In situ synthesis of highly-active Pt nanoclusters via thermal decomposition for high-temperature catalytic reactions

Abstract

Noble metal nanoclusters are highly desirable for achieving superior catalytic performance in diverse catalytic reactions. However, they usual suffer from fast surface oxidation before use, which substantially degrades their activities. Herein, we address, for the first time, an in situ preparation strategy towards well-dispersed, ligand-free Pt nanoclusters (1.28 nm) supported on carbon spheres (CS) for high-temperature catalytic reactions. The nanoclusters are formed through thermal decomposition of platinum(II) acetylacetonate concomitantly with the preheating of the reactor and directly serve as catalysts for the dehydrogenation of methylcyclohexane (MCH) without deoxidization pretreatment. Nearly 97% of MCH is converted to hydrogen and toluene over the 0.68 wt% Pt/CS catalyst at 320 °C, and the highest hydrogen evolution rate reaches a value of 575 mmol g_met⁻¹ min⁻¹. Besides, the reaction equilibrium is achieved much faster than using catalysts prepared ex situ that require deoxidization of Pt species by hydrogen generated during the initial stage of the reaction.

---

Introduction

Platinum nanoparticles (NPs) loaded on carbon materials are essential to diverse catalytic reactions such as oxygen reduction, dehydrogenation, hydrogen evolution reaction, and methanol oxidation, to name a few. However, due to the low abundance and high cost of Pt, many attempts have been put forward to lowering Pt loading, while keeping their high catalytic activities. One possible solution is to enhance the mass activity by increasing the active surface area (m_Pt² g_Pt⁻¹),¹ which could be realized by reducing NPs sizes. Consequently, small Pt NPs (<5 nm) well dispersed on carbon materials, without any stabilizing agents on their surfaces are highly anticipated. Wet impregnation is a well-established method for preparing ligand-free, well-dispersed small noble metal NPs on supports, however, suffering from large size distribution.² More seriously, the ex situ synthesized small noble metal NPs by wet impregnation are easily to be oxidized in air before use, which will remarkably reduce their catalytic activities.³

In contrast, the catalysts in situ synthesized in the reactor during or before the catalytic reaction proceeds could effectively avoid fast surface oxidization. For example, ligand-free Ru and Rh nanoclusters were synthesized in situ by reducing Ru and Rh ions during the catalytic hydrolysis of ammonia-borane.^4,5 Nevertheless, in situ synthesis of catalysts for high-temperature catalytic reactions has been rarely reported. We note that a number of previous studies have demonstrated that thermal decomposition of noble metal acetylacetonates could lead to the formation of small noble metal NPs. For instance, Pd NPs (3.5–7 nm) were prepared by thermal decomposition of palladium(II) acetylacetonate (Pd(acac)₂) in organic solution (i.e., trioctylphosphine, TOP).⁶ Highly dispersed Pt NPs supported on carbon nanotubes were synthesized by thermal decomposition of Pt(acac)₂ under Ar gas flow, and the carbonyl groups were considered responsible to the catalytic decomposition.⁷ Therefore, we speculate that, for high-temperature catalytic reactions, highly active Pt catalysts could be synthesized in situ via thermal decomposition of acetylacetonate during the preheating process of reactor, and directly catalyze the subsequent high-temperature catalytic reactions, effectively avoiding fast surface oxidation of noble metal NPs in air before actual use.

In this work, ligand-free, well-dispersed, ultrafine Pt nanoclusters loaded on carbon spheres were prepared in situ by thermal decomposition of Pt(acac)₂ concomitantly with the preheating of reactor and immediately served as catalyst for the following high-temperature catalytic dehydrogenation of liquid organic hydrides (LOH). The liquid organic hydrogen carriers, i.e., methylcyclohexane (MCH), can produce both pure hydrogen and an easily removable liquid product.⁸ Therefore the catalytic dehydrogenation of LOH is regarded as remarkably suitable for one-step production of pure hydrogen, which is a kind of promising energy for fuel-cell powered vehicles.^8–10 It is generally recognized that supported catalysts often exhibit different catalytic performance depending on the preparation of the catalysts.^11–13 The in situ synthesized catalysts in this work are highly efficient in hydrogen generation from dehydrogenation of MCH, without the need of any pre-reduction treatments. The reaction equilibrium is achieved much faster than using catalysts prepared ex situ that require deoxidization of Pt species by the generated hydrogen at the initial reaction stage. The proposed strategy could be easily generalized to other high-temperature catalytic reactions where Pt is the most efficient catalyst, such as catalytic oxidization of nitrogen monoxide in diesel exhaust,¹⁴ reforming of n-hexane,¹⁵ selective dehydrogenation of isobutane,¹⁶ oxidation of VOCs,¹⁷ hydrogenation of benzaldehydes and nitrobenzenes,¹⁸ dissociation of phenol,¹⁹ selective-chloronitrobenzene hydrogenation²⁰ and so on.

Experimental methods

Preparation of carbon spheres

The carbon spheres (CS) serving as catalyst supports were synthesized via pyrolysis of waste tire, following the same preparation procedures as in our previous work.²¹

Preparation of Pt(acac)₂/CS

1 g of CS was well suspended in 20 mL of deionized water under ultrasonication for 30 min. A certain amount of Pt(acac)₂ was dissolved in 30 mL of ethanol under ultrasonication and stirred for 30 min, and then was added dropwise into the suspension and stirred for 5 h to allow complete adsorption of Pt(acac)₂ molecules on the CS. Catalysts with different Pt loadings (0.18–0.68 wt%) were prepared by varying the initial amount of Pt(acac)₂.

Characterization

Transition electron microscope (TEM) images are obtained by using an FEI Technai G2 F20 transmission electron microscope equipped with a field-emission gun. Infrared spectra of samples were recorded using a Bruker Tensor 27 FT-IR spectrometer in the range of 400–4000 cm⁻¹. The X-ray photoelectron spectroscopy (XPS) spectra of samples were investigated by using a PHIQuantum 2000 scanning ESCA microprobe spectrometer. Thermal gravimetric analysis (TGA) (Netzsch STA 449F3) were performed at 5 °C min⁻¹ in argon atmosphere. ICP-MS (Agilent 7700X) was adopted to determine the Pt loadings of catalysts.

Catalytic tests

The catalytic dehydrogenation reaction of MCH was performed under atmospheric pressure in a fixed-bed flow reactor (10 mm inner diameter × 250 mm). In a typical reaction, 0.554 g Pt(acac)₂/CS catalyst was filled in the reactor without any pretreatments, and then the reactor was preheated at 320 °C (external temperature) for 30 min. Subsequently, MCH was injected to the reactor by using a micro feeder at a feed rate of 1.8 mL h⁻¹ under a nitrogen stream (5 mL min⁻¹) for 10 h. The nitrogen rate was controlled by a gas flow meter. The conversion of MCH was determined with a Rock GC 7800 gas chromatograph equipped with a flame ionization detector and an SE-54 column.

Results and discussion

In situ formation of Pt nanoclusters on CS at high temperatures

The Pt(acac)₂ adsorbed on carbon spheres (Pt(acac)₂/CS) with different Pt contents (0.18–0.68 wt%) determined by ICP-MS, in this study, serve directly as catalysts for the dehydrogenation of MCH at high temperatures (280–340 °C) without any pretreatments with H₂ flow. The decomposition temperature of Pt(acac)₂ complex on CS was determined by thermal analysis on Pt(acac)₂/CS with a relatively high Pt loading (2.83 wt%). The TGA data, as shown in Fig. 1, indicates a major weight loss event over the temperature range 240–300 °C, accompanied with a discernible endothermic peak in the differential scanning calorimeter (DSC) curve. The primary weight loss ∼2.8% is consistent with the weight fraction of the acac ligand in Pt(acac)₂/CS, which might be ascribed to the catalytically decomposition of Pt(acac)₂ on the surface of carbon spheres.⁷ Meanwhile, as seen in the FTIR spectra (Fig. S1†), the surface of carbon spheres is modified with carboxyl group (C=O; C–O), which could facilitate the dissociation of organic ligands and the crystallization of Pt NPs on carbon surfaces.²² The oxygen groups are also favourable to immobilize the NPs on carbon supports leading to considerably enhanced catalyst activity.²³ Therefore, it could be inferred that the decomposition of Pt(acac)₂ on CS would initiate at 240 °C, and the high temperatures (280–340 °C) for dehydrogenation reactions would allow Pt(acac)₂ on CS to decompose.

Fig. 1 TGA and DSC curves of Pt(acac)₂/carbon sphere (Pt(acac)₂/CS) composites with a Pt loading of 2.83 wt%.

In catalytic dehydrogenation reactions, the reactor is preheated at certain temperatures for 30 min. To further confirm a complete decomposition of Pt(acac)₂ and formation of Pt NPs on the CS supports during the preheating process of the reactor, the original Pt(acac)₂/CS composites were subjected to annealing at 320 °C for 30 min under the protection of nitrogen. The characteristic absorption bands for Pt(acac)₂ that are discernible in the FTIR spectrum of Pt(acac)₂/CS composites, completely disappear after the composites being annealed (Fig. S1†), suggesting a fully decomposition of Pt(acac)₂. The XPS Pt 4f core level of the annealed composites shows broad doublets peaks corresponding to Pt 4f_7/2 and 4f_5/2 (Fig. 2). After the XPS deconvolution, three-doublets binding energies (BE) were resolved which are assigned to the metallic (zero) oxidation state of Pt (BE = 71.50 and 74.80 eV), accounting for 31.8%, and the oxidation states, Pt(II) (BE = 72.40 and 75.70 eV) and Pt(IV) (BE = 73.09 and 76.39 eV).²¹ Table 1 lists the XPS peak positions and concentrations of all the Pt species. Noted that the XPS signals from Pt(II) locate at lower binding energies, as compared to that for Pt(acac)₂, which could be assigned to PtO or Pt(OH)₂. This suggests that after high temperature annealing of Pt(acac)₂/CS composites, Pt(acac)₂ completely disappears. Therefore, the XPS results verify again that the high-temperature annealing procedure could completely convert Pt(acac)₂ to the metallic Pt, while the oxidation states could originate from the oxidation of fine Pt nanoclusters in air before the XPS tests. The TEM images of the as-prepared samples show that fine nanoclusters form on the carbon supports with even distribution after annealing (Fig. 3a and b, S2 and S3a†), which would be beneficial to improve the catalytic activities for the dehydrogenation of cyclic hydrocarbon. The NPs display good crystallinity as seen in the high-resolution TEM (HRTEM) image (insets in Fig. 3a), where the lattice spacing of 0.23 nm corresponds to (111) crystal plane of cubic Pt crystals (ICDD-4-802). In order to precisely evaluate the size distribution of Pt NPs, we performed a statistical analysis on more than 300 NPs in the TEM images of the annealed sample with a higher Pt content (2.83 wt%). The Pt nanoclusters are well dispersed on the carbon supports with a higher density than those with low loadings (Fig. S3a†), and the size of nanoclusters is 1.28 ± 0.14 nm (Fig. S3b†). The EDX results also confirm the existence of Pt on the carbon spheres (Fig. S3c†).

Fig. 2 High-resolution Pt 4f XPS spectra for Pt(acac)₂, 2.83 wt% Pt(acac)₂/CS composites after being annealed at 320 °C for 30 min, and the original composites after catalytic dehydrogenation reaction at 320 °C. The experimental data are indicated by black dotted lines. The dashed lines for the metallic (zero) oxidation state (Pt⁰), Pt(II) as in PtO or Pt(OH)₂, Pt(IV) as in PtO₂, and Pt(II) as in Pt(acac)₂ are shown in blue, green, magenta, and red, respectively.

Table 1 Binding energies of Pt 4f and relative content of Pt species calculated from XPS spectra of Pt(acac)₂, 2.83 wt% Pt(acac)₂/CS composites after being annealed at 320 °C for 30 min, and the original composites after catalytic hydrogenation reaction at 320 °C

Catalyst	Pt species	Pt 4f7/2 (eV)	Pt 4f5/2 (eV)	Area (%)
a Pt^2+ as in Pt(acac)2.
After annealing	Pt^0	71.50	74.80	31.8
	Pt(II)	72.40	75.70	24.2
	Pt(IV)	73.09	76.39	44.0
After dehydrogenation	Pt^0	71.40	74.70	67.7
	Pt(II)	72.41	75.71	32.3
Pt(acac)2	Pt(II)^a	72.77	76.07	100

---

Fig. 3 TEM and HRTEM (insets) images of the Pt nanoclusters supported on CS with different Pt contents. (a) 0.45 wt% Pt(acac)₂/CS annealed at 320 °C for 30 min and (b) after hydrogenation. (c) 0.68 wt% Pt(acac)₂/CS annealed at 320 °C for 30 min and (d) after hydrogenation. All scale bars in the insets are 5 nm.

Catalytic activity for high-temperature catalytic dehydrogenation reaction

The above characterizations suggest that the Pt nanoclusters could be created on the carbon supports via in situ thermal decomposition of Pt(acac)₂ concomitantly with the preheating process of the high-temperature catalytic reactions and, thereafter, served as catalyst for the catalytic reactions. Herein, we chose the dehydrogenation of MCH as a model reaction to study the catalytic activities of these in situ synthesized Pt nanoclusters.

The conversion of MCH over catalysts with different Pt loadings at 320 °C are shown in Fig. 4a. As the Pt content doubles (from 0.18 wt% to 0.37 wt%) at lower loadings, the conversion also roughly doubles in value (from 41% to 87%). For catalysts with higher Pt loadings, the catalytic activity promotes much slower, i.e., the conversion of MCH over 0.37 wt% and 0.45 wt% Pt/CS are nearly identical, almost 87% MCH are converted into hydrogen and toluene, with nearly 100% selectivity towards dehydrogenation. The highest conversion of 97% is achieved for the 0.68 wt% Pt/CS catalyst at 320 °C. Meanwhile, the hydrogen evolution rate reaches a maximum value of 575 mmol g_met⁻¹ min⁻¹ over 0.37 wt% catalyst (Fig. 4b). In Table S1,† we compare the activity of the Pt-based catalyst prepared by the in situ method with those synthesized via the impregnation approach in the literatures.^21,24–26 It suggests that the in situ synthesized catalyst exhibits superior catalytic performance, since it could realize a considerably high hydrogen evolution rate with a much lower Pt loading. It should be also noted that, for all the prepared catalysts, the conversion of MCH increases slightly with the reaction process and reaches reaction equilibrium after reaction for 5 h, much faster than the ex situ prepared catalysts.²¹ Since most of the Pt ions have been already deoxidized by thermal decomposition of Pt(acac)₂ and form Pt nanoclusters during the preheating process, the in situ synthesized catalysts could certainly afford faster arrival to reaction equilibrium than the ones prepared ex situ that require further reduction of the oxidization states of Pt and thus consume some amount of hydrogen generated at the initial stage of reaction. The slight rising of conversion with reaction time could be attributed to the deoxidization of Pt species residue on the carbon supports.

Fig. 4 The conversion of MCH (a) and hydrogen evolution rates (b) at different reaction time at 320 °C over catalysts with different Pt contents. (c) The variation of conversion of MCH with reaction time over 0.45 wt% Pt/CS at different temperatures. (d) The conversion of MCH at different initial feed rates over 0.45 wt% Pt/CS at different temperatures.

After the catalytic reaction, the remaining catalysts were recollected and characterized by TEM. As shown in Fig. 3b and d, the Pt nanoclusters are still well-dispersed, in favor of long-term stability. The particles size is about 2.5 nm, larger than that of the annealed samples (1.28 nm), probably owning to the much longer heating time in the catalytic reaction (i.e., up to 10 h). Note that the Pt⁰ concentration is higher in the sample after dehydrogenation reaction than that in the annealed samples (Fig. S2,† Table 1), which could be ascribed to the slower oxidation of larger nanoclusters in air before XPS tests. The long-term catalytic stability of the catalyst with a Pt loading of 0.68 wt%, which shows the highest conversion efficiency of MCH, was also examined. The result suggests that the in situ synthesized catalyst is highly durable, since its catalytic activity is kept at almost the same level even after 72 h reaction (Fig. S4†).

Because the dehydrogenation is an endothermic reaction, the reaction temperature is expected to have a strong effect on the catalytic activities. As seen in Fig. 4c, nearly 100% of MCH are converted into hydrogen and toluene over the 0.45 wt% catalyst at 340 °C. However, as the reaction temperature drops, the conversion of MCH decreases significantly and reaches ∼50% at 280 °C. The temperature-dependent conversion (Fig. 4c) allows us to estimate the activation energy (E_a) for the dehydrogenation of MCH over the catalysts, based on the Arrhenius equation.²⁷ It was found that E_a = 47.1 kJ mol⁻¹ for the in situ prepared Pt/CS, which is much lower than most of the catalysts synthesized by conventional methods.^27–29 Meanwhile, it is very important for practical applications to obtain a high conversion at a high gas hourly space velocity (GHSV). The conversion of MCH over 0.45 wt% Pt/CS at higher GHSV, up to 5.4 mL h⁻¹ was also investigated at different temperatures. The results illustrate that the catalyst exhibits high catalytic activity even at high reactant feed rates (Fig. 4d).

Conclusions

In summary, we develop a novel strategy for in situ preparation of well-dispersed, ligand-free Pt nanoclusters loaded on carbon supports for high-temperature catalytic reactions. The formation of nanoclusters occurs in the process of reactor preheating, and the as-prepared pure Pt nanoclusters could, afterwards, directly catalyze the reactions, effectively avoiding fast surface oxidation of noble metal nanoclusters synthesized ex situ. The in situ prepared catalysts demonstrate high activities in the high-temperature catalytic dehydrogenation reaction. Consequently, our work suggests that the in situ synthesis approach could not only greatly simplify the preparation procedures of noble metal nanoclusters, but also, promote the catalytic activities for a variety of high-temperature catalytic reactions.

Acknowledgements

This work was supported by the National Key Basic Research Program of China (2014CB931703), Natural Science Foundation of China (51271129), Natural Science Foundation of Tianjin City (14JCYBJC17200, 08JCZDJC21400), Program for New Century Excellent Talents in University (NCET-13-0414).

Notes and references

1. S. Choi, M. Shao, N. Lu, A. Ruditskiy, H.-C. Peng, J. Park, S. Guerrero, J. Wang, M. J. Kim and Y. Xia, ACS Nano, 2014, 8, 10363.
2. H. Jiang and Q. Xu, Catal. Today, 2011, 170, 56.
3. M. Chandra and Q. Xu, J. Power Sources, 2007, 168, 135.
4. H. Liang, G. Chen, S. Desinan, R. Rosei, F. Rosei and D. Ma, Int. J. Hydrogen Energy, 2012, 37, 17921.
5. Q. Yao, Z.-H. Lu, Y. Jia, X. Chen and X. Liu, Int. J. Hydrogen Energy, 2015, 40, 2207.
6. S.-W. Kim, J. Park, Y. Jang, Y. Chung, S. Hwang and T. Hyeon, Nano Lett., 2003, 3, 1289.
7. C. Mercado-Zúñiga, J. R. Vargas-García, M. A. Hernández-Pérez, M. Z. Figueroa-Torres, F. Cervantes-Sodi and L. M. Torres-Martínez, J. Alloys Compd., 2014, 615, S538.
8. F. Alhumaidan, D. Cresswell and A. Garforth, Energy Fuels, 2011, 25, 4217.
9. F. Alhumaidan, D. Tsakiris, D. Cresswell and A. Garforth, Int. J. Hydrogen Energy, 2013, 38, 14010.
10. D. Sebastian, C. Alegre, L. Calvillo, M. Perez, R. Moliner and M. J. Lazaro, Int. J. Hydrogen Energy, 2014, 39, 4109.
11. J. D. Scholten, M. H. G. Prechtl and J. Dupont, ChemCatChem, 2010, 2, 1265.
12. M. Amende, C. Gleichweit, T. Xu, O. Hodert, M. Koch, P. Wasserscheid, H. P. Steinruck, C. Papp and J. Libuda, Catal. Lett., 2016, 146, 851.
13. M. Sobota, I. Nikiforidis, M. Amende, B. S. Zanon, T. Staudt, O. Hofert, Y. Lykharch, C. Papp, W. Hieringer, M. Laurin, D. Assenbaum, P. Wasserscheid, H. P. Steinruck, A. Gorling and J. Libuda, Chem.–Eur. J., 2011, 17, 11542.
14. J. Després, M. Elsener, M. Koebel, O. Kröcher, B. Schnyder and A. Wokaun, Appl. Catal., B, 2004, 50, 73.
15. A. Kwangjin, Q. Zhang, S. Alayoglu, N. Musselwhite, J.-Y. Shin and G. A. Somorjai, Nano Lett., 2014, 14, 4907.
16. G. Wang, C. Li and H. Shan, ACS Catal., 2014, 4, 1139.
17. H.-J. Joung, J.-H. Kim, J.-S. Oh, D.-W. You, H.-O. Park and K.-W. Jung, Appl. Surf. Sci., 2014, 290, 267.
18. H. Pan, X. Li, Y. Yu, J. Li, J. Hu, Y. Guan and P. Wu, J. Mol. Catal. A: Chem., 2015, 399, 1.
19. G. Li, J. Han, H. Wang, X. Zhu and Q. Ge, ACS Catal., 2015, 5, 2009.
20. P. Zhang, Y. Hu, B. Li, Q. Zhang, C. Zhou, H. Yu, X. Zhang, L. Chen, B. Eichhorn and S. Zhou, ACS Catal., 2015, 5, 1335.
21. C. Zhang, X. Liang and S. Liu, Int. J. Hydrogen Energy, 2011, 36, 8902.
22. C. Liang, W. Xia, M. van den Berg, Y. Wang, H. Soltani-Ahmadi, O. Schlüter, R. A. Fischer and M. Muhler, Chem. Mater., 2009, 21, 2360.
23. Q. Yao, Z.-H. Lu, Y. Jia, X. Chen and X. Liu, Int. J. Hydrogen Energy, 2015, 40, 2207.
24. N. Kariya, A. Kukuoka, T. Utagawa, M. Sakuramoto, Y. Goto and M. Ichikawa, Appl. Catal., A, 2003, 247, 247.
25. A. A. Shukla, P. V. Gosavi, J. V. Pande, V. P. Kumar, K. V. R. Chary and R. B. Biniwale, Int. J. Hydrogen Energy, 2010, 35, 4020.
26. A. Shukla, J. V. Pande and R. B. Biniwale, Int. J. Hydrogen Energy, 2012, 37, 3350.
27. Z. Wu, L. Luo, G. Deng, W. Xiong and Z. Zhong, Ind. Catal., 2010, 9, 45.
28. K. Oda, K. Akamatsu, T. Sugawara, R. Kikuchi, A. Segawa and S. Nakao, Ind. Eng. Chem. Res., 2010, 49, 11287.
29. Z. Kou, Z. Zhi, G. Xu, Y. An and C. He, Appl. Catal., A, 2013, 467, 196.

---

Footnotes

† Electronic supplementary information (ESI) available: FTIR characterization, additional TEM image, and EDS result. See DOI: 10.1039/c6ra04681a

‡ These authors contribute equally to this work.

---