- Email: [email protected]
Al2O3
Catalysis Communications 8 (2007) 1305–1309 www.elsevier.com/locate/catcom
Selective hydrogenation of m-dinitrobenzene to m-nitroaniline catalyzed by PVP-Ru/Al2O3 Songlin Zhao a
a,*
, Huading Liang a, Yafen Zhou
b
Department of Pharmaceutics and Chemical Engineering, Taizhou University, Linhai 317000, China b Department of Chemistry, China West Normal University, Nanchong 637002, China
Received 15 September 2006; received in revised form 19 November 2006; accepted 25 November 2006 Available online 1 December 2006
Abstract Selective hydrogenation of m-dinitrobenzene to m-nitroaniline (m-NA) catalyzed by polylvinylpyrrolidone stabilized Ru/Al2O3 (PVPRu/Al2O3) was studied experimentally. The effects of solvents, metal cation additives and reaction conditions were examined. The highest total yield of m-NA was obtained with 97.9% selectivity at 100% conversion when Sn4+ used as modifier (the molar ratio of m-DNB to catalyst was 477:1, the molar ratio of Sn4+ to ruthenium was 1:4) under suitable conditions. 2006 Elsevier B.V. All rights reserved. Keywords: Selective hydrogenation; m-Dinitrobenzene; m-Nitroaniline; Ruthenium; Metal cations
1. Introduction m-Nitroaniline (m-NA) is an important intermediate for the preparation of dyestuffs, mechanical and electronic corrosion inhibitor. The traditional routes for the chemical were done through: (1) Partial reduction of m-dinitrobenzene (m-DNB) with sodium disulfide or a metal system. (2) From aniline by nitration after acetylation, with subsequent removal of acetyl group by hydrolysis. These methods are hazardous to the environment and required high cost for the waste of disposal, and work-up of the reaction mixture is cumbersome. The catalytic hydrogenation of m-DNB to m-NA should be the method of choice for the preparation of m-NA from m-DNB, provided a high selectivity with respect to m-NA could be realized [1]. The catalytic hydrogenation of m-DNB (shown in Scheme 1) has been reported using catalysts based on both base transition metals such as nickel, copper [2–6] and noble metals such as palladium and platinum [1,7], but which generally leading to the formation of m-phenylenediamine
(m-PDA) [2–9], hence development of a catalytic system to achieve a high selectivity to m-NA is highly desirable. As an important member of the platinum group metals, ruthenium is well known as a novel catalyst for selective hydrogenation of aromatic ring to cycloalkenes, halonitroaromatics to aromatic haloamines, carbonyl group in the vicinity of conjugated or isolated double bonds to corresponding unsaturated alcohol [10]. However, to our knowledge, catalysts based on ruthenium have scarcely been used in selective hydrogenation of m-DNB to m-NA. Herein, an attempt was made to use polylvinylpyrrolidone stabilized Ru/Al2O3 (designated as PVP-Ru/Al2O3) as catalyst in the reaction experimentally. The effects of solvents, metallic cation additives and experimental parameters, such as reaction time, reaction temperature and hydrogen pressure were examined. 2. Experimental 2.1. Materials
*
Corresponding author. Tel.: +86 576 5285621; fax: +86 576 5177066. E-mail address: [email protected] (S. Zhao).
1566-7367/$ - see front matter 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.catcom.2006.11.033
PVP (average molecular weight 40,000) was supplied by BASF. Ruthenium chloride (RuCl3) was purchased from
1306
S. Zhao et al. / Catalysis Communications 8 (2007) 1305–1309 NO2
NH2
NH2
Catalysts + 3H2
Catalysts +3 H2
-2H 2O
NO2
m-DNB
-2H2 O
NO 2
NH2
m-NA
m-PDA
Scheme 1. Pathway of the hydrogenation of m-dinitrobenzene.
Kunming Institute of Precious Metals. The carrier c-alumina, 160–180 mesh, surface area 200 m2 g 1, was supplied by Shanxi Branch of China Aluminium Co., Ltd. Anhydrous metallic chlorides were prepared according to the literature [11]. Hydrogen (H2) with a purity of 99.99% was used for the catalytic experiments (supplied by Sichuan Tianyi Science & Technology Co., Ltd). m-Dinitrobenzene, methanol, ethanol, i-propanol, THF, toluene and ethyl acetate (Beijing Chemicals, analytical grade) were used without further purification. 2.2. Catalyst preparation The catalyst PVP-Ru/Al2O3 was prepared by the reported method [12]: The carrier c-Al2O3 was impregnated in aqueous solution of RuCl3 in the presence of PVP-40000 overnight. Thereafter, the sample was reduced by refluxed mixed alcohol (ethanol and isopropanol in presence of KOH) for 2.5 h, filtered and washed with distilled water for several times until the final filtrate was free from Cl by AgNO3 test, subsequently vacuum dried at 333 K for 24 h. The content of ruthenium is 2.1 wt.% (detected by IRIS Advantage ICP). 2.3. Catalyst characterization X-ray diffraction (XRD) studies were performed with a D/max-TTR instrument equipped with a Cu Ka source (k = 0.1542 nm) and in the range of 10–75 2h. The X-ray photoelectron spectrums (XPS) were recorded using a KRATOS CO XSAM800 spectrometer (Mg Ka X-ray radiation (1215 eV), operated at 10 mA and 12 kV). The vacuum of the chamber was 2 · 10 8 Pa during data acquisition. All binding energy values were referenced to carbon (C 1s 284.9 eV).
tion was set at 1000 rpm. m-DNB conversion and product selectivity were determined by GC-1890-II (Agilent), coupled with FID detector and quartz capillary column (Supelco, SE-30 capillary column, 30 m length · 0.53 mm i.d.), carrier gas, nitrogen. Reactants and products were identified by comparison with samples, and GC–MS coupling. 3. Results and discussion 3.1. The results of XRD and XPS study The XRD pattern of the support (Al2O3, Fig. 1a) was very broad and diffuse. The diffraction peaks at 2h = 31.78, 37.80, 45.35 and 66.81 due to c-Al2O3 were observed. The corresponding diffraction peaks of Al2O3 on PVP-Ru/Al2O3 catalyst XRD pattern (Fig. 1b) were broadened. The result could be explained in terms of strength of the metal–support interactions. Furthermore, a high and homogeneously dispersion of Ru species in catalyst did not allow the detection of the ruthenium or ruthenium oxide phases. Since the binding energy of Ru (around 281 eV) overlapped with that of C 1s (around 284 eV), it was difficult to resolve the tiny Ru peak out from the large peak of C 1s. Two Ru species were detected after a careful deconvolution: Ru0 at 280.3 eV and another corresponding to electron-deficient Ru specie (Rud+) at 281.4 eV. This indicates the incomplete reduction of Ru3+ to Ru0.
b
2.4. Catalytic tests Catalytic hydrogenation of m-DNB was conducted in a 60 mL stainless autoclave equipped with magnetic stirrer. The typical hydrogenation procedure was conducted as follows: The catalyst, m-DNB and solvent were added to the reactor. It was purged with H2 for three times and then H2 was introduced to the desired pressure. The reaction temperature was maintained by a thermostat. The reaction was started by switching the stirrer on. The speed of agita-
a
10
20
30
40 2θ
50
60
70
Fig. 1. XRD patterns of Al2O3 (a) and PVP-Ru/Al2O3 catalyst (b).
S. Zhao et al. / Catalysis Communications 8 (2007) 1305–1309
1307
3.2. The effect of agitation
3.4. Activity and stability of the catalyst
Higher agitation minimizes mass-transfer resistance during reactions. However, it largely depended on the size of the reactor, catalyst loading, substrate concentration and the volume of solvent. The primary study showed that, when the catalyst concentration was kept 2.52 mg/mL (12.6 mg catalyst and 5 mL solvent), the molar ratio of m-DNB to catalyst was 477:1, the agitation had little influent on the reaction rate and selectivity at speed above 900 r/min. Hence, the speed of agitation of all the reactions in this study was set at 1000 rpm.
In order to check for the activity of the catalyst, PVP-Ru/ Al2O3 and Ru/Al2O3 (used as a contrast catalyst prepared without PVP, ruthenium loading was 2.1% by weight) were used for the catalytic reaction under the same conditions. The stability of the catalysts was warranted by submitting the same catalysts to three subsequent recycles of this reaction. The obtained results are summarized in Fig. 2. It appeared clearly that the activity of PVP-Ru/Al2O3 catalyst was marginally higher (conversion, 93.9%) than that of Ru/Al2O3 (conversion, 73.6%). We suspected that the different preparation methods could bring about a difference in catalyst dispersity, which was responsible for the difference in activity. There were no appreciable loss in activity occurred over PVP-Ru/Al2O3 whereas the activity of Ru/Al2O3 decreased considerably with recycles of catalysts.
3.3. The effect of solvents Solvents are known to have a significant effect on the rates of catalytic hydrogenation [13], which include the solubility of hydrogen, the interaction of catalyst with reactants and products, and the competitive adsorption of solvents, etc. The reaction results in different solvents were listed in Table 1. It could be seen that the reduction rate of m-DNB over PVP-Ru/c-Al2O3 in various solvent decreased in the following order: methanol (93.9%) > ethanol (71.7%) > i-propanol (65.2%) > THF (63.6%) > Toluene (60.0%) > Cyclohexane (54.0%) > Ethyl acetate (44.2%) (The number in the parentheses is the conversion of m-DNB). Higher selectivity to m-NA was obtained in protic solvents of methanol (100%), ethanol (99.6%) and i-propanol (97%) compared to that in non-protic solvents of THF (85.8%), Toluene (92.0%), Cyclohexane (89.3%) and Ethyl acetate (92.3%) (The number in the parentheses is the selectivity to m-AN). Some intermediates such as azo- and azoxy-compounds were obviously formed in the non-protic solvents, which were similar to the hydrogenation of other aromatic nitro compounds [14]. Apparently, methanol was the best of all the solvents used for the selective hydrogenation of m-DNB to m-NA. Therefore, methanol was used as solvent in the further reactions.
3.5. The effect of temperature and H2 pressure As expected, with the rise of temperature, the reaction rate increased linearly (Fig. 3). When the temperature was increased from 343 to 373 K, the conversion of m-DNB in 3 h increased from 29.6% to 93.9%. The selectivity to m-NA was exclusive (100%). When the reaction temperature was enhanced to over 380 K, the conversion increased to 100%. H2 pressure effected the reaction in a similar way to that of the reaction temperature. It was observed that there was a marginal increase in the conversion of m-DNB when enhancing H2 pressure. As seen in Fig. 4. It was worth noting that temperature and pressure did not affect the selectivity but reaction rate. The selectivity to m-NA was exclusive. Another nitro-group was almost untouched when the total conversion of m-DNB was not beyond 95%. The significant decreases in selectivity were due to the high overall conversion of m-DNB
100 PVP-Ru/Al2O3
Solvent
Methanol Ethanol i-Propanol THF Toluene Cyclohexane Ethyl acetate
Conversion (%)
93.9 71.7 65.2 63.6 60.0 54.0 44.2
Selectivity (%) m-NA
m-PDA
Othersa
100 99.6 97.0 85.8 92.0 89.3 92.3
0 0.4 3.0 13.7 4.8 6.1 3.4
0 0 0 0.5 3.2 4.6 4.3
Reaction conditions: m-DNB, 1.25 mmol, Catalyst, 12.6 mg (The molar ratio of m-DNB and Ru is 477:1), 373 K, P(H2), (2.0 MPa), 3 h. Solvent, 5 mL. a Mainly azo- and azoxy-compounds.
Conversion (%)
Table 1 Effect of solvents on the hydrogenation of m-DNB over catalyst PVP-Ru/ Al2O3
80
60
40 Ru/Al2O3 20
0 0
1
2
3
Reaction cycle Fig. 2. Catalytic activity of PVP-Ru/Al2O3 and Ru/Al2O3 catalysts in the three subsequent reaction cycles. Reaction conditions were the same as those in Table 1. Methanol used as solvent.
1308
S. Zhao et al. / Catalysis Communications 8 (2007) 1305–1309
3.6. The effect of metal cations
100
Selectivity (%)
Conversion (%)
80
60
40
20
0 340
350
360
370
380
390
400
Temperature (K) Fig. 3. Effect of temperature. Reaction conditions: m-DNB, 1.25 mmol; catalyst, 12.6 mg (the molar ratio of m-DNB and Ru is 477:1), P(H2), (2.0 MPa), 3 h. Methanol, 5 mL. Symbols: r, conversion; ., m-NA; m, m-PDA.
100
Selectivity (%)
Conversion (%)
80
60
40
20
0 1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
Pressure (Mpa) Fig. 4. Effect of pressure. Reaction conditions: m-DNB, 1.25 mmol; catalyst, 12.6 mg (the molar ratio of m-DNB and Ru is 477:1), 373 K, 3 h. Methanol, 5 mL. Symbols were the same as those in Fig. 3.
because of the high reaction rate at high temperature (over 380 K) and pressure (over 2.5 MPa). No intermediates were detected.
The addition of the metal cations modifiers to the catalytic system can considerably modulate both the activity and selectivity of the catalysts. Liu et al. have reported that the metal cations, such as Li+, Fe3+ Co2+, Ni2+, and Zn2+, have significant effect on both the activity and the selectivity of PVP-Pt, PVP-Pd/Pt and PVP-Ru cluster catalysts [15–17]. It is meaningful to investigate such modification of metal cations on the PVP-Ru/Al2O3 catalyst system. The effect of metal cations on the selectivity hydrogenation of m-DNB over PVP-Ru/Al2O3 catalyst was observed under the conditions of 1.5 MPa H2 and 353 K. The metal cations were added to the reaction mixture by anhydrous metal chloride solution of methanol. The results are listed in Table 2. We could see that the conversion of m-DNB was 41.2% with 100% selectivity, when using neat PVPRu/Al2O3 alone as catalyst. Introduced Fe3+ could double the conversion (79.6%) of m-DNB without losing of selectivity. Upon introducing Co2+, Ni2+, and Li+ into the catalytic system, the conversion of m-DNB were all enhanced to 100%. However, the selectivity to m-NA decreased to 90.3%, 88.7% and 79.8%, respectively because of further hydrogenation of m-NA to m-PDA (selectivity were 6.9%, 8.4% and 16.2%, respectively) and the formation of some azo- and azoxy-compounds (total amount were 2.2%, 1.2% and 4.2%, respectively, which were obviously detected by GC) and some high boiling point products (which could not be detected by GC analysis). Hg2+ did not exhibit a very obvious influence on the catalytic activity (41.6% conversion) and selectivity (100% selectivity). It could be seen that there was marked difference in the influences of Sn4+ at different concentration. The conversion of m-DNB increased slightly (from 41.2% to 53.4%) when Sn4+ was added to the system with the molar ratio of Sn4+ and Ru = 1:1, however, ascended to 100% with the molar ratio of Sn4+ and Ru = 1:4. The selectivity to mNA (98.0% and 97.9%, respectively) was almost unaffected. Fig. 5 presents the time course of the hydrogenation of m-DNB over PVP-Ru/Al2O3 in the absence (curves A) and presence (curves B), respectively, of Sn4+ promoter
Table 2 Effect of metal cations on the hydrogenation of m-DNB over PVP-Ru/Al2O3 Entry
1 2 3 4 5 6 7 8
Catalytic systema
PVP-Ru/Al2O3 PVP-Ru/Al2O3-Fe3+ PVP-Ru/Al2O3-Co2+ PVP-Ru/Al2O3-Ni2+ PVP-Ru/Al2O3-Li+ PVP-Ru/Al2O3-Hg2+ PVP-Ru/Al2O3-Sn4+ PVP-Ru/Al2O3-Sn4+(1:4)
Conversion (%)
41.2 79.6 100 100 100 41.6 53.4 100
Selectivity (mol %) m-NA
m-PDA
Othersb
100 100 90.3 88.7 79.8 100 98.0 97.9
– –
– – 2.2 1.2 4.0 – 0.2 –
6.9 8.4 16.2 – 1.8 2.1
Reaction conditions: m-DNB, 1.25 mmol, catalyst, 1.26 · 10 2 g (the molar ratio of m-DNB and Ru is 477:1), 353 K, P(H2), 1.5 MPa, 3 h, total amount of methanol 5 mL. a Molar ratio Mn+:Ru = 1:1 expect Entry 8 is 1:4. b Mainly azo- and azoxy-compounds (some products with high boiling point could not be detected by GC analysis).
S. Zhao et al. / Catalysis Communications 8 (2007) 1305–1309 m-PDA
(curves A)
100
(curves B)
100
m-PDA
m-DNB
m-DNB 80
80
Composition (%)
Concentration (%)
1309
60
40
20
60
40
20
m-NA
m-NA 0
0 0
1
2
3
4
5
6
7
8
0
1
2
3
4
5
Reaction time (h)
Reaction time (h)
Fig. 5. Time course of the hydrogenation of m-DNB over PVP-Ru/Al2O3 in the absence (curves A) and presence (curves B) of Sn 4+ promoter. Reaction conditions were the same as those in Table 2.
(Molar ratio of Sn4+ to Ru was 1:4) under the given conditions. Curves A: in the initial stage (about 1.5 h) the reaction rate was very slow due to the activation stage of catalyst. Up to 98.3% conversion (4.75 h), the reaction was fully selective. Only a small amount of m-PDA was found (4.1%). When m-DNB was totally consumed (5 h), the reaction medium contained m-NA (93.1), m-PDA (6.9%). Later, m-NA was transformed into m-PDA in 2.5 h. Curves B: the reaction occurred in the same manner as with neat PVP-Ru/Al2O3 catalyst (curves A). However, a shorter stage of catalyst activation (about 0.75 h) and a higher rate were observed. m-DNB was totally consumed in 3 h. At the moment, the selectivity to m-NA was 97.9%, which was higher than that of using PVP-Ru/ Al2O3 alone as catalyst (selectivity was 93.1% when mDNB was consumed) and a smaller amount of m-PDA was formed (2.1%). The m-NA was subsequently transformed into m-PDA rapidly in 1.5 h. No azox-, nitrosoor azoxy-compounds were detected during the course of the reaction. The stronger adsorption affinity of m-DNB with catalyst than that of m-NA could be responsible for the preferred hydrogenation of m-DNB to m-NA in a system. It causes the mechanism of metal cations effect very complicated. However, it is generally believed that the interaction between metal cations and the ruthenium particles may change the electronic density of the catalyst active site, which in turn influences the catalytic performance. Furthermore, the coordination effect between the metal cations and amino group of the m-NA could bring an increase in the relative strength of adsorption between m-NA and m-DNB, thus caused the increase of the activity of the catalytic system and selectivity to m-NA at the same time [15]. A detailed and systematic investigation is needed to clear the mechanism. 4. Conclusion The selective hydrogenation of m-DNB to m-NA over PVP-Ru/Al2O3 was observed for the first time. It was
found that the solvents and metal cation additives affected not only the rate but also the selectivity. Protic solvents were more favorable for the reaction and the selectivity to m-NB than that of non-protic solvents. Metal cations, such as Co2+, Ni2+, Li+ and Sn4+, could markedly enhance the reaction rate. Temperature and pressure did not affect the selectivity but reaction rate. m-NA could be obtained at a very high selectivity. Maximum yield of m-NA (97.9%) was obtained in methanol and in the presence of a low concentration of the Sn4+ modifier. References [1] V. Khilnani, S.B. Chandalia, Org. Process Res. Dev. 5 (2001) 263. [2] M.M. Telkar, J.M. Nadgeri, C.V. Rode, R.V. Chaudhari, Appl. Catal. A: Gen. 295 (2005) 23. [3] Y.X. Liu, J.X. Chen, J.Y. Zhang, Chinese J. Catal. 3 (2003) 224. [4] A.T. Masenova, F.B. Bizhanov, Izv. Akad. Nauk Kaz. SSR Ser. Khim. 2 (1993) 32. [5] A. Ono, S. Terasakai, Y. Tsuruoka, Chem. Ind. 12 (1983) 477. [6] W.H. Jones, W.F. Benning, P. Davis, N.Y. Ann, Acad. Sci. 158 (1969) 471. [7] Z.K. Yu, S.J. Liao, Y. Xu, B. Yang, D.R. Yu, J. Mol. Catal. A: Chem. 120 (1997) 247. [8] H. Mizuta, T. Nishimura, M. Wada, T. Nagata, JP 05/331, 113, 1993. [9] H. Mizuta, T. Nishimura, M. Wada, T. Nagata, JP 06/009, 511, 1994. [10] P. Kluson, L. Cerveny, Appl. Catal. A: Gen. 128 (1995) 13, and the references wherein. [11] >.B.Rapzby, B.B.Ayuekod, Pure Chemical Reagents (Chinese Translation, S. Cao, Trans.), Higher Education Press, Beijing, 1989. [12] S.L. Zhao, J.R. Chen, X.J. Li, et al., Chinese J. Catal. 25 (1) (2004) 850. [13] R.A. Rajadhyaksha, S.L. Karwa, Chem. Eng. Sci. 41 (7) (1986) 1765. [14] J. Wisnial, M. Klein, Ind. Eng. Chem. Prod. Res. Dev. 23 (1984) 44. [15] X.L. Yang, H.F. Liu, Appl. Catal. A 164 (1997) 197. [16] M. Liu, W. Yu, H.F. Liu, J. Mol. Catal. A: Chem. 138 (1999) 295. [17] X.L. Yang, H.F. Liu, Z. Hao, J. Mol. Catal. A: Chem. 147 (1999) 55.
Selective hydrogenation of m-dinitrobenzene to m-nitroaniline catalyzed by PVP-Ru/Al2O3 Songlin Zhao a
a,*
, Huading Liang a, Yafen Zhou
b
Department of Pharmaceutics and Chemical Engineering, Taizhou University, Linhai 317000, China b Department of Chemistry, China West Normal University, Nanchong 637002, China
Received 15 September 2006; received in revised form 19 November 2006; accepted 25 November 2006 Available online 1 December 2006
Abstract Selective hydrogenation of m-dinitrobenzene to m-nitroaniline (m-NA) catalyzed by polylvinylpyrrolidone stabilized Ru/Al2O3 (PVPRu/Al2O3) was studied experimentally. The effects of solvents, metal cation additives and reaction conditions were examined. The highest total yield of m-NA was obtained with 97.9% selectivity at 100% conversion when Sn4+ used as modifier (the molar ratio of m-DNB to catalyst was 477:1, the molar ratio of Sn4+ to ruthenium was 1:4) under suitable conditions. 2006 Elsevier B.V. All rights reserved. Keywords: Selective hydrogenation; m-Dinitrobenzene; m-Nitroaniline; Ruthenium; Metal cations
1. Introduction m-Nitroaniline (m-NA) is an important intermediate for the preparation of dyestuffs, mechanical and electronic corrosion inhibitor. The traditional routes for the chemical were done through: (1) Partial reduction of m-dinitrobenzene (m-DNB) with sodium disulfide or a metal system. (2) From aniline by nitration after acetylation, with subsequent removal of acetyl group by hydrolysis. These methods are hazardous to the environment and required high cost for the waste of disposal, and work-up of the reaction mixture is cumbersome. The catalytic hydrogenation of m-DNB to m-NA should be the method of choice for the preparation of m-NA from m-DNB, provided a high selectivity with respect to m-NA could be realized [1]. The catalytic hydrogenation of m-DNB (shown in Scheme 1) has been reported using catalysts based on both base transition metals such as nickel, copper [2–6] and noble metals such as palladium and platinum [1,7], but which generally leading to the formation of m-phenylenediamine
(m-PDA) [2–9], hence development of a catalytic system to achieve a high selectivity to m-NA is highly desirable. As an important member of the platinum group metals, ruthenium is well known as a novel catalyst for selective hydrogenation of aromatic ring to cycloalkenes, halonitroaromatics to aromatic haloamines, carbonyl group in the vicinity of conjugated or isolated double bonds to corresponding unsaturated alcohol [10]. However, to our knowledge, catalysts based on ruthenium have scarcely been used in selective hydrogenation of m-DNB to m-NA. Herein, an attempt was made to use polylvinylpyrrolidone stabilized Ru/Al2O3 (designated as PVP-Ru/Al2O3) as catalyst in the reaction experimentally. The effects of solvents, metallic cation additives and experimental parameters, such as reaction time, reaction temperature and hydrogen pressure were examined. 2. Experimental 2.1. Materials
*
Corresponding author. Tel.: +86 576 5285621; fax: +86 576 5177066. E-mail address: [email protected] (S. Zhao).
1566-7367/$ - see front matter 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.catcom.2006.11.033
PVP (average molecular weight 40,000) was supplied by BASF. Ruthenium chloride (RuCl3) was purchased from
1306
S. Zhao et al. / Catalysis Communications 8 (2007) 1305–1309 NO2
NH2
NH2
Catalysts + 3H2
Catalysts +3 H2
-2H 2O
NO2
m-DNB
-2H2 O
NO 2
NH2
m-NA
m-PDA
Scheme 1. Pathway of the hydrogenation of m-dinitrobenzene.
Kunming Institute of Precious Metals. The carrier c-alumina, 160–180 mesh, surface area 200 m2 g 1, was supplied by Shanxi Branch of China Aluminium Co., Ltd. Anhydrous metallic chlorides were prepared according to the literature [11]. Hydrogen (H2) with a purity of 99.99% was used for the catalytic experiments (supplied by Sichuan Tianyi Science & Technology Co., Ltd). m-Dinitrobenzene, methanol, ethanol, i-propanol, THF, toluene and ethyl acetate (Beijing Chemicals, analytical grade) were used without further purification. 2.2. Catalyst preparation The catalyst PVP-Ru/Al2O3 was prepared by the reported method [12]: The carrier c-Al2O3 was impregnated in aqueous solution of RuCl3 in the presence of PVP-40000 overnight. Thereafter, the sample was reduced by refluxed mixed alcohol (ethanol and isopropanol in presence of KOH) for 2.5 h, filtered and washed with distilled water for several times until the final filtrate was free from Cl by AgNO3 test, subsequently vacuum dried at 333 K for 24 h. The content of ruthenium is 2.1 wt.% (detected by IRIS Advantage ICP). 2.3. Catalyst characterization X-ray diffraction (XRD) studies were performed with a D/max-TTR instrument equipped with a Cu Ka source (k = 0.1542 nm) and in the range of 10–75 2h. The X-ray photoelectron spectrums (XPS) were recorded using a KRATOS CO XSAM800 spectrometer (Mg Ka X-ray radiation (1215 eV), operated at 10 mA and 12 kV). The vacuum of the chamber was 2 · 10 8 Pa during data acquisition. All binding energy values were referenced to carbon (C 1s 284.9 eV).
tion was set at 1000 rpm. m-DNB conversion and product selectivity were determined by GC-1890-II (Agilent), coupled with FID detector and quartz capillary column (Supelco, SE-30 capillary column, 30 m length · 0.53 mm i.d.), carrier gas, nitrogen. Reactants and products were identified by comparison with samples, and GC–MS coupling. 3. Results and discussion 3.1. The results of XRD and XPS study The XRD pattern of the support (Al2O3, Fig. 1a) was very broad and diffuse. The diffraction peaks at 2h = 31.78, 37.80, 45.35 and 66.81 due to c-Al2O3 were observed. The corresponding diffraction peaks of Al2O3 on PVP-Ru/Al2O3 catalyst XRD pattern (Fig. 1b) were broadened. The result could be explained in terms of strength of the metal–support interactions. Furthermore, a high and homogeneously dispersion of Ru species in catalyst did not allow the detection of the ruthenium or ruthenium oxide phases. Since the binding energy of Ru (around 281 eV) overlapped with that of C 1s (around 284 eV), it was difficult to resolve the tiny Ru peak out from the large peak of C 1s. Two Ru species were detected after a careful deconvolution: Ru0 at 280.3 eV and another corresponding to electron-deficient Ru specie (Rud+) at 281.4 eV. This indicates the incomplete reduction of Ru3+ to Ru0.
b
2.4. Catalytic tests Catalytic hydrogenation of m-DNB was conducted in a 60 mL stainless autoclave equipped with magnetic stirrer. The typical hydrogenation procedure was conducted as follows: The catalyst, m-DNB and solvent were added to the reactor. It was purged with H2 for three times and then H2 was introduced to the desired pressure. The reaction temperature was maintained by a thermostat. The reaction was started by switching the stirrer on. The speed of agita-
a
10
20
30
40 2θ
50
60
70
Fig. 1. XRD patterns of Al2O3 (a) and PVP-Ru/Al2O3 catalyst (b).
S. Zhao et al. / Catalysis Communications 8 (2007) 1305–1309
1307
3.2. The effect of agitation
3.4. Activity and stability of the catalyst
Higher agitation minimizes mass-transfer resistance during reactions. However, it largely depended on the size of the reactor, catalyst loading, substrate concentration and the volume of solvent. The primary study showed that, when the catalyst concentration was kept 2.52 mg/mL (12.6 mg catalyst and 5 mL solvent), the molar ratio of m-DNB to catalyst was 477:1, the agitation had little influent on the reaction rate and selectivity at speed above 900 r/min. Hence, the speed of agitation of all the reactions in this study was set at 1000 rpm.
In order to check for the activity of the catalyst, PVP-Ru/ Al2O3 and Ru/Al2O3 (used as a contrast catalyst prepared without PVP, ruthenium loading was 2.1% by weight) were used for the catalytic reaction under the same conditions. The stability of the catalysts was warranted by submitting the same catalysts to three subsequent recycles of this reaction. The obtained results are summarized in Fig. 2. It appeared clearly that the activity of PVP-Ru/Al2O3 catalyst was marginally higher (conversion, 93.9%) than that of Ru/Al2O3 (conversion, 73.6%). We suspected that the different preparation methods could bring about a difference in catalyst dispersity, which was responsible for the difference in activity. There were no appreciable loss in activity occurred over PVP-Ru/Al2O3 whereas the activity of Ru/Al2O3 decreased considerably with recycles of catalysts.
3.3. The effect of solvents Solvents are known to have a significant effect on the rates of catalytic hydrogenation [13], which include the solubility of hydrogen, the interaction of catalyst with reactants and products, and the competitive adsorption of solvents, etc. The reaction results in different solvents were listed in Table 1. It could be seen that the reduction rate of m-DNB over PVP-Ru/c-Al2O3 in various solvent decreased in the following order: methanol (93.9%) > ethanol (71.7%) > i-propanol (65.2%) > THF (63.6%) > Toluene (60.0%) > Cyclohexane (54.0%) > Ethyl acetate (44.2%) (The number in the parentheses is the conversion of m-DNB). Higher selectivity to m-NA was obtained in protic solvents of methanol (100%), ethanol (99.6%) and i-propanol (97%) compared to that in non-protic solvents of THF (85.8%), Toluene (92.0%), Cyclohexane (89.3%) and Ethyl acetate (92.3%) (The number in the parentheses is the selectivity to m-AN). Some intermediates such as azo- and azoxy-compounds were obviously formed in the non-protic solvents, which were similar to the hydrogenation of other aromatic nitro compounds [14]. Apparently, methanol was the best of all the solvents used for the selective hydrogenation of m-DNB to m-NA. Therefore, methanol was used as solvent in the further reactions.
3.5. The effect of temperature and H2 pressure As expected, with the rise of temperature, the reaction rate increased linearly (Fig. 3). When the temperature was increased from 343 to 373 K, the conversion of m-DNB in 3 h increased from 29.6% to 93.9%. The selectivity to m-NA was exclusive (100%). When the reaction temperature was enhanced to over 380 K, the conversion increased to 100%. H2 pressure effected the reaction in a similar way to that of the reaction temperature. It was observed that there was a marginal increase in the conversion of m-DNB when enhancing H2 pressure. As seen in Fig. 4. It was worth noting that temperature and pressure did not affect the selectivity but reaction rate. The selectivity to m-NA was exclusive. Another nitro-group was almost untouched when the total conversion of m-DNB was not beyond 95%. The significant decreases in selectivity were due to the high overall conversion of m-DNB
100 PVP-Ru/Al2O3
Solvent
Methanol Ethanol i-Propanol THF Toluene Cyclohexane Ethyl acetate
Conversion (%)
93.9 71.7 65.2 63.6 60.0 54.0 44.2
Selectivity (%) m-NA
m-PDA
Othersa
100 99.6 97.0 85.8 92.0 89.3 92.3
0 0.4 3.0 13.7 4.8 6.1 3.4
0 0 0 0.5 3.2 4.6 4.3
Reaction conditions: m-DNB, 1.25 mmol, Catalyst, 12.6 mg (The molar ratio of m-DNB and Ru is 477:1), 373 K, P(H2), (2.0 MPa), 3 h. Solvent, 5 mL. a Mainly azo- and azoxy-compounds.
Conversion (%)
Table 1 Effect of solvents on the hydrogenation of m-DNB over catalyst PVP-Ru/ Al2O3
80
60
40 Ru/Al2O3 20
0 0
1
2
3
Reaction cycle Fig. 2. Catalytic activity of PVP-Ru/Al2O3 and Ru/Al2O3 catalysts in the three subsequent reaction cycles. Reaction conditions were the same as those in Table 1. Methanol used as solvent.
1308
S. Zhao et al. / Catalysis Communications 8 (2007) 1305–1309
3.6. The effect of metal cations
100
Selectivity (%)
Conversion (%)
80
60
40
20
0 340
350
360
370
380
390
400
Temperature (K) Fig. 3. Effect of temperature. Reaction conditions: m-DNB, 1.25 mmol; catalyst, 12.6 mg (the molar ratio of m-DNB and Ru is 477:1), P(H2), (2.0 MPa), 3 h. Methanol, 5 mL. Symbols: r, conversion; ., m-NA; m, m-PDA.
100
Selectivity (%)
Conversion (%)
80
60
40
20
0 1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
Pressure (Mpa) Fig. 4. Effect of pressure. Reaction conditions: m-DNB, 1.25 mmol; catalyst, 12.6 mg (the molar ratio of m-DNB and Ru is 477:1), 373 K, 3 h. Methanol, 5 mL. Symbols were the same as those in Fig. 3.
because of the high reaction rate at high temperature (over 380 K) and pressure (over 2.5 MPa). No intermediates were detected.
The addition of the metal cations modifiers to the catalytic system can considerably modulate both the activity and selectivity of the catalysts. Liu et al. have reported that the metal cations, such as Li+, Fe3+ Co2+, Ni2+, and Zn2+, have significant effect on both the activity and the selectivity of PVP-Pt, PVP-Pd/Pt and PVP-Ru cluster catalysts [15–17]. It is meaningful to investigate such modification of metal cations on the PVP-Ru/Al2O3 catalyst system. The effect of metal cations on the selectivity hydrogenation of m-DNB over PVP-Ru/Al2O3 catalyst was observed under the conditions of 1.5 MPa H2 and 353 K. The metal cations were added to the reaction mixture by anhydrous metal chloride solution of methanol. The results are listed in Table 2. We could see that the conversion of m-DNB was 41.2% with 100% selectivity, when using neat PVPRu/Al2O3 alone as catalyst. Introduced Fe3+ could double the conversion (79.6%) of m-DNB without losing of selectivity. Upon introducing Co2+, Ni2+, and Li+ into the catalytic system, the conversion of m-DNB were all enhanced to 100%. However, the selectivity to m-NA decreased to 90.3%, 88.7% and 79.8%, respectively because of further hydrogenation of m-NA to m-PDA (selectivity were 6.9%, 8.4% and 16.2%, respectively) and the formation of some azo- and azoxy-compounds (total amount were 2.2%, 1.2% and 4.2%, respectively, which were obviously detected by GC) and some high boiling point products (which could not be detected by GC analysis). Hg2+ did not exhibit a very obvious influence on the catalytic activity (41.6% conversion) and selectivity (100% selectivity). It could be seen that there was marked difference in the influences of Sn4+ at different concentration. The conversion of m-DNB increased slightly (from 41.2% to 53.4%) when Sn4+ was added to the system with the molar ratio of Sn4+ and Ru = 1:1, however, ascended to 100% with the molar ratio of Sn4+ and Ru = 1:4. The selectivity to mNA (98.0% and 97.9%, respectively) was almost unaffected. Fig. 5 presents the time course of the hydrogenation of m-DNB over PVP-Ru/Al2O3 in the absence (curves A) and presence (curves B), respectively, of Sn4+ promoter
Table 2 Effect of metal cations on the hydrogenation of m-DNB over PVP-Ru/Al2O3 Entry
1 2 3 4 5 6 7 8
Catalytic systema
PVP-Ru/Al2O3 PVP-Ru/Al2O3-Fe3+ PVP-Ru/Al2O3-Co2+ PVP-Ru/Al2O3-Ni2+ PVP-Ru/Al2O3-Li+ PVP-Ru/Al2O3-Hg2+ PVP-Ru/Al2O3-Sn4+ PVP-Ru/Al2O3-Sn4+(1:4)
Conversion (%)
41.2 79.6 100 100 100 41.6 53.4 100
Selectivity (mol %) m-NA
m-PDA
Othersb
100 100 90.3 88.7 79.8 100 98.0 97.9
– –
– – 2.2 1.2 4.0 – 0.2 –
6.9 8.4 16.2 – 1.8 2.1
Reaction conditions: m-DNB, 1.25 mmol, catalyst, 1.26 · 10 2 g (the molar ratio of m-DNB and Ru is 477:1), 353 K, P(H2), 1.5 MPa, 3 h, total amount of methanol 5 mL. a Molar ratio Mn+:Ru = 1:1 expect Entry 8 is 1:4. b Mainly azo- and azoxy-compounds (some products with high boiling point could not be detected by GC analysis).
S. Zhao et al. / Catalysis Communications 8 (2007) 1305–1309 m-PDA
(curves A)
100
(curves B)
100
m-PDA
m-DNB
m-DNB 80
80
Composition (%)
Concentration (%)
1309
60
40
20
60
40
20
m-NA
m-NA 0
0 0
1
2
3
4
5
6
7
8
0
1
2
3
4
5
Reaction time (h)
Reaction time (h)
Fig. 5. Time course of the hydrogenation of m-DNB over PVP-Ru/Al2O3 in the absence (curves A) and presence (curves B) of Sn 4+ promoter. Reaction conditions were the same as those in Table 2.
(Molar ratio of Sn4+ to Ru was 1:4) under the given conditions. Curves A: in the initial stage (about 1.5 h) the reaction rate was very slow due to the activation stage of catalyst. Up to 98.3% conversion (4.75 h), the reaction was fully selective. Only a small amount of m-PDA was found (4.1%). When m-DNB was totally consumed (5 h), the reaction medium contained m-NA (93.1), m-PDA (6.9%). Later, m-NA was transformed into m-PDA in 2.5 h. Curves B: the reaction occurred in the same manner as with neat PVP-Ru/Al2O3 catalyst (curves A). However, a shorter stage of catalyst activation (about 0.75 h) and a higher rate were observed. m-DNB was totally consumed in 3 h. At the moment, the selectivity to m-NA was 97.9%, which was higher than that of using PVP-Ru/ Al2O3 alone as catalyst (selectivity was 93.1% when mDNB was consumed) and a smaller amount of m-PDA was formed (2.1%). The m-NA was subsequently transformed into m-PDA rapidly in 1.5 h. No azox-, nitrosoor azoxy-compounds were detected during the course of the reaction. The stronger adsorption affinity of m-DNB with catalyst than that of m-NA could be responsible for the preferred hydrogenation of m-DNB to m-NA in a system. It causes the mechanism of metal cations effect very complicated. However, it is generally believed that the interaction between metal cations and the ruthenium particles may change the electronic density of the catalyst active site, which in turn influences the catalytic performance. Furthermore, the coordination effect between the metal cations and amino group of the m-NA could bring an increase in the relative strength of adsorption between m-NA and m-DNB, thus caused the increase of the activity of the catalytic system and selectivity to m-NA at the same time [15]. A detailed and systematic investigation is needed to clear the mechanism. 4. Conclusion The selective hydrogenation of m-DNB to m-NA over PVP-Ru/Al2O3 was observed for the first time. It was
found that the solvents and metal cation additives affected not only the rate but also the selectivity. Protic solvents were more favorable for the reaction and the selectivity to m-NB than that of non-protic solvents. Metal cations, such as Co2+, Ni2+, Li+ and Sn4+, could markedly enhance the reaction rate. Temperature and pressure did not affect the selectivity but reaction rate. m-NA could be obtained at a very high selectivity. Maximum yield of m-NA (97.9%) was obtained in methanol and in the presence of a low concentration of the Sn4+ modifier. References [1] V. Khilnani, S.B. Chandalia, Org. Process Res. Dev. 5 (2001) 263. [2] M.M. Telkar, J.M. Nadgeri, C.V. Rode, R.V. Chaudhari, Appl. Catal. A: Gen. 295 (2005) 23. [3] Y.X. Liu, J.X. Chen, J.Y. Zhang, Chinese J. Catal. 3 (2003) 224. [4] A.T. Masenova, F.B. Bizhanov, Izv. Akad. Nauk Kaz. SSR Ser. Khim. 2 (1993) 32. [5] A. Ono, S. Terasakai, Y. Tsuruoka, Chem. Ind. 12 (1983) 477. [6] W.H. Jones, W.F. Benning, P. Davis, N.Y. Ann, Acad. Sci. 158 (1969) 471. [7] Z.K. Yu, S.J. Liao, Y. Xu, B. Yang, D.R. Yu, J. Mol. Catal. A: Chem. 120 (1997) 247. [8] H. Mizuta, T. Nishimura, M. Wada, T. Nagata, JP 05/331, 113, 1993. [9] H. Mizuta, T. Nishimura, M. Wada, T. Nagata, JP 06/009, 511, 1994. [10] P. Kluson, L. Cerveny, Appl. Catal. A: Gen. 128 (1995) 13, and the references wherein. [11] >.B.Rapzby, B.B.Ayuekod, Pure Chemical Reagents (Chinese Translation, S. Cao, Trans.), Higher Education Press, Beijing, 1989. [12] S.L. Zhao, J.R. Chen, X.J. Li, et al., Chinese J. Catal. 25 (1) (2004) 850. [13] R.A. Rajadhyaksha, S.L. Karwa, Chem. Eng. Sci. 41 (7) (1986) 1765. [14] J. Wisnial, M. Klein, Ind. Eng. Chem. Prod. Res. Dev. 23 (1984) 44. [15] X.L. Yang, H.F. Liu, Appl. Catal. A 164 (1997) 197. [16] M. Liu, W. Yu, H.F. Liu, J. Mol. Catal. A: Chem. 138 (1999) 295. [17] X.L. Yang, H.F. Liu, Z. Hao, J. Mol. Catal. A: Chem. 147 (1999) 55.