Energy & Fuels 1998, 12, 41-48
41
CO2-Free Production of Hydrogen by Catalytic Pyrolysis of Hydrocarbon Fuel Nazim Z. Muradov Florida Solar Energy Center, 1679 Clearlake Road, Cocoa, Florida 32922 Received July 10, 1997. Revised Manuscript Received September 9, 1997X
All conventional options of hydrogen production from hydrocarbon fuel (primarily natural gas, NG), e.g., steam reforming (SR), partial oxidation, and autothermal reforming, involve CO2 production at some point in the technological chain of the process. Therefore, the main problem remains: how to produce hydrogen from hydrocarbon fuels without CO2 emission. The capture of CO2 from the SR process streams and its sequestration (underground or ocean disposal) are actively discussed in the literature. However, this method is energy intensive, poses uncertain ecological consequences, and still does not completely eliminate CO2 emission. Another approach is to decompose hydrocarbon fuels into hydrogen and carbon. The thermal decomposition of NG is a technologically simple one-step process without energy and material intensive gas separation stages and shows the potential to be a CO2-free hydrogen production process. The experimental results of the thermocatalytic decomposition of gaseous (methane and propane) and liquid (hexane, gasoline and diesel fuel) hydrocarbons over metal-oxide and carbon-based catalysts are presented. Although transition metal catalysts produce gas with high initial hydrogen concentration, their activity rapidly drops because of the surface deposition of carbon. Carbon-based catalysts offer certain advantages over metal catalysts, since there is no need for the carbon separation from the catalyst. Catalytic pyrolysis of light liquid hydrocarbons, including gasoline, over activated carbon yields hydrogen-rich gas (40-50 vol % H2). The decomposition of methane in binary mixtures with saturated and unsaturated hydrocarbons was also studied. It was found that the addition of small amounts of unsaturated hydrocarbons (e.g., acetylene) to methane noticeably increases the steady-state concentration of hydrogen in the effluent gas. The studied process can be the basis for the development of compact catalytic units for on-site production of hydrogen/ methane blends from NG and liquid hydrocarbon fuels at gas refueling stations. The concept can also be used for a CO2-free production of hydrogen for fuel cell applications.
Introduction It is generally accepted that in the near-to-medium term future hydrogen production will continue to rely on fossil fuels, primarily natural gas (NG). The advanced processes of hydrogen production (high-temperature water electrolysis, thermochemical and hybrid water-splitting cycles, photoelectrochemical and photobiological decomposition of water, solar/photovoltaic water electrolysis, etc.) do not yield significant reduction in hydrogen costs in the near future (10-20 years). For decades, steam reforming (SR) of NG has been the most efficient and widely used process for the production of hydrogen. Other conventional processes for hydrogen production from fossil fuels (partial oxidation of residual oil and coal gasification) are significantly more expensive than SR.1 The process basically represents a catalytic conversion of methane (a major component of the hydrocarbon feedstock) and water (steam) to hydrogen and carbon oxides and consists of three main steps: (a) synthesis gas generation, (b) water-gas shift reaction, and (c) gas purification (CO2 removal). Four Abstract published in Advance ACS Abstracts, November 1, 1997. (1) Steinberg, M.; Cheng, H. Proceedings of the 7th World Hydrogen Energy Conference, Moscow, 1988; Pergamon Press: New York; p 699. X
moles of hydrogen are produced in the reaction with half of it coming from the methane and the other half from water. The energy requirement per mole of hydrogen produced for the overall process is equal to 163/4 ) 40.75 kJ/mol H2. To ensure a maximum conversion of CH4 into the products, the process generally employs a steam/carbon ratio of 3-5, a process temperature of 800-900 °C, and a pressure of 35 atm.2 The thermal efficiency of the SR process is seldom greater than 50%.2 A steam reformer fuel usage is a significant part (up to 30-40%) of the total NG usage of a typical hydrogen plant. There is no byproduct credit (except for steam) for the process, and in the final analysis, it does not look environmentally benign because of large CO2 emissions. The total CO2 emission from the SR process reaches up to 0.35-0.42 m3 per each m3 of hydrogen produced. The perspectives of CO2 capture and sequestration (ocean or underground disposal) are actively discussed in the literature.3-5 According to ref 5, the capture and (2) Cromatry, B. Proceedings of the 9th World Hydrogen Energy Conference; M.C.I.: Paris, 1992; pp 13-22. (3) Nakicenovic, N. Energy Gases: The Methane Age and Beyond. IIASA, Working Paper-93-033; IIASA: Laxenburg, Austria, 1993; pp 1-23. (4) Blok, K.; Williams, R.; Katofsky, R.; Hendriks, C. Energy 1997, 22, 161-168.
S0887-0624(97)00114-X CCC: $15.00 © 1998 American Chemical Society Published on Web 01/12/1998
42 Energy & Fuels, Vol. 12, No. 1, 1998
disposal of CO2 (80-85% of CO2 captured from the concentrated streams of the SR process) add about 2530% to the cost of hydrogen produced by the SR of NG. It should be noted, however, that the capture, transportation, and sequestration of CO2 are energy intensive processes. Thus, according to ref 4, the electric energy consumption to pressurize CO2 to 80 bar (five-stage compression for a pipeline transportation) is equal to 281 kJe/kg CO2. The total of the electricity requirement for CO2 transportation (100-500 km) to the underground disposal site and its injection is estimated at approximately 2000 kJe/kg CO2. World average for CO2 emission associated with the electricity production is 0.153 kg of CO2 per each kW h produced.4 Thus, the amount of CO2 produced as a result of the capture of CO2 from the concentrated streams (after pressure swing adsorption, PSA, unit) of the SR process and its sequestration reaches up to 0.09-0.1 kg per kg of sequestrated CO2. In principle, CO2 can be captured also from the stack gases of the hydrogen plant (where it is presented in a highly diluted form) and sequestrated; however, the energy cost of this operation would be very high. For example, according to ref 6, the energy consumption associated with CO2 recovery from the stack gases by hot K2CO3 solutions amounts to 3000 kJ/kg CO2. In consequence, the total CO2 emissions from CO2 capture, transportation, and underground disposal could easily reach 0.25 kg CO2 per kg of sequestrated CO2. Thus, CO2 sequestration is an energy intensive process and, in the final analysis, does not completely eliminate CO2 emission. In addition to this problem, some uncertainties remain regarding the duration and extent of CO2 retention (underground or under the ocean) and its possible environmental effect. Thermal decomposition (TD) is another strategy for a CO2-free production of hydrogen from NG
CH4 f C + 2H2 + 75.6 kJ The energy requirement per mole of hydrogen produced (75.6/2 ) 37.8 kJ/mol H2) is somewhat less than that for the SR process. The process is slightly endothermic so that less than 10% of the heat of methane combustion is needed to drive the process. In addition to hydrogen as a major product, the process produces a very important byproduct: clean carbon. The process is environmentally compatible, since it produces relatively small amounts of CO2 (approximately 0.05 m3 per m3 of H2 produced, if CH4 is used as a fuel). It should be noted, however, that the process could potentially be completely CO2-free if a relatively small part of hydrogen produced (approximately 14%) is used as a process fuel. The decomposition of methane can also be carried out plasmochemically. In a paper,7 the authors advocated a plasma-assisted decomposition of methane into hydrogen and carbon. It was estimated that up to 1.9 kW h of electrical energy is consumed per one normal cubic meter of hydrogen produced. Since almost 80% of the (5) Audus, H.; Kaarstad, O.; Kowal, M. Proceedings of 11th World Hydrogen Energy Conference, Stuttgart; Scho¨n & Wetzel GmBH: Frankfurt am Main, 1996; pp 525-534. (6) Gritsenko, A. Cleaning of Gases from Sulfurous Compounds; Nedra: Moscow, 1985. (7) Fulcheri, L.; Schwob, Y. Int. J. Hydrogen Energy 1995, 20, 197202.
Muradov
Figure 1. Comparative assessment of different hydrogen production processes: (1) SR; (2) SR with CO2 sequestration (after PSA); (3) electrolysis; (4) PD; (5) TD (CH4-fuel); 6- TD (H2-fuel).
total world energy supply is based on fossil fuels,3 one can expect the electricity-driven hydrogen production processes to be among the major CO2 producers. A comparative assessment of the hydrogen production by SR, without and with CO2 (after PSA unit) sequestration, electrolysis, plasmochemical decomposition (PD), and TD (with CH4 and H2 as a process fuel options) of NG is depicted in Figure 1. The comparison is based on two very important parameters, which reflect the energetic and ecological features of the processes. The first parameter (E1) is equal to the total volume of NG consumed (both as a feedstock and a fuel) for the production of a unit volume of H2 (E1 ) NG/H2, m3/m3). The second parameter (E2) is equal to the total volume of CO2 produced from both the feedstock and fuel usage of NG per a unit volume of H2 produced (E2 ) CO2/H2, m3/m3). Evidently, the lesser are both E1 and E2 parameters, the better is the hydrogen production process. For the sake of simplicity and comparability, it was assumed that NG was the primary fuel (at the power plant) for the water electrolysis and PD of NG. In fact, this assumption leads to a rather conservative value for the E2 parameter, since the NG share of the total energy supply is only 19% and, what is more, NG produces 1.9 and 1.7 times less CO2 (per kW h produced) than oil and coal, respectively.3 The following conclusions can be extracted from Figure 1. (1) The processes with a large consumption of electric energy (water electrolysis, PD of NG) are characterized by the highest NG consumption and CO2 emission per unit of hydrogen produced. It should be noted, however, that this conclusion is based on the world average energy production scenario. Therefore, in countries with a large nonfossil fuel energy sector (hydroelectric, nuclear energy) both E1 and E2 parameters could be much lower. (2) SR with CO2 capture (after PSA unit) and sequestration produces 30% less CO2 emission than SR without CO2 sequestration. (3) SR with CO2 sequestration consumes 33% less NG than the TD process; however, it produces 5 times more CO2 emission.
CO2-Free Production of H2
(4) TD of NG is the only fossil fuel-based process that shows a real potential to be a completely CO2-free hydrogen production process. TD of NG is a technologically simple one-step process without energy and material intensive gas separation stages, while SR is a multistep and complex process. The technoeconomic assessment showed that the cost of hydrogen produced by TD of NG ($58/1000 m3 H2, with carbon credit) is somewhat lower than that for the SR process ($67/1000 m3 H2).1 Currently, the total world production of carbon black is close to 6 million tons per year, with prices varying in the range of hundreds to thousands of dollars per ton, depending on the carbon quality.7 Carbon black has a great market potential both in traditional (rubber industry, plastics, inks, etc.) and in new areas. For example, authors8 identified the metallurgical industry as a very promising market for carbon black. Carbon black is particularly valuable as a reducing reagent for the production of SiC and as a carbon additive/carburizer in steel industry. According to ref 8, the carbon black market for these applications in Europe currently approaches 0.5 million tons/year with the price level for the high-quality materials reaching $615 per ton. An interesting approach to the CO2 mitigation problem, which involves a methane decomposition reaction coupled with a methanol synthesis (Carnol process), was discussed in a recent publication.9 In this process, CO2 recovered from a coalfired power plant reacts with hydrogen produced by the methane decomposition. The resulting methanol is used as an automotive fuel with overall significant reduction in CO2 emission compared to that of the conventional energy systems. TD of NG is a well-known process. It has long been used for the production of carbon black with hydrogen being used as a supplemental gaseous fuel. The process was practiced batchwise using firebrick contact and required elevated temperatures (up to 1500 °C). Attempts have been made to use catalysts to reduce the maximum temperature of the decomposition of methane and light hydrocarbons. For example,10 the authors used alumina, silica-alumina, silica-magnesia, and other contacts at 800-1000 °C. The data on the catalytic decomposition of methane using Co, Cr, Ni, Fe, Pt, Pd, and Rh-based catalysts have also been reported in the literature.11,12 In all cases, carbon produced was burned off the catalyst surface to regenerate its initial activity. In this regard, these processes display no significant advantages over the conventional processes (for example, SR) because of large CO2 emissions. It was reported recently on the production of hydrogen and novel carbonous materials by the catalytic pyrolysis of methane over Ni and Ni-Cu catalysts.13 We studied (8) Gaudernack, B.; Lynum, S. Proceedings of 11th World Hydrogen Energy Conference, Stuttgart; Scho¨n & Wetzel GmBH: Frankfurt am Main, 1996; pp 510-523. (9) Steinberg, M. Proceedings of 11th World Hydrogen Energy Conference, Stuttgart; Scho¨n & Wetzel GmBH: Frankfurt am Main, 1996; pp 499-510. (10) Pohleny, J.; Scott, N. U.S. Patent No 3,284,161 (UOP), 1966. (11) Callahan, M. Proceedings of 26th Power Sources Symposium; PSC Publishers: Red Bank, NJ, 1974; p 181. (12) Pourier, M.; Sapundzhiev, C. Int. J. Hydrogen Energy 1997, 22, 429-433. (13) Parmon, V.; Kuvshinov, G.; Sobyanin, V. Proceedings of 11th World Hydrogen Energy Conference, Stuttgart; Scho¨n & Wetzel GmBH: Frankfurt am Main, 1996; pp 2439-2448.
Energy & Fuels, Vol. 12, No. 1, 1998 43
Figure 2. Maximum concentrations of H2 in H2-CH4 mixtures produced by methane decomposition over Ni- and Febased catalysts.
the thermocatalytic decomposition of methane over various catalysts and supports including Pt, Mo, Fe, Ni, and Co catalysts using pulse and continuous flow reactors.14-16 Results and Discussion Methane Decomposition over Metal-Oxide Catalysts. We determined the catalytic activity of Ni and Fe catalysts for methane decomposition reaction over a wide range of temperatures, 200-900 °C, at the atmospheric pressure using multisectional fixed-bed microreactors. It was found that Ni/alumina and Fe/alumina catalysts exhibited extremely high initial activity in the methane decomposition reaction. For example, in the presence of freshly reduced Ni- and Fe-catalysts the hydrogen was detected in the effluent gas at a temperature as low as 200 °C. Figure 2 depicts the maximum hydrogen concentrations detected in the effluent gases of methane decomposition process using Ni- and Fecatalysts over a wide range of temperatures. It is apparent that hydrogen concentrations in H2-CH4 mixtures are close to the equilibrium values, which is an indication of very active catalysts. Hereafter, we will use hydrogen concentration values (in vol %) on the “Y” axis of the graphs, which can be easily changed to methane conversion values via the following formula:
Kmethane )
[H2]
100
(200 - [H2])
where Kmethane is methane conversion (%) and [H2] is hydrogen concentration in the effluent gas (% by volume). Figure 3 depicts the typical kinetic curve of hydrogen production over alumina-supported Fe2O3 catalyst at 850 °C. Evidently, the process can be broken down into (14) Muradov, N. Energy and Environmental Progress-1; Veziroglu, N., Ed.; Nova Science: New York, 1991; pp 93-103. (15) Muradov, N. Int. J. Hydrogen Energy 1993, 18, 211-215. (16) Muradov, N. Proceedings of 11th World Hydrogen Energy Conference, Stuttgart; Scho¨n & Wetzel GmBH: Frankfurt am Main, 1996; pp 697-702.
44 Energy & Fuels, Vol. 12, No. 1, 1998
Figure 3. Thermocatalytic decomposition of methane over Fe2O3 (10 wt %)/Al2O3 catalyst at 850 °C.
four zones. Zones A and B correspond to the reduction of iron oxide to the metallic form (accompanied by the formation of carbon oxides) and Fe-catalyzed decomposition of methane, respectively. This was followed by the depletion of the catalytically active metal and carbide phase (zone C) and, finally, the steady-state methane decomposition (zone D). It should be noted that we did not observe an induction period when Feor Ni-oxide catalysts were prereduced by hydrogen to the metallic state prior to the methane decomposition reaction. The shape of the kinetic curves of methane decomposition over alumina-supported NiO catalyst was similar to that of the Fe catalyst, except the induction period (reduction to the metallic state) (zone A) was shorter and the time interval for the depletion of the catalytically active phase (zone C) was longer. We found it impossible to effectively restore the original catalytic activity of the deactivated catalyst through any mechanical or chemical method other than burning carbon off the catalyst surface. Methane Decomposition over Carbon-Based Catalysts. The use of carbon-based catalysts offers significant advantages over metal catalysts, since there is no need for the separation of carbon from the catalyst surface; the carbon produced builds up on the surface of the original carbon catalyst and can be continuously removed from the reactor (for example, using a fluidized bed reactor). There is a lack of information in the literature on the catalytic properties of various forms and modifications of carbon in the methane decomposition reaction. A simple experiment demonstrates that the carbon produced by the thermal decomposition of methane catalyzes the process. Figure 4 depicts the kinetic curve of methane decomposition over activated alumina at 850 °C in a fixed bed reactor. Alumina exhibited extremely low catalytic activity toward methane decomposition reaction, which explains the relatively long induction period (20 min). As carbon formed on the alumina surface, the rate of methane decomposition increased. At this stage of the process the shape of the kinetic curve was typical of an autocatalytic reaction. Over 2-2.5 h the rate of hydrogen production reached the maximum followed by a slow (more than 10 h) decrease in hydrogen concentration in the effluent gas and, finally, steady-state decomposition of methane.
Muradov
Figure 4. Thermal decomposition of methane over activated alumina at 850 °C.
This experiment clearly demonstrated the catalytic properties of carbon produced in the course of the methane decomposition reaction. However, the catalytic activity of carbon was quite low and obviously much less than that of Ni- and Fe- catalysts. These experimental results can be explained as follows. It is known that the initial rate of carbon production by the thermal decomposition of hydrocarbons depends on the nature of a support surface.17 As a surface is covered with carbon species, the rate of methane decomposition may increase or decrease, depending on the catalytic properties of the surface. With freshly reduced Ni- and Fecatalysts we always observed the decrease (after initial maximum) in the rate of methane decomposition reaction due to the blockage of the catalytically active metal carbide sites by carbon on the catalyst surface. Materials with a catalytically inert surface (like alumina) demonstrate quite different behavior. The total rate of the methane decomposition process is the sum of the rate of carbon nuclei formation and the rate of carbon crystallites growth. Interestingly, the activation energy of carbon nuclei formation (317 kJ/mol) is much higher than the activation energy of crystallite growth (227 kJ/ mol).17 This explains the relatively long induction period of hydrogen production over the alumina surface. The boost in hydrogen production rate after the induction period can be explained by the increase in both the concentration of carbon nuclei on the alumina surface and methane decomposition rate over relatively small carbon crystallites. This was followed by the growth of the existing carbon crystallites and, as a result, the reduction of the active surface area and gradual decrease in methane decomposition rate. We determined the catalytic activity of various carbon materials (graphite, carbon black, activated carbon, etc.) for methane decomposition reaction over the range of temperatures from 700 to 900 °C. Figure 5 depicts the experimental results of methane decomposition reaction in the presence of several carbon materials in a multisectional fixed bed reactor at 850 °C. The carbon black (Vulcan XC 72) and graphite demonstrated the lowest activity for the methane decomposition reaction. The (17) Tesner, A. The Kinetics of Carbon Black Production; VINITI: Moscow, 1987; pp 1-64 (in Russian).
CO2-Free Production of H2
Figure 5. Thermocatalytic decomposition of methane over different forms of carbon at 850 °C.
activated carbon produced from coconut shells displayed the highest initial activity among other forms of carbon, producing gas with hydrogen concentrations up to 7075 vol %. This, however, was followed by a dramatic drop in the catalytic activity and relatively quick (3 h) achievement of a steady-state production of gas with a very low hydrogen concentration. Another modification of the activated carbon (AX-21) demonstrated astoundingly different behavior compared to the coconutproduced activated carbon. The initial maximum of hydrogen production yield ([H2]max ) 48 vol %) was followed by a drop in hydrogen concentration to half its initial value, and then it stayed almost flat for several hours (to be exact, every hour decreasing by 0.5 vol %). In all cases, there were no methane decomposition products other than hydrogen and carbon and traces of ethane and ethylene detected in the effluent gas. The amount of carbon formed corresponded to the volume of hydrogen produced within the margin of error (5%). Poor performance of the graphite and carbon black catalysts can be explained by the structure and size of carbon crystallites; the closer the crystallites are to the graphite type structure, the lesser is their catalytic activity.17 The difference in behavior of two types of activated carbon in methane decomposition reaction is yet to be understood. Methane Decomposition in Binary Mixtures with Other Hydrocarbons. We conducted a series of experiments on methane decomposition in binary mixtures with saturated and unsaturated hydrocarbons over inert supports (activated alumina and quartz wool). The objective was to find out if carbon produced by the decomposition of other than methane hydrocarbon exhibits higher catalytic activity in methane decomposition reaction than the carbon formed from methane. Figure 6 depicts the results of the thermal decomposition of neat propane and methane (70 vol %)/propane binary mixtures over activated alumina at 850 °C. It can be seen that, unlike methane decomposition, propane decomposition reaction over alumina does not exhibit a noticeable induction period. The hydrogen production rate dropped over a period of approximately 1 h and reached the steady-state rate. The addition of propane to methane feedstock did not result in the desirable increase in a sustainable methane decomposition rate. Hydrogen concentration in the effluent gas
Energy & Fuels, Vol. 12, No. 1, 1998 45
Figure 6. Thermal decomposition of propane and methane (70 vol. %)/propane mixture over activated alumina at 850 °C.
Figure 7. Methane decomposition in the presence of acetylene at 850 °C. Support is quartz wool.
gradually decreased over almost 3 h until it reached the steady-state value. This result implies that the carbon crystallites produced from the same family of saturated hydrocarbons (alkanes) most probably are identical in size and structure and, therefore, display similar activity in the methane decomposition reaction. We observed quite different kinetics of methane decomposition when a small amount of acetylene was added to methane. Figure 7 depicts the kinetic curves of methane decomposition in the presence of 7 vol % of acetylene over quartz wool at 850 °C. Filled circles correspond to the steady-state decomposition of methane over quartz wool (with carbon deposits produced from methane). At some moment, acetylene was added to methane, which resulted in a steady-state production of gas (open circles) with hydrogen concentration significantly higher than that with neat methane. The data on acetylene (7 vol % in nitrogen) decomposition (triangles), which were collected after CH4/C2H2 decomposition, are also shown in Figure 7. The difference between the CH4/C2H2 decomposition curve and the arithmetical sum of CH4 and C2H2/N2 decomposition curves (dotted line) is significantly higher than the margin of error. Therefore, it can be concluded that the observed increase in hydrogen concentration is due to the catalytic effect of the carbon produced from acety-
46 Energy & Fuels, Vol. 12, No. 1, 1998
Figure 8. Thermocatalytic decomposition of methane over carbon produced from methane and acetylene at 850 °C. Support is activated alumina.
lene on the methane decomposition reaction. It should be noted that at 850 °C acetylene is less stable molecule than methane. The activation energies of the decomposition reaction for acetylene and methane are equal to 126 and 272 kJ/mol, respectively.17 The results of experiments with methane-acetylene binary mixtures clearly demonstrate that the carbon produced from acetylene is catalytically more active toward methane decomposition than one produced from methane. The following series of experiments supports this point of view. Figure 8 depicts the results of two consecutive experiments. The left curve (open circles) corresponds to the steady-state decomposition of methane over carbon produced from methane at 850 °C (support alumina). The right curve (filled circles) corresponds to methane decomposition over carbon produced from neat acetylene at 850 °C (support alumina). It can be seen (the right curve) that the hydrogen concentration in the effluent gas was highest at the beginning; however, over period of 1 h, hydrogen concentration dropped and approached the steady-state value (which was very close to that of the left curve). These experimental results can be explained in terms of the difference in the catalytic activity of carbon species produced from methane and acetylene. Owing to the smaller size of carbon crystallites produced from acetylene (5 nm, against 80 nm obtained from methane),17 their catalytic activity is higher than that of the methane-generated carbon crystallites (of course, a possible difference in the carbon crystallite structures cannot be ruled out). Therefore, at the beginning, methane decomposition over catalytically more active acetylene-produced carbon results in a higher hydrogen yield compared to the methane-produced carbon. As the larger crystallites of carbon are produced during the methane decomposition process, the reaction rate decreases and gradually reaches the steady-state value, typical of the methane-produced carbon. On the other hand, the decomposition of acetylene over carbon produced from methane yields a gas with hydrogen concentration lower than the steady-state concentration (Figure 9). As the smaller crystallites of carbon are formed from acetylene, the reaction rate increases and gradually reaches the steady-state value. It can be concluded from the experimental data discussed in this section that the addition of small amounts of unsatu-
Muradov
Figure 9. Thermocatalytic decomposition of acetylene over carbon produced from methane and acetylene at 850 °C. Support is activated alumina.
Figure 10. Catalytic pyrolysis of hexane over Fe2O3 (10 wt %)/Al2O3 catalyst at 650 °C.
rated hydrocarbons (e.g., acetylene) can noticeably increase the steady-state methane decomposition rate. Efforts are underway to determine the effect of other hydrocarbons (olefines, aromatics, etc.). Catalytic Pyrolysis of Liquid Hydrocarbons. From the thermodynamic point of view the decomposition (pyrolysis) of liquid hydrocarbons is more favorable than the decomposition of methane, since almost 1.5-2 times less energy is required to produce a unit volume of hydrogen. We conducted a series of experiments on the catalytic pyrolysis of a wide range of liquid hydrocarbons (pentane, hexane, octane, gasoline, and diesel fuel) using different metal-oxide and carbon-based catalysts. Figure 10 depicts the experimental results of the catalytic pyrolysis of hexane over aluminasupported iron-oxide catalyst at 650 °C. As was observed with methane, there was an induction period corresponding to the reduction of the iron oxide into the metallic form accompanied by the production of carbon oxides. After an induction period, there was a relatively short (20-30 min) period of hydrogen-rich (90 vol %) gas production followed by the decline in hydrogen concentration in the effluent gas and a steady-state gas production. The activated carbon (coconut) yielded quite a different distribution of products (Figure 11). In both
CO2-Free Production of H2
Figure 11. Catalytic pyrolysis of hexane over activated carbon (coconut) at 650 °C.
Energy & Fuels, Vol. 12, No. 1, 1998 47
Figure 13. Catalytic pyrolysis of diesel fuel over activated carbon (coconut) at 750 °C.
The studied process can be the basis for the development of compact catalytic units for on-site production of hydrogen/methane blends from NG and liquid hydrocarbon fuels at gas refueling stations. The concept can also be used for CO2-free production of hydrogen for fuel cell applications. Experimental Section
Figure 12. Catalytic pyrolysis of gasoline over activated carbon (coconut) at 750 °C.
cases, the steady-state production of the pyrolysis products was achieved over a period of 1.5-2 h. After 1-1.5 h we observed the production of small amounts of the dark liquid products. For example, after 2 h the yields of liquid products were 8 and 10 wt % (of the total amount of the hydrocarbon introduced) for Fe2O3/ alumina and activated carbon, respectively. Figure 12 depicts the results of gasoline pyrolysis over activated carbon (coconut) at 750 °C. In this experiment the effluent gas from the first reactor was directed into the second reactor (with activated carbon) operating also at 750 °C. This modification resulted in the production of gas consisting mainly of hydrogen and methane with a relatively small fraction of C2+ hydrocarbons (
