Jan B. C. Pettersson

The effects of fuel washing techniques on alkali release from biomass

The influence of different washing techniques on the alkali release during pyrolysis of biomass is studied. After washing and drying, samples of wheat straw, wood waste and cellulose are subjected to a constant heating rate in a N2 atmosphere, and the release rate of alkali compounds from the sample is measured continuously by a surface ionization technique. Alkali is released from the untreated biomass in two temperature intervals: (1) in connection with the pyrolysis process taking place at 200–500°C, and (2) from the material remaining after pyrolysis at temperatures above 600°C. Separate vacuum pyrolysis experiments show that the alkali release is dominated by potassium-containing compounds, with minor contributions from sodium-containing compounds. The effect of water washing of the biomass is compared with a more thorough acid leaching technique. In the temperature range 200–500°C, washing with water reduces the alkali emission from wood waste and wheat straw by 5–30%, while acid leaching is more effective and reduces the emission by around 70%. Above 600°C where the vaporization of alkali compounds from untreated wheat straws increases sharply, the washing procedures are sufficient for a reduction in the measured alkali release by more than 90%. Experiments with pure cellulose (ash content 0.07%) indicate that the washing methods are ineffective in removing alkali bound to the organic structure of the biomass. The results support the conclusion from earlier studies that relatively simple washing techniques can improve the combustion properties of biomass fuels with a high ash content. For fuels with a lower ash content like woody biomass, the concentration of alkali bound to the organic structure limits the effect of the washing techniques.

Pyrolysis of large wood particles : a study of shrinkage importance in simulations

Shrinkage models have been developed and included in a model for the pyrolysis of large wood particles. Shrinkage is modelled in three different ways: uniform shrinkage, shrinking shell and shrinking cylinders. These models and a reference model without shrinkage are compared with experimental data for mass loss versus time during pyrolysis of birch cylinders at different temperatures. In the experiments a wood particle was introduced into a pyrolysis furnace held at constant temperature. The particle mass and volume were recorded using a balance and a video camera. Uniform shrinkage slows down the pyrolysis whereas shrinking shell and cylinder models enhance the pyrolysis rate. The effect was sufficiently small to be neglected given the uncertainty about some wood physical properties.

Scattering of water from graphite: simulations and experiments

Abstract The scattering of water molecules from a graphite (0001) surface has been studied experimentally and using molecular dynamics simulations. Model potentials for the gas–solid interaction have been developed and tested by comparing experimental and simulated angular distributions for different translational energies (0.09–0.8 eV), incident angles (30–75°) and surface temperatures (300–1200 K). Good agreement is found for both scattering intensities and angular resolved average velocities, supporting the main features of the model potentials. The importance of trapping at low incident energy is discussed as well as the possibility of vibrational excitation at high collision energy. The latter phenomenon is investigated using purely classical trajectory calculations, a forced oscillator model and a wavepacket propagation scheme. The results indicate that meaningful estimates of the vibrational excitation can be obtained using classical simulations, supporting previously reported results on the CF 3 Br/graphite system [M.B. Nagard, N. Markovic, J.B.C. Pettersson, J. Chem. Phys. 109 (1998) 10350].

Angular resolved neutral desorption of potassium promoter from surfaces of iron catalysts

Abstract The angular dependence of neutral potassium emission in the form of ground-state atoms as well as Rydberg species is studied from a fused iron catalyst. The catalyst is of the type used for ammonia synthesis in so-called pre-reduced (metallic) condition. The angular distributions observed by surface ionization detection have a more peaked shape than the cosine distribution expected for thermal equilibrium. In the case of a catalyst sample used in the industrial process even a sharp peak on top of a cosine distribution is found. Using detection by field ionization, i.e. detection of Rydberg species only, a rather sharp lobe in the normal direction is found. A theoretical description of cluster formation outside the sample surface from atoms with velocity distributions characteristic for thermal equilibrium is used to interpret the results. The cluster formation is probably due to the long-range interaction between the Rydberg atoms formed on the surface, and the clusters are at least partially formed in an excited state. The cluster sizes contributing to the distributions are estimated from fits to the experimental results. The main cluster size observed with surface ionization detection is concluded to be quite small, containing just a few atoms. There also exist contributions of larger clusters of the size around 10–30 atoms in the case of the pre-reduced catalyst. The used catalyst also gives mainly small clusters, but it does not give clusters of the size 10–30 atoms. Both types of catalyst also give a small number, less than 5%, of very large clusters, with more than 100 atoms according to the model. The field ionization data for the pre-reduced catalyst are well matched by a single cluster size of approximately 30 atoms, which indicates that such clusters have a longer lifetime in the initial excited state than the small clusters.

Surface properties of water ice at 150–191K studied by elastic helium scattering

A highly surface sensitive technique based on elastic scattering of low-energy helium atoms has been used to probe the conditions in the topmost molecular layer on ice in the temperature range of 150-191 K. The elastically scattered intensity decreased slowly as the temperature was increased to about 180 K, followed by a rapid decrease at higher temperatures. An effective surface Debye temperature of 185+/-10 K was calculated from the data below 180 K. The changes in the ice surface above 180 K are interpreted as the onset of an anomalous enhancement of the mean square vibrational amplitude for the surface molecules and/or the onset of a limited amount of disorder in the ice surface. The interpretation is consistent with earlier experimental studies and molecular dynamics simulations. The observed changes above 180 K can be considered as the first sign of increased mobility of water molecules in the ice surface, which ultimately leads to the formation of a quasiliquid layer at higher temperatures. A small shift and broadening of the specular peak was also observed in the range of 150-180 K and the effect is explained by the inherent corrugation of the crystalline ice surface. The peak shift became more pronounced with increasing temperature, which indicates that surface corrugation increases as the temperature approaches 180 K. The results have implications for the properties and surface chemistry of atmospheric ice particles, and may contribute to the understanding of solvent effects on the internal molecular motion of hydrated proteins and other organic structures such as DNA.

Classical trajectory study of argon–ice collision dynamics

Classical trajectory simulations have been used to study Ar–ice Ih collisional energy transfer, trapping coefficients and scattering distributions for initial Ar kinetic energies between 0.1 and 2.0 eV, incident angles between 0 and 70° and surface temperatures between 0 and 300 K. Collisional energy transfer is extremely efficient due to substantial transfer of energy from the Ar atom to the ice surface over typically 2–4 gas-surface encounters, and the rapid dissipation of this energy away from the collision center, preventing energy transfer back to the Ar atom. This leads to large trapping coefficients over this range of Ar collision energies, incident angles and surface temperatures. Scattered gas atoms lose most of their initial kinetic energy and have broad angular distributions. The large trapping coefficients obtained for the Ar–ice collisions are expected to be found for similar reactions under stratospheric conditions (e.g., HCl–ice, HOCl–ice and ClONO2–ice).

Intermethod comparison of the particle size distributions of colloidal silica nanoparticles

Abstract There can be a large variation in the measured diameter of nanoparticles depending on which method is used. In this work, we have strived to accurately determine the mean particle diameter of 30–40 nm colloidal silica particles by using six different techniques. A quantitative agreement between the particle size distributions was obtained by scanning electron microscopy (SEM), and electrospray-scanning mobility particle sizer (ES-SMPS). However, transmission electron microscopy gave a distribution shifted to smaller sizes. After confirming that the magnification calibration was consistent, this was attributed to sample preparation artifacts. The hydrodynamic diameter, d h , was determined by dynamic light scattering (DLS) both in batch mode, and hyphenated with sedimentation field flow fractionation. Surprisingly the d h were smaller than the SEM, and ES-SMPS diameters. A plausible explanation for the smaller sizes found with DLS is that a permeable gel layer forms on the particle surface. Results from nanoparticle tracking analysis were strongly biased towards larger diameters, most likely because the silica particles provide low refractive index contrast. Calculations confirmed that the sensitivity is, depending on the shape of the laser beam, strongly size dependent for particles with diameters close to the visualization limit.

Scattering and trapping dynamics of gas-surface interactions: Theory and experiments for the Xe-graphite system

We report on molecular beam experiments and molecular dynamics simulations of xenon scattering with incident energies E=0.06−5.65 eV from graphite. The corrugation felt by an atom interacting with the surface is found to be influenced by both surface temperature, Ts, and E. Angular distributions are significantly broadened when Ts is increased, clearly indicating corrugation induced by thermal motion of the surface also at the highest E employed. Direct scattering dominates for high E, while trapping becomes important for kinetic energies below 1 eV. The coupling between atom translation and surface modes in the normal direction is very effective, while trapped atoms only slowly accommodate their momentum parallel to the surface plane. The very different coupling normal and parallel to the surface plane makes transient (incomplete) trapping-desorption unusually pronounced for the Xe/graphite system, and atoms may travel up to 50 nm on the surface before desorption takes place. The nonlocal and soft charac...

Chemical analysis of individual alkali-containing aerosol particles : Design and performance of a surface ionization particle beam mass spectrometer

A mobile particle beam mass spectrometer has been developed to measure the alkali metal content in individual submicron aerosol particles. The instrument employs an aerodynamic inlet system for efficient sampling of particles into vacuum, and the detection of individual particles is based on decomposition and surface ionization on a hot platinum surface. A boxlike design of the hot ionizing surface is shown to limit problems associated with particle bounce effects and incomplete ionization, and the decomposition/ionization process is not sensitive to detailed particle properties. High transmission efficiencies and quantitative determination of the alkali metal content in individual particles with diameters down to 14 nm are demonstrated. Experiments with particles doped with alkali salt show that the size range may be extended down to a few nanometers after further improvements of the inlet system. High size resolution can be achieved with the instrument for particle sizes down to tens of nanometers, as illustrated by the detection of multiply charged particles passing through a DMA. The robustness of the instrument makes it suitable for field measurement applications, and the technique is demonstrated at a 12 MW biomass combustion facility. The performance of the instrument and further refinements of the technique are discussed, and potential applications in field and laboratory studies are outlined.

Collision dynamics of large water clusters

Classical trajectory calculations of (H2O)n+(H2O)n collisions are carried out for n=125 and n=1000. We investigate energy redistribution and fragmentation behavior for relative collision velocities up to 3000 ms−1, impact parameters up to 4 nm, and initial cluster temperatures of 160 and 300 K. Three main scattering channels are identified; coalescence, stretching separation, and shattering collisions. For small impact parameters, low collision velocities produce coalesced clusters while high velocities yield shattering behavior. Large impact parameters combined with high velocities result in stretching separation collisions. A decreased internal temperature influences the dynamics by increasing the stability of the collision complex. The results for (H2O)125 and (H2O)1000 are comparable, although the smaller size allows individual molecules to have a larger influence on the overall behavior. We find good agreement between the cluster simulations and experimental data for water drops in the micrometer ran...