Catalysis Volume 23

A Review of Recent Literature

By James J. Spivey, Kerry M. Dooley

The Royal Society of Chemistry

Copyright © 2011 Royal Society of Chemistry
All rights reserved.
ISBN: 978-1-84973-142-3

Contents

Preface James J. Spivey, v,
Key catalyst types for the efficient refining of Fischer–Tropsch syncrude: alumina and phosphoric acid Arno de Klerk, 1,
Recent developments and achievements in partial oxidation of methane with and without addition of steam Roberto Lanza, Jorge A. Velasco and Sven G. Järås, 50,
Precious metal catalysts for deep hydrodesulfurization John N. Kuhn, 96,
Catalytic reactions studied by angle-resolved product desorption Tatsuo Matsushima and Kosuke Shobatake, 139,
Quantification of cluster size effect (structure sensitivity) in heterogeneous catalysis Dmitry Yu. Murzin and Valentin N. Parmon, 179,
Photocatalysis in green chemistry and destruction of very toxic compounds Vasile I. Parvulescu and Hermenegildo Garcia, 204,
Hierarchical zeolites: materials with improved accessibility and enhanced catalytic activity D. P. Serrano, J. Aguado and J. M. Escola, 253,
Catalysis involved in dimethylether production and as an intermediate in the generation of hydrocarbons via Fischer-Tropsch synthesis and MTG process Eduardo Falabella Sousa-Aguiar and Lucia Gorenstin Appel, 284,
Selective oxidation catalysis on rhenium-oxide catalysts Mizuki Tada, 316,

CHAPTER 1

Key catalyst types for the efficient refining of Fischer–Tropsch syncrude: alumina and phosphoric acid

Arno de Klerk

DOI: 10.1039/9781849732772-00001

Fischer–Tropsch syncrude is characterised by its high content of linear hydrocarbons (alkanes and alkenes) and oxygenates. Efficient refining of such syncrude requires catalyst types that are oxygenate and water tolerant, and that enable useful conversion pathways for chemicals and fuels production. This report deals with two key catalyst types, namely, alumina and phosphoric acid. Past and present industrial applications with syncrude are considered, as well as related catalysis, catalyst deactivation, reaction mechanisms and chemistry. This includes topics such as catalyst hydration, oxygenate adsorption, side-reactions involving oxygenates and competitive adsorption between oxygenates and hydrocarbons. These topics are also pertinent to biomass refining. Alumina catalysed isomerisation, alkylation and dehydration are covered in detail, as is phosphoric acid catalysed isomerisation, oligomerisation, alkylation and hydration.

1 Introduction

Conventional crude oil is at present the primary feed material for the production of most petrochemicals and transportation fuels. Like all non-renewable natural resources, it is finite. Concern over the future availability of crude oil has rekindled interest in the conversion of alternative carbon based energy sources, such as coal, natural gas, biomass and waste, into synthetic crude oil, which is referred to here as syncrude. One of the few industrially applied technologies is indirect liquefaction employing Fischer–Tropsch synthesis. This involves the conversion of the carbon feedstock into synthesis gas (a mixture of H2 and CO), which is then converted into a syncrude (a mixture of hydrocarbons and oxygenates) during Fischer–Tropsch synthesis.

A comparison of conventional crude oil and Fischer–Tropsch syncrude (Table 1) reveals important differences in composition. These differences have implications for catalysis, catalyst selection and the conversion processes that can be used to efficiently refine Fischer–Tropsch syncrude. It is consequently not surprising that the catalysts required to refine syncrude are different to those that are commonly employed for crude oil refining.

The oxygenates in syncrude poses catalysis challenges similar to that encountered during the catalytic conversion of biomass derived liquids. Syncrude is further characterised by a high concentration of compounds with linear skeletal structure and a large amount of 1-alkenes (linear a-olefins) in the lighter fractions. These attributes also affect catalyst selection and performance.

Fischer–Tropsch based facilities constructed since the 1990’s highlighted the importance of hydrocracking as technology for the upgrading of low temperature Fischer–Tropsch (LTFT) syncrude. However, refining of syncrude involves far more than just hydrocracking. In fact, hydrocracking is only of minor importance in the upgrading of high temperature Fischer–Tropsch (HTFT) syncrude. The catalysts relevant to the refining of Fischer–Tropsch syncrude has recently been reviewed. The purpose of the present work is to provide an in depth look at the chemistry and the performance of two of the key catalyst types needed for efficient refining of Fischer–Tropsch syncrude, namely, alumina and phosphoric acid.

2 Alumina

2.1 Alumina in Fischer–Tropsch refining

Alumina is often used as a support material for refining catalysts and it is also employed as the oxidic support for some Co-LTFT catalysts. Yet, in Fischer–Tropsch refining context, alumina plays a significant role as a catalyst per se.

One of the earliest applications of alumina was for deoxygenation and double bond isomerisation of syncrude with an associated improvement in fuel quality. The low octane number linear 1-alkenes are isomerised to higher octane number alkenes with an internal double bond. Gains in research octane number (RON) of 15 units and in motor octane number (MON) of 10 units have been reported. Such units were installed in the HTFT refinery in the USA and the HTFT-LTFT refinery in South Africa.

The catalyst was manufactured from bauxitez, rather than pure alumina. Bauxite may arguably be considered a silica-alumina material, but it contains little silica and its catalytic behaviour resembles that of alumina. In one of the units at the LTFT-HTFT refinery, an acidic clay (Al2O3· SiO2· n H2O) catalyst was employed. When a second coal-to-liquids HTFT refinery was constructed in South Africa, the bauxite catalyst was substituted by a rare earth exchanged Y-zeolite catalyst to deoxygenate and isomerise a C5–C6 syncrude fraction. The zeolite resulted in excessive cracking and ultimately led to the decommissioning of the unit due to poor performance. Alumina clearly outperformed a more active and modern zeolite catalyst for deoxygenation and double bond isomerisation of syncrude. Chain degradation through cracking was not desirable and the zeolite catalyst was too active for this application.

The dehydration of alcohols to produce alkenes is another application of alumina. This was industrially practised at the South African coal-to-liquids HTFT refinery, where mixed alcohols from the HTFT aqueous product was dehydrated over an η-alumina catalyst to produce mixed alkenes. The unit was later decommissioned because it became more profitable to sell the alcohols as chemicals than to refine the mixed alkenes to fuels. In a more recent but related application at the same refinery, a γ-alumina catalyst is employed for the production of 1-octene by the dehydration of 1-octanol.

Alumina is also employed as catalyst for the skeletal isomerisation of HTFT derived pentenes. The high temperature alumina based process (Iso-5) is tolerant of water and oxygenates that are present in the syncrude. In this application alumina was found to be better suited for syncrude conversion than a more sophisticated acidic, nonzeolitic molecular sieve catalyst (Pentesom), which was inhibited by oxygenate adsorption. Desorption of oxygenates in syncrude required a minimum operating temperature of 320 °C. Although the molecular sieve catalyst performed well above 320 °C, the operating range of 280–320 °C was inaccessible, thereby significantly reducing the operating window of the catalyst. This reduced the cycle length of catalyst operation from 10–12 months to 1–2 months. Alumina-based skeletal isomerisation technology was therefore selected for the South African coal-toliquids HTFT facility.

2.2 Alumina as catalyst

Alumina can be obtained in many crystallographic forms, of which γ-Al2O3 and η-Al2O3 are the most prominent alumina catalysts in the refining of Fischer–Tropsch syncrude. Despite the differences in crystallographic form, the surface hydration chemistries of these aluminas are similar, but the concentration of strong acid sites are different ([MATHEMATICAL EXPRESSION OMITTED]).

It is probably an understatement to say that the catalytic behaviour of alumina is complex. This is exemplified by the many conflicting reports in literature on the catalytic behaviour of alumina. Even in the limited body of literature on the alumina catalysed conversion of syncrude, examples of seemingly conflicting reports on the double bond isomerisation propensity of alumina can be found.

Catalysis pathways can be enabled or disabled depending on alumina preparation and pretreatment. This can be understood in terms of the following properties of alumina:

(a) Based on electrostatic considerations, as well as classical and quantum mechanical calculations, the most favourable surface of alumina is one terminated with a single plane of Al, which is almost in the same plane as the second-layer O.

(b) An alumina catalyst can be dehydrated or hydrated to change its activity. The activity may pass through a maximum with increasing dehydration, which is due to the competing effects of increasing activity, loss of surface area and the nature of the site pairs created on the alumina surface.

(c) The surface chemistry of alumina is affected by both its temperature history and water content, with water being present as adsorbed water and water being chemically incorporated into the alumina. The surface may also be modified by oxidation or reduction.

(d) The composition of an alumina catalyst can formally be represented as Al2O3·nH2O, where the H-content may vary in the range n = 0–0.6 without undermining the crystallographic form of the alumina. (A description has been proposed whereby the H enters the bulk and the O stays at the surface).

(e) The transfer of H between hydroxyl ions to form water and an oxide ion and vice versa can readily take place at 200 °C and even lower, but protons (H+) become mobile only above 400 °C.

(f) Despite the hydrogen-transfer ability of alumina, it is a poor catalyst for hydrogenation and dehydrogenation of hydrocarbons. Nevertheless, alumina is not completely inert with respect to either reaction. Hydrogenation requires two adjacent coordinatively unsaturated Al Lewis acid sites to enable the simultaneous adsorption of both carbon atoms involved in an unsaturated bond. Hydrogenation likewise requires H2 to interact with both carbon atoms simultaneously to enable desorption of the hydrogenated product. These are demanding requirements and alumina is consequently a very poor hydrogenation catalyst.

(g) No Brønsted acidity is apparent on alumina, although strong acidity by acid-base ion pairs develop. Adsorbed ammonia that is retained at temperatures above 400 °C has a different character ([MATHEMATICAL EXPRESSION OMITTED]).

(h) Progressive dehydration of alumina gradually transforms the surface. Hydroxyl groups that are initially adjacent to each other (infrared absorption 3200–3600 cm-1) are with progressive dehydration changed into isolated hydroxyl groups (infrared absorption 3700–3800 cm-1) having different configurations. As the surface is increasingly dehydrated (Fig. 1), the number of isolated hydroxyl groups passes through a maximum and the number of coordinatively unsaturated Al Lewis acid sites increases.

(i) Some surface hydroxyl groups are very difficult to remove and the removal of such hydroxyl groups creates high-energy strain sites.

(j) Rehydration of dried alumina with water occurs to a different extent depending on the temperature of rehydration. Rehydration at room temperature causes the water to be physically adsorbed. At a given temperature water readsorption and the associated heat release with such adsorption is rapid and complete within a short time span (5–10 min at 25 °C). The physisorbed water is progressively converted into surface hydroxyl groups as the temperature is increased. Rehydration at around 300 1C is required to convert most physisorbed water into surface hydroxyl groups on alumina, and hence change the catalytic behaviour of the alumina.

Catalysis by alumina is clearly very dependent on the level and nature of surface hydration of the working alumina catalyst. Conversion of water-free and oxygenate-free hydrocarbon feed materials will affect alumina differently to feed materials that contain water and/or oxygenates. Furthermore, the temperature at which the conversion takes place not only affects the kinetics of the reaction, but also the kinetics of catalyst hydration change.

When alumina is employed as a catalyst for hydrocarbon conversion, any water or oxygenates that may be present in the feed will act as catalyst modifiers (poisons, inhibitors or activators, depending on the reaction). This can be understood in the following way. Proton mobility needed for hydrocarbon conversion is achieved by pretreating (dehydrating) the alumina at temperatures of 400 °C or more. The effect of such pretreatment is lost if the catalyst is exposed to water or water forming substances at elevated temperatures.

Water and oxygenates are ever-present in Fischer–Tropsch syncrude. Syncrude conversion over alumina involves reactions of both hydrocarbons and oxygenates. It is therefore anticipated that the nature of a working alumina catalyst is continuously changed by the reaction medium.

2.2.1 Oxygenate adsorption on alumina. Infrared spectroscopy has been employed extensively to identify the nature of the adsorbed oxygenate species on alumina. Although these studies have focussed on alcohol adsorption, it revealed general oxygenate binding modes that are applicable to alcohols, carbonyls and carboxylic acids. The main adsorption modes that have been identified are shown in Fig. 2.

(a) Nondissociative hydrogen bonding of the alcohol O to the H of a hydroxyl group at the alumina surface (Fig. 2a). This is a physisorbed species and can be removed by degassing while heating to 80 °C.

(b) Nondissociative coordination of the alcohol O to an Al Lewis acid site on the alumina surface (Fig. 2b). The singly coordinated species has a characteristic δOH band at ~1280 cm-1 typical of a hydrogen bonded alcohol. Coordination with two adjacent Al Lewis acid sites is also possible. This type of adsorption can be considered a precursor state of dissociative chemisorption, which substantially take place below 100 °C.

(c) Dissociative chemisorption of the Lewis acid coordinated species, yields an alkoxide species, where the O is bonded to the Al Lewis acid site (Fig. 2c). The alkoxide species can be bonded to a single Al, with vC–O/vC–C band at ~1170 cm-1, or the alkoxide species can bridge Al sites, with vC–O/vC-C band at ~1130 cm-1 (when 2-propanol is the adsorbed alcohol). By heating to 200 °C, the alkoxide species is mostly converted.

(d) Dissociative chemisorption to form a carboxylate structure (Fig. 2d) takes place as the temperature is increased above 100 °C. The temperature threshold depends on the alumina pretreatment. The asymmetrical vCO-2 band at ~1570 cm-1 and accompanying weaker symmetrical vCO-2 band at ~1470 cm-1 are characteristic of carboxylates. The carboxylates are stable up to temperatures of at least 300 °C and often higher. It has been suggested that the carboxylate is formed via the chemisorbed alkoxide species, albeit not exclusively so.

The nature of the adsorbed oxygenate species is dependent on the alumina surface and specifically the nature of the adsorption site and its nearest neighbours. It is also dependent on the temperature history, which determines the degree of alumina dehydration and allows the transformation between different adsorption modes. The selectivity of different oxygenates reactions over alumina, such as dehydration as opposed to dehydrogenation, follows from these adsorption modes. The model by Peri is useful for the understanding of oxygenate conversion over alumina.

(Continues…)Excerpted from Catalysis Volume 23 by James J. Spivey, Kerry M. Dooley. Copyright © 2011 Royal Society of Chemistry. Excerpted by permission of The Royal Society of Chemistry.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.