Raney-type catalysts with superior electronic, magnetic, mechanical, and chemical properties have been extensively used in the petrochemical industry [1, 2, 3]. They are prepared by the rapid quenching technique to obtain a metastable and uniform alloy with an amorphous or nanocrysalline microstructure [4, 5, 6, 7, 8, 9, 10]. For instance, skeletal Ni alloy prepared by the alkali dissolution of the rapidly quenched Ni-Al alloy (RQ Ni-Al) exhibits remarkable performance in the hydrogenation of unsaturated organic compounds [4, 5, 6]. The hydrogenation activity, acid resistance, and magnetic property can be varied by the addition of appropriate heteroatoms, thus giving rise to a family of RQ Ni-Al based catalysts tailored precisely for a specific purpose [1, 2, 3, 10]. For example, the introduction of Fe can increase the magnetically susceptibility, which copes with the needs of the magnetic separation technique and magnetically stabilized bed [1, 10]. The adding of the corrosion- resistant Cr enables the catalyst to be used in an acidic reaction system [1, 11]. The thermal stability and specific surface area play key roles in enhancing the performance of Raney-type catalysts [2, 6, 7, 12, 13], which depend on the phase composition, crystallite size, pore size distribution, and the introduction of the heteroatom. Therefore, the fine structures of the RQ Ni-Al based catalyst is crucial for exploring the correlation between the microstructural features and active phase. Devred et al. [14] obseved the structure of Raney-type catalysts by vacuum transfer holder transmission electron microscopy (TEM) in the TEM and high resolution TEM (HRTEM) modes, and related them to the performance. However, there actually are only few reports on the fine structure issues. It is difficult to prepare the TEM specimen due to the spontaneous combustion that occurs when the sample is exposed in air. In addition, the magnetism of Raney-type catalysts also affect the characterization in TEM at the atomic level. Based on present knowledge of the Raney-type catalyst studied by X-ray diffraction (XRD) and X-ray photoelectron spectroscopy (XPS) [15, 16], it is of great interest to explore the detailed structure by advanced electron microscopy and associate techniques (e.g., energy dispersive X-ray spectroscopy (EDS) and electron energy loss spectroscopy (EELS)) at the atomic scale. In the present work, we investigated the morphology, phase composition, distribution of the constituent elements, and electron structure of an RQ Ni-Al alloy doped with Fe and Cr (RQ Ni-Al-FC) by advanced electron microscopy and spectroscopy. The microstructural comparison of the sample protected by a vacuum transfer TEM holder and that after passivation and after spontaneous combustion was studied and discussed. Our work provides valuble information for unraveling the catalytic mechanism and understanding the reaction pathway of RQ Ni-Al alloy based catalysts in chemical reactions.
The RQ Ni-Al-FC alloy ribbon was prepared by a single-roller melt-spinning method [6, 7, 12]. Then the ribbon was ground to 200 mesh and added slowly to a NaOH solution (5 mol/L) with vigorous stirring. The RQ Ni-Al-FC catalyst obtained was washed with distilled water and then with ethanol.
The combusted sample was prepared by putting the pristine RQ Ni-Al-FC sample in air. It was spontaneously combusted and then was collected for TEM characterization.
The pristine RQ Ni-Al-FC catalyst that was stored in ethanol was put into a furnace. The air in the sealed furnace was expelled by flowing Ar at room temperature (RT) for 0.5 h. Subsequently, the sample was heated to 110 °C and kept for 6 h to dry the sample in Ar. The sample was then cooled to RT. The flow rate of Ar was 200 mL/min. Finally, the passivation gas (0.5% O2/He, 100 mL/min) was passed through the furnace at RT for 11 h.
The RQ Ni-Al-FC sample was transferred into the microscope without contact with air by a vacuum transfer holder. The combusted and passivated RQ Ni-Al-FC samples were ultrasonically dispersed in ethanol, and then a drop of the solution was deposited onto a holey C/Cu TEM grid to be used for TEM characterization. A FEI Tecnai G2 F20 microscope equipped with Gatan Image Filter (GIF) operated at 200 kV was employed to conduct the structural investigation of the series of RQ Ni-Al-FC samples in both TEM and STEM modes.
A X-ray diffractometer (XRD; Cu-Kα) was used to verify the crystalline structure. The surface morphology of the RQ Ni-Al-FC sample was characterized by a FEI NOVA NANOSEM 450 operated at 15 kV.
The RQ Ni-Al-FC catalyst was prepared by the rapid quenching method, followed by alkaline extraction. Figure 1(a) shows the XRD pattern of the RQ Ni-Al-FC catalyst. Besides the diffraction peaks of the Ni nanocrystal (111), (200), and (220) planes as the main crystalline phase in the RQ Ni-Al-FC sample, the characteristic diffraction peaks of the Ni2Al3 (011), (110), (012), (202), and (122) planes can also be indexed. There was no other phase detected by XRD other than Ni and Ni2Al3. Figures 1(b) and (c) show the TEM image of RQ Ni-Al-FC sample together with the corresponding selective area electron diffraction (SAED) pattern. From the corresponding d values, we can identify these as showing the Ni and Ni2Al3 phases, which was consistent with the XRD results.
Figure 2 displays the scanning TEM-EDS (STEM-EDS) elemental maps and the EDS spectrum of a typical region in the RQ Ni-Al-FC sample. These showed that all constituent elements, Ni, Al, Fe, Cr, and O, were homogeneously distributed in this area. Some O existed in the RQ Ni-Al-FC catalyst, but the oxide was not detected by XRD. Furthermore, the small amount of Fe and Cr was quantified by EDS (Fig. 2(c)), giving the atomic ratio of Al:Cr:Fe:Ni = 11.03:1.87:1.11:85.99, which was consistent with the preparation target. To show the fine structures of the RQ Ni-Al-FC catalyst, TEM observations were conducted (Fig. 3). The nanocrystal particle size distribution (PSD) was uniform, and the average size was 6 nm (Fig. 3(a)). The lattice spacings measured in Figs. 3(b) and (c) showed an interplanar spacing of 2.04 Å determined as the (111) basal planes of crystalline Ni, and an interplanar spacing of 2.85 Å determined as the (011) basal planes of crystalline Ni2Al3, which was consistent with XRD results. No crystalline phase of Fe and Cr was found in the HRTEM images. These atoms would be doped in the Ni crystal lattice to substitute Ni atoms or existed as nanoclusters that were too small. In addition, a NiO nanocrystal was identified in Fig. 3(d), which was otherwise seldom seen in the sample. This result explained why there was a little amount of O detected by EDS, as shown in Fig. 2.
The existence and structural information of O, Al, Fe, Ni, and Cr in RQ Ni-Al-FC catalyst were also obtained by EELS. Figure 4 shows the O K, Al L, Fe L, Ni L, and Cr L edges. The energy loss from 180 to 500 eV showed no peak in this region, which indicated that there was no C in this catalyst. By reference to the reference metal or metal oxide for the peak position, peak shape and white line ratio in the EELS spectra [17, 18, 19, 20, 21, 22, 23], most of the Ni, Fe, and Cr existed as metal Ni, metal Fe, and oxide Cr, respectively. Furthermore, the quantitative analysis of the composition showed the same results as the EDS spectrum.
The microstructural information discussed above was investigated with the use of the vacuum transfer TEM holder. We also studied the combusted and passivated samples for comparing with the pristine RQ Ni-Al-FC sample. Much crystalline NiO was identified by XRD and SAED to exist in the combusted RQ Ni-Al-FC sample. The EELS spectrum also revealed that the O content increased in the sample that had undergone spontaneous combustion. Therefore, less useful information can be obtained because the original structure features have been destroyed in the combusted RQ Ni-Al-FC sample.
The structural features of the passivated RQ Ni-Al-FC sample were totally different from those of the combusted RQ Ni-Al-FC sample. The SAED pattern (Fig. 5) showed the intensity of the NiO diffraction ring was very weak and no NiO peak was detected by XRD pattern, which indicated that there was only a small amount of the NiO phase in the passivated RQ Ni-Al-FC sample. The HRTEM images displayed that the structure of the RQ Ni-Al-FC catalyst was almost same with that of the pristine one excepted for the formation of a thin layer of NiO on the surface, as shown in Fig. 6. Thus, the passivation method can also provide valuable information about the RQ Ni-Al-FC catalyst.
The surface morphology of the passivated RQ Ni-Al-FC sample in the SEM images showed a large number of pores uniformly distributed in the whole sample, which may allow more active sites to be exposed for the catalytic reaction. The TEM image further displayed that the particle size was small, uniform, and the same with that of the pristine sample. The STEM-EDS elemental mappings (Fig. 7) showed that the constituent elements, O, Al, Cr, Fe, and Ni, were homogeneously distributed in small aggregates and the outer part of big aggregates, but the distribution of Al was not uniform in the inner part of big aggregates. The reason was that the alkaline extraction was not strong enough to remove more inner Al. The EELS spectrum of the passivated RQ Ni-Al-FC sample (Fig. 8) revealed almost the same results with that of the RQ Ni-Al-FC sample. The increasing of the oxygen peak intensity was due to the NiO layer on the surface.
It is critical to illustrate the nature of the active phase of the RQ Ni-Al-FC catalyst. Based on the microscopic features discussed and its performance in catalytic reactions (e.g., adsorptive desulfurization of light oil distillates [24]), the Ni nanocrystal is the main phase exposed on the surface, which would be the catalytically active phase in the catalyst. The existence of a little bit of NiO would be due to oxidation in the ethanol during storing. The active Ni phase should be fixed and exposed on the catalyst surface, but it is not stable due to its magnetism. Therefore, the Ni2Al3 phase plays the role of a skeleton, and the Al must be extracted by alkaline extraction for exposing more active Ni when it is used as a catalyst. The doped Fe and Cr in the RQ Ni-Al-FC catalyst were highly dispersed, and may also be the active site in some reactions or can tune the fine structure of Ni to promote its performance, such as increasing the defects in the Ni phase.
The structural features of a RQ Ni-Al-FC catalyst were systematically investigated in terms of the crystal structure, electronic structure, and chemical composition. The passivation method can provide some crucial structural information on the phase composition, elemental distribution, and the sub-surface. Microscopy investigation by a vacuum transfer holder combined with the passivation method can reveal the microstructure information of air-sensitive Raney-type catalyst to contribute to exploring and understanding their synthesis- structure-performance relationship. The structural features of fresh and used RQ Ni-Al-FC catalyst should be studied in the future because this will help reveal the active site variations in catalytic reactions.
2015, Vol. 36 Issue (10): 1662-1667
PDF (1596 KB)
2015, Vol. 36 Issue (10): 1662-1667
PDF (1596 KB)