Effect of modi fied carrier fluoride on the performance of Ni–Mo/Al2O3 catalyst

Zhao Han · Zhuang Ma · Cheng‑Bo Wang · Chun‑Yu He · Ming Ke · Qing‑Zhe Jiang · Zhao‑Zheng Song

Abstract An alumina support was modi fied by fluorine via impregnation to investigate the effect of fluoride content on the reactivity of Ni–Mo/Al2O3 catalyst. The catalyst was characterized by X-ray diffraction, N2 adsorption–desorption (Brunauer–Emmett–Teller) isotherms, temperature-programmed desorption of ammonia, X-ray photoelectron spectroscopy and high-resolution transmission electron microscopy. Sulfur etheri fication performance of the catalyst was studied using a fixed-bed reactor.The results show that increasing fluoride content increases the pore volume and pore size but reduces the speci fic surface area. In addition, the degree of sul fidation of Ni first increases and then decreases. The amounts of strong acid and total acid also increase with increasing fluoride content. Performance evaluation of the catalyst reveals that the fluoride content has a minor effect on the thioetheri fication performance of the catalyst; however, an optimum fluoride content, which was determined to be 0.2%, can ensure lower ole fin saturation and an efficient diene selective hydrogenation.

Keywords Fluorine · Thioetheri fication · Fluorinated ammonia · Selective hydrogenation · NiMo/γ–Al2O3

1 Introduction

Automobile exhaust emissions are largely accountable for haze. One of the important measures to control automobile exhaust emissions is to improve the quality of automobile gasoline. Fluid catalytic cracking (FCC) gasoline, characterized by high sulfur and ole fin contents, is the main blending component of automotive gasoline in China. Therefore, the key to achieve cleaner gasoline in China is to reduce the sulfur and ole fin contents of FCC gasoline and to minimize its octane loss to the highest possible extent. At present, a clean production of catalytic gasoline is realized worldwide by the selective hydrodesulfurization technology (Huang et al. 2018a, b; Yu et al. 2016; Zhang et al. 2018), in which the thioetheri fication and diole fin selective hydrogenation(Miller et al. 2000; Toba et al. 2005) are the key processes.

The full-fraction FCC gasoline converts small molecules of mercaptans into larger thioethers through thioetheri fication. The light fractions with ultra-low sulfur and high ole fin contents, together with heavy fractions of high sulfur contents, were separated by recti fication. Following this, the heavy fractions were directed into the selective hydrodesulfurization device for deep desulfurization, during which the diole fin in the gasoline was also selectively hydrogenated in the thioetheri fication reaction. This prevented the in fluence of the diene on the downstream selective hydrodesulfurization catalyst, thereby maintaining its activity and stability(Frey 1998; Hearn 1996; Hearn et al. 2002). This process removes mercaptan sulfur from gasoline by sul fide etheri fication and prevents the saturation of light ole fin-rich components in the entire selective hydrodesulfurization process,resulting in large loss of octane number of gasoline (Nocca and Debuisschert 2002; Rock and Shorey 2003). Another outstanding advantage of using thioetheri fication to desulfurize mercaptan is that no alkali is needed in this process.This avoids any problems related to disposal of waste alkali residues, thereby rendering this process more environmentally friendly.

Nowadays, thioetheri fication catalysts mainly load NiO and MoO3on alumina supports. Modi fication of the catalysts has become an important strategy to further improve the catalytic activity, selectivity and service life of the catalysts(Hartwig 2008; Huang et al. 2018a, b; Liu and Szostak 2018;Rodriguez et al. 1999; Shen et al. 2015). Among the numerous modi fication strategies, synthesis of fluorine-modi fied γ-Al2O3was one of the most effective (Flego and Parker Jr 1999; Kwak et al. 2000; Lewis et al. 1989). A previous study(Rodriguez et al. 1999) has shown that the introduction of F in a catalyst support can change the acid content and acid distribution of the catalyst and is conducive to the dispersion and existence of metals in the catalyst support, as well as to the speci fic surface and pore structure of the support.F is highly electronegative, and hence, when Ni and Mo are loaded on the catalyst, electrons on the surface of adjacent metal atoms are attracted toward it, leaving the metals in an electron-de ficient state. These are available for the formation of metal sul fide during the vulcanization of the metals.In this study, the effect of carrier fluoride modi fication on the activity of thioetheri fication catalyst was investigated,together with the selectivity of the catalyst after the introduction of F.

2 Experiment

2.1 Preparation of catalyst

An appropriate amount of dilute nitric acid solution was added to the homogeneous mixture of pseudo-boehmite and sesbania powder and extruded by twin-screw extruder. Following this, it was dried in an oven at 110 °C for 8 h, placed in a muffle furnace at 550 °C for 4 h, ground and sieved. A 20–40-mesh sieve was selected to obtain the Al2O3carrier.The carrier was impregnated with different mass fractions of NH4F solution, such that the loading of F on the carrier was 0.2%, 0.4%, 0.6%, 0.8% and 1.0% of the total mass of the carrier. After this, the fluorine-modi fied Al2O3carrier was impregnated simultaneously with aqueous solutions of ammonium molybdate and nickel nitrate; the MoO3and NiO contents in the obtained sample were equal. After each impregnation step, the samples were first dried naturally,then dried at 110 °C for 8 h and then roasted at 550 °C for 4 h. Correspondingly, the samples were named F-1, F-2,F-3, F-4 and F-5. The untreated catalyst (without ammonia fluoride solution treatment) was used as blank sample and will be referred to as C-1.

2.2 Characterization of catalyst

X-ray diffraction (XRD): XRD-ray diffractometer (Bruker D8 Advance, Germany) was used for the phase analysis and characterization of the samples. The X-ray source was Cu Kα, tube voltage was 40 kV, tube current was 30 mA, scanning rate was 4° min−1, and scanning range was 5°–90°. The data were collected automatically by a computer.

N2adsorption–desorption (BET) isotherm: The BET isotherm was recorded on an ASAP 2020 automatic speci fic surface area and pore diameter distribution instrument from Micromeritics, USA. The adsorbed gas was a mixture of nitrogen and helium, in which nitrogen and helium were the adsorbed gas and carrier gas, respectively. The adsorption measurement was carried out at 77 K. The speci fic surface area was calculated by the multi-point BET method.The pore size was 0.35–500 nm, resolution of microporous region was 0.1 nm, and the minimum detection value of pore volume was 0.0001 cm3g−1.

Temperature-programmed desorption of ammonia(NH3-TPD): The automatic chemical adsorbent of Autochem II 2920 made in the USA was used. The sample quantity was 0.2 g, particle size was 20–40 mesh, and pretreatment was carried out at 500 °C for 30 min. Ar was the carrier gas at a flow rate was 40 mL min−1. Ammonia flow rate was 20 mL min−1; adsorption was measured at 100 °C for 30 min and temperature programmed at 10 °C min−1up to 800 °C.The outputs were simultaneously detected by thermal conductivity cell detector and mass spectrometry detector.

X-ray photoelectron spectroscopy (XPS): The Mo and Ni species on the surface of the sul fided catalyst were analyzed by K-α X-ray photoelectron spectrometer (Thermo Fisher).Using Al Kα X-ray as excitation source, a base vacuum of 3 × 10−9mbar was attained, and the spectra were corrected by electron binding with C 1 s peak (284.8 eV) of the contaminated carbon. XPS analysis of Mo and Ni 3dwas carried out in order to obtain the degree of sul fidation of the active metal of the catalyst.

High-resolution transmission electron microscopy(HRTEM): Tecnai G2 F20 (FEI, USA) high-resolution transmission electron microscope was used to observe the morphology (stripe length, number of layers and other parameters) of the active components on the catalyst surface.

2.3 Evaluation of sul fide properties

Thioetheri fication performance was evaluated on a selfassembled micro- fixed-bed reactor. The inner diameter of stainless steel reactor tube was 10 mm, and the length was 550 mm. The reactor adopts three-stage temperature control mode. The middle constant temperature section was filled with 5 mL catalyst, and the rest were partly filled with quartz sand. The raw material is simulated FCC gasoline, which is composed ofn-hexene, isoprene andn-hexane at mass fractions of 30%, 1% and 69%, respectively. Following this,200 µg g−1n-butyl mercaptan was added to the simulated oil.

Catalyst pre-sulfurization: H2and sul fide oil were introduced into the reaction system. The sul fide oil was petroleum ether containing 2 wt% CS2(90–120 °C). The pre-vulcanization conditions of the temperature-programmed pre-vulcanization process were as follows: pressure: 3 MPa, H2/oil volume ratio: 300:1, space velocity (weight hourly space velocity): 2 h−1. Vulcanization was conducted at 310 °C for 6 h. The reaction conditions were as follows: reaction pressure:3.0 MPa, H2/oil volume ratio: 6:1, liquid hourly space velocity:4 h−1, reaction temperature: 110 °C.

2.4 Sample analysis and data processing

The mercaptan content was determined by GB/T1792-88 potentiometric titration.

The compositions of FCC simulated oil before and after the reaction were quantitatively analyzed by SP-3420A gas chromatograph. The chromatographic data were recorded and processed by BF-2002 (Beijing Beifen-Ruili Analytical Instrument Group, China) chromatographic workstation.

The conversion formula of thiol conversion rate (XL), selective hydrogenation rate of diole fin (XE), ole fin double-bond isomerization ratio (XY) and ole fin saturation ratio (XB) in the reaction process are as follows:

In the above equations,WLandWL1represent the mass fraction of mercaptan in the sample before and after the reaction;WEandWE1represent the mass fraction of the diene in the sample before and after the reaction;WYandWY1are,respectively, the mass fraction ofn-hexene in the raw material before the reaction and mass fraction of isomeric ole fin after reaction;WS1is the mass fraction ofn-hexene in the sample after the reaction.

Fig. 1 XRD patterns of F–X series of catalysts: oxidized state (a) and sulfurized state (b)

3 Results and discussion

3.1 Catalyst characterization

3.1.1 XRD analysis

The series of fluoride-modi fied catalysts were characterized by XRD, and the results are shown in Fig. 1. It can be seen from Fig. 1a that the characteristic peaks of γ-Al2O3at 39.5°,45.9° and 66.9° correspond to the (311) (400) and (440)crystal planes (JCPDS-0425) (Li and Shao 2008). At the same time, it is found that with increasing fluoride content,the obvious characteristic peaks of NiO crystals appearing in the spectrum at 2θ values of 37.3°, 43.4°, 63.0° and 75.6°correspond to the (111), (200), (220) and (311) crystal faces,respectively, of NiO crystals. In addition, analysis of the spectrum of F-5 catalyst with JADE 5.0 software reveals that the peaks of NiAl2O4crystal appear at 27.02°, 44.98° and 65.55°, which indicates that with increasing fluoride content in the catalyst, the interaction between Ni and the carrier changes. This is detrimental to the dispersion of Ni on the catalyst. The aggregation of Ni causes the catalyst to adopt a spinel structure on the surface during firing, and this is not only harmful to the sulfuration of Ni, but can also reduce the dispersion of the active center of the catalyst, thereby affecting the activity of the catalyst.

Figure 1b shows the XRD spectra of the sul fide state of the catalyst. Compared with the oxidized state spectrum, the main characteristic diffraction peaks are the peaks of Al2O3carrier. Analysis by JADE 5.0 software reveals that there are no characteristic peaks of sul fide loaded with metal, and the spinel structure disappears after sul fidation, indicating that the metal on the surface of the carrier was uniformly dispersed after the catalyst was vulcanized, and no crystal aggregation occurred.

Table 1 Textural properties of the F–X series of catalysts

3.1.2 BET characterization

BET isotherms of the fluoride-modified catalysts were recorded, and the results are listed in Table 1.

Table 1 shows that the speci fic surface area of the catalysts decreases with increasing fluoride content, while the pore size and pore volume tend to increase. This may be because of the erosion of fluoride to the thin-walled structure between the channels, resulting in reduced speci fic surface area (Ebrahimynejad et al. 2014). In addition, the catalyst produced a large amount of AlF3after modi fication; AlF3has a smaller speci fic surface area (Benítez et al. 1996) than γ-Al2O3.

3.1.3 NH3‑TPD characterization

NH3-TPD characterization was carried out for the fluoridemodi fied catalysts, and the results are shown in Fig. 2.

Figure 2 shows the NH3-TPD spectrum of the sulfided state of the C-1 and F–X series of catalysts. In Fig. 2a, peaks in the range 20–200 °C, 200–350 °C and 350–550 °C correspond to weak, medium and strong acidic sites interacting with NH3molecules (Andonova et al. 2007). As the fluoride content in the catalyst increases, the number of weak acidic sites first increases slightly and then decreases, while the amounts of strong acid and total acid increase as a function of the fluoride content. Figure 2b shows the NH3-TPD spectrum of the F-1 catalyst before and 60 h after the reaction. As is apparent from the figure, the strong acid center in the catalyst disappears after the reaction, indicating that the strong acid center is masked during the reaction. At the same time, analysis of the peak area of the weak acid peak at 180 °C suggests that the weak acidity of the catalyst decreases significantly after the reaction; the total amount of acid also decreases significantly. This can be attributed to the fact that the interaction between fluoride and the exposed OH reduces the bond energy of OH,thereby increasing the acidity of the latter (Corma et al.,1985). Lewis acid sites exist on the surface of γ-Al2O3,and fluoride may replace oxygen atoms or OH bonds. At low fluoride contents, the F atom may replace the OH bond on the surface of γ-Al2O3and exist in the form of AlFi(OH)3−(i= 1–3) and free radicals. With increasing fluoride content, a large amount of AlF3will be generated (Miciukiewicz et al. 1989; Qu et al. 1998). However,the excess acid and acid strength of the catalyst will lead to the superposition reaction of olefins on the acid sites during the reaction. Excessive superposition reaction will affect the quality of oil and reduce the octane number of oil products. Excess acidity of the catalyst will lead to the formation of carbon deposits, and the pore channels of the catalysts would be blocked. This will accelerate the deactivation of the catalysts, which is not conducive to the sulfide etherification reaction.

3.1.4 XPS characterization

The series of fluoride-modi fied catalysts were characterized by XPS. Figure 3 shows the XPS spectra of Ni 2p3/2and Mo 3din the vulcanized state.

Ni 2p3/2spectrum was analyzed by means of XPS peak differentiation fitting analysis using XPSPEAK software(Fig. 3a). For the sul fided state of the three types of catalysts with a single load of Ni, the binding energies of the two species were 852.9, 856.7 and 858 eV. In analogy with the analysis by Lai et al. (Lai et al. 2013), it is estimated that the first peak corresponds to sul fided Ni species (NiS), and the latter two peaks correspond to two types of unvulcanized,oxidized Ni–Al spinel species.

The high electronegativity of F results in the attraction of electrons in the outer layer of the Ni, thereby decreasing the electron density in the outermost layer of Ni and increasing the binding energy (as seen in Fig. 3 and Table 2).For F-1, F-2 and F-3 catalysts, the degree of sul fidation of Ni increased, and the Ni3S2content in F-3 catalyst reached 35.7%, which was 8.8% higher than that of unmodi fied C-1catalyst. When the fluoride content in the catalyst carrier was further increased, the Ni3S2content decreased. According to the BET analysis, when the fluoride content is very high,the speci fic surface area of the catalyst will decrease, and the aggregation of Ni will cause the catalyst to form a spinel structure on the surface during calcination. This is not conducive to the vulcanization of Ni and reduces the Ni3S2content in the vulcanized catalyst.

Fig. 3 Ni XPS spectrum of the F–X series of catalysts Ni 2p3/2 (a) and Mo 3d (b)

Table 2 Binding energy and degree of sul fidation of Ni species on the catalyst surface

Figure 3b shows the XPS spectra of Mo 3dof the catalyst. Based on previous studies (Ding et al., 2009; Wang and Ozkan, 2005), the sul fided Mo 3dXPS peak differentiation fitting reveals that there are S 2 s peaks (226.3 eV)(Andonova et al., 2007) and three groups of Mo 3d5/2and Mo 3d3/2double peaks in the series of catalysts: The first group of double peaks appears at 229.1 eV (Mo 3d5/2) and 231.8 eV (Mo 3d3/2) and corresponds to Mo (IV), i.e., completely sul fided Mo 3dspecies. The second group of double peaks emerging at 230.1 eV (Mo 3d5/2) and 233.5 eV (Mo 3d3/2) could be assigned to Mo(V), i.e., molybdenum oxysul fide copolymer that has not been completely vulcanized(MoOxSy) (Ninh et al., 2011). The third group of double peaks at 231.2 eV (Mo 3d5/2) and 235.6 eV (Mo 3d3/2) can be ascribed to Mo (VI), i.e., oxidized Mo species that are not sul fided. Table 3 shows that the binding energies of Mo (IV)3d5/2and Mo (IV) 3d3/2increase slightly after fluoride modification. Compared with the C-1 catalyst, the binding energies of Mo (IV) 3d5/2and Mo (IV) 3d3/2of the F-5 catalyst increased by about 0.2 eV. At the same time, with increasing fluoride content in the catalyst, the degree of vulcanization of Mo increased. The degrees of sul fidation of Mo in the C-1 and F-5 catalysts are 58.3% and 64.2%, respectively,suggesting a 10% increase in the degree of sul fidation of Mo in the two catalysts.

3.1.5 HRTEM characterization

The sulfurized F–X series of catalysts were characterized by HRTEM.

Figure 4 shows the HRTEM images of the sul fided F–X catalysts. The MoS2structure can be clearly identi fied in the images. Ni could be hardly observed by HRTEM, because it is present in the corner of the MoS2stack. The layer number and length of MoS2fringes increased obviously with the addition of fluoride in the catalyst, which was consistent with the increase in molybdenum sul fidation with increasing fluoride content. The textural properties of the MoS2fringes as obtained by statistical analysis are shown in Table 4. The average length of the MoS2fringes of the unmodi fied C-1 catalyst was 4.08 nm, which increased to 4.83 nm for the F-1 catalyst, corresponding to a 15.5% increase. Meanwhile, the average length of the MoS2fringes increased with increasing fluoride content. MoS2fringe length of the F-5 catalyst was 5.81 nm, which was 42.4% higher than that of the F-1 catalyst. Data analysis of MoS2fringe layers showed that the number of MoS2fringe layers in catalyst increased with the addition of fluoride, resulting in new 4–5 layered multilayer fringes. A previous study (Andonova et al. 2007)showed that the angular position of multilayer NiMoS had strong hydrogenation activity, and it was the active center of the catalyst. More the number of layers, higher will be the hydrogenation activity of the catalyst. The increase in stacking layers will increase the number of active sites at the corners, thus improving the ole fin saturation performance and reducing the selectivity of the catalyst.

Table 3 Binding energy and degree of sul fidation of Mo species on the catalyst surface

3.2 Catalyst performance evaluation

The reaction of F–X series of catalysts was conducted primarily to investigate the mercaptan etheri fication performance of the catalyst, selective hydrogenation performance of diole fins, ole fin saturation and ole fin isomerization performance of the catalyst.

It is evident from Fig. 5 that the addition of fluoride to the catalyst has a minor effect on the thioetheri fication performance and the mercaptan content is maintained above 96%.At the initial stage of the reaction, the catalyst is strongly acidic. However, as the reaction progresses, the carbon deposits on the catalyst are substantial; the hydrogenation activity decreases sharply, and the ole fin isomerization and the selective hydrogenation performance of the diene of the catalyst also decrease. Higher the fluoride content of the catalyst, faster will be the decrease. Combined with the results of NH3-TPD analysis (Fig. 2), the change in acidity of the catalyst before and after the reaction is very obvious. The strong acidic center of the catalyst after the reaction is basically absent, and the amount of weak acid is also decreased,thereby decreasing the total amount of acid signi ficantly. It can be seen that if the fluoride content in the catalyst is very high, the activity of the catalyst rapidly decreases. The F-1 catalyst (low fluoride content) ensures good thioetheri fication and selective hydrogenation of the diole fin, lower olefin hydrogenation, improved selectivity and increased ole fin isomerization capacity. Thus, the reaction performance of the F-1 catalyst is superior to that of the unmodi fied C-1 catalyst.

Fig. 4 HRTEM images of sul fided F–X series of catalysts

Table 4 Textural properties of MoS2 on sul fided F–X series of catalysts

4 Conclusion

After modi fication by fluoride, the pore structure changes,the catalyst has a smaller speci fic surface area and the pore size and pore volume increase to some extent. Simultaneously, the acidity of the catalyst is also changed. The strong acid center appeared in the fluoride-modi fied catalyst, and the amount of strong acid and total acid increased with increasing fluoride content. XRD shows that when the fluoride content in the catalyst is very low, the metal has a good dispersion on the surface of the support, while at high fluoride content, a distinct spinel structure appears in the catalyst.

The introduction of F changes the interaction between Ni and Mo in the catalyst and the carrier, and the degree of sul fidation of Ni is increased at low concentrations and decreased at high concentrations. The degree of vulcanization of Mo increased with increasing fluoride content.The surface morphology of the catalyst changed signi ficantly after vulcanization. With the increase in fluoride content, the length of the stripe of MoS2increased gradually, and the number of stacking layers also increased to a certain extent. More five-layered stripes appeared in the F-5 catalyst.

The fluoride-modi fied catalyst showed strong hydrogenation performance at the initial stage of the reaction, but as the reaction progressed, the stronger acidic nature of the catalyst led to faster deactivation of carbon deposition. The strong acid center gradually disappeared, and the amount of acid decreased. When the reaction reaches equilibrium, the ole fin hydrogenation performance and the diene selective hydrogenation performance become weaker. The F-1 catalyst is a highly selective catalyst with less ole fin saturation, higher selectivity and less octane number loss. It also ensures good thioetheri fication and selective hydrogenation performance of the diole fin.

Fig. 5 Reaction performance of the F–X series of catalysts thiol conversion rate (a), ole fin saturation rate (b), ole fin isomerization rate (c) and selective hydrogenation rate of diene (d)

AcknowledgementsThis work has been supported by the National Natural Science Foundation of China (Grant No. 21776303).

Open AccessThis article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source,provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creat iveco mmons.org/licen ses/by/4.0/.