Recombinative Desorption Of Hydrogen Research Articles

Thermal Rates and High-Temperature Tunneling from Surface Reaction Dynamics and First-Principles.

Studying dynamics of the dissociative adsorption and recombinative desorption of hydrogen on copper surfaces has shaped our atomic-scale understanding of surface chemistry, yet experimentally determining the thermal rates for these processes, which dictate the outcome of catalytic reactions, has been impossible so far. In this work, we determine the thermal rate constants for dissociative adsorption and recombinative desorption of hydrogen on Cu(111) between 200 and 1000 K using data from reaction dynamics experiments. Contrary to current understanding, our findings demonstrate the predominant role of quantum tunneling, even at temperatures as high as 400 K. We also provide precise values for the reaction barrier (0.619 ± 0.020 eV) and adsorption energy (0.348 ± 0.026 eV) for H2 on Cu(111). Remarkably, the thermal rate constants are in excellent agreement with a first-principles quantum rate theory based on a new implementation of ring polymer molecular dynamics for reactions on surfaces, paving the way to discovering better catalysts using reliable and efficient computational methods.

Microstructure and hydrogen storage properties of MgH2/MIL-101(Cr) composite

In this paper, a new hydrogen storage composite based on magnesium hydride and metal-organic framework MIL-101(Cr) (an acronym for Material Institute Lavoisier) has been mechanically synthesized and investigated. The hydrogen sorption and desorption behavior in the composite was investigated for the temperature and pressure ranges of (593−653) K and (0−3) MPa, respectively. According to the pressure-composition-temperature isotherms of hydrogen sorption/desorption, the MgH2–5 wt%MIL-101(Cr) composite starts to absorb hydrogen at a lower pressure at 593 K. In addition, the hydrogen release peak for composite shifts to lower temperatures by about 140 K compared to the pure milled MgH2 during temperature programmed desorption at a heating rate of 6 K/min. The activation energy of hydrogen desorption from the composite is 120 ± 2 kJ/mol, which is 36% lower than the activation energy of hydrogen desorption from magnesium hydride (189 ± 2 kJ/mol). The enthalpy of hydrogen absorption was found to be 73 kJ/mol H2 and 60 kJ/mol H2 for milled MgH2 and MgH2–5 wt%MIL-101(Cr) composites, respectively. It is showed that the MIL-101(Cr) MOF structures decompose during ball milling and the chromium oxide nanoparticles form a core-shell structure with MgH2 particles. Thus, composite has better sorption and desorption properties than pure magnesium/magnesium hydride due to the nanoconfinement of MgH2. The chromium oxide nanoparticles act as active sites on the surface of the magnesium particles and facilitate the dissociative chemisorption or recombinative desorption of hydrogen. The main regularities of the phase transition in the magnesium-hydrogen system for the composite and magnesium hydride during dehydrogenation have been studied for temperatures in the range of 298–750 K. The in situ analysis of the phase transitions during the dehydrogenation of MgH2 and MgH2–5 wt%MIL-101(Cr) composite indicates a pronounced catalytic effect of the addition of MIL-101(Cr) to the Mg/MgH2. Based on the experimental results and ab initio calculations, a mechanism for the sorption and desorption of hydrogen by the composite has been proposed.

First principles rates for surface chemistry employing exact transition state theory: application to recombinative desorption of hydrogen from Cu(111).

We present first principles calculations of the reactive flux for thermal recombinative desorption of hydrogen from Cu(111). We follow a theoretical paradigm used successfully for gas phase reactions, where electronic structure theory (DFT-GGA) is combined with transition state theory (TST). Classical ab initio molecular dynamics trajectories initiated from a thermal distribution near the transition state provide dynamical corrections to the desorption rate. We use this to calculate and study the recrossing error of TST and to directly simulate thermal desorption experiments based on a high temperature permeation method. Transition state recrossing is strongly temperature dependent and is even important in a frozen Cu-atom model. It is not influenced by inclusion of electron-hole pair excitation at the level of the local density electronic friction approximation. We also present the kinetic energy resolved flux of desorbing H2 at elevated temperature. This provides a more direct way to compare first principles theory to experiment, with no need to invoke detailed balance.

Hydrogen flux of BCC and FCC PdCuAg membranes

First-principles calculations have been used to study the effects of Ag addition on adsorption and dissociation of H2 on BCC and FCC PdCu surfaces as well as hydrogen diffusion and recombinative hydrogen desorption through the PdCu membranes. It is found that the Ag addition makes it energetically difficult for the adsorption of H2 on PdCu surfaces and hydrogen diffusion through PdCu, while could help the recombinative desorption of H atoms from both BCC (110) and FCC (111) surfaces of PdCu. Moreover, substitution of Ag for Pd or Cu would impede or improve the dissociation of H2 on PdCu surface. Calculations also reveal that the overall hydrogen flux of BCC Pd8Cu8 (Pd8Cu7Ag) membranes is determined by the recombinative desorption and diffusion when the membrane thickness is smaller and bigger than 10.31 (4.73) μm, respectively. In addition, hydrogen diffusion is the dominant step of hydrogen permeation of FCC PdCu and PdCuAg as well as BCC Pd7Cu8Ag membranes. The present results not only agree well with experimental observations in the literature, but also deepen the understanding of the effect of Ag alloying on hydrogen permeation through PdCu membranes.

Hydrogen production by ammonia decomposition over Cs-modified Co3Mo3N catalysts

The Cs-modified Co3Mo3N catalysts were successfully prepared by a facile single-step decomposition of a mixture containing hexamethylenetetramine (HMTA) and corresponding metal salts under the flow of nitrogen at 700°C and their catalytic activity for ammonia decomposition was investigated. The Co3Mo3N phase was observed in the XRD pattern of as-prepared sample and acted as the active component for the reaction. The XPS analysis clarified the existence of Cs species, indicating that these species should be dispersed over Co3Mo3N. The addition of a small amount of Cs species significantly improved the catalytic performance of the Co3Mo3N catalyst. The further addition of Cs species had a negative effect on the catalytic activity for the reaction. 15NH3 isotopic studies revealed that the nitrogen species in Co3Mo3N were exchangeable with nitrogen atoms in gas phase, elucidating the mechanism of the reaction over the Co3Mo3N catalysts. The kinetics analysis indicated that the modification by a small amount of Cs species alleviated the negative effect of the hydrogen poisoning on the active sites of Co3Mo3N catalysts. The desorption behavior of hydrogen and nitrogen suggested that the Cs modification facilitated the recombinative desorption of hydrogen and nitrogen atoms from the active components, resulting in the improvement in the activity of Co3Mo3N. This improvement was due to the electronic modification of Co3Mo3N by the electron donation of Cs promoter.

Lattice kinetic Monte Carlo simulation of epitaxial growth of silicon thin films in H2/SiH4 chemical vapor deposition systems

The use of lattice kinetic Monte Carlo techniques for reproducing epitaxial growth of silicon using H2/SiH4 chemistry is described in this work. The implementation of the surface mechanisms is limited to silane and hydrogen adsorption, surface dissociation of higher order hydrides and recombinative hydrogen desorption. Good agreement between simulation and experimental results available in the literature was obtained for H2 desorption peak positions, equilibrium hydrogen coverages and Si growth rates. The formation of (111) and (113) planes is also reproduced, however no evidence of (131) was found even at 850° C, and (119) planes were not formed at the same position than actual grown structures.

Determination of the rotational population of H2 and D2 including high-N states in low temperature plasmas via the Fulcher-α transition

Vibrational and rotational excitation of the hydrogen molecule can significantly affect molecular reaction rates in low pressure low temperature plasmas, for example for the creation of H−/D− ions via the dissociative attachment process. In general, the rotational population in these discharges is known to be non-thermal with an overpopulation of states with high rotational quantum number N. In contrast to a sophisticated direct measurement of the rotational distribution in the XΣg+1,v=0 state, it is demonstrated that the determination can also be carried out up to high-N levels rather easily via optical emission spectroscopy utilizing the Fulcher-α transition of H2 and D2. The measured rotational populations can be described with a two-temperature distribution where the cold part reflects the population according to the gas temperature of the discharge. This has been verified by using the emission of the second positive system of nitrogen as independent gas temperature diagnostic. The hot part where the rotational temperature reaches several thousand Kelvin arises most probably from recombinative desorption of hydrogen at the discharge vessel wall where parts of the binding energy are converted into rotational excitation. Neglecting the hot population – what is often done when using the Fulcher-α transition as gas temperature diagnostic – can lead to a strong overestimation of Tgas. No fundamental differences in the rotational distributions between hydrogen and deuterium have been found, only the hot rotational temperature is smaller for D2 indicating an isotope-dependency of the recombinative desorption process.

Quantum Dynamical Simulations of the Femtosecond‐Laser‐Induced Ultrafast Desorption of H2 and D2 from Ru(0001)

We investigate the recombinative desorption of hydrogen and deuterium from a Ru(0001) surface initiated by femtosecond laser pulses. We adopt a quantum mechanical two-state model including three molecular degrees of freedom to describe the dynamics within the desorption induced by electronic transition (DIET) limit. The energy distributions as well as the state-resolved and ensemble properties of the desorbed molecules are analyzed in detail by using the time-energy method. Our results shed light on the experimentally observed 1) large isotopic effects regarding desorption yields and translational energies and 2) the nonequal energy partitioning into internal and translational modes. In particular, it is shown that a single temperature is sufficient to characterize the energy distributions for all degrees of freedom. Further, we confirm that quantization effects play an important role in the determination of the energy partitioning.

Electronic excitation and dynamic promotion of a surface reaction.

The mechanism of recombinative desorption of hydrogen from a Ru(0001) surface induced by femtosecond-laser excitation has been investigated and compared to thermally initiated desorption. For the laser-driven process, it is shown that hot substrate electrons mediate the reaction within a few hundred femtoseconds resulting in a huge isotope effect between H2 and D2 in the desorption yield. In mixed saturation coverages, this ratio crucially depends on the proportions of H and D. Deviations from second order desorption kinetics demonstrate that the recombination is dynamically promoted by excitation of neighboring, but nonreacting adatoms. A concentration dependent rate constant which accounts for the faster excitation of H versus D is proposed.

Dehydrogenation of ethylbenzene and isobutane over Ga- and Fe-containing mesoporous silicas

Incorporation of gallium and iron into MCM-41-type silicas leads to creation of active and selective catalysts for the dehydrogenation of hydrocarbons. This catalysts demonstrate remarkable stability with time on stream and do not suffer from repeated reaction–reoxidation cycles. Ga-MCM-41 is particularly efficient for the dehydrogenation of light alkanes. At 550 °C isobutane is converted into a mixture of C 2C 4 olefins containing 52% of isobutene, which corresponds to 32% of the equilibrium conversion into isobutene at this temperature. Fe-MCM-41 shows the best performance for ethylbenzene dehydrogenation into styrene, in particular, at 600 °C, the yield over this catalyst is 38% per pass and selectivity is 82%, which corresponds to 64% of the equilibrium value. The catalysts have been characterised by XRD, volumetry, infrared spectroscopy, ammonia temperature-programmed desorption (TPD) and temperature-programmed reduction (TPR). Catalytic activity is associated with the presence of nanoparticles of oxides, entrapped in the pores or embedded in the silica network, which generate reducible species able to split CH bonds and allow fast recombinative desorption of hydrogen and olefins.

Spectroscopic evidence for hydrogen diffusion through a several-nanometers-thick titanium carbonitride layer on silicon.

Multiple internal reflection Fourier transform infrared spectroscopy, together with other analytical techniques, was used to follow the diffusion of atomic hydrogen through a 10-nm-thick titanium carbonitride layer deposited onto a Si(100)-2x1 surface from tetrakis(dimethylamino)titanium as a chemical vapor deposition precursor. The recombinative desorption of hydrogen from the TiCN/Si interface was shown to coincide with the temperature range where most Ti-based diffusion barriers break down.

Thermal chemistry of toluene and benzene on Si(1 0 0)2×1 and modified surfaces

The effects of methyl substitution on the room-temperature (RT) adsorption and reactivity of toluene on Si(1 0 0)2×1 have been investigated by thermal desorption spectrometry (TDS), low energy electron diffraction (LEED), and Auger electron spectroscopy (AES). The similarities in the exposure dependence of the molecular desorption profiles of d 8-toluene and those of d 6-benzene suggest that both molecules have similar adsorption configurations on the (2×1) surface. In particular, three molecular desorption features are found and can be attributed to adsorption on the single-dimer, double-dimer, and defect sites. However, the higher coverage and much lower molecular desorption of toluene than those of benzene suggest that the majority of adsorbed toluene has undergone other surface reactions during the TDS experiment. In addition to the weak molecular desorption, strong recombinative hydrogen desorption at 820 K, characteristic of hydrogen coming from mono-hydride (Si–H) surface species, is observed for toluene. The lack of a correspondingly strong D 2 desorption peak for the RT exposure of d 6-benzene to Si(1 0 0)2×1 indicates that the deposition of D atoms results from hydrogen abstraction from the methyl group of d 8-toluene. Since a RT post-exposure of atomic hydrogen to Si(1 0 0)2×1 saturated with d 8-toluene does not appear to increase molecular desorption, the observed hydrogen abstraction therefore occurs upon adsorption and appears to be an irreversible process at RT. The difference in the D 2 desorption profiles between toluene and D 2 on Si(1 0 0)2×1 suggests, in addition to hydrogen desorption, other surface processes including condensation polymerization and/or dissociative desorption that occur at 750–950 K. The Si(1 0 0)2×1 surface is found to be more active in hydrogen abstraction and toluene dissociation than Si(1 1 1)7×7. For a RT exposure of d 8-toluene (d 6-benzene) to a sputtered Si(1 0 0) surface, the presence of an additional broad mass 4 desorption band at ∼600 K suggests the presence of a range of pathways by which D 2 evolves directly from the dissociation of adsorbed d 8-toluene. A RT exposure of O 2 to Si(1 0 0)2×1 pre-adsorbed with d 8-toluene (d 6-benzene) is found to induce a substantial increase in the mass 4 desorption. The increase in the dissociation can be attributed to either surface-mediated oxidation or an O 2-induced process that stabilizes the adsorbed molecules to a higher temperature at which dehydrogenation and dissociation occur.

Distribution of dangling bond pairs on partially hydrogen-terminated Si(100) surface observed by scanning tunneling microscopy

Hydrogen recombinative desorption from a Si 2H 2 monohydride dimer on the Si(100) surface regenerates a dangling bond (DB) pair on the dimer. In this paper, we investigated the spatial distribution of the regenerated DB pairs on the Si(100)-2×1:H and disilane-passivated Si(100) surface using elevated temperature scanning tunneling microscopy (HT-STM) in the temperature range between 590 and 622 K. Experimental results indicate that the ends of one-dimensional (1D) monohydride dimer islands are preferred sites for DB pairs and a repulsive interaction occurs between two neighboring DB pairs in the same dimer row.

Desorption of hydrogen from GaN(0001) observed by HREELS and ELS

The bare and hydrogenated GaN(0001) surfaces were characterized using high resolution electron energy loss spectroscopy (HREELS), electron energy loss spectroscopy (ELS), Auger electron spectroscopy (AES) and low energy electron diffraction (LEED). AES and LEED show that the GaN surface is clean but faceted. Changes in the surface electronic structure are characterized using ELS. Exposure to atomic hydrogen produces a merging of the double peak centered near 20 eV and a leveling of the region from 10 to 15 eV. These hydrogen-induced changes begin to decrease after briefly heating to 260°C and are completely reversed after heating to 380°C. Atomic hydrogen exposure produces HREELS loss peaks at 2580, 3280 and 3980 cm −1 due to combinations of the Ga–H stretching vibration at 1880 cm −1 and Fuchs–Kliewer phonons. Briefly heating to 260°C or above produces a decrease in the intensity of Ga–H vibrations. Heating to 380°C results in a complete disappearance of adsorbate vibrational peaks. Observation of desorbing molecular hydrogen by mass spectrometry confirms that the disappearance of Ga–H features in HREEL and ELS spectra is due to recombinative desorption of hydrogen. Overall this work demonstrates desorption of hydrogen from surface Ga sites between 250 and 450°C.

Thermal desorption of hydrogen from the diamond C(100) surface

Temperature-programmed desorption (TPD) is used to measure the kinetics of hydrogen desorption from the C(100) surface. Two major desorption channels, attributed to hydrogen desorption from (2×1) domains ( α sites) and disordered domain boundaries sites ( β sites) were observed. The change in the morphology of the C(100) surface upon hydrogen adsorption/desorption leads to a variation in the intensity ratio between the two hydrogen desorption peaks. The degradation results in smaller (2×1) domains, as evidenced by deteriorated LEED patterns in which the characteristic (2×1) half-order spots disappear. The (2×1) LEED pattern can be restored on the degraded C(100) surface by repeated hydrogenation–annealing cycles in the temperature range 300–1100 K. Fitting the experimental TPD data with simulated curves from the two-site model shows that desorption from the α site obeys first-order kinetics with a prefactor of 1.4±0.9×10 13 s −1 and an activation energy of 80.3±1.3 kcal mol −1, whereas the process giving rise to the β peak follows second-order kinetics with a prefactor of 2.3±0.9×10 13 s −1 (expressed in first-order units) and an activation energy of 75.1±0.5 kcal mol −1. Possible mechanisms for first-order recombinative hydrogen desorption from the C(100)-(2×1) surface are also discussed.

Structure and function of metal cations in light alkane reactions catalyzed by modified H-ZSM5

The rate of propane dehydrocyclodimerization to form C 6 aromatics is limited by a sequence of irreversible dehydrogenation reactions leading to propene, higher alkenes, dienes, trienes, and aromatics. Quasi-equilibrated acid—catalyzed cracking, oligometization, and cyclization reactions of alkene intermediates occur in sequence with these dehydrogenation reactions. Each dehydrogenation reaction is in turn limited by the rate of elementary steps that dispose of H-atoms formed in C-H bond activation steps. The rate of C-H bond activation, recombinative hydrogen desorption, and propane chemical conversion have been measured from the rates of isotopic redistribution and chemical conversion during reactions of C 3H 8/C 3/D 8and D 2/C 3/H 8 mixtures on H-ZSM5, Ga/H-ZSM5, and Zn/H-ZSM5. Isotopic studies show that C-H activation steps are fast during steady-state propane dehydrocyclodimerization on H-ZSM5, Ga/H-ZSM5, and Zn/H-ZSM5. Ga and Zn species increase the rates of propane chemical conversion, recombinative hydrogen desorption, and deuterium incorporation from D 2into reaction products. Disposal of hydrogen formed in C-H bond activation steps occurs by transfer of H-atoms to unsaturated species to form alkanes or to Ga and Zn species, which catalyze the recombinative desorption of H-atoms to form dihydrogen (H 2). The sequential release of several H-atoms during a propane dehydrocyclodimerization turnover limits the rate and selectivity of this reaction on H-ZSM5. In-situ X-ray absorption studies suggest that Ga and Zn species reside at cation exchange sites as monomeric cations and that recombinative desorption involve reduction—oxidation cycles of such cations during each dehydrocyclodimerization turnover. These monomeric species form directly during exchange of Zn ions from solution onto H-ZSM5. Ga 3+species, however, do not exchange directly from solution onto H-ZSM5, but instead form extrazeolitic Ga 2O 3 crystals. Ion exchange occurs during subsequent contact with propane or hydrogen at 700-800 K via vapor phase exchange of volatile Ga 1+ species.

Nonlinear optical investigations of the dynamics of hydrogen interaction with silicon surfaces

Optical second-harmonic generation (SHG) from silicon surfaces may be resonantly enhanced by dangling-bond-derived surface states. The resulting high sensitivity to hydrogen adsorption combined with unique features of SHG as an optical probe has been exploited to study various kinetical and dynamical aspects of the adsorption system H2/Si. Studies of surface diffusion of H/Si(111)7×7 and recombinative desorption of hydrogen from Si(111)7 × 7 and Si(100)2 × 1 revealed that the covalent nature of hydrogen bonding on silicon surfaces leads to high diffusion barriers and to desorption kinetics that strongly depend on the surface structure. Recently, dissociative adsorption of molecular hydrogen on Si(100)2×1 and Si(111)7×7 could be observed for the first time by heating the surfaces to temperatures between 550 K and 1050 K and monitoring the SH response during exposure to a high flux of H2 or D2. The measured initial sticking coefficients for a gas temperature of 300K range from 10−9 to 10−5 and strongly increase as a function of surface temperature. These results demonstrate that the lattice degrees of freedom may play a decisive role in the reaction dynamics on semiconductor surfaces.

Coverage dependence of the kinetics for H 2 desorption from Rh(111)

Coverage-dependent sticking probabilities and second-order rate constants for recombinative desorption of hydrogen from Rh(111) have been measured using molecular beam relaxation spectroscopy (MBRS) and time-resolved specular helium scattering. The sticking probability follows second-order Langmuir coverage dependence, with s 0 equal to 0.01 ± 0.005. Under isothermal and nearly isosteric conditions over the coverage range 0.2–0.7 ML, the second-order rate constant for desorption is essentially independent of hydrogen coverage, in contrast to kinetic parameters determined from thermal desorption spectra.

Reaction dynamics of molecular hydrogen on silicon surfaces

Experimental and theoretical results on the dynamics of dissociative adsorption and recombinative desorption of hydrogen on silicon are presented. Using optical second-harmonic generation, extremely small sticking probabilities in the range ${10}^{\mathrm{\ensuremath{-}}9}$-${10}^{\mathrm{\ensuremath{-}}5}$ could be measured for ${\mathrm{H}}_{2}$ and ${\mathrm{D}}_{2}$ on Si(111)7\ifmmode\times\else\texttimes\fi{}7 and Si(100)2\ifmmode\times\else\texttimes\fi{}1. Strong phonon-assisted sticking was observed for gases at 300 K and surface temperatures between 550 K and 1050 K. The absolute values as well as the temperature variation of the adsorption and desorption rates show surprisingly little isotope effect, and they differ only little between the two surfaces. These results indicate that tunneling, molecular vibrations, and the structural details of the surface play only a minor role for the adsorption dynamics. Instead, they appear to be governed by the localized H--Si bonding and Si--Si lattice vibrations. Theoretically, an effective five-dimensional model is presented taking lattice distortion, corrugation, and molecular vibrations into account within the framework of coupled-channel calculations. While the temperature dependence of the sticking is dominated by lattice distortion, the main effect of corrugation is a reduction of the preexponential factor by about one order of magnitude per lateral degree of freedom. Molecular vibrations have practically no effect on the adsorption/desorption dynamics itself, but lead to vibrational heating in desorption with a strong isotope effect. Ab initio calculations for the ${\mathrm{H}}_{2}$ interaction with the dimers of Si(100)2\ifmmode\times\else\texttimes\fi{}1 show properties of the potential surface in qualitative agreement with the model, but its dynamics differs quantitatively from the experimental results. \textcopyright{} 1996 The American Physical Society.

DYNAMICS OF HYDROGEN INTERACTIONS WITH Si(100) AND Si(111) SURFACES

Experimental and theoretical work probing the dynamics of dissociative adsorption and recombinative desorption of hydrogen at Si(100) and Si (111) surfaces is reviewed. Whereas molecular beam experiments demonstrate that molecular excitations do aid in overcoming a substantial activation barrier toward adsorption, desorbed molecules are found to have a total energy content only slightly above the equilibrium expectation at the surface temperature. A consistent interpretation of the ad/desorption dynamics is arrived at which requires neither a violation of microscopic reversibility nor defect-mediated processes. An essential element of this model is that surface atom relaxations play an essential role in the dynamics such that different portions of the potential energy hypersurface govern the results of adsorption and desorption experiments. The ‘lost’ energy, i.e. that portion of the activation energy not evident in the total energy of the desorbed molecules, is deposited in the surface coordinates where it is inaccessible to experiments that probe the desorbates final state.