The process of ethanol production depends on the raw materials used. A general simplified representation of these processes is shown in Fig. 3.3, and a brief description of different units of the process is presented in the rest of this chapter. It should be noted that if sugar substances, such as molasses and sugarcane juice, are used as raw materials, then milling, pretreatment, hydrolysis, and detoxification are not necessary.

Milling, liquefaction, and saccharification processes are usually nec­essary for production of fermentable sugar from starchy materials, while milling, pretreatment, and hydrolysis are typically used for ethanol pro­duction from lignocellulosic materials. Furthermore, a detoxification unit is not always considered, unless a toxic substrate is fed to the biore­actors

Membranes can also be used for ethanol purification. Reverse osmosis (RO), which employs membranes impermeable to ethanol and perme­able to water, can be used for purification of ethanol from water. Using a membrane permeable to ethanol but not to water is another approach [9]. Pervaporation, a promising membrane technique for separation of organic liquid mixtures such as azeotropic mixtures or near-boiling-point mixtures, can also be used for separation of these azeotropes [81, 83]. It involves the separation of ethanol-water azeotrope or near-azeotropic ethanol-water composition (from about 95 to 99.5 wt% ethanol) through water-permeable (or water-selective) membranes to remove the rest of the water from the concentrated ethanol solution [84].

The cetane number (CN), which is related to the ignition properties, is a prime indicator of fuel quality in the realm of diesel engines. It is con­ceptually similar to the octane number used for gasoline. Generally, a compound that has a high octane number tends to have a low CN and vice versa. The CN of a DF is related to the ignition delay (ID) time, i. e., the time between injection of the fuel into the cylinder and onset of ignition. The shorter the ID time, the higher the CN, and vice versa.

Standards have been established worldwide for CN determination, e. g., ASTM D613 in the United States, and internationally the International Organization for Standardization (ISO) standard ISO 5165. A long straight-chain hydrocarbon, hexadecane (C16H34; trivial name cetane, giving the cetane scale its name) is the high-quality stan­dard on the cetane scale with an assigned CN of 100. A highly branched isomer of hexadecane, 2,2,4,4,6,8,8-heptamethylnonane (HMN), a com­pound with poor ignition quality, is the low-quality standard with an assigned CN of 15. The two reference compounds on the cetane scale show that CN decreases with decreasing chain length and increasing branching. Aromatic compounds that are present in significant amounts in petrodiesel have low CNs but their CNs increase with increasing size of n-alkyl side chains [12, 13]. The cetane scale is arbitrary, and com­pounds with CN > 100 or CN < 15 have been identified. The American standard for petrodiesel (ASTM D975) prescribes a minimum CN of 40, while the standards for biodiesel prescribe a minimum of 47 (ASTM D6751) or 51 (European standard EN 14214). Due to the high CNs of many fatty compounds, which can exceed the cetane scale, the lipid combustion quality number for these compounds has been suggested [14].

The use of biodiesel reduces most regulated exhaust emissions from a diesel engine. The species reduced include carbon monoxide, hydro­carbons, and particulate matter (PM). Nitrogen oxide (NOx) emissions are slightly increased, however. When blending biodiesel with petrodiesel, the effect of biodiesel is approximately linear to the blend level. A report summarizing exhaust emissions tests with biodiesel is available [15], and other summaries are given in Refs. [16, 17].

The structure of the fatty esters in biodiesel affects the levels of exhaust emissions. When using a 1991-model, 6-cylinder, 345-bhp (257-kW), direct-injection, turbocharged, and intercooled diesel engine, NOx exhaust emission increased with increasing number of double bonds and decreas­ing chain length for saturated chains [18]. Although often a trade-off is observed between NOx and PM exhaust emissions, no trade-off has been observed in this work when varying the chain length [18]. The CN and density were correlated with emission levels [18]. However, emissions are likely affected by the technology level of the engine. When conducting tests on a 2003-model, 6-cylinder, 14-L, direct-injection, turbocharged, intercooled diesel engine with exhaust gas recirculation (EGR), no chain length effect has been observed for NOx exhaust emissions, although the level of saturation still played a significant role [19]. PM exhaust emissions were reduced to levels close to the US 2007 regulations required for ultra-low-sulfur petrodiesel fuel. Also, PM levels were lower than those for neat hydrocarbons which would be enriched in “clean” petrodiesel fuel [19]. In both studies [18, 19], NOx emissions of the sat­urated esters were slightly below those of the reference petrodiesel fuel.

For petrodiesel fuel, higher CNs have been correlated with reduced NOx exhaust emissions [20]. This correlation has led to efforts to improve the CN of biodiesel fuels by using additives known as cetane improvers [8]. Despite the inherent relatively high CNs of fatty compounds, NOx exhaust emissions usually increase slightly when operating a diesel engine on biodiesel, as mentioned above. The relationship between the CN and engine emissions is complicated by many factors, including the technology level of the engine. Older, lower-injection pressure engines are generally very sensitive to CN, with increased CN causing signifi­cant reductions in NOx emissions, due to shorter ID times and the result­ing lower average combustion temperatures. More modern engines that are equipped with injection systems that control the rate of injection are not very sensitive to CN [21-23].

Historically, the first CN tests were carried out on palm oil ethyl esters [24, 25], which have a high CN, a result confirmed by later stud­ies on many other vegetable oil-based DFs and individual fatty com­pounds. The influence of the compound structure on CNs of fatty compounds has been discussed in more recent literature [26], with the predictions made in that paper being confirmed by practical cetane tests [7-9, 13]. CNs of neat fatty compounds are given in Table 5.1. In sum­mary, the results show that CNs decrease with increasing unsaturation and increase with increasing chain length, i. e., uninterrupted CH2 moi­eties. However, branched esters derived from alcohols such as iso­propanol have CNs competitive with methyl or other straight-chain alkyl esters [9, 27]. Thus, one long, straight chain suffices to impart a high CN, even if the other moiety is branched. Branched esters are of interest because they exhibit improved low-temperature properties.

Recently, cetane studies on fatty compounds have been conducted using the Ignition Quality Tester™ (IQT) [9]. The IQT is a further, automated development of a constant volume combustion apparatus (CVCA) [28, 29]. The CVCA was originally developed for determining CNs more rapidly with greater experimental ease, better reproducibil­ity, reduced use of fuel, and therefore less cost than the ASTM D613 method utilizing a cetane engine. The IQT method, which is the basis of ASTM D6890, was shown to be reproducible and the results compet­itive with those derived from ASTM D613. Some results from the IQT are included in Table 5.1. For the IQT, ID and CN are related by the fol­lowing equation [9]:

CNiqt = 83.99 (ID — 1.512)-0658 + 3.54 7 (5.1)

In the recently approved method ASTM D6890, which is based on this technology, only ID times of 3.6-5.5 ms [corresponding to 55.3-40.5 DCN (derived CN)] are covered as the precision may be affected outside that range. However, for fatty compounds, the results obtained by using the IQT are comparable to those obtained by other methods [9]. Generally, the results of cetane testing for compounds with lower CNs, such as more unsaturated fatty compounds, show better agreement over various related literature references than the results for compounds with higher CNs, because of the nonlinear relationship [see Eq. (5.1)] between the ID time and the CN, which was observed previously [30]. Thus, small changes at shorter ID times result in greater changes in CN than at longer ID times. This would indicate a leveling-off effect on emissions such as NOx, as discussed above, once a certain ID time with corresponding CN has been reached as the formation of certain species depend on the ID time. However, for newer engines, this aspect must be modified as discussed above.

Most of the properties are similar, with differences of only 5-10%. Ethanol is superior to methanol as it has a wider ignition limit (3.5-17) than methanol (2.15-12.8). Its calorific value (CV) (26,880 kJ/kg) is con­siderably higher than methanol (19,740 kJ/kg).

Ethanol is a much more superior fuel for diesel engines as its cetane number is 8 compared to the cetane number of 3 for methanol. There are wide resources for manufacturing ethanol compared with methanol. Therefore, ethanol is widely used as SI engine fuel in many countries. Methanol is superior to ethanol in one respect: Its vaporization rate is much higher than ethanol. Therefore, mixing with air rapidly forms a uniformly vaporized mixture and also burns uniformly. One major drawback of methanol is that it creates vapor locks because of the higher vaporiza­tion rate. Properties of ethanol and methanol as compared with petrol are listed in Table 7.3.

Btu British thermal unit. Heat energy necessary to raise the temper­ature of 1 lb of water 1°F.

cal or gcal Calorie or gram calorie. Heat energy required to raise the temperature of 1 mL of water 1°C (from 15 to 16°C).

electron volt 1.6 X 10 12 erg = 1.6 X 10 19 J = 23.06 kcal/mol. Energy gained by an electron passing through a potential of 1 V.

foot • pound ft • lb. Work energy needed to raise 1 lb to a height of 1 ft = 0.138 kg • m.

force Correct force definition can be obtained from the second law of Newton stating that the inertia is disturbed by unbalanced force, which causes acceleration on a body directly proportional to the force (F) and inversely proportional to the mass of the body F = K mf (F = Kmf where m is mass and f is acceleration). If all are reduced to unity, the unit of force becomes pound foot per second per second (poundal) or gram cen­timeter per second per second (dyne).

joule Work energy to raise 1 kg to a height of 10 cm = 0.1 kg • m = 0.74 ft • lb.

joule (electrical) 0.239 cal. Energy developed when 1 C of electrons (10.364 X 10~6 mole) passes through a potential of 1 V.

kcal/einstein Energy of a mole of a photon (einstein) of wavelength (in ^) 28589.7 ^ 1 kcal/mol.

power Rate of doing work, P = W/t.

J/s = W or ft • lb/s

or horsepower = 550 ft • lb/s or 33,000 ft • lb/min.

107 ergs/s = 1 J/s = 0.239 cal/s 550 ft • lb/s = 33,000 ft • lb/min 746 W = 178 cal/s 1000 W = 1.34 hp

3.6 X 106 J = 860 kcal/h = 3413 Btu/h 1.356 W = 0.324 cal

quantum Wavelength ^eV = 1239.8 ^—

work, W Force X distance (both in the same direction on a body).

621D ^ 242He + 211H + 2n + 43.1 Mev

Several designs and modifications are suggested:

2P ^ e+ + y + D D + P ^ T + y 2T ^ He + 2P The fusion reaction, omnipresent in the sun, needs to be tried out: 2H1 ^ e+ + v + H2

where two protons fuse, and deuterium, positron, and neutrino are evolved; energy is evolved in two steps; four protons are annihi­lated for each helium formed. Much of the reaction mechanism is yet unknown, but the model shows great promise.

[2] Geothermal source: Other than volcanic or geyser origin at an 8000-ft depth of the earth’s crust, it is possible to obtain geother­mal steam at 2000oC, which can be used for producing electricity.

Hot dry rock (HDR) remains out of reach at present capability of drilling. But “heat mining,” as estimated by Los Alamos Scientific Laboratory, promises 1.2 cents/MJ compared to 2 cents/MJ from an oil-fired thermal plant ($34/bbl).

[3] Aerodynamic generations: Several models are available. Low — velocity windmills are also being used. Wind is stronger at upper atmosphere; array of floating windmills are also designed.

[4] Hydrodynamics: High hopes are created by some hydroelectric firms, who proclaim that power can be effectively generated by ocean waves and ocean currents.

[5] Magnetohydrodynamic generators: High-temperature com­bustion gas expands through a nozzle where ionized sodium is intro­duced and directed to a magnetic field and a moving conductor cuts the field, and an electromagnetic field (EMF) is produced.

[6] Oil shale and oil sand: Though of limited supply, these have not been fully explored.

[7]

[8]

[9]

1.10.1 Biomass

Imitating the coal-based process, biomass conversion has also been tried and looks promising. Main sources of biomass are agricultural, horti­cultural, and forest wastes. Municipal organic solid wastes (which are also plenty) are potential resources as well. Considering biomass as a renewable resource, the bioconversion may be pyrolytic, where biogas and bio-oil are the main products and yet the residue contains some calo­rie value which can be further utilized (as adsorbents, filter beds, chars, etc.). Supercritical conversion and superheated steam reformation of bio­mass are recent techniques. During 1990—1997, quite a few reports appeared in the literature showing success and promise of catalytic or uncatalytic reformation of biomass to hydrogen (almost to 18% v/v) without any char or residues.

Temperature ranges of 340-650oC, with pressures of 22-35 MPa, are cited with as low as 30-s residence time, through supercritical flow reac­tors. The raw materials are widely varying: water hyacinth, algae, bagasse, whole biomass, sewage sludge, sawdust, and other effluents rich in organic matters. In some efficient carbon bed-catalyzed reactors, other products (i. e., carbon mono — and dioxides and methane) were also detected.

Historically, yeasts have been the most commonly used microorganisms for ethanol production. Yeast strains are generally chosen among S. cere — visiae, S. ellypsoideuse, S. fragilis, S. carlsbergensis, Schizosaccharomyces pombe, Torula cremoris, and Candida pseudotropicalis. Yeast species which can produce ethanol as the main fermentation product are reviewed, e. g., by Lin and Tanaka [8].

Among the ethanol-producing yeasts, the “industrial working horse” S. cerevisiae is by far the most well-known and most widely used yeast in industry and research for ethanol fermentation. This yeast can grow both on simple hexose sugars, such as glucose, and on the disaccharide sucrose. S. cerevisiae is also generally recognized to be safe as a food additive for human consumption and is therefore ideal for producing alcoholic beverages and for leavening bread. However, it cannot fer­ment pentoses such as xylose and arabinose to ethanol [14, 31]. There have been several research efforts to genetically modify S. cerevisiae to be able to consume xylose [33, 48-50]. Several attempts have been made to clone and express various bacterial genes, which is necessary for fer­mentation of xylose in S. cerevisiae [51, 52]. It resulted in great success, but probably not enough yet to efficiently ferment xylose with high yield and productivity [32].

Alternatively, xylose is converted to ethanol by some other naturally occurring recombinant. Among the wild-type xylose-fermenting yeast strains for ethanol production, Pichia stipitis and C. shehatae have reportedly shown promising results for industrial applications in terms of complete sugar utilization, minimal by-product formation, low sensitivity to temperature, and substrate concentration. Furthermore, P. stipitis is able to ferment a wide variety of sugars to ethanol and has no vitamin requirement for xylose fermentation [2].

Olsson and Hahn-Hagerdal [20] have presented a list of bacteria, yeasts, and filamentous fungi that produce ethanol from xylose. Certain species of the yeasts Candida, Pichia, Kluyveromyces, Schizosaccharomyces, and Pachysolen are among the naturally occurring organisms. Jeffries and Kurtzman [53] have reviewed the strain selection, taxonomy, and genetics of xylose-fermenting yeasts.

Utilization of cellobiose is important in ethanol production from lig — nocellulosic materials by SSF. However, a few ethanol-producing microorganisms are cellobiose-utilizing organisms. The requirement for addition of ^-glucosidase has been eliminated by cellobiose utilization during fermentation, since presentation of cellobiose reduces the activity of cellulase. Cellobiose utilization eliminates the need for one class of cellulase enzymes [2]. Brettanomyces custersii is one of the yeasts iden­tified as a promising glucose — and cellobiose-fermenting microorganism for SSF of cellulose for ethanol production [54].

High temperature tolerance could be a good characterization for ethanol production, since it simplifies fermentation cooling. On the other hand, one of the problems associated with SSF is the different optimum temperatures for saccharification and fermentation. Many attempts have been made to find thermotolerant yeasts for SSF. Szczodrak and Targonski [55] tested 58 yeast strains belonging to 12 different genera and capable of growing and fermenting sugars at temperatures of 40-46oC. They selected several strains belonging to the genera Saccharomyces, Kluyveromyces, and Fabospora, in view of their capacity to ferment glucose, galactose, and mannose at 40oC, 43oC, and 46oC, respectively. Kluyveromyces marxianus has been found to be a suitable strain for SSF [56].

Crop description. Dipteryx odorata—commonly known as sarapia, tonka bean, amburana, aumana, yape, charapilla, and cumaru—belongs to the family Leguminacea and grows in tropical areas (see Fig. 4.13). Major producing countries are Guianas and Venezuela. The tonka bean is the seed of a large tree. The kernel contains up to 46% oil on a dry basis. Major fatty acid composition of oil includes palmitic acid (6.1%), stearic acid (5.7%), oleic acid (59.6%), and linoleic acid (51.4%) [77].

Main uses. The oil is used in perfumery and as a flavoring material. Tonka extracts are used in the tobacco industry to impart a particular aroma. Few attempts have been made to use it as a raw material to pro­duce biodiesel. Abreu et al. conducted methanolysis of cumaru oil using different homogeneous metal (Sn, Pb, and Zn) complexes as catalysts. They found that pyrone complexes of different metals are active for cumaru-oil transesterification reaction [122].

Ethanol (ethyl alcohol) as a transport fuel has attracted a lot of atten­tion because it is seen as a relatively cheap nonpetroleum-based fuel. It is produced to a large extent from biomass, which aids agricultural economies by creating a stable market. Ethanol, being a pure compound, has a fixed set of physical as well as chemical properties. This is in con­trast to petrol and diesel, which are mixtures of hydrocarbons [3].

The use of alcohol in spark ignition (SI) engines began in 1954 in countries like the United States, Germany, and France. During World Wars I and II, gasoline shortages occurred in France and Germany, and alcohol was used in all types of vehicles, including military planes. Nowadays, it is used with gasoline (a mixture) in the United States and has become a major fuel in Brazil.

Ethyl alcohol can be produced by fermentation of vegetables and plant materials. But in countries like India, ethanol is a strong candidate since they possess the agricultural resources for the production of ethyl alcohol. It is a more attractive fuel for India because the productive capacity from sugarcane crops is high, of the order 1345 L/ha. Earlier, this fuel was not used in automobiles due to low energy density, high pro­duction cost, and corrosion. The current shortage of gasoline has made it necessary to substitute ethanol as fuel in SI engines.

Any new fuel that is going to be introduced should be evaluated from the aspect of availability, renewability, safety, and cost adaptability to the existing engines’ performance, economy, and finally emission. A mas­sive research effort has been put into the study and analysis of all these aspects for ethanol, which is now an established, viable alternative fuel for IC engines. The comparative properties of ethanol with petrol and diesel are shown in Table 7.1.

The PEMFC uses a solid polymer membrane as an electrolyte. The main components of this fuel cell are an electron-conducting anode consist­ing of a porous gas diffusion layer as an electrode and an anodic cata­lyst layer; a proton-conducting electrolyte, a hydrated solid membrane; an electron-conducting cathode consisting of a cathodic catalyst layer and a porous gas diffusion layer as an electrode; and current collectors with the reactant gas flow fields (see Fig. 9.3).

In the PEMFC, platinum or platinum alloys in nanometer-size par­ticles are used as the electrocatalysts with NafionTM (a DuPont trade­mark) membranes [3, 10—12]. The polymer electrolyte membranes have some unusual properties: In a hydrated membrane, the negative ions are rigidly held within its structure and are not allowed to pass through.

Electric current

,-WV

Only the positive ions contained within the membrane are mobile and free to carry a positive charge through the membrane. The proton exchange membrane (PEM) is a good conductor of hydrogen ions (pro­tons), but it does not allow the flow of electrons through the electrolyte membrane. As the electrons cannot pass through the membrane, elec­trons produced at the anode side of the cell must travel through an external wire to the cathode side of the cell to complete the electrical circuit in the cell.

In the PEMFC, the positive ions moving through the electrolyte are hydrogen ions, or protons. Therefore, the PEMFC is also called a proton exchange membrane fuel cell. The polymer electrolyte membrane is also an effective gas separator; it keeps the hydrogen fuel separated from the oxidant air. This feature is essential for the efficient operation of a fuel cell.

The heart of a PEMFC is the membrane electrode assembly (MEA), consisting of the anode-electrolyte-cathode assembly that is only a few hundred microns thick [11].

Electrochemistry of PEM fuel cells. All electrochemical reactions consist of two separate reactions: an oxidation half reaction occurring at the anode and a reduction half reaction occurring at the cathode.

Oxidation half reaction: 2H2 ^ 4H+ + 4e~

Reduction half reaction: O2 + 4H+ + 4e~ ^ 2H2O

Overall cell reaction: 2H2 + O2 ^ 2H2O

The H2 half reaction. At the anode, hydrogen (H2) gas molecules diffuse through the porous electrode until they encounter a platinum (Pt) par­ticle. Pt catalyzes the dissociation of the H2 molecule into two hydrogen atoms (H) bonded to two neighboring Pt atoms; here each H atom releases an electron to form a hydrogen ion (H+). These H+ ions move through the hydrated membrane to the cathode while the electrons pass from the anode through the external circuit to the cathode, resulting in a flow of current in the circuit.

The O2 half reaction. The reaction of one oxygen (O2) molecule at the cathode is a four-electron reduction process that occurs in a multistep sequence. The catalysts capable of generating high rates of O2 reduction at relatively low temperatures (~80°C) appear to be the Pt-based expen­sive catalysts. The performance of the PEMFCs is limited primarily by the slow rate of the O2 reduction half reaction, which is many times slower than the H2 oxidation half reaction.

Electrolyte. The polymer electrolyte membrane is a solid organic poly­mer, usually poly-[perfluorosulfonic] acid. Atypical membrane material used in the PEMFC is Nation [11, 12]. It consists of three regions:

1. The teflon-like fluorocarbon backbone, hundreds of repeating —CF2—CF—CF2— units in length

2. The side chains, — O-CF2-CF-O-CF2-CF2-, which connect the molec­ular backbone to the third region

3. The ion clusters consisting of sulfonic acid ions, SO3~ H+

The negative ion SO3~ is permanently attached to the side chain and cannot move. However, when the membrane becomes hydrated by absorbing water, the hydrogen ion becomes mobile. Ion movement occurs by protons (H ) bonded to water molecules, hopping from one SO3 site to another within the membrane. Because of this mechanism, the solid hydrated electrolyte is an excellent conductor of hydrogen ions.

Electrodes. The anode and the cathode are separated from each other by the electrolyte, the PEM. Each electrode consists of porous carbon to which very small Pt particles are bonded. The porous electrodes allow the reactant gases to diffuse through each electrode to reach the catalyst. Both platinum and carbon are good conductors, so electrons are able to move freely through the electrode [13, 14].

Catalyst. The two half reactions occur very slowly under normal con­ditions at the low operating temperature (~80°C) of the PEMFC. Therefore, catalysts are needed on both the anode and cathode to increase the rates of each half reaction. Although platinum is a very expensive metal, it is the best material for a catalyst on each electrode.

The half reactions occurring at each electrode can occur only at a high rate at the surface of the Pt catalyst. A unique feature of Pt is that it is sufficiently reactive in bonding H and O intermediates, as required to facilitate the electrode processes, and is also capable of effectively releas­ing the intermediate to form the final product. The anode process requires Pt sites to bond H atoms when the H2 molecule reacts; next, these Pt sites release the H atoms, as follows:

2H+ + 2e~ ^ H2
H2 + 2Pt ^ 2(Pt-H)

2(Pt-H) ^ 2Pt + 2H+ + 2e~

This optimized bonding to H atoms (neither very weak nor very strong) is a unique property of the Pt catalyst. To increase the reaction rate, the catalyst layer is constructed with the highest possible surface area. This is achieved by using very small Pt particles, about 2 nm in diameter, resulting in an enormously large total surface area of Pt that is accessible to gas molecules. The original MEAs for the Gemini space program used 4 mg of platinum per square centimeter of membrane area (4 mg/cm2). Although the technology varies with the manufacturer, the total platinum loading has decreased from the original 4 mg/cm2 to about 0.5 mg/cm2. Laboratory research now uses platinum loadings of 0.15 mg/cm2. For catalyst layers containing Pt of about 0.15 mg/cm2, the thickness of the catalyst layer is ~10 ^m; the MEA with a total thickness ~200 ^m can generate more than half an ampere of current for every square centimeter of the MEA at a voltage of 0.7 V between the cath­ode and the anode [2, 3, 10-12]. Recently, scientists at Los Alamos National Laboratory, USA have developed a new class of hydrogen fuel cell catalysts that exhibit promising activity and stability. The cata­lysts, cobalt-polypyrrole-carbon (Co-PPY-XC72) composite, are made of low-cost metals entrapped in a heteroatomic-polymer structure.

The cell hardware. The hardware of the fuel cell consists of backing layers, flow fields, and current collectors. These are designed to maxi­mize the current that can be obtained from an MEA. The backing layers placed next to the electrodes are made of a porous carbon paper or carbon cloth, typically 100-300 ^m thick. The porous nature of the back­ing material ensures effective diffusion of the reactant gases to the cat­alyst. The backing layers also assist in water management during the operation of the fuel cell; too little or too much water can halt the cell operation. The correct backing material allows the right amount of water vapor to reach the MEA and keep the membrane humidified.

Carbon is used for backing layers because it can conduct the electrons leaving the anode and entering the cathode. Apiece of hardware, called a plate, is pressed against the outer surface of each backing layer. The plate serves the dual role of a flow field and current collector. The side of the plate next to the backing layer contains channels machined into the plate. The plates are made of a lightweight, strong, gas-impermeable, electron-conducting material; graphite or metals are commonly used, although composite material plates are now being developed. Electrons produced by the oxidation of hydrogen move through the anode, through the backing layer, and through the plate before they can exit the cell, travel through an external circuit, and reenter the cell at the cathode plate. In a single fuel cell, these two plates are the last of the compo­nents making up the cell.

In a fuel cell stack, current collectors are the bipolar plates; they make up over 90% of the volume and 80% of the mass of a fuel cell stack [11, 15, 16].

Water and air management. Although water is a product of the fuel cell reaction and is carried out of the cell during its operation, it is nec­essary that both the fuel and air entering the fuel cell be humidified. This additional water keeps the polymer electrolyte membrane hydrated. The humidity of the gases has to be carefully controlled, as too little water dries up the membrane and prevents it from conducting the H+ ions and the cell current drops. If the air flow past the cathode is too slow, the air cannot carry all the water produced at the cathode out of the fuel cell, and the cathode “floods.” Cell performance deteri­orates because not enough oxygen is able to penetrate the excess liquid water to reach the cathode catalyst sites. Cooling is required to maintain the temperature of a fuel cell stack at about 80°C, and the product water produced at the cathode at this temperature is both liquid and vapor.

Performance of the PEM fuel cell [3,11,16]. Energy conversion in a fuel cell is given by the relation:

Chemical energy of the fuel = electric energy + heat energy

Power is the rate at which energy (E) is made available (P = dE/dt, or AE = PAt). The power delivered by a cell is the product of the current (I) drawn and the terminal voltage (V) at that current (P = IV watts). In order to compute power delivered by a fuel cell, we have to know the cell voltage and load current. The ideal (maximum) cell voltage (E) for the hydrogen/air fuel cell reaction (H2 + 1/2O2 ^ H2O) at a specific tem­perature and pressure is calculated from the maximum electrical energy

AG

Wel = — AG = nFE or E = —

TP

where AG is the change in Gibbs free energy for the reaction, n is the number of moles of electrons involved in the reaction per mole of H2, and F (Faraday’s constant) = 96,487 C (coulombs = joules/volt). At a constant pressure of 1 atm, the change in Gibbs free energy in the fuel cell process (per mole of H2) is calculated from the reaction temperature (T) and from changes in the reaction enthalpy (H) and entropy (S).

AG = AH — T AS

= -285,800 J — (298 K)(-163.2 J/K) = -237,200 J

For the hydrogen-air fuel cell at 1 atm pressure and 25°C (298 K), the cell voltage is

AG

nF

As temperature rises from room temperature to the PEM fuel cell operating temperature (80°C or 353 K), the change in values of H and S is very small, but T changes by 55°C. Thus the absolute value of AG decreases. Assuming negligible change in the values of H and S,

AG = -285,800 J/mol — (353 K)(163.2 J/mol • K)

= -228,200 J/mol

Therefore,

Thus, for standard pressure of 1 atm, the maximum cell voltage decreases from 1.23 V at 25°C to 1.18 V at 80°C. An additional correc­tion is needed for using air instead of pure oxygen, and also for using humidified air and hydrogen instead of dry gases. This further reduces the maximum voltage from the hydrogen-air fuel cell to 1.16 V at 80°C and 1 atm pressure. With an increase in load current, the actual cell potential is decreased from its no-load potential because of irreversible losses, which are often called polarization or overvoltage (h). These origi­nate primarily from three sources:

■ Activation polarization (hact)

■ Ohmic polarization (hohm)

■ Concentration polarization (hconc)

The polarization losses result in a further decrease in actual cell volt­age (V) from its ideal potential E (V = E — potential drop due to losses). The activation polarization loss is dominant at low current density. This is because electronic barriers have to be overcome prior to current and ion flows. Activation polarization is present when the rate of an electro­chemical reaction at an electrode surface is controlled by sluggish elec­trode kinetics. Therefore, activation polarization is directly related to the rates of electrochemical reactions. In an electrochemical reaction with hact > 50 — 100 mV, activation polarization is described by a semi­empirical equation known as the Tafel equation:

ha“ = ІЗ) ln (t)

where a is the electron transfer coefficient of the reaction at the elec­trode (anode or cathode), and i0 is the exchange current density. The Tafel slope for the PEMFC electrochemical reaction is about 100 mV/decade at room temperature. Thus there is an incentive to develop electrocat­alysts that yield a lower Tafel slope [2, 3, 11, 12].

Ohmic losses occur because of the resistance to the flow of ions in the electrolyte and resistance to the flow of electrons through the electrode materials. Decreasing the electrode separation and enhancing the ionic conductivity of the electrolyte can reduce the ohmic losses. Both the electrolyte and fuel cell electrodes obey Ohm’s law; the ohmic losses can be expressed by the equation: hohm = iR, where i is the current flowing through the cell and R is the total cell resistance, which includes ionic, electronic, and contact resistance. See Figure 9.4.

Due to the consumption of reactants at the electrode by an electro­chemical reaction, the surrounding material is unable to maintain the

initial concentration of the bulk fluid and a concentration gradient is formed, resulting in a loss of electrode potential. Although several processes contribute to concentration polarization, at practical current densities, slow transport of reactants and products to and from the elec­trochemical reaction site is a major contributor to concentration polar­ization. The effect of polarization is to shift the potential of the electrode: For the anode,

and for the cathode,

kathode Ecathode I hcathode I

The net result of current flow in a fuel cell is to increase the anode potential and to decrease the cathode potential. This reduces the cell voltage. The cell voltage includes the contribution of the anode and cathode potentials and ohmic polarization. See Figure 9.5.

Fcell kathode Fanode iR; or

Fcell Ecathode ^cathode 1 (Eanode ^ |^anode 1) iR; or

Fcell Ecell |Ecathode 1 |^anode 1

where Ecell Ecathode Eanode

The goal of fuel cell developers is to minimize the polarization losses so that the Fcell approaches the Ecell by modifications to the fuel cell

Figure 9.5 PEM fuel cell voltage versus current density curve [3].

design by improvement in the electrode structures, better electrocata­lysts, more conductive electrolytes, thinner cell components, and so forth. It is possible to improve the cell performance by modifying the operat­ing conditions such as higher gas pressure, higher temperature, and a change in gas composition to lower the gas impurity concentration [3].