Development and electrochemical studies of gas diffusion electrodes for polymer electrolyte fuel cells

Electrochemical studies on low catalyst loading gas diffusion electrodes for polymer electrolyte fuel cells are reported. The best performance is obtained with an electrode formed from 20 wt% Pt/C, 0.4 mg Pt cm^–2 and 1.1 mg Nafion^® cm^–2 in the catalyst layer and 15% PTFE in a diffusion layer of 50 µm thickness, for both the cathode and the anode. However, it is also observed that the platinum requirement can be diminished to values close to 0.2 mg Pt cm^–2 in the cathode and 0.1 mg pt cm^–2 in the anode, without appreciably affecting the good characteristics of the fuel cell response. The experimental fuel cell data were analysed using theoretical models of the electrode structure and of the fuel cell system. It is seen that most of the electrode systems present limiting currents and some also show linear diffusion components arising from diffusion limitations in the gas channels and/or in the thin film of electrolyte covering the catalyst particles.     

Numerical study of shock initiation threshold of detonation burning in a monofuel gas suspension

Some results of numerical simulation of shock initiation of plane heterogeneous detonation waves in homogeneous monodisperse gas suspensions of a monofuel are reported. The influence of the initial size of monofuel particles on the critical (minimum) Mach number of the initiating shock wave is studied. The dependence of the critical (maximum) size of fuel particles on the initial relative mass content of the reacting dispersed phase is analyzed.Translated from Fizika Goreniya i Vzryva, Vol. 32, No. 4, pp. 110–112, July–August, 1996.     

Electrocatalytic oxidation of methanol and ethanol at carbon ceramic electrode modified with platinum nanoparticles

The use of carbon ceramic electrode (CCE) modified with platinum particles was studied for the electrocatalytic oxidation of methanol and ethanol by cyclic voltammetry and chronoamperometry. After preparation of a carbon ceramic as an electrode matrix by sol–gel technique, its surface was potentiostatically coated with Pt nanoparticles at −0.2 V vs. SCE in an aqueous solution of 0.1 M H₂SO₄ containing 0.002 M H₂PtCl₆. The electrocatalyst was characterized by XRD, SEM and cyclic voltammetry. The effective parameters on electrocatalytic oxidation of the alcohols, i.e. the amount Pt loadings, medium temperature and working potential limit in anodic direction were investigated and the results were discussed. This modified electrode showed an enhanced current density over the other Pt-modified electrodes making it more attractive for fuel cell applications.     

Co–SiO2 aerogel-coated catalytic walls for the generation of hydrogen

Cobalt–silica (Co–SiO₂) aerogel coatings were successfully grown on the walls of ceramic monolith channels, thus resulting in structured catalytic wall materials useful for gas–solid reactions. The preparation involved the synthesis in situ of SiO₂ gels by the sol–gel hydrolysis and condensation of tetraethylorthosilicate (TEOS), quenching at the gelation point, incorporation of Co by impregnation, and extraction of the solvent by supercritical drying. Characterization of the aerogel-coated monoliths revealed an excellent dispersion and homogeneity of the aerogel and good adherence properties. The catalytic performance of these materials in the ethanol steam reforming reaction addressed to obtain hydrogen was studied at atmospheric pressure by carrying out consecutive cycles at 473–773–473 K with a C₂H₅OH:H₂O  1:3 (molar) mixture and stability tests, and the results were compared with those obtained over monoliths prepared by conventional washcoating methods from Co–SiO₂ xerogel. Co–SiO₂ aerogel catalytic walls were about four times more active for hydrogen generation at 623 K than conventional monoliths. An unusual rapid activation of aerogel-coated monoliths was attained at 580–590 K, even after several cycles and oxidation treatments at 563–613 K, which was attributed to highly dispersed cobalt particles and higher effective diffusivity of reaction species due to high porosity and larger average pore size. The reproducible low-temperature activation of Co–SiO₂ aerogels supported on ceramic monoliths may be useful for the practical application of fuel reformers to the on-site generation of hydrogen from ethanol.     

Infrared spectroscopy studies of cyclohexene hydrogenation and dehydrogenation catalyzed by platinum nanoparticles supported on mesoporous silicate (SBA-15). Part 1: The role of particle size of Pt nanocrystals supported on SBA-15 silicate

Platinum nanoparticles were prepared from colloidal solution in the 5–16 nm range. SBA-15 mesoporous silica was impregnated with particles small enough to enter the 10 nm pores until 0.1 wt% loading was reached. Characterization of the catalysts was carried out by XRD, TEM and BET. Cyclohexene hydrogenation/dehydrogenation was monitored using a reaction cell that permitted infrared spectroscopy monitoring of the gas phase as well as the catalyst surface that was pressed in a wafer and inserted in the reactor. The surface hydroxyl groups on the mesoporous silica show shifts in the 3633–3705 cm^–1 range characteristic of the presence of cyclohexene, 1,3- and 1,4-cyclohexadienes. Reaction studies using 10 Torr of cyclohexene and 100 Torr of hydrogen in the 298–473 K range showed that hydrogenation occurs readily at room temperature while dehydrogenation takes place only at higher temperatures as expected. The small platinum nanoparticles carry out reactions at the highest rates while the largest size metal particles of the lowest. There is no apparent change of metal particle size during the reactions.     

High energy ball-milled Pt and Pt–Ru catalysts for polymer electrolyte fuel cells and their tolerance to CO

High energy ball milling, an industrially amenable technique, has been used to produce CO tolerant unsupported Pt–Ru based catalysts for the oxidation of hydrogen in polymer electrolyte fuel cells. Nanocrystalline Pt_0.5–Ru_0.5 alloys are easily obtained by ball-milling but their performances as anode catalysts are poor because nanocrystals composing the material aggregate during milling into larger particles. The result is a low specific area material. Improved specific areas were obtained by milling together Pt, Ru and a metal leacheable after the milling step. The best results were obtained by milling Pt, Ru, and Al in a 1:1:8 atomic ratio. After leaching Al, this catalyst (Pt_0.5–Ru_0.5 (Al₄)) displays a specific area of 38 m²g^–1. Pt_0.5–Ru_0.5 (Al₄) is a composite catalyst. It consists of two components: (i) small crystallites (4 nm) of a Pt–Al solid solution (1–3 Al wt%) of low Ru content, and (ii) larger Ru crystallites. It shows hydrogen oxidation performance and CO tolerance equivalent to those of Pt_0.5–Ru_0.5 Black from Johnson Matthey, the commercial catalyst which was found to be the most CO tolerant one in this study.     

Electrodeposited PtCo and PtMn electrocatalysts for methanol and ethanol electrooxidation of direct alcohol fuel cells

PtCo and PtMn electrocatalyst particles were successfully synthesized on Ti substrate by the electrodepostion method. PtCo particles deposited are star-shaped particles with size of 100–200 nm and very porous with many slices of 10 nm. On the other hand, PtMn particles are spherical and have no obvious conglomeration, and the particle is in the range of 100–200 nm. The results reveal that the effect of the incorporation of Co and Mn on the electrochemical active surface area of Pt nanoaprticles is very small. However, incorporation of trace Co and Mn in Pt (e.g., Pt₁₀₀₀Co and Pt₁₀₀₀Mn) has dramatic effect on the electrochemical oxidation reaction of alcohol. The mass specific peak current for the methanol oxidation in alkaline media is 49 mA cm⁻² and 39 mA cm⁻² on Pt₁₀₀₀₀Mn and Pt₁₀₀₀Co, which is three and two times higher, respectively, than that on pure Pt electrocatalyst nanoparticles. PtMn and PtCo electrocatalysts also show significant enhanced stability for methanol oxidation. However, the electrocatalytic enhancement of Co or Mn to Pt is relatively small for the electrooxidation reactions of ethanol in alkaline media.     

Competitive oxidation of formaldehyde and formate on amorphous copper-palladium-zirconium alloy electrodes in alkaline media

Raney-type Cu–Pd alloy electrodes were prepared from amorphous Cu–Pd–Zr ternary alloys by treatment with aq. HF, and competitive anodic oxidation reactions of HCHO and HCOO^– were studied on these electrodes in alkaline media. The initial HCHO oxidation product was HCOO^– even on Pd or Pd-rich alloy electrodes which should be more active to the HCOO^– oxidation than to HCHO. The product HCOO^– was oxidized only after a large decrease of the HCHO concentration in the electrolyte. The oxidation rate of HCOO^– was considerably lowered by the existence of even a small amount of HCHO, as well as by the introduction of CO. These results suggest that the HCHO electro-oxidation is accompanied by production of a surface contaminant such as adsorbed CO. The optimum nominal Pd atomic fraction in the Cu–Pd alloy electrodes suitable for the steady simultaneous oxidation of HCHO and HCOO^– in mixed solution was shown to be 0.25 and 0.4 in 1.0 M NaOH (M=moldm^–3) and 0.5 M K₂CO₃, respectively.     

Fe loading of a carbon-supported Fe–N electrocatalyst and its effect on the oxygen reduction reaction

Carbon-supported non-noble metal catalysts with Fe as the metal and tripyridyl triazine (TPTZ) as the ligand (Fe-TPTZ/C) were synthesized using a simple chemical method. How the Fe loading in this Fe-TPTZ/C catalyst affected the ORR activity was investigated using several Fe loadings: 0.2, 0.4, 0.7, 2.7, 4.7, 5.8 and 7.8 wt%. The as-prepared catalysts were then heat-treated at 800 °C in an N₂ environment to obtain catalysts of Fe–N/C. Energy dispersive X-ray spectroscopy (EDX) was used to identify the Fe–N/C catalysts. These Fe–N/C catalysts showed significant ORR activity improvement over the as-prepared Fe-TPTZ/C catalysts. The kinetics of the ORR catalyzed by the catalysts with different Fe loadings was studied using the rotating disk electrode technique. It was observed that a 4.7 wt% Fe loading yielded the best catalytic ORR activity. Regarding the overall ORR electron transfer number, it was found that as the catalyst's Fe loading increased, the overall ORR electron transfer number changed from 2.9 to 3.9, suggesting that increasing the Fe loading could alter the ORR mechanism from a 2-electron to a 4-electron transfer dominated process. The Tafel method was also used to obtain one important kinetic parameter: the exchange current density. A fuel cell was assembled using a membrane electrode assembly with 4.7 wt% Fe loaded Fe–N/C as the cathode catalyst, and the cell was tested for both performance and durability, yielding a 1000-h lifetime.     

Corrosion protection of steel in molten Li2CO3–K2CO3 and Na2CO3–K2CO3 mixtures in a hydrogen-containing atmosphere

The electrochemical behaviour of TiN-, TiN–AlN-, Cr- and CrN-coated 316L stainless steel in molten Li₂CO₃–K₂CO₃ and Na₂CO₃–K₂CO₃ melts in a reducing gaseous atmosphere (10% H₂–90% N₂) was studied using voltammetry and scanning electron microscopy combined with energy-dispersed X-ray analysis in the temperature range of 600–730 C. To facilitate the identification of the electrochemical reactions the voltammetric behaviour of stainless steel, titanium, nickel and gold was also investigated. Voltammetric characteristics obtained at AlN–TiN coated electrodes showed no anodic reactions at potentials more negative than that of CO^2– ₃ oxidation. Cr- and CrN-coated electrodes demonstrated a suppressed anodic dissolution after the first steady state voltammetric cycle. The voltammograms obtained for the other electrodes studied displayed the corresponding anodic metal-dissolution waves. TiN, AlN, Cr and CrN coatings seem to be the most promising as corrosion-resistant materials for the anodic compartments of molten carbonate fuel cells.     

Thermal stability and mechanical properties of SiC particulate-reinforced Si–C–N composites after heating at 2000 °C

A SiC particulate-reinforced Si–C–N ceramic composite was fabricated using the precursor impregnation and pyrolysis method, and its thermal and mechanical properties were analyzed. The weight loss of the composite was 5% after a heating at 2100 °C in Ar. The pores of the composite enlarged at and above 1700 °C in Ar due to the decomposition of the Si–C–N matrix. However, the composite retained mechanical properties such as strength and hardness after heating at 1700 °C. 88% of the original strength was remained after heating at 2000 °C for 10 h although the fabrication temperature was 1350 °C. The weight gain of the composite was 3.2% after an oxidation at 1450 °C for 30 min in air. The inner oxidation of the particulate-reinforced composites (PRC) was suppressed above 1400 °C due to the closure of the open pores by SiO₂. Consequently, the composite possessed excellent creep resistance at 1400 °C in air. The SiC/Si–C–N composite is a challenging candidate for the application at high temperature.     

A Highly Active Ag/Alumina Catalytic Converter for Continuous HC-SCR During Lean-Burn Conditions: From Laboratory to Full-Scale Vehicle Tests

A common-rail diesel vehicle was equipped with a full-scale Ag/alumina catalytic converter. The converter consisted of several Ag/alumina bricks, with free space in between each brick to fulfil important gas phase reactions. An oxidation catalyst was placed at the end of the converter to remove formed CO and unburned HC. High conversion levels of NO _x , around 60%, were recorded at several speeds and loads using additional HC (diesel) injection corresponding to 2–5% fuel penalty.     

Nitrogen fixation of and N transfer from cowpea, mungbean and groundnut when intercropped with maize

Grain legumes are used widely in intercropping systems. However, quantitative and comparative data available as to their N₂ fixation and N beneficial effect on the companion crop in intercropping systems are scarce. Hence, studies were conducted to ascertain the above when cowpea (Vigna unguiculata L.), mungbean (Vigna radiata L.) and groundnut (Arachis hypogaea L.) were intercropped with maize. The study was¹⁵N-aided and made outdoors in basins (30 L) filled with 38 kg of soil.¹⁵N labelling was effected by incorporating¹⁵N-tagged plant material or applying¹⁵N-labelled fertilizer along with sucrose to stabilize¹⁵N enrichment in the soil during the experimental period. Intercropped groundnut fixed the highest amount of nitrogen from the atmosphere (i.e. 552 mg plant^–1), deriving 85% of its N from the atmosphere. Intercropped cowpea and mungbean fixed 161 and 197 mg N plant^–1, obtaining 81% and 78% of their N content from the atmosphere, respectively. The proportion of N derived by maize from the associated legume varied from 7-11% for mungbean, 11–20% for cowpea and 12–26% for groundnut which amounted to about 19–22, 29–45 and 33–60 mg N maize plant^–1, respectively. The high nitrogen fixation potential of groundnut in dual stands and its relatively low harvest index for N have apparently contributed to greater N-benefical effect on the associated crop.     

Methane Emission from Rice Fields at Cuttack, India

Methane (CH₄) emission from rice fields at Cuttack (State of Orissa, eastern India) has been recorded using an automatic measurement system (closed chamber method) from 1995–1998. Experiments were laid out to test the impact of water regime, organic amendment, inorganic amendment and rice cultivars. Organic amendments in conjunction with chemical N (urea) effected higher CH₄ flux over that of chemical N alone. Application of Sesbania, Azolla and compost resulted in 132, 65 and 68 kg CH₄ ha^–1 in the wet season of 1996 when pure urea application resulted in 42 kg CH₄ ha^–1. Intermittent irrigation reduced emissions by 15% as compared to continuous flooding in the dry season of 1996. In the wet season of 1995, four cultivars were tested under rainfed conditions resulting in a range of emissions from 20 to 44 kg CH₄ ha^–1. Application of nitrification inhibitor dicyandiamide (DCD) inhibited while Nimin stimulated CH₄ flux from flooded rice compared to that of urea N alone. Wide variation in CH₄ production and oxidation potentials was observed in rice soils tested. Methane oxidation decreased with soil depth, fertilizer-N and nitrification inhibitors while organic amendment stimulated it. The results indicate that CH₄ emission from the representative rainfed ecosystem at the experimental site averaged to 32 kg CH₄ ha^–1 yr^–1.     

A microplot study of the fate of15N-labelled ammonium nitrate and urea applied at two rates to ryegrass in spring

Confined microplots were used to study the fate of¹⁵N-labelled ammonium nitrate and urea when applied to ryegrass in spring at 3 lowland sites (S1, S2 and S3). Urea and differentially and doubly labelled ammonium nitrate were applied at 50 and 100 kg N ha^–1. The % utilization of the¹⁵N-labelled fertilizer was measured in 3 cuts of herbage and in soil to a depth of 15 cm (soil_0–15).Over all rates, forms and sites, the % utilization values for cuts 1, 2, 3 and soil_0–15 were 52.4, 5.3, 2.4 and 16.0% respectively. The % utilization of¹⁵N in herbage varied little as the rate of application increased but the % utilization in the soil_0–15 decreased as the rate of application increased. The total % utilization values in herbage plus soil_0–15 indicated that losses of N increased from 12 to 25 kg N ha^–1 as the rate of N application was increased from 50 to 100 kg N ha^–1.The total % utilization values in herbage plus soil_0–15 over both rates of fertilizer N application were 84.1, 80.8 and 81.0% for urea compared with 74.9, 72.5 and 74.4% for all ammonium nitrate forms at S1, S2 and S3 respectively. Within ammonium nitrate forms, the total % utilization values in herbage plus soil_0–15 over both rates and all sites were 76.7, 69.4 and 75.7% for¹⁵NH₄NO₃, NH₄ ¹⁵NO₃ and¹⁵NH₄ ¹⁵NO₃ respectively. The utilization of the nitrate moiety of ammonium nitrate was lower than the utilization of the ammonium moiety.The distribution of labelled fertilizer between herbage and soil_0–15 varied with soil type. As the total utilization of labelled fertilizer was similar at all sites the cumulative losses due to denitrification and downward movement appeared to account for approximately equal amounts of N at each site.     

Fate of Fertilizer15N applied as urea and ammonium sulphate in opium poppy (Papaver somniferum L.) grown under greenhouse conditions

Laboratory incubation and greenhouse experiments were conducted to investigate the comparative effectiveness of urea and ammonium sulphate in opium poppy (Papaver somniferum L.) using¹⁵N dilution techniques. Fertilizer treatments were control (no N), 600 mg N pot^–1 and 1200 mg N pot^–1 (12 kg oven dry soil) applied as aqueous solution of urea or ammonium sulphate. Fertilizer rates, under laboratory incubation study were similar to that under greenhouse conditions. A fertilizer¹⁵N balance sheet reveals that N recovery by plants was 28–39% with urea and 35–45% with ammonium sulphate. Total recovery of¹⁵N in soil-plant system was 77–82% in urea. The corresponding estimates for ammonium sulphate were 89–91%. Consequently the unaccounted fertilizer N was higher under urea (18–23%) as compared to that in ammonium sulphate (9–11%). The soil pH increased from 8.2 to 9.4 with urea whereas in ammonium sulphate treated soil pH decreased to 7.3 during 30 days after fertilizer application. The rate of NH₃ volatilization, measured under laboratory conditions, was higher with urea as compared to the same level of ammonium sulphate. The changes in pH of soil followed the identical trend both under laboratory and greenhouse conditions.     

Distribution of acid extractable P and exchangeable K in a grassland soil as affected by long-term surface application of N, P and K fertilizers

Information on the fate and distribution of surface-applied fertilizer P and K in soil is needed in order to assess their availability to plants and potential for water contamination. Distribution of extractable P (in 0.03 M NH₄F + 0.03 M H₂SO₄ solution) and exchangeable K (in neutral 1.0 M ammonium acetate solution) in the soil as a result of selected combinations of 30 years (1968–1997) of N fertilization (84–336 kg N ha^–1), 10 years of P fertilization (0–132 kg P ha^–1), and 14 years of K fertilization (0 and 46 kg K ha^–1) was studied in a field experiment on a thin Black Chernozem loam under smooth bromegrass (Bromus inermis Leyss.) at Crossfield, Alberta, Canada. Soil samples were taken at regular intervals in October 1997 from 0–5, 5–10, 10–15, 15–30, 30–60, 60–90 and 90–120 cm layers. Soil pH decreased with N rate and this declined with soil depth. Increase in extractable P concentration in the soil reflected 10 years of P fertilization relative to no P fertilization, even though it had been terminated 20 years prior to soil sampling. The magnitude and depth of increase in extractable P paralleled N and P rates. The extractable P concentration in the 0–5 cm soil layer increased by 2.2, 20.7, 30.4 and 34.5 mg P kg^–1 soil at 84, 168, 280 and 336 kg N ha^–1, respectively. The increase in extractable P concentration in the 0–15 cm soil depth was 1.5 and 12.8 mg P kg^–1 soil with application of 16 and 33 kg P ha^–1 (N rate of 84 N ha^–1 for both treatments), respectively; and it was 81.6 and 155.2 mg P kg^–1 soil with application of 66 and 132 kg P ha^–1 (N rate of 336 N ha^–1 for both treatments), respectively. The increase in extractable P at high N rates was attributed to N-induced soil acidification. Most of the increase in extractable P occurred in the top 10-cm soil layer and almost none was noticed below 30 cm depth. Surface-applied K was able to prevent depletion of exchangeable K from the 0–90 cm soil, which occurred with increased bromegrass production from N fertilization in the absence of K application. As only a small increase of exchangeable K was observed in the 10–30 cm soil, 46 kg K ha^–1 year^–1 was considered necessary to achieve a balance between fertilization and bromegrass uptake for K. The potential for P contamination of surface water may be increased with the high N and P rates, as most of the increase in extractable P occurred near the soil surface.     

Extinction of dispersed heterogeneous systems

We consider extinction of various dispersed systems. Isolated boron particles and boron particles in gases are studied. Stability analysis of steady-state thermal regimes of reacting heterogeneous systems for the case of two parallel reactions on the reaction surface using the Frank-Kamenetskii method gives extinction conditions in oxygen-containing media. Curves of the extinction particle size versus the ambient temperature, oxidizer concentration, and, for particles in gases, also versus the oxidizer-to-fuel ratio are plotted. Approximate analytical calculations showed that the extinction process can be most actively controlled by varying the combustion temperature: a decrease in the latter increases the extinction particle size and decreases the completeness of fuel combustion. It is shown that at low ambient temperatures the extinction particle size for suspensions is larger than that for isolated particles. This effect is caused by a decrease in the oxidizer concentration during combustion of suspensions. At high temperatures, the role of this factor weakens.Translated from Fizika Goreniya i Vzryva, Vol. 32, No. 6, pp. 12–19, November–December, 1996.     

Study of a Li–Ni oxide mixture as a novel cathode for molten carbonate fuel cells by electrochemical impedance spectroscopy

Li–Ni oxide mixtures with high lithium content are considered to be an alternative cathode material for molten carbonate fuel cells (MCFCs). The electrochemical behaviour of Li_0.4Ni_0.6O samples has been investigated in a Li–K carbonate melt at 650 °C by electrochemical impedance spectroscopy as a function of immersion time and O₂ and CO₂ partial pressure. The impedance spectra have been interpreted using a transmission line model that includes contact impedance between reactive particles. The Li_0.4Ni_0.6O powder particles show structural changes due to high lithium leakage and low nickel dissolution from the reactive surface to the electrolyte during the first 100 h of immersion. After this time, the structure seems to be stable. The partial pressures of O₂ and CO₂ affect the processes of oxygen reduction and Li–Ni oxide oxidation. X-ray diffraction and chemical analysis performed on samples before and after the electrochemical tests have confirmed that the lithium content decreases. SEM observations reveal a reduction in grain size after the electrochemical tests.     

Characterization of the Pt-C bond of CO adsorbed by small Pt particles in a NaX zeolite by IR diffuse reflectance spectroscopy

The unusual properties of small (1–2 nm) Pt particles within a zeolite support are documented by diffuse reflectance infrared spectroscopy of the metal-carbon bond of linearly adsorbed CO. The wave number of the metal-carbon stretching frequency is found to be about 60 cm^–1 higher compared to larger particles (2–3 nm) or to the bulk metal. The stronger bond is attributed to a negative charge on the small metal particles.