Selective Reduction of Nitrite to Nitrogen with Carbon-Supported Pd

Sep 25, 2017 - Electrocatalytic Reduction of Nitrate to Nitrous Oxide by a Copper-Modified Covalent Triazine Framework. The Journal of Physical Chemis...
0 downloads 0 Views 2MB Size
Article Cite This: Ind. Eng. Chem. Res. 2017, 56, 11745-11754

pubs.acs.org/IECR

Selective Reduction of Nitrite to Nitrogen with Carbon-Supported Pd−AOT Nanoparticles A. M. Perez-Coronado,† L. Calvo,*,† J. A. Baeza,† J. Palomar,† L. Lefferts,‡ J. J. Rodriguez,† and M. A. Gilarranz† †

Sección Departamental de Ingeniería Química, Universidad Autónoma de Madrid, C/Francisco Tomás y Valiente 7, 28049 Madrid, Spain ‡ Catalytic Processes and Materials, MESA+ Institute for Nanotechnology, University of Twente, Enschede 7500AE, The Netherlands S Supporting Information *

ABSTRACT: The catalytic reduction of nitrite in water with hydrogen has been studied using a new strategy to control selectivity. The catalysts used are based on size-controlled Pd−AOT nanoparticles, synthesized via sodium bis[2-ethylhexyl] sulfosuccinate (AOT)/isooctane reverse microemulsion, supported on activated carbon. The most remarkable feature of the catalysts is the negligible selectivity toward ammonium, which is attributed to shielding of Pd NPs by AOT and blockage of the active centers related to ammonium generation. The shielding by AOT also reduces the activity of Pd−AOT NPs, which can be overcome by immobilization on carbon and thermal treatment at mild conditions (473 K, N2 atmosphere), although excessive removal of AOT results in higher ammonium production. Complete nitrite conversion with total selectivity to N2 was achieved at room temperature for the Pd−AOT/C catalysts at controlled pH media using CO2 as buffer agent. Moreover, the catalytic activity results at controlled pH show that the nitrite reduction reaction is not structure sensitive.

1. INTRODUCTION The catalytic reduction of nitrate with hydrogen has been suggested in the literature as a promising method, which was reported for the first time by Vorlop et al. in 1989.1,2 Nitrate reduction over metal catalysts proceeds principally in two reactions. The first step is the reduction of nitrate to nitrite, and the second one is the reduction of nitrite to nitrogen and/or ammonia, as shown in Scheme 1.

Hence, one of the major limitations in the catalytic reduction of nitrate is the production of harmful ammonium species, nitrite hydrogenation being the key step in that respect.4,5 In the nitrite reduction, selectivity depends on the concentrations of nitrite and H2 in the reaction medium, since a low nitrite/H2 ratio leads to higher selectivity to ammonium, according to the studies on adsorbed intermediates on the Pd surface in the literature.6,7 In semicontinuous reactions, where the H2 concentration remains over time while nitrite concentration decreases, there is a rise in selectivity to ammonium+ along with nitrite conversion and reaction time increase.6 Likewise, the activity and selectivity are influenced by the pH. It is important to note that the nitrite reduction consumes protons during the reaction, contributing to an increase in the pH value, which is unacceptable for drinking water. Likewise, an efficient pH

Scheme 1. Reaction Scheme in Nitrate and Nitrite Reduction According to Martı ́nez et al.3

Received: Revised: Accepted: Published: © 2017 American Chemical Society

11745

July 17, 2017 September 25, 2017 September 25, 2017 September 25, 2017 DOI: 10.1021/acs.iecr.7b02944 Ind. Eng. Chem. Res. 2017, 56, 11745−11754

Article

Industrial & Engineering Chemistry Research

were applied to partially remove the surfactant from the catalysts surface, in order to maximize the activity and minimize the selectivity to ammonium.

control is needed during the reaction, which also reduces the formation of ammonium, strongly favored under basic conditions.8,9 Thus, the use of CO2 for pH control is a common approach.10 The structure of the catalysts support also has an influence on nitrite reduction. D’Arino et al.11 investigated the combined effect of CO2 and the support porosity. They observed that with catalysts of low surface area and large pore size, OH− ions were neutralized in the proximity of the active phase thus hindering a rise in pH and reducing the selectivity to ammonium. Moreover, Strukul et al.12 found that the use of Pd/Al2O3 catalysts particles with a size over 0.5 mm favored the selectivity to ammonium. Different metals such as Pd, Pt, Ru, Rh, or Ir have been tested, Pd being the most active and selective to N2. A variety of supports has also been reported, including mainly alumina,13,14 silica,14,15 zeolites,16,17 pumice,15 and activated carbon (AC).18,19 AC has attracted attention due to its physical and chemical properties. The preparation of rationally designed catalysts of high performance is one of the classical challenges in catalysis.20,21 There are few studies about the mechanism of nitrite reduction. It has been well accepted that the turnover frequency (TOF) of the reaction is independent of the Pd particle size.22,23 However, some works24 indicated that there is a relationship between nanoparticles (NPs) size and the selectivity. Therefore, the methods to control NPs size have generated increasing attention for the elucidation of structure sensitiveness. One of these methods is the reverse microemulsion (water in oil, w/o) technique,25−29 which requires the use of water, a nonpolar solvent, a reducing agent, and a surfactant. The NPs size can be controlled using different water-to-surfactant (w0) and/or reducing agent-to-metal molar ratio.27 The surfactant agents used for preparation of NPs by microemulsion (ME) can be cationic, anionic, or nonionic.28,29 Among them, the anionic surfactant sodium bis-2-ethylhexyl sulfosuccinate (AOT) is one of the most used in the synthesis of metallic, metal sulfide, and metal oxide NPs in water/oil suspension. In general, NPs with high stability and good monodispersity can be obtained by the ME method.27 Nevertheless, the strong interaction between the AOT molecule and the NPs has been considered a drawback, because the surfactant remaining on the NPs surface after the synthesis reduces the available catalyst surface area and can affect the catalyst performance.30 Xiong et al.31 reported on the difficulty to remove AOT from NPs, even after thermal treatment at 673 K in air, mostly due to the presence of sodium and sulfur in the AOT molecule. In a previous work on the reduction of nitrite using unsupported Pd, NPs synthesized using an AOT-based w/o ME,5 we also observed the difficulty in removing AOT by purification with solvents. However, the shielding effect attributed to the interaction between Pd NPs and AOT was shown to be a very interesting tool to reduce the selectivity to ammonium, although at the expense of low activity. In the current work, size-controlled Pd NPs, synthesized by an AOT-based w/o ME method according to our previous work,5 were supported on AC with the aim of obtaining catalysts with more perspectives of application, without losing the concept of control of selectivity by the blockage of the surface of the Pd NPs with AOT. These catalysts are used as catalyst models in nitrite reduction. The effect of the surfactant remaining on the catalysts surface after the synthesis on catalytic performance and the interaction with the support are studied. Different thermal treatments and purification methods

2. EXPERIMENTAL SECTION 2.1. Material. Tetraamminepalladium(II) chloride monohydrate (Pd(NH3)4Cl2·H2O) (≥99%, Sigma-Aldrich Co.), isooctane (99.8%, Sigma-Aldrich Co.), and hydrazine hydrate solution (50−60%, Fluka) were used as Pd precursor, as oil medium to prepare water-in-oil ME, and as reducing agent, respectively. AOT, dioctyl sulfosuccinate sodium salt (98%, Sigma-Aldrich Co), was used as stabilizing agent and was vacuum-dried for 24 h at 333 K immediately before use. AC (SX PLUS, Norit) was used as catalysts support with a particle size less than 100 μm. Sodium nitrite (≥99%, Panreac) was used to prepare nitrite solutions. Demineralized bidistilled water was used throughout this work (Nihon Millipore Ltd.). 2.2. Preparation and Characterization of Pd NPs. The preparation of Pd NPs in ME was carried out through the reduction of Pd(NH3)4Cl2·H2O with hydrazine in AOT/ isooctane reverse micellar solution, as it was detailed in a previous work.28 The synthesis of Pd NPs was achieved by mixing equal volumes of two reverse micellar solutions prepared using different w0 values (3 and 7) and an AOT concentration in isooctane of 0.35 M. After 10 min of reduction, isooctane was evaporated in a rotary evaporator at 368 K and the NPs were purified from the excess surfactant by the addition of methanol or THF followed by centrifugation (this washing procedure was repeated three times). The Pd NPs synthesized were characterized by transmission electron microscopy (TEM) at 200 kV (JEOL, mod. JEM-2100). More than 100 measurements were taken for each sample from different TEM images to assess NP size. The Pd NPs synthesized via ME were supported on AC to prepare the catalysts. The AC was mixed with the Pd NPs ME in a rotary evaporator (Büchi) at 368 K until complete removal of methanol. The nominal content of Pd of the catalysts was varied between 0.5 and 2.5% (w, dry AC basis). Some of the catalysts prepared were subjected to thermal treatment during 2 h under N2 atmosphere (series N, temperature range 423−673 K) or H2 atmosphere (series NH, 473 K). A catalyst was prepared by incipient wetness impregnation (series IWI) and used as a blank representative of AOT-free Pd/AC catalysts. The IWI catalyst was prepared using an aqueous solution of Pd(NH3)4Cl2·H2O and a nominal Pd content of 1% (w, dry AC basis). The impregnation solution volume exceeded by 30% the pore volume of the support. Impregnation was followed by drying at room temperature for 2 h and overnight at 333 K. Finally, the catalyst was calcined in air atmosphere at 573 K and at the same temperature under 60 N mL·min−1 of H2 for 2 h. In some experiments the IWI catalyst was exposed to AOT to check if the interaction between Pd and AOT leading to selectivity control can be achieved by simple impregnation (poisoning) or if it is a characteristic of the ME synthesis method. The poisoning of the IWI catalyst was carried out by impregnation with a solution of AOT in ethanol at a AOT/Pd molar ratio of 640, and then the catalyst was washed with methanol. The porous structure of the AC supports and the catalysts was characterized by means of N2 adsorption−desorption at 77 K using a Micromeritics apparatus (Tristar II 3020 model). The samples were previously outgassed at 423 K and a residual pressure lower than 10−3 Pa. The BET and Dubinin− 11746

DOI: 10.1021/acs.iecr.7b02944 Ind. Eng. Chem. Res. 2017, 56, 11745−11754

Article

Industrial & Engineering Chemistry Research

which H2 was continuously fed at 50 N·mL·min−1 flow rate under vigorous stirring (500−700 rpm) in order to facilitate H2 distribution through the nitrite solution (150 mL and 50 mg NO2− L−1). The concentration of Pd in the reaction medium was 7.7 × 10−3 g·L−1. The reaction temperature, 303 K, was controlled by a thermostatic bath connected to the reactor jacket. In some experiments, CO2 was introduced to the reaction medium for buffering the solution at a pH ≈ 6. Different loads of catalyst and stirring velocities were also checked in preliminary experiments in order to confirm that the process takes place under kinetic control.5 No nitrite conversion was observed in the blank experiment performed using only AC as catalyst. Samples were withdrawn from the reactor at 0, 5, 15, 30 min and each hour until the end of the experiment (4 h). The catalyst was separated by filtration using 0.22 μm pore size PTFE filters. The samples were analyzed by ion chromatography (Metrohm 790 Compact IC Plus) using a Metrosep C4 column to analyze the ammonium concentration and a mixture of 1.7 mM HNO3 and 0.7 mM 2,6-pyridinedicarboxylic acid as stationary phase. A Metrosep A Supp 5 column and a mixture of 3.20 mM NaHCO3 and 1.00 mM Na2CO3 were used to analyze nitrite concentration. Routine duplicated analysis of nitrite− and ammoinum was carried out, showing always excellent reproducibility. The error associated with the ions determination by ion chromatography is 10 nm) is more important in the case of the thermal treatment in H2 atmosphere, despite the small differences in Pd NP size. The textural characterization of some selected Pd/C catalysts is summarized in Table 2. The AC used as support showed high SBET (980 m2/g) and micro- and mesopore volumes (0.309 and 0.344 cm3/g, respectively). In general, as the Pd load in the catalysts increased, the SBET and the micropore and mesopore

MeOH washing after AOT poisoning 11747

DOI: 10.1021/acs.iecr.7b02944 Ind. Eng. Chem. Res. 2017, 56, 11745−11754

Article

Industrial & Engineering Chemistry Research

Figure 1. Particle size distribution and TEM images of selected catalysts prepared with NPs synthesized at w0 = 7 and w0 = 3.

contributing to the decrease of the aforementioned SBET since that AOT is not degraded in N2 at 473 K, but AOT on NPs could be decomposed at lower temperature due to the catalytic effect of Pd (Figure S1). 3.2. Nitrite Reduction Tests. 3.2.1. Reaction at Uncontrolled pH. As a first approach, nitrite reduction with the 1% Pd/C catalysts synthesized via ME, IWI, and IWI followed by poisoning with AOT were tested at uncontrolled pH. The results are shown in Figure 2. Figure 2A depicts the evolution of nitrite conversion upon reaction time, and the selectivity to ammonium versus nitrite conversion can be seen in Figure 2B. All the catalysts tested at uncontrolled pH exhibited relatively low nitrite conversion (15−40%). The w3_1% catalyst showed lower activity than the IWI_1%, probably due to blockage of active centers by AOT. This would take place during the synthesis of the NPs and results in Pd−AOT interactions

Table 2. Textural Characteristics of Selected Catalysts SBET −1

As

Vmicropore

(m g )

(m g )

(cm g )

(cm3 g−1)

SX PLUS commercial w3_0.5%_N473 w3_0.5%_N473H w3_1%_N473 w3_2.5%_N473 w3_2.5%_N473H

977 884 860 621 567 633

329 308 299 260 196 242

0.309 0.273 0.266 0.169 0.175 0.184

0.344 0.327 0.320 0.277 0.212 0.257

2

−1

3

−1

Vmesopore

sample

2

volumes decreased, indicating partial pore blockage by the metal and the surfactant. A decrease in SBET was observed when the catalysts were treated at 473 K in N2 and H2, especially in the case of those with higher metal load. The changes can be interpreted in terms of a rearrangement of NPs during thermal treatment. Moreover, a partial removal of AOT can be also

Figure 2. Nitrite reduction experiments at uncontrolled pH for catalysts prepared using different procedures (7.7 mg Pd/L): (A) nitrite conversion vs time; (B) selectivity to ammonium vs nitrite conversion. 11748

DOI: 10.1021/acs.iecr.7b02944 Ind. Eng. Chem. Res. 2017, 56, 11745−11754

Article

Industrial & Engineering Chemistry Research

Figure 3. Nitrite reduction at uncontrolled pH with catalysts subjected to thermal treatment in N2 atmosphere (A1, B1) and thermal treatment in N2 atmosphere followed by reduction with H2 (A2, B2).

different from those taking place when a catalyst prepared by IWI is impregnated with AOT. This view is supported by the results obtained with catalyst IWI_1%AOT, where the poisoning with AOT did not produce a significant loss of activity and even showed a higher activity during the first minutes of reaction. On the other hand, the conversion rate observed for the supported NPs is higher than that previously reported for unsupported NPs.5 The NPs synthesized by ME can be expected to interact with the carbon support due to their hydrophobicity.35 However, when the support was oxidized with nitric acid to increase the content of oxygen and surface oxygen groups, thus turning it more polar, a lower activity of the catalyst was observed, which may indicate higher agglomeration of the NPs (Figure S2). The ammonium selectivity was also different for catalysts IWI_1% and IWI_1%AOT, indicating that at uncontrolled pH the generation of ammonium was higher for supported NPs that were poisoned by AOT during their synthesis. Therefore, shielding by AOT for the prevention of ammonium generation is conditioned by the pH of the reaction medium. Thus, the high pH values (8−9) characteristic of the uncontrolled pH runs causes a loss of the capability of w3_1% catalyst to control the selectivity. This behavior can be related to a lower interaction between Pd NPs and AOT at high pH due to the competition with hydroxyl ion. This competition would take

place preferentially at sites corresponding to low coordination atoms. The catalysts prepared by ME at w0 = 3 were treated at different temperatures between 423 and 673 K in N 2 atmosphere to remove AOT. These temperatures were selected according to the decomposition range of AOT observed by TGA in N2 atmosphere (530−580 K, see Supporting Information Figure S1). Some of the catalysts were also subjected to additional reduction with H2. The behavior of the catalysts after thermal treatment in N2 can be seen in Figure 3 panels A1 and B1. It can be observed that the treatment at mild temperature (423−523 K, Figure 4) increases the activity, with a peak at 423 K. At higher temperatures the decrease in activity can be attributed to deposition of AOT decomposition products on the NPs surface and sintering. Regarding ammonium selectivity, it increases substantially for the catalysts treated at the highest temperatures, where a higher removal of AOT can be expected. For these catalysts the generation of ammonium is favored even at low nitrite conversion. According to Scheme 1, this fact can be related to the hydrogenation and release of species adsorbed on the catalysts to produce ammonium. When the catalysts were subjected to reduction with H2 after thermal treatment in N2, a slight decrease of activity was observed (Figure 3 A2). This can be due to a higher sintering of 11749

DOI: 10.1021/acs.iecr.7b02944 Ind. Eng. Chem. Res. 2017, 56, 11745−11754

Article

Industrial & Engineering Chemistry Research

expected to reduce the number of uncoordinated atoms, both by reduction of electron-deficient atoms and also by rearrangement and sintering of the NPs. This may result in a lower availability of active centers involved in the generation of ammonium species.37,38. Regarding the selectivity (Figure 3 B2), the trend in the generation of ammonium is the same as that observed when catalysts subjected to thermal treatment were used. With the aim of exploring the effect of the Pd loading, catalysts with different nominal loads of Pd (0.5, 1, and 2.5% wt) were tested. Figure 5 summarizes nitrite conversion upon reaction time for a Pd concentration in the reaction medium of 7.7 mg L−1 (Figure 5A.1) and 19.25 mg L−1 (Figure 5A.2). At 7.7 mg Pd L−1, nitrite conversion range between 10 and 40 after 4 h while at 19.25 mg Pd L−1 conversion values within 20−80% were obtained. At the same Pd concentration in the reaction medium, the conversion of nitrite decreased at increasing Pd loading on the catalyst, which can be in part related to poorer dispersion of the metal phase and blockage of the porosity of the catalyst by agglomerations of NPs, as it can be inferred from the lower surface area, as shown in Table 2. Besides, the catalysts of higher nominal load showed a higher agglomeration of Pd NPs (see Supporting Information, Figure S3). In general, as the Pd nominal load in the catalysts increased, the selectivity to ammonium was higher (Figure 5 panels B.1 and B.2) especially at the higher concentration of Pd in the reaction medium (19.25 mg Pd L−1). This increase of ammonium generation with a metal nominal load has also been reported by other authors.39 3.2.2. Reaction at Controlled pH. Reaction tests at controlled pH (5−6) were carried out in order to study the catalysts behavior in a more favorable environment allowing higher activity and lower selectivity to ammonium. The evolution of nitrite conversion upon reaction time (Figure

Figure 4. Nitrite conversion at uncontrolled pH with catalysts subjected to thermal treatment in N2 atmosphere (reaction time = 4 h).

NPs. Interestingly, the selectivity to ammonium was also lower. The reduction of the catalyst can modify the ratio between Pdn+ and Pd0 species and also decrease the number of uncoordinated atoms on the NPs surface due to annealing and sintering. A decrease in the Pdn+/Pd0 ratio upon reduction with H2 is a wellknown phenomenon and can be assumed according to literature.36 It has to be taken into account that the NPs were reduced with hydrazine at room temperature, but the treatment with H2 took place at 473 K. Thus, the annealing resulting from the thermal treatment in the presence of H2 is

Figure 5. Nitrite reduction at uncontrolled pH with catalysts of different Pd nominal loading: (A1, B1), 7.7 mg Pd L−1; (A2, B2), 19.25 mg Pd L−1. 11750

DOI: 10.1021/acs.iecr.7b02944 Ind. Eng. Chem. Res. 2017, 56, 11745−11754

Article

Industrial & Engineering Chemistry Research

Figure 6. Nitrite reduction experiments at controlled pH for catalysts subjected to thermal treatment in N2 atmosphere (A1, B1) and thermal treatment in N2 atmosphere followed by reduction with H2 (A2, B2).

species give rise to N2, but NO can led to N2 or ammonium, in the last case through the formation of ammonium precursor of different stoichiometry: HxNyOz. Such species build up on the catalysts surface, and they are completely reduced to ammonium and released when nitrite conversion reaches ca. 100% values.22 The poisoning with AOT of the IWI catalysts do not show any effect in preventing ammonium formation, although it is delayed as well. Interestingly, ammonium selectivity is negligible for the catalysts prepared by ME, even at complete nitrite conversion. Thus, the high pH values (8−9) characteristic of the uncontrolled pH runs causes a loss of the capability of w3_1% catalyst to control the selectivity. This behavior can be related to a lower interaction between Pd NPs and AOT at high pH. In a previous work, a low generation of ammonium was reported for unsupported Pd NPs prepared by ME, which was attributed to the shielding effect of the AOT attached to the surface of the NPs.5 Therefore, thanks to the immobilization of the NPs on the carbon support the activity is enhanced while the shielding effect is maintained. The negligible production of ammonium at complete nitrite conversion denotes that building up of the nitrogen species evolving to ammonium on the surface of the NPs is prevented.22 This fact is of crucial significance since the main drawback of nitrate/nitrite reduction is the formation of ammonium. The catalysts reported in the current work could be coupled with others highly selective in the reduction of nitrate to nitrite, thus providing a useful catalytic system for the ammonium-free reduction of nitrate. It would be essential a study of the stability of the catalysts in continuous reactors in order to assess the applicability of the concept of the two-step process. The catalysts with Pd NPs prepared by ME and subjected to thermal treatment in N2 also yielded a higher reaction rate at

6A1) shows a significantly higher reaction rate for the catalysts prepared by IWI, yielding complete nitrite conversion in less than 1 h. The catalysts prepared by IWI and poisoned by impregnation with AOT showed lower activity but led to complete nitrite conversion within the reaction time tested. It can also be observed that the catalysts prepared by ME exhibit a higher activity in the case of the NPs synthesized at w0 = 3, that is, those of smaller size. In a previous work5 where unsupported NPs were prepared by ME synthesis and used as model catalysts, a faster conversion of nitrite was observed for the larger NPs (w0 = 7 and 12) which was explained by the apparent structure sensitiveness caused by AOT blockage of active centers. In the current work, a much faster nitrite conversion than for unsupported NPs was obtained, and the difference in activity between catalysts with NPs of different size is lower. In all the cases the surface of the NPs is covered by AOT, independently of the wo value used, that enables the control of the selectivity. Therefore, thanks to the immobilization of the NPs the negative effect of AOT on the activity is diminished. The control of agglomeration of NPs, which is likely to occur for unsupported ones in aqueous medium, can also play a positive role. Figure 6B1 shows a significant generation of ammonium with the catalyst prepared by IWI, delayed with respect to the reaction at uncontrolled pH, while selectivity to ammonium becomes significant only at almost complete nitrite conversion. This has been previously observed and explained in terms of the evolution of ammonium precursor species that build up on the catalyst surface once nitrite is almost completely converted.22 NO and N2O species are important intermediate species that are usually considered to occur as adsorb species on the surface of metal catalysts and only trace amounts of them would occur in the aqueous phase.3,40 N2O adsorbed 11751

DOI: 10.1021/acs.iecr.7b02944 Ind. Eng. Chem. Res. 2017, 56, 11745−11754

Article

Industrial & Engineering Chemistry Research

Figure 7. Nitrite reduction experiments at controlled pH for IWI catalyst and catalysts purified using different solvents.

tested at uncontrolled pH exhibited low nitrite conversion, especially in the case of the catalyst synthesized via ME, and high selectivity to ammonium. Pd−AOT NPs supported on AC showed complete nitrite conversion in the time range studied at controlled pH, although lower activity than for IWI catalysts was observed due to shielding by AOT. The thermal treatment of the catalysts increased activity due to modification and removal of AOT, and the catalytic activity was similar to that of different NPs size, evidencing nonstructure sensitiveness at controlled pH runs. Likewise, a higher reaction rate was observed for those catalysts based on Pd−AOT NPs supported AC purified with THF, showing the importance of metal−support interaction, and the role of solvents in facilitating this interaction. The catalysts based on Pd−AOT NPs showed a negligible production of ammonium when used at buffered pH. The production of ammonium was increased when the catalysts were subjected to thermal treatment or purification with THF, showing the role of AOT in the control of the selectivity through shielding of the metal surface and active centers. This remarkable feature opens the field for the development of highly selective catalysts for the reduction of nitrate and other reactions.

controlled than at uncontrolled pH (Figure 3A2 and Figure 6A2). The nitrite conversion rate in this case is almost equivalent to that observed for IWI-based catalysts, evidencing an increase in activity due to the thermal treatment, even though AOT cannot be removed from the catalyst by such treatment in N2 atmosphere at 423−523 K (Figure S1). Changes in the sulfonate group of AOT have been reported upon thermal treatment,41 which can provoke some rearrangement of AOT on the surface of the catalyst improving the availability of the active centers. On the other hands, the removal of AOT increased the selectivity to ammonium up to values equivalent to those observed for the IWI catalyst. Subsequent reduction of the Pd NPs with H2 led to a significant decrease of the catalytic activity (Figure 6A2), in contrast to the small differences that were observed in the uncontrolled pH experiments. Therefore, the role of the electro-deficient Pd species and the effect of low pH may counterbalance that of NPs sintering occurring upon reduction of the catalyst. To learn more on the role of the AOT on catalysts activity and selectivity, alternative purification of the NPs and catalysts by washing with THF was carried out. Figure 7 shows nitrite conversion upon reaction time at controlled pH (5−6) with those catalysts. The ones prepared with NPs washed with THF (w7 THF_1%) and by washing with THF after NPs impregnation (w7_1%THF) were more active than the equivalent catalyst washed with methanol (w7_1%). In particular, the highest reaction rate was obtained when the THF washing was performed after the impregnation of Pd NPs onto the support (w7_1%THF), suggesting that THF plays an important role on the metal−support interaction. In a previous work with unsupported Pd NPs5 we observed that NPs washed with THF were less active than those washed with MeOH, therefore lower removal of AOT can be expected from THF washing. Several authors have used THF to destabilize NP suspensions resulting from ME synthesis42,43 and to enhance the interaction with the support.44 Such interaction would reduce the shielding effect of AOT, therefore giving rise to a higher production of ammonia.



ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.iecr.7b02944. (Figure S1) TGA profiles of AOT, an AOT and isooctane mixture, and 5%Pd/C under N2 atmosphere; (Figure S2) (left) nitrite conversion versus time and (right) ammonium selectivity versus nitrite conversion for a catalysts supported on oxidizing carbon support (w7_1%CHNO3) and for a catalysts supported on carbon support not subjected to oxidation (w7_1%); (Figure S3) particle size distributions and TEM images of the nanoparticles samples obtained at different weight percent of active metals on support. The relation molar water-to-surfactant used is 3 (PDF)

4. CONCLUSIONS The catalytic reduction of nitrite in the aqueous phase was studied as an intermediate and critical step in the reduction of nitrate using size-controlled Pd−AOT NPs supported on AC. Supported Pd−AOT NPs showed higher activity than unsupported ones studied in a former work. All the catalysts



AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. 11752

DOI: 10.1021/acs.iecr.7b02944 Ind. Eng. Chem. Res. 2017, 56, 11745−11754

Article

Industrial & Engineering Chemistry Research ORCID

(17) Soares, O. S. G. P.; Marques, L.; Freitas, C. M. A. S.; Fonseca, A. M.; Parpot, P.; Ó rfão, J. J. M.; Pereira, M. F. R.; Neves, I. C. Mono and bimetallic NaY catalysts with high performance in nitrate reduction in water. Chem. Eng. J. 2015, 281, 411. (18) Soares, O. S. G. P.; Ó rfão, J. J. M.; Pereira, M. F. R. Activated carbon supported metal catalysts for nitrate and nitrite reduction in water. Catal. Lett. 2008, 126, 253. (19) Calvo, L.; Gilarranz, M. A.; Casas, J. A.; Mohedano, A. F.; Rodriguez, J. J. Denitrification of Water with Activated CarbonSupported Metallic Catalysts. Ind. Eng. Chem. Res. 2010, 49, 5603. (20) Yang, J.; Frøseth, V.; Chen, D.; Holmen, A. Particle size effect for cobalt Fischer−Tropsch catalysts based on in situ CO chemisorption. Surf. Sci. 2015, 648, 67. (21) Baeza, J. A.; Calvo, L.; Gilarranz, M. A.; Mohedano, A. F.; Casas, J. A.; Rodriguez, J. J. Catalytic behavior of size-controlled palladium nanoparticles in the hydrodechlorination of 4-chlorophenol in aqueous phase. J. Catal. 2012, 293, 85. (22) Zhao, Y.; Koteswara Rao, N.; Lefferts, L. Adsorbed species on Pd catalyst during nitrite hydrogenation approaching complete conversion. J. Catal. 2016, 337, 102. (23) Zhao, Y.; Baeza, J. A.; Koteswara Rao, N.; Calvo, L.; Gilarranz, M. A.; Li, Y. D.; Lefferts, L. Unsupported PVA- and PVP-stabilized Pd nanoparticles as catalyst for nitrite hydrogenation in aqueous phase. J. Catal. 2014, 318, 162. (24) Yoshinaga, Y.; Akita, T.; Mikami, I.; Okuhara, T. Hydrogenation of Nitrate in Water to Nitrogen over Pd−Cu Supported on Active Carbon. J. Catal. 2002, 207, 37. (25) Solanki, J. N.; Sengupta, R.; Murthy, Z. V. P. Synthesis of copper sulphide and copper nanoparticles with microemulsion method. Solid State Sci. 2010, 12, 1560. (26) Bedia, J.; Calvo, L.; Lemus, J.; Quintanilla, A.; Casas, J. A.; Mohedano, A. F.; Zazo, J. A.; Rodriguez, J. J.; Gilarranz, M. A. Colloidal and microemulsion synthesis of rhenium nanoparticles in aqueous medium. Colloids Surf., A 2015, 469, 202. (27) Perez-Coronado, A. M.; Calvo, L.; Alonso-Morales, N.; Heras, F.; Rodriguez, J. J.; Gilarranz, M. A. Multiple approaches to control and assess the size of Pd nanoparticles synthesized via water-in-oil microemulsion. Colloids Surfaces A Physicochem. Colloids Surf., A 2016, 497, 28. (28) Gustafsson, H.; Isaksson, S.; Altskär, A.; Holmberg, K. Mesoporous silica nanoparticles with controllable morphology prepared from oil-in-water emulsions. J. Colloid Interface Sci. 2016, 467, 253. (29) Kosydar, R.; Góral, M.; Gurgul, J.; Drelinkiewicz, A. The effect of support properties in the preparation of Pd size-controlled catalysts by “water-in-oil” microemulsion method. Catal. Commun. 2012, 22, 58. (30) Joo, S. H.; Cheon, J. Y.; Oh, J. Y. Current Trends of Surface Science and Catalysis; Springer: New York, 2014. (31) Xiong, L.; He, T. Synthesis and characterization of ultrafine tungsten and tungsten oxide nanoparticles by a reverse microemulsionmediated method. Chem. Mater. 2006, 18, 2211. (32) Calvo, L.; Gilarranz, M. A.; Casas, J. A.; Mohedano, A. F.; Rodriguez, J. J. Effects of Support Surface composition on the activity and selectivity of Pd/C catalysts in aqueous-phase hydrodechlorination reactions. Ind. Eng. Chem. Res. 2005, 44, 6661. (33) Calvo, L.; Gilarranz, M. A.; Casas, J. A.; Mohedano, A. F.; Rodriguez, J. J. Hydrodechlorination of 4-chlorophenol in aqueous phase using Pd/AC catalysts prepared with modified active carbon supports. Appl. Catal., B 2006, 67, 68. (34) Eriksson, S. Preparation of catalysts from microemulsions and their applications in heterogeneous catalysis. Appl. Catal., A 2004, 265, 207. (35) Lemus, J.; Bedia, J.; Calvo, L.; Simakova, I. L.; Murzin, D. Y.; Etzold, B. J. M.; Rodriguez, J. J.; Gilarranz, M. A. Improved synthesis and hydrothermal stability of Pt/C catalysts based on size-controlled nanoparticles. Catal. Sci. Technol. 2016, 6, 5196.

L. Calvo: 0000-0002-6329-0952 Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS The authors greatly appreciate financial support from the Spanish MICINN (CTQ2012-32821). The SusFuelCat project has received funding from the European Union’s Seventh Framework Programme for research technological development and demonstration under Grant Agreement No. 310490 (www. susfuelcat.eu).



REFERENCES

(1) Vorlop, K. D.; Tacke, T., 1st steps towards noble-metal catalyzed removal of nitrate and nitrite from drinking-water. Chem. Ing. Tech. 1989, 61, 836. (2) Soares, O.S.G.P; Ó rfão, J. J. M.; Ruiz-Martínez, J.; SilvestreAlbero, J.; Sepúlveda-Escribano, A.; Pereira, M. F. R. Pd-Cu/AC and Pt-Cu/AC catalysts for nitrate reduction with hydrogen: Influence of calcination and reduction temperatures. Chem. Eng. J. 2010, 165, 78. (3) Martínez, J.; Ortiz, A.; Ortiz, I. State-of-the-art and perspectives of the catalytic and electrocatalytic reduction of aqueous nitrates. Appl. Catal., B 2017, 207, 42. (4) Pintar, A.; Batista, J. Catalytic stepwise nitrate hydrogenation in batch-recycle fixed-bed reactors. J. Hazard. Mater. 2007, 149, 387. (5) Perez-Coronado, A. M.; Calvo, L.; Baeza, J. A.; Palomar, J.; Lefferts, L.; Rodriguez, J. J.; Gilarranz, M. A. Metal-surfactant interaction as a tool to control the catalytic selectivity of Pd catalysts. Appl. Catal., A 2017, 529, 32. (6) Pintar, A.; Berčič, G.; Levec, J. Catalytic liquid-phase nitrite reduction: kinetics and catalyst deactivation. AIChE J. 1998, 44, 2280. (7) Ebbesen, S. D.; Mojet, B. L.; Lefferts, L. In situ ATR-IR study of nitrite hydrogenation over Pd/Al2O3. J. Catal. 2008, 256, 15. (8) Soares, O. S. G. P.; Pereira, M. F. R.; Ó rfao, J. J. M.; Faria, J. L.; Silva, C. G. Photocatalytic nitrate reduction over Pd-Cu/TiO2. Chem. Eng. J. 2014, 251, 123. (9) Hörold, S.; Tacke, T.; Vorlop, K. Catalytical removal of nitrate and nitrite from drinking water: 1. Screening for hydrogenation catalysts and influence of reaction conditions on activity and selectivity. Environ. Technol. 1993, 14, 931. (10) Soares, O. S. G. P.; Jardim, E. O.; Reyes-Carmona, A.; RuizMartínez, J.; Silvestre-Albero, J.; Rodríguez-Castellón, E.; Orfão, J. J. M.; Sepúlveda-Escribano, A.; Pereira, M. F. R. Effect of support and pre-treatment conditions on Pt-Sn catalysts: application to nitrate reduction in water. J. Colloid Interface Sci. 2012, 369, 294. (11) D’Arino, M.; Pinna, F.; Strukul, G. Nitrate and nitrite hydrogenation with Pd and Pt/SnO2 catalysts: the effect of the support porosity and the role of carbon dioxide in the control of selectivity. Appl. Catal., B 2004, 53, 161. (12) Strukul, G.; Pinna, F.; Marella, M.; Meregalli, L.; Tomaselli, M. Sol-gel palladium catalysts for nitrate and nitrite removal for drinking water. Catal. Today 1996, 27, 209. (13) Palomares, A. E.; Franch, C.; Corma, A. Nitrates removal from polluted aquifers using (Sn or Cu)/Pd catalysts in a continuous reactor. Catal. Today 2010, 149, 348. (14) Garron, A.; Lázár, K.; Epron, F. Effect of the support on tin distribution in Pd−Sn/Al2 O 3 and Pd−Sn/SiO 2 catalysts for application in water denitration. Appl. Catal., B 2005, 59, 57. (15) Deganello, F.; Liotta, L. F.; Macaluso, A.; Venezia, A. M.; Deganello, G. Catalytic reduction of nitrates and nitrites in water solution on pumice-supported Pd-Cu catalysts. Appl. Catal., B 2000, 24, 265. (16) Nakamura, K.; Yoshida, Y.; Mikami, I.; Okuhara, T. Cu-Pd/bZeolites as Highly Selective Catalysts for the Hydrogenation of nitrate with hydrogen to harmless products. Chem. Lett. 2005, 34, 678. 11753

DOI: 10.1021/acs.iecr.7b02944 Ind. Eng. Chem. Res. 2017, 56, 11745−11754

Article

Industrial & Engineering Chemistry Research (36) Gómez-Sainero, L. M.; Seoane, X. L.; Fierro, J. L. G.; Arcoya, A. Liquid-phase hydrodechlorination of CCl4 to CHCl3 on Pd/carbon catalysts: nature and role of Pd active species. J. Catal. 2002, 209, 279. (37) Mélendrez, R.; Del Angel, G.; Bertin, V.; Valenzuela, M. A.; Barbier, J. Selective hydrogenation of carvone and o-xylene on Pd−Cu catalysts prepared by surface redox reaction. J. Mol. Catal. A: Chem. 2000, 157, 143. (38) Matatov-Meytal, U.; Sheintuch, M. The relation between surface composition of Pd−Cu/ACC catalysts prepared by selective deposition and their denitrification behavior. Catal. Commun. 2009, 10, 1137. (39) Jung, J.; Bae, S.; Lee, W. Nitrate reduction by maghemite supported Cu-Pd bimetallic catalyst. Appl. Catal., B 2012, 127, 148. (40) Shin, H.; Jung, S.; Bae, S.; Lee, W.; Kim, H. Nitrite reduction mechanism on a Pd surface. Environ. Sci. Technol. 2014, 48, 12768. (41) Kiciński, W.; Szala, M.; Bystrzejewski, M. Sulfur-doped porous carbons: Synthesis and applications. Carbon 2014, 68, 1. (42) Bonarowska, M.; Karpinski, Z.; Kosydar, R.; Szumelda, T.; Drelinkiewicz, A. Hydrodechlorination of CCl4 over carbon-supported palladium-gold catalysts prepared by the reverse “water-in-oil” microemulsion method. C. R. Chim. 2015, 18, 1143. (43) Trépanier, M.; Dalai, A. K.; Abatzoglou, N. Synthesis of CNTsupported cobalt nanoparticle catalysts using a microemulsion technique: Role of nanoparticle size on reducibility, activity and selectivity in Fischer−Tropsch reactions. Appl. Catal., A 2010, 374, 79. (44) Martínez, A.; Prieto, G. Breaking the dispersion-reducibility dependence in oxide-supported cobalt nanoparticles. J. Catal. 2007, 245, 470.

11754

DOI: 10.1021/acs.iecr.7b02944 Ind. Eng. Chem. Res. 2017, 56, 11745−11754