Preparation of alumina with different precipitants for the gas phase dehydration of glycerol and their characterization by thermal analysis

Article abstract of DOI:10.1007/s10973-020-09308-4

Mesoporous aluminas were prepared using different precipitants at different pH of precipitation and evaluated in the gas phase dehydration of glycerol. Samples were characterized by thermal analysis techniques (TGA, NH₃-TPD and TPO), specific surface area measurements, XRD and SEM. The gas phase dehydration of glycerol was performed in a fixed bed quartz tubular reactor at 773?K using a 10% glycerol aqueous solution. Thermogravimetric analysis has revealed the temperature regions of decomposition for the prepared alumina catalyst precursors. The specific surface area and the crystallinity of the samples were dependent on both the pH and the used precipitating agent. Samples precipitated with NaOH presented higher density of acid sites than the samples prepared with Na₂CO₃, regardless of precipitation pH. The catalytic properties of the prepared aluminas are mainly related to the specific surface area and to acidic characteristics. Conversions of glycerol above 85% were obtained for all samples. The selectivity for glycerol dehydration was strongly related to the amount and strength of acid sites. The best result for dehydration was obtained for samples prepared with NaOH and precipitated at pH = 5. These results are related to the higher specific surface area, greater amount of acid sites and the higher ratio of weak acid sites. TPO revealed the amount of carbon deposited on the catalysts. Samples that showed higher carbon formation also showed a higher production of light olefins, indicating that the formation of carbon is related to the formation of these byproducts. NH₃-TPD has shown the ratio of different acid sites on the surface of alumina samples that makes possible to estimate the correlation between the acidity and the catalytic properties.

Full text of DOI:10.1007/s10973-020-09308-4

Journal of Thermal Analysis and Calorimetry
https://doi.org/10.1007/s10973-020-09308-4
Preparation of alumina with diꢀerent precipitants for the gas phase
dehydration of glycerol and their characterization by thermal analysis
Dirleia S. Lima · Oscar W. Perez‑Lopez1
1
Received: 12 September 2019 / Accepted: 9 January 2020
©
Akadémiai Kiadó, Budapest, Hungary 2020
Abstract
Mesoporous aluminas were prepared using diferent precipitants at diferent pH oꢀ precipitation and evaluated in the gas
phase dehydration oꢀ glycerol. Samples were characterized by thermal analysis techniques (TGA, NH -TPD and TPO), spe-
3
ciꢁc surꢀace area measurements, XRD and SEM. The gas phase dehydration oꢀ glycerol was perꢀormed in a ꢁxed bed quartz
tubular reactor at 773 K using a 10% glycerol aqueous solution. Thermogravimetric analysis has revealed the temperature
regions oꢀ decomposition ꢀor the prepared alumina catalyst precursors. The speciꢁc surꢀace area and the crystallinity oꢀ the
samples were dependent on both the pH and the used precipitating agent. Samples precipitated with NaOH presented higher
density oꢀ acid sites than the samples prepared with Na CO , regardless oꢀ precipitation pH. The catalytic properties oꢀ the
2
3
prepared aluminas are mainly related to the speciꢁc surꢀace area and to acidic characteristics. Conversions oꢀ glycerol above
5% were obtained ꢀor all samples. The selectivity ꢀor glycerol dehydration was strongly related to the amount and strength
8
oꢀ acid sites. The best result ꢀor dehydration was obtained ꢀor samples prepared with NaOH and precipitated at pH=5. These
results are related to the higher speciꢁc surꢀace area, greater amount oꢀ acid sites and the higher ratio oꢀ weak acid sites.
TPO revealed the amount oꢀ carbon deposited on the catalysts. Samples that showed higher carbon ꢀormation also showed
a higher production oꢀ light oleꢁns, indicating that the ꢀormation oꢀ carbon is related to the ꢀormation oꢀ these byproducts.
NH -TPD has shown the ratio oꢀ diferent acid sites on the surꢀace oꢀ alumina samples that makes possible to estimate the
3
correlation between the acidity and the catalytic properties.
Keywords Dehydration oꢀ glycerol · Acrolein · Alumina synthesis · Co-precipitation · Acid strength
Introduction
An interesting product that can be obtained through glyc-
erol conversion is acrolein, which is currently produced by
the oxidation oꢀ propylene with a catalyst-based process on
bismuth molybdate [3, 4]. The production oꢀ acrolein ꢀrom
sustainable and renewable resources is recent. However, the
main obstacle to such industrial application (large scale) is
the economic issue related to the cost oꢀ reꢁned glycerol [4].
Crude glycerol ꢀrom biodiesel production usually contains
water. The direct use oꢀ this glycerol without the removal
oꢀ water is an advantage in the catalytic reaction to obtain
acrolein, in addition to bringing economic and environmen-
tal beneꢁts since it makes its puriꢁcation unnecessary and
also minimizes industrial water consumption. A promising
way oꢀ converting this aqueous crude glycerol solution is
by dehydration, which leads to the production oꢀ high-value
chemicals such as acrolein and hydroxyacetone, according
to reaction (1) [5]:
Glycerol is an alcohol obtained as a byproduct oꢀ biodiesel
production via the transesteriꢁcation reaction oꢀ vegetable
oils and animal ꢀats [1]. One way oꢀ adding value to residual
glycerol is through glycerochemistry, which studies the diꢀ-
ꢀerent chemical routes through which it is possible to recover
this low-quality glycerol and transꢀorm it into other prod-
ucts. Among the existing catalytic chemical routes, dehydra-
tion, oxidation, hydrogenolysis, etheriꢁcation, acetylation,
halogenation and steam reꢀorming oꢀ glycerol are the most
important [2].
*
Oscar W. Perez-Lopez
perez@enq.uꢀrgs.br
1
Laboratory oꢀ Catalytic Processes - PROCAT, Department
oꢀ Chemical Engineering, Federal University oꢀ Rio
Grande do Sul (UFRGS), Ramiro Barcelos Street, 2777,
Porto Alegre, RS CEP 90035-007, Brazil
Vol.:(0
1
123456789)
3

D. S. Lima, O. W. Perez-Lopez
it is known that other parameters, such as medium pH and
choice oꢀ precipitating agent, also have a great inꢂuence
on the synthesis process. It is thereꢀore very important that
these ꢀactors are also evaluated.
H
C
O
C
H
C
H
H
H
C
H
C
H
C
Acid site
H O
In this context, the objective oꢀ the present work is the
preparation oꢀ aluminas by the continuous precipitation
method, varying the pH oꢀ precipitation and the nature oꢀ
the precipitating agent, and its application in the dehydra-
tion oꢀ glycerol.
H
H
Acrolein
–
2
OH OH OH
Glycerol
O
H C
CH OH
2
3
Hydroxyacetone
(
1)
Acrolein obtained ꢀrom these processes can be used as
eedstock to produce acrylic acid and its derivatives, while
Experimental
ꢀ
hydroxyacetone is an intermediate ꢀor the production oꢀ pro-
panediols and can be employed to obtain acetic acid [6, 7].
Several solid acid catalysts have been tested ꢀor gas phase
dehydration oꢀ glycerol [8]. Some oꢀ them are: ion exchange
resins, sulꢀates, phosphates, heteropoly acids supported on
silica, alumina, active carbon and zeolites [9]. Although
widely used, the microporous nature oꢀ zeolites oꢀten has
transport limitations, particularly when large molecules are
involved, which adversely afects the catalytic perꢀormance
Catalyst preparation
The catalysts were prepared using the continuous precipi-
tation method. A solution oꢀ aluminum nitrate (Al(NO ) )
3 3
and diferent precipitating solutions (Na CO , NaOH and
2
3
KOH) were used. The precipitation was perꢀormed in a
CSTR (Continuous Stirred-Tank Reactor) jacketed reactor
by mixing the aluminum nitrate solution (0.5 M) with the
alkaline solution (1 M) containing the base, at ꢁxed tem-
perature and pH (333 K and 7.0± 0.1 or 5.0± 0.1, respec-
tively). The precipitated material was collected and submit-
ted to crystallization at 333 K ꢀor 4 h. Subsequently, it was
vacuum-ꢁltrated and washed with deionized water until a
conductivity lower than 50 μS was reached. The material
was oven-dried at 353 K ꢀor 12 h. Finally, it was calcined at
[
10]. Pathak et al. [11] concluded that the acidity oꢀ the
catalyst has a great efect on the conversion oꢀ glycerol to
the yield oꢀ liquid products: acetaldehyde, acrolein, ꢀormal-
dehyde and acetone. In the dehydration oꢀ glycerol, zeo-
lites and heteropoly acids are the mainly used catalysts [4].
However, while most oꢀ these catalytic systems lead to high
selectivity ꢀor acrolein with high total glycerol conversion,
−
1
8
73 K ꢀor 6 h with a heating rate oꢀ 10 K min and an air
ꢀ
ew maintain their catalytic properties ꢀor more than 5 to
−
1
ꢂow rate oꢀ 50 mL min .
1
0 h. Catalyst deactivation occurs mainly due to the exten-
sive deposition oꢀ coke on its surꢀace [12].
Due to its mechanical and physical properties, alumina
can be used as a catalyst or a catalytic support ꢀor numerous
chemical processes. Among the diferent types oꢀ alumina
Catalyst characterization
Thermogravimetric analysis (TG) was perꢀormed in a ther-
mobalance (TA Instruments, SDT-Q600) to evaluate the
thermal events due to heating. Totally, 10 mg oꢀ uncal-
cined sample was heated to 1073 K with a heating ramp oꢀ
used in catalysis, γ-Al O is the most used as a catalytic sup-
2
3
port due to its mechanical stability, moderately high speciꢁc
surꢀace area, sintering resistance over a wide temperature
range, and also ꢀor having a high degree oꢀ metallic disper-
sion [13]. Despite its properties, alumina has been used on
a smaller scale in this process, mainly as the support [8, 12,
−
1
−1
1
0 K min under 100 mL min oꢀ air ꢂow [25].
The XRD difractograms oꢀ the samples were obtained
using a Bruker D2 Phaser X-ray difractometer with Cu-kα
radiation [26]. Scanning electron microscopy (SEM) was
perꢀormed on an EVO MA10—Carl Zeiss equipment using
backscattered electrons at 10 kV, with samples coated with
a thin gold ꢁlm prior to the SEM analyses.
1
4, 15].
Industrially, alumina is mainly obtained by the Bayer pro-
cess which involves the extraction oꢀ alumina ꢀrom bauxite
using a hot sodium hydroxide solution under high pressure.
Gibbsite is precipitated ꢀrom the solution at temperatures
between 338 and 343 K and calcined to produce alumina
The speciꢁc surꢀace area oꢀ the catalysts was obtained
ꢀ
rom nitrogen adsorption/desorption measurements. Previ-
[
16]. On the other hand, there are diferent methods oꢀ syn-
ously, the samples were pre-treated at 573 K ꢀor a period
oꢀ 3 h under vacuum. The analysis was perꢀormed on a
Quantachrome analyzer, model NOVA 4200e. The spe-
ciꢁc surꢀace area oꢀ the samples was determined by the
BET (Brunauer–Emmett–Teller) multi-point method
thesis oꢀ γ-Al O reported in the literature [17–24]. One
2
3
oꢀ these methods is the continuous co-precipitation [13],
which allows obtaining a structure with more homogene-
ous crystals. Some oꢀ these studies evaluate the inꢂuence
oꢀ calcination temperatures to obtain these materials, but
[
27, 28]. The pore volume, average pore diameter and
1
3

Preparation of alumina with diꢀerent precipitants for the gas phase dehydration of glycerol…
pore size distribution were calculated with the BJH (Bar-
rett–Joyner–Halenda) method using the desorption data.
The acidity oꢀ the catalysts was determined by tempera-
Table 1 Samples nomenclature, preparation conditions, BET speciꢁc
surꢀace area, average pore diameter and pore volume
2
−1
3
Catalyst Precipitant
pH SBET/m g
D /nm V /cm g
p
p
ture-programmed desorption oꢀ ammonia (NH -TPD), per-
3
A7AP1
A7AP2
Na CO
NaOH
7.0
7.0 277
48
3.83
3.84
9.22
3.84
3.82
3.80
0.126
0.479
0.489
0.660
0.367
0.235
2
3
ꢀ
ormed in a SAMP3 multipurpose analysis system equipped
with a thermal conductivity detector (TCD). First, the sam-
−
1
A7AP12 Na CO +NaOH 7.0 323
2
3
ples were pre-treated at 373 K with 30 mL min oꢀ He
A7AP3
A5AP1
A5AP2
KOH
Na2CO3
NaOH
7.0 432
5.0 389
5.0 359
ꢀ
or 30 min. The ammonia adsorption step was perꢀormed at
−
1
3
73 K ꢀor a period oꢀ 30 min using 30 mL min oꢀ a mix-
ture containing 5 vol % ammonia in He [26, 28–30]. Aꢀter
that, a purge with He was perꢀormed ꢀor 30 min. Then, the
temperature was raised to 1023 K with a heating ramp oꢀ
−
1
1
0 K min under He ꢂow.
1
00
A7AP1
A7AP2
A7AP12
A7AP3
A5AP2
A5AP1
The amount oꢀ carbon deposited on the catalysts aꢀter
the reaction was determined by temperature-programmed
oxidation (TPO) in a thermobalance (TA Instruments
Q600). A 10 mg sample was heated under an air ꢂow rate
90
80
−
1
−1
oꢀ 100 mL min at a rate oꢀ 10 K min until 1023 K [26,
7
6
5
4
3
2
0
0
0
0
0
0
3
0, 31].
Catalytic activity
The catalytic activity oꢀ the samples was evaluated by the
conversion oꢀ glycerol in a tubular quartz micro-reactor.
Approximately, 0.1 g oꢀ catalyst (32–42 mesh oꢀ granu-
lometry) was disposed in a ꢁxed bed supported by quartz
wool. The reactor was heated under nitrogen ꢂow at a rate
373
473
573
673
773
873
973
1073
−
1
oꢀ 10 K min . Nitrogen was ꢀed by a mass ꢂow control-
ler (Sierra Instruments), and a solution oꢀ glycerol in water
Temperature/K
−
1
(
10% w w ) was ꢀed through a syringe-type pump (KD
Fig. 1 Mass variation obtained during the TG analysis
−
1
Scientiꢁc) at a ꢂow rate oꢀ 0.4 mL h . The catalytic tests
were carried out at a temperature oꢀ 773 K during 5 h. The
gaseous products oꢀ the reaction were analyzed every 1 h by
online gas chromatography (Varian 3600cx) using a Pora-
pak Q packed column at 473 K and nitrogen as the carrier
gas. Flame ionization (FID) and thermal conductivity (TCD)
detectors were used. The liquid products oꢀ the reaction were
continuously condensed and collected every 1 h oꢀ reaction
2
lowest speciꢁc surꢀace area (48 m g). The pH oꢀ precipita-
tion had also a signiꢁcant inꢂuence on the speciꢁc surꢀace
area oꢀ the obtained material. Samples prepared at pH 5
presented a larger speciꢁc surꢀace area than the samples pre-
pared at pH 7 independently oꢀ the used precipitating agent.
Samples prepared at pH 5 have a higher speciꢁc surꢀace
area than those prepared at pH 7 due to the type oꢀ precipi-
tate ꢀormed. In the case oꢀ precipitation perꢀormed at pH 7,
the precipitate ꢀormed is poorly crystallized leading to the
ꢀ
or ꢀurther analysis on a Varian 3600cx gas chromatograph
[
26, 30].
ꢀ
ormation oꢀ a pseudo-boehmite. For precipitation at pH 5,
Results and discussion
an amorphous precipitate is obtained which gives a micro-
crystalline boehmite oꢀ greater speciꢁc area. In addition,
the sample with larger speciꢁc surꢀace area (A7AP3) also
presented the largest pore volume.
Catalyst characterization
Table 1 presents the prepared samples, the nomenclature
adopted ꢀor each sample and also the results obtained ꢀor the
BET speciꢁc surꢀace area, average pore diameter and pore
volume. It is observed that among the inꢂuences oꢀ the pre-
cipitating agent, the sample synthesized with KOH (A7AP3)
Figure 1 shows the mass variation ꢀor the prepared sam-
ples during the thermogravimetric analysis (TGA). It can
be observed that, with the exception sample A7AP12, all
samples had the highest percentage oꢀ mass loss up to 473 K
oꢀ heating. This initial loss is linked to the removal oꢀ water
molecules ꢀrom the material. The decomposition ꢀor the
samples prepared with sodium carbonate occurred around
2
resulted in the highest speciꢁc surꢀace area (432 m g),
whereas the sample prepared with Na CO resulted in the
2
3
1
3

D. S. Lima, O. W. Perez-Lopez
5
73 K, whereas ꢀor the samples prepared with hydroxide the
observed that the A5AP2 sample showed a diferent behav-
ior with a rapid uptake at a low relative pressure (0–0.4 bar),
which indicates the presence oꢀ micropores in the structure.
This result is conꢁrmed by the pore size distribution shown
in Fig. 3ꢀ, where the A5AP2 sample has a narrow pore
distribution with most pores having diameters oꢀ less than
5 nm. A similar behavior was also observed in the material
developed by Chang et al. [34]. In addition, the shape oꢀ the
hysteresis indicates pores with narrow neck and broad body
as presented by Cychosz and Thommes [35].
decomposition extends to higher temperatures, above 673 K.
Figure 2 shows the nitrogen adsorption–desorption iso-
therms ꢀor all samples. It is observed that all samples pre-
sented type IV hysteresis (according to IUPAC classiꢁca-
tion) which is characteristic oꢀ a mesoporous material, as
was also observed in other studies [32, 33].
Figure 3 shows the pore size distribution results ꢀor all
samples, whereas Table 1 shows the average pore diameter
values. These results demonstrated that these materials are
mesoporous. The A7AP3 sample exhibits a bimodal pore
size distribution, which is in agreement with the high spe-
ciꢁc surꢀace area ꢀor this sample. However, ꢀrom Fig. 2ꢀ it is
Figure 4 shows the results oꢀ the X-ray difraction ꢀor the
synthesized samples. With the exception oꢀ sample A7AP1,
all samples presented reꢂections at 37.2°, 45.7° and 66.8°
A7AP1
A7AP2
(a)
(b)
100
350
9
8
7
6
5
4
3
2
1
0
0
0
0
0
0
0
0
0
0
3
00
250
00
2
Adsorption
Desorption
150
100
Adsorption
Desorption
5
0
0
0
0.2
0.2
0.2
0.4
0.6
0.8
1
1
1
0
0
0
0.2
0.2
0.2
0.4
0.6
0.8
1
1
1
Relative pressure, P/Po
Relative pressure, P/Po
A7AP12
A7AP3
(c)
(d)
3
3
2
2
1
1
50
00
50
00
50
00
500
4
4
50
00
350
3
2
2
00
50
00
Adsorption
Desorption
Adsorption
Desorption
150
00
1
50
50
0
0
0
0.4
0.6
0.8
0.4
0.6
0.8
Relative pressure, P/Po
Relative pressure, P/Po
A5AP1
A5AP2
(e)
(f)
3
2
2
1
1
00
50
00
50
00
200
1
1
1
1
80
60
40
20
100
Adsorption
Desorption
Adsorption
Desorption
80
60
4
2
0
0
0
50
0
0
0.4
0.6
0.8
0.4
0.6
0.8
Relative pressure, P/Po
Relative pressure, P/Po
Fig. 2 Nitrogen adsorption–desorption isotherms ꢀor samples calcined at 873 K: a A7AP1, b A7AP2, c A7AP12, d A7AP3, e A5AP1 and f
A5AP2
1
3

Preparation of alumina with diꢀerent precipitants for the gas phase dehydration of glycerol…
(a)
(b)
A7AP1
A7AP2
1
0
0
0
0
0
0
0
0
0
0
,0
,9
,8
,7
,6
,5
,4
,3
,2
,1
,0
1,0
0
0
0
0
,8
,6
,4
,2
0,0
0
5
10
15
20
25
30
35
0
0
0
5
5
5
10
10
10
15
20
25
25
25
30
30
30
35
Pore diameter/nm
Pore diameter/nm
(c)
A7AP12
(d)
A7AP3
1
1
0
0
0
0
,2
,0
,8
,6
,4
,2
1
0
0
0
0
0
,0
,8
,6
,4
,2
,0
0,0
0
5
10
15
20
25
30
35
15
20
35
Pore diameter/nm
A5AP1
Pore diameter/nm
A5AP2
(e)
(f)
1
,0
,9
,8
,7
,6
,5
,4
,3
,2
,1
,0
1
,2
,0
,8
,6
,4
,2
,0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
5
10
15
20
25
30
35
15
20
35
Pore diameter/nm
Pore diameter/nm
Fig. 3 Pore size distribution ꢀor samples: a A7AP1, b A7AP2, c A7AP12, d A7AP3, e A5AP1 and f A5AP2
1
3

D. S. Lima, O. W. Perez-Lopez
*
Al O
2
3
ammonia in this temperature range demonstrates that these
materials have weak and moderate acidity and no strong acid
sites, which was also observed by Zhong et al. [19].
For a better analysis oꢀ the results, deconvolution oꢀ the
*
*
*
obtained NH -TPD proꢁles was perꢀormed using the Gauss-
3
A7AP12
ian method. The results oꢀ this deconvolution can be seen in
Fig. 7 and the numerical data in Table 2.
A7AP3
A5AP2
A7AP2
The deconvolution results show the existence oꢀ two types
oꢀ sites ꢀor all the samples. The weak acid sites are charac-
terized by the peak in the lower-temperature region, between
4
83 and 513 K, and the moderate acid sites, represented by
the peak at higher temperatures, above 573 K. It is noted
that both the precipitating agent and the pH inꢂuenced the
acidic properties oꢀ the obtained material. Among the mate-
rials prepared at pH 7.0, the A7AP2 sample had the highest
density oꢀ total acid sites, these being predominantly weak
acid sites. In addition, the acid strength oꢀ both types oꢀ sites
was higher since the desorption temperatures in the decon-
volution were higher. The A7AP1 sample had the opposite
behavior, resulting in a lower density oꢀ total acid sites and
a predominant presence oꢀ moderate acid sites. This same
behavior was observed ꢀor the samples prepared at pH 5.0,
in which the A5AP2 sample had a higher total acidity, con-
sisting predominantly oꢀ weak acid sites, than the A5AP1
sample. It is evident that the precipitating agent used in the
synthesis has inꢂuence on the acidity oꢀ the ꢀormed material.
It was observed that the use oꢀ NaOH as a precipitant led to
the ꢀormation oꢀ materials with a higher acidity than those
obtained ꢀrom Na CO .
A5AP1
A7AP1
1
0
20
30
40
θ/°
50
60
70
2
Fig. 4 X-ray difractograms ꢀor samples calcined at 873 K
corresponding to transitional low-temperature alumina,
indicating that the used method was efective to obtain the
alumina materials [3, 13, 15, 21, 36]. The more amorphous
structure ꢀor the A7AP1 sample may explain the ꢀact that
this material had a signiꢁcant lower surꢀace area than the
other samples.
Figure 5 shows the SEM images ꢀor all the samples. There
are diferences in the morphology oꢀ the obtained materials.
The A7AP1 sample has a smooth surꢀace compatible with
the low surꢀace area and porosity oꢀ this material (Table 1).
In Fig. 5b, d, micrographs oꢀ the A7AP2 and A7AP3 sam-
ples, respectively, it is noted that the grain was ꢀormed ꢀrom
overlapping ꢂakes, which is a classical grain morphology oꢀ
pure alumina materials. In the case oꢀ the A7AP12 (Fig. 5c)
and A5AP1 (Fig. 5e) samples, each particle is an agglomer-
ate oꢀ crystals, most oꢀ them platy, with sharp edges. This
morphology is more evident ꢀor sample A5AP1. This pecu-
liar shape can be compared to that oꢀ a lettuce. The A5AP2
sample (Fig. 5ꢀ) has a ꢂat surꢀace with the presence oꢀ very
small pores which is in agreement with the pore size distri-
bution presented in Fig. 3ꢀ.
2
3
Catalytic activity
Figure 8 shows the conversion oꢀ glycerol during 5 h oꢀ
reaction at the temperature oꢀ 773 K ꢀor all the samples. The
obtained glycerol conversion values were higher than 80%
ꢀor all the samples during the time-on-stream.
For samples prepared at pH 7.0 with diferent precipi-
tants, it is observed that the glycerol conversion is inꢂu-
enced by both the speciꢁc surꢀace area and the acidity oꢀ
the catalyst. The lowest conversion obtained ꢀor the sam-
ple A7AP1 can be related to the low speciꢁc surꢀace area
and also to the lower acidity, when compared to the other
samples. In addition, the samples A7AP1 and A7AP12,
which presented the lowest values oꢀ density oꢀ acid sites,
were those that resulted in the lowest values oꢀ conversion,
Figure 6 shows the NH -TPD results ꢀor all the synthe-
3
sized samples. A broad desorption peak is observed ꢀor
all samples in the range oꢀ 423–723 K. The desorption oꢀ
1
3

Preparation of alumina with diꢀerent precipitants for the gas phase dehydration of glycerol…
Fig. 5 SEM images ꢀor the sam-
ples: a A7AP1 with magniꢁca-
tion oꢀ 5000 X; b A7AP2 with
magniꢁcation oꢀ 10,000 X; c
A7AP12 with magniꢁcation oꢀ
5
000 X; d A7AP3 with magni-
ꢁ
cation oꢀ 10,000 X; e A5AP1
with magniꢁcation oꢀ 5000 X;
and f A5AP2 with magniꢁca-
tion oꢀ 5000 X
indicating that more acidic materials give better results in
the catalytic conversion oꢀ glycerol, as it was also observed
by Kim et al. [37]. It is observed that alumina prepared using
NaOH (A7AP2) showed the best conversion oꢀ glycerol,
maintaining approximately 99% in the ꢁrst 3 h oꢀ reaction,
and remaining with a conversion oꢀ approximately 97% at
the end oꢀ 5 h oꢀ reaction. Samples prepared at pH 5.0 had
higher speciꢁc area and higher acidity and, consequently,
they showed a higher conversion than the respective samples
prepared at pH 7.0. It may be ꢀurther noted that the acidity oꢀ
the materials has an important efect on glycerol conversion,
which has also been observed by Pathak et al. [11].
These results can be considered promising considering
that alumina is used as support in most oꢀ the works [12, 15].
In addition, the conversion values obtained in the present
1
3

D. S. Lima, O. W. Perez-Lopez
led to the ꢀormation oꢀ hydroxyacetone, and this process
is catalyzed by Lewis acid sites. The A7AP1 sample also
showed high selectivity ꢀor hydroxyacetone (40%). On the
other hand, samples A7AP2 and A7AP3 were prominent
in the ꢀormation oꢀ light oleꢁns as byproducts, which were
identiꢁed as ethylene and propylene, as shown in Fig. 9.
In addition, the lower selectivity oꢀ the A7AP2 sample to
acrolein may be due to a more severe coking due to the
A7AP12
A7AP3
ꢀ
act that this material presents acid sites with higher acidic
A5AP2
A7AP2
A5AP1
strength [41]. The A5AP1 sample prepared with Na CO
2
3
at pH 5.0 also showed good selectivity in the ꢀormation oꢀ
hydroxyacetone and oleꢁns.
By analyzing the conversion results (Fig. 8) together with
the distribution oꢀ gaseous products (Fig. 9), all the sam-
ples showed good selectivity ꢀor the ꢀormation oꢀ acrolein
and hydroxyacetone, the latter with values between 30 and
5
0%. Furthermore, the inꢂuence oꢀ the pH oꢀ the reaction
A7AP1
medium, as well as the precipitating agent, on the yield oꢀ
the products is observed. Among the samples prepared at pH
3
73
473
573
673
773
873
Temperature/K
7
.0, samples A7AP1 and A7AP12 presented the best results
oꢀ acrolein and hydroxyacetone. For samples prepared at
pH 5.0, the sample A5AP2 synthesized with NaOH as the
precipitating agent presented the best dehydration products.
In contrast, the samples A7AP2 (NaOH) and A7AP3 (KOH),
which were prepared at pH 7.0, were also ꢀound to have a
good yield in light oleꢁns as byproducts oꢀ the reaction.
Sample A5AP1, which was synthesized with Na2CO3 at
pH 5.0, also showed higher ꢀormation oꢀ these byproducts.
Among the samples prepared with NaOH at diferent pH,
the higher production oꢀ light oleꢁns ꢀor the A7AP2 sample
may be due to the ꢀact that it has acid sites with higher acidic
strength than the A5AP2 sample, because the desorption
temperatures ꢀor A7AP2 were higher (Table 2).
Fig. 6 NH -TPD curves ꢀor samples calcined at 873 K
3
work are approximate to those already reported in the litera-
ture. According to the work oꢀ Zakaria et al. [38], the con-
version oꢀ glycerol to oleꢁns resulted in conversion values
oꢀ 77; 77.4 and 79.6% when Al O was used as a catalyst.
2
3
Further, in the study perꢀormed by Haider et al. [39], total
conversion in the glycerol dehydration was achieved using
ceria and zirconia graꢀting onto alumina (α and θ–δ phases)
as supports ꢀor silicotungstic acid, producing 85% selectiv-
ity to acrolein.
Figure 9 shows the average distribution oꢀ gaseous
products obtained aꢀter 5 h oꢀ reaction. It is observed that
the sample A5AP2 exhibited the highest values ꢀor acr-
olein (30%) and hydroxyacetone (40%), revealing high
selectivity ꢀor glycerol dehydration; the remaining 30%
are composed oꢀ light oleꢁns and other products ꢀormed
by secondary reactions. This result may be associated with
the high density oꢀ acid sites oꢀ this material (Table 2),
as well as to a larger amount oꢀ Lewis acid sites being
converted to Bronsted acid sites under the reaction con-
ditions interacting with vapor present in the system as
already observed in other studies in the literature [15,
Characterization after the reaction
Table 3, as well as Fig. 10, shows the results oꢀ the TPO
analysis ꢀor all the samples aꢀter the reactions.
The mass loss up to 473 K is due to the removal oꢀ
moisture and/or adsorbed volatile compounds, as shown in
Fig. 10a. From 473 K, samples A7AP2 and A7AP3 showed
a greater mass loss during the oxidation, which indicates
higher carbon ꢀormation during the catalytic tests. This
result may be related to high acidity, which provides more
coking reactions as well as greater oleꢁn ꢀormation. The
A5AP2 sample, which had the best perꢀormance in glycerol
dehydration, obtained the lowest carbon ꢀormation. This is
due to its higher percentage oꢀ weak acid sites (57%) shown
in Table 2. In addition, these results also show that the
4
0]. These products were also obtained in the study oꢀ
glycerol dehydration over Al O perꢀormed by Massa et al.
2
3
[
15], which obtained acrolein as the desired product and
hydroxyacetone as the main byproduct. The explanation
or this result was that the monodehydration oꢀ glycerol
ꢀ
1
3

Preparation of alumina with diꢀerent precipitants for the gas phase dehydration of glycerol…
Fig. 7 Deconvolution oꢀ the
A7AP2
A7AP1
NH -TPD curves ꢀor the pre-
3
pared samples
3
73
473
573
673
773
873
373
473
573
673
773
873
Temperature/K
Temperature/K
A7AP3
A7AP12
3
73
473
573
673
773
873
373
473
573
673
773
873
Temperature/K
Temperature/K
A5AP2
A5AP1
3
73
473
573
673
773
873
373
473
573
673
773
873
Temperature/K
Temperature/K
Table 2 Deconvolution oꢀ the NH -TPD curves ꢀor the catalysts
secondary reactions are responsible ꢀor the greater ꢀorma-
tion oꢀ coke during the process.
3
Catalyst Temperature/K
Relative ꢀraction oꢀ the Total
total sites/% acid sites/
From the oxidation temperatures shown in Fig. 10b, it is
observed that all samples ꢀormed graphitic carbon which
oxidized at higher temperatures, except ꢀor samples A7AP1
and A7AP12. The sample A7AP1 led to the ꢀormation oꢀ
light carbon, whereas the sample A7AP12 indicates the ꢀor-
mation oꢀ light carbon and graphitic carbon.
µmol g−
1
cat
1
st peak 2nd peak Weak acid
sites
Medium
acid sites
A7AP1
A7AP2
A7AP12 489.2
A7AP3
A5AP1
A5AP2
481.8
514.6
573.2
622.9
585.4
573.1
590.4
614.1
42.9
53.1
47.6
41.3
43.3
57.3
57.1
46.9
52.4
58.7
56.7
42.7
532.3
756.9
595.5
720.3
636.0
855.1
486.4
497.0
497.9
1
3

D. S. Lima, O. W. Perez-Lopez
Table 3 TPO results ꢀor all the samples aꢀter the reactions
1
00
A7AP2
A5AP1
A7AP3
9
9
9
9
9
8
8
8
8
8
8
6
4
2
0
8
6
4
2
0
Catalyst Mass loss/%
Total
mass
DTA peak
A5AP2
temperature/K
loss/%
Up to 473 K 473 to
1st peak 2nd peak
A7AP12
973 K
A7AP1 3.3
A7AP2 2.1
A7AP12 9.2
A7AP3 3.7
A5AP1 1.7
A5AP2 1.1
14.7
21.8
16.7
23.7
18.5
13.3
18.0
23.9
25.9
27.4
20.2
14.4
575
–
573
–
–
–
653
739
739
750
744
760
A7AP1
5
0
100
150
200
250
300
smallest speciꢁc surꢀace area. Samples prepared at a lower
pH resulted in materials with a higher speciꢁc area. All sam-
ples showed poor crystalline XRD pattern oꢀ transitional
low-temperature alumina, except ꢀor the sample prepared
with Na CO at pH=7, which was almost amorphous.
Time/min
Fig. 8 Conversion oꢀ glycerol versus reaction time
2
3
Samples precipitated with NaOH presented higher den-
sity oꢀ acid sites than samples prepared with Na2CO3, inde-
pendently oꢀ precipitation pH, whereas the sample prepared
with both precipitants showed an intermediate density oꢀ
acid sites. The strength oꢀ acid sites was predominantly weak
and moderate ꢀor all samples.
Conclusions
Mesoporous aluminas with diꢀꢀerent properties were
obtained by varying the precipitating agent and the pH oꢀ
precipitation. The speciꢁc surꢀace area, the crystallinity and
the acidic properties oꢀ the samples were dependent on both
the pH and the precipitating agent used.
The catalytic properties oꢀ the prepared aluminas ꢀor
the glycerol dehydration are mainly related to the speciꢁc
surꢀace area and to acidic characteristics. Conversions oꢀ
glycerol above 85% were obtained ꢀor all samples. A lower
The sample prepared with KOH had the largest BET
area, whereas the sample prepared with Na CO had the
2
3
Fig. 9 Distribution oꢀ
60.00
gaseous products ꢀor all
catalysts aꢀter 5 h oꢀ time-
on-stream (T=773 K; ꢂow
5
4
3
2
1
0.00
0.00
0.00
0.00
0.00
−
1
rate=0.4 mL h
)
0.00
A7AP1
A7AP2
A7AP12
A7AP3
A5AP2
A5AP1
Methane
Ethylene
Propylene
Acrolein
Hydroxyacetone
Other products
1
3

Preparation of alumina with diꢀerent precipitants for the gas phase dehydration of glycerol…
(
a)
References
1
00
A7AP1
A7AP2
A7AP12
A7AP3
A5AP2
A5AP1
1
.
Jang HS, Bae K, Shin M, Kim SM, Kim CU, Suh YW. Aromati-
zation oꢀ glycerol/alcohol mixtures over zeolite H-ZSM-5. Fuel.
9
9
8
8
7
7
5
0
5
0
5
0
2
014;134:439–47. https://doi.org/10.1016/j.ꢀuel.2014.05.086.
2
.
Ferreira P, Fonseca IM, Ramos AM, Vital J, Castanheiro JE. Acet-
ylation oꢀ glycerol over heteropolyacids supported on activated
carbon. Catal Commun. 2011;12:573–6. https://doi.org/10.1016/j.
catcom.2010.11.022.
3
4
.
.
Liu L, Ye XP, Bozell JJ. A comparative review oꢀ petroleum-
based and bio-based acrolein production. ChemSusChem.
2
012;5:1162–80.
Talebian-Kiakalaieh A, Amin NAS, Hezaveh H. Glycerol ꢀor
renewable acrolein production by catalytic dehydration. Renew
Sustain Energy Rev. 2014;40:28–59. https://doi.org/10.1016/j.
rser.2014.07.168.
3
73
473
573
673
773
873
973
1073
5
.
Talebian-Kiakalaieh A, Amin NAS. Thermo-kinetic and difu-
sion studies oꢀ glycerol dehydration to acrolein using HSiW-
Γ-Al O supported ZrO solid acid catalyst. Renew Energy.
Temperature/K
2
3
2
2
017;114:794–804.
(
b)
6
7
.
.
Katryniok B, Paul S, Bellière-Baca V, Rey P, Dumeignil F. Glyc-
erol dehydration to acrolein in the context oꢀ new uses oꢀ glycerol.
Green Chem. 2010;12:2079–98.
A7AP1
Ma T, Ding J, Shao R, Xu W, Yun Z. Dehydration oꢀ glycerol
to acrolein over Wells-Dawson and Keggin type phospho-
tungstic acids supported on MCM-41 catalysts. Chem Eng J.
A7AP2
A7AP12
A7AP3
A5AP2
2
017;316:797–806. https://doi.org/10.1016/j.cej.2017.02.018.
8
9
.
.
Suprun W, Lutecki M, Haber T, Papp H. Acidic catalysts ꢀor the
dehydration oꢀ glycerol: activity and deactivation. J Mol Catal A:
Chem. 2009;309:71–8.
A5AP1
de Oliveira AS, Vasconcelos SJS, de Sousa JR, de Sousa FF,
Filho JM, Oliveira AC. Catalytic conversion oꢀ glycerol to acr-
olein over modiꢁed molecular sieves: activity and deactivation
studies. Chem Eng. 2011;168:765–74. https://doi.org/10.1016/j.
cej.2010.09.029.
273 373 473 573 673 773 873 973 1073
1
0. Decolatti HP, Dalla Costa BO, Querini CA. Dehydration oꢀ glyc-
erol to acrolein using H-ZSM5 zeolite modiꢁed by alkali treatment
with NaOH. Microporous Mesoporous Mater. 2015;204:180–9.
https://doi.org/10.1016/j.micromeso.2014.11.014.
Temperature/K
Fig. 10 Mass variation (a) and temperature diference (b) during the
1
1
1. Pathak K, Reddy KM, Bakhshi NN, Dalai AK. Catalytic conver-
sion oꢀ glycerol to value added liquid products. Appl Catal A Gen.
TPO analysis ꢀor samples aꢀter the reaction
2
010;372:224–38.
2. Danov S, Esipovich A, Belousov A, Rogozhin A. Gas-phase dehy-
dration oꢀ glycerol over commercial Pt/γ-Al O catalysts. Chin
glycerol conversion was obtained ꢀor samples with a lower
surꢀace area and lower density oꢀ acid sites. The selectivity
2
3
J Chem Eng. 2015;23:1138–46. https://doi.org/10.1016/j.cjche
.
2015.03.012.
ꢀ
or glycerol dehydration was strongly related to the amount
1
1
3. Zhang L, Wu Y, Zhang L, Wang Y, Li M. Synthesis and char-
acterization oꢀ mesoporous alumina with high speciꢁc area via
coprecipitation method. Vacuum. 2016;133:1–6.
and strength oꢀ acid sites. The best result ꢀor dehydration
was obtained ꢀor the sample prepared with NaOH and pre-
cipitated at pH=5. These results are related to the higher
speciꢁc surꢀace area, greater amount oꢀ acid sites and the
higher ratio oꢀ weak acid sites.
4. Liu L, Wang B, Du Y, Zhong Z, Borgna A. Biꢀunctional Mo VO /
3
x
H SiW O /Al O catalysts ꢀor one-step conversion oꢀ glycerol
4
12 40
2
3
to acrylic acid: catalyst structural evolution and reaction path-
ways. Appl Catal B Environ. 2015;174–175:1–12. https://doi.
org/10.1016/j.apcatb.2015.02.032.
Samples that showed higher carbon ꢀormation also
showed a higher production oꢀ light oleꢁns, indicating that
the carbon production is related to the ꢀormation oꢀ these
byproducts. These results are related to the higher acidic
strength oꢀ the medium acid sites.
1
1
1
5. Massa M, Andersson A, Finocchio E, Busca G. Gas-phase
dehydration oꢀ glycerol to acrolein over Al O -, SiO -,
2
3
2
and TiO -supported Nb- and W-oxide catalysts. J Catal.
2
2013;307:170–84.
6. Yoshida MI, Silva VR, Pinto PCC, Sant’Anna SS, Silva MC, Car-
valho CF. Physico-chemical characterization and thermal analysis
data oꢀ alumina waste ꢀrom Bayer process. J Therm Anal Calorim.
Acknowledgements The authors are grateꢀul ꢀor the ꢁnancial support
provided by CAPES (Brazilian Agency ꢀor Improvement oꢀ Graduate
Personnel) and CNPq (National Council oꢀ Science and Technological
Development). The authors wish to thank LACER Laboratory ꢀor the
SEM analysis.
2
012;109:1429–33.
7. Kucza W, Obłąkowski J, Gajerski R, Łabuś S, Danielewski M,
Małecki A, et al. Synthesis and characterization oꢀ alumina- and
1
3

D. S. Lima, O. W. Perez-Lopez
zirconia-based powders obtained by the ultrasonic spray pyrolysis.
J Therm Anal Calorim. 2007;88:65–9.
30. Lima DS, Calgaro CO, Perez-Lopez OW. Hydrogen production
by glycerol steam reꢀorming over Ni based catalysts prepared by
diferent methods. Biomass Bioenergy. 2019;130:105358. https://
doi.org/10.1016/j.biombioe.2019.105358.
1
1
2
8. Liu Q, Wang A, Wang X, Zhang T. Morphologically controlled
synthesis oꢀ mesoporous alumina. Microporous Mesoporous
Mater. 2007;100:35–44.
31. Lima DS, Perez-Lopez OW. Conversão catalítica do etanol sobre
catalisadores suportados em ZSM-5. Ceramica. 2018;64:1–9.
32. Yaakob Z, Bshish A, Ebshish A, Tasirin SM, Alhasan FH. Hydro-
gen production by steam reꢀorming oꢀ ethanol over nickel catalysts
supported on sol gel made alumina: inꢂuence oꢀ calcination tem-
perature on supports. Materials. 2013;6:2229–39.
9. Zhong L, Zhang Y, Chen F, Zhang Y. Synthesis oꢀ mesoporous
alumina using a recyclable methylcellulose template. Microporous
Mesoporous Mater. 2011;142:740–4.
0. Shang X, Wang X, Nie W, Guo X, Zou X, Ding W, et al. A tem-
plate-ꢀree synthesis oꢀ mesoporous crystalline γ-alumina through
partial hydrolysis oꢀ aluminum nitrate aqueous solution. Mater
Lett. 2012;83:91–3.
33. Lee G, Kim I, Yang I, Ha JM, Na HB, Jung JC. Efects oꢀ the
preparation method on the crystallinity and catalytic activ-
2
2
2
1. Huang B, Bartholomew CH, Smith SJ, Woodꢁeld BF. Facile sol-
vent-deꢁcient synthesis oꢀ mesoporous γ-alumina with controlled
pore structures. Microporous Mesoporous Mater. 2013;165:70–8.
https://doi.org/10.1016/j.micromeso.2012.07.052.
ity oꢀ LaAlO perovskites ꢀor oxidative coupling oꢀ methane.
3
Appl Surꢀ Sci. 2018;429:55–61. https://doi.org/10.1016/j.apsus
c.2017.08.092.
34. Chang G, Shang Z, Yu T, Yang L. Rational design oꢀ a novel
indole-based microporous organic polymer: enhanced carbon
dioxide uptake via local dipole-π interactions. J Mater Chem A R
Soc Chem. 2016;4:2517–23.
2. Wu W, Wan Z, Chen W, Zhu M, Zhang D. Synthesis oꢀ
mesoporous alumina with tunable structural properties.
Microporous Mesoporous Mater. 2015;217:12–20. https://doi.
org/10.1016/j.micromeso.2015.06.002.
35. Cychosz KA, Thommes M. Progress in the physisorption char-
acterization oꢀ nanoporous gas storage materials. Engineering.
2018;4:559–66. https://doi.org/10.1016/j.eng.2018.06.001.
36. He X, Sun J, Xu X, Lv Z, Song J. Thermal analysis oꢀ phos-
phorus-modiꢀied boehmite nanosheets and isoelectric points
(IEP) oꢀ the corresponding γ-alumina. J Therm Anal Calorim.
2017;130:2249–55.
3. Aꢀshar Taromi A, Kaliaguine S. Synthesis oꢀ ordered
mesoporous Γ-alumina—efects oꢀ calcination conditions and
polymeric template concentration. Microporous Mesoporous
Mater. 2017;248:179–91. https://doi.org/10.1016/j.micro
meso.2017.04.040.
2
2
2
4. Ipek M, Toplan N, Toplan HO. Transꢀormation kinetics oꢀ θ- to
α-phase oꢀ alumina powders prepared ꢀrom diferent alumina salts
by chemical processing. J Therm Anal Calorim. 2016;125:645–9.
5. Masiero SS, Marcilio NR, Perez-Lopez OW. Aromatiza-
tion oꢀ methane over Mo-Fe/ZSM-5 catalysts. Catal Lett.
37. Kim YT, Jung KD, Park ED. Gas-phase Dehydration oꢀ glycerol
over supported silicotungstic acids catalysts. Bull Korean Chem
Soc. 2010;31:3283–90.
38. Zakaria ZY, Amin NAS, Linnekoski J. A perspective on cata-
lytic conversion oꢀ glycerol to oleꢀins. Biomass Bioenerg.
2013;55:370–85. https://doi.org/10.1016/j.biombioe.2013.02.014.
39. Haider MH, D’Agostino C, Dummer NF, Mantle MD, Gladden LF,
Knight DW, et al. The efect oꢀ graꢀting zirconia and ceria onto
alumina as a support ꢀor silicotungstic acid ꢀor the catalytic dehy-
dration oꢀ glycerol to acrolein. Chem A Eur J. 2014;20:1743–52.
40. Wang F, Dubois JL, Ueda W. Catalytic perꢀormance oꢀ vana-
dium pyrophosphate oxides (VPO) in the oxidative dehydration
oꢀ glycerol. Appl Catal A Gen. 2010;376:25–32. https://doi.
org/10.1016/j.apcata.2009.11.031.
2
009;131:194–202.
6. Lima DS, Perez-Lopez OW. Catalytic conversion oꢀ glycerol to
oleꢁns over Fe, Mo, and Nb catalysts supported on zeolite ZSM-5.
Renew Energy. 2019;136:828–36. https://doi.org/10.1016/j.renen
e.2019.01.051.
2
7. Rosset M, Sꢀreddo LW, Hidalgo GEN, Perez-Lopez OW, Féris
LA. Adsorbents derived ꢀrom hydrotalcites ꢀor the removal oꢀ
dicloꢀenac in wastewater. Appl Clay Sci. 2019;175:150–8. https
:
//doi.org/10.1016/j.clay.2019.04.014.
2
2
8. Denardin F, Perez-Lopez OW. Tuning the acidity and reducibility
oꢀ Fe/ZSM-5 catalysts ꢀor methane dehydroaromatization. Fuel.
41. Chai SH, Wang HP, Liang Y, Xu BQ. Sustainable production oꢀ
acrolein: investigation oꢀ solid acid-base catalysts ꢀor gas-phase
dehydration oꢀ glycerol. Green Chem. 2007;9:1130–6.
2
019;236:1293–300. https://doi.org/10.1016/j.ꢀuel.2018.09.128.
9. Calgaro CO, Perez-Lopez OW. Biogas dry reꢀorming ꢀor hydro-
gen production over Ni-M-Al catalysts (M=Mg, Li, Ca, La, Cu
Co, Zn). Int J Hydrogen Energy. 2019;44:17750–66. https://doi.
org/10.1016/j.ijhydene.2019.05.113.
Publisher’s Note Springer Nature remains neutral with regard to
jurisdictional claims in published maps and institutional aꢃliations.
1
3