Pyrolysis of a single wood particle: experimental and CFD-aided study toward the improvement of process engineering and product tailoring

Pyrolysis of a single wood particle experimental and CFD-aided study toward the improvement ofprocess engineering and product tailoring Przemysław Maziarka 0 1 0 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 1 0 1 0 1 0 1 0 1 0 0 1 0 1 0 1 0 1 0 1 0 0 1 0 1 0 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 0 1 1 0 1 1 1 1 0 1 0 1 0 1 1 0 1 0 0 0 0 0 0 0 0 0 0 1 1 0 1 1 0 1 0 1 0 1 0 0 1 0 1 0 1 1 0 1 0 0 1 0 1 0 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 0 1 0 1 0 1 0 1 1 0 1 0 1 0 1 0 1 0 0 1 0 1 0 1 0 1 0 1 0 0 1 0 1 0 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 1 0 1 0 1 0 1 0 1 0 1 0 1 0 0 1 0 1 0 1 0 1 0 1 0 0 1 0 1 0 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 1 0 1 0 1 0 1 0 1 1 0 1 1 0 0 0 1 0 1 0 1 0 0 1 0 1 0 1 0 1 0 1 0 0 1 0 1 0 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 0 1 0 1 0 1 0 1 1 0 1 0 1 0 1 0 1 0 0 1 0 1 0 1 0 1 0 1 0 0 1 0 1 0 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 1 0 1 0 1 0 1 0 1 “Wer nicht vorwärts geht, kommt zurück.“ (Eng. “He who moves not forward, goes backward.”) Johann Wolfgang von Goethe Faust: Part I, “Scene: A Study” In my recollection, the quote will always be remembered as “Kto stoi w miejscu, ten się cofa.“ (Eng. “One who stands still, goes backward.”) By Eugeniusz, my beloved dad used whenever it aligned with the context Promoters Prof. dr. ir. Frederik Ronsse (Main promoter) Department of Green Chemistry and Technology, Faculty of Bioscience Engineering, Ghent University, Ghent, Belgium Prof. Dr. Joan J. Manyà (Supporting promoter) Aragón Institute for Engineering Research (I3A), Thermochemical Processes Group, University of Zaragoza, Huesca, Spain Prof. Dr. Andrés Anca-Couce (Supporting promoter) Thermal and Fluids Engineering Department Universidad Carlos III de Madrid, Madrid, Spain Dean of Faculty Prof. dr. Els Van Damme Department of Biotechnology, Faculty of Bioscience Engineering Ghent University, Ghent, Belgium Rector of Ghent University Prof. dr. ir. Rik Van de Walle Department of Electronics and Information Systems, Faculty of Engineering and Architecture Ghent University, Ghent, Belgium DOCTORAL DISERTATION Pyrolysis of a single wood particle: experimental and CFD-aided study toward the improvement of process engineering and product tailoring Thesis submitted in fulfillment of the requirements for the degree of Doctor (Ph.D.) of Bioscience Engineering: Chemistry and Bioprocess Technology By Przemysław Maziarka Dutch translation of the title Pyrolyse van een enkel houtdeeltje: experimenteel en CFD-ondersteund onderzoek naar verbetering van procestechniek en productaanpassing Front Cover Original artwork by the author Back Cover The Atomium under construction at the World Expo, Brussels (1957), Dolf Kruger © De Waarheid. In de schaduw van de wederopbouw, Nederlands Fotomuseum Cite as Przemysław Maziarka (2025). Pyrolysis of a single wood particle: Experimental and CFD-aided study toward the improvement of process engineering and product tailoring (doctoral dissertation). Ghent University, Ghent, Belgium. ISBN number 9789463579018 All rights reserved. To reproduce any part of this document, prior written permission from the author or dissertation promoter(s) shall be obtained. Every other use is subjected to the copyrights laws. VI Acknowledgements Both of the following quotes perfectly capture the truth of my PhD journey over these years. The first, by John F. Kennedy, states: “We choose to go to the Moon in this decade and do the other things, not because they are easy, but because they are hard.” The second, from the Programmers’ Credo, says: “We do these things not because they are easy, but because we thought they were going to be easy.” The first reflects the strong mindset needed to overcome obstacles that were certain from day one. The second portrays the reality of the experience, although an#cipated mentally, the physical and emo#onal challenges oMen came as a surprise. The la9er relates to the daily struggles of maintaining momentum, long hours leading to exhaus#on, and naviga#ng complex prac#cal problems with a thin margin for error. In such moments, I was finding support in a proverb origina#ng from Burkina Faso, (introduced to me by the invaluable Stef!): “If you want to go fast, go alone. If you want to go far, go together.” From my perspec#ve, this work represents that “far,” which I could not have reached without the support of many – both those I knew before star#ng and those I met along the way. I sincerely thank you all for your kind support, for sharing your exper#se and guidance, and for offering moments to “log off” into reality – so cri#cally needed but oMen forgo9en when one is reaching “for the moon.” My deepest gra#tude goes to Prof. Frederik Ronsse. Throughout these years, you always made #me to support me with guidance when needed, while giving me the freedom to choose my own path and develop independence as a researcher. A sincere thank you! I also truly admire how you managed to maintain good spirits and coopera#on within the TCCB group, which is composed of unique researchers. AMer all these years, I clearly see how challenging and demanding such a task is. Kudos! Beyond the everyday support, I will warmly remember our shared trips to events during my PhD – from the kickoff in Huesca to, if I recall correctly, Edinburgh, with the #me-#ght finale :) I truly appreciate all your efforts and pa#ence, without which I would not be where I am today. Though my PhD took a bit longer than usual, I hope the result is worth it. Special and truly sincere thank you, Prof. Wolter Prins. Your detailed, exper#se-rich feedback consistently sharpened my researcher toolkit. Though your remarks oMen demanded extra effort, the results consistently showed their value. Your guidance certainly will not be forgo9en. The TCCB group will always bring me heartwarming memories. No ma9er how long I was away, the welcome was always kind and homelike, making me feel as if I had never leM. My one and only Jonas, dear desk colleague, your cheerful aRtude and unconven#onal sense of humor brightened even the most challenging days. I will cherish our memorable scien#fic discussions on porous carbons and the countless small amusements we shared during breaks – and how Dilani “bap#zed” you just before finishing her PhD :) Dear, oh dear Stef, thank you for everything since day one. I could write a million words about you. Unique personality, inimitable style, and true devo#on to science – that is Stef in my eyes. Adriana, your genuine approach and passion were a true guiding lighthouse for me. It was a great pleasure to meet and work with you. Thank you sincerely! I will always remember your devo#on to science and joyful smile – and victory dance with Jonas in the lab :) My project partner, Dilani, with whom I shared the struggles of studying abroad. I am grateful for your kindness and different perspec#ve, which broadened my worldview. There were also two shining stars, Tamara and Leya, who reminded me that while hard work is essen#al, life must also be enjoyed. I hope the “ghost” logging into your computer didn’t disturb your work too much :) I am grateful for the mentorship of our experienced colleague, Mehmet. Thank you for your wise advice on facing the challenges of doctoral research – some#mes served in a pintje-size por#ons, in a small pub near the opera :) I would also like to warmly thank all TCCB group members I have not men#oned by name. I am truly grateful to each of you and will always remember you fondly. Furthermore, I will cherish the unforge9able memories of all the members of Ghent University and the kind people of Ghent city, who were an inseparable part of this great journey. VIII My most significant apprecia#on goes to Prof. Joan J. Manyà, head of the GreenCarbon project funded by the European Commission, through whose courtesy I could begin my PhD journey. From my perspec#ve, the GreenCarbon project was an eye-opening experience that gave me great confidence in the scien#fic world, which I find invaluable. Dear Juan, my deepest thanks to you for accep#ng me for a posi#on in the project, as well as all your support and wise guidance throughout the project and our collabora#ve work. I am also profoundly grateful that you agreed to be the suppor#ng promoter of my thesis, fully aware of how demanding that role was. The most valuable experience of the project was collabora#ng with numerous young and senior researchers across Europe and beyond, all working toward common goals. I will especially cherish my coopera#on with Chris#an, with whom we co-created, as cita#ons indicate, a relevant scien#fic work. Chris#an, I am very thankful for the opportunity to work with you. Your “out-of-the-box” insights and prac#cal problem-solving approach changed my mindset. Another memorable collabora#on was (and s#ll is!) with Pablo, who has unparalleled scien#fic stamina and great devo#on to advancing science. Pablo, I truly appreciate our fruiTul coopera#on, which resulted in several interes#ng publica#ons. Furthermore, kind thank you for your support and help in #mes of need and brightening my day with your “funny stories” :), which allowed me to decompress. Special thanks also to Maciej, my long#me friend, with whom I shared numerous scien#fic, economic and some#mes pointless discussions :) I sincerely thank you for your con#nued support over the years! To all those I have not men#oned – I do not forget the shared struggles, and successes, and am grateful to have met you along the way. I would like to express my sincere thanks to Prof. Andrés Anca-Couce, whose scien#fic work I admired even before beginning my PhD. Dear Andrés, I am truly glad to have had the chance to meet you (especially in person). Thank you for all your valuable comments and insights, not only regarding single par#cle conversion but throughout our scien#fic collabora#on. These insights have helped shape my perspec#ve across several areas, and I value them highly. I am also very grateful that you agreed to become the suppor#ng promoter of my thesis. My deepest gra#tude goes to the examina#on commi9ee of my PhD thesis: Prof. Jan Pieters, Prof. Steven De Meester, Prof. Wolter Prins, Prof. Ondřej Mašek, and Dr. Axel Funke. I appreciate that you dedicated your #me to review my thesis, especially given its length (and struggles with being concise). I sincerely thank you for your efforts and for providing valuable feedback. From the bo9om of my heart, I would like to thank my parents – Ewelina and Eugeniusz – for their love, understanding, and uncondi#onal support throughout the long journey of earning my PhD. Drodzy rodzice, dziękuję Wam serdecznie za całą Waszą pomoc przez te wszystkie lata mojego doktoratu. Wiem, że czasem mogło nie być łatwo. Jestem Wam niezwykle wdzięczny za Wasze wsparcie i wyrozumiałość w tym wymagającym dla mnie czasie. Bez Was nie byłbym w stanie dotrzeć do tego miejsca, w którym jestem dziś. Z całego serca, dziękuję! I would also like to thank my beloved girlfriend, Martyna. My Love, above all, I sincerely thank you for your pa#ence and support during the long hours I had to devote to comple#ng the PhD. Kochanie, jestem Ci szczerze wdzięczny za Twoją miłość, wsparcie oraz bezgraniczną cierpliwość. Dziękuję! Finally, I am grateful to my closest family for their support, especially my dear sister Donata and brothers Radosław and Jarosław. Dziękuję serdecznie za Wasze wsparcie przez te wszystkie lata! As I cannot men#on everyone who supported me along the way, please accept my hearTelt thanks to each and every one of you. I appreciate all contribu#ons, no ma9er the size, as each one has helped build my path toward the final des#na#on – earning my PhD #tle. Thank you! John Steinbeck said – “It is good to have an end to journey toward, but it is the journey that ma9ers in the end.”. Every journey oMen begins with a ques#on – Will it be hard? But does that really ma9er? Mine was long, full of twists, turns, and challenges, yet immensely rewarding – filled with life-las#ng knowledge and experience. Most importantly, it was enriched by the valuable people I met along the way. I wish everyone a similarly fruiTul and perspec#ve-widening journey through their PhD! IX Abstract A brief review of the literature reveals that, historically, thermochemical conversion of biomass was an inseparable part of global development. As evidenced by historical records, charcoal served as the primary carbon source for metallurgical processes over the centuries. Moreover, the transforma#on of wood into liquid products played a crucial role in the development of the modern chemical industry, including the produc#on of dyes and synthe#c materials. Such informa#on contrasts with the contemporary public percep#on on biomass thermochemical conversion, which is largely associated with the use of charcoal as a fuel for outdoor cooking. Such a percep#on is not en#rely unexpected, as other thermochemically derived biomass products are rarely encountered in today’s market. Considering the focus of the Thesis, it is per#nent to consider ques#ons such as: “What happened throughout history that the role of thermochemical biomass conversion technologies became diminished?” and “Is it possible (or even necessary) to reverse this trend?” While these are indeed compelling and relevant inquiries, they fall outside the scope of this work, a thorough and objec#ve answer must be sought in other sources. Without knowing the precise answers to the above-men#oned ques#ons, nearly a century ago, scien#sts – and subsequently, society at large – began to recognize that the uncontrolled exploita#on of available resources would inevitably lead to their deple#on. Aside from the obvious problem of deple#on of resources such as materials or fuels (fossil fuels or freshwater), it became increasingly relevant that the atmosphere is also a resource with a limited capacity. Nevertheless, the realiza#on that the release of non-visible combus#on gases into the atmosphere will have long-term consequences took #me to fully permeate public awareness. A comprehensive overview of the origins of climate change awareness and the current approaches to mi#ga#on – including the role of biomass and bio-based char solu#ons – can be found in Chapter 1. Taking a step further, with the increase of society’s expecta#ons for reducing GHG emissions, the global market started seeking alterna#ve solu#ons that may reduce the overall nega#ve impacts more efficiently than currently employed technologies. Among these alterna#ves, bio-based subs#tutes have emerged as a promising op#on. Thermochemical conversion technologies, like biomass pyrolysis and gasifica#on, offer the capability to produce a range of bio-based products and can be scaled to meet market demand. These technologies already supply low-engineered bio-based products for wellestablished applica#ons, such as charcoal for energy use. However, market trends reveal a growing demand for more advanced, purpose-designed materials, par#cularly those with engineered porous structures like bio-based carbons for soil amendment or supercapacitor produc#on. Despite their poten#al, the industrial-scale supply of these advanced bio-based products to replace fossil-based materials s#ll remains very limited. Aside from economy-of-scale aspects, the main issue hindering novel bio-based materials from reaching their poten#al is the low repeatability of their proper#es, especially when aligned with the requirements of their end applica#ons. The problem can be dis#lled into a statement that it is s#ll impossible to make precise quan#ta#ve predic#ons of the final proper#es of the product (e.g., char water reten#on or electric double-layer capacitance) based only on known process parameters and basic feedstock proper#es. At this point, the path toward resolving this issue becomes increasingly complex and warrants more detailed discussion. Inves#ga#ons into the structural changes of the resul#ng materials, based on reactor-scale experiments, yield the most representa#ve insights in terms of real-scale produc#on systems. However, in studies on such a scale, the effect of the reactor needs to be considered, which is complex and challenging to assess. Addi#onally, extended trials on a reactor scale involve no#ceable costs, and their results may be specific to the par#cular system under inves#ga#on, limi#ng their applicability to broader, more universal cases. An alterna#ve approach involves studying the process at the singlepar#cle scale, where structural transforma#ons are examined in individual feedstock par#cles during conversion. In such studies, precise control over conversion condi#ons and par#cle boundary condi#ons is possible (i.e., thereby reducing or omiRng the effect of the reactor). Furthermore, interac#ons between par#cles can be neglected. With the reduced impact of the reactor on the XVI een enkel houtdeeltje kwan#ta#ef beschrijven. In hoofdstuk 5 werd de nadruk gelegd op de formuleringen die de thermofysische eigenschappen van hout en zijn verkoling in verband brengen met de procesomstandigheden, met name die welke verband houden met de structuuraiankelijke parameters. Een vereiste voor hoofdstuk 5 was dat het nieuw afgeleide formuleringen moest beva9en die ontbraken (voor zover deze konden worden afgeleid uit de beschikbare gegevens uit hoofdstuk 2 en hoofdstuk 4). Daarom worden in hoofdstuk 5, naast de bepalende en fundamentele vergelijkingen, aanvullende formuleringen gepresenteerd die zijn afgeleid uit experimentele gegevens. Er werd vastgesteld dat de beschikbare formuleringen mogelijks onvoldoende zijn, dus werden nieuwe func#es voorgesteld om de werkelijke dichtheid van verkoling bij verschillende temperaturen te beschrijven. Op basis van een vergelijkbare aanname werd de verandering van de poriëndiameter geformuleerd, die vermoedelijk relevant is voor de beschrijving van de permeabiliteit. Verder werd een geconsolideerde formulering van de veranderingen in thermische geleidbaarheid en geometrische vervorming voorgesteld. In hoofdstuk 6 wordt een stapsgewijze opbouw van een betrouwbaar CFD-model gepresenteerd, dat echter nog niet volledig is. Zoals in hoofdstuk 2 en hoofdstuk 5 is gebleken, moesten specifieke elementen van een CFD-model eerst objec#ef worden beoordeeld, omdat: (1) er in de literatuur verschillende formuleringen van hetzelfde fenomeen beschikbaar zijn en (2) beschikbare CFD-modellen eenvoudigweg zijn opgesteld en gevalideerd voor algemene procesomstandigheden of juist voor enge procesomstandigheden. Een dergelijke situa#e maakte het niet mogelijk om formuleringen aan te geven die zouden kunnen leiden tot de construc#e van een volledig CFD-model. De vereisten voor het geconstrueerde CFD-model waren: (1) het moet bestaan uit state-of-the-art formuleringen die tot de meest nauwkeurige voorspelling leiden, en indien deze niet beschikbaar zijn, moeten deze worden aangegeven, en (2) het moet worden gevalideerd tegen een brede waaier aan procescondi#es om aan te tonen dat het model niet scenariospecifiek is en op een algemene manier kan worden gebruikt. Daarom wordt in hoofdstuk 5 een stapsgewijs onderzoek gepresenteerd van de impact op de nauwkeurigheid van gesimuleerde resultaten als gevolg van de geïmplementeerde formulering van: (1) rich#ngsaiankelijkheid, dus isotropie en anisotropie in thermische geleidbaarheid en gasdoorlaatbaarheid, (2) formulering van het droogmodel (kine#sch, evenwicht en warmteafvoer), en (3) kine#sch schema voor de agraak van primaire biocomponenten, waarbij de presta#es van het Shafizadehen Chin-schema (eenvoudig), het Ranzi-schema (gedetailleerd) en het Ranzi-Anca-Couce (RAC)-schema (gedetailleerd) werden gecontroleerd. Op basis van de meest betrouwbare formuleringen werd het CFD-model vervolgens gevalideerd aan de hand van een uitgebreide dataset, die temperaturen in het bereik van 500-840 °C, deeltjesgroo9es in het bereik van 10-20 mm en twee deeltjesvormen (cilindrisch en bolvormig) omva9e. De valida#e vereiste enorme rekenkracht en werd daarom parallel uitgevoerd met behulp van een high-performance computersysteem. Aangezien het CFD-model bij de valida#e een bevredigende nauwkeurigheid en betrouwbaarheid liet zien, werd een onderzoek uitgevoerd naar de invloed van de verwarmingssnelheid op de omzeRng. Dit werd gedaan omdat de verwarmingssnelheid vaak wordt genoemd als een relevante parameter voor de omzeRng, maar tot nu toe geen van de studies de invloed ervan kwan#ta#ef heeM onderzocht. De belangrijkste observa#e uit het onderzoek was de vorming van de “buitenste thermische laag”, waar de verwarmingssnelheid onder intense thermische omstandigheden van de omzeRng dynamischer verandert dan in andere delen van het deeltje. Hieruit werd afgeleid dat een dergelijke laag voor hout doorgaans een dikte heeM van 2-2,5 mm, wat overeenkomt met experimentele overwegingen. Daarom werd numeriek aangegeven dat wanneer een deeltje met een hoge verwarmingssnelheid moet worden omgezet, de groo9e ervan in ten minste één rich#ng niet groter mag zijn dan 4-5 mm. Hoofdstuk 7 is het laatste hoofdstuk van het werk en heeM betrekking op het uiteindelijke doel van het werk, namelijk het uitgebreide CFD-model voor pyrolyse van een enkel houtdeeltje. De algemene eis voor het CFD-model was dat het model zo nauwkeurig en betrouwbaar mogelijk moest zijn in de prak#jk. Dat had in detail betrekking op (1) een nauwkeurige voorspelling van het resultaat van het proces over een breed scala aan procesomstandigheden met typisch beschikbare, ini#ële eigenschappen van hout en (2) het verkrijgen van een nauwkeurig resultaat bij de simula#e van de XVII structurele veranderingen van een deeltje, wat het knelpunt vormde van de beschikbare CFDmodellen. Het uitgebreide CFD-model dat in hoofdstuk 7 werd afgeleid, mogelijk als eerste CFD-model, kon een zeer nauwkeurige voorspelling van de relevante structurele eigenschappen van de uit hout akoms#ge koolstofdeeltjes aantonen. De gemiddelde voorspellingsfout over het genoemde temperatuurbereik van de bulkdichtheid was 31 ± 15 kg/m 3 en voor de porositeit was de voorspellingsfout 1.8 ± 1.1 vol.% in een temperatuurbereik van 300 °C tot 840 °C. Bovendien bedroeg de gemiddelde voorspellingsfout voor temperaturen tussen 400 °C en 840 °C voor koolopbrengst 2.8 ± 1.1 gew.%, voor opbrengst aan condenseerbare frac#e 6.5 ± 3.3 gew.% en voor opbrengst aan pyrolysegas 4.1 ± 1.9 gew.%. Een dergelijke presta#e van het CFD-model werd als bevredigend beoordeeld, vooral gezien het brede bereik van de gesimuleerde temperaturen. Bovendien bleek uit het onderzoek dat geometrische vervorming een onlosmakelijk kenmerk is van het CFD-model, aangezien dit een groot effect heeM op de implementa#e van de meer ontwikkelde thermische geleidbaarheid en formuleringen van permeabiliteit. Het was de bedoeling om een uitgebreid CFD-model te ontwikkelen om het ontwerpen van biogebaseerde houtskool met alleen kennis van procesparameters en basiseigenschappen van de oorspronkelijke biomassa mogelijk te maken of, indien dit niet mogelijk was, de huidige mogelijkheden uit te breiden. Als dit zou lukken, zou dit een posi#eve invloed hebben op de mogelijkheid om biogebaseerde (houts)kool aan te passen aan niet verder-gespecifieerde, nieuwe toepassingen (meestal op basis van structuurgerelateerde eigenschappen). Vanaf het begin was het onwaarschijnlijk dat het CFD-model volledig zou kunnen worden toegepast voor het op maat maken van biogebaseerde (houts)kool, omdat er nog onvoldoende gegevens beschikbaar zijn over de eigenschappen van kool en de veranderingen daarin, zelfs met de eenvoudigste procesparameters zoals temperatuur. Nie9emin bleek het ontwikkelde uitgebreide CFD-model in staat om het verloop en het resultaat van de pyrolyse van een enkel houtdeeltje op bevredigende wijze te voorspellen. Bovendien voorspelde het model met grote nauwkeurigheid de geometrische vervorming van een deeltje #jdens thermische agraak, dus de bulkdichtheid en porositeit, wat tot nu toe niet was bereikt. Dit is vooruitgang die van een enkel doctoraatsproefschriM kon worden verwacht, en die vooruitgang is ook geboekt. Naast het ontwikkelde model vormen de geconsolideerde herzieningen, observa#es en ontdekkingen (op basis van experimenten en numerieke berekeningen) in dit werk een solide basis voor toekoms#g onderzoek. Bovendien is de verstrekte informa#e gesystema#seerd en gebundeld, wat het gebruik ervan moet vergemakkelijken voor anderen die besluiten om de vooruitgang op dit gebied verder te ze9en. XVIII TABLE OF CONTENTS Acknowledgements VII Abstract IX Samenva8ng XIII TABLE OF CONTENTS XIX CHAPTER 1 1 1. Prologue 2 2. Role of biomass in the low-carbon economy 3 2.1. Greenhouse effect and decarboniza#on 3 2.2. Decarboniza#on with bioenergy and biofuels 4 2.3. Emissions reduc#on of chemical produc#on with biomass 6 2.4. Biomass-based carbon-nega#ve solu#ons 7 3. Adding value by the biomass processing of biomass 8 3.1. Biomass and its poten#al 8 3.2. Decarboniza#on with biomass 12 3.3. Biomass and its processing 15 4. Bio-based chars and their applica,on 21 4.1. Thermo-chemical conversion of bio-components 21 4.2. Pyroly#c technologies for the produc#on of bio-based char 25 4.3. Environmental and market aspects of bio-based chars 29 4.4. Applica#ons of bio-based chars 32 4.5. Relevance of porosity for bio-based chars 43 5. Modelling and biomass-derived chars 47 5.1. Needs for and profits from modeling 47 5.2. Mul#-scale models for biomass pyrolysis 50 5.3. Relevance of the single par#cle scale 56 6. The scope and goal of the Thesis 59 7. References for Chapter 1 63 CHAPTER 2 77 1. Introduc,on to Chapter 2 78 2. Mul,scale of wood structure 79 2.1. Tree anatomy and hierarchical wood structure 79 2.2. Dimensions of wood cell and wood cell wall 82 2.3. Wood cell wall and assembly of bio-components 83 2.4. Chemical composi#on and its assessment 85 TABLE OF CONTENTS XX 3. Impact of thermal degrada,on on the structure of wood 87 3.1. Geometric deforma#on 87 3.2. Structural changes of cell wall 88 3.3. Structural changes of lumen 91 3.4. Changes in molecular structure (up to 500 °C) 92 3.5. Organiza#on and ordering of bio-originated carbon structures 94 3.6. Changes in molecular structure (above 500 °C) 97 4. Density of wood and its char 99 4.1. True density of cell wall – wood 99 4.2. True density of cell wall – wood-derived char 101 4.3. Bulk density and porosity 103 4.4. Moisture, water types, and fiber satura#on point 106 4.5. Thermal treatment and hydrophilicity/hydrophobicity 108 5. Structure dependent thermo-physical parameters 110 5.1. Fluids transport through wood structure 110 5.2. Thermal degrada#on and changes in wood permeability 113 5.3. Heat transfer and thermal conduc#vity of wood 118 5.4. Thermal conduc#vity of char and its cell wall 123 6. Final remarks to Chapter 2 127 7. References for Chapter 2 128 CHAPTER 3 137 1. Introduc,on to Chapter 3 138 1.1. Novel uses of bio-based chars 138 1.2. Engineering of bio-based char’s structure 139 1.3. Reliable measurement issue 140 1.4. Reliable isotherm processing issue 141 1.5. Aim of the study presented in Chapter 3 143 2. Materials and methods 144 2.1 Bio-based chars prepara#on 144 2.2 Data from the literature (external data set) 144 2.3 Physisorp#on measurements 144 2.4 Isotherm analysis 145 3. Experimental results of the inves,ga,on 145 3.1 Isotherms obtained with nitrogen (77 K) 145 3.2 Assessment of N 2 isotherms with a pore non-specific model (BET) 147 3.3 Assessment of N 2 isotherms with pore specific models (HS-NLDFT and QSDFT) 150 3.4. Assessment of CO 2 isotherms with a pore specific model (GCMC) 152 4. Analysis of iden,fied micropore deforma,on phenomenon 155 4.1. The influence of measurement condi#ons on N 2 adsorp#on 155 4.2. Open hysteresis and adsorp#on-induced pore deforma#on 156 4.3. Molecular structure of bio-based chars and the pore deforma#on 158 TABLE OF CONTENTS XXI 5. Origin of the ar,ficial peak and mi,ga,on method 160 5.1. Hypothesis of a micropore structure collapse 160 5.2. Modified N 2 adsorp#on results assessment 162 6. Final remarks to Chapter 3 166 7. References for Chapter 3 167 CHAPTER 4 173 1. Introduc,on to Chapter 4 174 1.1. Specific context of Chapter 4 174 1.2. Aim of the study presented in Chapter 4 175 2. Materials and methods 176 2.1. Feedstock (beech wood par#cles) 176 2.2. Single-par#cle pyrolysis setup 176 2.3. Data prepara#on (smoothing, conver#ng, and averaging) 180 2.4. Analysis of released vapors and collec#on of bio-oil 184 2.5. Bio-oil sample prepara#on and GC-MS/FID measurements 185 2.6. Iden#fica#on of bio-oil components 186 2.7. Quan#fica#on with the Effec#ve Carbon Number (ECN) method 186 2.8. Dimensional changes of par#cles 187 2.9. Scanning electron microscopy (SEM) 189 2.10. Helium pycnometry 189 2.11. Gas adsorp#on with N 2 and CO 2 and data processing 189 2.12. Mercury intrusion porosimetry (MIP) 190 2.13. Assessment of hierarchical porous structure (micro-meso-macro) 191 3. Results of pyrolysis of singe beech wood par,cles 191 3.1. Temperatures and mass change profiles 191 3.2. Yields and release profiles of vapors 195 3.3. Composi#on of collected bio-oil 199 3.4. Mass and elemental balance 200 4. Beech wood chr assessment 203 4.1. General changes in wood-derived char structure 203 4.2. Analysis of char structure and morphology 205 4.3. General changes in the pore size distribu#on 207 4.4. Quan#ta#ve effects on specific pore size ranges 209 5. Tradeoffs and bio-based char tailoring for a given applica,on 212 6. Final remarks to Chapter 4 214 7. References for Chapter 4 215 CHAPTER 5 221 1. Introduc,on for Chapter 5 222 1.1. Numerical models for pyrolysis of a single wood par#cle 222 1.2. Base assump#ons and nota#ons in the study 224 TABLE OF CONTENTS XXII 2. Governing equa,ons 225 2.1. Mass conserva#on equa#ons: Solids 225 2.2. Mass conserva#on equa#ons: Single component in the gas mixture 226 2.3. Mass conserva#on equa#ons: Liquid and Bound water 226 2.4. Energy conserva#on equa#on 227 2.5. Mass change rates 228 2.6. Shape specifica#on and coordinate systems 228 2.7. Direc#onal dependence 229 3. Kine,c schemes for the thermal degrada,on of wood 230 3.1. Primary degrada#on: Simple – Single component compe##ve schemes 230 3.2. Primary degrada#on: Detailed – Ranzi kine#c scheme (Ranzi) 231 3.3. Primary degrada#on: Detailed – Ranzi-Anca-Couce scheme (RAC) 232 3.4. Secondary gas-phase degrada#on: Simple – One-step kine#cs schemes 233 3.4. Secondary gas-phase degrada#on: Detailed – Advanced schemes 234 4. Moisture and evapora,on models 234 4.1. Moisture content and satura#on 234 4.2. Drying model: Kine#c model (KIN) 235 4.3. Drying model: Heat sink model (HS) 236 4.4. Drying model: Equilibrium model (EQ) 237 4.5. Heat of water evapora#on 238 5. Auxiliary equa,ons and relevant parameters 238 5.1. Deriva#on of the auxiliary rela#ons 238 5.2. Conversion factor 239 6. Density of specific phases 239 6.1. True des#ny of dry wood and char 239 6.2. Gas pressure and true density of fluids 241 7. Fluid transport: Convec,on 242 7.1. Flow descrip#on fundamentals 242 7.2. Capillary pressure 243 7.3. Permeability of dry wood and char (intrinsic) 244 7.4. Gas and liquid water permeability in the wet wood 246 7.5 Viscosity of fluids 247 8. Fluid transport: Diffusion 249 8.1. Gas binary diffusion 249 8.2. Tortuosity factor 250 8.3. Bound water diffusion 251 9. Heat transfer: Specific heat capacity 252 9.1. Specific heat capacity of wood and chars 252 9.2. Specific heat capacity of fluids 253 TABLE OF CONTENTS XXIII 10. Heat transfer: Thermal conduc,vity 254 10.1. Thermal conduc#vity of wood 254 10.2. Thermal conduc#vity of wood-derived chars 258 10.3. Thermal conduc#vity of fluids 262 10.4. Heat transfer through radia#on 263 10.5. Effec#ve thermal conduc#vity 264 11. Geometric deforma,on 265 11.1. Generic assump#ons 265 11.2. Deforma#on coupled to mass loss 268 12. Boundary condi,ons 272 13. Final remarks to Chapter 5 273 14. References for Chapter 5 274 CHAPTER 6 281 1. Introduc,on to Chapter 6 282 1.1. Specific context of Chapter 6 282 1.2. Specific issues covered in Chapter 6 283 1.3. Aim of the study presented in Chapter 6 285 2. Founda,ons of established CFD models 286 2.1. General assump#ons 286 2.2. Applied equa#ons and boundary condi#ons 287 2.3. Drying models 289 2.4. Kine#c schemes of primary degrada#on 289 2.5. SoMware, solver configura#on and convergence 290 2.6. Mesh and maximal #me step selec#on 291 2.7. Hardware limita#ons and High-Performance Compu#ng (HPC) 293 3. Part 1: Direc,onal dependence and drying models 294 3.1. Founda#ons of CFD models used in Part 1 294 3.2. Influence of the direc#onal dependence of wood proper#es 295 3.3. Pre-selec#on of parameters for drying models 299 3.4. Accuracy of drying models 301 4. Part 2: Kine,c schemes of primary degrada,on 306 4.1. Founda#ons of CFD models used in Part 2 306 4.2. Predic#on accuracy of center temperature 306 4.3. Relevance of reac#on enthalpies in kine#c schemes 308 4.4. Yields of lumped products and compounds release profiles 310 5. Part 3: Reliability of a simula,on vs. temperature, par,cle size and shape 314 5.1. Founda#ons of CFD models used in Part 3 314 5.2. Simulated temperatures vs. a broad range of conversion parameters 316 5.3. Simulated yields vs. a broad range of conversion parameters 320 5.4. Hea#ng rate and outer thermal layer 323 TABLE OF CONTENTS XXIV 6. Guidelines for improvements in the future 326 7. Final remarks to Chapter 6 329 8. References for Chapter 6 330 CHAPTER 7 335 1. Introduc,on to Chapter 7 336 1.1. Specific context of Chapter 7 336 1.2. Specific issues covered in Chapter 7 337 1.3. Aim of the study presented in Chapter 7 338 2. Founda,ons of established CFD models 338 2.1. General assump#ons 338 2.2. Applied equa#ons and formula#ons 340 2.3. Primary and secondary gas-phase reac#on schemes 342 2.4. Ini#al condi#ons and boundary condi#ons 344 2.5. Setup of numerical computa#ons 345 2.6. Inves#ga#on strategy of the study 346 3. Case 1: Geometric deforma,on and predic,on accuracy 347 3.1. Temperature and mass loss profiles 347 3.2. Yield and composi#on of wood-derived chars 351 3.3. Predic#on of changes in geometry and porosity 352 4. Case 2: Impact of secondary gas-phase reac,ons 354 4.1. Secondary gas phase reac#ons and lumped yields 354 4.2. Yields of specific compounds and their release profiles 357 4.3. Intrinsic secondary reac#ons and soot forma#on 362 5. Case 3: Permeability formula,ons and fluxes of vapors 367 5.1. Formula#ons and dynamic changes in permeability values 367 5.2. Permeability formula#ons and conversion of a par#cle 369 6. Insights from results of Chapter 7 371 7. Final remarks to Chapter 7 374 8. References for Chapter 7 376 CHAPTER 8 379 APPENDICES 391 Appendix 1 for Chapter 3 392 A1.3.1. External data 392 A1.3.2. Micro-Vpore (HS-DFT) and SSA (BET) correla#on 396 A1.3.3. Comparison between HS-NLDFT and QSDFT method (N 2 adsorp#on) 396 A1.3.4. Results from GCMC method (CO 2 adsorp#on) 397 A1.3.5. Rela#onship between micro-V pore obtained with N 2 and CO 2 adsorp#on 398 A1.3.6. Pore size distribu#on (HS-NLDFT) from sequen#al gas adsorp#on (N 2 ) on RH chars 399 A1.3.7. Pore size distribu#on (GCMC) from sequen#al gas adsorp#on (CO 2 ) on RH chars 400 TABLE OF CONTENTS XXV A1.3.8. Comparison of the SSA derived from N 2 adsorp#on with different approaches 401 A1.3.9. References for Appendix 1 401 Appendix 2 for Chapter 4 402 A2.4.1. Auxiliary FT-IR results 402 A2.4.2. GC-MS/FID results and ECN method 410 A2.4.3. Pore size distribu#on obtained from MIP 419 A2.4.4. Pore size distribu#on obtained by merging data from all measurement techniques 420 A2.4.5. Tabulated values of mass and elemental balances 421 A2.4.6. Tabulated values of pore volumes and specific surface areas 423 Appendix 3 for Chapter 5 425 A3.5.1. List of compounds and grouping for detailed schemes 425 A3.5.2. Primary degrada#on scheme (Simple) – Shafizadeh and Chin 426 A3.5.3. Primary degrada#on scheme (Detailed) – Ranzi-Anca-Couce (RAC) scheme 426 A3.5.4. Secondary gas-phase reac#on scheme – (Simple) Mellin et al. with adjustment 428 A3.5.5. References for Appendix 3 428 Appendix 4 for Chapter 6 429 A4.6.1. Hea#ng rate distribu#on and measurements at specific points (Part 3) 429 A4.6.2. References for Appendix 4 431 AUTHORS CONTRIBUTIONS 433 ACADEMIC CURRICULUM 435 CHAPTER 1 7 The studies also indicate that at the current moment, the produc#on of bio-chemicals to subs#tute fossil-derived ones can be related to the increase in produc#on costs. Studies forecast that with the produc#on technology matura#on and forecasted upscaling of the produc#on, the scale factor will be able to bring the cost of bio-based and fossil-based chemicals to a similar range. However, the forecasted #me before reaching this cost-converging point may be overly op#mis#c.[34] The co-processing of fossil and biomass feedstocks is also indicated as a shortand mid-term pathway toward the decarboniza#on of producing fuels and chemicals. The solu#on relates to the use of mixtures of biomass-derived liquids (e.g., bio-oil from pyroly#c processes and bio-crudes from hydrothermal processes) and crude oil-derived liquids and their simultaneous conversion in conven#onal refining processes (hydrotrea#ng – HT and fluidized cataly#c cracking – FCC).[43-46] Un#l very recently, the industrial applica#on of co-processing solu#ons in Europe was hindered by inconclusive legisla#on on assessing the emission reduc#on stemming from the co-processing of biobased and fossil fuel mixtures. It posed a risk that the GHG emission reduc#on obtained by a company would not be accounted for objec#vely in the emissions trading system. That poses a significant investment risk as the implemented solu#on would not bring a financial profit for the company. The situa#on changed with the Delegated Regula#on 2023/1640 published in August 2023, which, among others, clarified legislatorial ma9ers on biofuels and biogases derived from coprocessing. Furthermore, the direc#ve provided a methodology for determining the bio-component of the final fuel obtained in co-processing. The new assessment methods, aside from the recognized but very expensive radiocarbon ( 14 C) tes#ng, enable the use of calcula#on methods based on mass or energy balances and yield to indicate the bio-content in products. With the new legisla#on, the accelera#on of co-processing implementa#on in exis#ng European refineries is forecasted, which may translate to a more efficient reaching of the GHG emission reduc#on targets.[47] 2.4. Biomass-based carbon-nega,ve solu,ons As indicated by the IPCC, reaching the goals of the Paris Agreement may not be possible only with the decarboniza#on of the economy (net-zero). Necessary ac#on also seems to be compensa#on for the emissions in the areas that will not meet the required goals and removal of the historically formed GHG emissions from the atmosphere to mi#gate the greenhouse effect (i.e., the so-called carbon-nega#ve solu#ons). Those solu#ons in the IPCC report are indicated as Carbon Dioxide Removal (CDR) solu#ons, which relate to technologies, prac#ces, and approaches that remove carbon dioxide from the atmosphere and store it over a long #me in other environmental reservoirs (land and oceans).[5, 9] On one end, the CDR solu#ons include rela#vely non-technology advanced solu#ons like afforesta#on, pasture management, and wetland restora#on. On the other end, the CDR solu#ons include technologically advanced and cost-intensive ones, with a major example being direct air carbon dioxide capture and storage (DACCS). Among the CDR op#ons, two solu#ons directly relate to sustainably sourced biomass. As biomass growth would lead to the removal of CO 2 from the atmosphere, biomass product sequestra#on would remove the carbon from circula#on. The first biomass-related CDR solu#on is the produc#on of energy from biomass (bioenergy) and sequestra#on of the products (i.e., emissions). The solu#on is also known as the Bioenergy Carbon Capture and Storage (BECCS), where the collected CO 2 from combus#on is directed to an underground deep geological forma#on for its long-term storage.[5, 9, 48] The second biomass-related CDR solu#on is the produc#on of biomass-derived chars (from forest, agriculture, industrial, and urban bio-wastes) and their sequestra#on in soil (carbon sink). Such a carbon sequestra#on system is also known as a biochar system and Biochar Carbon Removal (BCR), where the biomass-derived char is referred to as a biochar.[5, 49-51] The general difference between carbon sequestra#on using biochar and using biomass is the significantly higher stability of the carbonaceous structures in chars compared to biomass. That translates to an elevated ability of biochar to store carbon over a long period despite being affected by environmental factors. The IPCC report indicates (with high confidence) that the use of biochar in soil, aside from working as a direct carbon CHAPTER 1 8 sink, can lead to a posi#ve impact on soil health (improving water reten#on, aera#on, and microbiological ac#vity) and indirect carbon capture (through enhancement of biomass growth). However, the report also indicates that mi#ga#on and agronomic co-benefits depend strongly on the proper#es of biochar and the soil to which it is applied (also with high confidence).[5, 9] Biochar was recognized by the IPCC as the CDR measure rela#vely recently (Special Report from 2018).[52] At the #me, the status of biochar derived from bio-wastes was considered as a waste in view of the EU regula#on. Therefore, only biochar derived from primary biomass feedstock, such as wood and straw, was allowed for soil applica#on. The restric#on was related to the then-unclear rela#on between the use of biochar on soil quality and was advocated by worries of possible soil contamina#on (e.g., with poly-aroma#c hydrocarbons) when bio-waste was used as biochar feedstock. The net result was elevated costs of biochar applica#on as the feedstock for its produc#on was market products with other, more profitable uses in the cascade applica#ons, like non-contaminated wood (resul#ng in a low economic benefit from the implementa#on of biochar systems). Ul#mately, market en##es were discouraged from biochar produc#on and applica#on.[53, 54] A few years ago, the European Commission updated the regula#ons on fer#lizing products (Regula#on 2019/1009) with Regula#on 2021/2088. The regula#on update was made, amongst others, to promote fer#lizing products made from biomass wastes to enhance the implementa#on of carbon sequestra#on by biochar systems. The new regula#on broadly extended the feedstock base for bio-fer#lizers. Since the update, biochar can be produced from feedstock like forestry and agricultural residues, animal by-products, industrial biowastes (food produc#on wastes, organic wastes from the paper industry, etc.), biodegradable urban waste, and sewage sludge. Extending the feedstock base to bio-wastes without specific market applica#on and with prices lower than market-established biomass has the poten#al to reduce the costs of implemen#ng biochar systems. That relates especially to bio-waste feedstock with specific proper#es (e.g., high ash and nitrogen content), which does not impose high suitability for other applica#ons (like biomass-derived liquid fuels). On the contrary, biochar derived from biomass with the above-men#oned proper#es may show substan#al benefits upon applica#on in soil. Overall, due to the regula#on update, it is forecasted that the implementa#on of biochar systems will gain interest and become a relevant factor toward carbon neutrality.[53, 54] 3. Adding value by the biomass processing of biomass 3.1. Biomass and its poten,al The current impact of biomass on the low-emission economy can be derived from the global use of biomass in energy applica#ons. In accordance with the informa#on provided by the World Bioenergy Associa#on (WBA), in 2021, the global use of biomass for energy purposes accounted for 54 EJ.[55] The value includes 45 EJ derived from solid biomass and ca. 3 EJ derived from the municipal and industrial bio-wastes. With a rough assump#on of higher hea#ng values (HHVs) for both biomasses at 17 MJ/kg on a dry basis (1 EJ = ca. 58.8 Mt), in 2021, the global use of solid biomass for bioenergy purposes was ca. 2,650 Mt, and bio-wastes were ca. 175 Mt. The same report indicates that the installed industrial bioenergy capacity worldwide accounts roughly for the use of 5 EJ of biomass annually (ca. 250 Mt of dry solid biomass). From the comparison with the total global primary energy supply for 2021 (fossil + nuclear), which accounted for 614 EJ, it can be derived that biomass was responsible for ca. 9 % of the total produced primary energy. The trend indicates that the amount of energy derived from biomass has increased over the last decades. However, with the corresponding trend in the global energy produc#on increase, the share of bioenergy in the total mix rose only marginally.[55] The report prepared by the WBA accounts for biomass that is “actually used” without indica#ng the source. Therefore, the biomass data indicated in the report does not reflect “sustainable biomass”, so biomass that is sourced in a low-emission manner, considering the effect of sourcing on the soil carbon pool and the local ecosystems. CHAPTER 1 9 Contrarily, according to the IPCC, the projec#ons and es#ma#ons of the impact of using biomass on decarbonizing and GHG emission reduc#on relate only to the use of sustainable biomass. The biomass pool used for the es#ma#ons and forecasts nowadays is counted as the annual availability since the sustainability goal induces such an evalua#on. To some extent, the es#ma#on of the biomass pool in the environment resembles the es#ma#on of mineral reserves. In an assessment of a mineral reserve, the amount of the theore#cally es#mated reserve rarely matches the amount of the proven reserves. Also, the amount actually economically favorable to be mined is oMen much lower than theore#cally es#mated at the beginning. Contrarily to the mining nomenclature, the biomass pools and their annual availability are defined not as reserves but as poten#als, and types of poten#als are defined within generally assumed frames. The structure and rela#on between types of biomass poten#als are presented in Fig. 1.1, as it was ini#ally proposed by Ba#dzirai et al.[56] Fig.1.1. Division of the biomass poten#als in rela#on to availability, economy, sustainability, and feasibility of use.[56] The total available biomass (theore#cal poten#al) represents all biomass available globally, showing the highest poten#al among all other types of poten#al. However, it naturally accounts for biomass that cannot be sourced, as the harves#ng, collec#ng, and transpor#ng of every growing biomass is prac#cally impossible.[57, 58] Further, the amount of biomass that can be prac#cally sourced (technical poten#al) may be burdened with extensive sourcing costs, making it una9rac#ve to the market. Therefore, the biomass that can be used with sufficient economic profit accounts for the economic poten#al. The economical use of biomass depends on numerous factors, such as biomass quality and suitability toward specific use, local business environment, available quan#ty, price, and supply chain stability. Furthermore, the economic poten#al of biomass fluctuates due to changes in macroeconomic factors like fossil fuel prices.[57-60] Aside from economics, the other type of restric#on in biomass sourcing is sustainability (sustainable poten#al), which was derived from the fact that sourcing some biomass may lead to counterproduc#ve effects in terms of the environment. A straighTorward example is sourcing biomass from rainforests, which may be economically favorable, but it directly leads to a decrease in biodiversity. However, as a measure for GHG emission reduc#on, only the poten#al of sustainable biomass can be considered.[5] The poten#al of economic and sustainable biomass does not fully match, so their common poten#al is lower than the pool of each. Finally, marketa9rac#ve and sustainable biomass poten#al may be considered for a specific applica#on and conversion into a valuable product (implementa#on poten#al). Therefore, in a cri#cal assessment, the biomass poten#al aMer imposing all the restric#ons should be considered as the decarboniza#on measure in shortand mid-term scenarios. The implementa#on poten#al of biomass is complex and difficult to es#mate due to its dependency on several #me-unstable factors. So, the poten#al for implementa#on is rarely provided in the literature. For situa#ons when it is, data is usually provided for a specific applica#on, region, and #me as it relates to a par#cular case, and the generaliza#on of conclusions based on such data will induce a biased view. Therefore, in the es#mates, biomass poten#al is usually provided on a sustainable poten#al basis as such poten#al is feasible to be es#mated with an acceptable level of confidence. CHAPTER 1 10 The sustainable biomass poten#al is an es#mated value and not a value that is analy#cally measured. Therefore, in literature from the last decade, various values can be found depending on the es#ma#on method (and some#mes referring to different types of poten#al). Also, changes in the es#ma#on methodology took place over the years, which adds to the difficulty of comparing es#mates between reports from different years.[57, 61] The es#mates that can be found in the literature (from decades before), especially the long-term ones, rarely precisely match the situa#on they forecast. As generic flaws of economic forecasts of biomass and its deriva#ves can be men#oned: unpredictable changes in macro-economic factors (e.g., changes in the oil prices), the complexity of the biomass market (e.g., food vs. fuel debate, insufficiently stable poli#cs, and local dependencies) and inconsistencies in decisions of the market en##es (difference between stated and executed ac#ons by customers and industrial investment decisions). However, forecasts usually predict overall trends well, while overly op#mis#c predic#ons are also common. Therefore, for the cri#cal assessment, it seems advisable to base the es#ma#ons and reasoning on the available data regarding the ongoing situa#on. Furthermore, for deriving the assump#ons, it can be advised to rely on short-term forecasts from respectable ins#tu#ons with the highest confidence in the “low” scenarios, so rela#vely “pessimis#c” scenarios, as such (so far) have had the highest precision of predic#on. The more prac#cal issue with the assessment of sustainable biomass availability is the grouping of biomass from a specific origin (taxonomy), which differs between ins#tu#ons. On the one end, the (recent) reports under the supervision of the European Commission are based on the taxonomy provided in RED II (2018/2001), Annex IX A and B. On the other end, the reports from the USA are usually based on the taxonomy established by the Interna#onal Renewable Energy Agency (IRENA).[62] However, even in the reports using the same taxonomy as ones provided lately by EU-related organiza#ons, the es#ma#ons of biomass poten#als differ.[63-65] The es#mated maximum sustainable biomass poten#als for the EU (forecast for 2020 and 2030) and the same global poten#al (forecast for 2030), are presented in Fig. 1.2. For the overview purpose, the detailed taxonomies were simplified into 3 major groups: forestry (merged wood-related biomass, including forest and industrial wastes), agriculture (merged agriculture-related biomass, including manures, energy crops, and oil crops residues) and bio-waste (no changes, including urban and municipal bio-wastes). The es#ma#ons were based on the most recent from reputable ins#tu#ons. For the EU-related data, the values were presented as the average of the available data included in the reports from the Directorate-General for Research and Innova#on of the European Commission (DG R&I) from 2017, the Joint Research Centre (JRC) of the European Commission from 2019 and Imperial College London (ICL) from 2021, commissioned by the Oil Companies’ European Associa#on for Environment, Health and Safety in Refining and Distribu#on (Concawe).[63-65] The data for the world (forecast for 2030) were sourced from the report of the Interna#onal Renewable Energy Agency (IRENA) from 2014, despite the recent work by Errera et al. indica#ng this forecast as overly op#mis#c.[57, 60] Data from the report were adjusted to match the taxonomy used in EU-related reports. The results in Fig.1.2.A. represent the biomass availability in energy terms, while Fig. 1.2.B. reports the availability in mass terms. The HHV used for conversion was derived from the ICL report, where mass-averaged HHV of the forest biomass was 16.8 ± 0.4 MJ/kg, the agricultural biomass was 16.7 ± 0.4 MJ/kg, and the bio-wastes were 17.1 MJ/kg (one record). The figures do not contain data for aqua#c biomass, as the biomass is not present in the es#ma#on for 2020, and its availability starts in the es#ma#on for 2030 (and even then, with a rela#vely low share). The presented data requires addi#onal explana#on to grasp its meaning comprehensively. Finding accurate and reliable data regarding the poten#al (or actual supply) of suitable biomass in specific years remains elusive. The current legisla#on does not include strict and precise criteria for assessing sustainable biomass. That also relates to the RED II direc#ve, which currently suggests the verifica#on of sustainable biomass use by voluntary cer#fica#on schemes (divergent between countries due to na#onal legisla#on).[66] As the real data is not possible to obtain, it can be indicated that the es#ma#ons regarding sustainable biomass availability are values that currently cannot be verified objec#vely (under the falsifiability rules). It does not necessarily indicate that the deriva#on of the CHAPTER 1 11 es#mates had methodological flaws or did not implement the best prac#ces. However, a ques#on follows, whether the es#ma#ons on sustainable biomass poten#al are reliable for prac#cal use, considering the impossibility of their valida#on. Lack of objec#ve data verifica#on leads to an unusual situa#on in science where the new reports provide es#mates for upcoming #me intervals while the previous forecasts disappear from discussion (nor being validated). It may lead to a counter-logical situa#on, as presented in Fig. 1.2., where a change in the #me intervals is visible, while none of the data (nor the trend) is verifiable. Fig.1.2. Es#mates of the maximum sustainable biomass availability in the European Union with forecast for 2020 and 2030, and global with forecast for 2030; A. – availability in terms of energy (in EJ) and B. – availability in terms of the mass (in million metric tons of dry biomass).[60, 63-65] Heightened efforts were made to find any valida#on dataset (real data from any of the years 20182024) for the es#mates of the sustainable biomass poten#al, which included a search through scien#fic literature and ins#tu#onal reports. However, as the criteria for the assessment of sustainability are vague, the data for the annual availability of sustainable biomass (and its use) is unavailable. The issue is also highlighted in the IPCC report, which indicates that one of the most significant risks in the es#ma#on of biomass-related decarboniza#on in the long term is the uncertainty in the availability of sustainably sourced biomass.[7, 9] It can be presumed that in the future, more precise assessment criteria will be proposed by intergovernmental ins#tu#ons, e.g., ones based on the LCA (or its extension, the LCSA), which would enable quan#fica#on and provision of real data on the sustainable biomass poten#al. Overall, the current situa#on creates great confusion, as consecu#ve es#ma#ons related to biomass-based GHG emissions reduc#on rely on unverifiable es#ma#ons. CHAPTER 1 12 3.2. Decarboniza,on with biomass Despite the low confidence in the data regarding sustainable biomass availability, it enables to deriva#on of a few pieces of informa#on. Considering the total global energy supply (614 EJ, in 2021), the global sustainable biomass poten#al would have to rise 6-fold (from the forecast for 2030) to match the required energy demand.[55] It should be remembered that the value in Fig. 2 is already assessed as op#mis#c.[57, 60] Considering the current trends in the energy supply, a scenario where bioenergy will fulfill the total global energy demand can be considered as rela#vely improbable, while its presence in the mix is not restricted. Considering the EU, the energy produced in 2021 accounted for 52.6 EJ, whereas bioenergy cons#tuted 5.9 EJ (ca. 11.2% total energy produc#on).[67, 68] Therefore, the complete subs#tu#on of the energy demand with bioenergy in the EU can also be assessed as not likely to happen.[57] As pointed out by Diaz et al., biomass in the long-term scenario most probably will reach 15-20 % of the total global energy mix due to the available land-related restric#ons.[69] Despite bioenergy having a low poten#al to fully subs#tute fossil-derived energy, its presence in the total energy mix may be considered favorable, while without the data on the sustainability of the used biomass, it cannot be proven. Nonetheless, it needs to be pointed out that currently, there are no imposed restric#ons (aside from the possible par#al reduc#on of subsidies) against the use of nonsustainable biomass (economic poten#al). The data for economic biomass poten#al is scant, as it is difficult to es#mate. The available conserva#ve es#mate made by the WBA (from 2016) indicates that the global economic biomass poten#al for bioenergy produc#on (combined agriculture, forestry, and waste) will reach 150 EJ in 2035.[70] Therefore, it can be suspected that the economy currently uses ca. 1/3 of the economic biomass poten#al. The available biomass aside, for energy produc#on, may be directed to areas with limited op#ons in decarboniza#on, like avia#on fuel or chemical produc#on, which will lead to more precise decarboniza#on of the economy.[8] As sustainable biomass impacts the reduc#on of GHG emissions, further elabora#on below will relate to that poten#al, despite its, to some extent, ar#ficial character. IRENA forecast for 2030 (“low” scenario) indicates that globally, the woody biomass (wood, wood residues, and forest residues) and herbaceous biomass (straw, stover, energy crops, primarily plantbased waste) represent ca. 81 wt.% of the annually available biomass. So, it can be derived that those biomass types account for ca. 4,650 Mt/yr of sustainable biomass poten#al.[60] For the EU, the ILC report forecasts that in 2030, the share of the men#oned biomass groups will account for 79 wt.% of available biomass. That corresponds to ca. 410 Mt/yr of sustainable biomass poten#al.[65] The men#oned biomass types are referred to in general as the lignocellulosic biomass group. With a rough es#ma#on of 45 wt.% carbon content in lignocellulosic biomass, it can be derived that in the EU, the annual poten#al of sustainable carbon (for conversion) accounts for ca. 185 MtC/yr. Interes#ngly, the WBA es#mated that in 2021, the biomass for bioenergy produc#on (mainly lignocellulosic) accounted for ca. 151 MtC/yr in the EU (economic poten#al).[55] Considering another less sustainably restricted biomass poten#al, Artz et al. indicated that, as of 2015, European forests could provide woody biomass with an amount of 180 MtC/yr (theore#cal poten#al).[8] Furthermore, it was es#mated that on an annual basis, 80 wt.% of the theore#cal poten#al was economically u#lized (146 MtC/yr). In view of the cascade use of woody biomass (Fig. 1.3.), the study es#mated that the combined mass of woodbased residues from cascaded wood use was ca. 83 MtC/yr. Furthermore, through rough assump#ons, it was es#mated that in 2015, the fossil-based carbon embedded in the plaTorm chemicals or highvalue chemicals by the European chemical industry was ca. 64 MtC/yr.[8] With an applica#on of a crude conversion factor of biomass carbon into carbon embedded in chemicals at a value of 50 wt.% (conversion losses to CO 2 ), it was derived that wood waste has the ability to embed ca. 42 MtC/yr in biomass-derived conven#onal chemicals. Such a value may be insufficient to cover the carbon pool embedded in the chemical produc#on from fossil-based feedstock. However, if the es#ma#ons were applied to all currently used wood, such an amount should be sufficient (93 MtC/yr).[8] CHAPTER 1 13 Fig. 1.3. Simplified scheme of the cascade use of biomass to maximize its posi#ve impact on GHG emission reduc#on (Source: IEA Bioenergy).[71] Aside from the produc#on of chemicals, sustainable avia#on fuels (SAF) produc#on is also burdened with a cri#cal need for biomass, as its conversion is one of the few feasible routes to derive avia#on fuel. In accordance with the es#ma#ons provided by the European Union Avia#on Safety Agency (EASA) the demand for avia#on fuels by 2030 will account for ca. 46 Mt/yr.[72] The SAF blending mandate at the EU level imposed by Regula#on 2023/2405 (ReFuelEU Avia#on) induces a requirement for 5 wt.% of SAF content in avia#on fuel in 2030. The same regula#on imposes the increase of the share of SAF in avia#on fuel to 32 wt.% in 2040 and to 63 wt.% in 2050. Therefore, as of 2030, the demand for SAF can account for ca. 2.5 Mt/yr. Averaging the carbon content of the currently produced SAF, a reasonable assump#on can be pointed out as 85 wt.% (JET-A1 carbon content 83–90 wt.%).[25] Therefore, the carbon needed for SAF in 2030 would account for ca. 2.1 MtC/yr. With the use of the conserva#ve conversion factor of 50 wt.%, the required amount of biomass for SAF produc#on in 2030 could be es#mated at ca. 4.2 MtC. Considering the amounts of wood waste indicated in the previous paragraph, biomass can certainly cover the carbon needed for SAF produc#on. Assuming a constant demand for avia#on fuels (at the 2030 level), the SAF demand for 2040 and 2050 in rela#on to the blending regula#on would increase to ca. 13 MtC/yr and 34 MtC/yr, respec#vely. That would correspond to the demand for sustainable biomass in 2040 of 26 MtC/yr and in 2050 of 64 MtC/yr. Such amounts seem to be within the range of the forecasted sustainable biomass poten#al (for 2030), even assuming no increase in sustainable biomass availability in the future. Despite the available biomass poten#al, providing SAF to the market in the quan##es forecasted for 2040 and 2050 will be difficult. It is forecasted that SAF produc#on will be hindered due to complex issues (feedstock and product specificity, produc#on technology, and biomass supply chain), details of which can be found in relevant reports.[24, 25, 72] CHAPTER 1 14 From a broader perspec#ve, it can be derived that the whole world, including the EU, took a very ambi#ous route toward decarboniza#on (net zero GHG emissions in 2050). The route toward decarboniza#on leads through consecu#ve reduc#on of fossil-derived energy use and its replacement with low-carbon energy, like energy sourced from sustainable biomass. Furthermore, CDR solu#ons seem necessary as not every area of the economy can be (or will be sufficiently fast) en#rely decoupled from fossil-based resources. Therefore, aside from the decarboniza#on of chemical produc#on and avia#on transport, there is a strong need for CDR solu#ons that could be implemented in the shortand mid-term. As indicated in the IPCC report, biomass-related technologies such as biomass carbon capture and sequestra#on (BECCS) and biochar systems are among those solu#ons.[5, 9] The very recent EU regula#on 2024/1735 (Net-Zero Industry Act) from June 2024 indicates that by 2030, the EU industry will have to be able to capture 50 Mt CO 2 -eq./yr from all applied CDR solu#ons. The quota will increase in the consecu#ve decades to 280 Mt CO 2 -eq./yr in 2040 and 450 Mt CO 2 -eq./yr in 2050. In terms of the captured and sequestered carbon, for 2030, the annual capture would have to account for ca. 13.5 MtC/yr (76.5 MtC/yr in 2040 and 122.8 MtC/yr in 2050). As of 2021, the bioenergy produc#on in the EU accounted for 151 MtC/yr, which theore#cally could be sequestrated with BECCS technology, leading to sufficient carbon emission reduc#on (even for 2050). However, the biomass currently used is not accounted for as sustainable, so its use cannot be accounted for as net-zero, and sequestra#on of its CO 2 produced upon combus#on as carbon nega#ve (in full, while par#ally it is possible). Considering the use of sustainable lignocellulosic biomass poten#al (es#ma#on for 2030), the CDR emission goal with the BECSS could be achieved, leaving an excess of sustainable biomass in the amount of ca. 170 MtC/yr for other needs. The es#ma#on for CDR in 2050 (biomass poten#al as in 2030) is a bit less op#mis#c, as it indicates that available biomass poten#al will account only for 62 MtC/yr of sustainable biomass available. Such an amount may be insufficient to reach the set decarboniza#on goals. The scenario poses a worry that all the economic sectors cannot be simultaneously decarbonized and that such ac#on will lead to a significant (sustainable, lignocellulosic) biomass deficiency, especially considering the higher success rate and precision for the “low” scenario forecasts with respect to the increase in biomass availability. Biochar systems may be a way to provide addi#onal CDR in the case of an insufficient (sustainable, lignocellulosic) biomass supply. Biochar may be produced from biomass with a rela#vely low HHV and carbon content and elevated ash, sulfur, and nitrogen (manure, sewage sludges, and bio-wastes), which is unsuitable for other applica#ons (SAF, bio-chemicals). As the forecast for 2030 indicates, the poten#al of sustainable non-lignocellulosic biomass in the EU accounts for ca. 110 Mt/yr (manure, oil crops, and bio-wastes). As the biomass composi#on is characterized by high heterogeneity with a carbon content within the range between 20 wt.% and 40 wt.%, it can be es#mated that the non-lignocellulosic biomass pool contains between 22 MtC/yr and 44 MtC/yr. The factor for the carbon reten#on in biochar for non-lignocellulosic biomass can be roughly assumed to be 50 wt.%.[73] The assump#on would lead to the possibility of sequestering 11 MtC/yr to 22 MtC/yr the biochar derived from non-lignocellulosic biomass, which would also be sufficient to meet the CDR requirements for 2030. However, even with the aid from biochar systems, the CDR goal for 2050 s#ll would seem hard to reach. Simultaneously, the forecasted biomass shortage may accelerate CDR with other non-biomass-related solu#ons, like reforesta#on, which also counts for carbon sequestra#on.[5, 9] From a broader perspec#ve, it can be indicated that one of the greatest risks of reaching net-zero by the EU (and globally) may be basing the forecasted effects on es#ma#ons, which remain challenging to validate. Biomass use will remain necessary to achieve decarboniza#on, and its availability will play a cri#cal role in transport fuel and chemical produc#on (and forecasted CDR). A generic conclusion can be drawn that wood biomass cannot be considered for all applica#ons at once, as its resource has limita#ons, and that non-woody biomass (agricultural, bio-waste) should be strongly considered for applica#ons, especially to achieve the forecasted GHG emission reduc#on. The success related to biomass use lies in the readiness of the technologies for biomass conversion toward a required outcome. Therefore, in the upcoming decades, significant efforts are forecasted to be made concerning the development of low-carbon technologies, which may bring unforeseeable progress. CHAPTER 1 15 Despite the worrying prognosis, the ac#on toward decarboniza#on will lead to the reduc#on of GHG emissions, which is beneficial for the environment and social security. Moreover, a stronger coupling of the economy to biomass may bring tangible benefits through increasing the self-sustainability of the EU economy (bio-economy).[74-76] Therefore, the endeavor may lead to more profit than expected, regardless of not reaching the goal within the assumed #me. Also, the op#mis#c conclusion is that the European economy is s#ll far ahead of the situa#on regarding the route toward decarboniza#on. Therefore, as the beginning is the most challenging part, the prac#cal solu#ons may present themselves along the way if well-thought-out efforts are conducted in the upcoming years. 3.3. Biomass and its processing Biomass refers to renewable organic material that comes from plants and animals. As animal biomass does not cons#tute biomass considered for decarboniza#on, the focus will be on plant biomass. In a general perspec#ve, biomass and its deriva#ves, similar to crude oil and coal, can be considered on a few plaTorms, like materials, fuels, and chemicals. For example, wood can be mechanically converted into #mber (direct) or stepwise, chemically into paper (indirect) and, as a fuel, combusted (direct) or converted thermo-chemically into bio-oil (indirect). Another plaTorm on which biomass can be viewed, among others, is the sustainability of its sourcing, composi#on, and origin. The literature includes various divisions of biomass (e.g., standards EN 14961 and EN 15234) and bio-wastes (e.g., RED II taxonomy). The most common division of biomass is presented in Fig. 1.4, which is based on a division of biomass in view of its applica#on as feedstock for biofuel produc#on (for transporta#on).[16, 77] Fig.1.4. Simplified biomass division for biofuel produc#on in regard to its genera#on, content, and relevant groups. The biomass division into genera#ons assumes the following categories: 1 st Genera,on – represents agricultural products such as starch, oil, and sugar crops. The biomass included in the 1 st genera#on represents primary biomass that is mainly edible, so it can be used for food or animal feed produc#on. 2 nd Genera,on – is the widest group, as it includes almost all non-edible, terrestrial biomass. Groups such as lignocellulosic biomass (forestry and agricultural origin) and non-lignocellulosic biomass (manure, sludges, bio-wastes) can be dis#nguished within this genera#on. The division also relates to primary biomass (e.g., wood) and secondary biomass (e.g., residues from wood processing like sawdust). It usually relates to non-edible biomass composed of cellulose, hemicellulose, and lignin, which does not contain significant amounts of starch, saccharides, or oils. Such an assump#on is incomplete as the genera#on also includes the wastes and residues, which may contain high-calorific components (waste oils) and other components (like digestates, manures). CHAPTER 1 16 3 rd Genera,on – represents biomass ar#ficially grown in aqua#c environments (marine or land, oMen in non-arable areas). The genera#on includes algal biomass (microand macro-algae), characterized by intensive growth over a short #me period. The typical algae composi#on relates to high lipid and protein content. The biomass does not represent a significant poten#al as of the 2030 forecast, while in consecu#ve periods, it is es#mated to significantly add to biomass poten#al. As can be deduced from the provided descrip#on of biomass genera#ons, such a division incompletely organizes the biomass and induces vagueness. A clearer systema#c of biomass division was proposed by Basu and Kaushal, whose division is based on biomass type (primary and secondary) and groups related to biomass origin.[78] The division can addi#onally be extended with key indicators of biomass sub-groups relevant to its further conversion, like sustainability poten#al (SUS), higher hea#ng value (HHV), and moisture content. The sustainability poten#al was indicated subjec#vely in rela#on to the compe##on between food vs. fuel, the possibility of sustainable sourcing, and the poten#al for consecu#ve cascade use. The category low sustainability poten#al was indicated for the 1 st genera#on biomass, as its compe##on with the food and animal feed applica#on is evident. Furthermore, farming sugar, starch, and oil crops has a high requirement for water and fer#lizers (aside from the use of pes#cides and herbicides), which leads to soil deple#on in the long term. Finally, expanding farming areas for fuel purposes leads to deforesta#on and land-use change, which relates to risks of releasing the already sequestrated GHG and reducing biodiversity (e.g., through monoculture).[15, 79] The proposed division of the biomass with an indica#on of the key proper#es is presented in Fig. 1.5. Fig. 1.5. Detailed division of biomass with dis#nc#on on the primary and secondary biomass groups and an indica#on of the key proper#es of biomass subgroups. Despite the indicated flaws of the 1 st genera#on biomass, its use in the produc#on of biofuels is a common prac#ce, and technologies for their conversion into biofuels are commercially available (TRL 9). In 2021, to address the issues related to the 1 st genera#on of biofuels, the European Commission, through the RED II direc#ve, took the ini#a#ve to phase out food crop-based biofuels (conven#onal biofuels). It translates to the obliga#on of the European transport fuel suppliers to provide more sustainable biofuels, like advanced biofuels (biofuels from 2 nd genera#on biomass), CHAPTER 1 23 form of glucose derived from cellulose) and levomannosan (from the glucomannan part of hemicellulose). The depolymeriza#on of those polysaccharides also leads to the forma#on of furanic deriva#ves, like 5-HMF (from cellulose) and furfural (from both xylan and glucomannan parts from hemicellulose). The degrada#on through fragmenta#on of all bio-components is, to a different extent, responsible for the forma#on of compounds like acids (ace#c and formic acid), ketones (hydroxyacetone, acetaldehyde), and alcohols (mainly methanol).[116] It is worth highligh#ng that the bio-oil cons#tutes the mixture of the abovemen#oned organic compounds and water, so it is highly homogeneous regarding the compound groups it includes.[106, 137] Considering the composi#on of the pyrolysis gases (noncondensable), up to ca. 400 °C, the majority share of the evolved non-condensable gas cons#tutes CO 2 . Further degrada#on is responsible for the forma#on, among others, of CO and CH 4 , which are not prevailing pyrolysis gas components un#l roughly 700 °C. Therefore, in general considera#on, up to 500 °C, the pyrolysis gases represent a product with a rela#vely low high hea#ng value (ca. 5 MJ/kg).[112] The forma#on of char relates to the conversion pathway in condi#ons in which the ac#vated biocomponent does not receive sufficient energy to degrade through depolymeriza#on and fragmenta#on, or their products are unable to be released from the surface of ac#vated biocomponents (solid). The release of the evolved molecules (vapors) requires, among others, their evapora#on, which is an energy-demanding phenomenon, especially for larger molecules (having a higher boiling point). Like conven#onal evapora#on, evapora#on of molecules during the pyrolysis process depends on temperature and pressure during conversion. When insufficient energy is provided to form and release deriva#ves of bio-components, the ac#vated bio-components and un-released deriva#ves start to thermally degrade, releasing oxygen and hydrogen from their structure. The evolved vapors may also re-deposit on the remaining structures and then undergo degrada#on. Such a phenomenon may occur when conversion condi#ons include insufficient removal of vapors from the reac#on environment. Consequently, a new carbonaceous solid material, namely char, is produced, which is characterized by an elevated carbon content in comparison to the ini#al bio-component. Char that originates directly from the bio-component is called the primary char (primary charring route). Along with the increase of the pyrolysis temperature, the structures enclosed in the primary char undergo further degrada#on, which involves various bond cleavages. Through such a mechanism, among others, CO 2 , CO, and H 2 are released from the char, which leads to a further increase in the carbon content of the char, a process also termed char matura#on.[111, 116] The release of oxygen and hydrogen from the char structure induces the forma#on of carbon fused rings (aroma#c), from linear carbon chains (alipha#c), which further organize themselves into polyaroma#c structures (conjugated aroma#c rings). Overall, nuclear magne#c resonance (NMR) indicates that around 300 °C, the char structure is composed of carbon built out of 1–2 aroma#c rings. That number rises to 7–13 around 500 °C, while a further increase of the pyrolysis temperature brings the polyaroma#c structure of the char to ca. 30 aroma#c rings (at ca. 700 °C).[138, 139] Below 500°C, the bio-component-derived structures enclosed in char represent carbon ma9er in a disorganized manner. The enclosed polyaroma#c structures are loosely bonded and rela#vely small, which does not allow them to stack (form graphene-like structures). Around 500–600 °C, when the primary degrada#on of bio-components is completed, the solid ma9er formed in the pyrolysis process is represented only by their charred deriva#ves. As the process of the polyaroma#c structure growth and its stacking is temperature-related, with its increase, the char structure becomes composed of larger and be9er organized graphene-like units. However, the structure is not homogeneous, and such units are s#ll loosely bonded to each other. Around 600-700 °C, the graphene-like units start to merge, forming a stronger bonding in between and releasing the noncarbon elements from the edges of the stacks. Such a process is related to the intensive forma#on of hydrogen and methane, which further enriches the pyrolysis gas (however, with minor impact).[116] However, the graphene-like structures, despite being larger, s#ll have not developed a strong bond between units. Such a situa#on appears only around 900 °C and leads to the forma#on of the semi-graphite, crystalline char structure, which is CHAPTER 1 24 well and strongly bonded between units.[140, 141] The process of re-organiza#on of the char structure, in the literature, is called graphi#za#on, and some of the sources indicate that pyrolysis finishes when such a phenomenon starts (typically 700 °C or 900 °C).[139, 142] Other literature sources account for the pyrolysis process to finish with the comple#on of the primary degrada#on reac#on in biocomponents around 500 °C.[118, 119] This can create confusion, especially among early-stage researchers, where even within one literature source, the indicated temperature range for pyrolysis varies.[87] The Author favors the lower range to define pyrolysis (300-500°C), as its defini#on has a clear and objec#ve boundary (possible to validate experimentally). Overall, as the temperature boundaries are not strict, a principal approach seems unnecessary. When pyrolysis is conducted at a temperature above 500 °C, it has a non-negligible effect on the evolved vapors. This explicitly relates to the situa#on when the evolved vapors are exposed to temperatures exceeding 500 °C before they are evacuated from the reactor zone. The organic molecules are prone to further thermal degrada#on (thermal cracking), so their exposure to elevated temperatures (for a sufficiently long #me) leads to their decomposi#on. In such a phenomenon, the condensable molecules degrade, among others, to CO, CH 4, and H 2 , which majorly elevates their content in the pyrolysis gas (at the expense of the condensable product yield).[112] The degrada#on of the condensable molecules can also lead to their conversion into monocyclic aroma#c hydrocarbon deriva#ves (MAH) like benzene, toluene, and xylene (commonly known as BTX).[111, 116] It is explicitly relevant for larger molecules, like anhydrous saccharides and phenolic deriva#ves. The suscep#bility of specific molecules to thermal degrada#on is not homogeneous, so the conversion of all evolved condensable vapors does not necessarily occur just above 500 °C. However, the most prone compounds show suscep#bility from such a temperature onwards. Therefore, 500 °C is the commonly assumed threshold value above which secondary degrada#on of condensable vapors starts to occur.[111, 143] The presence of secondary degrada#on of vapors nega#vely impacts the yield of the condensable vapors (bio-oil), so its impact has to be mi#gated in solu#ons that aim for a high bio-oil yield. A special case of the secondary degrada#on of vapors is soot forma#on, also called secondary carbon/char forma#on. In favorable condi#ons (sufficiently high temperature and residence #me), the MAH compounds can undergo polycyclic condensa#on into polyaroma#c hydrocarbons (PAH), which can also grow further. When the PAH compounds grow sufficiently large, they may form a high molecular, carbon-rich, and non-vola#le par#culate ma9er, differen#a#ng as an aerosol or as a separate solid phase aMer deposi#on.[144-146] It is worth men#oning that secondary char forma#on may occur in the hot reac#on zone (of a reactor) as well as within a converted biomass par#cle if the par#cle is sufficiently large and the temperature sufficiently high.[147, 148] Char origina#ng from this PAH route adds to the overall solid yield and, when it occurs within a par#cle, may impact its porosity. The conversion pathway of bio-components is also related to induced thermal effects, such as heat release (exothermic reac#on) or heat absorp#on (endothermic reac#on). Furthermore, the heat effect is strongly bound to the reac#on rate, so a strong thermal effect will correspond to the case of rapid conversion. Studies that involve differen#al scanning calorimetry (DSC) are oMen used to analyze the thermal effect of bio-components and biomass during pyrolysis.[123, 149, 150] Conversion of hemicellulose (as xylan) leads to heat produc#on over a narrow temperature range corresponding to the peak of conversion (ca. 250–270 °C). The further decomposi#on of hemicellulose char also shows an exothermic effect, but not as high as the bio-polymer decomposi#on (up to 500 °C). Lignin degrada#on is also exothermic over a broad temperature range, with a no#ceable peak around 350– 400 °C, and the effect stays un#l 500 °C to a lesser extent. Cellulose degrada#on relates to a strong endothermic effect with a sharp peak at ca. 325 °C. Interes#ngly, the subsequent cellulose-char degrada#on has a slightly exothermic character. The thermal effects in the decomposi#on of each bio-cons#tuent may lead to a presump#on that some biomass could be converted without the addi#on of external heat to pyrolysis (or minimal to ini#ate the reac#on). DSC studies of biomass (woody and herbaceous) pyrolysis counter such presump#ons, as they unambiguously indicate the overall endothermic effect in biomass pyrolysis. The conversion of CHAPTER 1 25 biomass in the DSC indicates that un#l cellulose is converted, the thermal effect of conversion is endothermic, while it changes into slightly exothermic aMerwards, roughly around 400 °C. Overall, biomass conversion analyzed with DSC does not support the possibility of autothermal biomass pyrolysis induced only by the heat of the degrada#on reac#ons (endothermic combined thermal effects).[149, 151] The standard DSC measurements are conducted under specific condi#ons where the sample is strongly pulverized. At the same #me, the sample mass is small (a few milligrams), and pyrolysis is conducted in a crucible with an open lid, so the pyrolysis vapors can easily diffuse or be swept away from the reac#on zone. Under such condi#ons, biomass conversion is biased towards depolymeriza#on and fragmenta#on of its cons#tu#ng bio-components, while the primary charring reac#ons are reduced. DSC studies where the sample is converted in a closed crucible instead, so that pyrolysis vapors are retained in the crucible/reac#on zone, enhance the occurrence of primary charring reac#ons and indicate a discrepancy in results from those studies conducted with an open crucible.[151, 152] Those studies where char forma#on was enhanced (closed crucible) indicate that the exothermic effect is present before (hemicellulose-related) and aMer (lignin-related) the cellulose conversion, while the cellulose degrada#on s#ll induces a high endothermic effect. The overall thermal effect from DSC studies using biomass samples in a closed crucible indicates lower heat required for pyrolysis compared to studies using open crucibles, while both indicate the need for heat to drive the process. Aside from the endothermicity of pyrolysis (depolymeriza#on and fragmenta#on), the results indicate that depolymeriza#on and frac#ona#on pathways lead to a higher heat requirement than the char forma#on pathway. Therefore, with the enhancement of primary char forma#on, the overall heat requirement for the pyrolysis process can be reduced, but not en#rely.[111] 4.2. Pyroly,c technologies for the produc,on of bio-based char The following sec#on is a synthesis of several literature sources, where details regarding the technical and economic ma9ers of pyrolysis technology can be found.[84, 87-89, 101, 104, 105, 111-131] Therefore, the elabora#on will focus on a general overview with an indica#on of the common and dis#nc#ve features of the different pyrolysis technologies. Only specific informa#on will be supported by references to avoid using references redundantly. The previous Sec#ons used terms like slow and fast pyrolysis (e.g., Fig. 1.7), but their dis#nc#ve features have not yet been provided. The slow pyrolysis process historically precedes the fast pyrolysis process, as the former was already carried out in ancient #mes, and the la9er was developed in the late 1980s. However, technologically, slow pyrolysis on an industrial scale, precedes fast pyrolysis only by roughly 50–100 years, as the primary technological developments for slow pyrolysis took place between the 1880s and 1930s.[101] As both processes have undergone decades of development, various technological concepts have been provided, which cons#tute the basis for dis#nguishing or grouping the various pyrolysis technologies. Based on different criteria, divisions of the pyrolysis processes can be found in the literature. Considering the division, one suggested by Basu and Kaushal indicated grouping the processes based on the #me required to heat the fuel (biomass) to the pyrolysis temperature (t h ) and characteris#c pyrolysis reac#on #me (t r ), defined as the #me required to convert the biomass par#cle.[89] In such a division, the slow pyrolysis process is characterized by t h >> t r and the fast pyrolysis process is on the opposite. In theory, the division seems quan#ta#ve and possible to be applied objec#vely. In prac#ce, this division is vaguely applicable, as it requires prior knowledge of the thermo-physical characteris#cs of specific biomass, its dimensions and composi#on (or thermal degrada#on characteris#cs), and processing condi#ons (final temperature and heat influx to biomass). Another division, based on the hea#ng rate, also seems arbitrary.[112] For a more quan#ta#ve grouping, a prac#cal division can be the use of a threshold value of 1 °C/s (higher indicates fast pyrolysis, lower indicates slow pyrolysis) as suggested by AncaCouce.[111] In rela#on to informa#on commonly provided in the literature, a simplis#c and intui#ve division of the various pyrolysis processes can be found in Table 1.1. CHAPTER 1 26 By its defini#on, the fast pyrolysis process is a thermochemical process that aims to produce bio-oil from biomass.[106, 137] Therefore, its dis#nc#ve features correspond to the processing condi#ons that are able to meet this aim. To maximize the bio-oil yield in the pyrolysis process, the charring reac#ons need to be minimized, which relates to the use of a high hea#ng rate and short vapor residence #me in the high-temperature zone (reac#on zone). Fine par#cles (a few mm) are used to convert biomass in such condi#ons to ensure complete conversion within the required #me. Such a boundary condi#on is necessary, especially considering the low thermal conduc#vity of biomass and its char. As the pyrolysis process of fine par#cles is primarily endothermic, the energy for the process is supplied by the combus#on of the side products, namely char and evolved gases. As prac#ce indicates, side product combus#on provides sufficient heat needed for fast pyrolysis. The details of specific technologies developed for the fast pyrolysis process, like the commercially available technology developed by BTGBTL B.V., can be found, among others, in the review by Venderbosch and Prins.[137] Regarding Table 1.1, it is important to note that some literature sources suggest fast pyrolysis hea#ng rates may extend up to 1,000 °C/s (or even 10,000 °C/s). However, such high rates are not used in industrial fast pyrolysis prac#ce. The men#oned hea#ng rates represent “flash pyrolysis,” a theore#cal, extreme solu#on developed in the 1980s to push the boundaries of bio-oil produc#on. In prac#ce, bio-oil yields plateau at around 70–75 wt.% under op#mal condi#ons (process, feedstock), even as hea#ng rates increase beyond 100–200 °C/s. Therefore, applying these extreme hea#ng rates offers no significant economic advantage and introduces significant technical challenges and costs. Informa#on about a process called intermediate pyrolysis, which aims for simultaneous bio-oil and char produc#on, can also be found in the literature. Intermediate pyrolysis may be less efficient toward bio-oil yield (and higher water yield) than fast pyrolysis, while bio-oil s#ll cons#tutes the majority of its products. It usually refers to pyrolysis based on a screw reactor, where the heat is provided by a carrier (hot sand), and biomass residence #me is counted in seconds.[153] Such systems were further op#mized and upscaled toward fast pyrolysis technology, e.g., the Bioliq® process developed by KIT.[154] Therefore, making a separate dis#nc#on between fast and intermediate pyrolysis processes seems unnecessary. Table 1.1. Division of pyrolysis processes in rela#on to their targeted pyrolysis product and their dis#nc#ve features.[89,106,112,114,122,137] fast pyrolysis slow pyrolysis continuous (semi-) continuous (semi-) batch main product oil char (+oil) char (+oil) heating rate 10 1 – 10 2 °C/s 10 -1 – 10 0 °C/s 10 −3 – 10 -1 °C/s residence time seconds minutes/hours hours/days particle size µm – mm mm – cm cm – m energy recovery char+gas gas+oil / only gas none / gas+oil As the pyrolysis process in any variant does not aim for gas as its major product, the logical comple#on of the division is that the slow pyrolysis process aims for char produc#on. That would be only par#ally true, as some slow pyrolysis technologies focus on simultaneous char and bio-oil produc#on. Moreover, in the 20 th century, most established endeavors conver#ng wood in a slow pyrolysis process aimed mostly at liquid products (organics and tar), while the char was considered as a secondary product. However, as such technologies were not yet op#mized toward bio-oil produc#on (minimiza#on of charring pathway impact), and so despite the aim, the yields of bio-oil and char, in the best cases, were similar. The technological progress of the slow pyrolysis process started over a century ago, so through the years, its porTolio has been enriched by various systems, star#ng from small-scale solu#ons with a basic opera#on, finishing on large systems with complex product collec#on and heat recovery (while such systems are no longer common in opera#on as of today). CHAPTER 1 27 The literature also shows a division between carboniza#on and conven#onal (slow) pyrolysis processes. The former relates to char produc#on in basic systems (e.g., meilers, pits, and kilns), where a batch process takes days, is primarily uncontrolled, and resembles the solu#ons from centuries ago. The la9er refers to the technologically developed solu#on, mainly con#nuous and semi-con#nuous, where conversion takes place within minutes or hours. As both processes aim for char produc#on (slow pyrolysis), the dis#nc#on between carboniza#on and conven#onal slow pyrolysis technologies, especially based only on the characteris#cs of those technologies, seems incomplete and misleading. A more detailed and comprehensive division of slow pyrolysis technologies was proposed on Fig. 1.9. It should be pointed out that the division is generic and does not include the nuances of the technological developments of specific endeavors that used slow pyrolysis technologies. A more detailed overview of the opera#onal concepts, alignment of plants, and heat recupera#on with slow pyrolysis endeavors, including wood dis#lleries, can be found in the relevant literature.[101, 104, 105, 122, 124-131, 155] The most generic division of slow pyrolysis technologies can be based on the opera#on regime, as it has a non-negligible impact on the technical boundaries of a solu#on. One of such is the par#cle size, which in batch systems can reach par#cles measured in the meter scale, e.g., logs (less than 1 m), while con#nuous systems, e.g., those equipped with rota#ng elements, are restricted to use par#cles with cen#meter–millimeter size ranges. The produc#on regime is also related to the residence #me, which may account for minutes or hours for con#nuous systems (e.g., rotary drum, auger, Lambio9e, and Herreshoff systems) or semi-con#nuous systems (e.g., wagon system). On the other end are semi-batch systems (e.g., tandem/twin and Reichert retort system) and batch systems (various singular retorts and kilns), in which residence #me is counted in hours or days. Fig. 1.9. General division of the slow pyrolysis technologies based on produc#on regime, reactor material, hea#ng op#miza#on (recovery and recupera#on), technological possibility to recover bio-oil from evolved vapors, and retort alignment with exemplary technologies included.[101, 104, 105, 122, 124-131, 155] The ma9er of heat management in slow pyrolysis systems has always been of major importance, and the available solu#ons reflect the crea#vity of the inventors throughout the years. For simple systems like kilns, the heat to the process is provided by the direct addi#on of air into the reac#on zone, which induces the combus#on of evolved vapors and, to some extent, the derived char (apparent losses). Another commonly occurring solu#on is the collec#on of the evolved vapors, their consecu#ve combus#on in a separate unit (combustor), and then direc#ng hot flue gases into the reac#on zone (e.g., in batch retort systems, as well as wagon, Lambio9e, and Reichert systems). CHAPTER 1 28 The direct hea#ng op#on also includes solu#ons where the externally combusted gases heat an inert heat carrier (e.g., sand or metal pieces), which is further directed to the reac#on zone. Such solu#ons can be found in con#nuous systems, especially in auger systems. For systems working in such a regime, another common solu#on is the provision of heat by flue gases into the reac#on zone indirectly via the reactor wall (i.e., through a double-shelled reactor wall). In general, pyrolysis systems are constructed in such a way that the energy for the process is sourced from the ini#ally fed biomass, where the energy for the process is recovered from the side products. In terms of fast pyrolysis, the energy is usually recovered from char and evolved pyrolysis gases, while for slow pyrolysis, it stems from all evolved vapors (bio-oil and gas). For the la9er systems, solu#ons can also be found where the evolved vapors are condensed (par#al or complete bio-oil recovery), and which use only evolved non-condensable gases as a heat source. However, such a technological solu#on requires a precise op#miza#on of the heat recupera#on from the collected products and heat recovery from the exothermic part of the process, as the gases evolved in pyrolysis have a low HHV.[112] A cri#cal point of such heat management is that the feedstock needs to have rela#vely high HHV and low moisture content to enable a proper energy balance. For cases where low-quality biomass is converted, an addi#onal source of energy may be required (combus#on of char or providing other external fuels). An example of such a situa#on is conver#ng sewage sludge into char with the use of the Herreshoff system.[101] As slow pyrolysis is aimed at char produc#on, the technologies involved should be op#mized for char yield. Compared to bio-oil produc#on, a complete change of the biomass degrada#on mechanism towards char produc#on is impossible in prac#ce. Furthermore, if such a situa#on theore#cally could be achieved, it would induce technical problems with energy management in produc#on (no side products would be available to produce heat). Also, as historically proven, slow pyrolysis technologies are able to simultaneously derive bio-oil in the process (from wood), which may cons#tute addi#onal economic value through the produc#on of bio-fuels or bio-chemicals aside from the char. Produc#on of ace#c acid and tar aside from the char can s#ll be traced to ProFagus GmbH in Bodenfelde (Germany), and the Hemijska Industrija Des#lacija a.d. in Teslić (Bosnia).[87] Interes#ngly, both plants use technologies developed in the early 20 th century. The factors that influence the char yield in slow pyrolysis of biomass have already been deeply inves#gated and are directly coupled with the conversion mechanisms of bio-components (Sec#on 4.1). The factors include feedstock-related parameters like bio-composi#on, par#cle size, and alkali content, as well as process-related factors. The la9er includes the final biomass temperature and reactor temperature, the residence #me of evolved vapors in the hot reac#on environment, the hea#ng rate, and the pressure. The char yield is also influenced by the applied pyrolysis technology, as it directly translates to the process condi#ons.[101, 104, 105] The technologies related to batch pyrolysis processes show a variable char yield star#ng from 12–16 wt.% for basic and primi#ve earth kilns (with internal combus#on of products), up to 29–33 wt.% for the more advanced kiln processes like portable metal, brick, and Missouri-type kilns. However, in the men#oned batch technologies, the temperature control is rela#vely basic, as typically, the progress of the process is assessed by observing the color of evolving vapors. This leads to difficulty in accurately assessing the final temperature in pyrolysis, which is usually assumed to be around 400–450 °C. Also, it is not certain if the biomass feedstock is homogeneously converted, as these types of kilns allow for local overand under-hea#ng. Aside from the poor process control, the condi#ons within the batch process seem very favorable for the char forma#on route. In batch technologies, the process progresses very slowly (very low hea#ng rate), the biomass par#cles are rela#vely large (chunks and logs), and the removal of the vapors from the reac#on zone is strongly hindered (only through natural conversion, no use of a sweeping gas). Batch systems represent the least technologically developed systems, with poor control and op#miza#on of heat management. On the other end are semi-batch, semi-con#nuous, and fully con#nuous systems, where inventors no#ced the importance of op#mizing the process condi#ons to minimize material and energy losses. For semi-batch systems like tandem/twin and Reichert systems, CHAPTER 1 29 the char yield accounts for roughly 33–36 wt.%. The la9er system is addi#onally able to derive 20– 25 wt.% of organic liquid yield from wood without any addi#onal heat source. Further, semi-con#nuous systems, like the wagon system, represent the largest-scale produc#on technology, which historically was driven close to its op#mum with minimiza#on of the process heat losses. The char yield from wood in the wagon system accounted roughly for 33–38 wt.% with a simultaneous organic liquid yield of 20– 25 wt.%. Well-op#mized heat management can also be found in the ver#cal con#nuous systems like the Lambio9e and Herreshoff systems, where char yield from wood in the former is ca. 30–35 wt.%, and in the la9er is ca. 25–30 wt.% (with 20–22 wt.% organic liquid yield). The systems men#oned above are suitable for the conversion of rela#vely large biomass par#cles, with a strong preference toward wood logs (up to ca. 0.3 m), while the conversion of small and pulverized biomass with those technologies may lead to difficul#es in the prac#cal processing. Contrarily, rotary drums and screw reactors, so technologies working in con#nuous regimes with horizontal reactors, are suitable for being fed with smaller par#cle sizes. The use of smaller par#cles has pros and cons. From the cons side, the smaller par#cle size leads to a lower char yield, e.g., for auger reactors, the char yield is ca. 20–28 wt.%, and for rota#ng drum reactors, it is ca. 23–30 wt.%. From the pros side, the systems are well-suited for the conversion of residual biomass, which usually has a small par#cle size distribu#on (e.g., sawmill residue, shredded corn stalks, and olive stones). What can be addi#onally considered as a pro is that the horizontal con#nuous systems are characterized by a high bio-oil yield (40–60 wt.% possible for favorable feedstock and conversion condi#ons), which may significantly add to the economics of the process.[101] 4.3. Environmental and market aspects of bio-based chars The technology selected for char produc#on has a non-negligible impact on related GHG emissions. As pointed out in Sec#on 4.1., the pyrolysis process, aside from the evolu#on of CO 2 , relates to the forma#on of CO, CH 4 , light hydrocarbons, and low molecular weight organic deriva#ves. The la9er two combined are oMen referred to as non-methane vola#le organic components (NM-VOC), which include components like ethane, methanol, and ace#c acid.[156] Furthermore, the produc#on of char induces the release of aerosols of solid par#culate ma9er, with par#cle sizes up to 2.5 m (PM 2.5 ) and up to 10 m (PM 10 ). In the literature, combined aerosols of solid par#culate ma9er are referred to as total solid par#cles (TSP).[156] When released along with the flue gases, the aerosols have a nega#ve impact on the local environment and have harmful effects on human health. With the assump#on of the sustainable origin of biomass, CO 2 emi9ed during char produc#on can be assessed as neutral for GHG emissions (net-zero). However, the other gaseous products, whose global warming poten#al (GWP) is higher than CO 2 , certainly cannot be accounted for as neutral. The environmental impact of char produc#on is primarily related to basic batch processes, as such systems commonly do not have (or cannot have) implemented systems for the collec#on of evolved vapors and their combus#on for heat recovery. Therefore, the evolved vapors are directly released into the environment as flue gas. Aside from the GHG emission, the produc#on of char in earth kilns (meilers and pits) induces soil contamina#on, as the reac#on zone is not separated from the ground underneath. As Krzysztofik es#mated in 1968, the produc#on of 1 kg of wood char in an earth kiln relates to the release to the soil of ca. 15.6 MJ of heat energy, ca. 0.8 kg of tars, and ca. 2 kg of acidic water (aqueous mixture including ace#c acid and methanol).[157] Due to the economic (low char yield and high labor intensity) and environmental (soil pollu#on) issues, the use of metal kilns (portable with sealed bo9om) instead of earth kilns became suggested to char producers worldwide. One such ac#on was the propaga#on of the portable metal kilns called TPI kilns (developed by Tropical Products Ins#tute), which the FAO recommended as s#ll simple in use but a more efficient solu#on for char produc#on.[104] CHAPTER 1 30 Studies that assessed the air pollu#on related to char produc#on (emission of specific compounds) in rela#on to the used systems and feedstock can be found in the literature.[156, 158-160] To assess the GHG emissions indicated by sources in quan#ta#ve terms, it was assumed that the GWP of CO is 2.8, CH 4 is 27 (non-fossil), and NM-VOC is 6.8. The la9er value was derived as 25% of the CH 4 GWP, considering the uncertainty regarding the GWP of NM-VOC components. The comparison is presented in Fig. 1.10. Results of the comparison indicate that the produc#on of char with basic technologies relates to a surplus of GHG emissions in the range of 2.0–2.5 kg of CO 2 -eq. per kg of produced char (regardless of feedstock), related only to emi9ed, non-u#lized pyrolysis vapors. As presented in Fig. 1.10., the implementa#on of tar recovery has a non-negligible reducing effect on GHG emission from the process (ca. 1.6 kg/kg). Furthermore, energy recovery (collec#on and combus#on of the evolved vapors) can be indicated as the most effec#ve solu#on to reduce surplus emissions from char produc#on, where for batch systems, the surplus GHG emissions account roughly for 0.5 kg/kg, while for the con#nuous systems, they only account for roughly 0.2 kg/kg. The recent LCA study by Kavindi et al. indicates similar conclusions.[161] Moreover, the study suggests that char produc#on with energy recovery coupled with power produc#on (from heat excess) may even lead to nega#ve GHG emissions, even without sequestering the char-related carbon in soil (biochar system). However, for any scenario inves#gated (with and without energy recovery), the study indicated a notable increase in the nega#ve emissions from char produc#on when char is used in a biochar system. Fig. 1.10. Surplus GHG emission from the produc#on of char from biomass, with the division on used technology and ini#al feedstock (avg. – averaged, EPA – U.S. Environmental Protec#on Agency, IPCC – Intergovernmental Panel on Climate Change).[156, 158-160] The above-indicated surplus emissions relate to scenarios when biomass is sourced sustainably, while the current produc#on uses economically available biomass. Therefore, its use cannot necessarily be considered a feedstock with net zero emissions from the directly evolved CO 2 alone. That is relevant, especially when biomass sourcing may be burdened with harves#ng emissions and nega#vely impact the overall carbon pool (e.g., via deforesta#on).[162] The global charcoal produc#on for 2020 is shown in Fig. 1.11., with an indica#on of the top 20 producing countries.[163] It should be highlighted that the presented data is sourced from the Food and Agriculture Organiza#on of the United Na#ons (FAO) database, which accounts only for wood-derived char, as it is a commodity on the global market. As can be derived from Fig. 1.11., Brazil is the largest biomass-derived char producer (ca. 6.5 Mt/year). As the source from the 2000s indicated, roughly 90% of produc#on units in Brazil (mainly brick beehive kilns) were not equipped with energy recovery. Assuming par#al improvement, the remaining char produc#on with rather primi#ve technologies s#ll adds significantly to the GHG emissions despite using biomass as a feedstock.[158] 0,0 0,5 1,0 1,5 2,0 2,5 3,0 3,5 4,0 earth kiln beehive kiln metal retort avg. batch avg. batch (EPA) avg. batch (IPCC) avg. batch avg. batch avg. batch beehive kiln, TR avg. batch, ER (EPA) avg. continuous, ER (EPA) Surplus emission in char production [kg CO2-eq./kg char] – wood – rice husk – cocoa shells – corn cobs char feedstock production with tar recovery (TR) and energy recovery (ER) CHAPTER 1 31 Considering the world produc#on divided by con#nent, charcoal produced in Africa (especially in the sub-Saharan Africa region) accounts for roughly 65 wt.% of the worldwide produc#on (ca. 35 Mt/year). That produc#on is also burdened with the direct emission of GHG from the char produc#on process. As the literature indicates, the most commonly used systems in Africa are earth kilns and metal kilns, which are rarely equipped with energy recovery systems. Nonetheless, sources also indicate the gradual improvement in produc#on technology in recent years.[164, 165] In rela#on to feedstock origin, aMer decades of problems with the unsustainable biomass economy combined with deforesta#on, Brazil took ini#a#ves to reduce the impact of charcoal produc#on on na#ve forests.[166] However, in African countries, forest degrada#on (leading to deforesta#on) is s#ll relevant and originates in the unsustainable harves#ng of wood for fuel and char produc#on.[167] Interes#ngly, the con#nents where sustainable char produc#on could be conducted on a large scale (due to ecological and legisla#ve reasons) represent only 3 wt.% of global char produc#on (North America accounts for 2 wt.%, and Europe with 1 wt.%). Interes#ngly, the majority share of the produc#on in North America is by the USA, as e.g., annual produc#on in Canada only accounts for roughly 0.02 Mt/yr.[163] Fig. 1.11. Comparison of char produc#on in the top 20 producing countries (largest producers) and global structure of char produc#on, both for 2020 (data source: FAOSTAT).[163] Produc#on is strictly linked to char consump#on for the major char producing countries, so char export usually cons#tutes only a small part of their na#onal produc#on. In accordance with the data provided by the FAOSTAT, the global import-export in 2020, was ca. 2.8 Mt of char, which corresponds to ca. 5 wt.% of the annual global char produc#on (ca. 53 Mt).[163] The annual char consump#on by the European Union in 2019 accounted for roughly 1 Mt, while internal produc#on represented only ca. 25 wt.% of that amount. Therefore, the remaining 0.75 Mt had to be imported to mi#gate the demand. The sources of import indicated that ca. 50 wt.% of the import was coming from African countries, 20 wt.% from South American countries other than Brazil, and ca. 20 wt.% from Ukraine. As charcoal produc#on in the men#oned areas may be burdened with unsustainable biomass sourcing, using imported charcoal seems counterproduc#ve to achieving the GHG emissions goals. Furthermore, FAO indicated that charcoal imported from African countries and used in the EU is one of the cri#cal factors that is leading to the accelera#on of deforesta#on on the African con#nent.[167] CHAPTER 1 32 The market of bio-based chars in the EU is rela#vely non-transparent and poorly controlled. It leads to ethically ques#onable prac#ces like illegal import (not traceable in the FAOSTAT’s data), which involves both sides of the trade, re-export of char from unaccountable importers, “dissolving” illegal char (e.g., mixing chars with tropical and na#onal origin), combined with unreliable labeling of char products. As the ma9er is not the primary subject of the work, details regarding issues appearing on the EU char market can be found elsewhere.[164, 165, 168-170] The general perspec#ve on the ma9er can be derived from a conclusion of the socio-economic research made in 2023, by Rocchi et al., which reflects the rela#onship between market and sustainability – “Consumers are guided in choosing woody products for energy purposes (…) primarily by price, less by sustainability. This is also true for woody charcoal, which is mainly bought in different commercial categories of the retail market. Therefore, there is no connec#on between consumers and producers, and the price is the most relevant aspect of the purchase.”.[168] Concluding, it needs to be strongly highlighted that sustainable solu#ons based on biomass-derived chars are only possible if such solu#ons are favorable to the market, where the sustainability factor rarely plays a primary role. 4.4. Applica,ons of bio-based chars Literature research of biomass-derived char applica#ons leads to an enormous number of lengthy reviews, so finding relevant informa#on is a complex task. As an example, it can be men#oned that in the Scopus database for the last 5 years (between 2019 and 2024), the number of reviews, which included the combined terms ”biochar” and “applica#on”, resulted in more than 1,600 publica#ons (and >11,000 research papers). Therefore, finding relevant informa#on required more oriented research, which resulted in fewer reviews on a more specific subject. The generic conclusion derived from the literature search is that published reviews rarely allow for a clear deriva#on of quan#ta#ve effects (in char applica#ons, like in soil amendment) due to the structure of the provided informa#on (mainly elabora#ve and inconsistent comparisons). Also, the reviews most commonly the studies that inves#gated a specific use of char and rarely on larger-scale applica#ons.[171-177] Such an approach is understandable, as the reviews summarize informa#on on novel applica#ons typically assessed in a lab-scale environment. Nonetheless, assessing the feasibility of an indicated applica#on (technical, economic, and environmental) and forecas#ng the poten#al of char use on a large scale is not a straighTorward task. Regardless of the men#oned flaws, further elabora#on in this Sec#on is a synthesis of the informa#on included in the literature on the applica#on of biomass-based chars produced in the slow pyrolysis process. That includes applica#ons with and without prior/further altera#on of char proper#es. As referencing all relevant publica#ons is beyond prac#cal possibility, the referenced sources were subjec#vely selected concerning the reputa#on of a journal, cita#on number, and quality of provided informa#on. The elabora#on on the char applica#on provided in this Sec#on is focused on chars derived from lignocellulosic biomass, as that poten#al is forecasted to have the highest relevance in terms of economic availability and sustainable sourcing. The feedstock base assessment was expanded in terms of char applica#ons in soil (biochar), as the non-lignocellulosic biomass poten#al finds prac#cal relevance for such applica#ons as well. From a broader perspec#ve, the applica#ons presented were selected because of their ability to appear in the shortand mid-term, preferably on a larger scale, as such uses have a forecastable market appearance. In the most simplis#c terms, the applica#ons of chars produced in slow pyrolysis are primarily linked to physicochemical proper#es of the char, which in themselves depend on the proper#es of the ini#al biomass and condi#ons in which it was pyrolyzed. The literature is abundant with quan#ta#ve assessments of the rela#onship between physicochemical proper#es of char and slow pyrolysis process parameters in rela#on to the applied biomass, e.g., the rela#onship between carbon content and temperature of pyrolysis. Therefore, the elabora#on on such rela#ons will be omi9ed, and its details can be found elsewhere.[112, 121, 178-186] CHAPTER 1 39 precursor with an aqueous solu#on of the agent in a suitable ra#o. Next, the obtained pulp is dried and heated to the final temperature in a rotary kiln. The obtained product is further washed to remove the remaining ac#va#on agent (for ZnCl 2 – washing with acid and water, and for H 3 PO 4 – base and water) and dried. The use of ZnCl 2 provides carbon with highly suitable proper#es, while its produc#on is environmentally challenging (contamina#on with zinc), and the recovery of the ac#va#ng agent cons#tutes a considerable produc#on cost. Therefore, the use of ZnCl 2 is currently losing industrial interest.[252] As can be found in most of the literature, KOH is indicated as one of the primary agents for chemical ac#va#on.[176, 251, 257] Interes#ngly, a book chapter by Henning and von Kienle from 2020 indicates that chemically ac#va#ng agents other than ZnCl 2 and H 3 PO 4 did not a9ain industrial importance despite their presence in the literature.[252] Opposing informa#on is provided in the work of Menéndez-Díaz and Mar|n-Gullón from 2006 (Maxsorb TM by Kansai Coke and Chemicals Co.).[258] Its market presence is also confirmed in other sources.[259, 260] Furthermore, Wachtler et al. (cocontributors to the same book as Henning and von Kienle) indicate a high poten#al of the materials derived from the ac#va#on with KOH, especially in energy storage applica#ons.[261] Aside from the produc#on process, the final proper#es of the ac#vated carbon strongly relate to the proper#es of the ini#al biomass from which char was derived. For example, the physical ac#va#on of coconut shell char leads to carbons with a high volume of fine pores, while the ac#vated carbons from wood-derived char are characterized by open pores. Considering the differences between processes, it can be stated that the physical ac#va#on leads primarily to the forma#on of pores in the microporous region (< 2 nm), while the chemical ac#va#on relates to the forma#on of pores in both microporous (< 2 nm) and mesoporous regions (2–50 nm).[252] Among the indirect uses of biomass-derived char, the leading applica#ons are those based on the ac#vated carbons derived from it. In general, those applica#ons include all processes where the performance is based on a material with high porosity in the specific range of pore sizes. What is worth no#cing is that the current literature very oMen indicates the applica#on of biomass-derived char, while the publica#on content rarely refers to the use of nonac#vated char. Within the indirect applica#on of biomass-derived chars can be dis#nguished: A. Use as a catalyst aNer func,onaliza,on – biomass-derived char represents a material whose surface func#onal groups can be rela#vely easily modified for a required final applica#on. The commonly applied func#onaliza#on of char represents its sulfona#on, which includes post-treatment of a char with concentrated H 2 SO 4 (98 wt.%), fuming H 2 SO 4 , and gaseous SO 3 . The treatment leads to the forma#on of the -SO 3 H func#onal groups within the structure, which enables a char to work as a solid acid catalyst. The study by Kastner et al. indicates that the use of gaseous SO 3 may lead to the forma#on of a more reac#ve and selec#ve char in comparison to the one derived with concentrated H 2 SO 4 .[262] Considered the applica#on of the sulfonated chars (as a solid acid catalyst), these include biomass hydrothermal hydrolysis, sugar dehydra#on to furanic deriva#ves (e.g., 5-HMF), and biodiesel produc#on through the transesterifica#on process. For the la9er, the material has the highest poten#al in applica#on, as sulfonated chars indicate a high efficiency (yield of 99 wt.% at 65 °C) and sa#sfactory stability over #me.[257, 263] However, compared to tradi#onal methods, the cost of its applica#on currently limits its widespread use in industrial biodiesel produc#on. The produc#on of biodiesel through transesterifica#on from 1 st gen oils is forecasted to decrease in the upcoming decades, while its produc#on from 2 nd gen. oils (used cooking oils, e.g., UCO) shows the opposite trend. Therefore, sulfonated chars reaching an industrial produc#on scale may succeed in the la9er scenario. Details about char func#onaliza#on and its possible and novel applica#on as a catalyst can be found in the following literature.[176, 257, 262, 264] B. Use as a catalyst carrier – the literature is abundant with examples on the use of biomass-derived char and ac#vated carbon as catalyst carriers, as well as novel approaches to add a specific cataly#c ac#vity to the material for various processes.[176, 241, 252, 257, 264-268] The generic way to add a CHAPTER 1 40 specific cataly#c ac#vity to a material (i.e. char) is its impregna#on with a metal oxide or metal salt, which can be followed by a reduc#on/oxida#on of the incorporated cataly#c center to its final form (e.g., metallic form). The applica#on of biomass-derived carriers includes uses in tar cracking from biomass gasifica#on, where impregna#on with Ni, Co, Pd, Pt, and Rh is suggested to achieve relevant efficiency. The inorganic content in biomass-derived chars may be their virtue for applica#on as Ca, K, P, Fe, and Mg enhance the cleavage of carbon bonds (in tar cracking). Studies with biomass-derived chars indicate that the incorpora#on of Ni, Fe, and Co into their structure leads to their high efficiency in tar cracking. However, biomass-derived char catalysts also induce a need for a higher temperature of the process to match the efficiency of conven#onally used catalysts.[176] Ac#vated carbon is a wellestablished catalyst carrier for the selec#ve cataly#c reduc#on (SCR) of the nitrogen oxides (NO x ) from flue gas. Due to the similarity in structure, the use of biomass-derived char as the catalyst carrier was also inves#gated. As the results indicate, the char-derived catalysts are currently able to convert NO x contained in flue gas with 80 wt.% efficiency. The results are promising, but they s#ll do not match the performance of commercial catalysts based on ac#vated carbon (99 wt.% efficiency).[176] Generally stated, the use of biomass-derived char (without its prior ac#va#on) as a catalyst carrier depends majorly on its pore size distribu#on in the range relevant to the considered process. That in itself naturally relates to the ini#al biomass and pyrolysis process parameters, but the impact of the metal incorpora#on also needs to be taken into account (e.g., calcina#on and reduc#on/oxida#on of the catalyst sites). C. Use in adsorp,on (with and without func,onaliza,on) – ac#vated carbons have a long history in adsorp#on applica#ons, including solid-liquid and gas-solid processes. The liquid-solid processes include organic compound removal, like decoloriza#on in the sugar industry, and pollutant removal (heavy metals and organic compounds) in fresh water and wastewater treatment. Consecu#vely, the gas-solid processes include solvent recovery, gas separa#on processes like pressure swing adsorp#on (PSA), and pollutant removal from process gases. The adsorp#on mechanism in the liquid-solid process for ac#vated carbon is the same as for biomass-derived chars, so relevant details can be found in the few paragraphs above. It is worth repea#ng that for liquid-solid processes, the performance of a material is primarily linked to the pore size distribu#on and the size of the molecule that needs to be absorbed. Therefore, for the absorp#on applica#on, the porosity is relevant in the microporous (< 2 nm) and mesoporous range (2–50 nm).[252, 269, 270] For industrial gas-solid processes, the mechanism is mainly based on physical adsorp#on (reversible process – sorp#on and desorp#on), where the adsorbent is used to separate specific gases from their mixture or stream. Therefore, the cri#cal factor for a material is not to chemically bind the gases or any other compounds in the mixture rela#vely permanently (chemisorp#on). A surface func#onaliza#on tailored for a specific case can be applied to enhance the gas adsorp#on proper#es of the carbonaceous material (aside from the improvement of its pore size distribu#on characteris#cs). That includes acid treatment (H 2 SO 4 or H 3 PO 4 ) to enhance basic compound separa#on (NH 3 and vola#le amines) or a basic treatment (K 2 CO 3 or Na 2 CO 3 ) to enhance acidic compounds separa#on (HCl, SO 2 , H 2 S, NO 2 ).[252, 258] The gas separa#on process includes all kinds of non-polar molecules (vola#le solvents, light hydrocarbons). Furthermore, it includes the separa#on of mixtures of gases that do not condense at standard condi#ons, like CH 4 , H 2 , O 2 , N 2, and CO 2 , whose separa#on requires elevated pressure and lower temperatures for efficient separa#on (e.g., industrially with PSA). Considering the current decarboniza#on aims, biomass-derived ac#vated carbons find applica#on in carbon separa#on and sequestra#on (CCS) technologies, where the separa#on of CO 2 from process gas is the primary goal. As pointed out by Chen et al., an efficient CO 2 adsorbent, aside from the high porosity in the microporous region (< 2 nm), should include basic sites, delocalized π electrons in aroma#c rings, and unsaturated valencies to enhance the CO 2 sorp#on process.[271] A par#cular case of gas separa#on is the purifica#on of biogas, which cons#tutes a mixture of CH 4 (50– 70 vol.%) and CO 2 (30–50 vol.%), with a rela#vely high concentra#on of H 2 S (up to 10,000 ppm vol., so 1 vol.%).[272] In biogas separa#on, removing H 2 S from the gas mixture is crucial as it is highly toxic and CHAPTER 1 41 leads to corrosion upon combus#on. Furthermore, if biogas is considered for frac#ona#on toward biomethane and further compression into bio-CNG (Compressed Natural Gas) and bio-LNG (Liquefied Natural Gas), the removal of H 2 S needs to be conducted before the separa#on of CH 4 and CO 2 from the mixture. The adsorp#on of H 2 S is possible with carbonaceous adsorbents, while its adsorp#on process is highly irreversible (majorly driven by chemical bonding). As the frequent replacement of ac#vated carbon for CH 4 /CO 2 mixture separa#on represents a high opera#onal cost, current inves#ga#ons are driven to produce cheap, char-based adsorbents that would be able to adsorb H 2 S as a biogas pretreatment process. The use of pre-separa#on of H 2 S with func#onalized biomass-derived char could lead to a reduc#on in the cost of further biogas separa#on, while such a solu#on s#ll cons#tutes an opera#onal cost. Further details about gas-solid separa#on systems with biomass-derived adsorbents can be found in the following publica#ons.[271-275] E. Biomass-derived carbons in energy storage solu,ons – the applica#on of biomass-derived carbon materials (chars, but mainly ac#vated carbons) in energy storage solu#ons relates to their use in electrode produc#on.[276] Consecu#vely, biomass-derived carbons find applica#on in two major energy storage solu#ons: ba9eries and capacitors (specifically supercapacitors). The main difference between both energy storage systems is the mechanism of energy storage, which is the origin of further differences in characteris#cs and material requirements needed to achieve system-related performance. In the case of ba9eries, the energy is stored through chemical (redox) reac#ons occurring on and within the electrodes of a ba9ery, while closing a circuit induces reac#ons and corresponding electron flow (slow and steady charge and discharge). Supercapacitors, also called Electrochemical Double-Layer Capacitors (EDLCs), store energy as an electric charge at the electrode surface, so as a poten#al between two electrodes (separated by a liquid electrolyte). As the circuit closes, the poten#al stored between both electrodes is rapidly and unsteadily discharged (charging occurs in the opposite way).[276] Electrode material for each applica#on must fulfill rigorous criteria before it is accepted as usable for the solu#on, which relates to high safety standards, especially related to the use in ba9eries. Therefore, only well-proven materials, whose produc#on is characterized by reliability and consistency of achieved proper#es, can be considered for a large-scale applica#on. For the men#oned reason, only carbons derived from homogeneous and rela#vely clean biomasses (low ash and low heteroatom content) have been considered so far for use in energy storage solu#ons. The following elabora#on is a synthesis of informa#on derived from the following sources.[261, 276-287] In lithium-ion ba9eries (Li-ion), the carbon material (typically graphite) is used as the nega#ve electrode (anode), which stores lithium ions within its layered aroma#c sheet structure through a process known as intercala#on. Such a mechanism is possible as the spaces between layers are larger than the effec#ve radius of a lithium ion (Li + ). Ac#vated carbons (aMer treatment at temperatures of 1000–1400 °C) can be used as electrodes for Li-ion ba9eries, as their carbon structure has graphi#c centers that can intercalate Li-ions, thus serving the energy storage purpose.[261] However, their performance is usually insufficient, especially when compared against pure graphite, and suffers from large irreversible capaci#es (capacity losses during subsequent charge/discharge cycles) caused mainly by reac#ons occurring on the surface of the anode. The origin of these unwanted phenomena is related to the high porosity and large specific surface area of ac#vated carbons. Biomass-derived ac#vated carbons are thus not considered for Li-ion ba9ery anodes, but do appear to be a be9er fit for applica#on in “next-genera#on” ba9eries, namely the sodium-ion ba9eries (Na-ion). The lithium ion (Li + ) has a smaller radius than the sodium ion (Na + ), so the carbon anode material requires larger spaces between its graphite sheets to intercalate Na-ions. Biomass-derived carbons, due to their ini#al composi#on and upon produc#on, may form a specific organiza#on of the graphite layers with a larger distance in between (compared to graphite) that can properly accommodate sodium ions (Na + ). The use of biomass-derived carbons as anodes for Na-ion ba9eries has promising results, while the issues related to the high surface area of the ac#vated carbons remain the same. Another op#on for using biomass-derived carbons in ba9ery produc#on is their applica#on as a posi#ve electrode (cathode) in lithium-sulfur (Li-S) ba9eries. Sulfur is an insulator, so to obtain an CHAPTER 1 42 electronically conduc#ve electrode, it has to be distributed and well incorporated over a large surface of well-conduc#ng material. When sulfur is not well incorporated, it may lead to excessive dissolu#on of its reac#on intermediates, large volume expansion of electrodes during charging cycles, and in the end, poor cycle stability. Porous carbon materials with a high SSA, high chemical stability, and high conduc#vity are suitable materials for Li-S cathodes as their proper#es allow for appropriate distribu#on and incorpora#on of sulfur over the surface of the electrode. Therefore, biomass-derived ac#vated carbons are considered a well-suited material for this purpose, while their large-scale use has not emerged so far (same as for the Li-S ba9eries).[261, 282] From a general point of view, and aside from the forecasted applica#on in electrodes for Li-S ba9eries, biomass-derived carbons show rela#vely unfavorable proper#es compared to other carbonaceous materials in applica#ons as ba9ery electrodes. Consequently, this translates into a lower interest in using those materials in ba9ery produc#on. The energy stored by a supercapacitor depends on its capacitance and pseudo-capacitance. The former relates to the charge stored in the electrochemical double layer at the electrode, where the energy is stored in an electrosta#c field. This mechanism is responsible for the very rapid charge/discharge of a supercapacitor. The capacitance is directly related to the surface occupied by the electrolyte, which, in detail, represents the surface effec#vely accessible by the molecules used as electrolyte. The rela#on is oMen oversimplified to the statement that the capacitance is propor#onal to the specific surface area (SSA) of the electrode material and calculated from gas adsorp#on measurements using the BrunauerEmme9-Teller (BET) calcula#on method (usually referred to as the SSA BET ). The SSA values derived in such a way reflect the surface area of pores that may not all be accessible by electrolytes. Therefore, the BET-derived SSA value may refer to a value that does not effec#vely add to the capacitance of the material. Overall, it is advised to consider only the SSA of pores whose size is 2–3 #mes larger than the effec#ve diameter of the ions used as electrolytes (the effec#ve diameter of commonly used electrolytes ranges from 1.2 nm to 2.0 nm).[261, 277] The energy stored by the pseudo–capacitance mechanism refers to the energy stored by faradaic reac#ons (chemical reac#ons), where the mechanism is similar to the chemical reac#ons in ba9eries. Pseudo-capacitance depends on the chemical proper#es of the used electrolyte and its affinity to the surface of the used carbon, mainly specific surface groups. The release of charge stored as pseudo–capacitance is slower than charge stored by the capacitance. As storage through pseudo-capacitance is unavoidable in real systems, the voltammetry characteris#cs are distorted from the theore#cal ideal capacitance characteris#cs (perfect rectangle). The pseudo-capacitance may be perceived as a posi#ve feature, as it increases the overall energy storage by a supercapacitor. However, promo#ng pseudo-capacitance (e.g., by nitrogen doping) has to be approached with cau#on as it may finally lead to unsuitable charge/discharge characteris#cs of a supercapacitor. Contrary to the use in ba9ery systems, the carbons derived from biomass show advantageous proper#es to other carbon materials when used as electrodes in supercapacitors. The unique feature of the biomass-derived carbons is related to their hierarchical pore structure, where the micropores relate to the capacitance, while mesopores work as the interconnec#ng pores, which improve the diffusive transport of the electrolyte within the porous structure. Aside from advantageous porosity characteris#cs, biomass-based ac#vated carbons present suitable electrical conduc#vity and chemical and thermal resistance.[261] Furthermore, biomass-based material presents an environmental advantage over synthe#c and fossil-based materials, which may become a strong s#mulus leading to their industrial applica#on. The direct and indirect applica#ons of biomass-derived chars listed above represent the majority of their current applica#ons, as well as those that have a foreseeable #me for market appearance. That does not necessarily mean that the list of applica#ons will not extend in the future. The area of biomass-based chars has been developing intensively over the last decades, and the forecast is that it will con#nue, with the decarboniza#on aims as an accelera#ng s#mulus. The applica#ons of biomassbased chars and their deriva#ves require specific proper#es from the material, suited for a specific process. However, it can be no#ced that novel and advanced applica#ons like biochar, pollutant removal, catalysis, and energy storage are firmly anchored in the proper#es of a char related to its CHAPTER 1 43 porous structure and proper#es derived from its pore size distribu#on. Therefore, further progress in the men#oned applica#on is also related to progress with the iden#fica#on of factors that influence the pore size distribu#on in chars. Therefore, a sufficient understanding of pore forma#on to enable structure tailoring toward a specific use as a final goal can be indicated. 4.5. Relevance of porosity for bio-based chars In accordance with the defini#on provided by the Interna#onal Union of Pure and Applied Chemistry (IUPAC), the porosity of a material is defined as “the ra#o of the total pore volume to the volume of the par#cle or agglomerate”. In view of the defini#on, the pores of a par#cle relate to voids within a material structure enclosed within the described geometrical dimensions. Furthermore, the total surface area related to a porous par#cle can be dis#nguished between its external and internal surface. The former is defined as the “surface outside the pores”, while the la9er is defined as the “surface of all pore walls”, and it is commonly referred to as specific surface area (SSA). As the porosity in the material may be composed of pores with different sizes, the descrip#on of pore size is commonly made with a pore size distribu#on (PSD). The IUPAC differen#ates the following base ranges of pores: micropores (< 2 nm), mesopores (2–50 nm), and macropores (> 50 nm). The more detailed division provided by the IUPAC differen#ates the micropores into narrow micropores, also known as the ultramicropores (< 0.7 nm), and wide micropores, known as the super-micropores (0.7–2 nm). The IUPAC classifica#on of pores was developed to describe the proper#es of materials like adsorbents and catalysts, whose applica#ons relate mainly to pores in the range of 0.3–50 nm.[288] A significant difference in the pore classifica#on from the one provided by IUPAC can be found in the classifica#on from the Soil Science Society of America (SSSA). The men#oned classifica#on differen#ates the following pore categories: cryptopores (< 0.1 μm), ultra-micropores (0.1–5 µm), wide-micropores (5–30 μm), mesopores (30–75 μm), and macropores (> 75 μm). The commonly applied limit for the intrinsic pores is < 500 μm (< 0.5 mm), as larger pores should be accounted to the geometrical structure and not to the internal porosity. The SSSA classifica#on was derived to describe the proper#es of soils and minerals and differen#ate pore size ranges relevant to the proper#es of the men#oned materials.[289] As can be seen, both classifica#ons, IUPAC and SSSA, are vaguely compa#ble with each other, which may lead to confusion in data presenta#on.[290] The vague division may create significant confusion, especially with the descrip#on of proper#es of biomass-derived chars, as their porosity starts from IUPAC’s ultra-micropores (< 0.7 nm) and finishes with the SSSA’s macropores (< 500 μm). Another difficulty relates primarily to the descrip#on of char proper#es for soil applica#on, where both classifica#ons are relevant. A comparison of pore size ranges is presented in Fig. 1.13. to illustrate issues related to the relevance of specific pore sizes in a specific applica#on of biomassderived chars and corresponding phenomena. Moreover, in Fig. 1.13., pore size ranges of analy#cal characteriza#on techniques as well as both classifica#ons (IUPAC and SSSA) are presented to visually represent relevant ranges. To address the pore-related requirements for chars in soil applica#on, it is relevant to indicate the pore size range accessible to plants, fungi, and bacteria, as those relate to the bio#c effects that the char will exhibit in the soil. The uptake of nutrients and water in plants takes place by organs called hair roots, which further connect to the fine roots of a plant (and further). The diameter of hair roots usually varies between 5 and 20 μm.[291, 292] The organisms oMen connected to soil health are fungi, which form a symbio#c associa#on between a fungus and a plant (known as mycorrhiza). The basic cellular unit of filamentous structures of a fungus is called a hyphae, which has a similar role as hair roots in plants. The typical diameter range of fungal hyphae is between 0.5 µm and 20 µm.[293] The organisms present in the soil that can have both posi#ve and nega#ve effects on plants are bacteria, of which various strains are typically present in the soil in a size range from 0.2 μm to 10 μm. A bacterial strain that is nowadays oMen recommended as a pro-bio#c supplement for agricultural soils is the phosphorous solubilizing Bacillus megaterium, whose dimensions are ca. 4 μm in length and ca. 1.5 μm in diameter.[294] Considering the size of these organisms in the soil, it can be derived that pore sizes CHAPTER 1 44 between 0.5 μm and 50 μm can be indicated as the range that could serve as a habitat for these organisms. Studies indicate that pores with sizes between 5 μm and 75 μm are the ones related to water reten#on in soil. However, the water retained in the pores between 30 μm and 75 μm is usually considered as water available to the plant (as it is not strongly bound to the soil). The water retained in smaller pores, despite its presence in soil, usually does not account for plant available water as it is strongly bonded to soil aggregates, so it is basically unavailable for living organisms. The larger pores (above 75 μm) in biomass-derived chars are primarily responsible for draining the soil from water (opposite effect to water reten#on).[225, 295] Therefore, using char to increase water reten#on in soil (or its draining if needed) has to be carefully tailored as the char applica#on may bring an effect opposite to the intended one. The ma9er is complicated as for biomass, especially wood, data on the ini#al structure is available, while porosity data for chars is scant in literature.[133, 134, 296, 297] The la9er causes difficul#es in predic#ng char behavior in soils. Nutrient reten#on and immobiliza#on of heavy metals in the soil mainly relates to porosity in the pore size range between 0.5 nm and 50 nm, so in pores 1000x smaller than those relevant for water reten#on and microorganism habitats. Details about the rela#on between pore size, nutrient reten#on, and metal immobiliza#on in soil were provided in Sec#on 4.4, so further elabora#on in this paragraph will be omi9ed. Industrial applica#ons (wastewater treatment, gas separa#on, applica#on as catalyst support, and as electrode material) of chars mainly relate to pores with sizes between 0.5 nm and 50 nm. However, a specific applica#on requires a tailored porous material, aside from other physicochemical proper#es necessary for the applica#on. As the details about the industrial applica#ons were provided in Sec#on 4.4, further details in this paragraph will be omi9ed to avoid unnecessary redundancy. The measurement techniques presented in Fig. 1.13. relate to the most commonly used methods for assessing pore size distribu#on and pore structure of biomass-derived chars. Basic informa#on about these methods can be found in the publica#on made by Brewer et al. and further details in the dedicated literature, among others, in the book by Lowell et al. [298, 299] In Fig. 1.13., the pore size ranges based on gas sorp#on isotherm measurement by the 3 commonly used gases (N 2 , CO 2 , and Ar) were indicated. Regardless of the measurement and data processing protocols, the use of a specific gas relates to specific condi#ons of the measurement, which may further translate to the obtained result. The most commonly used nitrogen adsorp#on is conducted at 77 K, which induces very low mobility of the gas molecules and slow gas diffusion. The diffusion problem leads to a poor assessment of pores smaller than 0.7 nm with nitrogen. To avoid the diffusion issue, common prac#ce suggests star#ng from a larger pressure (so induce a lower vacuum, e.g., 10 -4 –10 -3 p/p 0 ).[300] Measurement with CO 2 is conducted at 273 K (ca. 200 K higher than for N 2 ), which overcomes the problems with the gas mobility issue. Therefore, the method is preferred for detailed measurement of the smallest pores (up to the size of the effec#ve radius of CO 2 ). The drawback of this method is that the commonly applied upper pressure (up to 1 bar) allows for assessing pores only with sizes up to ca. 1.5 nm. It is uncommon to use CO 2 under pressure for gas adsorp#on measurement, but by doing so, the measurement range could be extended to ca. 10 nm. The use of Ar faces a problem similar to that of N 2 measurement. Nonetheless, the use of 87K for Ar measurement has a non-negligible effect on gas mobility, which allows for measuring pores below 0.7 nm with fewer prac#cal problems than with N 2 .[299, 301] Although the methods for pore size measurement are well established, a significant issue with pore size measurement in biomass-derived chars can be no#ced in the literature. Such a presump#on is supported by the results contained in the work by Bachmann et al. The study included a round-robin measurement of SSA between 5 different laboratories, which showed differences up to 200 m 2 /g for the same sample of char (20 % devia#on from the average). As the systema#za#on of the SSA measurement was not pre-agreed upon between par#cipa#ng laboratories, the lack of a standardized protocol seems to be the key issue iden#fied in the study.[302] A few other outcomes commonly appearing in the literature regarding gas adsorp#on measurement of biomass-derived chars lead to a few more significant concerns. The first one is a high devia#on between the measurements.[298] CHAPTER 1 45 Nonetheless, this systemic problem is difficult to trace in a quan#ta#ve manner, as gas adsorp#on is perceived as a very precise but lengthy/expensive measurement, so its results are rarely replicated. The second issue relates to the devia#on in the pore volume of biomass-derived chars obtained through adsorp#on with N 2 as compared to CO 2 . It is especially worrying as every gas adsorp#on should lead to a similar outcome regardless of the applied probe gas, as was proven by Ravikovitch et al.[301] Therefore, results for biomass-derived chars should be comparable, especially for the same pore size range but the literature indicates otherwise.[303-306] Fig.1.13. Pore size ranges in rela#on to classifica#on (IUPAC and SSSA), relevancy to the applica#on in soil and industrial applica#ons, and characteriza#on techniques with dis#nc#ons on the methods for pores assessment and structure assessment. CHAPTER 1 46 Third, the most common visible feature of biomass-based chars is the dras#c rise in the SSA value derived from N 2 adsorp#on for chars produced at 500–700 °C and a similarly dras#c drop in SSA for chars produced at even higher temperatures. Furthermore, this apparent SSA op#mum with respect to pyrolysis temperature is inconsistent even for the same biomass, e.g., wood.[121, 139, 305] Interes#ngly, SSAs of the chars derived specifically for each bio-component do not indicate the appearance of a drop in value with pyrolysis temperature.[307] The literature oMen provides a hypothesis to support the appearance/disappearance of such a peak in SSA, which indicates the (micro- )pore structure collapse in pyrolysis carried out above 700 °C as the main reason.[308-311] Two issues with the explana#on can be found, where the first is related to the inconsistent appearance of the SSA peak for the same biomass type. The other issue relates to the lack of such a peak in SSA within the results derived from CO 2 adsorp#on, which is a more precise assessment method for the micropore structure assessment.[306] Overall, so far, the aforemen#oned issues have not been resolved. Therefore, the rela#on between pyrolysis temperature and the SSA of biomass-derived chars oMen looks similar in the literature to the one provided in Fig. 1.14.[182] Interes#ngly, in the literature, scien#fic studies have been found that use such an inconclusive database for the construc#on of a predic#ve model.[312] As pointed out in the previous and in this Sec#on, the majority of the novel applica#ons of char require a reliable assessment of porosity in the microporous region (IUPAC classifica#on). Therefore, the inability to reliably es#mate the impact of basic produc#on parameters like pyrolysis temperature on a key applica#on-related parameter is a cri#cal issue, hindering further progress. Fig. 1.14. The rela#on between the specific surface area (SSA) of biomass-derived chars (various feedstock groups) and the pyrolysis temperature. Source: Ippolito et al.[182] CHAPTER 1 47 5. Modelling and biomass-derived chars 5.1. Needs for and profits from modeling The issue highlighted in the last paragraph of Sec#on 4.5 is one example that needs to be addressed to accelerate the development of novel applica#ons related to biomass-derived chars. The generic issue in the field is oMen highlighted as the “knowledge gap”, which can be translated as the lack of certainty regarding char proper#es based only on known ini#al parameters (related to feedstock and pyrolysis process condi#ons). This knowledge gap is further amplified by the variability in proper#es between different biomasses and even within the same biomass group. For example, lignocellulosic biomass, whose overall chemical composi#on (in terms of bio-components) shows similari#es, differs significantly in terms of biomass structure (e.g., wood, straw, coconut shells) and mineral ma9er content and composi#on. However, even for the same biomass feedstock like wood, the results can also differ due to fluctua#ons in composi#on and structure between species and even within the same species (e.g., growth condi#ons can have an impact on mineral ma9er composi#on). When variance related to processing parameters in pyrolysis and in analy#cal methods for characterizing chars is added to the equa#on, the ma9er becomes complex, and the “knowledge gap” widens even further. Finding solu#ons and answers in the engineering field can be done through three fundamental pathways: theore#cal methods, experimental inves#ga#ons, and numerical simula#ons.[313] The theore#cal approach allows for understanding the phenomena occurring within the inves#gated system. However, the theore#cal outcomes usually do not provide prac#cally applicable answers. Furthermore, the real system is complex, so a direct analy#cal solu#on is beyond possible, and simplifying the problem statement quickly leads to highly unreliable outcomes. The more reliable and prac#cally applicable solu#on is experimental inves#ga#on, usually focused on a specific case. Aside from being most reliable, experimental inves#ga#ons have their limita#ons. It should be highlighted that biomass processing to char is a dynamic process, which takes place in rela#vely small en##es (par#cles whose size starts from m), which impose limita#ons on what is prac#cally possible to be measured. As the most commonly appearing issues related to finding solu#ons experimentally can be generalized: (1) measurements in real objects (e.g. biomass par#cles) are oMen difficult or even impossible due to the small dimensions, (2) measurements do not represent the condi#ons in a real system, e.g., downscaling the reactor technology into laboratory scale changes the processing condi#ons and (3) the experimental data matrix may become too large to be prac#cally assessed due to required #me and costs, e.g., feedstock parameters × processing condi#ons × assessment of parameters (feedstock, products, process data) × replica#ons.[313] The third way, numerical simula#ons toward finding prac#cal solu#ons, is a rela#vely novel way, as its relevance in engineering prac#ce dates back to the 1970s.[314] Solving problems numerically presents several advantages, while it also has its limita#ons. The benefits can be grouped as: (1) a low unit cost per test/inves#ga#on, (2) enabling conduc#ng proof-of-concept (PoC) at the very beginning of the project (low sunk cost in case of failure), and (3) enabling assessing parameters (prac#cally) impossible to assess through experiments.[315] The men#oned advantages can have a crucial impact on the development of new technological solu#ons by reducing the economic costs of their development and lowering the risks from the beginning of the development trajectory. The generic impact of numerical simula#ons in a given project is presented on Fig. 1.14. If the simula#on is sufficiently detailed and mimics the real system well, there is a possibility of inves#ga#ng and valida#ng new correla#ons and theories through large and detailed databases of process history. That, in consequence, would enable the iden#fica#on of cri#cal points within the system (bo9lenecks) and propose solu#ons for their mi#ga#on. The most significant concern with numerical solu#ons is the “if”, so the certainty that the constructed numerical representa#on (model) of a real system is sufficiently reliable and mimics the real system. In more generic terms, the ques#on is “if a simula#on outcome is true”, in the meaning of whether the obtained results will confirm to a real system and to what extent.[314] CHAPTER 1 48 Fig. 1.14. Comparison of a project cost distribu#on over project #me with and without the use of numerical simula#ons.[315] To elaborate on the limita#ons of numerical simula#ons, one needs to start with a systema#c overview of the terms related to the field. A model is the mathema#cally described (by algorithms and equa#ons) representa#on of a real system. The model contains a descrip#on of a specific system and scenario of applica#on, including, among others, data about geometry and proper#es, as well as the ini#al and boundary condi#ons (of a system and subjects within). A simula#on is a result of performing a test on the model, so the response of the model to the implemented condi#ons. The term numerical means that the mathema#cal descrip#ons on which the model is based will be translated through informa#cs into a numerical language, known by a numerical tool (a computer) to perform calcula#ons (computa#ons).[315, 316] Models are always only a representa#on of a real system, so they are based on simplifica#ons and approxima#ons. Those can include subjects that are or are not irrelevant (needed or not needed to be included in a model), while the impact and assignment of a subject to the appropriate category can only be made through a simula#on and its valida#on. The modeling field has a very strictly defined terminology, to precisely dis#nguish “real” from “simulated”, as otherwise it would lead to significant confusion and, further, to far-reaching consequences. Details about the nomenclature can be found in the book by Patrick J. Roache, who is a pioneer of the applica#on of numerical simula#on in engineering.[314] Two major terms, verifica#on and valida#on (commonly known as V&V), were a point of seman#c debate through the years in the modeling field, which led to their very precise defini#on.[317] In simple terms, verifica#on refers to the assessment of rela#ons implemented into a model (“solving the equa#ons in the right way”). It includes checking the model code to find if everything is well implemented. Consecu#vely, valida#on refers to the assessment of the simula#ons provided by a model and its agreement with a real system (“solving the right equa#ons”). It includes assessing the accuracy of predic#ons derived from a model, which is mainly made by comparing simulated results with ones experimentally obtained. Valida#on also relates to assessing model components (e.g., submodels) in view of their impact on the simulated results compared to the real system. In further elabora#on, the term valida#on will oMen be used as it relates to assessing the reliability and performance of a model. A model must be validated to be assessed as reliable to represent the real system well. The statement seems obvious, but it has far-reaching consequences in terms of assessing which model elements are or are not reliable. Such an assessment is not straighTorward, as devia#ons between simulated and experimental results can originate from the model as well as the measurements. A theore#cal example can be a case when the model provides an accurate result, but it is compared with a wrongly measured parameter. Then, the model can be wrongly assessed as unreliable despite the origin of the fault being elsewhere. CHAPTER 1 55 However, the Eulerian-Eulerian approach for modeling the conversion of par#cles above mm size may introduce a significant devia#on from reality. In such cases, the result of the simula#on is burdened with considerable inaccuracy. The applicability of this simplifica#on is usually based on assessing two non-dimensional numbers: the Biot number (Bi) and the Pyrolysis number (Py). The la9er is also called the reversed thermal Thiele modulus.[111, 336-338] Those numbers describe a specific thermal conversion regime within a given biomass par#cle based on par#cle size and proper#es. The thermal regime reflects the most relevant thermal phenomena that drive the conversion in a par#cle, which can be differen#ated by chemical reac#ons, intra-par#cle and extra-par#cle heat exchange. A par#cle can be assigned to one of the four following thermal regimes: pure kine#c, thermally thin, thermal wave, and thermally thick, with respect to the non-dimensional numbers.[111, 336-338] The simplifica#on through the Eulerian-Eulerian approach is the most valid for par#cles in the pure kine#c regime, whose size should not exceed 1 mm in any direc#on.[111] The conversion of par#cles in the thermally thin regime is also driven mainly by reac#on kine#cs, but external heat transfer also begins to play a relevant role. Due to their rela#vely small size, par#cles do not show high thermal or internal pressure gradients during conversion. Applying the Eulerian-Eulerian simplifica#on for par#cles in the thermally thin regime is not advised, but it does not represent a cri#cal error in the model. For this regime, the dilu#on of the solid phase also has to be considered in assessing the applicability of the simplifica#on. Regardless of par#cle size, if the solid phase is highly concentrated, the Eulerian-Eulerian approach also does not represent the real system well, so using other descrip#ons is advised (CFDDEM, CFD-DPM, or CFD-DDPM). The internal and external heat transfer is the primary driver of the conversion of par#cles assigned to the thermal wave regime. Addi#onally, a significant temperature and pressure gradient within a par#cle starts to be formed during its conversion. The main assump#on of the thermal wave regime is that the conversion of the par#cle takes place in a thin front (i.e., a layer where the conversion front thickness strives to 0). Such an approach opens the possibility of a par#al simplifica#on of describing conversion in a single par#cle within a reactor scale model without introducing a cri#cal error. In prac#ce, the conversion of a single par#cle is being made by implemen#ng a simplified conversion model, like the unreacted shrinking core model or the layer model.[111, 113, 339] Overall, the descrip#on of par#cle conversion in the thermal wave regime requires implemen#ng the EulerianLagrangian approach to dis#nguish each par#cle separately. The last dis#nguished thermal regime in par#cle conversion is the thermally thick regime. This relates to cases where internal heat transfer has the largest share in the control of the conversion of a par#cle. It refers mainly to the conversion of rela#vely large par#cles, which are characterized by the appearance of significant temperature and pressure gradients during conversion. No s#ff threshold value exists for a par#cle to be assigned to the thermally thick regime. Literature suggests that the thermally thick regime should be considered when a par#cle is described by a Bi number higher than 40 or 100 and a thermal Thiele modulus (1/Py) higher than 100 or 1400.[111] The conversion of a par#cle in the thermally thick regime is the most complex and computa#onally demanding case for a reactor model. When inves#gated, simplifica#on related to the reactor system may be applicable (e.g., DDPM or DEM), while no simplifica#on regarding single par#cle conversion is advisable. Therefore, a descrip#on of a single-par#cle model in the reactor model needs to be complete for the thermally thick regime. The field of reactor modeling of biomass pyrolysis, especially using the Lagrangian approach to describe solid behavior, is rela#vely young, as it requires sufficiently high computa#onal power. Nonetheless, excellent examples of the use of CFD-DEM (with various extents of simplifica#on) can be found in the literature, which reflects the progress in the field.[340-349] Furthermore, reviews enclosed in recent literature provide systema#zed knowledge about the mul#-scale approach in pyrolysis modeling and the implementa#on of solid and fluid descrip#on in reactor scale models.[350-356] The analysis of the literature on reactor scale modeling indicates that major interest is related to modeling biomass fast pyrolysis in fluidized bed systems. In such reactor models, the biomass par#cle size does not exceed 3 mm, and the major interest is devoted to bio-oil yield. Reactor models, which apply to systems that aim to produce biomass-derived chars, on the other hand, are in the minority (fixed-bed reactors or CHAPTER 1 56 rotary reactors). Regardless, those studies bring relevant progress in understanding char forma#on and reactor impact on the process. An excellent example of progress in this area is the work of Lu et al., who used a glued-sphere CFD-DEM with 3D intra-par#cle models to represent the conversion of nonspherical biomass par#cles.[357] Literature is also rich with conclusions, reading the challenges that need to be addressed to accelerate further progress. Among those, a common one is the sugges#on to improve single par#cle models, as those relate to cri#cal process characteris#cs in the reactor model. In detail, the conclusion usually relates to a need to improve models describing biomass degrada#on (kine#c models), which is a key element of reactor models aiming for high bio-oil produc#on. Interes#ngly, in the literature related to the produc#on of biomass-derived char, a common conclusion is that further improvement should be made in modeling changes related to the (micro-)porosity. However, such a conclusion is rarely observed in reactor modeling studies. Such a situa#on is likely caused by the available literature, which is biased towards reactor modeling for bio-oil produc#on. 5.3. Relevance of the single par,cle scale As highlighted in Sec#ons 4.5 and 5.2, two major issues can be iden#fied aMer inves#ga#ng the available literature. One is the lack of predic#ng porosity changes within a par#cle during the pyrolysis process. As the single-par#cle model uses the smallest scale that accounts for par#cle structure, it is advisable to begin from that scale. The second commonly highlighted issue is the need for higher precision of single par#cle models that are to be embedded in the reactor scale, which is mainly addressed by improving imperfec#ons of kine#c models. However, that might be an oversimplifica#on, and the appropriate scale to find a solu#on to the issue may also be the single par#cle scale. Kine#c models, both types, simple and advanced models like the Ranzi model, are primarily based on mass-loss data obtained in TGA.[111] The methodology of measuring in a TGA is based on using the material in a powder-like form in a rela#vely low amount, which translates to biomass conversion in the pure kine#c regime. As pointed out in Sec#on 4.1, biomass conversion in such a state favors the conversion pathway toward bio-oil forma#on while it hinders the char forma#on pathway. However, conversion of biomass par#cles on the reactor scale does not necessarily have to occur in the same thermal regime (i.e., larger par#cles in slow pyrolysis systems). From a generic standpoint, biomass conversion in fast pyrolysis systems will have more similari#es to TGA measurement than conversion of rela#vely large par#cles in slow pyrolysis systems. Therefore, for modeling pyrolysis processes aiming at bio-oil produc#on, the issues with kine#c models will relate to slightly different problems than those in char produc#on. In the case of bio-oil produc#on, two major issues are highlighted as key sources of inaccuracy in predic#ng bio-oil yield and composi#on. The first are the imperfec#ons of available kine#c reac#on schemes.[347] However, those have their origin in the balance between accuracy and complexity (e.g., the number of compounds and reac#ons included), where the la9er significantly influences the computa#onal burden. The second most commonly highlighted issue is the impact of minerals naturally present in the biomass (e.g., alkali and alkaline earth metals – AAEMs).[111] As the mineral ma9er induces a cataly#c effect during the degrada#on of bio-components, omiRng their presence in pyrolysis (kine#c) models has a severe consequence on predic#on accuracy, which is especially true for modeling agricultural residual biomass, which is rich in AAEMs. Overall, improving the descrip#on of a single par#cle model aiming at bio-oil produc#on does not seem cri#cally relevant, as improvement should be focused on the kine#c models. The men#oned issues should be solved regardless of the aimed product (char or bio-oil), as they would improve the accuracy of any model of biomass pyrolysis at any scale. The issue related to the predic#on inaccuracy becomes more complicated when a kine#c model (developed for the primary kine#c regime) is applied to model par#cle conversion in another regime, especially in the thermally thick regime. In the thermally thick regime, the conversion pathway toward char forma#on is favored over the conversion toward bio-oil. This issue was men#oned in Sec#on 4.1. CHAPTER 1 57 and details on the issue are highlighted in the review by Anca-Couce.[111] Among the available kine#c models, one that has the ability to implement the impact of elevated char forma#on in thermally thick regimes is the RAC scheme. This scheme introduces an adjustable parameter “” defined as the share of degrada#on toward char due to the condi#ons of par#cle conversion. It also includes an altera#on of the heat (enthalpy) of reac#on, as such an impact was observed experimentally, among others, by Rath et al.[151] The value of the adjustable “” parameter is posi#vely correlated to conversion factors of bio-components like pyrolysis temperature, hea#ng rate, and vapor residence #me within a given par#cle. Those factors are dependent on both par#cle size and process condi#ons. Unfortunately, a lack of data on “excess” char forma#on with respect to conversion condi#ons restricts the deriva#on of a quan#ta#ve correla#on. Therefore, the parameter “”, so far, is set a priori and can be further calibrated to a scenario through itera#ve calibra#on (to improve the precision of a simula#on). As such an approach is subjec#ve and a posterior, the implementa#on of the “” parameter should be improved toward a more objec#ve implementa#on. The precise experimental assessment of the “” parameter in prac#ce is not a straighTorward task. The key element to accurately assess excessive charring is the impact of the hea#ng rate on the conversion pathway, which is inhomogeneous within a par#cle conver#ng in the thermally thick regime. This most commonly relates to a high value for the conversion rate in the outer layer of a par#cle and a rela#vely low value for the conversion rate in the inner volume (which stems from differences in the thermal conduc#vity of wood and char). Contrarily, aMer the conversion, the char yield is measured as an aggregate for the whole par#cle without dis#nguishing loca#on within a par#cle. A proposed solu#on may be coupling an experimental study with a single par#cle model, as the la9er has the ability to dis#nguish between different loca#ons within a par#cle. In the end, such an approach could help with a reliable deriva#on of the “” parameter, with the restric#on that the employed single par#cle model would have to be sufficiently validated, and all bio-components would have to be reliably implemented. In general, no solu#on to this issue has been provided so far. Therefore, the inability to reliably account for excessive charring can be pointed out as one of the key factors that lead to a rela#vely high inaccuracy in model predic#on when conversion of rela#vely large par#cles is simulated with currently available kine#c models. Considering this lack of accuracy in current kine#c models, these models cannot be straighTorwardly applied in reactor-scale models. That, in the end, has a hindering effect on op#mizing pyrolysis reactor designs aiming for biomass-derived char produc#on. The ma9er of reliably predic#ng changes in the porosity in a par#cle is beyond the scope of a kine#c model, while degrada#on of bio-components has a non-negligible impact on the origin of pore development in chars. Regardless of the kine#c model, the changes in par#cle porosity need to be implemented in a single par#cle model to have a suitable predic#on of its overall conversion. In the literature, several studies have focused on establishing a reliable single-par#cle model.[143, 358-366] Among those, few studies have incorporated the changes in par#cle geometry as part of the model, which does not necessarily mean that they accounted for changes in the porosity of a par#cle.[364366] This relates to an issue that, despite the abundance of experimental data on basic proper#es of a char (e.g., elemental composi#on), the number of studies that provide experimental data on changes in the geometrical shape and/or density is scant.[367, 368] Furthermore, there are no studies in the literature that have inves#gated single-par#cle conversion coupled with assessing changes in par#cle geometry and density/porosity (as well as changes in its pore size distribu#on). In terms of the microporosity predic#on, the geometrical changes of a par#cle during pyrolysis may not seem cri#cal. However, the change in par#cle size as induced in pyrolysis has a non-negligible impact on intrapar#cle heat transfer. With scant experimental data and modeling studies, the forma#on of pores is s#ll insufficiently understood. Therefore, it cannot be ruled out that changes in par#cle geometry, and subsequent changes in heat transfer, do not have a relevant impact on pore size distribu#on. The change in par#cle geometry may also have an impact on other phenomena related to par#cle conversion. As probable impacts can be men#oned, changes in the intrapar#cle concentra#on and reten#on #me of evolved vapors due to shortening of the escape pathway of vapors and an increase in CHAPTER 1 58 gas permeability within par#cles due to pore widening. Those factors also may have a non-negligible impact on char forma#on, so indirectly on the pore size distribu#on. The literature does not include a study that inves#gated the men#oned issue experimentally or numerically, so the ques#on of whether changing par#cle geometry in pyrolysis is a relevant phenomenon in view of pore size distribu#on remains open. Aside from the geometrical changes occurring in biomass par#cles during pyrolysis, single-par#cle models presented in the literature do not include even a basic representa#on of changes in porosity and pore size distribu#on in rela#on to process condi#ons. A possible reason for this is rela#vely straighTorward (Fig. 1.14) and relates to a lack of reliable experimental data that would allow for deriving and valida#ng such quan#ta#ve rela#ons. Therefore, without consistent data, there is no founda#on on which the rela#on between conversion process condi#ons or biomass proper#es on the one end and, pore size distribu#on or corresponding SSA on the other end, could be derived. As pointed out in Sec#on 4.5., the issue is widely known in literature and usually refers to pores mainly making up the SSA (i.e., pores in micro-mesoporous size range). However, it is oMen not men#oned that the issue also relates to the macroporous region, which is relevant for char applica#on in soil. The aforemen#oned omission is because of only recent interest in such applica#ons of biomass-derived chars. Nonetheless, an improvement in the field can be observed, especially with studies based on microtomography (µ-CT), which aim to precisely assess the par#cle’s porous structure in the macroporous region.[369, 370] From a generic perspec#ve, it can be concluded that reliable data for the whole range of pore size distribu#on (micro-meso-macro) needs to be experimentally derived before being implemented into a single-par#cle model. Then, the simula#on of porosity change within a single par#cle model can be made, and an inves#ga#on will become possible. A reliable model of a single par#cle conversion has to accurately predict the outcome of its conversion, while the conversion process is transient and occurs unevenly over the volume of a par#cle (especially for larger par#cles assigned to a thermally thick regime). Therefore, precision in the predic#on of transient states of a par#cle has a non-negligible relevance to the overall outcome, as well as to the understanding of the mechanisms underlying par#cle conversion. Naturally, the reliability of such dynamic models requires valida#on datasets, which include #me-based changes of relevant parameters. From a review of single-par#cle models available in the literature, it can be derived that valida#on of such models is conducted with the use of several experimentally derived datasets.[143, 358, 359, 362, 371-374] The datasets include relevant dynamic parameters like, mass loss and temperature (in the par#cle center and at its surface), so a valida#on of a single par#cle model with their use is possible. However, the data is usually limited to the two major parameters men#oned above. Therefore, data on dynamic vapor release profiles, which are cri#cal for improving singlepar#cle models (and the kine#c model implemented within) are missing from the literature. Without such vapor release profiles, the current available data is only par#ally suitable for valida#ng more advanced kine#c models (like the Ranzi and RAC schemes). This situa#on hinders assessing the accuracy and reliability of single par#cle models as well as does not allowing for a cri#cal assessment of imperfec#ons in implemented kine#c models and their poten#al improvements. From a general perspec#ve, a robust and accurate single-par#cle model of biomass par#cle pyrolysis has great poten#al to accelerate the understanding of pore forma#on in biomass-derived chars. Furthermore, if reliably developed, it could also improve the descrip#on of reac#on kine#cs in biomass pyrolysis. Both improvements would lead to more suitable tailoring of the pyrolysis process toward required products, simultaneously reducing the uncertainty and business risks for implemen#ng biooil and biochar systems. With a great aim towards decarboniza#on and widening the applica#on of biomass-derived chars, addressing the issues hindering current progress in single par#cle pyrolysis modeling can be assessed as very relevant. Especially, since generic goals have been drawn, issues have been clearly defined and are widely known. CHAPTER 1 59 6. The scope and goal of the Thesis In view of the Thesis scope, the generic goal was to cri#cally inves#gate the current state-of-the-art in the field of single biomass par#cle pyrolysis, iden#fying obstacles and knowledge gaps, and to provide solu#ons to build a founda#on enabling future development. The forecasted outcome was to help #ghten the “knowledge gap” in predic#ng pore-related proper#es of biomass-derived chars produced in the slow pyrolysis process. If that goal is to be fully achieved, it should accelerate further finding a reliable rela#on between biomass-derived char produc#on and its final applica#on, unlocking the applica#on poten#al of biomass-derived chars. Nonetheless, naive would be the assump#on that this work will provide a comprehensive and broadly validated tool for predic#ng the pore structure during pyrolysis for a global range of cases (i.e., variability in biomass feedstocks, pyrolysis reactor systems, and char applica#ons). As was pointed out in previous sec#ons, the research field is very wide, so providing an objec#vely precise tool for the whole field is out of reach within one doctoral Thesis. As highlighted in Sec#on 5.1, both experimental and modeling work should be done in pairs to derive a meaningful outcome. Therefore, besides the modeling part, the work also included a robust experimental inves#ga#on. The founda#on laid in this work includes two major pillars. The first pillar relates to a cri#cal inves#ga#on of changes in porosity-related proper#es during the process of biomass slow pyrolysis, of which insufficient understanding has hindered development for a long #me. The second pillar is related to the development of a numerical tool, which, through the selec#on of model components on an objec#ve basis and extensive valida#on, would allow for reliable simula#on of the biomass (slow) pyrolysis process. The la9er especially refers to (dynamic) proper#es or parameters that are difficult to assess experimentally during the pyrolysis process. The scope of the work relates to the produc#on of biomass-derived chars, so the decision behind selec#ng slow pyrolysis was an obvious decision. Then, the scope was made narrower regarding wood as the feedstock. The reasons behind this feedstock selec#on were: (1) wood has rela#vely homogeneous composi#on and structure, (2) wood is characterized by a low concentra#on of heteroatoms and cataly#c mineral ma9er, of which the impact on pyrolysis is also rela#vely unknown, (3) having a fairly decent availability of experimental data regarding wood processing through slow pyrolysis, (4) having good forecasts of the sustainable woody biomass poten#al, which is sufficient to impact the GHG emission reduc#on goals, and (5) there is an established market for wood-derived char and its growth stemming from applica#ons where its pore structure is a key requirement, has been predicted. What should be highlighted is that if the selected feedstock would not comply with the first three men#oned restric#ons, a lack of available data would impede further model development. Therefore, the decision to start construc#ng a comprehensive tool for biomass pyrolysis using wood seemed appropriate. In view of the second pillar, the model had to be able to reliably simulate changes in pore-related proper#es, so the single-par#cle scale was selected as the lowest possible scale. That prompted the use of CFD soMware for construc#ng the model and running numerical simula#ons, and the commercial soMware COMSOL TM was selected as it represents an efficient and user-friendly op#on. The work was divided into two major parts, experimental and numerical, where each was progressing stepwise to ensure that each consecu#ve element of the work was based on a solid and cri#cally assessed founda#on. The research chapters of the Thesis are based on peer-reviewed manuscripts that have already been published in high-impact journals and as book chapters. Therefore, the informa#on covered by each chapter is informa#on that was verified and accepted as reliable by the scien#fic community. The structure of the Thesis is provided below, with an indica#on of the objec#ves and mo#va#ons of each chapter (ra#onale) and an indica#on of the published manuscript on which the chapter was based. The generic overview of the Thesis structure is also provided in Fig. 1.19. CHAPTER 1 60 Chapter 1. General introduc,on (this chapter) This chapter contains an elabora#on on the currently established goal of society worldwide toward the reduc#on of GHG emissions and the role of biomass and biomass-derived chars in this endeavor. Furthermore, it provides a perspec#ve on biomass processing technologies with an emphasis on the slow pyrolysis process and applica#ons of biomass-derived chars in view of their property requirements. It ends with an overview of the methods for enhancing progress in the field with numerically aided tools. Experimental part Chapter 2. Structure-dependent proper,es of wood and wood-derived chars – a cri,cal overview of experimental studies The literature overview indicates a great abundance of reviews related to the composi#on of wood and its char. However, the number of reviews related to thermo-physical and structural proper#es is rela#vely low. As wood conversion toward char is significantly related to its thermo-physical proper#es, a cri#cally assessed database to derive reliable parameters needs to be established. The general goal of the work in this Thesis will require such specific data, so the crea#on of a suitable database was needed. Chapter 2 serves as a more directed review, the outcome of which was needed to find reliable input data and thermo-physical rela#ons for implementa#on into a single-par#cle model, as well as data for its valida#on. The content of Chapter 2 was redraMed from the book chapter: P. Maziarka, F. Ronsse and A. Anca-Couce, “Review on modelling approaches based on computa!onal fluid dynamics for biomass pyrolysis systems”, in: Z. Fang, R. L. Smith, L. Xu (Eds.), “Produc!on of Biofuels and Chemicals with Pyrolysis”, Springer Singapore (2020).[375] Chapter 3. Toward reliable assessment of microand meso-structures of biomass-derived chars The generic issue with predic#ng changes in the microand mesoporosity-related proper#es of biomass-derived chars is the low reliability of experimentally obtained data. Chapter 3 provides an inves#ga#on of the causes of low reliability of experimental porosity data through a cri#cal review of the experimental measurement methods and data processing procedures, as well as theore#cal founda#ons of the char pore structure. Furthermore, it proposes, in detail, a reliable solu#on for measuring the microand mesoporosity of biomass-derived chars (including specific surface area and pore volume). The solu#on was derived from own experimental work and validated with literature data. The content of Chapter 3 was redraMed from two publica#ons: (1) P. Maziarka, C. Wurzer, P. J. Arauzo, A. Dieguez-Alonso, O. Mašek and F. Ronsse, “Do you BET on rou!ne? The reliability of N 2 physisorp!on for the quan!ta!ve assessment of biochar’s surface area”, Chemical Engineering Journal (2021) and (2) P. Maziarka, P. Sommersacher, X. Wang, N. Kienzl, S. Retschitzegger, W. Prins, N. Hedin and F. Ronsse, “Tailoring of the pore structures of wood pyrolysis chars for poten!al use in energy storage applica!ons”, Applied Energy (2021).[376, 377] Chapter 4. Experimental study of single wood par,cle pyrolysis with a focus on vapor release characteris,cs and structural changes of char The experimental work presented in Chapter 4 had two aims. The primary aim was to find a correla#on between pyrolysis temperature (300–900°C), par#cle length (Ø8×10 mm and Ø8×16 mm), and structural parameters of wood-based char par#cles. The comprehensive assessment of structure included changes in geometry, surface morphology, porosity, and pore size distribu#on through the whole porosity range (micro-meso-macro), using dedicated analy#cal methods. The second aim was to tackle inaccuracies in experimental data from a single par#cle experiment, which included dynamic changes in basic parameters like temperature and mass loss. CHAPTER 1 61 The study also aimed to construct a broad experimental dataset from single par#cle conversion experiments, which, aside from the basic proper#es, included #me-dependent vapor composi#on (based on 14 quan#fied pyrolysis vapor compounds) and the composi#on of collected bio-oils (quan#ta#vely analyzed with GC-MS/FID). The collected experimental dataset was further used to validate the newly developed single-par#cle models in subsequent chapters in this Thesis. The content of Chapter 4 was redraMed from the two publica#ons: (1) P. Maziarka, N. Kienzl, A. Dieguez-Alonso, V. Fierro, A. Celzard, P. J. Arauzo, N. Hedin, W. Prins, A. Anca-Couce, J. J. Manyà and F. Ronsse, “Part 1 – Impact of pyrolysis temperature and wood par!cle length on vapor cracking and char porous texture in rela!on to the tailoring of char proper!es”, Energy Fuels (2024) and (2) P. Maziarka, P. Sommersacher, X. Wang, N. Kienzl, S. Retschitzegger, W. Prins, N. Hedin and F. Ronsse, “Tailoring of the pore structures of wood pyrolysis chars for poten!al use in energy storage applica!ons”, Applied Energy (2021).[377, 378] Modeling part Chapter 5. Modeling of thermal degrada,on of a single par,cle – a cri,cal overview of components of single par,cle models In the studies related to single par#cle modeling, various rela#ons that are implemented into such models can be found. The aim of Chapter 6 was to synthesize available descrip#ons of phenomena in a homogeneous manner. Furthermore, literature data from Chapter 2 and experimental results derived in Chapter 4 were used to validate and adjust the auxiliary components of single par#cle models. That included a cri#cal assessment of the thermal conduc#vity descrip#on and establishment of the rela#on for permeability based on data from pore size distribu#ons. The systema#zed descrip#on was used in the consecu#ve modeling chapters. The content of Chapter 6 was redraMed from the book chapter: P. Maziarka, F. Ronsse and A. Anca-Couce, “Review on modelling approaches based on computa!onal fluid dynamics for biomass pyrolysis systems”, in: Z. Fang, R. L. Smith, L. Xu (Eds.) “Produc!on of Biofuels and Chemicals with Pyrolysis”, Springer Singapore (2020).[375] Chapter 6. CFD models of pyrolysis of a single wood par,cle – Valida,on and performance assessment of components Single par#cle models, found in literature, usually consist of one of the available descrip#ons of relevant phenomena which explicitly describe biomass anisotropy, drying of a biomass par#cle, and the implemented chemical reac#on scheme. Chapter 6 consists of a stepwise compara#ve inves#ga#on, which aims to objec#vely prove which of the possible decisions (in terms of which model components to include and which ones not) leads to the most precise and reliable outcome of a singlepar#cle model. Furthermore, the described inves#ga#on includes valida#on of the model over broad literature data (including temperature, par#cle shape, and size) to confirm the reliability of the developed single-par#cle model. The presented study addi#onally inves#gated the distribu#on of interpar#cle hea#ng rate during the pyrolysis process to iden#fy its impact on the overall conversion. The content of Chapter 7 was redraMed from the publica#on: P. Maziarka, A. Anca-Couce, W. Prins and F. Ronsse “A meta-analysis of thermo-physical and chemical aspects in CFD modelling of pyrolysis of a single wood par!cle in the thermally thick regime”, Chemical Engineering Journal (2022).[379] CHAPTER 1 62 Chapter 7. Comprehensive CFD model for pyrolysis of a single wood par,cle – novel extensions and the model capabili,es for tailoring of the process Chapter 7 was the outcome of work described in all previous chapters and presents the newly developed comprehensive model for the precise predic#on of the products from single par#cle pyrolysis. The established model described in Chapter 7 was based on previously assessed and validated components of the model (Chapter 6). Furthermore, the model was extended with geometrical deforma#on, which allowed to include the par#cle shrinking during pyrolysis, allowing for a real-#me inves#ga#on of density and porosity changes in a par#cle. The stepwise compara#ve inves#ga#on described in Chapter 7 includes an assessment of modeling par#cle shrinking, secondary cracking of evolved pyroly#c vapors, and the impact of the par#cle permeability descrip#on. The performance of the model was validated with the experimental dataset as described in Chapter 4. The content of Chapter 7 was redraMed from the publica#on – P. Maziarka, N. Kienzl, A. Dieguez-Alonso, W. Prins, P. J. Arauzo, Ø. Skreiberg, A. Anca-Couce, J. J. Manyà, and F. Ronsse “Part 2 – Tailoring of pyroly!c char proper!es with a single-par!cle CFD model with focus on the impact of shrinking, vapor cracking, and char permeability” Energy Fuels (2024).[380] Chapter 8. General summary, conclusions, and outlook This chapter is the last chapter of the thesis, which includes a synthesis of the main developments obtained from Chapters 2 to 8. Moreover, it contains the major conclusions of the conducted work, which include a cri#cal evalua#on of their meaning in rela#on to the predic#on of pore-related proper#es of biomass-derived chars and the possibili#es of their modeling. Chapter 8 finishes with conclusions and an outlook on the results presented in this Thesis. As Chapter 1 provides an extensive introduc#on, the specific introduc#ons provided in each chapter were significantly reduced to avoid unnecessary redundancy. The specific introduc#ons in each chapter contain only the most relevant informa#on related to the issue elaborated within each chapter. At the end of each chapter, a list of the references used was provided to ease finding relevant sources for readers. Appendices to the chapters are provided at the end of the Thesis. The Appendices are named “Appendix X for Chapter Y”, where the “X” represents the number of the appendix, and the “Y” represents the number of chapters to which the appendix corresponds. The Sec#ons, Tables, and Figures in the Appendix were referred to in the text with the le9er “A” to indicate their loca#on. Fig. 1.19. The generic overview of the Thesis structure. CHAPTER 1 63 7. References for Chapter 1 [1] H. Le Treut, R. Somerville, U. Cubasch, Y. Ding, C. Mauritzen, A. Mokssit, T. Peterson, M. Prather, Historical Overview of Climate Change, in: D.Q. S. Solomon., M. Manning, Z. Chen, M. Marquis, K.B. Averyt, M. Tignor and H.L. Miller (Ed.) Climate Change 2007: The Physical Science Basis. Contribu#on of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change (AR4), Cambridge University Press, Cambridge, 2007, pp. 93-127. [2] J.R. Fleming, Global Warming and Anthropogenic CO2, The Callendar Effect: The Life and Work of Guy Stewart Callendar (1898–1964), the Scien#st Who Established the Carbon Dioxide Theory of Climate Change, American Meteorological Society, Boston, MA, 2007, pp. 65-87. [3] The Intergovernmental Panel on Climate Change, History of the IPCC, 2024. [4] Greenhouse Gas Protocol, IPCC Global Warming Poten#al Values (v 2.0, Aug 2024), (2024) 1-10. [5] G.-J. Nabuurs, R. Mrabet, A. Abu Hatab, M. Bustamante, H. Clark, P. Havlík, J. House, C. Mbow, K.N. Ninan, A. Popp, S. Roe, B. Sohngen, S. Towprayoon, Chapter 7: Agriculture, Forestry, and Other Land Uses (AFOLU), in: J.S. P.R. Shukla, R. Slade, A. Al Khourdajie, R. van Diemen, D. McCollum, M. Pathak, S. Some, P. Vyas, R. Fradera, M. Belkacemi, A. Hasija, G. Lisboa, S. Luz, J. Malley (Ed.) Climate Change 2022: Mi#ga#on of Climate Change. Contribu#on of Working Group III to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change (AR6), Cambridge University Press, Cambridge, 2022, pp. 747-860. [6] J. Rissman, C. Bataille, E. Masanet, N. Aden, W.R. Morrow, N. Zhou, N. Ellio9, R. Dell, N. Heeren, B. Huckestein, J. Cresko, S.A. Miller, J. Roy, P. Fennell, B. Cremmins, T. Koch Blank, D. Hone, E.D. Williams, S. de la Rue du Can, B. Sisson, M. Williams, J. Katzenberger, D. Burtraw, G. Sethi, H. Ping, D. Danielson, H. Lu, T. Lorber, J. Dinkel, J. Helseth, Technologies and policies to decarbonize global industry: Review and assessment of mi#ga#on drivers through 2070, Applied Energy 266 (2020) 114848. [7] S. Dhakal, J.C. Minx, F.L. Toth, A. Abdel-Aziz, M.J. Figueroa Meza, K. Hubacek, I.G.C. Jonckheere, K. Yong-Gun, G.F. Nemet, S. Pachauri, X.C. Tan, T. Wiedmann, Chapter 2: Emissions trends and drivers, in: J.S. P.R. Shukla, R. Slade, A. Al Khourdajie, R. van Diemen, D. McCollum, M. Pathak, S. Some, P. Vyas, R. Fradera, M. Belkacemi, A. Hasija, G. Lisboa, S. Luz, J. Malley (Ed.) Climate Change 2022: Mi#ga#on of Climate Change. Contribu#on of Working Group III to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change (AP6), Cambridge University Press, Cambridge, 2022, pp. 215-294. [8] W. Arts, I. Storms, J. Van Aelst, B. Lagrain, B. Verbist, J. Van Orshoven, P.J. Verkerk, W. Vermeiren, J.-P. Lange, B. Muys, B.F. Sels, Feasibility of wood as a renewable carbon feedstock for the produc#on of chemicals in Europe, Biofuels, Bioproducts and Biorefining 18 (2024) 365-377. [9] L. Clarke, Y.-M. Wei, A. De La Vega Navarro, A. Garg, A.N. Hahmann, S. Khennas, I.M.L. Azevedo, A. Löschel, A.K. Singh, L. Steg, G. Strbac, K. Wada, Chapter 6: Energy systems, in: J.S. P.R. Shukla, R. Slade, A. Al Khourdajie, R. van Diemen, D. McCollum, M. Pathak, S. Some, P. Vyas, R. Fradera, M. Belkacemi, A. Hasija, G. Lisboa, S. Luz, J. Malley (Ed.) Climate Change 2022: Mi#ga#on of Climate Change. Contribu#on of Working Group III to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change (AP6), Cambridge University Press, Cambridge, 2022, pp. 613-746. [10] G. Ahamer, Why Biomass Fuels Are Principally Not Carbon Neutral, Energies 15 (2022) 9619. [11] G. Ahamer, How to Compute Whether Biomass Fuels Are Carbon Neutral, C 10 (2024) 48. [12] K. Navare, W. Arts, G. Faraca, G.V.d. Bossche, B. Sels, K.V. Acker, Environmental impact assessment of cascading use of wood in bio-fuels and bio-chemicals, Resources, Conserva#on and Recycling 186 (2022) 106588. [13] P. Jaramillo, S. Kahn Ribeiro, P. Newman, S. Dhar, O.E. Diemuodeke, T. Kajino, D.S. Lee, S.B. Nugroho, X. Ou, A. Hammer Strømman, J. Whitehead, Chapter 10: Transport, in: J.S. P.R. Shukla, R. Slade, A. Al Khourdajie, R. van Diemen, D. McCollum, M. Pathak, S. Some, P. Vyas, R. Fradera, M. Belkacemi, A. Hasija, G. Lisboa, S. Luz, J. Malley (Ed.) Climate Change 2022: Mi#ga#on of Climate Change. Contribu#on of Working Group III to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change (AR 6), Cambridge University Press, Cambridge, 2022, pp. 1049-1160. [14] S. Tsemekidi Tzeiranaki, M. Economidou, P. Bertoldi, C. Thiel, G. Fontaras, E.L. Clemen#, C. Franco De Los Rios, The impact of energy efficiency and decarbonisa#on policies on the European road transport sector, Transporta#on Research Part A: Policy and Prac#ce 170 (2023) 103623. [15] A. Muscat, E.M. de Olde, I.J.M. de Boer, R. Ripoll-Bosch, The ba9le for biomass: A systema#c review of food-feed-fuel compe##on, Global Food Security 25 (2020) 100330. [16] S.K. Sikdar, F.T. Princio9a, Advances in carbon management technologies biomass u#liza#on, manufacturing, and electricity management - Volume 2, CRC Press, Boca Raton, 2022. [17] T. Abbasi, S.A. Abbasi, Biomass energy and the environmental impacts associated with its produc#on and u#liza#on, Renewable and Sustainable Energy Reviews 14 (2010) 919-937. [18] A. Fadiga, L.M.D.F. Ferreira, J.F. Bigo9e, Decarbonising mari#me ports: A systema#c review of the literature and insights for new research opportuni#es, Journal of Cleaner Produc#on 452 (2024) 142209. [19] F. ZanobeR, G. Pio, S. Jafarzadeh, M. Muñoz Or#z, V. Cozzani, Decarboniza#on of mari#me transport: Sustainability assessment of alterna#ve power systems, Journal of Cleaner Produc#on 417 (2023) 137989. [20] T.I. Simonsen, N.D. Weiss, S. van Dyk, E. van Thuijl, S. Tjalfe Thomsen, Progress towards biofuels for marine shipping, in: S.T. Thomsen (Ed.), IEA Bioenergy: Task 39, Online, 2021, pp. 1-89. [21] C.-w.C. Hsieh, C. Felby, Biofuels for the marine shipping sector IEA Bioenergy: Task 39, Online, 2017, pp. 1-86. [22] M.J. Watson, P.G. Machado, A.V. da Silva, Y. Saltar, C.O. Ribeiro, C.A.O. Nascimento, A.W. Dowling, Sustainable avia#on fuel technologies, costs, emissions, policies, and markets: A cri#cal review, Journal of Cleaner Produc#on 449 (2024) 141472. [23] B.W. Kolosz, Y. Luo, B. Xu, M.M. Maroto-Valer, J.M. Andresen, Life cycle environmental analysis of ‘drop in’ alterna#ve avia#on fuels: a review, Sustainable Energy & Fuels 4 (2020) 3229-3263. CHAPTER 1 64 [24] J. O’Malley, C. Baldino, Availability of biomass feedstocks in the European Union to meet the 2035 ReFuelEU Avia#on SAF target, Interna#onal Council on Clean Transporta#on (ICCT), Online, 2024, pp. 1-15. [25] S.S. van Dyk, Jack, Progress in Commercializa#on of Biojet /Sustainable Avia#on Fuels (SAF), Technologies and policies, IEA Bioenergy: Task 39, Online, 2024, pp. 1-88. [26] T.D. Nielsen, J. Hasselbalch, K. Holmberg, J. Stripple, Poli#cs and the plas#c crisis: A review throughout the plas#c life cycle, WIREs Energy and Environment 9 (2020) e360. [27] E.A.R. Zuiderveen, K.J.J. Kuipers, C. Caldeira, S.V. Hanssen, M.K. van der Hulst, M.M.J. de Jonge, A. Vlysidis, R. van Zelm, S. Sala, M.A.J. Huijbregts, The poten#al of emerging bio-based products to reduce environmental impacts, Nature Communica#ons 14 (2023) 8521. [28] E. de Jong, H. S#chnothe, G. Bell, H. Jørgensen, I. de Bari, J. van Haveren, J. Lindorfer, Bio-Based Chemicals - A 2020 Update, IEA Bioenergy: Task 42, Online, 2020, pp. 1-79. [29] Interna#onal Organisa#on for Standardisa#on (ISO), ISO 14044:2006 - Environmental Management—Life Cycle Assessment—Requirements and Guidelines, Interna#onal Organisa#on for Standardisa#on, Geneve, 2006, pp. 1-46. [30] Interna#onal Organisa#on for Standardisa#on (ISO), ISO 14040:2006 - Environmental management — Life cycle assessment — Principles and framework, Interna#onal Organisa#on for Standardisa#on, Geneve, 2006, pp. 1-46. [31] C. Choe, J.A. Moon, J. Gu, A. Lee, H. Lim, Life cycle sustainability assessment for sustainable energy future: A short review on opportunity and challenge, Current Opinion in Green and Sustainable Chemistry 50 (2024) 100974. [32] D. Costa, P. Quinteiro, A.C. Dias, A systema#c review of life cycle sustainability assessment: Current state, methodological challenges, and implementa#on issues, Science of The Total Environment 686 (2019) 774-787. [33] I.A. Bashmakov, L.J. Nilsson, A. Acquaye, C. Bataille, J.M. Cullen, S. de la Rue du Can, M. Fischedick, Y. Geng, K. Tanaka, Chapter 11: Industry, in: J.S. P.R. Shukla, R. Slade, A. Al Khourdajie, R. van Diemen, D. McCollum, M. Pathak, S. Some, P. Vyas, R. Fradera, M. Belkacemi, A. Hasija, G. Lisboa, S. Luz, J. Malley (Ed.) Climate Change 2022: Mi#ga#on of Climate Change. Contribu#on of Working Group III to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, 2022, pp. 1161-1244. [34] Ó. Ögmundarson, M.J. Herrgård, J. Forster, M.Z. Hauschild, P. Fantke, Addressing environmental sustainability of biochemicals, Nature Sustainability 3 (2020) 167-174. [35] R. Fu, L. Kang, C. Zhang, Q. Fei, Applica#on and progress of techno-economic analysis and life cycle assessment in biomanufacturing of fuels and chemicals, Green Chemical Engineering 4 (2023) 189-198. [36] K. Huang, X. Peng, L. Kong, W. Wu, Y. Chen, C.T. Maravelias, Greenhouse Gas Emission Mi#ga#on Poten#al of Chemicals Produced from Biomass, ACS Sustainable Chemistry & Engineering 9 (2021) 14480-14487. [37] M.G. Davidson, S. Elgie, S. Parsons, T.J. Young, Produc#on of HMF, FDCA and their derived products: a review of life cycle assessment (LCA) and techno-economic analysis (TEA) studies, Green Chemistry 23 (2021) 3154-3171. [38] J. Wang, S. You, Z. Lu, R. Chen, F. Xu, Life cycle assessment of bio-based levoglucosan produc#on from co9on straw through fast pyrolysis, Bioresource Technology 307 (2020) 123179. [39] P.P. Van Uytvanck, B. Hallmark, G. Haire, P.J. Marshall, J.S. Dennis, Impact of Biomass on Industry: Using Ethylene Derived from Bioethanol within the Polyester Value Chain, ACS Sustainable Chemistry & Engineering 2 (2014) 1098-1105. [40] S. Roy, A. Cherevotan, S.C. Peter, Thermochemical CO2 Hydrogena#on to Single Carbon Products: Scien#fic and Technological Challenges, ACS Energy Le9ers 3 (2018) 1938-1966. [41] J. Artz, T.E. Müller, K. Thenert, J. Kleinekorte, R. Meys, A. Sternberg, A. Bardow, W. Leitner, Sustainable Conversion of Carbon Dioxide: An Integrated Review of Catalysis and Life Cycle Assessment, Chemical Reviews 118 (2018) 434-504. [42] A. Vilén, P. Laurell, R. Vahala, Compara#ve life cycle assessment of ac#vated carbon produc#on from various raw materials, Journal of Environmental Management 324 (2022) 116356. [43] P.L. Cruz, D. Iribarren, J. Dufour, Modeling, simula#on and life-cycle assessment of the use of bio-oil and char in conven#onal refineries, Biofuels, Bioproducts and Biorefining 14 (2020) 30-42. [44] T.M. Lammens, Effect of Various Green Carbon Tracking Methods on Life Cycle Assessment Results for Fluid Cataly#c Cracker Co-processing of Fast Pyrolysis Bio-oil, Energy & Fuels 36 (2022) 12617-12627. [45] C. Wang, R. Venderbosch, Y. Fang, Co-processing of crude and hydrotreated pyrolysis liquids and VGO in a pilot scale FCC riser setup, Fuel Processing Technology 181 (2018) 157-165. [46] X. Han, H. Wang, Y. Zeng, J. Liu, Advancing the applica#on of bio-oils by co-processing with petroleum intermediates: A review, Energy Conversion and Management: X 10 (2021) 100069. [47] S. van Dyk, J. Su, J.D. Mcmillan, J. Saddler, Poten#al synergies of drop-in biofuel produc#on with further co-processing at oil refineries, Biofuels, Bioproducts and Biorefining 13 (2019) 760-775. [48] H. Lund, I.R. Skov, J.Z. Thellufsen, P. Sorknæs, A.D. Korberg, M. Chang, B.V. Mathiesen, M.S. Kany, The role of sustainable bioenergy in a fully decarbonised society, Renewable Energy 196 (2022) 195-203. [49] D. Chiaramon#, J. Lehmann, F. Berru#, P. Giudicianni, H. Sanei, O. Masek, Biochar is a long-lived form of carbon removal, making evidence-based CDR projects possible, Biochar 6 (2024) 81. [50] S. Fawzy, A.I. Osman, N. Mehta, D. Moran, A.a.H. Al-Muhtaseb, D.W. Rooney, Atmospheric carbon removal via industrial biochar systems: A techno-economic-environmental study, Journal of Cleaner Produc#on 371 (2022) 133660. [51] S. Fawzy, A.I. Osman, H. Yang, J. Doran, D.W. Rooney, Industrial biochar systems for atmospheric carbon removal: a review, Environmental Chemistry Le9ers 19 (2021) 3023-3055. [52] H. de Coninck, A. Revi, M. Babiker, P. Bertoldi, M. Buckeridge, A. Cartwright, W. Dong, J. Ford, S. Fuss, J.-C. Hourcade, D. Ley, R. Mechler, P. Newman, A. Revokatova, S. Schultz, L. Steg, T. Sugiyama, Chapter 4: Strengthening and implemen#ng the global response, in: P.Z. V. Masson-Delmo9e, H.-O. Pörtner, D. Roberts, J. Skea, P.R. Shukla, A. Pirani, W. Moufouma-Okia, C. CHAPTER 1 71 [226] S.H. Bai, N. Omidvar, M. Gallart, W. Kämper, I. Tahmasbian, M.B. Farrar, K. Singh, G. Zhou, B. Muqadass, C.-Y. Xu, R. Koech, Y. Li, T.T.N. Nguyen, L. van Zwieten, Combined effects of biochar and fer#lizer applica#ons on yield: A review and meta-analysis, Science of The Total Environment 808 (2022) 152073. [227] A. El-Naggar, S.S. Lee, J. Rinklebe, M. Farooq, H. Song, A.K. Sarmah, A.R. Zimmerman, M. Ahmad, S.M. Shaheen, Y.S. Ok, Biochar applica#on to low fer#lity soils: A review of current status, and future prospects, Geoderma 337 (2019) 536-554. [228] M.W.I. Schmidt, M.S. Torn, S. Abiven, T. Di9mar, G. Guggenberger, I.A. Janssens, M. Kleber, I. Kögel-Knabner, J. Lehmann, D.A.C. Manning, P. Nannipieri, D.P. Rasse, S. Weiner, S.E. Trumbore, Persistence of soil organic ma9er as an ecosystem property, Nature 478 (2011) 49-56. [229] Y. Ding, Y. Liu, S. Liu, Z. Li, X. Tan, X. Huang, G. Zeng, L. Zhou, B. Zheng, Biochar to improve soil fer#lity. A review, Agronomy for Sustainable Development 36 (2016) 36. [230] E. Tenic, R. Ghogare, A. Dhingra, Biochar—A Panacea for Agriculture or Just Carbon?, Hor#culturae 6 (2020) 37. [231] J. Hammond, S. Shackley, M. Prendergast-Miller, J. Cook, S. Buckingham, V.A. Pappa, Biochar field tes#ng in the UK: outcomes and implica#ons for use, Carbon Management 4 (2013) 159-170. [232] S. Jeffery, D. Abalos, M. Prodana, A.C. Bastos, J.W. van Groenigen, B.A. Hungate, F. Verheijen, Biochar boosts tropical but not temperate crop yields, Environmental Research Le9ers 12 (2017) 053001. [233] L. Ye, M. Camps-Arbestain, Q. Shen, J. Lehmann, B. Singh, M. Sabir, Biochar effects on crop yields with and without fer#lizer: A meta-analysis of field studies using separate controls, Soil Use and Management 36 (2020) 2-18. [234] G. Agegnehu, A.K. Srivastava, M.I. Bird, The role of biochar and biochar-compost in improving soil quality and crop performance: A review, Applied Soil Ecology 119 (2017) 156-170. [235] Y. Jiang, T. Li, X. Xu, J. Sun, G. Pan, K. Cheng, A global assessment of the long-term effects of biochar applica#on on crop yield, Current Research in Environmental Sustainability 7 (2024) 100247. [236] J. Wang, L. Shi, L. Zhai, H. Zhang, S. Wang, J. Zou, Z. Shen, C. Lian, Y. Chen, Analysis of the long-term effec#veness of biochar immobiliza#on remedia#on on heavy metal contaminated soil and the poten#al environmental factors weakening the remedia#on effect: A review, Ecotoxicology and Environmental Safety 207 (2021) 111261. [237] K.N. Palansooriya, S.M. Shaheen, S.S. Chen, D.C.W. Tsang, Y. Hashimoto, D. Hou, N.S. Bolan, J. Rinklebe, Y.S. Ok, Soil amendments for immobiliza#on of poten#ally toxic elements in contaminated soils: A cri#cal review, Environment Interna#onal 134 (2020) 105046. [238] Y. Gong, D. Zhao, Q. Wang, An overview of field-scale studies on remedia#on of soil contaminated with heavy metals and metalloids: Technical progress over the last decade, Water Research 147 (2018) 440-460. [239] L. He, H. Zhong, G. Liu, Z. Dai, P.C. Brookes, J. Xu, Remedia#on of heavy metal contaminated soils by biochar: Mechanisms, poten#al risks and applica#ons in China, Environmental Pollu#on 252 (2019) 846-855. [240] F.U. Haider, C. Liqun, J.A. Coulter, S.A. Cheema, J. Wu, R. Zhang, M. Wenjun, M. Farooq, Cadmium toxicity in plants: Impacts and remedia#on strategies, Ecotoxicology and Environmental Safety 211 (2021) 111887. [241] J. Wang, S. Wang, Prepara#on, modifica#on and environmental applica#on of biochar: A review, Journal of Cleaner Produc#on 227 (2019) 1002-1022. [242] D.C.C.d.S. Medeiros, C. Nzediegwu, C. Benally, S.A. Messele, J.-H. Kwak, M.A. Naeth, Y.S. Ok, S.X. Chang, M. Gamal ElDin, Pris#ne and engineered biochar for the removal of contaminants co-exis#ng in several types of industrial wastewaters: A cri#cal review, Science of The Total Environment 809 (2022) 151120. [243] X. Tan, Y. Liu, G. Zeng, X. Wang, X. Hu, Y. Gu, Z. Yang, Applica#on of biochar for the removal of pollutants from aqueous solu#ons, Chemosphere 125 (2015) 70-85. [244] D. Mohan, A. Sarswat, Y.S. Ok, C.U. Pi9man, Organic and inorganic contaminants removal from water with biochar, a renewable, low cost and sustainable adsorbent – A cri#cal review, Bioresource Technology 160 (2014) 191-202. [245] W. Xiang, X. Zhang, J. Chen, W. Zou, F. He, X. Hu, D.C.W. Tsang, Y.S. Ok, B. Gao, Biochar technology in wastewater treatment: A cri#cal review, Chemosphere 252 (2020) 126539. [246] T.G. Ambaye, M. Vaccari, E.D. van Hullebusch, A. Amrane, S. R#mi, Mechanisms and adsorp#on capaci#es of biochar for the removal of organic and inorganic pollutants from industrial wastewater, Interna#onal Journal of Environmental Science and Technology 18 (2021) 3273-3294. [247] G. Enaime, A. Baçaoui, A. Yaacoubi, M. Lübken, Biochar for Wastewater Treatment—Conversion Technologies and Applica#ons, Applied Sciences 10 (2020) 3492. [248] N.A. Qambrani, M.M. Rahman, S. Won, S. Shim, C. Ra, Biochar proper#es and eco-friendly applica#ons for climate change mi#ga#on, waste management, and wastewater treatment: A review, Renewable and Sustainable Energy Reviews 79 (2017) 255-273. [249] E. Rosales, J. Meijide, M. Pazos, M.A. Sanromán, Challenges and recent advances in biochar as low-cost biosorbent: From batch assays to con#nuous-flow systems, Bioresource Technology 246 (2017) 176-192. [250] D. Kołodyńska, J. Krukowska, P. Thomas, Comparison of sorp#on and desorp#on studies of heavy metal ions from biochar and commercial ac#ve carbon, Chemical Engineering Journal 307 (2017) 353-363. [251] X. Xiong, I.K.M. Yu, L. Cao, D.C.W. Tsang, S. Zhang, Y.S. Ok, A review of biochar-based catalysts for chemical synthesis, biofuel produc#on, and pollu#on control, Bioresource Technology 246 (2017) 254-270. [252] K.-D. Henning, H. von Kienle, Ac#vated Carbon, Industrial Carbon and Graphite Materials, Wiley-VCH, Weinheim, 2021, pp. 491-531. [253] V.R. Deitz, History of commercial absorbents in rela#on to the suger refining industry, in: U.N.B.o. Standards (Ed.) Bibliography of Solid Adsorbents: An Annota#ve Bibliographical Survey of the Scien#fic Literature on Bone Char, Ac#vated CHAPTER 1 72 Carbons, and Other Technical Solid Adsorbents, for the Years 1900 to 1942 US Na#onal Bureau of Standards, Washington D.C., 1944, pp. IX - LX. [254] R.K. Wildner, Die Geschichte der Ak#vkohle: Von der Holzkohle im Altertum zum anspruchsvollen Technologieprodukt der heu#gen Industrie, AVM – Akademische VerlagsgemeinschaM München, München, 2014, pp. 1-170. [255] R. Child, Coconut shells as an industrial raw material IV. Coconut shell charcoal: (A) Commercial, Current Science 13 (1944) 245-250. [256] R. Child, Coconut shells as an industrial raw material IV. Coconut shell charcoal: (B) Ac#vated carbon, Current Science 16 (1947) 5-8. [257] X. Cao, S. Sun, R. Sun, Applica#on of biochar-based catalysts in biomass upgrading: a review, RSC Advances 7 (2017) 48793-48805. [258] J.A. Menéndez-Díaz, I. Mar|n-Gullón, Chapter 1 Types of carbon adsorbents and their produc#on, in: T.J. Bandosz (Ed.) Interface Science and Technology, Elsevier2006, pp. 1-47. [259] N.P. Stadie, J.J. Purewal, C.C. Ahn, B. Fultz, Measurements of Hydrogen Spillover in Pla#num Doped Superac#vated Carbon, Langmuir 26 (2010) 15481-15485. [260] A.S. Alsaman, E.M.M. Ibrahim, M.S. Ahmed, A.M. Farid, E.S. Ali, A.A. Askalany, Experimental inves#ga#on of innova#ve composite Maxsorb III adsorbent for cooling and water desalina#on, Journal of Cleaner Produc#on 418 (2023) 138047. [261] M. Wachtler, O. ÖRnger, R. Schweiss, Carbon and Graphite for Electrochemical Power Sources, Industrial Carbon and Graphite Materials, Wiley-VCH, Weinheim, 2021, pp. 379-455. [262] J.R. Kastner, J. Miller, D.P. Geller, J. Locklin, L.H. Keith, T. Johnson, Cataly#c esterifica#on of fa9y acids using solid acid catalysts generated from biochar and ac#vated carbon, Catalysis Today 190 (2012) 122-132. [263] M.D. Kos#ć, A. Bazargan, O.S. Stamenković, V.B. Veljković, G. McKay, Op#miza#on and kine#cs of sunflower oil methanolysis catalyzed by calcium oxide-based catalyst derived from palm kernel shell biochar, Fuel 163 (2016) 304-313. [264] T. Do Minh, J. Song, A. Deb, L. Cha, V. Srivastava, M. Sillanpää, Biochar based catalysts for the abatement of emerging pollutants: A review, Chemical Engineering Journal 394 (2020) 124856. [265] M.J. Valero-Romero, M.Á. Rodríguez-Cano, J. Palomo, J. Rodríguez-Mirasol, T. Cordero, Carbon-Based Materials as Catalyst Supports for Fischer–Tropsch Synthesis: A Review, Fron#ers in Materials 7 (2021). [266] H. Jüntgen, Ac#vated carbon as catalyst support: A review of new research results, Fuel 65 (1986) 1436-1446. [267] Y. Shen, Chars as carbonaceous adsorbents/catalysts for tar elimina#on during biomass pyrolysis or gasifica#on, Renewable and Sustainable Energy Reviews 43 (2015) 281-295. [268] E. Auer, A. Freund, J. Pietsch, T. Tacke, Carbons as supports for industrial precious metal catalysts, Applied Catalysis A: General 173 (1998) 259-271. [269] F. Çeçen, Water and Wastewater Treatment: Historical Perspec#ve of Ac#vated Carbon Adsorp#on and its Integra#on with Biological Processes, in: F.Ç.a.Ö. Aktaş (Ed.) Ac#vated Carbon for Water and Wastewater Treatment, Wiley-VCH Verlag & Co. KGaA, Weinheim, 2011, pp. 1-11. [270] C. Goel, S. Mohan, P. Dinesha, CO2 capture by adsorp#on on biomass-derived ac#vated char: A review, Science of The Total Environment 798 (2021) 149296. [271] Y. Chen, X. Zhang, W. Chen, H. Yang, H. Chen, The structure evolu#on of biochar from biomass pyrolysis and its correla#on with gas pollutant adsorp#on performance, Bioresource Technology 246 (2017) 101-109. [272] I. Angelidaki, L. Treu, P. Tsapekos, G. Luo, S. Campanaro, H. Wenzel, P.G. Kougias, Biogas upgrading and u#liza#on: Current status and perspec#ves, Biotechnology Advances 36 (2018) 452-466. [273] A.G. Georgiadis, N.D. Charisiou, M.A. Goula, Removal of Hydrogen Sulfide From Various Industrial Gases: A Review of The Most Promising Adsorbing Materials, Catalysts 10 (2020) 521. [274] K.P. Gopinath, D.-V.N. Vo, D. Gnana Prakash, A. Adithya Joseph, S. Viswanathan, J. Arun, Environmental applica#ons of carbon-based materials: a review, Environmental Chemistry Le9ers 19 (2021) 557-582. [275] A.K. Vuppaladadiyam, M.K. Jena, I.G. Hakeem, S. Patel, G. Veluswamy, A.V. Thulasiraman, A. Surapaneni, K. Shah, A cri#cal review of biochar versus hydrochar and their applica#on for H2S removal from biogas, Reviews in Environmental Science and Bio/Technology 23 (2024) 699-737. [276] A.G. Pandolfo, A.F. Hollenkamp, Carbon proper#es and their role in supercapacitors, Journal of Power Sources 157 (2006) 11-27. [277] E. Frackowiak, F. Béguin, Carbon materials for the electrochemical storage of energy in capacitors, Carbon 39 (2001) 937-950. [278] P.R. Bandaru, H. Yamada, R. Narayanan, M. Hoefer, Charge transfer and storage in nanostructures, Materials Science and Engineering: R: Reports 96 (2015) 1-69. [279] T.A. Centeno, O. Sereda, F. Stoeckli, Capacitance in carbon pores of 0.7 to 15 nm: a regular pa9ern, Physical Chemistry Chemical Physics 13 (2011) 12403-12406. [280] W. Gu, G. Yushin, Review of nanostructured carbon materials for electrochemical capacitor applica#ons: advantages and limita#ons of ac#vated carbon, carbide-derived carbon, zeolite-templated carbon, carbon aerogels, carbon nanotubes, onionlike carbon, and graphene, WIREs Energy and Environment 3 (2014) 424-473. [281] M. Inagaki, H. Konno, O. Tanaike, Carbon materials for electrochemical capacitors, Journal of Power Sources 195 (2010) 7880-7903. [282] A. Benítez, J. Amaro-Gahete, Y.-C. Chien, Á. Caballero, J. Morales, D. Brandell, Recent advances in lithium-sulfur ba9eries using biomass-derived carbons as sulfur host, Renewable and Sustainable Energy Reviews 154 (2022) 111783. CHAPTER 1 73 [283] J. Wang, P. Nie, B. Ding, S. Dong, X. Hao, H. Dou, X. Zhang, Biomass derived carbon for energy storage devices, Journal of Materials Chemistry A 5 (2017) 2411-2428. [284] E.E. Miller, Y. Hua, F.H. Tezel, Materials for energy storage: Review of electrode materials and methods of increasing capacitance for supercapacitors, Journal of Energy Storage 20 (2018) 30-40. [285] A. Ghosh, Y.H. Lee, Carbon-Based Electrochemical Capacitors, ChemSusChem 5 (2012) 480-499. [286] M. Bartoli, M. Giorcelli, P. Jagdale, M. Rovere, A. Tagliaferro, A Review of Non-Soil Biochar Applica#ons, Materials 13 (2020) 261. [287] X. Li, J. Zhang, B. Liu, Z. Su, A cri#cal review on the applica#on and recent developments of post-modified biochar in supercapacitors, Journal of Cleaner Produc#on 310 (2021) 127428. [288] M. Thommes, K. Kaneko, V. Neimark Alexander, P. Olivier James, F. Rodriguez-Reinoso, J. Rouquerol, S.W. Sing Kenneth, Physisorp#on of gases, with special reference to the evalua#on of surface area and pore size distribu#on (IUPAC Technical Report), Pure and Applied Chemistry, 2015, pp. 1051. [289] Soil Science Society of America (SSSA), Glossary of soil science terms, Soil Science Society of America (SSSA), Madison, 2010. [290] R.V. Santos, M.A.A. Mendes, C. Alexandre, M.R. Carro9, A. Rodrigues, A.F. Ferreira, Assessment of Biomass and Biochar of Mari#me Pine as a Porous Medium for Water Reten#on in Soils, Energies 15 (2022) 5882. [291] I.J. Bingham, D. Robinson, ROOT DEVELOPMENT | Root Growth and Development, in: B. Thomas (Ed.) Encyclopedia of Applied Plant Sciences, Elsevier, Oxford, 2003, pp. 1115-1123. [292] C.A. Meisner, K.J. Karnok, Root Hair Occurrence anld Varia#on with Environment, Agronomy Journal 83 (1991) 814-818. [293] R.A. Zabel, J.J. Morrell, Chapter Three - The characteris#cs and classifica#on of fungi and bacteria, in: R.A. Zabel, J.J. Morrell (Eds.) Wood Microbiology (Second Edi#on), Academic Press, San Diego, 2020, pp. 55-98. [294] B. Bunk, A. Schulz, S. Stammen, R. Münch, M.J. Warren, D. Jahn, R. Biedendieck, A short story about a big magic bug, Bioengineered Bugs 1 (2010) 85-91. [295] S. Adhikari, W. Timms, M.A.P. Mahmud, Op#mising water holding capacity and hydrophobicity of biochar for soil amendment – A review, Science of The Total Environment 851 (2022) 158043. [296] A. Pfriem, M. Zauer, A. Wagenführ, Altera#on of the pore structure of spruce (Picea abies (L.) Karst.) and maple (Acer pseudoplatanus L.) due to thermal treatment as determined by helium pycnometry and mercury intrusion porosimetry, Holzforschung 63 (2009) 94-98. [297] M. Plötze, P. Niemz, Porosity and pore size distribu#on of different wood types as determined by mercury intrusion porosimetry, European Journal of Wood and Wood Products 69 (2011) 649-657. [298] C.E. Brewer, V.J. Chuang, C.A. Masiello, H. Gonnermann, X. Gao, B. Dugan, L.E. Driver, P. Panzacchi, K. Zygourakis, C.A. Davies, New approaches to measuring biochar density and porosity, Biomass and Bioenergy 66 (2014) 176-185. [299] S. Lowell, J.E. Shields, M.A. Thomas, Characteriza#on of porous solids and powders surface area, pore size and density, Springer, Dordrecht, 2011. [300] J. Jagiello, C. Ania, J.B. Parra, C. Cook, Dual gas analysis of microporous carbons using 2D-NLDFT heterogeneous surface model and combined adsorp#on data of N2 and CO2, Carbon 91 (2015) 330-337. [301] P.I. Ravikovitch, A. Vishnyakov, R. Russo, A.V. Neimark, Unified Approach to Pore Size Characteriza#on of Microporous Carbonaceous Materials from N2, Ar, and CO2 Adsorp#on Isotherms, Langmuir 16 (2000) 2311-2320. [302] H.J. Bachmann, T.D. Bucheli, A. Dieguez-Alonso, D. Fabbri, H. Knicker, H.-P. Schmidt, A. Ulbricht, R. Becker, A. Buscaroli, D. Buerge, A. Cross, D. Dickinson, A. Enders, V.I. Esteves, M.W.H. Evangelou, G. Fellet, K. Friedrich, G. Gasco Guerrero, B. Glaser, U.M. Hanke, K. Hanley, I. Hilber, D. Kalderis, J. Leifeld, O. Masek, J. Mumme, M.P. Carmona, R. Calvelo Pereira, F. Rees, A.G. Rombolà, J.M. de la Rosa, R. Sakrabani, S. Sohi, G. Soja, M. Valagussa, F. Verheijen, F. Zehetner, Toward the Standardiza#on of Biochar Analysis: The COST Ac#on TD1107 Interlaboratory Comparison, Journal of Agricultural and Food Chemistry 64 (2016) 513-527. [303] G. Sigmund, T. Hüffer, T. Hofmann, M. Kah, Biochar total surface area and total pore volume determined by N2 and CO2 physisorp#on are strongly influenced by degassing temperature, Science of The Total Environment 580 (2017) 770-775. [304] A. Mukherjee, A.R. Zimmerman, W. Harris, Surface chemistry varia#ons among a series of laboratory-produced biochars, Geoderma 163 (2011) 247-255. [305] V. Gargiulo, A. Gomis-Berenguer, P. Giudicianni, C.O. Ania, R. Ragucci, M. Alfè, Assessing the Poten#al of Biochars Prepared by Steam-Assisted Slow Pyrolysis for CO2 Adsorp#on and Separa#on, Energy & Fuels 32 (2018) 10218-10227. [306] J. Garrido, A. Linares-Solano, J.M. Mar#n-Mar#nez, M. Molina-Sabio, F. Rodriguez-Reinoso, R. Torregrosa, Use of nitrogen vs. carbon dioxide in the characteriza#on of ac#vated carbons, Langmuir 3 (1987) 76-81. [307] D. Chen, K. Cen, X. Zhuang, Z. Gan, J. Zhou, Y. Zhang, H. Zhang, Insight into biomass pyrolysis mechanism based on cellulose, hemicellulose, and lignin: Evolu#on of vola#les and kine#cs, elucida#on of reac#on pathways, and characteriza#on of gas, biochar and bio-oil, Combus#on and Flame 242 (2022) 112142. [308] A.C. Lua, J. Guo, Prepara#on and characteriza#on of chars from oil palm waste, Carbon 36 (1998) 1663-1670. [309] G. Yang, Z. Wang, Q. Xian, F. Shen, C. Sun, Y. Zhang, J. Wu, Effects of pyrolysis temperature on the physicochemical proper#es of biochar derived from vermicompost and its poten#al use as an environmental amendment, RSC Advances 5 (2015) 40117-40125. [310] S. Yoo, S.S. Kelley, D.C. Tilo9a, S. Park, Structural Characteriza#on of Loblolly Pine Derived Biochar by X-ray Diffrac#on and Electron Energy Loss Spectroscopy, ACS Sustainable Chemistry & Engineering 6 (2018) 2621-2629. CHAPTER 1 74 [311] B.C. Chaves Fernandes, K. Ferreira Mendes, A.F. Dias Júnior, V.P. da Silva Caldeira, T.M. da Silva Teófilo, T. Severo Silva, V. Mendonça, M. de Freitas Souza, D. Valadão Silva, Impact of Pyrolysis Temperature on the Proper#es of Eucalyptus WoodDerived Biochar, Materials 13 (2020) 5841. [312] D. He, Y. Luo, B. Zhu, Feedstock and pyrolysis temperature influence biochar proper#es and its interac#ons with soil substances: Insights from a DFT calcula#on, Science of The Total Environment 922 (2024) 171259. [313] M. Schäfer, Computa#onal Engineering - Introduc#on to Numerical Methods, 2nd Ed. ed., Springer-Verlag, Cham, 2022. [314] P.J. Roache, Fundamentals of Verifica#on and Valida#on, Hermosa Publishers, Socorro, 2009. [315] G. Dubois, Modeling and simula#on: challenges and best prac#ces for industry, CRC Press, Boca Raton, 2018. [316] O. Balci, Guidelines for successful simula#on studies, in: O. Balci (Ed.) 1990 Winter Simula#on Conference Proceedings, IEEE, New Orleans, 1990, pp. 25-32. [317] American Ins#tute of Aeronau#cs and Astronau#cs (AIAA), Guide for the verifica#on and valida#on of computa#onal fluid dynamics simula#ons, American Ins#tute of Aeronau#cs and Astronau#cs, Reston, 1998. [318] M.C. Kennedy, A. O'Hagan, Bayesian calibra#on of computer models, Journal of the Royal Sta#s#cal Society: Series B (Sta#s#cal Methodology) 63 (2001) 425-464. [319] F. Shafizadeh, Pyrolysis and Combus#on of Cellulosic Materials, in: M.L. Wolfrom, R.S. Tipson (Eds.) Advances in Carbohydrate Chemistry, Academic Press1968, pp. 419-474. [320] F. Shafizadeh, P.P.S. Chin, Thermal Deteriora#on of Wood, in: I.S. Goldstein (Ed.) Wood Technology: Chemical Aspects, American Chemical Society, Washington D.C., USA, 1977, pp. 57-81. [321] F. Thurner, U. Mann, Kine#c inves#ga#on of wood pyrolysis, Industrial & Engineering Chemistry Process Design and Development 20 (1981) 482-488. [322] J. Lédé, Cellulose pyrolysis kine#cs: An historical review on the existence and role of intermediate ac#ve cellulose, Journal of Analy#cal and Applied Pyrolysis 94 (2012) 17-32. [323] A.E.S. Green, M. Zanardi, Cellulose pyrolysis and quantum chemistry, Interna#onal Journal of Quantum Chemistry 66 (1998) 219-227. [324] B. Hu, B. Zhang, W.-l. Xie, X.-y. Jiang, J. Liu, Q. Lu, Recent Progress in Quantum Chemistry Modeling on the Pyrolysis Mechanisms of Lignocellulosic Biomass, Energy & Fuels 34 (2020) 10384-10440. [325] M. Jalili, A. Anca-Couce, N. Zobel, On the Uncertainty of a Mathema#cal Model for Drying of a Wood Par#cle, Energy & Fuels 27 (2013) 6705-6717. [326] P.N. Ciesielski, G.M. Wiggins, J.E. Jakes, C.S. Daw, Simula#ng Biomass Fast Pyrolysis at the Single Par#cle Scale, in: R.C. Brown, K. Wang (Eds.) Fast Pyrolysis of Biomass: Advances in Science and Technology, The Royal Society of Chemistry2017, pp. 0. [327] S.R.A. Kersten, X. Wang, W. Prins, W.P.M. van Swaaij, Biomass Pyrolysis in a Fluidized Bed Reactor. Part 1: Literature Review and Model Simula#ons, Industrial & Engineering Chemistry Research 44 (2005) 8773-8785. [328] H.K. Versteeg, W. Malalasekera, An introduc#on to computa#onal fluid dynamics: The finite volume method, 2nd Ed. ed., Pearson/Pren#ce Hall, Harlow, 2010. [329] L.D. Landau, E.M. Lifshit—z ˜, Fluid Mechanics: Course of Theore#cal Physics, Pergamon, Amsterdam, 2013, pp. 1-539. [330] G.K. Batchelor, An introduc#on to fluid dynamics, 14th Ed. ed., Cambridge University Press, Cambridge, 2013. [331] O.C. Zienkiewicz, R.L. Taylor, J.Z. Zhu, The finite element method: Its basis and fundamentals, Bu9erworth-Heinemann, Oxford, 2013, pp. 1-714. [332] C. Thornton, Granular dynamics, contact mechanics and par#cle system simula#ons: A DEM study, Springer, Cham, 2015, pp. 1-195. [333] N. Hamid Reza, Coupled CFD-DEM Modeling: Formula#on, Implementa#on and Applica#on to Mul#phase Flows, Wiley, Newark, 2016, pp. 1-433. [334] M. Sakai, How Should the Discrete Element Method Be Applied in Industrial Systems?: A Review, KONA Powder and Par#cle Journal 33 (2016) 169-178. [335] C. Kloss, C. Goniva, G. Aichinger, S. Pirker, Comprehensive DEM-DPM-CFD Simula#ons-Model Synthesis, Experimental Valida#on and Scalability, 7th Interna#onal Conference on CFD in the Minerals and Process IndustriesMelbourne, Australia, 2009, pp. 1-7. [336] A.D. Paulsen, M.S. Me9ler, P.J. Dauenhauer, The Role of Sample Dimension and Temperature in Cellulose Pyrolysis, Energy & Fuels 27 (2013) 2126-2134. [337] D.L. Pyle, C.A. Zaror, Heat transfer and kine#cs in the low temperature pyrolysis of solids, Chemical Engineering Science 39 (1984) 147-158. [338] G. SriBala, H.-H. Carstensen, K.M. Van Geem, G.B. Marin, Measuring biomass fast pyrolysis kine#cs: State of the art, WIREs Energy and Environment 8 (2019) e326. [339] R. Mehrabian, S. Zahirovic, R. Scharler, I. Obernberger, S. Kleditzsch, S. Wirtz, V. Scherer, H. Lu, L.L. Baxter, A CFD model for thermal conversion of thermally thick biomass par#cles, Fuel Processing Technology 95 (2012) 96-108. [340] J. Wiese, F. Wissing, D. Höhner, S. Wirtz, V. Scherer, U. Ley, H.M. Behr, DEM/CFD modeling of the fuel conversion in a pellet stove, Fuel Processing Technology 152 (2016) 223-239. [341] X. Gao, J. Yu, L. Lu, W.A. Rogers, Coupling par#cle scale model and SuperDEM-CFD for mul#scale simula#on of biomass pyrolysis in a packed bed pyrolyzer, AIChE Journal 67 (2021) e17139. [342] B. Jaeger, E. Illana Mahiques, S. Wirtz, V. Scherer, Pyrolysis of Spherical Wood Par#cles in a Packed Bed – Comparison between Resolved and Unresolved Discrete Element Method/Computa#onal Fluid Dynamics, Chemical Engineering & Technology 46 (2023) 2120-2132. CHAPTER 1 75 [343] A. Funke, R. Grandl, M. Ernst, N. Dahmen, Modelling and improvement of heat transfer coefficient in auger type reactors for fast pyrolysis applica#on, Chemical Engineering and Processing - Process Intensifica#on 130 (2018) 67-75. [344] J. Wang, X. Ku, S. Yang, Simula#on of Biomass Pyrolysis in a Rotary Drum by Coupling CFD-DEM with a One-Dimensional Thermally Thick Model, Energy & Fuels 36 (2022) 3665-3679. [345] R. Houston, O. Oyedeji, N. Abdoulmoumine, Detailed biomass fast pyrolysis kine#cs integrated to computa#onal fluid dynamic (CFD) and discrete element modeling framework: Predic#ng product yields at the bench-scale, Chemical Engineering Journal 444 (2022) 136419. [346] X. Gao, L. Lu, M. Shahnam, W.A. Rogers, K. Smith, K. Gaston, D. Robichaud, M. Brennan Pecha, M. Crowley, P.N. Ciesielski, P. Debiagi, T. Faravelli, G. Wiggins, C.E.A. Finney, J.E. Parks, Assessment of a detailed biomass pyrolysis kine#c scheme in mul#scale simula#ons of a single-par#cle pyrolyzer and a pilot-scale entrained flow pyrolyzer, Chemical Engineering Journal 418 (2021) 129347. [347] O.A. Oyedeji, M. Brennan Pecha, C.E.A. Finney, C.A. Peterson, R.G. Smith, Z.G. Mills, X. Gao, M. Shahnam, W.A. Rogers, P.N. Ciesielski, R.C. Brown, J.E. Parks Ii, CFD–DEM modeling of autothermal pyrolysis of corn stover with a coupled par#cleand reactor-scale framework, Chemical Engineering Journal 446 (2022) 136920. [348] B. Wang, J. Dai, S. Li, Y. Lin, M. Patrascu, X. Gao, Coupling coarse-grained DEM-CFD and intrapar#cle model for biomass fast pyrolysis simula#on and experiment valida#on, AIChE Journal 70 (2024) e18393. [349] S. Liu, A. Ahmadi-Senichault, A. Ben-Abdelwahed, H. Yao, J. Lachaud, Experimental inves#ga#on and DEM-CFD analysis of Darcy–Forchheimer flows in randomly packed bed systems of wood par#cles, Interna#onal Journal of Heat and Mass Transfer 235 (2024) 126229. [350] D.D. A9anayake, F. Sewerin, S. Kulkarni, A. Dernbecher, A. Dieguez-Alonso, B. van Wachem, Review of Modelling of Pyrolysis Processes with CFD-DEM, Flow, Turbulence and Combus#on 111 (2023) 355-408. [351] M.B. Pecha, J.I.M. Arbelaez, M. Garcia-Perez, F. Chejne, P.N. Ciesielski, Progress in understanding the four dominant intra-par#cle phenomena of lignocellulose pyrolysis: chemical reac#ons, heat transfer, mass transfer, and phase change, Green Chemistry 21 (2019) 2868-2898. [352] M.B. Pecha, E. Ramirez, G.M. Wiggins, D. Carpenter, B. Kappes, S. Daw, P.N. Ciesielski, Integrated Par#cleand ReactorScale Simula#on of Pine Pyrolysis in a Fluidized Bed, Energy & Fuels 32 (2018) 10683-10694. [353] P.N. Ciesielski, M.B. Pecha, A.M. La9anzi, V.S. Bharadwaj, M.F. Crowley, L. Bu, J.V. Vermaas, K.X. Steirer, M.F. Crowley, Advances in Mul#scale Modeling of Lignocellulosic Biomass, ACS Sustainable Chemistry & Engineering 8 (2020) 3512-3531. [354] H. Luo, X. Wang, X. Liu, X. Wu, X. Shi, Q. Xiong, A review on CFD simula#on of biomass pyrolysis in fluidized bed reactors with emphasis on par#cle-scale models, Journal of Analy#cal and Applied Pyrolysis 162 (2022) 105433. [355] H. Ström, H. Luo, Q. Xiong, Perspec#ves on Par#cle–Fluid Coupling at Varying Resolu#on in CFD-DEM Simula#ons of Thermochemical Biomass Conversion, Energy & Fuels 38 (2024) 17179-17190. [356] S. Golshan, R. Sotudeh-Gharebagh, R. Zarghami, N. Mostoufi, B. Blais, J.A.M. Kuipers, Review and implementa#on of CFD-DEM applied to chemical process systems, Chemical Engineering Science 221 (2020) 115646. [357] L. Lu, X. Gao, M. Shahnam, W.A. Rogers, Simula#ons of biomass pyrolysis using glued-sphere CFD-DEM with 3-D intrapar#cle models, Chemical Engineering Journal 419 (2021) 129564. [358] W.C. Park, A. Atreya, H.R. Baum, Experimental and theore#cal inves#ga#on of heat and mass transfer processes during wood pyrolysis, Combus#on and Flame 157 (2010) 481-494. [359] H. Bennadji, K. Smith, S. Shabangu, E.M. Fisher, Low-Temperature Pyrolysis of Woody Biomass in the Thermally Thick Regime, Energy & Fuels 27 (2013) 1453-1459. [360] P.N. Ciesielski, M.F. Crowley, M.R. Nimlos, A.W. Sanders, G.M. Wiggins, D. Robichaud, B.S. Donohoe, T.D. Foust, Biomass Par#cle Models with Realis#c Morphology and Resolved Microstructure for Simula#ons of Intrapar#cle Transport Phenomena, Energy & Fuels 29 (2015) 242-254. [361] A. Anca-Couce, P. Sommersacher, R. Scharler, Online experiments and modelling with a detailed reac#on scheme of single par#cle biomass pyrolysis, Journal of Analy#cal and Applied Pyrolysis 127 (2017) 411-425. [362] A. Atreya, P. Olszewski, Y. Chen, H.R. Baum, The effect of size, shape and pyrolysis condi#ons on the thermal decomposi#on of wood par#cles and firebrands, Interna#onal Journal of Heat and Mass Transfer 107 (2017) 319-328. [363] I. Haberle, N.E.L. Haugen, Ø. Skreiberg, Drying of Thermally Thick Wood Par#cles: A Study of the Numerical Efficiency, Accuracy, and Stability of Common Drying Models, Energy & Fuels 31 (2017) 13743-13760. [364] G. Gen#le, P.E.A. Debiagi, A. Cuoci, A. Frassolda#, E. Ranzi, T. Faravelli, A computa#onal framework for the pyrolysis of anisotropic biomass par#cles, Chemical Engineering Journal 321 (2017) 458-473. [365] F. Li, K. Wu, K. Yang, Z. Ge, J. Feng, H. Zhang, A comprehensive pyrolysis model for lignocellulosic biomass par#cles with a special emphasis on the anisotropic characteris#cs, Fuel 341 (2023) 127635. [366] M. Bellais, Modelling of the pyrolysis of large wood par#cles (Ph.D. Thesis), Department of Chemical Engineering and Technology, KTH - Royal Ins#tute of Technology, Stockholm, Sweden, 2007, pp. 97. [367] K.O. Davidsson, J.B.C. Pe9ersson, Birch wood par#cle shrinkage during rapid pyrolysis, Fuel 81 (2002) 263-270. [368] M. Bellais, K.O. Davidsson, T. Liliedahl, K. Sjöström, J.B.C. Pe9ersson, Pyrolysis of large wood par#cles: a study of shrinkage importance in simula#ons☆, Fuel 82 (2003) 1541-1548. [369] M.F. Crowley, H. Sitaraman, J. Klinger, F. Usseglio-Vire9a, N.E. Thornburg, N. Brunhart-Lupo, M.B. Pecha, J.H. Dooley, Y. Xia, P.N. Ciesielski, Measurement of Transport Proper#es of Woody Biomass Feedstock Par#cles Before and AMer Pyrolysis by Numerical Analysis of X-Ray Tomographic Reconstruc#ons, Fron#ers in Energy Research 10 (2022). [370] V. Haag, K. Dremel, S. Zabler, Volumetric imaging by micro computed tomography: a suitable tool for wood iden#fica#on of charcoal, IAWA Journal 44 (2022) 210-224. CHAPTER 1 76 [371] G. Gauthier, T. Melkior, M. Grateau, S. Thiery, S. Salvador, Pyrolysis of cen#metre-scale wood par#cles: New experimental developments and results, Journal of Analy#cal and Applied Pyrolysis 104 (2013) 521-530. [372] H. Bennadji, K. Smith, M.J. Serapiglia, E.M. Fisher, Effect of Par#cle Size on Low-Temperature Pyrolysis of Woody Biomass, Energy & Fuels 28 (2014) 7527-7537. [373] H. Lu, W. Robert, G. Peirce, B. Ripa, L.L. Baxter, Comprehensive Study of Biomass Par#cle Combus#on, Energy & Fuels 22 (2008) 2826-2839. [374] H. Lu, E. Ip, J. Sco9, P. Foster, M. Vickers, L.L. Baxter, Effects of par#cle shape and size on devola#liza#on of biomass par#cle, Fuel 89 (2010) 1156-1168. [375] P. Maziarka, F. Ronsse, A. Anca-Couce, Review on Modelling Approaches Based on Computa#onal Fluid Dynamics for Biomass Pyrolysis Systems, in: Z. Fang, R.L. Smith Jr, L. Xu (Eds.) Produc#on of Biofuels and Chemicals with Pyrolysis, Springer Singapore, Singapore, 2020, pp. 373-438. [376] P. Maziarka, C. Wurzer, P.J. Arauzo, A. Dieguez-Alonso, O. Mašek, F. Ronsse, Do you BET on rou#ne? The reliability of N2 physisorp#on for the quan#ta#ve assessment of biochar’s surface area, Chemical Engineering Journal 418 (2021) 129234. [377] P. Maziarka, P. Sommersacher, X. Wang, N. Kienzl, S. Retschitzegger, W. Prins, N. Hedin, F. Ronsse, Tailoring of the pore structures of wood pyrolysis chars for poten#al use in energy storage applica#ons, Applied Energy 286 (2021) 116431. [378] P. Maziarka, N. Kienzl, A. Dieguez-Alonso, V. Fierro, A. Celzard, P.J. Arauzo, N. Hedin, W. Prins, A. Anca-Couce, J.J. Manyà, F. Ronsse, Part 1─Impact of Pyrolysis Temperature and Wood Par#cle Length on Vapor Cracking and Char Porous Texture in Rela#on to the Tailoring of Char Proper#es, Energy & Fuels 38 (2024) 9751-9771. [379] P. Maziarka, A. Anca-Couce, W. Prins, F. Ronsse, A meta-analysis of thermo-physical and chemical aspects in CFD modelling of pyrolysis of a single wood par#cle in the thermally thick regime, Chemical Engineering Journal 446 (2022) 137088. [380] P. Maziarka, N. Kienzl, A. Dieguez-Alonso, W. Prins, P.J. Arauzo, Ø. Skreiberg, A. Anca-Couce, J.J. Manyà, F. Ronsse, Part 2─Tailoring of Pyroly#c Char Proper#es with a Single Par#cle CFD Model with a Focus on the Impact of Shrinking, Vapor Cracking, and Char Permeability, Energy & Fuels 38 (2024) 9772-9793. CHAPTER 2 Structure-dependent proper#es of wood and wood-derived chars – a cri#cal overview of experimental studies Chapter abstract Chapter 2 provides a cri#cal data review focused on structure-dependent proper#es of wood and wood-derived chars relevant to CFD modeling of the thermochemical conversion of wood to bio-based chars. The goal is to consolidate reliable experimental data and iden#fy knowledge gaps to improve overall understanding in the field. Efforts are also made to highlight bo9lenecks in the models for wood pyrolysis and char forma#on, which are relevant at both par#cle and reactor scales. Proper#es covered in Chapter 2, focus majorly on bio-composi#on, moisture, densi#es, porosity, permeability, and thermal conduc#vity. A mul#scale structural framework of wood is described in detail – from macroscopic to molecular levels – to analyze how thermal degrada#on affects proper#es at each scale and to connect structure with performance. Furthermore, geometric changes and cell wall thinning in response to heat are discussed in rela#on to shrinking and mechanical stress forma#on. A9en#on is paid to the evolu#on of the molecular structure of bio-based chars in two temperature ranges: 300–500 °C and beyond 500 °C (including descrip#on of aroma#za#on and polycondensa#on). Experimental data from microscopy, porosimetry, pycnometry, XRD, as well as molecular simula#ons of transi#ons at the nanoscale, are integrated to characterize dynamic changes in char structure. Thermal conduc#vity is highlighted as a cri#cal yet under-characterized parameter. It exhibits a clear anisotropic trend (structure-dependency) and undergoes a non-linear evolu#on during pyrolysis (composi#on-dependency), the la9er driven by carbon matrix reorganiza#on above 600 °C. Chapter 2 states that reliable quan#ta#ve data remains sparse and inconsistent, represen#ng a key limita#on to CFD model development. Although no universal correla#ons are proposed in Chapter 2, it establishes the founda#on for building a structure-dependent modeling framework of par#cle-scale conversion. The review will con#nue in Chapter 5, where quan#fica#ons of these aforemen#oned correla#ons are proposed. Chapter re-draNed from: P. Maziarka, F. Ronsse and A. Anca-Couce, “Chapter 13 – Review on modelling approaches based on computa!onal fluid dynamics for biomass pyrolysis systems”, in: Z. Fang, R. L. Smith, L. Xu (Eds.), “Produc!on of Biofuels and Chemicals with Pyrolysis”, Springer Singapore (2020) CHAPTER 2 78 1. Introduc,on to Chapter 2 As pointed out in Chapter 1 (Sec#on 3.1.), lignocellulosic biomass of forest origin has the highest pool, which can be used for carbon sequestra#on (Carbon Direct Reduc#on – CDR). Addi#onally, due to the low ash, sulfur, and nitrogen content, wood is the most suitable feedstock for thermochemical conversion to biofuels or carbon materials. From a modeling perspec#ve, the availability of data related to wood and its deriva#ves is the largest among lignocellulosic biomasses. Therefore, wood is also an outstanding candidate feedstock from which a reliable predic#on model could be derived. On the other hand, experimental data related to wood and the products of its thermochemical conversion (explicitly thermo-physical proper#es) are usually sca9ered among various sources. This hinders the consolida#on of available knowledge into a uniform mechanism reflec#ng changes in the proper#es of wood and its deriva#ves during thermochemical conversion. A comprehensive overview of the proper#es of wood and derived products and their dynamic change throughout the thermochemical conversion process should enable loca#ng missing links for the construc#on of a uniform model (data and mechanisms). Consequently, an overview could highlight a pathway to mi#gate the knowledge gaps and the missing experimental data, enabling the construc#on of a comprehensive and uniform model predic#ng changes in wood and its products during conversion. Several comprehensive books on wood processing technology toward market products (construc#on #mber, paper) can be found in the literature.[1-12] Experimental data related to physical and thermochemical proper#es of wood in rela#on to its typical applica#ons can be found in those. Among available data, composi#on-related parameters are burdened with the least uncertainty (true density, specific heat). Contrarily, parameters related to the wood structure (pore size distribu#on and permeability), or to both (thermal conduc#vity, bulk density), are burdened with high uncertainty. The origin of such a situa#on is related to wood structure varia#on between species, between specimens, and even within a specific tree part. Furthermore, due to the natural characteris#cs of wood, structuredependent parameters show anisotropic proper#es, which add to their uncertainty. The known data of wood-derived char parameters is significantly smaller than that of wood. Moreover, the proper#es of chars are related not only to the ini#al wood proper#es, but also to their thermal processing condi#ons, where the la9er impacts both, composi#on and structure in a case-specific way. Therefore, the uncertainty of char parameters is high, especially as the number of variables that impact char proper#es is broad and current data is limited. The changes in the wood/char composi#on and structure are con#nuous during conversion (dynamic), while in most cases, the available datasets reflect the processes in a discrete manner within a small range of inves#gated process parameters. Furthermore, the most frequently reported parameters are those that are easy to measure (composi#on-based), while those that are difficult to measure or require specific equipment are very rarely provided (structure-based). Due to the men#oned issues, deriving the descrip#on of changes in wood/char proper#es during thermochemical conversion in a uniform manner seems currently out of reach. This chapter comprehensively summarizes the chemical and thermo-physical data available in the literature, which are most relevant in terms of wood pyrolysis on the single par#cle scale. Considering uncertain#es, the focus will be on the structure-dependent parameters, as those are the least detailed in the literature but have a cri#cal impact on changes in the proper#es of wood/char during thermal degrada#on. That, in detail, relates to the bio-components composi#on of the wood cell wall and proper#es of wood and char, including true and bulk density, permeability, and thermal conduc#vity. The detailed mathema#cal formula#on and correla#ons of the men#oned parameters and other relevant rela#ons will be omi9ed in this Chapter as they will be provided in Chapter 5. CHAPTER 2 79 2. Mul,scale of wood structure 2.1. Tree anatomy and hierarchical wood structure In most generic terms, wood is a porous, bio-polymer composite material with a complex hierarchical structure. The structure of wood is inherited from the remains of a once-living tree #ssue, where the anatomy of the original plant organism is manifested at every length scale. Inter-species differences, such as branching pa9erns in trees or stem thicknesses, are visually apparent at the macroscale. At the microscale, the dominant structural feature of biomass is imparted by the cellular arrangement of the #ssue.[13-16] A division of the wood structure in rela#on to its hierarchical characteris#cs is presented in Fig. 2.1. The structure of wood has been a subject of detailed studies for over a century, with correspondingly robust literature (recent and relevant references).[1-12] To avoid diving into extensive details, in this Chapter, only key elements of the wood structure will be highlighted, with an emphasis on their consecu#ve impact on thermochemical conversion. However, wood represents a material with a hierarchical structure, so informa#on on each scale needs to be included as any comprehensive model of thermochemical conversion requires an inherently mul#scale approach. Fig. 2.1. Hierarchal structure of wood: a – coniferous tree, b – trunk sec#on from a pine tree, c – scanning electron microscopy (SEM) image of the yellow pine #ssue, d – transmission electron microscopy (TEM) image of the ultrastructure wood cell wall, e – schema#c of the molecular arrangement of lignocellulose, f – schema#c depic#on of an amorphous lignin polymer and a cellulose elementary fibril decorated with hemicellulose.[16] A growing tree can be divided into two major domains: the shoot and the roots. Roots are the underground structures responsible for water and mineral nutrient uptake, mechanical anchoring, and #ssue for storage of metabolized biochemicals. The shoot is the aboveground part of a tree, made up of the trunk, branches, and leaves. The trunk comprises various plant structures organized in concentric bands (cf. Fig. 2.1. e). CHAPTER 2 80 From the outside of the trunk to the inside are located the outer bark, the inner bark, vascular cambium, sapwood, heartwood, and the pith. The role of the outer bark is to provide mechanical protec#on to the soMer inner bark and to limit the evapora#ve water loss. The inner bark is the #ssue through which photosynthesis products (simply, monosaccharides) are transported from the leaves to the roots and the loca#ons where the tree grows. The further inner part is called the vascular cambium, which is a layer between the inner bark and sapwood. The sapwood is the younger part of a trunk, through which the transport of water (in the form of sap) from the roots to the leaves takes place. The older and non-conduc#ve part of the trunk is heartwood, which oMen has a dis#nc#ve darker color due to the accumula#on of tree-specific biochemical metabolites. The pith at the center of the trunk is the remnant of the growth of the trunk, before wood was formed, and does not have a dis#nc#ve role in tree growth. The vascular cambium is a thin layer of cells that produces, by means of cell divisions, wood (xylem) toward the inside (sapwood) and bark (phloem) to the outside (inner bark). Therefore, the vascular cambium is a part of a tree that adds cell layers around the trunk and increases its diameter. The newly formed tree cells produced together over a discrete #me interval are known as growth increments or, more commonly, growth rings. It is worth highligh#ng that in areas where dis#nct, regular seasonality exists, the annual growth increments of the trees are more dis#nguishable and visible as annual rings. However, in areas where the seasonal change of weather is less pronounced, the annual growth increments are less visible but dis#nguishable under a microscope. Lower in hierarchy, wood is composed of discrete cells interconnected and arranged into an integrated and con#nuous system from root to twig. The wood cells typically have a greater length than width and are organized into two separate systems of cells, depending on direc#on, namely the axial system and the radial system. The general representa#on of both systems is shown in Fig. 2.2. Fig. 2.2. Types of cell systems and specific cell structures presented in hardwoods and soMwoods (Source: Encyclopedia Britannica). CHAPTER 2 87 Despite its maturity, the data from Sjöström shows a highly similar range of the content of biocomponents in wood, so it s#ll represents a reliable source of data. As is shown by data presented in Table 2.2., the total bio-composi#on oMen does not match exactly 100 wt.%. However, the difference usually does not exceed 5 wt.% despite the presented data being derived using different assessment methods. The experimental data clearly indicate that the difference in each bio-component content can vary by a few wt.% even for the same species. Moreover, the values can be slightly different for the same tree, e.g., for sapwood and heartwood. From the other end, for a predic#ve thermochemical model, even a few wt.% differences in bio-composi#on have the ability to significantly affect the conversion and product yields.[41] The available data allow the average content of each biocomponent content to generally match the modeled case, but such content seems par#ally useful for the valida#on of models of the thermochemical conversion of wood, especially at scales lower than the reactor scale. Therefore, it is of high relevance to reliably assess the bio-composi#on of modeled wood so as not to introduce elevated uncertainty and obtain reliable outcomes from the simula#on of its thermal degrada#on. 3. Impact of thermal degrada,on on the structure of wood 3.1. Geometric deforma,on The reduc#on of the wood par#cle volume during thermal degrada#on is a well-established phenomenon, commonly known as par#cle shrinking. The studies also unambiguously indicate that the reduc#on in the size of a par#cle is not equal for every dimension. The representa#on of direc#ons for a cylindrical wood par#cle is shown in Fig. 2.8. The general tendency is that the most significant reduc#on occurs for the tangen#al (T) direc#on and then for the radial (R), while the lowest reduc#on is for the longitudinal (L) direc#on.[42-47] Fig. 2.8. Representa#on of the direc#ons of a cylindrical par#cle with indicated annual growth rings. The exemplary dimensional and volumetric change of a birch wood par#cle during thermal degrada#on between 350 °C and 900 °C from the study by Davidsson and Pe9ersson is presented in Fig. 2.9. As their data indicates, the size in the T-direc#on may reduce, by up to 45%, in the R-direc#on up to 35% and in the L-direc#on up to 20%. Therefore, the corresponding max. volume reduc#on of a par#cle can reach up to 60–65 vol.%.[42] The dimensional change originates from the thermal degrada#on of the bio-components that cons#tute the wood cell wall. As indicated in Chapter 1 (Sec#on 4.1.), the majority of hemicellulose decomposes up to 250 °C and cellulose up to 350 °C. The lignin, which decomposes through a soMened state (semi-liquid state), decomposes over a broader range of temperatures from 250–300 °C to 450–500 °C. Above the la9er temperature range, the further dimensional change of a par#cle is related to the thermal re-organiza#on of the carbonaceous ma9er. CHAPTER 2 88 As results from Bryne and Nagle indicate, above 1000 °C, no meaningful change in the dimensions of a wood par#cle takes place.[44] Aside from the wood cells arrangement, the alignment of the biocomponents in the call wall layers is suspected as an addi#onal parameter, which causes the devia#on in the dimensional changes between direc#ons.[42] Addi#onally, as indicated by Cu9er et al., the dimensional change is related to the hea#ng rate during the conversion, which was later confirmed in the study by Pa9anotai et al.[46, 48] Interes#ngly, the result from Davidsson and Pe9ersson indicates that above 85 wt.% mass loss, the dimension in the radial and tangen#al direc#on increases. This behavior can be translated as an increase in the par#cle volume (so-called “swelling).[42] The appearance of an increase in par#cle volume is inconsistent with other studies for the same range of mass loss/temperature.[43, 45, 46] Fig. 2.9. Dimensional and volumetric changes of a birch wood par#cle in rela#on to the mass loss (top, leM – longitudinal direc#on; top, right – radial direc#on; bo9om, leM – transverse direc#on; bo9om, right – volume).[42] 3.2. Structural changes of cell wall The effect of cell wall thinning during the thermal degrada#on of a wood structure is also well recognized in the literature.[46, 49] Such a phenomenon is presented in Fig. 2.10 and Fig. 2.11. The former figure was composed of video frames from a thin slice of poplar in thermal degrada#on between 25 °C and 500 °C, conducted with the microscope hot stage.[50] As can be observed from the figure, degrada#on of structural polysaccharides (hemicellulose up to 250 °C and cellulose up to 350 °C) leads to a reduc#on of the cell wall thickness, which translates to an increase in the lumen diameter, evident for the fiber structures. The rise of the temperature to 425 °C and then to 500 °C, which corresponds to degrada#on of lignin (par#ally and completely, respec#vely), induces further cell wall thinning, now visible as the vessel diameter increases. 0,8 0,9 1,0 0,6 0,7 0,8 0,9 1,0 L-dimension retention [-] 900 °C 350 °C 0,6 0,7 0,8 0,9 1,0 0,6 0,7 0,8 0,9 1,0 R-dimension retention [-] 350 °C 900 °C 0,6 0,7 0,8 0,9 1,0 0,6 0,7 0,8 0,9 1,0 T-dimension retention [-] mass loss [-] 350 °C 900 °C 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1,0 0,6 0,7 0,8 0,9 1,0 volume retention [ - ] mass loss [-] 350 °C 900 °C CHAPTER 2 89 Simultaneously with cell wall thinning, bio-components are converted into carbonaceous ma9er, which is less elas#c than the ini#al cell wall. It can be observed that structural changes induce mechanical tensions, which, if sufficiently high, may lead to the genera#on of structural discon#nui#es (“cracks”).[51] Such an outcome can be observed in the last three frames presented in Fig. 2.10. From the complete video, it can be addi#onally observed that thermal stretching and shrinking of the structure take place, which can be assessed as elevated in the conducted measurement due to a lack of mechanical boundaries imposed by the small thinness of the poplar wood sample.[50] Fig. 2.10. Selected video frames of a thin slice of poplar wood thermal degrada#on between 25 °C and 500 °C, conducted with the microscope hot stage.[50] SEM images from thermal degrada#on between 25 °C and 550 °C of a spruce wood par#cle with a few mm thicknesses are presented in Fig. 2.11., indica#ng a more realis#c behavior of the cell wall compared to a thin slice.[52] The degrada#on of bio-components occurs there in the same temperature ranges, while the degrada#on outcome reveals addi#onal informa#on about the structural change mechanism. Up to 325 °C, the cell walls get thinner, while they keep the ini#al geometric shape (crosssec#on) of the lumina (Fig. 2.11., b, c). Consecu#ve temperature increase con#nues the thinning of the cell wall (Fig. 2.11. d, e, f), while the cross-sec#on of each single lumen changes its shape. That, in a cumula#ve manner, is responsible for the reduc#on of the dimension of a par#cle (shrinking). The change in the lumina cross-sec#on can be directly observed by comparing Fig. 2.11. c and 2.11. d. The figures below dis#nctly indicate differences in behavior between earlywood (thin-walled and widelumen cells) and latewood (thick-walled and narrow-lumen cells). As indicated by Rinta-Paavola et al., thermal degrada#on of both cell wall forma#ons occurs with different shrinking extent, e.g., for 400 °C, the latewood shrinking in radial/tangen#al direc#on was 18 %, while for earlywood, it was 27 %.[52] The heterogeneous cell size reduc#on is presumed to be the main cause of mechanical stresses on the border between two forma#ons (at the annual growth ring). The final impact of mechanical stresses can be observed as the discon#nuity of the material (cracking). The appearance of discon#nui#es can be no#ced by comparison of Fig. 2.11. e and 2.11. f, which corresponds to the degrada#on at 475 °C and 550 °C, respec#vely. As the study by Davidsson and Pe9ersson showed, the maximum dimensional reduc#on in radial and tangen#al direc#on of a wood par#cle appears around 500 °C. Therefore, it can be presumed that the annual growth rings are under the strongest mechanical tensions at such temperatures. Aside from the annual growth rings, in Fig. 2.11. f, the appearance of cracks is also visible along the radial line, which can be considered as a ray cell structure. Those structures are natural devia#ons from the homogeneous (axial) structure, so the appearance of structure discon#nui#es due to the accumula#on of stresses in those forma#ons is understandable. From a general perspec#ve, discon#nui#es in the wood structure occur in places where irregulari#es are present, as those forma#ons accumulate mechanical stresses. Furthermore, as such irregulari#es are of natural occurrence, the appearance of cracks within the wood structure can be assumed as an unavoidable phenomenon of wood thermal degrada#on. CHAPTER 2 90 Fig. 2.11. SEM images from the exact loca#on on a spruce wood par#cle aMer thermal degrada#on at different temperatures: a – 25 °C (without thermal treatment); b – 250 °C; c – 325 °C, d – 400 °C, e – 475 °C, and f – 550 °C (scale bar in all images – 200 µm).[52] The thinning of the cell wall is one of the factors that reduces the mechanical endurance of a cell wall, which enables the appearance of discon#nui#es in its structure. The other factor is related to chemical changes occurring within the cell wall components during thermal degrada#on (charring of biocomponents), which affects the mechanical proper#es of the cell wall. An extensive study on the subject was conducted by Zickler et al., who provided a rela#on between degrada#on temperature and microscopic mechanical proper#es of a spruce wood cell wall (reduced elas#c modulus, hardness, and indenta#on duc#lity index – IDI).[53] The IDI is especially relevant as it refers to the ability of a material to deform plas#cally before fracture, so it quan#fies the ability of a structure to create discon#nuity (cracking) upon mechanical stress. The value of IDI is within the range of 0 and 1, where a bri9le material shows a value below 0.3 (materials that are prone to crack under mechanical stress), while duc#le materials show a value above 0.5 (materials that transform mechanical stress into plas#c deforma#on without cracking).[54-56] CHAPTER 2 91 Results of Zickler et al. show that the IDI of the charred wood cell wall drops from 0.8 to 0.1 between 200 °C and 450 °C and stays at that value with further increase in temperature. Therefore, the thermal degrada#on of bio-components composing the wood cell wall changes its proper#es from duc#le to very bri9le at 450 °C. Above 550 °C, the IDI index keeps the value of ca. 0.1, while the elas#c modulus and hardness of materials increase exponen#ally to 700 °C.[53] Therefore, up to 550 °C, the cell wall is bri9le and prone to fracturing under low mechanical stress. Then, the material toughness increases, so it gains the ability to endure higher mechanical stress loads. While it is s#ll bri9le, it can frac#onate under strong, impact stress factors. The change in the microscopic mechanical proper#es of the woodderived char cell wall translates further to the macroscopic proper#es of the wood-derived char structure.[57] A study by Kumar et al. presents the rela#on of crushing strength and impact strength index of eucalyptus and acacia wood chars produced up to 1200 °C. The study shows that chars from both woods encounter a drop in endurance to mechanical stress up to 500–600 °C (minimum value). Then, the mechanical endurance of wood-derived chars increases along with the temperature.[58] 3.3. Structural changes of lumen Lumina in the wood structure are primarily related to the porosity of a par#cle, with a pore size distribu#on characteris#c for a specific wood. Pore size distribu#on corresponding to the lumina size can be assessed experimentally with mercury porosimetry (intrusion) and by (micro-) X-ray computed tomography (-CT or XCT).[59] The la9er method recently gained a lot of interest as it allows for computa#onal-aided quan#ta#ve analysis and visual reconstruc#on of the structure.[15, 60-62] Nonetheless, mercury porosimetry represents a rela#vely straighTorward measurement method with long-established and proven reliability for the assessment of various porous materials. The common range of assessed pore size with mercury porosimetry starts from 2–3 nm (max. 414 MPa overpressure) and reaches pores up to 500 m (few kPa overpressure).[63-67] Pore size distribu#on in the lumina size range for woods was a ma9er of research in several studies.[68-74] SoMwoods are characterized by unimodal distribu#on, where the distribu#on peak can be found between 0.2–0.3 m and 2–3 m. It is worth highligh#ng that pit forma#ons in untreated soMwood can cause bo9lenecks during mercury intrusion, which results in biased distribu#on toward peak maxima at lower pore sizes. Moderate thermal treatment (e.g., 180 °C) unlocks bo9lenecks and shiMs the distribu#on to a more realis#c one. While present for soMwoods, hardwoods do not indicate the presence of such an issue.[73] Hardwoods show bimodal pore size distribu#on, where the first peak is oMen located at 0.5–2 m and the second peak at 20–40 m. Those peaks represent the characteris#cs of the cell wood structure of each wood type: soMwood – tracheids with homogeneous lumen diameter and hardwood – vessels and fibers with more variable sizes (cf. Table 2.1). Aside from the larger structures, the pore size distribu#on of wood indicates the presence of pores with a size range lower than 0.1 m, which is related to the presence of so-called “microvoids” in the structure. Those include the endpoints of lumina channels as well as pit forma#ons (up to 0.01 m). Pores with a smaller size represent vacancies in the cell wall structures (bio-component assembly), which also can be found under the name “nanovoids”.[71, 72] Overall, the structure of a specific wood is primarily dependent on the wood type (soMwood/hardwood), secondary to species and genus, and ter#ary to the sample loca#on within the trunk (sapwood, heartwood) and tree growth condi#ons. Therefore, the measurement of wood obtained from two different trees of the same species, genus, and loca#on in the trunk and growing in the same region should result in a rela#vely similar distribu#on. As pointed out in Sec#on 3.2., the thermal degrada#on of wood leads to changes in the dimensions of the lumina, which significantly impacts the pore size distribu#on of the obtained wood-derived char. However, the literature includes only a few studies that focused on the quan#ta#ve assessment of the impact of thermal treatment on the pore size distribu#on in wood.[43, 46, 75-79] The most generic and straighTorward observa#on from available studies indicates that the thinning of the cell walls increases the diameter of the lumina and the overall volume of pores within a considered structure. Furthermore, with the increase in the severity of the thermal treatment, the loca#ons of peaks in the CHAPTER 2 92 pore size distribu#on shiM toward higher values. That is especially visible for peaks at 2–3 m and at 20–40 m in the wood structure. From the results presented by Kameyama et al., it can be deduced that if the ini#al structure contained pores in the size range of 20–40 m, the thermal degrada#on in sufficiently high temperatures may lead to the forma#on of pores with a size above 100 m. Therefore, if a size range above 100 m is excluded from the measurement, the obtained results can be biased and do not reflect the complete pore size distribu#on.[78] The available results do not allow for a more comprehensive assessment of the changes in pore size distribu#on with an increase in the degrada#on temperature, nor for a more precise quan#fica#on of the phenomenon. Moreover, the impact of par#cle shrinking on the pore size distribu#on is not detailed in any of the available literature sources. Considering the tailoring of wood-derived chars for soil applica#ons, the lumina-range pore size distribu#on relates to the majority of its required proper#es, so addressing this issue in future studies seems relevant. 3.4. Changes in molecular structure (up to 500 °C) The macroscopic changes of the wood structure occurring during thermal conversion are directly related to the behavior of its bio-components and their deriva#ves, with thermal degrada#on leading to the forma#on of carbonaceous molecular structures (further referred to as char structure). The carbonaceous molecular structures derived from bio-components degrada#on change their characteris#cs with temperature, so a wood-derived char cannot be considered a uniform material regardless of the conversion condi#ons. In the literature, studies can be found focusing on the development of biomass-originated carbonaceous structures with temperature, which allows for deriving uniform mechanisms of changes.[57, 80-96] The inves#ga#on of changes in carbonaceous molecular structures is not a straighTorward task, and comprehensive deriva#on of the mechanism requires synthesizing results from several measurement techniques, such as: Fourier transform infrared spectroscopy (FTIR), Raman spectroscopy, X-ray diffrac#on (XRD), near-edge X-ray absorp#on fine structure spectroscopy (NEXAFS), wide-angle X-ray sca9ering (WAXS, 5° – 60°), small-angle X-ray sca9ering (SAXS, 0.1° – 5°) and nuclear magne#c resonance spectroscopy (NMR) methods using ¹H isotope and 13 C isotope. Those measurements are usually complemented with the assessment of the elemental composi#on (elemental analysis), X-ray photoelectron spectroscopy (XPS), and transmission electron microscopy analysis (TEM). Only a combina#on of results from the men#oned methods provides sufficient informa#on that can be translated into a consistent descrip#on of a mechanism of molecular changes.[57, 91, 92] Recent progress in computa#onal science, combined with progress in understanding the molecular mechanisms, allows for implemen#ng molecular dynamics (MD) methods to comprehensively simulate char structure development on a molecular level.[87, 88] One can ask, why such high effort is directed to the subject. The ma9er is of high importance, as molecular structures and their changes relate directly to the thermal, physical, mechanical, and chemical proper#es of biomass-derived chars, so which directly translate to their applicability and tailoring their proper#es for novel applica#ons (chemical and electro-chemical). Sources men#oned above provided the state-of-the-art descrip#on of molecular changes, while this and the following Sec#ons focus only on the main mechanisms and outcomes of the changes in carbonaceous molecular structure. The descrip#on will refer to the pathway of degrada#on of biocomponents, which does not describe the release of compounds with low molecular weight. Therefore, it will focus on changes in the organic structure that remain, as a solid aMer the thermal degrada#on of bio-components. That behavior is primarily related to degrada#on condi#ons when a rela#vely low hea#ng rate does not favor bio-component fragmenta#on and deriva#ves evapora#on (slow pyrolysis condi#ons). Furthermore, the focus will be devoted to changes occurring from 350–400 °C to 800–900 °C, as such a temperature range is typically associated with char produc#on from biomass. Thermal degrada#on up to ca. 300 °C induces mainly dehydra#on and decarboxyla#on of less stable polysaccharide-based structures, which leads to their anhydrous forms. Up to 300 °C, it refers primarily to hemicelluloses and includes chain scission and depolymeriza#on combined with the release of more CHAPTER 2 93 vola#le anhydrous hemicellulose deriva#ves. The unreleased part of the anhydrous hemicellulose undergoes further carboniza#on through dehydra#on and decarboxyla#on. It is coupled with ring opening and re-configura#on/re-polymeriza#on of hemicellulose deriva#ves, which, in the end, form carbonaceous ma9er. Up to 300 °C, cellulose does not undergo severe degrada#on, while its structure becomes affected. This relates primarily to breaking intermolecular and intramolecular hydrogen bonds between cellulose chains and par#al dehydra#on of their surface-located hydroxyl groups. Similarly, lignin does not decompose severely up to 300 °C, while its least stable groups undergo degrada#on. The la9er refers to breaking side chains and linkages between lignin units (alipha#c and oxygen bonds) through dehydrata#on and decarboxyla#on pathways. Moreover, lignin between ca. 150 °C and 250 °C becomes soM and obtains semi–fluid proper#es (maximum fluidity at ca. 225 °C). The lignin soMening is related to its glass phase transi#on caused by the scission of covalent bonds and re-configura#on of branches in thermally treated lignin, which unblocks movement.[97, 98] A temperature rise to 300 °C causes lignin solidifica#on, so a re-gaining of its solid ma9er proper#es. In contrast to polysaccharides, lignin degrada#on mainly occurs through free radical reac#ons. When ac#ve free radicals containing benzene rings are formed, they can interact with other molecules to condense into larger structures. As the hemicellulose degrada#on does not lead majorly to char forma#on, the mass loss of wood up to 300 °C corresponds to its disappearance. So, quan#ta#vely, the mass loss up to 300 °C roughly corresponds to the ini#al content of hemicellulose, while the case-specific amount of hemicellulosederived char adds to the varia#on in mass loss results. Addi#onally, to the mass loss, the ini#al degrada#on of cellulose and lignin adds up, which becomes significant above 250 °C. Together, it roughly adds to ca. 40 wt. % loss of ini#al mass up to 300 °C. The elemental composi#on is affected by the occurred degrada#on, which corresponds to a drop in the H/C ra#o from 1.6 to 1.1 and a drop in the O/C ra#o from 0.6–0.7 to 0.4. Nonetheless, the oxygen and hydrogen content in the structure is s#ll rela#vely high.[88, 99] As the extensive forma#on of fused benzene rings (aroma#za#on of the structure) does not occur at such a low temperature, the largest aroma#c parts of the structure contain ca. 1–2 benzene rings. Their presence is mainly related to the ini#al structure of lignin (and possible radical-based fusing of 2 coalescent rings).[100, 101] Overall, at 300 °C, the wood cell wall is composed of the ini#ally thermally affected cellulose (with reduced hydrogen bonding) and lignin (with reduced cross-linking) with a minor addi#on of the charred hemicellulose deriva#ves. The #ghtness of the biocomponent assembly should be reduced due to the lack of hemicellulose. However, it can be suspected that lignin fills the formed vacancies as its soMening occurs (also indicated by the par#cle shrinking at the macroscale). An increase of the conversion temperature above 300 °C has far more severe consequences on the ini#al structure composi#on and its charring, as between 300 °C and 500 °C, complete degrada#on of cellulose and lignin into their carbonaceous deriva#ves occurs. Star#ng from the cellulose, its main degrada#on occurs around 330 °C and corresponds to the complete disintegra#on of its crystallinity.[82] The degrada#on pathway of cellulose is based on the scission of glycosidic bonds and the complete disappearance of the stabilizing hydrogen bonds. Simultaneously, vola#le cellulose deriva#ves that did not evaporate from the structure undergo thermal degrada#on. This occurs primarily by the elimina#on of oxygen from its structure through dehydrata#on, decarboxyla#on, and decarbonyla#on pathways, which enable ring opening and molecular re-organiza#on. The la9er is related to the structure aroma#za#on, so the forma#on of aroma#c carbon (sp 2 ) rings from the alipha#c carbon (sp 3 ) parts. Alipha#c carbon aroma#za#on starts at 400 °C, while at such a temperature, its rate is rela#vely low and does not accelerate un#l 500–600 °C. More severe aroma#c re-organiza#on is related to polycondensa#on, so a fusion of aroma#c ring structures into larger polyaroma#c sheets, but it does not occur significantly un#l 600–700 °C.[94, 95, 102] For cellulose, between 400 °C and 500 °C, cellulose-derived char structure increases its aroma#c carbon content at the expense of alipha#c carbon.[86] At 500 °C, the cellulose-derived char begins to resemble the ligninderived char (for the same temperature), where the single rings and a few fused-ring structures are cross-lined through alipha#c carbon bridges and oxygen bridges. Nonetheless, the carbon structure s#ll shows a high degree of disorderedness and a significant share of alipha#c carbon.[81] The lignin CHAPTER 2 94 deriva#ves between 300 °C and 500 °C undergo extensive thermal degrada#on through the dehydrata#on pathway and fragmenta#on of the methoxy groups. The la9er occurrence is responsible for the forma#on of methanol and methane during the thermal degrada#on of lignin. From 450 °C, a severe scission of the bonds present in the ini#al lignin structure (alipha#c and oxygen bridges) begins, which starts from the most suscep#ble (ether bonds, oxygen bridges) to the most thermally resistant (alipha#c bonds between benzene rings).[103] Unbinding func#onal groups of the lignin structure, along with the appearance of free radicals, effec#vely promotes the fusion of aroma#c rings into larger agglomerates.[81, 84] In a more generic view, from 300 °C to 500 °C, further mass loss in wood-derived char takes place, and at the end of the inves#gated temperature range, the remaining mass cons#tutes ca. 30 wt.% of the ini#al mass (a drop by ca. 30 wt.% from 300 °C).[88, 99] The remaining mass is mainly originated from charred structures of lignin, while cellulose and hemicellulose-originated chars represent a minority.[81] The detailed shares are case-specific, as the extent of the degrada#on of each biocomponent to the char is strictly related to the condi#ons of the thermal degrada#on. Reac#ons and degrada#on pathways that occur between 300 °C to 500 °C cause consecu#ve removal of the oxygen and hydrogen from the char structure, which translates to a further reduc#on of the H/C ra#o to 0.3– 0.4 and the O/C ra#o to ca. 0.10–0.15 (at 500 °C).[88, 99] Along with the increase in the degrada#on temperature, the forma#on of aroma#c ring structures begins in place of the alipha#c carbon structures. That corresponds to the rise of the size of the fused benzene ring agglomerates, whose size increases from 1–2 rings at 300 °C, though 4–5 rings at 400 °C, to ca. 7–13 rings at 500 °C. The number of rings composing the agglomerate is strictly related to the ini#al composi#on of wood and conversion condi#ons, while the forma#on of the larger fused-rings agglomerates at 500 °C is mainly caused by the presence of lignin-deriva#ves.[94, 100, 101, 103] Overall, at 500 °C, the ini#al bio-polymer structure is completely degraded and replaced by the newly formed carbonaceous structure. Then, the char structure does not show any resemblance in proper#es to the ini#al wood from which it was formed. 3.5. Organiza,on and ordering of bio-originated carbon structures The descrip#on of the molecular structure of wood-derived char formed above 500 °C requires the introduc#on of a few addi#onal concepts and their clarifica#on. The representa#ve graphite structure in rela#on to informa#on that can be obtained with XRD analysis is shown in Fig. 2.12. Star#ng from the planar (2D) view, fused aroma#c rings in carbonaceous structures represent a graphene layer, which is defined as a single-atom-thick sheet of hexagonally arranged carbon atoms with sp 2 hybridiza#on (aroma#c carbon). In the literature, such a layer can also be found under names like: polyaroma#c structures, polycyclic aroma#c structure, aroma#c layer, polycyclic aroma#c hydrocarbon (PAH)-type structure, or graphite-like layer, as such names had been proposed before the science coined the term “graphene”.[57] The size of the graphene layer is typically described by layer coherence length (La), which corresponds to an in-plane number of aroma#c rings of a layer and is quan#fied by the (100) peak in the XRD spectrum. The other 2D plane of the molecular carbon structure is the out-of-plane dimension, which is perpendicular to the in-plane dimension. It describes the alignment of graphene layers, also known as “stacking of layers”, so an order of layers. The relevance of the out-of-plane dimension of graphene layer stacks gains significance with the presence of sufficiently large graphene layers, which allows for their forma#on. To describe the out-of-plane dimensions, two parameters are used: (1) interlayer spacing (d) represen#ng the average distance between two adjacent planes, and (2) average height of carbon layers (Lc) represen#ng the overall thickness of graphene stacks, where both parameters are quan#fied by the (002) peak in the XRD spectrum.[57, 104] The 3D organiza#on of carbon material is a combina#on of arrangements in the in-plane and out-ofplane dimensions, where graphite is used as an ul#mate benchmark. It is worth highligh#ng that the literature indicates the relevance of the (101) peak, which can be no#ced sparingly in the XRD spectrum for bio-derived carbons produced above 500–600 °C.[85] The (101) peak is usually assigned as the graphi#c structure (stacked graphene layers), while it is only par#ally true as its presence precisely CHAPTER 2 95 indicates well-ordered graphite (with hexagonal (AB) stacking – 2H structure or rhombohedral (ABC) stacking – 3R structure). However, carbonaceous material requires treatment at temperatures star#ng from ca. 2000 °C to achieve permanent forma#on of such well-organised stackings. Furthermore, the long-range forma#on of well-organized structures is only possible when the material has a sufficiently low content of heteroatoms (O, S, N, Cl) and inorganic ma9er, as these introduce defects in crystalline structures and prevent the op#mal merger of layers and stacks.[57] Fig. 2.12. Visualiza#on of crystalline graphi#c structure, to which base turbostra#c carbon stacking and basic structural unit (BSU) defini#ons were derived: La 100 – graphene layers size; d 002 – distance between two adjacent graphene layers; Lc 002 – thickness of a graphene stack formed from graphene layers. Addi#onally, the incorpora#on of heteroatoms (especially oxygen) into the carbon structure induces the forma#on of non-hexagonal rings within the graphene layer (mainly pentagonal and heptagonal). When present, such a varia#on within the structure leads to the curving of the layer and hinders the perfect alignment of layers into a well-ordered graphite structure.[57, 91] Therefore, it is important to dis#nguish imperfect graphene stacking from perfect graphi#c stacking in the elabora#on, as each term refers to significantly different stacking pa9erns. The appearance of the (101) peak in bio-derived carbons (below 2000 °C) indicates the presence of the disordered stacking of graphene layers, so a stack of randomly orientated non-ideal graphene layers called the turbostra#c structure, which can reflect as the (101) peak in the XRD spectrum. The turbostra#c structure is a dis#nc#ve structural element of the biomass-originated carbonaceous structures, which in literature is usually referred to as “turbostra#c carbon stacking” or “basic structural unit (BSU)”, where the la9er will be preferred in this work. An exemplary representa#on of carbonaceous material with the stacking of distorted layers is presented in Fig. 2.13. In the literature, various terms for stacks of disordered graphene layers can be found, like: graphi#c crystallites, turbostra#c crystallites, parallel-layer groups, graphi#c microcrystallites, nanocrystallites, conduc#ve nanoclusters, stacked polyaroma#c structures, or stacked graphene.[57] The conversion of wood up to 500 °C leads to the forma#on of its carbonaceous layers of sufficient size (number of rings) to enable their interlayer stacking. The composi#on of char up to 500 °C s#ll indicates rela#vely high oxygen content and alipha#c carbon content (sp 3 ), which translates to a carbon structure enriched with small graphene layers strongly cross-linked by alipha#c and oxygen bridges. Such structural organiza#on impacts the characteris#cs and further growth of the graphene layers and influences layer stacking pa9erns. In detail, the presence of oxygen (incorporated, in links and func#onal groups) and alipha#c carbon (cross-links between graphene layers and overall share within the structure), leads to the curving of layers and hinders the merger of layers and their parallel stacking.[57, 91] CHAPTER 2 96 As indicated by Oberlin, a higher content of oxygen in the carbonaceous structure is directly related to the reduc#on of mobility of BSUs (due to extensive cross-linking).[105] Furthermore, the oxygen in the ini#al material enhances the release of hydrogen from the structure through the dehydra#on pathway. The hydrogen in the material plays a cri#cal role in the merger of layers and the forma#on of stacks, as hydrogen increases a layer's mobility when present at the edges of a layer/stack. With sufficient mobility (low oxygen, high hydrogen), layers are not restricted to merging in the most energe#cally favored configura#on (large and flat layers) and well-organized stacking pa9erns (graphite stacking). Addi#onally, the high mobility of carbon structure elements enables the op#mal merging of BSUs into long-range agglomerates (graphite structure).[105] Overall, when layers and BSUs are sufficiently small (low number of aroma#c rings) and the stabilizing effect from alipha#c and oxygen bonds is weak, the structure has a high ability to re-organize.[57, 105] Fig. 2.13. LeM – Visualisa#on of the carbon material structures from a molecular assembly consis#ng of 15 distorted graphene stacks (3 layers); right – simplified schema#c represen#ng the stacking of distorted graphene stacked structures with their dis#nc#ve elements.[57] In rela#on to wood-derived chars, alipha#c and oxygen cross-links stabilize the structure and reduce the mobility of layers and BSUs. That leads to sub-op#mal shape and mergers of layers, and further to imperfect stacking. The carbonaceous structure has the ability to improve its organiza#on with temperature, but such re-organiza#on needs to be allowed in rela#on to the spa#al arrangement. Oppositely, mergers between aroma#c structures have high thermal stability and, when they appear, permanently impact the carbonaceous structure. Therefore, when imperfect layers and BSUs are formed in the carbonaceous structure, and aroma#c links between those are established, their presence restricts further improvement of carbon structure alignment into long-range, well-organized structures. In the end, the forma#on of a well-organized carbon structure, despite thermal treatments at elevated temperatures (e.g., 2000 °C), does not enable the structure to reach perfect ordering. The carbon materials, which, due to the characteris#cs of the ini#al material, follow such a structure development, are known as “non-graphi#zing carbons” or “hard carbons”. It explicitly refers to the inability of the carbon material to form a long-range, well-organized structure regardless of the severity of the treatment. However, it does not imply that the turbostra#c stacks cannot be organized into smallscale graphi#c structures, while their presence is local (short-range) and heterogeneously distributed within the carbon material.[57] In opposi#on stands the “graphi#zing carbons” or “soM carbons”, which do not encounter the hindering effect of reduced carbon element mobility during the structure forma#on. Therefore, carbonaceous materials derived from feedstocks unburdened from the impact of a high content of alipha#c carbon and heteroatoms have the ability to form a long-range, wellorganized graphite structure upon thermal treatment. CHAPTER 2 103 The changes in true density above 1000 °C are not presented in Fig. 2.16., but the literature indicates that the treatment above 1000 °C leads to a decrease in the true density of the wood-derived chars.[44] It might be seen as counterintui#ve, since the further temperature increase should lead to a higher structure polycondensa#on and be9er organiza#on toward a graphi#c structure, resul#ng in the true density even closer to the value of elemental carbon. The presumed cause of the drop in the true density above 1000 °C is related to the encapsula#on of the voids with the progress of graphene layers’ growth and their merger.[127-129] The encapsula#on of voids may impose a strong restric#on of access to pores by helium, leading to problems with its diffusion. As helium would not fill those restricted voids, the encapsulated empty volumes will lower the overall density derived from the measurement. On the macroscopic scale, it was observed that grinding of the char produced above 1000 °C increases its true density value, which advocates for the postulated hypothesis.[44] 4.3. Bulk density and porosity Among the proper#es of wood, bulk density is the most extensively measured. In Fig. 2.17. sta#s#cal distribu#ons of the bulk density of selected wood species are presented.[1] The key parameters that affect the density of wood relate primarily to factors as: tree species and genus, and secondarily to: growing condi#ons, geographic loca#on, tree age, and loca#on of #ssue within a tree (earlywood/latewood). A comprehensive revision of the bulk density of woods can be found in the book by Niemz et al.[12] The bulk density of a specific wood sample can differ from the generally assumed average due to the abovemen#oned factors and should be measured for a specific experimental work. Nonetheless, for generic considera#ons, an assump#on of the average value for an inves#gated case should not introduce a cri#cal error to an es#ma#on. Fig. 2.17. Frequency distribu#on of the bulk density of selected wood species with an indica#on of average bulk density value and the number of samples (measured on dry basis).[1] The bulk density is defined as the mass of a wood par#cle in a representa#ve volume, which includes cell wall structure and enclosed voids (pores). The generic overview of the rela#on between the share of cell walls and voids and the bulk density of various wood species is presented in Fig. 2.18. The la9er relates to lumens and microvoids, which correspond to structural elements such as (from largest to smallest): tracheids/vessels, pits, spaces between fibers, and vacancies in the assembly of biocomponents. The bulk density, also known as specific gravity, is typically measured for oven-dried wood, while the rela#on between the bulk density of wood and its moisture content is well documented in the literature.[4-6, 11, 12] 0 2 4 6 8 10 12 14 16 250 350 450 550 650 750 850 950 1050 frequency [%] bulk density [kg/m3] spruce 455 kg/m3 (n = 1692) pine 489 kg/m3 (n = 2918) oak 647 kg/m3 (n = 994) beech 671 kg/m3 (n = 1778) CHAPTER 2 104 Fig. 2.18. Generic rela#on between the share of cell wall substance on the bulk density of various wood species.[12] The sum of voids within a considered volume of a wood par#cle is commonly known as porosity. As the wood cell wall density has a rela#vely constant value and gas within voids has a negligible density in comparison to the cell wall, a straighTorward correla#on between porosity and bulk density of a wood sample can be derived. The experimental valida#on of such correla#on was provided by Plötze and Niemz and is presented in Fig. 2.19.[72] Porosity is a macro-property of wood, while more detailed characteris#cs of the porosity and its rela#on to lumina and microvoids within the wood structure can be obtained with pore size distribu#on analysis. Results from Plötze and Niemz indicate that the volume-relevant porosity in hardwood starts from pores larger than 10 nm, while for soMwood it starts from pores larger than ca. 50 nm.[72] As pointed out in Sec#on 4.1, the measurement of the true density of the wood cell wall with mercury porosimetry may result in a significant error. As shown in Fig. 2.19., assessing the sample porosity and bulk density with mercury porosimetry also introduces an error in the results. However, it should be highlighted that the introduced error relates to a case when the bulk density/porosity is calculated using the true density value derived from mercury porosimetry. Fig. 2.19. Correla#on between wood par#cle porosity and the bulk density, with values of bulk density derived by helium pycnometry and mercury porosimetry.[72] Changes in the bulk density of wood-derived char par#cles, which take place during the thermal degrada#on of a wood par#cle, originate from the occurrence of three major phenomena. The first relates to changes in the geometric dimensions of a par#cle with degrada#on temperature, which impacts the volume of the considered par#cle. The second derives from the thinning of the walls of 20 30 40 50 60 70 80 300 400 500 600 700 800 900 1000 1100 1200 porosity [vol.%] bulk density [kg/m3] He pycnometry Hg porosimetry CHAPTER 2 105 the cell wall along with its degrada#on and charring, which translates to changes in the shares in volume between solids and voids within the volume considered. The third impacTul phenomenon is the change in the true density of the cell wall, which relates to the degrada#on of bio-components and further organiza#on of the derived carbonaceous material. Therefore, forecas#ng changes in the bulk density of wood-derived char with degrada#on temperature remains a difficult task. The influence of the above-men#oned phenomena adds to the complexity of the reliable measurement of the bulk density of the wood-derived char, in comparison to wood. In the case of wood, the bulk density can be derived from the mass measurement (known mass) of precisely cut par#cles with easyto-measure dimensions (known volume) without introducing a significant error. In the case of woodderived char, the overall par#cle geometry can be complex as it is affected by uneven shrinking and the appearance of cracking (randomized forma#on), which is also formed within the considered geometrical shape of a par#cle. If addi#onal voids created by cracking and distorted geometric shapes are not accounted for properly, it can induce a high error in the bulk density measurement. Therefore, straighTorward geometric measurement and re-calcula#on of its results into a par#cle volume are not recommended. Brewer et al. indicated that a reliable bulk density assessment (for a par#cle volume assessment) can be conducted using the mercury porosimetry technique (at the ini#al pressure) or an envelope density measurement apparatus.[59] The other suggested method is the water immersion method in accordance with ASTM D2395.[122] Furthermore, as the true density of the wood-derived char changes with temperature, the assessment of the porosity in a reliable manner requires knowing both, bulk density and the true density. It should be no#ced that in contrast to the wood cell wall, the char cell wall contains a no#ceably higher volume of pores with a size below 2 nm. Therefore, assessing the true density measurement of a char par#cle should be conducted with helium pycnometry rather than mercury porosimetry, as the la9er will introduce far higher error for char than for wood. Only a few studies can be found in the literature that simultaneously provide a value of true and bulk density of the same wood-derived char, especially in a temperature-sequenced manner. The available data is presented in Fig. 2.20., where the values of the bulk density and porosity of char were normalized to the ini#al value of the origina#ng wood.[53, 59, 122, 130, 131] It should be highlighted that part of the sourced data (e.g., Zickler et al. [53]) does not include measured true density, so the averaged true density of char for a specific temperature was sourced from Fig. 2.16. While such an approach is theore#cally reasonable, its reliability depends on the quality of the rela#on between true density and temperature, so the presented porosity results can be burdened with a bias. Fig. 2.20. Change of normalized bulk parameters (to the value at 25 °C) of wood-derived chars with the temperature of thermal degrada#on: leM – bulk density, right – porosity (wood species: S – spruce, P – pine, B – beech, E – eucalyptus, M – mesquite).[53, 59, 122, 130, 131] 0,4 0,5 0,6 0,7 0,8 0,9 1,0 100 200 300 400 500 600 700 800 900 normalized bulk density [ - ] temperature [°C] Zickler et al.,(2006) (S) Brewer et al.,(2014) (M) Somerville et al.,(2015) (B) Dufourny et al.,(2019) (S) Dufourny et al.,(2019) (E) Li et al.,(2021) (S) Li et al.,(2021) (P) 1,0 1,2 1,4 1,6 1,8 2,0 100 200 300 400 500 600 700 800 900 normalized porosity [ - ] temperature [°C] Zickler et al.,(2006) (S) Brewer et al.,(2014) (M) Somerville et al.,(2015) (B) Dufourny et al.,(2019) (S) Dufourny et al.,(2019) (E) Li et al.,(2021) (S) Li et al.,(2021) (P) CHAPTER 2 106 As it can be derived from Fig. 2.20., the trend in bulk density and porosity with temperature shows a similarity between studies, including different wood species. In terms of the bulk density, on average, from all referenced studies, up to 400 °C, a drop in the density value by ca. 35 wt.% can be no#ced. Further temperature increases to 600 °C seem to reduce the bulk density even further, while the magnitude thereof is study-specific. Then, up to 900 °C, studies started to show a similar trend, where the bulk density increases following temperature. However, the magnitude s#ll remains study-specific. In terms of porosity, up to 400 °C, its value increases by ca. 20–25 vol.% for all referenced inves#ga#ons. The trends between studies start to vary above 400 °C. Part of the data (also within the same study) indicates that porosity above 400 °C remains rela#vely constant regardless of the thermal degrada#on temperature increase. The other part of the data indicates that the porosity increases with the temperature of the thermal treatment. As pointed out in Sec#ons 3.3 and 4.2., changes in the true density and the extent of cell wall thinning with temperature occur rela#vely homogeneously for wood species. Therefore, the change in the volume of the par#cle may be pointed out as the factor that, to the greatest extent, introduces devia#on in bulk density and porosity results. Considering that the progress of a par#cle changing its geometry with increasing temperature above 400–500 °C may be present, but it does not necessarily have to be (Sec#on 3.1.). Therefore, two extreme cases of changes in the volume of a par#cle with a temperature above 400–500 °C can be derived. In one scenario, the consecu#ve shrinking is not present above 400–500 °C, which translates to an increase in the porosity of par#cles only due to cell wall thinning. On the other end is a scenario where the shrinking progresses with the temperature. That corresponds to the occurrence of both phenomena, cell wall thinning and decrease of lumina diameter, which can achieve an effec#ve balance and lead to apparent constant porosity. For both theore#cal scenarios, the balance between the increase of the true density of the cell wall and the characteris#c behavior of porosity drives the effec#ve change in bulk density. In prac#ce, there can be a devia#on from the men#oned scenarios, which relates to the appearance of structural discon#nui#es (cracks within the considered par#cle volume) and strong distor#on of a par#cle shape. As both can influence the assessment of the considered volume (typically an increase), it can cause an eleva#on of the obtained porosity and a decrease of the bulk density. The risk of occurrence of such a bias is explicitly related when the bulk density is measured without the use of the appropriate methods. Fig. 2.20. Indicates the generic trend of changes in the bulk density and porosity, while it strongly highlights the mul#-dependence of the aforemen#oned men#oned parameters. Moreover, influencing phenomena are also related to the ini#al wood proper#es (e.g., species-related lumen characteris#cs, earlywood /latewood) as well as the conversion parameters aside from the temperature (explicitly hea#ng rate). Therefore, the final change in bulk density and porosity remains elusive to a9empt to model a generic scenario. Interes#ngly, a specific trend for soMwood and hardwood seems to be no#ceable in Fig. 2.20. However, a conclusion cannot be drawn as the uncertainty of the presented data is high and the number of studies is rela#vely low to support such a far-reaching statement. 4.4. Moisture, water types, and fiber satura,on point The ma9er of moisture and water content in wood has already been intensively inves#gated over several decades. An extensive summary of the rela#onship between water and wood, as well as its drying, can be found in the relevant literature.[4-6, 11, 12, 132] Therefore, in this Sec#on, the elabora#on on the subject will highlight only key elements related to wood thermal degrada#on. From the most general view, water in the wood can be present in two primary forms. The first is the water associated with the wood cell wall, commonly called “bound water”. The second is the water that fills the lumina of wood, called “liquid water” or “free water”. Both forms of water in wood naturally occur within the tree #ssue during its growth. Fig. 2.21. presents a generic overview of the forms of water present in wood #ssue. CHAPTER 2 107 The assembly of wood cell wall bio-components represents a very #ght structure, which is impenetrable by water molecules, where, especially, the crystalline cellulose fibers are impenetrable. Therefore, bound water relates to the water molecules that are a9ached to the surface of the biocomponents through hydrogen bonds formed between a water molecule and a hydroxyl group present on the surface of bio-components (water molecules have a high affinity to hydroxyl groups). Hydrogen bonding of a water molecule to a bio-component reduces its mobility within the structure, so transport of bound water occurs only through diffusion. The literature indicates that water has a preference toward hydroxyl groups located on the surface of cellulose microfibrils, so the cellulose content relates significantly to the presence of bound water in the wood cell wall.[12] Therefore, the presence of cellulose in the cell wall structure is one factor that enhances the hydrophilic proper#es of wood. In prac#ce, this can be observed by rapidly re-absorbing the moisture from the environment when wood is water-depleted upon drying. When the cell wall surface is saturated with water molecules (water molecules occupy all hydroxyl groups) and the concentra#on of water vapor in the environment is sufficiently high (ca. 99.0–99.9% rela#ve humidity), water can accumulate in the wood structure through another mechanism. That explicitly relates to capillary condensa#on, which causes the filling of empty voids in the wood structure with liquid/free water. As expected, the filling starts from the smallest presented voids (nanoand microvoids), progresses through slit-shaped ends of lumens and pit forma#ons, and ends with lumens (tracheids/vessels).[12] Fig. 2.21. Schema#c presenta#on of the rela#on between water content, wood lumen, and wood cell wall (in the circle): leM – oven-dry wood; middle – moist wood with the water content below the fiber satura#on point (FSP); right – freshly-cut, wet wood with the water content higher than the FSP.[12] Both water forms added together represent the water content of wood, which is more commonly known as the moisture content. The maximum amount of water that the cell wall can bind is called the fiber satura#on point (FSP), and it was introduced first by Tiemann in 1906.[133] When the moisture content of wood is equal to its FSP, all water contained in the wood is in the bound form. Further increase of the moisture content, above the FSP, leads to the appearance of water in the form of capillary water (free water). An overview of the FSP for different species and its value measured using different methods can be found in the book by Niemz.[12] In the literature, two commonly used values of the FSP for wood can be found: 30 wt.% proposed by Stamm and 40 wt.% proposed by Skaar, where the former is more oMen applied.[6, 134] Furthermore, in the literature, there is a consensus on the density of the capillary water, whose value is typically indicated as 1000 kg/m 3 . The opposite can be observed for the density of the bound water. One way of its descrip#on was introduced by Stamm in 1938, who postulated that the bound water density changes with the moisture content of wood and varies between 1000 kg/m 3 at FSP (30 wt.% moisture content) and 1300 kg/m 3 for oven-dry wood (0 wt.% moisture content).[4, 12, 135] Stamm argued that the hydrogen bonds between hydroxyl groups and water molecules have a strength comparable to those occurring during chemisorp#on, which causes a volume contrac#on and a denser packing of the water molecules.[71] Another way is to assume that the bound water density has a constant value of 1000 kg/m 3 , which was indicated experimentally by Kellogg and Wangaard in 1968 and Skaar in 1972.[6, 113] CHAPTER 2 108 Regardless of the scien#fic validity of each, the la9er is the more commonly applied op#on.[11, 136140] As the main reason for the popularity of the la9er, the simplifica#on of the drying descrip#on can be pointed out, as the rela#on between bound water density and FSP would addi#onally induce a need for a valid correla#on for the heat of evapora#on of bound water as a func#on of process parameters. 4.5. Thermal treatment and hydrophilicity/hydrophobicity The ma9er of wood drying is highly relevant for #mber produc#on and related industries, as conducted inappropriately drying can affect #mber quality by damaging its structure and inducing irregular bends. The la9er originates from the fact that a moisture reduc#on in wood below its FSP (oven drying) induces a direc#on-dependent dimensional shrinking, which, in total, can reach up to 15 vol.% (reaching MC = 0 wt.%).[12] Therefore, the wood drying process for #mber is conducted in a #mely manner, which is cri#cal to not introduce excessive stress within the structure due to the release of moisture.[141-143] The drying of wood is typically conducted at temperatures below 100 °C and is driven by the diffusion of water between moist wood and an unsaturated environment (e.g., warm air), rather than rapid water evapora#on. Contrarily, for the thermochemical conversion of wood, the ma9er of dimensional deforma#on has negligible relevance, while the drying efficiency, understood by the #mely reaching of sufficiently low moisture content, is of the essence. Therefore, wood drying before the thermal conversion process is typically conducted at temperatures above 100 °C. At such condi#ons, diffusion does not represent the prevailing drying mechanism, while evapora#on (boiling off) of water dominates.[144-147] It can be pointed out that the experimental studies on the thermal conversion of wood are usually based on oven-dried material, so moisture removal and shrinking related to drying are typically excluded from the study scope. However, when the conversion relates to air-dried wood (moisture up to the FSP), the dimensional change related to drying may be of relevance. Then, the inclusion of a wood drying descrip#on along with related phenomena into the overall model descrip#on may be necessary to comprehensively understand changes and the cumula#ve impact of the conversion. Hydrogen bonds between water molecules and hydroxyl groups are thermally unstable. Therefore, the ability of a cell wall to bind water changes with temperature, which affects the value of the FSP.[5] As indicated by Stamm and Nelson, between 20 °C and 120 °C, the FSP decreases by approximately 0.1 wt.% per temperature increase by 1 degree.[148] Therefore, an increase in temperature has the ability to transform the bound water into free/capillary water. Such a phenomenon occurs only to a certain extent, as in any type of drying, the temperature of water in wood does not exceed 120 °C (temperature/pressure rela#on), because it undergoes evapora#on before any further temperature rises. Thermal conversion of wood, which starts to take place above 150–200 °C, includes degrada#on of its bio-components, so it significantly affects the rela#onship between cell wall and water. Addi#onally, changes in the structure (e.g., cell wall thinning) affect the capillary condensa#on of the water vapors within voids. While capillary condensa#on is related directly to structural changes of wood, its ability to bind water is affected by changes caused by the thermal degrada#on of biocomponents. In the la9er, the highest relevance is by the presence of hydroxyl groups on the cell wall surface.[149-151] The overview of the changes in the hydroxyl groups with conversion temperature and its rela#on to the binding of water is presented in Fig. 2.22. As men#oned in the Sec#on above, virgin wood is characterized by a high concentra#on of hydroxyl groups on the cell wall surface, which originate from all bio-components, while the key impact has those origina#ng from cellulose.[85] Up to 300 °C (torrefac#on), the thermal degrada#on mainly affects hemicellulose. Therefore, at 300 °C, wood s#ll contains a high share of the mainly untreated cellulose in the cell wall structure, which translates to reten#on of high concentra#ons of hydroxyl groups. In general, it can be stated that torrefied wood retains its ini#al hydrophilic proper#es, understood as the ability to re-absorb and bind water passively. [Document text truncated for crawler view.]