Kinetic Analysis of Construction, Renovation, and Demolition (CRD) Wood Pyrolysis Using Model-Fitting and Model-Free Methods via Thermogravimetric Analysis

Abstract

The pyrolysis of non-recyclable construction, renovation, and demolition (CRD) wood waste is a complex thermochemical process involving devolatilization, diffusion, phase transitions, and char formation. CRD wood, a low-ash biomass containing 24–32% lignin, includes both hardwood and softwood components, making it a viable heterogeneous feedstock for bioenergy production. Thermogravimetric analysis (TGA) of CRD wood residues was conducted at heating rates of 10, 20, 30, and 40 °C/min up to 900 °C, employing model-fitting (Coats–Redfern (CR)) and model-free (Ozawa–Flynn–Wall (OFW), Kissinger–Akahira–Sunose (KAS), and Friedman (FM)) approaches to determine kinetic and thermodynamic parameters. The degradation process exhibited three stages, with peak weight loss occurring at 350–400 °C. The Coats–Redfern method identified diffusion and phase interfacial models as highly correlated (R² > 0.99), with peak activation energy (E_a) at 30 °C/min reaching 114.96 kJ/mol. Model-free methods yielded E_a values between 172 and 196 kJ/mol across conversion rates (α) of 0.2–0.8. Thermodynamic parameters showed enthalpy (ΔH) of 179–192 kJ/mol, Gibbs free energy (ΔG) of 215–275 kJ/mol, and entropy (ΔS) between −60 and −130 J/mol·K, indicating an endothermic, non-spontaneous process. These results support CRD wood’s potential for biochar production through controlled pyrolysis.

1. Introduction

Construction, renovation, and demolition (CRD) activities generate huge volumes of waste. Particularly, countries where the construction of residential, commercial, and industrial spaces utilizes substantial quantities of wood fall under this case. According to [1] over 500 million tons (Mt) of CRD waste was generated annually across Canada, the USA, and Europe. The management and disposal of CRD wood residues create numerous environmental and economic challenges. This also includes furniture parts that are organically contaminated wood fibers without a promising scope for reuse [2]. Such wood contains treated, tinted, or composite materials infused with chemicals like polyaromatic hydrocarbons (PAHs), heavy metals, paints, lubricants, and resins, making them non-recyclable and deteriorating in quality [3]. Especially, chromated copper arsenate (CCA) remnants in CRD wood waste disposal could face major hurdles in terms of disposal due to rising restrictions [4]. A common, non-advisable practice followed for their disposal is usually landfilling, which causes problems when these toxic substances from CRD wood waste leach into the surrounding soil and groundwater, posing serious risks to human and ecosystem health. The anaerobic decomposition of organic fractions in wood waste within landfills can produce harmful greenhouse gases (GHGs) like methane and carbon dioxide, contributing to the climate crisis [5]. Inadequate monitoring, a lack of public awareness, the absence of strict regulations for landfilling, and the economic/technological burden inflicted on recycling technologies for CRD wood allow handlers to look for easy disposal methods. Even simple sorting is complicated due to the large variability in wood composition with geography and provincial sourcing practices. These externalities exacerbate an already growing strain on limited landfill spaces. Hence, alternative waste management technologies are needed to pave the way for valorizing CRD wood residues to bioenergy, biochemicals, and bioproducts.

Waste forest wood, whether sequestered directly from the forests or sourced from CRD sites, is typically a low-value feedstock with a scarce probability of being reused or recycled. The economic value of such wastes could be improved using efficient thermochemical conversion technologies such as torrefaction, gasification, or pyrolysis [6]. Pyrolysis has immense potential in serving as a recovery pathway for refused wood residues, mainly due to its distinct product stream, moderate operation conditions, and relatively simpler operation/instrumentation than gasification [7]. Upon depolymerization of biomass in low or no oxygen, the transformation yields three major products with high energy and oxygen content: solid biochar, liquid pyrolysis oil or biooil, and pyrolysis gas [8]. Depending upon the application under focus, pyrolysis can either be carried out at low or high temperatures. For instance, low-temperature pyrolysis (<500 °C) produces cyclic, stable aromatics as in biochar, whereas temperatures > 600 °C produce pyrolysis gases that are partly condensable into bio-oil [9]. Bio-oil can serve as a foundation for producing platform chemicals or can be upgraded to fuel intermediates [10]. Pyrolysis gas, commonly referred to as energy-rich syngas, made of CO, CO₂, H₂, and other light C₁–C₃, can also serve as a base for liquid synfuels or can be combusted in boilers/furnaces for energy generation, surrounding it from a resource recovery perspective [11]. Biochar may serve as a partial substitute for fossil coal and carbon black in hard-to-abate industrial sectors such as metallurgy, automobiles, construction, agriculture, soil remediation, biogas production, air and water purification [12,13,14,15]. For a complex feedstock like CRD wood, pyrolysis may even trap all inherent hazardous components and immobilize them within the biochar product. But these biochars can eventually cause toxicity concerns because they may contain original PAHs from the wood feed, as well as volatile organic compounds (VOCs), dioxins, and free radicals that are persistent [16].

For safe pyrolysis, other non-wood contaminants like nails, plastics, concrete, glass, and rubber have to be removed without fail during the pre-treatment stages to avoid complications with reactor operation and efficient biomass conversion. On this note, for smooth conversion, CRD wood biomass has to be uniformly size-treated via shredding, grinding, or milling to reduce the particle size so that heat transfer, bulk density, and energy density limitations are subdued [17]. Whenever biochar is leveraged to address environmental or industrial challenges, its stability determines the resistance toward chemical, thermal, or microbial degradation while enabling it to act as a carbon sink [18]. The nature of feedstock and pyrolysis parameters like temperature, heating rate, and biomass residence time (BRT) govern this stability factor through the physicochemical properties of biochar, as understood from research by [19]. The final quality and intended applications depend on how the biomass feedstock is pre-treated, what the process optimization steps are, and the scale of production.

Biopolymers in biomass, such as cellulose, hemicellulose (together called holocellulose), and lignin contents, have specific decomposition pathways that take place at different temperatures and different stages. Solid-gas phase reactions involving multiple reactive species make the prediction of pyrolysis reaction modeling cumbersome [20]. To understand what pyrolysis does to CRD wood biomass, getting to know the kinetics and thermodynamics is beneficial. Kinetics describes the nature and type of chemical reactions occurring during pyrolysis and the energy required to form intermediate complexes and products, whereas thermodynamics describes how energy enters and leaves a system during these chemical reactions. Usually, pyrolysis may contain single-step reactions proceeding individually or multiple competitive biomass breakdown reactions that proceed simultaneously or sequentially, underscoring the intricacies of this process [21]. Yield; ion exchange capacities; and influence of ash, pore size, volatile content, and carbon arrangement in biochar are invariably dictated by the thermal degradation profiles too. Therefore, careful consideration of the various temperatures, heating rates, conversion rates, and selecting appropriate kinetic models is very important to forecast pyrolysis behavior. Thermogravimetric analysis (TGA) is a systematic and structured method used to study the chemical kinetics of biomass pyrolysis. According to [22], for fitting experimental data, assumptions based on different theoretical models are used via model-fitting methods, whereas the biomass conversion rate with the associated temperature required for that conversion at different heating rates is applied to the experimental TGA data via model-free or isoconversional methods (used interchangeably). Kinetic parameters such as activation energy and thermodynamic parameters like enthalpy (ΔH), entropy (ΔS), and Gibbs free energy (ΔG), can be calculated at each heating rate employed to study the influence of pyrolysis process parameters on biomass decomposition [23].

To the best of our knowledge, there has not been a detailed study on the reaction kinetics and thermodynamics of CRD wood waste pyrolysis. Hence, two unique aspects are undercovered here. The first novelty element in this work surrounds the utilization of an underexplored, heterogeneous, and real-world biomass feedstock in the form of CRD wood rejects, laden with organic and inorganic contaminants, to disseminate practical findings on its thermal degradation behavior. The second novel element focuses on extracting kinetic and thermodynamic parameters from the CRD wood TGA weight loss data under multiple heating rates, comparing the results yielded from different models. So, this dual emphasis on feedstock physicochemical/biochemical composition and robust pyrolysis modeling framework could result in meaningful contributions to the field of sustainable waste material management.

To assist in answering these questions, the objectives of our study will be to (a) succinctly explain the theory behind model-fitting and model-free techniques with necessary equations along with the benefits and drawbacks of adapting to these methods; (b) investigate the physicochemical characteristics of biomass, and understand its surface chemical make-up; and (c) analyze the thermochemical degradation reactions via TGA at different heating rates and calculate kinetic and thermodynamic parameters. These measures would facilitate understanding of a robust pyrolysis processes using CRD wood as a substrate for biochar production.

To engage with the aforesaid objectives, the idea behind each model is explored initially, followed by carrying out a compositional, proximate, ultimate, and metal analysis of the CRD wood waste to understand the physicochemical properties of the feedstock. Next, the surface chemistry of CRD wood was also studied using Fourier Transform Infrared (FTIR) spectroscopy to project its composition and structural framework. Subsequently, TGA experiments using CRD wood biomass at four different heating rates—10, 20, 30, and 40 °C/min—were carried out, and the data were cleaned and curated for analysis according to the methods under consideration. Firstly, regarding model-fitting methods, the Coats–Redfern (CR) approximation was utilized. For isoconversional methods, the Ozawa–Flynn–Wall (OFW), Kissinger–Akahira–Sunose (KAS), and Friedman (FM) models were used. Based on the computed activation energy, pre-exponential factor, and other constants, the thermodynamic parameters were calculated. Finally, a prognosis of matching suitable reaction models to specific biomass conversion rates was also attempted.

2. Theory, Materials, and Methods

All analysis were performed using the same CRD wood waste, instrumentation, methods, and by the same technician, like what was elucidated by [3]. The characterization techniques are explained here again for clarity.

2.1. Foundational Derivation

Different biomass sources exhibit different reactions due to variability in composition [24]. Therefore, an exact determination of reaction kinetics during biomass thermal decomposition is almost impossible [25]. To understand the pyrolytic conversion of CRD wood into biochar, TGA of this biomass is carried out at same or different heating rates in the presence of nitrogen as an inert carrier gas. An effective TGA test can mimick crucial process conditions such as the temperature, BRT, heating rate, and feed particle size, which are usually interdependent operational factors in a pyrolysis process that help in critically determining the composition of resulting products [26]. Through a kinetic study, the activation energy of a feedstock can be computed to evaluate its reactivity, estimate its conversion efficiency, and assist in scale-up considerations to larger pyrolyzers during the design phase [27].

Rate of reaction for biomass in an isothermal condition is given as

$${\displaystyle \frac{\mathrm{d}\mathrm{\alpha }}{\mathrm{d}\mathrm{t}}}=\mathrm{k}\mathrm{\ast }\mathrm{f}(\mathrm{\alpha })$$

(1)

Here, k is the reaction rate constant, which is temperature-dependent; dα/dt is the change in mass of biomass taken for the analysis with respect to time (also referred to as biomass conversion rate); and f(α) is a differential function representing reaction models and governed by different reaction schemes presented in Table 1.

$$\mathrm{k}=({\displaystyle \frac{\frac{\mathrm{d}\mathrm{\alpha }}{\mathrm{d}\mathrm{t}}}{\mathrm{f}\left(\mathrm{\alpha }\right)}})$$

(2)

In the above equation, α, which represents a fractional change in mass of biomass reactant, can be represented as

$$\mathrm{\alpha }={\displaystyle \frac{{(\mathrm{x}}_{\mathrm{o}}-{\mathrm{x}}_{\mathrm{t}})}{{(\mathrm{x}}_{\mathrm{o}}-{\mathrm{x}}_{\mathrm{f})}}}$$

(3)

Here, x_o is the initial mass of biomass sample used for the kinetic study by TGA, x_f is the final mass of sample at the end of TGA, and x_t is the mass of sample in the TGA cycle at a given time t. Decomposition rate of biomass as a function of temperature can be construed using the Arrhenius equation as follows:

$$\mathrm{k}=\mathrm{A}\mathrm{\ast }\exp ({\displaystyle \frac{-{\mathrm{E}}_{\mathrm{a}}}{\mathrm{R}\mathrm{T}}})$$

(4)

In the above expression, A is the pre-exponential or frequency factor (s⁻¹ or min⁻¹), which represents the extent of molecular collisions in the wake of thermal decomposition at a given time and temperature; E_a is the activation energy required to cleave the bonds between feedstock biopolymers, namely cellulose, hemicellulose, and lignin (J/mol); R is the universal gas constant (8.3145 J/mol·K); and T is the temperature at which a particular stage of conversion proceeds (K). Upon substituting the reaction rate constant k from Equation (2) within the global one-step Arrhenius expression in Equation (4), we arrive at

$${\displaystyle \frac{\mathrm{d}\mathrm{\alpha }}{\mathrm{d}\mathrm{t}}}=\mathrm{A}\mathrm{\ast }\exp \left({\displaystyle \frac{-{\mathrm{E}}_{\mathrm{a}}}{\mathrm{R}\mathrm{T}}}\right)\mathrm{\ast }\mathrm{f}(\mathrm{\alpha })$$

(5)

For a specific heating rate (K/min or K/s adopted during TGA, the temperature of the system is ramped up with respect to time and is, hence, a non-isothermal process. Note: K/min is also the same as °C/min due to the same magnitude with respect to a common time frame. An increase in K by one unit is identical to an increase in °C by one unit.

$$\mathrm{\beta }={\displaystyle \frac{\mathrm{d}\mathrm{T}}{\mathrm{d}\mathrm{t}}} (\mathrm{or}) \mathrm{d}\mathrm{t}={\displaystyle \frac{\mathrm{d}\mathrm{T}}{\mathrm{\beta }}}$$

(6)

By inserting the above heat rate expression and substituting for dt in Equation (5), we arrive at

$$\mathrm{\beta }\left({\displaystyle \frac{\mathrm{d}\mathrm{\alpha }}{\mathrm{d}\mathrm{T}}}\right)=\mathrm{A}\mathrm{\ast }\exp \left({\displaystyle \frac{-{\mathrm{E}}_{\mathrm{a}}}{\mathrm{R}\mathrm{T}}}\right)\mathrm{\ast }\mathrm{f}(\mathrm{\alpha })$$

(7)

Equation (7) can be re-written as

$$\left({\displaystyle \frac{\mathrm{d}\mathrm{\alpha }}{\mathrm{d}\mathrm{T}}}\right)=\left({\displaystyle \frac{\mathrm{A}}{\mathrm{\beta }}}\right)\mathrm{\ast }\exp \left({\displaystyle \frac{-{\mathrm{E}}_{\mathrm{a}}}{\mathrm{R}\mathrm{T}}}\right)\mathrm{\ast }\mathrm{f}(\mathrm{\alpha })$$

(8)

This serves as the basic expression that could be leveraged for model-fitting and model-free methods to be built further.

2.2. Model-Fitting and Model-Free TGA Methods Rooting from Foundational Equations

Generally, to perform a kinetic analysis and calculate corresponding parameters, TGA can incorporate two approaches: model-fitting and model-free methods, which can detail a thermal decomposition process [28].

2.2.1. Model-Fitting Methods

As the name suggests, model-fitting method assumes that the experimental data obtained from any thermal decomposition reaction should fit into a pre-theorized model [29]. This is where the predicted term f(α) as in the aforementioned equations comes into play: to describe the relationship between the degree of biomass conversion or α and the overall rate of reaction, where only a single heating rate is sufficient. As outlined in Table 1, various forms of f(α) can take shape according to the type of reaction mechanism chosen for analysis that provide the output values of E_a and A [25,26,27,30].

Chemical reactions can be classified into first, second, and third order based on how the reaction rate relates to the concentrations of reactants. First-order reactions are reliant on the concentration of a single reactant, second-order reactions involve two reactants, and third-order reactions may include combinations of three reactants or elevated powers of one reactant. In biomass pyrolysis, biopolymers such as cellulose and lignin act as reactants, enabling numerous simultaneous reactions within the biomass. This study emphasizes these three orders, as higher-order reactions are uncommon and necessitate substantial energy to surpass activation energy barriers for multiple reactants to effectively collide. Furthermore, the diffusion mechanism pertains to the rate-limiting step regarding the movement of reactants, intermediates, and products through a medium [31] which can be characterized as one-dimensional, two-dimensional, or three-dimensional diffusion. One-dimensional diffusion occurs in a singular direction, like within linear, constrained porous channels. Two-dimensional diffusion takes place at surfaces or boundaries, while three-dimensional diffusion involves movement in all spatial directions, typical of bulk solid regions. In systems related to biomass pyrolysis, mass and heat transfer are affected by these types of diffusion, which are influenced by reaction time, residence time, and temperature. Phase interfacial reactions occur at the junctions between solids and liquids or gases, influenced by aspects such as adsorption, diffusion, and desorption processes [32]. The rate of these reactions depends on factors such as surface area, composition, particle size, residence time, temperature, and the presence of reactive atmospheres. In the process of converting biomass to biochar, heat transfer mechanisms, namely, convection from the carrier gas and conduction from the surrounding biomass, are crucial. This conversion can result in the creation of secondary char and pyrolytic volatiles through random nucleation and growth mechanisms, where surface superstructures evolve into new phases that may display isotropic or anisotropic qualities [33]. This transformation might depend on how long surface clusters remain on the biochar and the temperature conditions that facilitate repolymerization and polycondensation reactions, akin to the processes seen in hydrothermal carbonization (HTC)-derived biochar [34]. Given its carbonaceous nature, biochar remains inherently reactive and continues to undergo various phase interfacial reactions even after its production, which is evident by its tendency to self-ignite when exposed to heat or air adsorption.

Table 1. Probable reaction mechanisms exhibited (non-exhaustive) during biomass pyrolysis. Adapted from equations in [25,26,27,35]. Here, g(α) is the integral function of f(α).

Reaction Model	Type/Mechanism	Denotation	f(α)	g(α)
Chemical	First-order	F1	1 − α	−ln(1 − α)
	Second-order	F2	(1 − α)2	(1 − α)−1 − 1
	Third-order	F3	(1 − α)3	[(1 − α)−2 − 1]/2
Diffusion	One-dimensional	D1	0.5α	α2
	Two-dimensional	D2	[−ln(1 − α)]−1	(1 − α)ln(1 − α) + α
	Three-dimensional Jander (J)	D3	1.5(1 − α)2/3[1 − (1 − α)1/3]−1	[1 − (1 − α)1/3]2
	Three-dimensional Ginstling-Brounshtein (GB)	D4	1.5[(1 − α)−1/3 − 1]−1	(1 − 2α/3) − (1 − α)2/3
Interfacial Phase	One-dimensional	R1	1	α
	Two-dimensional	R2	2(1 − α)1/2	1 − (1 − α)1/2
	Three-dimensional	R3	3(1 − α)2/3	1 − (1 − α)1/3
Nucleation and growth	Two-dimensional	A2	2(1 − α)[-ln(1 − α)]1/2	[−ln(1 − α)]1/2
	Three-dimensional	A3	3(1 − α)[−ln(1 − α)]2/3	[−ln(1 − α)]1/3

Table 1. Probable reaction mechanisms exhibited (non-exhaustive) during biomass pyrolysis. Adapted from equations in [25,26,27,35]. Here, g(α) is the integral function of f(α).

Reaction Model	Type/Mechanism	Denotation	f(α)	g(α)
Chemical	First-order	F1	1 − α	−ln(1 − α)
	Second-order	F2	(1 − α)2	(1 − α)−1 − 1
	Third-order	F3	(1 − α)3	[(1 − α)−2 − 1]/2
Diffusion	One-dimensional	D1	0.5α	α2
	Two-dimensional	D2	[−ln(1 − α)]−1	(1 − α)ln(1 − α) + α
	Three-dimensional Jander (J)	D3	1.5(1 − α)2/3[1 − (1 − α)1/3]−1	[1 − (1 − α)1/3]2
	Three-dimensional Ginstling-Brounshtein (GB)	D4	1.5[(1 − α)−1/3 − 1]−1	(1 − 2α/3) − (1 − α)2/3
Interfacial Phase	One-dimensional	R1	1	α
	Two-dimensional	R2	2(1 − α)1/2	1 − (1 − α)1/2
	Three-dimensional	R3	3(1 − α)2/3	1 − (1 − α)1/3
Nucleation and growth	Two-dimensional	A2	2(1 − α)[-ln(1 − α)]1/2	[−ln(1 − α)]1/2
	Three-dimensional	A3	3(1 − α)[−ln(1 − α)]2/3	[−ln(1 − α)]1/3

Arrhenius Model

As seen in Equation (8), it is important to note that heating rate is brought in here and matters mostly when the temperature profile, as a function of time, is combined with the Arrhenius expression during non-isothermal reactions. This can be seen in complex reaction models to theoretically simulate real-world thermal depolymerization processes like pyrolysis, gasification, or combustion, where how much time a sample is subjected to a particular temperature can alter the conversion rate. Such models also introduce the element of non-equilibrium dynamics, such as complex heat-transfer mechanisms. On the other hand, the standalone Arrhenius equation assumes a steady-state reaction where the temperature during a conversion process is held constant (isothermal) with negligible changes. Even though this model assumes a system under equilibrium, it is not used here to imply that pyrolysis by itself, is a reversible process. However, it is highlighted here just as a relative empirical approximation that could be used to discern the critical role of heating rate as in a non-isothermal and non-equilibrium system and also to arrive at a comparison between kinetic parameters derived from the two cases. Equation (8) is simplified as

$${\displaystyle \frac{\left(\frac{\mathrm{d}\mathrm{\alpha }}{\mathrm{d}\mathrm{t}}\right)}{\mathrm{f}\left(\mathrm{\alpha }\right)}}=\mathrm{A}\mathrm{\ast }\exp \left({\displaystyle \frac{-{\mathrm{E}}_{\mathrm{a}}}{\mathrm{R}\mathrm{T}}}\right)$$

(9)

Here, by applying the natural logarithms on both sides, we conclude with the expression of the basic Arrhenius model:

$$\mathrm{l}\mathrm{n}\left({\displaystyle \frac{\left(\frac{\mathrm{d}\mathrm{\alpha }}{\mathrm{d}\mathrm{t}}\right)}{\mathrm{f}\left(\mathrm{\alpha }\right)}}\right)=\mathrm{l}\mathrm{n}\left(\mathrm{A}\right)+{\displaystyle \frac{1}{\mathrm{T}}}\left({\displaystyle \frac{-{\mathrm{E}}_{\mathrm{a}}}{\mathrm{R}}}\right) (\mathrm{or}) \mathrm{l}\mathrm{n}\left(\mathrm{k}\right)=\mathrm{l}\mathrm{n}\left(\mathrm{A}\right)+{\displaystyle \frac{1}{\mathrm{T}}}\left({\displaystyle \frac{-{\mathrm{E}}_{\mathrm{a}}}{\mathrm{R}}}\right)$$

(10)

Plotting $\mathrm{l}\mathrm{n}\left({\displaystyle \frac{\left(\frac{\mathrm{d}\mathrm{\alpha }}{\mathrm{d}\mathrm{t}}\right)}{\mathrm{f}\left(\mathrm{\alpha }\right)}}\right)$ or ln(k) versus (1/T) results in a straight line with a negative slope of (−E_a/R), while the y-axis intercept can be used to directly calculate A [36]. With respect to $\mathrm{l}\mathrm{n}\left({\displaystyle \frac{\left(\frac{\mathrm{d}\mathrm{\alpha }}{\mathrm{d}\mathrm{t}}\right)}{\mathrm{f}\left(\mathrm{\alpha }\right)}}\right)$, different reaction models and the equations for calculating their corresponding f(α) are described in Table 1.

Coats-Redfern (CR) Model

The CR method is a model-fitting technique used in chemical kinetics to study thermal decomposition of solid biomass matter, irrespective of woody or non-woody nature [37]. It is widely used by researchers for analyzing the pyrolysis of lignocellulosic materials. The CR model functions as an integral of Equation (1). To understand how we land at the CR expression and for calculation clarity, [26] state that a primary model function for biomass pyrolysis kinetics can be interpreted as

$$\mathrm{f}\left(\mathrm{\alpha }\right)={(1-\mathrm{\alpha })}^{\mathrm{n}}$$

(11)

where n is the order of reaction (1, 2, 3, …, n). Also, unlike the Arrhenius model, the CR model does not assume a steady-state or isothermal model and can function effectively by following different heating rates (β). So, by substituting Equation (11) in Equation (8),

$$\left({\displaystyle \frac{\mathrm{d}\mathrm{\alpha }}{\mathrm{d}\mathrm{T}}}\right)=\left({\displaystyle \frac{\mathrm{A}}{\mathrm{\beta }}}\right)\mathrm{\ast }\exp \left({\displaystyle \frac{-{\mathrm{E}}_{\mathrm{a}}}{\mathrm{R}\mathrm{T}}}\right)\mathrm{\ast }{(1-\mathrm{\alpha })}^{\mathrm{n}}$$

(12)

Upon rearranging this equation, it can be formulated as

$$\left({\displaystyle \frac{\mathrm{d}\mathrm{\alpha }}{{(1-\mathrm{\alpha })}^{\mathrm{n}}}}\right)=\left({\displaystyle \frac{\mathrm{A}}{\mathrm{\beta }}}\right)\mathrm{\ast }\exp \left({\displaystyle \frac{-{\mathrm{E}}_{\mathrm{a}}}{\mathrm{R}\mathrm{T}}}\right)\mathrm{\ast }\mathrm{d}\mathrm{T}$$

(13)

Now, introduction of the term g(α), i.e., the integral conversion of f(α) linking the extent of reaction to temperature and time, takes precedence, which is based on a specific reaction mechanism encountered by biomass during pyrolysis. Integrating within the limits of α = 0 and α = α at T = 0 and T = T, respectively [35,38] generates the following:

$$\mathrm{g}\left(\mathrm{\alpha }\right)={\int }_0^{\mathrm{\alpha }}\left({\displaystyle \frac{\mathrm{d}\mathrm{\alpha }}{{(1-\mathrm{\alpha })}^{\mathrm{n}}}}\right)={\int }_0^{\mathrm{T}}\left({\displaystyle \frac{\mathrm{A}}{\mathrm{\beta }}}\right)\mathrm{\ast }\exp \left({\displaystyle \frac{-{\mathrm{E}}_{\mathrm{a}}}{\mathrm{R}\mathrm{T}}}\right)\mathrm{\ast }\mathrm{d}\mathrm{T}$$

(14)

After integrating Equation (14) on the LHS first,

$${\displaystyle \frac{1-{(1-\mathrm{\alpha })}^{1-\mathrm{n}}}{1-\mathrm{n}}}=\left({\displaystyle \frac{\mathrm{A}}{\mathrm{\beta }}}\right){\int }_{{\mathrm{T}}_0}^{\mathrm{T}}\exp \left({\displaystyle \frac{-{\mathrm{E}}_{\mathrm{a}}}{\mathrm{R}\mathrm{T}}}\right)\mathrm{\ast }\mathrm{d}\mathrm{T}$$

(15)

Following the recommendations by [35], after integration on RHS, Equation (15) becomes

$${\displaystyle \frac{1-{(1-\mathrm{\alpha })}^{1-\mathrm{n}}}{1-\mathrm{n}}}=\left({\displaystyle \frac{\mathrm{A}\mathrm{R}{\mathrm{T}}^2}{\mathrm{\beta }{\mathrm{E}}_{\mathrm{a}}}}\right)\mathrm{\ast }\left(1-{\displaystyle \frac{2\mathrm{R}\mathrm{T}}{{\mathrm{E}}_{\mathrm{a}}}}\right)\mathrm{\ast }\exp \left({\displaystyle \frac{-{\mathrm{E}}_{\mathrm{a}}}{\mathrm{R}\mathrm{T}}}\right)$$

(16)

Rearranging Equation (16),

$${\displaystyle \frac{1-{(1-\mathrm{\alpha })}^{1-\mathrm{n}}}{{\mathrm{T}}^2(1-\mathrm{n})}}=\left({\displaystyle \frac{\mathrm{A}\mathrm{R}}{\mathrm{\beta }{\mathrm{E}}_{\mathrm{a}}}}\right)\mathrm{\ast }\left(1-{\displaystyle \frac{2\mathrm{R}\mathrm{T}}{{\mathrm{E}}_{\mathrm{a}}}}\right)\mathrm{\ast }\exp \left({\displaystyle \frac{-{\mathrm{E}}_{\mathrm{a}}}{\mathrm{R}\mathrm{T}}}\right)$$

(17)

As shown by [38], assuming that $\left(1-{\displaystyle \frac{2\mathrm{R}\mathrm{T}}{{\mathrm{E}}_{\mathrm{a}}}}\right)\ll 1$ and can be neglected, Equation (17) can be reframed as

$${\displaystyle \frac{1}{{\mathrm{T}}^2}}\mathrm{\ast }{\displaystyle \frac{1-{(1-\mathrm{\alpha })}^{1-\mathrm{n}}}{(1-\mathrm{n})}}=\left({\displaystyle \frac{\mathrm{A}\mathrm{R}}{\mathrm{\beta }{\mathrm{E}}_{\mathrm{a}}}}\right)\mathrm{\ast }\exp \left({\displaystyle \frac{-{\mathrm{E}}_{\mathrm{a}}}{\mathrm{R}\mathrm{T}}}\right)$$

(18)

Applying natural logarithms on both sides of Equation (18), we get

$$\mathrm{l}\mathrm{n}\left({\displaystyle \frac{1}{{\mathrm{T}}^2}}\mathrm{\ast }{\displaystyle \frac{1-{(1-\mathrm{\alpha })}^{1-\mathrm{n}}}{(1-\mathrm{n})}}\right)=\mathrm{l}\mathrm{n}\left({\displaystyle \frac{\mathrm{A}\mathrm{R}}{\mathrm{\beta }{\mathrm{E}}_{\mathrm{a}}}}\right)+{\displaystyle \frac{1}{\mathrm{T}}}\left({\displaystyle \frac{{-\mathrm{E}}_{\mathrm{a}}}{\mathrm{R}}}\right)$$

(19)

Albeit, the above equation is valid for n≠1. According to [39], pyrolysis of lignin-based solids is a first-order reaction, where n = 1. Moreover, according to [40], CR method adopts the basis of a single-stage first-order reaction during thermochemical breakdown of a solid fuel. Thus, Equation (19) becomes

$$\mathrm{l}\mathrm{n}\left({\displaystyle \frac{-\mathrm{l}\mathrm{n}(1-\mathrm{\alpha })}{{\mathrm{T}}^2}}\right)=\mathrm{l}\mathrm{n}\left({\displaystyle \frac{\mathrm{A}\mathrm{R}}{\mathrm{\beta }{\mathrm{E}}_{\mathrm{a}}}}\right)+{\displaystyle \frac{1}{\mathrm{T}}}\left({\displaystyle \frac{{-\mathrm{E}}_{\mathrm{a}}}{\mathrm{R}}}\right)$$

(20)

By inserting the integral function once again in Equation (20), we can simplify the expression as

$$\mathrm{l}\mathrm{n}\left({\displaystyle \frac{\mathrm{g}\left(\mathrm{\alpha }\right)}{{\mathrm{T}}^2}}\right)=\mathrm{l}\mathrm{n}\left({\displaystyle \frac{\mathrm{A}\mathrm{R}}{\mathrm{\beta }{\mathrm{E}}_{\mathrm{a}}}}\right)+{\displaystyle \frac{1}{\mathrm{T}}}\left({\displaystyle \frac{{-\mathrm{E}}_{\mathrm{a}}}{\mathrm{R}}}\right)$$

(21)

For different reaction models (chemical reaction, diffusion, phase interfacial, nucleation, and growth) and associated mechanisms, the corresponding value of g(α) can be used from Table 1 to calculate kinetic parameters. Plotting $\mathrm{l}\mathrm{n}\left({\displaystyle \frac{\mathrm{g}\left(\mathrm{\alpha }\right)}{{\mathrm{T}}^2}}\right)$ versus $\left({\displaystyle \frac{1}{\mathrm{T}}}\right)$ gives a straight line with a negative slope equivalent to $\left({\displaystyle \frac{{-\mathrm{E}}_{\mathrm{a}}}{\mathrm{R}}}\right)$. With the known value of E_a, it is possible to calculate A using the y-axis intercept $\mathrm{l}\mathrm{n}\left({\displaystyle \frac{\mathrm{A}\mathrm{R}}{\mathrm{\beta }{\mathrm{E}}_{\mathrm{a}}}}\right)$.

Four key shortcomings of model-fitting methods are as follows: (a) the value of E_a is subjected to heavily depend on the reaction model assumed, when in reality, there is significant possibility of a process to manifest itself via several other models; (b) secondly, if the wrong model is chosen, the reliability of the measured kinetic parameter value can be questionable; (c) thirdly, a model-fitting method like CR may function harmoniously only at one specific heating rate at a given time where different reaction mechanisms can be tested under one roof for satisfactoriness rather than several heating rates simultaneously, as in iso-conversional models; (d) fourthly, despite providing an overview of E_a under many assumed models, it is a highly time-consuming method. Henceforth, while using model-fitting methods, it is advisable to assume as many reaction models as possible to help in quantitative analysis of E_a from more than a single perspective. In fact, it may be safe to say that convoluted processes such as pyrolysis can exhibit a myriad of simultaneous (parallel), sequential or both types of reactions at once. When using this approach, it could be possible to also narrow down on the ideal model using correlation coefficient (R²) and simple cross-comparison.

2.2.2. Model-Free or Iso-Conversional Methods

Model-free kinetic methods do not assume that a chemical reaction progresses as per a pre-selected model under only a single heating rate. It considers the possibility of addressing complex chemical reactions at multiple heating and conversion rates [41]. Here, different values of α (in the range of 0–1) are used to calculate E_a without the necessity for g(α) intervention, thereby offering greater flexibility for the method [42]. It is here that common slow pyrolysis heating rates are employed (10–40 °C/min in our case) in the presence of nitrogen to detect variations in E_a at each stage of α. Different values of α are obtained in different phases of pyrolysis, due to which kinetic parameters under variable heating rates need to be compared and fitted separately one after the other [39]. In other words, we will be plotting for one α value (e.g., α = 0.1) at all four heating rates (10, 20, 30, and 40 °C/min), proceed to the next conversion value (e.g., α = 0.2) at the four heating rates, and so on till α = 1. Variations in extracted E_a, for instance, may indicate that biomass decomposition in a pyrolysis-like setting encompasses an interplay of many different reaction mechanisms and not just one that improves overall dependability and robustness of the derived values of E_a. Although it is not necessary to use g(α) to calculate E_a, model-free methods need this function to calculate A and still depend upon a specific reaction model and mechanism. This is, without a doubt, a notable drawback of model-free methods that accentuates a limitation in fully characterizing both kinetic parameters without the role of assumptions. Also, model-free methods are criticised since they are too simplistic and general in describing a complex process like pyrolysis [26].

There are different model-free methods used in practice of approximating E_a of pyrolysis reactions, like FM, KAS, OFW, Vyazovkin (VZ), and distributed activation energy model (DAEM), that offer high simplicity during calculations. Among these, FM, KAS, and OFW are used in this study as representative methods. They are able to analyze complex reactions as a function of biomass conversion value, α. Using the expressions studied by [43,44], the following

Table 2. Model-free or isoconversional methods.

Model-Free Method	Equation		Procedure for Plotting
OFW	$\mathrm{l}\mathrm{n}\left(\mathrm{\beta }\right)=\mathrm{l}\mathrm{n}\left({\displaystyle \frac{\mathrm{A}{\mathrm{E}}_{\mathrm{a}}}{\mathrm{R}\mathrm{g}\left(\mathrm{\alpha }\right)}}\right)-2.315-0.457{\displaystyle \frac{{\mathrm{E}}_{\mathrm{a}}}{\mathrm{R}\mathrm{T}}}$	(1)	Plot $\mathrm{l}\mathrm{n}\left(\mathrm{\beta }\right)$ versus ${\displaystyle \frac{1}{\mathrm{T}}}$ to obtain a straight line with a negative slope of $-0.457{\displaystyle \frac{{\mathrm{E}}_{\mathrm{a}}}{\mathrm{R}}}$ to calculate Ea. Using this and an assumed reaction model/mechanism for g(α), A can be calculated from the y-axis intercept, $\mathrm{l}\mathrm{n}\left({\displaystyle \frac{\mathrm{A}{\mathrm{E}}_{\mathrm{a}}}{\mathrm{R}\mathrm{g}\left(\mathrm{\alpha }\right)}}\right)$.
KAS	$\mathrm{l}\mathrm{n}\left({\displaystyle \frac{\mathrm{\beta }}{{\mathrm{T}}^2}}\right)=\mathrm{l}\mathrm{n}\left({\displaystyle \frac{\mathrm{A}{\mathrm{E}}_{\mathrm{a}}}{\mathrm{R}\mathrm{g}\left(\mathrm{\alpha }\right)}}\right)-{\displaystyle \frac{{\mathrm{E}}_{\mathrm{a}}}{\mathrm{R}\mathrm{T}}}$	(2)	Plot $\mathrm{l}\mathrm{n}\left({\displaystyle \frac{\mathrm{\beta }}{{\mathrm{T}}^2}}\right)$ versus ${\displaystyle \frac{1}{\mathrm{T}}}$ to obtain a straight line with a negative slope of $-{\displaystyle \frac{{\mathrm{E}}_{\mathrm{a}}}{\mathrm{R}}}$ to calculate Ea. Using this and an assumed reaction model/mechanism for g(α), A can be calculated from the y-axis intercept, $\mathrm{l}\mathrm{n}\left({\displaystyle \frac{\mathrm{A}{\mathrm{E}}_{\mathrm{a}}}{\mathrm{R}\mathrm{g}\left(\mathrm{\alpha }\right)}}\right)$.
FM	$\mathrm{l}\mathrm{n}\mathrm{\beta }\left({\displaystyle \frac{\mathrm{d}\mathrm{\alpha }}{\mathrm{d}\mathrm{T}}}\right)=\mathrm{l}\mathrm{n}\left[{\left(\mathrm{A}\mathrm{f}\right(\mathrm{\alpha }\left)\right)}^{\mathrm{n}}\right]-{\displaystyle \frac{{\mathrm{E}}_{\mathrm{a}}}{\mathrm{R}\mathrm{T}}}$ (or) $\mathrm{l}\mathrm{n}\left({\displaystyle \frac{\mathrm{d}\mathrm{\alpha }}{\mathrm{d}\mathrm{t}}}\right)=\mathrm{l}\mathrm{n}\left[{\left(\mathrm{A}\mathrm{f}\right(\mathrm{\alpha }\left)\right)}^{\mathrm{n}}\right]-{\displaystyle \frac{{\mathrm{E}}_{\mathrm{a}}}{\mathrm{R}\mathrm{T}}}$	(3)	Plot $\mathrm{l}\mathrm{n}\mathrm{\beta }\left({\displaystyle \frac{\mathrm{d}\mathrm{\alpha }}{\mathrm{d}\mathrm{T}}}\right)$ or $\mathrm{l}\mathrm{n}\left({\displaystyle \frac{\mathrm{d}\mathrm{\alpha }}{\mathrm{d}\mathrm{t}}}\right)$ versus ${\displaystyle \frac{1}{\mathrm{T}}}$ to obtain a straight line with a negative slope of $-{\displaystyle \frac{{\mathrm{E}}_{\mathrm{a}}}{\mathrm{R}}}$ to calculate Ea. Using this and an assumed reaction model/mechanism for f(α), A can be calculated from the y-axis intercept, $\mathrm{l}\mathrm{n}\left[{\left(\mathrm{A}\mathrm{f}\right(\mathrm{\alpha }\left)\right)}^{\mathrm{n}}\right]$.

illustrates the expression for these methods, values of constants/variables, and the procedure for plotting to evaluate kinetic parameters.

2.3. Thermodynamic Study

During pyrolysis, thermodynamic parameters such as ΔH, ΔS, and ΔG are calculated using the following equations adapted from [45].

$$∆\mathrm{H}={\mathrm{E}}_{\mathrm{a}}-\mathrm{R}{\mathrm{T}}_{\mathrm{m}}$$

(22)

$$∆\mathrm{G}={\mathrm{E}}_{\mathrm{a}}-\mathrm{R}{\mathrm{T}}_{\mathrm{m}}\mathrm{l}\mathrm{n}\left({\displaystyle \frac{{\mathrm{K}}_{\mathrm{B}}\times {\mathrm{T}}_{\mathrm{m}}}{\mathrm{h}\mathrm{A}}}\right)$$

(23)

$$∆\mathrm{S}=\left({\displaystyle \frac{∆\mathrm{H}-∆\mathrm{G}}{{\mathrm{T}}_{\mathrm{m}}}}\right)$$

(24)

In the above equations, T_m is the maximum thermal decomposition temperature of CRD wood biomass that was obtained from DTG curves at each heating rate (10–40 °C/min). E_a is the activation energy derived using isoconversional models of OFW, KAS, and FM at each heating rate and conversion stage. Pre-exponential factor, A, was also calculated from the respective plot using the intercept. K_B is the Boltzmann constant of 1.381 $\times$ 10⁻²³ J/K, and h is the Plank constant, 6.626 $\times$ 10⁻³⁴ J/s.

Equation (25) is however, a general and simplified expression that could be used to calculate the thermodynamic parameters, and it may not cover all intricacies posed by multi-step biomass pyrolysis reactions. Rather, it conservatively represents a general overview of the wood matrix decomposition. If over-simplification of pyrolysis occurs, especially by neglecting critical energy barriers, activation energy calculations could be erroneous, inevitably influencing entropies since they are very small.

2.4. CRD Wood Biomass Sample Preparation for Characterization and Kinetic Analysis

Non-recyclable CRD wood sourced from “BRQ Fibre et Broyure” in Trois-Rivières, Québec, was delivered in pails to the Innofibre facility Figure 1A. After collecting, the CRD wood underwent mechanical sieving using a Labtech vibrational sieving unit to separate it into different particle sizes (>4.5 cm till >0.3 cm) after screening and removal of contaminants (Figure 1B). In fact, no washing was carried out due to the risk of creating contaminated wastewater from the leaching of impurities present in the CRD wood. Once the size separation was completed, the selected biomass fraction in Figure 1C was dried for an entire night at 105 °C. Given that sieving only led to particle sizes of up to 0.3 cm, it was necessary to mill the CRD wood further. A Retsch SM300 mill (3 kW motor and 100–3000 rpm), featuring a high-speed tungsten carbide rotor and screens with openings as small as 500 µm, or a Thomas Wiley mill with a similar mesh size, was utilized to produce fine particles, as shown in Figure 1D, from the biomass for subsequent analysis.

2.5. Physicochemical Characterization

Proximate analysis was employed to evaluate ash content, volatile carbon (VC), and fixed carbon (FC) in biomass. For the ash measurement, 1 g of dry biomass fines was placed into a crucible and heated in a muffle furnace. Beginning at room temperature, the sample was heated to 106 °C at a rate of 5 °C/min and maintained at this temperature for 1 h to ensure that most residual moisture is removed. Then, the temperature was increased at the same rate from 106 °C to 550 °C to decompose labile extractives and organics under controlled heat transfer. At 550 °C, the wood sample was held for a duration of 4–6 h for further degradation of any combustible biomass components. The sample was then allowed to cool to room temperature in a desiccator to prevent any hygroscopic effects, after which the ash was weighed to ascertain the proportion of biomass remaining. For volatile carbon, 1 g of dry biomass fines was subjected to heating at 900 °C for 7 min in a sealed crucible, after which it was cooled and weighed to quantify the carbon lost. VC is utilized in certain contexts in this study to specifically refer to labile carbon, distinguishing it from other volatile components. FC was calculated using the relation FC = 100 − (VC + ash) as weight percent dry basis (wt%DB), where VC and ash are also represented as wt%DB [46].

For the ultimate analysis, an Elementar Vario Macro Cube examined 1–200 mg of dry biomass fines for elements such as carbon, hydrogen, nitrogen, and sulfur, utilizing helium (purity > 99.996%) as the carrier gas and oxygen (purity > 99.996%) for combustion up to 1200 °C. The oxygen content was determined by calculating the difference from 100%, using the percentages of all individual elements and ash, as given in [$\mathrm{o}\mathrm{x}\mathrm{y}\mathrm{g}\mathrm{e}\mathrm{n}=100-(\mathrm{c}\mathrm{a}\mathrm{r}\mathrm{b}\mathrm{o}\mathrm{n}+\mathrm{h}\mathrm{y}\mathrm{d}\mathrm{r}\mathrm{o}\mathrm{g}\mathrm{e}\mathrm{n}+\mathrm{n}\mathrm{i}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{g}\mathrm{e}\mathrm{n}+\mathrm{s}\mathrm{u}\mathrm{l}\mathrm{f}\mathrm{u}\mathrm{r}+\mathrm{a}\mathrm{s}\mathrm{h})]$. It is derived from [47].

The higher heating value (HHV) of CRD wood biomass was determined using a similar approach outlined by [48]. This approach correlated the calorific value with elemental composition obtained from ultimate analysis of the biomass sample. For metal detection in biomass, an Agilent Technologies 4210 Microwave Plasma Atomic Emission Spectrometer (MP-AES) was utilized, which featured a nitrogen fuel source and an SPS 4 autosampler. This apparatus allowed for the simultaneous analysis of multiple elements by aerosolizing the biomass sample into a nitrogen plasma to produce monoatomic ions. The device guaranteed accurate measurements of the inorganic content by completely digesting the organic materials found in the biomass.

2.6. Composition Analysis

The presence of extractives, structural carbohydrates, and lignin in CRD wood biomass were analyzed according to National Renewable Energy Laboratory (NREL, United States of America Department of Energy) procedures as described in [49,50] respectively.

2.7. FTIR

The distribution of surface functional groups in biomass was examined using an Agilent Technologies Cary 630 spectroscope, with Microlab 5.8 and Microlab Expert 1.3 for operation. About 1 g of dry biomass fines was placed on a diamond crystal surface that was regularly cleaned with ethanol. The analysis was performed in transmittance mode over a spectral range of 500–4000 cm⁻¹ to determine the chemical bonds present in the material. Baseline and peak corrections were applied to the obtained spectra, followed by signal smoothing.

2.8. TGA

In order to examine the weight loss pattern of biomass in relation to temperature, a PerkinElmer TGA 8000 analyzer was utilized. This device was also able to assess proximate analysis parameters such as moisture, VC, FC, and ash. Only 8 mg of biomass fines in their original, non-dried state were needed per vial for the analyzer. Since TGA was leveraged to analyze the pyrolysis kinetics of CRD wood, the process gas used during heating was compressed nitrogen with a purity level exceeding 99.996%. The system was capable of accurately reaching temperatures up to 1200 °C with adjustable heating rates. For this research, 10, 20, 30, and 40 °C/min were employed, reaching a maximum temperature of 900 °C to investigate the transformation of biomass to biochar. Differential thermogravimetric (DTG) curves were generated by differentiating the weight loss features from the original thermogram in relation to time and plotting them against temperature. This led to the identification of different peaks indicating the points of maximum weight loss or melting of the sample during TGA, along with the corresponding temperatures, providing insights into biomass chemical composition and hydrocarbon structure.

3. Results and Discussion

3.1. Physicochemical Properties of CRD Wood—Comparative Analysis with Other Biomass Feedstock

The proximate and ultimate analysis parameters of CRD wood biomass were analyzed, as shown in

Table 3. Physicochemical properties of CRD wood in comparison with different biomass sources. References are mentioned below the characterization results in the table.

Biomass	Pinewood	Rice Husk	Beech	Wheat Straw	Miscanthus *	Madhuca Longifolia	Poplar Sawdust *	Palm Kernel *	MSW	CRD Wood *
Proximate analysis
Moisture (wt%)	4.01	10.10	15.20	11.63	---	5.98	---	---	2.75	---
Volatiles (wt%)	64.37	52.30	84.87	65.32	83.30	71.26	83.60	79.68	51.13	82.35
FC (wt%)	24.89	23.40	14.53	15.17	14.40	18.24	15.40	16.78	37.86	16.48
Ash (wt%)	6.73	14.20	0.60	7.88	2.30	4.50	1.00	3.54	8.26	1.17
Ultimate analysis
C (wt%)	54.53	34.99	49.38	44.12	47.50	47.20	49.47	43.84	76.26	49.88
H (wt%)	6.56	4.58	6.17	6.34	6.10	5.90	5.89	6.13	6.88	6.12
S (wt%)	0.09	---	0.01	---	0.10	1.20	0.05	0.06	0.28	0.10
N (wt%)	0.94	1.95	0.28	0.63	1.00	3.60	1.40	3.11	3.32	0.99
O (wt%)	31.14	34.18	43.55	39.99	45.30	42.10	43.12	46.86	13.25	41.74
Van-Krevelen parameters
H/C	1.44	1.57	1.50	1.72	1.54	1.50	1.43	1.68	1.07	1.47
O/C	0.43	0.73	0.66	0.68	0.72	0.67	0.65	0.80	0.13	0.63
Calorific value
HHV (MJ/kg)	22.33	14.09	19.3	15.29	15.82	---	19.50	16.83	29.70	20.28
Reference	[51]	[52]	[53]	[54]	[55,56]	[57]	[58]	[59,60]	[61]	Present study

* Expressed in wt%DB.

. The elemental composition, particularly carbon content in CRD wood, resembled hardwoods like beech and poplar, reaching 49–50% despite containing softwood residues in the bale. In terms of carbon content, MSW outperformed all the feedstocks, with a carbon content of 76.26% due to a heterogeneous composition involving plastics and other hydrocarbon-rich residues. Oxygen (41.74%) and hydrogen (6.12%) were also equivalent to the ranges in hardwood species. These bolster the role of CRD wood biomass in bioenergy applications for the production of biofuels, biochemicals, and bioproducts. Low sulfur (from gypsum dust, if any) and nitrogen content in CRD wood also means that the release of NOx and SOx could be starved during pyrolysis. For proximate parameters, CRD wood is a low-ash biomass at 1.17% relative to high-ash agricultural residues like rice husk at 14.20% or medium-ash sources like MSW (8.26%).

However, as delineated in Figure 2, the metal content in CRD wood, mainly alkaline and alkaline earth metals (AAEM) (86–87%), was very high which may result in secondary reactions by acting as indirect catalysts, leading to excess loss of carbon or technical snags in reactor equipment through slagging, fouling, or bed agglomeration [62,63]. However, among AAEM, alkaline earth metals like Ca, Ba, and Mg constituted 59.96%, whereas alkaline metals made up just 27.03%. So, there could be a possibility where the melting of ash may be pushed toward relatively higher temperatures since the presence of divalent species dominated this AAEM fraction. On the contrary, the other possibility with high AAEM could be secondary char formation [64]. AAEM, when present with woody biomass under pyrolysis temperatures of 350–750 °C, may increase the yield of biochar through secondary reactions [65]. This holds true for an AAEM-rich feedstock like CRD wood. Additionally, CRD wood contains some heavy metallic contaminants like Cu, Cr, and mild levels of As. Other heavy metals detected were mainly Zn and Pb. All these metals are part of preservatives, paints, varnishes, roofings, and surface coatings in conditioned CRD wood that are usually part of residential buildings, electricity poles, and handrails dating back to the 1960s–1990s. The VC in CRD wood (82.35%) matches VC in poplar (83.60%) and beech (84.87%), confirming highly decomposable fractions. Also, due to high volatiles and low ash, woody biomass could produce biochar product with high FC [66]. Also, due to the lower ash content that minimizes the presence of non-combustible materials, the HHV of CRD wood biomass is relatively high, at 20.28 MJ/kg, rendering it suitable for bioenergy applications [67]. On the contrary, a high ash rice husk had a low HHV at 14.09 MJ/kg. Wheat straw with moderate ash also had a low HHV of 15.29 MJ/kg. Another pattern observed is that as volatiles increase, HHV decreases due to the non-availability of stable carbon. This was true for MSW (51.13%) and pinewood (64.37%) relative to CRD wood (82.35%), where the increase in volatiles decreased HHV from 29.70 MJ/kg, 22.33 MJ/kg, and 20.28%, respectively.

3.2. CRD Wood Compositional Analysis

To achieve a representative and robust analysis of CRD wood biomass received, samples from four different sections of the pail, post-sample preparation steps, were subjected to compositional analysis. Percentages of extractives, holocellulose, and lignin are listed in

Table 4. Biochemical composition of CRD wood lots from the same bale.

Biomass Constituents	Lot 1	Lot 2	Lot 3	Lot 4
Hot Water Extractives (wt%)	3.2	2.0	3.1	2.7
Ethanol Extractives (wt%)	3.7	3.0	2.8	4.0
Total extractives (wt%)	6.9	5.0	5.9	6.7
Insoluble Lignin (wt%)	19.7	20.5	26.1	25.3
Total Lignin (wt%)	24.9	26.2	30.2	29.7
Glucan (wt%)	50.9	44.2	46.9	53.1
Xylan (wt%)	19.2	23.2	8.0	10.1
Arabinan (wt%)	0.0	0.1	0.9	1.3
Galactan (wt%)	0.6	0.8	2.5	2.4
Mannan (wt%)	2.0	2.3	9.7	10.0
Total holocellulose (wt%)	72.7	70.6	68.0	76.9

, which may explain the nature of biomass transformation to biochar. Extractives soluble in hot water and ethanol ranged from 2 to 3.5% and 2.5 to 4%, respectively, which are usually eliminated as volatiles. Since total lignin varied from 24 to 31%, it is conclusive that CRD wood is composed of both hard and softwood and may be called "wood waste" that has a comparable lignin content, recorded at 28.9%, as shown by [68]. As shown by [69], biochar produced from agriculture wastes like buckwheat husk (holocellulose: 73.72%, lignin: 24.73%), spent coffee grounds (holocellulose: 31.58%, lignin: 23.65%), hazelnut shell (holocellulose: 62.31%, lignin: 28.30%), and waste grass biomass like hemp (holocellulose: 25.81%, lignin: 20.69%) and miscanthus (holocellulose: 63.11%. lignin: 25.48%) were also compared. The greater the lignin, the greater the structural complexity due to the presence of aromatics, and the greater the biochar yields [70]. However, the lignin content in biochar is not a solely conclusive indicator of the mechanical strength/integrity and yield of biochar [71,72]. However, in lignin, if softened and melted to form condensed aromatic structures, the overall bulk density and resistance to breakage may increase [62]. The carbon content and stability of biochar could be largely conserved as well. The total holocellulose content in CRD wood varied from 68 to 77%. With six-carbon sugars like glucose (44–54%), galactose (0.5–2.5%), and mannose (2–10%), and five-carbon sugars like xylose (8–24%) and arabinose (0–1.5%), hemicellulose decomposition in CRD wood to pyrolysis gases at lower temperatures is plausible. The overall contribution to char formation is low for hemicellulose. Cellulose decomposition could favor both condensed volatiles and intermediates (high temperatures) and some char (lower temperatures) formation. This is determined by the temperature, BRT, and heating rate employed during the process [73].

3.3. FTIR Spectroscopy Analysis

The FTIR spectral distribution of the CRD wood biomass is shown in Figure 3. A high concentration of specific functional groups in the CRD wood biomass feed is denoted by strong transmittance peaks of infrared light through the sample, which demonstrate specific vibrational patterns: stretching or bending. Spectral band assignments from [74,75] were used to match the transmittance peaks in CRD wood biomass to relevant functionalities. Firstly, the distinct peak at 3320 cm⁻¹ indicates O-H stretching vibrations, as in moisture, carboxylic acids, alcohols, and phenols. Secondly, the distinct peak at 2892 cm⁻¹ could signify C-H stretching in aliphatic hydrocarbons, as in alkanes that are subunits of most biomass polymers. Thirdly, the spectral band at 2003 cm⁻¹ highlights stretching vibrations in alkynes (C≡C) that may be ascribed to polyaromatics with maximal carbon condensation. Next, a carbonyl stretch (C=O) in aldehydes, ketones, esters, and carboxylic acids appears around 1730 cm⁻¹, which stems from hemicellulose in CRD wood. This contributes to the loss of biomass carbon as CO and CO₂ during the onset of pyrolysis. A subdued peak at 1503 cm⁻¹ and 1265 cm⁻¹ means there are faint aromatic C-C and aromatic C-O ring stretching actions, respectively, amidst the biomass carbon skeleton. Methyl, methylene, or methoxy groups of biomass lignin also exhibit these vibrations owing to the presence of dense organic residues in biomass. A sharp band at 1022 cm⁻¹ may be due to C-O stretching in aliphatic ethers (C-O-R) and alcohols (C-O-H) that are key building blocks of carbohydrates/polysaccharides. The 799 cm⁻¹ band could be an extractive in biomass, showcasing the C-H bend in alkynes, phenyl rings, or aromatics, followed by the band at 664 cm⁻¹ stemming from C-H vibrations in alkenes.

3.4. TGA Analysis and Effect of Heating Rate

The thermal stability of CRD wood biomass was verified by TGA and DTG curves. The weight loss exhibited by biomass under four different heating rates, 10, 20, 30, and 40 °C/min, along with their stages of thermal degradation under an inert nitrogen atmosphere mimicking pyrolysis, were studied, as shown in Figure 4. In the TGA plot in Figure 4A, it is evident that an increase in the heating rate pushes the degradation temperature slightly higher (to the right), as reported by [76]. This is because an increase in the heating rate means that there is less time for heat to permeate from the surface to the core of the biomass particle, and that too, only non-linearly. So, the weight loss encountered is restricted as a surface effect rather than occurring throughout the particle and is hence low. At a lower heating rate, an inverse phenomenon, where there is gradual heat transfer that devolatilizes biomass constituents to fruition over an ample time frame, is clear. The biomass’s outer layer and the core inside reach almost similar temperatures, allowing a linear thermal behavior. This is validated by the maximum weight loss (72.73%) between 200 and 400 °C for the heating rate of 10 °C/min, followed by 69.45%, 64.95%, and 60.85% for the heating rates of 20, 30, and 40 °C/min, respectively. However, despite an increase in the heating rate by every 10 °C/min, the profile for degradation curves is unchanged, as concurred by [77].

If we look more closely at the DTG peaks for these heating rates, weight loss happens in three stages. The first stage occurs < 150 °C, where evaporation of moisture and other surface-bound volatiles occurs [78]. Around 5–7% weight loss happened here. As seen in Figure 4B,C, within the second stage from 200 to 400 °C, the melting temperatures of biomass (T_m) for each heating rate was observed to increase with an increase in the heating rate (367.41 °C for 10 °C/min to 389.62 °C for 40 °C/min), validating our earlier finding that greater BRT at that specific temperature means that there is adequate time for biomass to experience loss of structural integrity by efficiently exposing its biopolymers for cleavage, signifying an ongoing pyrolysis process [79]. This region encompasses extractives, VOCs, hemicellulose, and cellulose breakdown, respectively, with peak decomposition mainly between 350 and 400 °C. Since cellulose is a straight-chain, crystalline polymer, it needs greater energy to collapse and, hence, disintegrates after relatively less stable hemicellulose degrades to a large extent. Biomass carbon (e.g., carbon loss via CO₂ and CO) and heteroatom (e.g., H₂S, SO₂, and NH₃) rejection reactions from holocellulose, followed by the simultaneous heat uptake to initiate destabilization of lignin aromatics (e.g., phenols), are key reactions that increase the yield of gas and condensible volatiles. In the conversion (α) plot in Figure 4D, we can see that 80% (α = 0.8) is completed around 400 °C. In this second stage, profound mass loss of 60–73% took place. Beyond this point, the third stage of decomposition is rather slow and proceeds well-spaced out due to complex lignin fractions and other high-molecular-weight compounds. Moreover, since CRD wood contains noticeable AAEM and other metals/inorganics in its ash fraction, it could be stable up to a certain level and delay breakdown. This is when char generation peaks due to both primary and secondary pyrolysis reactions. It is safe to say that no additional peaks were observed, which means that significant degradation of inorganic material may not be accounted for. At the end of the process, about 20% of the initial mass of CRD wood remains. All in all, there was another interesting observation. At higher heating rates of 30 and 40 °C/min, decomposition peaks appear to overlap, which indicates that the degradation of biomass polymers may not be completely separated processes and could occur sharply within a narrow temperature range and as synchronous steps. As concluded by [76,80], the thermal decomposition patterns remained fairly consistent across all the heating rates tested.

3.5. Kinetic Analysis of CRD Wood Pyrolysis

3.5.1. Model-Fitting Method

Being a model-fitting technique, the CR method fits TGA data of biomass pyrolysis to various kinetic models used for calculating the activation energies and pre-exponential factors. The calculations in

Table 5. Activation energy (E_a), pre-exponential factor (A), and model correlation (R²) of different pyrolysis reaction models (albeit, non-exhaustive list) evaluated using the CR method.

Reaction Model	Type/Mechanism	Heating Rate = 10 °C/min			Heating Rate = 20 °C/min			Heating Rate = 30 °C/min			Heating Rate = 40 °C/min
		Activation Energy, Ea (kJ/mol)	Pre-Exponential Factor, A (min−1)	R2 from Plot	Activation Energy, Ea (kJ/mol)	Pre-Exponential Factor, A (min−1)	R2 from Plot	Activation Energy, Ea (kJ/mol)	Pre-Exponential Factor, A (min−1)	R2 from Plot	Activation Energy, Ea (kJ/mol)	Pre-Exponential Factor, A (min−1)	R2 from Plot
Chemical	First-order (F1)	65.82	5.0873 × 104	0.9882	67.34	1.0329 × 105	0.9888	66.40	1.0953 × 105	0.9881	66.40	1.3118 × 105	0.9864
	Second-order (F2)	94.62	3.0881 × 107	0.9559	94.62	6.1763 × 107	0.9563	96.89	9.7105 × 107	0.9563	95.55	6.9599 × 107	0.9524
	Third-order (F3)	129.70	6.4717 × 1010	0.9217	129.70	1.29434 × 1011	0.9219	132.90	2.1212 × 1011	0.9219	131.05	1.2162 × 1011	0.9172
Diffusion	One-dimensional (D1)	98.06	1.2944 × 107	0.9992	98.06	2.5889 × 107	0.9993	100.22	3.9199 × 107	0.9993	99.07	2.8758 × 107	0.9997
	Two-dimensional (D2)	110.51	1.0282 × 108	0.9993	110.51	2.0565 × 108	0.9996	112.97	3.1433 × 108	0.9996	111.67	2.1457 × 108	0.9992
	Three-dimensional J (D3)	125.69	6.4219 × 108	0.9958	125.69	1.2844 × 109	0.9963	128.54	1.9907 × 109	0.9963	127.04	1.2459 × 109	0.9949
	Three-dimensional GB (D4)	115.52	6.8969 × 107	0.9986	115.52	1.3794 × 108	0.9989	118.12	2.1179 × 108	0.9989	116.75	1.4021 × 108	0.9981
Phase interfacial	One-dimensional (R1)	43.95	3.3309 × 102	0.9990	43.95	6.6618 × 102	0.9990	44.93	9.9444 × 102	0.9990	44.26	9.6170 × 102	0.9995
	Two-dimensional (R2)	54.03	1.7353 × 103	0.9977	54.03	3.4705 × 103	0.9977	55.25	5.2249 × 103	0.9982	54.56	4.7584 × 103	0.9970
	Three-dimensional (R3)	57.76	2.7237 × 103	0.9953	57.76	5.4473 × 103	0.9953	59.08	8.2295 × 103	0.9958	58.24	7.3348 × 103	0.9943
Nucleation and growth	Two-dimensional (A2)	27.83	16.139	0.9845	27.83	3.2278 × 101	0.9845	28.48	48.360	0.9853	27.92	49.806	0.9819
	Three-dimensional (A3)	15.17	0.799	0.9784	15.17	1.5984	0.9784	15.53	2.3829	0.9795	15.14	2.6205	0.9739

surrounded all models for the four different heating rates, i.e., 10–40 °C/min. The correlation factor (R²) was used to judge the probability of each reaction model being interplayed. No marked increase in activation energy existed between different heating rates, although it was maximum for 30 °C/min, as shown in Figure 5.

For chemical reaction models (F₁, F₂, and F₃), an increase in reactants necessitated either higher temperature or reaction time because of a high activation energy demand. This is to overcome any heat or mass transfer limitations during pyrolysis, which is mainly attributed to the rampant molecular collisions within CRD wood biomass contributing to complex, overlapping reactions. It ranged from 65–68 kJ/mol for F₁ to 129–133 kJ/mol for F₃ as the heating rate increased, indicating energy-intensive heterogeneous structures competing in the reaction. This is supported by a concomitant increase in the pre-exponential factor as well. However, the average R² is between 0.92 and 0.98, suggesting a relatively lower correlation of these reaction models w.r.t. their counterparts. Due to the highest activation energies recorded for F₃ (more than any other reaction model tested), it may be possible to assign its role for higher conversion stages (α > 0.6), where a thorough rearrangement of biochar’s chemical structure is probable, as concluded by [81].

For the diffusion model, activation energies collectively ranged from 98 to 129 kJ/mol, with a dominant effect from linear (D₁), planar (D₂), and spherical (D₃) diffusion mechanisms. These models dominate the thermochemical data of CRD wood, as extrapolated in the findings by [82]. The average R² is > 0.99, highlighting their pivotal part in thermal decomposition, where the migration of volatiles and pyrolysis products could happen through porous channels and at the boundary layers. Diffusion was deemed suitable for coconut shell feedstock with a comparable activation energy of 68.9 kJ/mol, granting an appreciable bioenergy potential [83]. A high correlation factor for the diffusion model was also shown for feedstocks like wood sawdust and wheat straw [84]. In the case of any energy applications for CRD wood, D₁, D₂, and D₃ models may govern the diffusion of ambient air/oxygen and volatile matter that initiate ignition and proceed till final combustion. This concept was also proposed by [85]. As in [86], the D₃ model showed the highest activation energy for pyrolysis of MSW, which could contain easily decomposable constituents. Meanwhile, in the CRD wood D₃ model (125–129 kJ/mol with R² > 0.99), it corresponded to a high activation energy for carrying forward diffusion within constricted spaces, especially upon clogging of porous channels with residual tar. This may also mean that surface-based diffusion has lower resistance and takes place more freely until an intermediate or product layer forms that impedes any progress. Due to high activation energies overall, F₃, D₁, D₂, and D₃ models can be attributed to the second (200–400 °C) and third (400–600 °C) stages of the TGA cycle, where the majority of disintegration reactions happen for CRD wood pyrolysis. An average activation energy of up to 84 kJ/mol was obtained by [87] for almond shell pyrolysis using the CR method, where the decomposition reactions occur in the range of 150–550 °C as well. All these results were commensurate with CRD wood.

The occurrence of phase interfacial reactions (R₁, R₂, and R₃) are also explainable due to a high correlation of >0.99. The results here are comparable to [88], where D₃ and R₃ majorly drove the biomass decomposition along with F₁, F₂, …F_n. This means that decomposition of biomass could begin from the surface toward the less-reacted, unconverted core. As a result of constant interaction between reactants and products at an interphase, product layers are formed and sequentially shift from the outside to the inside of the biomass particle. This is why biochar production is favored at lower heating rates since it allows sufficient time for this gradual conversion. The activation energy here was lower than chemical reaction and diffusion models, at 43–60 kJ/mol. Finally, less energy (15–30 kJ/mol) was needed for secondary pyrolytic intermediates and product formation through A₂ and A₃ nucleation mechanisms (R² = 0.97–0.99). For this mechanism, inherent AAEM and other inorganics may have coincidentally aided in secondary charring and volatile formation by behaving as catalysts.

As depicted in Figure 5, a notable difference in activation energies existed only at a heating rate of 30 °C/min, which may signify maximum conversion at this point, unlike the ideal TGA curves, where thermal decomposition temperatures increased closely (367–390 °C) with a step-wise increase in the heating rate owing to a reduced depolymerization time, especially near the less conductive core. The same trends were observed for all other models, too. [77] reported a similar finding, where activation energies increased till a specific heating rate, after which they stabilized or dropped. However, this could be the case only till the maximum conversion is attained, after which activation energies may decrease. This was also concurred with by [35]. On the whole, since each reaction model has a different but reasonably high correlation factor (R² ≥ 0.92) and variable activation energies, it is difficult to assign the most suitable pathway, mainly because the pyrolytic decomposition of biomass could be viewed as a simultaneous process rather than being assumed as an individualistic system. This was exactly proved by [37].

3.5.2. Model-Free Methods

Model-free or isoconversional methods like OFW, KAS, and FM did not rely on any pre-determined reaction models for computing the variable activation energies of biomass pyrolysis reactions [55]. Instead, the biomass conversion rate, or α, was relied upon for thermal degradation studies. However, as outlined by [89], isoconversional models have limitations wherein they assume only a single-step reaction and could neglect the role of competing reactions. On the other hand, the kinetic analysis of CRD wood using the CR method proved that competing reaction models are plausible for a thermochemical process like pyrolysis. A kinetic analysis, whether performed using model-fitting or model-free methods, considers biomass particles that weigh between 6 and 8 mg, i.e., very small. Henceforth, the temperature on the surface can reach the center of the same particle within a short time frame, disregarding any heat transfer limitations. However, in actual pyrolysis reactors, the particle sizes of biomass can increase from a few cm to several inches, and even to logs. This is where the importance of heat conductivity is compromised, but in reality, the temperature requirements rise drastically to satiate high activation energies for biomass conversion. Thus, a TGA-based study may mostly function depending upon kinetics only and could often overlook the particle size, heating rate, composition of biomass (the percentages of holocellulose and lignin), and the externalities caused by them. In Figure 6A,B, KAS and OFW plots together seem to slightly differ from the FM plot.

For reference, the calculated kinetic parameters with the model correlation for each conversion stage are listed in

Table 6. Kinetic parameters of CRD wood pyrolysis computed using OFW, KAS, and FM isoconversional methods.

Conversion Rate, α	OFW Method			KAS Method			FM Method
	Activation Energy, Ea (kJ/mol)	Pre-Exponential Factor, A (min−1)	R2 from Plot	Activation Energy, Ea (kJ/mol)	Pre-Exponential Factor, A (min−1)	R2 from Plot	Activation Energy, Ea (kJ/mol)	Pre-Exponential Factor, A (min−1)	R2 from Plot
0.1	8.85	49.51005146	0.414	135.06	2.08822 × 1012	0.599	144.09	2.9616 × 1014	0.637
0.2	172.87	8.40043 × 1014	0.990	176.05	1.6275 × 1015	0.987	185.71	2.15137 × 1017	0.988
0.3	176.61	7.47933 × 1014	0.997	174.12	4.35321 × 1014	0.998	185.27	8.42214 × 1016	0.998
0.4	178.65	6.27054 × 1014	0.996	179.28	6.85439 × 1014	0.996	189.54	1.17277 × 1017	0.997
0.5	179.77	4.95313 × 1014	0.997	177.02	2.79223 × 1014	0.999	188.26	6.43149 × 1016	0.999
0.6	178.33	2.6664 × 1014	0.999	178.91	2.85704 × 1014	0.997	189.77	6.86145 × 1016	0.998
0.7	179.97	2.79517 × 1014	0.999	178.44	1.98226 × 1014	0.998	189.92	6.3583 × 1016	0.998
0.8	184.86	5.32708 × 1014	0.999	185.41	5.66593 × 1014	0.997	196.47	2.06678 × 1017	0.997
0.9	522.38	7.94663 × 1039	0.927	470.58	4.06828 × 1035	0.901	482.38	1.04521 × 1038	0.905

. For this tabulation, α > 0.8 and α < 0.2 were not considered due to relatively low R² (0.41–0.64 for α = 0.1; 0.90–0.93 for α = 0.9) and non-linearity. For other conversion rates, R² is >0.99, except for α = 0.2, under KAS and FM methods (R² = 0.98). A high correlation means that activation energies derived at these conversion rates using the three methods are accurate and dependable. The activation energies under OFW, KAS, and FM methods for α = 0.2–0.8 ranged from 172 to 185 kJ/mol, 176 to 186 kJ/mol, and 185 to 197 kJ/mol, respectively. This work discovered very minimal differences between overall activation energy requirements, along with an increase in heating and conversion rates, which may suggest fairly quick chemical reactions for the formation of activated intermediates, as stated by [90]. An α = 0.8 showed the highest activation energy since the decomposition in this phase may have occurred against highly stable aromatics in lignin and resolute inorganics/AAEM in ash.

FC in CRD wood biomass at 16–17% could also be targeted for decomposition during this stage. A significant difference in activation energy between each conversion rate was not observed since they varied very narrowly, just like the TGA curves. Moreover, there was no defined increase in activation energy, and it more or less adopted a partly sinusoidal pattern. These fluctuations could be due to complex multi-step reactions, as discussed by [27]. We may, hence, infer that 82–83% of VC in CRD wood biomass’s holocellulose component could be targeted from α = 0.2–0.7. A momentary jump in E_a could indicate that the VC in biomass was either from heavier lignin bridges or was part of an organo-metallic linkage (e.g., Fe-O-C; AAEM-O-C), which demands high energy for cleavage. These are confirmed by FTIR and SEM analyses, which show the presence of a myriad of functional groups and metal content co-existing with biomass carbon and manifesting different degradation behaviors. Since the activation energy crosses 170 kJ/mol even at an initial conversion rate of 0.2, we may conclude that the overall rate of the reaction was slow. A summary plot of all isoconversional methods for conversion rates between 0.2 and 0.8 is shown in Figure 7.

CRD wood may, thus, be an ineffective direct fuel source because it needs a high amount of energy to start reacting. Converting it into biochar will help ease conductivity and heat transfer issues, resulting in faster reaction times.

3.6. Thermodynamic Parameters

For thermodynamic parameters, α < 0.2 and α > 0.8 are not considered because of low correlation factors encountered during the calculation of activation energies. As tabulated in

Table 7. ΔH, ΔG, and ΔS calculated for each heating rate according to every isoconversional method.

	Ea and A Estimated from OFW Method
Conversion Rate, α	Heating Rate = 10 °C/min Tm = 367.41			Heating Rate = 20 °C/min Tm = 382.48			Heating Rate = 30 °C/min Tm = 386.85			Heating Rate = 40 °C/min Tm = 389.62
	ΔH (kJ/mol)	ΔG (kJ/mol)	ΔS (J/mol·K)	ΔH (kJ/mol)	ΔG (kJ/mol)	ΔS (J/mol·K)	ΔH (kJ/mol)	ΔG (kJ/mol)	ΔS (J/mol·K)	ΔH (kJ/mol)	ΔG (kJ/mol)	ΔS (J/mol·K)
0.1	3.52	−109.52	176.48	3.40	−112.44	176.68	3.36	−113.28	176.73	3.34	−113.82	175.99
0.2	167.54	216.73	−76.79	167.42	217.64	−76.60	167.38	217.90	−76.55	167.36	218.07	−76.23
0.3	171.28	219.86	−75.83	171.16	220.75	−75.64	171.12	221.00	−75.58	171.10	221.17	−75.26
0.4	173.32	220.96	−74.36	173.20	221.82	−74.17	173.16	222.08	−74.11	173.14	222.23	−73.80
0.5	174.44	220.82	−72.40	174.32	221.66	−72.21	174.28	221.90	−72.15	174.26	222.05	−71.85
0.6	173.00	216.08	−67.25	172.88	216.85	−67.06	172.84	217.07	−67.00	172.82	217.20	−66.72
0.7	174.64	217.97	−67.65	174.52	218.74	−67.45	174.48	218.96	−67.40	174.46	219.10	−67.12
0.8	179.53	226.30	−73.01	179.40	227.14	−72.81	179.37	227.39	−72.76	179.35	227.54	−72.45
0.9	517.05	872.53	−554.95	516.93	880.64	−554.76	516.89	883.00	−554.71	516.87	884.49	−552.39
	Ea and A Estimated from KAS Method
Conversion Rate, α	Heating Rate = 10 °C/min Tm = 367.41			Heating Rate = 20 °C/min Tm = 382.48			Heating Rate = 30 °C/min Tm = 386.85			Heating Rate = 40 °C/min Tm = 389.62
	ΔH (kJ/mol)	ΔG (kJ/mol)	ΔS (J/mol·K)	ΔH (kJ/mol)	ΔG (kJ/mol)	ΔS (J/mol·K)	ΔH (kJ/mol)	ΔG (kJ/mol)	ΔS (J/mol·K)	ΔH (kJ/mol)	ΔG (kJ/mol)	ΔS (J/mol·K)
0.1	129.73	146.98	−26.93	129.61	147.14	−26.31	129.57	147.18	−26.14	129.55	147.21	−26.03
0.2	170.72	223.44	−82.29	170.60	224.43	−80.40	170.56	224.71	−79.87	170.54	224.89	−79.54
0.3	168.79	214.48	−71.33	168.66	215.30	−69.69	168.63	215.54	−69.23	168.61	215.69	−68.94
0.4	173.95	222.06	−75.10	173.83	222.94	−73.38	173.79	223.19	−72.89	173.77	223.36	−72.59
0.5	171.70	215.02	−67.64	171.57	215.79	−66.08	171.54	216.01	−65.64	171.51	216.15	−65.37
0.6	173.58	217.03	−67.83	173.46	217.80	−66.27	173.42	218.02	−65.83	173.40	218.17	−65.55
0.7	173.11	214.61	−64.79	172.98	215.33	−63.30	172.95	215.54	−62.88	172.93	215.68	−62.62
0.8	180.08	227.18	−73.52	179.96	228.03	−71.83	179.92	228.28	−71.35	179.90	228.44	−71.06
0.9	465.26	768.12	−472.81	465.13	774.99	−461.94	465.10	776.98	−458.88	465.07	778.25	−456.96
	Ea and A Estimated from FM Method
Conversion Rate, α	Heating Rate = 10 °C/min Tm = 367.41			Heating Rate = 20 °C/min Tm = 382.48			Heating Rate = 30 °C/min Tm = 386.85			Heating Rate = 40 °C/min Tm = 389.62
	ΔH (kJ/mol)	ΔG (kJ/mol)	ΔS (J/mol·K)	ΔH (kJ/mol)	ΔG (kJ/mol)	ΔS (J/mol·K)	ΔH (kJ/mol)	ΔG (kJ/mol)	ΔS (J/mol·K)	ΔH (kJ/mol)	ΔG (kJ/mol)	ΔS (J/mol·K)
0.1	138.76	182.40	−68.13	138.64	183.18	−67.93	138.60	183.40	−67.88	138.58	183.54	−67.59
0.2	180.38	259.11	−122.90	180.26	260.71	−122.71	180.22	261.17	−122.65	180.20	261.47	−122.14
0.3	179.94	253.67	−115.11	179.81	255.15	−114.91	179.78	255.58	−114.86	179.75	255.86	−114.38
0.4	184.22	259.71	−117.86	184.09	261.24	−117.67	184.06	261.68	−117.61	184.03	261.96	−117.12
0.5	182.94	255.23	−112.86	182.81	256.68	−112.67	182.77	257.10	−112.62	182.75	257.37	−112.14
0.6	184.44	257.08	−113.40	184.32	258.54	−113.21	184.28	258.96	−113.15	184.26	259.23	−112.68
0.7	184.60	256.83	−112.77	184.47	258.28	−112.58	184.44	258.70	−112.52	184.41	258.97	−112.05
0.8	191.14	269.65	−122.57	191.01	271.25	−122.38	190.98	271.71	−122.32	190.95	272.00	−121.81
0.9	477.05	809.47	−518.94	476.93	817.04	−518.75	476.89	819.23	−518.69	476.87	820.62	−516.53

, for OFW, KAS, and FM methods, ΔH did not vary largely with heating rates. Likewise, a lower influence of increasing conversion rates on ΔH also meant that the difference in energy requirement between reactants and the intermediates is lower, which can aid in faster formation of activated complexes. Concomitantly, positive values of ΔH indicated endothermic (exothermic if ΔH < 0) reaction steps with heat absorption for biomass conversion, which is in agreement with the findings by [91]. ΔH was 179–180 kJ/mol, 179–181 kJ/mol, and 190–192 kJ/mol for OFW, KAS, and FM methods. This is in line with the findings for another woody biomass (Ficus Nitida), as discussed by [92]. ΔG, which describes the spontaneity of a reaction or a system, indicated a minor increasing trend with positive values for all heating rates, conversion rates, and methods. This means that the process was not spontaneous (only if ΔG < 0) and needed an external energy source to drive all parallel, sequential, and competing reactions. A ΔG of 215–230 kJ/mol was observed for both OFW and KAS methods between 0.2 and 0.8 conversion rates, which represented consistent results. However, for the FM method, ΔG ranged a bit higher, from 250 to 275 kJ/mol. This is in agreement with the studies mentioned by [93].

A chemical reaction like pyrolysis usually supports greater disorderliness and randomness, represented by the system’s entropy (ΔS > 0), where a solid biomass material breaks down into volatiles, gases, and oil, which are relatively fluid and harbor greater molecular freedom. On the contrary, as shown in Table 7, the ΔS values for OFW (−60 to −80 J/mol·K), KAS (−60 to −85 J/mol·K), and FM (−110 to −130 J/mol·K) non-isothermal methods are largely negative (ΔS < 0), which may mean that biomass samples under consideration could have lower molecular freedom and need more energy to embrace further structural transformations during pyrolysis, as reflected by [2]. Low ΔS could also hint that biomass may have undergone very gradual physical/chemical transformations, consistent with [45]. On the thermodynamic forefront, negative ΔS could arise from the formation of a more ordered or stable activated complex when some decomposition reactions unfolded, as indicated by [94], who used a heterogeneous and moderate-ash biomass feedstock (wheat straw). In other words, structural rearrangements of biomass in its intermediate state could have introduced a certain degree of resistance to degradation. Given the highly heterogeneous nature of CRD wood feedstock, composed of variable lignin percentages and painted or treated surfaces laden with diverse organic and inorganic constituents, localized chemical reordering could be plausible. This could lead to a limited molecular motion and thus, negative ΔS. Interestingly, negativity in ΔS decreases with an increase in α till a certain point, signifying growing randomness in the system, before it increases yet again at higher α—possibly due to the influence of ash matter or the formation of stable metal–organic carbon complexes under high-temperature decomposition.

3.7. Summary

To summarize, the kinetic analysis of CRD wood pyrolysis utilizing model-fitting and model-free methods provided notable insights into the topic of the thermochemical decomposition of complex heterogeneous biomass. This work revealed that a reasonable lignin content and low ash content make waste CRD wood a promising feedstock for pyrolytic biochar production. The decomposition process exhibited three distinctive breakdown phases: till 150 °C, 200–400 °C, and 400–600 °C, during which where peak weight loss occurred between 350 and 400 °C. Under the four different heating rates (10–40 °C/min) tested, the model-fitting CR method derived activation energies ranging from 65 to 133 kJ/mol for the chemical reaction models (F₁, F₂, and F₃), 98 to 129 kJ/mol for the diffusion reaction models (D₁, D₂, D₃, and D₄), 43 to 60 kJ/mol for phase interfacial reaction models (R₁, R₂, and R₃), and 15 to 29 kJ/mol for random nucleation and growth reaction models (A₂, A₃). A peak in activation energy was seen, especially at 30 °C/min, for all schemes. Model-free or isoconversional OFW, KAS, and FM methods yielded activation energies between 172 and 196 kJ/mol for the conversion rates of 0.2–0.8. Thermodynamic evaluations indicated positive values of ΔH and ΔG, signifying an endothermic and non-spontaneous pyrolysis process.

Subsequent work in the domain of CRD wood pyrolysis kinetics and thermodynamics proffers understated research gaps. Firstly, CRD wood from different regions could be tested to understand the effect of feedstock variability with geography, demolition practices, and pre-processing/pre-treatment steps on thermal decomposition behavior. Secondly, different heating rates under 10 °C/min can be experimented with to discriminate TGA over significantly slower pyrolysis conditions and their role in intrinsic heat transfer for a given biomass particle size. Another interesting area of work could be to estimate the kinetic and thermodynamic behavior of CRD wood co-pyrolysis with other organic substrates. On the whole, the present manuscript is important for readers since it addresses the pressing issue of CRD wood waste management and also offers a viable valorization pathway via pyrolysis. Through the kinetic and thermodynamic analysis, this work contributes to some nascent knowledge on the interplay of pyrolysis-independent variables, like the heating rate and temperature, the biochemical composition of CRD wood, and their synergistic influence on its strucutral breakdown.

4. Conclusions

TGA was employed to assess the thermal stability and kinetic behavior of CRD wood biomass under pyrolysis conditions. Different heating rates were utilized to analyze its thermochemical decomposition, revealing that the degradation is very complex, harbors multiple reaction steps, and could be jointly influenced by the feedstock’s biochemical and surface chemical composition. To calculate the kinetic parameters of pyrolysis, the weight loss data from TGA were ascribed to model-fitting and model-free/isoconversional methods, which enabled the calculation of activation energies: the former method dependent on specific assumptions pertaining to different reaction models, while the latter method without any assumptions of such reaction schemes and only functioning on the basis of biomass conversion levels (α). The CR method’s perspective ranked pyrolytic transformation based on the degree of fit or correlation obtained by ascribing TGA data to each model equation. The highest R² (consistently >0.99) was, however, only achieved for the diffusion and phase interfacial models, thereby bolstering their criticality in biomass pyrolysis. Higher heating rates during TGA also shifted the peak melting temperatures of biomass constituents to higher temperatures, necessitating optimization toward the direction of slow pyrolysis with low heating rates and long BRT. This could make sure there is gradual yet steady heat transfer from the surface to the core of every biomass particle, aiding complete conversion. On the other hand, the model-free OFW, KAS, and FM methods offered simple and time-effective calculations of pyrolysis kinetic parameters. Simple and straightforward estimations could be drawn for most biomass conversion stages (α = 0.2–0.8), where the highest R² (mostly >0.99) was obtained. Extrapolating activation energies to specific energy demands across the pyrolytic transformation of CRD wood was possible and also corroborated with findings from similar works. Overall, despite a detailed outlook, experimental TGA data were well-aligned with kinetic models. Even thermodynamically speaking, these TGA tests that mimicked pyrolysis-like conditions proved that CRD wood thermal degradation is an energy-intensive process. The TGA results could be crucial for scaling up the pyrolysis of CRD wood since they offer inputs on the decomposition kinetics of a biomass material within thermochemical settings. Moreover, insights on the stability, reactivity, and disintegration of biomass polymers at specific conditions of temperature, heating rate, and BRT could be obtained, which are essential for designing pyrolysis systems.

As much as the benefits of TGA-based kinetic parameter approximations have been discussed, acknowledging their limitations also remains vital. Firstly, these tests were conducted with meagre amounts of feedstock (8–10 mg) under controlled laboratory settings that may not represent pilot, demonstration, or industrial-scale settings that are largely dynamic environments. Secondly, the particle size of biomass and the role of heat transfer within the particle differ largely between such settings. Thirdly, nuances in larger pyrolysis systems, such as the residence and eventual settling of volatiles on biomass, the cracking of volatiles and their role in carbon polymerization to PAHs, and the interplay of any vacuum/suction effects across the reactor, may not be achieved within TGA set-ups. Lastly, based on the different methods used to calculate kinetic parameters, especially model-fitting methods like CR, there could be inaccuracies posted due to the selection of wrong reaction models, assumptions, and corresponding findings derived that may fail to capture some intricacies of pyrolysis at higher scales.