Introduction

Metal aluminate materials have been studied for many applications including magnetic materials, pigments, sensors, catalysts, and carriers [1]. They can be used as the carrier for catalysts due to their resistance to high temperatures, low surface acidity, and strong interaction with the noble metals, which provide chemical and physical stability [2].

Among the metal aluminates, nickel aluminates have widespread commercial value in the catalytic applications ranging from methane/steam and methanol reforming to hydrocarbon cracking, dehydrogenation, hydrodesulfurization, and hydrodenitrogenation [3].

High surface area NiAl2O4 is of prime importance for the catalytic purposes. In general, high surface area materials will have the small particle size. Hence, the synthesis of nanosized nickel aluminate nanoparticles is worth studing [4–6].

Interest in the synthesis of nanoparticles has increased due to their different properties when compared to the corresponding bulk material. Nickel aluminate (NiAl2O4) is a mixed cation oxide with normal spinel structure, where Al occupies the octahedral sites and Ni occupies the tetrahedral sites. Industrial application of this material is mainly based on its stable structure at high temperature and catalytic features. This form of nickel is more resistant to deactivation by coke formation in steam reforming than when deposited on an inert support [7].

A metal aluminate has various structures due to different cation distributions. Nickel aluminate spinel is an example with partially inverse 2–3 spinels. In many spinels such as nickel aluminate, the degree of inversion depends on the calcination temperature and the preparation procedure [8, 9].

The aluminate preparation has two main steps, preparation of a solid solution and calcination to shape the crystal structures [10]. Several preparation methods such as sol–gel, co-precipitation, solvothermal, impregnation, and mechanical mixing of metal oxides have effects on the preparation of solid solution. The microwave heating and combustion route are effective on the formation of crystal structures during calcination [1, 2].

The selection of appropriate preparation method of aluminate depends on its applications. In catalytic and support applications, the surface properties are important and in many studies sol–gel, co-precipitation and impregnation were used. In Table 1 the application of nickel aluminate as catalyst or catalyst carrier is presented. The nickel aluminates were prepared by various Ni/Al ratios, preparation methods, and calcination temperatures.

Table 1 Some studies on the nickel aluminate preparation as catalyst and carrier
Full size table

Using low calcination temperature and Ni/Al ratio, alumina phases beside the aluminates were obtained. Usually alumina completely converts to metal aluminates at temperature higher than 1000 °C [10].

In this study, four samples were prepared by sol–gel at Ni/Al ratios 0, 0.25, 0.5, and 0.75. A sample with Ni/Al molar ratio equal to 0.75 was prepared by impregnation. The porosity (BET), structure (XRD), morphology (SEM and TEM), and surface acidity (NH3-TPD) of samples were evaluated and compared.

Experimental

Materials and methods

Commercial spherical γ-Al2O3 (BET: 210.13 m2/g) was supplied by Sasol Co. The nickel nitrate hexahydrate (Ni(NO3)2·6H2O), aluminium nitrate nonahydrate (Al(NO3)2·9H2O), urea (CH4N2O), and ethanol (99.5 %, C2H5OH) were purchased from Merck Co. and used as precursors.

In the sol–gel method, the desired amounts of aluminium and nickel nitrate were dissolved in ethanol in the reflux flow closed system at 70–75 °C for 16 h. Then, urea solution in the distilled water was added to adjust pH of solution. The mixture was rested for 48 h to obtain neutral gel. A sample were prepared by using two aging times, 24, 48 h. The obtained products were stepwise calcined by 3 °C/min from room temperature to 500 °C and continue heating at 5 °C/min to 1100 °C and hold at this temperature for 3 h. Four samples were prepared by the Ni/Al ratio equal to 0, 0.25, 0.5, and 0.75, namely SG0, SG0.25, SG0.5, and SG0.75, respectively.

In the impregnation method, the nickel salt was placed on the heater to be melted. The melting point of hydrated nickel nitrate is at 45 °C. Desired amount of alumina was placed in an oven at 120 °C for 1 h to remove physical adsorbed water and then placed in the oil bath at 110–120 °C. The melted precursors were impregnated on the alumina for 30 min in the oil bath. A sample was prepared in the Ni/Al = 0.75 and was called IM0.75. The sample was dried in the oven at 120 °C for 16 h and then stepwise calcined in the electrical furnace by 1 °C/min to 300 °C and remained for 2 h then by 3 °C/min to 500 °C for 1 h and by 5 °C/min to 1100 °C for 3 h.

Characterization

The X-ray diffraction (XRD) patterns of prepared supports were carried out using an X-ray diffractometer (Model GNRMPD 3000) with Cu-Kα radiation at 40 kV and 30 mA. The crystal size was estimated by Scherrer equation (Eq. 1) [1].

$$d = \frac{B\lambda }{\beta \cos \left( \theta \right)}$$
(1)

where d is the average crystallite size of the phase under investigation, B is the Scherrer constant (0.89), λ is the wave length of X-ray beam used, β is the full width half at maximum (FWHM) of diffraction and θ is the Bragg’s angle.

The surface area of the samples was determined by N2 physisorption using an adsorption analyzer type Quanta chrome Nova 2200. The BET measurements, nitrogen adsorption–desorption isotherms, and pore structure analysis were conducted at 77 K. The pore size distribution was determined by the Barrett–Joyner–Halenda (BJH) method.

Temperature-programmed desorption (TPD) measurements of NH3 were performed in BELCAT A. Ammonia adsorption was carried out at 60 °C in a mixture of NH3 (2.5 mL/min) and He (47 mL/min) flows for 1 h, then the product was exposed to He for 30 min at 60 °C to remove all the physically adsorbed species before starting the temperature program. The samples adsorbed ammonia at 60 °C heated up to 700 °C with a heating rate of 10 °C/min. The amount of acidic sites on the catalyst surface was calculated based on desorption amount of ammonia. Desorption signals were monitored by using a TCD detector.

The images of samples were prepared with a scanning and transmission electron microscope. The SEM apparatus used was VEGA II TESCAN. The cross section of each spherical pellet was polished and covered with Au layer and studied with a scanning electron microscope. The measurement was carried out under high vacuum and with an electron beam having a voltage of 15 kV. The TEM images were obtained from Zeiss EM900 by 80 kV.

The metal content of the samples was determined by inductively coupled plasma (ICP-OES simultaneous instrument, model: VISTA-PRO, Varain).

Results and discussion

The XRD patterns of sol–gel samples with different Ni/Al ratios and the impregnated sample are shown in Fig. 1. The sharp peaks of sol–gel samples in comparison with impregnation method related to the existing higher crystalline structures.

Fig. 1
figure 1

XRD patterns of samples prepared by a sol–gel, b impregnation

Full size image

In all sol–gel samples only the NiAl2O4 phase (NiO-Al2O3, ICDD No. 10-0339) by the cubic structure of nickel aluminate was defined, but for IM0.75, the metastable phase NiAl10O16 (NiO-5Al2O3, ICDD No. 371292) by a monoclinic structure existed as well because the crystallization mechanism was different for sol–gel and the impregnation method [11]. The main differences in the aluminate structures are due to the cation distribution in the octahedral and tetrahedral interstices of the array [17]. The nickel aluminate prepared by sol–gel at high calcinations temperature in the other reports confirmed that only NiAl2O4 phase was formed [11, 16].

Calcination at high temperature resulted in penetration of Ni atoms maybe occupying the tetrahedral sites in NiAl2O4 matrix via a solid-state reaction as the limited solid solution of Ni was in NiAl2O4 phase [11]. A lower amount of NiAl2O4 was formed in the impregnation in comparison with the sol–gel method since the diffusion limitation was added to the solubility limitation. In the sol–gel method, homogenous Ni atom distribution was expected [10].

Pure alumina (SG0) was in the alpha phase. The mixture of alumina and nickel aluminate was formed from SG0.25. By increasing the amount of nickel in the samples, the characteristic peaks of nickel aluminates were enlarged, and most were in SG0.5; therefore, the highest amount of nickel aluminate was in this sample. The estimative amounts of phases in the samples are given in Table 2.

Table 2 Crystal sizes and estimated amount of each phase of sol–gel made samples
Full size table

In the sol–gel made samples, the XRD patterns were not overlapping unlike the impregnation-made sample. The crystal sizes calculated using the Scherrer equation (Eq. 1) were in the nanorange 17––35 nm and are reported in Table 2 with a 5 % probability error [18]. The crystal sizes of alpha alumina were larger than nickel aluminate because of their differences in unit cell structures. In the nickel aluminate and alumina mixture (SG0.25), the crystal sizes of alumina were larger than pure alumina (SG0).

The values of the BET surface area and amount of Ni in each sample are shown in Table 3. In the sol–gel made samples, the surface areas were not significantly different and were quite low, ranging between 7.7 and 12.4 m2/g, but IM0.75 had higher surface area (67.2 m2/g) in comparison with SG0.75. The surface area of sol–gel samples was similar to those reported by Mekasuwandumrong et al. [11]. Additing Ni atoms into the gamma alumina via impregnation has reduced the surface area, suggesting that the impregnated Ni blocked some pores in the support [16, 19]. SG0.5 had the largest surface area among the sol–gel samples.

Table 3 Amount of nickel and BET surface characteristic
Full size table

Figure 2 shows the N2 adsorption–desorption isotherms of the studied samples. The nickel aluminate samples' isotherms were similar to type IV of the IUPAC classification [20]. The maximum N2 adsorption at the saturation pressure was for IM0.75 and among sol–gel samples was for SG0.5, because of their surface area. The pure alumina, SG0, had an isotherm similar to type III of IUPAC classification, which represented macropores materials.

Fig. 2
figure 2

N2 adsorption–desorption of a different Ni/Al ratio, b impregnation and sol–gel method

Full size image

The hysteresis loops suggested that the nickel aluminate samples had mesoporous characters. The hysteresis loop was similar to type H1 for SG0.5 and H4 for SG0.25, SG0.75, respectively. These shape variations were related to different pore shapes [20].

The pore size distributions calculated by the BJH method are given in Fig. 3. SG0.25, SG0.5, and IM0.75 samples displayed a bimodal pore size distribution with one-peak maximums around 4, 5, and 3 nm and another larger, broader peak maximum around 6, 9, and 12 nm, respectively, hence most of the pores were around the lower limit of mesopores range. This type of pore size distribution for Ni-aluminate prepared by sol–gel was reported earlier [16]. The pore size distribution of SG0 was broad and tailed up to 100 nm. This sample had impressive numbers of macropores.

Fig. 3
figure 3

Pore size distribution of samples a at different Ni/Al ratios, b impregnation and sol–gel methods

Full size image

The IM0.75 had larger pore distribution than SG0.75 (Fig. 3b). Higher crystallinity structure in SG0.75 was caused to disturbed pores and void spaces were diminished and the lower pore diameter and surface area was obtained.

Aging time during gelation is the parameter that may affect some properties of aluminates. SG0.5 was prepared by two aging times, 24 and 48 h. In Fig. 4 the effect of this time on the N2 adsorption–desorption isotherms and pore size distributions is considered. The amount of adsorbed N2 at the saturation pressure was the same. This means that the surface area was not different but the hysteresis loop was become thinner by increasing the aging time so the micropores may be changed. Fig. 4b confirms that increasing aging time changed the pore size distribution from bimodal to unimodal and most of micropores were destroyed.

Fig. 4
figure 4

a N2 adsorption–desorption and b pore size distribution of sol–gel-made samples prepared at 24 and 48 h aging time

Full size image

The images of sol–gel-made samples obtained by TEM are given in Fig. 5. The particles sizes of pure alumina were larger than nickel-containing samples because the structures were different, and the presence of Ni atoms in NiAl2O4 inhibited the growth of crystals [11]. The cubic particles of nickel aluminates are seen obviously in images. The particle size values were in accordance with the crystal sizes that were calculated by XRD analysis, so the particles were well crystallized. The TEM images illustrate the formation of aggregates of nanostructures of nickel aluminate.

Fig. 5
figure 5

TEM images of sol–gel-made samples a SG0, b SG0.25, c SG0.5, and d SG0.75

Full size image

The backscatter SEM pictures of the IM0.75 (Fig. 6) reveal the good distribution of Ni atoms in the whole of alumina particle. Tiny bright particles represent the Ni atoms. Some pores in the cross section image were seen to confirm the BET analysis.

Fig. 6
figure 6

SEM image of nickel aluminate prepared by impregnation method

Full size image

The NH3 temperature programmed desorption profiles of the sol–gel and impregnation made samples are given in Fig. 7. The peak areas of the samples obtained from sol–gel were quite lower in comparison with the impregnation made sample because of their different surface area.

Fig. 7
figure 7

NH3-TPD graph of prepared samples

Full size image

Typically, the acidity of α-Al2O3 is much lower than γ-alumina phase [21]. Transition aluminas are a defect spinel and possess substantial acidity. The nickel aluminates had lower acidity than alpha alumina as seen in Table 4. In general, spinel-type mixed oxides are crystalline materials and are not expected to possess surface hydroxyl groups, especially at the drying conditions used in the TPD experiments.

Table 4 NH3-TPD data of prepared samples
Full size table

The Ni atoms decreased the surface acidity of alumina in quantity and intensity. For the sol–gel made sample, desorption peaks were small and the profiles became almost flat. Two desorption peaks corresponding to different acidic sites were observed. The peaks at maximum around 190 °C for alpha alumina (SG0) shifted to lower temperature between 125 and 145 °C so the intensity of weak acid sites was decreased. The peak at maximum 587 °C for SG0 shifted to 523, 583, and 508 °C for SG0.25, SG0.5, and SG0.75 and the intensity of strong acid sites was decreased as well. The amount of desorbed ammonia that revealed the quantity of acid sites from 0.126 mmol/g had reached to the range of 0.104–0.054 mmol/g.

The IM0.75 sample had a moderate amount of weak acid sites (T = 142 °C, peak I in Fig. 7) in comparison to SG0.75. This was consistent with the amount of NH3 desorbed, which was 0.204 and 0.076 mmol/g, respectively, for the low and high temperature peaks of IM0.75 [15].

The surface acidity density has the better sense for comparing the amount of surface acid sites per reachable surface area. When the acidities were expressed per unit area, the number of acid sites in a distinct area decreased by increasing the amount of nickels. IM0.75 had the lowest density of surface acidity in comparison with the sol–gel-made samples.

Conclusion

Nickel aluminates were synthesized by sol–gel and impregnation methods. In the sol–gel method only NiAl2O4 was formed, but with the impregnation method, the metastable phase, NiAl10O16, is obtained as well. The impregnation-made sample has the highest surface area and lowest surface acidity density. By increasing the Ni/Al ratio, the density of surface acidity decreased. The higher surface area in sol–gel-made samples is related to pure nickel aluminate.