Floating Catalyst Chemical Vapor Deposition (FCCVD) is an innovative process for low-carbon hydrogen production and carbon sequestration in advanced materials. Unlike steam reforming of methane, which generates high carbon dioxide (CO2) emissions, FCCVD enables the direct pyrolysis of methane, producing in molecular hydrogen and solid carbon, avoiding the release of greenhouse gases. A key aspect of this technology is the possibility of converting carbon into carbon nanotubes (CNTs), materials with unique electrical, thermal and mechanical properties. Thanks to their high strength, rigidity and lightness, CNTs find applications in strategic sectors such as electronics, aerospace and automotive, contributing to improve structural and energy efficiency. This process also stands out for its scalability and continuous operation, thanks to the use of fluidized catalysts, which allow CNTs to form within the reactor and aggregate towards the outlet, where they can be continuously spun into fibers or films. The absence of direct CO2 emissions, combined with the production of high-tech materials, therefore makes the FCCVD method a promising strategy for the industrial production of turquoise hydrogen and advanced materials with broad application potential. Its implementation could mark a crucial step towards the de-carbonization of industrial processes and the transition to sustainable chemistry. Despite its potential, the industrialization of this method still faces significant challenges. The complex interplay between gas-phase reactions and catalytic growth processes makes it difficult to control the quality and yield of CNTs, limiting process predictability and efficiency. Although parameters such as temperature and the concentration of catalytic precursors (Fe, S) have been identified as determinants for the structure of CNTs, there is a lack of a reliable quantitative model that can correlate operating conditions with product yield and quality. The evolution of catalytic nanoparticles (NPs) represents a further critical issue, as it is strongly influenced by process variables. Their evolution and growth along the reactor makes the correlation between the size distribution of NPs and the final properties of CNTs complex, hindering process control and maximization of catalytic activity. The competition between CNT growth and amorphous carbon formation is a further limitation. Although the formation of carbon by-products is inevitable in FCCVD processes, their impact on the yield and quality of CNTs has not yet been quantitatively described. From a multi-scale modeling perspective, currently available kinetic models do not provide a fully integrated and comprehensive representation of the complex network of chemical processes governing the growth of catalytic nanoparticles and the solid carbonaceous product. Moreover, their validation remains limited, as it has not been conducted on a sufficiently large and representative dataset. This limitation is further aggravated by the scarcity of accurate and well-characterized quantitative data in the literature, which are essential for a deeper understanding of the phenomena involved in the process. Consequently, the lack of a robust data foundation hinders the development of reliable kinetic models capable of rigorously describing the intricate network of reactions and interactions that characterize the FCCVD process. These limitations highlight the need for innovative approaches to deepen the understanding of CNT growth mechanisms and facilitate process optimization for large-scale implementation. In this study, an integrated experimental and modelling approach was adopted to develop a \textit{semi-detailed kinetic model} to describe key phenomena observed in the FCCVD process. The aim is to provide a predictive tool capable of reproducing system behavior and assessing the impact of operating conditions on process performance, thus supporting industrial-scale optimization. From an experimental perspective, an extensive campaign was conducted at the \textit{IMDEA Materials Institute}, where systematic experiments were carried out to analyze the effect of operating conditions on CNT productivity and quality. Material characterization through Raman spectroscopy and transmission electron microscopy (TEM) on samples obtained within a sufficiently comprehensive operating window provided well-characterized quantitative data. These data supported the development of the kinetic framework and its comparison with model predictions, ultimately leading to the validation of the proposed mechanism. The developed kinetic framework, modular and hierarchical, consists of two main components. The first is a semi-detailed model for CNT growth, describing the catalytic decomposition of hydrocarbons on iron nanoparticles and the subsequent incorporation of carbon into the nanotube structure. The second is a heterogeneous model for pyrolytic carbon deposition, which accounts for the potential lateral growth of less organized carbon layers, triggered by the deposition of heavier gas-phase species on CNT walls. These two surface models, coupled with the gas-phase model previously developed and validated by the CRECK Modeling Group at Politecnico di Milano, provide a comprehensive description of the evolution of both gas and solid phases in the FCCVD process. The global kinetic framework, written with the Surface CHEMKIN formalism, has been implemented in a simplified reactor model, implemented in OpenSMOKE++, to perform process simulations and compare the modelling predictions with experimental data, validating the model on the basis of the total carbon yield and the crystallinity of the material (ratio ID/IG in the Raman spectra). In parallel, the role of iron nanoparticles, which are essential for the growth of CNTs, was studied. A dedicated model for the formation and evolution of NPs was developed, based on a discrete section approach (DSM), and the predictions of this model were compared with an exhaustive set of data collected in a dedicated experimental campaign where we measured the size distribution of NPs along the reactor using SMPS-DMA (Scanning Mobility Particle sizer and Differential Mobility Analyzer). The integration of these experimental and modelling approaches provided an estimate of the evolution of the catalytic surface along the process as a key parameter for FCCVD simulations. Future developments should focus on integrating the NP model within the overall framework to track the size distribution of catalytic nanoparticles in greater detail and directly correlate it to the characteristics of the synthesized CNTs. This improvement would enable a more comprehensive and predictive representation of nanotube growth, in line with experimental evidence. Moreover, a more exhaustive study in relation to the role of the catalytic precursors iron and sulfur, and their interactions with other process variables, is essential to improve the understanding of the synthesis mechanisms and further refine the proposed kinetic model. To the best of our knowledge, the developed kinetic framework represents the first semi-detailed model capable of predicatively correlating operating conditions with the productivity and structural quality of CNTs in thermo-catalytic methane pyrolysis. The experimental validation demonstrated the model's ability to satisfactorily describe the interactions between process variables and their effect on the properties of the produced material. This work can be considered a significant advance in the predictive modelling of FCCVD, providing concrete tools for process optimization with a view to industrial-scale implementation. The integration of future proposed developments may further strengthen the framework, contributing to innovation in the scalable and sustainable production of advanced carbon materials.
Il Floating Catalyst Chemical Vapor Deposition (FCCVD) è un processo innovativo per la produzione di idrogeno a basso impatto carbonico e il sequestro del carbonio in materiali avanzati. A differenza del reforming del metano con vapore, che comporta elevate emissioni di anidride carbonica (CO2), FCCVD consente la pirolisi diretta del metano, producendo idrogeno molecolare e carbonio solido, senza emissione diretta di gas serra. Un aspetto chiave di questa tecnologia è la possibilità di convertire il carbonio in nanotubi di carbonio (CNT), materiali dalle proprietà elettriche, termiche e meccaniche distintive. Grazie alla loro elevata resistenza, rigidità e leggerezza, i CNT trovano applicazione in settori strategici come elettronica, aerospazio e automobilistica, contribuendo a migliorare l’efficienza strutturale ed energetica. Questo processo si distingue inoltre per la sua scalabilità e continuità operativa, resa possibile dall’uso di catalizzatori fluidizzati, che favoriscono la formazione dei CNT lungo il reattore e la loro aggregazione verso l’uscita, dove possono essere filati continuamente sotto forma di fibre o film. L’assenza di emissioni dirette di CO2, unita alla produzione di materiali ad alto valore tecnologico, rende quindi il metodo FCCVD una strategia promettente per la produzione industriale di idrogeno turchese e di materiali avanzati con ampio potenziale applicativo. La sua implementazione potrebbe rappresentare un passo cruciale verso la decarbonizzazione dei processi industriali e la transizione verso una chimica sostenibile. Nonostante il suo potenziale, l’industrializzazione di questo metodo presenta ancora sfide significative. La complessa interazione tra reazioni in fase gas e processi di crescita catalitica rende difficile il controllo della qualità e della resa dei CNT, limitando la prevedibilità e efficienza del processo. Sebbene parametri come la temperatura e la concentrazione dei precursori catalitici (Fe, S) siano stati identificati come determinanti per la struttura dei CNT, manca un modello quantitativo affidabile in grado di correlare le condizioni operative con la resa e la qualità del prodotto. L’evoluzione delle nanoparticelle catalitiche (NP) rappresenta un’ulteriore criticità, poiché è fortemente influenzata dalle variabili di processo. La loro evoluzione e crescita lungo il reattore rende complessa la correlazione tra la distribuzione dimensionale delle NP e le proprietà finali dei CNT, ostacolando il controllo del processo e la massimizzazione dell'attività catalitica. La competizione tra crescita dei CNT e formazione di carbonio amorfo rappresenta un’ulteriore limitazione. Sebbene la formazione di sottoprodotti carboniosi sia inevitabile nei processi FCCVD, il loro impatto sulla resa e sulla qualità dei CNT non è ancora stato descritto quantitativamente. Dal punto di vista della modellazione multi-scala, i modelli cinetici attualmente disponibili non offrono una rappresentazione integrata e completa della complessa rete di processi chimici che governano la crescita delle nanoparticelle catalitiche e la formazione del prodotto solido carbonioso. Inoltre, la loro validazione risulta limitata, poiché non è stata condotta su un set di dati ampio e rappresentativo. Questa carenza è aggravata dalla scarsa disponibilità, in letteratura, di dati quantitativi accurati e ben caratterizzati, fondamentali per comprendere in modo approfondito i fenomeni coinvolti nel processo. Di conseguenza, l’assenza di una base dati solida compromette lo sviluppo di modelli cinetici affidabili, capaci di descrivere con rigore le reazioni e le interazioni che caratterizzano il FCCVD. Queste limitazioni sottolineano la necessità di approcci innovativi per approfondire la comprensione dei meccanismi di crescita dei CNT e facilitare l'ottimizzazione del processo in vista della sua implementazione su larga scala. In questo studio è stato adottato un approccio integrato sperimentale e modellistico volto allo sviluppo di un modello cinetico semi-dettagliato per descrivere i fenomeni chiave osservati nel processo FCCVD. L’obiettivo è fornire uno strumento predittivo in grado di riprodurre il comportamento del sistema e valutare l’impatto delle condizioni operative sulle performance del processo, supportando così l’ottimizzazione su scala industriale. Dal punto di vista sperimentale, è stata condotta un’ampia campagna presso il centro di ricerca IMDEA Materials, in cui esperimenti sistematici hanno analizzato l’effetto delle condizioni operative su produttività e qualità dei CNT. La caratterizzazione dei materiali mediante spettroscopia Raman e microscopia elettronica a trasmissione (TEM) su campioni ottenuti in una finestra operativa ampia e rappresentativa ha permesso di raccogliere dati quantitativi ben caratterizzati. Questi dati hanno supportato lo sviluppo del framework cinetico, consentendo il confronto con le predizioni modellistiche e la validazione del meccanismo proposto. Il framework cinetico sviluppato, modulare e gerarchico, si articola in due componenti principali. La prima è un modello semi-dettagliato per la crescita dei CNT, che descrive la decomposizione catalitica degli idrocarburi sulle nanoparticelle di ferro e la successiva incorporazione del carbonio nella struttura del nanotubo. La seconda è un modello eterogeneo per la deposizione di carbonio pirolitico, che permette di descrivere una possibile crescita laterale di strati di carbonio meno organizzati, innescata dal deposito di specie gassose più pesanti sulle pareti dei CNT. Questi due modelli superficiali, accoppiati al modello in fase gas già sviluppato e validato dal CRECK Modeling Group del Politecnico di Milano, forniscono una descrizione complessiva dell’evoluzione delle fasi gassosa e solida nel processo FCCVD. Il framework cinetico globale, scritto con il formalismo Surface CHEMKIN, è stato implementato in un modello di reattore semplificato, implementato in OpenSMOKE++, per effettuare simulazioni di processo e confrontare le previsioni modellistiche con i dati sperimentali, validando il modello sulla base della resa totale del carbonio e della cristallinità del materiale (rapporto ID/IG) negli spettri Raman). Parallelamente, è stato studiato il ruolo delle nanoparticelle di ferro, essenziali per la crescita dei CNT. È stato sviluppato un modello dedicato alla formazione ed evoluzione delle NP, basato su un approccio a sezioni discrete (DSM). Le previsioni di questo modello sono state confrontate con un set esaustivo di dati sperimentali, raccolti in una campagna dedicata in cui la distribuzione dimensionale delle NP è stata misurata lungo il reattore mediante SMPS-DMA (Scanning Mobility Particle Sizer e Differential Mobility Analyzer). L’integrazione di questi approcci sperimentali e modellistici ha fornito una stima dell'evoluzione della superficie catalitica lungo il processo, quale parametro fondamentale per le simulazioni FCCVD. I futuri sviluppi di questo lavoro dovrebbero concentrarsi sull’integrazione del modello NP all’interno del framework globale, al fine di tracciare con maggiore dettaglio la distribuzione dimensionale delle nanoparticelle catalitiche e correlarla direttamente alle caratteristiche dei CNT sintetizzati. Questo miglioramento consentirebbe una rappresentazione più completa e predittiva della crescita dei nanotubi, in linea con le evidenze sperimentali. Inoltre, uno studio più esaustivo in relazione al ruolo dei precursori catalitici ferro e zolfo, e sulle loro interazioni con le altre variabili di processo risulta essenziale per migliorare la comprensione dei meccanismi di sintesi e affinare ulteriormente il modello cinetico proposto. Al meglio della nostra conoscenza, il framework cinetico sviluppato rappresenta il primo modello semi-dettagliato capace di correlare in modo predittivo le condizioni operative con la produttività e la qualità strutturale dei CNT nella pirolisi termo-catalitica del metano. La validazione sperimentale ha dimostrato la capacità del modello di descrivere in modo soddisfacente le interazioni tra le variabili di processo e il loro effetto sulle proprietà del materiale prodotto. Questo lavoro può considerarsi un avanzamento significativo nella modellazione predittiva del FCCVD, fornendo strumenti concreti per l’ottimizzazione del processo in vista di un’implementazione su scala industriale. L’integrazione dei futuri sviluppi potrà rafforzare ulteriormente il framework, contribuendo all’innovazione nella produzione sostenibile e scalabile di materiali carboniosi avanzati.
A modular and hierarchical kinetic modeling framework for CNT synthesis in floating catalyst CVD reactors
Giudici, Clarissa
2024/2025
Abstract
Floating Catalyst Chemical Vapor Deposition (FCCVD) is an innovative process for low-carbon hydrogen production and carbon sequestration in advanced materials. Unlike steam reforming of methane, which generates high carbon dioxide (CO2) emissions, FCCVD enables the direct pyrolysis of methane, producing in molecular hydrogen and solid carbon, avoiding the release of greenhouse gases. A key aspect of this technology is the possibility of converting carbon into carbon nanotubes (CNTs), materials with unique electrical, thermal and mechanical properties. Thanks to their high strength, rigidity and lightness, CNTs find applications in strategic sectors such as electronics, aerospace and automotive, contributing to improve structural and energy efficiency. This process also stands out for its scalability and continuous operation, thanks to the use of fluidized catalysts, which allow CNTs to form within the reactor and aggregate towards the outlet, where they can be continuously spun into fibers or films. The absence of direct CO2 emissions, combined with the production of high-tech materials, therefore makes the FCCVD method a promising strategy for the industrial production of turquoise hydrogen and advanced materials with broad application potential. Its implementation could mark a crucial step towards the de-carbonization of industrial processes and the transition to sustainable chemistry. Despite its potential, the industrialization of this method still faces significant challenges. The complex interplay between gas-phase reactions and catalytic growth processes makes it difficult to control the quality and yield of CNTs, limiting process predictability and efficiency. Although parameters such as temperature and the concentration of catalytic precursors (Fe, S) have been identified as determinants for the structure of CNTs, there is a lack of a reliable quantitative model that can correlate operating conditions with product yield and quality. The evolution of catalytic nanoparticles (NPs) represents a further critical issue, as it is strongly influenced by process variables. Their evolution and growth along the reactor makes the correlation between the size distribution of NPs and the final properties of CNTs complex, hindering process control and maximization of catalytic activity. The competition between CNT growth and amorphous carbon formation is a further limitation. Although the formation of carbon by-products is inevitable in FCCVD processes, their impact on the yield and quality of CNTs has not yet been quantitatively described. From a multi-scale modeling perspective, currently available kinetic models do not provide a fully integrated and comprehensive representation of the complex network of chemical processes governing the growth of catalytic nanoparticles and the solid carbonaceous product. Moreover, their validation remains limited, as it has not been conducted on a sufficiently large and representative dataset. This limitation is further aggravated by the scarcity of accurate and well-characterized quantitative data in the literature, which are essential for a deeper understanding of the phenomena involved in the process. Consequently, the lack of a robust data foundation hinders the development of reliable kinetic models capable of rigorously describing the intricate network of reactions and interactions that characterize the FCCVD process. These limitations highlight the need for innovative approaches to deepen the understanding of CNT growth mechanisms and facilitate process optimization for large-scale implementation. In this study, an integrated experimental and modelling approach was adopted to develop a \textit{semi-detailed kinetic model} to describe key phenomena observed in the FCCVD process. The aim is to provide a predictive tool capable of reproducing system behavior and assessing the impact of operating conditions on process performance, thus supporting industrial-scale optimization. From an experimental perspective, an extensive campaign was conducted at the \textit{IMDEA Materials Institute}, where systematic experiments were carried out to analyze the effect of operating conditions on CNT productivity and quality. Material characterization through Raman spectroscopy and transmission electron microscopy (TEM) on samples obtained within a sufficiently comprehensive operating window provided well-characterized quantitative data. These data supported the development of the kinetic framework and its comparison with model predictions, ultimately leading to the validation of the proposed mechanism. The developed kinetic framework, modular and hierarchical, consists of two main components. The first is a semi-detailed model for CNT growth, describing the catalytic decomposition of hydrocarbons on iron nanoparticles and the subsequent incorporation of carbon into the nanotube structure. The second is a heterogeneous model for pyrolytic carbon deposition, which accounts for the potential lateral growth of less organized carbon layers, triggered by the deposition of heavier gas-phase species on CNT walls. These two surface models, coupled with the gas-phase model previously developed and validated by the CRECK Modeling Group at Politecnico di Milano, provide a comprehensive description of the evolution of both gas and solid phases in the FCCVD process. The global kinetic framework, written with the Surface CHEMKIN formalism, has been implemented in a simplified reactor model, implemented in OpenSMOKE++, to perform process simulations and compare the modelling predictions with experimental data, validating the model on the basis of the total carbon yield and the crystallinity of the material (ratio ID/IG in the Raman spectra). In parallel, the role of iron nanoparticles, which are essential for the growth of CNTs, was studied. A dedicated model for the formation and evolution of NPs was developed, based on a discrete section approach (DSM), and the predictions of this model were compared with an exhaustive set of data collected in a dedicated experimental campaign where we measured the size distribution of NPs along the reactor using SMPS-DMA (Scanning Mobility Particle sizer and Differential Mobility Analyzer). The integration of these experimental and modelling approaches provided an estimate of the evolution of the catalytic surface along the process as a key parameter for FCCVD simulations. Future developments should focus on integrating the NP model within the overall framework to track the size distribution of catalytic nanoparticles in greater detail and directly correlate it to the characteristics of the synthesized CNTs. This improvement would enable a more comprehensive and predictive representation of nanotube growth, in line with experimental evidence. Moreover, a more exhaustive study in relation to the role of the catalytic precursors iron and sulfur, and their interactions with other process variables, is essential to improve the understanding of the synthesis mechanisms and further refine the proposed kinetic model. To the best of our knowledge, the developed kinetic framework represents the first semi-detailed model capable of predicatively correlating operating conditions with the productivity and structural quality of CNTs in thermo-catalytic methane pyrolysis. The experimental validation demonstrated the model's ability to satisfactorily describe the interactions between process variables and their effect on the properties of the produced material. This work can be considered a significant advance in the predictive modelling of FCCVD, providing concrete tools for process optimization with a view to industrial-scale implementation. The integration of future proposed developments may further strengthen the framework, contributing to innovation in the scalable and sustainable production of advanced carbon materials.| File | Dimensione | Formato | |
|---|---|---|---|
|
Clarissa_Giudici_PhD_Thesis_2105.pdf
non accessibile
Dimensione
33.52 MB
Formato
Adobe PDF
|
33.52 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/238598