Development of selective iron-based fischer-tropsch catalysts to light olefins

Abstract

Light olefins (alkenes) are among the key chemicals that are globally most produced from crude oil in amounts exceeding 200 million tons per year. They are hydrocarbons with at least one carbon - carbon double bond (C2-C4) namely, ethylene(C2H4), propylene(C3H6), and butylene(C4H8). Lower olefins (light olefins) are intermediates for the synthesis of a wide range of products such as solvents, polymers, drugs, detergents, and cosmetics. There are three types of olefins: alpha (also called ethylene molecules), beta, and gamma. Carbon-carbon double bond is located at the beginning, in the middle, and at the end of the olefin chain in alpha, beta, and gamma types, respectively. Currently, commercial light olefin production is mainly based on steam cracking of a broad range of hydrocarbon feedstock including naphtha, gas oil, condensates, ethane and propane. However, the production of lower olefins by steam cracking is one of the most energy-consuming processes of the chemical and petrochemical industry. The oil reserves are expected to be depleted at faster pace as the oil consumption surpasses the conventional oil production. As the conventional easy-reached oil reserves deplete, attempts are being made to use unconventional oil reserves for oil production. The extraction and upgrading of oil from unconventional oil reserves, however, may be expensive and involve release of higher amounts of CO2 release in comparison to conventional reserves. CO2, with its green-house effect, is widely claimed to be responsible for climate change and there is rapidly growing global awareness in this respect which leads to more and more stringent regulations about CO2 emissions. Therefore, many countries are searching for alternatives to reduce their reliance on imported crude oil and refined products and to comply with CO2 regulations. The new alternative fuels for olefin production are coal, natural gas and biomass. Light olefins may be produced from the synthesis gas (CO/ H2) obtained from gasification of these fuels by direct Fischer-Tropsch-to-Olefins (FTO) process. FTO is a catalytic process and the most crucial and critical issue of this process is using proper and effective catalyst(s). Although there are plenty of research available in literature focusing on FTO process, there still exists lack of a proper catalytic process to be used commercially in FTO. In this work, the aim was to make an effort to produce light olefins in direct unconventional way via FTS by synthesizing iron-based catalysts with different promoters and supports that can show high FTO performance, means high CO conversion and stability with time on stream, high selectivity to light olefins, and low selectivity to methane and CO2. In other words, the aim was to narrow the wide hydrocarbon range produced by FTS to C2-C4 olefins. To reach the goal of study, the catalysts have been synthesized in different routes and with promoters. Their performance has been evaluated via catalytic tests and catalytic activity-structure relation have been investigated. In this term, iron-based catalysts have been prepared both by precipitation and impregnation techniques. Precipitation route has been tuned as well by changing the alkalinity of the precipitation environment. To synthesize the first set of bulk catalysts, nitrate salt solutions of iron and zinc as prepared in a stoichiometry of Fe:Zn=2 have been co-precipitated with NH4OH (AH). Sodium has been incorporated to Fe.Zn precipitate by different routes; use of sodium nitrate during co-precipitation reaction (AH route) or its subsequent impregnation on co-precipitated Fe.Zn catalyst (AH-I route). Alternatively, Na2CO3 (SC) was employed instead of NH4OH (AH) for the initial precipitation to investigate the role of the precipitant and its effect on catalyst surface basicity in terms of Fischer-Tropsch activity. The basicity of the precipitate Fe.Zn (SC) has been altered by changing the number of washing cycles as well and impregnated with a sodium precursor for further basicity. The last route has been called as SC-I. In addition to Na, Cu and K promoters have been impregnated to Fe.Zn precipitate as well. For the impregnation route, activated carbon (AC) and nitrogen-doped AC have been used as support material. Fe:Zn of 2 with alkali promoters has been chosen since it resulted in high olefin selectivity for unsupported catalysts. Activated carbon (AC) and its nitrogen doped form have been used as support. Activated carbon has been treated with N-containing chemicals namely, HNO3, NH3 and urea in order to create nitrogenous surface functional groups over AC (Chemical modification of surface). The so-formed supports were denoted as AC-N1, AC-N2, and AC-N3, respectively upon treatment with HNO3, NH3, and urea. Co-impregnation has been applied as by first dissolving the metal salts in stoichiometric amounts in a minimum amount of water and then by wetting AC support with the metal salt solutions. All catalysts have been calcined, reduced and tested in a high pressure fixed-bed reactor to investigate their catalyst activity and performance in Fischer-Tropsch synthesis to light olefins (FTO). Test results were interpreted together with the characterizations such as BET surface areas via N2 adsorption, crystal phase identification by x-ray diffraction (XRD), elemental analysis by inductively coupled plasma (ICP-OES), thermal stability of supports using (TGA) analysis, morphological investigation by scanning electron microscopy (SEM), reducibility characteristics of active phases using H2-TPR , the basicity of Na promoted bulk catalysts by CO2-TPD, and SEM-EDS mapping to observe metal distribution in Na promoted catalysts and a carbon supported catalyst. As total alkalinity of precipitation affects hydrolysis and influences the composition of the intermediate hydrolytic complexes, the final features of metal hydroxide precipitates might be induced with the precipitation conditions. This was proved by the observed change on the textural properties of zinc ferrites such as total surface area, crystal size and morphology under different alkaline precipitation environment. Improved conversion due to the facilitated CO dissociation over basic sites and concomitant deactivation possibly through fouling might be interpreted as both the number of basic sites and strength were determinant on the final catalytic behavior. Na provides a surface with high electron density that leads to intensification of CO dissociation and adsorption. However, although this is a favorable effect, it has limitation in terms of alkali content of the catalyst and its dispersion. As mentioned before, there is an optimum basicity which ensures a balance between CO conversion to CHx and the rate of hydrocarbon chain growth and its termination. If this balance alters, long chain hydrocarbon may form and cover the catalyst surface which can block the active sites. Depleted surface vacancies might suppress the rate of CO dissociation with time on stream and result in high deactivation. In this case, the hydrocarbon distribution does not change significantly but deactivation dominates. When surface is covered with high C content, surface H deficiency may result in a decrease in hydrogenation activity that might end up with poor paraffin and methane selectivity. Alkali metals may also improve re-adsorption of olefinic intermediates which may further polymerize to C5+ species. Therefore, both C2-C4 olefin light olefin and C5+ selectivity increase in the presence of alkali promoter, sodium or potassium. However, potassium as with more alkalinity strength led to more coke formation over the catalytic surface, e.g. a total carbon content of 17% on the spent ⁓3%K-2Fe.Zn(SC) catalyst and thereby, a fast decrease in CO conversion from 88% to 55% has been observed. Copper (Cu) is a widely accepted promoter for facilitated reduction of iron oxides and it improves the catalyst stability when the reduction occurs at lower temperatures. Temperature programmed reduction (TPR) profile of the catalysts have shown the improved dispersion upon copper addition. AC and N-doped AC supported Fe.Zn catalysts have shown high and satisfactory FTO performance. High surface area of AC whether in N-doped form or not has provided improved catalytic stability. Supported catalysts in the same Fe:Zn ratio seemed to be less deactivated in comparison to the bulk catalysts. For all supported catalysts in Fe: Zn: P=2:1:0.2 molar ratios where P=Na and K, high olefin selectivity of ⁓ 45-50% and high CO conversion of ⁓89-93%(stability) have been achieved. Using AC as the support of the catalyst, dispersion and hence the number of active sites have been increased and accessibility to each single active site has been improved as compared to the case in bulk catalysts. In conclusion, the hydrocarbon product distribution in the presence of alkali has been altered towards high C2-C4 olefin and C5+ selectivity values. The strength and homogeneity of surface basicity seemed effective in case of bulk catalysts. Highly active and selective bulk catalysts can be prepared by changing the precipitation conditions. Copper was seen to stabilize the catalytic activity by improving the dispersion and reducing the reduction temperatures of metal oxides. Activated carbon appeared as a suitable support for well dispersion of active sites. Its surface nature has been modified with nitrogen and slight changes in catalytic performance has been noticed. Improved thermal stability upon nitrogen doping was the only point to be remarked. All in all, Fe to Zn ratio of 2 as in zinc ferrite spinel crystals were active catalysts for Fischer-Tropsch reaction and reaction selectivity have been directed to light olefins when appropriately doped with alkali metals.