasites by
controlled incorporation of Mg through ion exchange and pre-
cipitation of extraframework MgO clusters at varying loadings.
The catalytic performance of these catalysts was compared in
the conversion of C2 and C4 aldehydes to value-added prod-
ucts. The product distribution depends on both the level of
acetaldehyde conversion and the fraction of magnesium as ex-
traframework species. These species form rather uniform and
highly dispersed nanostructures that resemble nanopetals.
Specifically, the sample containing Mg only in the form of ex-
changeable Mg2+ ions has much lower activity than those in
which a significant fraction of Mg exists as extraframework
MgO. Both the (C6 +C8)/C4 and C8/C6 ratios increase with addi-
tional extraframework Mg at high acetaldehyde conversion
levels. These differences in product distribution can be attrib-
uted to 1) higher basicity density on the samples with extrafra-
mework species, and 2) enhanced confinement inside the zeo-
lite cages in the presence of these species. Additionally, the
formation of linear or aromatic C8 aldehyde compounds de-
pends on the position on the crotonaldehyde molecule from
which abstraction of a proton occurs. In addition, catalysts
with different confinement effects result in different C8
products.
Introduction
Although biorefineries can produce bioethanol on a large
scale,[1] the low return on investment, which is characteristic of
commodities, challenges their economic profitability and ulti-
mately the widespread use of bioethanol as a transportation
fuel. Therefore, chemical conversion of bioethanol into more
valuable commodity chemicals is a highly desirable option for
its utilization.[2] The integration of energy and chemical produc-
tion in biorefineries by combining existing fermentation facili-
ties and chemical industrial production seems to be a crucial
aspect in the development of the technology.[3]
their selectivity to C4 products to minimize excessive conden-
sation to ꢀC6 products. In this scenario, zeolites appear to be
potentially attractive catalysts to maximize the desirable prod-
uct due to their enhanced shape and size selectivity. Adjusta-
ble molecular confinement when using zeolites as catalysts is
a unique tool offered by zeolites to enhance selectivity. Previ-
ously, we investigated the aldol condensation of acetaldehyde
over faujasite zeolites with varying silicon/aluminum ratios and
different types of alkali-metal cations.[21] We found that the C4/
(C6 +C8) product ratio depended on a variety of factors, such
as strength, density, and accessibility of basic sites. It was also
shown that base-catalyzed condensation (C4 and ꢀC6 prod-
ucts) dominated at temperatures around 2308C, whereas acid-
catalyzed acetalization prevailed below 1808C. Because high
reaction temperatures may lead to side products and faster de-
activation, we investigate herein the modification of faujasite
zeolites to improve their activity at low temperatures.
For example, the conversion of ethanol to crotonaldehyde is
a reaction that has been investigated extensively.[4–8] Crotonal-
dehyde is a highly reactive a,b-unsaturated aldehyde that can
be used to produce the industrially valuable crotyl alcohol
through chemoselective hydrogenation.[3] The two main steps
in the conversion of ethanol into crotonaldehyde include dehy-
drogenation to acetaldehyde and aldol condensation to the C4
product. The latter step is conducted on either acid or basic
catalysts.[9–20] One of the key requirements of these catalysts is
Due to its high basicity, MgO is an effective basic catalyst for
aldol condensation, but, in general, it tends to generate exces-
sive over-condensation products. For instance, acetaldehyde
on MgO results in low C4/(C6 +C8) product ratios as the conver-
sion increases. Therefore, herein, we have investigated novel
compositions of MgO-modified faujasites, which exhibit the
desirable combination of the required basic properties of MgO
for the aldol reaction with the confinement benefits of the fau-
jasite zeolites, which inhibit excessive condensation. This work
is motivated by numerous previous investigations that have
demonstrated the incorporation of metal or metal oxide clus-
ters inside zeolite cavities.[22–32] Herein, we have used several
[a] Dr. L. Zhang, Dr. T. N. Pham, Dr. D. Santhanaraj, Prof. T. Sooknoi,
Dr. Q. Tan, Z. Zhao, Prof. D. E. Resasco
School of Chemical, Biological, and Materials Engineering
University of Oklahoma, Norman, OK 73019 (USA)
E-mail: resasco@ou.edu
[b] Dr. J. Faria
Abengoa Research, C/Energía Solar no. 1
Palmas Altas, Seville 41014 (Spain)
Supporting Information for this article can be found under http://
dx.doi.org/10.1002/cssc.201501518.
ChemSusChem 2016, 9, 736 – 748
736
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Full Papers
MgO-modified faujasites as catalysts in acetaldehyde/ethanol
mixtures and characterized them with a variety of characteriza-
tion techniques.
sulting nominal charge ratio was 1.14, which indicated the
presence of extraframework MgO in the sample (MgO-KNaX-
1.14). Likewise, when a 1m solution of Mg2+ was used in the
ion-exchange procedure, the nominal charge ratio further in-
creased to 1.32; an indication of a larger amount of excess
MgO in this sample (MgO-KNaX-1.32). These numbers might
slightly deviate if some of the exchanged Mg ions are in the
form of Mg(OH)+. However, because the sample is calcined
after the incorporation of Mg, any Mg in the form of hydrox-
ides (Mg(OH)+ or Mg(OH)2) is probably converted into Mg2+ or
MgO.
Previously,[21] we described how the silicon/aluminum ratio
and type of alkali cation affected the selectivity to C4 products.
Herein, we additionally analyze how to tune the selectivity of
different C8 products when using the C4 product crotonalde-
hyde as a feed. The C8 products are mainly 2,4,6-octatrienal
and tolualdehyde isomers. Both can have commercial interest.
For instance, 2,4,6-octatrienal is a key intermediate for the syn-
thesis of 2,4,6-octatrienoic acid and recent reports indicate
that 2,4,6-octatrienoic acid is an effective UV radiation blocker
that can be used in sunscreen products.[33] Likewise, tolualde-
hyde isomers are versatile starting materials for the synthesis
of phthalic and terephthalic acids, which are used as inter-
mediates in the production of phthalic anhydride and polyes-
ter polyethylene terephthalate (PET).[34] Only a few studies in
the literature describe the synthesis of octatrienal and tolual-
dehyde by aldol condensation. We have found that excessive
condensation to >C8 products can be inhibited, if the corre-
sponding C8 alcohols are formed through the Meerwein–Pon-
dorf–Verley (MPV) hydrogen-transfer reaction. Therefore, in this
part of the study, isopropanol was chosen as the solvent be-
cause it was a better hydrogen donor than ethanol.
Characterization of zeolites MgO-KNaX
SEM images of the KNaX and MgO-KNaX series of zeolites are
shown in Figure 1. As expected from the charge analysis de-
scribed above, the MgO-KNaX-0.99 sample with no excess
MgO shows an overall surface morphology similar to that of
KNaX, which exhibits only a somewhat rougher external sur-
face than that of the parent zeolite. As shown below, XRD and
NMR spectroscopy results indicate that the zeolite structure of
this sample (MgO-KNaX-0.99) remains almost intact. In con-
trast, the MgO-KNaX-1.14 sample, with excess MgO, appears
very different (Figure 1c). The surface of the sample is covered
Results and Discussion
Table 1 shows the conditions used to incorporate MgO into
the zeolite samples, as well as the contents of MgO in each
zeolite. The concentration of Mg2+ in the solution was adjust-
ed to obtain the desired amount of MgO, which was measured
by inductively coupled plasma optical emission spectroscopy
(ICP-OES), along with those of the other elements in the zeolite
(Si, Al, K, Na). As a result, the nominal positive/negative charge
ratio could be calculated as (K+, Na+, 2Mg2+)/Al to determine
the fraction of Mg as exchangeable Mg2+ or MgO. Because
this charge ratio should be unity, any number larger than 1.0
indicates that a fraction of Mg is in the form of excess MgO.
For example, when using a 0.01m solution of Mg2+, the result-
ing charge ratio was 0.99, which indicated that in this sample
(MgO-KNaX-0.99) it was possible that most of the Mg was in
the form of exchangeable Mg2+ ions, with no excess MgO. In
contrast, when the exchange solution was 0.1m Mg2+, the re-
Figure 1. SEM images of a) KNaX, b) MgO-KNaX-0.99, c) MgO-KNaX-1.14, and
d) MgO-KNaX-1.32.
Table 1. Preparation and resulting composition of MgO-incorporated zeolite catalysts.
Catalyst[a]
Concentration of Mg2+
for ion exchange (M)
Alkaline wash
Mg [wt%]
Si/Al
K+/Na+
Mg2+/Na+
(K+, Na+)/Al
(K+, Na+, 2Mg2+)/Al
KNaX
N/A[b]
0.01
N/A[b]
KOH
0.0
1.3
3.1
3.6
3.6
4.2
1.09
1.09
1.06
1.04
1.03
1.02
0.32
0.49
1.40
1.22
1.17
31.5
0.0
0.21
0.96
0.74
0.72
17.4
0.99
0.72
0.64
0.66
0.67
0.64
0.99
0.99
1.14
1.10
1.10
1.32
MgO-KNaX-0.99
MgO-KNaX-1.14
1st cycle catalyst[c]
2nd cycle catalyst[d]
MgO-KNaX-1.32
0.10
KOH
N/A[b]
N/A[b]
1.0
N/A[b]
N/A[b]
KOH
[a] The name code of the MgO-incorporated zeolite catalysts is based on the total positive to negative charge ratio indicated in the last column. [b] N/A=
not applicable. [c] The first cycle catalyst was obtained by washing the used MgO-KNaX-1.14 catalyst with acetone followed by drying in an oven. [d] The
second cycle catalyst was obtained by washing the used first cycle catalyst after the second cycle test with acetone followed by drying in an oven.
ChemSusChem 2016, 9, 736 – 748
www.chemsuschem.org
737
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Full Papers
by a high density of nanostructures, which resemble nanope-
tals, as recently named by Koh et al.[35] Similarly, the MgO-
KNaX-1.32 sample with even higher excess MgO also has
a high density of nanopetals, but this sample also exhibits
other plate-like structures (Figure 1d). As shown below, this is
an indication of MgO nanoparticles coating the surface.
a shift in the position of the signal. The sample with higher Mg
content (MgO-KNaX-1.32) showed a slight shift of the tetrahe-
dral 27Al signal from d=62 to 60 ppm, as well as the appear-
ance of a new signal at d=9 ppm, which could be assigned to
amorphous alumina.[37] In parallel, the corresponding 29Si NMR
spectrum for the MgO-KNaX-1.32 sample showed a dominant
signal at d=À88 ppm, instead of the strong one at d=
À85 ppm coupled with a weak one at d=À89 ppm observed
in the spectra of both MgO-KNaX-0.99 and À1.14. These two
29Si signals (d=À85 and À88/À89 ppm) are due to Si(OAl)4
and Si(OAl)3(OSi)1 species, respectively.[41] This behavior is in
agreement with the changes observed in the 27Al spectra and
indicates a minor modification of the zeolite structure in the
MgO-KNaX-1.32 sample due to the high concentration of the
alkaline solution (1m) used in the preparation of this sample.
As observed in Table 1, the Si/Al ratio in MgO-KNaX-1.32 is
1.02, which is very close to the original value of 1.09 in the
parent KNaX, which indicates that any structural modification
by KOH is insignificant compared with those observed in con-
ventional desilication treatments. Therefore, the intensity drop
observed in the XRD pattern of this sample (Figure 2) is
a result of two factors, namely, coating of the external surface
by MgO nanostructures and structural modification during
preparation. We describe our SEM/TEM/energy-dispersive X-ray
spectroscopy (EDX) results for a sliced specimen of the MgO-
KNaX-1.32 sample prepared by a focused ion beam (FIB) in the
Supporting Information. This technique uses a fine FIB to
create very precise cross sections of a sample for subsequent
imaging by SEM and TEM. The FIB allows thin slices of the zeo-
lite sample to be produced; these can be scanned in different
zones to determine whether there is an accumulation of mag-
nesium on the external surface of the crystallites. Although the
resolution of EDX is not enough to probe the nature of mag-
nesium species inside and outside the zeolite, it is clear that
magnesium is in both the internal and external regions. Nota-
bly, in the synthetic procedure used, the original KNaX zeolite
was first exchanged with Mg2+ ions. Because K+ and Na+ ions
are located inside the supercages of the parent KNaX sample,
the Mg2+ ions introduced by the ion-exchange method should
also be inside the pores. In the subsequent step, a KOH wash
is conducted to precipitate Mg2+ as Mg(OH)2, which is then
calcined to form MgO. At this point, nanopetals are formed on
the external surface, as clearly shown by SEM. The FIB/SEM/
EDX results show that the region of the external surface con-
tains all cations (Mg, Si, and Al) and, although there are regions
The XRD patterns and NMR spectra were obtained to exam-
ine structural changes of the zeolite framework. The results are
illustrated in Figures 2 and 3 (see below), respectively. As
shown in Figure 2, only diffraction peaks due to the faujasite
Figure 2. XRD patterns of zeolite samples and commercial MgO.
zeolite are observed for all MgO-KNaX samples, with no diffrac-
tion peaks that could be ascribed to MgO. This result suggests
that the MgO particles are either too small in size or in concen-
tration to be detected by XRD, which is expected. As shown in
Table 1, none of the three MgO-modified samples contain
more than 4 wt% of Mg. The 27Al and 29Si spectra for MgO-
KNaX-0.99 agree with those of KNaX reported in Ref. [36],
which again demonstrates that MgO-KNaX-0.99 retains an un-
changed zeolite structure. Interestingly, the intensity of all dif-
fraction peaks decreased as the MgO content increased from
1.14 to 1.32.
The XRD intensity drop for the MgO-KNaX-1.14 is likely to be
due to the presence of MgO nanopetals coating the zeolite,
since no shift in the NMR signals or appearance of new signals
was detected (Figure 3). Any structural collapse would gener-
ate a change in the octa-/tetrahedral signal ratio or at least
Figure 3. a) 27Al and b) 29Si MAS NMR spectra of zeolite samples.
ChemSusChem 2016, 9, 736 – 748
www.chemsuschem.org
738
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Full Papers
that are enriched in Mg, it is not possible to determine wheth-
er the observed nanopetal features are made entirely of MgO
or contain some Si or Al. Because Mg2+ ions were initially
inside the pores, it is reasonable to believe that part of the
MgO species are also located inside the pores.
The growth of Mg(OH)2 nanostructures on zeolite A has
been previously reported.[35,38] For example, Koh et al. found
that the precipitation of petal-like nanostructures was con-
trolled by the pH of the solution.[35] If the pH is above the iso-
electric point of Mg(OH)2, the Mg(OH)2 nuclei are negatively
charged and attract Na+ ions to decrease the surface energy.
Irregularly shaped nanoparticles are formed on the surface. On
the other hand, with lower pH, Mg(OH)2 nanopetals are ob-
served. Herein, different morphology observed for the MgO-
KNaX samples was controlled by the concentration of the solu-
tion of KOH.
The nitrogen adsorption–desorption isotherms and pore size
distribution curves are shown in Figure S1 in the Supporting
Information. The porosity characteristics of the zeolite samples
are shown in Table 2. The values of the pore volumes for the
Figure 4. TPD profiles of CO2 over zeolite catalysts and commercial MgO.
Table 2. Porosity characteristics of the zeolite catalysts.
sorption peaks start emerging at higher temperatures. On
MgO-KNaX-1.14, in addition to the low-temperature desorption
peak described above, two new ones centered at about 250
and 3258C become apparent. As the MgO loading increases
for sample MgO-KNaX-1.32, these peaks become more promi-
nent. They are clearly correlated with a higher basicity due to
MgO incorporated into the sample. Based on the correspond-
ing areas for this sample, the one centered at 1008C represents
51% of the total desorbed CO2. Pure MgO displayed two clear
desorption peaks at temperatures above 3508C, which sug-
gested that the Mg-modified faujasites displayed moderate ba-
sicity that could be controlled by the Mg loading. This poten-
tial for tailoring the basic properties of the zeolite opens up in-
teresting possibilities for base-catalyzed reactions.
Catalyst
Vmicro
[cm3 gÀ1
Vmeso
[cm3 gÀ1
]
Vmeso/Vmicro
]
KNaX
0.288
0.212
0.217
0.069
0.058
0.237
0.049
0.065
0.068
0.087
0.102
0.072
0.170
0.149
0.313
1.261
1.759
0.378
MgO-KNaX-0.99
MgO-KNaX-1.14
1st cycle catalyst[a]
2nd cycle catalyst[b]
MgO-KNaX-1.32
[a] First cycle catalyst was obtained by washing the used MgO-KNaX-1.14
catalyst with acetone followed by drying in an oven. [b] Second cycle cat-
alyst was obtained by washing the used first cycle catalyst after the
second cycle test with acetone followed by drying in an oven.
parent KNaX zeolite demonstrate its microporous characteris-
tics, along with the presence of a small extent of mesopores.
In comparison, all three MgO-modified zeolite samples exhibit
smaller Vmicro values together with some enhanced mesoporosi-
ty in the 5–15 nm range. It is clear that the KOH washes con-
ducted on these samples cause a small extent of framework
dissolution and generation of mesopores. The presence of
a larger number of K+ ions would also account for the ob-
served decrease in Vmicro (see the K+/Na+ ratio in Table 1). In-
terestingly, MgO-KNaX-0.99 presents a similar Vmeso/Vmicro ratio
to that of KNaX, whereas the Vmeso/Vmicro ratios in MgO-KNaX-
1.14 or -1.32 are almost twice as large as that of KNaX. The dif-
ference is ascribed to the MgO nanostructures created on the
surface of the Mg-modified samples. Similarly, Lydon et al.
found an increased external surface area due to Mg(OH)2 nano-
structures extending out from the surface of zeolite A.[38]
Production of C4 compounds from acetaldehyde
Table 3 shows the catalytic performance of the different cata-
lysts investigated for the reaction of acetaldehyde in ethanol.
In all runs, the volume ratio of acetaldehyde/ethanol was kept
at 1/10, by incorporating acetaldehyde (2 mL, 71.6 mmol of
total C atoms) into ethanol (20 mL). The reaction was carried
out at 1808C for 3 h. As mentioned above, the activity of each
sample was evaluated in terms of the amount of products ob-
tained rather than the disappearance of acetaldehyde, which
was highly volatile and could be additionally produced from
the conversion of ethanol solvent. Because acid-catalyzed reac-
tions are reversible, only condensation products are listed for
comparison of activity.
Under the test conditions, only C4 products formed over
KNaX, but the total yields produced were only 1.5 mmol. At
the other extreme, MgO with a higher strength and density
basic sites, yielded about 48 mmol of condensation products.
In this case, crotyl alcohol was the dominant product, as
a result of a high MPV activity of MgO. However, the selectivity
The strength and density of basic sites were assessed by
temperature programmed desorption (TPD) of adsorbed CO2.
As shown in Figure 4, the KNaX and MgO-KNaX-0.99 samples
have similar TPD profiles, with a main desorption centered at
about 1008C. However, as the Mg loading increases, clear de-
ChemSusChem 2016, 9, 736 – 748
www.chemsuschem.org
739
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim