Formation of 1,3-butadiene from ethanol in a two-step process using modified zeolite-β catalysts

Article abstract of DOI:10.1016/j.apcata.2016.01.026

Catalysts for the selective formation of 1,3-butadiene out of ethanol in a continuous two-stage gas-phase fixed bed reactor setup were developed. In the first stage, 20% acetaldehyde yield was aspired. This mixture was directly fed into the second stage facilitating enhanced selectivity to 1,3-butadiene. In the first stage, AgSiO₂ and CuSiO₂ were studied. Especially the copper based catalyst proved to be a selective and long-term stable system. In the second stage we tested different zeolite-β based catalyst systems which were modified in terms of their acidic and basic behaviour. We were able to reach 1,3-butadiene selectivities up to 72% at rather low temperatures of 573 K. Furthermore we could draw a direct correlation between the catalyst's acidic and basic properties and their catalytic performance. The physical properties were analysed via X-ray diffraction, nitrogen physisorption, TGA-DSC and HRTEM. The chemical composition was determined by ICP-AES and SEM-EDX measurements. The investigation of the acidic and basic properties was realized by pyridine adsorption infrared spectroscopy and temperature programmed desorption with ammonia and carbon dioxide, respectively.

Full text of DOI:10.1016/j.apcata.2016.01.026

Accepted Manuscript
Title: Formation of 1,3-butadiene from ethanol in a two-step
process using modified zeolite-␤ catalysts
Author: Alexander Klein Kristina Keisers Regina Palkovits
PII:
S0926-860X(16)30026-6
DOI:
Reference:
http://dx.doi.org/doi:10.1016/j.apcata.2016.01.026
APCATA 15741
To appear in:
Applied Catalysis A: General
Received date:
Revised date:
Accepted date:
18-11-2015
20-1-2016
21-1-2016
Please cite this article as: Alexander Klein, Kristina Keisers, Regina
Palkovits, Formation of 1,3-butadiene from ethanol in a two-step process
using modified zeolite-rmbeta catalysts, Applied Catalysis A, General
http://dx.doi.org/10.1016/j.apcata.2016.01.026
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

Formation of 1,3-butadiene from ethanol in a two-step process using modified zeolite-β catalysts
Alexander Klein, Kristina Keisers and Regina Palkovits*
Institut für Technische und Makromolekulare Chemie, RWTH Aachen University,
Worringerweg 2, D-52074 Aachen (Germany)
Phone: +49-(0)241-8026497
email: palkovits@itmc.rwth-aachen.de

Graphical abstract
Highlights:
-
Cu\SiO2 presents an active and long term stable catalyst for dehydrogenation of ethanol to
acetaldehyde
-
-
Selectivity towards 1,3-butadiene increases with increasing effective acidity
Selectivity towards the dehydration products ethylene and diethyl ether and ethanol
conversion decreases with rising acidity.

Abstract
Catalysts for the selective formation of 1,3-butadiene out of ethanol in a continuous two-stage
gas-phase fixed bed reactor setup were developed. In the first stage, 20% acetaldehyde yield was
aspired. This mixture was directly fed into the second stage facilitating enhanced selectivity to
1,3-butadiene. In the first stage, Ag\SiO2 and Cu\SiO2 were studied. Especially the copper based
catalyst proved to be a selective and long-term stable system. In the second stage we tested
different zeolite-β based catalyst systems which were modified in terms of their acidic and basic
behaviour. We were able to reach 1,3-butadiene selectivities up to 72% at rather low temperatures
of 573 K. Furthermore we could draw a direct correlation between the catalyst’s acidic and basic
properties and their catalytic performance. The physical properties were analysed via X-ray
diffraction, nitrogen physisorption, TGA-DSC and HRTEM. The chemical composition was determined
by ICP-AES and SEM-EDX measurements. The investigation of the acidic and basic properties was
realized by pyridine adsorption infrared spectroscopy and temperature programmed desorption with
ammonia and carbon dioxide, respectively.
Keywords: butadiene; ethanol; catalysis; heterogeneous; zeolite; ion exchange; acidity; basicity
Introduction
Since the depletion of fossil feedstock such as oil and natural gas is a known truth, there is not only a
need finding bio-based fuels, but also ways to produce chemicals from renewable resources. Due to
the increasing use of bioethanol as a biofuel in the last decade, the production volume increased as
well.[1] As a consequence the ethanol price dropped. Because of political efforts such as E10 in
Europe, a further growth in the next years is expected. Therefore, ethanol will be available as a low-
cost feedstock which cannot only be used as a fuel, but also as a promising platform chemical to
produce different bulk chemicals.[2] One important bulk chemical, 1,3-butadiene, which is mainly
used as a monomer in polymerization can be produced from ethanol.[3] Traditionally 1,3-butadiene
is produced as a side-product in the steam cracking process.[4] Due to the hydraulic fracturing boom
in the USA more and more shale gas is used as a starting material in this process. As a result the ratio
of ethylene to butadiene increased and the industry faces a lack of 1,3-butadiene.[5] Producing
1,3-butadiene out of ethanol seems to be an interesting option.
In the first part of the last century two processes, the Ostromislensky[6] and the Lebedev
process,[7,8] for the production of 1,3-butadiene from ethanol were developed. The main difference
in these two methods is, that in the Lebedev process ethanol is converted in one step while
Ostromislensky split the reaction in two parts. First ethanol is dehydrogenated to acetaldehyde,
which reacts in the second part with co-added ethanol to 1,3-butadiene. The generally accepted
mechanism includes a succession of dehydrogenation, dehydration, aldol condensation and
Meerwein-Ponndorf-Verley reaction steps.[9,10] Ethylene and diethyl ether formed by dehydration
of ethanol are considered to be the main side products. In line, the catalyst’s acidic and basic
properties, in terms of number and strength, are the crucial factors in this reaction. Besides, the
formation of other C4 components, e.g. butenes, ketones, alcohols etc., is reported. These species are

formed via several different routes, wherefore it is difficult to predict the effects of different
chemical properties.[11] Recently, a paper from Chieregato et al.[12] queried this mechanism.
Instead of an intermolecular hydrogen transfer via a Meerwein-Ponndorf-Verley mechanism, they
propose a reaction between acetaldehyde and an activated ethanol species (the corresponding
carbanion) to form crotyl alcohol.
As mentioned, two reaction routes, the Lebedev and the Ostromislensky process are known. Besides
a commercialization by Union Carbide in 1943[13] and some following patents and publications,
[9,14] where standard copper dehydrogenation catalysts were used in the first step, all subsequent
publications focus on the one-step process or directly add acetaldehyde to the ethanol feed.[11]
Various materials were tested as catalysts including hydrotalcites[15] or sepiolites[16] as well as
aluminium and silica based materials. The most promising results were achieved with silica as
support, with either non-ordered or ordered structures.[11] Several research groups tried to improve
the catalytic performance by adding a promotor, varying the preparation method or the ethanol to
acetaldehyde ratio.[17–22] Copper and silver were used to dehydrogenate ethanol to
acetaldehyde.[17, 23–25] To promote the aldol coupling and the Meerwein-Ponndorf-Verley
reaction, different metal oxides were introduced, such as Cr2O3[17,23,26,27], Ta2O5 and
ZrO2,[9,14,24,28–31] HfO2,[17,32] Co2O3[23] and especially MgO.[18–21,23,24,33,34] Jones et al. and
Lewandowski et al. described, that ZnO catalyses aldol coupling and the dehydrogenation
step.[24,28] Alkaline metal oxides were included into the material to poison a certain amount of
acidic centres.[28,34] Makshina et al.[23] and Janssens et al.[19] tested different metals and reached
with Ag up to 63% selectivity towards 1,3-butadiene. Other groups could either increase selectivity or
ethanol conversion. Exceptions represent the results of Onishi et al.[34] and more recently the
research of Baerdemaeker et al.[32]. Onishi et al. reached 87% selectivity at full conversion applying
a K2O doped MgO-SiO2 catalyst. Baerdemaeker reported 70% butadiene selectivity at almost full
conversion with a bimetallic catalyst comprehending Zn and Hf on silica. Except the two above
mentioned studies with very high yields, the formation of side products, such as ethylene, diethyl
ether, butenes and other C4 compounds, remains as the major challenge. Angelici et al.[21]
Investigated the influence of different preparation methods. They observed a significant influence of
the preparation method of the catalyst on the catalytic behaviour. In a further publication, they
correlated the preparation method and the resulting morphology and acidic and basic properties of
their catalysts with the formation of the side products ethylene and diethyl ether, respectively.[35]
Other authors could improve the catalytic performance by exchanging the support by either ordered
silica such as SBA-15, MMS and KIT-6[30] or crystalline zeolite-β structures.[36] Almost all of the
mentioned metals and supports were used with pure ethanol as well as with co-added acetaldehyde.
The ethanol to acetaldehyde ratio is of course of high importance. Different ratios were applied in
literature ranging from 2 - 4.[9, 14, 17, 18, 28, 31] In general, a higher ratio in this range leads to
higher yields and avoids coke formation. By using 100% ethanol the reaction performance drastically
deteriorates.[11]
Several studies identified the dehydrogenation of ethanol as rate determining step in this
reaction.[9,37] Additionally, superior results were observed upon addition of acetaldehyde to the
feed. Consequently, the two-step process bears a high potential. In a first step, 20% ethanol is
converted to acetaldehyde. Subsequently, the resulting ethanol/acetaldehyde mixture is fed into the
second reactor. As a result, not only the catalysts and the reaction conditions of each step can be
optimized independently, but rather the impact of the catalysts properties can be resolved for both
steps. In the first step, the dehydrogenation of ethanol to acetaldehyde, a highly selective and

especially long-term stable catalyst system is needed. Therefore we used, based on the Union
Carbide process and other publications, different dehydrogenation metals, such as copper and silver.
The main focus of this study lies on the investigation of the second step. Since it is assumed that the
acidic-basic properties are crucial in this reaction, several research groups tried to correlate these
properties with the catalytic performance. Despite a number of studies in this direction, an approach
utilizing a generalized preparation procedure to allow access to a set of materials with varying
acid/base properties has not been presented yet.[21,35]
In this study, we aspire to influence the acid/base-properties while keeping the physical properties of
the catalysts as unaltered as possible. Sushkevich et al. confirmed the potential of fully Lewis acidic
Zr-β materials to form 1,3-butadiene.[36,38] Preliminary pyr-IR measurements confirmed Lewis
acidity to be also dominant for zeolite-β, especially at higher modules[39]. Therefore, we applied
zeolite-β catalysts of different modules as supports and modified them by impregnation and ion
exchange, to introduce basic materials, such as MgO and ZnO to reduce the overall acidity. As a
result, we were able to synthesize a series of catalysts with comparable physical properties, which
differ in acidity over a broad range facilitating correlations between the acidic-basic behaviour of the
catalysts and their catalytic performance.
Experimental Section
2.1 Materials
SS61138 silica from Saint-Gobain was used as purchased without further treatment. The three
different zeolite-β structures H-β11, H-β67 and H-β280 were kindly provided by Clariant AG (referred
.
to as HCZB25, HCZB150 and HCZB650) and used as obtained. Mg(NO3)2 6H2O (99%, Sigma Aldrich),
.
.
.
Zn(NO3)2 6H2O (99%, Sigma Aldrich), Ni(NO3)2 6H2O (99%, Sigma Aldrich), Cu(NO3)2 6H2O (99%, Sigma
.
.
Aldrich), Al(NO3)3 9H2O (99%, Sigma Aldrich), Ca(NO3)2 4H2O (99%, Sigma Aldrich), CsNO3 (99%, Sigma
Aldrich), KNO3 (99%, Sigma Aldrich) and AgNO3 were used for the preparation of the different
catalysts. Ethanol (100%, Sigma) was used as starting material and benzene (99.9%, Sigma Aldrich) as
an internal standard for GC analysis. A certified gas mixture from Praxair (1 mol-% ethylene,
0.25 mol-% diethyl ether, 1 mol-% 1,3-butadiene, 0.25 mol-% acetaldehyde, 0.1 mol-% ethanol) was
used for GC calibration.
2.2 Catalyst preparation
.
Cu(NO3)2 6H2O and Ag(NO3), respectively, were supported by incipient wetness impregnation on SiO2
(Saint-Gobain, SS61138). The required amount of metal salt was diluted in 0.85 mL deionized water
and added afterwards dropwise to the support, dried at least for 8 h at 373 K and calcined at 823 K
for 5 h (10 K.min-1) under static air.
The second stage catalysts were synthesised by incipient wetness impregnation. Therefore, the
.
desired amount of Mg(NO3)2 6H2O was diluted in 0.5 mL of deionized water and added dropwise to
the support. The resulting materials were dried at 393 K for at least 8 h and calcined at 673 K for 5 h
(4 K.min-1) under static air. The second metal oxide was supported according to the same procedure.

The catalysts were labelled as MxOy(m)-MgO(n)\support, where m and n stand for the respective
metal loading with respect to the support.
Introducing different counter ions into the zeolite beta structure was done by liquid ion exchange.
H-β280 was stirred for 1 h in 20 mL.g-1 of an aqueous 0.5 M metal nitrate solution at 353 K. The
material was recovered via filtration, washed several times with deionized water and dried at 393 K.
The procedure was repeated twice and the material was finally calcined at 813 K for 5 h (1 K.min-1)
under static air. The resulting zeolite was labelled as M-β280 (M = K+, Cs+, Ca2+).
2.3 Catalyst characterization
X-ray diffraction patterns were recorded with a Siemens D5000 diffractometer using Cu-K-α with
λ = 1.542 Å, operating at 30 kV. It was measured between 2θ = 3 ° and 90 ° with a step size of
0.02 °.s-1 at 298 K.
ICP-AES measurements were conducted with a Spectroflame D from Spectro in Argon 4.6. Around
30 mg of the sample were dissolved in 5 mL hydrofluoric acid (40%) and then diluted in 45 mL of
distilled water.
Nitrogen sorption experiments were conducted with a Quadrasorb SI from Quantachrome. Samples
were activated at 393 K under vacuum for 8 h prior to measurements at 77 K. The total specific
surface area was calculated by using the Brunauer-Emmett-Teller (BET) method at partial pressures
from p/p0 = 0.05 – 0.2. The outer and micropore area, as well as the micropore volume were
determined by the t-plot method at p/p0 = 0.2 – 0.5. The total pore volume was determined at
p/p0 = 0.99.
TGA-DSC measurements were recorded with a STA 409 from NETZSCH. Samples were heated in air
from 293 K to 1273 K with a heating ramp of 5 K.min-1.
HRTEM images were done with a transmission electron microscope JEOL JEM-2200FS FEG TEM/STEM
equipped with an objective lens aberration corrector CETCOR at 200 kV. Electron probe diameter was
1 nm.
SEM images were obtained at 10 kV on a scanning electron microscope JEOL JSM7000F (2006) with
field emission cathode; lateral resolution 1.2 nm at 30 kV; 3 nm at 1 kV. EDX-mapping experiments
were conducted with an attached EDAX Pegasus EDX/EBSD unit.
Pyridine adsorption infrared spectroscopy was conducted at a Nicolet Protégé 460 FTIR in order to
determine the nature of the acidic sites. Prior to the experiment, all samples, pressed to
self-supporting disks, were dried in vacuum at 623 K for 2 h. Afterwards the sample was cooled down
to 313 K and the background was measured. Then, pyridine was adsorbed on the surface for 15 min.
The temperature was raised to 423 K for 10 min to desorb weakly bound pyridine and purge the cell.
After cooling down again to 313 K the spectrum was recorded.
To determine the acidic and basic properties of the catalysts, NH3- and CO2-TPD measurements were
performed on a Porotec TPDRO 1100 Series. Around 0.12 g of each sample was placed in the quartz
glass tube. The samples were dried in nitrogen (20 mL.min-1) at 623 K for 1 h. Afterwards the sample
was cooled down to 373 K and the gas was switched to NH3 (3% in N2) and CO2, respectively, for 1 h.
After purging the reactor with nitrogen for 30 min, the measurement was conducted by heating the

sample up to 1173 K with a ramp of 20 K.min-1 and a He flow of 30 mL.min-1. The quantification of the
number of acidic or basic centres was based on calibration curves obtained by injecting different
amounts of NH3 or CO2 and the assumption that only one molecule could adsorb on one active
centre. Blanc tests without prior adsorbing any probe molecules show no signals over the entire
temperature range.
2.4 Catalytic experiments
The catalytic test reactions were conducted in a continuous gas phase setup. Liquid ethanol was
mixed with nitrogen and evaporated prior entering the first reactor. The catalyst was placed on a
metal frit in a stainless steel tube reactor (Ø = 6 mm). In the first stage, 0.400 g CuO\SiO2 or Ag\SiO2
was used (sieved to 250 – 500 μm). The copper catalyst was reduced in a hydrogen stream prior to
the experiment (20 mL.min-1, 2 h, 373 K). The silver equivalent was used as synthesized. For the
second stage, the catalyst volume was kept constant (0.85 mL) to ensure a constant contact time of
2 s. After heating the reactor up to the desired temperature in a nitrogen stream (24 mL.min-1),
0.002 mL.min-1 ethanol were fed into the reactor. First analysis of the reaction mixture was
performed after 2 h to ensure steady state conditions utilizing a GC-FID equipped with two
consecutive columns (BR-SWax and BR-1). Calibration was carried out with a certified gas mixture
containing the main reaction components, such as ethylene, diethyl ether, 1,3-butadiene, ethanol
and acetaldehyde. 1 Vol.-% benzene was added to the ethanol feed as an internal standard. In all
experiments, except two, carbon balances > 90% were obtained. Carbon balances above 100% are
within the experimental error. The different catalytic numbers were calculated with the respective
formula:
푐퐸푡푂퐻,푖푛
푐표푛푣푒푟푠푖표푛 = 푋 = 1 −
푐퐸푡푂퐻,표푢푡
푐푖
푠푒푙푒푐푡푖푣푖푡푦 = 푆푖 =
푐푡표푡푎푙 − (푐퐸푡푂퐻 + 푐퐴푐퐻
)
푦푖푒푙푑 = 푌 = 푆푖 ∙ 푋
푖
Results and Discussion
3.1 1st stage: Ethanol to acetaldehyde
The selective dehydrogenation of ethanol to acetaldehyde is necessary before testing the conversion
of an ethanol/acetaldehyde mixture to 1,3-butadiene in the second stage. Based on several
studies[17,23–25], including the first commercial approach of Union Carbide[27], Ag\SiO2 and
Cu\SiO2 were investigated as potential catalysts for this transformation.
3.1.1 Catalyst characterization
The catalyst’s textural parameters are summarized in Table 1 and show trends as expected. The
unmodified silica support has a specific surface area (SSA) of 230 m².g-1. After impregnation,
calcination and reduction, the SSA of the actual catalysts decreases. This decrease in SSA is more
pronounced in case of the silver catalyst. The catalysts reveal SSA of 203 m².g-1 and 195 m².g-1,
respectively. The total pore volume (Vp) drops from 0.85 cm³.g-1 to 0.77 cm³.g-1 for both catalysts.
X-ray diffraction (XRD) analyses confirm the characteristic reflexes of copper and silver, respectively.
̅
All indexed reflexes correspond to the space group Fm3m. Reflexes of copper oxide (space group

C2/c) vanish completely after the reduction step prior to the catalytic testing (Figure 1). The metal
loading determined by ICP-AES ranges between 9.8 - 10.3 wt.-%.
3.1.2 Catalytic activity
Both materials were screened under the same conditions from 423 K - 623 K (Figure 2). It is obvious
that Cu\SiO2 is a more active catalyst compared to supported silver. Preliminary test reactions with
different acetaldehyde concentrations, such as 0%, 15%, 20%, 25% and 35%, exhibited, that 20%
acetaldehyde leads to the best 1,3-butadiene yields in the latter reaction. The aspired 20%
acetaldehyde content of the ethanol feed can be achieved at low temperatures (around 463 K)
without any by-products. Also higher yields of acetaldehyde up to 80% (523 K) with almost 100%
selectivity are possible. Applying Ag\SiO2, the product feed composition looks rather different. To
achieve 20% conversion, a reaction temperature of around 550 K is necessary. Furthermore,
significant side product formation occurs. Since the overall two-stage process requires a constant
feed composition from the first stage, long-term stability of the catalyst is essential. Therefore, both
catalysts were tested over 90 h time on stream (tos) at 20% initial yield of acetaldehyde (Figure 3).
While there is almost no deactivation with Cu\SiO2 (around 2.5%), Ag\SiO2 deactivates comparatively
strong. In both cases, activity appears to reach a plateau; however, future studies certainly need to
further elaborate this point. XRD analyses of the spent catalysts show unaltered reflexes with equal
intensities and widths. Hence, there is no sintering of the metal component at these temperatures.
Thermogravimetric analysis of the spent catalysts show for both catalysts a defined weight loss of
1.7% for the copper and 2.2% for the silver catalyst in the temperature range of 650 – 850 K. Along
with a corresponding exothermic signal in DSC, it seems that there are small amounts of carbon
species on the catalyst surface. The high temperature of around 770 K suggests high molecular
carbon species. Nevertheless the amount of carbon on the catalysts is rather low, which is probably
due to the low reaction temperatures.[40,41] Since the Cu\SiO2 meets the requested conditions
almost perfectly, it has been used in further experiments.
In contrast to our study, Angelici et al. used CuO doped MgO\SiO2 catalysts to improve the material’s
catalytic activity and selectivity.[21] They proposed a higher acetaldehyde formation as origin of
improved catalytic results. To confirm if either Cu or CuO is the active species, we performed the
same reaction with CuO\SiO2 not reduced prior to the experiment. After 4 h tos, the spent catalyst
was analysed via X-ray diffraction. The pattern exclusively exhibits reflexes indicating pure Cu0.
Furthermore, at the beginning of the reaction, ethanol conversion is almost zero and rises until a
plateau is reached (Figure 4). The increasing activity can be correlated with the reduction degree of
the copper species. Both observations confirm that CuO is reduced to Cu0 in situ, which presents the
active species oxidizing ethanol to acetaldehyde.
3.2 2nd stage: Ethanol/acetaldehyde to 1,3-butadiene
In the second stage of this process, the resulting mixture of 20% acetaldehyde and remaining ethanol
is converted to 1,3-butadiene. Therefore, we prepared a series of zeolite-β based materials. With
these systems, it is possible to influence the acidity over a broad range by either changing the
Si/Al-module, the counter cation, adding a second dopant or varying MgO loading. Based on a
corresponding material set, a correlation of the acidic behaviour and the catalytic results becomes
possible.

3.2.1 Catalyst characterization
All materials were analysed regarding their structural properties and their chemical composition by
nitrogen physisorption and ICP-AES measurements. Table 2 summarizes the results. X-ray diffraction
patterns in a range from 2 = 3 – 50 ° of the different materials prior to catalytic tests are
summarized in Figure 5. No sample shows reflexes above 2 = 50 °. All patterns exhibit the two
zeolite β’s main reflexes at diffraction angles of 2 = 7.8 and 22.5 °, respectively. All indexed reflexes,
which are more or less pronounced in the different patterns, can be assigned to the assumed
structure.[42] Figure 5a shows the diffraction patterns of the commercial zeolite-β materials.
Increasing the module from 11 to 67 leads to higher intensities of the reflexes, which can be
correlated with a higher crystallinity. H-β280, however, shows lower intensities than H-β67. In this
context Camblor et al. studied the influence of the aluminium content in zeolite-β.[43] They
proposed that a higher Al-content can be correlated to the presence of more defects in the structure
decreasing crystallinity. Besides the aluminium content, post treatment of the zeolites has a strong
impact on the structure. Dealumination with an acid for instance leads to more defects in the
structure. This could present a potential explanation for a lower reflex intensity in case of H-β280.
Figure 5b and 5c show the X-ray patterns of the different impregnated protonated catalysts. No
reflexes of MgO or another dopant are visible. Since ICP-AES emphasises the presence of magnesium,
the present species appears to be X-ray amorphous. However, reflex intensities clearly decrease for
higher metal oxide loading. With additional doping of the MgO\zeolite-β samples, there is a further
pronounced decrease in reflex intensity. In case of Al2O3, there are just the two main reflexes visible
at 2 = 7.8 and 22.5 °. Figure 5d shows the impact of exchanging the proton by alkaline and earth
alkaline cations, respectively. For the alkaline exchanged zeolites there is a small increase of the
FWHM visible, which is more pronounced in case of Cs+. By adding Ca2+, the reflexes broaden
drastically. The overall increase is probably due to the enhancement of defects mentioned before
due to the exchange process. The stronger effect of the Ca2+ cations could be related to its higher
charge compared to the proton and the alkaline cations. However, the zeolite phase is still present in
all these samples.
The material’s textural parameters were determined by nitrogen physisorption experiments. The
results are summarized in Table 2. Regarding un-modified zeolites, a correlation between reflex
intensity and the corresponding surface area becomes evident. Upon impregnation and doping, the
materials lose more and more surface area and the pore volume decreases. Impregnating cation
exchanged zeolites with 10wt.-% MgO, a pronounced loss in specific surface area compared to
MgO\H-β280 occurs. A partial incorporation of magnesium into the zeolite structure, which is less
significant in case of the exchanged materials, is one potential explanation.
To examine the microscopic structure of the different materials, they were analysed by HRTEM and
SEM-EDX (Figure 6 and 7). H-β280 was not stable in the electron beam. After adding 10wt.-%
magnesium oxide to the sample, analysis became possible. Albeit indirectly, this presents a hint for
partial incorporation of Mg into the structure improving the material stability. To get an idea of the
local distribution of the different elements, SEM-EDX-mapping experiments were conducted.
Selected results are illustrated in Figure 7. It is evident that Mg as well as the respective second metal
is homogeneously distributed over the entire support. In the case of aluminium, the amount is
almost too low for analysis.

The determination of the nature of the acidic sites of the zeolite supports were measured by pyridine
adsorption infrared spectroscopy (Figure 8). In the spectra of H-β67 and H-β280 the absence of an
adsorption band at 1540 cm-1 show, that not Brønsted acidity is detectable in these materials. The
other bands could be assigned to Lewis acidic sites (L) and hydrogen bonded pyridine (H).[39,44] The
shift of the bands comparing the spectra of both materials were also observed in literature .[39,44] In
case of H-β11 there ae clear bands at 1540 cm-1 and 1640 cm-1. This reveals that this material also
shows Brønsted acidic behaviour.
To quantify the number of acidic and basic sites, NH3- and CO2-TPD measurements were conducted.
In all cases, there are two main regions. One at very low temperatures of around 450 K and another
one, depending on the material, in the area of 750 - 1150 K. Referring to Katada et al.,[45] the signal
around 450 K is related to physisorbed species and is therefore not considered in further discussions.
The second region can be related to chemisorbed ammonia and carbon dioxide, respectively and is
used to determine the number of acidic as well as basic sites. The results of quantitative TPD
analyses are summarised in Table 4. As a measure for the effective acidity, we chose the ratio of the
total amounts of acidic to basic sites in the respective material. It is obvious that the pure zeolites
exhibit the highest ratio. By lowering the module, the effective acidity decreases. All different further
modifications lead to a more or less pronounced decrease in acidity. The strongest effect could be
observed in case of metal cation exchanged zeolites. Figure 9 shows selected temperature
programmed desorption profiles with ammonia (a + b) as well as carbon dioxide (c + d). Besides the
number of acidic and basic sites, also distinct differences in their strength and distribution become
evident. While the two pure zeolites H-β11 and H-β280 show just one signal at 712 and 1038 K,
H-β67 reveals a much broader signal corresponding of a broader distribution of acidic centres of
different strength (Figure 9a). Additionally, these three samples exhibit almost no basic sites (9c).
.
After impregnation with Mg(NO)3 6H2O and subsequent calcination, the amount of acidic as well as
basic sites increases. While the distribution of acidic sites keeps broad, MgO\H-β150 and MgO\β280
possess a sharp signal around 750 K indicating a very narrow distribution of basic sites of comparable
strength. The same effect is visible by exchanging β-zeolites with K+ and Cs+, respectively. Especially
for MgO\K-β280, only one type of acidic site appears to be present with a defined ammonia
desorption at 746 K. TGA-DSC measurements show no significant heat flow up to 1173 K, which
excludes that this observation is due to any structural changes in the material. A further doping with
an additional metal oxide introduces new basic and acidic functionalities into the respective material
over a broad temperature range.
With these different characterization methods we can conclude, that we were able to prepare a
series of materials, which differ in their acidic-basic behaviour over a broad range. The physical
properties, such as the SSA, the pore volume and the structure themselves are certainly influenced
by the different modifications, but keep in the same range. Thus, studying the catalytic behaviour of
this material set allows further insights into the influence of material properties such as acidity.
3.2.2 Catalytic results
The discussed materials were studied as catalysts in the production of 1,3-butadiene from the
ethanol/acetaldehyde ratio. Besides ethanol conversion, we focussed our interest on the four main
components appearing in this reaction. 1,3-butadiene, ethylene and diethyl ether, which are
produced by dehydration of ethanol, and other butenes. With these components, we generally cover
85 – 100% of the total peak area in the resulting chromatograms. C4 – C6 hydrocarbons occur as

further products in small quantities. The catalytic results including the carbon balances are
summarised in Table 3.
The primal experiment was conducted with pure H-β11 as catalyst (entry 1). This leads to almost 90%
selectivity to the dehydration products ethylene and diethyl ether at full conversion. The desired
product is only visible in traces. The use of a zeolite-β with a module of 67 shows comparable results
(entry 2). A further increase to a Si/Al ratio of 280 mainly decreases the conversion to 68% at
constant selectivity (entry 3).
With the addition of 10wt.-% magnesium oxide to the respective zeolites, 1,3-butadiene selectivity
improves (entries 7 – 9). With 22% yield of 1,3-butadiene, MgO\H-β280 facilitated the best result.
Surprisingly MgO\H-β67 shows with 22% by far the highest selectivity towards other butenes of all
tested catalyst systems. Additionally, in this run we received a carbon balance of just 84%. Entry 10
exhibits the drastic influence of the addition of acetaldehyde. With a pure ethanol feed, the
conversion drops to 15% and 75% of the converted ethanol is just dehydrated to ethylene and
diethyl ether, respectively.
To increase the low conversion of 34%, this catalyst system was doped with a redox active
component, such as ZnO and NiO, respectively (entries 15 and 16). These components were already
used in the one-stage catalytic conversion of ethanol to 1,3-butadiene. Test reactions reveal that
both materials have also a positive influence in the two-stage approach. Especially ZnO keeps the
selectivity towards 1,3-butadiene high and the amount of dehydration products low. Additionally,
ethanol conversion increases from 34 to 49%. The use of Al2O3 appears not to be beneficial. Even
though the conversion increases to almost 60% the selectivity to the desired product drops to 46%
(entry 17).
In the next step, the zeolites were exchanged by Ca2+, K+ and Cs+, respectively. Entries 4 – 6 show the
catalytic results obtained for β280 in its cationic form. In the case of Cs-β280, a low conversion of 5%
was observed. As a result, the absolute amounts of products, which are detected, are low as well and
consequentially the error is rather high. Furthermore, the four components just cover 29% of the
total peak area emphasising the formation of undesired by-products. Ca2+ and K+ materials
performed better than the protonated zeolite. While the 1,3-butadiene selectivity increases to 10
and 17%, respectively, the selectivity towards ethylene and diethyl ether and the conversion
decrease. The amount of formed butenes remains unaltered at a low level.
With the addition of 10wt.-% MgO to these materials, their catalytic behaviour improves. With
MgO\K-β280 and MgO\Cs-β280, the best results are reached facilitating 72 and 71% 1,3-butadiene
selectivity. Interestingly the conversion is not affected by this modification. A doping with ZnO
presents an interesting direction for further studies.
In a final step the influence of the amount of magnesium oxide doping on the H-β280 support was
studied. By bisecting the magnesium oxide amount to 5%, a slightly higher activity (41% conversion)
and a 7% increase in selectivity towards dehydration products can be observed (entry 15).
Interestingly, 1,3-butadiene formation is not affected. A further lowering of the amount follows a
drastic deterioration of the catalytic performance (entry 14)

3.2.3. Correlation between the catalytic results and the catalyst’s basic and acidic properties.
Several authors proposed a relationship between the acidic and basic properties of catalysts and
their catalytic performance. Recently Angelici et al. published a correlation of their results with the
preparation method and the resulting amount of acidic and basic sites in combination with their
respective strength. They proposed that a limited amount of strong basic sites and an intermediate
amount of acidic ones gives the best results.[35] Based on these investigations, we designed a
material set to decouple the acidic and basic properties from the preparation method. As a result,
the morphology and textural parameters remain comparable and their influence should be
minimized. In Figure 10, the ratio of the number of acidic and basic sites is correlated to the
selectivities towards the main products of the reaction: 1,3-butadiene, ethylene, diethyl ether and
other butenes (10a –10c). In Figure 10d, the dependency of the ethanol conversion is illustrated.
Leaving the selectivity to butenes aside, there are clear relationships between the respective reaction
parameters and the ratio of acidic to basic sites. Ca-β280, K-β280 and Cs-β280 are not included due
to the high experimental error induced by the low ethanol conversion and the resulting low amounts
of identified products. The results of the other materials clearly emphasise the importance of the
relative ratio of acidic and basic sites. Decreasing the effective acidity, the formation of the two main
side products ethylene and diethyl ether is more and more supressed (Figure 10b), while the
selectivity of 1,3-butadiene increases (Figure 10a). Contrariwise, decreasing acidity the conversion
decreases as well (Figure 10d). Considering the fact that the conversion follows the opposite trend,
an optimization of 1,3-butadiene production based on these material classes appears to be limited.
Importantly, a proper control of the acid-base properties of potential catalysts based on modified
zeolites is indispensable.
Conclusion
In this study we presented a two-stage process producing 1,3-butadiene out of ethanol. In the first
stage, we observed the partial conversion of ethanol to acetaldehyde on silica supported copper and
silver catalysts. With Cu\SiO2, 20% acetaldehyde yield with 100% selectivity at relatively low
temperatures of 463 K could be achieved. At these temperatures, catalyst deactivation could be
suppressed and the catalyst did not show any change of activity or selectivity for 90 h time on
stream. Furthermore, we could prove that CuO is reduced to Cu0 in situ and that this species
catalyses the dehydrogenation of ethanol. In the second stage, the resulting mixture of ethanol and
acetaldehyde was converted to 1,3-butadiene. A set of zeolite-β based catalysts was designed aiming
for variation of the concentration and relative ratio of acidic and basic sites for materials of
comparable textural properties. We varied the zeolite’s module, added magnesium oxide and a
second metal oxide as a dopant and exchanged the proton by alkaline and earth alkaline metals,
respectively. MgO(10)\K-β280 facilitated the highest 1,3-butadiene selectivity of 72%. Nevertheless,
the best yield of 33% was observed with a ZnO doped MgO\H-β280 catalyst. A clear relationship
between the effective acidity and the catalyst performance could be established. Surprisingly, no
correlation between the formation of butenes and the amount of acidic and basic centres occurred.
The presented results allow comparable yields as achieved in other studies. However, the reaction
temperature could be decreased below 600 K.

Acknowledgement
This study was made possible by the German Research Foundation DFG (PA 1689/4-1). The authors
gratefully acknowledge this financial support. Furthermore, we thank Clariant AG for providing
different zeolites.
References
[1] S. Gonzalez-Garcia, L. Luo, M.T. Moreira, G. Feijoo, G. Huppes, Biomass Bioenergy 36 (2012)
268–279.
[2] J.J. Bozell, G.R. Petersen, Green Chem. 12 (2010) 539–554.
[3] C. Angelici, B.M. Weckhuysen, Bruijnincx, Pieter C. A., ChemSusChem (2013).
[4] R. Dittmeyer, K. Winnacker, L. Küchler (Eds.), Methodische Grundlagen // Chemische Technik:
Prozesse und Produkte, 5th ed., Wiley-VCH, Weinheim, 2004 // 2004-2006.
[5] Bruijnincx, Pieter C. A., B.M. Weckhuysen, Angew. Chem., Int. Ed. 52 (2013) 11980–11987.
[6] I. Ostromislensky, J. Russ. Phys.-Chem. Soc. (1915) 1472.
[7] S.V. Lebedev, GB 331482 (1930).
[8] S.V. Lebedev, Zh. Obshch. Khim. 3 (1933) 698–717.
[9] W.J. Toussaint, J.T. Dunn, D.R. Jackson, J. Ind. Eng. Chem. (Washington, D. C.) 39 (1947) 120–
125.
[10] H.E. Jones, E.E. Stahly, B.B. Corson, J. Am. Chem. Soc. 71 (1949) 1822–1828.
[11] E.V. Makshina, M. Dusselier, W. Janssens, J. Degreve, P.A. Jacobs, B.F. Sels, Chem. Soc. Rev.
(2014) 7917–7953.
[12] A. Chieregato, J. Velasquez Ochoa, C. Bandinelli, G. Fornasari, F. Cavani, M. Mella,
ChemSusChem 8 (2015) 377–388.
[13] R. F. Goldstein, A. L. Waddams, The petroleum chemicals industry, E.&F.N. Spon LTD, London,
1967.
[14] W.J. Toussaint, J.T. Dunn, US 2421361 (1947).
[15] M. Leon, E. Diaz, A. Vega, S. Ordonez, A. Auroux, Appl. Catal., B 102 (2011) 590–599.
[16] Y. Kitayama, A. Michishita, J. Chem. Soc., Chem. Commun. (1981) 401–402.
[17] B.B. Corson, H.E. Jones, C.E. Welling, J.A. Hinckley, E.E. Stahly, J. Ind. Eng. Chem. (Washington, D.
C.) 42 (1950) 359–373.
[18] S. Kvisle, A. Aguero, Sneeden, R. P. A., Appl. Catal. 43 (1988) 117–131.
[19] W. Janssens, E.V. Makshina, P. Vanelderen, F. De Clippel, K. Houthoofd, S. Kerkhofs, J.A.
Martens, P.A. Jacobs, B.F. Sels, ChemSusChem (2014) n/a.
[20] H. Niiyama, S. Morii, E. Echigoya, Bull. Chem. Soc. Jap. 45 (1972) 655–659.
[21] C. Angelici, Velthoen, Marjolein E Z., B.M. Weckhuysen, Bruijnincx, Pieter C. A., ChemSusChem 7
(2014) 2505–2515.
[22] V.L. Sushkevich, I.I. Ivanova, V.V. Ordomsky, E. Taarning, ChemSusChem 7 (2014) 2527–2536.
[23] E.V. Makshina, W. Janssens, B.F. Sels, P.A. Jacobs, Catal. Today 198 (2012) 338–344.
[24] M. Lewandowski, G.S. Babu, M. Vezzoli, M.D. Jones, R.E. Owen, D. Mattia, P. Plucinski, E.
Mikolajska, A. Ochenduszko, D.C. Apperley, Catal. Commun. 49 (2014) 25–28.

[25] G. Natta, R. Rigamonti, Chim. Ind. (Milan, Italy) 29 (1947) 195–200.
[26] G. Natta, R. Rigamonti, Chim. Ind. (Milan, Italy) 29 (1947) 239–43.
[27] W. Szukiewicz, US 2357855 (1944).
[28] M.D. Jones, C.G. Keir, I.C. Di, Robertson, Ruth A. M., C.V. Williams, D.C. Apperley, Catal. Sci.
Technol. 1 (2011) 267–272.
[29] V.L. Sushkevich, I.I. Ivanova, V.V. Ordomsky, E. Taarning, ChemSusChem (2014).
[30] H.-J. Chae, T.-W. Kim, Y.-K. Moon, H.-K. Kim, K.-E. Jeong, C.-U. Kim, S.-Y. Jeong, Appl. Catal., B
150-151 (2014) 596–604.
[31] V.V. Ordomskiy, V.L. Sushkevich, I.I. Ivanova, WO 2012/015340 (2012).
[32] T. de Baerdemaeker, M. Feyen, U. Muller, B. Yilmaz, F.-S. Xiao, W. Zhang, T. Yokoi, X. Bao, H.
Gies, De Vos, Dirk E., ACS Catal. 5 (2015) 3393–3397.
[33] O. Larina, P. Kyriienko, S. Soloviev, Catal Lett 145 (2015) 1162-1168.
[34] R. Ohnishi, T. Akimoto, K. Tanabe, J. Chem. Soc., Chem. Commun. (1985) 1613–1614.
[35] C. Angelici, Velthoen, Marjolein E. Z., B.M. Weckhuysen, Bruijnincx, Pieter C. A., Catal. Sci.
Technol. 5 (2015) 2869–2879.
[36] V.L. Sushkevich, I.I. Ivanova, E. Taarning, Green Chem 17 (2015) 2552–2559.
[37] W.M. Quattlebaum, W.J. Toussaint, J.T. Dunn, J. Am. Chem. Soc. 69 (1947) 593–9.
[38] V.L. Sushkevich, D. Palagin, I.I. Ivanova, ACS Catal 5 (2015) 4833–4836.
[39] Y. Miyamoto, N. Katada, M. Niwa, Microporous Mesoporous Mater. 40 (2000) 271–281.
[40] J.M. Church, H.K. Joshi, Ind. Eng. Chem. 43 (1951) 1804–11.
[41] L. Kocurkova, D. Tomanova, M. Kraus, Chem. Prum. 30 (1980) 71–4.
[42] Treacy, M. M. J., Higgins, J. B., Editors., Collection of Simulated XRD Powder Patterns for
Zeolites, Elsevier, 2001.
[43] M.A. Camblor, A. Corma, A. Mifsud, J. Pérez-Pariente, S. Valencia, in: Hakze Chon, Son-Ki Ihm
and Young Sun Uh (Ed.), Studies in Surface Science and Catalysis Progress in Zeolite and
Microporous Materials Preceedings of the 11th International Zeolite Conference, Elsevier, 1997,
pp. 341–348.
[44] Parry, E. P., J. Catal. 2 (1963) 371–9.
[45] N. Katada, M. Niwa, Catal. Surv. Asia 8 (2004) 161–170.

Figure 1: X-ray diffraction patterns of Ag\SiO2, Cu\SiO2 and CuO\SiO2 with indexing of the respective
phases.
Figure 2: Product stream distribution for Cu\SiO2 and Ag\SiO2, respectively. Reaction conditions: 0.4 g
catalyst, ethanol (liquid phase) flow of 0.002 mL.min-1, nitrogen flow of 24 mL.min-1 (GHSV = 96 h-1,
contact time = 2 s).
Figure 3: Stability test at 20% conversion. Reaction conditions: 0.4 g catalyst, ethanol (liquid phase)
flow of 0.002 mL.min-1, nitrogen flow of 24 mL.min-1 (GHSV = 96 h-1, contact time = 2 s).
Figure 4: Dehydrogenation of ethanol to acetaldehyde with CuO\SiO2 at different temperatures.
Reaction conditions: 0.4 g catalyst, ethanol (liquid phase) flow of 0.002 mL.min-1, nitrogen flow of
24 mL.min-1 (GHSV = 96 h-1, contact time = 2 s).
Figure 5: X-ray diffraction patterns of the different catalysts used in the 2nd stage with indexing the
zeolite phase.
Figure 6: HRTEM images of MgO(10)\H-β280.
Figure 7: SEM-EDX images: a) MgO(10)\H-β280c; b) ZnO-MgO\H-β280; c) NiO-MgO\H-β280;
Al2O3-MgO\H-β280
Figure 8: FTIR spectra of the three different supports H-β11, H-β67 and H-β280 after adsorption of
pyridine. The spectra are taken at 313 K in the region of 1700 – 1400 cm-1 (L = Lewis acid site;
B = Brønsted acid site; H = hydrogen bonded pyridine).
Figure 9: Temperature programmed desorption profiles: a) Variation of the module (NH3); b)
Modification of H-β280 (NH3); c) Variation of the module (CO2); d) Modification of H-β280 (CO2).
Figure 10: Correlation curves: a) 1,3-butadiene selectivity; b) dehydration products selectivity; c)
butene selectivity; d) ethanol conversion. Grey data points represent the not considered samples
Ca β280, K-β280 and Cs-β280.

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7
a)
fig.7
Al
Al
Al
Mg
Mg
Mg
Mg
Si
Si
Si
Si
b)
c)
Zn
Ni
d)
Al

Figure 8

Figure 9

Figure 10

Table 1: Catalyst parameters 1st stage: Metal loading, specific surface area and pore volume.
Entry Material Metal content / Specific surface Total pore
wt.-% [a]
-
Cu\SiO2 9.8
Ag\SiO2 10.3
area / m².g [b]
230
203
195
volume / cm³.g
0.85
0.77
0.77
1
2
3
SiO2
[a] determined by ICP-AES; [b] determined by using the BET method
Table 2: Catalyst parameters 2nd stage: Metal loadings, specific surface area and pore volume.
Entry Material
MgO /
Dopant / Specific surface Total pore
wt.-% [a] wt.-% [a] area / m².g [b]
volume / cm³.g
0.76
1
H-β11
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
460
562
533
482
538
505
319
225
443
266
253
249
519
440
349
340
358
2
H-β67
0.67
0.90
0.83
0.81
0.79
0.51
0.34
0.69
0.59
0.70
0.69
0.81
0.70
0.59
0.63
0.58
3
H-β280
4
Ca-β280
5
K-β280
6
Cs-β280
7
8
9
MgO\H-β11
MgO\H-β67
MgO(10)\H-β280
MgO\Ca-β280
MgO\K-β280
MgO\Cs-β280
MgO(1)\H-β280
MgO(5)\H-β280
11.63
9.01
9.37
9.58
9.79
9.21
0.88
4.77
10
11
12
13
14
15
16
17
ZnO-MgO\H-β280 10.30
NiO-MgO\H-β280 9.31
Al2O3-MgO\H-β280 9.01
4.61
4.82
4.98
[a] determined by ICP-AES; [b] determined by using the BET method

Table 3: Catalytic results 2nd stage: Selectivities, conversion and carbon balances.
Entry Material
1,3-butadiene Dehydration
/ %
Butenes Conversion C-balance
products / % [a]
/ %
8
5
1
2
4
3
12
22
14
2
/ %
100
100
68
32
10
5
61
50
34
15
/ %
99
1
H-β11
2
1
4
10
19
17
40
40
65
13
88
93
92
86
60
9
35
30
12
75
2
H-β67
104
101
99
101
101
96
84
98
98
3
H-β280
4
Ca-β280
5
K-β280
6
Cs-β280
7
8
9
10
MgO\H-β11
MgO\H-β67
MgO(10)\H-β280
MgO(10)\H-β280
(0% Acetaldehyde)
MgO\Ca-β280
MgO\K-β280
MgO\Cs-β280
MgO(1)\H-β280
MgO(5)\H-β280
ZnO-MgO\H-β280
NiO-MgO\H-β280
11
12
13
14
15
16
17
18
60
72
71
18
66
64
57
22
7
7
69
19
16
33
39
7
5
4
12
8
7
8
11
44
35
34
54
41
49
55
59
90
93
92
104
84
91
90
92
Al2O3-MgO\H-β280 46
[a] ethylene + diethyl ether

Table 4: Catalyst parameters 2nd stage: Acidic and basic properties determined by temperature programmed desorption.
Ent Material
ry
nacidic Tmax nacidic
Tmax nacidic Tmax nbasic
/
Tmax nbasic
/
Tmax nbasic
/
Tmax nacidi
/
/ K
/
/ K
/
/ K
mmol / K
/g
mmol / K
/g
mmol / K
/g
c/nba
sic
mmol
/g
mmol
/g
mmol
/g
1
2
H-β11
H-β67
0.105 712
0.014 807
7.5
5.2
0.024 667 0.022 934 0.024 113 0.004 667 0.009 876
9
3
4
5
H-β280
Ca-β280
K-β280
0.065 103
8
0.029 958
0.013 999
5
0.035 903 0.012 110
0.62
0.73
5
0.004 789 0.029 103
9
0.003 808 0.021 973
6
7
8
9
Cs-β280
0.003 768 0.015 968
0.060 729 0.115 817
0.034 744 0.067 848
0.023 986
0.031 754 0.043 864
0.004 755 0.035 773 0.027 905 1.5
0.78
2.4
MgO\H-β11
MgO\H-β67
MgO(10)\H-
β280
0.038 751 0.006 810 0.006 943 0.005 743 0.042 767 0.008 921 0.90
10
11
12
13
MgO\Ca-β280 0.058 752 0.005 807 0.003 941 0.045 749 0.031 868
0.87
1.1
1.1
MgO\K-β280
MgO\Cs-β280
MgO(1)\H-
β280
0.063 746
0.011 731 0.045 771
0.012 722 0.048 763
0.002 779 0.016 923
0.006 768 0.061 840
0.057 949 0.025 108
0
4.6
14
15
16
17
MgO(5)\H-
β280
ZnO-MgO\H-
β280
NiO-MgO\H-
β280
0.045 771 0.004 807 0.006 953 0.007 742 0.036 764 0.016 920 0.93
0.032 757 0.016 801 0.009 934 0.007 743 0.021 776 0.012 912 1.4
0.027 759 0.002 856 0.054 917 0.082 610
1.0
Al2O3-MgO\H- 0.043 761 0.079 816 0.006 117 0.003 756 0.044 772 0.038 905 1.5
β280
3