AuPd Bimetallic Nanocrystals Embedded in Magnetic Halloysite Nanotubes: Facile Synthesis and Catalytic Reduction of Nitroaromatic Compounds

In this research, a facile and effective approach was developed for the preparation of well-designed AuPd alloyed catalysts supported on magnetic halloysite nanotubes (HNTs@Fe3O4@AuPd). The microstructure and the magnetic properties of HNTs@Fe3O4@AuPd were confirmed by transmission electron microscopy (TEM), high resolution TEM (HRTEM), energy-dispersive X-ray spectroscopy (EDS), and vibrating sample magnetometry (VSM) analyses. The catalysts, fabricated by a cheap, environmentally friendly, and simple surfactant-free formation process, exhibited high activities during the reduction of 4-nitrophenol and various other nitroaromatic compounds. Moreover, the catalytic activities of the HNTs@Fe3O4@AuPd nanocatalysts were tunable via adjusting the atomic ratio of AuPd during the synthesis. As compared with the monometallic nanocatalysts (HNTs@Fe3O4@Au and HNTs@Fe3O4@Pd), the bimetallic alloyed HNTs@Fe3O4@AuPd nanocatalysts exhibited excellent catalytic activities toward the reduction of 4-nitrophenol (4-NP) to 4-aminophenol. Furthermore, the as-obtained HNTs@Fe3O4@AuPd can be recycled several times, while retaining its functionality due to the stability and magnetic separation property.


Introduction
Noble metal nanoparticles (NPs), owing to their remarkable chemical and physical properties, have attracted significant attention in recent decades due to their potential use in electronics [1], optical devices [2], fuel cells [3], chemical sensors [4], catalysis [5][6][7] and biological materials [8][9][10]. Among these metals, gold nanoparticles (Au NPs) are frequently used for various catalytic reactions because Au catalyst is completely stable and active under mild condition [11,12]. For instance, Suvith et al. reported that Au NPs can be utilized for the catalytic degradation of methylene blue [13]. Toru Murayama et al. reported that nanoparticulate gold catalysts supported on niobium oxides (Nb 2 O 5 ) are effective catalysts for CO oxidation [14]. Furthermore, bimetallic NPs have gained much attention due to the bi-functional properties generated from the monometallic component, with applications in optics [15], fuel cells [3], and catalysis [16][17][18]. Many studies demonstrated that the addition of palladium to gold or silver leads to more effective and high catalytic performance in reactions as compared to monometallic NPs. For example, an AuPd-MnO x /ZIF-8-rGO nanocatalyst prepared by a facile wet-chemical strategy showed highly efficient catalytic activity in

Structural and Morphology Characterization
To investigate the formation of HNT@Fe 3 O 4 @AuPd nanocomposited catalysts, X-ray diffraction (XRD) analyses were conducted. Figure 1 represents the XRD patterns of HNT coated by Fe 3 O 4 , Au-Pd alloyed NPs, monometallic Au or Pd NPs, respectively. The exhibited characteristic diffraction peaks at 2θ = 12 • , 20.3 • , 24.9 • originated from the HNTs for each sample, which demonstrate that the original inner structure of halloysite is not damaged during the whole preparation of HNT@Fe 3 O 4 @AuPd nanocatalysts. The observed XRD peaks at 2θ = 30.1 • , 35.5 • , 43.1 • , 53.5 • , 57.1 • can be indexed to the planes of fcc Fe 3 O 4 (JCPDS-19-0629) [47]. The monometallic HNT@Fe 3 O 4 @Au and HNT@Fe 3 O 4 @Pd composited nanoparticles exhibit wide diffraction peaks which well match to fcc structure of bulk Au (JCPDS-65-8601) [48] and bulk Pd (JCPDS-65-2867) [49]. The weak diffractions detected for Au and Pd from XRD patterns of the above bimetallic and monometallic nancatalysts indicate the formation of small NPs [50]. Furthermore, as compared to the (111) diffraction peak of Pd NPs in HNT@Fe 3 O 4 @Pd, the diffraction peaks of HNT@Fe 3 O 4 @AuPd shift to lower 2θ values towards the Au(111) peak, which is due to the increase of the lattice parameters and the formation of the crystalline Au-Pd alloy nanoparticles on HNT@Fe 3 O 4 nanorods [19,51].

Structural and Morphology Characterization
To investigate the formation of HNT@Fe3O4@AuPd nanocomposited catalysts, X-ray diffraction (XRD) analyses were conducted. Figure 1 represents the XRD patterns of HNT coated by Fe3O4, Au-Pd alloyed NPs, monometallic Au or Pd NPs, respectively. The exhibited characteristic diffraction peaks at 2θ = 12°, 20.3°, 24.9° originated from the HNTs for each sample, which demonstrate that the original inner structure of halloysite is not damaged during the whole preparation of HNT@Fe3O4@AuPd nanocatalysts. The observed XRD peaks at 2θ = 30.1°, 35.5°, 43.1°, 53.5°, 57.1° can be indexed to the planes of fcc Fe3O4 (JCPDS-19-0629) [47]. The monometallic HNT@Fe3O4@Au and HNT@Fe3O4@Pd composited nanoparticles exhibit wide diffraction peaks which well match to fcc structure of bulk Au (JCPDS-65-8601) [48] and bulk Pd (JCPDS-65-2867) [49]. The weak diffractions detected for Au and Pd from XRD patterns of the above bimetallic and monometallic nancatalysts indicate the formation of small NPs [50]. Furthermore, as compared to the (111) diffraction peak of Pd NPs in HNT@Fe3O4@Pd, the diffraction peaks of HNT@Fe3O4@AuPd shift to lower 2θ values towards the Au(111) peak, which is due to the increase of the lattice parameters and the formation of the crystalline Au-Pd alloy nanoparticles on HNT@Fe3O4 nanorods [19,51]. The TEM images of HNTs@Fe3O4 and HNTs@Fe3O4@AuPd are shown in Figure 2. These images confirm that the nanocatalysts have been successfully fabricated and the HNTs still maintain their rod-like morphology during the preparation process. From Figure 2a, we can see the Fe3O4 nanoparticles (around 150 nm) are attached on the HNTs surface without obvious aggregation. AuPd bimetallic nanoparticles (diameter below 5 nm) have been successfully formed on the surface of HNTs@Fe3O4 by simply mixing the HAuCl4 and K2PdCl4 at room temperature. As shown in Figure  2b,c, there are many small nanoparticles adhered to the HNTs or Fe3O4 surface after the coreduction and there are no isolated AuPd catalysts outside the HNTs@Fe3O4. High-resolution TEM (HRTEM) (Figure 2d) and size distribution images indicate that the average size of AuPd bimetallic is 3.61 ± 0.28 nm. The chemical composition and elemental distribution of HNTs@Fe3O4@AuPd were The TEM images of HNTs@Fe 3 O 4 and HNTs@Fe 3 O 4 @AuPd are shown in Figure 2. These images confirm that the nanocatalysts have been successfully fabricated and the HNTs still maintain their rod-like morphology during the preparation process. From Figure 2a, we can see the Fe 3 O 4 nanoparticles (around 150 nm) are attached on the HNTs surface without obvious aggregation. AuPd bimetallic nanoparticles (diameter below 5 nm) have been successfully formed on the surface of HNTs@Fe 3 O 4 by simply mixing the HAuCl 4 and K 2 PdCl 4 at room temperature. As shown in Figure 2b,c, there are many small nanoparticles adhered to the HNTs or Fe 3 O 4 surface after the coreduction and there are no isolated AuPd catalysts outside the HNTs@Fe 3 O 4 . High-resolution TEM (HRTEM) (Figure 2d) and size distribution images indicate that the average size of AuPd bimetallic is 3.61 ± 0.28 nm. The chemical composition and elemental distribution of HNTs@Fe 3 O 4 @AuPd were investigated by EDS-mapping and the energy-dispersive X-ray spectroscopy (EDX). HAADF-STEM measurements ( Figure 3) were carried out to investigate the elemental spatial distribution of Pd and Au. Here, it is necessary to mention that for the ultrafine AuPdNPs, while it is really difficult to get their HAADF-STEM results. Therefore, AuPd NPs with some aggregation were found by the HAADF-STEM test. The dark-field images indicate a abundant deposition of AuPd nanocatalysts on the magnetic support. Clearly, from Figure 3, Si, Al and Fe atoms are distributed on HNTs@Fe 3 O 4 , whilst Pd and Au atoms are distributed uniformly over the entire HNTs@Fe 3 O 4 @AuPd nanocatalysts, which provide a strong proof on the formation of alloy decorated nanocatalyst [51]. Furthermore, the typical HRTEM image (Figure 2d) of the AuPd NPs indicated that the interplanar spacing of the particle lattice is 0.229 nm, which approximated both the (111) lattice spacing of face centered cubic (fcc) Pd (0.223 nm) and Au (0.234 nm), and this may indicate that line defects are formed [52]. From the EDX image (Figure 4a), we can observe the existence of Si, Al, Fe, Au and Pd elements, which can also prove the successful deposition of Au and Pd on the HNTs@Fe 3 O 4 catalysts carrier. According to the ICP-AES analyses, the content of the total metal supported on the HNTs@Fe 3 O 4 @AuPd nanocatalysts is 2.81 wt. % and the atomic ratio of Au:Pd is 0.402:0.598, which is consistent with the designed ratio. As shown in Figure 4b, the magnetic properties of HNTs@Fe 3 O 4 and HNTs@Fe 3 O 4 @AuPd at room temperature demonstrate that the decreasing magnetization values can be assigned to the loading of non-magnetic AuPd nanoparticles. However, the current magnetism is still strong enough for the recycling and reusing of the catalysts during the whole catalytic experiments. investigated by EDS-mapping and the energy-dispersive X-ray spectroscopy (EDX). HAADF-STEM measurements ( Figure 3) were carried out to investigate the elemental spatial distribution of Pd and Au. Here, it is necessary to mention that for the ultrafine AuPdNPs, while it is really difficult to get their HAADF-STEM results. Therefore, AuPd NPs with some aggregation were found by the HAADF-STEM test. The dark-field images indicate a abundant deposition of AuPd nanocatalysts on the magnetic support. Clearly, from Figure 3, Si, Al and Fe atoms are distributed on HNTs@Fe3O4, whilst Pd and Au atoms are distributed uniformly over the entire HNTs@Fe3O4@AuPd nanocatalysts, which provide a strong proof on the formation of alloy decorated nanocatalyst [51]. Furthermore, the typical HRTEM image (Figure 2d) of the AuPd NPs indicated that the interplanar spacing of the particle lattice is 0.229 nm, which approximated both the (111) lattice spacing of face centered cubic (fcc) Pd (0.223 nm) and Au (0.234 nm), and this may indicate that line defects are formed [52]. From the EDX image (Figure 4a), we can observe the existence of Si, Al, Fe, Au and Pd elements, which can also prove the successful deposition of Au and Pd on the HNTs@Fe3O4 catalysts carrier. According to the ICP-AES analyses, the content of the total metal supported on the HNTs@Fe3O4@AuPd nanocatalysts is 2.81 wt. % and the atomic ratio of Au:Pd is 0.402:0.598, which is consistent with the designed ratio. As shown in Figure 4b, the magnetic properties of HNTs@Fe3O4 and HNTs@Fe3O4@AuPd at room temperature demonstrate that the decreasing magnetization values can be assigned to the loading of non-magnetic AuPd nanoparticles. However, the current magnetism is still strong enough for the recycling and reusing of the catalysts during the whole catalytic experiments.

Catalytic Activities of Bimetallic or Monometallic Au/Pd Modified Nanocatalysts in the Reduction of 4-NP
The catalytic activities of as-prepared bimetallic and monometallic nanocatalysts were tested by the catalysts for the reduction of 4-NP with an excess of NaBH4, as depicted in Figure 5. The conversion from 4-NP to 4-AP can be easily monitored by the UV-vis spectroscopy. Generally, the UV-Visible spectra of 4-NP solutions have a distinct spectral profile with a maximum absorption peak at 317 nm in neutral or acidic situation, which is presented in Figure 5a. After addition of NaBH4, the absorption band shifts to 400 nm immediately, accompanied with the color change from light yellow to bright yellow. This peak shift is due to the formation of 4-nitrophenolate ion as the alkalinity of the solution increased [39]. Without the addition of the catalysts, the maximum absorption peak at 400 nm can not be altered, indicating that there is no reduction reaction, which is caused by the high kinetic barrier between 4-nitrophenolate ion and BH4 − ion.
After the addition of a small amount of HNT@Fe3O4@Au40Pd60 into the 4-NP and NaBH4 mixed solution, the color of the solution changed from bright yellow to colorless quickly, signalling the completion of the reaction. Time-dependent absorbance spectra of this reaction shows the intensity of the absorption spectra at 400 nm gradually decrease along with a concomitant increase of the 300 nm peak of 4-aminophenol, revealing the formation of 4-AP [51]. After 8 min, the peak at 400 nm is no longer observed, indicating that the reduction reaction is completed (Figure 5b). We proposed that BH4 − would adsorb onto the surface of HNT@Fe3O4@Au40Pd60 and divert activated hydrogen to the AuPd active centre to form a metal hydride complex [53]. Meanwhile, 4-NP can also simultaneously stick to the surfaces of HNT@Fe3O4@Au40Pd60 via chemical adsorption [54]. In the above factors, the transformation of adsorbed hydrogen species to 4-NP can reduce the nitro groups to amino groups, which is strongly supported by previous results [55,56].

Catalytic Activities of Bimetallic or Monometallic Au/Pd Modified Nanocatalysts in the Reduction of 4-NP
The catalytic activities of as-prepared bimetallic and monometallic nanocatalysts were tested by the catalysts for the reduction of 4-NP with an excess of NaBH4, as depicted in Figure 5. The conversion from 4-NP to 4-AP can be easily monitored by the UV-vis spectroscopy. Generally, the UV-Visible spectra of 4-NP solutions have a distinct spectral profile with a maximum absorption peak at 317 nm in neutral or acidic situation, which is presented in Figure 5a. After addition of NaBH4, the absorption band shifts to 400 nm immediately, accompanied with the color change from light yellow to bright yellow. This peak shift is due to the formation of 4-nitrophenolate ion as the alkalinity of the solution increased [39]. Without the addition of the catalysts, the maximum absorption peak at 400 nm can not be altered, indicating that there is no reduction reaction, which is caused by the high kinetic barrier between 4-nitrophenolate ion and BH4 − ion.
After the addition of a small amount of HNT@Fe3O4@Au40Pd60 into the 4-NP and NaBH4 mixed solution, the color of the solution changed from bright yellow to colorless quickly, signalling the completion of the reaction. Time-dependent absorbance spectra of this reaction shows the intensity of the absorption spectra at 400 nm gradually decrease along with a concomitant increase of the 300 nm peak of 4-aminophenol, revealing the formation of 4-AP [51]. After 8 min, the peak at 400 nm is no longer observed, indicating that the reduction reaction is completed (Figure 5b). We proposed that BH4 − would adsorb onto the surface of HNT@Fe3O4@Au40Pd60 and divert activated hydrogen to the AuPd active centre to form a metal hydride complex [53]. Meanwhile, 4-NP can also simultaneously stick to the surfaces of HNT@Fe3O4@Au40Pd60 via chemical adsorption [54]. In the above factors, the transformation of adsorbed hydrogen species to 4-NP can reduce the nitro groups to amino groups, which is strongly supported by previous results [55,56].

Catalytic Activities of Bimetallic or Monometallic Au/Pd Modified Nanocatalysts in the Reduction of 4-NP
The catalytic activities of as-prepared bimetallic and monometallic nanocatalysts were tested by the catalysts for the reduction of 4-NP with an excess of NaBH 4 , as depicted in Figure 5. The conversion from 4-NP to 4-AP can be easily monitored by the UV-vis spectroscopy. Generally, the UV-Visible spectra of 4-NP solutions have a distinct spectral profile with a maximum absorption peak at 317 nm in neutral or acidic situation, which is presented in Figure 5a. After addition of NaBH 4 , the absorption band shifts to 400 nm immediately, accompanied with the color change from light yellow to bright yellow. This peak shift is due to the formation of 4-nitrophenolate ion as the alkalinity of the solution increased [39]. Without the addition of the catalysts, the maximum absorption peak at 400 nm can not be altered, indicating that there is no reduction reaction, which is caused by the high kinetic barrier between 4-nitrophenolate ion and BH 4 − ion.
After the addition of a small amount of HNT@Fe 3 O 4 @Au 40 Pd 60 into the 4-NP and NaBH 4 mixed solution, the color of the solution changed from bright yellow to colorless quickly, signalling the completion of the reaction. Time-dependent absorbance spectra of this reaction shows the intensity of the absorption spectra at 400 nm gradually decrease along with a concomitant increase of the 300 nm peak of 4-aminophenol, revealing the formation of 4-AP [51]. After 8 min, the peak at 400 nm is no longer observed, indicating that the reduction reaction is completed (Figure 5b). We proposed that BH 4 − would adsorb onto the surface of HNT@Fe 3 O 4 @Au 40 Pd 60 and divert activated hydrogen to the AuPd active centre to form a metal hydride complex [53]. Meanwhile, 4-NP can also simultaneously stick to the surfaces of HNT@Fe 3 O 4 @Au 40 Pd 60 via chemical adsorption [54]. In the above factors, the transformation of adsorbed hydrogen species to 4-NP can reduce the nitro groups to amino groups, which is strongly supported by previous results [55,56].
For a comprehensive elucidation of the catalytic property of the HNT@Fe 3 O 4 @Au 40 Pd 60 and the synergetic effects of Au and Pd catalytic architecture in the reduction of 4-NP, we investigated the catalytic behaviour of monometallic HNT@Fe 3 O 4 @Au, HNT@Fe 3 O 4 @Pd, and bimetallic HNT@Fe 3 O 4 @Au 40 Pd 60 with different molar compositions of Au versus Pd under the identical conditions (the amount of each catalyst is controlled at 5 µg). As shown in the time-dependent UV-vis absorption spectra of Figure 5c,d, the time for HNT@Fe 3 O 4 @Pd to finish the reduction is slight shorter than that for HNT@Fe 3 O 4 @Au, while the time of the two catalysts used are obvious longer than that for HNT@Fe 3 O 4 @Au 40 Pd 60 . These results indicate that the catalytic activity of magnetic HNT-based AuPd alloyed bimetallic nanocatalysts is evidently enhanced as compared to the Au or Pd monometallic nanocatalysts. As a matter of fact, the catalytic activities of the HNT@Fe 3 O 4 @AuPd nanocomposite are tunable by altering the Au versus Pd atomic ratios. The catalytic reactions own the highest reaction rate when the atomic ratio of the alloyed catalysts reaches Au 40 Pd 60 , which can be understood in Figure 6a. For a comprehensive elucidation of the catalytic property of the HNT@Fe3O4@Au40Pd60 and the synergetic effects of Au and Pd catalytic architecture in the reduction of 4-NP, we investigated the catalytic behaviour of monometallic HNT@Fe3O4@Au, HNT@Fe3O4@Pd, and bimetallic HNT@Fe3O4@Au40Pd60 with different molar compositions of Au versus Pd under the identical conditions (the amount of each catalyst is controlled at 5 μg). As shown in the time-dependent UV-vis absorption spectra of Figure 5c,d, the time for HNT@Fe3O4@Pd to finish the reduction is slight shorter than that for HNT@Fe3O4@Au, while the time of the two catalysts used are obvious longer than that for HNT@Fe3O4@Au40Pd60. These results indicate that the catalytic activity of magnetic HNT-based AuPd alloyed bimetallic nanocatalysts is evidently enhanced as compared to the Au or Pd monometallic nanocatalysts. As a matter of fact, the catalytic activities of the HNT@Fe3O4@AuPd nanocomposite are tunable by altering the Au versus Pd atomic ratios. The catalytic reactions own the highest reaction rate when the atomic ratio of the alloyed catalysts reaches Au40Pd60, which can be understood in Figure 6a. To further evaluate the catalytic activities of the above different nanocatalysts, a pseudo-first order kinetics equation was applied, which is due to the amount of BH4 − in the catalytic system is more excessive than that of 4-NP and the reaction rate is not affected by the concentration of BH4 − . The experiment was carried out under the same conditions and the rate constant (k) is calculated as: kapt = ln(Ct/C0), where C0 is the initial absorbance of the reagents at the maximum absorption wavelength, Ct is the absorbance of reagents at maximum absorption wavelength under different time t and Kap is the apparent rate constant. Figure 6a shows the approximately linear relationships of ln(At/A0) vs. reaction time in the reaction catalyzed by the as synthesized nanocatalysts. It can be clearly seen that the reactions follow the pseudo-first-order reaction kinetics. Hence, the apparent rate constants obtained from the slopes of the linearly plots are 0.520, 0.389, 0.216, 0.023, and 0.174 min −1 for the HNT@Fe3O4@Au40Pd60, HNT@Fe3O4@Au67Pd33, HNT@Fe3O4@Au50Pd50, HNT@Fe3O4@Au and HNT@Fe3O4@Pd, respectively. It is obvious that all the HNT@Fe3O4@Au40Pd60 nanocatalysts have the highest catalytic ability and its reaction rate is about 22.61 and 2.98 times To further evaluate the catalytic activities of the above different nanocatalysts, a pseudo-first order kinetics equation was applied, which is due to the amount of BH 4 − in the catalytic system is more excessive than that of 4-NP and the reaction rate is not affected by the concentration of BH 4 − .
The experiment was carried out under the same conditions and the rate constant (k) is calculated as: k ap t = ln(C t /C 0 ), where C 0 is the initial absorbance of the reagents at the maximum absorption wavelength, C t is the absorbance of reagents at maximum absorption wavelength under different time t and K ap is the apparent rate constant. Figure 6a shows the approximately linear relationships of ln(A t /A 0 ) vs. reaction time in the reaction catalyzed by the as synthesized nanocatalysts. It can be clearly seen that the reactions follow the pseudo-first-order reaction kinetics. Hence, the apparent rate constants obtained from the slopes of the linearly plots are 0.520, 0.389, 0.216, 0.023, and 0. Furthermore, the scope of the catalytic activity of HNT@Fe 3 O 4 @Au 40 Pd 60 in the reduction for other nitroaniline derivates was also investigated. All the catalytic conditions of these analogues are same as that of 4-NP and the reaction progress of the nitroaniline was monitored by UV-vis absorption spectrometry. The data of reaction time and the conversion were calculated and are listed in Table 1, revealing our catalysts have excellent catalytic activities with perfect yield toward nitrophenol and nitroaniline derivatives regardless of the types and position of the substituents. Four kinds of nitroaniline can be reduced within 8 min with a conversion more than 99%. Interestingly, when the HNT@Fe 3 O 4 @Au 40 Pd 60 was used to reduce the nitrotoluene analog substrate, it displayed a lower activity than those of nitroaniline and nitrophenol, which indicated that the reaction processes of nitrotoluene derivatives are more complicated than those of nitroaniline and nitrophenol derivatives [57].
Nanomaterials 2017, 7, 333 7 of 12 faster than that of the monometallic Au or Pd modified nanocatalysts, which is accordance with the results obtained from the time-dependent UV-vis absorption spectra. Furthermore, the scope of the catalytic activity of HNT@Fe3O4@Au40Pd60 in the reduction for other nitroaniline derivates was also investigated. All the catalytic conditions of these analogues are same as that of 4-NP and the reaction progress of the nitroaniline was monitored by UV-vis absorption spectrometry. The data of reaction time and the conversion were calculated and are listed in Table 1, revealing our catalysts have excellent catalytic activities with perfect yield toward nitrophenol and nitroaniline derivatives regardless of the types and position of the substituents. Four kinds of nitroaniline can be reduced within 8 min with a conversion more than 99%. Interestingly, when the HNT@Fe3O4@Au40Pd60 was used to reduce the nitrotoluene analog substrate, it displayed a lower activity than those of nitroaniline and nitrophenol, which indicated that the reaction processes of nitrotoluene derivatives are more complicated than those of nitroaniline and nitrophenol derivatives [57].

Reusability of the HNT@Fe3O4@Au40Pd60
It is well known that the reusability is a main advantage of using a heterogeneous catalyst rather than a homogeneous catalyst. As the Fe3O4 NPs have remarkable magnetic properties [58], the catalysts were recovered by simple magnetic separation with the help of a magnetic field after completion of the reduction reaction and then washed with water for reusing in the next cycle. The reusability of the HNT@Fe3O4@Au40Pd60 were studied for the reduction of 4-nitrophenol with NaBH4. As shown in Figure 6b, even after five successive recycle, HNT@Fe3O4@Au40Pd60 still displayed excellent conversion efficiency without significant loss of its activity. Besides the catalytic activity and reusability, the stability of the catalysts is also a very important issue for industrial applications. To examine the stability of the nanocatalysts, the reduction reaction was repeated using composited catalysts for more than three months. The conversion of 4-NP was similar after the catalysts are  It is well known that the reusability is a main advantage of using a heterogeneous catalyst rather than a homogeneous catalyst. As the Fe 3 O 4 NPs have remarkable magnetic properties [58], the catalysts were recovered by simple magnetic separation with the help of a magnetic field after completion of the reduction reaction and then washed with water for reusing in the next cycle. The reusability of the HNT@Fe 3 O 4 @Au 40 Pd 60 were studied for the reduction of 4-nitrophenol with NaBH 4 . As shown in Figure 6b, even after five successive recycle, HNT@Fe 3 O 4 @Au 40 Pd 60 still displayed excellent conversion efficiency without significant loss of its activity. Besides the catalytic activity and reusability, the stability of the catalysts is also a very important issue for industrial applications. To examine the stability of the nanocatalysts, the reduction reaction was repeated using composited catalysts for more than three months. The conversion of 4-NP was similar after the catalysts are prepared three months later, suggesting that the catalysts were stable and can be used within three months without an obvious loss of activity [59].

Preparation of HNT@Fe 3 O 4
HNT coated by Fe 3 O 4 were synthesized through a solvothermal reaction according to a previously described method with some modification [60]. Briefly, 1 g of FeCl 3 ·6H 2 O was added to 30 mL of ethylene glycol. A clear yellow solution was obtained after sonication for 3 min. Then, 2.7 g of sodium acetate (NaAc) and 0.75 g PEI were added to the above solution and stirred for another 0.5 h. Subsequently, 0.3 g HNT was added. When the mixed solution was ultrasonically dispersed for 3 h to form a homogeneous dispersion, the mixture solution was transferred to a Teflon-lined stainless-steel autoclave (50 mL capacity). The autoclave was sealed and maintained at 200 • C. After reaction for 8 h, the autoclave was naturally cooled to ambient temperature. The obtained HNT@Fe 3 O 4 black magnetic particles were collected by an external magnetic field and washed four times with ethanol and deionized water in sequence. The product was then dried in vacuum at 60 for 12 h.

Catalytic Reduction of Nitrobenzene
The reduction of 4-NP with NaBH 4 was chosen to examine the catalytic activity and reusability of the HNTs@Fe 3 O 4 @Au 40 Pd 60 catalysts. In a typical procedure, 1 mL of 4-NP (0.1 mM) and 1.0 mL of fresh NaBH 4 (10 mM) was taken in a quartz cuvette, followed by addition of 20 µL of catalysts (0.25 mg/mL) to the mixture. The reaction solution was immediately monitored using UV-vis spectrophotometer at 1 min interval. The color of the solution changed gradually from yellow to colorless. The catalytic reduction of other nitroaniline were conducted under the same condition of 4-NP. Following the similar procedures, HNT@Fe 3 O 4 @Au, HNT@Fe 3 O 4 @Pd, and HNT@Fe 3 O 4 @Au x Pd y (each of 5 µg) were also used as catalysts for the reduction of 4-NP.

Characterization
Transmission electron microscopy (TEM), high resolution transmission electron microscopy (HRTEM), and the energy dispersive spectra (EDS) were determined by a Tecnai-G2-F30 at acceleration voltages of 200 kV. X-ray diffraction (XRD) measurements were carried out on a X'pert PRO X-ray power diffractometer using Cu Ka radiation of 1.5406 A (40 kV, 30 mA). Magnetization measurements were performed on a vibrating sample magnetometry (VSM, LAKESHORE-7304, Westerville, OH, USA) at room temperature. The UV measurement was finished on a Shimadzu UV-240 spectrophotometer. Au, Pd contents of the samples were determined by inductively coupled plasma-atomic emission spectroscopy (ICP-AES) using an IRIS Advantage ER/S spectrophotometer.

Conclusions
In conclusion, we fabricated magnetically separable and well dispersed bimetallic Au 40 Pd 60 alloyed nanoparticles on HNT@Fe 3 O 4 by a facile method. The HNT@Fe 3 O 4 @Au 40 Pd 60 nanocatalysts were explored for the reduction of organic pollutants in an aqueous medium. When compared with the monometallic HNT@Fe 3 O 4 @Au and HNT@Fe 3 O 4 @Pd nanocatalysts prepared under the same conditions, the AuPd alloyed HNT@Fe 3 O 4 @Au 40 Pd 60 nanocatalysts exhibited the much better catalytic performance, which may be ascribed to the synergistic effects of Pd and Au. Furthermore, the HNT@Fe 3 O 4 @Au 40 Pd 60 catalysts can be easily separated from the reaction system with the help of an external magnet and reused several cycles without significant loss of the activity. The as-obtained hybrids may become ideal recyclable catalysts for the reduction of aromatic nitro compounds owing to their stability and efficient magnetism.