- Email: [email protected]
Green route for the chlorination of nitrobenzene
G Model APCATA-13877; No. of Pages 8
ARTICLE IN PRESS Applied Catalysis A: General xxx (2012) xxx–xxx
Contents lists available at SciVerse ScienceDirect
Applied Catalysis A: General journal homepage: www.elsevier.com/locate/apcata
Green route for the chlorination of nitrobenzene Marilyne Boltz a , Márcio C.S. de Mattos b , Pierre M. Esteves b,∗ , Patrick Pale a , Benoit Louis a,∗∗ a Laboratoire de Synthèse Réactivité Organique et Catalyse (LASYROC), UMR 7177, Institut de Chimie, Université de Strasbourg, 1 rue Blaise Pascal, 67000 Strasbourg Cedex, France b Universidade Federal do Rio de Janeiro, Instituto de Química, Av. Athos da Silveira Ramos 149, CT Bloco A, Cidade Universitária, 21941-909 Rio de Janeiro, Brazil
a r t i c l e
i n f o
Article history: Received 12 July 2012 Received in revised form 21 August 2012 Accepted 17 September 2012 Available online xxx Keywords: Chlorination Nitrobenzene Trichloroisocyanuric acid (TCCA) Zeolite Green chemistry
a b s t r a c t A new green chlorination process of deactivated aromatics has been developed, being environmentalfriendly and allowing the continuous chlorination of 1.7 kg nitrobenzene/kg catalyst per day. The triple novelty consists of using a non-conventional chlorination agent, the trichloroisocyanuric acid (TCCA, C3 N3 O3 Cl3 ), along with solid acid catalysts (mainly zeolites) in a continuous flow reactor system. Different zeolites and solid acids have been tested in the chlorination of nitrobenzene, chosen as a model deactivated aromatic substrate. HUSY zeolite was found as the more promising catalyst for performing the chlorination of nitrobenzene, with good conversions (39–64%) at high selectivity toward monochlorinated products (90–99%). Finally, it is worthy to note that HUSY zeolite could be reused for at least five successive runs. © 2012 Elsevier B.V. All rights reserved.
1. Introduction Chloroarenes are valuable starting molecules in fine chemistry, for the synthesis of dyes, bio-active compounds such as pesticides or pharmaceuticals [1,2]. Unfortunately, the conventional industrial methods used for the chlorination of aromatics usually produce mixtures of regioisomers, difficult to separate, thus raising the cost for the industry [3]. Indeed, the utilization of Cl2 , a toxic gas, as chlorine source produces a large quantity of non-recyclable wastes. Another disadvantage is the use of toxic acid catalysts such as Lewis acids (aluminum chloride or boron trifluoride), which are consumed and need to be neutralized after the reaction. In parallel, strong Brønsted acid catalysts such as H2 SO4 remain very corrosive and generate a lot of salts as co-products. During the past decades, researchers focused on the development of more efficient and selective processes for the chlorination of arenes [4–7]. For instance, Esteves et al. [7a] developed a methodology for the chlorination of deactivated arenes using trichloroisocyanuric acid (TCCA, C3 N3 O3 Cl3 ) in a superelectrophilic medium.
∗ Corresponding author. Tel.: +55 2125627444. ∗∗ Corresponding author. Tel.: +33 368851344. E-mail addresses: [email protected] (P.M. Esteves), [email protected] (B. Louis).
TCCA is a stable and inexpensive solid, easily available in pool supplies, being thus frequently used as swimming-pool disinfectant and bleaching agent. It is an efficient chlorine source due to its high chlorine content, which can be up to 45.5% in weight, allowing a priori a higher atomic efficiency than its N-chloro analogs [7b]. TCCA was successfully used for chlorination of electron-rich arenes [7b], alkenes [8] and carbonyl compounds [9], preparation of N-chloro substrates [10,11] and in diverse oxidation reactions [12]. Nowadays, organic chemistry is increasingly moving toward green chemistry, where both homogeneous and heterogeneous catalysis are at the core of this concept. In heterogeneously catalyzed reactions, zeolites represent the key solid acid materials in petrochemistry [13,14]. In addition, thanks to their internal porosity, zeolites provide a highly organized channel structure and strong acidity. The combination of these properties led them to exhibit high conversions and excellent selectivities in targeted reactions [15,16]. The chlorination of nitrobenzene is a very important process for the production of series of useful compounds in dye chemistry [17]. It is worthy to study the chlorination of nitrobenzene in order to produce valuable intermediates for the synthesis of azo dyes. The aim of the present study is therefore to develop a new chlorination process of deactivated aromatics, being more environmental-friendly by the use of solid acids instead of liquid acids. In order to achieve this goal, we set up both a catalytic and a
0926-860X/$ – see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.apcata.2012.09.021
Please cite this article in press as: M. Boltz, et al., Appl. Catal. A: Gen. (2012), http://dx.doi.org/10.1016/j.apcata.2012.09.021
G Model APCATA-13877; No. of Pages 8
ARTICLE IN PRESS M. Boltz et al. / Applied Catalysis A: General xxx (2012) xxx–xxx
2
Scheme 1. Chloration reaction of deactivated arenes over solid acid catalysts.
continuous process to chlorinate nitrobenzene (chosen as a model deactivated arene) while using TCCA as a chlorination agent and different solid acids as catalysts (Scheme 1). 2. Experimental 2.1. Reagents and catalysts Trichloroisocyanuric acid (98%, Aldrich) and nitrobenzene (99%, Merck) were used as received. Several commercial zeolites were employed: H-Y (Aldrich), H-BETA (Zeolyst), H-MOR (Zeolyst), HEMT (IS2M, Mulhouse), H-USY (Zeolyst), H-FER (Petrobras). Prior to use, these zeolites were activated by calcination at 550 ◦ C overnight under air. In addition, H-USY zeolite has been chemically modified to investigate the influence of its Lewis acidity. H-USY zeolite (200 mg) was suspended in a 0.1 M solution of Na2 H2 -EDTA (17.4 mL, 2 equiv.). The mixture was heated at 62 ◦ C and stirred for 24 h. Afterwards, the solution was cooled down to room temperature, filtered over a nylon membrane and thoroughly washed with distilled water (60 mL). The solid was reintroduced in the flask and a 1 M solution of NH4 Cl (20 mL) was added. The solution was then heated at 80 ◦ C and stirred for 24 h. Finally, the solution was filtered, washed with distilled water (60 mL) and dried at 105 ◦ C in an oven. According to van Bokhoven et al. [18], extra-framework aluminum (EFAl) was completely removed from pristine FAU material. 2.2. Brønsted acid sites titration Our home-developed H/D exchange isotope technique was used to quantify the number of Brønsted acid sites present in solid acids [19,20]. Fig. 1 presents the experimental set-up used to perform this
Fig. 2. Experimental setup to perform chlorination in a continuous flow system.
titration method. The catalyst (300 mg) was first activated under dry nitrogen flow (40 mL/min) at 450 ◦ C for 1 h to desorb water. The deuteration of the catalyst was carried out at 200 ◦ C by flowing, through the catalyst, nitrogen (40 mL/min), previously bubbled at room temperature through a U-shaped tube containing D2 O (about 0.05 g), during 1 h. Dry nitrogen was then sweeping through the sample during 60 min to remove the excess of D2 O. After this purge, the titration of O-D sites was performed by back-exchanging the deuterium present on the solid surface with distilled water (3% of H2 O in N2 stream) at 200 ◦ C for 1 h. During this step the partially exchanged water named Hx ODy , composed by H2 O, D2 O and HDO, was collected in a cold U-tube trap at −117 ◦ C. Collected Hx ODy was then weighted and allowed to react with trifluoroacetic anhydride, used in a two-fold excess. The acid solution obtained was then transferred to a NMR tube for analysis. Spectra were recorded on a Bruker AM400 spectrometer, after addition of a CDCl3 (10 wt%)/CHCl3 mixture used as reference. Finally, an accurate quantification of the H/D content in the different samples was achieved by the integration of CF3 COOH(D) and CH(D)Cl3 on both 1 H and 2 H spectra. 2.3. Continuous chlorination process
Fig. 1. Experimental setup to perform the H/D isotope exchange technique.
All the reactions were performed in a glass flow system with a cylindrical reactor. The gas flow was regulated by means of Brooks 5850E mass flow controllers and the dry nitrogen flow was set to 100 mL/min for each experiments. Fig. 2 shows the experimental set-up to perform these catalytic tests in a continuous manner. The reactions were carried out by diluting the catalytic bed (TCCA and catalyst) in an amorphous silica (Grace, USA) matrix to insure the same height for all catalyst beds. To investigate the influence of the solid acid structure on its catalytic performance, we have decided to work at iso-site conditions. Indeed, all the catalytic tests were performed by keeping a constant number of Brønsted acid sites for the catalysts, which was set to 0.44 mmol H+ . Table 1
Please cite this article in press as: M. Boltz, et al., Appl. Catal. A: Gen. (2012), http://dx.doi.org/10.1016/j.apcata.2012.09.021
G Model APCATA-13877; No. of Pages 8
ARTICLE IN PRESS M. Boltz et al. / Applied Catalysis A: General xxx (2012) xxx–xxx
Table 1 Brønsted acidity of the solid acids used in the chlorination process. Solid acids
Si/Al
mmol H+ /g
H-MOR H-FER H-EMT H-BETA H-ZSM-5 (zeolyst) H-Y (Aldrich) H-Y (UOP) H-USY Silica–alumina H3 PW12 O40 Cs2 PW12 O40 SAC-13 Ga-SBA-15
5 10.5 6.5 13.6 15 2.2 5.0 2.8 3.6 – – – –
1.90 2.26 2.38 1.07 1.04 5.34 2.20 3.90 3.68 0.91 0.32 1.00 1.87
presents the number of Brønsted acid sites titrated by H/D exchange for all solid acids used for the chlorination reaction. The solid acid catalyst (0.44 mmol H+ ), trichloroisocyanuric acid (0.15 mmol) and silica matrix (17 mmol, 1 g) were blended closely by grinding. The mixture was then transferred into the cylindrical reactor and the reactor was fixed to the set-up. The catalytic bed was first dried under dry N2 flow at 150 ◦ C for 30 min to desorb the water present in the void volume of the zeolite. Then, nitrobenzene was supplied in its gaseous state by sweeping a dry N2 flow through a stripping U-shaped reactor containing liquid nitrobenzene (Fig. 2). Hence, this dry nitrogen flow saturated with nitrobenzene’s vapor pressure was allowed to pass through the catalytic bed during 5 h. The products were trapped at −196 ◦ C and recovered downstream to the reactor with toluene (4 mL). They were analyzed by gas chromatography (HP 5890 Series II) equipped with a capillary column (PONA, 50 m). Retention times were compared with standards and used to characterize the different reaction products. The degree of conversion and the selectivity toward the different products were calculated by taking into account the response factor of the substrate (nitrobenzene) and those from the products (mono-, di- and tri-chlorinated aromatics) through the use of an external standard (styrene). 3. Results and discussion 3.1. Methodology In preliminary experiments, we initially performed the reaction between TCCA and nitrobenzene by changing only one parameter between each experiment. Firstly, we performed a screening of the different solid acid catalysts (working under iso-Brønsted site conditions). The effects of pore topologies and sizes along with the quantity of Brønsted acid sites were investigated among the different solid acids. Furthermore, we also decided to refine the number of catalysts tested to find adapted reaction conditions by analyzing the effect of temperature on the performance of the more promising catalyst. In parallel, the influence of Lewis acidity (EFAl) has been evaluated for chosen “optimal” catalyst. Finally, the chlorination process was further optimized by modifying different reaction parameters, as quantities of TCCA, the weight hourly space velocity (WHSV) while varying the catalyst mass. Likewise, the potential reuse of the catalyst was also tested. 3.2. Behavior of the different solid acid catalysts Fig. 3 shows the degree of nitrobenzene conversion and the selectivity to mono-chloronitrobenzenes among the different catalysts. It is noteworthy that the meta-chloronitrobenzene was always found as prominent isomer, i.e.; >90% with respect to the
3
ortho- and para-chloronitrobenzene, as a priori expected for the nitro electron attracting group. It seems that the zeolite pore size acts as a limiting factor for this reaction. Whilst no reaction occurred over H-FER zeolite (∼4 A˚ pore aperture), nitrobenzene reacted with TCCA over medium pore H˚ ZSM-5 zeolite, which possesses a rather narrow pore size of 5.5 A, affording 21% conversion. In contrast, 39% of nitrobenzene was con˚ This tends to support an verted over H-USY (pore size of 7.4 A). effect of higher accessibility of those large organic molecules to the catalytic sites in H-USY zeolite [21]. On the other hand, the chlorination reaction appears to be positively influenced by the channel architecture of a zeolite framework. Indeed, the SAC-13 catalyst (Nafion-H supported over MCM-41), being a polymer-grafted on mesoporous silica, only led to a 5% nitrobenzene conversion (Fig. 4). Likewise, gallium-doped mesoporous silica (Ga-SBA-15) [22] also led to poor catalytic performance (conversion of 7%) with respect to zeolites. Following the same trend, the selectivity toward monochlorination is also governed by the zeolite porosity. In parallel, the beneficial effect of microporosity has been further evidenced with the use of strongly acidic polyoxometalate H3 PW12 O40 . Whereas, this non porous solid acid led to an appreciable degree of nitrobenzene conversion (14%), in contrast the selectivity drastically diminished with respect to zeolites in favor of dichlorination products. However, Cs2 HPW12 O40 that exhibits a higher SSA value (123 m2 /g) and porosity with respect to its protonic counterpart (5 m2 /g) led to 2.6 fold increase in activity. The same behavior was observed with amorphous silica–alumina, which converted only 5% of nitrobenzene probably due to the absence of zeolite’s highly organized pore architecture. In addition, a non-conventional acid catalyst, sulfated tin dioxide, was also tested. This SnO2 /SO4 2− catalyst, which possesses exclusively Lewis acid sites and almost no porosity, was not a good candidate at all, since no activity could be observed. Besides, a rather surprising difference in reactivity between H-Y and H-USY zeolites was observed (Fig. 3). H-USY zeolite is usually obtained by dealumination via a high temperature steaming of pristine H-Y zeolite [23]. The dealumination process consists of removing aluminum atoms from the zeolite framework and thus to create new extra-framework aluminum species (EFAl) present either as amorphous species (pseudo Al2 O3 ) [24] or aluminum species exhibiting exalted Lewis acidity (AlOOH, AlO+ species, etc.) [25]. This phenomenon is beneficial for the chlorination process since the conversion of nitrobenzene was drastically improved to 39% with H-USY compared to HY zeolite (16%). To summarize these preliminary tests, one can observe that the optimal catalyst for performing the chlorination of nitrobenzene remains the H-USY zeolite. Indeed, the faujasite structure allows catalyzing the chlorination of nitrobenzene by combining a high selectivity in monochlorinated compounds (99%) at an acceptable degree of nitrobenzene conversion. 3.3. Influence of reaction temperature Fig. 5 shows the results of the chlorination reaction at different reaction temperatures. As ‘a priori’ expected, a raise in the temperature led to higher nitrobenzene conversion, but at the expense of the selectivity in monochlorinated products. The decrease in selectivity is due to the formation of di- and tri-chlorinated products. For instance, the selectivity toward monochlorination was diminished to 63% at 200 ◦ C. Selectivities in di-chlorinated and trichlorinated products were 32% and 5%, respectively. Among the products of di-chlorination, 1,4-dichloro,2-nitrobenzene was formed as major isomer (71%) whilst 1,2-dichloro,4-nitrobenzene represented nearly 28% selectivity among the dichlorinated products. It is therefore worthy to mention that the second chlorination
Please cite this article in press as: M. Boltz, et al., Appl. Catal. A: Gen. (2012), http://dx.doi.org/10.1016/j.apcata.2012.09.021
G Model APCATA-13877; No. of Pages 8
ARTICLE IN PRESS M. Boltz et al. / Applied Catalysis A: General xxx (2012) xxx–xxx
4
Fig. 3. Nitrobenzene conversion and selectivity to monochlorination products over different zeolite catalysts. Conditions: nTCCA = 0.15 mmol; nH+ = 0.44 mmol; reaction duration = 5 h; temperature = 150 ◦ C.
(involving deactivated meta-chloronitrobenzene) is favored in ortho and para positions of the chlorine atom. The higher selectivity in 1,4-dichloro,2-nitrobenzene major isomer might be explain by lower steric hindrance than 1,2-dichloro,4-nitrobenzene isomer. Based on these results, we have arbitrary chosen the temperature of 150 ◦ C as an optimal temperature, since a good compromise between a high selectivity and a reasonable conversion of nitrobenzene was achieved. 3.4. Influence of the Lewis acidity In order to highlight the effect of Lewis acidity in H-USY zeolite, we have performed a treatment to remove extra-framework aluminum species (EFAl) by complexing them with EDTA. The catalyst was then analyzed by 27 Al MAS NMR to assess the disappearance of the characteristic signal attributed to octahedral
species (EFAl) at nearly 0 ppm (Fig. 6). Two distinct signals were detected over pristine HUSY zeolite (before EDTA treatment): one located at 50 ppm and one broad near 0 ppm (Fig. 6a). These signals correspond, respectively, to framework aluminum species and to extra-framework aluminum (EFAl) species. The later species represent about 29% of the total aluminum present within the zeolite. Fig. 6b, corresponding to the zeolite HUSY after one EDTA treatment, shows the quasi-disappearance of the EFAl signal with only 12% of the remaining aluminum present as EFAl in the zeolite channels. It appears therefore that one EDTA treatment alone did not remove all the EFAl species in the H-USY. As a control experiment, no EFAl could be detected in the H-Y spectrum (Fig. 6c), all the aluminum cations are present in the framework as Tatoms. Furthermore, this dealumination treatment had another impact on the structural properties of H-USY zeolite. The complexation of EFAl species led to intense leaching of those species, thus reducing
Please cite this article in press as: M. Boltz, et al., Appl. Catal. A: Gen. (2012), http://dx.doi.org/10.1016/j.apcata.2012.09.021
G Model APCATA-13877; No. of Pages 8
ARTICLE IN PRESS M. Boltz et al. / Applied Catalysis A: General xxx (2012) xxx–xxx
5
Fig. 4. Comparison in terms of nitrobenzene conversion and selectivity toward monochlorination between HUSY zeolite and non-zeolitic solid acids. Conditions: nTCCA = 0.15 mmol; nH+ = 0.44 mmol; reaction duration = 5 h; temperature = 150 ◦ C.
pore obstruction, thus introducing also a high level of mesoporosity inside the FAU structure. Hence the diffusion of products/reactants can be seriously improved [26,27]. Fig. 7 presents the catalytic data obtained with the H-USY zeolite treated and non-treated with EDTA. One can observe that the presence or not of EFAl species strongly influence the catalytic performance of H-USY zeolite. Indeed, the EDTA treatment of H-USY led to a significant increase in nitrobenzene conversion, up to 64% while the selectivity in monochlorinated products remained slightly affected. This phenomenon could be due to the higher accessibility of the reactants to the active sites within the pores, emptied from their alumina aggregates [24]. Hence, the molecules easily access to the active sites, thus raising the conversion of nitrobenzene to mono- and also to di-chlorinated products. To summarize, Lewis acidity did not play a direct role in the chlorination reaction but the creation of mesoporosity in the
H-USY (by the removal of EFAl fragments) enhanced the catalytic performance. 3.5. Optimization of the reaction conditions To complete our study, we have further optimized the reaction conditions by changing the quantities of TCCA and H-USY zeolite (WHSV alteration). Fig. 8 shows the results obtained for different quantities of HUSY catalyst. We realized these experiments in comparison to the reference, where the ratio between TCCA and H-USY was 1/3 (based on the number of Brønsted acid sites). It is important to note that the value 0.15 mmol corresponds to nTCCA /nBrønsted acid sites = 1. Hence, this implies that 3 chlorine atoms could be potentially involved in the reaction (nCl /nH+ = 3). Based on the results depicted in Fig. 8, it can be concluded that the optimal amount of catalyst for reaching maximal conversion corresponds to a nTCCA /nBrønsted acid sites = 1
Please cite this article in press as: M. Boltz, et al., Appl. Catal. A: Gen. (2012), http://dx.doi.org/10.1016/j.apcata.2012.09.021
ARTICLE IN PRESS
G Model APCATA-13877; No. of Pages 8
M. Boltz et al. / Applied Catalysis A: General xxx (2012) xxx–xxx
6
Conversion Selectivity in monochlorinated products
100
[%]
75
50
25
0 120
140
160
180
200
Temperature, [°C]
Fig. 7. Comparison in terms of nitrobenzene conversion and selectivity toward monochlorination between HUSY zeolite and HUSY treated by EDTA. Conditions: nTCCA = 0.15 mmol; nH+ = 0.44 mmol; reaction duration = 5 h; temperature = 150 ◦ C.
Fig. 5. Influence of reaction temperature on the chlorination process in terms of nitrobenzene conversion and selectivity in monochlorinated products. Conditions: nTCCA = 0.15 mmol; nH+ = 0.44 mmol; reaction duration = 5 h.
100
75
[%]
ratio (0.15 mmol H+ ). This ratio allowed the conversion of 24% of nitrobenzene with an excellent selectivity toward monochlorinated compounds of 95%. In parallel, the quantity of TCCA was also varied (Fig. 9). It is noteworthy that a raise in the TCCA amount led to an increase in the conversion of nitrobenzene while keeping a high selectivity toward monochlorinated products. These experiments confirmed that the optimal ratio between TCCA and the number of Brønsted acid sites of the zeolite was 1 (0.45 mmol TCCA, referring to nCl /nH+ = 3). These new conditions allowed converting 53% of nitrobenzene with a high selectivity of 94% toward monochlorination products. Scheme 2 rationalizes the Brønsted acid-catalyzed activation of one TCCA molecule within the zeolite pores and cavities, thus
Conversion Selectivity in monochlorinated products
50
25
0 0,0
0,1
0,2
0,3
0,4
0,5
Number of Bronsted acid sites, [mmol] Fig. 8. Influence of the HUSY catalyst loading on the nitrobenzene conversion. It is worthy to mention that 0.15 mmol H+ corresponds to nTCCA /nBrønsted acid sites = 1 (or nCl /nH+ = 3). Conditions: nTCCA = 0.15 mmol; reaction duration = 5 h; temperature = 150 ◦ C.
Conversion Selectivity in monochlorinated products
100
[%]
75
50
25
0 0,1
0,2
0,3
0,4
0,5
Quantity of TCCA, [mmol]
Fig. 6. 27 Al MAS NMR spectra: (a) HUSY zeolite before EDTA treatment; (b) HUSY zeolite after EDTA treatment; (c) HY zeolite.
Fig. 9. Effect of TCCA quantity on the chlorination of nitrobenzene over HUSY zeolite. The equimolar ratio between the number of Brønsted acid sites and TCCA (nCl /nH+ = 3) corresponds to 0.45 mmol. Conditions: nH+ = 0.44 mmol; reaction duration = 5 h; temperature = 150 ◦ C.
Please cite this article in press as: M. Boltz, et al., Appl. Catal. A: Gen. (2012), http://dx.doi.org/10.1016/j.apcata.2012.09.021
G Model APCATA-13877; No. of Pages 8
ARTICLE IN PRESS M. Boltz et al. / Applied Catalysis A: General xxx (2012) xxx–xxx
7
Scheme 2. Brønsted acid-catalyzed activation of TCCA within the zeolite framework.
raising the electrophilicity of Cl-atom. Hence, the electrophilic aromatic substitution remains therefore favored on the aromatic ring of nitrobenzene. To complete our study in a green chemistry context, we have also evaluated the recyclability of our “optimal” H-USY catalyst by performing five successive runs without changing the catalyst bed. The catalyst was treated at 550 ◦ C under air during one night after Conversion Selectivity in monochlorinated products
100
4. Conclusion
75
[%]
reaction to remove all organics from its surface. After calcination, new TCCA (100 mg) was added and mixed with the catalyst. Fig. 10 shows that after the first run the catalyst lost about 20–25% in terms of relative activity but remained stable during the four consecutive runs. Furthermore, it is worthy to mention that the catalyst held a constant and high selectivity toward monochlorinated compounds. Further studies are currently under progress to evaluate numerous kinetic parameters as the different reaction orders relative to the reactants, the activation energy and the detailed mechanism of this catalytic chlorination reaction.
50
25
0
run 1
run 2
run 3
run 4
run 5
Fig. 10. Recyclability of HUSY catalyst during five consecutive runs. Conditions: nTCCA = 0.15 mmol; nH+ = 0.44 mmol; reaction duration = 5 h; temperature = 150 ◦ C.
The present work describes the development of a green route to perform the chlorination of nitrobenzene (model deactivated aromatic) in the presence of zeolites and trichloroisocyanuric acid. We were able to setup a continuous process for the chlorination reaction, allowing the production of 1.7 g of chlorination products of nitrobenzene per gcatalyst per day, along with a high selectivity in monochlorinated products. A first generation of FAU zeolite catalyst was designed, combining the appropriate channel/cavity structure, with some mesoporosity, and a strong Brønsted acidity. Indeed, this H-USY zeolite exempted from its EFAl species led to achieve a 64% conversion at a 85% selectivity toward monochlorinated products. Finally, the potential reuse of the catalyst has been demonstrated but with an irreversible loss in activity after the first run (only), which remained stable after the next consecutive runs.
Please cite this article in press as: M. Boltz, et al., Appl. Catal. A: Gen. (2012), http://dx.doi.org/10.1016/j.apcata.2012.09.021
G Model APCATA-13877; No. of Pages 8
ARTICLE IN PRESS M. Boltz et al. / Applied Catalysis A: General xxx (2012) xxx–xxx
8
Acknowledgements The authors are grateful to the Agence Nationale de la Recherche (ANR) for supporting the ANR-10-JCJC-0703 project (SelfAsZeo). The technical assistance of M. Gaëtan Lutzweiler for electronic maintenance was highly appreciated. PME and MCSM acknowledge CNPq for a fellowship. References [1] H.-G. Franck, J.W. Stadelhofer, Industrial Aromatic Chemistry, Springer-Verlag, Berlin, Heidelberg, 1988, pp. 224–226. [2] J.I. Kroschwitz, M. Howe-Grant, C.A. Treacy, L.J. Humphreys, Encyclopedia of Chemical Technology, vol. 6, Wiley, New York, 1993, pp. 109–113. [3] H.-G. Franck, J.W. Stadelhofer, Industrial Aromatic Chemistry, Springer-Verlag, Berlin, Heidelberg, 1988, pp. 218–220. [4] K. Smith, M. Butters, W.E. Paget, D. Goubet, E. Fromentin, B. Nay, Green Chem. 1 (1999) 83–90. [5] G.K.S. Prakash, T. Mathew, D. Hoole, P.M. Esteves, Q. Wang, G. Rasul, G.A. Olah, J. Am. Chem. Soc. 126 (2004) 15770–15776. [6] G.F. Mendonca, R.R. Magalhaes, M.C.S. de Mattos, P.M. Esteves, J. Braz. Chem. Soc. 16 (2005) 695–698. [7] (a) G.F. Mendonca, M.R. Senra, P.M. Esteves, M.C.S. de Mattos, Appl. Catal. A 401 (2011) 176–181; (b) G.F. Mendonc¸a, M.C.S. de Mattos, Quim. Nova 31 (2008) 798–801. [8] (a) S.D.F. Tozetti, L.S. de Almeida, P.M. Esteves, M.C.S. de Mattos, J. Braz. Chem. Soc. 18 (2007) 675–677; (b) G.F. Mendonc¸a, A.M. Sanseverino, M.C.S. de Mattos, Synthesis (2003) 45–48.
[9] G.A. Hiegel, K.B. Peyton, Synth. Commun. 15 (1985) 385–392. [10] L. De Luca, G. Giacomelli, Synlett (2004) 2180–2184. [11] (a) L. De Luca, G. Giacomelli, G. Nieddu, Synlett (2005) 223–226; (b) A. Shiri, A. Khoramabadi-Zad, Synthesis (2009) 2797–2801. [12] U. Tilstam, H. Weinmann, Org. Process Res. Dev. 6 (2002) 384–393. [13] T.F. Degnan Jr., Top. Catal. 13 (2000) 349–356. [14] J. Weitkamp, Solid State Ionics 131 (2000) 175–188. [15] N.Y. Chen, W.E. Garwood, F.G. Dwyer, Shape Selective Catalysis in Industrial Applications, Marcel Dekker, New York, Basel, 1989, pp. 203–204. [16] B. Louis, G. Laugel, P. Pale, M. Maciel Pereira, ChemCatChem 3 (2011) 1263–1272. [17] H.E. Fierz-David, L. Blangey, Fundamental Processes of Dye Chemistry, Interscience Publishers, New York, 1949, pp. 111–115. [18] B. Xu, S. Bordiga, R. Prins, J.A. van Bokhoven, Appl. Catal. A 333 (2007) 245–253. [19] (a) B. Louis, S. Walspurger, J. Sommer, Catal. Lett. 93 (2004) 81–84; (b) S. Walspurger, B. Louis, Appl. Catal. A 336 (2008) 109–115. [20] B. Louis, A. Vicente, C. Fernandez, V. Valtchev, J. Phys. Chem. C 115 (2011) 18603–18610. [21] P. Kuhn, P. Pale, J. Sommer, B. Louis, J. Phys. Chem. C 113 (2009) 2903–2910. [22] Z. El Berrichi, B. Louis, J.P. Tessonnier, O. Ersen, L. Cherif, M.J. Ledoux, C. PhamHuu, Appl. Catal. A 316 (2007) 219–225. [23] S. Altwasser, J. Jiao, S. Steuernagel, J. Weitkamp, M. Hunger, Stud. Surf. Sci. Catal. 154B (2004) 1212–1215. [24] C.R. Moreira, M.H. Herbst, P.R. de la Piscina, J.-L.G. Fierro, N. Homs, M.M. Pereira, Micropor. Mesopor. Mater. 115 (2008) 253–260. [25] C. Mirodatos, D. Barthomeuf, J. Chem. Soc. Chem. Commun. (1981) 39–40. [26] F. Ocampo, H.S. Yun, M.M. Pereira, J.P. Tessonnier, B. Louis, Cryst. Growth Des. 9 (2009) 3721–3729. [27] C.H. Christensen, K. Johannsen, E. Toernqvist, I. Schmidt, H. Topsoe, C.H. Christensen, Catal. Today 128 (2007) 117–122.
Please cite this article in press as: M. Boltz, et al., Appl. Catal. A: Gen. (2012), http://dx.doi.org/10.1016/j.apcata.2012.09.021
ARTICLE IN PRESS Applied Catalysis A: General xxx (2012) xxx–xxx
Contents lists available at SciVerse ScienceDirect
Applied Catalysis A: General journal homepage: www.elsevier.com/locate/apcata
Green route for the chlorination of nitrobenzene Marilyne Boltz a , Márcio C.S. de Mattos b , Pierre M. Esteves b,∗ , Patrick Pale a , Benoit Louis a,∗∗ a Laboratoire de Synthèse Réactivité Organique et Catalyse (LASYROC), UMR 7177, Institut de Chimie, Université de Strasbourg, 1 rue Blaise Pascal, 67000 Strasbourg Cedex, France b Universidade Federal do Rio de Janeiro, Instituto de Química, Av. Athos da Silveira Ramos 149, CT Bloco A, Cidade Universitária, 21941-909 Rio de Janeiro, Brazil
a r t i c l e
i n f o
Article history: Received 12 July 2012 Received in revised form 21 August 2012 Accepted 17 September 2012 Available online xxx Keywords: Chlorination Nitrobenzene Trichloroisocyanuric acid (TCCA) Zeolite Green chemistry
a b s t r a c t A new green chlorination process of deactivated aromatics has been developed, being environmentalfriendly and allowing the continuous chlorination of 1.7 kg nitrobenzene/kg catalyst per day. The triple novelty consists of using a non-conventional chlorination agent, the trichloroisocyanuric acid (TCCA, C3 N3 O3 Cl3 ), along with solid acid catalysts (mainly zeolites) in a continuous flow reactor system. Different zeolites and solid acids have been tested in the chlorination of nitrobenzene, chosen as a model deactivated aromatic substrate. HUSY zeolite was found as the more promising catalyst for performing the chlorination of nitrobenzene, with good conversions (39–64%) at high selectivity toward monochlorinated products (90–99%). Finally, it is worthy to note that HUSY zeolite could be reused for at least five successive runs. © 2012 Elsevier B.V. All rights reserved.
1. Introduction Chloroarenes are valuable starting molecules in fine chemistry, for the synthesis of dyes, bio-active compounds such as pesticides or pharmaceuticals [1,2]. Unfortunately, the conventional industrial methods used for the chlorination of aromatics usually produce mixtures of regioisomers, difficult to separate, thus raising the cost for the industry [3]. Indeed, the utilization of Cl2 , a toxic gas, as chlorine source produces a large quantity of non-recyclable wastes. Another disadvantage is the use of toxic acid catalysts such as Lewis acids (aluminum chloride or boron trifluoride), which are consumed and need to be neutralized after the reaction. In parallel, strong Brønsted acid catalysts such as H2 SO4 remain very corrosive and generate a lot of salts as co-products. During the past decades, researchers focused on the development of more efficient and selective processes for the chlorination of arenes [4–7]. For instance, Esteves et al. [7a] developed a methodology for the chlorination of deactivated arenes using trichloroisocyanuric acid (TCCA, C3 N3 O3 Cl3 ) in a superelectrophilic medium.
∗ Corresponding author. Tel.: +55 2125627444. ∗∗ Corresponding author. Tel.: +33 368851344. E-mail addresses: [email protected] (P.M. Esteves), [email protected] (B. Louis).
TCCA is a stable and inexpensive solid, easily available in pool supplies, being thus frequently used as swimming-pool disinfectant and bleaching agent. It is an efficient chlorine source due to its high chlorine content, which can be up to 45.5% in weight, allowing a priori a higher atomic efficiency than its N-chloro analogs [7b]. TCCA was successfully used for chlorination of electron-rich arenes [7b], alkenes [8] and carbonyl compounds [9], preparation of N-chloro substrates [10,11] and in diverse oxidation reactions [12]. Nowadays, organic chemistry is increasingly moving toward green chemistry, where both homogeneous and heterogeneous catalysis are at the core of this concept. In heterogeneously catalyzed reactions, zeolites represent the key solid acid materials in petrochemistry [13,14]. In addition, thanks to their internal porosity, zeolites provide a highly organized channel structure and strong acidity. The combination of these properties led them to exhibit high conversions and excellent selectivities in targeted reactions [15,16]. The chlorination of nitrobenzene is a very important process for the production of series of useful compounds in dye chemistry [17]. It is worthy to study the chlorination of nitrobenzene in order to produce valuable intermediates for the synthesis of azo dyes. The aim of the present study is therefore to develop a new chlorination process of deactivated aromatics, being more environmental-friendly by the use of solid acids instead of liquid acids. In order to achieve this goal, we set up both a catalytic and a
0926-860X/$ – see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.apcata.2012.09.021
Please cite this article in press as: M. Boltz, et al., Appl. Catal. A: Gen. (2012), http://dx.doi.org/10.1016/j.apcata.2012.09.021
G Model APCATA-13877; No. of Pages 8
ARTICLE IN PRESS M. Boltz et al. / Applied Catalysis A: General xxx (2012) xxx–xxx
2
Scheme 1. Chloration reaction of deactivated arenes over solid acid catalysts.
continuous process to chlorinate nitrobenzene (chosen as a model deactivated arene) while using TCCA as a chlorination agent and different solid acids as catalysts (Scheme 1). 2. Experimental 2.1. Reagents and catalysts Trichloroisocyanuric acid (98%, Aldrich) and nitrobenzene (99%, Merck) were used as received. Several commercial zeolites were employed: H-Y (Aldrich), H-BETA (Zeolyst), H-MOR (Zeolyst), HEMT (IS2M, Mulhouse), H-USY (Zeolyst), H-FER (Petrobras). Prior to use, these zeolites were activated by calcination at 550 ◦ C overnight under air. In addition, H-USY zeolite has been chemically modified to investigate the influence of its Lewis acidity. H-USY zeolite (200 mg) was suspended in a 0.1 M solution of Na2 H2 -EDTA (17.4 mL, 2 equiv.). The mixture was heated at 62 ◦ C and stirred for 24 h. Afterwards, the solution was cooled down to room temperature, filtered over a nylon membrane and thoroughly washed with distilled water (60 mL). The solid was reintroduced in the flask and a 1 M solution of NH4 Cl (20 mL) was added. The solution was then heated at 80 ◦ C and stirred for 24 h. Finally, the solution was filtered, washed with distilled water (60 mL) and dried at 105 ◦ C in an oven. According to van Bokhoven et al. [18], extra-framework aluminum (EFAl) was completely removed from pristine FAU material. 2.2. Brønsted acid sites titration Our home-developed H/D exchange isotope technique was used to quantify the number of Brønsted acid sites present in solid acids [19,20]. Fig. 1 presents the experimental set-up used to perform this
Fig. 2. Experimental setup to perform chlorination in a continuous flow system.
titration method. The catalyst (300 mg) was first activated under dry nitrogen flow (40 mL/min) at 450 ◦ C for 1 h to desorb water. The deuteration of the catalyst was carried out at 200 ◦ C by flowing, through the catalyst, nitrogen (40 mL/min), previously bubbled at room temperature through a U-shaped tube containing D2 O (about 0.05 g), during 1 h. Dry nitrogen was then sweeping through the sample during 60 min to remove the excess of D2 O. After this purge, the titration of O-D sites was performed by back-exchanging the deuterium present on the solid surface with distilled water (3% of H2 O in N2 stream) at 200 ◦ C for 1 h. During this step the partially exchanged water named Hx ODy , composed by H2 O, D2 O and HDO, was collected in a cold U-tube trap at −117 ◦ C. Collected Hx ODy was then weighted and allowed to react with trifluoroacetic anhydride, used in a two-fold excess. The acid solution obtained was then transferred to a NMR tube for analysis. Spectra were recorded on a Bruker AM400 spectrometer, after addition of a CDCl3 (10 wt%)/CHCl3 mixture used as reference. Finally, an accurate quantification of the H/D content in the different samples was achieved by the integration of CF3 COOH(D) and CH(D)Cl3 on both 1 H and 2 H spectra. 2.3. Continuous chlorination process
Fig. 1. Experimental setup to perform the H/D isotope exchange technique.
All the reactions were performed in a glass flow system with a cylindrical reactor. The gas flow was regulated by means of Brooks 5850E mass flow controllers and the dry nitrogen flow was set to 100 mL/min for each experiments. Fig. 2 shows the experimental set-up to perform these catalytic tests in a continuous manner. The reactions were carried out by diluting the catalytic bed (TCCA and catalyst) in an amorphous silica (Grace, USA) matrix to insure the same height for all catalyst beds. To investigate the influence of the solid acid structure on its catalytic performance, we have decided to work at iso-site conditions. Indeed, all the catalytic tests were performed by keeping a constant number of Brønsted acid sites for the catalysts, which was set to 0.44 mmol H+ . Table 1
Please cite this article in press as: M. Boltz, et al., Appl. Catal. A: Gen. (2012), http://dx.doi.org/10.1016/j.apcata.2012.09.021
G Model APCATA-13877; No. of Pages 8
ARTICLE IN PRESS M. Boltz et al. / Applied Catalysis A: General xxx (2012) xxx–xxx
Table 1 Brønsted acidity of the solid acids used in the chlorination process. Solid acids
Si/Al
mmol H+ /g
H-MOR H-FER H-EMT H-BETA H-ZSM-5 (zeolyst) H-Y (Aldrich) H-Y (UOP) H-USY Silica–alumina H3 PW12 O40 Cs2 PW12 O40 SAC-13 Ga-SBA-15
5 10.5 6.5 13.6 15 2.2 5.0 2.8 3.6 – – – –
1.90 2.26 2.38 1.07 1.04 5.34 2.20 3.90 3.68 0.91 0.32 1.00 1.87
presents the number of Brønsted acid sites titrated by H/D exchange for all solid acids used for the chlorination reaction. The solid acid catalyst (0.44 mmol H+ ), trichloroisocyanuric acid (0.15 mmol) and silica matrix (17 mmol, 1 g) were blended closely by grinding. The mixture was then transferred into the cylindrical reactor and the reactor was fixed to the set-up. The catalytic bed was first dried under dry N2 flow at 150 ◦ C for 30 min to desorb the water present in the void volume of the zeolite. Then, nitrobenzene was supplied in its gaseous state by sweeping a dry N2 flow through a stripping U-shaped reactor containing liquid nitrobenzene (Fig. 2). Hence, this dry nitrogen flow saturated with nitrobenzene’s vapor pressure was allowed to pass through the catalytic bed during 5 h. The products were trapped at −196 ◦ C and recovered downstream to the reactor with toluene (4 mL). They were analyzed by gas chromatography (HP 5890 Series II) equipped with a capillary column (PONA, 50 m). Retention times were compared with standards and used to characterize the different reaction products. The degree of conversion and the selectivity toward the different products were calculated by taking into account the response factor of the substrate (nitrobenzene) and those from the products (mono-, di- and tri-chlorinated aromatics) through the use of an external standard (styrene). 3. Results and discussion 3.1. Methodology In preliminary experiments, we initially performed the reaction between TCCA and nitrobenzene by changing only one parameter between each experiment. Firstly, we performed a screening of the different solid acid catalysts (working under iso-Brønsted site conditions). The effects of pore topologies and sizes along with the quantity of Brønsted acid sites were investigated among the different solid acids. Furthermore, we also decided to refine the number of catalysts tested to find adapted reaction conditions by analyzing the effect of temperature on the performance of the more promising catalyst. In parallel, the influence of Lewis acidity (EFAl) has been evaluated for chosen “optimal” catalyst. Finally, the chlorination process was further optimized by modifying different reaction parameters, as quantities of TCCA, the weight hourly space velocity (WHSV) while varying the catalyst mass. Likewise, the potential reuse of the catalyst was also tested. 3.2. Behavior of the different solid acid catalysts Fig. 3 shows the degree of nitrobenzene conversion and the selectivity to mono-chloronitrobenzenes among the different catalysts. It is noteworthy that the meta-chloronitrobenzene was always found as prominent isomer, i.e.; >90% with respect to the
3
ortho- and para-chloronitrobenzene, as a priori expected for the nitro electron attracting group. It seems that the zeolite pore size acts as a limiting factor for this reaction. Whilst no reaction occurred over H-FER zeolite (∼4 A˚ pore aperture), nitrobenzene reacted with TCCA over medium pore H˚ ZSM-5 zeolite, which possesses a rather narrow pore size of 5.5 A, affording 21% conversion. In contrast, 39% of nitrobenzene was con˚ This tends to support an verted over H-USY (pore size of 7.4 A). effect of higher accessibility of those large organic molecules to the catalytic sites in H-USY zeolite [21]. On the other hand, the chlorination reaction appears to be positively influenced by the channel architecture of a zeolite framework. Indeed, the SAC-13 catalyst (Nafion-H supported over MCM-41), being a polymer-grafted on mesoporous silica, only led to a 5% nitrobenzene conversion (Fig. 4). Likewise, gallium-doped mesoporous silica (Ga-SBA-15) [22] also led to poor catalytic performance (conversion of 7%) with respect to zeolites. Following the same trend, the selectivity toward monochlorination is also governed by the zeolite porosity. In parallel, the beneficial effect of microporosity has been further evidenced with the use of strongly acidic polyoxometalate H3 PW12 O40 . Whereas, this non porous solid acid led to an appreciable degree of nitrobenzene conversion (14%), in contrast the selectivity drastically diminished with respect to zeolites in favor of dichlorination products. However, Cs2 HPW12 O40 that exhibits a higher SSA value (123 m2 /g) and porosity with respect to its protonic counterpart (5 m2 /g) led to 2.6 fold increase in activity. The same behavior was observed with amorphous silica–alumina, which converted only 5% of nitrobenzene probably due to the absence of zeolite’s highly organized pore architecture. In addition, a non-conventional acid catalyst, sulfated tin dioxide, was also tested. This SnO2 /SO4 2− catalyst, which possesses exclusively Lewis acid sites and almost no porosity, was not a good candidate at all, since no activity could be observed. Besides, a rather surprising difference in reactivity between H-Y and H-USY zeolites was observed (Fig. 3). H-USY zeolite is usually obtained by dealumination via a high temperature steaming of pristine H-Y zeolite [23]. The dealumination process consists of removing aluminum atoms from the zeolite framework and thus to create new extra-framework aluminum species (EFAl) present either as amorphous species (pseudo Al2 O3 ) [24] or aluminum species exhibiting exalted Lewis acidity (AlOOH, AlO+ species, etc.) [25]. This phenomenon is beneficial for the chlorination process since the conversion of nitrobenzene was drastically improved to 39% with H-USY compared to HY zeolite (16%). To summarize these preliminary tests, one can observe that the optimal catalyst for performing the chlorination of nitrobenzene remains the H-USY zeolite. Indeed, the faujasite structure allows catalyzing the chlorination of nitrobenzene by combining a high selectivity in monochlorinated compounds (99%) at an acceptable degree of nitrobenzene conversion. 3.3. Influence of reaction temperature Fig. 5 shows the results of the chlorination reaction at different reaction temperatures. As ‘a priori’ expected, a raise in the temperature led to higher nitrobenzene conversion, but at the expense of the selectivity in monochlorinated products. The decrease in selectivity is due to the formation of di- and tri-chlorinated products. For instance, the selectivity toward monochlorination was diminished to 63% at 200 ◦ C. Selectivities in di-chlorinated and trichlorinated products were 32% and 5%, respectively. Among the products of di-chlorination, 1,4-dichloro,2-nitrobenzene was formed as major isomer (71%) whilst 1,2-dichloro,4-nitrobenzene represented nearly 28% selectivity among the dichlorinated products. It is therefore worthy to mention that the second chlorination
Please cite this article in press as: M. Boltz, et al., Appl. Catal. A: Gen. (2012), http://dx.doi.org/10.1016/j.apcata.2012.09.021
G Model APCATA-13877; No. of Pages 8
ARTICLE IN PRESS M. Boltz et al. / Applied Catalysis A: General xxx (2012) xxx–xxx
4
Fig. 3. Nitrobenzene conversion and selectivity to monochlorination products over different zeolite catalysts. Conditions: nTCCA = 0.15 mmol; nH+ = 0.44 mmol; reaction duration = 5 h; temperature = 150 ◦ C.
(involving deactivated meta-chloronitrobenzene) is favored in ortho and para positions of the chlorine atom. The higher selectivity in 1,4-dichloro,2-nitrobenzene major isomer might be explain by lower steric hindrance than 1,2-dichloro,4-nitrobenzene isomer. Based on these results, we have arbitrary chosen the temperature of 150 ◦ C as an optimal temperature, since a good compromise between a high selectivity and a reasonable conversion of nitrobenzene was achieved. 3.4. Influence of the Lewis acidity In order to highlight the effect of Lewis acidity in H-USY zeolite, we have performed a treatment to remove extra-framework aluminum species (EFAl) by complexing them with EDTA. The catalyst was then analyzed by 27 Al MAS NMR to assess the disappearance of the characteristic signal attributed to octahedral
species (EFAl) at nearly 0 ppm (Fig. 6). Two distinct signals were detected over pristine HUSY zeolite (before EDTA treatment): one located at 50 ppm and one broad near 0 ppm (Fig. 6a). These signals correspond, respectively, to framework aluminum species and to extra-framework aluminum (EFAl) species. The later species represent about 29% of the total aluminum present within the zeolite. Fig. 6b, corresponding to the zeolite HUSY after one EDTA treatment, shows the quasi-disappearance of the EFAl signal with only 12% of the remaining aluminum present as EFAl in the zeolite channels. It appears therefore that one EDTA treatment alone did not remove all the EFAl species in the H-USY. As a control experiment, no EFAl could be detected in the H-Y spectrum (Fig. 6c), all the aluminum cations are present in the framework as Tatoms. Furthermore, this dealumination treatment had another impact on the structural properties of H-USY zeolite. The complexation of EFAl species led to intense leaching of those species, thus reducing
Please cite this article in press as: M. Boltz, et al., Appl. Catal. A: Gen. (2012), http://dx.doi.org/10.1016/j.apcata.2012.09.021
G Model APCATA-13877; No. of Pages 8
ARTICLE IN PRESS M. Boltz et al. / Applied Catalysis A: General xxx (2012) xxx–xxx
5
Fig. 4. Comparison in terms of nitrobenzene conversion and selectivity toward monochlorination between HUSY zeolite and non-zeolitic solid acids. Conditions: nTCCA = 0.15 mmol; nH+ = 0.44 mmol; reaction duration = 5 h; temperature = 150 ◦ C.
pore obstruction, thus introducing also a high level of mesoporosity inside the FAU structure. Hence the diffusion of products/reactants can be seriously improved [26,27]. Fig. 7 presents the catalytic data obtained with the H-USY zeolite treated and non-treated with EDTA. One can observe that the presence or not of EFAl species strongly influence the catalytic performance of H-USY zeolite. Indeed, the EDTA treatment of H-USY led to a significant increase in nitrobenzene conversion, up to 64% while the selectivity in monochlorinated products remained slightly affected. This phenomenon could be due to the higher accessibility of the reactants to the active sites within the pores, emptied from their alumina aggregates [24]. Hence, the molecules easily access to the active sites, thus raising the conversion of nitrobenzene to mono- and also to di-chlorinated products. To summarize, Lewis acidity did not play a direct role in the chlorination reaction but the creation of mesoporosity in the
H-USY (by the removal of EFAl fragments) enhanced the catalytic performance. 3.5. Optimization of the reaction conditions To complete our study, we have further optimized the reaction conditions by changing the quantities of TCCA and H-USY zeolite (WHSV alteration). Fig. 8 shows the results obtained for different quantities of HUSY catalyst. We realized these experiments in comparison to the reference, where the ratio between TCCA and H-USY was 1/3 (based on the number of Brønsted acid sites). It is important to note that the value 0.15 mmol corresponds to nTCCA /nBrønsted acid sites = 1. Hence, this implies that 3 chlorine atoms could be potentially involved in the reaction (nCl /nH+ = 3). Based on the results depicted in Fig. 8, it can be concluded that the optimal amount of catalyst for reaching maximal conversion corresponds to a nTCCA /nBrønsted acid sites = 1
Please cite this article in press as: M. Boltz, et al., Appl. Catal. A: Gen. (2012), http://dx.doi.org/10.1016/j.apcata.2012.09.021
ARTICLE IN PRESS
G Model APCATA-13877; No. of Pages 8
M. Boltz et al. / Applied Catalysis A: General xxx (2012) xxx–xxx
6
Conversion Selectivity in monochlorinated products
100
[%]
75
50
25
0 120
140
160
180
200
Temperature, [°C]
Fig. 7. Comparison in terms of nitrobenzene conversion and selectivity toward monochlorination between HUSY zeolite and HUSY treated by EDTA. Conditions: nTCCA = 0.15 mmol; nH+ = 0.44 mmol; reaction duration = 5 h; temperature = 150 ◦ C.
Fig. 5. Influence of reaction temperature on the chlorination process in terms of nitrobenzene conversion and selectivity in monochlorinated products. Conditions: nTCCA = 0.15 mmol; nH+ = 0.44 mmol; reaction duration = 5 h.
100
75
[%]
ratio (0.15 mmol H+ ). This ratio allowed the conversion of 24% of nitrobenzene with an excellent selectivity toward monochlorinated compounds of 95%. In parallel, the quantity of TCCA was also varied (Fig. 9). It is noteworthy that a raise in the TCCA amount led to an increase in the conversion of nitrobenzene while keeping a high selectivity toward monochlorinated products. These experiments confirmed that the optimal ratio between TCCA and the number of Brønsted acid sites of the zeolite was 1 (0.45 mmol TCCA, referring to nCl /nH+ = 3). These new conditions allowed converting 53% of nitrobenzene with a high selectivity of 94% toward monochlorination products. Scheme 2 rationalizes the Brønsted acid-catalyzed activation of one TCCA molecule within the zeolite pores and cavities, thus
Conversion Selectivity in monochlorinated products
50
25
0 0,0
0,1
0,2
0,3
0,4
0,5
Number of Bronsted acid sites, [mmol] Fig. 8. Influence of the HUSY catalyst loading on the nitrobenzene conversion. It is worthy to mention that 0.15 mmol H+ corresponds to nTCCA /nBrønsted acid sites = 1 (or nCl /nH+ = 3). Conditions: nTCCA = 0.15 mmol; reaction duration = 5 h; temperature = 150 ◦ C.
Conversion Selectivity in monochlorinated products
100
[%]
75
50
25
0 0,1
0,2
0,3
0,4
0,5
Quantity of TCCA, [mmol]
Fig. 6. 27 Al MAS NMR spectra: (a) HUSY zeolite before EDTA treatment; (b) HUSY zeolite after EDTA treatment; (c) HY zeolite.
Fig. 9. Effect of TCCA quantity on the chlorination of nitrobenzene over HUSY zeolite. The equimolar ratio between the number of Brønsted acid sites and TCCA (nCl /nH+ = 3) corresponds to 0.45 mmol. Conditions: nH+ = 0.44 mmol; reaction duration = 5 h; temperature = 150 ◦ C.
Please cite this article in press as: M. Boltz, et al., Appl. Catal. A: Gen. (2012), http://dx.doi.org/10.1016/j.apcata.2012.09.021
G Model APCATA-13877; No. of Pages 8
ARTICLE IN PRESS M. Boltz et al. / Applied Catalysis A: General xxx (2012) xxx–xxx
7
Scheme 2. Brønsted acid-catalyzed activation of TCCA within the zeolite framework.
raising the electrophilicity of Cl-atom. Hence, the electrophilic aromatic substitution remains therefore favored on the aromatic ring of nitrobenzene. To complete our study in a green chemistry context, we have also evaluated the recyclability of our “optimal” H-USY catalyst by performing five successive runs without changing the catalyst bed. The catalyst was treated at 550 ◦ C under air during one night after Conversion Selectivity in monochlorinated products
100
4. Conclusion
75
[%]
reaction to remove all organics from its surface. After calcination, new TCCA (100 mg) was added and mixed with the catalyst. Fig. 10 shows that after the first run the catalyst lost about 20–25% in terms of relative activity but remained stable during the four consecutive runs. Furthermore, it is worthy to mention that the catalyst held a constant and high selectivity toward monochlorinated compounds. Further studies are currently under progress to evaluate numerous kinetic parameters as the different reaction orders relative to the reactants, the activation energy and the detailed mechanism of this catalytic chlorination reaction.
50
25
0
run 1
run 2
run 3
run 4
run 5
Fig. 10. Recyclability of HUSY catalyst during five consecutive runs. Conditions: nTCCA = 0.15 mmol; nH+ = 0.44 mmol; reaction duration = 5 h; temperature = 150 ◦ C.
The present work describes the development of a green route to perform the chlorination of nitrobenzene (model deactivated aromatic) in the presence of zeolites and trichloroisocyanuric acid. We were able to setup a continuous process for the chlorination reaction, allowing the production of 1.7 g of chlorination products of nitrobenzene per gcatalyst per day, along with a high selectivity in monochlorinated products. A first generation of FAU zeolite catalyst was designed, combining the appropriate channel/cavity structure, with some mesoporosity, and a strong Brønsted acidity. Indeed, this H-USY zeolite exempted from its EFAl species led to achieve a 64% conversion at a 85% selectivity toward monochlorinated products. Finally, the potential reuse of the catalyst has been demonstrated but with an irreversible loss in activity after the first run (only), which remained stable after the next consecutive runs.
Please cite this article in press as: M. Boltz, et al., Appl. Catal. A: Gen. (2012), http://dx.doi.org/10.1016/j.apcata.2012.09.021
G Model APCATA-13877; No. of Pages 8
ARTICLE IN PRESS M. Boltz et al. / Applied Catalysis A: General xxx (2012) xxx–xxx
8
Acknowledgements The authors are grateful to the Agence Nationale de la Recherche (ANR) for supporting the ANR-10-JCJC-0703 project (SelfAsZeo). The technical assistance of M. Gaëtan Lutzweiler for electronic maintenance was highly appreciated. PME and MCSM acknowledge CNPq for a fellowship. References [1] H.-G. Franck, J.W. Stadelhofer, Industrial Aromatic Chemistry, Springer-Verlag, Berlin, Heidelberg, 1988, pp. 224–226. [2] J.I. Kroschwitz, M. Howe-Grant, C.A. Treacy, L.J. Humphreys, Encyclopedia of Chemical Technology, vol. 6, Wiley, New York, 1993, pp. 109–113. [3] H.-G. Franck, J.W. Stadelhofer, Industrial Aromatic Chemistry, Springer-Verlag, Berlin, Heidelberg, 1988, pp. 218–220. [4] K. Smith, M. Butters, W.E. Paget, D. Goubet, E. Fromentin, B. Nay, Green Chem. 1 (1999) 83–90. [5] G.K.S. Prakash, T. Mathew, D. Hoole, P.M. Esteves, Q. Wang, G. Rasul, G.A. Olah, J. Am. Chem. Soc. 126 (2004) 15770–15776. [6] G.F. Mendonca, R.R. Magalhaes, M.C.S. de Mattos, P.M. Esteves, J. Braz. Chem. Soc. 16 (2005) 695–698. [7] (a) G.F. Mendonca, M.R. Senra, P.M. Esteves, M.C.S. de Mattos, Appl. Catal. A 401 (2011) 176–181; (b) G.F. Mendonc¸a, M.C.S. de Mattos, Quim. Nova 31 (2008) 798–801. [8] (a) S.D.F. Tozetti, L.S. de Almeida, P.M. Esteves, M.C.S. de Mattos, J. Braz. Chem. Soc. 18 (2007) 675–677; (b) G.F. Mendonc¸a, A.M. Sanseverino, M.C.S. de Mattos, Synthesis (2003) 45–48.
[9] G.A. Hiegel, K.B. Peyton, Synth. Commun. 15 (1985) 385–392. [10] L. De Luca, G. Giacomelli, Synlett (2004) 2180–2184. [11] (a) L. De Luca, G. Giacomelli, G. Nieddu, Synlett (2005) 223–226; (b) A. Shiri, A. Khoramabadi-Zad, Synthesis (2009) 2797–2801. [12] U. Tilstam, H. Weinmann, Org. Process Res. Dev. 6 (2002) 384–393. [13] T.F. Degnan Jr., Top. Catal. 13 (2000) 349–356. [14] J. Weitkamp, Solid State Ionics 131 (2000) 175–188. [15] N.Y. Chen, W.E. Garwood, F.G. Dwyer, Shape Selective Catalysis in Industrial Applications, Marcel Dekker, New York, Basel, 1989, pp. 203–204. [16] B. Louis, G. Laugel, P. Pale, M. Maciel Pereira, ChemCatChem 3 (2011) 1263–1272. [17] H.E. Fierz-David, L. Blangey, Fundamental Processes of Dye Chemistry, Interscience Publishers, New York, 1949, pp. 111–115. [18] B. Xu, S. Bordiga, R. Prins, J.A. van Bokhoven, Appl. Catal. A 333 (2007) 245–253. [19] (a) B. Louis, S. Walspurger, J. Sommer, Catal. Lett. 93 (2004) 81–84; (b) S. Walspurger, B. Louis, Appl. Catal. A 336 (2008) 109–115. [20] B. Louis, A. Vicente, C. Fernandez, V. Valtchev, J. Phys. Chem. C 115 (2011) 18603–18610. [21] P. Kuhn, P. Pale, J. Sommer, B. Louis, J. Phys. Chem. C 113 (2009) 2903–2910. [22] Z. El Berrichi, B. Louis, J.P. Tessonnier, O. Ersen, L. Cherif, M.J. Ledoux, C. PhamHuu, Appl. Catal. A 316 (2007) 219–225. [23] S. Altwasser, J. Jiao, S. Steuernagel, J. Weitkamp, M. Hunger, Stud. Surf. Sci. Catal. 154B (2004) 1212–1215. [24] C.R. Moreira, M.H. Herbst, P.R. de la Piscina, J.-L.G. Fierro, N. Homs, M.M. Pereira, Micropor. Mesopor. Mater. 115 (2008) 253–260. [25] C. Mirodatos, D. Barthomeuf, J. Chem. Soc. Chem. Commun. (1981) 39–40. [26] F. Ocampo, H.S. Yun, M.M. Pereira, J.P. Tessonnier, B. Louis, Cryst. Growth Des. 9 (2009) 3721–3729. [27] C.H. Christensen, K. Johannsen, E. Toernqvist, I. Schmidt, H. Topsoe, C.H. Christensen, Catal. Today 128 (2007) 117–122.
Please cite this article in press as: M. Boltz, et al., Appl. Catal. A: Gen. (2012), http://dx.doi.org/10.1016/j.apcata.2012.09.021