Adjacent Pt Nanoparticles and Sub-nanometer WO x Clusters Determine Catalytic Isomerization of C 7 H 16

Open AccessCCS ChemistryRESEARCH ARTICLE5 Aug 2022Adjacent Pt Nanoparticles and Sub-nanometer WOx Clusters Determine Catalytic Isomerization of C7H16 Bin Zhang, Wei Zhou, Jie Zhang, Zirui Gao, Danyang Cheng, Lipeng Tang, Xingwu Liu, Yueqin Song, Chunyang Dong, Yao Xu, Jie Yan, Mi Peng, Huizhen Liu, Mark Douthwaite, Meng Wang and Ding Ma Bin Zhang Beijing National Laboratory for Molecular Sciences, College of Chemistry and Molecular Engineering and College of Engineering, and BIC-ESAT, Peking University, Beijing 100871 Beijing National Laboratory for Molecular Sciences, CAS Key Laboratory of Colloid and Interface and Thermodynamics, Institute of Chemistry, Chinese Academy of Sciences, Beijing 100190 Google Scholar More articles by this author , Wei Zhou Beijing National Laboratory for Molecular Sciences, College of Chemistry and Molecular Engineering and College of Engineering, and BIC-ESAT, Peking University, Beijing 100871 Google Scholar More articles by this author , Jie Zhang Beijing National Laboratory for Molecular Sciences, College of Chemistry and Molecular Engineering and College of Engineering, and BIC-ESAT, Peking University, Beijing 100871 Google Scholar More articles by this author , Zirui Gao Beijing National Laboratory for Molecular Sciences, College of Chemistry and Molecular Engineering and College of Engineering, and BIC-ESAT, Peking University, Beijing 100871 Google Scholar More articles by this author , Danyang Cheng Beijing National Laboratory for Molecular Sciences, College of Chemistry and Molecular Engineering and College of Engineering, and BIC-ESAT, Peking University, Beijing 100871 Google Scholar More articles by this author , Lipeng Tang Beijing National Laboratory for Molecular Sciences, College of Chemistry and Molecular Engineering and College of Engineering, and BIC-ESAT, Peking University, Beijing 100871 Google Scholar More articles by this author , Xingwu Liu Beijing National Laboratory for Molecular Sciences, College of Chemistry and Molecular Engineering and College of Engineering, and BIC-ESAT, Peking University, Beijing 100871 Google Scholar More articles by this author , Yueqin Song Petroleum Processing Research Center, East China University of Science and Technology, Shanghai 200237 Google Scholar More articles by this author , Chunyang Dong Beijing National Laboratory for Molecular Sciences, College of Chemistry and Molecular Engineering and College of Engineering, and BIC-ESAT, Peking University, Beijing 100871 Google Scholar More articles by this author , Yao Xu Beijing National Laboratory for Molecular Sciences, College of Chemistry and Molecular Engineering and College of Engineering, and BIC-ESAT, Peking University, Beijing 100871 Google Scholar More articles by this author , Jie Yan Beijing National Laboratory for Molecular Sciences, College of Chemistry and Molecular Engineering and College of Engineering, and BIC-ESAT, Peking University, Beijing 100871 Google Scholar More articles by this author , Mi Peng Beijing National Laboratory for Molecular Sciences, College of Chemistry and Molecular Engineering and College of Engineering, and BIC-ESAT, Peking University, Beijing 100871 Google Scholar More articles by this author , Huizhen Liu Beijing National Laboratory for Molecular Sciences, CAS Key Laboratory of Colloid and Interface and Thermodynamics, Institute of Chemistry, Chinese Academy of Sciences, Beijing 100190 Google Scholar More articles by this author , Mark Douthwaite Cardiff Catalysis Institute, School of Chemistry, Cardiff University, Cardiff CF10 3AT Google Scholar More articles by this author , Meng Wang *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] Beijing National Laboratory for Molecular Sciences, College of Chemistry and Molecular Engineering and College of Engineering, and BIC-ESAT, Peking University, Beijing 100871 Google Scholar More articles by this author and Ding Ma *Corresponding authors: E-mail Address: m.w[email protected] E-mail Address: [email protected] Beijing National Laboratory for Molecular Sciences, College of Chemistry and Molecular Engineering and College of Engineering, and BIC-ESAT, Peking University, Beijing 100871 Google Scholar More articles by this author https://doi.org/10.31635/ccschem.021.202101454 SectionsSupplemental MaterialAboutAbstractPDF ToolsAdd to favoritesDownload CitationsTrack Citations ShareFacebookTwitterLinked InEmail Processes for the isomerization of light alkanes have been commercialized; however, the isomerization of paraffins (CnH2n+2, n ≥ 7) remains a challenge. On mesoporous tungsten-zirconia catalyst supported Pt catalysts (Pt/WZrOx), initial isomerization productivity of 5249 moli-C7/molPt/h was obtained for n-heptane reforming at 275 °C and 5 bar of hydrogen. The catalyst possessed quasi-monolayer and uniformly distributed WOx clusters (diameter of ∼0.5 nm), and Pt nanoparticles were preferentially deposited in close vicinity to these WOx sub-nanometer monolayers. The initial productivity over the best Pt/WZrOx catalyst was 18 times higher than that of a corresponding physical mixture (0.8 wt % Pt/ZrO2 + WZrOx), highlighting the significance of intimacy between the Pt and WOx species. The isomerization could not be triggered until the acid density of the catalysts reached a threshold because of the necessity of close proximity between adequate acid sites and adjacent Pt metal sites. Meanwhile, the adjacency increased the metallicity of the supported Pt; thus, enabling easier activation of C–H bonds of C7H16. Download figure Download PowerPoint Introduction Bifunctional catalysts, containing both metal and acid sites, have been utilized extensively for both research and commercial applications since Oblad et al.1 and Weisz2 first uncovered the catalytic potential of such materials in the 1950s. In the years proceeding, such materials have shown highly effective catalytic activities for a number of industrial processes, including the hydroisomerization of light naphtha (C5–C6),3–5 the hydrocracking of heavy oils (C9–C18),6–8 hydrolytic hydrogenation,9–11 the hydrodeoxygenation of petrochemicals,12 and biomass conversion,13–15 among others.The isomerization of linear paraffins is an established route for the production of high-octane-value gasoline (C5–C11). Both acid catalysts and bifunctional metal–acid catalysts have been shown to be effective for such reactions. However, most often, poorer activity and product selectivity to isomers are exhibited on monofunctional acid catalysts.16,17 But the addition of metal sites and co-feeding of hydrogen leads to significantly better catalytic performance.18,19 The metal component acts as “an engine” and could initiate the dehydrogenation of alkanes, the reaction of which must occur to push forward C–C skeletal isomerization. To date, processes for the isomerization of light C5 and C6 feedstocks have been commercialized but are not considered sufficiently effective for the isomerization of longer chain alkanes such as heptane (n-C7). This is due to the fact that long-chain alkanes are more susceptible to undergoing C–C cleavage, which limits the desired selectivity to their corresponding isomers (i-C7). Given that this undesirable pathway is often promoted by intensive heating, it motivates researchers to design bifunctional catalysts which are more efficient at lower temperatures. Previous reports have suggested that the performance of metal–acid catalysts in hydroisomerization reactions could be influenced by several factors, including metal-to-acid ratio, the intrinsic properties of metal,20,21 the strength and density of acid sites,22 and the distance between metal and acid sites.17,23 The influence of metal–acid proximity remains a mystery. Pioneer work by Weisz,24 suggested that the highest catalytic performance was achieved when the metal and acid sites were closer together. However, more recently, research has indicated that higher proximity is not always beneficial. Seminal work by de Jong and co-workers12 found that the highest isomerization yield was achieved for conversion of C10 and C19 paraffins when the acid sites were nanoscale away to the metal sites; their most recent work further discussed this viewpoint in length.23 Regalbuto and co-workers16 found that a single-site Pt/SA catalyst possessing the atomic intimacy of Pt atoms and acidic H+ sites was poorly active and selective for heptane isomerization, compared with Pt nanoparticle catalysts that have metal–acid proximity of nanoscale or micrometer. Further, several studies investigated the use of metal/zeolite catalysts and have claimed that the intimacy of metal and acid sites has minimal impact on their ability to catalyze hydroisomerization reactions.19,25,26 Höchtl et al.26 reasoned that almost no improvement could be achieved as long as the metal and acid support were in direct contact in the catalyst bed. Unfortunately, most of the works from the literature considered metal, acid sites, or their distance as independent variates. It is, however, worth noting that proximate metal and acid sites likely influence the other component’s metallicity (electron-donor ability) or acidity (acid properties) significantly, thereby varying the ability of C–H activation and C–C rearrangement or cleavage.16,27–29 Since both metallicity and acidity play pivotal roles in catalytic hydroisomerization, their changes arising from their proximity must be taken into account instead of reasoning they are completely independent factors. Since the reaction mechanism on Pt/WZrOx catalysts has been carefully investigated, 30–32 the acid sites and metal sites could be comprehensively characterized,33–37 which revealed that the acid-metal intimacy was controllable.16,17,27,38 WZrOx was chosen as the acidic support, and Pt was the engine used to investigate the influence of adjacency on the properties of acid and metal and the hydroisomerization performance. In this work, we found that the intimacy of Pt/WZrOx catalyst showed significant effects on catalytic reforming of n-heptane. We designed a series of catalysts containing acid (WZrOx), metal (Pt/WO3, Pt/ZrO2), metal–acid with different intimacy from different loadings of metal or acid (Pt/WZrOx), and distant metal–acid (Pt/ZrO2 + WZrOx) catalysts, for the catalysis of heptane reforming. A clear spatial distribution of Pt and WOx monolayers was visualized by aberration-corrected high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) and energy dispersive X-ray (EDX) map. The capability of activating the first C–H of heptane molecules was carefully investigated by isotopic temperature-programmed surface reaction (i-TPSR) experiments. The accurate acid concentrations of Pt/WZrOx catalysts were elegantly performed by pyridine titration experiments combined with steady-state methanol dehydration. A clear conditional relationship between the reactivities and acid concentrations was obtained from Pt/WZrOx catalysts when Pt loading was fixed. Experimental Methods Materials Zirconium oxynitrate hydrate [ZrO(NO3)2·xH2O, >99.5%] was obtained from Aladdin Co., Ltd. (Beijing, China). Ammonium metatungstate hydrate [(NH4)6H2W12O40 xH2O, >99.9%] was obtained from STREM Chemicals, Inc. (Beijing, China). Chloroplatinic acid (H2PtCl6·6H2O, >99.9%] was obtained from Aladdin Co., Ltd. Heptane, 2-methylhexane, 3-methylhexane, 2,2-dimethylpetane, 2,3-dimethylpetane, 2,4-dimethylpetane, 3,3-dimethylpetane, 2-ethylpetane, and 2,2,3-trimethylbutane (C7H16, GR) were obtained from Aladdin Co., Ltd. Hexane, 3-methylpentane, and 2-methylpentane (C6H14, GR) were obtained from Aladdin Co., Ltd. Pentane and 2-methylbutane (C5H12, GR) were obtained from Aladdin Co., Ltd. Butane and 2-methylpropane (C4H10), propane (C3H8), ethylene (C2H6), and methane (CH4) (>99.9%) were obtained from Beijing Haikeyuanchang Practical Gas Co., Ltd. (Beijing China). Catalyst preparation Synthesis of WZrOx support The four preparation methods in this work are as follows.34 Ammonia sol–gel method: A given amount of zirconium oxynitrate and ammonium metatungstate (13 wt % W loading in the assumed formation of WOx-ZrO2 or 16.6% WO3 loading) were dissolved separately in 100 mL water until the solution became transparent. The two solutions were mixed dropwise under vigorous stirring at a constant pH of 10.5 through slow addition of ammonia water. The suspended solution was stirred for another 4 h, then filtered, washed with deionized water until the pH value of the filtrate was 7. The sol–gel obtained was dried for 4 h at 60 °C in a vacuum oven, then 2 h at 120 °C in the oven, and calcined at 700 °C for 4 h with a heating rate of 2 °C/min. The final sample was denoted as WZrOx. For the WZrOx catalyst throughout the work, the W weight loading was 13 wt %, unless otherwise specified. The other three WZrOx samples with different W loading (1%, 6%, and 10%) were prepared using the same procedures mentioned above, except for the amount of ammonium metatungstate. Oxalic acid co-precipitation (CP): As stated above, only the precipitation agent and solvent were substituted by oxalic acid and ethanol, respectively, and other procedures stayed the same. Incipient wetness impregnation (IWI): ZrO2 support was first synthesized by the above ammonia sol–gel method and then was impregnated by a given amount of ammonium metatungstate aqueous solution by the incipient wetness method. The procedure of drying and calcination stayed the same, as indicated above. Alkaline hydrothermal (HT) synthesis: As stated in the sol–gel method, the solution mixture was transferred to the HT autoclave for 24 h. The other steps in the procedure stayed the same. Preparation of Pt/WZrOx catalysts The WZrOx support was dispersed in 10 mL of deionized water (pH 7), and the given chloroplatinic acid solution with a final Pt loading (0.2–5 wt %) was impregnated on the support. Then the powder was dried at 60 °C in a vacuum oven overnight and calcined at 350 °C for 2 h with a heating rate of 2 °C/min. Preparation of Pt/WO3, Pt/ZrO2, and physical hybrid catalysts WO3 support was prepared from ammonium metatungstate through calcination at 700 °C for 4 h with a heating rate of 2 °C/min. ZrO2 support was prepared using the above-mentioned sol–gel method. Then Pt was impregnated to these supports as stated in the last subsection. Catalyst characterization Powder X-ray diffraction (XRD) patterns were collected in the 2θ angle range of 20–80° on a Rigaku D/MAX-PC 2500 powder X-ray diffractometer (Beijing, China) using Cu Kα radiation (λ = 1.5406 Å). The accelerating voltage and current were 40 kV and 100 mA, respectively. Brunauer–Emmett–Teller (BET) surface areas of catalyst samples were measured on an ASAP 2020 analyzer (Micromeritics, Beijing, China) by N2 physisorption at its typical boiling point (−196 °C) after the sample evacuation (<2.66 Pa) at 120 °C for 6 h. The element content of the catalysts was determined using inductively coupled plasma atomic emission spectrometry (ICP-AES), in which all samples were dissolved in aqua regia at 80 °C for 2 h for the tests. HAADF-STEM and EDX elemental mapping measurements were carried out on a double aberration-corrected FEI Titan Cubed Themis G2 (Beijing, China) operated at 300 kV and equipped with an XFEG (a unique brightness module) gun and Bruker Super-X EDS detectors (Beijing, China). X-ray photoelectron spectroscopy (XPS) measurements were carried out on an Axis Ultra imaging photoelectron spectrometer (Beijing, China) equipped with Al Kα (1486.7 eV) quartz monochrome source. The binding energies were corrected by setting the C 1s signal of 284.8 eV as reference. Solid-state NMR experiments were performed on a Bruker AVANCE III 400 MHz spectrometer (Beijing, China) using a 4 mm magic-angle-spinning (MAS) probe. Pyridine-FTIR experiment was conducted at room temperature in the in situ transient platform CRCP-7070-A (Beijing, China). CO-/hydrogen physicochemsorption experiments were performed on Micromeritics Autochem II 2920 physio-chemisorption analyzer (Beijing, China) with a standard operation to measure the dispersion of Pt, following the manufacturer instrumental instructions. The productivity was normalized on surface Pt based on CO chemosorption unless otherwise specified. We performed temperature-programmed desorption with ammonia (NH3-TPD) on TP-5080B equipped with a thermal conductivity detector (TCD; Tianjin Xianquan Company Ltd., Beijing, China). Isotopic TPSR experiments were performed on a temperature-programmed micro-fixed bed reactor equipped with a mass spectrometer (MS; OmniStar, Pfeiffer, Beijing, China). Typically, a 50 mg sample was loaded in a quartz tube. The sample was pretreated in Ar (20 mL/min) at 300 °C for 60 min with a heating rate of 10 °C/min then cooled down in Ar until the baseline signal in MS spectra was steady at 20 °C. At that time, D2 gas (20 mL/min; or D2 saturated with heptane vapor at 20 °C) was introduced into the reactor and kept for ∼30 min at 20 °C. Next, the sample was subjected through a temperature program from 20 to 300 °C with a heating rate of 10 °C/min in the same atmosphere and kept for a period at 300 °C. The products were detected using MS. Pyridine titration experiments combined with methanol dehydration were conducted in a tubular flow reactor. The experimental apparatus was schematically diagrammed in Supporting Information Figure S1. In a standard run, 80–500 mg of catalysts were ground, pelleted within 20–40 mesh, then loaded to the reactor tube. The catalysts were subjected to a pretreatment process via calcination at 550 °C for 0.5 h in Ar, then cooled down to the testing temperature of 150 °C and maintained in the remaining processing time in Ar. The methanol saturator was maintained at −5 °C with Ar flow at 46 mL/min. Until the consuming rate of methanol dehydration reached a constant (∼30 min), the other 2 mL/min of Ar was passed through the other mixture saturator of cyclohexane(l)/pyridine(l) (V/V 1/50), maintained at −9 °C. The experiment was completed when the concentration of pyridine reached an equilibrium (∼90 min). The tail gas in the whole process was analyzed online continuously by gas chromatography (GC; Agilent 7890A, Beijing, China). The GC programming involved heating, testing, and cooling in one run, which was controlled within 5 min over catalysts with 1–13 wt % W. The initial conversion of methanol dehydration was controlled <5% by changing the catalyst loading weight (80–500 mg). Reaction evaluation The catalytic reaction was carried out in a fixed bed stainless steel reactor with an 8-cm-diameter glass liner. In a standard run, 100 mg of catalyst was loaded to the bed, and the remaining volume of the reactor tube (∼8 cm in height) was filled with quartz beads. The catalyst was used directly without any treatment (granulation or activating procedures). The hydrogen pressure was 5 bar with a flow rate of 15 mL min−1 unless otherwise specified. Once the system was stabilized, the heptane was passed through the reactor at a precise liquid flow rate of 0.01 mL/min. The temperature increased stepwise, and the reaction was allowed to equilibrate for 1 h before product sampling. Every reaction point was tested three times. The effluent gas from the reactor was analyzed online by GC (Agilent 7890A) equipped with an Agilent HP-PONA column (Agilent, Beijing, China) connected with a flame ionization detector. All the products were vaporized and detected by all-on-line GC with PONA column, and the product distributions were calculated using the normalization method, as reported in the literature.23,39 All the C1–C7 components containing 10 products (C1–C6), 8 isomers of C7, and heptane reactant were detected and well separated on GC spectra ( Supporting Information Figure S2). Results and Discussion The WZrOx support was synthesized via a sol–gel method with ammonia,34 using ammonium metatungstate and zirconium oxynitrate as precursors. XRD patterns confirmed that the material comprised pure tetragonal zirconia (t-ZrO2). No diffraction patterns indicative of WO3 crystals were observed, highlighting its high dispersion throughout the material ( Supporting Information Figure S3). Additional WZrOx materials were prepared by CP, IWI, and the alkaline HT method.34 The W loading was fixed at 13 wt % for most cases to provide sufficient acid concentration, while lower W loadings were prepared for further analysis. Noticeable reflections, indicative of monoclinic zirconia (m-ZrO2) and tungsten trioxide (WO3) crystallites, were observed when the material was synthesized by CP, IWI, and HT ( Supporting Information Figure S3). Given that crystallographic WO3 has previously been demonstrated to have no acidity and isomerization activity,40 only WZrOx materials prepared by the sol–gel method were selected for further studies. Subsequently, a series of Pt/WZrOx catalysts (0.2–5 wt % Pt) were prepared by wet impregnation with chloroplatinic acid. Then the catalysts were probed using HAADF-STEM and EDX maps. Resolving WOx and Pt species clearly on the atomic scale by microscopy was often challenging due to the close atomic mass between W and Pt.41 However, in our micrographs obtained with 0.8 wt % Pt/WZrOx catalyst, the WOx sub-nanometer clusters and Pt nanoparticles were fully resolved (Figure 1). The tungsten-oxo clusters, with a mean particle diameter of 0.5 nm (±0.1 nm), were clearly shot on the surface of zirconia support (Figures 1a and 1b). A line intensity profile (Figures 1b and 1c) showed that the majority of these clusters existed as monolayers in thickness, but a few bilayer clusters were observed in the boundary. Through consideration of the W loading (13 wt %) and the surface area of the support (50 m2/g), the apparent density of tungsten was estimated to be 8.6 W atoms/nm2 (Details see Supporting Information Table S1). This was slightly higher than the theoretically calculated value, which assumed that the tungsten-oxo clusters were fully mono-dispersed (6.4 W atoms/nm2). Thus, our result supported our observation that WOx species existed primarily as monodispersed monolayers and a few bilayers. Raman spectroscopy further confirmed that the tungsten-oxo clusters were present ( Supporting Information Figure S4c). Strong bands indicative of tungsten-oxo clusters were observed; the bands at 1016 and 822 cm−1 were characteristic of terminal W=O stretching and W–O–W vibrations in WOx clusters, respectively. Also, a strong band at 890 cm−1 was observed, ascribed to W–O–Zr vibrations. The band at 639 cm−1 was characteristic of t-ZrO2 vibration; there was no observable m-ZrO2. Figure 1 | Electron microscopy characterization of 0.8 wt % Pt/WZrOx catalyst. HAADF-STEM images of (a) quasi-monolayer dispersed sub-nanometer tungsten-oxo clusters and (b) Pt nanoparticles ∼4-nm diameter. (c) Line intensity profile across WOx clusters on the periphery (1) and bulk (2) of the support. (d) HAADF-STEM images and EDX elemental maps of Pt, W, Zr, and O. Download figure Download PowerPoint Figure 1d and Supporting Information Figure S5 illustrate that the supported Pt nanoparticles (ca. 4 nm in diameter) are intimately associated with the WOx clusters. Elemental mapping highlighted that W and Zr were well distributed throughout the sample after impregnation with Pt. The other Pt/WZrOx catalysts, with varying Pt loadings (0.2–5 wt % Pt), were also well-characterized ( Supporting Information Figures S6–S8). In the 0.2 wt % Pt catalyst, no Pt nanoparticles were observed during the whole shooting, indicating that Pt species, most likely, highly dispersed ( Supporting Information Figure S6). On the contrary, much larger Pt particles (∼20 nm) were evident in the 5 wt % Pt catalyst ( Supporting Information Figure S7). We also noted in STEM images that ZrO2 support possessed an abundance of mesoscale channels ( Supporting Information Figure S6). The mesoporosity was confirmed by observing hysteresis in an N2-sorption experiment ( Supporting Information Figure S8). Almost no changes were observed in the mesoporosity and surface area of Pt/WZrOx catalysts as the Pt loading increased from 0 to 2 wt % inferred from Supporting Information Figure S8 and Supporting Information Table S1. Previous work has suggested that a minimum metal/acid ratio of 0.03 is required for the acid-catalyzed skeletal rearrangement to be a rate-determining step.26,42 In the 0.8 wt % Pt/WZrOx catalyst, the metal-to-acid ratio was determined to be 0.25, based on CO chemisorption (8.2 μmolPt/gcat) and pyridine titration (32.8 μmolPy/gcat). Thus, according to previous suggestions, this catalyst possessed an eligible ratio of metal/acid to study the acid-catalyzed isomerization. Furthermore, the intimate metal and acid sites and the mesoporosity (important for limiting any internal diffusion) suggested that it aligned well with Guisnet’s interpretation of an “ideal” bifunctional catalysts for alkane isomerization.17 Further, the Pt/WZrOx catalyst was assessed for its ability to catalyze heptane reforming under a weight hour space velocity (WHSV) of 4.08 gn-C7·gcat·h−1 in a fixed-bed reactor. The results from these experiments are presented in Figure 2a. The conversion of heptane increased from 4.2% at 200 °C to 51.1% at 275 °C, after which it appeared to level off. The selectivity to isomerization dropped as the temperature was increased since a higher temperature promoted undesirable C–C cleavage. Nevertheless, at 275 °C, an isomerization selectivity of 81.4% was observed. Additional catalytic experiments over Pt/WZrOx catalysts with Pt loadings of 0.2–5 wt % were next conducted ( Supporting Information Figure S9). Under the same conditions (4.08 gn-C7·gcat·h−1, 275 °C), the moderate Pt loading catalysts, 0.8 or 2 wt %, exhibited the highest isomerization rate, compared with much lower or higher Pt loading ( Supporting Information Figure S9a). The catalytic performances of these catalysts are shown in Supporting Information Figure S9b. The 5 wt % Pt/WZrOx catalyst presented a lower selectivity to isomerization than other Pt loadings, suggesting that a heavier hydrocracking or hydrogenolysis reaction occurred due to poor dispersion of large-size Pt particles. The conversion of heptane increased as the Pt loadings were increased; until the Pt loading was increased to 2 wt %, the conversion reached 81.1%. Increasing the Pt loading beyond 2 wt % led to no further improvements in performance. Over the range of catalysts and assessed conditions, the highest isomerization yield to i-C7 observed was 55.7% over the 2 wt % Pt/WZrOx catalyst at 275 °C. The associated isomerizing productivity of this catalyst was comparable, and in some cases, superior to the state-of-the-art Pt catalysts presented in the open literature ( Supporting Information Table S2). The apparent activation energy of 0.8 wt % Pt catalyst was also determined to be the lowest, 92.6 kJ/mol, among these catalysts ( Supporting Information Figure S10). Furthermore, this catalyst was shown to be stable for 22 h on stream, while the isomers’ yield was maintained ( Supporting Information Figure S11). Figure 2 | (a) Heptane reforming over 0.8 wt % Pt/WZrOx during 200–300 °C. Reaction conditions: WHSV 4.08 gn-C7·gcat·h−1, n(H2)/n(oil) 10, and 5 bar H2. (b) Effects of metal–acid proximity on isomerization of heptane. The initial productivity was measured at a low conversion (<15%) at 275 °C. The conversion (right axis) was measured in the conditions of (a). All the Pt loadings are 0.8 wt %. (c) NH3-TPD profiles of catalysts tested on equal catalyst weight. (d) XPS of Pt 4f and W 4f regions. Download figure Download PowerPoint To assess how the proximity of the acid and metal sites influenced catalytic performance, a series of additional catalysts were prepared and tested for heptane reforming under similar conditions (Figure 2b). The Pt/ZrO2 catalyst was prepared with the same synthetic procedure as Pt/WZrOx catalyst, except for the addition of tungsten precursors. In reactions where only one of either the metal or acid components were used (WZrOx, Pt/WOx, and Pt/ZrO2), the initial productivity was exceptionally poor, while Pt/WZrOx presented extremely high productivity up to 5.2 × 103 moli-C7/molPt/h. In this case, the Pt metal was responsible for catalyzing the initial dehydrogenation, with the subsequent skeletal isomerization occurring on adjacent acid sites. The Pt/WO3 catalyst was prepared by calcination of the tungsten precursor in a muffle furnace. It was well-crystallize