catalysts
catalysts
Proefschrift
Ter verkrijging van de graad van doctor aan de Technische Universiteit Delft
op gezag van de Rector Magnificus prof.dr.ir. J.T.Fokkema, voorzitter van het College van Promoties,
in het openbaar te verdedigen op vrijdag 4 juni om 13.00 uur door
François DEVRED
Diplome d'études approfondies de structure et dynamique des systèmes réactifs, Université des Sciences et Techniques de Lille, France
Toegevoegd promotor: Dr. P.J. Kooyman
Samenstelling promotiecommissie: Restor Magnificus, voorzitter
Prof. dr. H.W.Zandbergen Technische Universiteit Delft, promotor Dr. P.J. Kooyman Technische Universiteit Delft
Prof dr. F. Kapteijn Technische Universiteit Delft Prof.dr. J. J.A. Moulijn Technische Universiteit Delft Prof. dr. ir. L. Katgerman Technische Universiteit Delft Prof.dr. J. Geus Utrecht Universiteit
Dr. C. Bayense Engelhart
Printed and distributed by: DUP Science Dup Science is an imprint of
Delft University Press P.O. BOX 98 2600 GM Delft The Netherlands Telephone: +31 15 278 5678 Telefax: +31 15 278 5706 E-mail: info@library.tudelft.nl ISBN: 90-407-2503-9
Copyright © 2004 by François DEVRED
All rights reserved. No part of the material protected by this copyright notice may be reproduced or utilized in any forms or by any means, electronic or mechanical, includind photocopying, recording or by any information storage and retrieval system, without written permission from the publisher: Delft University Press.
Chapter 1
1
General IntroductionChapter 2
17
Instrumentation
Chapter 3
37
Genesis of the active phase in Raney-type nickel catalysts
Chapter 4
65
On the transformation of Ni-Al alloy into Raney-type nickel catalysts
Chapter 5
81
Raney-type nickel catalysts: promoted systems
Chapter 6
109
Melt spinning: an alternative route to prepare Ni-Al alloy for Raney-type nickel catalysts
Chapter 7
123
Preparation of Raney-type nickel catalysts for TEM: the ultramicrotomy technique
Summary – Samenvatting
135
Acknoledgements
Curriculum vitae
General introduction
In this chapter, a general introduction about Raney-type nickel catalysts is given, based on available literature. The main advantage of Raney-type catalysts is that they do not require a reduction step prior to use as do conventional catalysts. The activation process determines the structure and properties of Raney-type nickel catalysts, but the structure and the composition of the starting alloy also play an important role in the performance of the final catalyst. Some discrepancies can be found in the literature concerning the microstructural changes of the solid phase leading to the final catalyst. Thus, even though it has been invented as long ago as 1925 by Murray Raney, the characterization of Raney-type nickel catalysts and the relation between its structural properties and catalytic performance are still of interest in research.
1. Characterization in catalysis research
Van Santen identified three levels of research in catalysis [1]: - The macroscopic level.
- The mesoscopic level. - The microscopic level.
The macroscopic level is the world of reaction engineering, test reactors and catalyst beds. Questions concerning the catalyst deal with such aspects as activity per unit volume, mechanical strenght and shape (monoliths, extrudates, spheres or powders). The mesoscopic level, which includes much characterization work, comprises kinetic studies, activity per unit surface area and correlation between the composition and the structure of a catalyst and its catalytic performance. Finally, the microscopic level is the one that deals with the details on adsorption at the surface, reaction mechanisms, theoretical modeling and surface science.
The industrial view on catalyst characterization is different. Developing an active, selective and stable catalyst is the main goal. In order to accomplish this, identification of the structural properties that discriminate between an efficient and a less efficient catalyst is required. Information from characterization techniques will help to establish a correlation between catalytic performance on the one hand and catalyst composition, particle size, shape and pore dimensions on the other. This is extremely useful in catalyst development even if such a correlation may not give fundamental insight into how the catalyst operates in molecular detail.
To simplify, one could say that industrial research mainly deals with macroscopic and mesoscopic levels whereas fundamental research would mainly focus on the microscopic level. An intensive cooperation between experts in catalysis and materials science is the best way to ensure a successful outcome to any research project in catalysis. In the research described in this thesis, Raney-type nickel catalysts have been extensively studied in a project that includes fundamental study as well as cooperation with industrial partners (see framework of the thesis in this chapter).
2. Raney-type nickel catalysts
2.1 Introduction
After graduating as a mechanical engineer from the University of Kentucky in 1909, Murray Raney joined the Lookout Oil and Refining Company in Tennessee in 1915. He was responsible for the installation of electrolytic cells for the production of hydrogen, which was used in the hydrogenation of vegetable oils. At the time, the industry used a nickel catalyst
that was prepared by hydrogen reduction of supported nickel oxide. Believing that better catalysts could be produced, Raney formed his own research company in 1921. In 1924, he produced a 50-50 wt % nickel silicon alloy which he treated with sodium hydroxide solution to produce a greyish metallic solid. Testing for the hydrogenation of cottonseed oil, he found the activity of his catalysts to be five times higher than the best catalyst then in use. A patent was issued in December 1925 [2]. Subsequently, Raney produced an even more efficient catalyst by leaching a 50-50 wt % nickel aluminum alloy in aqueous sodium hydroxide. A patent application was filed in 1926 [3]. In 1963, Raney sold his business to W.R. Grace and Co. whose Davison Division still produces and markets a wide range of these catalysts. As the name “Raney Ni” is protected by registered trademark, “skeletal”, “spongy” or “Raney-type” nickel is used to refer to this type of catalysts. Even though other alloy systems were considered (including iron [4], cobalt [5] and copper [6]), it is of interest to note that nowadays, the 50-50 wt % nickel aluminum alloy is used for the major part of the world production of Raney-type nickel catalyst. Also Raney-type Co and Cu are commercially produced. For specific applications, a third metal can be added to the binary alloy to promote catalytic performance. The most common promoters used for Ni are Cr [7], Mo [8], Zn [9], Fe [4], Cu [10] and Co [11].
2.2 The starting alloy
2.2.1 The binary Ni-Al phase diagram
Figure 1: The Ni-Al phase diagram.
The phase diagram of Ni-Al alloys [12] is complex, but the region over 60 wt% nickel is not of interest because the corresponding intermetallic compound cannot give the catalyst
A 50-50 wt % composition B C D E
[13]. The commercial alloy contains 50 wt % nickel in the form of NiAl3, Ni2Al3 and of eutectic Al-Ni Al3 [13-15]. The NiAl phase is difficult to detect and its presence in small amounts cannot be excluded for commercial alloys [14].
In the area that concerns starting alloys for Raney-type nickel, the following reactions occur [12]:
Table 1: Special points of the Ni-Al system.
Reaction (B)* L + NiAl ¤ Ni
2Al3
(C) L + Ni2Al3 ¤ NiAl3
(D) L ¤ NiAl3 + Al
Composition (at.% Ni) 26.9 42 40 15 36.8 25 2.7 25 0.01 T ( °C ) 1133 854 639 Reaction type Peritectic Peritectic Eutectic *
the letters indicate the area of interest in the phase diagram.
The solidification pathway from the liquid phase to the solid phase is illustrated in figure 1. During the cooling down, the liquid will have the composition of the thick line. For a 50-50 wt % nickel aluminum alloy (A), the first phase to solidify under natural cooling is NiAl. It occurs around 1350 °C (B). At 1133 °C this, this phase will react with the liquid to give Ni2Al3 and a liquid (C). Since no NiAl is found in our alloy samples, a complete reaction of NiAl with the liquid must have occurred. Thus, from 1133 °C to 854 °C, only Ni2Al3 is crystallized. At 854 °C, part of the Ni2Al3 can react with the liquid to give NiAl3(D). Between 854 °C and 639 °C, NiAl3 is crystallized. At 639 °C, the formation of a small amount of Al occurs due to the eutectic Al-NiAl3 (E).
2.2.2 Preparation of the alloy
Alloys are prepared commercially and in the laboratory by melting the nickel and aluminum under inert atmosphere and quenching the resultant melt. Then, the alloy is crushed to obtain a particle size required for particular applications. The solidification method has a large influence on the alloy structure and on the phase composition [16,17]. In the literature different methods are described to prepare the starting alloy.
Conventionally, nickel-aluminum alloys were obtained by melting under argon in an arc furnace or under air in an induction furnace [14]. The starting materials were electrolytic nickel and aluminum of purity 99.99%. The alloys were annealed at a temperature from
830°C to 940°C for 25 h to 30 h in argon, crushed and sieved. The final catalysts were obtained from powders of particle size 60-200 m.
The starting alloys can also be prepared by co-fusion of the elements [18,19]. Annealing has been performed at different temperatures below the peritectic point (1133 °C in the Ni-Al system) and for different times (from two days to one month) in order to obtain the thermodynamical equilibrium.
Russian workers have developed another technique, mechanochemical alloying [17], as a promising preparation method for Raney-type nickel catalysts. In this method, the alloys are made in a planetarium-type ball mill or in an attritor with nickel carbonyl and aluminum powder as starting materials. In the uncooled attritor, an alloy with near-equilibrium composition is formed by an exothermal reaction, caused by local heating and melting of aluminum. Mechanical alloying can also be performed in a cooled planetarium mill. In this case, the alloy formation is close to a diffusion type, whereas the phase composition is far from equilibrium. The resulting alloys are believed to be much more active in specific processes. Another advantage is that the starting materials do not necessary have to be fresh metal powders, allowing use of spent catalysts.
2.2.3 Preparation of single phases
The literature gives three main methods to prepare single phases: - The electromagnetic technique of phase decantation. - Slow bulk cooling.
- The melt spinning process.
NiAl3 and Ni2Al3, which decompose peritectically at 854°C and 1133°C, have been obtained by the electromagnetic technique of phase decantation [20-22]. The starting compositions are chosen in such a way that NiAl3 (or Ni2Al3) forms by primary crystallization. The alloys are heated to the temperature where solid NiAl3 (or Ni2Al3) exists in equilibrium with a liquid phase; the temperature is maintained for a time long enough to allow the coarsening of the solid crystals and their decantation by electromagnetic stirring is induced by medium frequency induction heating. After cooling, the small amount of liquid enclosed between the solid crystals is dissolved using HCl solution. The experimental conditions used for each compound are summarized in table 2.
Table 2: Experimental conditions of the alloy preparation [21].
Composition of the starting mixture At%
Ni Al 13 87 25 75 Electromagnetic phase decantation 650 °C – 8 h 860 °C – 8 h HCl dissolution 72 h 96 h Intermetallic phase obtained NiAl3 Ni2Al3
The preparation of pure NiAl3 was described in detail [13] and it involves a drop in temperature from a melt containing 28 wt % nickel heated at 1200°C. The crucible is cooled rapidly to 860 °C and then slowly to 760 °C while needles of NiAl3 are formed. The remaining liquid is then eliminated and the needles are etched by a 2 % sodium hydroxide solution, which removes the highly reactive eutectic and leaves pure NiAl3. In contrast to NiAl3, the Ni2Al3 phase tolerates a slight variation in Ni content (from 56 to 60 wt %) compared to the theoretical composition (52 wt %). A crucible containing the metals with the above composition is heated at 1600°C in vacuum and cooled slowly [13].
The third method used to prepare single phases is melt spinning, which yields microcrystalline alloys [16,18,23]. This technique consists of induction melting in a crucible, which has a nozzle (diameter 0.5 to 1 mm) through which the molten metal is ejected onto a cooled copper surface. The castings are performed under helium/argon atmosphere. The temperature of the liquid metal before ejection is kept at about 1550°C. The extruded molten metal in contact with the casting surface solidifies as a ribbon. The melt-spinning preparation process appears to be quite simple but the large number of parameters involved makes it relatively complex. The influence of the ejection pressure, the rolling velocity, the angle of the nozzle with respect to the cold surface and the atmosphere in which the experiment is carried out, are important factors affecting the resulting morphology of the ribbon. It has been demonstrated for instance that the ejection pressure has an influence only on the width of the ribbon whereas the angle of ejection will have an influence on the thickness of the ribbon. Performing melt spinning under ambient atmosphere will lead to formation of defects at the surface due to air bubbles, which are trapped during the solidification. The influence of the parameters of the preparation process has been described in detail [24].
Figure 2: The melt spinning process [25].
2.2.4 Properties of single phases
Commercial Raney-type nickel precursor alloy contains about 50 wt % nickel and 50 wt % aluminum. These alloys are composed of several phases: NiAl3, Ni2Al3 and an Al-NiAl3 eutectic [13-15,26]. The structure and the unit cell parameters of each phase are given in table 3.
Table 3: Crystallographic structure of single phases.
compound Space group
Lattice parameter (Å) a b c Al Fm3m 4.04 NiAl3 Pbnm 4.81 6.61 7.36 Ni2Al3 P3ml 4.03 4.9 NiAl Pm3m 2.88 Ni Fm3m 3.52
The eutectic and NiAl3 are very reactive towards hydroxide and easily lose the aluminum to give the skeletal nickel. Ni2Al3 reacts more slowly with hydroxide solution but aluminum can be removed at 50 °C and the material is completely decomposed in boiling
alkali. Catalysts prepared from commercial alloy, pure NiAl3 and pure Ni2Al3 have about the same activity [13,27].
The commercial alloy generally contains a small amount of the eutectic and somewhat more Ni2Al3 than NiAl3. The exact composition of the alloy depends not only on the Ni:Al ratio but also on the thermal conditions used in the preparation of the alloy [16,17,26].
2.3 Preparation of the catalyst
The usual preparation procedure involves the addition of the alloy to a sodium hydroxide solution held at a specific temperature. This not only removes the aluminum but also generates an atmosphere of hydrogen that serves to activate the nickel catalyst. After the leaching is complete, the excess base and the dissolved aluminate produced by the reaction are removed by an aqueous washing process. The active and highly pyrophoric catalyst is then stored in water or in alcohol [38].
The general leaching reaction is given by [29,30]: 2(M-Al)(s) + 2OH
-(aq) + 6H2O(l) Æ 2M(s) + 2Al(OH)
-4(aq) + 3H2(g) (eq. 1)
The dissolution of aluminum in aqueous sodium hydroxide may be represented more simply. The aluminum is removed by means of sodium hydroxide solution in the form of aluminate:
2Al + 2OH
+ 2H2O Æ 2AlO2 -
+ 3H2 (eq. 2)
To keep the aluminate in solution, it is necessary to use an excess of concentrated alkali, which contains 20 to 30 wt % NaOH. The aluminates have a tendency to hydrolyze when the pH decreases and they precipitate in the form of Bayerite:
2AlO2 -
+ 4H2O Æ Al2O3.3H2O + 2OH –
(eq. 3)
The Bayerite may block the pores of the catalyst and considerably reduce the performance of the catalyst.
There are three different ways to combine the starting alloys and sodium hydroxide [15]: in the first method, the most common one, an excess of sodium hydroxide solution is put into a flask and the alloy is added very slowly in order to have a gentle reaction. At the other extreme, the concentrated alkali solution is added stepwise to a suspension of alloy particles in distilled water [32]. The second method is easier to control and more convenient
to prepare large quantities of catalyst. The third way aims to completely eliminate the aluminum and to obtain nickel-rich catalysts [13,30-32]. The alloy is added slowly to 25 wt % sodium hydroxide as in the first method. When the hydrogen evolution stops, the suspension is heated to the boiling point for 30 min to 3 h, decanted and washed with a highly concentrated solution of sodium hydroxide. A second 2 h treatment using boiling 25 wt % sodium hydroxide is performed. Then the suspension is washed again with hot sodium hydroxide solutions of decreasing concentrations. Finally, it is stored in 1N sodium hydroxide solution.
Independently from the sodium hydroxide concentration, the temperature of leaching is of great importance for the first two methods. It has been clearly shown that high temperature preparations (above 100 °C) give catalysts that have different properties from those prepared under mild conditions (temperatures under 50 °C) [18]. It must be noted that under mild conditions, it is necessary to use 40 wt % sodium hydroxide solution to decompose the Ni2Al3. However, many factors are important and reproducible preparations require careful attention. The best method to have rigorously similar samples is to take them from one large batch of Raney-type nickel catalyst.
2.4 Properties of Raney-type nickel catalyst
The main advantage of Raney-type catalysts is that they do not require a reduction step prior to use as do conventional catalysts, which are in the form of the oxide of the active metal supported on a carrier. They have high activity since the BET surface area is essentially the metal surface area. The high metal content provides good resistance to catalytic poisoning. Also, the density of the material is relatively high, making separation of catalyst and product mixture relatively simple by sedimentation. The high thermal conductivity and magnetic properties of Raney-type catalyst are further advantages. Raney-type catalysts are kept under liquid to protect them from oxidation, since they are pyrophoric. This pyrophoric character makes handling difficult. The main disadvantage however, besides the pyrophoric character, is the lack of reproducibility of the preparation of the catalysts. The preparation procedure is relatively simple, but has so many parameters (i.e. leaching time, temperature, concentration of the sodium hydroxide solution), which all affect the performance of the final catalyst, that manufacturing a catalyst with predictable catalytic properties in a certain reaction is hard to accomplish batch after batch. Typical Raney-type Ni properties are listed in table 4.
Table 4: properties of Raney-type nickel catalysts.
Properties:
Total surface area (BET) 40 – 120 m2 /gcat Metallic surface 85 – 90 % Particle size 10 – 250 mm Crystallite size 2.5 – 15 nm Pore diameter 2 – 8 nm Pore volume 0.05 – 0.2 cm3 /g Typical composition (Ni/Al/Al2O3) 85/5/10
Apparent density 6 to 7.5 kg/m3
The activation process determines the structure and properties of Raney-type nickel catalysts, but the structure and the composition of the starting alloy also play an important role in the performance of the final catalyst [30]. Unfortunately, many discrepancies can be found in the literature on the properties of Raney-type nickel catalysts in relation to their preparation. For instance, a controversy still exists concerning the microstructural changes of the solid phase leading to the final catalyst. Gros et al. [18] suggest that leaching of Ni2Al3 occurs stepwise via the formation of intermediate phases, which are not present in the phase diagram. Presnyakov et al. [33] propose a sequence in which only the stable phases as present in the phase diagram of Ni-Al are formed due to continuous loss of aluminum during the leaching process. Delannay [20] reports that the formation of NiAl occurs by diffusion of nickel atoms into the precursor Ni2Al3 phases. In two recent publications, quasi in-situ XRD in combination with HREM was used to study the evolution of the microstructure of Raney-type nickel catalysts during leaching. Knies et al. [34] found that the leaching process leads to an intermediate non-equilibrium body-centered cubic phase of disordered nickel and aluminum atoms, which exists over a wide range of compositions. In this case, further loss of aluminum leads to a rearrangement of the remaining nickel atoms to form the face-centered cubic structure of metallic nickel. Rong Wang et al. [35] studied the leaching of a rapidly solidified Ni-Al alloy. They propose that Ni2Al3 transforms into metallic nickel as an advancing interface type process with Ni3Al2 as possible intermediate phase. The above mentioned discrepancies call for more investigations concerning the changes that occur in the precursor alloy during the leaching process. In chapters 3 and 4 of this thesis, the mechanism of transformation of the starting alloy into Raney-type nickel catalyst is studied.
2.5 Promoted Raney-type nickel catalysts
The addition of a second component to metal catalysts is widely used in order to enhance activity and/or selectivity. In the case of skeletal nickel catalysts, it is a simple procedure to add small amounts of a third metal during the alloy preparation stage. Although other metals have been used in laboratory studies, the most common metals used to promote skeletal nickel catalysts employed industrially are Fe, Cr and Mo. It is possible to use copper and cobalt but these two elements can also directly give a catalyst from a Cu-Al or Co-Al starting alloy (without nickel). They are known as Raney-type copper [5] and Raney-type cobalt [6]. Montgomery [8] has published a detailed study of the functional group activity of promoted Raney-type nickel catalysts. He prepared catalysts by leaching alloy powder of the type Al (58 wt %), Ni (37-42 wt %), M (0-5 wt %), where M = Co, Cr, Cu, Fe, Mo, in aqueous sodium hydroxide at 50 °C. Of the metals tested, molybdenum was found to be much more effective than the other metals for the hydrogenation of butyronitrile and acetone (table 4). In the case of the hydrogenation of sodium p-nitrophenolate, relative activities were comparable between the different promoted catalysts. It was found that the optimum level of promoter present in the precursor alloy was Cr = 1.5wt %, Mo = 2.2 wt %, Cu = 4.0 wt % and Fe = 6.5 wt % (table 5).
Table 5: Effect of metallic promoters on the hydrogenation of organic compounds [8].
Promoter (M) M wt% in alloy Reactant Relative activitya Mo Cr Fe Cu 2.2 1.5 6.5 4.0 Butyronitrile “ “ “ 6.5 3.8 3.3 2.9 Mo Cr Fe Cu 2.2 1.5 6.5 4.0 Acetone “ “ “ 2.9 1.5 1.3 1.7 Mo Cr Fe Cu 1.5 1.5 6.5 4.0 Sodium p-nitrophenolate “ “ “ 1.7 1.6 2.1 1.3 a
Ratio of reaction rate for promoted catalyst to reaction rate over unpromoted Raney-type nickel.
2.6 Applications
Raney-type nickel catalysts are applied in a large number of important industrial processes in both chemical and pharmaceutical industry, mostly hydrogenation reactions. In table 6, some typical industrial reactions are listed.
Table 6 : Industrial processes involving Raney-type catalysts [36].
Reaction/Functional Group Typical Reactant Product Application
Hydrogenation/Nitro 2,4-Dinitrotoluene 2,4-Toluenediamine Polyurethane
Hydrogenation/Diene 1,5,9-Cyclododecatriene Cyclododecane Nylon-6,12
Hydrogenation/Carbonyl 2-Ethylhexanal 2-Ethylhexanol Plastics
Glucose Sorbitol Sweetener
Hydrogenation/Nitrile Adiponitrile Hexamethylenediamine Nylon-6,6
Succinonitrile 1,4-Diaminobutane Nylon-4,6
Hydrogenation/Alkyne 1,4-Butynediol 1,4-Butanediol THF
Hydrogenation/Aromatic Benzene Cyclohexane Polyamides
Aminolysis/Alcohol 1,6-Hexanediol Hexamethylenediamine Nylon-6,6
Alkylation/Amine Dodecylamine Dimethyldodecylamine Surfactants
3. Framework of this thesis
Delft University of Technology has many contacts with industrial companies. Among these industries are catalyst manufacturers and catalyst users. In the past, signals were given from both parties that concerned the complexity of Raney-type Ni catalysts. The basics of the preparation procedure are relatively simple, but many parameters influence the performance of the final catalyst. As a consequence, the manufacturing of a reproducible catalyst with predictable catalytic properties for a specific reaction is hard to accomplish. Moreover, although a large amount of information on Raney-type Ni can be found in literature, still a fundamental study of the complete life cycle (from alloy to spent catalyst) of Raney-type catalysts is lacking or many different parameters have been used, so that general conclusions are hard to draw. It was for this reason that Delft University of Technology approached Engelhard (Raney-type Ni producer) and DSM (Raney-type Ni user) to set-up a research project, partly described in this thesis. The other part of the project is described in the thesis of Bram Hoffer [37] and concerns catalytic performance and kinetics. It was the goal of the project to predict the catalytic properties of this type of catalysts by studying the complete process of preparation and resulting catalytic properties. The elucidation of structure – activity relationships was the main goal of this project. The work was performed
at the National Centre for High-Resolution Electron Microscopy and the Reactor & Catalysis Engineering department, both at Delft University of Technology.
4. Outline of the thesis
Chapter 2 deals with a presentation of the main characterization techniques used in this thesis. The principle and a description of the technique are given in the case of X-ray Diffraction, Electron Microscopy and X-ray Photoelectron Spectroscopy. Chapters 3 and 4 focus on the mechanism of transformation of the starting alloy into Raney-type nickel catalysts. Different leaching times and different concentrations of sodium hydroxide solution were used. The hydrogenation of succinonitrile was used as test reaction. As promoters are commonly used in industry, chapter 5 deals with characterization, activity, selectivity and stability of commercial promoted Raney-type nickel catalysts. They are compared with several homemade catalysts using the hydrogenation of glucose to sorbitol as test reaction. In chapters 6 and 7, two different approaches in the preparation process and in the characterization process of Raney-type nickel catalysts are the subject of discussion. In chapter 6, a rapid solidification technique, melt- spinning, is used to prepare the starting alloy. In chapter 7, ultramicrotomy is used to prepare specimens of Raney-type nickel catalysts for Transmission Electron Microscopy.
References
[1] R.A. Van Santen, Theoretical Heterogeneous Catalysis, World Scientific Singapore, 1991, ISBN 981-02-0384-5.
[2] M. Raney, US patent 1,563,787 (1925). [3] M. Raney, US patent 1,628,191 (1927). [4] M. Raney, US Patent 1,915,473 (1933). [5] B.V. Aller, J. Appl. Chem. 7 (1957) 130.
[6] L. Fauconnau, Bull. Soc. Chim. France 5 (1938) 58. [7] R. Paul, Bull. Soc. Chim. France 13 (1946) 208.
[8] S.R. Montgomery, Catalysis of organic reactions, W.R. Moser (Editor) Dekker, New York, (1981) 383.
[9] W.F Mardsen, M.S. Wainwright and J.B. Friedrich, Ind. Eng. Chem. Prod. Res. Dev. 19 (1980) 551.
[10] J.A. Stanfield and P.E. Robbins, Actes Congr. Inter. Catal. (2nd
[11] J.P. Orchard, A.D. Tomsett, M.S. Wainwright and D.J. Young, J. Catal. 84 (1983) 189.
[12] M.F. Singleton, J.L. Murray, P. Nash, Binary alloys phase diagrams, Vol 1, T.B. Massalki (Editor) ASM international (1990) 183, ISBN 0-87170-404-8.
[13] R. Sassoulas and Y. Trambouze, Bull. Soc. Chim. France 5 (1964) 985.
[14] A.B. Fasman, V.F. Timofeeva, V.N. Rechkin and Y.F. Klyuchnikov, Kyn. and Khem. 13 (1971) 1513.
[15] J. Freel, W.J.M. Pieters and R.B. Anderson, J. Catal.14 (1969) 247.
[16] S. Hamar-Thibault, J. Gros, J.C. Joud, J. Masson, J.P. Damon, J.M. Bonnier, Preparation of catalysts V, G. Poncelet, P.A. Jacobs, P. Grange and B. Delmon (Editors), Elsevier Science Publishers B.V., Amsterdam (1991) 601, ISBN 0-444-88616-8.
[17] A.B. Fasman, S.D. Mikhailenko, O.T. Kalinina, E.Y. Ivanov and G.V. Golubkova, Preparation of catalysts V, G. Poncelet, P.A. Jacobs, P. Grange and B. Delmon (Editors), Elsevier Science Publishers B.V., Amsterdam (1991) 601, ISBN 0-444-88616-8.
[18] J. Gros, S. Hamar-Thibault and J.C. Joud, Surf. Interf. Anal. 11 (1988) 611. [19] J. Masson, P. Cividino and J. Court, J. Mol. Cat. A: Chem. 111 (1996) 289. [20] F. Delannay, Reactivity of solids, 2 (1986) 235.
[21] M. Khaidar, C. Allibert and J. Driole, Mat. Res. Bull. 17 (1982) 329.
[22] S. Sane, J.M. Bonnier, J.P. Damon and J. Masson, Appl. Catal. 9 (1984) 69. [23] J. Gros, S. Hamar-Thibault and J.C. Joud, J. Mat. Sci., 24 (1989) 2987. [24] P. Fournier and M. Henry, RGE, (1983) 314.
[25] J. Matthys, Melt spinning and strip casting: research and implementation, Warrendale, 1992, ISBN 0-87339-183-7.
[26] J. Freel, W.J.M. Pieters and R.B. Anderson, J. Catal. 16 (1970) 281.
[27] M.L. Bakker, D.J. Young and M.S. Wainwright, J. Mat. Sci. 23, (1988) 3921 [28] J.H.P. Tyman, Chem. Ind. (1964) 404.
[29] P. Fouilloux, Appl. Catal. 8 (1983) 1.
[30] M.S. Wainwright, Handbook of Heterogeneous catalysis, Vol 1, G. Ertl, H. Knozinger and J. Weitkamp (Editors), VCH Publ., New York (1997), ISBN 3-527-29212-8. [31] S. Nashumira and Y. Urushibara, J. Chem. Soc. Jap. 20 (1957) 199.
[32] P. Fouilloux, G.A. Martin, A.J. Renouprez, B. Moraweck, B. Imelik and M. Prettre, J. Catal. 25 (1972) 212.
[33] A.A. Presniakov, K.T. Chernousova, T. Kabiev, A.B. Fasman and T.T. Bocharova, J. Appl. Chem. USSR (Engl. Transl.), 40 (1967) 929.
[34] S. Knies, G. Miehe, M. Rettenmayr and D.J. Ostgard, Z. Metallkd. 92 (2001) 596. [35] Rong Wang, L. Zhilong and K. Tsun, J. Mat. Sci. 36 (2001) 5649.
[36] C. De Bellefon, Act. Chim. 4 (2000) 71.
[37] B. Hoffer, Tuning Raney-type nickel and supported Ni catalysts for commercial hydrogenation reactions, PhD thesis, Delft university press, Delft, 2003. ISBN-90-6464-2680.
Instrumentation
In this chapter, a short description of the main characterization techniques used in this thesis is given. For X-ray diffraction, electron microscopy (SEM and TEM coupled with EDX elemental analysis) and XPS, the principle of the technique and a description of the apparatus are presented. To understand what was the use of those techniques for the research described in this thesis, some examples are also given. As Raney-type catalysts are pyrophoric, a special vacuum transfer holder for transmission electronic microscopy has been used.
1. Introduction
There are many ways to obtain information on the physico-chemical properties of materials [1-3]. For example, one can irradiate a sample with rays and study how the X-rays are diffracted (X-ray diffraction, XRD), or one can study the energy distribution of electrons that are emitted from the sample as a result of the photoelectric effect (X-ray photoelectron spectroscopy, XPS). It is good to realize that, although many techniques undoubtedly provide valuable results, the most useful information always comes from a combination of several techniques. In this chapter, a description of the main characterization techniques used in this thesis (XRD, TEM, SEM and XPS) is given.
X-ray diffraction [4] is one of the oldest and most frequently applied techniques in catalyst characterization. In catalysis, it is mainly used to identify crystalline phases present in heterogeneous catalysts by means of lattice structural parameters and to obtain an indication of particle sizes.
Seeing atomic detail at the surface is the ideal for catalytic chemist. What can be seen with a light microscope is limited by the rather long wavelength of visible light (a few hundred nanometers), which does not permit detection of features smaller than one micrometer. Electron beams (wavelenght of about 2 pm) offer better opportunities [5,6]. Development over the last 40 years of electron microscope and especially Scanning Electron Microscopy (SEM) and Transmission Electron Microscopy (TEM) permit magnifications of the order of one million times and achieve resolution of 0.1 nm in the case of TEM. The main difference between SEM and TEM is that SEM sees contrast due to the topology and composition of a surface, whereas the electron beam in TEM projects all information on the material it encounters in a two-dimensional image. Both techniques have become very popular in catalyst characterisation.
X-ray Photoelectron Spectroscopy (XPS) is one of the most frequently used techniques in catalysis. It yields information on the elemental composition and the oxidation state at the surface where the catalytic reaction will occur. It involves the energy analysis of photoelectrons emitted from a sample generated from core level shells under impact from X-rays, usually Al or Mg Ka.
In this chapter, the principle of these techniques and the instrumentation required will be described along with the relevant information obtained: phase identification and sample composition.
2. X-Ray Diffraction
2.1 Production of X-rays
X-rays are electromagnetic radiation with typical photon energies in the range of 100 eV - 100 keV. For diffraction applications, only short wavelength X-rays (hard X-rays) in the range of a few angstroms to 0.1 angstrom (1 keV - 120 keV) are used. Because the wavelength of X-rays is comparable to the size of atoms, they are ideally suited for probing the structural arrangement of atoms and molecules in a wide range of materials. The energetic X-rays can penetrate deeply into the materials and provide information about the bulk structure.
X-rays are produced generally by either X-ray tubes or synchrotron radiation. In an X-ray tube, which is the primary X-ray source used in laboratory X-ray instruments, X-rays are generated when a focused electron beam accelerated across a high voltage field bombards a stationary or rotating solid target. As electrons collide with atoms in the target and slow down, a continuous spectrum of X-rays is emitted. The high energy electrons eject inner shell electrons from atoms through ionization processes. When a free electron fills the shell, an X-ray photon with an energy characteristic of the target material is emitted. Common targets used in X-ray tubes include Cu and Mo, which emit 8 keV and 14 keV X-rays with corresponding wavelengths of 1.54 Å and 0.8 Å, respectively. (The energy E of an X-ray photon and its wavelength are related by the equation E = hc/l, where h is Planck's constant and c the speed of light).
In recent years, synchrotron facilities have become widely used as preferred sources for X-ray diffraction measurements. Synchrotron radiation is emitted by electrons or positrons travelling at near light speed in a circular storage ring. These powerful sources, which are thousands to millions of times more intense than laboratory X-ray tubes, have become indispensable tools for a wide range of structural investigations and have brought advances in numerous fields of science and technology.
2.2 Bragg’s law
Figure 1: Scheme from which Bragg’s law is derived.
dsinq
Perhaps the most dramatic progress in understanding crystal structure came with the discovery of X-rays. In 1912, the German physicist Max von Laue (1879-1960) demonstrated that crystals would diffract X-rays, thus proving that minerals posses a regular and repeating internal arrangement of atoms. By 1914, W.H. Bragg (1862-1942) and his son W.L. Bragg (1890-1971) in Cambridge, England used X-rays to determine the structure of minerals. The equation they derived in doing so has now become known as Bragg’s Law:
nl = 2dsinq (1)
! Where
n is a positive integer (order of reflection); l is the wavelength of the X-rays;
d is the distance between layers of atoms in the crystal;
q is the angle between the incoming X-rays and the normal to the reflecting lattice plane.
When a monochromatic X-ray beam with wavelength l is projected onto a crystalline material at an angle q, diffraction occurs only when the distance traveled by the rays reflected from successive planes differs by a complete number n of wavelengths.
By varying the angle q, the Bragg's Law conditions are satisfied by different d-spacings in crystalline materials. Plotting the angular positions and intensities of the resultant diffracted lines of radiation produces a pattern which is characteristic of the crystalline material. Where a mixture of different phases is present, the resultant diffractogram is formed by addition of the individual patterns.
2.3 Diffractometer
The essential features of a diffractometer are shown in Figure 2. A powder specimen C, in the form of a plate, is supported on a table H, which can be rotated about an axis O perpendicular to the plane of the drawing. X-rays diverge from this source and are diffracted by the specimen to form a convergent diffracted beam, which comes to a focus at the slit F and then enters the detector. A (divergent slit) and B (receiving slit) are special slits which define and collimate the incident and diffracted beams. The monochromator or filter is usually placed in a special holder (not shown) in the diffracted, rather than the incident, beams; a monochromator or filter in the diffracted beam not only serves its primary function (suppression of Kb radiation) but also decreases background radiation originating in the specimen. The receiving slits and detector are supported on the carriage E, which may be
rotated about the axis O and whose angular position 2q may be read on the diffractometer circle. The supports E and H are mechanically coupled so that a rotation of the detector through 2q degrees is automatically accompanied by rotation of the specimen through q degrees. This coupling ensures that the angle of incidence (q1) on the flat specimen always equals the diffracted angle (q2), and both equal half the total angle of diffraction (2q), an arrangement necessary to preserve focusing conditions. Modern automated diffractometers generally collect data with the detector and sample set at a large number of fixed angles spaced by an angular increment of the order of 0.01 degree; the length of time counted and the size of the angular increment are controlled through software.
Figure 2: Construction of a diffractometer.
2.4 Phase identification
A given crystalline material always produces a characteristic diffraction pattern, whether that material is present in the pure state or as one constituent of a mixture of phases. Qualitative analysis for a particular phase is accomplished by identification of the pattern of that phase. Thus, a collection of diffraction patterns for a great many substances allows identification of an unknown material by recording its diffraction pattern and then locating in the file of known patterns the one which matches the pattern of the unknown material exactly.
The task of building up a collection of known patterns was initiated by Hanawalt, Rinn and Frevel at Dow Chemical Company. They obtained and classified diffraction data on some 1000 different substances. It appeared that those data were of great potential value to a wide range of industries. In 1941, several societies, including the American Society for Testing and Materials, began to cooperate in acquiring diffraction data. Since 1969 this activity has been carried out by the Joint Committee on Powder Diffraction Standards (JCPDS) which in 1978 was renamed the International Center for Diffraction Data (ICDD). Approximately 300 scientists from around the world participate in the Center. In 1995, the Powder Diffraction File (PDF) contained nearly 62,000 diffraction patterns. The substances included are elements, inorganic compounds, minerals, organic compounds and organometallic compounds. Figure 3 gives an example of phase identification. The commercial alloys (A and B) used in this thesis as starting material in the production of Raney-type nickel catalysts are 50-50 wt % Ni-Al alloys. By means of XRD, three phases are found to be present in these alloys: NiAl3, Ni2Al3 and Al, identified with the PDF.
Figure 3: Diffractograms of two (A and B) 50-50 wt % Ni-Al alloys used in the preparation process of Raney-type nickel catalysts. u represents Ni2Al3 main lines, t Al main lines, the other lines belong
to NiAl3.. 2.5 Crystallite size
If the path differences between X-ray photons scattered by the first two planes of atoms differs only slightly from an integral number of wavelengths, then the plane of atoms scattering X-rays exactly out of phase with the photons from the first plane will lie deep within the crystal. If the crystal is so small that this plane of atoms does not exist, then
15 25 35 45 55 2q [0] In te nsity [A.U.] {101} {001} A-alloy B-alloy
complete cancellation of all scattered X-rays will not result. It follows that with the decrease of crystal size, Bragg conditions will be less stringent and thus reflections will broaden. The result is that very small crystals cause broadening (a small angular divergence) of the diffracted beam. A treatment of this problem will lead to the Scherrer equation, which is used to estimate the size of very small particles from the measured width of their diffraction peak.
L B cosq = 0.9l (2)
Where
L is the size of the crystallite in Å; l the X-ray wavelength in Å;
q is the Bragg angle for the reflection;
B is the observed peak width at half maximum intensity for any given reflection.
3. Electron microscopy
3.1 Electrons
Figure 4: Interaction between the primary electron beam and the sample.
Electrons have characteristic wavelenghts of less than an angstrom, and come close seeing atomic detail. Figure 4 shows that interaction between the primary electron beam (1 kV to 1.5 kV) leads to a number of detectable signals:
- Depending on sample thickness, a fraction of electrons passes through the sample without suffering energy loss. As the attenuation of the beam depends on density and thickness, the transmitted electrons form a two-dimensional projection of the sample (TEM, HRTEM).
- Electrons are diffracted by particles if these are favorably oriented. This may lead to dark field images as well as crystallographic information (TEM).
- Electrons can collide with atoms in the sample and be backscattered (SEM).
- Auger electrons and X-rays are formed in the relaxation of core-ionized atoms (AES, EDX). - Electrons excite characteristic vibrations in the sample, and these can be studied by analyzing the energy loss suffered by the primary electrons (EELS).
- Many electrons lose energy in cascade of consecutive inelastic collisions. Most of the secondary electrons emitted by the sample had their last loss process in the surface region (SEM).
Thus, the interaction of the primary beam with the sample can yield a lot of information on morphology, crystallography and chemical composition.
3.2 Transmission Electron microscopy (TEM) 3.2.1 Instrumentation
In addition to a source of electrons (electron gun), a TEM is a complex assembly of magnetic lenses, several apertures, a sample holder and an image recording/viewing system. The magnetic lenses can be grouped into those of the illumination system between the electron gun and the sample and those of the imaging system after the sample. Typically one finds two condenser lenses in the illumination system and three lenses in the imaging system. Consider first the illumination system, shown at the top of figure 5. The conventional TEM mode adjusts the condenser lens to illuminate the sample with a nearly parallel beam (convergence angle a < 10-4
rad compared to electron diffraction angles on the order of 10-2 rad). Deflection coils are often used to control beam position or angle. The beam strikes the specimen and parts of it are transmitted. This transmitted portion is focused by the objective lens into an image. The image is passed down the column through the intermediate and projector lenses, being enlarged all the way. The image strikes the phosphor imaging screen (or CCD camera) and light (or electron current) is generated, allowing the user to see the image. The darker areas of the image represent those areas of the sample that fewer electrons were transmitted through (they are thicker or denser). The lighter areas of the image represent those areas of the sample that more electrons were transmitted through (they are thinner or less dense).
Figure 5: Ray diagram for viewing the diffraction pattern (a) and the image (b) of the sample.
In TEM one can switch between imaging the sample (figure 5b) and viewing its diffraction pattern (figure 5a) by changing the strength of the intermediate lens. To see the diffraction pattern, the intermediate lens is adjusted to focus on the back focal plane of the objective lens; i.e. the back focal plane of the object lens acts as the object plane for the intermediate lens. In the imaging mode, the intermediate lens is adjusted so that its object plane is the image plane of the objective lens.
3.2.2 Imaging modes, diffraction mode, elemental analysis
3.2.2.1 Bright Field (BF) and Dark Field (DF) Imaging
This primary imaging mode takes advantage of mass contrast or diffraction contrast to image the internal microstructure of materials. It is commonly used to image grain and defect structures (i.e. dislocations, voids, stacking faults and twins) within materials. Precipitates or inclusions are also easily observed by this technique. The range of features that can be resolved by this type of imaging is hundreds of µm to 1 nm.
Similar in purpose to the BF technique, the DF imaging mode makes use of the specific Bragg diffracted electrons to image the region from which they originated. This allows the experimenter to link diffraction (i.e., crystallographic) information with specific regions or phases in the sample.
3.2.2.2 High Resolution electron microscopy (HREM)
This highly specialized imaging mode is used to directly resolve the atomic structure of crystalline materials. By orienting regions of the sample along specific directions through thin (< 10 nm) crystals, it is possible to resolve individual columns of atoms of many materials. For example, when used in conjunction with imaging simulation calculations, it is possible to quantitatively deduce the structure of interfaces and defects.
3.2.2.3 Selected Area Diffraction pattern (SAED)
It is possible to view and record the electron diffraction pattern from selected areas of the specimen as small as ~ 1 µm by placing an aperture in the image plane and then projecting the diffraction pattern of that image onto the recording plane. Nanodiffraction is a special form of SAED in which the emphasis is on obtaining diffraction patterns from regions of the specimen about 1 nm or less in diameter. Unless a field-emission gun (FEG) is used, the intensity in a beam 1nm in diameter is too small to be useful. Electron beam diameters may be as small as 0.2 nm. The primary purpose of electron diffraction techniques is to identify the crystal structure of the materials under investigation.
3.2.2.4 Convergent beam electron diffraction (CBED)
By forming a small diameter (~ 1.0 nm) beam on the surface of the specimen with a convergent angular range of illumination, the conventional SAED pattern will contain diffraction discs (instead of spots). Within these discs, additional detail can be seen that is
related to the crystallographic structure. At higher scattering angles, higher order Laue circles can be observed. In the thicker regions of the specimen, Kikuchi lines will also be seen. These diffraction details reveal additional three-dimensional crystallographic and symmetry information about materials.
3.2.2.5 Energy dispersive X-ray (EDX)
Figure 6: EDX system [7].
Characteristic X-rays are produced when high-energy electrons interact with the specimen. These X-rays are detectable using a SiLi detector. Figure 6 shows an EDX system. When X-ray photons are captured by the detection crystal they create electron-hole pairs. These electron-hole pairs are further converted to voltage pulse by a charge-to-voltage converter (preamplifier). The signal is further amplified and shaped by a linear amplifier and finally passed to a computer X-ray analyzer (not shown in the figure) where the data are displayed as a histogram of intensity by voltage. The key to understanding how an energy-dispersive spectrometer (EDX) works is to recognize that each voltage pulse is proportional to the energy of the incoming X-ray photon.
Figure 7 gives an example of an EDX spectrum: the identification of Ni2Al3 in the commercial alloy used to produce Raney-type nickel catalysts. The resulting spectra can be quantitatively analyzed for the presence and amount of element(s) present in the sample. EDX is hard to quantify for low-Z elements.
Figure 7: EDX analysis, identification of Ni2Al3 in a 50-50 wt % Ni-Al alloy. 3.3 Scanning Electron Microscopy
The signals most commonly used in SEM are secondary electrons, backscattered electrons and X-rays.
3.3.1 Secondary electrons
Secondary electrons are predominantly produced by the interactions between energetic incident electrons and weakly bonded conduction-band electrons in metals or valence electrons of insulators and semiconductors. There is a great difference between the amount of energy contained by incident electrons compared to the specimen electrons and because of this, only a small amount of kinetic energy can be transferred to the secondary electrons. An electron detector is used in SEM to convert the radiation of interest into an electrical signal for manipulation and display by signal processing electronics. Most SEM's are equipped with an Everhart-Thornley (E-T) detector. It works in the following manner: The scintillator material is struck by an energetic electron. This collision produces photons which are conducted by total internal reflection in a light guide to a photomultiplier. These photons are now in the form of light so they can pass through a vacuum environment and a quartz glass window. The photon is then converted back into an electron current where a positive bias attracts the electrons and collects them so that they are detected and form an image.
3.3.2 Backscattered electrons
Although secondary electron images are obtained most frequently in SEM, backscattered electron images also provide important information. Backscattered electrons vary in their amount and direction with the composition, surface topography, crystallinity and magnetism of the specimen. The contrast of a backscattered electron image depends on (1) the backscattered electron generation rate, which in turn depends on the mean atomic number of the specimen and (2) angle dependence of backscattered electrons at the specimen surface. The backscattered electron image contains two types of information: one on specimen composition and the other on specimen topography. To separate these two types of information, a paired semiconductor detector is provided symmetrically with respect to the optical axis. Addition of the signals of both detectors gives a composition image while subtraction gives a topography image. And with composition images of crystalline specimens, the difference in crystal orientation can be obtained as the so-called "channeling contrast," by utilizing the advantage that the backscattered electron intensity changes to a large extent before and after Bragg's condition.
The generation region of backscattered electrons is larger than that of secondary electrons, namely, several tens of nm. Therefore, backscattered electrons give poorer spatial resolution than secondary electrons. But because they have a higher energy than secondary electrons, they are less influenced by charging and specimen contamination.
3.3.3 EDX
As for transmission electron microscopy, SEM is often equipped with an EDX analysis system. The principle of EDX analysis has been described in section 3.2.2.5 of this chapter.
4. XPS
4.1 Principle
By absorbing a photon, an atom gains an amount of energy equal to hn. It then releases an electron to regain its original stable energy state. The released electron retains all the energy from the incident photon. It can then escape from the atom, and even further from matter and kinetic energy keeps it moving. With XPS, incident photons usually carry an energy ranging from 1 to 2 keV. The relatively high level of the incident energy causes the matter to release an electron from an atomic internal shell. Consequently, there will be some
atoms lacking electrons in the internal shells from which photoelectrons have been released. To recover from this ionized state the atom can emit another electron.
Figure 8: Photoemission process.
The principle of the conservation of energy allows us to write the energy balance equation, valid for the absorption of a photon carrying an energy of hn (figure 8):
hn = E kinetic + E binding + j work function (3)
Where
h is Planck’s constant;
n is the frequency of the exciting radiation;
E kinetic is electron kinetic energy of the photoelectron when leaving the specimen; Ebinding is electron binding energy inside the atom;
jwork function is the work function of the spectrometer.
Since the work function can be compensated artificially, it is eliminated, giving the binding energy as follows:
E Binding = hn - E Kinetic (4)
Once the photoelectrons are emitted out of the sample surface, a positive charge zone will be quickly established in the sample surface. As a result, the sample surface acquires a positive potential (varying typically from several volts to tens of volts) and the kinetic energies of core electrons are reduced by the same amount C.
E Binding = hn - (E Kinetic – C) (5)
It can be seen that the surface charging results in the shift of the XPS peaks to higher binding energy. In this case, the binding energy has to be calibrated with an internal reference peak. The C 1s peak from contamination, with a binding energy of 284.8 eV, is commonly used as the reference for calibration. In order to neutralize the surface charge during data acquisition, a low-energy electron flood gun is used to deliver electrons to the sample surface. The electron flood gun can be tuned to provide the correct current to push the XPS peaks back to the real position.
The core electron of an element has a unique binding energy, which can be described as a "fingerprint" for the element. Thus almost all elements except for hydrogen and helium can be identified by measuring the binding energy of their core electron. Furthermore, the binding energy of the core electron is very sensitive to the chemical environment of element. The same atom can be bonded to different chemical species, leading to changes in the binding energy of its core electron. The variation of binding energy results in the shift of the corresponding XPS peak, ranging from 0.1 eV to 10 eV. This "chemical shift" can be used to study the chemical state of element in the surface.
Since the number of photoelectrons corresponding to a specific element is dependent on the atomic concentration of that element in the sample, XPS is not only used to identify the elements but also to quantify the chemical composition.
4.2 Instrumentation
The basic requirements for an XPS experiment are illustrated in figure 9. X-rays from an aluminum (Al Ka = 1486.3 eV) or magnesium (Mg Ka = 1253.6 eV) cathode strike the sample and eject electrons from it. The electrons are focused with a lens system (not shown) and then, to increase resolution, delayed before they enter the analyser. An electron energy analyser (which can disperse the emitted electrons according to their kinetic energy, and thereby measure the flux of emitted electrons of a particular energy). There are many different designs of electron energy analyser but the preferred option for photoemission experiments is a concentric hemispherical analyser (CHA) which uses an electric field between two hemispherical surfaces to disperse the electrons according to their kinetic energy. The number of impulses received by the detector in a given time interval, the count rate, is plotted as ordinate against the binding energy (deduced from the kinetic energy using equation 4) in eV to give the XPS spectrum. A high vacuum environment is required to enable the emitted photoelectrons to be analysed without interference from gas phase collisions.
Figure 9: Construction of an XP spectrometer. 4.3 Line identification.
Figure 10: XPS spectrum of Raney-type nickel catalyst prepared by leaching the starting alloy during 5 min at 80 °C with a 20 wt % sodium hydroxide solution
In XPS, one measures the intensity of photoelectrons N(E) as a function of their kinetic energy. The XP spectrum, however, is usually a plot of N(E) versus the binding energy. Figure 10 shows the XP spectrum of a home-made Raney-type nickel catalyst (A 5 min leached at 80 °C with a 20 wt % sodium hydroxide solution, see also chapter 3 of this thesis). In addition to the expected photoelectron peaks in figure 10, the spectrum also contains peaks due to Auger electrons. Auger peaks can be recognized by recording the spectrum at two different X-ray energies: the XPS peaks appear at the same binding energy, while the Auger peaks will shift on the binding energy scale. This is the main reason why X-ray sources often contain a dual anode of Mg and Al. By varying the X-X-ray energy, Auger peaks are identified and overlap between XPS and Auger peaks can be avoided. Photoelectron peaks are labelled according to the quantum level from which the electron originates. As the spin may be up (s = +1/2) or down (s = -1/2) for p, d and f levels, these levels become split upon ionization. Thus photoelectron peaks from core levels come in pairs (doublet) except for the s level, which gives a single peak. In case of a doublet, the difference in energy is called spin-orbit splitting. Figure 11 shows the Ni 2p area of a NiAl single crystal. A difference of 17.4 eV between the Ni 2p3/2 and Ni 2p1/2 is characteristic of metallic nickel (Ni2+
in NiO would have given a difference of 18 eV).
In general, interpretation of the XP spectrum is most readily accomplished first by identifying those lines that are almost always present (specifically those of carbon and oxygen), then by identifying major lines and associated weaker lines, and lastly by identifying the remaining weak lines. Most modern, commercially available spectrometers have peak identification algorithms within their data reduction packages. Data such as binding energies and chemical shift doublet separation are available in the Handbook of X-ray photoelectron spectroscopy [8] and also on the internet at the NIST X-X-ray Photoelectron Spectroscopy Database website [9].
Figure 11: Ni2p area of a NiAl single crystal: the doublet separation.
Ni2p
Ni2p1/2
Ni2p3/2
17.4 eV
880 870 860 850
4.4 Quantitative analysis
For many XPS investigations, it is important to determine the relative concentrations of the various constituents. Methods have been developed for quantifying the XPS measurements using peak area and peak height sensitivity factors. The method using peak area sensitivity factors is the more accurate.
The number of photoelectrons in a specific peak is given by:
I = nfsqlyAT (6)
Where
n is the number of atoms of the element per cm3
of the sample; f is the X-ray flux in photons.cm-2
.sec-1 ;
s is the photoelectric cross section for the atomic orbital of interest in cm2;
q is an angular efficiency factor for the instrumental arrangement based on the angle between the photon path and detected electron;
y is the efficiency in the photoelectric procees for formation of photoelectrons of the normal photoelectron energy;
l is the mean free path of the photoelectrons in the sample;
A is the area of the sample from which photoelectrons are detected; T is the detection efficiency for electrons emitted from the sample. From equation (6):
n = I / fsqlyAT (7)
The denominator in equation (7) can be defined as the atomic sensitivity factor, S. If we consider a strong line from each of two elements, then:
Values of S are based on empirical data [10], which have been corrected for the transmission function of the spectrometer. Thus, for any spectrometer it is possible to develop a set of relative S values for all the elements. Ratios between two different elements can thus be determined.
n1
n2 I2/S2 I1/S1 =
5. The vacuum transfer holder (TEM)
Raney-type nickel catalysts are extremely pyrophoric. This means that for each experiments, for each techniques used, the home-made catalysts must not be exposed to air. The preparation process for each technique will be described in the different chapters of this thesis. However, preparing the sample is only the first step in our experimental process. As Raney-type nickel catalysts are easily oxidised by exposure to air, the transfer of the sample to the desired characterization equipment is very important. For XRD and XPS, it is common to work under protective atmosphere, but for TEM, the use of a special vacuum transfer holder [11] is necessary. By assuring exclusion of oxygen during the transfer, it made our TEM experiments relevant. The vacuum transfer holder is shown in figure 12. The specimen area is introduced the glovebox (O2 and H2O less than 1 ppm) via an airlock similar to the specimen microscope airlock (figure 12a). The sample is prepared inside the glovebox and the microscope grid is fixed in the vacuum transfer holder. Still in open position (figure 12b), the specimen area is placed in the airlock chamber (under argon) of the glovebox. The chamber is evacuated during 10 min. By a screw system, the vacuum transfer holder can be closed (figure 12c) from the outside. Once it has been closed, the airlock chamber is filled with argon and the transfer holder can be completely removed and carried safely to the microscope.
Figure 12: a: View of the vacuum transfer holder, the specimen compartment is inside the glovebox; b: open position; c: closed position.
a
References
[1] J.W. Niemantsverdriet (Editor), Spectroscopy in Catalysis, second edition, Wiley-VCH, Weinheim, 2000. ISBN 3-527-30200X.
[2] G. Ertl, H. Knozinger, J. Weitkamp (Editors), Handbook of Heterogeneous Catalysis, VCH, Weinheim, 1997, ISBN 3-527-29212-8.
[3] J.L.G Fierro (Editor), Spectroscopic Characterization of Heterogeneous Catalysis, Elsevier, Amsterdam, 1990, ISBN 0-444-88242-1.
[4] B.D. Cullity, S.R. Stock (Editors), Elements of X-ray Diffraction, third edition, Prentice Hall, Upper Saddle River, 2001, ISBN 0-201-61091-4.
[5] D.B. Williams, C.B. Carter (Editors), Transmission Electron Microscopy, a Textbook for Materials Science, Plenum, New York, 1996, ISBN 0-306-45247-2.
[6] S. Amelinckx, D. Van Dyck, J. Van Landuyt, G. Van Tedeloo (Editors), Handbook of Microscopy: Methods 1, VCH, Weinheim, 1997, ISBN 3-527-29280-2.
[7] I. Goldstein, E. Dale, P. Echlin, D. C. Joy, A.D. Romig, E. Charles, C. Fiori, E. Lifshin (Editors), Scanning Electron Microscopy and X-Ray Microanalysis, 2nd Edition, Plenum Press, New York,2003, ISBN 0-306-47292-9.
[8] J.F. Moulder, W.F. Stickle, P.E. Sobol, K.D. Bomben, J. Chastain (Editors), Handbook of X-Ray Photoelecton Spectroscopy, Perkin Elmer Corporation, Eden Prairie, 1992 ISBN, 0-962-70262-5.
[9] http://srdata.nist.gov/xps/
[10] C.D. Wagner, Surf. Inter. Anal. 3 (1981) 211.
[11] H.W. Zandbergen, P.J. Kooyman and A.D. van Langeveld, Electron Microscopy 1998, proceedings ICEM 14, Cancun, Mexico, 31 August to 4 September 1998, Symposium W, Volume II, 491.
Genesis of the active phase in Raney-type nickel Catalysts
The parameters involved in the preparation process of Raney-type nickel catalysts have a large influence on the properties of the resulting catalyst. The morphology and microstructure of two commercial 50-50 wt % nickel aluminum alloys (A and B) have been studied. As expected from the phase diagram, NiAl3, Ni2Al3 and a small amount of metallic aluminum were observed. The distribution of the two main phases consists of a core of Ni2Al3 with a shell of NiAl3. Even if the composition of the starting alloys was the same, the amounts of the different phases are different: more Ni2Al3 and less metallic aluminum were present in alloy B. Treatment with an aqueous sodium hydroxide solution (leaching process) leads to Raney-type nickel catalyst. Different leaching times were used in order to follow an evolution in the morphology of the catalysts. In this way, intermediate phases can be studied using quasi
in-situ techniques: XRD, XPS and HREM. The activation process is very fast. The reaction of
Ni-Al alloy (50-50 wt %) with sodium hydroxide leads to nanometer-sized metallic nickel particles (7-8 nm) within five minutes. Simultaneously, two aluminum-rich oxidic phases, Bayerite and nickel aluminum oxide hydrate, are formed and these require a longer leaching time to be dissolved. Once these two Al-rich oxidic phases have disappeared, the metallic nickel particles become even smaller (3- 4 nm) to give the final Raney-type nickel catalysts. XPS shows that most of the aluminum is present at the surface as Al3+
species for short leaching time and essentially as metallic aluminum for longer leaching time. A strong correlation exists between the catalytic performance and the amount of Al-rich phases.
1. Introduction
Raney-type nickel catalysts [1] are widely used for hydrogenation reactions both in industry as well as in the laboratory. Raney-type nickel catalyst is prepared by leaching a nickel aluminum alloy (usually 50 – 50 wt %) using an alkali solution at temperatures between room temperature and 100 °C [2]. The resulting highly pyrophoric catalyst is envisaged to contain mainly metallic nickel, but depending on the preparation conditions also varying amounts of aluminum and alumina (formed during leaching) are present.
The use of Raney-type nickel catalysts has many advantages over supported transition metal catalysts. The production process is relatively easy and inexpensive. Since the Raney-type catalysts are already in the metallic form, they do not have to be reduced at elevated temperatures (which can be as high as 250 °C for supported Ni catalysts). Raney-type nickel catalysts are denser than supported catalysts, making catalyst separation from the liquid product after reaction relatively easy. Ni is much cheaper and, due to the large active metal area, not as sensitive to poisoning as noble metals like Pt, Pd and Ru.
As already mentioned in chapter 1, many discrepancies can be found in the literature on the properties of Raney-type nickel catalysts in relation to their preparation. For instance, a controversy still exists concerning the microstructural changes of the solid phase leading to the final catalyst. Several mechanisms for the evolution of the nickel structure can be found in the literature:
- Gross et al. [3] suggested the formation of intermediate phases, which are not present in the phase diagram.
- Presniakov et al. [4] proposed the nucleation and growth of NiAl and Ni3Al. - Delannay [5] reports that the formation of NiAl occurs by diffusion of dissolved
nickel atoms into the precursor Ni2Al3 phases.
- Knies et al. [6] claimed that the leaching process leads to an intermediate non-equilibrium body-centered cubic phase of disordered nickel and aluminum atoms, which exists over a wide range of compositions.
- Rong Wang et al. [7] reported that Ni2Al3 transforms into metallic nickel as an advancing interface type process with Ni3Al2 as possible intermediate phase.
Next to the bulk structure of the resulting Raney-type catalysts, the surface (where the catalytic reactions occur) is of particular interest. The bulk structure can be studied using X-ray diffraction (XRD), whereas the surface can be investigated using X-X-ray photoelectron spectroscopy (XPS). For a detailed characterization of the structure, high-resolution transmission electron microscopy (HREM) is a very useful technique. As Raney-type
