The effect of the nature of vanadium species on
benzene total oxidation
R. S. G. Ferreira, P. G. P. de Oliveira and F. B. Noronha
,
Laboratório de Catálise, Instituto Nacional de Tecnologia, Av. Venezuela 82, CEP 20081-310 Rio de
Janeiro, Brazil
Abstract
The nature of the vanadium species present on V2O5/Al2O3 catalysts was investigated by using solid
51
state V NMR, diffuse reflectance spectroscopy (DRS), X-ray diffraction (XRD) and temperature
51
5+
programmed reduction (TPR). V NMR and DRS analyses indicated the presence of V in tetrahedral
symmetry at low vanadium loading. A surface polymeric vanadium species and/or the bulk crystalline
V2O5 were mainly observed at high vanadium loading as also detected by XRD. The positions of the
absorption edges determined through the UV–VIS spectra allowed distinguishing between various
tetrahedral symmetries. After TPR, the average oxidation state of vanadium depended on the vanadium
content. The nature of vanadium species was related to the catalyst behavior on the benzene oxidation
reaction. The catalysts containing high vanadium content were more active suggesting that a high amount
4+
of V is responsible for the higher activity.
Author Keywords: Benzene total oxidation; V2O5/Al2O3 catalysts; VOC removal
Article Outline
1. Introduction
2. Experimental
2.1. Catalyst preparation
51
2.2. V NMR spectroscopy
2.3. Diffuse reflectance spectroscopy (DRS)
2.4. X-ray diffraction (XRD)
2.5. Temperature programmed reduction
2.6. Benzene oxidation
3. Results
3.1. NMR
3.2. DRS
3.3. XRD
3.4. TPR
3.5. Catalytic activity
4. Discussion
4.1. The structure of vanadium species
4.2. The state of vanadium after reduction
4.3. The role of the different vanadium species on benzene oxidation
5. Conclusions
Acknowledgements
References
1. Introduction
Recently, higher concerns have been raised about the control of pollutants
like volatile organic compounds (VOCs) mainly due to the more strictly
environmental regulations on several countries [1 and 2].
VOCs involve a wide range of different compounds, such as oxygenates
(alcohols, aldehydes, ketonas), aromatic hydrocarbons and halogen
hydrocarbons [3].
The catalytic oxidation is an effective technology for removal of VOCs,
presenting several advantages over absorption or thermal incineration
processes [4 and 5]. However, the catalytic destruction of VOCs exhibit
specific problems characteristic of these processes, such as low
concentrations of VOCs in the feed stream; complex composition of the
mixture of VOCs in the feed and high space velocities [6]. Therefore, the
development of catalysts for oxidation of a particular mixture of VOCs is an
important task.
Benzene is a VOC present in different industries, such as petrochemical
(benzene alkylation to ethyl benzene; reaction of benzene and long chain
olefins to produce linear alkyl benzene; toluene disproportionation to
benzene and xylenes; naphtha aromatization to benzene); paint and
coating industries and steel manufacture. In the literature, several catalysts
have been used for the total oxidation of benzene [7, 8, 9, 10 and 11].
Generally, noble metal supported catalysts are preferred for the
combustion of VOCs due to their higher activity. Papaefthimiou et al. [7]
studied the catalytic oxidation of benzene isolated or in the presence of
ethyl acetate and butanol on different group VIII metals supported on
alumina. Pt and Pd supported on alumina were the most active catalysts.
Furthermore, in the presence of ethyl acetate or butanol, the total oxidation
of benzene was achieved at higher temperatures than benzene alone.
However, metal oxides can be more active than noble metal catalyst in
function of the nature of the VOC [6]. The deep oxidation of benzene,
ethylacetate and hexane was carried out on MnO2 and Pt/TiO2 catalysts.
The removal of ethylacetate was easier on the metal oxide than the noble
metal catalyst. On the other hand, benzene was oxidized at lower
temperatures on Pt/TiO2 catalyst. Moreover, the metal oxide catalysts are
more resistant to poisoning by compounds containing sulfur or chlorine
[12]. This is an interesting feature for the treatment of streams from steel
manufacturing processes, which contains aromatic compounds in the
presence of sulfur.
In the literature, there are several works about the mechanism of benzene
total oxidation on noble metal catalysts [10, 13 and 14]. Barresi and Baldi
[10] proposed a kinetic model based on irreversible adsorption of benzene
and nonequilibrium adsorption of oxygen over different sites. A Mars-van
Krevelen mechanism was used to explain the oxidation of benzene,
toluene and xylene on a platinum catalyst [13]. However, there is scarce
information about the mechanism of benzene oxidation on metal oxide
catalyst. Vassileva et al. [11] performed the total oxidation of benzene on
30% V2O5/Al2O3 catalyst. The catalytic activity was correlated to the
valency state of vanadium present. A redox mechanism was proposed to
explain these results.
The aim of this work was to study the benzene total oxidation on
V/Al2O3 catalysts with different vanadium loading. The catalysts were
51
characterized using solid-state V nuclear magnetic resonance (NMR),
diffuse reflectance spectroscopy (DRS), X-ray diffraction (XRD) and
temperature-programmed reduction (TPR) in order to determine the nature
of the vanadium species. The role of the different vanadium species on the
benzene total oxidation was proposed.
2. Experimental
2.1. Catalyst preparation
Al2O3 (Engelhard, AL-3916P) was calcined in air at 823 K for 16 h (BET
2
area=180 m /g). V/Al2O3 samples were prepared by Al2O3 impregnation
with an aqueous solution of ammonium methavanadate (Baker Analyzed).
The samples were dried at 353 K under vacuum and calcined under airflow
at 773 K, for 16 h. The samples prepared and their metal contents are
given in Table 1.
Table 1. Vanadium content for V/Al2O3 samples
2.2.
51
V NMR spectroscopy
The
51
V NMR spectra were obtained on a Bruker DRX 300 (7.05T)
spectrometer operating at 78.86 MHz using simple pulses under static
conditions. The spectra were referenced to VOCl3.
2.3. Diffuse reflectance spectroscopy (DRS)
The diffuse reflectance spectra were recorded between 200 and 2000 nm
on an UV–VIS NIR spectrometer (Cary 5 — Varian) equipped with an
integrating sphere (Harrick). Al2O3 was used as the reference. The
condensation degree of the vanadium species present on the
V/Al2O3 calcined samples was calculated through the same procedure
used by Weber [15] to reference molybdenum compounds and supported
molybdenum catalysts. DRS analyses of 5 and 20% V/Al2O3 catalysts after
reaction were also performed. In this case, the sample was reduced under
pure H2 at 773 K, for 1 h, and the reaction was carried out at 573 K, for 3 h,
by using a mixture containing air and benzene (AGA, 482 ppm of
benzene). Then, the catalyst was exposed to air before the DRS
measurements.
2.4. X-ray diffraction (XRD)
XRD has been carried out in a Philips diffractometer PW 1410 using Cu Kα
radiation (λ=1.54050 Å) and Ni filter. The X-ray tube was operated at 40 kV
and 30 mA and the X-ray diffractogram was scanned with a step size of
0.02° (2 θ) from 10 to 80° (2 θ) and counting time of 2.5 s per step.
2.5. Temperature programmed reduction
TPR experiments were performed in a conventional apparatus, as
described previously [16]. Before reduction, the catalysts were heated at
423 K in flowing nitrogen for 0.5 h. Then, a mixture of 4.7% hydrogen in
nitrogen flow was passed through the sample and the temperature was
raised from 298 up to 1173 K at a heating rate of 10 K/min.
2.6. Benzene oxidation
The oxidation of benzene was performed in a flow micro-reactor at
atmospheric pressure. In order to have the same amount of vanadium in
each run (12 mg of vanadium), the mass of the catalyst was varied
between 60 and 240 mg. The sample was reduced under hydrogen at
773 K, for 1 h. After reduction, the reaction was carried out from 373 to
773 K in intervals of 20 K. The reaction mixture consisted of air/benzene at
a flow rate of 30 ml/min (AGA; 482 ppm of benzene). Analyses of the
products were carried out by on line gas chromatography (FID) with a
column Porapak Q (6 m, carrier gas: H2). A gas chromatograph equipped
with a thermal conductivity detector and a molecular sieve and Porapak
QS columns were also used for CO and CO2 analysis.
3. Results
3.1. NMR
51
The V NMR spectra of 1, 5 and 20% V/Al2O3 samples are shown in Fig.
1. The 1% V/Al2O3 exhibited only one peak at −501 ppm whereas the 5%
V/Al2O3 displayed a peak around −418 ppm. The 20% V/Al2O3 sample
showed a peak at −341 ppm.
Fig. 1. NMR spectra of 1, 5 and 20% V/Al2O3 samples.
3.2. DRS
The DRS spectra of V/Al2O3 calcined samples with different vanadium
loading are presented in Fig. 2. The spectra of 1% V/Al2O3 sample showed
a charge transfer band at 288 nm. This band was shifted towards higher
wavelength by increasing the vanadium loading. Furthermore, a new band
appeared around 371 nm on the 5% V/Al2O3 sample, which was shifted to
429 nm on the 20% V/Al2O3.
Fig. 2. DRS spectra of V/Al2O3 samples containing different vanadium
content.
From the spectra in the UV–VIS region of the samples, the positions of the
absorption edges (Fig. 3) were determined by the interception of the
straight line fitted through the low energy side of the curve
2
[F(R∝)·hν] versus hν, where F(R∝) is the Kubelka-Munk function
and hν the energy of the incident photon. The edge energies obtained for
the V/Al2O3 calcined samples were presented in Table 2.
Fig. 3. (a) DRS spectra of and 1% V/Al2O3 sample. (b) Determination of the
position of the absorption edge by the interception of the straight line fitted
2
through the low energy side of the curve [F(R∝)·hν] vs. hν.
Table 2. Edge energies and polymerization degree calculated for the
V/Al2O3 samples with different vanadium content
After reaction, the DRS spectra of 5 and 20% V/Al2O3 catalyst are shown
in Fig. 4. The spectrum of 5% V/Al2O3 catalyst under reaction conditions is
very similar to the one of the calcined sample, exhibiting the charge
transfer band at 328 nm. On the other hand, the DRS spectrum of 20%
V/Al2O3 catalyst presented a band at 327 nm and a new broad band
around 800 nm.
Fig. 4. DRS spectra of 5 and 20% V/Al2O3 sample after reaction at 573 K,
for 3 h.
3.3. XRD
The X-ray diffraction patterns of Al2O3 support and V/Al2O3 sample are
shown in Fig. 5. The diffraction patterns of 5% V/Al2O3 sample displayed
only the Bragg lines characteristics of Al2O3 phase. On the 10%
V/Al2O3 sample, the lines corresponding to V2O5 were detected. The peaks
characteristic of V2O5 were well defined on the 20% V/Al2O3 sample.
Fig. 5. X-ray diffraction patterns of (a) Al2O3; (b) 5% V/Al2O3; (c) 10%
*
V/Al2O3; (d) 20% V/Al2O3; ( ) position of the Bragg peaks for V2O5.
3.4. TPR
The TPR profiles of the bulk V2O5 and V/Al2O3 catalysts are presented
in Fig. 6. The reduction profile of bulk V2O5 exhibited three peaks at 933,
984 and 1116 K. The TPR profile of 1% V/Al2O3 catalyst showed only one
peak at 862 K while the 5% V/Al2O3 catalyst displayed a broad peak
around 791 K. The 10 and 20% V/Al2O3 catalysts showed peaks at higher
reduction temperatures (819 and 869 K, respectively). The reduction
degree calculated from H2consumption during TPR is presented in Table 3.
Fig. 6. TPR profiles of the V/Al2O3 sample: (a) V2O5; (b) 1% V/Al2O3; (c)
5% V/Al2O3; (d) 10% V/Al2O3; (e) 20% V/Al2O3.
Table 3. Hydrogen consumption and reduction degree calculated from TPR
analysis
3.5. Catalytic activity
The results of benzene conversion as a function of reaction temperature on
V/Al2O3 catalysts are shown in Fig. 7. Basically, the 20% V/Al2O3 catalyst
was the most active in all ranges of temperature, achieving total
conversion at 673 K. The following order of activity was observed: 20%
V/Al2O3>10% V/Al2O3>5% V/Al2O3. It is important to stress that CO2 was
the only product formed. The reaction rates calculated from these curves
are presented inTable 4. In order to allow a comparison of the catalytic
behavior of the different catalysts, the reaction rates were obtained at low
conversion. Increasing the vanadium loading increased the reaction rate.
Fig. 7. Benzene conversion as a function of reaction temperature: (▪) 5%
V/Al2O3; (•) 10% V/Al2O3; (□) 20% V/Al2O3.
Table 4. Benzene conversion and reaction rates at 525 K calculated for the
V/Al2O3 samples with different vanadium content
4. Discussion
4.1. The structure of vanadium species
The NMR analyses were performed in order to determine the nature of
vanadium species on the V/Al2O3 samples. In the literature [17, 18 and 19],
the peaks on NMR spectra have been attributed to different vanadia
symmetries. Eckert and Wachs [17] attributed to dimeric form of vanadium
in tetrahedral symmetry the peak around −550±30 ppm on V/Al2O3. The
peak at −300 ppm corresponded to vanadium species in distorted
octahedral structure, similar to bulk V2O5, or superficial polymeric species.
Eon et al. [18] observed the presence of two peaks at −310 and −540 ppm
on the NMR spectra of V/Al2O3 samples. The authors suggested that they
were due to octahedral and tetrahedral species, respectively. Koranne et
al. [19] used the NMR technique on the characterization of
V/Al2O3 samples containing different vanadium content. Generally, the
NMR spectra exhibited two peaks around −300 and −600 ppm. The peak
at −600 ppm was the most important for samples with low vanadium
content while the peak at −300 ppm was predominant on the
V/Al2O3 samples containing high vanadium concentration. According to
Koranne et al. [19], the 2.0 and 4.8% V2O5/Al2O3 samples would have only
tetrahedral species whereas the octahedral vanadium species or polymeric
species would appear on the samples with high vanadium content. This
modification of vanadium species as a function of the content has been
associated with the degree of surface coverage of the support. Therefore,
isolated vanadates are present at low coverage while the increase of the
coverage led to the appearance of octahedral polymeric species.
In our work, the 1% V/Al2O3 sample presented only tetrahedral vanadium
species. The increase of the vanadium content led to the polymerization of
vanadium species represented by the shift of the peak to −418 ppm on the
5% V/Al2O3 sample. The 20% V/Al2O3 sample contained only superficial
polymeric species or octahedral species as bulk V2O5.
The DRS analyses are in agreement with the NMR results. According to
the literature [18], the absorption band around 420–480 nm corresponded
5+
to V in octahedral symmetry whereas the band at 300 nm is attributed to
5+
V in tetrahedral symmetry. Two bands at 625 and 770 nm are
4+
characteristic of the d–d transition of V in octahedral symmetry.
5+
The 1% V/Al2O3 sample exhibited only tetrahedral V species. The
increase of vanadium loading led to the appearance of octahedral species.
On the 20% V/Al2O3 sample, the band at 429 nm has been also observed
on the bulk V2O5, suggesting the formation of crystalline V2O5 species.
The shift of the band position as vanadium content increase depends on
condensation degree or cluster size [20 and 21]. The condensation degree
of the V/Al2O3 samples was calculated based on Cruz and Eon [22]. They
determined a correlation between edge energy and number of nearest
vanadium neighbors for different model compounds ( Table 5).
Table 5. Number of nearest neighbors, vanadium species and edge
position of vanadium model compounds from reference [22]
From the correlation obtained for the model compounds, the condensation
degree of V/Al2O3 samples was determined (Table 2). The condensation
degree increased while the vanadium loading increased which agreed well
with the literature [15, 18 and 19]. Thus, the shift in the position of the
bands suggested degree of polymerization enhanced. The 1%
V/Al2O3 presented, basically, isolated vanadate species. On the 5%
V/Al2O3 sample was observed the appearance of polymerized structures,
such as linear vanadate species. At high vanadium content, polymeric
species similar to octahedral bulk V2O5 were formed. Then, the DRS
analysis allowed one to distinguish between various tetrahedral
symmetries observed on the V/Al2O3 samples, which is very difficult to
obtain using NMR technique.
The XRD results also revealed the increase of the polymerization degree.
Increasing the vanadium content, the lines corresponding to V2O5 phase
appeared. In our work, the peaks characteristic of V2O5 phase were
detected on the 10% V/Al2O3 sample. Koranne et al. [19] also observed the
formation of V2O5 crystalline on a 14% V/Al2O3 sample.
4.2. The state of vanadium after reduction
Koranne et al. [19] also observed the presence of three peaks on the
reduction profile of bulk V2O5. They ascribed these peaks to the reduction
on different steps: V2O5→V6O13→V2O4→V2O3.
On the V/Al2O3 catalysts, the increase of vanadium content shifted the
maximum of the peak to higher temperature. As previously observed the
bulk V2O5is reduced at high temperature. Then, these shifts could be
explained by the increase of the condensation degree while vanadium
loading increased, as revealed by NMR and DRS analyses. These results
are in agreement with the literature [23].
After TPR, the vanadium average oxidation state on the 1 and 5%
5+
3+
V/Al2O3 catalysts corresponded to the complete reduction of V →V . On
the other hand, the catalyst containing higher vanadium content revealed
3+
4+
the presence of a mixture of V and V .
Roozeboom et al. [23] observed a complete reduction of V2O5 to V2O3 on
alumina supported catalysts. However, Koranne et al. [19] detected a
5+
4+
reduction stoichiometry of V →V on V/Al2O3 catalysts.
4.3. The role of the different vanadium species on benzene
oxidation
Our results of total benzene oxidation on V/Al2O3 catalysts suggested that
the catalytic activity could be correlated with vanadium content.
After the catalytic test, an important fraction of vanadium is still present as
4+
V on the catalyst containing 20% of vanadium, as it can be seen from the
appearance of the band around 800 nm (Fig. 4). On the other hand, the
4+
band corresponding to V was not observed on the 5% V/Al2O3 catalyst. It
is important to stress that the catalyst with a higher amount of V
the high benzene activity (Table 4).
4+
exhibited
Eon et al. [18] studied the influence of the degree of condensation of
vanadium on the catalytic properties of Al2O3 supported vanadium oxides
for the oxidative dehydrogenation of propane. The catalytic results showed
that as the vanadium content decreased, the selectivity to C3H6 increased,
whereas the selectivity towards CO2 decreased. DRS analysis of the
4+
catalysts after reaction revealed a higher fraction of V species on the
catalysts containing higher vanadium content. A correlation between the
CO2 selectivity and the nature of vanadium species could therefore be
established. The catalyst most selective to CO2 formation was the one with
4+
the highest amount of V species.
It is suggested that the same type of active site for CO2 formation would be
present on our catalysts for the total oxidation of benzene. In fact, our DRS
and catalytic tests are in agreement with these results. The catalysts
4+
containing a high fraction of V species (10 and 20% V/Al2O3) were the
most active for the oxidation of benzene to CO2. This more important
4+
amount of V species was due to the presence of polymeric species
similar to octahedral bulk V2O5, which was more difficult to reduce.
5. Conclusions
51
5+
V NMR and DRS analyses indicated the presence of V
in tetrahedral
symmetry at low vanadium loading. A surface polymeric vanadium species
and/or the bulk crystalline V2O5 were mainly observed at high vanadium
loading. The positions of the absorption edges determined through the
UV–VIS spectra allowed distinguishing between various tetrahedral
symmetries. After TPR, the vanadium average oxidation state on the 1 and
5+
3+
5% V/Al2O3catalysts corresponded to the complete reduction of V →V .
On the other hand, the catalyst containing higher vanadium content
5+
4+
revealed the presence of a mixture of V and V . The results of total
benzene oxidation on V/Al2O3 catalysts suggested that the catalytic activity
could be correlated with vanadium content. The catalysts containing high
vanadium content were more active suggesting that a high amount of
V
4+
is responsible for the higher activity on the total benzene oxidation.
Acknowledgements
The authors would like to thank Rosane A.S. San Gil (IQ/UFRJ) and
Ricardo Aderne (NUCAT/PEQ/COPPE/UFRJ) for the analyses of NMR
and DRS, respectively. F.B. Noronha and R.S.G. Ferreira thank CNPq and
FAPERJ for financial support.
References
1. G. Parkinson, Chem. Eng. (1991) 37.
2. I.D. Dobson. Progress Org. Coatings 24 (1994), p. 55. Abstract |
PDF (727 K) | View Record in Scopus | Cited By in Scopus (6)
3. J.N. Armor. Appl. Catal. B 1 (1992), p. 221. Abstract |
PDF (2675
K) | View Record in Scopus | Cited By in Scopus (276)
4. J.J. Spivey. Ind. Eng. Chem. Res. 26 (1987), p. 2165. Full Text via
CrossRef | View Record in Scopus | Cited By in Scopus (385)
5. D.R. van der Vaart, W.M. Vatvuk and H. Wehe. J. Air Waste Manage.
Assoc. 41 (1991), p. 92. View Record in Scopus | Cited By in Scopus (38)
6. C. Lahousse, A. Bernier, P. Grange, B. Delmon, P. Papaefthimiou, T.
Ionnides and X. Verykios. J. Catal. 178 (1998), p. 214. Abstract |
PDF
(174 K) |View Record in Scopus | Cited By in Scopus (104)
7. P. Papaefthimiou, T. Ioannides and X.E. Verykios. Appl. Catal. B:
Environ. 13 (1997), p. 175. Abstract |
PDF (909 K) | View Record in
Scopus | Cited By in Scopus (122)
8. P. Papaefthimiou, T. Ioannides and X.E. Verykios. Appl. Catal. B:
Environ. 15 (1998), p. 75. Abstract |
PDF (1426 K) | View Record in
Scopus | Cited By in Scopus (90)
9. S.K. Gangwal, M.E. Mullins, J.J. Spivey, P.R. Caffrey and B.A.
Tichenor. Appl. Catal. 36 (1988), p. 231. Abstract |
PDF (994 K) | View
Record in Scopus | Cited By in Scopus (64)
10. A. Barresi and G. Baldi. Ind. Eng. Chem. Res. 33 (1994), p. 2964. Full
Text via CrossRef | View Record in Scopus | Cited By in Scopus (46)
11. M. Vassileva, A. Andreev, S. Dancheva and N. Kotsev. Appl.
Catal. 49 (1989), p. 125. Abstract |
PDF (1074 K) | View Record in
Scopus | Cited By in Scopus (20)
12. R.M. Heck, R.J. Farrauto, Catalytic Air Pollution Control: Comercial
Technology, Van Nostrand Reinhold, New York, 1995, p. 147.
13. K.T. Chuang, S. Cheng and S. Tong. Ind. Eng. Chem. Res. 31 (1992),
p. 2466. Full Text via CrossRef | View Record in Scopus | Cited By in
Scopus (31)
14. A. Barresi and G. Baldi. Chem. Eng. Sci. 47 (1992), p.
1943. Abstract |
PDF (1046 K) | View Record in Scopus | Cited By in
Scopus (25)
15. R.S. Weber. J. Catal. 151 (1994), p. 318.
16. F.B. Noronha, M. Primet, R. Frety and M. Schmal. Appl.
Catal. 78 (1991), p. 125. Abstract |
PDF (932 K) | View Record in
Scopus | Cited By in Scopus (47)
17. H. Eckert and I.E. Wachs. J. Phys. Chem. 93 (1989), p. 6796. Full
Text via CrossRef | View Record in Scopus | Cited By in Scopus (245)
18. J.G. Eon, R. Olier and J.C. Volta. J. Catal. 145 (1994), p.
PDF (643 K) | View Record in Scopus | Cited By in
318. Abstract |
Scopus (152)
19. M.M. Koranne, J.G. Goodwin and G. Marcelin. J. Catal. 148 (1994), p.
369. Abstract |
PDF (758 K) | View Record in Scopus | Cited By in
Scopus (116)
20. M. Fournier, C. Louis, M. Che, P. Chaquin and D. Masure. J.
Catal. 119 (1989), p. 400. Abstract | Article |
PDF (1085 K) | View
Record in Scopus |Cited By in Scopus (104)
21. D. Masure, P. Chaquin, C. Louis, M. Che and M. Fournier. J.
Catal. 119 (1989), p. 415. Abstract | Article |
PDF (804 K) | View
Record in Scopus |Cited By in Scopus (27)
22. A.M.A. Cruz and J.G. Eon. Appl. Catal. A: Gen. 167 (1998), p.
203. Article |
PDF (512 K)
23. F. Roozeboom, M.C. Mittelmeijer-Hazeleger, J.A. Moullijn, J. Medema,
V.H.J. de Beer and P.J. Gellings. J. Phys. Chem. 84 (1980), p. 2783. Full
Textvia CrossRef | View Record in Scopus | Cited By in Scopus (167)