Introduction

Nanomaterials are of topical interest, because of their fascinating properties which are different from the corresponding bulk materials. Due to their unique properties, nanomaterials are employed in electronic, optical, catalytic, coating, medical and sensor applications [1–3]. A plethora of research work has been cited on the synthesis of Ni nanoparticles by various methods such as chemical reduction [4], radiolytic reduction [5], UV photolysis [6], solvent extraction reduction [7], microemulsion technique [8], polyol reduction method [9] and sonochemical decomposition [10]. Among these methods, chemical reduction appears to be the most promising method for the large-scale nanoparticle production, although simple methods for preparing metal-supported nanocatalysts under mild and inexpensive conditions have still not been definitely addressed. Among the chemical reduction reagents, utilization of NaBH4 leaves Na+ ions and B in the catalysts that affect the reaction while the utilization of Na2HPO4 leaves both Na+ ions and P in the catalyst which also influences the catalytic activity. In our present study, we used hydrazine as a reducing agent in the presence of the base at room temperature for 6 h which on decomposition produces H2 as well as N2. The liberated N2 is enough to act as inert atmospheric coverage to the reaction vessel.

Recently, supported nickel catalysts were reported for various organic transformations such as hydrogenation [11–13], steam reforming reactions [14], reductive amination of alcohols [15], hydrodechlorination [16, 17], partial oxidation [18] and dry reforming of methane [19]. Among, hydrogenation is of great interest to both industry and academia.

The synthesis of aniline is generally done by the nitrobenzene (NB) hydrogenation. Aniline is an important raw material for the synthesis of more than 300 different end products industrially, aniline is mainly used for the production of methylene diphenyl diisocyanate (MDI), a key compound in the production of polyurethane rigid and/or semi-rigid foams and elastomers. It is also used as a solvent, and has been used as an antiknock compound for gasoline [20, 21].

Aniline also finds application in various end-user markets including construction, rubber products, transportation, consumer, adhesives/sealants, packaging, agriculture, textiles, coatings, photography, dyes and pharmaceuticals. A predominant proportion of the worldwide aniline production is utilized in the manufacture of polyurethanes, which are used in a wide array of applications including footwear, insulation and furniture.

The global aniline market is projected to reach 6.2 million tons by the year 2015, due to the increasing demand from various end-user markets. In particular, the rising demand from MDI, the chief ingredient in polyurethane foam, is expected to fuel the consumption of aniline [22].

The present paper highlights the synthesis of Ni catalysts by reductive deposition method compared to impregnation method and as an efficient catalyst for the hydrogenation of NB at atmospheric pressure in the vapour phase.

Experimental

Catalyst preparation

Materials used in this research work are Ni(NO3)2·6H2O (M/s. LOBA Chemie, India, 98 %), hydrazine hydrate solution (M/s. LOBA Chemie, India, 80 %), sodium hydroxide (M/s. Sigma Aldrich, USA) and deionized water in the preparation of all the solutions. The nano-sized supported nickel particles were synthesized by dissolving requisite quantities of Ni(NO3)2·6H2O in deionized water and placing the respective supports in a vessel. The pH of the solution was then increased to 10 by adding a mixture of sodium hydroxide and hydrazine hydrate solution. The solution is stirred for 6 h. During the stirring, we have observed a change in the colour of the solution from green to violet to colourless and then to grey/black nickel nanoparticles with the evolution of N2 gas. The liberated N2 is enough to act as an inert atmospheric blanket. In order to avoid the expansion of volume of N2 during the stirring, precaution has been taken in selecting a flask of bigger volume. The resultant catalysts were designated as Ni/TiO2 (R) and Ni/ZrO2 (R), respectively. In comparison, we have also prepared the catalysts by impregnation method with similar Ni loading. They were designated as Ni/TiO2 (I) and Ni/ZrO2 (I), respectively. The Ni loading in all the catalysts was maintained at 5 wt%.

Catalyst characterization

All the above catalysts were characterized by XRD, BET method, H2 pulse decomposition, TPR and TEM. Details of all the above techniques and procedures have already been described in our previous publication [11–13]. The synthesized catalysts were digested in aqua regia and conducted AAS (M/s. Perkin-Elmer A-300) for Ni content.

Activity studies

Nitrobenzene hydrogenation was carried out in a continuous flow fixed bed reactor operated under atmospheric pressure. In each catalytic run, about 1 g of the catalyst (crushed to 18/25 BSS sieve to eliminate mass transfer limitations) diluted with an equal amount of silica beads to maintain the isothermal conditions, was positioned between two layers of quartz wool at the centre of the reactor comprising 14 mm internal diameter and 300 mm length. The upper portion of the reactor was filled with silica beads that served both as a preheater and a mixer for the reactants. Reaction temperature was continuously monitored by a thermocouple inserted in a thermowell which was placed within the catalyst bed. The liquid reactant was delivered to the reactor from a glass/Teflon air-tight syringe and Teflon tube using a microprocessor-controlled infusion pump (M/s. B. Braun Co, Germany) at a required flow rate. NB and ultra pure H2 were maintained at a WHSV ≈ 1 (h−1) with inlet NB flow = 1 ml h−1 and H2/NB molar ratio = 4.

In a blank test, passage of NB reactant along with a stream of H2 through an empty reactor or over the support alone did not result in any appreciable NB conversion. The product mixture was cooled in an ice-cooled trap and the analyses were performed on a gas chromatograph (GC-17A, M/s. Shimadzu Instruments Co, Japan) equipped with a programmed split injector and a flame ionization detector, with a OV-1 (30 m length 0.53 mm i.d., 0.33 mm film thickness) capillary column and the product components were confirmed by Shimadzu-17A gas chromatograph equipped along with the mass spectrum QP-8080.

Results and discussion

Hydrazine is normally used as a reducing agent. It can be oxidized depending upon the medium. In acid medium, it easily undergoes oxidation to NH3 while in case of the basic medium it easily undergoes oxidation to N2 and H2. Hence, it is necessary to use the basic medium to produce N2 and H2. For this purpose, we have chosen NaOH as base. The H2 produced is further utilized for the reduction of nickel precursors which are present in the solution.

In particular, the decomposition of hydrazine takes place in two pathways, which are given as follows:

  1. (i)

    Complete decomposition (basic medium)

    N2H4 → N2 (g) + 2H2 (g)

  2. (ii)

    Incomplete decomposition (acid medium)

    3 N2H4 → 4NH3 + 2N2 (g)

X-ray diffraction studies

The XRD patterns of all the catalyst samples are displayed in Fig. 1. In case of Ni/ZrO2 catalyst, it is reported that peaks at 2θ = 30.3° and 35.14° reveal the presence of indices (111) and (200) planes, respectively corresponding to tetragonal zirconia phase [23]. Major peak at 2θ = 28.3° reveals the part of the sample containing the monoclinic phase of zirconia [24]. It is interesting to note that the incorporation of di-positive or tetra-positive metal ions into the zirconia promotes the transition of tetragonal to monoclinic phase [25, 26]. Furthermore, the pH of the solution during the synthesis was maintained at 10. pH also influences the transformation of zirconia phase [24]. In case of Ni/TiO2 catalyst, the peaks at 2θ = 25.32°, 37.84°, 48.07°, 53.95°, 55.10°, 62.16°, 70.34° and 75.12° reveal the presence of indices (101), (004), (200), (105), (211) (Tetragonal TiO2, Space group 141/amd 141)) corresponding to titania (anatase) (ICDD. No. 84-1286). The peaks at around 2θ = 27.54°, 37.06° and 43.18° are corresponding to rutile phase of TiO2. The peaks at around 2θ = 22.87°, 34.09° and 49.51° are attributed to NiTiO3 species [27, 28]. Moreover, no peaks corresponding to Ni species were observed and one cannot rule out the formation of Ni species with smaller crystallite sizes (<4 nm), which is the detection limit of XRD technique. From this one can conclude that the Ni species in the catalysts are in highly dispersed state.

Fig. 1
figure 1

XRD patterns of Ni catalysts a Ni/ZrO2 (R), b Ni/ZrO2 (I), c Ni/TiO2 (R), d Ni/TiO2 (I). t Tetragonal, m Monoclinic, A Anatase, R Rutile and @ NiTiO3

Full size image

N2 physisorption studies

The surface area, pore volume and pore size of the catalysts are shown in Table 1. The N2 adsorption–desorption isotherms of all the catalyst samples are shown in Fig. 2. The surface areas of the catalysts prepared by the reductive deposition precipitation are smaller than that of impregnation method. The surface area of the catalysts prepared by the reductive deposition precipitation is half of the surface area of the catalysts prepared by impregnation method. Although the surface area of the catalysts prepared by reductive deposition method is smaller compared to that of the impregnation method, the catalysts show good yields towards the aniline. The domination of t-ZrO2 could partially contribute to the higher surface area of the Ni/ZrO2 catalyst prepared by impregnation method [29, 30]. Although the surface area of Ni/TiO2 (R) is lower compared to Ni/TiO2 (I), the higher catalytic performance was noticed which could be attributed to the presence of supreme number of surface Ni species (Table 2).

Table 1 Physicochemical properties of Ni catalysts prepared by impregnation and reductive deposition–precipitation method
Full size table
Fig. 2
figure 2

N2 adsorption–desorption isotherms of Ni catalysts a Ni/ZrO2(I), b Ni/ZrO2 (R), c Ni/TiO2(I), d Ni/TiO2(R)

Full size image
Table 2 H2 uptake, dispersion, metal area and particle size of Ni catalysts
Full size table

Temperature-programmed reduction studies

Typical profiles of TPR are presented in Fig. 3. For Ni/TiO2 (I), the reduction peaks at temperatures in the range of 360–480 °C are probably due to the reduction of bulk NiO and NiO with significant interaction with the TiO2 surface [31]. In the TPR patterns of Ni/ZrO2 (I), reduction peaks at a T max of about 443 and 503 °C are observed. The former peak can be assigned to the reduction of relatively bulk NiO species and the latter is assumed to be interacted species with ZrO2 [32]. Presence of strong chemical bonds between the Ni metal and the ZrO2 has been revealed based on a DFT calculation by Beltran [33]. The reduction peak in the TPR pattern at 306 °C is attributed to the reduction of bulk NiO [33] while the reduction peaks at 420–450 °C were related to NiO x species that are having strong interaction with ZrO2 [29, 34]. The strong interaction between the nickel metal and the zirconia could be probably due to an increased interfacial area [29, 34]. While in the case of reductive precipitation catalysts, no significant consumption of hydrogen was observed in their TPR profiles which clearly indicates that the Ni is in its metallic state.

Fig. 3
figure 3

TPR profiles of Ni catalysts a Ni/TiO2 (I), b Ni/ZrO2 (I), c Ni/TiO2 (R), d Ni/ZrO2 (R)

Full size image

H2 pulse chemisorption studies

H2 chemisorption results are compiled in Table 2. Ni dispersion, and its particle size are determined based on the assumption that each Ni atom chemisorbs one H-atom. The number of surface Ni species on the TiO2-supported catalysts is more compared to that of the ZrO2 support and the activity is in good agreement with the H2 chemisorption results. The metal dispersion follows a trend: Ni/ZrO2 (I) < Ni/ZrO2 (R) < Ni/TiO2 (I) < Ni/TiO2 (R). This clearly indicates that the metal dispersion depends not only on the preparation method but also on the nature of the support. The smaller particles and higher dispersion are generally useful for the catalytic hydrogenation of NB (Fig. 4).

Fig. 4
figure 4

SEM images of Ni catalysts a Ni/TiO2(R), b Ni/TiO2 (I), c Ni/ZrO2(R), d Ni/ZrO2 (I)

Full size image

Microscopic studies

The SEM images of catalysts prepared by reductive deposition–precipitation method show the presence of well-dispersed Ni particles than in the impregnated catalysts. The greater nucleation in the impregnation method leads to agglomeration of metal nanoparticles.

TEM images of Ni/TiO2 and Ni/ZrO2 catalysts prepared by impregnation and reductive deposition method are presented in Fig. 5. All the catalysts show highly dispersed amorphous Ni nanoparticles. This is in concurrent with the XRD wherein no peaks corresponding to metallic Ni nanoparticles were noticed. Furthermore, H2 pulse chemisorption also supports the results such as particle size i.e., below the detection limit of XRD.

Fig. 5
figure 5

TEM images of Ni catalysts a Ni/TiO2 (I), b Ni/TiO2 (R), c Ni/ZrO2(I), d Ni/ZrO2 (R)

Full size image

XPS studies

The XPS profiles of catalyst samples are displayed in Fig. 6. XPS is a useful technique in order to know the chemical state of Ni. All the reduced samples show two major peaks observed corresponding to the core level of Ni 2P3/2 and Ni 2P1/2 transitions at binding energies (BE) of 852.5 and 869.7 eV, which is due to the presence of Nickel in the metallic state. The XPS profile of Ni/TiO2 (I) seems to be different which might be due to the presence of unreducible NiO species (partial surface oxidation prior to performing the XPS) during the sample transfer.

Fig. 6
figure 6

XPS patterns of reduced Ni catalysts a Ni/TiO2 (I), b Ni/ZrO2 (I), c Ni/TiO2 (R), d Ni/ZrO2 (R)

Full size image

Catalytic activity studies

Effect of temperature

The catalysts prepared by reductive deposition method are found to be efficient for the production of aniline than for the impregnated catalysts (Fig. 7). Moreover, among all the catalysts Ni/TiO2 (R) catalyst prepared by reductive deposition method is found to be an efficient catalyst due to high dispersion of Ni with smaller Ni particles (Table 2). In addition, the XRD studies, TEM and H2 pulse chemisorption results also support the presence of smaller Ni particles in this catalyst. Among all the temperatures, 250 °C is found to be an optimum temperature for yielding the aniline at cent percent conversion. Because of the reducible nature of TiO2, strong metal-support interaction (SMSI) prevails between the Ni and TiO2. It is noteworthy to mention that the Ni is held strongly by reducible TiO2 which results in the formation of few NiTiO3 species as evidenced by the XRD pattern (Fig. 1). That’s why Ni dispersion and thereby nitrobenzene conversion is higher with this catalyst than in the Ni/ZrO2 (R). Besides, greater number of H2 uptake was found in the TiO2-supported Ni catalysts compared to others which might be ascribed to the hydrogen spill-over phenomenon usually found in TiO2-supported metal catalysts [35].

Fig. 7
figure 7

Effect of support and preparation method on the NB conversion over Ni-based catalysts

Full size image

Effect of time on stream

The stability of Ni/TiO2 (R) and Ni/ZrO2 (R) catalysts were prepared by reductive deposition method which can be observed from the Fig. 8, during the time on stream for 10 h at 250 °C. The conversion of NB over Ni/TiO2 (R) catalyst is decreases slightly during the course of time, which may be due to the condensation of reaction intermediates (coke formation) or water released during the reaction. Although the comparable activity of NB was noticed over Ni/ZrO2 (R) at 250 °C, drastic decrease in activity was observed during time on stream study due to the sintering of Ni particles. The higher sintering of Ni/ZrO2 (R) could be ascribed to the lower metal-support interaction. While the mild deactivation of Ni/TiO2 (R) catalyst is owing to higher metal-support interaction. The selectivity towards the aniline is always cent percentage over both the catalysts during the time on stream study. In our previous publication [11], we observed the catalytic performance of Ni/SBA-15 catalyst, during the time on stream for 10 h which remain constant due to hydrophobic nature of SBA-15, while in case of Ni/MgO catalyst, drastic decline in the catalytic performance of NB hydrogenation was observed due to the water adsorption on the MgO support. The adsorbed water may promote brucite–periclase transformation of MgO in case of MgO-supported Ni catalyst. The formation of coke is also another main cause for the deactivation of catalyst during the course of the reaction. Lin et al. [36] studied the NB hydrogenation over Ni/TiO2 and observed a decline in the activity in subsequent cycles due to the hydrophilic nature of catalyst. In order to avoid the hydrophilic nature, the catalyst was coated with carbon which ultimately decreased the hydrophilic nature of catalyst and found to improve the performance of the catalyst. To know the details of deactivation of the present catalytic systems, we have performed H2 pulse chemisorption for spent catalysts and the results are presented in Table 3. The results suggest that the decrease in Ni dispersion and increase in average particle size in Ni/ZrO2(R) spent catalyst is due to sintering of Ni particles which is believed to be responsible for the drastic decrease in catalytic activity. Whereas in the case of Ni/TiO2 (R) spent catalysts, we observed an insignificant decrease in Ni dispersion and almost the same Ni particle size when compared to Ni/TiO2 (R) fresh catalysts which might be caused by high metal-support interactions. Besides, Corma et al., also reported that the deactivation of catalytic system can also happen through the condensation of reaction intermediates [37]. The mild deactivation observed in the case of Ni/TiO2 (R) catalyst might be due to the condensation of reaction intermediate over the catalyst surface and also its hydrophilic nature.

Fig. 8
figure 8

Time on stream study on Ni/ZrO2 (R) and Ni/TiO2 (R) catalysts

Full size image
Table 3 H2 Uptake, dispersion, and particle size of spent Ni catalysts
Full size table

The turnover frequency (TOF) of NB on each catalyst is defined as the number of moles of NB converted per one surface Ni atom per second [12]. The data are presented in Table 2. Among all the catalysts, the Ni/ZrO2 (R) catalyst showed greater TOF than the other catalysts. Although the catalyst is showing greater TOF, drastic decrease in conversion was observed during the time on stream presumably due to the agglomeration of Ni nanoparticles (as evidenced from the H2 chemisorption of the used catalyst). While no such agglomeration was noticed on Ni/TiO2 (R) catalyst, it could be ascribed to the strong metal-support interaction.

Table 4 highlights the superior performance of the present catalytic system when compared to previously published catalysts for NB hydrogenation in vapour phase conditions. Higher NB conversion (96.5 %) with nearly 100 % selectivity towards aniline at 275 °C was reported over Ru/SBA-15 catalyst at atmospheric pressure when compared to the SiO2- and Al2O3-supported Ru catalysts, which clearly shows the advantage of SBA-15 support [38]. Among the Pd-supported MgO–Al2O3 (hydrotalcite) catalysts with different Pd contents, a catalyst with 0.5 wt% Pd exhibited very high NB conversion (97 %) with nearly 100 % selectivity towards AN at atmospheric pressure and at 225 °C [39]. On the other hand, this catalyst suffered activity loss during the time on stream studies, even though, the Pd particle size in the spent catalyst is more or less same as that in the reduced catalyst. The deactivation might, therefore, be due to the water released during the NB hydrogenation. A similar observation was reported over 1 % Pd/MgO–Al2O3 catalyst [40]. In our earlier publication, we reported the role of water over Ni/MgO catalyst on the NB hydrogenation activity [11]. It was reported in another publication [41] that higher NB conversion was observed when NB hydrogenation is coupled with 1–4 butanediol dehydrogenation over Cu/MgO catalysts at 250 °C. However, these catalysts also suffered deactivation during time on stream study. It was reported that among the SiO2-supported Cu catalysts, Cu in combination with Cr and Mo species exhibited good lifetime for the NB hydrogenation reaction [41].

Table 4 Comparison of nitrobenzene hydrogenation activities of 5 wt% Ni/TiO2 (R) catalyst with the reported catalysts
Full size table

Conclusions

In summary, the catalysts prepared by reductive precipitation method are found to be efficient for the hydrogenation of NB. Among all the catalytic systems, Ni/TiO2 (R) catalyst shows good conversion of NB as well as selectivity towards aniline which is attributed to the high Ni dispersion having smaller Ni particles along with strong metal-support interaction. However, decrease in conversion during the time on stream analysis is ascribed to the condensation of reaction intermediate on the catalyst surface and also TiO2 hydrophilic nature. Further efforts will be underway to get efficient catalytic system for consistent catalytic performance for longer lifetime to meet the current demand for aniline.