Following the train of components of Figure 2.1 and given the potential options for gasification, gas cleaning and conditioning, synthesis and separation, many routes to produce methanol from biomass can be imagined. The authors have previously analyzed the techno-economic performance of methanol from wood through 6 concepts, which will be recapitulated here. At the end of the section, results will be placed into broader perspective with other literature and with fossil gasoline and diesel.

Selection of Concepts

Some concepts chosen resemble conventional production of methanol from nat­ural gas, making use of wet gas cleaning, steam reforming, shift, and a solid-bed methanol reactor. Similar concepts have previously been analyzed by Katofsky (1993). Advanced components could offer direct or indirect energy benefits (liquid phase-methanol synthesis, hot gas cleaning), or economic benefits (autothermal reforming). Available process units are logically combined so the supplied gas composition of a unit matches the demands of the subsequent unit, and heat leaps are restricted if possible. The following considerations play a role in selecting concepts:

1. The IGT direct oxygen fired pressurized gasifier, in the normal and maximized H2 option, and the Battelle indirect atmospheric gasifier are considered for synthesis gas production because they deliver a medium calorific nitrogen undiluted gas stream and cover a broad range of gas compositions.

2. Hot gas cleaning is only sensible if followed by hot process units like reforming or (intermediate temperature) shifting. Hot gas cleaning is not applied after atmospheric gasification since the subsequent pres­surization of the synthesis gas necessitates cooling anyway.

3. For reforming fuel gas produced via an IGT gasifier, an autothermal reformer is chosen, because of higher efficiency and lower costs. The high hydrogen yield possible with steam reforming is less important here since the H2:CO ratio of the gas is already high. The BCL gasifier, however, is followed by steam reforming to yield more hydrogen.

4. Preceding liquid-phase methanol synthesis, shifting the synthesis gas composition is not necessary since the reaction is flexible toward the gas composition. When steam is added, a shift reaction takes place in the reactor itself. Before gas-phase methanol production the composi­tion is partially shifted and because the reactor is sensitive to CO2 excess, part of the CO2 is removed.

5. After the methanol passes through once, the gas still contains a large part of the energy and is expected to suit gas turbine specifications. The same holds for unreformed BCL and IGT gases, which contain energy in the form of C2+ fractions. When the heating value of the gas stream does not allow stable combustion in a gas turbine, it is fired in a boiler to raise process steam. The chemical energy of IGT+ gas is entirely in hydrogen and carbon monoxide. After once through meth­anol production, the gas still contains enough chemical energy for combustion in a gas turbine.

6. Heat supply and demand within plants are to be matched to optimize the overall plant efficiency.

7. Oxygen is used as oxidant for the IGT gasifier and the autothermal reformer. The use of air would enlarge downstream equipment size by a factor 4. Alternatively, oxygen-enriched air could be used. This would probably give an optimum between small equipment and low air sep­aration investment costs.

These considerations led to a selection of 6 conversion concepts (see Table 2.3). The six concepts selected potentially have low-cost and/or high-energy efficiency. The concepts are composed making use of both existing commercially available technologies, as well as (promising) new technologies.

Genetic modification to improve alfalfa over the past century has increased resistance to several diseases and pests and widened the range of environmental adaptation of the crop by producing varieties that differ in fall dormancy and winter hardiness. Most improvements in forage quality of alfalfa have occurred through changes in harvest management and production practices. Alfalfa pro­duced as feed for ruminant livestock is harvested frequently at early maturity when the leaf to stem ratio is high, producing hay that is high in protein and easily digested. Maximum forage yield, which occurs at later maturity stages in alfalfa, is usually sacrificed in order to produce high-quality hay. For competitive use of alfalfa as a biofuel feedstock, research is needed to develop alfalfa germ — plasm and management strategies that yield more biomass (both leaf and stem) with minimal production costs.

Marquez-Ortiz et al. (1999) reported that individual stem diameter was her­itable and controlled by additive genetic effects and suggested that selection for larger stems in alfalfa was feasible. Volenec et al. (1987) found that selection for high yield per stem may be an effective means to increase forage yield, but plants may have less digestible, larger stems. Germplasms from southern Europe referred to as Flemish types are a genetic source for large stem size and resistance to foliar diseases in alfalfa, but display early maturity, lack winter hardiness, and are susceptible to root and crown diseases (Barnes et al., 1977).

The effects of plant population or density on stem, leaf and total forage yield have been well documented in alfalfa. As alfalfa plant densities increase, annual forage yield per land area unit increases, but yield of individual alfalfa stems and number of stems per plant decreases (Cowett and Sprague, 1962; Rumbaugh 1963). Hansen and Krueger (1973) reported that higher plant densities produced finer stems, decreased root and crown weights and increased leaf drop due to shading. Volenec et al. (1987) stated that stem diameter and nodes per stem decreased as plant density increased and that shoot weight was an important component of plant weight, especially at high plant densities. Decreasing plant density to approximately 45% (180 plants m-2) of that conventionally used in alfalfa hay production stands (450 plants m-2) and delaying harvest until the green pod stage maximized leaf and stem yield in four unrelated alfalfa germplasms (Figure 5.1). The reduced plant density decreased plant-to-plant competition for light, water, and nutrients, which minimized leaf drop caused by shading. Delay­ing harvest until late flower to green pod maturity stages increased stem yield and maximized total forage yield (Lamb et al., 2003).

Plant Density
Plants / m2

Although many E85 vehicle demonstrations have been made using off-road vehi­cles such as airplanes,10 snowmobiles,11 boats, and all-terrain vehicles, there are not currently any significant numbers of these vehicles operating on E85.

material compatibility

Some materials that are commonly used with gasoline-powered vehicles are not compatible with E85. These materials are degraded when in contact with E85 and cause leaks or fuel system contamination.13 Fortunately, there are many alternatives for these materials. Also, limited duration contact with E85 in many of these materials has shown no detrimental effects. Most degradation requires long-term contact with E85.

E85 can be used in both four-stroke and two-stroke spark-ignited engines. Four-stroke engines are widely used in on-road vehicles because they generally offer better emissions and fuel consumption than two-stroke spark-ignited engines. In countries with strict air-pollution standards, even most motorcycles generally employ this type of engine. The strict emissions standards are also contributing to more widespread use of four-stroke engines for off-road vehicle use, such as snowmobiles and all-terrain vehicles (ATVs).

Most four-stroke spark-ignited engines currently available today introduce the fuel into the air intake system, not directly into the cylinder. This means that the fuel will come into contact with the materials used in the intake manifold of the engine. Fuel also comes into contact with the engine cylinders and the fuel induction and storage systems of these engines.

Two-stroke engines are lighter and often have better power-to-weight ratios than four-stroke spark-ignited engines and are, therefore, often used in smaller vehicles or in cases where weight is a major design consideration.

Developing countries still widely use two-stroke spark-ignited engines in vehicles due to their lower costs and smaller sizes. Two-stroke spark-ignited engines complete a mechanical cycle in two strokes of the piston, or one engine revolution. These types of engines do not have separate intake and exhaust processes. Because of this, these engines produce power every revolution, leading to smaller, lighter engines. Unfortunately, this also leads to higher tailpipe emis­sions and problems with bypass, where raw fuel and air pass through the engine unburned. To help combat this, many two-stroke spark-ignited engines use crank­case compression to improve scavenging efficiency and to reduce bypass.

This means that in two-stroke spark-ignited engines fuel not only comes into contact with the fuel storage and delivery systems, the intake, and the engine cylinders, but also the engine crankcase and even the exhaust manifold. Further, residual fuel is left in the crankcase after the engine is stopped, leading to potential long-term exposure.

Unfortunately, to save weight, most of these engines use aluminum exten­sively in their blocks, leading to potential long-term corrosion problems. As described later, hard-anodized aluminum has been shown to be resistant to E85 degradation. At this time, the long-term use of E85 in off-road vehicles with two — stroke engines has not been studied.

Over the last decade, there has been substantial research on immobilizing enzymes at electrode surfaces for use in biofuel cells [12,14-15]. These immobilization strategies have been successful at increasing biofuel cell lifetimes to 7-10 days

[15]. Therefore, one of the main obstacles that is still plaguing enzyme-based biofuel cells is the ability to immobilize the enzyme in a membrane at the electrode surface that will extend the lifetime of the enzyme and form a mechan­ically and chemically stable layer, while not forming a capacitive region at the electrode surface.

The problem associated with bioelectrodes as reported in the literature is ineffective techniques for enzyme immobilization [16]. The most common tech­niques used are sandwich [17] or wired [16,18]. However, sandwich and wired techniques still leave the enzyme exposed to the matrix, so the enzyme’s three­dimensional configuration can change due to the harsh physical and chemical forces resulting in the loss of optimal enzymatic activity [14,16,18,19].

To solve these issues and offer a more stable enzyme immobilization, researchers have employed a micellar polymer (Nafion®). Nafion® is a perfluori — nated ion exchange polymer that has excellent properties as an ion conductor and has been widely employed to modify electrodes for a variety of sensor and fuel cell applications. The molecular structure of Nafion® is shown in Figure 12.2. Nafion® is a cation exchange polymer that has superselectivity against anions. Nafion® also preconcentrates cations at the electrode surface and serves as a protective coating for the electrode. A simple approach to obtain selective elec­trodes is performed by solvent casting of the Nafion® polymer directly onto the electrode surface. Nafion® can be employed for enzyme immobilization in three different ways by employing either the wired technique, sandwich technique, or entrapment technique [20].

It is extremely important to remember that even during the summer, when farms in the Northeast are at peak production, the Northeast still imports 95% of its food. During the winter, this number increases to 98% (7). As a result, 95-98% of all money spent on food in the Northeast leaves the Northeast, creating a tremendous cash flow out of region. At the same time, this represents a huge market that is virtually untapped by local growers. These local food producers have the advantage of not having to transport their product into the area. There­fore, promoting local production of food will save on the associated costs of long­distance hauling of food and the fuel associated with transportation. Environmen­tal impact will also be reduced as a result.

Reserve Food Supply

Additionally, the reserve food supply for the entire Northeast (i. e., all the food on supermarket shelves and in warehouses) will supply the food needs of its inhabitants for only 3 days. This constitutes a terrible vulnerability for the entire region, should anything such as terrorism interrupt this constant influx of food. This project, which promotes the decentralization of fuel production, encourages the expansion of local food production, and may potentially save family farms, will address all of these community problems.

The second phase of this project is composed of the following subsystems:

1. Energy plantation — gasification

2. Growth of fish feed from plants

3. Compost-based aquaponic greenhouses

4. Duplication in inner cities

Carlo N. Hamelinck

(currently working with Ecofys b. v. Utrecht,

The Netherlands)

Andrn P. C. Faaij

(Utrecht University, Copernicus Institute of Sustainable Development and Innovation, Utrecht, The Netherlands)

CONTENTS

Introduction………………………………………………………………………………………………………. 8

Technology……………………………………………………………………………………………………….. 9

Overview……………………………………………………………………………………………….. 9

Pretreatment………………………………………………………………………………………….. 9

Gasification…………………………………………………………………………………………. 10

IGT Gasifier……………………………………………………………………………….. 10

BCL Gasifier……………………………………………………………………………… 12

Oxygen Supply………………………………………………………………………….. 13

Gas Cleaning and Contaminant Limits………………………………………………. 13

Raw Gas versus System Requirements……………………………………… 13

Tar Removal……………………………………………………………………………… 15

Wet Gas Cleaning………………………………………………………………………. 17

Dry/Hot Gas Cleaning……………………………………………………………….. 19

Gas Conditioning………………………………………………………………………………… 20

Reforming………………………………………………………………………………….. 20

Water Gas Shift…………………………………………………………………………. 22

CO2 Removal…………………………………………………………………………….. 23

Methanol Synthesis…………………………………………………………………………….. 25

Fixed-Bed Technology………………………………………………………………. 26

Liquid-Phase Methanol Production…………………………………………… 27

Options for Synergy………………………………………………………………………………………… 28

Electricity Cogeneration by Combined Cycle…………………………………….. 28

Natural Gas Cofiring/Cofeeding…………………………………………………………. 29

Black Liquor Gasification…………………………………………………………………… 29

Other Biofuels via Gasification…………………………………………………………… 30

Hydrogen…………………………………………………………………………………… 30

Fischer-Tropsch (FT) Diesel……………………………………………………….. 30

Methanol to Diesel…………………………………………………………………….. 31

Methanol to Gasoline………………………………………………………………… 31

Dimethyl Ether (DME)……………………………………………………………….. 31

Techno-Economic Performance……………………………………………………………………… 32

Selection of Concepts…………………………………………………………………………. 32

Modeling Mass and Energy Balances………………………………………………… 33

Costing Method………………………………………………………………………………….. 36

Results………………………………………………………………………………………………… 37

Conclusions……………………………………………………………………………………………………… 44

References……………………………………………………………………………………………………….. 45

INTRODUCTION

Methanol (CH3OH), also known as methyl alcohol or wood alcohol, is the sim­plest alcohol. It can be used as a fuel, either as a blend with gasoline in internal combustion engines[2] or in fuel cell vehicles.[3] Also, methanol has a versatile function in the chemical industry as the starting material for many chemicals.

Methanol is produced naturally in the anaerobic metabolism of many varieties of bacteria and in some vegetation. Pure methanol was first isolated in 1661 by Robert Boyle by distillation of boxwood. In 1834, the French chemists Dumas and Peligot determined its elemental composition. In 1922, BASF developed a process to convert synthesis gas (a mixture of carbon monoxide and hydrogen) into methanol. This process used a zinc oxide/chromium oxide catalyst and required extremely vigorous conditions: pressures ranging from 300-1000 bar, and temperatures of about 400°C. Modern methanol production has been made more efficient through the use of catalysts capable of operating at lower pressures. Also the synthesis gas is at present mostly produced from natural gas rather than from coal.

In 2005, the global methanol production capacity was about 40 Mtonne/year, the actual production or demand was about 32 Mtonne (Methanol Institute 2005). Since the early 1980s, larger plants using new efficient low-pressure technologies are replacing less efficient small facilities. In 1984, more than three quarters of
world methanol capacity was located in the traditional markets of North America, Europe, and Japan, with less than 10 percent located in “distant-from-market” developing regions such as Saudi Arabia. But from that time most new methanol plants have been erected in developing regions while higher cost facilities in more developed regions were being shut down. The current standard capacities of conventional plants range between 2000 and 3000 tonnes of methanol per day. However, the newest plants tend to be much larger, with single trains of 5000 tonnes/day in Point Lisas, Trinidad (start-up in 2004), 5000 tonnes/day in Dayyer, Iran (start-up in 2006), and 5000 tonnes/day in Labuan, Malaysia (start construc­tion in 2006).

Methanol produced from biomass and employed in the automotive sector can address several of the problems associated with the current use of mineral oil derived fuels, such as energy security and greenhouse gas emissions.

This chapter discusses the technology for the production of methanol from biomass. For a selection of concepts, efficiencies and production costs have been calculated.

Structure and Composition of the Corn Kernel

Corn kernels contain, by weight, approximately 70% starch, 9% protein, 4% fat and oil, and 9% fiber on a dry basis [10]. Most corn grown for ethanol production is #2 yellow dent corn, so named because of the indentation in the top of the dried kernel. Energy is stored in the seed in the form of starch and oil, which are segregated to the endosperm and germ, respectively (Figure 4.1). Different pro­teins are contained in the endosperm, germ, and tip cap. The gluten protein fraction is found in the endosperm, bonded to the starch. The seed contents are protected by a waxy coat and fibrous outer layer (the pericarp). Fiber is also present in the germ and tip cap.

Ethanol yield potential varies among corn hybrids [11, 12], and also depends on agronomic practices and environmental factors. Corn hybrids are being devel­oped and marketed specifically for enhanced ethanol production, and seed corn is labeled for sale with high extractable starch (for wet milling) or high ferment­able starch content (for dry grind ethanol processing). Total starch content and total extractable starch content do not necessarily correlate with ethanol yields obtained in dry grind processing of the whole kernel; instead, high-performing varieties have been identified by empirical testing. Grain from high ethanol-

yielding hybrids reportedly results in ethanol yields up to 4% higher than the yield from mixed commercial grain, representing an additional one to two million dollars to a 40-million-gallon-per-year dry grind ethanol facility [13]. Spectro­scopic methods using near-infrared (NIR) technology have been developed for use at ethanol plants to predict the ethanol yield potential of samples of whole corn kernels [14].

Solvent-forming species, including C. acetobutylicum and C. beijerinckii, are mesophilic, growing best between 30° and 40°C. The pH varies during the fermentation and can drop from an initial value of 6.8-7.0 to about 5.0-4.5 (acidogenesis) and can also rise up to 7.0 later in the fermentation (solventoge — nesis). It has been suggested that the switch to solvent production is an adaptive response of the cell to the low medium pH resulting from acid production (Bahl et al., 1982).

Solventogenic clostridia can be grown on simple media such as ground corn, molasses, whey permeate, or on semidefined and defined media. When semi — defined and defined media are used, a wide array of vitamins and minerals are required in addition to a carbohydrate source. Clostridia can utilize a wide range of carbohydrates. C. acetobutylicum and C. beijerinckii can utilize starch, hexoses, pentoses, and cellobiose. Currently, those clostridia that are able to utilize cellu­lose directly produce little or no solvents. Recently, attempts have been made to express cellulase genes in the solventogenic clostridia.

The uptake of carbohydrates in the solventogenic clostridia is achieved by a phosphoenolpyruvate (PEP)-dependent phosphotransferase system (PTS). This mechanism involves simultaneous uptake and phosphorylation of substrate that results in the conversion of glucose to glucose-6-phosphate, which is subsequently metabolized to pyruvate via the Embden-Meyerhof-Parnas (EMP) pathway (Mitchell, 2001). Fructose is converted to fructose-1-phosphate and enters the EMP pathway upon conversion to fructose 1,6-bisphosphate. D-xylose is con­verted to D-xylulose by the xylose isomerase enzyme and the metabolism pro­ceeds by a phosphorylation reaction. The reaction is catalyzed by xylulokinase, which results in the formation of D-xylulose-5-phosphate. The pentose phosphate pathway utilizes enzymes transaldolase and transketolase to convert D-xylulose — 5-phosphate to glyceraldehyde-3-phosphate and fructose-6-phosphate (Singh and Mishra, 1995). The glyceraldehyde-3-phosphate and fructose-6-phosphate enter the EMP pathway leading to the formation of pyruvate. The ability of solvento — genic clostridia to metabolize these sugars is important when corn is considered as the starting material for fermentation, as all of these sugars can be derived from corn or corn coproducts.

Solvent producing clostridia metabolize substrates in a biphasic fermentation fashion. During the first phase, acid intermediates (acetic and butyric acids), hydrogen, and a large amount of ATP are produced. In the second phase, butanol, acetone, and ethanol are produced, and hydrogen and ATP production decrease (Jones and Woods, 1986). CO2 is produced during both phases of growth—two moles are produced from each mole of glucose metabolized to pyruvate—but CO2 production in the solventogenic phase is higher as an additional mole is produced for every mole of acetone produced. The simplified overall fermentation pathway is given in Figure 6.1.

During the acidogenic phase, cells typically grow exponentially due to the high amount of ATP (3.25 mol/mol of glucose) being produced (Jones and Woods, 1986). The enzymes phosphate acetyltransferase and acetate kinase convert acetyl-CoA to acetate and, analogously, phosphate butyltransferase and butyrate kinase convert butyryl-CoA to butyrate during this phase of growth. The pH of the fermentation broth decreases as butyric and acetic acids accumulate. The acetic and butyric acids produced during the fermentation may be freely perme­able to the cell membrane and these acids equilibrate the internal (bacterial) and fermentation broth pH. Both reduction of pH and accumulation of acetate and butyrate have been associated with triggering solventogenesis (Jones and Woods, 1986).

The solventogenic phase is typically associated with stationary phase. ATP production is reduced to 2 mol/mol of glucose during this phase. The fermentation intermediates (acetic and butyric acids) are reassimilated and converted into acetone and butanol. It has been suggested that butyric and acetic acids are reassimilated by the action of the enzyme acetoacetyl-CoA:acetate/butyrate:CoA transferase (Andersch et al., 1983). This enzyme catalyzes the reaction that transfers CoA from acetoacetyl-CoA to either acetate or butyrate. Acetate is converted to acetyl-CoA, which can be converted to acetone, butanol, or ethanol. Butyrate is converted to butyryl-CoA, which can only be used to produce butanol. This is because there is no metabolic pathway to regenerate acetyl-CoA from butyryl-CoA. When CoA is removed from acetoacetyl-CoA, acetoacetate is pro­duced, which can be transformed directly into acetone and CO2 by acetoacetate decarboxylase.

The central core of both the acidogenic and solventogenic pathways is the series of reactions that produces butyryl-CoA from acetyl-CoA. Thiolase con­denses two molecules of acetyl-CoA into one molecule of acetoacetyl-CoA. Acetoacetyl-CoA is reduced to 3-hydroxybutyryl-CoA by hydroxybutyryl-CoA dehydrogenase. From this, crotonyl-CoA is formed by dehydration, catalyzed by

crotonase. The carbon-carbon double bond in crotonyl-CoA is reduced with NADH to produce butyryl-CoA. This last step is catalyzed by butyryl-CoA dehydrogenase (Bennett and Rudolph, 1995).

The accumulation of both acids (butyrate and acetate) and solvents (acetone, butanol, and ethanol) in the fermentation broth is toxic to the microorganism and eventually causes cell death. The shift to solventogenesis is effective in extending the fermentation, but the butanol produced eventually reaches toxic levels. The presence of butanol in the membrane increases membrane fluidity and destabilizes the membrane and membrane-associated processes (Jones and Woods, 1986). The
maximum amount of solvents (total acetone, butanol, and ethanol) that the cell can tolerate is 20 gL-1 (Maddox, 1989). This limits the amount of glucose that can be fermented in batch culture to 60 gL-1 because using a higher concentration of glucose would result in incomplete substrate utilization due to butanol toxicity. Many studies today are focused on overcoming the butanol toxicity issue, whether by developing a more butanol tolerant microorganism or by selectively removing butanol from the fermentation broth.

The development of modern fuel cells has been driven by the need to generate clean and efficient electrical power for different applications. The first demon­stration of a fuel cell was described by William Grove in 1839. Grove was inspired by the observation that electrolysis of water produces hydrogen and oxygen gas. He ran the process in reverse by feeding oxygen to a Pt cathode and hydrogen to a Pt anode, where all electrodes were immersed in a common sulfuric acid bath. Several such cells were connected in series to generate voltage that was measured as shown in Figure 9.2. This “gas voltaic battery” was of little practical value.

Notable work done in the late 19th century by Ludwig Mond and Charles Langer aimed to produce a working fuel cell run on air and industrial coal gas. They were the first to suggest the use of “stacks” of cells with manifolds to deliver fuel and oxidant streams. William White Jaques, who is generally credited with coining the term “fuel cell,” replaced Grove’s sulfuric acid electrolyte with a phosphoric acid electrolyte. All of these systems fell short of producing practical power plants.

Francis T. Bacon, direct descendant of the renowned and similarly named philosopher, was first in developing a truly useful fuel cell power plant in 1959. It was a 5-kW system used to power a welding machine. It used nickel electrodes and an alkaline KOH electrolyte. Later that same year, Harry Karl Ihrig demon­strated the first fuel cell-powered vehicle, an Allis Chalmers tractor, powered by 1008 cells split into 112 stacks comprising a 15-kW power source [1,2,3]. Bacon’s fuel cell design was the product of more than a quarter century of effort on his part and was the basis for rapid development of fuel cells during the “space race” between the United States and the Soviet Union. Fuel cell development culmi­nated in the use of fuel cells in the Gemini and Apollo missions during the late 1960s and early 1970s [1,2,3]. It is interesting to note that during the earlier Mercury missions and Gemini missions 1 through 4, batteries were used for power. In Gemini 5 and later missions, power was generated by polymer elec­trolyte fuel cells (PEFCs). In the Apollo missions, PEFCs were replaced by the alkaline fuel cell design because of performance problems caused by oxygen crossover and PEM instability.

The early 1970s to the present can arguably be thought of as the modern era of fuel cell development. A modern fuel cell design generally falls into one of five categories: alkaline fuel cell (AFC), polymer electrolyte fuel cell (PEFC), phosphoric acid fuel cell (PAFC), molten carbonate fuel cell (MCFC), and solid oxide fuel cell (SOFC). All of the categories generate considerable attention in the scientific and patent literature.

Evolution of the modern DMFC is intimately linked to the development of the modern low-temperature PEFC. PEFCs fueled with a pure hydrogen or refor­mate stream have received the majority of attention in the literature as the top candidate for power systems ranging from the 10-3 to 105 W needed for power plants. The recent historical development of the PEFC follows a bifurcated path of parallel development of electrocatalysts and PEMs. DMFCs are under increas­ing consideration as an attractive alternative to PEFCs because of the inherent economic and technological limitations of hydrogen production and storage [1,2,3,4,5].

A review of ethanol reactions over the surface of supported noble metal catalysts has recently been published [46]. The noble metal-based catalysts most widely studied in ethanol steam reforming are those based on Pd, Pt, Ru, and Rh, and their behavior also depends, in this case, on the support. Comparative studies of g-Al2O3-supported catalysts showed Rh to be the most active metallic phase [47,48]. Taking into account that ethylene and methane were the main by-prod­ucts, the performance of different metals in the reformation of ethylene and methane is a key aspect to take into account. Rh turned out to be the most active in ethylene steam reforming, whereas Pd was almost inactive. In this context, the selectivity to ethylene of an Rh/Al2O3 catalyst showed a maximum at 973-1023 K, and then dropped at higher temperatures because ethylene reforming took place [47]. To optimize the hydrogen production by steam reforming of bioethanol on Rh/Al2O3 catalyst, high temperature and long contact times (high reactor volume/volumetric flow rate ratios) are required [8]. Although Rh is highly active in hydrogenation and its presence may reduce coke formation, under steam reforming conditions the catalyst deactivates by sintering and coke formation. The introduction of a small amount of O2 (0.4 vol%) has been proposed to reduce coke formation by combustion of carbonaceous species forming during the reac­tion. However, this combustion could be responsible for the formation of larger metal particles and consequently of the decrease in activity [8].

Several papers have been published in which the reforming of ethanol is carried out over Rh on supports other than alumina, namely, CeO2, ZrO2, and derived systems [29,49]. A high yield in H2 is found for catalysts containing ZrO2. This is related to the available oxygen on the surface, which participates in the WGS reaction [49].

As for palladium catalysts, it has recently been shown that the steam reform­ing of ethanol can be effectively carried out over a commercial Pd/g-Al2O3 catalyst, which does not produce ethylene as a by-product [23]. At 473-623 K, ethanol is dehydrogenated to acetaldehyde, which is decomposed to CH4 and CO. Then, at temperatures higher than 733 K, CH4 can be reformed as a function of the H2O/ethanol ratio.

Other promising systems for the steam reforming of ethanol could contem­plate the concurrence of different catalysts, with an appropriate active phase to catalyze each step of the total process. A two-layer fixed-bed reactor containing Pd/C and a Ni-based catalyst has been proposed to produce COx:H2 mixtures from ethanol steam reforming [50]. Ethanol can be converted to carbon oxides, CH4 and H2 over the Pd/C catalyst (608 K). Then, over the second layer containing the Ni-based catalyst, methane can be reformed with steam (923-1073 K) [50].