Ammoxidation of 3-picoline to nicotinonitrile using silica-supported VCrO catalysts

Article abstract of DOI:10.1007/s11164-012-0691-x

A series of vanadium-chromium oxide (VCrO) catalysts supported on silica was prepared by wetness impregnation method with different Cr/V molar ratios from 0.2 to 1.0. These catalysts were characterized by XRD, TG, temperature-programmed desorption of ammo

Full text of DOI:10.1007/s11164-012-0691-x

Res Chem Intermed (2013) 39:1353–1361
DOI 10.1007/s11164-012-0691-x
Ammoxidation of 3-picoline to nicotinonitrile using
silica-supported VCrO catalysts
•
Feng Jiang Ruiping Wei Lijing Gao
•
•
•
Guomin Xiao Lei Niu
Received: 17 April 2012 / Accepted: 13 June 2012 / Published online: 30 June 2012
Ó Springer Science+Business Media B.V. 2012
Abstract A series of vanadium–chromium oxide (VCrO) catalysts supported on
silica was prepared by wetness impregnation method with different Cr/V molar
ratios from 0.2 to 1.0. These catalysts were characterized by XRD, TG, temperature-
programmed desorption of ammonia (NH3-TPD), X-ray photoelectron spectra
(XPS), and Raman spectroscopy, and their catalytic activity was evaluated in the
ammoxidation of 3-picoline (3-PIC) to nicotinonitrile (NN). The results of XRD,
TG, Raman, and XPS confirmed that the active components on the silica surface
were mainly amorphous V2O5 and CrVO4. The results of NH3-TPD showed that
acidity of the catalysts decreased with the increase of Cr/V ratio. Catalytic results
revealed that acidity of the catalysts was closely related to the catalytic perfor-
mance. Low acidity gave low conversion of 3-PIC and high NN selectivity. Fur-
thermore, the conversion of 3-PIC increased with rise in reaction temperature, and
the selectivity of NN was slightly influenced by the 3-PIC conversion. Therefore,
among the catalysts (Cr/V ratio was 0.2, 0.4, 0.6) tested, Cr/V-0.6 catalyst retained
the lowest acidity and exhibited the highest selectivity and yield of NN in the
ammoxidation of 3-PIC.
Keywords Ammoxidation ꢀ 3-Picoline ꢀ Nicotinonitrile ꢀ VCrO catalysts
Introduction
Nicotinonitrile (NN) is an essential chemical intermediate, which is mainly applied
to the large-scale production of nicotinic acid and nicotinamide [1, 2]. Both of these
compounds are considered to be the components of the vitamin B complex and are
F. Jiang ꢀ R. Wei ꢀ L. Gao ꢀ G. Xiao (&) ꢀ L. Niu
School of Chemistry and Chemical Engineering, Southeast University,
Nanjing 211189, Jiangsu, People’s Republic of China
e-mail: xiaogmseu@gmail.com
123

1354
F. Jiang et al.
used extensively as feed additives. NN is usually produced by ammoxidation of
3-picoline (3-PIC), for ammoxidation is considered as an economic and eco-friendly
process and water is the only byproduct in ideal case [1–4].
During the past 50 years, thousands of patents and papers have been reported in
the literature on different ammoxidation reactions [5]. It is believed that vanadium-
based catalysts are generally efficient in the ammoxidation of 3-PIC, such as VPO
[6], V2O5/Al2O3–ZrO2 [7], and VTiO [8]. Acidity is an important property that
governs in particular the selectivity and activity of ammoxidation catalysts. For the
vanadium-based catalysts, the increase of the acidity resulted in the enhancement of
the conversion of 3-PIC [9, 10]. Some researchers claimed that the introduction of
chromium in the V2O5/TiO2 catalyst could increase the acidity of catalysts, which in
turn enhanced the conversion of 3-PIC to nicotinic acid [11]. Actually, catalysts of
vanadium–chromium mixed oxides have shown attractive performance in the
ammoxidation of o-xylene and methylbenzyl chlorides [12, 13]. On the other hand,
silica support has been proved to have an attractive performance in the
ammoxidation of aromatics for its large surface area and stability at high reaction
temperature [13, 14]. However, there were no reports about the ammoxidation of
3-PIC with silica-supported vanadium–chromium catalyst.
In the present work, we have prepared VCrO/SiO2 catalysts with V2O5 and CrO3,
which were rarely used as sources of vanadium and chromium. Then, the VCrO/
SiO2 catalysts were applied to catalyze the ammoxidation of 3-PIC to produce
nicotinonirile. In addition, the molar ratio of Cr/V was investigated to throw light on
its effects on the catalytic properties, which might fill in the blank space in the study
of the catalyst class of vanadium–chromium mixed oxides.
Experimental
Preparation of VCrO/SiO2 catalyst
A series of VCrO/SiO2 catalysts (labeled as Cr/V-n, n refers to Cr/V molar ratio;
n = 0.2, 0.4, 0.6, 0.8, 1.0) were prepared by the incipient wetness impregnation
method. Requisite amounts of V2O5 (purity [99 %) and CrO3 (purity [99 %)
dissolved in aqueous oxalic acid solution were impregnated on to the particle form
of SiO2 support to achieve desired Cr/V molar ratios. The total content of V2O5 and
CrO3 was maintained at 10 wt%. The samples were dried at 110 °C for 10 h and
calcined at 500 °C for 6 h in air and then sieved to a desired fraction (20–40 mesh).
Catalytic reaction of 3-PIC ammoxidation to NN
Catalytic experiments were performed in a fixed bed, stainless steel reactor (inner
diameter = 12 mm) with 10 ml of catalyst packed. The flow rates of NH3 and Air
were maintained by mass flow controllers. The mixture of water and 3-PIC was
injected by a syringe pump and vaporized in a preheating zone provided on the top
of the catalyst bed. In general, the following reaction conditions were applied:
GHSV = 2,088 h-1, molar ratio: 3-PIC:Air:NH3:H2O = 1:22:4:5. The temperature
123

Ammoxidation of 3-picoline to nicotinonitrile
1355
of the catalyst bed was measured during the reaction using a thermocouple inside
the reactor. The products were collected in a cold trap and analyzed by GC equipped
with FID using a capillary column (SE-54). Conversion of 3-PIC and selectivity of
NN (S-NN) were calculated on the basis of carbon balance as follows. Yield of NN
(Y-NN) was calculated by multiplying conversion of 3-PIC and S-NN
moles of 3-picoline reacted
Conversion of 3-picoline ¼
moles of 3-picoline supplied
moles of NN formed
Selectivity of NN ¼
moles of 3-picoline reacted
Characterization of VCrO/SiO2 catalyst
XRD patterns were recorded on a Rigaku D/MAX 2000 instrument with Cu Ka
˚
(k = 1.5406 A) radiation. Thermal analysis was carried out using a simultaneous
DSC–TGA instrument (SDT Q600, TA, USA) and performed at a uniform heating
rate of 20 °C min-1 under air atmosphere. Raman spectra were measured by a
LabRAM HR800 instrument with excitation from an Ar laser operating at 514 nm.
X-ray photoelectron spectra (XPS) were recorded on an ESCALAB MKII unit with
Al Ka radiation (E = 1,486.6 eV) at the base pressure of the UHV chamber. The
temperature-programmed desorption of ammonia (NH3-TPD) was performed to
scrutinize the acidic properties of the catalysts. Before adsorption with ammonia,
the catalyst sample (150 mg) was pretreated for 1 h in a helium stream at 500 °C.
After cooling to 100 °C, NH3 was injected until saturation. The ammonia desorption
profile was detected by GC 5890 at a heating rate of 10 °C min-1 from room
temperature to 500 °C under 10 ml min-1 helium flow.
Results and discussion
XRD results
XRD patterns of fresh samples were shown in Fig. 1. All of them showed just a
single prominent diffraction peak at 2h = 22° assigned to disordered silica
materials except Cr/V-1.0. This suggested that chromium and vanadium species
deposited during the preparation were either X-ray amorphous in nature or the
crystals were less than 4 nm in size. However, Cr/V-1.0 catalyst revealed additional
XRD lines at 2h = 25°, 34°, 36°, 50°, and 55° assigned to Cr2O3 (JCPDS PDF#
38-1379).
Thermal analysis
Result of thermal analysis of V/Cr-0.4 catalyst (uncalcined) is presented in Fig. 2. It
was evident that there were three stages of weight loss of catalysts. The first one was
from room temperature to 180 °C, which was clearly ascribed to desorption of
123

1356
F. Jiang et al.
Fig. 1 XRD patterns of
(a) SiO2, (b) Cr/V-1.0,
(c) Cr/V-0.4, and (d) Cr/V-0.2
Fig. 2 TGA and DTG curves of uncalcined Cr/V-0.4
physically adsorbed water, about 4 %. Such a content of water was expected because
the catalyst was prepared from aqueous solution. The second one which occurred at
around 240 °C was undoubtedly due to the decomposition of vanadyl oxalate formed
by the reaction of oxalic acid and vanadium pentoxide [15]. The last stage was very
small and dependent on the second one lying at around 300 °C, which was probably
due to the formation of CrVO4 by the replacement of V with Cr in the bridging V–O–V
of V2O5. However, none of the three stages of weight loss belonged to the three major
thermal events of CrO3 [16]. Such a phenomenon was anticipated because CrO3 was
reduced to Cr3? species by oxalic acid in the preparation process.
Raman spectroscopy
Raman spectra of the support and catalysts were presented in Fig. 3. Pure SiO2 did
not show any intense Raman bands but there was a weak band at *493 cm-1 in the
123

Ammoxidation of 3-picoline to nicotinonitrile
1357
Fig. 3 Raman spectra of VCrO/
SiO2 catalysts
200–1,100 cm-1 region. The broad Raman band of the catalysts at *965 cm-1
should correspond to V–O–V functionality of the surface polymeric vanadium
species [17, 18]. The Raman band at 1,034 cm-1 was assigned to the terminal V=O
bond of the dehydrated surface vanadia species [17]. Crystalline V2O5 vibrations at
141, 191, 406, 526, 695, and 993 cm-1 were not present in the samples [19], which
suggested that V2O5 was amorphous on the silica. These results were in agreement
with the XRD results. The Raman bands at 917, 877, 760, and 655 cm-1 were
characteristic of the CrVO4 species [20, 21], which was in agreement with the TG
results. Moreover, the Raman band presented at 550 cm-1 of the high chromium
loaded sample (Cr/V-1.0) could be assigned to the crystalline Cr2O3 [22, 23], which
was consistent with the XRD data.
X-ray photoelectron spectroscopy
XPS was used to determine the oxidation states and the surface compositions of the
fresh catalysts. The position of V5? lay at 517.2 and 517.7 eV with respect to the
reference value of C 1s at 285 and 284.6 eV, respectively [24, 25]. As shown in
Table 1, the binding energies of V 2p3/2 (517.5, 517.5 and 517.5 eV) with respect to
the reference value of C 1s (284.9, 284.7, and 284.8 eV) indicated that only one
type of vanadium oxide with a ?5 oxidation state was present in the fresh samples.
The differences between the XPS signals of Cr 2p3/2 and O 1s binding energies were
around 45.3 eV, clearly indicating the presence of Cr3? on the surface of the fresh
catalysts [17]. In addition, the surface atomic Cr/V ratios were higher than the
corresponding values of the bulk, suggesting a surface enrichment of the chromium
species, which was attributed to the existence of CrVO4 [21]. In general, the active
components on the silica surface were mainly amorphous CrVO4 and V2O5 in the
low molar ratio of Cr/V (0.2–0.6). The surface atomic Cr/V ratios continually
increased to 0.82 close to 1.0 (the value of Cr/V molar ratio in CrVO4) with the rise
of Cr/V ratio from 0.2 to 0.6. As a result of that, the concentration of the surface
CrVO4 was increasing with the increase of Cr/V ratio from 0.2 to 0.6.
123

1358
F. Jiang et al.
Cr/V-0.6
Table 1 XPS results of the
fresh samples
Sample
Cr/V-0.2
Cr/V-0.4
0.62
Surface atomic ratios
Cr/V
0.33
0.82
Atomic surface concentration (%)
Cr 2p
0.20
0.60
0.32
0.52
59.5
10.9
0.36
0.44
61.0
9.5
V 2p
O 1s
56.9
12.0
C 1s
Bending energy (BE)
Cr 2p3/2
V 2p3/2
Si 2p
578.3
517.5
103.9
284.9
533.1
578.7
517.4
103.8
284.7
533.2
578.2
517.7
103.9
284.8
533.1
C 1s
O 1s
NH3-TPD results
The NH3-TPD was performed to scrutinize the acidic properties of the catalysts. The
results of NH3-TPD (Fig. 4) showed a decrease in the amount of acid sites along
with the increase of Cr/V ratio from 0.2 to 0.6. NH3-TPD of SiO2 was also tested
and compared with the catalysts of the VCrO loaded. It was observed that the
supported catalysts had higher amounts of acid sites but weaker strength of acid
sites compared with the silica support. The amount of acid sites estimated from the
NH3-TPD results varied depending on the Cr/V ratio. However, the temperature at
the top peak of NH3-TPD was almost constant with all the supported catalysts,
suggesting that the strength of the acid site was not substantially affected by the
Cr/V ratio. Since the amount of CrVO4 increased with the increase of Cr/V ratio as
confirmed by the XPS results, it was deduced that the increase of Cr/V ratio resulted
in the increase amount of CrVO4, which in turn reduced the acidity in this VV–CrIII
composite oxide, and the acidity of CrVO4 was lower than V2O5.
Effect of Cr/V molar ratio on the catalytic performance of VCrO/SiO2 catalysts
The research on the influence of the Cr/V molar ratio on the catalytic performance
of VCrO/SiO2 catalysts was carried out at atmospheric pressure and 360 °C.
Variation of conversion of 3-PIC (X-3-PIC), S-NN and Y-NN as a function of Cr/V
molar ratio was depicted in Fig. 5. Obviously, the molar ratio of Cr/V showed a
strong influence on the catalytic behavior of the catalysts. The conversion of 3-PIC
remarkably decreased with the rise of Cr/V ratio. X-3-PIC for Cr/V-0.2 was roughly
fourfold that for Cr/V-1.0. It may be noted that S-NN increased considerably from
51 to 87 % when the Cr/V ratio increased from 0.2 to 0.6, and then at a slower rate
up to 93 % as the Cr/V ratio further increased to 1.0. The highest yield of 15.7 %
was obtained when the value of Cr/V ratio was 0.4, the corresponding 3-PIC
conversion was 19.5 % and NN selectivity was 80.5 %.
123

Ammoxidation of 3-picoline to nicotinonitrile
1359
Fig. 4 NH3-TPD profiles of
samples
Fig. 5 Catalytic performance
of VCrO/SiO2 catalysts with
different Cr/V molar ratios
These observations indicated that the low Cr/V ratio resulted in the high amount
of acidity which led to the high activity of the catalyst, and reduced the S-NN. It was
known that acid sites were necessary for the initial adsorption of aromatic substrate
on the catalyst surface and then forming p-complexation with the catalyst surface
through p-electrons of aromatic ring following H-abstraction. After ammoxidation
reaction, however, the interaction between acid sites and p electrons of the nitrile
product also hindered its fast desorption and made it prone for total oxidation to
COx. Therefore, the high amount of acidity had enhanced the activity but reduced
the selectivity of the catalysts.
Correlations between conversion of 3-PIC and Y-NN and S-NN
To study the influence of conversion upon the selectivity, three of the active
catalysts (Cr/V-0.2, Cr/V-0.4, Cr/V-0.6) were tested at 360–400 °C. Figure 6
depicts a NN selectivity and yield versus 3-PIC conversion plot. Two sets of results
can be observed: on the one hand the selectivity of all catalysts seemed to be
123

1360
F. Jiang et al.
Fig. 6 NN selectivity (open
symbol) and yield (closed
symbol) versus 3-PIC
conversion (360–400 °C
temperature window); closed
diamonds Cr/V = 0.2, closed
triangles Cr/V = 0.4, closed
circles Cr/V = 0.6
independent with conversion levels of 3-PIC between 20 and 57 %, while on the
other hand, the selectivity of higher Cr/V ratio was still higher than that of the lower
Cr/V ratio at the same conversion level of 3-PIC. Therefore, the Cr/V-0.6 displayed
better performance than the Cr/V-0.4 and Cr/V-0.2. The lower amount of acidity
caused by higher Cr/V ratio had a promotional effect on the Y-NN in ammoxidation
of 3-PIC. Similar results were also obtained by Roy et al. [26].
Effect of temperature on the catalytic performance of Cr/V-0.4 catalyst
The study on the influence of temperature was carried out using the Cr/V-0.4
catalyst at atmospheric pressure and 360–400 °C. Variation of S-NN, Y-NN, and
X-3-PIC as a function of reaction temperature is illustrated in Fig. 7. The results
revealed that S-NN appeared to be independent on the reaction temperature and
maintained at about 75 % throughout the temperature range from 360 to 400 °C.
The conversion of 3-PIC was observed to increase continually from 20 to 56 %.
Consequently, the Y-NN has increased gradually from 16 to 41 % along with the
increase in temperature. Similar observations have also been made by Kalevaru
et al. [27]. The enhancement in the catalytic activity might be attributed to the
Fig. 7 Effect of temperature on
the catalytic performance
of Cr/V-0.4
123

Ammoxidation of 3-picoline to nicotinonitrile
1361
increase in the intrinsic activity of the existing active sites and also to the creation of
more active sites with rise in reaction temperature.
Conclusions
In summary, the VCrO/SiO2 catalysts were prepared by the incipient wetness
impregnation method with the Cr/V molar ratio between 0.2 and 1.0. XRD results
showed that the active contents were amorphous on the silica surface. Raman, TG,
and XPS analyses revealed that the main active contents were CrVO4 and V2O5, and
the content of CrVO4 increased with the increase of Cr/V molar ratio from 0.2 to
0.6. Results of NH3-TPD indicated a decrease in acidity of catalysts with the rise of
Cr/V ratio. Catalytic experiments revealed that low acidity of catalysts resulted in
low conversion of 3-PIC and high S-NN. Moreover, the increase of reaction
temperature resulted in the increase of conversion of 3-PIC, and the S-NN was
negligibly affected. Consequently, the Cr/V-0.6 catalyst displayed the best
performance in the form of higher selectivity, and Y-NN compared to others
(X-3-PIC = 54 %, Y-NN = 43 %) for its lowest acidity among the catalysts evaluated.
Acknowledgment The authors thank the Natural Science Foundation of China (nos. 21076044 and
20906013) for financial support.
References
1. C. Janke, J. Radnik, U. Bentrup et al., ChemCatChem. 1, 485–491 (2009)
2. B. Manohar, B.M. Reddy, J. Chem. Technol. Biotechnol. 71, 141–146 (1998)
3. A. Wickman, A. Andersson, Appl. Catal. A 391, 110–117 (2011)
4. A. Martin, B. Lucke, Catal. Today 57, 61–70 (2000)
5. A. Martin, V.N. Kalevaru, ChemCatChem 2, 1504–1522 (2010)
6. V.N. Kalevaru, N. Madaan, A. Martin, Appl. Catal. A 391, 52–62 (2011)
7. K.V.R. Chary, C.P. Kumar, D. Naresh et al., J. Mol. Catal. A 243, 149–157 (2006)
8. F. Chen, F.M. Zhang, W.C. Fang et al., Zhejiang Shifan Daxue Xuebao 13, 316–320 (2008)
9. B.D. Raju, K.V. Narayana, S.K. Masthan et al., Catal. Lett. 84, 27–30 (2002)
10. S.K. Masthan, K.V. Narayana, V.V. Rao et al., Catal. Commun. 3, 173–178 (2002)
11. W.J. Yuan, H.F. Lu, B. Zhang et al., Fenzi Cuihua 25, 322–327 (2011)
12. G.Y. Xie, A.Q. Zhang, C. Huang, Res. Chem. Intermed. 36, 969–973 (2010)
13. G.Y. Xie, C. Huang, Jingxi Huagong 24, 790–793 (2007)
14. H.F. Lu, Y. Zhou, H.F. Huang et al., Chin. J. Catal. 26, 101–105 (2005)
15. M. Gu, W. Li, J.L. Lv et al., Petrochem. Technol. 33, 568–569 (2004)
16. N.E. Fouad, J. Therm. Anal. Calorim. 60, 541–547 (2000)
17. M.A. Banares, M.V. Martinez-Huerta, G. Deo et al., J. Phys. Chem. C 111, 18708–18714 (2007)
18. S.B. Xie, E. Iglesia, A.T. Bell, J. Phys. Chem. B 105, 5144–5152 (2001)
19. M.A. Banares, M.V. Martinez-Huerta, X. Gao et al., Catal. Today 118, 279–287 (2006)
20. H.J. Tian, I.E. Wachs, L.E. Briand, J. Phys. Chem. B 109, 23491–23499 (2005)
21. I.E. Wachs, L.E. Briand, J.M. Jehng et al., Catal. Today 78, 257–268 (2003)
22. H.X. Dai, J.G. Deng, L. Zhang et al., Catal. Today 164, 347–352 (2011)
23. H. Dai, L. Zhang, Y. Zhao et al., Catal. Today 131, 42–54 (2008)
24. T. Dogu, G. Karamullaoglu, Ind. Eng. Chem. Res. 46, 7079–7086 (2007)
25. G. Silversmit, D. Depla, H. Poelman et al., J. Electron Spectrosc. Relat. Phenom. 135,
167–175 (2004)
26. S.K. Roy, P. Dutta, L.N. Nandi et al., J. Mol. Catal. A 223, 211–215 (2004)
27. V.N. Kalevaru, B.D. Raju, V.V. Rao et al., Appl. Catal. A 352, 223 (2009)
123