RICERCA ITALIANA     

New redox catalysts for new reactor technologies.

Università degli Studi di Napoli "Federico II"

Abstract

The research program " New Redox Catalysts for New Reactor Technologies" is based on different elements of possible innovation that are:
1) The improvement of the catalysts preparation methods, in order to obtain a more dispersed active component with a better control on: the redox and acid-base properties of the active site and the characteristics of the chemical environment of the active site. In particular, vanadiumm, molibdenum, wolframium and chromium based catalysts, that have already shown good activity and selectivity in the oxidative dehydrogenation (ODH) of hydrocarbons to the corresponding olefins, will be studied. Innovative methods of catalysts preparation will be developed, based on: a) grafting metal alkoxides on the surface of oxides rich of hydroxyls; b) flame hydrolysis of the precursors, followed by deposition by dip-coating on a proper support, such as cordieritic honeycombs pre-treated with a proper primer and/or on porous membranes; c) impregnation following the "equilibrium adsorption" technique.
2) The use of innovative reactors such as redox-decoupling reactors and membrane or honeycomb-type catalytic reactors. Catalysts for these reactors will be prepared by using the traditional methods of preparation by impregnation and the new methods based on alkoxide grafting or on deposition by dip-coating.
3) The employment of sophisticated techniques of surface characterisation, in order to define better the redox and acid base properties of the active site affecting activity and selectivity.
4) The use of suitable test reactions, to evaluate the catalysts performances. These test reactions, that are the oxidative dehydrogenation (ODH) of respectively propane, butane and ethylbenzene to the corresponding olefins are also reactions of great industrial interest, any reaction product being a building block originating many other industrial products. In particular, for these reactions the short contact times, without introduction of inacceptable pressure drop along the catalyst bed, allowed by the honeycomb reactor, permit a substantial improvement of selectivity to the olefin, which constitutes the desired intermediate product. Innovation in this field is of extreme interest for the industry. <<<

Principal Investigator

Elio SANTACESARIA Università degli Studi di NAPOLI "Federico II"

Research Objectives

This Research Program has different objectives, both scientific and technologic. However, one of the main objectives is to improve the performance of heterogeneous redox catalysts, aiming at approaching the feasibility of the oxidative dehydrogenation (ODH) industrial processes.
It is widely believed by the researchers active in this field that the improvement of the performance can be achieved by coupling the development of new catalytic systems with new reactor configurations, which may help in overcoming the limits which are intrinsic in conventional technologies. Specifically, one key aspect for the improvement of selectivity in redox catalysis is the achievement of a control of the redox properties of the catalyst, and of the nature of the oxygen species which develop in the reaction environment, as a consequence of the activation of molecular oxygen. This justify the study of the catalysts performances in different test reactions such as: the ODH of propane, butane and ethylbenzene to the corresponding olefins and the ODH of methanol and/or ethanol to the corresponding aldheydes. The firstly mentioned reactions occur at high temperatures (400-500°C), while, the last ones occur at much lower temperature (150-200°C), probably, involving different catalytic sites.
The study of the ODH of the hydrocarbons will be oriented to evaluate:
1) the effect of the catalysts preparation methods on the catalytic performances by adopting, in particular new preparation techniques based on:
(a) the grafting technique;
(b) the flame hydrolysis and the dip-coating techniques, in order to prepare in both cases well dispersed catalysts in which the active sites have a controlled chemical environment;
(c) the evaluation of the effect of the support in the original form or properly modified on the surface.
2) the effect of the adopted reactor, by comparing the performance of the traditional tubular continuous reactor with those of redox-decoupling reactor, of catalytic membrane reactor and of honeycomb reactor.
All the reactions that will be used as model, or test reactions, for determining the performance of the prepared catalysts and of the innovative reactors are perfectly suitable for the scope. However, the ODH of propane will be considered as a reference reaction for testing different reactor systems. It must be pointed out, at last, that all the considered reactions are of extreme interest for the industry. Therefore, satisfactory results in the catalysts performance will open perspectives for the industrial development of the corresponding processes. In addition, the achievement of the general objective of the research for an improvement of the redox catalysts performance will generate new technological objectives, for developing new industrial processes for obtaining with high selectivity propene, butenes, isobutene from the ODH of the corresponding alkanes. <<<

Timescale

24 months

National and international background

The industrial applications of the oxidative/reductive catalysis, often referred to as redox catalysis, are several. Amongst others, of particular relevance are the oxidation and ammoxidation of alkylaromatics, the oxidation and ammoxidation of olefins, the oxychlorination of olefins, the oxidation of n-butane to maleic anhydride. There are consolidated technologies, widely employed at industrial level, for which however there remain wide margins for considerable improvement. Specifically, one of the most important targets is the achievement of improved selectivity to the product of interest. In fact, in all cases the reactions of total combustion, or the formation of other by-products, lower the selectivity to the desired product. What said above is also valid in the case of a reaction which has not yet found industrial application, i.e., the oxidative dehydrogenation (ODH) of light alkanes to the corresponding olefins (1-11).
The production of olefins, building blocks of the petrochemical industry, is currently realised by the strongly endothermal processes of catalytic dehydrogenation, steam-cracking and catalytic cracking. These processes are carried out at considerably high temperatures, because of thermodynamic constraints, and therefore problems arise due to i) the need for construction materials which have to withstand severe operative conditions, ii) catalysts which easily undergo deactivation due to coke formation, thus needing frequent regeneration, and iii) very high energetic costs, due to the need of supplying heat to a system operating at high temperature. All this leads to processes characterised by very high investment and operative costs. Therefore, there is a considerable interest in the chemical industry to develop new technologies, able to overcome the thermodynamic constraints, through operation with an oxidizing agent able to abstract hydrogen from the saturated reactant molecule and to yield water, through an exothermal reaction not limited by thermodynamic equilibrium. A significant example is represented by the oxidative dehydrogenation of paraffins, carried out in the presence of molecular oxygen. Other advantages are the possibility to operate at temperatures lower than 500°C, thus with construction materials less expensive than those employed for endothermal processes, and to limit the deactivation phenomena of the catalyst due to coke formation. The main disadvantage of oxidative dehydrogenation processes, where water is coproduced instead of hydrogen, is just the loss of the valuable hydrogen. However, oxidative dehydrogenation might result particularly convenient with respect to traditional technologies for the specific production of olefins as intermediates in the production of other chemicals. For instance, a oxidehydrogenation process might be easily integrated with down-stream processes which transform the olefin, particularly when the down-stream process is an oxidative one.
The interest for the reaction of paraffins oxidehydrogenation is witnessed by the high number of scientific publications and patents issued in recent years.
One of the key factors for improving the selectivity in the redox processes is the control, under the reaction conditions, of the redox properties of the catalyst surface, that is the control on the type of oxygen ions formed in the reaction conditions as a consequence of the oxygen activation. A significant improvement in the performance could be obtained by coupling the development of catalysts prepared by new methods with the use of new reactors.
However, the main challenge remain (a) the development of an efficient catalytic system for this reaction, able to catalyse the transformation with high selectivity and productivity, (b) the development of reactor technologies able to take advantage of the control of the paraffin-to-oxygen ratio in the feed, which is one of the most important parameters in the control of catalytic performance. This is also useful to increase the safety aspects of the process, avoiding the formation of explosive mixtures, and thus increasing the flexibility of the process. (c) the use of reactors allowing to reduce substantially the contact time, favouring the selectivity to the desired intermediate, without excessive formation of total combustion byproducts.
As for the catalytic systems, the following aspects must be considered:
(i) the nature and dispersion of the active component and the chemical environment surrounding the active site.
(ii) the redox and acid-base properties of the active site and the effect of these properties on activity and selectivity.
(iii) The influence of the support, as a consequence of its physical and chemical characteristics.

Some of these aspects have already been studied for the catalysts prepared by the traditional way, i.e. by impregnation and data are available in the literature. Catalysts and supports prepared by grafting alkoxides have been less studied.
The preparation of new catalysts and supports can be performed by grafting a metal alkoxide on the hydroxyls-rich surface of an oxide [12-25]. It is possible to cover completely a surface with a monolayer of the alkoxide by putting the support in contact with a solution containing an excess of the alkoxide. The obtained solid, after the grafting reaction, can be washed with the alkoxide solvent and the residual alkoxide groups on the surface can be removed by hydrolysis with steam (steaming) or by combustion with air at higher temperature (calcinating). In this way the surface of the original support is completely modified as for the chemical acid-base and/or redox properties, according to the type of alkoxide used. It is also possible to cover the surface with a submonolayer, obtaining in this way more or less dispersed catalysts with catalytic sites possessing particular acid-base or redox properties. Moreover, it is possible to repeat more times the sequence of operations: grafting, steaming and calcination, in order to obtain a multi-layers coating. We can so obtain a new better support with surface chemical properties completely different with respect to the original one, but largely retaining mechanical and morphological properties. By this procedure it has been obtained, for example, SiO2 coated with a multilayer of TiO2 [15,21,26,27] having a specific surface area of about 300 m2/g, that of TiO2 prepared via precipitation being, normally, not higher than 100 m2/g. Moreover, the first support is much more resistant to sintering. In the catalyst and support preparation by grafting, the alkoxide to be used for the anchorage, can be opportunely modified before the operation in order to: (i) acquire a greater solubility in an apolar solvent by exchanging, for example, one or more alkoxide group with other having a longer alkyl chain; (ii) increase or decrease the electronic density on the metal; (iii) favour the molecular aggregation of the precursor with a partial hydrolysis; (iv) obtain alkoxides of two different metals. With the described technique it is possible to change the properties of the catalytic precursor in a tailored way, aiming at obtaining a final catalytic site with the appropriate acid-base and/or redox properties.
The preparation of catalysts by grafting alkoxides on oxides surfaces is not a common practice, despite the availability, at low cost, of many metal alkoxides and the facility of synthesising many others [28-30]. Few papers have been published on the subject [12-25] showing that catalysts prepared by grafting have a peculiar and often very surprising behaviour, compared with that of catalysts prepared by the traditional techniques. In particular we observed that catalysts obtained by grafting vanadyl-tri-isopropoxide on silica, coated with a multilayer of TiO2, are more active and selective in the SCR (Selective Catalytic Reduction) of NOx with NH3, than catalysts of the same type, prepared by impregnation of a vanadium salt on the same TiO2/SiO2 support [23]. On the contrary, catalysts obtained by grafting vanadyl-tri-isopropoxide on silica, partially coated with TiO2, dispersed on the surface in a quantity corresponding to a sub-monolayer, are more selective in the ODH of propane to propene and of isobutane to isobutene, than the catalysts obtained by grafting vanadyl-tri-isopropoxide on silica coated with a multilayer of TiO2 [31,32]. More selective are also the catalysts obtained by reacting preliminarly mixtures of vanadium and titanium alkoxides with stochiometric amounts of water, to favour the aggregation of the two different alkoxides and grafting then the obtained compound directly on silica [31,32]. These two last experimental observations suggest the perspective of preparing a catalyst for a given reaction with the optimal redox and acid-base properties already balanced in the alkoxide precursor to be anchored on a support.
The preparation of catalysts by flame hydrolysis allows the formation of an extended class of different oxide mixtures, in nanosize particles, fully crystalline and thermally very stable [33-37], which whitstand effectively the sinterisation caused by the relatively high reaction temperature of the dehydrogenative oxidation. These oxides can then be anchored, by the dip-coating technique, to an oxidic porous membrane or to a ceramic multichannel honeycomb-shaped support, through a proper primer [38], so to obtain a dispersed catalyst highly resistant to the sintering caused by the high temperature and allowing to operate with very short contact times. Short contact times favour a better selectivity in partial oxidation reactions, such as those studied in the present research project.
The traditional way of redox catalysts preparation by impregnation, although well studied for several aspects, is still under discussion for the surface species reducibility as a function of condensation degree, monolayer surface density, evaluated on the basis of geometrical consideration, and promoters action. Therefore, the traditional approach to catalysts preparation is not only a useful reference for any other innovative approach experimented, but also it can originate significant improvements in catalysts performance, by understanding the points still unexplained.
As for the use of new reactor technologies, it must be pointed out that up to now the development of the oxidative dehydrogenation of alkanes has been limited by the low yield to the olefin, since in the presence of molecular oxygen and for high conversion of the alkane, the contribution of the consecutive reaction of olefin combustion becomes considerable, due to its high reactivity. A new reactor technology, which might allow to overcome the problems associated to high reactivity of the olefins towards combustion, is the so-called redox technology (or redox-decoupling) (39-51). The redox technology is based on the fact that the oxido-reductive cycle is indeed mediated by the catalyst, which receives the electrons from the organic substrate and gives to the latter the ionic oxygen (to be inserted into the molecule or to form water), thus being at the same time reduced. Molecular oxygen then receives the electrons from the reduced catalyst, yielding again the O2- species, which is incorporated in the catalyst, with recovery of the original oxidation state of the latter. On the basis of this cycle, a process has been proposed in which the steps are carried out separately, in different vessels. In one reactor the hydrocarbon is put in contact with the oxidized catalyst, to form the desired product of oxidation, and then the reduced catalyst is conveyed to a second reactor, the regenerator, in which it is put in contact with air and reoxidized. In such a way, it is theoretically possible to obtain an improvement of selectivity to the product of interest, since the hydrocarbon and molecular oxygen are never put in contact. A further advantage of this process is the intrinsic safety, since there is no presence of explosive mixtures.
Another favourable possibility is the use of membrane catalytic reactors. Over the last few years the use of membrane reactors has generated increasing interest, either because the separation and reaction functions can be integrated into a single operation [52], or because of their flexibility in controlling the reactant distribution in the reaction environment. The membrane reactor is therefore an optimum candidate for the optimisation and industrial development of selective oxidation processes in the heterogeneous phase and in particular where it is necessary to control the stoichiometry of the reactants in the reaction zone [53,54], or distribute oxygen along the reactor, making the concentration and temperature profiles more favourable and achieving a greater selectivity [55]. Membrane reactors therefore offer a flexibility in their use, by allowing managing the two compartments of the reactor and also by keeping separate these two compartments. The membrane can be used as a porous medium, containing the traditional fixed-bed. This is known as the Inert Membrane Reactor, where the membrane distributes oxygen on the catalytic bed, thus avoiding 'hot spot' formation along the reactor axis and forming explosive mixtures [55]. Another system is represented by the Catalytic Membrane Reactor, where the catalyst is present as a thin layer (of the order of some microns) on the membrane surface or dispersed inside it. In membrane reactors the control of feeding mode allows to limit the evolution of consecutive reactions and therefore to obtain high selectivity at industrially interesting conversions.
At last the honeycomb reactor, in which the catalyst is spread as a thin layer onto the walls of a multichannel honeycomb-shaped ceramic support, permits to reduce substantially the contact time of the reacting mixture with the catalyst. The short contact times prevent the deep oxidation of the desired olefin to the combustion end products, so improving substantially the selectivity of the process.
The model reactions chosen, that is the ODH of propane, butane and isobutane to the corresponding olefins, are very suitable for testing the performance of the prepared catalysts. The ODH of propane is also very suitable to evaluate the performance of the mentioned innovative reactors. These reactions are of great interest for the possible industrial development of the corresponding processes if activity and selectivity are sufficiently high. As many information are available in the literature about the mentioned ODH reactions, we have a good base for verifying the value of the results achieved in the proposed research. <<<