Diane R. Milburn

Activity and selectivity of precipitated iron Fischer-Tropsch catalysts

Abstract Low-temperature (230°C), slurry phase Fischer-Tropsch synthesis (FTS) was conducted with precipitated iron-silicon catalysts under industrially relevant conditions (flow=3.1 Nl h −1 g-Fe −1 , H 2 :CO=0.7, P=1.31 MPa). The effects of activation gas (hydrogen, carbon monoxide, or syngas) and promoters (potassium and copper) on activity and selectivity were explored. Optimum potassium promotion was 4–5 at%, relative to iron. Promotion with copper lowered the reduction temperature and increased FTS activity, regardless of the activation gas used. Carbon monoxide activation gave the highest activity for a 100Fe/ 4.4Si/5.2K catalyst (atomic percent, relative to iron) while syngas activation was superior for a 100Fe/4.4Si/2.6Cu/5.2K catalyst. Selectivity of the FTS product was not affected by the activation gas employed or copper promotion; however, potassium promotion increased wax and alkene selectivity. A syngas activated 100Fe/4.4Si/2.6Cu/5.2K catalyst gave the best overall performance at 230°C. Alkene selectivity was >75% for the C 2 -C 11 fraction, methane selectivity was 12 + selectivity was >70 wt%.

Conversion of 2-octanol over nickel-alumina, cobalt-alumina, and alumina catalysts

Abstract Structural characterization of CoAl and NiAl coprecipitated catalysts has been accomplished using techniques such as X-ray diffraction, DTA, TGA, TG-MS, BET surface area and pore volume. The catalyst samples were prepared by coprecipitation of cobalt (or nickel) and aluminum from their aqueous salt solutions. The interaction of Co (or Ni) with aluminum has been investigated by X-ray diffraction as a function of calcination temperature. The results indicate that the interaction between Co and Al can lead to the formation of a normal spinel of CoAl 2 O 4 at a very low temperature such as 473 K. On the other hand the interaction between Ni and Al is not sufficient to form well crystallized NiAl 2 O 4 species at low temperatures but nickel-aluminate formation was seen above 973 K. The activity and selectivity of these catalysts for 2-octanol conversion are dependent on the reaction temperature, liquid hour space velocity (LHSV) as well as on the calcination temperature. The dehydrogenation activity of nickel based catalysts is attributed to the presence of nickel oxide whereas the dehydration activity is due to the spinel phase. The differences in the dehydration activity between alumina and nickel-alumina is mostly because of the incomplete formation of spinel phase (i.e., low temperature calcined samples). Catalytic activity of cobalt-alumina catalysts is comparable with that of alumina catalysts towards 2-octanol conversion.

Fischer-Tropsch synthesis: Impact of potassium and zirconium promoters on the activity and structure of an ultrafine iron oxide catalyst

Slurry phase Fischer-Tropsch synthesis was conducted with an ultrafine iron oxide catalyst promoted with either 0.5 at% K or 1.0 at% Zr or both. Pretreatment in CO yielded higher conversions and a more stable catalyst than activation in hydrogen or synthesis gas. Hydrogen pretreatment of K promoted catalysts and synthesis gas activation in general were less effective. Mössbauer spectroscopy and XRD showedχ-Fe5C2 and ε′-Fe2.2C were formed during pretreatment in CO and did not depend on promoters present. Catalysts pretreated in H2 were reduced to metallic Fe and Fe3O4; promotion with K and Zr decreased the extent of reduction. Hydrogen pretreated catalysts, promoted with K, lost surface area and carbided rapidly under synthesis conditions. Activation in synthesis gas reduced all catalysts to Fe3O4. Subsequent synthesis did not affect the phase present for the unpromoted and Zr promoted catalysts while those promoted with K formed χ-Fe5C2 and ε′-Fe2.2C. It is concluded that pretreatment type is more important to the catalyst activity during the early period of synthesis than the impact of promotion with K and/or Zr and that changes in the bulk composition of iron catalysts do not necessarily correlate with changes in activity.

PtSO42−ZrO2 catalysts. Correlation of catalytic activity with SO42− XPS data

Increasing activity for the conversion of n-hexadecane is observed for 0.6% PtSO42−ZrO2 catalysts with increasing time of activation in air at 500°C (773 K). XPS analysis suggests that, rather than a change in the crystal phase, this increase is due to a loss of water and/or an increase in surface concentration of SO42−. It is noted that there is a reasonably good correlation between the relative surface sulfate concentration and relative conversion of n-hexadecane with increasing heating time, although there are minor differences. In spite of the many assumptions that were made to generate the two curves showing the relative XPS peak area and the catalytic activity, there is surprising agreement between the two. It is therefore inviting to relate the catalytic activity for n-hexadecane conversion to that of the sulfate group. Furthermore, it appears that the active site related to the sulfate group is not one that contains water in its make-up.

XPS investigation of an iron/manganese/sulfated zirconia catalyst

Abstract A sample of Fe–Mn–SO 4 2− –ZrO 2 has been heated at 500°C in air for 98 h. At intervals, the sample was evacuated and transferred without atmospheric exposure to an XPS chamber. As noted with Pt–SO 4 2− –ZrO 2 , the O 1s peak resolved to a doublet; one of these peaks is interpreted to result by dehydration of the sulfate group. Following the 98 h air treatment, the sample was treated at 150°C at 1 atm in flowing hydrogen for a total of 78 h. The XPS spectra, obtained at intervals during the heating in hydrogen, showed that both Fe and Mn remained in an oxidized state.

Sulfated zirconia: attempt to use n-butylamine to measure acidity

Efforts to measure Bronsted acidity using the chemisorption of n-butyl amine and quantifying the number of Bronsted sites as being equal to the butenes and the ammonia formed during temperature-programmed heating was unsuccessful. Rather, it was observed that even at ambient temperature, chemisorption of n-butyl amine was accompanied by oxidation/reduction reactions that produced water, CO2 and oxides of sulfur. The results suggest that oxidation/reduction reactions initiate the conversion of hydrocarbons by this catalyst at low temperatures.

Promoted iron Fischer-Tropsch catalysts: characterization by thermal analysis

Abstract The effects of metal and alkali promoters on precipitated iron oxide Fischer-Tropsch catalyst precursors are examined using thermal gravimetry (TG) and differential thermal analysis (DTA). A distinct exotherm corresponding to what appears to be the transformation from FeOOH to Fe 2 O 3 is observed for unpromoted and promoted iron oxides when the promoters ionic radius is less than approximately 0.7A. The heat released during the exotherm is similar for unpromoted and promoted iron oxides corresponding to this size range. For promoters having ionic radii larger than 0.7A, the exotherm occurs over a broad temperature range. Silica exhibits a unique propensity to stabilize against this transition, and appears to be present as two species. Increased loadings of Si cause a shift from the sharp, narrow temperature range exotherm to the broad transition observed for promoters with larger ionic radii.

Promoted iron Fischer-Tropsch catalysts: characterization by nitrogen sorption

Abstract The effects of promoters on BET surface area and porosity of precipitated iron catalysts are examined. A series of metal oxide promoted iron catalysts was prepared by coprecipitation of the nitrate salts to produce 6% by weight of promoter oxide. Surface areas and pore volumes tend to decrease with increasing promoter ionic radius although there is essentially no change in the shape of the nitrogen isotherm or pore size distribution. Addition of alkali (K) or alkaline-earth (Ca) metals to silica or alumina promoted iron catalysts also results in a decrease in surface area and pore volume with increasing metal loading. The presence of alkali or alkaline-earth metals appears to impede crystallization during heating.

Porosity of Silicas: Comparison of Nitrogen Adsorption and Mercury Penetration

Publisher Summary This chapter discusses the porosity of silicas: comparison of nitrogen adsorption and mercury penetration. Nitrogen adsorption and mercury penetration are two of the most common methods for obtaining information about the porosity of solids. Unfortunately, experimental realities limit the region where the two methods are both applicable to pore sizes of ca. 10–30 nm. Both nitrogen adsorption and mercury penetration provide a direct measure of pore size distribution and pore volume. However, these direct measures do not permit a determination of the morphology of the materials from pressure–volume relationships. Nitrogen adsorption–desorption isotherms were obtained with a Quantachrome Autosorb 6 instrument. Prior to analysis, samples were outgassed for several hours at about 10 -3 torr and 200°C. Surface areas are calculated from the linear form of the BET equation. The model utilized for pore size distribution calculations is a packed particle model, assuming each primary spherical particle is in contact with six neighboring spheres. Mercury penetration curves were generated from pressure–volume measurements from 0 to 60,000 psia using a Quantachrome Autoscan 60 instrument. The surface areas are calculated using the Rootare–Prenzlow equation.

Mechanism for coking of coal liquefaction catalysts involving basic nitrogen compounds, sodium and catalyst acid sites

A simple mechanistic model is presented to explain the deposition of sodium and coke on hydrotreating catalysts used in the Wilsonville, AL coal liquefaction pilot plant. The model incorporates data from four process configurations, two NiMo/Al2O3 catalysts, and two coal types. In this model, “coke” takes the form of basic nitrogen compounds which compete with sodium for acid sites on the catalyst surface. Elemental analyses of the spent catalysts allow calculation of average stoichiometries and approximate desorption rate constants for the basic nitrogen compounds, which are consistent with hydrotreater feed stream analyses and laboratory studies.