Influence of Pluronic^® P123 Addition in the Synthesis of Bulk Ni Promoted MoS₂ Catalyst. Application to the Selective Hydrodesulfurization of Sulfur Model Molecules Representative of FCC Gasoline

Abstract

A way to improve hydrotreatment processes is to enhance the intrinsic activity of Ni or Co promoted MoS₂ catalysts that are commonly used in such reactions. The aim of this work was to investigate the impact of the presence of Pluronic^® P123 as a structuring agent during the synthesis of Ni promoted MoS₂ catalysts (named NiMoS) in water at room temperature. A series of analyses, i.e., X-ray diffraction (XRD), chemical analysis, inductively coupled plasma mass spectrometry (ICP-MS), nitrogen adsorption-desorption isotherms, transmission electron microscopy (TEM) and X-ray photoelectron spectroscopy (XPS), helped in characterizing the NiMoS-P123 and NiMoS catalysts, the latter being prepared in the absence of polymer. Both compounds contained MoS₂ phase (~85 atomic% considering Mo atoms), a similar amount of mixed Ni-Mo-S phase (40–50% considering Ni) and some amount of NiS and Ni-oxidized impurity phases. The main differences between the two catalysts were a much larger specific surface area (126 m²·g⁻¹ instead of 31 m²·g⁻¹) and a better dispersion of the active phase as shown by the lower slab stacking (2.7 instead of 4.8) for NiMoS-P123, and the presence of C in NiMoS-P123 (9.4 wt.% instead of 0.6 wt.%), indicating an incomplete decomposition of the polymer during thermal treatment. Thanks to its larger specific surface area and lower slab stacking and therefore modification of active Mo site properties, the compound prepared in the presence of Pluronic^® P123 exhibits a strong increase of the catalytic activity expressed per Mo atom for the transformation of 3-methylthiophene. Such improvement in catalytic activity was not observed for the transformation of benzothiophene likely due to poisonous residual carbon which results from the presence of Pluronic^® P123 during the synthesis.

1. Introduction

Worldwide growth of the global mobility of people and merchandises, along with industrial activity, have resulted in a severe increase in airborne pollutants concentration in the last decades. To counterbalance this, the production of cleaner fuels used for on-road and off-road transportations has been imposed to refiners. In particular, the reduction of sulfur content in gasoline and diesel cuts through hydrodesulfurization (HDS) processes have become essential to meet worldwide regulations [1,2]. This has consequently stimulated research and innovations in hydrotreatment processes and heterogeneous catalysts.

Conventional hydrotreating catalysts are based on Group 6 transition metal sulfides (MoS₂) usually promoted by Group 9 and 10 transition metals such as nickel or cobalt [3]. They are traditionally prepared by incipient wetness impregnation of supports such as alumina [3], zeolites [4], or graphitic carbons [5], by an aqueous solution of metal precursors and then activation by a sulfidation step. However, this preparation method often leads to a poor dispersion of the metal and limited metal loadings in contradiction with requirements for higher activity in hydrodesulfurization reactions. In order to improve the intrinsic activities of supported sulfide active phases, strategies based on a better control of the physicochemical properties—such as the morphology, size, and stacking of the MoS₂ slabs, electronic properties, or surface composition have been extensively described in literature [6].

A different approach for the development of highly active catalysts comes with the preparation of bulk hydrotreating Ni or Co promoted MoS₂ catalysts (named NiMoS or CoMoS) in order to maximize their activity per volume of reactor. This synthesis usually proceeds in two steps, i.e., the synthesis of mixed NiMo oxide phases by co-precipitation followed by a sulfidation step [7,8,9,10]. Alternatively, the direct synthesis of sulfide phases with thio precursors (i.e., without the sulfidation step of metal oxide intermediates) has been also described in the literature [11,12,13]. Noteworthy, in both cases, the obtained catalysts suffer from low specific surface areas. In order to get larger specific surface areas, the structuration of bulk MoS₂ catalysts with organic compounds such as surfactants, polymers, or ionic liquids can be considered [6]. For instance, several studies used polymers such as poly(vinyl pyrrolidone) (PVP) or poly(ethylene glycol) (PEG), which led to a decrease of the MoS₂ stacking [14,15] and to an increase of the specific surface area [16]. Under hydrothermal conditions, the MoS₂ synthesis templated by surfactants or polymers may also lead to different morphologies such as nano or microspheres, nanoflowers, or hexagonal particles [17,18,19,20]. Interestingly, the addition of organic compounds during the synthesis of Ni (or Co)-promoted MoS₂ catalysts has been less studied. Some papers report the synthesis, usually in hydrothermal conditions, of molybdenum oxide phases promoted by Ni, in the presence of polymers such as PVP [21,22] or PEG [23] but a further sulfidation step was required. In these studies, the presence of nickel sulfide phases has been observed in addition to the mixed Ni-Mo-S phase. Furthermore, although the oxide materials could present a specific surface area up to 144 m²·g⁻¹, the sulfide catalysts displayed much lower surface areas, from 8 to 20 m²·g⁻¹. Actually, the direct synthesis of the sulfide phase from sulfur precursors in the presence of organic structuring agents has been scarcely reported. Genuit et al. [24] reported the preparation of efficient bulk Co(Ni)MoS catalysts obtained with nonionic surfactants (poly(ethylene oxide) oligomers with aryl-alkyl tail), in a mixed water/ethylene glycol solution, leading to a specific surface area of 122 m²·g⁻¹. For the hydrothermal synthesis of Ni-promoted MoS₂ in the presence of decalin, Yoosuk et al. [25] have shown that by increasing the temperature to 375 °C and the pressure up to 2.8 MPa, a very high specific surface area up to 250 m²·g⁻¹ could be obtained. Recently, we have reported a first series of data on the preparation of Ni-promoted MoS₂ catalysts in the presence of an ionic liquid and their physicochemical characterization has shown different morphologies and an increase of the specific surface area depending on the amount of ionic liquid [26].

The present paper focuses on the consequences of the presence of a polymeric surfactant on the catalytic activity of a bulk NiMoS catalyst prepared by “soft chemistry” route in water. To do so, a commercial and cheap polymer, Pluronic^® P123, was used to act as a steric stabilizer. This amphiphilic water-soluble poly(ethylene oxide)-b-poly(propylene oxide)-b-poly(ethylene oxide) triblock copolymer (HO(CH₂CH₂O)₂₀(CH₂CH(CH₃)O)₇₀(CH₂CH₂O)₂₀H), able to self-assemble in aqueous media, will interact with the water-insoluble MoS_x formed during the reaction. Physicochemical characterizations based on nitrogen adsorption–desorption isotherms, X-ray diffraction, X-ray photoelectron spectroscopy, and transmission electron microscopy were carried out. Finally, the catalytic performances of bulk NiMoS were investigated for the transformation of two sulfur model molecules—3-methylthiophene and benzothiophene—representative of Fluid Catalytic Cracking (FCC) gasoline.

2. Results and Discussion

2.1. Catalysts Characterizations

The synthesis described in this paper is a very easy and very fast reaction performed in water at ambient temperature and pressure. The reaction directly leads to the sulfide catalysts. Indeed, the reaction between nickel nitrate and ammonium thiomolybdate is instantaneous and leads to a colloidal unstable black suspension according to the following suggested reaction [24]: (NH₄)₂MoS₄ + Ni(NO₃)₂.6H₂0 → NiMoS₄ + 2NH₄NO₃ + 6H₂O. The presence of Pluronic^® P123 above its critical micellar concentration does not alter the kinetic of the reaction and a sterically stabilized black suspension is obtained as soon as the two precursor solutions are mixed.

Chemical analyzes of the materials before and after thermal treatment at 320 °C (under a reductive atmosphere of argon and 5% H₂) have been performed and are reported in

Table 1. Impact of the presence of P123 on bulk NiMoS characteristics (specific surface area, elemental analysis (wt.%)) before and after thermal treatment at 320 °C.

Sample	SBET (m2·g)	C*	H*	S*	Ni**	Mo**	Ni/Mo**
NiMoS a		0.6	0.1	34.8
NiMoS b	31	0.6	0.1	34.6	17.6	34.6	0.5
NiMoS-P123 a		26.9	5.1	21.5
NiMoS-P123 b	126	9.4	1.1	30.9	12.9	27.6	0.5

* Determined by elemental analysis, ** Determined by inductively coupled plasma mass spectrometry ICP-MS, (a) before thermal treatment at 320 °C, (b) after thermal treatment at 320 °C.

. The amount of sulfur after thermal treatment varies from 35 wt.% for NiMoS to 31 wt.% for NiMoS-P123. The amount of nickel determined by inductively coupled plasma mass spectrometry (ICP) is higher for NiMoS (17.6 wt.%) than for NiMoS-P123 (12.9 wt.%). However, the Ni/Mo ratio is similar (≈0.5). As expected, only traces of C or H residue are visible in the absence of polymer. In contrast, when Pluronic^® P123 is added, the carbon and hydrogen content are equal to 26.9 and 5.1 wt.%, respectively, before the thermal treatment. This corresponds to a H/C weight ratio of 0.19 very close to the theoretical value of the Pluronic^® P123 (theoretical H/C weight ratio of 0.17). After thermal treatment at 320 °C, the carbon and hydrogen contents decrease to 9.4 and 1.1 wt.%, respectively, giving a H/C weight ratio of 0.11, which indicates that the organic residue does not correspond to Pluronic^® P123 anymore. The remaining of a carbonaceous phase can be explained by an incomplete decomposition of the polymer trapped within the materials, during the heat treatment in a reductive H₂(5%)/Ar atmosphere. In an attempt to remove all carbonaceous residue, heat treatment at 500 °C has been performed. The elemental analysis showed similar result with 9.5 wt.% of C. It must be noticed that all the synthesis of NiMoS in the presence of organic compounds (solvent, surfactant, or polymer) described in the literature also reported the presence of similar amounts of carbon after thermal treatment [17,27,28,29].

The nitrogen adsorption–desorption curves of the NiMoS and NiMoS-P123 materials obtained after thermal treatment at 320 °C are represented in Figure 1. The isotherm featured for the NiMoS catalyst prepared without polymer is the signature of inter-particle capillary condensation (type II according IUPAC classification). On the other hand, the isotherm obtained for the NiMoS-P123 catalyst prepared with Pluronic^® P123 corresponds to type IV isotherm characteristic of mesoporous materials [30]. Furthermore, the hysteresis loop can be attributed either to pore-blocking/percolation in a narrow range of pore necks or to cavitation-induced evaporation [31]. The pore size distribution of the NiMoS-P123 catalyst (Figure 2) obtained by the BJH method from the desorption branch shows a narrow peak centered at 3.8 nm with a pore volume of 0.16 cm³·g⁻¹. Considering the t-plot from Harkins and Jura law, it is possible to assert that no microporosity is evidenced in NiMoS-P123 catalysts but very few microporosities corresponding to 4 m²·g⁻¹ should not be excluded in NiMoS. The specific surface areas of the catalysts have been determined by the BET method and are reported in Table 1. The addition of Pluronic^® P123 generated specific surface area of 126 m²·g⁻¹, four times higher than the surface of the NiMoS catalyst prepared without polymer (31 m²·g⁻¹). Several authors have reported the large increase of the specific surface area of MoS₂ by using polymer such as polyvinyl pyrrolidone (92 m²·g⁻¹) [16] or organic compound such as propylene carbonate (between 140 and 190 m²·g⁻¹) [17]. Most of the studies on the synthesis of Ni-promoted MoS₂ catalysts concern the synthesis of oxides which are further sulfided just before the catalytic evaluations. In those cases, the specific surface area of the oxide phases could be increased by the use of PVP up to 144 m²·g⁻¹ as reported by Liu [22]. But very small specific surface areas were obtained for the final NiMoS catalysts: less than 20 m²·g⁻¹. To our knowledge, Genuit et al. [24] are the only ones who reported a specific surface area of 122 m²·g⁻¹ for NiMoS catalyst prepared in one step and at ambient pressure without further sulfidation. This NiMoS catalyst has been obtained in the presence of Tergitol^® surfactant in a mixture of water and ethylene glycol. The mechanism responsible of the increase of the surface area is not clearly detailed in the literature but the role of the organic phase seems to be the steric stabilization of the nuclei to prevent the particle growth. In this study, the addition of an amphiphilic polymer not only leads to a higher specific surface area but also to the formation of a mesoporosity. This may be due to an increase of the disorder of the MoS₂ slabs by the polymer. Furthermore, the remaining carbonaceous species due to an incomplete removal of the polymer during the washing and the thermal treatment steps may contribute to the formation of a mesoporous composite MoS₂-C phase.

X-ray diffraction analyses have been performed on the NiMoS materials obtained after thermal treatment at 320 °C and the spectra are reported in Figure 3. The catalysts exhibited broad diffraction peaks, indicating a very poorly crystallized MoS₂ structure, particularly when the polymer is added (NiMoS-P123). Without polymer, a second phase corresponding to nickel sulfide NiS is also observed. The formation of nickel sulfide phase has already been observed in several synthesis of NiMoS catalysts [21,26,32]. This may be due to the reaction between the sulfur expelled during the thermal treatment and some Ni that is not used to form the mixed Ni-Mo-S phase. Yoosuk [25] claimed that the presence of a small amount of secondary phases Ni₃S₂ and NiS, for atomic ratio of Ni/(Ni+Mo) lower than 0.5, had no apparent detrimental effect on the conversion of dibenzothiophene (DBT) and 4,6-dimethyldibenzothiophene (4,6-DMDBT). This has been confirmed by Liu [21].

It is also very interesting to notice that the presence of an organic phase such as a polymer in our case, or a surfactant in the case of Xu, [28] led to a decrease or to the absence of the X-ray diffraction peaks attributed to the nickel sulfide phase. The disorder due to the organic phase may thus increase the amount of the Ni-Mo-S phase and/or lead to Ni_xS_y nano-crystallites that are too small to be characterized by X-ray diffraction method. To address this question and gain further insights in their chemical composition, the synthesized materials were analyzed by X-ray photoelectron spectroscopy (XPS). According to the chemical analysis presented in Table 1 and to the Mo 3d, Ni 2p, and S 2p XPS spectra decomposition shown in Figure 4, Figure 5, Figure 6 and Figure 7, the co-existence of MoS₂, NiMoS, NiS, and Ni_ox phases was confirmed. Indeed, the various MoS₂ (Mo 3d_5/2 BE = 229.3 eV; S 2p_3/2 BE = 162.2 eV), NiS (Ni 2p_3/2 BE = 853.2 eV), Ni_ox (Ni 2p_3/2 BE = 856.1 eV), and mixed phase Ni-Mo-S (Ni 2p_3/2 BE = 854.3 eV) species were well identified. Quantitative analysis of XPS data helped in evaluating the S/Mo and Ni/(Ni+Mo) atomic ratios as well as the sulfidation rate of molybdenum (proportion of MoS₂ (%)) and the amount of Ni-Mo-S mixed phase (

Table 2. Quantitative XPS analysis of the bulk catalysts after thermal treatment at 320 °C and sulfidation at 400 °C.

			Mo Fraction (rel.%)			Ni Fraction (rel.%)
Catalyst	S/Mo	Ni/(Ni+Mo)	Mo5+	Mo6+	MoS2	NiMoS	NiS	Niox	(Ni/Mo)slab*
NiMoS	2.5	0.37	9	5	86	48	35	17	0.56
NiMoS-P123	2.3	0.28	11	5	84	38	21	41	0.45

* promoter ratio in the active phase slab determined using (Ni/Mo)_slab = $\frac{C_{NiMoS}}{C_{Mo{S}_2}}$ where C_X is the absolute concentration of Ni(Mo) in the NiMoS(MoS₂) species (atomic%).

). The MoS₂ phase (S/Mo ratio between 2.3 and 2.5) was indeed formed in a large amount (around 85%), similar for the two catalysts. The Ni/(Ni+Mo) atomic ratios values ranging between 0.3 and 0.4 are closed to the one of the conventional NiMo/Al₂O₃ catalyst used in hydrotreatment processes. The mixed NiMoS phase was formed in a large amount (around 40–50 %). On the whole, the synthesis method proposed in this work led to an amount of Ni-Mo-S mixed phase in the material similar to the one observed for conventional supported catalyst (45%) [13]. While no oxidation of sulfur and subsequent formation of sulfates were noticed in the XPS spectra, the relative amount of Ni oxidized impurity phases, Ni_ox (41%) in the NiMoS-P123 sample was significantly higher than that of the reference NiMoS sample (17%). The increase of the Ni_ox phase for NiMoS-P123 may be explained by the release of oxygen due to the degradation of the polymer during the thermal treatment at 320 °C. In the case of NiMoS-P123, this increase in the Ni_ox phase was essentially counterbalanced by a decrease in the proportion of the Ni_xS_y phases. This also confirmed the works of Xu and al. [28] suggesting the formation of small Ni_xS_y crystals non detectable by X-ray diffraction.

Notably, the peak corresponding to the MoS₂ (002) plan at 2θ ≈ 14° in the X-ray diffraction patterns of NiMoS samples decreased drastically with the addition of Pluronic^® P123 (Figure 3). The height of the peak being linked to the number of stacked slabs, this decrease is attributed to a decrease in the stacking of the MoS₂ slabs for the NiMoS-P123 [33]. The size of the particles can be estimated using the Scherrer formula. Owing to the small peak in the case of NiMoS-P123, the particles diameter size has only been estimated for NiMoS and was found to be around 3 nm (

Table 3. MoS₂ particles size determined by XRD, stacking of the MoS₂ layers determined by XRD and TEM images, average length of the layers determined by TEM and average ratio edge-to-corner of MoS₂ slabs.

	XRD		TEM
Catalyst	S a (nm)	St b	St b	L c (nm)	Lmin (nm)	Lmax (nm)	(fe/fc)Mo
NiMoS	3	6	4.8	3.8	2.2	9.3	4.4
NiMoS-P123	-	-	2.7	3.3	1.1	10.2	3.7

^a MoS₂ particle size (S), ^b average stacking (St), ^c average length (L), ratio edge-to-corner of a MoS₂ slab ((f_e/f_c)_Mo).

). The position of the (002) peak, 2θ ≈ 14°, corresponds to an interspace layer of 6.3 Å and thus to a stacking of 5–6. The decrease of the crystallinity of the MoS₂ phase and the low stacking due to the addition of polymer such as PEG or PVP have been depicted by several authors [14,15,16,18] and was said to be consistent with an expansion of the interlayers by Wu [15]. However, in the present work, it is difficult to point the peak at 14° precisely and thus to attest to any displacement on the basis of X-ray diffraction patterns.

The interlayer of the NiMoS slabs was also measured using TEM images (Figure 8) and corresponds to 6.3 Å for NiMoS, which is consistent with the X-ray diffraction results and matches the (002) basal plane of crystalline MoS₂. Regarding the NiMoS-P123 catalyst, an interlayer of 6.6 Å was measured, which indicates a small expansion. The length and stacking of MoS₂ slabs were also estimated from TEM analysis (Figure 9 and Figure 10, respectively) while the average length and stacking are reported in Table 3. As expected from the XRD analyses, the stacking is low and the average length of the layer is short: 3.8 nm for NiMoS and 3.3 nm for NiMoS-P123. The addition of Pluronic^® P123 during the synthesis led to a decrease of the average stacking from 4.8 without polymer down to 2.7 layers in the presence of polymer. Furthermore, the distribution of the stacking differs between the two samples. Indeed, for NiMoS-P123, 77% of the observed MoS₂ particles have 2 or 3 layers and almost 10% have monolayer. But for NiMoS, the distribution is centered at 4–5 layers without any monolayers. In the literature, most works report a stacking higher than 5 for the syntheses of bulk catalysts carried out without any structuring agent and a decrease of the MoS₂ stacking or even no stacking for the syntheses with the addition of an organic compound; however, most of these studies dealt with the synthesis of MoS₂ and not NiMoS [15,29]. The ratio edge-to-corner of a MoS₂ slab ((f_e/f_c)_Mo) has been calculated (Table 3) from TEM images according to Nikulshin [34]. This ratio is slightly lower for NiMoS-P123 (3.7) than for NiMoS (4.4), in agreement with the decrease of the average slab length.

2.2. Catalytic Performances in the Transformation of Model Molecules Representative of FCC Gasoline

Both NiMoS-P123 and NiMoS catalysts were evaluated in the hydrodesulfurization of 3-methylthiophene (3MT) and benzothiophene (BT).

Data related to the transformation of 3MT are shown in Figure 11 and Figure 12, and in

Table 4. Transformation of sulfur model compounds (3MT, BT). Activity per gram, per square meter and per Mo atom (T = 250 °C; P = 20 bar; H₂/HC = 360 NL/L, isoconversion: around 30%).

	3MT			BT
	mmol·gcata−1 ·h−1	10−2 mmol ·m²·h−1	10−21 mmol·atMo−1·h−1	mmol·gcata−1·h−1	10−2 mmol·m²·h−1	10−21 mmol·atMo−1·h−1
NiMoS	2.2	7.1	1.2	4.2	13.6	2.2
NiMoS-P123	3.1	2.5	2.2	3.7	2.9	2.6

and

Table 5. Transformation of 3MT. Comparison of selectivities (T = 250 °C; P = 20 bar; H₂/HC = 360 NL/L, isoconversion: around 30%) (2MB2N: 2-methylbut-2-ene, 2MB1N: 2-methylbut-1-ene, 3MB1N: 3-methylbut-1-ene, 3MTHT: 3-methyltetrahydrothiophene, i-C5/=C5 = isopentane/(2MB2N + 2MB1N + 3MB1N)).

Catalyst		iC5 %	3MB1N %	2MB1N %	2MB2N %	iC5/=C5 %
NiMoS		11.7	8.4	18.4	61.5	0.13
NiMoS-P123		13.2	5.7	18.2	62.9	0.15

Hydrodesulfurization (HDS): iC5 + 3MB1N + 2MB1N + 2MB2N.

. The NiMoS-P123 sample synthesized in the presence of P123 exhibited an increase of around 30% in the catalytic activity per gram for the transformation of 3MT although it corresponds to a decrease of activity when taking into account its specific surface area (Table 4). This indicated that the improvement was not the only consequence of the increase in specific surface area but also to a modification of the properties of the molybdenum active sites. As a matter of fact, the intrinsic activity (per Mo atom) is doubled when the polymer is present during the synthesis. For the two catalysts, the conversion increases with the contact time (Figure 11) and for a same contact time (0.25 s), the conversion is higher for NiMoS-P123. For the whole range of conversion, the product distribution is similar for both catalysts (Figure 12). The olefins are the main products with a majority of 2MB2N (2-methybut-2-ene) and small amounts of 2MB1N and 3MB1N. When the conversion increases, the amount of isopentane (i-C5)—the final hydrogenation product—increases and becomes the main product for conversion close to 100%. This mainly corresponds to a decrease in the formation of 2MB2N. These results are in agreement with the well-established reaction scheme where the HDS of 3MT involved consecutive reactions with olefins as intermediates [35]. No modification of selectivities is noticed between the two materials (Table 5).

Data related to the transformation of BT are shown in Figure 13 and Figure 14, and in Table 4 and

Table 6. Transformation of BT. Comparison of selectivities (T = 250 °C; P = 20 bar; H₂/HC = 360NL/L, isoconversion: around 30%) (DHBT: dihydrobenzothiophene, EB: ethylbenzene).

Catalyst	DHBT %	EB %
NiMoS	3.9	96.1
NiMoS-P123	6.8	93.2

. Conversely to data obtained for the transformation of 3MT, the NiMoS catalyst has higher activities per gram and per square meter than the sample prepared with P123 in the transformation of BT (Figure 13, Table 4). However, the activity per molybdenum atom remained similar. As shown in Figure 14 for NiMoS catalyst, two products are formed, according to the reaction scheme previously established [35]: dihydrobenzothiophene (DHBT) corresponding to the partial hydrogenation of BT and ethylbenzene (EB) as the HDS product, this one being the main product.

When comparing the two catalysts (Table 4), it appears that BT is clearly more reactive than 3MT over the NiMoS catalyst, and has a similar reactivity to 3MT over NiMoS-P123 catalyst. The first result agrees with data reported in the literature and dealing with the reactivity of the two sulfur molecules in presence of CoMoS/Al₂O₃ [35]. It has been explained by a stronger adsorption of BT compared to that of 3MT. The second data is more surprising. Some differences in structure and/or morphology of the two catalysts due to the presence or not of polymer during the synthesis must be at the bottom of the understanding.

If taking into account the uncertainty on the data (20% for XPS), both chemical and XPS analysis show that the Ni/Mo ratio and quantity of active Ni-Mo-S phase are similar in the two catalysts (Table 1 and Table 2). However, some differences between the two compounds are observed. Upon the addition of Pluronic^® P123, the specific surface area of the compound increased from 31 to 126 m²·g⁻¹, the average slab stacking decreased from 4.8 to 2.7 and finally, a large amount of C remained trapped in the compound prepared in presence of polymer, even after an heat treatment at 320 °C (9.4 wt.% for NiMoS-P123 compared to 0.6 wt.% for NiMoS) (Table 1). The two tested sulfur molecules differ in their adsorption properties and in the way their transformation preferentially proceeds.

Indeed, as reported previously for the transformation of 2MT [36], 3MT involves mainly C-S bond ruptures and can be adsorbed and reacts in all active sites located on the edges (sulfur and metallic) of MoS₂ slabs. The main way involved in BT transformation is the hydrogenation way to form dihydrobenzothiophene as an intermediate, followed by C–S bond breaking to produce ethylbenzene. In this case, a flat adsorption of BT could be involved on metallic edges or on rim sites in agreement with the rim-edge model [37]. It has already been reported in the literature that the presence of carbonaceous phase in a catalyst could decrease the accessibility of some active sites, especially the ones involved in the hydrogenation way [36,38,39]. Therefore, the presence of carbon due to an incomplete decomposition of Pluronic^® P123 is particularly detrimental in the case of BT. It is a poison of the most favorable sites for its transformation and might also be at the origin of steric hindrance for these large molecules.

In the case of 3MT, the decrease in active hydrogenating sites might be counterbalanced by an increase in the number of accessible active sites due to the much larger specific surface area and decrease in average slab packing. Its smaller size compared to BT might also less suffer from steric hindrance due to carbon.

On the whole, the presence of P123 during the synthesis results in a markedly increase of the transformation of 3MT (a 30% increase for the activity per gram and a doubled intrinsic activity (per Mo atom)). Whereas, the catalytic activities per gram and per Mo atom for the transformation of BT change only by ±15%, whether the polymer was present or not during the synthesis. Such results are interesting in the perspective of an industrial use where FCC feed comprises several sulfur molecules. As a matter of fact, it was shown recently [35] in a model feed representative of an FCC feed, that the transformation of 3MT was inhibited by the presence of BT. In the case of NiMoS-P123, its improved activity for 3MT might therefore allow an easier sulfur removal when this catalyst is used. It should then be easier to reach the level of sulfur requested by regulations (10 ppm of sulfur).

3. Materials and Methods

3.1. Catalyst Preparation

The triblock copolymer Pluronic^® P123 (EO₂₀-PO₇₀-EO₂₀; EO: ethylene oxide; PO: propylene oxide), nickel nitrate Ni(NO₃)₂·6H₂O (99.99%), ammonium tetrathiomolybdate (NH₄)₂MoS₄ (99.97 %), and ethanol were purchased from Sigma Aldrich (Sigma Aldrich, St. Louis, MO, USA). All products were used without further purification. H₂ (5%)/Ar (99.95%) was provided by Linde gas (Linde gas, Munich, Germany).

NiMoS catalysts were prepared as follows: (NH₄)₂MoS₄ (4 mmol) and Ni(NO₃)₂·6H₂O (4 mmol) were separately dissolved in 40 mL of pure H₂O or in 40 mL of an aqueous solution of Pluronic^®P123 (0.2 mmol). The two solutions were then mixed with gentle magnetic stirring for 30 s, at room temperature and under an air atmosphere. The reaction mixture was then stored for 12 h at room temperature and the particles were separated by centrifugation (40,000 rpm, 4 h) using a Beckman Coulter Optima XPN ultracentrifuge with a 45 TI rotor and 70 mL PC tubes. Three washings with 80 mL of water were performed (centrifugation at 40,000 rpm, 4 h), followed by a final washing step with 80 mL of ethanol. The particles were finally recovered by centrifugation (20,000 rpm, 4 h) in 94 mL PP tubes. All the products were then dried under primary vacuum for 12 h at T = 25 °C. Finally, the powder was heated under an atmosphere of 5% hydrogen in argon from room temperature to 320 °C at 5 °C.min⁻¹, held 2 h at 320 °C, and cooled down to room temperature at the same rate under argon. This temperature has been chosen as a compromise: At a lower temperature, the decomposition of Pluronic^® P123 would decrease (potentially increasing the carbon residue), whereas at a higher temperature, the number of MoS₂ layers would increase. The catalysts were then stored under vacuum before use.

3.2. NiMoS Characterizations

The specific surface areas were determined by nitrogen adsorption–desorption using a Micromeritics tristar apparatus after a night of degassing at 120 °C (Micromeritics, Norcross, GA, USA). Elemental analysis (sulfur, hydrogen, nitrogen, and carbon amounts) was carried out using an Elementar Vario Micro Cube (Elementar, Langenselbold, Germany) while the molybdenum and nickel amounts were measured by inductively coupled plasma mass spectrometer (ICP-MS) using a SPECTRO ARCOS ICP-AOS analyzer (SPECTRO Analytical Instruments, Kleve, Germany). X-ray powder diffraction (XRD) patterns were recorded with a Malvern Panalytical X’pert PRO X-ray diffractometer by using Cu radiation (λ = 1.5406 Å) (Malvern Panalytical, Royston, U.K.). A JEOL JEM2100F (JEOL, Tokyo, Japan) transmission electron microscope (TEM) was used to characterize the MoS₂ phase (length and stacking).

XPS spectra were recorded using a KRATOS AXIS ultra spectrometer (Kratos Analytical, Manchester, U.K.) equipped with a (150 W) Al Kα monochromatic source (hν = 1486.6 eV). Before analysis, catalysts were sulfided at 400 °C during 10 h by a mixing of H₂S (10 molar %)/H₂ at atmospheric pressure and stored in Schlenk under argon to avoid oxidation resulting in the formation of sulfates at the surface. They were identified with reference samples drawn from the handbook of X-ray photoelectron spectroscopy [40], NIST X-ray Photoelectron Spectroscopy Database (NIST Standard Reference Database 20, Web Version 3.4, Gaithersburg, MD, USA). The calibration has been made with the carbon peak of contamination at 284.6 eV. For each catalyst, the metal and sulfur peaks were identified according to their binding energies [41,42]. The elemental surface composition of the catalysts, and therefore, the sulfur/metal atomic ratio (S/Me) and the active phase formation were determined from the area of the metal and sulfur peaks (the uncertainty of the value is around 20%).

3.3. Catalytic Performances

As previously reported [35], catalytic activity measurements were carried out in a fixed bed reactor at 523 K under a total pressure of 2 MPa with a ratio H₂/feed of 360 NL/L. First, the catalyst was sulfided in situ under H₂S/H₂ flow (10 mol% H₂S) for 10 h at 673 K at atmospheric pressure. These operating conditions correspond to those of the industrial process [36].

Catalytic performances of the NiMoS bulk catalysts were measured for the transformation of a single component feed including a sulfur model molecule alone (corresponding to 1000 ppm S) (0.3 wt.% of 3MT or 0.42 wt.% of BT) in n-heptane. The sulfur model feeds were injected in the reactor by a HPLC Gilson pump (307 series, pump’s head: 5 cm³). 3-methylthiophene (98% purity), benzothiophene (95% purity), and n-heptane (>99% purity) were purchased from Sigma-Aldrich (Sigma Aldrich, St. Louis, MO, USA). They were used without further purification. Hydrogen sulfide (10 mol% in mixture with H₂) and hydrogen was purchased from Air Liquide (Air Liquide, Paris, France).

The different partial pressures of the reactants, H₂S and H₂ introduced are reported in

Table 7. Partial pressures (kPa) of the different compounds for the sulfidation step and the transformation of the different feeds.

Pressure (kPa)	Sulfidation	3MT	BT
PH2S	10	0	0
P3MT	0	200	0
PBT	0	0	200
PH2	90	1360	1360
PnC7	0	637	637
PTOT	100	2000	2000

. n-heptane was not converted under these experimental conditions. No catalyst deactivation was observed for all the experiments.

The reaction products were injected on-line by means of an automatic sampling valve into a Varian gas chromatograph equipped with a PONA capillary column and a flame ionization detector as in previous works [35]. Desulfurized products, resulting from the transformation of 3-methylthiophene and benzothiophene are designated as HDS products. Regarding the transformation of the 3-methylthiophene, HDS products (mainly pentanes and pentenes) are the main observed products according to the reaction scheme described in the literature [36] (Scheme 1).

The transformation of BT led mainly to ethylbenzene as HDS product and also to tetrahydrobenzothiophene (THBT) corresponding to the partially hydrogenation of BT (Scheme 2).

The contact time is defined as the ratio between the total amount of feed and the mass of catalyst. 3MT and BT reactivities are defined as the number of moles of products formed per hour and per gram of catalyst and were calculated at conversion lower than 30% in a differential regime.

4. Conclusions

The influence of addition of polymer Pluronic^® P123 during the synthesis of bulk NiMoS catalyst on the structure, morphology, and catalytic properties has been investigated. It has been shown that, while no change in the Ni/Mo ratio and amount of Ni-Mo-S active phase could be observed, the compound prepared in the presence of polymer has a much larger specific surface area, a lower average slab stacking, and a higher amount of C than the catalyst prepared in the absence of polymer.

These differences affected the catalytic properties of the two tested sulfur molecules, 3MT, and BT, in different ways. The presence of carbon in NiMoS-P123 was detrimental to the catalytic activity for BT transformation, probably due to some steric effect and the poisoning of active sites favorable to hydrogenation, the main transformation way for BT. In the case of 3MT which can be adsorbed on all active sites, the negative effect of residual carbon due to an incomplete decomposition of Pluronic^® P123 is probably counterbalanced by an increase in the number of accessible active sites due to the large specific surface area and low average slab stacking in NiMoS-P123. As a result, the catalytic activity, either per gram or Mo atoms of NiMoS-P123 is higher for the 3MT transformation than the activity of NiMoS.