Ni 2 P/Al 2 O 3 hydrodesulfurization catalysts prepared from hypophosphite under a nitrogen atmosphere

Abstract

A novel method for preparing Ni₂P/Al₂O₃(L) catalysts in an N₂ atmosphere by decomposition of hypophosphites was proposed, and Ni₂P/Al₂O₃(T) catalyst was synthesized by the temperature programmed reduction method in a H₂ atmosphere for comparison. These prepared catalysts were washed with deionized water to remove impurities. The X-ray diffraction, N₂-adsorption specific surface area measurements, CO uptake, and X-ray photoelectron spectroscopy were applied to characterize these catalysts. The activities of the Ni₂P/Al₂O₃ catalysts prepared with the two different methods were tested in the dibenzothiophene hydrodesulfurization reaction.

Introduction

With the increasingly stringent environmental regulations in recent years, hydrodesulfurization (HDS), as the most used method to remove sulfur (S) in the petroleum refining industry, has attracted much attention.^1,2 The efficiency of the HDS method was highly influenced by the applied catalysts. ³ Thus, the academic and industrial communities have tested a series of catalysts, such as noble metal catalysts, Mo or MoS₂ catalysts, and transition metal phosphides catalysts. Among these, Ni₂P with high activity and easy access has been reported as the most promising catalyst for the HDS process. ⁴

The traditional method for Ni₂P preparation was a temperature programmed reduction (TPR) in an H₂ atmosphere.^5,6 However, a heat treatment method with hypophosphite in an N₂ atmosphere has roused great interest. ⁷ Compared with the Ni₂P catalyst synthesized with hypophosphites in an N₂ atmosphere, the synthesis of Ni₂P from hypophosphites in H₂ atmosphere was expensive and complex.

In this article, a simple and novel heat treatment method for synthesizing Ni₂P catalyst was proposed. The Ni₂P/Al₂O₃(L) catalyst was synthesized by direct heating to the preparation temperature with NaH₂PO₂ and NiCl₂·6H₂O in a flow of N₂ . For comparison, Ni₂P/Al₂O₃(T) catalyst was prepared with NaH₂PO₂ in an H₂ atmosphere with the TPR method. The effects of the preparation method on the structure and dibenzothiophene (DBT) hydrodesulfurization (HDS) catalytic activity of the prepared catalysts and washed catalysts were investigated.

Experimental

Preparation of catalysts

The catalyst precursor was prepared via impregnating nickel chloride (NiCl₂·6H₂O) and sodium hypophosphite (NaH₂PO₂) on an Al₂O₃ support (initial P/Ni molar ratio = 2, Ni = 10 wt.%). In a typical process, 0.44 g of NiCl₂·6H₂O and 0.60 g of NaH₂PO₂ were dissolved in 20 mL of deionized water. The 2.0-g Al₂O₃ support was impregnated with the above solution for 12 h and then evaporated at 80°C. The resulting precursor was reduced by different methods. The precursor obtained by directly heating to 300°C and being kept for 2 h in an N₂ atmosphere was denoted as Ni₂P/Al₂O₃(L). This precursor was heated to 300°C at a rate of 2°C/min and kept for 2 h in flowing H₂ (100 mL/min) to form the catalyst denoted as Ni₂P/Al₂O₃(T). Both catalysts were passivated at room temperature in an O₂/N₂ mixture (1.0 mol% of O₂) at 20 mL/min for 1 h. The catalysts were then washed with deionized water 3 times to remove impurities. The washed catalysts were denoted as Ni₂P/Al₂O₃(L-W) and Ni₂P/Al₂O₃(T-W), respectively. Washing the catalysts could expose much more active site and thus increase the surface area, which is beneficial to reveal the intrinsic reaction.

Characterization of catalysts

X-ray diffraction (XRD) data were collected on a D/max-2200 PC X-ray diffractometer using Cu Kα radiation. The 2θ scans covered the range 10–80°. The typical physicochemical properties were analyzed using a Quantachrome Autosorb-1-MP analyzer. CO chemisorption uptake was performed in a Micromeritics ASAP 2010 apparatus. X-ray photoelectron spectroscopy (XPS) spectra were recorded with a ESCALAB MKII spectrometer equipped with a hemispherical analyzer using monochromatic Mg Kα radiation (true = 1253.6 eV).

Catalytic activities

The catalytic activity was tested with an HDS reaction in a fixed-bed reactor with 0.5 wt.% DBT in decalin. Catalytic activity was measured at 320°C, 4.0 MPa, hydrogen/oil ratio of 500 (V/V), and weight hourly space velocities (WHSVs) of 8.0 h⁻¹. The products were analyzed by an Agilent-6890N gas chromatograph equipped with a flame ionization (FID) detector using a commercial HP-5 column.

Results and discussion

XRD

Figure 1 shows the XRD patterns of the Al₂O₃ support, Ni₂P/Al₂O₃(L) and Ni₂P/Al₂O₃(T) catalysts. The broad peaks at 2θ = 37.5°, 45.5°, and 66.1° belong to Al₂O₃. The sharp peak at 40.6° belongs to Ni₂P (PDF: 03-0953). No other peak related to Ni or P appeared, indicating that both methods could produce a pure Ni₂P active phase. The average size of the Ni₂P crystallite (D_c) calculated with the Scherrer equation is shown in Table 1. The average size of Ni₂P/Al₂O₃(L) catalyst was 18.4 nm, that is smaller than that for Ni₂P/Al₂O₃(T) catalyst (22.6 nm). This result indicated that the heat treatment method forming a Ni₂P phase with the strongly reducing gas PH₃ was beneficial to produce a smaller average crystallite size. ⁸

Figure 1.

XRD patterns of the support and catalysts.

Table 1.

Textural characterization of the support and catalysts.

Sample	SBET (m2 g−1)	Vp (cm3 g−1)	dp (nm)	Dc (nm)	CO uptake (μmol g−1)
Al2O3 support	205	0.47	9.2	—	—
Ni2P/Al2O3(L)	121	0.20	6.6	18.4	27
Ni2P/Al2O3(T)	116	0.25	8.6	22.6	27
Ni2P/Al2O3(L-W)	140	0.28	8.0	—	34
Ni2P/Al2O3(T-W)	120	0.27	9.0	—	32

BET

Table 1 shows the Brunauer–Emmett–Teller (BET) results for the Al₂O₃ support and catalysts. The surface area (S_BET) and pore volume (V_p) of the Al₂O₃ support were 205 m² g⁻¹ and 0.47 cm³ g⁻¹, respectively; however, the S_BET and V_p values of the Ni₂P/Al₂O₃ catalysts decreased significantly because of the blockage of pores by loading the nickel phosphide phase. The S_BET of the Ni₂P/Al₂O₃(L) was 121 m² g⁻¹, higher than that for the Ni₂P/Al₂O₃(T) catalyst (116 m² g⁻¹). This result may be caused by the run over of PH₃—releasing pores blockage. The S_BET and V_p of Ni₂P/Al₂O₃(L) and Ni₂P/Al₂O₃(T) showed an increase after washing. What deserves our attention was that the S_BET of Ni₂P/Al₂O₃(L) catalyst had a significant increase from 121 to 140 m² g⁻¹after washing. These results indicate that the support pores had been blocked by the reaction by-products such as P₂O₃ and/or H₃PO₃, which were dissolved in the washing process to increase S_BET and V_p. ⁹

CO uptake

The CO molecules are mainly adsorbed at Ni-active sites; therefore, the CO uptake measurements were applied to “titrate” the number of Ni-active sites exposed on the catalyst surface and their dispersion on the catalysts.^10,11 The CO uptakes of the catalysts are listed in Table 1. The CO uptakes of Ni₂P/Al₂O₃(L) and Ni₂P/Al₂O₃(T) were both 27 μmol g⁻¹, which indicated that the two catalysts prepared from heat treatment and the TPR method had exposed similar Ni sites. Also, the CO uptakes of the two catalysts had increased significantly after washing. The CO uptakes of Ni₂P/Al₂O₃(L-W) and Ni₂P/Al₂O₃(T-W) were 34 and 32 μmol g⁻¹, respectively. This result may be caused by excluding excess phosphorus atoms as well as increasing S_BET after washing.

XPS

The XPS spectra of the Ni (2p) and the P (2p) regions of the washed catalysts are shown in Figure 2. The surface compositions and the corresponding binding energies are reported in Table 2. In all samples, the spectra showed Ni^δ+ (0 < δ < 2) and P^δ− (0 < δ < 1) species associated with the Ni₂P phase. The binding energies of the Ni 2p (856.5 eV) and P 2p (134.7 eV) belong to Ni²⁺ and P⁵⁺ species, respectively. The peaks at 852.6 and 129.2 eV are attributed to Ni^δ+ (0 < δ < 2) and P^δ− (0 < δ < 1) species, respectively. ¹² We found that the peak intensity of the Ni^δ+ and P^δ− for Ni₂P/Al₂O₃(L-W) was higher than that for Ni₂P/Al₂O₃(T-W), which suggested that the exposure of the Ni₂P phase on Ni₂P/Al₂O₃(L-W) was higher than that for Ni₂P/Al₂O₃(T-W). This result was in agreement with the CO uptake discussed in the “CO uptake” section.

Figure 2.

XPS spectra of the catalysts.

Table 2.

Spectral parameters obtained by XPS analysis.

Sample	Binding energy (eV)					Superficial atomic ratio P/Ni
	Ni 2p3/2			P 2p
	Niδ+	Ni2+	Satellite	P5+	Pδ–
Ni2P/Al2O3(L-W)	852.6	856.5	862.0	134.7	129.2	1.59
Ni2P/Al2O3(T-W)	852.6	856.5	862.0	134.7	129.2	1.97

XPS: X-ray photoelectron spectroscopy.

The surface P/Ni atomic ratios calculated from XPS analyses are shown in column 7 on Table 2. The theoretical P/Ni ratio was 1/2 for the two catalysts. But the catalysts exhibited higher P/Ni than 1/2, which was caused by the enrichment of phosphorous on the surface of the catalysts. The superficial P/Ni ratio of Ni₂P/Al₂O₃(L-W) catalyst was 1.59, which was lower than that of Ni₂P/Al₂O₃(T-W). This result may be caused by the release of PH₃ during the heat treatment process. ¹³ This was in accordance with the CO uptake results and peak intensities.

HDS activity

The catalytic activity and stability test of the catalysts were evaluated with the HDS of DBT, and the results are given in Figure 3. The DBT conversion over the Ni₂P/Al₂O₃(L) and Ni₂P/Al₂O₃(T) were 83.9% and 81.4% after the stream became stable, respectively. The activities increased after washing; those for Ni₂P/Al₂O₃(L-W) and Ni₂P/Al₂O₃(T-W) were 94.5% and 86.1%, respectively. This result may be caused by the better dispersion and smaller size of the Ni₂P phase; on the other hand, the higher surface area and lower surface P-rich were also beneficial to the catalytic activity.

Figure 3.

HDS activity of the catalysts.

The HDS selectivity over the catalysts is shown in Figure 4. The products detected were biphenyl (BP) and cyclohexylbenzene (CHB). The selectivity for BP was much higher than that for CHB. According to the literature, ¹⁴ the BP is produced from DBT desulfurization (DDS) pathway at Ni(I) sites, and CHB is converted from DBT hydrogenation (HYD) pathway at Ni(II) sites. Therefore, the Ni sites on the prepared catalysts were mainly Ni(I) sites. The heat treatment method in an N₂ atmosphere could lead to many more Ni(I) sites than the TPR method. These activity reaction results have great significance to the design and modification of Ni₂P catalysts.

Figure 4.

HDS selectivity of the catalysts.

Conclusion

Ni₂P/Al₂O₃(L) catalyst with high HDS activity has been prepared in an N₂ atmosphere by decomposition of NaH₂PO₂ with NiCl₂·6H₂O. XRD analysis has shown that the Ni₂P/Al₂O₃(L) catalyst has developed a pure Ni₂P phase. The crystallite size of Ni₂P/Al₂O₃(L) was smaller than that of Ni₂P/Al₂O₃(T) catalyst according to Scherrer’s equation. BET results showed that Ni₂P/Al₂O₃(L) exhibited a larger surface area than Ni₂P/Al₂O₃(T) catalyst. Also the surface area shows an obvious increase after washing with deionized water. XPS results indicated that the enrichment of P on the surface decreases on Ni₂P/Al₂O₃(L-W) catalyst, which exposed more nickel atoms, leading to a higher CO uptake. The highest DBT HDS activity was 94.5%, as achieved over Ni₂P/Al₂O₃(L-W) at a condition of 320°C, 4.0 MPa, H₂/oil = 500 (V/V), and WHSV = 8.0 h⁻¹.