A high-performance Pt–Co bimetallic catalyst with polyethyleneimine decorated graphene oxide as support for hydrolysis of ammonia borane

Abstract

A series of Pt_xCo_1−x bimetallic nanoparticles (NPs) were deposited on polyethyleneimine (PEI) decorated graphene oxide (GO) by a simple co-reduction method. The PEI molecules facilitated the uniform distribution of bimetallic NPs (∼2.3 nm) on GO. PEI–GO/Pt_0.17Co_0.83 showed extraordinary catalytic properties: a total turnover frequency of 377.83 mol_H2 min⁻¹ mol⁻¹_metal, a hydrogen generation rate of 111.28 L_H2 min⁻¹ g⁻¹_metal at 298 K, an activation energy of 51.6 kJ mol⁻¹ and good recyclability (∼80% of the original activity after five cycles). The synergistic effect between Pt and Co, and the small size of these NPs play important roles in improving their catalytic properties. Such low-cost, recyclable bimetallic catalysts will enable many practical applications in portable devices and fuel cells.

---

Introduction

Hydrogen is a clean, effective energy carrier that can meet the increasing demands for energy. It can be used in fuel cells and portable devices^1–3 that require high hydrogen capacity materials and high real-time production efficiency, and has thus created a strong interest in recent years.^4–6 Ammonia borane (NH₃BH₃, AB) is one of the most promising chemical hydrogen storage materials, and offers many advantages, such as low molecular weight, high hydrogen capacity (19.6 wt%), non-toxicity, high water solubility and chemical stability.^7,8 Compared with thermolysis, hydrolysis of AB can be completed at lower reaction temperatures and dehydrogenation occurs at faster rates.⁹ Also, it can release hydrogen in a sustained manner when proper catalysts are employed.¹⁰ This makes performance optimization of the catalysts a critical prerequisite for improving dehydrogenation efficiency of AB hydrolysis.

Noble metals^11,12 and non-noble metals^13–15 have long been studied for the catalytic hydrolysis of AB. However, noble metal catalysts usually suffer from resource and cost limitations whereas non-noble metal catalysts only have moderate activity and stability. Compared to monometallic catalysts, much better selectivity and catalytic activity have recently been demonstrated for bimetallic catalysts.^16,17 In bimetallic systems two effects are found to play important roles. The first one is the geometric effect, in which the coordination of one metal to the other provides new active sites. The other effect is the electronic effect, for which the addition of one metal alters the electron properties of the other metal because of electron transfer. It is difficult to distinguish these two effects, and both often operate synergistically.¹⁸ Because of such synergistic effects, bimetallic catalysts can show significantly greater catalytic activity than can monometallic catalysts, even at lower concentrations.¹⁹ Controlling the morphology of bimetallic nanoparticles (NPs) is important for optimizing their performance, and different structures such as core–shell structures, heterostructures and alloy structures have been designed to make the most of the bimetallic synergistic effect. The form of bimetallic systems also improves the stability of catalysts. For example, Jiang et al.²⁰ synthesized Au–Ni@SiO₂ catalysts that showed better catalytic activities than did the corresponding monometallic catalysts. Wen et al.²¹ prepared hexagonal NiCo–Pt nanoplate catalysts to achieve remarkably improved catalytic activities. Xu et al.²² demonstrated improved catalytic activity and recyclability of PdPt cubic NPs. These bimetallic catalysts with controlled composition, size and morphology can find practical applications for hydrolysis of AB.

It should be noted, however, that particles with high specific surface areas (SSA), such as small bimetallic NPs, have a high propensity to aggregate. To optimize their catalytic activity and recyclability, employment of a suitable support is necessary for preventing aggregation while still maximizing their SSA and thus, their catalytic activity. Graphene, a two-dimensional carbon film with single-atom thickness, has emerged as a promising candidate because of its high SSA and extraordinary properties.^23,24 Also, it may interact directly with metal NPs to enhance the electron-transfer efficiency and catalytic activity.²⁵ So far, there have been many studies dedicated to developing graphene-based catalysis.^26,27 However, the size and spatial distributions of the resulting metal NPs on graphene are often not well controlled.²⁸ To overcome this shortcoming, various organic functional polymers have been used to tune the morphology and spatial distribution of metal NPs deposited on graphene. The functional groups of the polymers can immobilize metal ions, leading to good dispersion of metal NPs after reduction.^29,30 In our previous study,³¹ branched polyethyleneimine (PEI) was found to optimize the morphology and enhance the performance of metal NPs deposited on graphene oxide (GO), which greatly helps increase the activity of the catalyst.

In this work we have made full use of the synergistic effect between Pt and Co and the uniform dispersion of small-sized metal NPs on PEI decorated GO to significantly improve the activity and stability of catalysts. We synthesized a series of Pt_xCo_1−x bimetallic NPs deposited on PEI–GO and examined their catalytic activity and recyclability for AB dehydrogenation. We found that their catalytic performance shows a “volcano shape” relationship with changing Pt/Co ratios. The activity of the composite catalyst at a Pt/Co ratio of 0.2, i.e., PEI–GO/Pt_0.17Co_0.83, was observed to be superior to all the other reported Pt-based bimetallic catalysts, exhibiting a total turnover frequency (TOF) value of 377.83 mol_H2 min⁻¹ mol⁻¹_metal, a maximum hydrogen generation rate of 111.28 L_H2 min⁻¹ g⁻¹_metal at room temperature and good recyclability. Thus, this catalyst is suitable for further applications in fuel cells and portable devices.

Experimental section

Chemicals

Graphite powder 8099200 (120 μm) was purchased from Qingdao BCSM. 98% H₂SO₄, 65% HNO₃, 30% H₂O₂, 96.0% sodium borohydride and 98% H₂PtCl₆·6H₂O were obtained from Sinopharm Chemical Reagent Company. Sodium nitrate was obtained from Shanghai Qiangshun Chemical Co., KMnO₄ was obtained from Shanghai Zhenxing Chemical Company, PEI (M_w = 600, branched) and cobalt(II) acetate tetrahydrate were purchased from Alfa Aesar, and ammonia borane complex (97%) was obtained from Aldrich. All chemicals were used as received. Deionized water was used in all experiments.

Characterization

Transmission electron microscopy (TEM, Jeol JEM-2100F and Tecnai G² TF20 Twin, both operating at 200 kV) was used to observe the morphology of PEI–GO/Pt_xCo_1−x at different magnifications. All the samples for imaging were prepared by depositing aqueous dispersions (∼0.2 mg mL⁻¹) on holey copper grids. Fourier transform infrared spectroscopy (FT-IR) was performed on a Thermo-Nicolet NEXUS 670 spectrometer at room temperature over a frequency range of 3750–750 cm⁻¹. Thermogravimetric analysis (TGA) was conducted with a PerkinElmer Pyris 1 TGA from ambient temperature to 800 °C at a heating rate of 20 °C min⁻¹ with nitrogen as a purge gas. X-ray photoelectron spectra (XPS) was carried out on an Kratos AXIS Ultra DLD system with monochromatic Al Kα radiation (hv = 1486.6 eV). X-ray diffraction (XRD) experiments were carried out using a PANalytical X'Pert PRO X-ray diffractometer with Cu Kα radiation (λ = 1.54 Å) at a scanning rate of 14° min⁻¹ in a 2θ range from 30° to 80°.

Preparation of the catalysts

GO was prepared by the Hummers' method and freeze-dried before use. Then GO (50 mg) was dissolved in deionized water (10 mL) with ultrasonic mixing, and then deionized water (10 mL) containing PEI (1 g) was slowly added. The mixture was stirred magnetically at 60 °C for 24 h to synthesize PEI–GO. The mixture was then purified by centrifugation and freeze-dried to constant weight. PEI–GO supported catalysts were prepared by the deposition of metal NPs on PEI–GO. Typically, as-prepared PEI–GO (0.5 mg) was dissolved in deionized water (10 mL) with ultrasonic mixing in a two-necked round bottomed flask, to which Co(Ac)₂·4H₂O (2.5 μmol) and H₂PtCl₆·6H₂O (0.5 μmol) were then added. The mixture was stirred for 15 min under a N₂ atmosphere. Then NaBH₄ (5 mg) was added into the mixture to synthesize the PEI–GO/Pt_0.17Co_0.83 catalyst. The collected product was washed with deionized water, ethanol and dried in a two-necked flask under the a nitrogen atmosphere.

To carry out the comparative study, the concentration of H₂PtCl₆·6H₂O was fixed in all experiments, while changing the concentration of Co(Ac)₂·4H₂O to produce normalized Pt/Co ratios at 1 : 1, 1 : 2.5, 1 : 5, 1 : 10 and 1 : 15. The ratio of PEI–GO to metal precursor (0.5 mg PEI–GO to 3 μmol metal precursor) was also fixed to avoid a possible influence on the deposition of NPs. This design led to the synthesis of PEI–GO/Pt_0.5Co_0.5, PEI–GO/Pt_0.29Co_0.71, PEI–GO/Pt_0.17Co_0.83, PEI–GO/Pt_0.09Co_0.91 and PEI–GO/Pt_0.06Co_0.94 catalysts. Monometallic PEI–GO/Pt and PEI–GO/Co were also prepared by the same method, except that a higher concentration of metal precursor was used (10 μmol for Pt and 50 μmol for Co).

Catalytic hydrolysis of AB

The as-prepared PEI–GO/Pt_xCo_1−x was dispersed into an 8 mL aqueous solution in a two-necked round-bottom flask. AB (1.1 mmol) dissolved in deionized water (2 mL) was injected into the solution with a syringe under stirring (1200 rpm). The reaction was carried out under ambient conditions. The volume of the H₂ generated was monitored with a gas burette. To study the effect of reaction temperature, the AB dehydrogenation experiment was performed at 15, 25, 35 and 45 °C.

Recycling tests of the catalysts

After the dehydrogenation reaction was completed, the catalyst was collected at the bottom of the flask with a magnet and the supernatant solution was carefully removed. After washing, deionized water (8 mL) was added to the flask and then deionized water (2 mL) containing AB (1.1 mmol) was injected into the solution for recyclability tests. The same process was repeated five times.

Results and discussions

The PEI–GO support was prepared by a modified method from our previous work.³¹ In brief, GO was first dispersed in water followed by dropping desired amounts of PEI (M_w = 600) aqueous solution. After stirring for 24 h at 60 °C, the color of the suspension changed from brown to dark grey, indicating a partial reduction of GO.³² A large number of amine groups on the branched PEI enabled them to be adsorbed or covalently linked to GO. Because of the presence of carboxyl groups, GO can bind with positively charged PEI molecules using electrostatic interactions with the amine groups and form amide bonds.³³

As shown in Fig. 1a, the weight of GO decreased sharply at about 200 °C, which is attributed to the thermal removal of adsorbed water and parts of the oxygen-containing groups. For PEI, its very large weight loss between 300–400 °C originates from the decomposition of its polymer skeleton. By comparison, PEI–GO underwent a weight loss of about 40% in this temperature range and it was speculated that a significant amount of PEI (∼40%) was deposited on GO. In Fig. 1b, a small shoulder around 1660 cm⁻¹ would originate from amide bonds of PEI–GO, suggesting that parts of the PEI molecules form covalent linkage with GO. XPS was utilized to identify the chemical composition of PEI–GO. In Fig. 1d the C_1s core level is fitted with five Gaussian peaks at 284.6, 285.6, 286.8, 287.8 and 289.0 eV, which can be ascribed to C–C, C–N, C–O, C=O and N–C=O species, respectively. The existence of a N–C=O species means that PEI molecules are confined by GO through covalent bonding interactions. It is seen that the C–N species are the major functional groups (41.4%) because of the deposition of PEI molecules after the reaction. This observation is in agreement with the N_1s spectrum in Fig. 1e, where ∼80.0% of the amine groups (at 399.0 eV) are observed for PEI–GO. This large number of amine groups enables the coordination interaction between the metal ions and the amine groups, so that most of the NPs could be confined within the polymer framework during the subsequent fast reduction, and result in the uniform distribution of NPs on PEI–GO.

Fig. 1 (a) TGA curves and (b) FT-IR spectra of GO, PEI and the PEI–GO composite; XPS spectra of (c) PEI–GO and the corresponding (d) C_1s, (e) N_1s and (f) O_1s core levels for PEI–GO.

TEM was then used to observe the morphology of the as-prepared PEI–GO/Pt_0.17Co_0.83. Fig. 2a shows the microscopic structure of PEI–GO/Pt_0.17Co_0.83 where obvious folds and corrugations are visible, indicating that PEI covers the GO surface.³¹ Fig. 2b displays numerous dark Pt and Co NPs that have been deposited on PEI–GO. They are all distributed compactly, which helps to prevent the aggregation of graphene nanosheets. Fig. 2c clearly shows the sizes of metal NPs. The size distribution of metal NPs was calculated from at least 100 NPs, which is shown in Fig. 2d. Because most of the metal ions are strongly restricted within the PEI layers (heterogeneous nucleation effect), it is hard for homogeneous nucleation to take place in the bulk solution, which results in the formation of small metal NPs (∼2.3 nm). The uniform dispersion of Pt and Co NPs on PEI–GO is also confirmed by the STEM dark field image in Fig. 2e. Because Co (Z = 27) and Pt (Z = 78) have a much higher contrast than low Z-number elements such as C, O, N, the bright spots in the dark field image reflect the location of metal NPs on PEI–GO. The corresponding elemental mapping images are shown in Fig. 2f and g, where the red spots represent Co NPs and green spots represent Pt NPs. The ratio of Pt to Co is calculated as ∼1 : 4 according to Fig. 2h, which is close to the feed ratio (1 : 5). This is consistent with the speedy nucleation and growth of metal NPs (<10 s) under a strong reductant with the help of the support. Based on these results, it was concluded that Pt and Co are well dispersed in PEI–GO/Pt_0.17Co_0.83, and it was expected to yield an improved performance for catalytic hydrolysis of AB at room temperature. XRD patterns of PEI–GO/Pt_xCo_1−x with different x values are shown in Fig. S1.† There are no obvious peaks in the 2θ range from 30° to 80°, indicating the amorphous character of these metal NPs.

Fig. 2 Microscopy of PEI–GO/Pt_0.17Co_0.83. (a and b) TEM and (c) high resolution transmission electron microscopy images at different magnifications; (d) size distribution for metal NPs; (e–h) scanning transmission electron microscopy (STEM) dark field image, elemental mapping images and corresponding energy-dispersive X-ray spectrum.

Fig. 3a shows the volume of H₂ generated using different catalysts, and also demonstrates the advantage of using PEI–GO as a support and provides evidence for the synergistic effect between the two metals. For Pt_0.17Co_0.83 without support, it took 9 min or more to complete the dehydrogenation. GO/Pt_0.17Co_0.83 showed an improved catalytic activity, completing the reaction in about 5 min. This is ascribed to the fact that GO improves the dispersion of the small metal NPs, leading to increased catalytic activity. In the absence of GO, PEI/Pt_0.17Co_0.83 also showed good catalytic activity, but slightly lower than that of PEI–GO/Pt_0.17Co_0.83. To determine the role of GO and PEI in the catalytic hydrolysis of AB, a recyclability test of PEI/Pt_0.17Co_0.83 was carried out, as shown in Fig. S2.† The amine groups of PEI molecules can immobilize metal ions, and can tune the size of metal NPs and increase the catalytic activity. But with PEI alone as a support, the recyclability of the catalyst is restricted, because of the aggregation and encapsulation of the metal NPs. The role that GO plays in tuning the morphology of PEI and metal NPs is important, resulting in increased recyclability and better catalytic activity. The advantage of GO compared with other carbon materials is also shown in Fig. S4.† Compared with other carbon materials such as activated carbon and multi-wall carbon nanotubes, GO is better at tuning the morphology of PEI and metal NPs. For PEI–GO/Pt_0.17Co_0.83, the reaction ended in about 3 min, implying a three-fold increase in catalytic activity with the addition of PEI–GO. Apparently, the uniform distribution of small metal NPs (∼2.3 nm) on PEI–GO is responsible for this enhancement because of the significantly improved accessibility of the SSAs. In addition, the synergistic effect between Pt and Co could be another key factor for the increased catalytic activity. As shown in Fig. 3b, monometallic PEI–GO/Pt and PEI–GO/Co catalysts took ∼3 min and ∼5 min to complete the dehydrogenation, respectively, even though their precursor concentrations are 20 times higher than the bimetallic counterparts. The low catalytic activity of a physical mixture of PEI–GO/Pt and PEI–GO/Co also demonstrates the existence of the synergistic effect between Pt and Co as shown in Fig. S5.†

Fig. 3 Hydrogen generation from AB with different catalysts at 25 ± 0.5 °C: (a) Pt_0.17Co_0.83, GO/Pt_0.17Co_0.83, PEI/Pt_0.17Co_0.83 and PEI–GO/Pt_0.17Co_0.83. (b) PEI–GO/Pt and PEI–GO/Co. (c) PEI–GO/Pt_xCo_1−x with (1) x = 0.5; (2) x = 0.29; (3) x = 0.17; (4) x = 0.09, and (5) x = 0.06 and (d) the corresponding plots of TOF versus x (x = 1, 0.5, 0.29, 0.17, 0.09, 0.06 and 0) for PEI–GO/Pt_xCo_1−x. [AB] = 110 mM. For monometallic PEI–GO/Pt and PEI–GO/Co, [Pt] = 1 mM and [Co] = 5 mM; whereas for bimetallic catalysts, [Pt] = 0.05 mM. [Co] is calculated from the normalized Pt/Co ratios. Volume of the dispersion is 10 mL.

To determine the optimal ratio of Pt/Co for PEI–GO/Pt_xCo_1−x, a series of composite catalysts with a constant Pt content were prepared. Their catalytic activities show a volcano shaped change with increasing Pt/Co ratios, as shown in Fig. 3c and d, and reach a peak value at a Pt/Co ratio of 0.2. Further increasing the Co content ([Co]) leads to a sharp decrease in activity. It appears that when [Co] is low, the synergistic effect is not a dominating factor, but with an increase of [Co], the synergistic effect gradually becomes dominant and thus greatly improves the catalytic activity. When the Pt/Co ratio exceeds 0.2, a large amount of Co NPs may form in the vicinity of Pt NPs and result in the occurrence of some core–shell like structures. This in turn may block active surface sites of Pt NPs and decrease the catalytic activity, because the activity of Co is lower than that of Pt.⁴³

Among our catalysts, PEI–GO/Pt_0.17Co_0.83 exhibits the highest catalytic activity for AB dehydrogenation, with a TOF value of 377.83 mol_H2 min⁻¹ mol⁻¹_metal and a H₂ generation rate of 111.28 L_H2 min⁻¹ g⁻¹_metal. This is an extremely high catalytic activity and may fulfill the application requirements for fuel cells and portable devices. Compared with other reported noble metal-based catalysts (see Table 1) and among Pt-based bimetallic catalysts, our catalyst shows the highest activity, to the best of our knowledge. It is believed that this extraordinary performance originates from the synergistic effect between Pt and Co NPs and the effect of the small size of the NPs. Next, PEI–GO/Pt_0.17Co_0.83 was chosen to evaluate the kinetics, activation energy and recyclability of this class of catalysts.

Table 1 Comparison of catalytic activity and activation energy of various noble metal-based bimetallic catalysts for the hydrolysis of AB

Catalysts	TOF^a molH2 min^−1 mol^−1metal	Activation energy kJ mol^−1	Ref.
a For the core–shell structure, only the shell metals are taken into consideration of TOF, while all the metal is considered for the other structures.
PEI–GO/Pt0.17Co0.83	377.83	51.6	This study
Ni0.33@Pt0.67/C	166.91	33.0	34
Co0.32@Pt0.68/C	147.63	41.5	35
PdPt cubic NPs	50.02	21.8	22
Pt/CeO2/RGO	48	—	36
Ru@Ni/graphene	45.27	36.6	25
Pd@Co/graphene	37.44	—	37
Ni16Co80/Pt4	30.77	45.7	38
Pt0.65Ni0.35	28.82	39	39
Co35Pd65/C	22.7	27.5	19
RuCo/γ-Al2O3	12.69	52	40
Pt/γ-Al2O3	222.22	21	41
Pt-MIL	211.27	40.7	42

---

Studying the kinetics of AB hydrolysis provides valuable information on what experimental factors control the rate of hydrogen generation. The effect of the amount of metal on the hydrogen generation rate was investigated by measuring the hydrolysis of 110 mM AB solution at 25 °C with varying metal concentrations (0.3, 0.6, 0.9 and 1.2 mM) while keeping other factors constant. The corresponding results are presented in Fig. 4a. The hydrogen generation rate r for different amounts of metal in the catalysts was calculated from the linear part of each curve. For clarity, a plot of ln(r) versus ln([M]) is re-plotted in Fig. 4b. It is seen that ln(r) changes almost linearly with ln([M]), and the slope obtained is 0.84, suggesting that the PEI–GO/Pt_0.17Co_0.83 catalyzed AB hydrolysis is nearly a first-order reaction when compared to the metal concentration. This is consistent with previous reports,⁴⁴ indicating that the hydrogen generation rate is controlled by the surface reaction. Therefore, given that the apparent kinetic rate constant is proportional to the total surface area of all metal NPs,⁴⁵ the increased active sites on small metal NPs (∼2.3 nm) are expected to show better catalytic performance for hydrolysis dehydrogenation of AB.

Fig. 4 (a) The hydrolysis of AB by PEI–GO/Pt_0.17Co_0.83 at different metal concentrations, at 25 ± 0.5 °C and [AB] = 110 mM and (b) the corresponding plot of H₂ generation rate (ln(r)) versus ln([metal]). The volume of the dispersion is 10 mL.

To study the effect of temperature on AB hydrolysis, a series of AB hydrolysis experiments were carried out between 15 °C and 45 °C. As shown in Fig. 5a, the hydrogen generation rate was found to rapidly increase with the increase of temperature from 15 °C to 45 °C. ln(r) versus 1/T is re-plotted in Fig. 5b, which gives an apparent activation energy of 51.6 kJ mol⁻¹ according to the Arrhenius formula. This value is much lower than those of monometallic catalysts (e.g., 62 kJ mol⁻¹ for Co/γ-Al₂O₃ (ref. 46) and 56 kJ mol⁻¹ for Pd/zeolite⁴⁷) and bimetallic catalysts (e.g., 52.7 kJ mol⁻¹ for Ni–Ru nanocomposites⁴⁸), but still higher than other catalysts reported, as listed in Table 1. This implies that our catalysts may work much better at high temperatures, which is important for those applications in portable devices and fuel cells that often require medium operating temperatures.

Fig. 5 (a) The hydrolysis of AB catalyzed by PEI–GO/Pt_0.17Co_0.83 at different temperatures with [AB]/[metal] = 366.7, and (b) the corresponding plots of ln r versus the reciprocal absolute temperature 1/T. Volume of the dispersion is 10 mL.

Excellent stability and recyclability are important qualities for a catalyst for AB hydrolysis. Fig. 6a shows the catalytic activity of PEI–GO/Pt_0.17Co_0.83 after several cycles. The rate of hydrogen production remained almost constant at 3 mol H₂ per mol AB, but the completion time gradually increased with increasing cycles, indicating decreased catalytic activity. After the fifth cycle, the PEI–GO/Pt_0.17Co_0.83 still retained ∼80% of the initial catalytic activity. Fig. 6b shows the dispersion of metal NPs after the fifth cycle. There was no significant change in size and shape of the metal NPs, implying that the reduced activity was not because of the aggregation of these NPs. XRD patterns of PEI–GO/Pt_0.17Co_0.83 also show that no reaction induced crystalline material appeared during the recycling catalytic reaction, indicating the stable dispersion of metal NPs (Fig. S6†). To identify the valence state of Co and Pt in the catalyst, an XPS study of PEI–GO/Pt_0.17Co_0.83 before and after the catalytic hydrolysis of AB, was also conducted and the results are shown in Fig. 7. More than half of the Co was oxidized before the catalytic reaction, because of the exposure to air during the preparation of the XPS sample (Fig. 7a). According to Fig. 7c, the content of oxidized Co increased from 60% to ∼80% after the catalytic reaction, indicating the occurrence of Co oxidation during the AB hydrolysis.⁴⁹ In Fig. 7b, the peaks of zero valent Pt emerged at 71.5 eV and 74.7 eV for Pt 4f_7/2 and 4f_5/2. After the catalytic reaction, the peaks slightly shifted to 70.9 eV and 74.1 eV (Fig. 7d), which are closer to that of monometallic Pt (70.9 eV for 4f_7/2 and 75.1 eV for 4f_5/2).⁵⁰ This shift may have arisen from the oxidation of Co and the reduced synergistic effect between Pt and Co.²¹ However, Pt remained in the zero valence state after the catalytic reaction. These results indicate that during the catalytic reaction, Pt was stable, but Co was partially oxidized. Therefore, the decreased catalytic activity during the catalytic reaction would be caused mainly by the oxidation of Co, following the weakened synergistic effect between Pt and Co. Although somewhat reduced, the catalytic activity of PEI–GO/Pt_0.17Co_0.83 still remained high even after several cycles. This relative stability of activity makes this catalyst suitable for many practical applications.

Fig. 6 (a) Hydrogen generation from the hydrolysis of AB catalyzed by PEI–GO/Pt_0.17Co_0.83 at sequential runs at 25 ± 0.5 °C with [AB]/[metal] = 366.7, and (b) TEM image of PEI–GO/Pt_0.17Co_0.83 after five cycles. The volume of the dispersion is 10 mL.

Fig. 7 XPS spectra of Co_2p and Pt_4f (a and b) before and (c and d) after the catalytic hydrolysis of AB.

Conclusions

We have demonstrated the applicability of PEI–GO supported Pt_xCo_1−x composite catalysts for the hydrolytic dehydrogenation of AB. The PEI adsorbed on GO facilitated the control of the size and the uniform dispersion of metal NPs. Because of the anchoring effect of PEI to precursor metal ions, the resulting metal NPs were small, with typical dimensions of about 2.3 nm. By tuning the molar ratio of Pt/Co precursors, a series of PEI–GO/Pt_xCo_1−x composite catalysts were synthesized, among which PEI–GO/Pt_0.17Co_0.83 exhibited extraordinary catalytic activity compared to its monometallic counterparts. It is proposed that two key factors are responsible for the improved activity: the synergistic effect between the two metals and the small particle sizes (high SSA) of these metal NPs. We observed a volcano shaped relationship with varying Pt/Co ratios, as well as a sharp decrease in catalytic activity after the [Pt]/[Co] exceeded 0.2, which suggests a significant dependence of the synergistic effect on the ratios of the components of the bimetallic catalysts. In the catalytic kinetics of AB hydrolysis, a first-order reaction is believed to originate from the large SSAs of bimetallic NPs, given that the apparent kinetic rate constant is strictly proportional to the total surface areas of all metal NPs. The unprecedented TOF value (377.83 mol_H2 min⁻¹ mol⁻¹_metal), high hydrogen generation rate (111.28 L_H2 min⁻¹ g⁻¹_metal) and moderate activation energy (51.6 kJ mol⁻¹) exhibited by PEI–GO/Pt_0.17Co_0.83 in our experiments show that this novel catalyst is highly competitive compared to all other noble metal-based bimetallic catalysts. Also, it was shown that the PEI–GO/Pt_0.17Co_0.83 catalyst has a good magnetic recyclability and retained ∼80% of the initial activity after five cycles with no significant change in morphology. As a result, we believe that the high efficiency and low cost of PEI–GO/Pt_xCo_1−x are advantages that make it attractive for many practical applications such as portable devices and fuel cells.

Acknowledgements

We thank Ms Limin Sun and Ms Qianqian Hu (Shanghai Jiao Tong University) for their assistance with the XPS. This work was supported by National Basic Research Program of China (Grant no. 2011CB605702), NSF of China (Grant nos. 50573014, 50773012 and 51173027), Shanghai Nanotechnology Program (Grant no. 1052nm00400) and Shanghai Basic Research Program (grant no. 14JC1400600).

Notes and references

1. R. J. Keaton, J. M. Blacquiere and R. T. Baker, J. Am. Chem. Soc., 2007, 129, 1844–1845.
2. S. C. Amendola, S. L. Sharp-Goldman, M. S. Janjua, M. T. Kelly, P. J. Petillo and M. Binder, J. Power Sources, 2000, 85, 186–189.
3. Y. Kojima, Y. Kawai, H. Nakanishi and S. Matsumoto, J. Power Sources, 2004, 135, 36–41.
4. G. A. Deluga, J. R. Salge, L. D. Schmidt and X. E. Verykios, Science, 2004, 303, 993–997.
5. X. B. Chen, S. H. Shen, L. J. Guo and S. S. Mao, Chem. Rev., 2010, 110, 6503–6570.
6. C. W. Hamilton, R. T. Baker, A. Staubitz and I. Manners, Chem. Soc. Rev., 2009, 38, 279–293.
7. U. B. Demirci and P. Miele, Energy Environ. Sci., 2009, 2, 627–637.
8. F. H. Stephens, V. Pons and R. T. Baker, Dalton Trans., 2007, 2613–2626.
9. G. Wolf, J. Baumann, F. Baitalow and F. P. Hoffmann, Thermochim. Acta, 2000, 343, 19–25.
10. M. Chandra and Q. Xu, J. Power Sources, 2006, 156, 190–194.
11. S. Akbayrak, S. Gençtürk, İ. Morkan and S. Özkar, RSC Adv., 2014, 4, 13742–13748.
12. S. S. Mal, F. H. Stephens and R. T. Baker, Chem. Commun., 2011, 47, 2922–2924.
13. F. Y. Qiu, Y. L. Dai, L. Li, C. C. Xu, Y. N. Huang, C. C. Chen, Y. J. Wang, L. F. Jiao and H. T. Yuan, Int. J. Hydrogen Energy, 2014, 39, 436–441.
14. Y. W. Yang, Z. H. Lu, Y. J. Hu, Z. J. Zhang, W. M. Shi, X. S. Chen and T. T. Wang, RSC Adv., 2014, 4, 13749–13752.
15. J. M. Yan, X. B. Zhang, S. Han, H. Shioyama and Q. Xu, Inorg. Chem., 2009, 48, 7389–7393.
16. A. K. Singh and Q. Xu, ChemCatChem, 2013, 5, 652–676.
17. L. Yang, J. Su, X. Y. Meng, W. Luo and G. Z. Cheng, J. Mater. Chem. A, 2013, 1, 10016–10023.
18. F. Tao, Chem. Soc. Rev., 2012, 41, 7977–7979.
19. D. H. Sun, V. Mazumder, Ö. Metin and S. H. Sun, ACS Nano, 2011, 5, 6458–6464.
20. H. L. Jiang, T. Umegaki, T. Akita, X. B. Zhang, M. Haruta and Q. Xu, Chem.–Eur. J., 2010, 16, 3132–3137.
21. M. Wen, S. Q. Zhou, Q. S. Wu, J. Y. Zhang, Q. N. Wu, C. X. Wang and Y. Z. Sun, J. Power Sources, 2013, 232, 86–92.
22. A. J. Amali, K. Aranishi, T. Uchida and Q. Xu, Part. Part. Syst. Charact., 2013, 30, 888–892.
23. S. Garaj, W. Hubbard, A. Reina, J. Kong, D. Branton and J. A. Golovchenko, Nature, 2010, 467, 190–193.
24. K. S. Novoselov, A. K. Geim, S. V. Morozov, D. Jiang, Y. Zhang, S. V. Dubonos, I. V. Grigorieva and A. A. Firsov, Science, 2004, 306, 666–669.
25. N. Cao, J. Su, W. Luo and G. Z. Cheng, Int. J. Hydrogen Energy, 2014, 39, 426–435.
26. P. Yang, S. Y. Jin, Q. Z. Xu and S. H. Yu, Small, 2013, 9, 199–204.
27. F. Y. Qiu, L. Li, G. Liu, Y. J. Wang, C. H. An, C. C. Xu, Y. N. Xu, Y. Wang, L. F. Jiao and H. T. Yuan, Int. J. Hydrogen Energy, 2013, 38, 7291–7297.
28. M. Fang, Z. X. Chen, S. Z. Wang and H. B. Lu, Nanotechnology, 2012, 23, 085704.
29. H. Bai, C. Li and G. Q. Shi, Adv. Mater., 2011, 23, 1089–1115.
30. G. Goncalves, P. A. A. P. Marques, C. M. Granadeiro, H. I. S. Nogueira, M. K. Singh and J. Grácio, Chem. Mater., 2009, 21, 4796–4802.
31. X. H. Zhou, Z. X. Chen, D. H. Yan and H. B. Lu, J. Mater. Chem., 2012, 22, 13506–13516.
32. Y. Zhang, B. Chen, L. M. Zhang, J. Huang, F. H. Chen, Z. P. Yang, J. L. Yao and Z. J. Zhang, Nanoscale, 2011, 3, 1446–1450.
33. Y. Zhang, B. Chen, L. Zhang, J. Huang, F. Chen, Z. Yang, J. Yao and Z. Zhang, Nanoscale, 2011, 3, 1446–1450.
34. X. J. Yang, F. Y. Cheng, J. Liang, Z. L. Tao and J. Chen, Int. J. Hydrogen Energy, 2011, 36, 1984–1990.
35. X. J. Yang, F. Y. Cheng, Z. L. Tao and J. Chen, J. Power Sources, 2011, 196, 2785–2789.
36. X. Wang, D. P. Liu, S. Y. Song and H. J. Zhang, Chem.–Eur. J., 2013, 19, 8082–8086.
37. J. Wang, Y. L. Qin, X. Liu and X. B. Zhang, J. Mater. Chem., 2012, 22, 12468–12470.
38. M. Wen, Q. N. Wu, J. Peng, Q. S. Wu and C. X. Wang, J. Colloid Interface Sci., 2014, 416, 220–226.
39. X. J. Yang, F. Y. Cheng, J. Liang, Z. L. Tao and J. Chen, Int. J. Hydrogen Energy, 2009, 34, 8785–8791.
40. G. P. Rachiero, U. B. Demirci and P. Miele, Int. J. Hydrogen Energy, 2011, 36, 7051–7065.
41. M. Chandra and Q. Xu, J. Power Sources, 2007, 168, 135–142.
42. A. Aijaz, A. Karkamkar, Y. J. Choi, N. Tsumori, E. Rönnebro, T. Autrey, H. Shioyama and Q. Xu, J. Am. Chem. Soc., 2012, 134, 13926–13929.
43. G. Z. Chen, S. Desinan, R. Rosei, F. Rosei and D. L. Ma, Chem.–Eur. J., 2012, 18, 7925–7930.
44. Ö. Metin and S. Özkar, Int. J. Hydrogen Energy, 2011, 36, 1424–1432.
45. C. Z. Zhu, P. Wang, L. Wang, L. Han and S. J. Dong, Nanoscale, 2011, 3, 4376–4382.
46. Q. Xu and M. Chandra, J. Alloys Compd., 2007, 446–447, 729–732.
47. M. Rakap and S. Özkar, Int. J. Hydrogen Energy, 2010, 35, 1305–1312.
48. S. C. Amendola, S. L. Sharp-Goldman, M. S. Janjua, N. C. Spencer, M. T. Kelly, P. J. Petillo and M. Binder, Int. J. Hydrogen Energy, 2000, 25, 969–975.
49. U. B. Demirci and P. Miele, Phys. Chem. Chem. Phys., 2010, 12, 14651–14665.
50. B. V. Crist, On-Screen PDF Handbook of Monochromatic XPS spectra, in The Elements and Native Oxides, XPS International, Inc, 1999, vol. 1.

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c4ra05607h

‡ These authors contributed equally to the work.

---