1: History of Catalysis

Without catalysis, life is not possible and, in that sense, catalysis is as old as life. The oldest catalytic processes used by humans are the production of wine, beer, alcohol and acetic acid, all through the fermentation of carbohydrates. This technology was already practised earlier than 2000 years BC. Compared to biocatalysts, the utilisation of chemocatalysts in industrial production is recent. Over the years, catalysis has developed into a broad discipline. History includes a wealth of theories and achievements that have changed society.¹  Since catalysis as a whole is a topic too broad, we will not try to completely cover all historic developments. We will mention the numerous successful multicomponent catalyst systems as a chronological list presented in a table. In this chapter, the history of catalysis will be presented with a flavour of practical applications and a focus on heterogeneous (chemo)catalysis.

The name ‘catalysis’ was coined by Berzelius in 1836.²  He concluded that besides ‘affinity’, a new force was operative, a ‘Catalytic Force’. The word ‘catalysis’ stems from the Greek: καταλυση, which means loosening, allow to move downwards freely. In the time of Berzelius, the term ‘affinity’ was used; however, on a molecular level, no understanding existed of reaction rates. Although insight was limited, catalysis as a new tool was booming.

When we limit ourselves to catalytic production processes of chemicals, the first commercial application was the production of sulphuric acid in the mid-18th century. The history of the production of sulphuric acid is interesting.¹  In the Middle Ages, it was synthesised in the laboratory in glass equipment by burning sulphur with nitric acid in humid air. In 1746, lead was introduced as a construction material and the so-called lead chamber process allowed commercial production. In 1793, Clément and Desormes discovered that the quantity of nitric acid could be reduced if additional air was provided: a catalytic process was born. They realised that SO₂ was converted into SO₃ by air and that nitrous vapours were not consumed in a stoichiometric reaction but they were only intermediates. Remarkably, this process was a homogeneously catalysed process: in aqueous solution, SO₂ is oxidised to sulphuric acid with nitrogen oxides functioning as O-transfer agents. A specific characteristic of the process is that, at high sulphuric acid concentrations, stable nitrosyl sulfate compounds are formed making it impossible to produce sulphuric acid in high concentration and purity. As early as 1831, a process was patented in which the oxidation was catalysed by a solid catalyst, Pt, allowing the production of the generally desired concentrated sulphuric acid. A commercial application, however, was strongly delayed due to practical difficulties, the major one being the low stability of the catalyst. It should be noted that elemental sulphur was initially the raw material. Later, pyrite was used because of its lower price. However, processing of pyrite ore was associated with considerable amounts of impurities, which acted as catalyst poisons. Not surprisingly, the Pt crystals were more sensitive to poisons than nitrogen oxides. This is a major reason why it took so long before a heterogeneously catalysed process became feasible. Nowadays, all processes for the production of sulphuric acid are heterogeneous. Vanadium-based catalysts have largely replaced Pt catalysts as they are cheaper and less susceptible to poisoning. Vanadium is present as a liquid salt in the pores of a porous solid carrier. At present, the raw material for the production of sulphuric acid is mainly elemental sulphur again, strongly reducing the catalyst deactivation. Sulphur is not expensive as it is a by-product from the hydrotreatment of oil fractions.

Around the turn of the 18th century, the influence of metal and metal oxides on decomposition reactions was a hot topic. An example is the decomposition of ethanol. In the presence of copper or iron, carbon and a ‘flammable gas’ were produced, whereas in the presence of pumice stone, ethene and water were observed. In retrospect, catalyst selectivity was demonstrated.

Good insight into the struggle to find an interpretation of observed reaction rates is given by the work Humphry Davy performed at the beginning of the 18th century in the development of a miner’s safety lamp.³  He discovered that coal gas, which contains relatively a lot of methane, burns in the presence of hot platinum and palladium, whereas copper, silver and gold did not exhibit any activity. From the observation that the active metals had to be hot, he concluded that the function of the active metals was heating the reaction mixture. The difference in the behaviour of the two sets of metals was attributed to the difference in their heat capacity and thermal conductivity. Later, it was found that finely dispersed platinum is active already at room temperature. Davy was not right in his explanation, but he discovered that reaction between two gases took place on a metal surface, while this surface did not change chemically.

In 1834, Faraday proposed that in catalysis the reactants have to adsorb simultaneously at the surface, but he did not explain the catalytic mechanism.⁴  Berzelius did not give an explanation, but he nicely generalised the observations in a simple description. Later, Ostwald gave the definition that a catalyst does not influence the thermodynamic equilibrium of a reaction, but that the rates are influenced. Berzelius and Faraday proved to be right. Clearly, the 19th century was an important period for catalysis. In a sense, the Nobel Prize of Sabatier for his work on catalytic hydrogenation was the recognition of the importance of this productive period. Sabatier formulated the principle that the reaction intermediates formed at the surface of a catalytic material should have intermediate stability. When too stable, they would not decompose; when too unstable, they would not be formed. This molecular view of the catalytic reaction, not as a single reaction but as a cycle of reaction steps, in which intermediate complexes between a catalyst and a reagent are formed and then decay, was particularly modern. Sabatier’s principle is the formulation of the molecular basis of catalytic action and complements Ostwald’s physicochemical view.

Table 1.1 gives a survey of the fascinating history of successes in the application of catalysis. The data refer to activities on at least pilot plant scale.

The early industrial catalysis applications were dominated by the development of production processes for organic chemicals, but inorganic chemical production was also important. Early inorganic processes were the production of sulphuric acid, as mentioned above, and the conversion of HCl into Cl₂.

The development of the process for the synthesis of NH₃ is an enlightening example of a successful R&D project.⁵  It was based on the systematic, scientific search for a catalyst that would perform optimally under the envisaged practical conditions. Reading the story of the development of this process is, one century later, still inspiring.

Initially, the source of ammonia was coke oven gas, containing typically 1–5% ammonia, and Chile saltpetre, as the large deposits of NaNO₃ in Chile are called. It was recognised as early as the turn of the 20th century that insufficient ammonia was available to cover the agricultural demand. In addition, ammonia was used in increasing amounts to manufacture explosives at the beginning of the First World War. Particularly in Germany, extensive research efforts were made to synthesise ammonia directly from N₂. Non-catalysed routes were discovered and even commercialised despite their inefficiency, but the breakthrough was the development of a catalytic process.

In 1905, Haber succeeded in producing ammonia catalytically at a temperature of 1293 K and a pressure of 1 bar. The yield was a few percent. He extrapolated his data to lower temperatures and concluded that a temperature of 520 K would be the maximum temperature for a commercial process (at 1 bar). This was the first application of chemical thermodynamics to catalysis and, not surprisingly, accurate thermodynamic data were not available for a broad range of conditions. Haber concluded that the commercial development of a practical process was a hopeless undertaking. Later, in independent work, Nernst concluded that the thermodynamic data of Haber were not correct. On his turn, Haber reinvestigated his data and concluded that a high-pressure route had to be followed. He tried many catalysts and found osmium and uranium to be promising catalysts. He patented a process with a yield of ammonia at 175 atmospheres and 550 °C. The recycle system he developed worked well, see Figure 1.1. In fact, this lay-out is still used nowadays. He approached BASF and they decided on a large development program, in which Bosch was in charge of the scale-up.

The scale-up study was carried out in a systematic, modern way. Three main challenges had to be faced. The feedstock, hydrogen and nitrogen, had to be produced at the lowest cost possible at that stage. A good catalyst had to be developed, the reactor had to be scaled up and the hardware had to be designed and constructed.

Systematic studies were carried out to find a good catalyst. Iron catalysts got special attention because it was known that iron catalyses the decomposition of ammonia, the reverse of the synthesis reaction. It was found that iron alone was only slightly active but its activity could be promoted or worsened by additives. Over 10 000 catalysts were prepared and over 4000 were tested in a kind of high-throughput experimentation program. The catalyst developed at that time is in essence the one still used today.

Regarding the hardware, the major challenge was to develop a reactor that could withstand the harsh reaction conditions (combination of high temperature and high pressure). High-strength carbon steel had to be used because of the high reaction pressures, but this steel was corroded by H₂ under those severe reaction conditions. In careful experiments, Bosch discovered that the reason was the decarbonisation of the carbon steel by H₂ at high temperature. He designed a reactor that contained an outer wall of high-carbon steel and an internal wall of low carbon steel. The outer wall was protected by cooling it with the feed gas.

Very pure H₂ (from water electrolysis) was used in the pilot studies. For a commercial plant this was not practical. What feedstock and what process could enable the development of a commercial process? Up to then, biomass had been the dominant feedstock. In the 19th century, coal became the major raw material. In fact, it was the basis for the industrial revolution. A satisfactory process for the production of H₂ was found in a combination of coal gasification and the water–gas shift reaction:

Figure 1.2 places the work of Haber and Bosch in historical perspective. It shows the energy input per amount of product over time. The big step was the Haber–Bosch process, later innovations allowed modest but consistent improvements. Haber received the Nobel Prize for his work on the synthesis of ammonia and later Bosch, together with Bergius, were awarded a Nobel Prize for their achievements in high-pressure technology.

A spin-off of the successful development of the ammonia synthesis was the expertise in high-pressure processes and the availability of synthesis gas that was sufficiently pure for catalytic processes. Processes based on high-pressure hydrogenation reactions, such as the methanol synthesis and the Fischer–Tropsch synthesis, became in principle feasible. In fact, the same team that developed the ammonia synthesis process at BASF also developed the first process for the production of methanol. In the period between 1930 and the Second World War, coal continued to be the main feedstock for the production of chemicals and fuels, although oil and natural gas were playing an increasingly important role, in particular in the U.S.

Germany was the Walhalla of coal utilisation processes. A major application of coal was the production of the coke needed in large-scale blast furnace processes. The production of coke was a mild pyrolysis process and, besides coke, large quantities of cracking products were produced. Acetylene was a main component and it was seen as a valuable base chemical, for instance, for the production of ethene and vinylchloride. Parallel to the emergence of a synthesis gas based process industry, the high-pressure catalytic hydrogenation of coal was attempted.⁷  Not surprisingly, catalyst deactivation was a major challenge. Sulphur appeared to be responsible in many cases. The solution was found in two-stage processing. First, a liquid phase hydrogenation was carried out in a slurry of coal particles (oil produced from coal) with a highly dispersed catalyst. In the second stage, conventional catalyst hydrogenation was carried out in a fixed bed reactor. As early as 1924, it was known that sulphides of Mo, W, Co and Fe were suitable catalysts not poisoned by the heteroatoms (S, N, O) in the feedstock.^7,8  Much later, during the development of hydrotreating processes in the oil refinery industry, the same catalysts were used. Thus, chemical plants using coal as raw material thrived with a wealth of catalytic processes, in particular in Germany. This work can be considered the origin of catalytic reaction and reactor engineering as we now know it.

Particularly in the U.S., and besides coal, oil and natural gas played a significant role as early as the 1920s. An intense debate arose about what the best option was, coal or oil? From an engineering point of view, oil is preferred because processing liquids (and gases) is much easier than working with solids. However, compared to coal, crude oil is very unreactive, in particular the paraffin fraction. The situation changed when thermal and later catalytic cracking processes were developed.

In 1930–1960, an impressive number of petroleum-based catalytic processes were developed. The major drive came from the rapidly expanding market for transport liquids with satisfactory properties. For instance, gasoline with high octane numbers was in demand. Catalytic cracking shifted the product spectrum of oil refinery towards the desired boiling point range.⁹  It is striking that, in catalytic cracking, process and catalyst development was carried out in such close harmony. Initially, AlCl₃ solutions were used, leading to enormous technical and environmental problems of corrosion and pollution. Subsequently, solid catalysts were used in fixed beds and later in fluidised beds. The discovery of zeolites and their high activity enabled the use of riser reactor technology. This cracking technology, referred to as FCC (Fluid Catalytic Cracking), has had an enormous impact.

Although straight-run oil fractions may be in the desired boiling point range for gasoline application, their characteristics do not allow their use in practice. The octane number is too low and exhaust emissions are unacceptable, in particular those of acid rain components. The octane number can be increased by ‘catalytic reforming’ (major reactions occurring are isomerisation and dehydrogenation to aromatics) catalysed by Pt.¹⁰  Since this catalyst is poisoned by sulphur, a large industry has evolved producing hydrodesulphurisation catalysts. A fortunate side effect of the fact that the gasoline produced is essentially sulphur-free is that an unprecedented reduction in emissions of (gasoline utilising) cars could be realised: catalytic converters were quickly developed in which all of the important emissions of gasoline powered cars were eliminated (CO, NO_x, and hydrocarbons) to a large extent. At present, it is recognised that diesel is very attractive as automotive fuel. Processes are being developed to increase the production of diesel in refinery (hydrocracking), in combination with tailored hydrodesulphurisation processes.

Natural gas deserves special attention. In steam reforming of natural gas, the synthesis gas produced is very pure, making it an excellent feedstock for catalytic processes.¹¹  Because of its high purity, synthesis gas based on methane reforming has excellent quality as raw material for catalytic processes. A large variety of bulk chemicals are produced catalytically from syngas, examples being the ammonia and methanol processes mentioned above. In the 1960s, at Imperial Chemical Industries (ICI), a breakthrough finding was the development of highly active catalysts leading to the so-called low-pressure process. This is a good example of the often encountered situation that a new catalyst leads to completely new technology. The route to low-cost and rather pure syngas is also the basis for ‘gas-to-liquid’ technology, referred to as the Fischer–Tropsch synthesis.^12,13 

Although the major drive for the enormous impact of oil applications are transport fuels, an efficient petrochemical industry has emerged in parallel. Analogous to the coal-based synthesis gas route to chemicals, oils and natural gas can be used via synthetic routes. For instance, the first large-scale methanol processes were coal based, but very similar technology is used nowadays based on oil and, in particular, on natural gas-based synthesis gas. In the interbellum period in Germany, a lot of research was carried out aimed at using synthesis gas from coal for producing chemicals. Otto Roelen in the late 1930s discovered that alkenes react with syngas provided an appropriate catalyst is present. This process is called hydroformylation (originally, the ‘Oxo process’).¹⁴  With propene and syngas as reactants, n- and i-butyraldehyde are formed. After World War II, the reaction was studied again. It was discovered that a non-supported Co catalyst was the catalyst, although HCo(CO)₄ appeared to be the active complex formed in the liquid phase. This complex was successfully used but the pressure had to be very high (200–450 bar) to ensure acceptable catalyst stability. In the 1960s, a more stable catalyst, HCo(CO)₃P(n-C₄H₉)₃ was developed. Research continued and catalysts based on Rh were developed that were very active, stable and selective. In modern processes, not only Rh but also Co catalysts are still used for higher alkenes with internal double bonds, which are rather inert compared to terminal alkenes. The reason is the favourable isomerisation activity of Co catalysts.

A success story of homogeneous catalysis is the production of acetic acid.^14,15  Acetic acid was originally (and still is) produced by fermentation. In 1916, the first industrial process was commercialised, in which acetic acid was produced by oxidation of acetaldehyde. This was replaced by the catalytic oxidation of naphtha or n-butane because of the low cost of the reactants. These processes involved radical reactions and the selectivity was limited, in particular for the butane/naphtha oxidation process.

The manufacture of acetic acid by carbonylation of methanol was described as early as 1913. However, the combination of a corrosive strongly acidic reaction mixture (HI!) and harsh reaction conditions (high temperature and pressure) made it very hard to find suitable construction materials. Only after the development of appropriate materials, commercial production was feasible. The process was optimised by BASF at the beginning of the 1950s. The catalyst precursor was a mixture of cobalt iodide and hydrogen iodide in water. The first commercial methanol carbonylation plant did not come on stream until 1963. The pressure was >500 bar. The reactor was built of Hastelloy. In 1968, Monsanto introduced a novel highly active and selective Rh/HI catalyst that could operate at much lower pressure, 30–60 bar. This process was commercialised only two years later. Despite the success of this process, the search for new catalysts continued. In 1966, BP introduced a new catalyst system based on Ir/HI and promoted by Ru. This new process had several advantages. The need for expensive corrosion-resistant materials was still a point of concern and Ti was mostly used.

A work horse process of the petrochemical industry is thermal cracking, often referred to as ‘steam cracking’, in which the principal reactions are the conversion of relatively unreactive alkanes into much more reactive alkenes. Large amounts of light alkenes are formed, with ethene as the main product. Valuable coproducts are formed such as butadiene or pyrolysis gasoline, with benzene as the main constituent. These compounds were an important basis for the flourishing petrochemical industry by making available a set of so-called base chemicals. Since the 1930s, when the petrochemical industry started to take shape, ethene almost completely replaced coal-derived ethyne, and is now the world’s largest volume building block in the chemical industry. It should be noted that the boundary between fuels and chemicals is not strict: for instance, the FCC process in oil refinery is an important source for the base chemical propene, and part of the high molecular weight products of the naphtha cracker go into the gasoline pool. The majority of the reactions involved are catalytic processes. Dedicated catalytic processes for the selective dehydrogenation of specific alkanes, for instance propane or isobutene, have been developed (ref. 15, see Table 1.1).

An option for the production of light alkenes from natural gas was developed by converting methanol in shape-selective zeolites, the so-called methanol-to-olefin process.¹⁶  This process was developed by the same group that developed the methanol-to-gasoline process, also based on shape-selective zeolite catalysts.

In answer to environmental legislation in the 1980s, various gasoline additives were developed, in particular methyl-tert-butyl ether (MTBE). Besides conventional fixed bed technology, a novel technology, catalytic distillation, was introduced by combining the reaction with the major separation step under the same conditions.

The metathesis of alkenes is one of the very few fundamentally novel and important organic reactions discovered since World War II.^17,18  It involves the conversion of alkenes to produce new alkenes. Formally, double bonds are broken with the simultaneous formation of new ones, as shown in Scheme 1.1.

An early commercial process was the metathesis of propene, in which an equimolar mixture of ethene and butene is formed (TriOlefinProcess).

Linear 1-alkenes are important feedstocks for the chemical industry. They were initially produced by thermal cracking of waxes (high-molecular-weight alkanes) or by dehydration of 1-alcohols. Since the mid-1970s, C₄–C₁₈ alkenes were produced by oligomerisation of ethene, most of them based on Ziegler-type catalysts. An innovative process at the time was the Shell Higher Olefin Process (SHOP), which is a combination of ethene oligomerisation and metathesis.¹⁴  The SHOP oligomerisation process uses a Ni-complex resulting in high yields and nearly exclusively linear 1-alkenes. The chain length varies from C₄ to C₄₀, with an even number of C-atoms and with a statistical Anderson–Flory–Schultz distribution, roughly C₄–C₈ 40%, C₁₀–C₁₈ 40% and C₂₀₊ 20%. The desired fraction is C₁₀–C₁₈, whereas the value of the light fraction is low and for the heavy alkenes there is no market. It would thus be great to combine these latter fractions and to produce alkenes with the average chain length of C₁₀–C₁₈. This can be done by metathesis, provided that the 1-alkenes are converted into internal alkenes (1-alkenes always give ethene as a product, which is highly undesired!). Thus, the process encompasses three subprocesses: oligomerisation by a homogeneous catalyst, and isomerisation and metathesis, both by heterogeneous catalysts. After oligomerisation, the light and the heavy products are recycled and subjected to isomerisation and metathesis. The higher alkenes are recycled to extinction.

The polymerisation of alkenes by organometallic catalysts is one of the most important processes in the chemical industry with enormous impact on society.^19,20  In the 1950s, Ziegler and Natta discovered titanium- and vanadium-based catalysts capable of polymerising ethene at relatively low pressure compared to radical polymerisation processes. The product of the new process had a high density and was called high-density polyethene (HDPE), in contrast to the low-density polyethene (LDPE) produced in the radical polymerisation process. At about the same time, researchers at Phillips Petroleum Co. discovered that catalysts of chromium supported on silica were also capable of producing HDPE at moderate temperature and pressure. One of the inventors was Robert Banks, the same researcher who discovered the metathesis of alkenes! Phillips Cr-based catalysts are less active than Ziegler–Natta catalysts and therefore limited to the production of polyethene. Commercial production of HDPE started in the late 1950s for both catalyst systems. The HDPE produced with these catalysts is linear without branching. Homogeneous catalysts had also been known since the beginning but they failed to polymerise propene. Clearly, the field was dominated by heterogeneous catalysts. However, in the 1970s, the group of Kaminsky discovered a family of Zr-metallocene based organometallic catalysts that exhibited orders of magnitude higher rates. The field of catalytic polymerisation is still very much at the centre of research and development programs. Many new materials have been commercialised in the last decades.

Selective catalytic oxidation plays a major role in the chemical industry. Examples are the production of acetaldehyde and ethylene oxide, EO, propene oxidation to propene oxide, PO, and propene oxidation to acrolein and to acrylonitrile.^21–23 

Originally, acetaldehyde was produced from ethyne by addition of water. When the oil/natural gas era started, ethene was the preferred feedstock. This process became feasible by the discovery of the ‘Wacker’ process. The observation that ethene forms acetaldehyde in aqueous palladium chloride solution dates back to the end of the 19th century. In this reaction, Pd²⁺ is reduced to Pd, but catalysis does not take place as O₂ is not able to re-oxidise the Pd formed. In the 1950s, Wacker Chemie and Hoechst developed a process in which CuCl₂ re-oxidised the Pd catalyst. The catalytic cycle was closed because Cu⁺ (formed in the Pd re-oxidation) readily re-oxidises to Cu²⁺ in O₂ or air.

The direct oxidation of ethene to the epoxide with O₂ is known since the 1930s, when Ag/Al₂O₃ was found to catalyse this highly desired reaction. The catalyst performance has improved over time in a rather empirical way. The current commercial Ag catalysts are promoted by alkali metals, such as cesium, and chlorine. The alkali metals are added during catalyst production, but chlorine has to be added as organic chlorides, such as 10–40 ppm vinyl chloride, continuously during operation.

Up till now, an analogous process for the production of propene oxide has not been found. Using Ag-based catalysts leads predominantly to total combustion. The chlorohydrine process and the hydroperoxide processes were until recently the only commercial processes for propene oxide production. The latter is preferred because of the lower environmental impact. The hydroperoxide processes are based on the peroxidation of an alkane to an alkyl hydroperoxide, which subsequently reacts with propene, producing propene oxide and an alcohol. The various commercial processes differ in the hydrocarbon used and the fate of the alcohol formed. For instance, in the Styrene Monomer–Propene Oxide (SMPO) process, ethyl benzene is oxidised to ethylbenzene hydroperoxide, which reacts with propene to produce propene oxide and α-phenyl ethanol. The latter is dehydrated to produce styrene. Most processes use either a homogeneous tungsten, molybdenum or vanadium catalyst, or a heterogeneous titanium-based catalyst for the epoxidation reaction. A new related process is based on the use of cumene, where the alcohol produced is converted back over a copper–chromium catalyst. The special feature of this process is that no stoichiometric amount of a co-product is formed. A recent development is the production of propene oxide using hydrogen peroxide. The advent of titanosilicate TS-1 opened the possibility of carrying out heterogeneously catalysed epoxidations in polar solvents (water, alcohol) using H₂O₂ as oxidant. Although commercially available hydrogen peroxide is no alternative as an oxidising agent because of its high cost, the process is viable if the production is integrated with the propene oxide production step.

In the 1940s, another selective propene oxidation process, the formation of acrolein catalysed by Cu, was discovered at Shell:

In the 1950s, at SOHIO, the potential of Bi-molybdates that afforded a great improvement over the Cu catalysts was discovered. In the 1950s and early 1960s, the catalysts were further successfully developed up to pilot plant stage. These catalysts were the basis for modern commercial processes. Research on the kinetics of this reaction led to the formulation of the ‘Mars-van Krevelen’ mechanism. The catalyst functions as a reservoir of O-atoms. In the first step, a lattice O reacts with an incoming propene molecule and, in a second step, lattice O is replenished by gas phase O₂. This appears to be a basic way of operation of many reducible catalyst systems.

At SOHIO, a related process was developed involving the oxidation of propene but in the presence of ammonia. With the same family of Bi-molybdate catalysts, acrylonitrile is formed:

The reaction is called ‘ammoxidation’. This new process was already used at commercial scale in the 1960s. Prior to this invention, the BASF acetylene/HCN route was the leading commercial process. Nowadays, all acrylonitrile is essentially produced according to the SOHIO technology. With time, the catalyst was further developed by changing the composition of the heteropoly- and isopoly-acid salts. The optimal catalysts are multicomponent catalysts containing several transition metals. The mechanical strength has to be high for technical reasons; therefore, silica is used as the support.

In the late 1950s and the beginning of the 1960s, a completely new field in heterogeneous catalysis emerged. At that time, it was generally realised that the air quality was poor. Acid rain was seen as a global problem and the energy sector – power plants, domestic heating and the transport sector (gasoline and later diesel powered engines) – was blamed. Legislation was for decades a major driver for developments in catalysis. In 1966, California was the first state to introduce limits of emission for passenger cars. Initially, reduction of emission levels was achieved by engine modifications; catalysts were tested but deactivation was dominant for most of the metal based catalysts. Later, further reductions in emission were required and engine modifications alone did not afford the required results. Catalysis paved the way to a solution.²⁴  The breakthrough was the development of noble metal based catalysts in combination with the elimination of lead from gasoline. The process was limited to the reduction of the emissions of CO and hydrocarbons by oxidation. Later, also the reduction of NO_x was required. The three-way catalyst system combined with engine control essentially solved the problem of the emissions of gasoline-powered cars. It is fair to mention that, besides impressive successes in catalyst development, the development of catalytic reactors was crucial in this success. Structured catalyst supports became the state-of-the-art reactor bodies.^25,26  It is estimated that, up to now, more than 1 billion monoliths, the most widely used structured reactors, have found their way to the market.

In the 1960s and 1970s, it was generally realised that acid rain was an urgent problem. NO_x played a major role, and abatement of NO_x was high on the agenda. The main sources of NO_x are nitric acid plants and power plants. These points were thus first addressed. Although total decomposition to the elements is thermodynamically possible, the rates are so low that this reaction is not feasible in practice. Reaction with a reductant was the only practical solution. Owing to the presence of considerable amounts of O₂, this process had to be selective. The process is referred to as ‘Selective Catalytic Reduction (SCR)’. Reaction with NH₃ or urea give satisfactory results and the main catalysts are based on vanadia on titania.²⁷ 

Compared to gasoline engines, the treatment of diesel exhaust gas is more difficult due to (i) the presence of particulates and (ii) the presence of O₂ in the several percent range.²⁸  Since the beginning of the 1990s, oxidation catalysts have been applied for the oxidation of CO and hydrocarbons by Pt or Pd catalysts. Particulates are trapped in filters with periodical regeneration. The regeneration is realised by occasionally raising the temperature of the filter enabling combustion of the trapped soot. Much research has been done in order to create a catalytic oxidation process enabling a certain degree of continuous soot removal. This can be realised by adding catalysts (Fe, Cu complexes) to the fuels. The metals of the added catalysts end up in the soot particles and function as oxidation catalysts (also known as ‘fuel-borne catalysts’). In all systems, the pressure drop over the filter is monitored and, when clogging occurs, a thermal regeneration step is initiated by a temperature rise. An appealing idea is the development of a two stage system consisting of an oxidation monolith and a classical soot trap. The oxidation monolith catalyses the oxidation of the NO present into NO₂. NO₂ is a strong oxidant and converts the soot into CO/CO₂. A tough challenge is the removal of NO_x. Like before, selective reduction with ammonia or urea does work. However, for passenger cars, SCR units are rather large. An alternative has been developed, called the ‘NO_x storage and reduction (NSR)’ system. These systems rely on a three-way catalyst used for gasoline cars, but modified with an oxide forming a nitrate that is stable under oxidising conditions and unstable under reducing conditions (N₂ is formed). In this case, the lifetime is the challenge.

A special sector where catalysis has contributed to is the production of fine chemicals. Although the production of fine chemicals is two orders of magnitude lower than that of the production of bulk chemicals, the annual amount of waste is similar for both sectors.²⁹  A major difference between the two sectors is the use of catalysts. In the case of fine chemical production, the extensive use of catalysts is less well developed. Fortunately, also in this sector, the number of successful catalytic processes is increasing. A convincing example is the production of ibuprofen.²⁹  The old process needed six steps. Five of these steps with substantial formation of salts have been replaced by two steps, viz., a heterogeneously catalyzed hydrogenation and a homogeneously catalysed hydroformylation with an atom efficiency of 100% without the formation of waste.

Table 1.1 gives a good overview of the rich history of the development of industrial catalytic processes.

The process industry and the relevant catalytic processes are profoundly influenced by the available raw materials. Originally, biomass was dominant. In the 19th century, coal became the major raw material. In fact, it was the basis for the industrial revolution. Later, petroleum took over this position and, nowadays, we see an increasingly important role for natural gas and, very recently, shale oil and gas. Acetic acid, for example, was originally produced by fermentation of biomass, subsequently from acetylene (produced from coal) and later from ethylene (from petroleum fractions or natural gas). At present, it is produced from methanol, which is produced from natural gas or coal. It is interesting to note that the fermentation process continues to be important. Another example is butanol, which from the 1920s on was produced together with acetone through the fermentation of corn. In the 1960s, this process was replaced by the hydroformylation of propylene. Thus, originally, biomass was the raw material but later petroleum-based propylene became the main feedstock. Today, we see many efforts to produce butanol again from biomass (Figure 1.3)!

At the time when coal was the major feedstock, ammonia was produced from coke oven gas. Later, it was produced in large-scale processes from synthesis gas, obtained by steam reforming of natural gas. Vinyl chloride was originally produced by the addition of HCl to acetylene. A more convenient process was the route via oxychlorination of ethylene, producing dichloroethane, followed by dehydrochlorination. Toluene, like ammonia, was a by-product of the coal industry. Nowadays, it is recovered from the product stream of the catalytic reforming of naphtha.

All these processes are catalytic, and changes in the raw material markets have been one of the driving forces for the development of novel catalytic processes. Similar developments have generally occurred for most other products.

With fossil fuel reserves dwindling and the massive usage of fossil fuels being experienced as a burden to the environment, the focus has shifted to the use of renewable feedstocks and recycling of materials and products. This has been (and is) an enormous challenge, as new synthetic routes have to be developed and be competitive from an economic viewpoint. This change in focus has been a major driver in catalysis Research & Development.

In the biotechnology industry, fermentation is extensively used for the production of beer, wine, etc. In an early stage, it was realised that this technology had promise in producing the so-called bioethanol for the transport sector. The production of bioethanol is the largest volume of biofuels in the market. By adapting the microorganisms, biobutanol can be produced. Butanol is more attractive than ethanol because it is less corrosive, has a higher energy density and can be mixed with gasoline or diesel.

The production of biodiesel is done by chemo-catalytic processes. Biodiesel, defined as a mixture of methyl or ethyl esters of fatty acids, has also been introduced on a significant scale in the market.³¹  The first processes for the production of biodiesel were based on transesterification reactions catalysed in the homogeneous phase by basic catalysts (mainly NaOH or NaOMe) and, more recently, by Zn/Al mixed oxides on alumina. An alternative process is the catalytic hydrodeoxygenation (HDO) by conventional hydrotreating catalysts.

Just as the petrochemical industry is based on a relatively small number of so-called base chemicals derived from oil components, a bio-based industry can be envisioned based on a small number of building blocks, the so-called platform molecules, derived from biomass components. The products derived from platform molecules may be the same as those derived from oil-derived chemicals, or different with the same function, or they may even be completely new. Glycerol, the coproduct from transesterification, is an obvious candidate platform molecule. DuPont produces 1,3-propanediol by fermentation of glucose. An important product is polyester, produced by polycondensation of 1,2-propanediol and terephthalic acid. Succinic acid is produced from glucose by fermentation. Several bacteria can be used but, recently, it was claimed that a yeast-based fermentation process is superior (DSM-Roquette). In view of the high OH-functionality of biomass, it is not surprising that reduction and dehydration are often encountered reactions. Examples where heterogeneous catalysis is the applied technology are the reduction of glucose to sorbitol (RaNi, Ru) with H₂, the dehydration of ethanol to ethylene and the dehydration of xylose to levulinic acid (acid catalysts). Hexose sugars can be dehydrated to form furan compounds such as hydroxymethylfurfural (HMF), a versatile platform molecule.³²  For instance, catalytic oxidation to a dicarboxylic acid gives a molecule that is used in the production of polyesters. Instead of the dehydration of hexoses, an intermediate hydrogenation step to sorbitol allows the hydrogenation into sorbitan and isosorbide, also with high potential in polyester type applications.³³ 

The majority of the processes in Table 1.1 are still of practical importance. Of course, they have been improved tremendously, but the principles and the types of catalysts applied in many cases are strikingly similar. Such information is very valuable for the improvement of existing processes or the development of novel ones, even if a sound interpretation is not (yet) available. Historical empiric information plays the main role in the practice of catalyst selection. However, increasingly, theory gives a basis to catalyst selection and improvement.

The so-called Sabatier’s principle provides first-principle leads to catalyst improvement. According to this principle, catalyst reactivity is maximal when the interaction of the catalyst with the reagents has an intermediate optimum value. In the catalytic reaction, the intermediates are adsorbed on the catalysts. The catalytic reaction cycle is closed by desorption of the products, which regenerates the catalytic sites. In terms of the reactants, when the interaction between reactants and catalyst is weak, no reaction takes place as the reagents are not activated. For very strong interactions, the reactant molecules block the catalytic sites, because they do not desorb. The catalytic reaction rate will be the highest for an intermediate strength of adsorption. This results in the generally observed volcano plots, where different catalyst rates of reaction are plotted against a catalyst reactivity parameter, such as the heat of adsorption, as shown in Figure 1.4.

A good illustration of the potential of computational methods in catalyst screening for the reaction

is given in Figure 1.5. The activation energy of CO dissociation is plotted versus the CO dissociation energy for a series of transition metals. Figure 1.5b presents the corresponding catalytic activities, showing a clear volcano shape, fully in agreement with Sabatier’s principle.

This chapter has been written emphasising the discoveries leading to practical processes, most of them corresponding to the last 100 years. It is enlightening to complement this discussion with the relevant Nobel Prizes awarded in this period.

These ongoing fundamental discoveries have made catalysis an essential technology and an important science that will undoubtedly increase in significance in the coming century.