Periodic Reporting for period 2 - BioSCoPe (Impact of Biofuels on the Oxidation Stability and Combustion Pollutants of Heavy Duty and Jet Fuels)
Reporting period: 2023-03-01 to 2024-08-31
Several transport sectors rely on propulsion systems that remain difficult to replace with electric motors. The most emblematic case is the aviation sector, where the use of electric motors for long-haul jet aircraft is unthinkable for a long time to come with current battery technologies. Indeed, the power needed for take-off and to supply the energy required for a long-haul flight would require a battery weight of around 1 ton per passenger, which leads to a total battery weight that remains prohibitive. Other sectors, such as road freight, face similar challenges in making the transition to 100% electric motors, due to the heavy weight of batteries, which would take up the entire payload, and the long recharging times of such batteries. For these sectors, which are crucial to our societies and account for a significant proportion of total CO2 emissions, the most effective short-term solution for reducing their carbon footprint lies in the use of biofuels. In the context of regulated production criteria and use in transport, biofuels can be, renewable, locally produced and directly usable in today's engines and logistics infrastructures. The variety of biofuel source materials leads to a great diversity of molecules, which have in common the fact of containing oxygen atoms in their chemical structure, in addition to their carbon skeletons that are also present in fossil fuels. This difference has consequences for the properties of biofuels, whether used as pure or in blends with conventional fuels. One of the critical fuel properties that is degraded in biofuel is their resistance to liquid-phase oxidation. This ubiquitous phenomenon is the main cause of fuel aging, as the fuel is in contact with air throughout its life cycle, from storage to combustion. Resistance to aging is therefore a crucial criterion to ensure that the composition of the fuel does not change, leading to changes in essential physical properties and ultimately to the formation of gums and solid deposits that can clog the fuel supply system. The consequences of these phenomena are engine malfunctions, in terms of energy efficiency, pollutant formation and can lead to safety problems, especially in aeronautics. The industrial solution currently used to increase the fuel resistance to ageing is the use of antioxidant additives, but the chemical structure of biofuel poses a challenge to the effectiveness of these conventional additives. Another consequence of ageing, totally neglected in the literature, is the possibility of new pollutants during aged fuel combustion. The BioSCOPE project aims to better understand the oxidation of (bio)fuels in the liquid phase and its consequences on combustion efficiency and pollutants, in order to predict and ultimately control these synergistic processes, with new antioxidants, throughout the entire use chain, from storage to combustion. To achieve this goal, innovative experimental and simulation theories and tools must be developed, and their elaboration is at the core of the work of the project.
The BioSCOPE project comprises 4 Work Packages (WPs), designed to understand and simulate the ageing of fuels and biofuels, and their blends, as well as the effect of this ageing on combustion. The work carried out since the start of the project aims to develop and validate new experimental and numerical methods to achieve this objective, which remains a considerable challenge at the present time. The first work package is devoted to experiments, while work packages 2 to 4 are dedicated to the development of software capable of simulating the aging and combustion of these fuels. The challenges and work carried out to meet them, during the first 30 months of the project, are detailed below for the experimental and numerical parts, respectively.
WP1's goal is to developp and validate the experimental setup. The main challenge in studying the chemistry of (bio)fuel oxidation reactions and their rates in the liquid phase is to decouple physical effects from chemical effects. Liquid-phase oxidation kinetics experiments need to be free from a greater number of physical effects than gas-phase experiments. A well-defined reactor for liquid-phase chemical oxidation kinetics must be homogeneous in temperature and composition (by agitation), and not be limited by the rate of oxygen transfer from gas to liquid. The reactor must therefore be capable of withstanding high oxygen pressures and high temperature conditions, up to 200°C. All these conditions must be achieved in a reactor made of a chemically inert material, whose walls do not catalyze fuel reactions and for long residence times in the reactor, typically hours. we have developed a reactor capable of fulfilling all the conditions for being kinetically well defined using the tools of micro-fluidics. A novel microchannel reactor made of silicon-glass was developed in the project. It can reach temperatures of up to 227°C and pressures of up to 40 bar. Within the reactor, a steady segmented gas-liquid flow (alternating liquid slugs and gas bubbles) is established, ensuring homogeneous mixing in the liquid sluge and no oxygen transfer limitations. An important challenge is that residence times of several hours can be achieved on a 10 cm diameter silicon chip. Analysis of oxidation products is achieved in situ with Raman spectroscopy through the glass wafer sealing the reactor and at the exit of the reactor with HPLC. Since the main oxidation products, hydroperoxides, are fragile and cannot be measured by classical gas chromatography (GC), we have developed a unique method in the literature, based on high-performance liquid chromatography (HPLC) coupled to a post-column reactor where hydroperoxides are transformed into diiodine, which is detected by UV-visible absorption. This rapid method enables us to analyze separate the hydroperoxides and the fractions of fuel aged and analyze them with micro-GC.
The experimental set-up design enables the use of detailed kinetic mechanisms to simulate the oxidation process. Work packages 2 to 4 are devoted to the rigorous development of these kinetic mechanisms. In WP2 and WP3 new theories and numerical tools have been develop for the high-throughput calculations of accurate thermodynamic and kinetic data for liquid phase oxidation models. To rigorously simulate the oxidation of gas or liquid fuels, models that contain hundreds of species (molecules and radicals) and thousands of elementary reactions, are needed. The data required in these models are the thermochemical data for each species and the rate constants for each reaction. These thousands of data require fast and reliable estimation methods, as it is impossible to obtain so much experimental data, over wide temperature and pressure ranges. For gas-phase kinetic models, methods are available that can rapidly generate accurate thermochemical data. It is possible to adapt them to the liquid-phase by using a solvation free energy correction. The main challenge of this adaptation lies in the fact that, while the gas-phase data do not depend on the medium in which the combustion takes place, these quantities change for each (bio)fuel or their mixtures in the liquid phase. Since data are scarce in the literature, this issue is a major bottleneck for the detailed simulations of ageing of fuels. We overcame this challenge combining the power of equation of states to describe thermodynamic quantities in the liquid, gas and supercritical phases, with that of quantum chemistry for its ability to apply to free radicals. Our methodology avoids the computation times that would be impossible if quantum chemical methods were to be applied to thousands of reactions. Prior to the BioSCOPE project, it was not possible to apply an equation of state to a short-lived species such as a radical. The work carried out on thermochemical data has been extended to transition states and enables us to calculate instantaneously, in a way unique in the literature, the thermochemical and kinetic data of molecules and radicals in a pure or blended liquid or supercritical (bio)fuel, as a function of temperature. These methods are currently being coded for coupling to our automatic generator of detailed gas-phase combustion kinetic mechanisms (WP4). This will allow the software to generate models for combustion and oxidation of liquids. As the methods from WP2 and WP3 correct the gas phase data, the latter need to be highly accurate. We have developed a method called tabulated model transition states (TMTS), which enables us to extrapolate rate constants calculated very precisely on small chemical structures, with computationally time-consuming quantum chemical calculations, to large chemical structures for which such calculations are virtually impossible. This new method implies to account for all the possible combinations that can occur in the model transition states, for a given type of reaction, and leads to a very large number of cases that cannot be treated by hand. In this work, we have created a code that automatically generate the TS structures, creates the input files of a quantum chemistry code and extract the relevant data to compute the rate constants.
WP1's goal is to developp and validate the experimental setup. The main challenge in studying the chemistry of (bio)fuel oxidation reactions and their rates in the liquid phase is to decouple physical effects from chemical effects. Liquid-phase oxidation kinetics experiments need to be free from a greater number of physical effects than gas-phase experiments. A well-defined reactor for liquid-phase chemical oxidation kinetics must be homogeneous in temperature and composition (by agitation), and not be limited by the rate of oxygen transfer from gas to liquid. The reactor must therefore be capable of withstanding high oxygen pressures and high temperature conditions, up to 200°C. All these conditions must be achieved in a reactor made of a chemically inert material, whose walls do not catalyze fuel reactions and for long residence times in the reactor, typically hours. we have developed a reactor capable of fulfilling all the conditions for being kinetically well defined using the tools of micro-fluidics. A novel microchannel reactor made of silicon-glass was developed in the project. It can reach temperatures of up to 227°C and pressures of up to 40 bar. Within the reactor, a steady segmented gas-liquid flow (alternating liquid slugs and gas bubbles) is established, ensuring homogeneous mixing in the liquid sluge and no oxygen transfer limitations. An important challenge is that residence times of several hours can be achieved on a 10 cm diameter silicon chip. Analysis of oxidation products is achieved in situ with Raman spectroscopy through the glass wafer sealing the reactor and at the exit of the reactor with HPLC. Since the main oxidation products, hydroperoxides, are fragile and cannot be measured by classical gas chromatography (GC), we have developed a unique method in the literature, based on high-performance liquid chromatography (HPLC) coupled to a post-column reactor where hydroperoxides are transformed into diiodine, which is detected by UV-visible absorption. This rapid method enables us to analyze separate the hydroperoxides and the fractions of fuel aged and analyze them with micro-GC.
The experimental set-up design enables the use of detailed kinetic mechanisms to simulate the oxidation process. Work packages 2 to 4 are devoted to the rigorous development of these kinetic mechanisms. In WP2 and WP3 new theories and numerical tools have been develop for the high-throughput calculations of accurate thermodynamic and kinetic data for liquid phase oxidation models. To rigorously simulate the oxidation of gas or liquid fuels, models that contain hundreds of species (molecules and radicals) and thousands of elementary reactions, are needed. The data required in these models are the thermochemical data for each species and the rate constants for each reaction. These thousands of data require fast and reliable estimation methods, as it is impossible to obtain so much experimental data, over wide temperature and pressure ranges. For gas-phase kinetic models, methods are available that can rapidly generate accurate thermochemical data. It is possible to adapt them to the liquid-phase by using a solvation free energy correction. The main challenge of this adaptation lies in the fact that, while the gas-phase data do not depend on the medium in which the combustion takes place, these quantities change for each (bio)fuel or their mixtures in the liquid phase. Since data are scarce in the literature, this issue is a major bottleneck for the detailed simulations of ageing of fuels. We overcame this challenge combining the power of equation of states to describe thermodynamic quantities in the liquid, gas and supercritical phases, with that of quantum chemistry for its ability to apply to free radicals. Our methodology avoids the computation times that would be impossible if quantum chemical methods were to be applied to thousands of reactions. Prior to the BioSCOPE project, it was not possible to apply an equation of state to a short-lived species such as a radical. The work carried out on thermochemical data has been extended to transition states and enables us to calculate instantaneously, in a way unique in the literature, the thermochemical and kinetic data of molecules and radicals in a pure or blended liquid or supercritical (bio)fuel, as a function of temperature. These methods are currently being coded for coupling to our automatic generator of detailed gas-phase combustion kinetic mechanisms (WP4). This will allow the software to generate models for combustion and oxidation of liquids. As the methods from WP2 and WP3 correct the gas phase data, the latter need to be highly accurate. We have developed a method called tabulated model transition states (TMTS), which enables us to extrapolate rate constants calculated very precisely on small chemical structures, with computationally time-consuming quantum chemical calculations, to large chemical structures for which such calculations are virtually impossible. This new method implies to account for all the possible combinations that can occur in the model transition states, for a given type of reaction, and leads to a very large number of cases that cannot be treated by hand. In this work, we have created a code that automatically generate the TS structures, creates the input files of a quantum chemistry code and extract the relevant data to compute the rate constants.
The state of the art has been exceeded in many experimental and modeling aspects of the project. These significant advances were necessary to meet the scientific challenges to the study of (bio)fuel aging and its consequences on combustion.
From an experimental point of view, the development of a set-up where the oxidation of fuel in the liquid phase can be modeled as an ideal reactor that is chemically inert, homogeneous in temperature and concentration, not limited by the mass transfer of oxygen gas, capable of reaching high temperatures and pressures, and with long residence times is a result unprecedented in the literature. This reactor takes advantage of the benefits of micro-fluidics to achieve these goals, and the maintenance of Taylor gas-liquid flow over residence times in hours in a microchannel of over 3 m etched on a 10 cm diameter wafer, with the use of reactive oxygen in large quantities, constitutes a breakthrough that goes beyond the state of the art. The development of an analytical tool - an HPLC chain coupled to a post-column reactor - makes it possible to analyze very finely and precisely, rapidly and over a wide range of concentrations, the main oxidation products in the liquid phase, namely hydroperoxides. These compounds are fragile and cannot be quantified by conventional methods such as gas chromatography. Our HPLC analytical device, designed by the Bioscope project team, is unique in the world and outperforms all other methods in the literature. By the end of the project, further advances beyond the state of the art are expected on the experimental side, such as the coupling of the liquid oxidation device to combustion experiments, which has never been achieved in the literature.
On the modeling side, which aims to develop methods for generating liquid-phase oxidation kinetic models from combustion models, several advances beyond the state of the art have been made. We have developed a novel method capable of calculating liquid-phase thermochemical data for molecules and radicals that couples an equation of state and a solvating model from theoretical chemistry. The predictive power of equations of state has been applied for the first time in the literature to transient species such as free radicals and transition states. This approach generates highly accurate data at high throughput. The use of the equation of state makes it possible to calculate the thermochemical quantities of molecules and radicals in a mixture of (bio)fuels in the liquid, real gas or supercritical phase, which is a major and unprecedented advance. This approach will be extended to transition states by the end of the project. A Python code is being developed to perform these calculations, and will be coupled to our automatic kinetic model generation software, which will allow precise correction of all generated gaseous thermos-kinetic data. Work on improving the estimation of gas kinetic data currently generated by the software, using the tabulated model transition state method (TSMT), also represents a breakthrough that goes beyond the state of the art. The TSMT approach allows variations in the chemical structures of the reactants to be taken into account when estimating rate constants, and achieves unprecedented accuracy by enabling the results of high-level quantum chemistry calculations to be extrapolated to larger structures, for which high-level quantum calculations are impossible. This TSMT approach has been developed and applied to alkyl radical isomerization reactions, and will be extended to other types of reactions encountered in radical oxidation mechanisms: initiation, metathesis, beta-scission. All these advances will be implemented in the kinetic mechanism generation software, which will be capable, for the first time in the literature, of producing detailed kinetic models of combustion and oxidation in the liquid or supercritical phase, of pure (bio)fuels or mixtures, and of explicitly including antioxidant reactions. The novel models generated will be validated on the original experimental data obtained in the project.
From an experimental point of view, the development of a set-up where the oxidation of fuel in the liquid phase can be modeled as an ideal reactor that is chemically inert, homogeneous in temperature and concentration, not limited by the mass transfer of oxygen gas, capable of reaching high temperatures and pressures, and with long residence times is a result unprecedented in the literature. This reactor takes advantage of the benefits of micro-fluidics to achieve these goals, and the maintenance of Taylor gas-liquid flow over residence times in hours in a microchannel of over 3 m etched on a 10 cm diameter wafer, with the use of reactive oxygen in large quantities, constitutes a breakthrough that goes beyond the state of the art. The development of an analytical tool - an HPLC chain coupled to a post-column reactor - makes it possible to analyze very finely and precisely, rapidly and over a wide range of concentrations, the main oxidation products in the liquid phase, namely hydroperoxides. These compounds are fragile and cannot be quantified by conventional methods such as gas chromatography. Our HPLC analytical device, designed by the Bioscope project team, is unique in the world and outperforms all other methods in the literature. By the end of the project, further advances beyond the state of the art are expected on the experimental side, such as the coupling of the liquid oxidation device to combustion experiments, which has never been achieved in the literature.
On the modeling side, which aims to develop methods for generating liquid-phase oxidation kinetic models from combustion models, several advances beyond the state of the art have been made. We have developed a novel method capable of calculating liquid-phase thermochemical data for molecules and radicals that couples an equation of state and a solvating model from theoretical chemistry. The predictive power of equations of state has been applied for the first time in the literature to transient species such as free radicals and transition states. This approach generates highly accurate data at high throughput. The use of the equation of state makes it possible to calculate the thermochemical quantities of molecules and radicals in a mixture of (bio)fuels in the liquid, real gas or supercritical phase, which is a major and unprecedented advance. This approach will be extended to transition states by the end of the project. A Python code is being developed to perform these calculations, and will be coupled to our automatic kinetic model generation software, which will allow precise correction of all generated gaseous thermos-kinetic data. Work on improving the estimation of gas kinetic data currently generated by the software, using the tabulated model transition state method (TSMT), also represents a breakthrough that goes beyond the state of the art. The TSMT approach allows variations in the chemical structures of the reactants to be taken into account when estimating rate constants, and achieves unprecedented accuracy by enabling the results of high-level quantum chemistry calculations to be extrapolated to larger structures, for which high-level quantum calculations are impossible. This TSMT approach has been developed and applied to alkyl radical isomerization reactions, and will be extended to other types of reactions encountered in radical oxidation mechanisms: initiation, metathesis, beta-scission. All these advances will be implemented in the kinetic mechanism generation software, which will be capable, for the first time in the literature, of producing detailed kinetic models of combustion and oxidation in the liquid or supercritical phase, of pure (bio)fuels or mixtures, and of explicitly including antioxidant reactions. The novel models generated will be validated on the original experimental data obtained in the project.