Graphene can be synthesized entirely in the gas phase within microwave-assisted reactors operating at atmospheric pressure. Although these systems are sustained by plasmas with extremely high local temperatures, graphene formation occurs downstream where chemical kinetics govern molecular growth. A one-dimensional plug-flow model coupled with a sectional aerosol framework is used to evaluate how different detailed gas-phase chemical mechanisms influence graphene formation from an ethanol precursor. Five mechanisms commonly used for polycyclic aromatic hydrocarbon (PAH) chemistry—ABF, DLR, CALTECH, KAUST, and CRECK—are compared with experimental measurements of graphene yield and Feret diameter. The mechanisms predict very different onsets of graphene formation. Notably, the KAUST mechanism, despite its unrealistic assumption of irreversible PAH growth, reproduces experimental trends most closely. This outcome suggests that the plasma environment maintains a chemically frozen state where large PAHs behave as effectively irreversible species. Comparison between kinetic and equilibrium calculations confirms that PAH concentrations in the post-plasma region exceed equilibrium predictions by 18–20 orders of magnitude. Because the model itself does not include plasma physics, this kinetic–equilibrium disparity provides indirect, but not exclusive, evidence that plasma-driven processes push the system far from chemical equilibrium and enable the rapid molecular growth required for graphene formation. These findings explain why equilibrium models fail to predict graphene synthesis and demonstrate that model discrepancies can expose hidden nonequilibrium mechanisms.

Soot formation in combustion processes significantly impacts pollutant emissions and energy efficiency. Transient laminar flames, characterized by periodic forcing, provide a controlled environment to study the dynamics of soot formation and oxidation. While previous studies have explored soot behaviour in steady and forced flames, the interplay between different forcing mechanisms and soot formation processes remains insufficiently understood. Additionally, advanced modelling approaches are needed to capture the transient evolution of soot in such systems. This study aims to investigate soot formation mechanisms in transient laminar diffusion flames under periodic forcing, leveraging detailed experiments and numerical simulations to enhance understanding and predictive capabilities. While prior studies have examined soot formation in acoustically forced flames, this study expands the analysis by incorporating mechanical oscillations, a forcing mode that simultaneously perturbs fuel and oxidizer. This work identifies distinct soot formation pathways by comparing different forcing mechanisms and introduces mixture fraction histories as a new approach to capturing transient soot dynamics in non-steady flames. Forcing significantly influenced soot formation, with acoustic and mechanical perturbations enhancing surface growth processes. Numerical simulations captured trends in soot dynamics, revealing key differences between forcing mechanisms. Temporal soot histories provided new insights into precursor availability and formation rates in non-steady environments. The results underscore the value of transient laminar flames as proxies for studying turbulent combustion and offer a robust foundation for improving soot models in complex environments. These findings also inform the development of machine learning tools for predictive combustion modelling.

Validation is a vital part of any computational fluid dynamics study. Validation is done by comparing the computed results to the experimental measurements performed in a known configuration. In sooting flames, such a comparison is non-trivial since the measurements have a wide range of uncertainty due to difficulties in directly measuring soot volume fractions. This work introduces a different way to verify the computations using a software package called SootImage. In this proposed methodology, comparisons are not made directly to the computed properties of interest (soot volume fraction and temperature). Instead, a post-processing procedure is performed to obtain an image of the flame based on the computed properties. This reconstructed image is compared to an actual image of the flame being studied. The algorithm of the image reconstruction utilized within SootImage is presented in detail. Finally, the usage of SootImage is demonstrated on a co-flowing, laminar ethylene/air diffusion flame.

This study presents a numerical approach for simulating the internal morphology of carbon black particles formed in a pyrolysis reactor. The simulation process involves a two-step approach using population balance models (PBM) and detailed population balance models (DPBM), solved via sectional and stochastic methods, to simulate the arrangement of primary particles within aggregates to allow determination of their fractal dimension (FD). The outcome is a novel introduction of simulating real flow reactors that for the first time provides the fractal dimension of particles as an output, rather than an assumed input. The results of this study have practical implications for optimizing the synthesis processes in carbon black production. The effects of various production parameters, including aggregation efficiency, temperature, pressure, and acetylene concentrations, on the fractal dimension values of carbon black particles are examined. It is observed that higher temperatures lead to the formation of larger fractal shapes with lower fractal dimensions and larger primary particle diameters. Moreover, increased reactor pressure and higher aggregation efficiency enhance the formation of carbon black aggregates, but also have a time-based effect with higher compactness at longer residence times. The time-based effect reveals the importance of sintering, where high loads of small particles enhance the overall sintering of the aggregates. These findings provide insights into the interplay between temperature, pressure, and particle morphology, highlighting the dynamic nature of carbon black nanoparticles and their response to synthesis process conditions.

This paper examines the evolution of particle size and morphology of plasma-synthesized carbonaceous nanoparticles for various reactants, reactant concentrations, reactor temperatures, and positions within the reactor. Aerosol particles were characterized at various positions downstream from the plasma zone by spatially-resolved thermophoretic sampling and transmission electron microscopy, Raman spectroscopy, and by in situ time-resolved laser-induced incandescence. Depending on the carbon-bearing reactant (ethanol or toluene) and its concentration, either pure few-layer graphene (FLG) or soot-like particles, or a mixture of both, were generated. The initial carbon nucleation has been found to commence less than 12.4 cm downstream from the plasma nozzle. In the case of FLG formation, particles show an increasing level of crumpling with increasing distance downstream from the plasma zone. The process was numerically simulated with a 1D plug-flow reactor model employing a joint population balance model for graphitic and amorphous-like particles (FLG and soot-like, respectively). The simulation includes modified versions of inception, hydrogen-abstraction-carbon-addition (HACA), and polycyclic-aromatic hydrocarbon (PAH) adsorption models adapted from a pre-existing soot model. The simulations showed that the ratio of HACA/PAH adsorption determined by post-plasma C 2 H 2 is the main factor that affects the soot-to-FLG ratio. This is also consistent with experimental observations as reducing the rate at which carbon precursor is supplied to the reactor leads to less C 2 H 2 in the post-plasma zone, which results in less PAH adsorption, and consequently suppressed soot formation in favor of FLG synthesis. (c) 2023 The Combustion Institute. Published by Elsevier Inc. All rights reserved.

The effect of ammonia (NH3) addition on soot growth and inception is investigated in laminar co-flow NH3-ethylene (C2H4) diffusion flames. By comparing C2H4 flames with increasing% of NH3 by volume for the same carbon flow and flame height, the impact of NH3 addition on soot formation is identified. Experimental measurements of flame temperature, soot volume fraction, and primary particle sizes and number densities are compared with a numerical two-dimensional co-flow flame model to evaluate the contribution of NH3 to soot reduction in C2H4 flames. Experimental and numerical results show a significant reduction in the amount of soot formed and the diameter and the number of the primary particles suggesting reduced rates of soot growth and inception due to NH3 addition. The numerical model provides the understanding of how NH3 inhibits soot formation in the flame by analyzing the soot formation pathways and investigating the concentration of relevant gas phase species in the flame. While the impact of NH3 addition on the growth of soot particles is not fully captured by the model, the NH3’s reductive effect on the number of soot particles formed is well predicted. The results suggest that NH3 reduces soot by suppressing the concentration of methyl radicals which are responsible for forming odd numbered carbon species such as propargyl and cyclopentadiene. These species are found to be important contributors to the formation of large polycyclic aromatic compounds which are precursors of soot inception.

Physical clustering of Polycyclic Aromatic Hydrocarbon (PAH) molecules is subject to many studies as a part of understanding the complex clustering behavior and formation of carbon nanoparticles, such as soot and carbon black. A new fully reversible PAH clustering (FRPC) model is introduced to account for the full reversibility of PAH inception and molecular adsorption. Instead of assessing the performance of the new FRPC in a flame simulation as is typically done, a simplified system is utilized such that PAH clustering can be isolated from all other particle processes. The chosen simplified system consists of a 0D homogeneous reactor of PAHs at certain temperatures and has previously been investigated using molecular dynamics. The results from the FRPC model display excellent performance when compared to the molecular dynamic results. It is found that reversible rates play a significant role for smaller PAHs and at higher temperatures where the conditions are closer to an equilibrium between the gas-phase and condensed-phase PAHs. It is demonstrated that efficiency-based models are ill-suited to model PAH clustering due to the lack of explicitly modeling the reverse rates of clustering. The newly developed FRPC model represents the first step in a piece-wise development of a carbon nanoparticle formation model.

Fuel dilution is commonly used to reduce soot emissions from combustion engines or control carbon black morphology during fuel pyrolysis. Dilution has both physical and chemical effects on soot formation; however, the impact of fuel dilution on each soot formation process has not been fully understood yet. This study investigates the differing effects of dilution with H2 and N2 on particle inception and surface growth by hydrogen abstraction carbon addition (HACA) and polycyclic aromatic hydrocarbon (PAH) adsorption in ethylene/air laminar coflow diffusion flames. Applicable correlations are introduced to predict the conversion of carbon to soot to an accuracy of 96.5% at different dilution ratios. Both diluents result in significant reductions in HACA and PAH adsorption, up to 42% and 68%, respectively, at 40% dilution, with greater reductions for dilution with N2. For replacement cases, more reduction comes from a considerable change in the flame length, resulting in reduced high-temperature particle residence time. Combined, these effects significantly reduce HACA and PAH adsorption by up to 85% and 60%, respectively, for mixed replacement. Reduced acetylene and PAH concentrations for both gases, respectively, reduce HACA and PAH adsorption rates. Earlier in the flame, reduced PAH formation in the inception zone of the flame results in lower inception rates, producing fewer primary particles for subsequent growth processes. Comparing the two diluents, N2 reduces acetylene and PAH concentrations by 12% and 18%, respectively, more than H2, and delays PAH formation in the flame, which delays inception and surface growth.

The present work examines two-dimensional effects during nanoparticle formation in laminar flow reactors using two approaches: a 2D model and a simplified 1D model. The investigated cases consist of carbon nanoparticle synthesis via ethylene pyrolysis in laminar flow. The simplified 1D simulations are conducted using the 1D plug flow solver Nano PFR, while the 2D simulations have been performed using CoFlame in an axisymmetric configuration. Both models utilize a sectional population balance model coupled with gas-phase chemistry to simulate detailed particle formation processes. The effects of radial diffusion of the 2D approach are analyzed in a reference case of an isothermal reactor by comparing a diffusion model against a no-diffusion model. Results show that radial diffusion sufficiently mixes the gas phase species along the radius to approximate plug flow. While radial diffusion effects are apparent in particle formation, it is not nearly enough to develop a uniform radial profile for the primary particle and agglomerate number densities and volume fraction. Next, experimental data from the literature are simulated, examining the suitability, differences, and limitations between the 1D and 2D domains. Both models perform similarly in prediction accuracy for the gas-phase species, including fuel pyrolysis and PAH formation. Moreover, by integrating particle mass across several regions (i.e. radial integration approach), the 1D model can achieve accuracy on par with the 2D model for particle formation predictions. Both models can adequately capture SMPS measurements for particle volume fraction, number densities, and size distributions. Ultimately, within the range of experimental conditions typically found in current laminar flow reactor studies for carbon nanoparticle formation, a modified 1D approach can sufficiently describe the gas-phase chemistry and particle formation processes with lower computational costs than the 2D approach. (c) 2021 The Combustion Institute. Published by Elsevier Inc. All rights reserved.

In the last two decades several mechanisms have been proposed for the growth of PAHs during combustion. Studies have suggested that the growth of PAHs of different structures should be analyzed carefully to improve our understanding of PAH formation during combustion. In this study, a wide of range of PAHs of different structures have been measured along the centreline of a coflow diffusion flame of ethylene using GC/MS, allowing for a previously not possible detailed analysis of potential formation routes. The discussed potential global empirical pathways can help in the development of accurate and detailed chemical mechanisms in the future. The experimental results show that the difference in reactivity of the sites of a PAH where growth takes place limits the formation of benzenoid PAHs at high temperatures. This difference in reactivity also leads to the formation of PAHs with a five-membered ring at high temperatures. The recombination of PAH radicals may not be favoured in this flame which explains the small amounts of alkyl-bridged PAHs at high temperatures. PAHs with an alkyl sidechain are favoured only along the free edge. Alkyl substituted PAHs are formed in low amounts and decrease with an increase in temperature. Methylene-bridged PAHs can lead to faster growth of large PAHs and become significant at intermediate temperatures. The target flame has also been simulated using state-of-the-art models. The simultaneous measurements of both PAHs and soot allowed for a comprehensive assessment of the chemical mechanisms and soot aerosol dynamics. The assessment suggests that both the chemical mechanisms and the soot aerosol dynamics need further improvements. The results confirm the importance of accounting for PAH structure when developing kinetic mechanisms and their coupled soot formation models. A complete database comprising of PAHs, soot and temperature is generated for future model validations.

The effects of elevated reactant temperatures on soot formation in a laminar coflow ethylene flame were experimentally and numerically investigated. Ethylene flames at the reference reactant temperatures (both air and fuel), Tr, of 300K (LT), 473K (MT), 673K (HT), and 713K (UHT) were established. In the experiment, soot volume fractions (fv), primary particle diameters (dp), and soot (flame) temperatures (TF) were measured. The flames were also simulated by the CoFlame code with the Conjugate Heat Transfer (CHT) condition. The experimental results show that elevating Tr positively affects soot formation. The increase in the maximum fv is greater on the wing pathline (∼1.9 times) than on the centerline (∼1.2 times) when the adiabatic flame temperature increases by ∼100K. An analysis of the experimental and numerical data suggests that soot formation is promoted by enhanced soot surface growth. The numerical simulation reveals that PAH (polycyclic aromatic hydrocarbon) adsorption, which is a function of PAH concentration, becomes important at high Tr as its mass contribution increases from ∼50% to ∼70%. This may be attributed to early fuel pyrolysis within the fuel tube.

There are several processes to occur in soot formation and destruction for which some in the growth regime require a better understanding. In this work, a consistent surface reactivity model, developed in recent years, has been implemented across various sooting laminar flames at varying pressures. The surface reactivity function proposed by Khosousi and Dworkin [1] is employed in the present study. It is based on the temperature history of soot particles. As the functionally dependent model has been derived and validated for atmospheric pressure flames, there are discrepancies between simulation and experiment that can be observed as pressures vary. One reason for these discrepancies could be explained by the fact that chemical reaction rates for the soot growth mechanism at atmospheric combustion do not adequately characterize the kinetics at higher pressures. Based on a recently published study [2], the elementary reaction rates that compose the Hydogen-Addition-Carbon-Abstraction soot surface growth mechanism depend on pressure and an empirical pressure scaling factor to account for this pressure dependence has been introduced. It has been determined that after applying the new empirical pressure scaling factor for the soot growth mechanism, the performance of the functionally dependent surface reactivity model improves in the wing regions of the flame for pure-ethylene flames; however, there is minimal change on the wings for the nitrogen-diluted flames. Additionally, the quantity for soot concentration along the centerline of all flames is nearly independent of the surface reactivity model chosen and needs further investigation. For the flames investigated, it is concluded that pressure dependent HACA rates do not alleviate all discrepancies between the numerical and experimental results and that further work is required.

Exhaust After-Treatment (EAT) systems are necessary for automotive powertrains to meet stringent emission standards. Computational modelling has been applied to aid designing EAT systems. Models with global kinetic mechanisms are often used in practice, but they cannot accurately predict the behaviour of after-treatment devices under a wide range of conditions. In this study, a numerical EAT model with rigorous treatment of the catalytic chemistry is proposed to investigate the impact of the configuration of individual devices in the EAT system; one of the key design decisions. The performance of the proposed model is first critically assessed against experimental and simulation data from the literature before being applied to design a multi-device EAT system for a diesel engine. The target EAT system is composed of a diesel oxidation catalyst (DOC), an ammonia-based selective catalytic reduction (NH3-SCR) device and a diesel particulate filter (DPF). The steady state behaviour of various EAT designs under operating conditions across the engine map are examined. The DOC-DPF-SCR layout is found to be more beneficial than the alternative DOC-SCR-DPF for the specific engine studied. Furthermore, the DPF-front system is more robust with respect to changes in emission regulations. Flux analysis is applied to study the chemical interaction in the SCR and explain the disadvantage of the SCR-front system. In addition, it is demonstrated in the study that future catalyst investigations should consider more realistic feed compositions.

I confirm that I request a copy of this work for my individual use only and not on behalf of someone else. I understand that I am not authorised to quote, translate, publish, archive, or share with a third party the whole or any part of the work without obtaining explicit consent from the author

The data-set contains the soot particle size distribution along the centreline of n-heptane/toluene fed co-flow diffusion flames. The samples from within the flame were taken with a newly developed quartz sampling probe (technical drawing attached) and passed to a Differential Mobility Spectrometer (DMS500).

A new model has been developed to describe the size-dependent effects that are responsible for transient particle mass (PM) and particle number (PN) emissions observed during experiments of the active regeneration of Diesel Particulate Filters (DPFs). The model uses a population balance approach to describe the size of the particles entering and leaving the DPF, and accumulated within it. The population balance is coupled to a unit collector model that describes the filtration of the particles in the porous walls of the DPF and a reactor network model that is used to describe the geometry of the DPF. Two versions of the unit collector model were investigated. The original version, based on current literature, and an extended version, developed in this work that includes terms to describe both the non-uniform regeneration of the cake and thermal expansion of the pores in the DPF. Simulations using the original unit collector model were able to provide a good description of the pressure drop and PM filtration efficiency during the loading of the DPF, but were unable to adequately describe the change in filtration efficiency during regeneration of the DPF. The introduction of the extended unit collector description enabled the model to describe both the timing of particle breakthrough and the final steady filtration efficiency of the hot regenerated DPF. Further work is required to understand better the transient behaviour of the system. In particular, we stress the importance that future experiments fully characterise the particle size distribution at both the inlet and outlet of the DPF.

Nucleation of incipient soot or carbon black nanoparticles has significant scientific and industrial importance because it influences the particle size distribution, morphology and composition, hence its health and environmental impact as well as functional properties. Reversible Polycyclic Aromatic Hydrocarbon (PAH) Clustering (RPC) with van der Waals forces (vdW) was proposed recently as opposed to Irreversible PAH Clustering (IPC) for soot nucleation (Eaves et al., Proc. Combust. Inst, 35 (2015) 1787) to relax the assumption of stable dimer formation with physical bonds at flame temperatures. Here, the necessity of considering chemical bond formation between PAHs in a dimer for reducing soot nucleation reversibility in ethylene coflow diffusion flames with a wide range of nitrogen dilution ratios is demonstrated. An RPC model with Chemical Bond Formation (RPC-CBF) is developed and its performance is compared to those of RPC and IPC models. Only the RPC-CBF model with low reversibility for PAH addition on the surface of soot primary particles can predict soot volume fraction, average primary particle diameter, primary particle number density and the bimodal size distribution of soot agglomerates on the flame centerline within experimental uncertainty. While the IPC model overpredicts soot concentrations by a factor of five with increased dilution, the RPC model underpredicts it by more than two orders of magnitude for the flame with 68% nitrogen dilution. The RPC model fails to predict the observed bimodality of agglomerate particle size distribution in flames with low dilution because its predicted nucleation rate is much weaker compared to that of growth. Accounting for chemical bond formation for dimers is essential to form enough nuclei with increasing dilution and temperature. Also, low reversibility for PAH addition is required to predict the balance between nucleation and surface growth, hence the average size of soot primary particles. Crown Copyright (C) 2018 Published by Elsevier Inc. on behalf of The Combustion Institute. All rights reserved.

The evolution of primary soot particles is studied experimentally and numerically along the centreline of a co-flow laminar diffusion flame. Soot samples from a flame fueled with C2H4 are taken thermophoretically at different heights above the burner (HAB), their size and nano-structure are analysed through TEM. The experimental results suggest that after inception, the nascent soot particles coagulate and coalesce to form larger primary particles (similar to 5 to 15 nm). As these primary particles travel along the centreline, they grow mainly due coagulation and condensation and a layer of amorphous hydrocarbons (revealed by HRTEM) forms on their surface. This amorphous layer appears to promote the aggregation of primary particles to form fractal structures. Fast carbonisation of the amorphous layer leads to a graphitic-like shell around the particles. Further graphitization compacts the primary particles, resulting in a decrease of their size. Towards the flame tip the primary particles decrease in size due to rapid oxidation. A detailed population balance model is used to investigate the mechanisms that are important for prediction of primary particle size distributions. Suggestions are made regarding future model development efforts. Simulation results indicate that the primary particle size distributions are very sensitive to the parameterization of the coalescence and particle rounding processes. In contrast, the average primary particle size is less sensitive to these parameters. This demonstrates that achieving good predictions for the average primary particle size does not necessarily mean that the distribution has been accurately predicted. (C) 2018 The Combustion Institute. Published by Elsevier Inc. All rights reserved.

Model guided application (MGA) combining physico-chemical internal combustion engine simulation with advanced analytics offers a robust framework to develop and test particle number (PN) emissions reduction strategies. The digital engineering workflow presented in this paper integrates the kinetics & SRM Engine Suite with parameter estimation techniques applicable to the simulation of particle formation and dynamics in gasoline direct injection (GDI) spark ignition (SI) engines. The evolution of the particle population characteristics at engine-out and through the sampling system is investigated. The particle population balance model is extended beyond soot to include sulphates and soluble organic fractions (SOF). This particle model is coupled with the gas phase chemistry precursors and is solved using a sectional method. The combustion chamber is divided into a wall zone and a bulk zone and the fuel impingement on the cylinder wall is simulated. The wall zone is responsible for resolving the distribution of equivalence ratios near the wall, a factor that is essential to account for the formation of soot in GDI SI engines. In this work, a stochastic reactor model (SRM) is calibrated to a single-cylinder test engine operated at 12 steady state load-speed operating points. First, the flame propagation model is calibrated using the experimental in-cylinder pressure profiles. Then, the population balance model parameters are calibrated based on the experimental data for particle size distributions from the same operating conditions. Good agreement was obtained for the in-cylinder pressure profiles and gas phase emissions such as NOx. The MGA also employs a reactor network approach to align with the particle sampling measurements procedure, and the influence of dilution ratios and temperature on the PN measurement is investigated. Lastly, the MGA and the measurements procedure are applied to size-resolved chemical characterisation of the emitted particles.