Diesel fuel from biomass

Abstract

Several processes are currently under development to produce diesel fuels starting from biomass. Most of them are described and compared. Particular attention has been paid to processes currently studied in eni laboratories. Problems are highlighted, as well as some issues for which further research is advisable. Eventually, however, it is most likely that the success of one or the other of these technologies will depend on several factors, including the availability and the quality of the feedstock, the complexity of the process, and the quality of the final biofuel.

---

Carlo Perego

Carlo Perego is the director of the eni corporate Research Centre in Novara, Italy. He received the Laurea in Industrial Chemistry in 1978 from the State University of Milan and the Master in R&D Management at the Bocconi University of Milan. In 1980 he joined Montedison R&D Centre in Bollate. In 1990 he joined the corporate research centre of eni company near Milan, where he worked in the catalysis group. In 2007 he moved to his current position, dealing with new research programs devoted to solar energy, biomass and biofuels and environmental technologies. He has been a chairman of Industrial Chemistry Division of Italian Chemistry Society. He is a member of Catalysis Commission of International Zeolite Association. He authored more than 60 patents and 80 papers, mainly concerning catalysis and environmentally friendly catalytic processes.

Marco Ricci

Marco Ricci graduated in Chemistry at the Università “La Sapienza” in Roma in 1979. In the same year, he joined the Istituto eni Donegani where he is currently the head of the Sustainable Chemistry unit. In 1984 he was a Research Associate at the Laboratoire de Chimie de Coordination in Toulouse (France) with Dr Bernard Meunier. His research interests focus on selective oxidation, synthetic models of redox enzymes, homogeneous catalysis, and energy from renewables. He has been awarded the Oscar Masi prize for the industrial innovation (1991), the eni Innovation Award, and the National Innovation Prize (both 2011). He holds 30 patents and has authored 50 scientific papers.

---

Western economy, society and culture are largely founded on the availability of cheap energy, mostly produced by burning fossil fuels: coal, oil, and natural gas. Although it is unlikely that this situation will radically change within a few decades, we are facing increasing concern about it, basically due to a couple of major reasons. On the one side, the observed climate change, particularly global warming, is widely held as a consequence of the increase of atmospheric CO₂ concentration due to fossil fuel combustion. On the other side, people are increasingly aware that fossil fuels reserves are huge but not unlimited.

Expansion and social progress is however boosting a further increase of energy demand, particularly in the transportation sector which, in developed countries, accounts for most of this increase. This is the reason why carbon-neutral biofuels are largely considered a possible way to satisfy the energy demand without dramatically increasing the CO₂ content of the atmosphere.

Biofuels production basically started late in the 1980s when Brazil, and soon after USA, started mass production of ethanol (bioethanol). In the 1990s, several European countries started biodiesel manufacture. Currently, bioethanol and biodiesel are usually referred to as first generation biofuels.

Ethanol is obtained by fermentation of carbohydrates, the primary products of photosynthesis, with a technology which, for ethanol diluted aqueous solutions, was basically already known in protohistoric times (ca. 2000 B.C.) as mentioned, e.g., in the Bible (Genesis 9, 20–24) and the Odyssey (Book IX). Many centuries later, Henry Ford proved that ethanol is a valuable fuel for cars, largely used today as a gasoline component, either as such or as its tert-butyl ether derivative, bio-ETBE.

Biodiesel, in its turn, is produced by exploiting the capability of plants to transform carbohydrates into all the other molecules they need, including triglycerides, i.e. lipids, oils and fats. Triglycerides can undergo a transesterification reaction with methanol with formation of a mixture of fatty acids methyl esters to be used in mixture with (or even instead of) mineral diesel.

Both ethanol and biodiesel are produced from traditional crops, usually grown for food or animal feed purposes. This generated a hot debate about a possible food vs. fuel conflict, i.e. about the risk of diverting farmland or crops for biofuels production at the expense of food supply. A possible solution to this conflict could be offered by the development of second generation biofuels, manufactured from agricultural and forest residues and from ligno-cellulosic non-food energy crops, possibly grown on poor lands, currently not exploited for food or feed production. The value of these second generation biofuels is acknowledged, for instance, by the European Community which, in its RED directive, states that “For the purposes of demonstrating compliance with national renewable energy obligations…, the contribution made by biofuels produced from wastes, residues, non-food cellulosic material, and ligno-cellulosic material shall be considered to be twice that made by other biofuels”.¹

Since in the European market gasoline consumption is continuously reducing whereas diesel demand is steadily increasing,² the following discussion will mainly focus on the use of biomass as raw material for the production of diesel fuels.

Mineral diesel basically results from refining and fractional distillation of crude oil between 160 and 380 °C at atmospheric pressure and is mostly formed by mixtures of paraffins containing between 12 and 20 carbon atoms per molecule. In these mixtures, linear paraffins are particularly appreciated due to their high cetane number, i.e. their excellent ignition performances. The most obvious biological source of linear alkyl chains of the appropriate length is provided by the fatty acids which occur in the vegetable oils and animal fats. This is why research on biodiesel, and later its production, started with simple chemical modifications of these raw materials.

Diesel biofuels from lipids: Fatty Acids Methyl Esters (FAME)

Vegetable oils are basically complex mixtures of lipids which, in their turn, are mainly triglycerides, i.e. esters of glycerol with both saturated and unsaturated fatty acids. In principle, most oils could be directly used as fuels for diesel engines, as already recognized by Rudolf Diesel himself who, at the Paris Expo of 1900, showed an engine running on peanut oil. The scientific optimism of those years, celebrated by the most famous Excelsior ballet (1881), prompted Diesel to develop a visionary idea of the use of vegetable oils as a possible way to the development of poorest countries: “Regions devoid of coal and petroleum may find abundant source of power in the use of this engin (sic in the text) and some vegetable or nut oil”.³ Today, however, the use of vegetable oils as such is hampered by a number of drawbacks, including high viscosity and scales formation. So, it is largely preferred to subject vegetable oils to a transesterification reaction with methanol, affording glycerol and a mixture of fatty acids methyl esters. This mixture is usually referred to as FAME (Fatty Acids Methyl Esters) or, simply, as biodiesel (reaction (1)).

As already mentioned, Europe played a pivotal role in the development of the biodiesel industry. The first commercial biodiesel plant went into operation in 1991 in Austria. Then, for several years both plants’ capacity and actual production and consumption experienced extremely high average annual growth close, e.g., to 50% in the 2002–2006 period. This explosive development, however, resulted in some overcapacity, and further growth is expected to slow down. Europe, however, still remains the dominant player in the biodiesel industry and, within Europe, the main players are, in the order, Germany, France, Spain, and Italy. In 2010, the European production of biodiesel was 9.57 Mt.⁴ In the meanwhile, other regions led by North America and Asia had also started to develop their own biodiesel industries and today the European share can be estimated at about 45–50%.

Current European regulation (directive 2009/30/EC on fuel quality⁵) allows for a maximum of 7% of FAME in diesel fuel, a mixture usually referred to as B7. Blends above this level require guarantees by car manufacturers and, possibly, amended servicing schedules, e.g. including more frequent oil and filter changes. The use of 100% biodiesel (B100) may require more drastic modifications, for example, of seals or piping. On the other hand, biodiesel marketing did not require significant and expensive changes in the distribution system.

The biodiesel properties are determined, to some extent, by the characteristics of the oil used as raw material: length of the fatty acids chains, content of carbon–carbon double bonds, free fatty acids and moisture content, etc. Most important is the low price of the feedstock, since it accounts for 80% or more of the biodiesel production cost, and its large availability, constant as possible all over the year. Few vegetable oils are actually exploited and, among them, rapeseed oil is by far the most important, followed by soybean, sunflower, and palm oil.

Transesterification is a reversible reaction requiring a catalyst, which can be either a strong acid or a strong base.⁶ In commercial plants, basic catalysts are used since, compared to acidic ones, they are more active (up to 4000 times⁷) and pose fewer corrosion problems. Homogeneous catalysts such as sodium methoxide or potassium or sodium hydroxide are most frequently used. The transesterification is usually carried out at 60 °C at high methanol/oil ratio, so as to obtain high conversion. FAME yield can be higher than 99.5%. Nevertheless, the process still has a few drawbacks. On one hand, the presence of free fatty acids and moisture in the feed causes the formation of soaps distributed among the two phases, with some catalyst deactivation and loss. As a consequence, the process can only accept a restricted range of feedstock, typically with less than 0.5% of free fatty acids and 0.2% of moisture. On the other hand, at the end of the reaction the basic catalyst must be neutralised, and the resulting salts are difficult to remove from glycerol to get it with a suitable purity. In order to solve these drawbacks, an obvious challenge was the search for an efficient heterogeneous catalyst.^6,8,9 Several problems, however, have been met in the search for a suitable heterogeneous transesterification catalyst. So, for instance, most heterogeneous catalysts are less active than homogeneous ones on a weight basis (because they have less active sites per mass unit) and higher temperatures (up to 200–250 °C) and pressures are required. Several catalysts are poisoned by the small amounts of free fatty acids which are usually present in the feedstock. In some cases, stability problems also occur: Na-ETS-10 proved to be quite active, but its performances were affected by some leaching of sodium.⁸ Despite all these difficulties, IFP (Institut Francais du Petrole) developed an industrial process (Esterfip-H™) which is currently commercialized by Axens.^10,11

This new, continuous process is catalyzed by a zinc aluminate mixed oxide. The reaction section includes two fixed-bed reactors fed with methanol and vegetable oil. In order to shift the methanolysis equilibrium, after each reactor esters and glycerol are separated in a settler. After complete methanol removal by evaporation, the biodiesel is purified in an adsorber to remove any dissolved glycerol. The Esterfip-H process provides very high yields (up to 99.8%) due to the absence of ester loss due to soap formation and, compared with the homogeneously catalyzed processes, it avoids any chemicals consumption and waste stream. Furthermore, it affords a very pure (>98%), salt-free glycerol, thus greatly reducing problems connected with its purification.¹⁰

The first plant based on the Esterfip-H process was started by Diester Industries in Sète (Southern France) in 2006. In the following year, a second plant with the same technology was started up in Sweden.

A few enzymes, typically lipases, also catalyze the transesterification. In principle, they offer several advantages. Lipases are not deactivated by free fatty acids: in contrast, they are able to catalyze, at the same time, their esterification with methanol and the transesterification reaction.^8,12 Co-produced glycerol is of good quality, only polluted by minimal amounts of water or inorganic materials. On the other hand, enzymes are relatively expensive and can be deactivated by several feed impurities.

However, no matter of its production process, traditional biodiesel presents several disadvantages from the performance point of view such as limited oxidative stability (largely due to unsaturated fatty acids chains), poor cold properties, high solvency (resulting in some degradation of rubber and elastomers and filters plugging), formation of deposits at injector tips, and higher NO_x emission than conventional diesel. Furthermore, biodiesel production is accompanied by the co-production of glycerol (ca. 9% by weight) and any increase of biodiesel production could result in a glut in the relatively small glycerol market. For these reasons, a radical innovation has been proposed in the field of biomass-derived diesel fuels, consisting of the extensive hydrogenation of the vegetable oils to hydrocarbon mixtures. This technology will be dealt with in the following section.

Diesel biofuels from lipids: hydrotreating of vegetable oils

The extensive hydrotreating of vegetable oils is a way to produce a synthetic fuel that, compared with mineral diesel, has a similar chemical composition and similar (or even better) chemical–physical properties, and can be easily blended with conventional refinery streams. As a consequence, with respect to FAME, both the process and the product are better integrated in the existing refinery infrastructure and fuel distribution system. Furthermore, any co-production of glycerol is avoided.

The hydrogenation of vegetable oil to diesel fuel has been studied for several years.¹³ Since 1992 the Saskatchewan Research Council and Arbochem investigated the use of refinery technology to convert vegetable oils into a product resembling diesel fuel and found that the use of a medium severity hydroprocess yielded a product with high cetane values (55–90). The hydroprocessing technology was tested in a pilot facility using both tall oil and canola oil.¹⁴ More recently several companies have been developing proprietary processes for the triglycerides hydroprocessing including Neste Oil (NexBTL™ process),¹⁵ Petrobras,¹⁶ BP, Conoco–Phillips, Dynamic Fuels, Haldor Topsoe, Axens (Vegan™ process), and UOP–eni (Ecofining™ process). In the following, a few details will be provided about the latter technology.¹⁷

UOP and eni started a collaborative research effort in 2005 to develop jointly a process largely based on hydrotreating technologies already widely in use in refineries. According to this process, transformation of vegetable oils into diesel fuel occurs in two different steps. The first one is a catalytic hydrotreatment in which any carbon–carbon double bond initially present in the fatty acids chains is saturated and, at the same time, oxygen is removed from the triglyceride molecules, taking advantage of three competitive, simultaneous reactions: hydrodeoxygenation, decarboxylation, and decarbonylation, which remove oxygen as water, CO₂, or CO, respectively (reaction (2)).

The glycerol moieties of the triglyceride molecules afford propane which, when the process is run in a refinery, is easily recovered as a valuable co-product. At the same time, the linear carbon chains arising from the fatty acids moieties are transformed into linear paraffins. Hydrodeoxygenation affords hydrocarbons with the same number of carbon atoms of the parent fatty acid, while decarboxylation and decarbonylation produce paraffins with one carbon less. The reaction is usually run above 310 °C, in the presence of bimetallic catalysts based on mixtures of cobalt (or nickel) and molybdenum oxides supported on alumina. Under these conditions, the overall reaction is quite exothermic: ΔH depends on the number of double carbon–carbon bonds initially present in the fatty acids chain and can vary between ca. 400 kJ kg⁻¹ for the saturated stearic acid and ca. 1750 kJ kg⁻¹ for linolenic acid. Complete conversion of vegetable oil is observed even for short contact times.

The extent of hydrodeoxygenation, decarboxylation, and decarbonylation depends on the catalyst and on the reaction conditions. Decarboxylation and decarbonylation roughly occur at similar rates, as witnessed by the CO and CO₂ yields.¹⁸ Since vegetable oils only contain fatty acids with an even number of carbon atoms, the ratio between the amounts of C₁₇ and C₁₈ hydrocarbons in the products mirrors the ratio between the extent of decarboxylation plus decarbonylation and hydrodeoxygenation. The C₁₇/C₁₈ ratio increases with the temperature, while the extent of hydrodeoxygenation increases with hydrogen pressure.¹⁶ There are several reasons for pushing decarboxylation: hydrogen consumption is limited to the amount needed by carbon–carbon double bonds and, moreover, water is not produced, thereby preserving catalyst performance. However, the loss of carbon in the form of CO₂ and CO reduces the overall hydrocarbon yield: from 1 kg of tristearin, 0.856 kg of octadecane can be produced by hydrodeoxygenation, compared with 0.809 kg of heptadecane obtained by decarboxylation.

At the end of the first step of the process, the only liquid products are linear paraffins obtained in 99% volume yield. Co-products and by-products are mainly in the gas phase: propane, water, and CO/CO₂. As already stated, linear paraffins are very much appreciated in diesel fuels due to their excellent ignition performances. Unfortunately, however, they are also characterized by poor cold properties (e.g., cloud point and pour point), which severely limit the volume which can be blended with mineral diesel. Therefore a proper compromise between the amounts of linear and branched hydrocarbons is to be realized. To this aim, the reaction mixture produced in the first hydrotreatment is subjected to an isomerization step to produce a diesel fuel rich in isoparaffins. This second step is usually run under a hydrogen atmosphere (hydroisomerization) in the presence of a suitable bi-functional catalyst, with a metal loaded on an acidic carrier. Typical acidic supports studied for the hydroisomerization reaction are amorphous oxides or their mixture (e.g. HF-treated Al₂O₃, SiO₂–Al₂O₃, ZrO₂–SO₄²⁻), zeolites (Y, Beta, Mordenite, ZSM-5, ZSM-22), silicoaluminophosphates (SAPO-11, SAPO-31, SAPO-41), or mesoporous materials (MCM-41, Al-MCM-41).¹⁹ Unfortunately, besides hydroisomerization, these catalysts also promote undesired cracking reactions. In order to reduce the cracking extent, a proper combination of porosity and mild acidity is needed. Porosity can be tailored using proper synthetic parameters: MSA, an amorphous silica–alumina with controlled porosity in the region of mesoporous and a mild acidity, turned out to be one of the best supports for this purpose after being loaded with a noble metal such as palladium, platinum, or nickel.²⁰

The process is very flexible to the feedstock and can be fed with conventional edible oils (rapeseed, soybean, palm, etc.), but also with inedible ones (e.g., jatropha or camelina), with tallow and other animal fats, and even with waste cooking oils, hence turning out to be a bridge between first and second generation biodiesel. A simplified flow diagram of the Ecofining process is shown in Fig. 1.

Fig. 1 Simplified flow diagram of the Ecofining process.

The process output, referred to as Green Diesel, has better properties compared to those of biodiesel and even of mineral diesel. Particularly, compared to FAME, it is more stable and does not contain oxygen. So, it also has a higher heat value, close to that of mineral diesel (43 MJ kg⁻¹ for ultra low sulphur diesel, 38 for FAME, 44 for Green Diesel). It has virtually no sulphur or aromatics, and its cetane numbers (70–90) are higher than those of mineral diesel (51) or of FAME (50–65).

Life cycle analysis (LCA) has been performed on the compared productions of mineral diesel, conventional biodiesel, and Green Diesel. Both biodiesel and Green Diesel allow substantial reduction of fossil resources consumption per energy unit in the fuel. The larger reduction is for production of Green Diesel from soybean or palm oil: up to 84–90% respectively. Also greenhouse gas emissions are greatly reduced for both biodiesel and Green Diesel and, again, the latter (produced from soybean) performs better than conventional biodiesel.²¹

So, hydrotreating of vegetable oils (or even of possible waste animal fats) is a new, sustainable route to produce premium quality diesel fuels. The technology has been developed into several different processes and the first plants are already running. In Porvoo (Finland), two different plants came on stream in 2007 and 2009 with a total capacity of 380 kt per year (thousand metric tons per year). Two more plants, both with 800 kt y⁻¹ of capacity, were then built in Singapore (started up in 2010) and in Rotterdam (started up in September 2011).²² All these plants are based on the NexBTL™ process of Neste Oil. On the other hand, Valero is building a plant for the production of Green Diesel in Norco, Louisiana, near its St. Charles refinery site, with a capacity of ca. 415 kt per year (137 000 000 gal per year). This project will be the first full-scale application of its kind in the US to use the UOP–eni Ecofining process.²³

Both biodiesel and Green Diesel productions, however, rely upon vegetable oils obtained from oleaginous crops with relatively poor productivities: the most productive crop is the oil palm which can afford ca. 5 t oil per ha per year. Rapeseed, the most common raw material for the European biodiesel industry, only produces 0.6–1.2 t oil per ha per year. Soybean and sunflower productivities are even lower.

To overcome this limitation, two very different routes are currently pursued. On one hand, attention has been focused on the most productive photosynthetic organisms of our planet, which are also good at producing oils: algae and, particularly, microalgae. On the other hand, compared with oleaginous crops, ligno-cellulosic ones can produce huge amounts of biomass, up to 80 t per ha per year for Brazilian sugar cane or 40–50 t per ha per year for European giant reed (Arundo donax). So, an obvious challenge is the development of second generation biodiesel manufactured by the exploitation of this ligno-cellulosic material, rather than that of valuable edible oils. These two approaches will be discussed in the following.

Diesel biofuels from lipids: microalgae as raw material

Several algal species are able to accumulate oil in their cells, up to 50% or more of their dry weight. So, these algae can be considered true oleaginous crops. As already mentioned, microscopic algae are the most productive photosynthetic organisms on the planet. The reason for their high productivity is that they complete an entire growth cycle every few days. Our knowledge about the algal growth in high-density systems, either open ponds or photobioreactors, is rapidly increasing. Dry-matter productivities of 100–130 t per ha per year have been already relatively established.²⁴ At odds with traditional crops, algae do not need farmland nor freshwater, since they can be grown in sea water or even in waste water, either domestic or industrial, which can be even purified upon their growth. So, any competition with food production is greatly reduced.

For these reasons, possible biofuel production from algae gained much attention in the last few years. Nevertheless, several major problems are still to be solved in order to develop a reliable process to produce an algal oil which would be used, e.g., to feed a biodiesel or a Green Diesel plant:

• algal growth and oil production do not usually occur simultaneously. Lipids are synthesized and stored (as oil) as an energy reserve under particular stress conditions, usually poor availability of nitrogen in the culture medium. Under these conditions, however, algae are not able to grow. So, algae either grow or produce lipids. To get, at the same time, high biomass production and high lipids content still remains an elusive objective;

• algae only grow up to a maximum concentration which can be very low: usually 1% or so. Above this concentration, competition for sun light prevents further growth of the cells. As a consequence, greatly diluted streams have to be managed;

• algal cells include large amounts of endocellular water which is difficult to remove to get a high dry matter content. Water removal, if needed (as usually it is), can be very energy-demanding and expensive;

• finally, algal lipids are often tightly bonded to cell membranes and their recovery is all but easy.

Several companies are working on these issues. Particularly, in order to get first-hand experience on these problems and, possibly, to solve a few of them, the eni Refining and Marketing Division built a pilot plant unit in a Sicily refinery, to grow microscopic, autochthonous algae in a one hectare pond. The microalgae are cultivated in waste waters, and fed with a CO₂ stream coming from a partial oxidation plant of the refinery. The lipids withdrawn from the algae are perfectly suitable for Green Diesel production.²⁵

Diesel biofuels from carbohydrates: fermentation to microbial lipids

The routes to biodiesel so far discussed all start from lipids, i.e. from the biomolecules which are more similar to standard diesel components. In the last few years, however, different approaches have been explored to produce biodiesel starting from carbohydrates which are definitely available in much larger amounts than lipids.

The first of these approaches still use lipids as intermediates, taking advantage from particular microorganisms that are able to grow on carbohydrates and to produce and accumulate large amounts of lipids. As a matter of fact, several heterotroph microorganisms, especially yeasts but also bacteria or fungi, are able to grow on different sugar sources and then, under proper fermentation conditions, to accumulate lipids up to 70% of their dry weight. These yeasts can be grown in relatively small reactors and fed with sugars obtained from the hydrolysis of ligno-cellulosic biomass. A few yeasts, particularly, are able to exploit not only glucose produced from the cellulose hydrolysis, but also other hexoses and even the pentoses arising from the hemicellulose. So, farmland, or even marginal fields, can be used not to grow low-yield oleaginous crops (such as rapeseed or soybean) but to produce substantial amounts of ligno-cellulosic biomass to be hydrolyzed in order to get sugars to feed the oleaginous yeast. In this way, one hectare of low-quality land, cultivated with giant reed, would afford enough biomass and sugars to feed yeasts eventually able to produce 8 oil t y⁻¹, a higher productivity than that obtained from palm tree. The recovered oil is fully suitable to be processed, e.g., by an Ecofining plant to afford high quality Green Diesel (Fig. 2).²⁶

Fig. 2 Production of microbial oil, a suitable feed for the Ecofining process.

Should the raw material be a high yield ligno-cellulosic crop not grown in any competition with edible crops (e.g., on poor marginal fields), resulting hydrocarbons would be a second generation biofuel. Currently, this route is actively pursued by eni: the growth of several oleaginous yeasts (mainly, strains from Rhodotorula²⁷ and Lipomyces genera) has been already successfully tested and a few of them have been grown in fermentors with volumes up to 200 l. According to press releases, Neste Oil and, to a lesser extent, UOP and BP together with Martek Biosciences are also active in the field.

Diesel biofuels from carbohydrates: chemical conversion of carbohydrates to hydrocarbons

A completely different approach to the biodiesel production starting from carbohydrates was pioneered by James Dumesic in a seminal paper.²⁸ According to this route, pentoses and hexoses are dehydrated to furfural and hydroxymethylfurfural (HMF), respectively. These aldehydes are further reacted, taking advantage of their capability to undergo aldol condensation, including crossed condensation with acetone, affording C₈ to C₁₅ oxygenated intermediates which are eventually hydrodeoxygenated to get C₈–C₁₅ liquid alkanes. For instance, HMF upon hydrogenation and subsequent aldol self-condensation affords a C₁₂ product eventually subjected to hydrodeoxygenation to get C₁₂ alkane. Alternatively, HMF can undergo crossed aldol-condensation with acetone, and then possibly again with HMF, allowing us to produce, upon hydrodeoxygenation, C₉ and C₁₅ alkanes (Fig. 3).²⁶

Fig. 3 Proposed scheme for the formation of alkanes from C₆ sugars according to Dumesic and co-workers.²⁸

At least part of this approach, including base catalyzed aldol condensation to increase the molecular weight of sugar-derived intermediates, is possibly being exploited on a demonstration scale within the Bioforming® process by Virent Inc. According to press releases, however, the plant is producing much gasoline but little diesel. A simplified scheme of the Bioforming process is shown in Fig. 4.

Fig. 4 Simplified scheme of Virent's Bioforming process.²⁹

A similar approach was developed by Avelino Corma starting from 2-methylfuran (sylvan), a C₅ synthon which can be easily produced from furfural, in its turn arising from the hydrolysis of hemicellulose. Sylvan undergoes a trimerization reaction (aqueous sulfuric acid; 60 °C), selectively affording a C₁₅H₁₈O₃ compound which can then be hydrodeoxygenated (50 bar; 120 to 350 °C; Pt on carbon and Pt on TiO₂ as catalysts) to 6-butyl undecane, a valuable component for diesel fuels (Fig. 5).³⁰ The choice of an unusual raw material such as sylvan is justified not only due to its hydrophobic character that allows its spontaneous separation from water and avoids any need of water distillation, but also because, in its alkylation reactions, polymer formation is inhibited since one of the two reactive α-positions is blocked by the unreactive methyl group.

Fig. 5 Production of C₁₅ alkanes from C₅ sugars according to Corma and co-workers.³⁰

Diesel fuels can also be prepared from γ-valerolactone, obtained from biomass-derived carbohydrates via levulinic acid. γ-Valerolactone undergoes decarboxylation at elevated pressures (36 bar) over a silica/alumina catalyst to afford CO₂ and butenes which can then be oligomerized to C₈–C₁₆ olefins.³¹ Alternatively, γ-valerolactone can be hydrogenated to valeric acid which, in turn, can be transformed into pentyl valerate, which proved to be a valuable component for diesel fuel.³²

Biofuels from thermal treatment of the whole biomass: liquefaction and pyrolysis bio-oils

The approaches so far described to produce diesel fuels from renewable raw materials rely upon the selective transformation of a biomass fraction, either the lipids or the carbohydrates, into fatty acids methyl esters or paraffins. Other completely different approaches, however, can be envisaged, based on completely unselective thermal treatments aiming at transforming the whole biomass into complex mixtures of very different organic compounds with fairly high heat value (bio-oils) or, alternatively, into a mixture of hydrogen and carbon monoxide (syngas, or bio-syngas when obtained from biomass) which can then be used to build hydrocarbons through the well known Fischer–Tropsch reaction.

In the following, the production of bio-oils will be dealt with. The synthesis of hydrocarbons from bio-syngas through the Fischer–Tropsch reaction (biomass to liquids processes, usually referred to as BtL processes) will be the subject of the next paragraph.

Bio-oils can be mainly obtained through two different technologies: pyrolysis and liquefaction.⁸ Their composition and quality depend on the nature of the biomass and the operating conditions: temperature, pressure, heating rate, and residence time. Bio-oils have greater energy densities than raw biomass, typically by a factor of 2 to 6–7 times,³³ so the conversion of a biomass in a bio-oil should simplify handling, transportation and storage, increasing the feasibility for large-scale bioenergy facilities.^33,34

A few typical properties of pyrolysis and liquefaction bio-oils are reported in Table 1.

Table 1 Typical properties of pyrolysis and liquefaction bio-oils (from wood and municipal sorted waste, respectively) and of heavy fuel oil

Property	Wood pyrolysis oil^a	Municipal sorted waste liquefaction oil^b	Heavy fuel oil^a
a From ref. 8. b Original data, unpublished. c Lower heating value, MJ kg^−1.
Moisture content, wt%	15–30	<5	0.1
pH	ca. 2.5
Density	1.2	ca. 1.02	0.94
Elemental composition, wt%
Carbon	54–58	70–75	85
Hydrogen	5.5–7.0	9–10	11
Oxygen	35–40	10–12	1
Nitrogen	0–0.2	3–6	0.3
Ash	0–0.2	<0.1	0.1
Higher heating value, MJ kg^−1	16–19	35^c	40
Viscosity @ 50 °C, cP	40–100	500–1000	180
solids, wt%	0.2–1	Not detected	1

---

Pyrolysis is the thermal degradation of relatively dry biomass in the absence of oxygen. It results in the production of solid (charcoal), liquid (bio-oil) and gaseous products. Depending on the operating conditions, the pyrolysis process can be arranged into three categories: conventional (slow heating rate, long residence time), fast (high temperature range, heating rate 10–200 K s⁻¹, residence time 0.5–10 s) and flash pyrolysis (heating rate > 1000 K s⁻¹, residence time <0.5 s, very fine particles as feed).

Conventional pyrolysis is traditionally used for charcoal production. On the other hand, when liquids are the preferred products, the pyrolysis is run at high temperatures (400 up to 900 °C) for very short contact times (seconds or less), under kinetic rather than thermodynamic control. This is the reason why pyrolysis bio-oils are quite unstable and difficult to store. Furthermore, they still have high oxygen content (35–40%) and a content of water as high as 15–30%, derived from both the original biomass moisture and the occurrence of dehydration reactions (Table 1). So, they are usually water soluble and not miscible with most organic solvents, nor with conventional diesel fuel and major problems are encountered in directly using them for diesel or gasoline blending.³⁵ Yields, on the other hand, are quite good: liquids account for up to 60–75% of the dry weight of the starting biomass. By-products are char (10–15%) and gas (5–10%). It is also worthwhile to note that the presence of water, although reducing the bio-oil heating value and flame temperature, also reduces its viscosity.

Liquefaction,⁸ in its turn, is run at milder temperatures (250–350 °C) for longer reaction times (0.5–2 h), in the presence of a liquid phase, most often provided by the constituent water of the raw biomass. So, liquefaction is particularly suitable for wet biomass, such as domestic organic wastes and sewage sludges, saving the energy costs of drying the feedstock before its transformation. The choice to have a liquid, aqueous phase at 300 °C or so implies the need to work under significant pressure, up to ca. 180 bar. Typical residence times are 10–90 minutes.³⁶ Under these conditions, biomass is transformed into hydrophobic bio-oils which, compared with pyrolysis ones, are more stable and have a much lower oxygen content (Table 1). So, they also have a higher heat value and, usually, are water insoluble and more miscible with mineral diesel and with organic solvents. Yields are lower than that of pyrolysis and liquids account for 40–55% of the dry weight of the starting biomass, still retaining, however, 75–80% of its initial heat value. Some gas (15–20%, mostly CO₂) and small amounts of char (10–15%) also form, while a significant part of the starting biomass (20–30%) results in water soluble by-products. As a consequence, the process waste water often has a too high COD to be disposed in conventional water treatment plants. A possible solution to this problem is the use of process water as a feed to grow microorganisms, possibly oleaginous ones, which, in their turn, can then be fed to the liquefaction reaction. In this way, the bio-oil yield is slightly improved and, at the same time, the COD of the process water is greatly reduced, so it can be disposed in a conventional treatment plant (Fig. 6).³⁷

Fig. 6 Simplified production and possible uses of bio-oil from domestic organic waste.

From a chemical point of view, both pyrolysis and liquefaction bio-oils are complex mixtures containing, e.g., alcohols, aldehydes, acids, esters, ketones, sugars, lignin derived phenols, oxygen- and/or nitrogen-containing heterocycles, etc. Their composition is the result of very complex processes in which several reactions occur: dehydration and decarboxylation are probably the main pathways to reduce oxygen content, but hydrogen transfer reactions, aromatization, depolymerization, condensation and, in the presence of water, hydrolysis also occur, together with significant isomerization.

Hopefully, several bio-oils may be directly exploited as fuel in robust engines, possibly for electric energy production rather than for transportation. Alternatively, bio-oils can be up-graded to produce high quality fuels. Several strategies have been envisaged for their up-grading, including cracking (e.g. FCC), decarboxylation, and hydrotreating/hydrodeoxygenation.³⁸ So, for instance, UOP already announced the construction of a demonstration unit to convert cellulosic biomass by the RTP (Rapid Thermal Processing) technology; the produced liquids will then be up-graded into high-quality, renewable gasoline, diesel and jet fuels using proprietary hydroprocessing technologies. The bio-oil up-grading, however, is still a field in which further research is needed.

Diesel biofuels from thermal treatment of the whole biomass: BtL processes

As already mentioned, BtL processes are based on the decomposition of the whole biomass into a mixture of hydrogen and carbon monoxide (bio-syngas) which is then used as raw material to synthesize hydrocarbons through the Fischer–Tropsch reaction. Biofuels produced by BtL processes have very good properties, basically not different from (or even better than) those of diesel and gasoline obtained by oil refining.

The main steps of a BtL process include thermal decomposition of the biomass into syngas (gasification), the syngas clean up, and the Fischer–Tropsch reaction, which affords the synthetic fuel and which is possibly followed by a hydrocracking step. The overall energy efficiency (i.e., fuel heat value/feed heat value) can be as high as 40–50%. Similar technologies are already used to produce huge amounts of synthetic fuels from coal (coal to liquid, CtL, processes) or natural gas (gas to liquid, GtL, processes).

BtL processes: syngas production and purification

Gasification is a process in which the biomass reacts with air (or oxygen) and/or steam to produce a bio-syngas that mostly contains H₂, CO, CO₂, and water. The gasification process is a complex combination of reactions occurring in the solid, liquid, and gas phases including pyrolysis, partial oxidation, combustion, and the water gas shift (WGS, reaction (3)):

CO + H₂O → CO₂ + H₂ (3)

 ΔH₂₉₈ = −41 kJ mol⁻¹

The process is typically carried out at high temperature (800–1000°C), under low to moderate pressure (1–20 bar). The H₂/CO ratio in the outlet gas can range between 0.5 and 1.8, depending on the biomass composition, the gasifying agent (air, oxygen, or steam) and the gasification technology. Different gasification reactors have been developed, including updraft gasifier, downdraft gasifier and fluidized-bed gasifier.⁸

A very critical step in the whole process is the syngas clean-up. Biomass contains significant amounts of nitrogen, sulphur and chlorine and, therefore, undesired contaminants, such as H₂S, COS, nitrogen compounds (mainly ammonia and hydrogen cyanide), and HCl are produced, along with H₂ and CO and also with some tars and particulates.

Possible catalysts involved in the following FT reaction show very different tolerances to these impurities and, particularly, any FT process fed with bio-syngas requires a complex sequence of gas cleaning steps.

Sulphur compounds are the most critical to be controlled, since both hydrogen sulphide and COS are strong poisons, quickly forming catalytically inactive metal sulphides not only with any FT catalyst but also with the nickel catalyst of the reforming unit which can be included to convert into CO and H₂ the methane and light hydrocarbons which are always present in the bio-syngas and represent a significant part of its heat value.

Desulfurization of bio-syngas may be obtained by chemical or physical adsorption.

In chemical adsorption, a base reacts with the acid gases to form some complexes that, changing pressure and temperature, can in turn dissociate to release the acid gases. The most used bases are alkanolamines, particularly monoethanolamine (MEA) and methyl-diethanolamine (MDEA).

Processes based on physical adsorption use a solvent to adsorb acid gases by dissolution, typically at subzero temperature. Acid gases can then be released from the solvent by pressure reduction or temperature change.

Selexol™ (UOP) and Rectisol™ (Linde and Lurgi) are the most widely used physical processes, respectively using, as solvents, the dimethyl ether of polyethylene glycol (DMPEG) and methanol. As a matter of fact, about 75% of the world's syngas produced from oil residue, coal and waste is purified by the Rectisol™ process. Solubilities of H₂S and COS in methanol, under process conditions, allow a sulphur removal below 0.1 ppmv. Carbon dioxide is also removed.

BtL processes: the Fischer–Tropsch reaction and the subsequent hydrocracking step

The purified bio-syngas is then used to feed a Fischer–Tropsch section. The Fischer–Tropsch reaction (FT) was discovered in the 1920s by Franz Fischer and Hans Tropsch at the Kaiser-Wilhelm-Institut für Kohlenforschung in Mülheim an der Ruhr (Germany).³⁹ Starting from CO and H₂, the FT reaction affords hydrocarbons according to reaction (4) which is basically the reverse of a steam reforming reaction:

nCO + 2nH₂ → C_nH_2n + nH₂O (4)

 ΔH₂₉₈ = ca. −150 kJ mol⁻¹ CO

Reaction (4) is definitely exothermic. Its thermochemistry slightly depends upon the value of n but, as n increases, ΔH₂₉₈ quickly levels at ca. −150 kJ mol⁻¹ CO (Fig. 7).

Fig. 7 Calculated ΔH₂₉₈ of reaction (4) per mole of CO as a function of the number of carbon atoms in the product.

Linear alkenes and alkanes (presumably arising from alkenes hydrogenation) are the main products, but several other reactions also occur, including methane formation by CO hydrogenation (reaction (5))

CO + 3H₂ → CH₄ + H₂O (5)

 ΔH₂₉₈ = −206 kJ mol⁻¹

the formation of oxygenated products (mainly alcohols, reaction (6))

nCO + 2nH₂ → C_nH_2n+1OH + (n − 1)H₂O (6)

the Boudouard reaction (reaction (7))

2CO → C + CO₂ (7)

 ΔH₂₉₈ = −172 kJ mol⁻¹

some coke formation (reaction (8))

CO + H₂ → C + H₂O (8)

 ΔH₂₉₈ = −131 kJ mol⁻¹

and the water gas shift (reaction (3)).

The core of any FT process is its catalyst. Only a few metals show catalytic activity in the FT synthesis. The reaction is generally assumed to start with the adsorption of CO on the catalyst surface where it reacts with adsorbed hydrogen to afford a methylene group which, in its turn, is responsible for the C–C chain growth. Probably, at least two reaction paths co-exist for the initial steps of the FT reaction: one in which carbon monoxide is adsorbed in a dissociative way (i.e., its C–O bond is cleaved before any possible reaction with hydrogen), and another in which some hydrogenation by adsorbed hydrogen atoms precedes the C–O cleavage (Fig. 8).⁴⁰

Fig. 8 Possible initial steps in FT catalysis. H_ads denotes some form of hydrogen adsorbed on the catalyst surface.

So, a good catalyst should adsorb both CO, possibly in a dissociative way, and H₂. Furthermore, since metal oxide formation is always possible under FT conditions either by dissociative CO absorption or by metal reaction with co-produced water, the metal oxide should be easily reduced under the reaction conditions. With this respect, most early transition metals are not good FT catalysts because, despite their favourable CO adsorption, they form very stable oxides that are not reduced under FT conditions. Iridium, platinum and palladium adsorb CO in a non-dissociative manner, while metals of groups 11 and 12 hardly adsorb it: none of them is an effective FT catalyst. The specific activities (i.e., the reaction rates per unit surface area of metal) of most of the metals of the former Group VIII (all but osmium) were assessed under comparable conditions, and ruthenium proved to be the most active catalyst.⁴¹ As expected on the basis of the classical work by Sabatier (1902), nickel (and, even to a larger extent, palladium) showed very high selectivity towards methane formation, obviously a feature not appreciated in a FT catalyst.⁴⁰ Osmium was successively evaluated,⁴² but turned out to be ca. 100 times less active than ruthenium.⁴³ So, the best FT catalysts are based on iron, cobalt and ruthenium, with nickel, rhodium, and osmium (and, possibly, rhenium)⁴⁴ being moderately active (Fig. 9).⁴⁵

Fig. 9 Catalytic activity of transition metals in the FT reaction. Blue denotes early transition metals able to adsorb CO in a dissociative way; they, however, show poor or no capability to adsorb H₂ and their oxides are not reduced under usual FT conditions: accordingly, these metals are not active FT catalysts. Orange denotes late transition metals and a few main group elements which show poor or no CO adsorption and no FT activity. Yellow shaded elements have good H₂ adsorption capability and reducible oxides; they, however, adsorb CO in a non-dissociative way and, as a matter of fact, are poor FT catalysts. Brilliant green denotes the best FT catalysts. Nickel, rhodium, osmium and possibly rhenium (pale green) are moderately active.⁴⁵

However, FT industrial plants require huge amounts of catalyst and ruthenium is too rare and expensive to be used on this scale. As a matter of fact, cobalt and iron are the only metals of choice for industrial applications. Iron is obviously cheaper than cobalt but, to select between them, a key issue is the carbon feedstock. Iron is a good water gas shift catalyst and, for this reason, is particularly suitable for hydrogen-poor syngas, such as those obtained from coal or biomasses. Cobalt performs better with an almost stoichiometric ratio between hydrogen and carbon monoxide, so it is preferred when the carbon feedstock is natural gas. Alternatively, cobalt can also be used with hydrogen-poor syngas, provided that the H₂/CO molar is adjusted by a WGS unit between the gasification and the FT reactors.

Apart from methane (which usually forms in amounts higher than expected), the FT products distribution follows the Anderson–Schulz–Flory model.^40,46 So, the FT output is always a complex mixture of products ranging from methane to waxes formed by high molecular weight linear paraffins. A proper choice of catalyst and reaction conditions allows tuning, to some extent, the composition of the final mixture, but it is impossible to force the process to produce selectively a well defined range of products, i.e. middle distillates. So, the best strategy to maximize diesel production is to select conditions which allow obtaining the maximum amount of long chain linear paraffins that can be then fed to a hydrocracking stage, transforming them into a most valuable fuel.

When evaluating the possibility to build up BtL plants, it should be underlined that FT plants are quite expensive. Integrated BtL plants will be obviously even more expensive, since they also require the gasification step and the complex purification train for the bio-syngas. Significant savings, however, can be obtained by realizing rather huge plants. Commonly used estimates agree on figures of 15–30 000 bpd (barrels per day) as the best choice for a BtL plant or, in more traditional units, 750–1500 kt per year. To feed a 750 kt per year BtL plant (energy efficiency 0.3; fuel heat value of 37.8 GJ t⁻¹) with giant reed (Arundo donax; dry biomass productivity: 40 t per ha per year with a heat value of 17.4 GJ t⁻¹) harvested with an efficiency of 0.8, a circular area with a radius of ca. 23.2 km is needed. This value is not discouraging: as a matter of fact, a significant pilot experience of a BtL plant has been done in Germany by Choren-Shell. According to press releases and presentations by the company, the Alpha plant by Choren produced 200 t per year of top-quality diesel fuel (SunDiesel™) with typical yields, based on the dry biomass, around 20%. The company, however, shut down the plant and discontinued its activities in July 2011. Other companies which have been developing commercial Fischer–Tropsch technologies, mainly devoted to GtL applications (e.g. Sasol, Shell, BP, COP, eni-IFP/Axens, ExxonMobil, Statoil, Rentech, Syntroleum), are in a favourable position for BtL projects.

Conclusions

There are several reasons to expect an increasing demand of biofuels in the next future. Particularly, several routes can be envisaged to produce diesel biofuels.

Hydroprocessing of vegetable oils to green diesel seems to be a promising alternative to FAME biodiesel and definitely allows the production of a very high quality diesel fuel, with high flexibility with respect to the feedstock.

Sugars, obtained from lignocellulosic biomass via a proper hydrolysis treatment, can be fermented to second generation lipids and then transformed into green diesel. The combination of high productivity crops and efficient fermentation conditions provides lipid productivities larger than those of conventional oleaginous crops. Even larger productivity might be achieved by microalgae.

Bio-oils obtained from biomasses by thermal treatments like pyrolysis or liquefaction are also very interesting. Their main drawback is represented by the poor quality of the oil requiring severe upgrading processing, still to be fully developed.

BtL seems to be a promising route for the production of synthetic diesel from ligno-cellulosic stuff. The fuel is of superb quality, but some questions are still open, i.e. plant capacity vs. biomass availability and logistics.

Eventually, however, the success of one or the other of these technologies will depend on several factors, including the availability and the quality of the feedstock, the complexity of the process, and the quality of the final biofuel.

Notes and references

1. European Union, RED Directive 2009/28/EC on the promotion of the use of energy from renewable sources, 23 April 2009, Article 21. Available at http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2009:140:0016:0062:en:PDF. Last accessed 6 March 2012.
2. See, e.g., European Petroleum Industry Association (Europia), 2010 Activity Report. Available at http://europia.com/Content/Default.asp?PageID=391. Last accessed 8 March 2012.
3. Chem. Abs. 1912, 1984.
4. http://www.ebb-eu.org/stats.php. Last accessed 7 May 2012.
5. Available at http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2009:140:0088:0113:EN:PDF. Last accessed 20 March 2012.
6. A. Sivasamy, K. Y. Cheah, P. Fornasiero, F. Kemausuor, S. Zinoviev and S. Miertus, ChemSusChem, 2009, 2, 278–300.
7. M. W. Formo, J. Am. Oil Chem. Soc., 1954, 31, 548–559 and reference therein.
8. G. W. Huber, S. Iborra and A. Corma, Chem. Rev., 2006, 106, 4044–4098 and references therein.
9. C. Perego and A. Bosetti, Microporous Mesoporous Mater., 2011, 144, 28–39.
10. L. Bournay, D. Casanave, B. Delfort, G. Hillion and J. A. Chodorge, Catal. Today, 2005, 106, 190–192.
11. E. Freund, in Sustainable Industrial Processes, ed. F. Cavani, G. Centi, S. Perathoner and F. Trifirò, Wiley-VCH, Weinheim, 2009, pp. 439–448.
12. For the use of enzymes to catalyze a closely related reaction see: E. M. Usai, E. Gualdi, V. Solinas and E. Battistel, Bioresour. Technol., 2010, 101, 7707–7712.
13. See, e.g. J. Gusmão, D. Brodzki, G. Djéga-Mariadassou and R. Frety, Catal. Today, 1989, 5, 533–544.
14. M. Stumborg, A. Wong and E. Hogan, Bioresour. Technol., 1996, 56, 13–18.
15. J. Jakkula, V. Niemi, J. Nikkonen, V. M. Purola, J. Myllyoja, P. Aalto, J. Lahtonen and V. Alopaeus, US Patent Appl., 2004/0230085 to Neste Oil and others.
16. J. R. Gomes, US Patent Appl. 2006/0186020 to Petrobras.
17. F. Baldiraghi, M. Di Stanislao, G. Faraci, C. Perego, T. Marker, C. Gosling, P. Kokayeff, T. Kalnes and R. Marinangeli, in Sustainable Industrial Processes, ed. F. Cavani, G. Centi, S. Perathoner and F. Trifirò, Wiley-VCH, Weinheim, 2009, pp. 427–438.
18. G. W. Huber, P. O'Connor and A. Corma, Appl. Catal., A, 2007, 329, 120–129.
19. H. Deldari, Appl. Catal., A, 2005, 293, 1–10 and references therein; A. Corma, A. Martinez, S. Pergher, S. Peratello, C. Perego and G. Bellussi, Appl. Catal., A, 1997, 152, 107–125; V. Calemma, S. Peratello and C. Perego, Appl. Catal., A, 2000, 190, 207–218; V. Calemma, S. Peratello, F. Stroppa, R. Giardino and C. Perego, Ind. Eng. Chem. Res., 2004, 43, 934–940.
20. C. Perego, L. Sabatino, F. Baldiraghi and G. Faraci, World Patent 2008/58664 to eni s.p.a. and UOP LLC.
21. T. Kalnes, T. Marker and D. R. Shonnard, Int. J. Chem. React. Eng., 2007, 5, A48.
22. Neste, 2011. Neste Oil starts up renewable diesel plant in Rotterdam. Available at: http://www.dieselnet.com/news/2011/09neste.php. Last accessed 2 March 2012.
23. http://www.hydrocarbonprocessing.com/Article/2815857/Valero-JV-selects-Dresser-Rand-to-supply-rotating-equipment-on-renewable-diesel-project.html. Last accessed 7 May 2012.
24. J. B. K. Park, R. J. Craggs and A. N. Shilton, Bioresour. Technol., 2011, 102, 35–42; A. Sun, R. Davis, M. Starbuck, A. Ben-Amotz, R. Pate and P. T. Pienkos, Energy, 2011, 36, 5169–5179.
25. E. N. D'Addario, F. Capuano, E. D'Angeli and R. Medici, World Patent 2010/416 to eni s.p.a.
26. D. Bianchi, G. Franzosi and A. Romano, World Patent 2010/46051 to eni s.p.a; G. Franzosi, D. Bianchi, F. Pizza, D. Cucchetti, C. Compagno and S. Galafassi, It. Pat. Appl. MI2010A 1951 to eni s.p.a.
27. S. Galafassi, D. Cucchetti, F. Pizza, G. Franzosi, D. Bianchi and C. Compagno, Bioresour. Technol., 2012, 111, 398–403.
28. G. W. Huber, J. N. Cheda, C. J. Barrett and J. A. Dumesic, Science, 2005, 308, 1446–1450.
29. Adapted from http://newenergyandfuel.com/wp-content/uploads/2009/06/Virents-Bioforming-Process-Chart.png. Last accessed 19 April 2012.
30. A. Corma, O. de la Torre, M. Renz and N. Villandier, Angew. Chem., Int. Ed., 2011, 50, 2375–2378.
31. J. Q. Bond, D. M. Alonso, D. Wang, R. M. West and J. A. Dumesic, Science, 2010, 327, 1110–1114.
32. J.-P. Lange, R. Price, P. M. Ayoub, J. Louis, L. Petrus, L. Clarke and H. Gosselink, Angew. Chem., Int. Ed., 2010, 49, 4479–4483.
33. P. C. Badger and P. Fransham, Biomass Bioenergy, 2006, 30, 321–325.
34. J. G. Rogers and J. G. Brammer, Biomass Bioenergy, 2009, 33, 1367–1375.
35. A. Shihadeh and S. Hochgreb, Energy Fuels, 2000, 14, 260–274; Q. Zhanga, J. Changa, T. Wanga and Y. Xua, Energy Convers. Manage., 2007, 48, 87–92.
36. A. A. Peterson, F. Vogel, R. P. Lachance, M. Froling, M. J. Antal and J. W. Tester, Energy Environ. Sci., 2008, 1, 32–65.
37. A. Bosetti, D. Bianchi, G. Franzosi and M. Ricci, World Patent 2011/30196 to eni s.p.a.
38. M. Stocker, Angew. Chem., Int. Ed., 2008, 47, 9200–9211.
39. F. Fischer and H. Tropsch, Chem. Ber., 1926, 59, 830–831; F. Fischer and H. Tropsch, Chem. Ber., 1926, 59, 832–836.
40. P. M. Maitlis and V. Zanotti, Chem. Commun., 2009, 1619–1634.
41. M. A. Vannice, J. Catal., 1975, 37, 449–461.
42. M. Leconte, A. Theolier, D. Rojas and J. M. Basset, J. Am. Chem. Soc., 1984, 106, 1141–1142.
43. N. Bungane, C. Welker, E. van Steen, J. R. Moss and M. Claeys, Z. Naturforsch., 2008, 63b, 289–292.
44. V. Ponec, Catal. Today, 1992, 12, 227–254.
45. C. Perego, R. Bortolo and R. Zennaro, Catal. Today, 2009, 142, 9–16.
46. P. J. Flory, J. Am. Chem. Soc., 1936, 58, 1877–1885.

---