Enhanced catalytic properties of Pd nanoparticles by their deposition on ZnO-coated ceramic membranes

Abstract

An efficient and reusable catalyst was developed by loading palladium (Pd) nanoparticles on a ceramic membrane modified with a ZnO nano-coating. The microstructures of the as-prepared Pd-loaded ceramic membranes were characterized by FESEM, EDS, ICP, XPS and HRTEM. Their catalytic properties in liquid-phase p-nitrophenol reduction to p-aminophenol were evaluated. A comparative study was also made with the Pd nanoparticles deposited onto the ceramic membrane without any modification. The XPS and HRTEM results indicate that the ZnO coating can provide strong Pd–Zn interactions on the ceramic membrane, resulting in higher catalytic stability with similar catalytic activity as compared to that without modification. It was found that the leaching and agglomeration of Pd nanoparticles should be responsible for the deactivation of Pd-loaded ceramic membranes. This work would aid the development of membrane catalysts with higher catalytic performance.

---

1. Introduction

Palladium as an outstanding catalyst has been widely used for organic synthesis reactions, e.g., hydrogenation, dehydrogenation and C–C coupling reactions.^1–4 However, industry always meets a trade-off scenario: i.e., the nano-sized particles required for high reactivity versus the difficulty in fine catalyst separation and recovery. Although the homogeneous catalysts usually own high catalytic activities, they are difficult to recover from the reaction mixture.⁵ Conversely, the heterogeneous catalysts are advantageous over the homogeneous catalysts because of their better stability and ease in separation.⁶ Therefore, many researchers have focused on developing preparation methods for highly dispersed Pd-based heterogeneous catalysts with excellent catalytic efficiency.^7–11

To develop Pd catalysts with excellent catalytic performance for a wide range of reactions, lots of attempts have been made to immobilize Pd nanoparticles on a variety of supports, e.g., carbon,^12,13 membrane,¹⁴ silica,¹⁵ hydroxyapatite,¹⁶ zeolites,¹⁷ MOFs¹⁸ and organic polymers.¹⁹ Among these supports, membranes have been considered as the good choice because the as-prepared catalysts deposited on membranes can be easily separated from the reaction medium in contrary to free powder catalysts.²⁰ As the catalyst support, the ceramic membranes such as Al₂O₃, TiO₂ and ZrO₂ have some unique advantages, i.e., superior chemical stability, long life, and excellent mechanical strength as compared to polymeric membranes.^21–24

The nature of metal active sites determines the catalytic properties of membrane-based catalysts, which is significantly influenced by the preparation methods and the properties of membrane surface.^25–27 In our previous works,^28–30 several preparation methods were developed to achieve membrane-based catalysts with better catalytic performance. A flow-through method was developed for the deposition of Pd nanoparticles on a ceramic membrane, and the Pd nanoparticles could be deposited both on the membrane surface and in the membrane pores by the flow-through method, leading to an increased loading amount of Pd nanoparticles and an enhanced p-nitrophenol conversion.²⁹ The hollow ceramic fibers, which can provide more area for the deposition of Pd nanoparticles at the same volume of membrane module, were used for Pd immobilization. A higher Pd loading was obtained, and the pore structure of hollow ceramic fiber was beneficial for the reactants to diffuse onto the catalyst surface, making the catalytic activity higher.³⁰

The surface modification plays an important role in preparing catalysts. The surface modification of membrane with γ-amino-propyltriethoxy silane (3-APTS) could improve the dispersion of Pd nanoparticles and the Pd adhesion on ceramic membranes, resulting in enhanced catalytic activity and stability.²⁸ Recently, the surface modifications based on nano-technology have been reported. To reduce the membrane fouling in treating oil–water emulsion, Chang et al.³¹ modified the Al₂O₃ microfiltration membrane with nano-sized Al₂O₃ grains by an in situ hydrolysis method. Veronovski et al.³² used TiO₂ nano-coating to obtain regenerated cellulose surfaces with self-cleaning properties. Wu et al.³³ enhanced the cycling stabilities of electrodes by modifying the spinel LiMn_1.5Ni_0.5O₄ cathode with MgF₂ nano-coating. Due to the advantages of nano-coating technology, the influence of nano-coating on the membrane-based catalysts should be investigated aiming at enhancing the catalytic properties of Pd nanoparticles.

This work attempts to fabricate Pd-loaded ceramic membranes with enhanced catalytic performance by altering the surface properties of the ceramic membranes with ZnO nano-coating using a sol–gel method. The as-fabricated Pd-loaded ceramic membrane was extensively characterized by FESEM, EDS, ICP, XPS and HRTEM, and its catalytic properties were evaluated in the reduction of p-nitrophenol to p-aminophenol with sodium borohydride (NaBH₄). A comparative study was also performed with Pd nanoparticles deposited on a ceramic membrane without modification.

2. Experimental section

2.1. Catalyst preparation

An experimental setup, shown in Fig. 1, was designed and built for the preparation of Pd-loaded ceramic membranes with a flow-through method.²⁹ The apparatus consisted of a glass solution storage tank, a feed peristaltic pump, and a membrane module. The solution storage tank had a working volume of 1 L and was equipped with an external jacket for controlling temperature. The peristaltic pump (Baoding Longer Precision Pump Co., Ltd., China) was applied to pump the solution into the membrane module. The membrane module (14 mm outer diameter and 13 mm inner diameter) was also equipped with an external jacket for temperature control. The tubular Al₂O₃ ceramic membrane (12 mm in outer diameter, 8 mm in inner diameter and 6 cm in length) was fixed in the membrane module with sealing washers. This ceramic membrane (designated as CM) was provided by Nanjing Jiusi High-Tech Co. Ltd., China and used as the starting material. The ceramic membrane had a symmetric pore structure and a nominal pore size of 3 μm.

Fig. 1 Experimental setup for the preparation of Pd-loaded ceramic membranes.

As shown in Fig. 2, the procedure for preparing Pd-loaded ceramic membranes mainly included two steps: the nano-coating modification by ZnO and the loading of Pd nanoparticles. At the first step, the ceramic membrane was immersed in 40 mL of a 0.01 M solution of anhydrous zinc acetate in a mixture of absolute ethyl alcohol and ethylene glycol monomethyl ether with a volume ratio of 1 at room temperature for 1 h. Then, the ceramic membrane were heated to 600 °C with a ramp rate of 1 °C min⁻¹, and maintained at 600 °C for 1 h before returning to room temperature at 1 °C min⁻¹; designated as ZnO–CM. At the second step, the modified membrane was impregnated with 100 mL of Pd(OAc)₂ solution (0.015 M) in acetone at 30 °C for 12 h. After impregnation, the support surface was changed to be yellowish, and the Pd²⁺ ions were reduced by 100 mL of a hydrazine hydrate (N₂H₄·H₂O) alkaline solution (0.015 M) at 30 °C for 1 h to form metallic palladium particles. During the reduction, the color of the membrane surface became black, indicating the formation of metallic palladium nanoparticles on the membrane. During the second step, the solution was pumped from the outer surface of the ceramic membrane to the inner surface and then back to the solution storage tank; the flux through the ceramic membrane was controlled at 0.7 mL cm⁻² min⁻¹. The as-fabricated Pd-loaded ceramic membrane is marked as Pd/ZnO–CM. For comparison, the ceramic membrane without modification was impregnated with Pd(OAc)₂, and reduced with hydrazine hydrate under the same conditions as mentioned above, and the obtained Pd-loaded ceramic membrane is named as Pd/CM.

Fig. 2 Procedure for preparing Pd-loaded ceramic membranes.

2.2. Catalyst characterization

Pd content was measured by inductively coupled plasma emission spectroscopy (ICP-AES, Optima 7000 DV). For ICP analyses, the samples were digested in 10% (v/v) nitric acid solution at 60 °C for 1 h. The measurements were performed at 340.458 nm, with a Pd standard. The surface morphology of the supports was analyzed by field-emission scanning electron microscopy (FESEM, Hitachi S-4800). The distribution of Pd nanoparticles was observed by using energy-dispersive X-ray spectroscopy (EDS, NORAN System Six). The element composition and chemical state of the samples scraping off the membrane were investigated by X-ray photoelectron spectrometry (XPS, Thermo ESCALAB 250) with a monochromatized Al Kα radiation (hν = 1486.6 eV) at 15 kV. The residual pressure in the analysis chamber was about 10⁻¹⁰ mbar. The C 1s signal (284.8 eV) was used to calibrate the binding energies. The morphology and particle size of the Pd nanoparticles were examined by high-resolution transmission electron microscopy (HRTEM, JEM-2010) at 200 kV. For HRTEM observations, the samples were prepared by scraping off the membrane as well, applying sonication in ethanol for 10 min, and then depositing the material onto carbon-coated copper grids.

2.3. Catalytic performance evaluation

In order to test the catalytic properties of as-fabricated Pd-loaded ceramic membranes, we adopted the p-nitrophenol reduction to p-aminophenol with NaBH₄ as the model reaction, as depicted in Scheme 1.

Scheme 1 Reduction of p-nitrophenol to p-aminophenol with NaBH₄.

The catalytic reaction was also performed in the setup shown in Fig. 1. For each test, 1.25 g (9 mmol) of p-nitrophenol and 1.0 g (26.5 mmol) of NaBH₄ were dissolved in 500 mL of an ethanol–water mixture with a volume ratio of 1 : 49 in the solution storage tank, and the system was heated with circulating water at 30 °C. When the p-nitrophenol was completely dissolved, the pump was turned on. The solution was forced to flow through the pores of the Pd-loaded ceramic membrane and then back into the solution storage tank, and the reaction started. After 1 h of reaction, the Pd-loaded ceramic membrane was removed from the membrane module, thoroughly washed with ethanol, and dried at room temperature for the next test. The products were collected at set intervals from the outlet of the reactor and analyzed by a high-performance liquid chromatography (HPLC) system (Agilent 1200 series, USA) equipped with a diode array detector (DAD) and an auto sampler. Chromatographic separations were performed at 35 °C using a ZORBAX Eclipse XDB-C18, 5 μm, 4.6 mm × 250 mm column. A mobile phase composed of 80% methanol and 20% water at a flow rate of 1 mL min⁻¹ was used. To evaluate the Pd leaching degree during the reaction, the Pd amount in the reaction mixture was determined using inductively coupled plasma optical emission spectroscopy (ICP-AES, Optima 7000 DV).

3. Results and discussion

3.1. Characterization of Pd-loaded ceramic membranes

Fig. 3 shows the FESEM images of the surface morphology of the Pd-loaded ceramic membranes. Compared to the bare surface of CM (Fig. 3a), some changes in the surface morphology are observed for Pd/CM (Fig. 3b), which are attributed to the adherence of Pd nanoparticles on the surface. Fig. 3c highlights that the ceramic membrane can be successfully modified with ZnO nano-coating. It is seen from the inset that the particle size of ZnO is about 300 nm. Yang et al. deposited ZnO nanoparticles on Si substrates by a sol–gel method as the seed layers for the growth of ZnO nanorod arrays and found that the diameter of ZnO was 35–60 nm.³⁴ The larger particle size of ZnO in this study is probably caused by higher zinc acetate concentration. The concentration of zinc acetate adopted was 0.01 M as presented in the Section 2, significantly higher than the value of 0.005 M in the work of Yang and co-workers.³⁴ No obvious difference can be observed by comparing parts (c) and (d) of Fig. 3, which might be due to the small particle size of Pd nanoparticles. Similar results can be observed from the cross-section of Pd-loaded ceramic membranes, as shown in Fig. 4.

Fig. 3 Surface FESEM images of (a) CM, (b) Pd/CM, (c) ZnO–CM and (d) Pd/ZnO–CM.

Fig. 4 Cross-section FESEM images of (a) CM, (b) Pd/CM, (c) ZnO–CM and (d) Pd/ZnO–CM.

The EDS analysis was used to analyze the distribution of Pd nanoparticles on the membrane, and the scan area and time (90 s) were the same for all the samples. The ZnO nanoparticles are uniformly distributed on both the membrane surface and cross-section (Fig. 5a and 6a), in agreement with the FESEM characterization. The EDS analyses in parts (b) and (c) of Fig. 5 and 6 show that many Pd nanoparticles adhere on the membrane surface and cross-section, and there is no obvious difference between the two types of Pd-loaded ceramic membranes, indicating that ZnO nano-coating has no significant influence on the dispersion of Pd nanoparticles. It is also observed from Fig. 5c and 6c that the Pd nanoparticles can be uniformly distributed on the ZnO-modified ceramic membrane, suggesting that there is no obvious difference in the distribution of Pd nanoparticles on the ZnO surface and the membrane surface.

Fig. 5 EDS analyses of (a) the distribution of Zn on ZnO–CM surface, (b) the distribution of Pd on Pd/CM surface and (c) the distribution of Pd on Pd/ZnO–CM surface.

Fig. 6 EDS analyses of (a) the distribution of Zn on ZnO–CM cross-section, (b) the distribution of Pd on Pd/CM cross-section and (c) the distribution of Pd on Pd/ZnO–CM cross-section.

The Pd contents in the as-prepared Pd-loaded ceramic membranes were measured by ICP. According to the ICP analysis, the Pd content is 0.103 mg per cm² of membrane area (corresponding to 2.33 mg of palladium mounted in the reactor for the catalytic reactions) for Pd/ZnO–CM, and similar palladium content, 0.119 mg per cm² of membrane area (corresponding to 2.69 mg of palladium mounted in the reactor for the catalytic reactions) is determined for Pd/CM. The results are in agreement with the EDS analyses.

The XPS analysis was employed to confirm the presence of ZnO and Pd in the as-fabricated Pd/ZnO–CM. For the observation, the sample was prepared by scraping off the membrane. Fig. 7 shows the XPS survey spectrum of the powder taken from Pd/ZnO–CM, and the result for Pd/CM is also presented for comparison. Pd is detected in both types of Pd-loaded ceramic membranes, whereas Zn is only found in Pd/ZnO–CM. The results confirm the presence of the ZnO nano-coating on the Pd/ZnO–CM and the loading of Pd nanoparticles on the membrane. The detected Al and O elements come from the membrane. As presented in Fig. 8, the Pd 3d_5/2 and Pd 3d_3/2 electronic states for Pd/CM are observed at 335.60 and 340.70 eV, respectively, which are the characteristic values for Pd(0) species.³⁰ In contrast, Pd/ZnO–CM shows photoelectron peaks corresponding to both Pd(0) and Pd(II) species (Fig. 9). The Pd 3d_5/2 and Pd 3d_3/2 electronic states for Pd(0) are at 335.60 and 340.90 eV, respectively. Peaks are also detected at 337.10 and 342.60 eV corresponding to the Pd 3d_5/2 and Pd 3d_3/2 electronic states for Pd(II) species,²⁸ respectively. The presence of Pd(II) species might be due to the interaction between Pd and ZnO, which is beneficial for the enhancement of adherence of Pd nanoparticles on the membrane. Table 1 lists the binding energies of some elements. For Pd/ZnO–CM and Pd/CM, no obvious difference is found in the binding energy of Al 2p. Similar phenomenon is observed for O 1s. Instead, the Pd 3d binding energy of Pd/ZnO–CM is significantly higher than that of Pd/CM, thus showing that the signal cannot be properly fitted to a single component and strongly implying that the Pd species in Pd/ZnO–CM are already in multi-chemical states such as metallic Pd and PdZn alloys. Barrios et al. found that the formation of PdZn alloys could lead to an increase in the binding energy of Pd metal by XPS studies of Pd–Zn interaction.³⁵ Furthermore, a shift to lower binding energy (by 0.5 eV) is observed for Zn 2p in Pd/ZnO–CM as compared to the reported data,^35,36 also indicating the interaction of Pd–Zn. As it has been reported by several authors that the alloying of Pd and Zn increases the binding energy of Pd but reduces the binding energy of Zn,^35–37 which is consistent with the results in the present study.

Fig. 7 XPS survey spectra of the power taken from (a) Pd/ZnO–CM and (b) Pd/CM.

Fig. 8 Pd 3d XPS spectrum of the power taken from Pd/CM.

Fig. 9 Pd 3d XPS spectrum of the power taken from Pd/ZnO–CM.

Table 1 Binding energies of some elements in Pd/ZnO–CM and Pd/CM

Elements	Pd/ZnO–CM	Pd/CM
Al 2p	74.09	74.07
O 1s	531.1	531.05
Pd 3d	336.19	335.46
Zn 2p	1021.67	—

---

The HRTEM was carried out to analyze the morphology and particle size of Pd nanoparticles in the two types of Pd-loaded ceramic membranes. The samples were taken from the membrane surfaces. As displayed in Fig. 10a, the Pd nanoparticles can be uniformly distributed on the ZnO-modified membrane, and the mean diameter of the Pd nanoparticles is about 2.8 nm. Similarly, the Pd nanoparticles with an average size of 3.3 nm can also be highly dispersed on the membrane without modification (Fig. 10b). The characteristic lattice fringes of 2.2 Å spacing confirm the (111) planes of the face-centered-cubic Pd(0) structure, as shown in the HRTEM images.³⁰ The analyses of plane spacing of the metal crystallites of the Pd/ZnO–CM indicate the presence of the PdZn alloy (d = 2.26 Å, corresponding to the (1 1 1) diffracting planes),³⁵ in accordance with the XPS studies.

Fig. 10 TEM images of the power taken from (a and c) Pd/ZnO–CM and (b and d) Pd/CM.

3.2. Catalytic performance of Pd-loaded ceramic membranes

The catalytic reduction of p-nitrophenol to p-aminophenol with NaBH₄ (Scheme 1) was performed as a mode reaction for testing the catalytic properties of the as-fabricated Pd-loaded ceramic membranes.

It is seen from Fig. 11 that the p-nitrophenol conversion significantly increases with time irrespective of the type of Pd-loaded ceramic membrane, similar to our previous work.²⁹ Apparently, the p-nitrophenol conversion for Pd/ZnO–CM is almost the same as Pd/CM under the same reaction conditions. According to the EDS, TEM and ICP characterization as discussed above, the Pd dispersion and loading are similar for Pd/ZnO–CM and Pd/CM, which should be responsible for their same reactivities. In addition, the Pd leaching of the Pd-loaded ceramic membrane is considerable as discussed in the following ICP analysis, and the leaching Pd species may catalyze the p-nitrophenol reduction. In order to verify the presumption, the following experiment was carried out. When the p-nitrophenol reduction reaction was performed for 60 min, the Pd-loaded ceramic membrane was removed from the reactor system, and the reduction reaction was carried out for another 60 min. It is observed (data not shown here), the p-nitrophenol conversion remains stable after the remove of Pd-loaded ceramic membrane, indicating that the leaching Pd species plays a negligible role and the p-nitrophenol reduction may proceed over the Pd catalyst surface in a heterogeneous fashion.^38,39 In the present work, after 60 min of reaction, the p-nitrophenol conversion can reach about 60%, lower than the value of 100% in our previous one.²⁹ The difference should be mainly related with the adopted ceramic membrane. In this study, as presented in the Experimental section, the length of ceramic membrane is 6 cm, obviously smaller than the value of 18 cm in our previous one,²⁹ leading to less membrane area for the deposition of Pd nanoparticles. As a result, the Pd mounted in the reactor for the catalytic reaction is less, corresponding to the lower p-nitrophenol conversion.

Fig. 11 Variation of p-nitrophenol conversion with time: (a) Pd/ZnO–CM and (b) Pd/CM.

To investigate the catalytic stabilities of the as-fabricated Pd-loaded ceramic membranes, a series of catalytic reaction cycles were carried out. In this study, the catalytic stability is expressed by the ratio of the p-nitrophenol conversion at 60 min after a certain number reaction cycles to that at the first reaction cycle. Fig. 12 shows the correlation between the relative decrease degree in the p-nitrophenol conversion and the number of catalytic reaction cycle. It is noted that, both Pd/ZnO–CM and Pd/CM are undergoing deactivation during each reaction cycle, but the deactivation degree is obviously different. After six continuous reaction cycles, the Pd/ZnO–CM suffers about 19.46% deactivation, whereas the Pd/CM suffers about 33.85% deactivation. These results highlight that the catalytic stability of Pd/ZnO–CM is superior to that of Pd/CM. The interaction between Pd and ZnO as presented by the XPS and HRTEM characterization (Fig. 9 and 10) make the Pd nanoparticles not easy leach from the ZnO surface. Furthermore, ZnO is stable on the membrane as discussed below. Thus, the Pd nanoparticles adhered on the ZnO surface do not easy leach from the ZnO-modified ceramic membrane during the reaction cycles, leading to a superior catalytic stability of Pd/ZnO–CM. To verify the assumption, the recovered Pd-loaded ceramic membranes were characterized by ICP, XPS and FESEM. The ICP analysis in Table 2 reveals that, after six continuous reaction cycles, the Pd amount in Pd/ZnO–CM is 0.0903 mg per cm² of membrane area and it is 87.7% of the initial palladium content in fresh Pd/ZnO–CM. However, the Pd content decreases significantly to 0.0792 mg per cm² of membrane area in Pd/CM, corresponding to 66.6% of the initial content in fresh Pd/CM. As given in Fig. 13, the Pd peaks for the recovered Pd-loaded ceramic membranes become weaker than those of the fresh ones shown in Fig. 7, especially for Pd/CM, which should be due to the falling off of Pd nanoparticles, in agreement with the analysis of ICP. The FESEM images in Fig. 14 show that many ZnO particles still adhere on the membrane surface and cross-section for the recovered Pd/ZnO–CM, similar with the fresh ones (Fig. 3d and 4d), indicating the good stability of ZnO on the membrane during the reaction cycles. These results confirm that the modification of membrane with ZnO nano-coating is beneficial in improving the adhesion of Pd nanoparticles on the membrane and its catalytic stability. The results also suggest that Pd leaching is one of the main reasons responsible for the decrease in the catalytic activity of Pd-loaded ceramic membrane. Similar results were also reported in the literatures.^18,19,29,30 It is found by correlating the catalyst deactivation and Pd leaching that both do not match well, and the degree of catalyst deactivation is larger as compared to Pd leaching. For instance, 87.7% of the initial palladium is present in the recovered Pd/ZnO–CM, but the catalyst maintains only 80.54% of its initial catalytic activity. This finding illustrates that, not only the Pd leaching but also other parameters such as morphology evolution and species change of the Pd nanoparticles may lead to the catalyst deactivation. Hence, the evolution of Pd morphology should be investigated to explain the decrease in the catalytic activity. The Pd nanoparticles on the Pd-loaded ceramic membranes after six catalytic reaction cycles were examined with HRTEM as presented in Fig. 15. Compared to the fresh membrane catalysts (Fig. 10), for the recovered Pd/ZnO–CM and Pd/CM, the obvious agglomeration of Pd nanoparticles on the membrane can be observed, resulting in larger particle size and lower catalytic activity. Kuhn et al.⁴⁰ also suggested that the isolated Pd sites or small Pd ensembles were more active for the conversion compared to Pd surface ensembles containing contiguous Pd atoms. According to these analyses, we can conclude that the deactivation of Pd-loaded ceramic membrane is mainly caused by the leaching and agglomeration of Pd nanoparticles. Compared to our previous work,²⁹ the deactivation degree is higher in this study, which may be mainly caused by the surface modification method. For the former, the ceramic membrane was modified with an aminofunctional silane of N-(2-aminoethyl)-3-aminopropyltrimethoxysilane (AAPTS). While for the latter, the ceramic membrane was modified with ZnO nano-coating. In comparison with the ZnO nano-coating, the modification with an aminofunctional silane may be more effective for the adhesion of Pd nanoparticles on the ceramic membrane, leading to a higher catalytic stability.

Fig. 12 Catalytic stability investigations of (a) Pd/ZnO–CM and (b) Pd/CM.

Table 2 ICP analysis of palladium content

Samples	Palladium content (mg cm^−2)	Palladium leaching (%)
Pd/ZnO–CM (fresh)	0.103	—
Pd/CM (fresh)	0.119	—
Pd/ZnO–CM (used 6 times)	0.0903	12.3
Pd/CM (used 6 times)	0.0792	33.4

---

Fig. 13 XPS survey spectra of the power taken from (a) Pd/ZnO–CM and (b) Pd/CM after six catalytic reaction cycles.

Fig. 14 FESEM images of the surface (a) and cross-section (b) of Pd/ZnO–CM after six catalytic reaction cycles.

Fig. 15 TEM images of the power taken from (a) Pd/ZnO–CM and (b) Pd/CM after six catalytic reaction cycles.

The HPLC analysis shows that the selectively to p-aminophenol can reach 100% in each case (data not shown here), indicating that the as-fabricated Pd-loaded ceramic membranes have high catalytic selectivity in the p-nitrophenol reduction to p-aminophenol, in good agreement with our previous works.^28–30

4. Conclusions

In this work, an alumina tubular membrane was modified with ZnO nano-coating, and Pd nanoparticles were then deposited on the modified membrane to fabricate a Pd-loaded ceramic membrane. Additional separation steps for the Pd nanoparticles were not needed. The as-fabricated Pd-loaded ceramic membrane displays significantly higher catalytic stability with similar catalytic activity in the p-nitrophenol reduction to p-aminophenol compared to the Pd-loaded ceramic membrane without any modification. The modification of membrane with ZnO nano-coating can enhance the interaction between the Pd nanoparticles and membrane, and make the Pd nanoparticles not easily fall off from the membrane during the reaction cycles, resulting in a superior catalytic stability. The experimental results of this study are encouraging, and further study is under progress to improve the catalytic performance of the as-fabricated Pd-loaded ceramic membrane, especially for the catalytic stability, by optimizing the synthesis conditions of ZnO nano-coating.

Acknowledgements

Financial supports from the Jiangsu Natural Science Foundation for Distinguished Young Scholars (BK20150044), the National Natural Science Foundation (91534110, 21306081, 21125629), the Natural Science Foundation of the Higher Education Institutions of Jiangsu Province (14KJB530004), the Foundation from State Key Laboratory of Materials-Oriented Chemical Engineering (ZK201402, ZK201407), and the Technology Innovation Foundation for Science and Technology Enterprises in Jiangsu Province (BC2015008) of China are gratefully acknowledged.

References

1. T. Zhang, B. Li, X. F. Zhang, J. S. Qiu, W. Han and K. L. Yeung, Microporous Mesoporous Mater., 2014, 197, 324–330.
2. X. F. Zhou, Y. H. Wei, X. H. Zhao, X. Li, Y. Zhang, B. Li and J. J. Liu, Colloids Surf., A, 2015, 482, 416–421.
3. T. Zhang, X. F. Zhang, X. J. Yan, L. Lin, H. O. Liu, J. S. Qiu and L. Y. King, Catal. Today, 2014, 236, 41–48.
4. K. P. Song, P. Liu, J. Y. Wang, L. Pang, J. Chen, I. Hussain, B. Tan and T. Li, Dalton Trans., 2015, 44, 13906–13913.
5. H. Chen, W. L. Dai, R. H. Gao, Y. Cao, H. X. Li and K. N. Fan, Appl. Catal., A, 2007, 328, 226–236.
6. F. F. Karbach, J. R. Severn and R. Duchateau, ACS Catal., 2015, 5, 5068–5076.
7. M. Beauperin, R. Smaliy, H. Cattey, P. Meunier, J. Ou, P. H. Toy and J. C. Hierso, ChemPlusChem, 2015, 80, 119–129.
8. T. Iwai, R. Tanaka, T. Harada and M. Sawamura, Chem.–Eur. J., 2014, 20, 1057–1065.
9. M. Beaupérin, R. Smaliy, H. Cattey, P. Meunier, J. Ou, H. Patrick, P. H. Toy and J. C. Hierso, Chem. Commun., 2014, 50, 9505–9508.
10. J. Liu, H. Q. Yang, F. Kleitz, Z. G. Chen, T. Y. Yang, E. Strounina, G. Q. Lu and S. Z. Qiao, Adv. Funct. Mater., 2012, 22, 591–599.
11. L. R. S. Lara, A. D. Zottis, W. C. Elias, D. F. Jr, C. E. M. de Campos, J. J. S. Acuña and J. B. Domingos, RSC Adv., 2015, 5, 8289–8296.
12. J. X. Yang, X. M. Tan, Y. Wang and X. H. Wang, J. Porous Mater., 2013, 20, 501–506.
13. M. Martin-Martinez, A. Álvarez-Montero, L. M. Gómez-Sainero, R. T. Baker, J. Palomar, S. Omar, S. Eser and J. J. Rodriguez, Appl. Catal., B, 2015, 162, 532–543.
14. A. G. Gil, Z. T. Wu, D. Chadwick and K. Li, Chem. Eng. Sci., 2015, 137, 364–372.
15. B. Basu and S. Paul, Appl. Organomet. Chem., 2013, 27, 588–594.
16. A. Indra, C. S. Gopinath, S. Bhaduri and G. K. Lahiri, Catal. Sci. Technol., 2013, 3, 1625–1633.
17. H. Q. Yang, Y. W. Wang, Y. Qin, Y. Z. Chong, Q. Z. Yang, G. Li, L. Zhang and W. Li, Green Chem., 2011, 13, 1352–1361.
18. M. Bagherzadeh, F. Ashouri, L. Hashemi and A. Morsali, Inorg. Chem. Commun., 2014, 44, 10–14.
19. A. L. Isfahani, I. Mohammadpoor-Baltork, V. Mirkhani, A. R. Khosropour, M. Moghadam and S. Tangestaninejad, Eur. J. Org. Chem., 2014, 10, 5603–5609.
20. K. C. D. Figueiredo, V. M. M. Salim and C. P. Borges, Catal. Today, 2008, 133, 809–814.
21. X. W. Chen, L. Hong, Y. F. Xu and Z. W. Ong, ACS Appl. Mater. Interfaces, 2012, 4, 1909–1918.
22. L. M. Corneala, M. J. Baumann, S. J. Masten, S. H. R. Davies, V. V. Tarabara and S. Byun, J. Membr. Sci., 2011, 369, 182–187.
23. C. Zhang, R. Ran, G. H. Pham, K. Zhang, J. Liu and S. M. Liu, RSC Adv., 2015, 5, 5379–5386.
24. D. Z. Han, J. H. Wu, Z. F. Yan, K. Zhang, J. Liu and S. M. Liu, RSC Adv., 2014, 4, 19999–20004.
25. C. Mateos-Pedrero, H. Silva, D. A. P. Tanaka, S. Liguori, A. Iulianelli, A. Basile and A. Mendes, Appl. Catal., B, 2015, 174, 67–76.
26. J. F. Wu, S. Y. Shan, V. Petkov, B. Prasai, H. Cronk, P. Joseph, J. Luo and C. J. Zhong, ACS Catal., 2015, 5, 5317–5327.
27. H. Y. Na, L. Zhang, H. X. Qiu, T. Wu, M. X. Chen, N. Yang, L. Z. Li, F. B. Xing and J. P. Gao, J. Power Sources, 2015, 288, 160–167.
28. R. Z. Chen, Y. G. Jiang, W. H. Xing and W. Q. Jin, Ind. Eng. Chem. Res., 2011, 50, 4405–4411.
29. H. Y. Li, H. Jiang, R. Z. Chen, Y. Wang and W. H. Xing, Ind. Eng. Chem. Res., 2013, 52, 14099–14106.
30. R. Z. Chen, Y. G. Jiang, W. H. Xing and W. Q. Jin, Ind. Eng. Chem. Res., 2013, 52, 5002–5008.
31. Q. B. Chang, J. E. Zhou, Y. Q. Wang, J. M. Wang and G. Y. Meng, Desalination, 2010, 262, 110–114.
32. N. Veronovski, A. Rudolf, M. S. Smole, T. Kreze and J. Gersak, Fibers Polym., 2009, 10, 551–556.
33. Q. Wu, X. P. Zhang, S. W. Sun, N. Wan, D. Pan, Y. Bai, H. Y. Zhu, Y. S. Hu and S. Dai, Nanoscale, 2015, 7, 15609–15617.
34. L. L. Yang, Q. X. Zhao, M. Willander and J. H. Yang, J. Cryst. Growth, 2009, 311, 1046–1050.
35. C. E. Barrios, M. V. Bosco, M. A. Baltanas and A. L. Bonivardi, Appl. Catal., B, 2015, 179, 262–275.
36. K. Dumbuya, R. Denecke and H. P. Steinrück, Appl. Catal., A, 2008, 348, 209–213.
37. S. T. Liu, K. Takahashi, K. Fuchigami and K. Uematsu, Appl. Catal., A, 2006, 299, 58–65.
38. Y. B. Huang, S. J. Liu, Z. J. Lin, W. J. Li, X. F. Li and R. Cao, J. Catal., 2012, 292, 111–117.
39. W. Zhong, H. L. Liu, C. H. Bai, S. J. Liao and Y. W. Li, ACS Catal., 2015, 5, 1850–1856.
40. M. Kuhn, M. Lucas and P. Claus, Ind. Eng. Chem. Res., 2015, 54, 6683–6691.

---