Exploring Perhydro-Benzyltoluene Dehydrogenation Using Sulfur-Doped PtMo/Al₂O₃ Catalysts

Abstract

This study investigates the dehydrogenation of perhydrobenzyltoluene, a Liquid Organic Hydrogen Carrier (LOHC), using sulfur-doped bimetallic PtMo/Al₂O₃ catalysts. Based on previous research that highlighted the superior performance of PtMo catalysts over monometallic Pt catalysts, this work focuses on minimizing byproduct formation, specifically methylfluorene, through sulfur doping. Catalysts with low platinum content (<0.3 wt.%) were synthesized using the wet impregnation method by varying sulfur concentrations to study their impact on catalytic activity. Characterization techniques, including CO–DRIFT and CO–TPD, revealed the role of sulfur in selectively blocking low-coordinated Pt sites, thus improving selectivity and maintaining high dispersion. Catalytic tests revealed that samples with ≥0.1 wt.% sulfur achieved up to a threefold reduction in methylfluorene formation compared to the unpromoted PtMo/Al₂O₃ sample, with a molar fraction below 2% at 240 min. In parallel, these samples reached a degree of dehydrogenation (DoD) above 85% within 240 min, demonstrating that improved selectivity can be achieved without compromising catalytic performance.

1. Introduction

Utilizing hydrogen for green energy storage represents a promising solution to address the fluctuations in energy production from renewable sources and mismatches in energy consumption. However, the transition to a CO₂-free economy centers on the creation of an efficient hydrogen distribution network [1,2,3]. Presently, hydrogen storage and transportation technologies face significant limitations. Conventional methods require high pressures and low temperatures due to hydrogen’s low energy density. Moreover, transporting hydrogen through existing gas network infrastructures presents challenges such as leakage, limiting the feasibility of a hydrogen-based energy system [4,5,6].

Over the past decade, various hydrogen storage technologies have been explored [7,8]. Among them, the chemical storage of hydrogen, particularly using Liquid Organic Hydrogen Carriers (LOHCs), has emerged as a promising option for long-distance and large-scale applications. LOHC technology involves reversible hydrogen chemical storage utilizing organic homocyclic and heterocyclic compounds [7,8,9]. During hydrogenation, these compounds undergo saturation of their organic C=C bonds, transforming hydrogen lean molecules into hydrogen-rich molecules. Subsequently, hydrogen can be released, restoring the original molecule’s physicochemical properties [10,11].

To date, numerous molecules have been investigated, with toluene [12,13], decalin [14], diphenylmethane [15,16], biphenyl [16,17,18], benzyltoluene [19,20,21], dibenzyltoluene [20,22,23,24,25,26], and N-ethylcarbazole [20,26,27] being major examples. Each class of molecules possesses its own advantages and disadvantages [28,29,30,31]. However, a crucial factor determining their practical application lies in the development of robust catalysts that facilitate hydrogen storage and release processes under mild conditions, while maintaining high reaction rates and minimizing byproduct formation. In this regard, the endothermic nature of the dehydrogenation reaction and the energy input required for this step represent the main challenges in catalyst design [32,33].

As previously mentioned, the dehydrogenation of LOHCs is a catalytic process. Among the different catalytic approaches, metal-based heterogeneous catalysts stand out over homogeneous systems due to their broader operational applicability, improved catalytic stability, and easier separation and recyclability [30,34]. Compared to homogeneous systems, they also enable more scalable and sustainable operation. Among the various metals and noble metals [35,36], platinum remains the most widely used, owing to its high activity and selectivity in C–H bond cleavage, and is widely reported in the literature as benchmark systems [37].

Furthermore, metal oxides are recognized for their thermal stability and their ability to promote dehydrogenation through hydrogen spillover [38,39,40,41]. Consequently, Pt/Al₂O₃ is commonly used as a catalyst in LOHC dehydrogenation processes [13,18,21,42,43]. However, the selectivity of Pt-based catalysts is greatly influenced by Pt particle size, support acidity, and metal–support interactions. Modisha and Bessarabov [23] reported cracking and cyclization reactions of aromatic molecules on the acidic sites of the support, while Auer et al. highlighted the impact of Pt particle size on low-coordinated sites, affecting byproduct formation [44]. Modisha et al. studied the behavior of Pt/Al₂O₃-coated foams in an unstirred tank reactor and a fixed-bed reactor, demonstrating the possible process intensification of LOHC dehydrogenation technology [45]. Given that LOHC technology relies on the repeated utilization of organic compounds through multiple dehydrogenation and hydrogenation cycles, the formation of byproducts directly affects the physicochemical properties of the carrier and compromises reusability and process efficiency [46,47]. To address this issue, Auer et al. demonstrated the efficacy of sulfur incorporation into Pt/Al₂O₃ catalysts, reducing byproduct formation while enhancing catalytic activity [48].

In this study, we examine perhydrobenzyltoluene dehydrogenation (H12-BT), focusing on byproduct formation during the reaction. Based on our previous work [49] in which the incorporation of molybdenum as a promoter in PtMo/Al₂O₃ catalysts significantly enhanced the dehydrogenation activity of benzyltoluene, achieving higher conversion and selectivity compared to a monometallic 1 wt.% Pt/Al₂O₃ sample prepared by the same method, we selected a Mo-promoted Pt catalyst as the base system for further optimization. The same study also revealed that catalysts with higher activity led to an increased formation of methylfluorene, an undesired byproduct generated from highly dehydrogenated benzyltoluene species (H0-BT). While previous studies on PtMo/Al₂O₃ catalysts have demonstrated high activity for benzyltoluene dehydrogenation, they often suffer from the formation of undesirable by-products such as methylfluorene. In this work, we propose sulfur incorporation as a strategy to suppress these side-reactions while preserving catalytic performance. Moreover, catalyst formulations were developed with significantly reduced platinum content (<0.25 wt.%), below literature values, aiming to decrease both catalyst cost and environmental impact. These advances represent a step forward toward more selective and sustainable dehydrogenation catalysts. Moreover, catalyst formulations were developed with a significantly reduced platinum content (<0.25 wt.%), below literature values [50], aiming to (i) synthesize bimetallic PtMo catalysts with low platinum content, using molybdenum as a catalytic promoter; (ii) reduce the platinum loading in the developed catalysts below 0.25 wt.%, in order to improve materials with a competitive noble metal content if compared with those reported in the literature; and (iii) investigate the effect of sulfur incorporation on catalytic performance, particularly regarding the formation of methylfluorene as a by-product during the dehydrogenation reaction, with the aim of minimizing its formation and supporting the potential for benzyltoluene recyclability in future hydrogen storage and release cycles. These advances represent a step forward toward more selective and sustainable dehydrogenation catalysts with lower cost and environmental impact.

2. Results

2.1. Catalyst Metal Content and Metal Dispersion

The effect of sulfur incorporation was studied by preparing a series of catalysts containing 0.2 wt.% Pt and 0.5 wt.% Mo, while varying the sulfur content from 0 wt.% to 0.2 wt.%. Based on our previous results, a Mo loading close to 0.5 wt.% was shown to be effective as a promoter for perhydrobenzyltoluene dehydrogenation, and was therefore selected for this study [49]. As a novel approach, we reduced the noble metal content to 0.2 wt.% Pt, lower than the values typically reported in the literature for this reaction [24,41,51], while maintaining a consistent Pt:Mo atomic ratio to enable a clear evaluation of sulfur’s influence. The XRF results confirm the successful incorporation of metals during catalyst preparation, with measured metal contents closely matching the nominal values. Importantly, the Pt:Mo atomic ratio remained consistent. This consistency in Pt ratios ensures that the influence of sulfur incorporation can be analyzed without interference from variations in the catalyst’s overall composition (

Table 1. Metal contents and Pt:Mo and Pt:S atomic ratios determined by XRF.

Catalyst	Pt wt.%	Mo wt.%	S wt.%	Pt:Mo	Pt:S
PtMo-0S	0.20	0.45	0	0.22	0
PtMo-0.05S	0.23	0.52	0.06	0.22	0.61
PtMo-0.10S	0.23	0.52	0.11	0.22	0.34
PtMo-0.15S	0.17	0.43	0.15	0.19	0.19
PtMo-0.20S	0.21	0.50	0.20	0.22	0.17

).

Due to the low metal incorporation, no significant differences were observed in the shape of the N₂ physisorption isotherms among the samples, with adsorption–desorption profiles characteristic of mesoporous γ-Al₂O₃ (N₂ isotherms are included in the Supplementary Materials). Regarding the catalysts’ surface area and pore volume, slight differences were noted when small amounts of sulfur (<0.1 wt.%) were incorporated, showing a minor decrease in both properties. However, these differences became less pronounced for catalysts containing higher sulfur contents (see

Table 2. Specific surface area (m²/g), pore volume, and pore size calculated using BET and BJH methods.

Sample	BET, m2/g	Pore Volume, cm3/g	Pore Radius, nm
PtMo-0S	249.37	0.86	3.91
PtMo-0.05S	219.11	0.81	3.92
PtMo-0.10S	208.47	0.75	3.93
PtMo-0.15S	233.23	0.83	4.83
PtMo-0.20S	231.81	0.82	3.91

). These results suggest that at low sulfur contents, metal nanoparticles (NPs) are incorporated into the catalyst pores, whereas at higher sulfur contents, increased metal–support and metal–metal interactions result in a more superficial incorporation of these elements.

The metal dispersion was calculated by CO-pulse chemisorption using a CO:Pt = 1 chemisorption stoichiometry [52,53]. The catalysts exhibit high Pt dispersion, with values above 70%. For the samples with 0.05% S and 0.1% S, which have the same Pt content, a slightly lower dispersion value is observed for the catalyst with 0.1% S, suggesting that sulfur incorporation may reduce metal dispersion. However, although the differences in Pt content across the samples are minimal, this variability complicates determining whether sulfur definitively impacts CO chemisorption (see

Table 3. CO chemisorption results.

Sample	CO, µmol/g	Pt, µmol/g	D, %
PtMo-0S	8.93	10.3	87.1
PtMo-0.05S	8.93	11.8	75.8
PtMo-0.10S	8.49	11.8	72.0
PtMo-0.15S	8.49	8.7	97.4
PtMo-0.20S	9.38	10.8	87.1

).

STEM images confirm the presence of platinum NPs, with mean particle sizes below 1.5 nm. However, as sulfur content increased, a more pronounced platinum agglomeration was observed, resulting in larger NPs. This suggests that sulfur weakens the Pt–support interactions, reducing the stabilization of smaller NPs and promoting the formation of larger clusters (Figure 1). Unfortunately, accurately determining the Pt distribution from STEM images to compare with CO chemisorption results is challenging due to the small particle sizes, as assuming spherical particles introduces considerable error though the calculations [54].

Figure 2 presents the HAADF maps for the unpromoted PtMo sample and the catalyst containing 0.2 wt.% sulfur. In the unpromoted sample, Pt is homogeneously distributed over the support, whereas the sulfur-containing catalyst (PtMo-0.2S) shows some Pt agglomeration (red dots). Although the PtMo-0.2S sample exhibited more intense contrast in the HAADF images, differentiating between molybdenum and sulfur in the mapping is challenging due to the overlap of the molybdenum Lα and sulfur Kα signals [55]. This spectral overlap, combined with the low overall metal content in the samples, limits the signal-to-background ratio and prevents the clear identification of sulfur distribution in the presence of molybdenum. For a complete comparison, HAADF maps of all synthesized samples are included in the Supplementary Materials (Figure S3).

2.2. Metal Oxide Reducibility

The H₂-TPR results illustrate that as the sulfur content in the catalysts increases, the intensity of the signal around 920 K rises due to the reduction of SO₄²⁻ to H₂S [56]. Additionally, it is observed that with higher sulfur content, the signal shifts to lower temperatures, indicating weaker interactions of sulfate groups with the metal/support (Figure 3a). This displacement can also be justified by a closer vicinity of SO₄²⁻ groups with the reduced Pt0 when a higher sulfur quantity was incorporated, as the dissociative adsorption of hydrogen and subsequent spillover onto the support can promote the reduction of close sulfur species at lower temperatures [57,58]. Moreover, Figure 3b reveals two distinct temperature reduction ranges for Pt, reflecting the varying strengths of interaction between Pt and the Al₂O₃ support. The first range (420–550 K) corresponds to weaker Pt-Al₂O₃ interactions (α-PtO₂), while the second range (600–750 K) indicates stronger Pt-Al₂O₃ interactions (β-PtO₂) [59]. In a recent publication, it was confirmed that the presence of molybdenum oxides helped reduce the platinum at lower temperatures [49]. Furthermore, these results show that sulfur groups enhance this effect even more.

2.3. Dehydrogenation Activity

Following the H₂-TPR results, before reaction tests, the samples were reduced ex situ at 673 K, to maximize the presence of Pt⁰ species without compromising the stability of the sulfur groups. Then, catalyst performance was evaluated at the same dehydrogenation conditions of 533 K, 1 bar, 1200 rpm, and 9.7 g_LOHC/g_cat.

The activity results indicated that catalytic activity increased with the addition of sulfur, especially during the first two hours of reaction, except for the sample containing 0.2 wt.% S. After four hours, all samples showed a similar final degree of dehydrogenation (DoD), with values exceeding 80%. Among these, the catalysts with the highest sulfur contents (0.15 wt.% and 0.2 wt.%) achieved the highest dehydrogenation levels, close to 85% (Figure 4a). During the reaction, while the addition of 0.05 wt.% sulfur did not significantly reduce the methylfluorene formation, increasing the sulfur content led to a marked decrease, particularly for the sample with the highest sulfur content, which showed more than a threefold reduction in methylfluorene formation compared to the unpromoted PtMo catalyst (Figure 4b).

Regarding H12-BT conversion, all catalysts tested achieved similar values above 95% after 4 h of reaction. However, during the first two hours, notable differences were observed between the undoped PtMo sample and the sulfur-doped samples. The addition of sulfur increased H12-BT conversion, particularly when 0.15 wt.% sulfur was incorporated (Figure 5a). This effect was not observed for the sample with 0.2 wt.% S, which showed a slight decrease in conversion, as previously reflected in the DoD (Figure 4a). Conversely, the selectivity values for this catalyst indicate that, despite its lower reaction yield during the first two hours, it achieved the maximum selectivity after 4 h (Figure 5b).

Although the incorporation of sulfur has been shown to effectively reduce methylfluorene formation, achieving complete conversion of the intermediate product (H6-BT) remains a challenge for maximizing dehydrogenation yield. Notably, while H12-BT concentrations vary across different samples, the consumption of H6-BT does not follow a direct two-step reaction sequence (H12-BT → H6-BT → H0-BT). The formation of H0-BT increases over time, whereas the H6-BT concentration decreases only gradually (Figure 6). The kinetic model based on the sequential two-step pathway showed that, particularly during the first two hours of reaction, this model underestimated the concentration of H0-BT and overestimated the accumulation of H6-BT compared to the experimental data. This discrepancy suggests that H12-BT may not exclusively convert through the H6-BT intermediate but may also undergo a direct dehydrogenation to H0-BT (H12-BT → H0-BT), as previously reported by Wang et al. [60]. The system of equations used in the kinetic model (Equations (S1)–(S9)) and the corresponding simulation results (Figure S1) are provided in the Supplementary Materials document.

2.4. Dehydrogenation Pathway

Based on this assumption, the dehydrogenation of perhydrobenzyltoluene was adjusted to a system of four reactions, including an additional reaction for the formation of methylfluorene (MF) via benzyltoluene dehydrogenative cyclization (H0-BT → MF) [49]. For this model, all reactions were assumed to follow first-order kinetics, and the formation of intermediate products, such as H4-BT and H10-BT, was considered negligible [19]. Since the reaction was conducted at atmospheric pressure with hydrogen being continuously removed from the reactor, the concentration of hydrogen gas was excluded from the kinetic equations.

The dehydrogenation process starts with the hydrogen-rich molecule H12-BT being converted into the partially dehydrogenated H6-BT. Additionally, H12-BT is directly dehydrogenated to the fully dehydrogenated target molecule H0-BT as previously mentioned (Equations (1) and (2)):

k₁
H12-BT -> H6-BT

(1)

k₄
H12-BT -> H0-BT

(2)

The H12-BT consumption rate is described by Equation (3), where k₁ is the rate constant for the conversion of H12-BT to H6-BT, and k₄ is the rate constant for the direct conversion of H12-BT to H0-BT.

$$-{\displaystyle \frac{\mathrm{d}\left[\mathrm{H}12\mathrm{˗}\mathrm{B}\mathrm{T}\right]}{\mathrm{d}\mathrm{t}}}={\mathrm{k}}_1·\left[\mathrm{H}12\mathrm{˗}\mathrm{B}\mathrm{T}\right]+{\mathrm{k}}_4·\left[\mathrm{H}12\mathrm{˗}\mathrm{B}\mathrm{T}\right]$$

(3)

Next, the partially dehydrogenated H6-BT is consumed in a subsequent step (Equation (4)):

k₂
H6-BT -> H0-BT

(4)

The H6-BT consumption rate is expressed as (Equation (5)):

$${\displaystyle \frac{\mathrm{d}\left[\mathrm{H}6\mathrm{˗}\mathrm{B}\mathrm{T}\right]}{\mathrm{d}\mathrm{t}}}={\mathrm{k}}_1·\left[\mathrm{H}12\mathrm{˗}\mathrm{T}\right]-{\mathrm{k}}_2·\left[\mathrm{H}6\mathrm{˗}\mathrm{B}\mathrm{T}\right]$$

(5)

Finally, the byproduct methylfluorene is considered to form only through the further dehydrogenation of H0-BT [49] (Equation (6)):

k₃
H0-BT -> MF

(6)

Thus, the rates of H0-BT and methylfluorene formation are given by Equations (7) and (8):

$${\displaystyle \frac{\mathrm{d}[\mathrm{H}0-\mathrm{B}\mathrm{T}]}{\mathrm{d}\mathrm{t}}}={\mathrm{k}}_2·\left[\mathrm{H}6-\mathrm{B}\mathrm{T}\right]+{\mathrm{k}}_4·\left[\mathrm{H}12-\mathrm{B}\mathrm{T}\right]-{\mathrm{k}}_3·\left[\mathrm{H}0-\mathrm{B}\mathrm{T}\right]$$

(7)

$${\displaystyle \frac{\mathrm{d}\left[\mathrm{M}\mathrm{F}\right]}{\mathrm{d}\mathrm{t}}}={\mathrm{k}}_3·\left[\mathrm{H}0\mathrm{˗}\mathrm{B}\mathrm{T}\right]$$

(8)

After defining the kinetic model, the system of differential equations was solved using MATLAB 9.13.0 R2022b. Then, the rate constants were then optimized by minimizing the sum of squared errors between the experimental data and the model predictions to a value lower than 5 × 10⁻³ (Figure 7). The obtained values and their corresponding errors are thoroughly detailed in the Supplementary Materials.

The reaction pathway for the dehydrogenation of H12-BT was significantly influenced by the presence of sulfur. As sulfur content increases (>0.1 wt.%), the reaction rate of the second dehydrogenation step (H6-BT → H0-BT) increases, indicating an enhanced ability of those catalysts to convert partially dehydrogenated H6-BT. However, the direct conversion from H12-BT to H0-BT was increasingly inhibited, especially at higher sulfur loadings, which suggests that sulfur blocks the more reactive but less selective Pt sites (Figure 8). This selective suppression of direct conversion aligns with the improved selectivity observed, where methylfluorene formation is minimized. These catalytic systems, especially at higher sulfur contents, promote a more controlled dehydrogenation, resulting in better overall performance with respect to both activity and selectivity.

2.5. Effect of Sulfur Doping on Bimetallic PtMo/Al₂O₃ Catalyst Activity

The differences in Pt nanoparticles’ functionality after sulfur incorporation into PtMo/Al₂O₃ were evaluated using CO-FTIR. In Figure 9, four bands for the CO absorbance on Pt can be distinguished. The signal at higher wavenumber values (2144–2154 and 2222 cm⁻¹) is linked to CO bonding with the remaining Pt²⁺ post-catalyst activation [61,62]. The predominant peak at 2064 cm⁻¹ corresponds to CO linearly bound to the Pt terraces, while the broad shoulder at 2042 cm⁻¹ on the lower-energy side of this band is ascribed to CO adsorbed on low-coordinated Pt edges and/or corners [63,64,65]. Furthermore, a signal is observed between 1837 and 1841 cm⁻¹ with a shoulder between 1789 and 1791 cm⁻¹ attributed to CO bridge-bonding two adjacent platinum terraces [66,67].

In Figure 9, the signal of CO bonded to Pt crystal faces at 2090–2080 cm⁻¹ linked to the presence of large crystals is not observed, confirming the presence of small Pt nanoparticles on the prepared catalysts [55,56]. Additionally, while small, the contribution of bridge-bonded CO to Pt sites explains the discrepancy between the particle size observed in STEM images and the Pt distribution calculated by CO pulse chemisorption. The CO bridge-bonded to Pt has a chemisorption stoichiometry of 2:1, different to the 1:1 stoichiometry assumed for particle dispersion calculations (Table 3), leading to an underestimation of the number of Pt active sites, especially for the catalysts with low sulfur contents (<0.15 wt.%).

After sulfur incorporation, the contribution of low-coordinated sites on Pt decreased, proving sulfur’s selective attachment to these sites and limiting CO access [48] (Figure 10). Additionally, the signal for linearly bonded CO on Pt terraces shifts slightly to higher wavenumbers due to the electron-withdrawing effect of sulfur. This effect reduces the electron density on Pt atoms, weakening the back-donation of electron density from Pt to the anti-bonding orbitals of CO. Consequently, the C–O bond is strengthened, resulting in a higher stretching frequency [68,69,70]. It is also remarkable that MoO₃ contributes to this electron-withdrawing effect, as shown in Figure 11, where the intensity of the signal between 1950 and 2100 cm⁻¹ increased and shifted the peak for adsorbed CO on the Pt edges and corners to higher wavenumbers.

To complement these results, the desorption behavior of CO on Pt/Al₂O₃ and PtMo/Al₂O₃ catalysts was investigated using CO-TPD, focusing on the effect of sulfur incorporation. The desorption profiles deconvoluted using the Gaussian method reveal three distinct regions: a low-temperature range (300–400 K), attributed to weakly adsorbed CO via van der Waals interactions [60]; an intermediate range (400–700 K); and a high-temperature range (>700 K), both associated with CO moderately and strongly interacting with Pt active centers (Figure 12) [71].

The most notable difference is observed in the desorption of CO at intermediate temperatures (500–700 K): while the monometallic Pt/Al₂O₃ sample shows a broad peak in this temperature range, the incorporation of molybdenum decreased this peak intensity and shifted it to higher temperatures. This shift was even more pronounced in the sample doped with 0.2 wt.% S, where the desorption peak shifted by 9 K compared to the bimetallic PtMo/Al₂O₃ sample without sulfur. This effect is ascribed to the role of molybdenum and sulfur in modifying the catalyst surface. Consistent with the CO-FTIR results, sulfur and molybdenum block low-coordinated active sites on Pt, reducing the availability of these highly reactive sites for CO adsorption. Consequently, the remaining exposed Pt sites are predominantly high-coordinated terrace sites, which are less reactive for CO adsorption and require higher temperatures for CO desorption. This alteration creates a more selective surface for the reaction.

In Figure 13, the evolution of hydrogen released (mmol) per mmol of Pt introduced into the reaction medium is presented. This figure shows that the sample doped with 0.15 wt.% S exhibited outstanding efficiency in dehydrogenation, likely due to its slightly lower platinum content. For the other sulfur-doped samples with similar Pt contents, the H₂ desorption profiles during the first two hours are comparable, with lower amounts of H₂ released than the one observed for the undoped PtMo catalyst. Notably, the catalyst containing 0.2 wt.% S shows a significant difference: while the other catalysts nearly reach a plateau after two hours, this catalyst continues increasing the amounts of hydrogen released. This effect is attributed to its enhanced ability to dehydrogenate the intermediate H6-BT, as discussed in Section 2.4. These results confirm that the incorporation of sulfur limits H0–BT adsorption by inducing electron deficiency on the Pt surface. Simultaneously, it regulates adsorption/desorption dynamics, ensuring the availability of active sites for catalytic turnover, consistent with reported in situ DRIFT studies [50].

Since the primary effect of sulfur is to block low-coordinated Pt sites, as Pt particle size increases, the number of low-coordinated centers decreases, meaning less sulfur is required to achieve the same effect [44,72]. In the case of the PtMo-0.15S sample, the sulfur content was lower than the optimal level for the catalyst’s particle size (<1.5 nm). While this sample exhibited higher dehydrogenation productivity, the sulfur incorporation was insufficient in effectively blocking these Pt active sites, resulting in a higher formation of methylfluorene. In contrast, the PtMo-0.2S sample showed lower dehydrogenation activity during the first two hours of reaction. This performance was attributed to a more extensive coverage of low-coordinated Pt sites, which are suggested to play a key role in the direct dehydrogenation pathway from fully hydrogenated H12-BT to fully dehydrogenated H0-BT, as discussed in Section 2.4 ‘Dehydrogenation pathway’. To further investigate Pt–S interactions, preliminary XPS analyses were conducted. However, due to the low metal content and the resulting high noise-to-signal ratio, the data were inconclusive, and no additional analyses were performed.

The PtMo-S catalysts presented in this work have been proven to be highly efficient in the dehydrogenation of perhydrobenzyltoluene, reaching yields comparable with catalysts reported with higher metal loadings (>2 wt.% Pt) [73,74]. Smaller particle sizes enable a reduction in Pt content by increasing particle dispersion. However, smaller particle sizes make the catalysts less open to modification through the sulfur doping strategy, requiring higher sulfur quantities to suppress undesired reactions and carrying the risk of poisoning Pt active sites due to excessive sulfur content.

3. Materials and Methods

3.1. Catalyst Preparation

The catalysts were prepared using the wetness impregnation method, with sequential incorporation of the metal loadings as described in previous work [75]. First molybdenum was added to the γ-Al₂O₃ (1/8″ Alfa Aesar) using (NH₄)₆Mo₇O₂₄·4H₂O (Merk > 99.0%), as a metal precursor. After calcination at 773 K (2 K/min) for 4 h, the resulting powder was milled to a particle size below 250 µm. Platinum was then added using (NH₃)₄Pt(NO₃)₂ (Aldrich, Polk County, MO, USA, 99.995%) along with sulfur (NH₄)₂SO₄ into the Mo/Al₂O₃ powder, followed by calcination at 673 K (2 K/min) for 4 h. The detailed procedure is included in the Supplementary Materials.

3.2. Catalyst Characterization

After catalyst preparation, the metal content was determined using X-ray fluorescence (XRF) with a PANalytical AXIOS sequential wavelength-dispersive X-ray fluorescence spectrometer. This device is equipped with a Rh tube and three detectors: gas flow, scintillation, and Xe sealing.

Catalyst textural properties were obtained from the N₂ adsorption–desorption isotherms using an Autosorb iQ3 (Quantachrome, Boynton Beach, FL, USA). Surface areas were calculated by the BET method, and the average pore size and total pore volume were calculated by the BJH from desorption points of the isotherms.

Metal dispersion was measured by CO pulse-chemisorption using an Autochem II (Micromeritics, Norcross, GA, USA). Prior to analysis, the samples were reduced at 673 K for 1 h under a 5% H₂/Ar (vol./vol.) gas mixture (Air Liquide, Paris, France).

Particle size and dispersion were analyzed via STEM-HAADF with a Schottky X-FEG electron emission microscope (FEI Titan Cubed G2 60-300, Thermo Fisher Scientific, Waltham, MA, USA).

The reducibility of the catalyst species was investigated through temperature-programmed reduction (H₂–TPR) on an Autochem II (Micromeritics, USA) under a reducing atmosphere of 5% H₂/Ar (vol./vol.) (Air Liquide).

The functionality of the Pt nanoparticles and the influence of sulfur doping were studied using CO Diffuse Reflectance Infrared Fourier Transform Spectroscopy (DRIFT) with a VERTEX 70v FT-IR spectrometer, equipped with a Harrick ATC temperature controller (Thermo Fisher Scientific, Waltham, MA, USA).

Finally, the CO desorption behavior of the catalysts was evaluated by temperature-programmed desorption (CO–TPD) using an Autochem II (Micromeritics, Norcross, GA, USA). Samples were reduced in situ under a 5% H₂/Ar (vol./vol.) atmosphere (Air Liquide) before being saturated with CO via pulses of 10% CO/He. Physisorbed CO was subsequently removed during a 10 min purge prior to conducting the CO desorption analysis.

Additional information regarding equipment and procedures is included in the Supplementary Materials.

3.3. Dehydrogenation Tests

Dehydrogenation experiments were performed in a 100 mL three-neck flask reactor. For each experiment, the reactant, perhydrobenzyltoluene (H12–BT), was incorporated in the reactor which was then purged with N_2. The reactor was heated, and upon reaching the desired operation temperature (533 K), the reactant was stirred with the catalyst to start the reaction. Samples of the reaction mixture were collected at intervals through a septum using a syringe, while the released hydrogen was quantified using a flow meter placed at the reactor outlet. Liquid samples were analyzed by GC–MS (Agilent 7890–A), and the degree of dehydrogenation (DoD), conversion of H12–BT, and selectivity to fully dehydrogenated benzyltoluene (H0–BT) were calculated using Equations (9), (10) and (11), respectively.

$$\mathrm{D}\mathrm{o}\mathrm{D} \left(\%\right)= {\displaystyle \frac{{\mathrm{n}}_{\mathrm{H}2,\mathrm{m}\mathrm{á}\mathrm{x}.}-{\mathrm{n}}_{\mathrm{H}2,\mathrm{t}=\mathrm{t}}}{{\mathrm{n}}_{\mathrm{H}2,\mathrm{m}\mathrm{á}\mathrm{x}.}}}·100$$

(9)

$${\mathrm{C}\mathrm{o}\mathrm{n}\mathrm{v}\mathrm{e}\mathrm{r}\mathrm{s}\mathrm{i}\mathrm{o}\mathrm{n}}_{\mathrm{H}12-\mathrm{B}\mathrm{T}}= {\displaystyle \frac{{\mathrm{n}}_{\mathrm{H}12-\mathrm{B}\mathrm{T},\mathrm{t}=0}-{\mathrm{n}}_{\mathrm{H}12-\mathrm{B}\mathrm{T},\mathrm{t}=\mathrm{t}.}}{{\mathrm{n}}_{\mathrm{H}12-\mathrm{B}\mathrm{T},\mathrm{t}=0.}}}·100$$

(10)

$${\mathrm{S}\mathrm{e}\mathrm{l}\mathrm{e}\mathrm{c}\mathrm{t}\mathrm{i}\mathrm{v}\mathrm{i}\mathrm{t}\mathrm{y}}_{\mathrm{H}0-\mathrm{B}\mathrm{T}}= {\displaystyle \frac{{\mathrm{n}}_{\mathrm{H}0-\mathrm{B}\mathrm{T},\mathrm{t}=\mathrm{t}.}}{{\mathrm{n}}_{\mathrm{H}12-\mathrm{B}\mathrm{T},\mathrm{t}=0}-{\mathrm{n}}_{\mathrm{H}12-\mathrm{B}\mathrm{T},\mathrm{t}=\mathrm{t}.}}}·100$$

(11)

4. Conclusions

In this study, low-PGM (<0.3 wt.% Pt) PtMo-S catalysts were successfully prepared using the wetness impregnation method. The catalysts exhibited highly dispersed nanoparticles homogeneously distributed across the support. The introduction of sulfur proved to selectively block low-coordinated Pt sites, resulting in fewer but more selective active centers, thereby enhancing the catalytic performance for perhydrobenzyltoluene dehydrogenation. Notably, the catalyst containing 0.2 wt.% sulfur demonstrated superior conversion of the partially dehydrogenated intermediate (H6–BT) and significantly reduced methylfluorene formation, achieving a threefold reduction compared to the unpromoted PtMo/Al₂O₃ catalyst. Sulfur also promoted the formation of larger Pt clusters, affecting the catalytic productivity. However, due to simultaneous variations in particle size and sulfur content, it was not possible to isolate the individual contribution of these factors to catalytic activity in this study. Nonetheless, both factors and the results demonstrate the excellent performance of these catalysts for LOHC applications, achieving high activity with minimal metal usage and low environmental impact.

These findings lay the groundwork for future research focusing on continuous operation, where the stability and dehydrogenation performance of the catalysts will be evaluated under varying operational conditions and the potential of sulfur-modified PtMo catalysts in sustainable and efficient LOHC systems will be explored.