H. Marsh

Preparation of activated carbon cloths from viscous rayon. Part I. Carbonization procedures

Viscose rayon cloth was carbonized under a wide range of experimental conditions, including heating rate, isothermal stage near the temperature of decomposition, and preoxidation in air. Slower carbonization rates produce increases in yield, strength and micropore volume of the char, lower reactivity against gasification, and micropores with narrower entrances. The inclusion of an isothermal stage at temperatures near the beginning of the pyrolytic decomposition leads to a lower presence of constrictions in the char. Preoxidation with air produces an increase in carbonization yield and in breaking load; the presence of oxygen surface groups in the char are responsible for its higher reactivity to gasification.

Pyrolysis of petroleum residues: II. Chemistry of pyrolysis

Abstract Pyrolysis of two petroleum residues, R1 and R2 of different aromaticities, were studied at 0.1 and 1.0 MPa, 420–480°C, for soak times up to 12 h. Pyrolysis was followed by extracting resultant semicokes, sequentially, with heptane, toluene and N-methyl-pyrrolidinone to obtain four fractions HS, TS, NMPS and NMPI. Development of the anisotropic mesophase was monitored by polarised light optical microscopy. The extraction and microscopy methods are critically compared in monitoring pyrolysis progress. Differences between measured growths of the extraction fractions and anisotropy are explained. Growth in aromaticity in the HS and TS was followed using 1H-NMR and FTIR (CHal/CHar), and in NMPS and NMPI fractions by FTIR. Vapour pressure osmometry estimated average molecular weights of four fractions from R1. Aromaticities in the final semicokes from R1 and R2, were very similar, despite initial differences in the parent feedstocks. Growth in constituent molecular size probably occurs via a consecutive reaction mechanism. The early stages of pyrolysis involve dealkylation reactions. Early mesophase still contains significant amounts of aliphatic side-chains, which are reduced as pyrolysis proceeds. Rates of transfer of HS to TS to NMPS to NMPI for R1 and R2 are compared. The paper reports in considerable detail aspects of the chemistry of these processes.

Preparation of activated carbon cloths from viscous rayon. Part II: physical activation processes

Activated carbon cloths were prepared from viscous rayon by a two-step activation process (carbonization followed by activation) and a single-step activation process (simultaneous carbonization and activation), and were compared in terms of gasification rate, development of porosity and changes in breaking strength. Direct activation of the precursor takes place with a higher rate than char activation, the final products of the former process exhibiting slightly lower micropore volume and surface area. The effect of using carbon dioxide or steam as the activating agent was also analyzed, the former producing a continuous development of narrow microporosity and slight widening only above 30% activation, and the latter a widening of micropores as from the early stages of the process.

Pyrolysis of petroleum residues : I. Yields and product analyses

Abstract This study examines the pyrolysis of three petroleum pitch residues of different aromaticities, (R1), (R2) and (R3), varying the experimental parameters of pressure, temperature and soak time. The overall objective is to provide further detailed information of factors which influence formation of anisotropy or mesophase in resultant semicokes. Pressure was varied during the progress of a pyrolysis. Yields of gases, liquids and semicokes were obtained. Gases were analysed by gas chromatography, the liquids by simulated distillation and 1 H-NMR, and the semicokes by elemental analysis and FTIR. For the semicokes from R1, yields are dominantly a function of pressure, with little influence of temperature and soak time. For semicokes from R2, yields are dominantly a function of pressure and temperature, with little influence of soak time. For semicokes from R3, yields are dominantly a function of temperature and soak time, with little influence of pressure. The use of simulated distillation and pressure release, at reaction temperatures, provides additional information about mechanisms of the pyrolysis reactions.

Delayed coking : Industrial and laboratory aspects

Abstract Delayed coking is a thermal process to convert petroleum residues to a solid coke material. Processes occurring in a delayed coker are complicated and attempts have been made, at the laboratory level, to simulate industrial delayed coking. Although the latter studies are useful, it is impossible to scale-down to the laboratory level. Industrial delayed coking is a turbulent process and such movements cannot be simulated easily in the laboratory. Of industrial importance are the multiphase systems i.e. volumes of unreacted isotropic pitch residue, transported through the bulk, fluid anisotropic mesophase, so creating ordering into acicular structures in the vicinity of the multiphase systems. Four petroleum residues were analysed chemically. Pyrolyses were carried out under pressures of up to 1.0 MPa. Complete mass balances were obtained and the semicokes examined by optical microscopy. Feedstocks for delayed cokers can be blends of petroleum residues, some of which can produce considerable amounts of volatile materials. Volatile evolution, at the optimum operating condition of the delayed coker, can bring about improvements in resultant coke quality. In industrial delayed coking it is important not only to consider the chemistry of the feedstocks, but attention must also be given to the physico-chemical aspects of coker operation.

Influence of pressure variations on the formation and development of mesophase in a petroleum residue

Abstract The objective of this study is to search for a method, involving control over the pressures within a pyrolyzing system, to produce non-coalesced mesophase spheres of required sizes. Semicokes were studied from pyrolyses of a petroleum residue, R1, carried out in a tube reactor, 440–445°C, 1.0 MPa pressure, and soak times 1.0–2.0 h. In some pyrolyses, pressures were reduced from 1.0 to 0.1 MPa and the pyrolysis continued at the same reaction temperature at 0.1 MPa. These experiments are termed depressurizations. Semicokes were examined by optical microscopy to measure mesophase content as well as distributions of size of mesophase spherules and sphere concentration (number of spheres mm−2). High-pressure (1.0 MPa) pyrolyses favour the coalescence of mesophase spheres to form bulk mesophase. Initially formed spheres do not grow beyond a certain size indicating that rates of coalescence exceed rates of growth of the spheres. Depressurization causes a loss of volatile matter with a proposed resultant increase in viscosity of the remaining isotropic phase. Some volatile material may also be removed from the mesophase spheres themselves, again promoting a viscosity increase. As a result of this, the spheres increase their diameter and there is less production of bulk mesophase. With depressurization, coalescence is not stopped entirely but is reduced considerably in extent.

Surface topography of oxidized HOPG by scanning tunneling microscopy

Abstract Scanning tunneling microscopy (STM) has been used to image the surface structure of gasified highly oriented pyrolytic graphite (HOPG). Surfaces of HOPG were oxidized in air at 650 °C by TGA to 5 and 30 wt% burnoffs. The pits that were formed following this oxidation were examined by STM. Removal of carbon atoms adjacent to the vacancy in the same plane as well as removal of carbon atoms in planes below the surface were observed. The initial oxidation resulted in the formation of circular nm-sized pits. After oxidation to a depth of a few basal planes the nm-sized pits elongated into channels. Upon further oxidation, these elongated pits and channels collapsed to form hexagonal μm-sized pits (of the order of several μm). Steep μm-sized pit walls that are nearly vertical were observed in some images. The cross-sectional analyses of these μm-sized pits showed deeper intrusion at the sides of the pit rather than at the center of the pit floor (a few hundred basal planes), giving the pit floor a dome-like appearance. Because oxidation in the z-direction was expected to be much slower than in the x-y-direction the steepness of the pits was unexpected. One possible explanation is that the pit is formed from oxidation on a very deep screw dislocation or at the micro-crystallite grain boundaries. A second explanation is that naturally occurring vacancies in the basal planes allow oxygen to react with many layers relatively quickly. A third explanation could be the association with small sized impurities (catalysts) on the surface. However, scans were made of the unoxidized HOPG prior to oxidation and no defects were imaged. A cross-sectional analysis of the initial circular nm-sized pits showed the presence of these surface-oxide complexes.

Pyrolysis of petroleum residues: III. Kinetics of pyrolysis

Abstract The kinetics of mesophase formation for three petroleum residues of different aromaticities (R1, R2 and R3) have been followed by optical microscopy (anisotropic content ‘A’) and solvent extraction (insolubility in heptane (HI), toluene (TI) and 1-methyl-2-pyrrolidinone (NMPI)). Kinetics of mesophase formation followed as development of HI, TI, NMPI and ‘A’, approximate to apparent first order kinetics. Deviations from Arrhenius plots are observed. These could be caused by: a) depletion of reactants; b) effect of consecutive reactions, c) effect of reversible reactions, d) submicron mesophase in the isotropic phase and e) mesophase solubility. Activation energies for the three feedstocks are in close agreement. Values are between 160 and 270 kJ mol−1 and are a function of the experimental method (extraction with different solvents and optical microscopy). A consecutive reaction model has also been applied for R1 and R2 pyrolyses. Experimental values adjust reasonably to calculated values applying this method. Both R1 and R2 follow this model, with calculated activation energies of 240±20 kJ mol−1.