Вы находитесь на странице: 1из 10

Applied Catalysis A: General 401 (2011) 46–55

Contents lists available at ScienceDirect

Applied Catalysis A: General


journal homepage: www.elsevier.com/locate/apcata

Study on the preparation of Cu/ZnO catalyst by sol–gel auto-combustion method


and its application for low-temperature methanol synthesis
Lei Shi a , Kai Tao a , Ruiqin Yang b , Fanzhi Meng a , Chuang Xing a , Noritatsu Tsubaki a,c,∗
a
Department of Applied Chemistry, School of Engineering, University of Toyama, Gofuku 3190, Toyama 930-8555, Japan
b
Zhejiang Provincial Key Lab. for Chem. & Bio. Processing Technology of Farm Products, School of Biological and Chemical Engineering, Zhejiang University of Science &
Technology, Hangzhou 310023, PR China
c
JST, CREST, Sanbancho 5, Chiyoda-ku, Tokyo 102-0075, Japan

a r t i c l e i n f o a b s t r a c t

Article history: A series of the as-burnt and the burnt Cu/ZnO catalysts were prepared by a sol–gel auto-combustion
Received 9 December 2010 method using metal nitrates with the mole ratio of Cu/Zn = 1/1 (noted as M, M = Cu + Zn) and citric acid
Received in revised form 16 April 2011 (noted as CA). When the xerogels were burnt in the argon atmosphere, H2 and CH4 which were came
Accepted 27 April 2011
from the decomposition of the citric acid, were the reducing agents and were used in the redox process
Available online 10 May 2011
for synthesizing metallic Cu from Cu2+ in the chelated compound. The XRD patterns revealed that all the
as-burnt catalysts with different M/CA molar ratios were converted into pure Cu and ZnO species. TPR
Keywords:
analysis of the as-burnt catalyst illustrated that almost no hydrogen was consumed. It proved that Cu2+ in
Low-temperature
Methanol synthesis
the chelated compound was absolutely reduced to metallic Cu in the as-burnt catalyst. The effects of M/CA
Auto-combustion molar ratio on the properties of catalysts were studied by TG-DTA, FT-IR, Raman spectrum, XRD, SEM-
Cu/ZnO EDS, BET, and N2 O chemisorption techniques. The activity of the as-burnt catalysts without reduction
Metallic catalysts was investigated for low-temperature methanol synthesis from syngas containing CO2 using ethanol as a
promoter at 443 K and 5.0 MPa for 12 h. The total carbon conversion increased with increasing the content
of citric acid and reached a maximum for the as-burnt Cu/ZnO catalyst C0.8 with M/CA = 1/0.8, and then
decreased. The variation trend was in accordance with that of the copper (Cu0 ) surface area. Comparing
with the burnt catalyst C0.8-air after reduction, the methanol selectivity of the as-burnt Cu/ZnO catalyst
was much lower owing to lower hydrogenation activity of the ethyl formate. The as-burnt catalyst C0.8
was also used in continuous low-temperature methanol synthesis at 443 K and 5.0 MPa for 40 h. The
total carbon conversion was stable after 15 h and no obvious deactivation during 40 h reaction, but the
methanol selectivity was still not high.
© 2011 Elsevier B.V. All rights reserved.

1. Introduction The present authors proposed a new method of low temperature


synthesis of methanol from CO/CO2 /H2 on Cu/ZnO based catalyst
Methanol was widely used as a feedstock for the chemical indus- using alcohol as a catalyst and solvent, by which methanol was
tries and at the same time as an alternative fuel which was used in produced at 443 K and 5.0 MPa [4–7]. As the used alcohol solvent
fuel cells [1–3]. Methanol was industrially produced under high- contained a small amount of water, it was considered that this
temperature and high pressure (523–573 K, 8.0–10 MPa), mainly new low-temperature process could directly use low-grade syngas
using Cu/ZnO catalysts developed by ICI Co. However, the efficiency containing CO2 and H2 O, without purification.
of methanol synthesis was severely limited by thermodynamics It was generally known that the particle size, surface area, metal-
because methanol synthesis was an extremely exothermic reaction. lic surface area and composition of the catalyst were important
Therefore, developing a low-temperature process for methanol factors to the catalytic activity. A linear relationship between the
synthesis will greatly reduce the production cost and high one-pass catalytic activity and the metallic Cu surface area of Cu/ZnO cata-
CO conversion will become available at low temperature. lyst was proved by Chinchen et al. [8]. To increase furthermore the
activity and selectivity of Cu/ZnO based catalyst for the new low-
temperature methanol synthesis, its preparation method should
∗ Corresponding author at: Department of Applied Chemistry, School of Engineer-
be improved. In recent years, various methods were developed to
ing, University of Toyama, Gofuku 3190, Toyama 930-8555, Japan.
prepare Cu/ZnO based catalysts with large surface area and copper
Tel.: +81 76 445 6846; fax: +81 76 445 6846. surface area, which were believed to have a high catalytic activity
E-mail address: tsubaki@eng.u-toyama.ac.jp (N. Tsubaki). for methanol synthesis from CO (CO2 ) hydrogenation [6,9,10].

0926-860X/$ – see front matter © 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.apcata.2011.04.043
L. Shi et al. / Applied Catalysis A: General 401 (2011) 46–55 47

Mixing of metal nitrate and citric acid

Ammonia Solution

pH value 7, transparent blue solution

Refluxing at 353 K for 4 h

Evaporating at 343 K

Dark blue viscous gels

Drying at 393K for 40h

Porous polymeric xerogel

Calcining at 523-723 K in argon, Calcining at 523-723K in air,


auto-combustion auto-combustion

Porous and loose pure Cu/ZnO catalyst Porous and loose CuO/ZnO

Fig. 1. Schematic flow chart of the catalysts prepared by a sol–gel auto-combustion method.

Recently, the sol–gel auto-combustion technique had emerged (wt.%) to reach a pH value of 7. Stirring and refluxing at 353 K for
as an attractive method for the production of high purity and crys- 4 h ensured the citric acid completely chelating with the metal ions.
talline oxide powders [11–17]. For the first time, Jiang et al. had Subsequently, the neutralized solution was evaporated at 343 K
used citric acid–nitrate auto-combustion method to prepare pure on a hot plate with continuously stirring, until dark blue viscous
metals and alloys [18]. Wolf et al. presented a glycine–nitrate com- xerogels were formed. Then the xerogels were dried at 393 K for
bustion method to prepare nanopowders of pure nickel [19] and 40 h. Next, the xerogels was treated in two ways, respectively. One
copper–nickel alloys [20]. For the first time, we improved this novel way was that the xerogels were put into a ceramic tube oven and
technology to directly prepare metallic catalyst. And for the first the temperature was increased up to about 493 K in the argon
time we used the as-burnt catalyst in the catalytic reaction without atmosphere, leading to the ignition of the xerogels. The xerogels
further reduction. This technique was based on a chemical sol–gel burnt in a self-propagating combustion manner until all xerogels
process combined with subsequent combustion process. An aque- completely burnt out to form a loose product. After that, the tem-
ous solution containing the desired metal salts and organic fuel perature was continuously increased to 723 K and maintained for
formed a gel through a sol–gel process, and then the gel was ignited 1 h in the argon atmosphere. The as-burnt catalysts were directly
to combust, giving a voluminous and fluffy product with a large used for low-temperature methanol synthesis without hydrogen
surface area. In this process, citric acid was generally used as a reduction. Another way was that the xerogels with M/CA = 1/0.8
complexant to form a homogeneous precursor gel. (noted as C0.8-air ) was put into a muffle oven and the temperature
There were few reports [21–23] on the combustion methods of was increased up to about 493 K in the air atmosphere, leading
the preparation of CuO based catalysts for methanol synthesis and to the ignition of the xerogels. After that, the temperature was
reforming of methanol [24]. However, the reported CuO based cat- continuously increased to 723 K and maintained for 1 h in the air
alysts had to be further reduced with hydrogen at high temperature atmosphere. The precursors were then reduced by a flow of 5%
to obtain metallic Cu. The extra cost for reduction energy and engi- hydrogen in nitrogen at 523 K for 10 h and successively passivated
neering made the whole catalyst preparation process inefficient. by 1% oxygen diluted by nitrogen. As for a reference, the as-burnt
In the present work, the as-burnt Cu/ZnO catalysts were pre- C0.8 which was also reduced by a flow of 5% hydrogen in nitrogen
pared by one-step sol–gel auto-combustion method, and applied at 523 K for 10 h and successively passivated by 1% oxygen diluted
directly without hydrogen reduction in low-temperature methanol by nitrogen was noted as C0.8-reduction .
synthesis reaction from the syngas containing CO2 .

2.2. Characterization of catalysts


2. Experimental
The thermal decomposition behaviour of the xerogels was char-
2.1. Preparation of catalysts acterized by thermogravimetic and differential thermal analysis
(DTA/TGA 60, Shimadzu) at a heating rate 10 K/min from 293 to
Schematic flow chart of the catalyst prepared by a sol–gel 1173 K in the argon atmosphere.
auto-combustion method was shown in Fig. 1. Analytical-grade XRD patterns were obtained on a Rigaku RINT 2400 X-ray
Cu(NO3 )2 ·3H2 O (≥99%), Zn(NO3 )2 ·6H2 O (≥99%) and citric acid were diffractometer in the air atmosphere, using monochromatic Cu K␣
used as raw materials. The metal nitrates with the mole ratio of radiation, scanning 2 from 20◦ to 80◦ , and at 40 kV and 40 mA.
Cu/Zn = 1/1 (noted as M, M = Cu + Zn) and citric acid (noted as CA) The Cu crystalline average size was calculated by Scherrer formula:
were firstly dissolved in the distilled water according to different D = K/(ˇcos ), where D was the averaged dimension of crystallites,
M/CA molar ratios 1:0.5, 1:0.8, 1:1 and 1:1.5, denoted as C0.5 , C0.8 , K was the Scherrer constant 0.89,  was the wavelength of X-ray
C1 and C1.5 . The solutions were adjusted using ammonia of 28% (Cu K␣ = 0.154 nm), and ˇ was the width of the peak at half height.
48 L. Shi et al. / Applied Catalysis A: General 401 (2011) 46–55

Fig. 2. The flow-type batch configuration for low-temperature methanol synthesis.

Infrared spectrum was recorded in the air atmosphere with a were firstly poured into the autoclave. The air inside was purged
Shimadzu Fourier Transform Infrared Spectrophotometer (FTIR- by the reactant gas, Ar/CO/CO2 /H2 = 3.13/33/5.23/58.64, then the
8400). pressure in the reactor was raised to 5.0 MPa at 298 K and the reac-
Micro-Raman analysis of specimens was performed in the air tion took place at 443 K for 12 h. The stirring speed was fixed at
on a multichannel bench Renishaw InVia Reflex spectrometer 2000 rpm to prevent the diffusion controlled regime. A continu-
coupled with a Peltier-cooled CCD detector. The instrument was ous methanol synthesis (as shown in Fig. 2) was also conducted at
equipped with a 514 nm diode laser for excitation, with a maxi- 443 K and 5.0 MPa for 40 h, using a flow-type batch reactor with
mum output power of 100 mW and the power of the laser could 80 ml inner volume and with 3 g catalyst as well as 40 ml ethanol.
be decreased to 10% of maximal output power. The spectrum was All products were determined on GC–MS (Shimadzu GCMS 1600). A
centred to 520 cm−1 over a time of 1 s with the diffraction grat- TCD (Shimadzu GC-320) was used to analyze gaseous products and
ing of 1200 grooves/mm. The samples were scanned from 800 to a FID (Shimadzu GC-8A) was used to analyze liquid products. The
2000 cm−1 (the spectrum region that provided the most valuable CO, CO2 , total carbon conversions and the selectivity of the liquid
data on the microstructure of the amorphous carbon). All samples products were calculated as follows:
were measured five times to check the reproducibility of spectrum
(CO/Ar in feed − CO/Ar in effluent)
and to enhance the accuracy of the data by accumulation. CO conv. = 100 × (1)
(CO/Ar in feed)
TPR experiments were carried out by catalyst analyzer BELCAT-
B, using 0.035 g burnt catalysts with 5% hydrogen diluted by (CO2 /Ar in feed − CO2 /Ar in effluent)
CO2 conv. = 100 × (2)
argon at atmospheric pressure. Before reduction, the catalysts were (CO2 /Ar in feed)
heated at 323 K in flowing argon for 2 h. Then, the 5% hydrogen
a b
(50 ml/min) was passed over the catalyst sample, and the tem- Total carbon conv. = CO conv. × + CO2 conv. × (3)
perature was linearly raised from 298 K to 723 K at a heating rate (a + b) (a + b)
of 5.0 K/min. Water, formed during the reduction, was trapped by (a, b were the contents of CO, CO2 in the feed gas)
molecular sieves 3A. The effluence gas was analyzed by a thermal The selectivity of the liquid products was calculated as follows:
conductivity detector (TCD), with argon used as a reference.
Pi yield
The physical structures of the catalysts, as well as compositional Pi selectivity = 100 ×  (4)
analysis and surface morphology observation, were measured by Pi yield
scanning electron microscopy (SEM) (JEOL, JSM-6360 LV) with (Pi was the mol number of each liquid product)
energy-diffusive X-ray spectroscopy (EDX) (JEOL, JED-2300). The
samples for this analysis were firstly coated with a platinum layer 3. Results and discussion
on the surface by an auto fine coater (JEOL, JFC-1600).
The copper (Cu0 ) surface area (SCu ) was determined by a nitrous 3.1. TG-DTA analysis of the citric acid and the xerogels with
oxide (N2 O) pulse method as described elsewhere [6]. different M/CA molar ratios to study the auto-combustion process
The surface area of the catalysts was determined by N2 physical
desorption at 77 K, using a Micromeritics NOVA 1000 SP surface Thermal behaviour of the xerogels with different molar ratios
area and porosimetry analyzer. Samples were degassed at 473 K of metal nitrates (noted as M, M = Cu + Zn) to citric acid (noted
prior to acquiring the adsorption isotherm. The surface areas were as CA) was investigated by TG-DTA measurements in the argon
calculated according to the BET method, using 0.164 nm2 as the atmosphere and in the air atmosphere. The results of the TG-DTA
cross-sectional area of the nitrogen molecule. analysis were shown in Fig. 3. The DTA plots of Fig. 3A presented
two endothermic peaks at 440 and 483 K, and one exothermic
2.3. Catalytic activity tests peak at 743 K. The endothermic peak at 440 K could be assigned
to the melting of the citric acid, and there was no weight loss.
A batch reactor with 80 ml inner volume and a stirrer was used to Another endothermic peak at 483 K came from the decomposi-
study the catalytic activity. The catalyst of 1 g and 40 ml of ethanol tion of citric acid [25], releasing CH4 , H2 , H2 O and CO2 , resulting
L. Shi et al. / Applied Catalysis A: General 401 (2011) 46–55 49

400 400
1.0 A 300 1.0 B
TG TG 300
0.8 200 0.8 200

Weight loss

Weight loss
0.6 100 0.6 100

μV
DTA

μV
DTA
0.4 0 0.4 0
-100 -100
0.2 0.2
-200 -200
0.0 0.0
400 600 800 1000 400 600 800 1000
Temperature (K) Temperature (K)
400 400
1.0 C 300 1.0 D 300
TG TG
0.8 200 0.8 200
Weight loss

Weight loss
0.6 100 0.6 100

μV
DTA

μV
DTA
0.4 0 0.4 0
-100 -100
0.2 0.2
-200 -200
0.0 0.0
400 600 800 1000 400 600 800 1000
Temperature (K) Temperature (K)
400 400
1.0 E 1.0 F
TG 300 TG 300
0.8 200 0.8 200
Weight loss

Weight loss
0.6 100 0.6 DTA 100
μV

μV
DTA
0.4 0 0.4 0
-100 -100
0.2 0.2
-200 -200
0.0 0.0
400 600 800 1000 400 600 800 1000
Temperature (K) Temperature (K)

Fig. 3. TG-DTA curves of (A) citric acid, and the xerogels burnt in argon with the different M/CA molar ratios of (B) 1:0.5, (C) 1:0.8, (D) 1:1, (E) 1:1.5 and the xerogel burnt in
air with M/CA molar ratio of (F) 1:0.8.

in a large weight loss. The exothermic peak arose from the pyrol- cated carbonic residues left in the as-burnt catalysts were further
ysis of the residue organics at higher temperature. At last, 10% proved by FT-IR, Raman and SEM-EDS.
weight of carbonic residue was left, as observed from the TG As well known, the redox process was strongly influenced by
curve. the ratio of oxidant to reductant. Decreasing the content of reduc-
The DTA plots in Fig. 3B–F presented three endothermic peaks, tant made the redox process more severe. The fierceness and the
and two or three exothermic peaks. The first endothermic peak at flame temperature of the combustion process were strongly influ-
about 323 K could be assigned to the evaporation of the remained enced by the original M/CA molar ratio [13–15]. As shown in Fig. 3B,
water in the gel. The second endothermic peak at 393 K arose from the combustion process was the most severe for the catalyst C0.5
desorption of chemically adsorbed water. The third endothermic with M/CA = 1/0.5 than others, consequently there was only one
peak at 443 K came from the decomposition of NH4 NO3 produced sharp exothermic peak and the residue organics decomposed at the
by pH adjustment of the solution. All the DTA plots in Fig. 3B–F pre- same time. With increasing the content of the citric acid, the flame
sented one sharp exothermic peak at about 493 K in a very narrow temperature was lower, and the combustion was much gentler. As
temperature range and at the same time the TG plots showed an compared in Fig. 3C–E, the second exothermic broad peak became
abrupt weight loss, which indicated that the combustion process larger gradually. It was considered that more complicated carbonic
of xerogels happened as a severe redox reaction wherein citric acid residue which was derived from the pyrolysis of the residue organ-
acted as reductant, and nitrate ions and Cu2+ in the chelated com- ics, was left in the as-burnt catalyst with increasing the content of
pound acted as oxidant. During this combustion process, a large citric acid.
amount of gases such as H2 , H2 O, CH4 , NO, CO2 , NH3 , NO2 were The DTA plots of the xerogel burnt in the air atmosphere with
liberated according to the literatures [18,26]. H2 and CH4 were M/CA = 1/0.8 were compared in Fig. 3F. The DTA plots presented
the reducing agents that could be used in the redox process for three exothermic peaks. One sharp exothermic peak at about 493 K
obtaining metallic Cu from Cu2+ in the chelated compound. The in a very narrow temperature range was derived from the combus-
whole reduction process could be separated in four steps: (1) the tion process in the air atmosphere. As O2 was a kind of oxidant and
decomposition of NH4 NO3 at about 440 K, some NO3 − still left in the participated in the redox reaction during the combustion process,
xerogels; (2) generation of CH4 and H2 by the decomposition of cit- the xerogel burnt in the air atmosphere was more severe than that
ric acid at the beginning of 483 K; (3) severe redox process between in argon (as compared in Fig. 3C). The second exothermic peaks at
CH4 –H2 and NO3 − at about 493 K; (4) reduction of metallic Cu from about 590 K arose from the amorphous carbon oxidation which was
Cu2+ in the chelated compound by CH4 and H2 . in accordance with the experiment results reported by Liu et al. [27]
The second exothermic broad peaks from about 650 to 800 K in and Pranda et al. [28]. Amorphous carbon oxidation was a sustain-
Fig. 3B–E arose from the pyrolysis of the residue organics. Compli- ing heat releasing process in the air. The third exothermic peak at
50 L. Shi et al. / Applied Catalysis A: General 401 (2011) 46–55

A G-band
D-band
2376 1770 820
B 1620

Intensity Counts
3180 3034 1400 1384
Transmittance

820 A
C 3180 2376 1770 1620 1384 B
1720 1400
D 3034 C
2376 1770 1384 820 D
1720 E
3034 1400
2376 1770 820
1384 F
3180 1720 1620
3034 1400
4000 3500 3000 2500 2000 1500 1000 500 800 1000 1200 1400 1600 1800 2000
Wavenumber (cm-1) Raman Shift (cm-1)

Fig. 4. FT-IR spectrum of the xerogels with different M/CA molar ratios of (A) 1:0.5, Fig. 6. Raman analysis of the as-burnt catalysts (A) C1.5 , (B) C1 , (C) C0.8 , (D) C0.5 and
(B) 1:0.8, (C) 1:1 and (D) 1:1.5. the burnt catalysts (E) C0.8-air , (F) C0.8-reduction after 10 h reduction by 5% hydrogen.

about 653 K was derived from the pyrolysis of the residue organics NO3 − provided an in situ oxidizing environment for the decompo-
in air. sition of the citric acid.
There were almost no bands in Fig. 5A. It implied that lit-
3.2. FT-IR analysis of the xerogels with different M/CA molar tle carbonic residue left in the as-burnt catalyst C0.5 . Bands from
ratios and the burnt catalysts about 1100–1650 cm−1 in Fig. 5B, 1100–1700 cm−1 in Fig. 5C, and
1100–1700 cm−1 in Fig. 5D were presented. The bands at 1428,
The FT-IR spectrum of the xerogels with different M/CA molar 1534, 1551 and 1582 cm−1 were attributed to carbonates species
ratios was compared in Fig. 4. A broad absorption band at around (COOO) [32,33], and those at about 1650 cm−1 to C–O stretching
3180 cm−1 indicated the presence of adsorbed moisture [29]. The vibration of bicarbonate [33]. It was clear that with increasing
band at 2376 cm−1 was attributed to the absorption of CO2 [29,30]. the initial content of citric acid in the xeogels, the FT-IR bands
The bands at 820 and 1384 cm−1 were assigned to the NO3 − and of the as-burnt catalyst increased gradually. It inferred more car-
the band at 3034 cm−1 was assigned to O–H group [13,29,30]. Two bonic residues left in the as-burnt catalysts with the increased
bands occurred near 1620 and 1400 cm−1 , representing the asym- content of citric acid. There were two small bands at about 1350 and
metric stretching and symmetric stretching vibrations for carboxyl 1550 cm−1 which were attributed to the symmetric and asymmet-
ions (COO− ), while the bands at 1720 and 1770 cm−1 were from the ric stretches of (OCO) in a carboxylate species [32–35] that came
free carboxylic group [13,31]. Citric acid had two types of carboxyl from the imcomplete decomposition of citric acid in Fig. 5E. The
groups, one inner carboxyl group (1770 cm−1 ) and two terminal bands in Fig. 5E were smaller than those in Fig. 5B. It proved that
free carboxyl groups (1720 cm−1 ). There were two bands at 1720 less carbonic residues were left in the catalyst C0.8-air than C0.8 , and
and 1770 cm−1 in Fig. 3B–D, but no band at 1720 cm−1 in Fig. 4A, the pyrolysis of residue organics was relatively complete in air than
indicating that two terminal free carboxyl groups were completely in argon.
chelated in the xerogel with M/CA = 1/0.5. A small peak and a bigger peak at 2235 cm−1 were observed
Fig. 5 shows the FT-IR spectrum of the as-burnt catalysts with in Fig. 5C and D. Some papers [36–38] reported that the bands
different M/CA molar ratios and the burnt catalyst C0.8-air after in the 2200–2260 cm−1 region possibly arised from isocyanate
reduction. It was clearly observed that there were no NO3 − (820 and/or nitrile intermediates formed by the reaction of NOx with
and 1384 cm−1 ) bands and carboxyl bands (1720 and 1770 cm−1 ) CH4 . And at higher temperature, these surface species were hardly
for all the burnt catalysts. These findings clarified that all the NO3 − detectable. The results were accordance with our experiment phe-
and carboxyl group took part in the reaction of combustion and the nomenon. In the auto-combustion process, a large amount of gases
such as H2 , H2 O, CH4 , NO, CO2 , NH3 , NO2 were liberated. NO and
NO2 with CH4 might form isocyanate and/or nitrile intermediates
A and adsorbed on the surface of the as-burnt catalysts. Decreasing
the content of the citric acid, the combustion process was more
severe and the flame temperature was higher, which was already
B proved by TG-DTA and XRD patterns. High temperature might lead
to the decomposition of isocyanate and/or nitrile intermediates. So
16501550 1100 there were no bands at 2235 cm−1 in the FT-IR spectrum of the as-
Transmittance

C burnt catalyst C0.5 and C0.8 because of higher flame temperature. A


2235 small peak and a bigger peak at 2235 cm−1 were observed only in
D 1700 1576 1100 Fig. 5C and D.

2235
1750 3.3. Raman analysis of the as-burnt catalysts and the burnt
1576 1100 catalysts after reduction
E

1550 1350 The Raman spectrum of the as-burnt catalysts was shown in
4000 3500 3000 2500 2000 1500 1000 500 Fig. 6A–D in the region from 800 to 2000 cm−1 . Two relatively
broad Raman bands at about 1340 and 1590 cm−1 which were
Wavenumber (cm-1)
corresponded to the D (disordered) and G (graphitic) bands of
Fig. 5. FT-IR spectrum of the as-burnt catalysts (A) C0.5 , (B) C0.8 , (C) C1 , (D) C1.5 and the amorphous carbon [39–42], were exhibited in Fig. 6A–C. With
the burnt catalyst (E) C0.8-air after reduction. decreasing the content of citric acid, the intensity of the bands
L. Shi et al. / Applied Catalysis A: General 401 (2011) 46–55 51

Table 1
Cu The EDS analysis of the as-burnt catalysts with different M/CA molar ratios.
ZnO
Catalysts Surface element content (atomic ratio %)a

C O Cu Zn
A
C0.5 22.54 15.26 32.82 29.38
B C0.8 29.62 20.47 26.54 23.37
C1 44.30 17.10 22.03 16.23
C C1.5 58.54 10.14 16.35 14.97
D C0.8-air 8.80 32.00 29.60 29.60
C0.8-reduction 29.03 19.98 25.58 25.41

E a
Determined by element analysis.

F
The XRD patterns of the catalyst C0.8-air after reduction for 10 h
30 40 50 60 70 80 by a flow of 5% hydrogen was compared in Fig. 7E. Cu crystallite size
2θ (º) of the catalyst C0.8-air which was presented in Table 2, was a little
larger than the as-burnt catalyst C0.8 . The result was in accordance
Fig. 7. XRD patterns of the as-burnt catalysts (A) C0.5 , (B) C0.8 , (C) C1 , (D) C1.5 and with the DTA plots in Fig. 3C and F. When the xerogel was burnt
the burnt catalysts (E) C0.8-air , (F) C0.8-reduction after 10 h reduction by 5% hydrogen. in the air, O2 was as a kind of oxidant participating in the redox
process. More oxidant taking part in the combustion process made
at 1340 and 1590 cm−1 decreased gradually in Fig. 6A–C. There the flame temperature higher. Higher temperature resulted in the
were almost no bands in Fig. 6D. It was clear that the amount of larger Cu particle size. Cu crystallite size of the catalyst C0.8-reduction
the amorphous carbon left on the surface of the as-burnt catalysts which was reduced by 5% hydrogen for 10 h at about 523 K was
decreased with the increased M/CA molar ratios. The Raman spec- nearly the same as that of the as-burnt catalyst C0.8 without reduc-
trum of the burnt catalysts C0.8-air after reduction was compared in tion. It inferred that the metallic Cu in the as-burnt catalyst C0.8 was
Fig. 6E. No bands were found, proving that the amorphous carbon generated in the auto-combustion process and the reduction of the
was completely burnt in the air atmosphere. The Raman spectrum as-burnt catalyst C0.8 at 523 K could not lead to the increasing of Cu
of the as-burnt catalyst C0.8-reduction after reduction was almost the crystallite size.
same to that of the as-burnt catalyst C0.8 , suggesting that the amor-
phous carbon could not be eliminated by 5% hydrogen during 10 h 3.5. SEM-EDS analysis of the as-burnt catalysts and the burnt
reduction at about 523 K. catalysts after reduction

3.4. XRD analysis of the as-burnt catalysts and the burnt catalysts SEM micrographs of the as-burnt catalysts with different M/CA
after reduction molar ratios were compared in Fig. 8A–D. All the as-burnt catalysts
exhibited highly porous microstructure, but the morphology was
Fig. 7A–D shows the XRD patterns of the as-burnt catalysts with different. The as-burnt catalyst C0.5 in Fig. 8A showed a foamy fila-
different M/CA molar ratios. All diffraction peaks of the as-burnt ment and spongy in nature owing to higher flame temperature and
catalysts were indexed to Cu and ZnO phase. There was no impu- abundant gases releasing in a split second. For the as-burnt cata-
rity observed. These findings evidently indicated that the CuO phase lyst C0.8 in Fig. 8B, the sample had a large amount of small pores,
was absolutely reduced to metallic Cu by H2 and CH4 produced from whereas the as-burnt catalyst C1 in Fig. 8C exhibited more com-
the combustion process, and the xerogels had been converted into pact structure than C0.8 on the surface. The as-burnt catalyst C1.5
Cu/ZnO catalysts in the combustion process. Cu crystallite size of in Fig. 8D presented ordered polymerization on the surface and a
the catalysts with different M/CA molar ratios calculated by Scher- large pore microstructure in the body, which might be caused by the
rer formula was presented in Table 2. With the increased content excess amount of the citric acid. The unchelated citric acid changed
of citric acid (CA), Cu particle size decreased gradually. It could be to quasi-solid state at 440 K, and then surrounded the xerogels. The
found that Cu particle sizes were closely related to the M/CA moral gases were not easy to be liberated in the combustion process from
ratio, which played an important role on the flame temperature the xerogels because of the surface coverage by the citric acid with
in the auto-combustion process. Decreasing the content of the cit- quasi-solid state, finally forming larger pores. As compared from
ric acid made the flame temperature higher and the combustion Fig. 8A to D, the surface color of the as-burnt catalysts changed
process more severe. Higher temperature resulted in the larger Cu darker gradually, which was in accordance with the Raman results.
particle size. These results were in accordance with combustion With the increased content of the citric acid, more amorphous car-
fierceness degree illustrated by the TG-DTA curves in Fig. 3. bon was left on the surface of the as-burnt catalysts. The color of
Table 2
Batch reaction for methanol synthesis using the as-burnt Cu/ZnO catalysts.

Catalysts SBET a (m2 /g) SCu b (m2 /g) Cuc Conversion (%) Selectivity (%) Yield (%)

Dia. (nm) CO CO2 Ctotal CH3 OH HCOOC2 H5 CH3 OH HCOOC2 H5

C0.5 6.4 1.70 68 24.6 53.2 28.6 3.4 96.6 1.0 10.2
C0.8 26.4 7.69 36 63.1 78.1 65.2 10.6 89.4 6.9 58.3
C1 31.5 6.17 31 58.4 77.7 61.1 11.5 88.5 7.0 54.1
C1.5 27.7 2.80 25 42.5 67.1 45.6 15.7 84.3 7.2 38.4
C0.8-air 18.4 4.93 45 59.2 79.3 62.0 60.4 39.6 37.4 24.5
C0.8-reduction 25.6 7.32 37 60.5 76.4 62.7 10.6 89.4 6.6 56.05

Reaction conditions: T = 443 K, P = 5.0 MPa, catalyst weight: 1 g, ethanol solvent: 40 ml, stirring speed: 2000 rpm, reaction time: 12 h, syngas:
Ar/CO/CO2 /H2 = 3.13/33/5.23/58.64.
a
Determined by N2 physical adsorption–desorption at 77 K.
b
Determined from N2 O pulse chemisorption.
c
Calculated by Scherrer formula.
52 L. Shi et al. / Applied Catalysis A: General 401 (2011) 46–55

Fig. 8. SEM micrographs of the as-burnt catalysts (A) C0.5 , (B) C0.8 , (C) C1 , (D) C1.5 and the catalysts (E) C0.8-air , (F) C0.8-reduction after 10 h reduction by 5% hydrogen.

the as-burnt catalysts also changed from wine red to dark gray with of the carbonic residues left in C0.8-air after reduction significantly
the increased content of citric acid. For the catalyst C0.8-air in Fig. 8E, decreased comparing with the as-burnt catalyst C0.8 . Because a
the sample had a large amount of small pores, and the distribu- large amount of amorphous carbon was burnt in the air atmo-
tion of the small pores was relatively uniform. The SEM image of sphere, as the second exothermic peak shown in Fig. 3F. There was
C0.8-reduction in Fig. 8F was nearly the same with that in Fig. 8B, full almost no amorphous carbon left in the catalyst C0.8-air which was
of small pores. already proved by Raman analysis.
The EDS analysis of the as-burnt catalysts with different M/CA
molar ratios was listed in Table 1. A part of carbonic residues stayed 3.6. TPR analysis of the as-burnt catalyst C0.8 and the burnt
on the as-burnt catalyst surface, which was associated with the con- catalyst C0.8-air
tent of the citric acid. Increasing the content of citric acid resulted
in the increased content of carbonic residues which were derived The TCD signal in terms of hydrogen consumption from H2 -TPR
from the pyrolysis of the residue organics on the surface, and the of the burnt catalyst was shown in Fig. 9. There was only one peak
decreased content of Cu and ZnO. This trend implied that more car- for the burnt catalyst C0.8-air before reduction in Fig. 9A at about
bonic residues were left on the surface of the as-burnt catalysts 508 K, which was corresponded to the reduction of CuO to metallic
with increasing the initial content of citric acid in the xerogels. Cu. The shape of the peak was narrow and symmetrical, indicated
These results were in accordance with those in FT-IR and Raman a homogeneous particle size distribution [43,44]. It implied that
analysis. The EDS analysis of the catalyst C0.8-reduction was nearly CuO phase of the burnt catalyst C0.8-air dispersed homogeneously,
the same with the as-burnt catalyst C0.8 . It inferred that Cu2+ in the which was in accordance with the SEM image with uniform small
chelated compound was absolutely reduced to metallic Cu during pores. The TPR profiles of the catalyst C0.8-air after 10 h reduction
the combustion process in the argon atmosphere with M/CA = 1/0.8. by a flow of 5% hydrogen was shown in Fig. 8B. There was only
This result was accordance with XRD result. However, the amount a little broad peak at about 470 K, which was derived from small
L. Shi et al. / Applied Catalysis A: General 401 (2011) 46–55 53

Consumption of H2 (a.u.)

A
B
C

300 350 400 450 500 550 600 650 700


Temperature (K)

Fig. 9. TPR profiles of the burnt catalysts (A) C0.8-air before reduction, (B) C0.8-air after
reduction, and (C) as-burnt catalyst C0.8 . Fig. 10. The mechanism for the synthesis reaction of low-temperature methanol
from syngas (CO/CO2 /H2 ) using ethanol as promoter (B was the as-burnt catalysts
Cu/ZnO).
quantities of CuO oxidized by 1% oxygen. The TPR profiles of the as-
burnt catalyst C0.8 in Fig. 9C was nearly the same as that in Fig. 9B.
It inferred that all the Cu phase in the as-burnt catalyst C0.8 which synthesis. On the other hand, it was notable that the methanol
was burnt in the argon atmosphere was almost reduced to metallic selectivity increased gradually with the increased content of cit-
Cu. This result was in accordance with that of XRD and EDS. ric acid. This trend was closely related to the crystallite size change
of Cu in Table 2. It was considered that the methanol selectivity
3.7. Activity of the catalysts was associated with the crystallite size of Cu. Smaller crystallite
size of Cu was in favor of methanol formation, exhibiting higher
The activity of the as-burnt catalysts and the burnt catalysts hydrogenation ability, although its SCu value was smaller as shown
reduced by a flow of 5% hydrogen in nitrogen at 523 K for 10 h in Table 2, partly owing to the higher carbonic residues coverage as
was investigated for the low-temperature methanol synthesis from shown in Table 1 and Fig. 5D.
syngas containing CO2 using ethanol as a promoter at 443 K and The activity and the selectivity of the burnt catalysts after reduc-
5.0 MPa for 12 h. The synthetic route was a new method of low- tion were also compared in Table 2. The characterization results and
temperature methanol synthesis [4,45] which allowed using solid the batch reaction records of the burnt catalyst C0.8-reduction were
catalysts, as well as syngas containing CO2 and H2 O directly from an similar to those of the as-burnt catalyst C0.8 . All the CuO phase was
industrial reformer or biomass gasifier without deep purification, reduced to metallic Cu by CH4 and H2 during the auto-combustion
and could be operated at significantly low temperature. process in the argon atmosphere. But for the burnt catalyst C0.8-air
Characterization results and the batch reaction records of the after reduction, the methanol selectivity was remarkably promoted
as-burnt Cu/ZnO catalysts were shown in Table 2. It was shown from 10.6% to 60.4%, although the total carbon conversion was
that the BET surface areas of the as-burnt catalysts increased with slightly lower than that of the as-burnt catalyst C0.8 . The proposed
increasing the citric acid content at first and reached a maximum mechanism [5,7,32] for the low-temperature methanol synthesis
for the as-burnt catalysts C1 with M/CA = 1/1, and then decreased from syngas (CO/CO2 /H2 ) using ethanol as promoter (B was the as-
furthermore. With increasing the content of citric acid, there were burnt catalysts Cu/ZnO) was shown in Fig. 10. Cu+ and ZnO were the
more gases liberated out which led to the porous structure, but active sites for the formation of formate and ethyl formate, but only
the excess citric acid resulted in surplus of the carbonic residues, Cu0 was active for hydrogenation of the ethyl formate. As proved
which coated on the as-burnt catalysts and caused the decreasing by FT-IR and EDS, there were more carbonic residues left in the as-
of the surface areas. The variation trend of the BET surface areas burnt catalyst C0.8 than in C0.8-air . Carbonic residues might influence
of the as-burnt catalysts was similar with that of Cu crystalline the activity of ethyl formate hydrogenation. The activity of the as-
change except the as-burnt C1.5 . The as-burnt catalysts were con- burnt Cu/ZnO catalysts for the ethyl formate hydrogenation should
sisted of metallic Cu, ZnO and a part of carbonic residues which be improved in the low-temperature methanol synthesis reaction
were already proved by FT-IR and Raman analysis. The BET surface in the future study.
areas were influenced by Cu crystalline size and the amount of the As listed in Table 2, the as-burnt catalysts and the burnt catalysts
carbonic residues. Increasing the initial content of the citric acid in were more active for methanol synthesis from CO2 hydrogenation,
the xerogels, more carbonic residues were left in the as-burnt cata- although the content of CO2 was only 5.23%. As had been proved
lysts, which led to the decreased BET surface areas of the catalysts. [48–50], a metallic Cu–Zn surface alloy was a catalytically active
There was a balance point between the content of citric acid and species for the methanol synthesis from CO2 and H2 . The high activ-
Cu crystalline size, at which, the as-burnt catalyst had the biggest ity of CO2 hydrogenation prepared by auto-combustion method
BET surface area. might be owing to the high temperature treatment during com-
Only CO and CO2 were the carbon-containing gas-phase prod- bustion process to form Cu–Zn alloy, as the combustion method
ucts, and only methanol and ethyl formate were the liquid-phase here was conventionally used in producing the alloys and metal
products. It could be found that the total carbon conversion oxide powders.
increased with increasing the content of citric acid and reached The continuous reaction results for low-temperature methanol
a maximum for the as-burnt catalysts C0.8 with M/CA = 1/0.8, and synthesis at 443 K and 5.0 MPa for 40 h using the as-burnt catalyst
then decreased furthermore. The variation trend was similar to that C0.8 were presented in Fig. 11 and Table 3. The conversion of CO,
of the copper (Cu0 ) surface area. This phenomenon was in good CO2 and the total carbon were stable after 15 h. The total carbon
accordance with the literature data [8,46,47], because higher Cu TOF was 1.6 × 10−3 s−1 calculated by N2 O pulse method. The selec-
surface area provided more active sites, and Cu0 was the active site tivity of methanol and ethyl formate was calculated by analyzing
for the rate-determining step in this low-temperature methanol the liquid products which were the plus of those in the cold trap
54 L. Shi et al. / Applied Catalysis A: General 401 (2011) 46–55

Table 3
Continuous reaction result for low-temperature methanol synthesis with the as-burnt catalyst C0.8 .

Catalyst Conversion (%) TOFa (10−3 s−1 ) Selectivity (%) Yield (%)

CO CO2 Ctotal CH3 OH HCOOR CH3 OH HCOOR

C0.8 18.4 −1.7 15.5 1.6 39.6 51.4 6.1 8.0

Reaction conditions: T = 443 K, P = 5.0 MPa, catalyst weight: 3 g, solvent: ethanol 40 ml, stirring speed: 2000 rpm, reaction time: 40 h, flow rate: 20 ml/min, syngas:
Ar/CO/CO2 /H2 = 3.13/33/5.23/58.64.
a
Based on N2 O pulse method. TOF was the total carbon turnover frequency.

20 much lower. The xerogels burnt in the air atmosphere was more
severe than that in argon, as O2 was a kind of oxidant and par-
15 ticipated in the redox reaction during the combustion process. The
Conversions (%)

CO 2 exothermic peak of the amorphous carbon oxidation was also found


10
CO in DTA curves. The SEM images illustrated that all the catalysts
5 C total exhibited highly porous microstructure, but the morphology was
different. EDS, FT-IR and Raman analysis showed that more car-
0
10 bonic residues and amorphous carbon were left in the as-burnt
20 30 40 catalysts with the increased content of the citric acid. There were
-5
no amorphous carbon and a little carbonic residue left in the burnt
-10 catalyst C0.8-air . Cu particle size calculated by Scherrer formula was
influenced by the flame temperature, and became lager in the as-
Time on stream (40 h)
burnt Cu/ZnO catalysts with increasing the M/CA molar ratios. The
Fig. 11. Variations of CO, CO2 , and total carbon conversions with reaction time for BET surface areas of the as-burnt catalysts which were influenced
the continuous synthesis of methanol. by Cu crystalline size and the amount of the carbonic residues,
increased with increasing the citric acid content at first and reached
a maximum for the as-burnt catalysts C1 with M/CA = 1/1, and then
and the autoclave after the reaction. Fig. 11 expressed the time-on- decreased furthermore.
stream activity change of the reaction during 40 h. The analysis on The methanol synthesis reactions were conducted using the
the effluent gas after the trap exhibited that only CO, CO2 , and H2 as-burnt Cu/ZnO catalysts without further reduction from syngas
exist and no alcohols and esters appeared. At the initial stage of the containing CO2 using ethanol as promoter at 443 K and 5.0 MPa
reaction, as 40 ml dead volume of the reactor and 5 ml cold trap was for 12 h. The total carbon conversion increased by decreasing the
filled of pressurized feed gas, the apparent conversions were low M/CA molar ratio and reached a maximum for the as-burnt cata-
but increased gradually. After 15 h, the conversions were stable. CO lyst C0.8 with M/CA = 1/0.8, and then decreased furthermore. The
conversion was about 17%, while CO2 conversion was about −2%. variation trend was in accordance with that of the copper (Cu0 )
The total carbon conversion was about 15%. At the initial 15 h, the surface area. Comparing with the burnt catalyst C0.8-air after reduc-
CO2 conversion dropped to minimum −12%, and then increased tion, the methanol selectivity of the as-burnt Cu/ZnO catalyst was
again to about −2%. The negative CO2 conversion meant that CO much lower according to lower hydrogenation activity of the ethyl
was first converted to CO2 through the water-shift reaction, and formate. The as-burnt catalyst C0.8 was used in continuous low-
then CO2 hydrogenation to methanol. The balance between CO2 temperature methanol synthesis at 453 K and 5.0 MPa for 40 h. The
formation and formate formation was reached after 15 h. total carbon conversion was stable after 15 h and no obvious deac-
tivation during 40 h reaction, but the methanol selectivity was also
4. Conclusion not high.
This novel auto-combustion method [18] which was used to
A series of the as-burnt and the burnt Cu/ZnO catalysts after prepare pure metals and alloys, was for the first time introduced
reduction were prepared by a sol–gel method using metal nitrates to prepare metallic catalysts without further reduction. This novel
with the mole ratio of Cu/Zn = 1/1 and citric acid. During the auto- auto-combustion method will open a new window to prepare
combustion process in the argon atmosphere, a large amount of metallic catalysts, especially the catalysts which was difficult to
gases such as H2 , H2 O, CH4 , NO, CO2 , NH3 , NO2 were liberated. be reduced at high temperature.
H2 and CH4 were the reducing agents that could be used in the
redox process for synthesizing metallic Cu from Cu2+ in the chelated References
compound. The XRD patterns revealed that all the as-burnt cata-
lysts with different M/CA molar ratios were converted into pure Cu [1] H.S. Liu, C.J. Song, L. Zhang, J.J. Zhang, H.J. Wang, D.P. Wilkinson, J. Power Sources
and ZnO species. TPR analysis of the as-burnt catalyst illustrated 155 (2006) 95–101.
[2] J.P. Breen, J.R.H. Ross, Catal. Today 51 (1999) 521–533.
that almost no hydrogen was consumed. It proved that Cu2+ was [3] M.S. Wainwright, D.L. Trimm, Catal. Today 23 (1995) 29–42.
absolutely reduced to metallic Cu in the as-burnt catalyst. [4] L. Fan, Y. Sakaiya, K. Fujimoto, Appl. Catal. A 180 (1999) L11–L13.
The effects of M/CA molar ratio on the properties of catalysts [5] R.Q. Yang, Y. Zhang, Y. Iwama, N. Tsubaki, Appl. Catal. A 288 (2005) 126–133.
[6] R.Q. Yang, X.C. Yu, Y. Zhang, W.Z. Li, N. Tsubaki, Fuel 87 (2008) 443–450.
were studied by TG-DTA, FT-IR, Raman, XRD, SEM-EDS, BET and
[7] T.S. Zhao, K. Zhang, X.R. Chen, Q.X. Ma, N. Tsubaki, Catal. Today 149 (2010)
N2 O chemisorption techniques. The results of the TG-DTA in argon 98–104.
showed that the fierceness and the flame temperature of the auto- [8] G.C. Chinchen, K.C. Waugh, D.A. Whan, Appl. Catal. 25 (1986) 101–107.
[9] Y.L. Zhang, Q. Sun, J.F. Deng, D. Wu, S.Y. Chen, Appl. Catal. A 158 (1997) 105–120.
combustion process were strongly influenced by the initial M/CA
[10] Y. Nitta, O. Suwata, Y. Ikeda, Y. Okamoto, T. Imanaka, Catal. Lett. 26 (1994)
molar ratio in the xerogels. The auto-combustion process was con- 345–354.
sidered as a severe redox process, wherein citric acid acted as [11] Q. Xiao, Z.C. Si, Z.M. Yu, G.Z. Qiu, Mater. Sci. Eng. B 137 (2007) 189–194.
reductant, and nitrate ions and Cu2+ in the chelated compound [12] A.S. Mukasyan, P. Dinka, Adv. Eng. Mater. 9 (2007) 653–657.
[13] Y.Y. Li, L.H. Xue, L.F. Fan, Y.W. Yan, J. Alloys Compd. 478 (2009) 493–497.
acted as oxidant. Increasing the content of citric acid, the com- [14] J. Chandradass, M.H. Kim, D.S. Bae, J. Alloys Compd. 470 (2009) L9–L12.
bustion process was much gentler and the flame temperature was [15] J.L. Liu, W. Zhang, C.J. Guo, Y.W. Zeng, J. Alloys Compd. 479 (2009) 863–869.
L. Shi et al. / Applied Catalysis A: General 401 (2011) 46–55 55

[16] Y.J. Hao, Q.Y. Lai, D.Q. Liu, Z. Xu, X.Y. Ji, Mater. Chem. Phys. 94 (2005) 382–387. [35] A. Yee, S.J. Morrison, H. Idriss, J. Catal. 191 (2000) 30–45.
[17] S. Vivekanandhan, M. Venkateswarlu, N. Satyanarayana, J. Alloys Compd. 462 [36] D. Pietrogiacomi, M.C. Campa, V. Indovina, Catal. Today 155 (2010) 192–198.
(2008) 328–334. [37] T. Montanari, O. Marie, M. Daturi, G. Busca, Appl. Catal. B 71 (2007) 216–222;
[18] Y.W. Jiang, S.G. Yang, Z.H. Hua, H.B. Huang, Angew. Chem. Int. Ed. 48 (2009) M. Ahrens, O. Marie, P. Bazin, M. Daturi, J. Catal. 271 (2010) 1–11.
1–4. [38] M. Huuhtanen, T. Kolli, T. Maunula, R.L. Keiski, Catal. Today 75 (2002) 379–384.
[19] A. Kumar, E.E. Wolf, A.S. Mukasyan, AIChE J. (2010), doi:10.1002/aic.12416. [39] K.H. Kwok, W.K.S. Chiu, Carbon 41 (2003) 673–680.
[20] A. Kumar, E.E. Wolf, A.S. Mukasyan, AIChE J. (2011), doi:10.1002/aic.12537. [40] A. Kromka, J. Breza, M. Kadlečıı̌ková, J. Janıı̌k, F. Balon, Carbon 43 (2005)
[21] X.M. Guo, D.S. Mao, S. Wang, G.S. Wu, G.Z. Lu, Catal. Commun. 10 (2009) 425–429.
1661–1664. [41] U. Ritter, P. Scharff, C. Siegmund, O.P. Dmytrenko, N.P. Kulish, Y.I. Prylutskyy,
[22] X.M. Guo, D.S. Mao, G.Z. Lu, S. Wang, G.S. Wu, J. Catal. 271 (2010) 178–185. N.M. Belyi, V.A. Gubanov, L.I. Komarova, S.V. Lizunova, V.G. Poroshin, V.V. Shla-
[23] J.R. Jensen, T. Johannessen, S. Wedel, H. Livbjerg, J. Catal. 218 (2003) 67–77. patskaya, H. Bernas, Carbon 44 (2006) 2694–2700.
[24] A. Kumar, A.S. Mukasyan, E.E. Wolf, Appl. Catal. A 372 (2010) 175–183. [42] M. Veres, S. Tóth, M. Füle, M. Koós, J. Non-Cryst. Solids 352 (2006) 1348–1351.
[25] L.C. Pathak, T.B. Singh, S. Das, A.K. Verma, P. Ramachandrarao, Mater. Lett. 57 [43] J. Agrell, H. Birgersson, M. Boutonnet, I. Melián-Cabrera, R.M. Navarro, J.L.G.
(2002) 380–385. Fierro, J. Catal. 219 (2003) 389–403.
[26] J.G. Huang, H.R. Zhuang, W.L. Li, Mater. Res. Bull. 38 (2003) 149–159. [44] J. Agrell, M. Boutonnet, I. Melián-Cabrera, J.L.G. Fierro, Appl. Catal. A 253 (2003)
[27] J.H. Zhao, Z.Y. Liu, D.K. Sun, J. Catal. 227 (2004) 297–303. 201–211.
[28] P. Pranda, K. Prandovı̌, V. Hlavacek, Fuel Process. Technol. 61 (1999) 211–221. [45] N. Tsubaki, M. Ito, K. Fujimoto, J. Catal. 197 (2001) 224–227.
[29] J. Chandradass, K.H. Kim, Adv. Powder Technol. 21 (2010) 100–105. [46] G.C. Chinchen, P.J. Denny, D.G. Parker, M.S. Spencer, D.A. Whan, Appl. Catal. 30
[30] Z.X. Wei, L. Wei, L. Gong, Y. Wang, C.W. Hu, J. Hazard. Mater. 177 (2010) (1987) 333–338.
554–559. [47] W.X. Pan, R. Cao, D.L. Roberts, G.L. Griffin, J. Catal. 114 (1988) 440–446.
[31] W.D. Li, J.Z. Li, J.K. Guo, J. Eur. Ceram. Soc. 23 (2003) 2289–2295. [48] T. Fujitani, J. Nakamura, Appl. Catal. A 191 (2000) 111–129.
[32] R.Q. Yang, Y.L. Fu, Y. Zhang, N. Tsubaki, J. Catal. 228 (2004) 23–35. [49] W.B. Pearson, A Handbook of Lattice Spacing of Metals and Alloys, Pergamon,
[33] R. Srivastava, D. Srinivas, P. Ratnasamy, J. Catal. 233 (2005) 1–15. Oxford, 1973, p. 601.
[34] A. Yee, S.J. Morrison, H. Idriss, J. Catal. 186 (1999) 279–295. [50] T. van Herwijnen, W.A. De Jong, J. Catal. 34 (1974) 209–214.

Вам также может понравиться