Turning Construction, Renovation, and Demolition (CRD) Wood Waste into Biochar: A Scalable and Sustainable Solution for Energy and Environmental Applications

Abstract

This study investigates the pyrolysis of construction, renovation, and demolition (CRD) wood waste to produce biochar, with a focus on its robustness, scalability, and characterization for energy and environmental applications. Pyrolysis conditions, including the temperature, biomass residence time (BRT), and feedstock mass, were varied to evaluate their effects on biochar properties. High-temperature biochars (B800) showed the highest fixed carbon (FC) (87%) and thermostable fraction (TSF) (96%) and the lowest volatile carbon (VC) (9%), with a high carbon content (92%), a large BET surface area (300 m²/g), and a high micropore volume (0.146 cm³/g). However, the hydrogen (0.9%) and oxygen (2.2%) content, Van-Krevelen parameters (H/C: 0.1; O/C: 0.02), and biochar yield (21%) decreased with increasing temperature. Moderate-temperature biochars (B600) have balanced physicochemical properties and yields, making them suitable for adsorption applications. Methyl orange dye removal exceeded 90% under the optimal conditions, with B600 fitting well with the Freundlich isotherm model (R² = 0.97; 1/n = 0.5) and pseudo-second-order kinetic model (R² = 1). The study highlights biochar’s suitability for varied applications, emphasizing the need for scalability in CRD wood pyrolysis.

1. Introduction

Discarded construction, renovation, and demolition (CRD) wood waste presents environmental risks due to its heterogeneous composition and metal contamination [1]. Disposal practices vary with construction trends, legislation, and location, while traditional methods, like landfilling and incineration, raise issues of space scarcity, leachate pollution, odor, and greenhouse gas emissions. Pyrolysis, a thermal decomposition process that takes place in low-oxygen environments, offers a feasible alternative for CRD wood waste management, yielding a multifaceted energy-dense product stream encompassing bio-oil, pyrolysis gas, and biochar. Moreover, pyrolysis can also be optimized by varying process conditions as slow, fast, and flash stages depending upon the aforementioned product of interest [2]. This valorization technique reduces the waste volume and immobilizes pollutants in biochar, supporting circular and sustainable waste management [3].

Biochar, a carbon-rich solid derived from pyrolysis, finds utility in soil remediation, carbon sequestration, wastewater treatment, metallurgy, and construction as a partial coal/coke substitute [4]. Emerging applications include sustainable agriculture [5], biopolymer fillers [6], and supercapacitor components [7]. Its properties are governed by feedstock and process parameters, like temperature, biomass residence time (BRT), heating rate, feedstock mass, and particle size [8,9]. Feedstocks, like CRD wood, differ from residues, like agricultural manure or municipal waste, in their biochemical and inorganic makeup. High-ash biomass (e.g., agri-waste) introduces alkaline and alkaline earth metals (AAEMs), such as Na, K, Ca, Mg, and Ba, which promote CO₂ release, lower carbon retention, and reduce biochar yields [10]. Above 600 °C, biochar becomes stable, aromatic, and graphitized, ideal for carbon sequestration [10], while milder temperatures produce surface-charge-rich reactive biochar suited for adsorption and soil remediation. However, key surface functionalities decline beyond 500 °C due to decarbonylation, decarboxylation, and other de-functionalization reactions [11]. Activation via acids, alkalis, steam, or CO₂ may alter the surface area, pore volume, and reactivity for adsorption [12], but many biochars perform well without such treatment methods [13].

Methyl orange, a synthetic anionic azo dye, is common in the textile, pulp and paper, and chemical industries [14]. The pulp and paper demonstration plant involved in this study uses methyl orange dye as one of the key pH indicators and for adjusting the acidity–basicity of hydrolysis and pulp solutions within the paper-manufacturing unit’s operations, so its wastewater contains notable levels of the dye. Due to its carcinogenic and ecotoxic effects, persistent methyl orange in effluents raises environmental concerns [15]. Therefore, this dye has been chosen as the model compound for adsorption studies. Although activated carbon is the benchmark adsorbent, its high emissions from its fossil coke substrate, activation requirements, and transport from remote locations drive the search for sustainable alternatives [16]. While biochar has previously been studied for dye removal [17], CRD wood biochar remains largely underexplored. Adjustable porosity, surface area, surface functional groups, and carbon stability are four important properties necessary for biochar to be an effective adsorbent in water treatment to remove organic/inorganic pollutants [18]. The nature of the biomass feedstock, the pyrolysis parameters, and any chemical, thermal, or physical activation techniques govern these properties [19]. Activated biochars, though effective, are less feasible for single-use applications due to the cost of regeneration. Hence, using raw industrial residues, like CRD wood, is promising [20]. The adsorption efficiency also depends on the pollutant concentration, adsorbent dose, temperature, pH, co-ions, and contact time, influencing mechanisms such as electrostatic interactions, π-π bonding, hydrogen bonding, and pore filling [19,21,22]. If all these conditions are satisfied, adsorption is thermodynamically spontaneous and endothermic [23].

This study investigates raw CRD wood biochar for methyl orange adsorption without chemical or physical activation. Given CRD wood’s end-of-life and low-cost nature, the spent biochar may be immobilized in concrete or asphalt for long-term pollutant containment and the elimination of desorption risks. Soil use, however, would require attention to leaching risks. Avoiding chemicals, reactive gases, or energy-intensive steps, this approach enables on-site production, use, and disposal with minimal cost and impact. Adsorption tests determine the best-performing biochar, selected solely from pyrolysis experimental outcomes.

The novelty of this work lies in applying a design of experiments (DOE) framework to study the influence of temperature, BRT, and feed mass on CRD-wood-derived biochar via slow pyrolysis in a laboratory horizontal tube furnace reactor—an underrepresented area in the current literature—building on our prior work [24]. The same CRD wood was later processed under chosen identical conditions in a semi-pilot horizontal rotary retort-furnace reactor to assess scalability potential. Unique to this study is the combined evaluation of CRD wood biochar for its physicochemical suitability in carbon sequestration, soil amendment, and metallurgy (as a reductant), alongside its performance in non-activated dye adsorption. Kinetic, isotherm, and thermodynamic modeling add depth, offering critical insights into its practical environmental and industrial applications.

2. Materials and Methods

2.1. Sample Preparation

CRD wood residues sourced from BRQ Fibre et Broyure, Trois-Rivières, Quebec, Canada, with an initial moisture content of 18%, were received in cartons (Figure 1A) and oven-dried overnight at 105 °C until moisture content fell below 1 wt% (0.18 wt% in this case). This low moisture level facilitates faster pyrolysis, enhances heat transfer, and reduces reactor energy demand. A known mass of dried wood was subjected to vibrational sieving (LABTECH unit) for ≥5 min to determine particle size distribution. Non-wood contaminants of varying sizes were manually removed throughout. The mass retained on each sieve was recorded to calculate the average particle size. Biomass with particle size >2.5 cm was milled using a Thomas Wiley mill model 4 to achieve a manageable size range of 3 mm–2.5 cm (Figure 1B) for pyrolysis in both reactors. For characterization, the same mill (Figure 1C) with a screen for the smallest pore size was used to generate fines of both biomass (Figure 1D) and pyrolytic biochar (Figure 1E).

2.2. Slow Pyrolysis Setup and Design of Experiments

Response Surface Methodology (RSM) was chosen to optimize CRD wood pyrolysis due to its effectiveness in modeling systems with few variables (temperature, BRT, feedstock mass, as in the current study), providing dependable polynomial equations and robust statistical validation (R², ANOVA, p-values) for easy interpretation. Its low experimental demand makes it a practical, cost/time-efficient option for resource-intensive processes like pyrolysis. RSM also ensures reproducibility, replicability, and transparency in process optimization studies—a vast understanding has already been established. In this work, slow pyrolysis followed a DOE framework using Stat-Ease’s Design Expert v25.0 RSM tool with central composite design (CCD), varying the three aforesaid variables across multiple runs to assess their individual and interactive effects on key biochar properties. The conditions are summarized in

Table 1. Experimental conditions chosen for pyrolysis.

Coded Independent Variable	Actual Independent Variable	Coded and Actual Values for the Five Levels
		−α (Lowest)	−1 (Low)	0 (Mid-Point)	1 (High)	α (Highest)
A	Pyrolysis temperature (°C)	400	500	600	700	800
B	BRT (min)	30	45	60	75	90
C	Mass of feedstock (g)	50	100	150	200	250

, with model rationale and validation outlined in Supplementary Figure S1 and Supplementary Table S1. Performance was evaluated using statistical tools, regression coefficients, and multiple plot types (surface, contour, predicted versus actual, and perturbation plots), serving as key performance indicators (KPIs).

Two reactors were used for pyrolysis. The first, a horizontal tube furnace reactor (“Thermo Scientific Lindberg Blue M™ 1100 °C”, Waltham, MA, USA) (Figure 2, left), followed the design in [24]. A 2.5-inch-diameter, 69-inch-long steel tube reactor (5.55 L) holding up to 400 g CRD wood was inserted into the furnace cavity. The system reached set pyrolysis temperatures within 15 min, with three K-type thermocouples for zone monitoring and one in the reaction zone. Nitrogen with a flow rate of 1–5 L/min and a high-suction vacuum pump maintained inert conditions. The drawn-out gases are fed directly to a combustion chamber of the in-house pilot pyrolyzer to meet a small portion of its energy demand in tandem with natural gas and pilot pyrolysis gas. To prevent the carry-over of dust and other particulates from the reaction zone, a steel mesh filter is used.

The second, a horizontal rotary retort-furnace reactor (Figure 2, right), used the same thermocouple configuration, including one inside the retort to track thermal inertia. Thermal inertia can be defined as the reactor’s resistance to an increase in temperature caused by non-uniform heating patterns and loss of heat via the reactor walls due to inefficient insulation. As a result, the refractory bricks of the furnace heat up to the set-point temperature, but this heat is not transferred adequately to the retort, and thus, the biomass inside experiences delayed heat transfer. Adequate heat transfer is essential to minimize unconverted biomass percentage as inferred from [25]. In fact, the influence of thermal inertia on material energy balance during pyrolytic thermal decomposition in both these reactor configurations is also addressed in Supplementary Tables S4 and S5. Quartz wool insulated the retort ends, and the same nitrogen flow and vacuum system enabled volatile removal. The retort processed ~3 kg of CRD wood per batch, with rotational mixing promoting uniform heating and minimizing secondary reactions from volatile condensation.

2.3. Physicochemical and Morphological Characterization of Biochar

All characterizations were performed by the same technician using the same CRD wood and instrumentation as in [24]. This section outlines the equipment used to assess biochar’s carbon structure, degree of aromatization, thermal and oxidative stability, and permeance. Selected analyses were also conducted on the original CRD wood biomass.

Proximate analysis was performed using 1 g of dried biochar fines in a muffle furnace, heated at 5 °C/min from room temperature to 550 °C and held for 2 h to measure ash content. Volatile carbon (VC) was determined per ISO 562 by heating 1 g of dried biochar fines in a lidded quartz crucible at 900 °C for 7 min. Fixed carbon (FC) was calculated by the difference [VC + FC + ash = 100]—all values are reported as weight percentage dry basis (wt% DB). Elemental analysis used an Elementar Vario Macro Cube (Langenselbold, Germany) with a helium carrier and oxygen combustion at 1200 °C; oxygen was calculated as [O = 100 − (C + H + N + S + ash)]. Thermogravimetric analysis (TGA) was performed on 8 mg of undried biochar using a PerkinElmer TGA 8000 (Waltham, MA, USA) ramped at 10 °C/min to 900 °C under air to assess thermal and oxidative decomposition. Correcting the thermograms for moisture and ash is key to calculating the TGA R50 recalcitrance index, i.e., the ratio of temperatures at which 50% weight loss in a biochar material happens to that of a reference graphite material. Surface functionalities were identified by Fourier Transform Infrared (FTIR) spectroscopy (Agilent Cary 630 spectrometer, Santa Clara, CA, USA) using diamond/germanium crystal and transmittance spectra (500–4000 cm⁻¹) processed in ORIGIN. A Raman spectrometer (ThermoScientific DXR3) with a 532 nm laser, 7 mW power, and 60 s exposure time was used over a scan range of 800–1800 cm⁻¹ to quantify the I_D/I_G ratio for assessing carbon structural disorder or the degree of amorphous-to-crystalline carbon transitions in biochar. The Brunauer–Emmett–Teller (BET) surface area and micropore volume were measured on a Micromeritics Tristar II (Norcross, GA, USA) using helium/nitrogen as a carrier gas. Metal content was analyzed via a Microwave Plasma–Atomic Emission Spectrometer (MP-AES) (Agilent 4210) after acid digestion and nitrogen plasma nebulization of biochar. The pH was determined by mixing biochar (1:10 w/v) with demineralized water, shaking, resting for 30 min, and measuring with a bench-top pH analyzer. Bulk density was calculated from the mass-to-volume ratio in a leveled, pre-weighed steel container. Biochar yield from pyrolysis was expressed as the mass percentage relative to the initial biomass feed taken. The higher heating value (HHV) was estimated using elemental data from [26,27]. Scanning Electron Microscopy–Energy-Dispersive X-ray spectroscopy (SEM-EDX) (Hitachi SU-70, 15 kV, vacuum, Tokyo, Japan) was used to image dry biochar fines at 100×, 250×, 500×, and 1000× to detect surface elements—primarily C, O, Fe, and AAEMs—and the porous morphologies of biochar.

3. Results and Discussion

3.1. CRD Wood Pyrolysis in Horizontal Tube Furnace Reactor

Five biochar samples produced between 400 and 800 °C were randomly selected (B400–B800) for characterization and scalability tests to better assess the influence of pyrolysis parameters. Nonetheless, all data from

Table 2. Results of slow pyrolysis runs from the DOE model. Bulk density and HHV are discussed separately as plots (Figure 3E,F). Five biochars at five different pyrolysis temperatures of 400, 500, 600, 700, and 800 °C with mixed BRT/mass of feedstock conditions are considered for future characterizations. These are labeled as B400, B500, B600, B700, and B800 (highlighted in gray).

Run	Independent Variables for RSM			Elemental Parameters					Porosity		Biochar Output	Proximate Parameters			Thermal and Chemical Stability Indices
															TGA	Van-Krevelen		Carbon Strength
	A: Temperature (°C)	B: BRT (min)	C: Mass of Feedstock (g)	N (wt%)	C (wt%)	H (wt%)	S (wt%)	O (wt%)	BET Surface Area (m2/g)	Micropore Volume (cm3/g)	Yield (%)	VC (wt%DB)	FC (wt%DB)	Ash (wt%DB)	R50	H/C	O/C	TSF (%)
CRD wood	N.A	N.A	N.A	0.990	49.880	6.123	0.101	41.736	N.A	N.A	N.A	83.480	15.350	1.170	0.391	1.473	0.628	15.534
1 (B400)	400	60	100	0.780	65.320	4.293	0.442	26.145	4.238	0.002	57.800	26.384	70.596	3.020	0.513	0.789	0.300	73.271
2	600	90	200	0.670	86.180	2.193	0.210	7.417	308.703	0.139	21.800	15.748	80.922	3.330	0.570	0.305	0.065	86.061
3	600	30	200	1.070	83.660	2.329	0.213	6.288	309.032	0.132	20.700	17.427	76.133	6.440	0.568	0.334	0.056	80.501
4 (B500)	500	75	200	0.630	79.430	2.933	0.177	13.110	195.278	0.095	26.900	25.837	70.443	3.720	0.544	0.443	0.124	73.169
5	600	60	100	0.910	82.370	2.391	0.241	8.578	255.854	0.123	21.300	18.473	76.017	5.510	0.603	0.348	0.078	80.132
6	500	75	150	0.870	78.340	2.910	0.180	13.840	179.077	0.090	26.400	24.312	71.828	3.860	0.561	0.446	0.132	74.783
7	500	90	150	0.720	77.600	2.746	0.218	12.546	260.562	0.114	26.500	24.247	69.583	6.170	0.545	0.425	0.121	72.452
8	600	45	50	0.700	86.720	2.034	0.123	5.613	334.887	0.149	19.200	14.132	81.058	4.810	0.632	0.281	0.049	86.794
9	700	90	200	0.940	88.760	1.379	0.205	4.026	315.809	0.148	23.300	12.151	83.159	4.690	0.587	0.186	0.034	90.002
10 (B800)	800	75	200	0.890	91.740	0.870	0.168	2.232	299.554	0.146	21.100	9.169	86.731	4.100	0.647	0.114	0.018	96.191
11 (B700)	700	75	200	0.950	88.340	1.413	0.253	2.914	293.195	0.138	23.600	11.640	82.230	6.130	0.588	0.192	0.025	89.295
12	500	90	250	0.510	82.000	2.825	0.151	11.344	229.033	0.108	26.600	23.857	72.973	3.170	0.546	0.413	0.104	76.032
13 (B600)	600	90	250	0.430	87.510	2.209	0.123	6.168	323.776	0.145	21.900	15.037	81.403	3.560	0.577	0.303	0.053	86.817
14	700	45	150	1.630	85.880	1.352	0.332	2.136	217.947	0.118	25.200	13.903	77.427	8.670	0.638	0.189	0.019	82.996
15	400	90	200	1.000	73.770	3.351	0.151	17.598	4.034	0.002	31.200	32.542	63.328	4.130	0.534	0.545	0.179	65.274
16	600	60	150	1.140	72.180	1.786	0.215	2.649	247.457	0.112	25.600	22.723	55.247	22.030	0.647	0.297	0.028	57.678

N.A: Not analyzed.

were incorporated into the DOE model for statistical analysis.

3.1.1. Physicochemical Characterization

CRD wood shares some similar physicochemical properties with hardwoods like poplar and beech, including comparable carbon, oxygen, hydrogen, and ash content. However, it contains slightly higher sulfur, likely from gypsum residues in drywall fractions present in the initial heterogeneous feedstock. CRD wood also differs from lignin-rich softwoods like pine, which have higher carbon and lower oxygen due to their inherent structural aromaticity. This is implicit from CRD wood’s higher O/C, which confirms its holocellulose richness—characteristic of hardwoods with oxygen-based functional groups such as hydroxyls and carboxyls.

The results from all 16 slow pyrolysis runs in the horizontal tube furnace reactor are presented in Table 2. Pearson’s coefficient of correlation (r) and model p-value were used as statistical indicators to analyze the influence of key process variables. The carbon content rose by 40–41% from B400 to B800, strongly correlated with pyrolysis temperature (r = 0.92; p-value = 0.0002). Concurrent declines in oxygen and hydrogen yielded lower O/C (r = −0.88; p-value < 0.0001) and H/C ratios (r = −0.95; p-value < 0.0001), indicating aromatic stability and biochar maturity. For soil amendment, especially in acidic or metal-contaminated soils, biochars like B400 and B500 are preferable due to their higher O/C, associated with increased surface negativity and CEC [28,29]. The BET surface area peaked at 600 °C before falling below 300 m²/g (r = 0.83; p-value = 0.001), likely from micropore wall collapse and mesopore formation. This aligns with [30], where a higher pyrolysis temperature and a longer BRT enlarged the pore size. Biochars at 600 °C exhibited a high surface area (320–340 m²/g) and micropore volume (0.112–0.149 cm³/g), supporting adsorption. Such porous structures also enhance soil fertility and water retention in arid zones [31]. FC increased by 360% from raw CRD wood to B400, demonstrating the positive role of even the slightest of carbonization. From B400 to B800, FC rose further to 86.73% (r = 0.95; p-value = 0.0099), driven by the decomposition of volatile organics, as confirmed by a 65% drop in VC (r = −0.96; p-value < 0.0001).

Other biochar properties were also significantly affected by the pyrolysis temperature. The ash content in CRD wood biochar was comparable to that in other woody biomass types, specifically pinewood (0.40–1.80 wt%), poplar (3.45 wt%), and softwood pellets (0.61 wt%), classifying CRD wood as a low-ash feedstock [32]. B700 showed the highest ash content (6.13%), which declined to 4.10% in B800, likely due to the volatilization of unstable inorganics. The biochar yield decreased from 57.8% (B400) to 21.1% (B800) (r = −0.78). At 400 °C, incomplete pyrolysis left behind unconverted woody material, inflating the yield; a 53% drop at 500 °C confirmed B400 overestimation. The HHV rose from 20.28 MJ/kg (CRD wood) to 32.74 MJ/kg (B800), reflecting increased carbon and reduced volatiles (r = 0.87). This enhances energy density, benefiting combustion and high-temperature applications. The bulk density increased from 270 to 451 kg/m³ from B400 to B800 (r = 0.99), aligning with [31]. Higher heat exposure induced compaction, crystallite ordering, and volume shrinkage [33], although [34] reported the opposite. They also noted declining ion exchange and reactivity at higher temperatures, possibly due to surface charge loss and carbon localization. The resulting structured FC improves thermal and electrical conductivity, supporting its use in metallurgy and charge storage.

As noted in [35], polyaromatic hydrocarbon (PAH) content in high-temperature biochars may decline if the carrier gas flow is sufficient to prevent volatile condensation and uncontrolled polymerization. Rather than adjusting the nitrogen flow, this study used a high-suction vacuum pump to minimize the gas residence time. According to [36], low-temperature biochars prioritize yield and contain fewer PAHs, lowering ecotoxicity risks. However, higher pollutant desorption can elevate ecotoxicity. Future work will assess priority PAHs such as Phenanthrene, Naphthalene, Pyrenes, and their benzo-derivatives [37].

Effect of Particle Size on Biomass Thermal Decomposition and Biochar Properties

The biomass particle size in pyrolysis affects conversion, reactivity, surface area, and composition. In this study, CRD wood particles ranged from >3 mm to <2.5 cm in the horizontal tube furnace reactor. This range enabled thorough volatilization under slow heating, ensuring uniform heat penetration from the boundary layer towards the core of each particle. As [38] reported, small to intermediate particles with low moisture favor efficient heat transfer and yield. The observed 20–30% average biochar yield aligns with biomass pyrolysis studies at similar temperatures, confirming the suitability of the particle size. Successful conversion was evident from the crisp texture and uniformity of biochar, supporting micro- and mesopore development [39]. According to [40], densified, small-particle biochar may serve as a metallurgical reductant, enhancing strength, density, and controlled reactivity under high temperatures. Some biochar particles showed reduced surface area due to micro- to mesoporous structures, which can benefit wastewater treatment and nutrient recovery [41]. Larger pores may aid in heavy metal entrapment. For carbon sequestration, porosity is less critical than FC, VC, carbon content, and stability indices. High-temperature CRD wood biochars exhibited both high stability and enhanced surface characteristics. Their large surface area, balanced by low H/C, O/C, and hydrophobicity, balances reactivity towards ambient reactive species, moisture, or air. Conversely, larger particles may offer structural integrity for shear-intensive industrial processes [41].

Visual descriptions of these data are provided in Figure 3A–D.

3.1.2. Evaluating the Robustness of the RSM Model—Statistical Analysis, Contour/3-D Plots, and Predicted Versus Actual Value Distribution

To evaluate DOE robustness, four biochar properties—surface area, micropore volume, H/C, and O/C—were selected. After 16 runs, interdependence between pyrolysis parameters and these properties was modeled using a polynomial equation, with coefficients, ANOVA (p-value < 0.05 indicating model significance), and model fit indicating significance. Statistical results are summarized in

Table 3. Statistical significance results—equation coefficients, significance of each independent variable (p-values) on the nature of the chosen response variables, and model-fit data are shown. Cells shaded in the color gray are not accounted for due to the type of applicable model.

Response Variable	A	B	C	AB	AC	BC	A2	B2	C2	Intercept	Model Type	Fit (R2)	Lack of Fit
BET surface area (m2/g)	84.6827	57.4696	−23.4390	4.2486	−12.0859	−125.5880	−46.6398	10.6473	209.0550	230.5000	Quadratic	0.9724	Insignificant
p-value	0.0010	0.0194	0.3150	0.8474	0.7916	0.0230	0.0034	0.2764	0.0119
Micropore volume (cm3/g)	0.0425	0.0240	−0.0086	−0.0006	−0.0061	−0.0478	−0.0203	0.0016	0.0806	0.1130	Quadratic	0.9753	Insignificant
p-value	0.0004	0.0214	0.3844	0.9532	0.7546	0.0358	0.0031	0.6873	0.0184
H/C	−0.2017	−0.0224	0.0303	−0.0075	0.1497	0.0153				0.3136	Two-factor interaction	0.9769	Insignificant
p-value	<0.0001	0.1099	0.1850	0.7145	0.0038	0.4522
O/C	−0.0829	−0.0040	0.0121	−0.0162	0.1069	0.0112				0.0613	Two-factor interaction	0.9326	Insignificant
p-value	<0.0001	0.6894	0.4763	0.3198	0.0059	0.4779

, with detailed data provided in Supplementary Tables S2 and S3. Next, surface and contour plots of two response variables—BET surface area and Van-Krevelen H/C index—are also described as examples with reference to the influence of independent pyrolysis variables. Finally, for all four aforesaid response or dependent variables, a comparison of theoretical and actual experimental results was performed to validate the predictability of the model.

RSM’s CCD tool recommended quadratic models for BET surface area and micropore volume and two-factor interaction models for H/C and O/C. In all cases, synergistic effects of pyrolysis variables were evident. The pyrolysis temperature was significant for all responses, identifying it as the most influential parameter. The BRT significantly affected only the BET surface area and micropore volume; an excessive BRT reduced both, likely due to pore rupture and coalescence. The feedstock mass alone showed no significant effect (p-value > 0.05) but influenced the BET surface area and micropore volume through interactions and squared terms (Figure 4iv). A higher feed mass requires an extended BRT to ensure complete pyrolysis and maintain yield. For H/C and O/C, feed mass and temperature governed volatilization and surface group rearrangement (Figure 4xii). Smaller samples at higher temperatures facilitated uniform heating and promoted oxygen/hydrogen elimination, favoring carbon retention. In contrast, a higher feed mass restricted heat/mass transfer, extending volatile residence time and causing condensation on biochar surfaces, forming C–O–C and C=O groups [42], which increase uncontrolled reactivity and reduce suitability for carbon sequestration and metallurgical use. Secondary cracking reactions may also cause carbon loss as light gases [43]. To mitigate these effects, a higher inert gas flow, an extended BRT, and elevated temperatures are necessary.

The model fit (R² > 0.90 for all responses) confirmed a strong correlation and high predictive accuracy. This is evident in the predicted-versus-actual plots in Figure 5, where points cluster near the origin line, validating model selection. In the following DOE equations, (−) indicates antagonistic, and (+) indicates synergistic effects.

$$\begin{array}{c}\mathbf{B}\mathbf{E}\mathbf{T} \mathbf{s}\mathbf{u}\mathbf{r}\mathbf{f}\mathbf{a}\mathbf{c}\mathbf{e} \mathbf{a}\mathbf{r}\mathbf{e}\mathbf{a}\\ =\hfill & 84.6827\mathrm{A}+57.4696\mathrm{B}-23.4390\mathrm{C}+4.2486\mathrm{A}\mathrm{B}-12.0859\mathrm{A}\mathrm{C}-125.5880\mathrm{B}\mathrm{C}\\ -\hfill & 46.6398{\mathrm{A}}^2+10.6473{\mathrm{B}}^2+209.0550{\mathrm{C}}^2+230.5000\hfill \end{array}$$

(1)

Slow pyrolysis DOE model Equation (1)

$$\begin{array}{c}\mathbf{M}\mathbf{i}\mathbf{c}\mathbf{r}\mathbf{o}\mathbf{p}\mathbf{o}\mathbf{r}\mathbf{e} \mathbf{v}\mathbf{o}\mathbf{l}\mathbf{u}\mathbf{m}\mathbf{e}\hfill \\ =\hfill & 0.0425\mathrm{A}+0.0240\mathrm{B}-0.0086\mathrm{C}-0.0006\mathrm{A}\mathrm{B}-0.0061\mathrm{A}\mathrm{C}-0.0478\mathrm{B}\mathrm{C}-0.0203{\mathrm{A}}^2\hfill \\ +\hfill & 0.0016{\mathrm{B}}^2+0.0806{\mathrm{C}}^2+0.1130\hfill \end{array}$$

(2)

Slow-pyrolysis DOE model Equation (2)

$${\displaystyle \frac{\mathbf{H}}{\mathbf{C}}}=-0.2017\mathrm{A}-0.0224\mathrm{B}+0.0303\mathrm{C}-0.0075\mathrm{A}\mathrm{B}+0.1497\mathrm{A}\mathrm{C}+0.0153\mathrm{B}\mathrm{C}+0.3136$$

(3)

Slow-pyrolysis DOE model Equation (3)

$${\displaystyle \frac{\mathbf{O}}{\mathbf{C}}}=-0.0829\mathrm{A}-0.0040\mathrm{B}+0.0121\mathrm{C}-0.0162\mathrm{A}\mathrm{B}+0.1069\mathrm{A}\mathrm{C}+0.0112\mathrm{B}\mathrm{C}+0.0613$$

(4)

Slow-pyrolysis DOE model Equation (4)

Choosing the best or optimal biochar production conditions is ultimately dependent upon the application under focus. Based on our findings, temperature can be underscored as the deciding pyrolysis parameter with the most drastic influence on biochar properties. This can be confirmed by the r, p-value, and perturbation analysis (relatively steeper slopes observed for pyrolysis temperature) obtained from the DOE model.

Figure 5. Predicted (y-axis) versus experimental/actual (x-axis) values for (A) BET surface area; (B) micropore volume; (C) H/C; and (D) O/C. Squares outlined in dark black represent B400–B800, spread along the reference line. All colored points represent the different DOE points.

Figure 5. Predicted (y-axis) versus experimental/actual (x-axis) values for (A) BET surface area; (B) micropore volume; (C) H/C; and (D) O/C. Squares outlined in dark black represent B400–B800, spread along the reference line. All colored points represent the different DOE points.

3.1.3. Other Characterizations

Thermogravimetric Analysis (TGA)

The TGA and DTG curves (Figure 6A,B) show the weight loss patterns for CRD wood and its biochars, influenced by pyrolysis conditions and feedstock composition. CRD wood (~70–75% holocellulose, 20–30% lignin) exhibited three distinct degradation phases [44]. In Phase I (0–150 °C), dehydration and extractive removal caused 2.38% weight loss in CRD wood, and 1.29–1.64% for B400–B800. Perturbations in B600–B800 likely resulted from moisture/labile species adsorption during cooling/storage. Phase II (200–400 °C) showed a major mass loss in raw wood (66.95%) due to cellulose and hemicellulose breakdown, releasing CO, CO₂, CH₄, and volatiles. The release of these species in particular was observed when the internal reactor temperature was between 200–325 °C and 325–400 °C. Key intermediates like carboxylic acids, furfural, oxygenates, levoglucosan, and aldehydes formed pyrolysis oil. CRD wood’s first peak occurred at 344.28 °C. Biochar weight loss in this phase decreased from 27.53% (B400) to 1.87% (B800), indicating reduced volatile content and higher charring with increased heat treatment severity. B400 displayed a minor peak at 337.44 °C, and graphitized biochars showed a shift in decomposition to higher temperatures, consistent with [45].

Phase III (400–600 °C) involved lignin breakdown into phenolics and aromatics. CRD wood had a second peak at 481.76 °C (27.80% loss), while B400 peaked at 502.74 °C (66.81% loss). For B500–B800, decomposition occurred between 513.91 and 597.13 °C, indicating greater aromatic condensation. B500–B700 lost 88.17–90.39% mass; B800 lost only 56.10% due to stable C–C, C=C, and C≡C bonds from dehydrogenation, cyclization, and aromatization. Between 600 and 900 °C, B800 showed 37.46% additional loss from the breaking of these bonds, requiring higher activation energy [46,47]. This reflects the turbostratic transformation of amorphous carbon to hyper-stable graphitized forms. Other biochars showed <2.5% loss in this range, indicating prior conversion. Residual mass at 900 °C corresponds to recalcitrant ash/inorganics. TGA confirms that high-temperature biochars possess traits ideal for carbon sequestration and metallurgical use.

TGA R50

Before calculating the R50 recalcitrance index, TGA curves were corrected for moisture and ash, as shown in Figure 6C. The R50 index increased with pyrolysis temperature, from 0.39 (CRD wood) and 0.51 (B400) to 0.65 (B800), as shown in Figure 6D. This highlights its direct relationship with temperature and Van-Krevelen ratios. B800 showed the highest thermal stability due to high carbonization, FC, and low VC, consistent with [48]. Based on [49], B400–B800 fall into the minimally degradable biochar category (0.5 ≤ R50 < 0.70), suitable for long-term soil carbon sequestration. Such stability also benefits metallurgical use, where high resistance to stress and breakage is crucial. As noted in [50], higher pyrolysis temperatures can form stable organo-metal complexes (e.g., Al₂O₃–O–C), enhancing durability—likely in CRD wood biochars as they contain Al. However, high R50 also suggests low reactivity, making these biochars less ideal for soil amendments, where surface interaction matters. Ultimately, while R50 indicates overall degradation resistance, it does not reveal the structural arrangement of organic carbon (amorphous-to-graphitic transformations in particular).

FTIR Spectroscopy

FTIR analysis of CRD wood and B400–B800 assessed chemical changes via functional group presence and transformation under varying pyrolysis conditions. Shifts in functional group intensity reflected structural evolution and its impact on biochar properties. Peaks in the 500–4000 cm⁻¹ range (Figure 7) were analyzed using spectral databases from [51,52].

Hydrogen-bonded –OH vibrations (3200–3600 cm⁻¹) indicated moisture in CRD wood, with decreasing intensity from CRD wood to B800 due to thermal dehydration. Similar diminishing perturbations in the 3020–3080 cm⁻¹ band, linked to aromatic C–H vibrations, were most pronounced in CRD wood, reflecting high amorphous carbon undergoing aromatic transformation. Minor ramps in B400–B600 suggest gradual cyclization, which disappears at higher temperatures as aromatization stabilizes. However, ref. [53] noted that these vibrations alone do not define graphitization. In the 2840–3020 cm⁻¹ region, aliphatic C–H vibrations diminished progressively from CRD wood to B800 due to dehydration, demethylation, and condensation of terminal –CH₃ and –CH₂ groups from depolymerizing holocellulose [54], aligning with [52]. This implies incomplete decomposition in B400 and B500, consistent with their high VC, O/C, and H/C ratios.

In the 2100–2260 cm⁻¹ region, C≡C alkyne stretching associated with condensed polyaromatics increased from B400 to B800, correlating positively with the pyrolysis temperature [51]. C≡N stretching (~2250 cm⁻¹) also intensified, indicating growing nitrile group presence and enhanced structural rigidity. Between 1740 and 2100 cm⁻¹, increasing C=C=C allene stretches from B400 to B800 signaled strain propagation during graphitization due to adjacent dienes, as reported in [51]. C=O stretches, linked to lignin breakdown, became more pronounced at higher temperatures. Lignin-derived carbonyls (1610–1740 cm⁻¹) diminished from B400 to B800 due to degradation under thermal severity [53]. These carbonyls are products of aldehydes, acids, ketones, and esters [51,52].

C=C ring stretching (1510–1610 cm⁻¹) was most intense in B400–B600, marking holocellulose breakdown and structural deprotonation, which increased biochar hydrophobicity. The 1440–1510 cm⁻¹ band reflected early-stage aromatic skeletal vibrations with oxygen and hydrogen loss. C–OH stretching and O–H bending (1280–1440 cm⁻¹) contributed to aromatic backbone formation. A decline in 1020–1280 cm⁻¹ peaks from CRD wood to B800 indicated decomposition of hemicellulose and cellulose, as seen through weakening C–O, C–OH, C–H, and C–O–C skeletal vibrations [54,55]. These labile groups in B400 and B500 contribute to higher CEC [56]. Finally, peaks in 700–840 cm⁻¹, related to surface acidic functionalities, were eliminated at higher pyrolysis temperatures.

SEM-EDX Spectroscopy

SEM analysis of B400 [Figure 8A–C], B600 [Figure 8D–F], and B800 [Figure 8G–I] at 250×, 500×, and 1000× magnifications, respectively, was conducted without gold sputtering. Higher pyrolysis temperatures yielded partly smooth surfaces, while lower temperatures resulted in pronounced roughness with residual cell walls and fibers. B400 micrographs revealed minimal porosity, as only hemicellulose and partial cellulose decomposition occurred, with lignin largely intact. This led to sporadic macropores, as noted in [38], making B400 suitable for soil amendments due to its water retention, aeration, and microbial compatibility [57]. Low-temperature chars like B400 have also been used as additives to boost methane yields in anaerobic digestion [58]. Surface oxidative residues contribute to reactivity, supporting their use as solid fuels for energy purposes [59].

At intermediate pyrolysis temperature, B600 developed more defined pores [Figure 8F], following lignin softening, melting, and breakdown after holocellulose decomposition [60]. Depolymerization and monomer cross-linking increased solid char yield [61]. Rapid volatile release and collapsing cell walls expanded ordered porosities, causing shrinkage and growth of FC-rich layers, particularly with a longer BRT [59]. As the temperature approached 600 °C, the pore sizes diminished [62]. Its surface area and thermal maturity support carbon sequestration, metallurgical use, and the adsorption of dyes and heavy metals. The loss of surface acidic groups renders B600 alkaline, enabling acidic soil remediation [59]. At higher temperatures, B800 exhibited smooth, honeycomb-like structures [Figure 8A–I], indicating complete carbonization and surface volatile removal. High carbon content, TSF (low VC/FC ratio), and low H/C and O/C ratios promote the formation of stable graphitic sheets ideal for energy materials (e.g., supercapacitors, electrodes) and carbon sequestration. This trend with rising temperature was confirmed by [63]. From [64], pore expansion likely arises from micropore wall cracking and mesopore coalescence, reducing surface area and micropore volume in B700 and B800—a finding already discussed under the Physicochemical Characterization Section.

Pore blockage may occur due to increasing ash and AAEM concentrations, which reduce active surface sites up to a certain temperature. Beyond this, less stable monovalent AAEMs like Na and K volatilize at higher temperatures, aiding pore reaming. SEM-EDX spectra [Figure 9] confirm the presence of AAEMs on biochar surfaces. As shown in Figure 9A–C, surface carbon increased from 81.6% (B400) to 95.3% (B600) and 96.2% (B800), as also noted in [63]. Conversely, surface oxygen content dropped by ~80%, revealing an inverse relation with pyrolysis temperature. EDX peaks indicated trace levels (0.2–0.4%) of AAEMs (Ca, K, Na), which may form salts (carbonates, oxides, chlorides), elevate pH, and enhance CEC for electrostatic interactions [65]. These elements were dispersed across the biochar, including near or within pores [66]. Sulfur (probably from gypsum drywall in CRD wood residues) appeared at 0.1% in B400 but volatilized at higher temperatures as SOx. Overall, biochar morphology is highly tunable based on feedstock composition (with or without pretreatment) and pyrolysis conditions—directly influencing its applicability in the environmental, energy, and agricultural sectors.

Raman Spectroscopy

Raman spectroscopy was employed to detect structural transitions from amorphous to crystalline carbon in biochars produced at pyrolysis temperatures ranging from 400 to 800 °C. Figure 10 illustrates the spectral bands, with the I_D/I_G index—i.e., the ratio of the intensity of deformations or disorderliness in amorphous carbon (D-band) to that of the quantity of graphitic carbon structures (G-band)—used as a quantitative measure of graphitization. A higher I_D/I_G ratio implies greater structural disorder or deformation in amorphous carbon, indicating a progressive transformation toward turbostratic carbon arrangements under thermal stress [67]. In contrast, highly crystalline carbonaceous materials like graphite or graphene exhibit low I_D/I_G values, reflecting minimal lattice defects [68].

For biochars B400-B800, the I_D/I_G index ranged from 0.78 to 0.97, supporting the conclusion that increasing heat treatment intensifies defects arising from incomplete graphitization. This observation is consistent with [68], confirming that pyrolysis induces restructuring in disordered carbon domains. As noted by [69], these defect-rich zones, particularly in low-temperature biochars, harbor functional groups such as hydroxyls and carboxylic acids, which contribute beneficially to adsorption tendencies. Notably, the I_D/I_G ratio tends to rise with the temperature until most disordered carbon reorganizes into sp²-bonded graphitic layers. At this stage, the trend may plateau or reverse, signaling nearly complete graphitization. The authors of [67] also showed that incorporating chemical activators like sulfur and boron into biochar can facilitate this transition by reducing the activation energy required for lattice rearrangement, enabling graphitization even at sub-1000 °C temperature ranges, significantly below the 2000 °C typically required for full transformation.

Once graphitization reaches a maximum, biochar becomes ideal for carbon sequestration, even in sensitive biotic or abiotic environments. Such thermally stable materials are also promising candidates for catalytic applications due to enhanced electron transfer capabilities [45]. Furthermore, their low oxygen content and high mechanical strength render them suitable as partial coke substitutes in high-temperature metallurgical operations, such as blast furnaces, where it is advised that premature reactivity be avoided. Under these elevated conditions, labile inorganic impurities gradually volatilize, reducing ash and volatile matter. In fact, as reported in [50], coal-derived gangue has been co-pyrolyzed with lignocellulosic biomass to improve I_D/I_G values (0.5–0.8), showing enhanced structural evolution with temperature.

Metal Content

AAEMs in biochars were concentrated up to 600 °C at 92.28%, after which they may have decreased due to volatilization. Relative to CRD wood, AAEM content is higher in all biochars due to shrinkage in biomass volume (loss of volatiles) and the breakdown of organo-metallic linkages with increasing temperatures [70]. Non-volatile trace metals like Fe, Cu, Cr, Zn, Ni, Pb, Co, and Cd are present in varying percentages, which could be explained by advanced treatments that these CRD wood residues were subjected to initially (e.g., chromated copper arsenates or ACC) during their applications in regional construction projects, such as houses, electricity poles, piers, and bridge components. That is why treated wood, as in CRD wood waste, if openly burnt for energy purposes, will be classified as hazardous in nature according to the United States Environment Protection Agency (USEPA) [1]. From [71], the upper and lower (permissible) limits of some metallic constituents in biochar are very important to consider. In B400–B800 evaluated here, it is conclusive that hazardous metal content is within the prescribed range and meets the stipulations set forth by renowned biochar standards like those of the International Biochar Initiative (IBI), European Biochar Certificate (EBC), Australia New Zealand Biochar Industry Group (ANZBIG), Biochar Quality Mandate (BQM), and Singapore standard SGS). Metals below the detection limits in CRD wood and its biochars were Hg, As, and Se. Also, Fe was the most abundant metal (non-AAEM entity) in CRD wood (16,086 mg/kg) and its biochars (from 15,546 mg/kg in B400 to 16,699 mg/kg in B800) due to the presence of embedded remnants of nuts, bolts, nails, and joints, once part of the feedstock before disassembly or demolition. The metal analysis of all tested biochars has been tabulated below in

Table 4. Metal content in B400–B800 and the permissible limit set by the IBI.

Sample	Al	Ba	Ca	Cd	Co	Cr	Cu	Fe	K	Mg	Mn	Mo	Na	Ni	Pb	V	Zn	Total	AAEM	AAEM
	mg/kg	mg/kg	mg/kg	mg/kg	mg/kg	mg/kg	mg/kg	mg/kg	mg/kg	mg/kg	mg/kg	mg/kg	mg/kg	mg/kg	mg/kg	mg/kg	mg/kg	mg/kg	mg/kg	%
CRD wood	12,043	1165	149,284	2	69	86	613	16,086	38,043	17,566	5587	24	37,685	5	561	21	1389	280,229	243,742	86.98
B400	7707	1949	163,961	2	58	79	5777	15,546	49,507	17,794	6799	21	37,901	41	381	17	1484	309,026	271,113	87.73
B500	4273	1554	143,017	4	57	112	300	8861	36,641	15,503	7649	<2	44,478	17	450	11	1924	264,852	241,194	91.07
B600	3505	1834	165,480	11	51	179	237	11,898	68,529	20,299	7576	<2	41,471	92	296	13	1038	322,507	297,612	92.28
B700	8856	1900	161,627	9	39	179	231	10,387	38,942	15,529	3883	11	23,800	104	210	24	1316	268,940	241,797	89.91
B800	9680	3889	167,379	14	60	318	2297	16,699	44,660	20,019	8303	19	26,990	82	225	27	76	300,738	262,938	87.43
IBI (low)				0.3		15	40					10		10	10		150
IBI (high)				39		1200	6000					75		600	500		7400

.

3.1.4. Choosing a Potential Adsorbent Material for Dye Removal Tests from Contaminated Water—B600

B600 was selected as the optimal adsorbent for the removal of methyl orange from aqueous solutions. Although another biochar produced at the same temperature range (Run 8) exhibits slightly higher surface area and micropore volume, B600 surpasses it in elemental composition (carbon, hydrogen, oxygen), FC, and molar ratios (H/C, O/C), all of which are critical indicators of a biochar’s adsorption potential. These parameters influence not only hydrophilicity and chemical/thermal stability but also the ability to form electrostatic, covalent, and hydrogen bonds through surface functional groups, making B600 more suitable for adsorption of pollutants in water across variable temperature and pH conditions. This aligns with findings from [72], which emphasize that surface area alone does not dictate adsorption performance. Since methyl orange is a polar dye that readily dissolves in polar solvents like water, it is essential that the adsorbent’s surface possesses adequate polarity to interact favorably. B600’s relatively high H/C ratio suggests a moderate positive surface charge, which may reduce electrostatic repulsion between negatively charged functional groups on the biochar and the anionic dye, enhancing the overall adsorption efficiency.

In contrast, biochars like B700 and B800, though more thermally mature, have lower H/C and O/C ratios, indicating increased hydrophobicity. These may be better suited for adsorbing non-polar or hydrophobic contaminants in wastewater. However, for a polar molecule like methyl orange, B600 offers a balanced profile of surface chemistry and porosity. Additionally, B600 retained a significant surface area (323.78 m²/g)—the second highest across the entire DOE and impressively achieved without any post-pyrolysis activation (chemical, thermal, physical, or mechanical). While it is still lower than surface areas seen in activated carbons, this value is considerable for raw biochar. Given the molecular size of methyl orange (<10 nm), adsorption via dense micro- to mesoporous channels is feasible, similar to mechanisms seen in mineral-based adsorbents [73]. There is a possibility that if a pore-filling mechanism takes place during adsorption, and if the porosity of biochar is too high, desorption of the dye molecules may occur!

3.2. CRD Wood Pyrolysis in Horizontal Rotary Retort-Furnace Reactor—Scale-Up from Horizontal Tube Furnace Reactor

The same batch of contaminant-separated, dried, and size-reduced CRD wood (3 mm–2.5 cm) was subjected to scaled-up pyrolysis in a horizontal rotary retort-furnace reactor. The exact temperature and BRT conditions listed in the DOE for B400–B800 were replicated to enable an approximate comparison with the horizontal tube furnace reactor experiments. These samples were selected to represent a broad pyrolysis temperature range, which critically influences biochar properties. The results are tabulated in

Table 5. Yield, proximate, and elemental analysis results for B400-B800 in the horizontal rotary retort-furnace reactor.

Sample	N (wt%)	C (wt%)	H (wt%)	S (wt%)	Ash (wt%DB)	O (wt%)	H/C	O/C	VC (wt%DB)	FC (wt%DB)	TSF (%)	Yield (%)
B400	0.650	78.380	2.887	0.184	2.640	15.259	0.442	0.146	26.739	70.621	72.536	37.700
B500	0.380	80.580	2.489	0.089	4.120	12.342	0.371	0.115	23.451	72.429	75.541	32.100
B600	1.150	81.690	2.410	0.372	4.050	10.328	0.354	0.095	17.992	77.958	81.249	29.500
B700	0.580	83.790	2.149	0.114	5.970	7.397	0.308	0.066	14.858	79.172	84.199	27.900
B800	0.740	85.610	2.042	0.133	5.820	5.655	0.286	0.050	12.613	81.567	86.608	26.200

Note: Only proximate and elemental analyses were conducted for scaled-up biochars. All advanced characterizations were reserved only for samples produced in the horizontal tube furnace reactor, as in the former case.

.

Compared to the previous experiments, yields were more reliable this time, particularly for B400, which underwent complete pyrolysis. Similar yields for 400 °C (B400: 37.7%) and 600 °C (B600: 29.5%) were also reported in [74]. However, the carbon content in B600–B800 was 5–7% lower, despite a strong positive correlation with temperature (r = 0.99). This is likely due to the larger feedstock mass (3 kg) used in the rotary retort, compared to 50–250 g in the horizontal tube furnace reactor, indicating heat transfer limitations despite rotation. Additional factors could include (a) carbon loss from the secondary cracking of volatiles due to low inert gas flow and (b) ambient air infiltration causing mild combustion. Still, ash levels remained consistent with tube furnace tests. Slightly elevated H/C and O/C ratios for B600–B800 suggest some residual surface volatiles, yet both indices remain well within accepted limits, indicating good carbonization and aromaticity with negative correlations with temperature (O/C: r = −0.99; H/C: r = −0.98).

As noted in [71], CRD wood biochars produced at higher pyrolysis temperatures meet superior-grade criteria, with O/C < 0.2 and H/C < 0.4, indicating low VC, high FC, strong aromatization, and long-term stability for soil and carbon sequestration. Their possible resistance to oxidation, hydrophilic interactions, and microbial degradation also aligns with Van-Krevelen parameters. The IBI states that it is now better to use the index H/C_org in place of H/C since carbon content as a whole will also include inorganics like carbonates in the ash fraction and not be selectively describing the organic carbon aromatics of interest in biochar, which is the baseline condition for any thermochemical conversions to have taken place relative to the original biomass feedstock [37]. Proximate values showed FC increasing by 15% and VC decreasing by 53% from B400 to B800, with r = 0.98 and −0.99, respectively. TSF rose by 19% (r = 0.99), indicating enhanced thermal stability. HHV values (29–32 MJ/kg) aligned with those in earlier tests, confirming suitability for energy and metallurgical applications.

Overall, scalability experiments showed good relatability in terms of proximate values, elemental properties, and a similar dominant effect of pyrolysis temperature. Future studies entailing pilot-scale conversions can thus be recommended, where at least 15–20 kg of CRD wood can be treated at once via batch pyrolysis.

3.3. The Effect of Feedstock Composition and Pyrolysis Process Parameters on Biochar Properties—Comparison to Similar Works

Biochar’s physicochemical properties depend heavily on biomass composition and pyrolysis conditions, both of which must be optimized for applications like carbon sequestration, metallurgy, and pollutant adsorption. Lignocellulosic biomass, such as forest residues, CRD wood, and crop waste, yields stable, carbon-rich biochars due to its lignin, cellulose, and hemicellulose content, which decompose over a broad temperature range [75]. Lignin-rich feedstocks can exceed 80% carbon, while cellulose-rich ones yield more volatiles [76]. In contrast, algae-derived biochars are typically unstable, with low carbon and surface area due to high N, P, S, and inorganic contents [77]. Co-pyrolysis, combining weak feedstocks like biosolids with lignocellulosic materials, improves the elemental balance, porosity, and surface area (up to 300–400 m²/g) while preserving minerals [78].

Among process variables, temperature is critical. Above 500 °C, the biochar yield drops, but carbon content, aromaticity, hydrophobicity, and surface area improve [79], as seen in B600–B800. Around 700 °C, structural ruptures from surface treatments increase porosity for gas adsorption [80]. Lower temperatures (<400 °C) preserve nutrients and functional groups, enhancing water retention [81,82]. The surface area can rise from 150 m²/g at 300 °C to over 500 m²/g at 700 °C [83,84]. Micropores predominate at 450–600 °C, while mesopores form at temperatures >600 °C due to bulk volatile release, aligning with our surface area trends and favoring water filtration or catalysis applications [85]. Despite the reduced surface area at 700–800 °C, high FC, low VC, and thermal stability make such biochars suitable for metallurgical use. BRT also plays a key role. An extended BRT promotes uniform heat transfer, deeper carbonization, and stable aromatization [86], reducing H/C ratios to <0.3 [80] and O/C to <0.2 at 700 °C [80,87]. High carbon yields (>85%) and surface areas (~500 m²/g) are also reported under these conditions. A long BRT enhances FC while reducing volatiles, improving stability [88,89]. In contrast, a short BRT (<30 min) yields reactive but less stable chars. Higher pyrolysis temperatures also induce alkaline pH, aiding acidic soil remediation [81]. Long BRTs in slow-heating fluidized reactors further improve gas–solid interaction and morphology and subdue premature volatile condensation [90]. The feedstock mass impacts heat transfer and carbonization. Smaller masses, especially under low heating rates, moderate temperatures, and long BRTs, enhance porosity and moisture retention, benefiting adsorption and soil applications [91,92,93,94]. Larger masses risk uneven heating, incomplete devolatilization, and inferior structure. Therefore, understanding the interplay between feedstock traits and pyrolysis conditions is key to tailoring biochar for targeted outcomes.

3.4. Methyl Orange Batch Adsorption Experiments—An Application for CRD Wood Biochar (B600) from the DOE

As a practical application derived from the DOE runs, adsorption experiments were conducted to assess the removal of the industrial dye methyl orange from contaminated water. Among the produced biochars, B600 was selected as the adsorbent due to its optimal combination of morphological features, surface chemistry, and thermal stability. These attributes made B600 a suitable candidate for water treatment, and it was tested across varying dosage levels of the pollutant dye during the adsorption trials.

3.4.1. Adsorption Standard Determination

To begin with, the methyl orange indicator with a concentration of 0.05 w/v (i.e., 0.05 g per 100 mL or 0.5 g/L) was sourced from the institute’s pulp and paper analytical laboratory. Using its molar mass of 327.33 g/mol, the molarity of the stock solution was calculated as 1.5 mM. Reference solutions of 10 mL total volume were then prepared at varying concentrations—0.001, 0.01, 0.0175, 0.025, and 0.0375 mM—by applying the dilution formula C₁V₁ = C₂V₂, where C₁ is the original concentration (1.5 mM), V₁ is the volume of stock to be used, C₂ is the target concentration, and V₂ is the final volume (10 mL). The absorbance values of each solution were measured using a UV–visible spectrophotometer (HACH—DR 6000) at 464 nm under alkaline conditions (pH ≥ 4.4, orange color) and at 506 nm under acidic conditions (pH ≤ 3.2, pink color). The absorbance data were fitted to Beer–Lambert’s law, yielding the intercept, molar absorptivity coefficient, and an excellent linear correlation, with R² = 0.9995.

3.4.2. Batch Adsorption Experiments

For the adsorption experiments (key steps shown in Figure 11), 50 mL test solutions with varying initial concentrations of methyl orange were prepared using deionized water to eliminate interference from competing ions. All tests were conducted in 250 mL reaction flasks placed on a magnetic stirrer (300 rpm) with temperature control. Two concentrations, 0.025 mM (8 ppm) and 0.5 mM (164 ppm), were selected based on prior studies (Supplementary Table S7). Other process parameters were varied to evaluate dye removal efficiency: pH (1–9, adjusted using 0.1 M H₂SO₄ and 0.1 M NaOH), temperature (20–50 °C or 293–323 K), contact time (up to 24 h), and B600 biochar dosage (1–10 mg/mL). At set intervals, solutions were filtered using 0.45 µm syringe filters to separate B600 particles; vacuum filtration with papers of similar porosity was also applied. UV-vis absorbance readings were then compared against the standard calibration curve to determine residual methyl orange concentrations, removal percentages, and adsorption capacity (q_e). Each experiment was repeated three times, and average values are reported.

The fundamental equations governing dye adsorption estimation methods, particularly removal percentage, q_e, and kinetic/equilibrium/thermodynamic models, can be referred to in Supplementary Equations (S13)–(S34).

3.4.3. Optimization of Adsorption Parameters and Results

Before conducting adsorption isotherm and kinetic studies, two parameters—pH and B600 dosage—were optimized. Based on Supplementary Table S7, the initial methyl orange concentration tested was approximately 150 ppm, consistent with values reported in [95]. Accordingly, 164 ppm was selected for the test solution due to its close range. Similarly, the maximum temperature (50 °C or 323 K) and contact time (24 h) were chosen after reviewing comparable studies involving biochar adsorbents derived from pyrolyzed biomass for methyl orange removal (Supplementary Table S7). The 24 h duration was intentionally selected to extend the observation window and capture potential saturation behavior, if any.

(A) Effect of pH

The pH range of 1–9 (values: 1, 3, 5, 7, 9) was tested to assess dye removal based on methyl orange’s affinity to biochar. Since biochars produced at higher pyrolysis temperatures tend to have alkaline pH and negatively charged surfaces, they are less favorable for adsorbing anionic dyes like methyl orange unless surface protonation occurs. A high AAEM content in CRD wood, as per [96], likely contributed to B600’s alkaline pH (9.7) through the presence of Ca, Mg, K, and Na salts, carbonates, and oxides and the loss of acidic surface groups with increasing temperature. B600’s surface area, micropore volume, Van-Krevelen molar ratios, FC, and thermal stability further supported its selection for adsorption tests, especially under acidic conditions and elevated temperatures. A low pH enhances the protonation and anion exchange capacity (AEC) of the biochar surface, promoting adsorption [97]. Alternatively, acid activation can also be used to shift surface charges positively [98]. The electrostatic attraction mechanism is further explained by the pHpzc of B600, measured at 9.02 (Figure 12A); below this pH, B600’s surface carries positive charges that attract negatively charged dye molecules. Above pHpzc, the surface becomes negative and repels methyl orange. Despite the high pHpzc, maximum dye removal occurred at solution pH 1–3.

Confirming Changes to B600 Surface Chemical Composition Post-pH Modification

To confirm surface-level adsorption after pH modification across methyl orange concentrations (8–164 ppm), FTIR spectra of B600 post-adsorption were analyzed (Figure 12B). Most spectral features of native B600 remained unchanged, though three subtle differences were observed: (a) a mild drop in peak intensity between 1020 and 1280 cm⁻¹ suggests the masking of oxygenated groups (C–O, C–O–C), possibly due to interaction with positively charged sulfonic acid groups of methyl orange under acidic conditions; (b) slight peaking between 1280 and 1440 cm⁻¹ at higher dye concentrations may reflect O–H bending from alcohols and phenols formed via heteroatom abstraction by methyl orange; and (c) the emergence of peaks in the 1610–1740 cm⁻¹ range indicates C=C and amide C=O stretching, pointing to π–π interactions between biochar aromatics and methyl orange’s benzene rings—absent in native B600. As noted in [99,100], acidic media can also promote hydrogen bonding and π–π interactions beyond electrostatics. Conversely, at higher pH, hydroxyl groups dominate the B600 surface, potentially competing with methyl orange and increasing repulsion on the now negatively charged surface [101].

The concept of electrostatic attraction is evident from Figure 13A, where methyl orange removal peaked at pH 1 (93.53%) and declined progressively at higher pH levels: 88.64% (pH 3), 60.56% (pH 5), 48.09% (pH 7), and 33.72% (pH 9). Similar behavior was observed in [102], indicating that mass transfer and equilibrium uptake (q_e) are enhanced in acidic media. Methyl orange’s structure supports this: at low pH, it can carry a Na⁺-derived positive charge and an additional HSO₃⁺ charge (in contrast to its sulfonate SO₃⁻ form at higher pH) [16]. This facilitates a charge transfer mechanism with B600, where surface protonation at low pH promotes effective adsorption of the anionic dye. However, since pH 1 is too harsh for practical use, pH 3, where dye removal was still high, was selected for subsequent adsorption experiments.

(B) Effect of Adsorbent Dosage

Methyl orange removal was studied at B600 dosages of 1, 3.5, 5, 7.5, and 10 mg/mL. Maximum dye removal (93.48%) occurred at 10 mg/mL due to abundant adsorption sites, with a decreasing trend observed at lower dosages: 89.24% (7.5 mg/mL), 72.33% (5 mg/mL), 46.04% (3.5 mg/mL), and 25.89% (1 mg/mL), as shown in Figure 13B. Similar behavior was noted with Congo red dye in [101]. However, despite higher removal efficiency, q_e decreased at higher dosages due to a diminished concentration gradient and less driving force for diffusion. With excessive binding sites and limited dye per unit mass, many sites remain unutilized or overlap, reducing accessibility and the surface area [103,104,105]. Compared to other studies, as shown in Supplementary Table S7, this work used a 100-fold higher biochar loading, leading to significantly lower q_e. For instance, at 10 g/L (10 mg/mL) and 164 ppm dye, q_e is low. However, reducing the dosage to 1 g/L could yield qe ≈ 146 mg/g, and at 0.5 g/L, qe ≈ 293 mg/g—making the results comparable to others. Hence, the adsorbent dosage is inversely related to q_e but directly related to dye removal efficiency. Increasing the adsorbate–adsorbent ratio or using lower, optimized dosages with high-binding-energy sites can enhance performance.

(C) Effect of Initial Dye Concentration—Adsorption Isotherms

As the methyl orange concentration increased from 8 to 164 ppm, its removal dropped from >99% to 89% (Figure 13C), consistent with the findings from [106], which employed another azo dye called ‘Acid Orange 7’. This suggests that at higher concentrations, not all B600 sites have a strong affinity, leading to weaker interactions and dye aggregation. As noted in [107,108], adsorption plateaus beyond the optimal concentration since only high-energy sites remain effective. At low concentrations, strong driving forces ensure rapid adsorption, but these weaken with saturation. Similar trends were observed in [109]. High dye and adsorbent levels can also cause mass transfer resistance and fewer collisions, limiting efficiency despite increased q_e. Lowering the B600 dosage at high dye concentrations may improve q_e and adsorption kinetics, as seen in [110].

Adsorption isotherms using Langmuir, Freundlich, and Temkin models were applied to describe dye-adsorbent equilibrium interactions, similar to [111,112,113]. Freundlich and Langmuir showed strong fits with R² = 0.97 and 0.94, respectively (Figure 14A,B;

Table 6. Adsorption isotherm parameters.

Average	Langmuir Isotherm					Freundlich Isotherm				Temkin Isotherm
qe,exp (mg/g)	qm (mg/g)	qe,cal (mg/g)	KL (L/mg)	RL	R2	KF [(mg/g)(L/mg)1/n]	1/n	qe,cal (mg/g)	R2	B (J/mol)	KT (L/mg)	qe,cal (mg/g)	R2
5.52	5.00	3.56	3.33	0.002–0.036 from 164–8 mg/L	0.94	2.92	0.50	5.15	0.97	2.25	8.79	5.52	0.59

).

The Freundlich model indicates multilayer adsorption on heterogeneous B600 sites, supported by 1/n = 0.5 (favorable, since 0 < 1/n < 1) and K_F = 2.92, with reference to [114,115,116,117]. The low q_e and K_F are due to the high dosage (10 g/L) and increasing dye concentration. Nevertheless, model reliability is high (RMSE = 0.216, χ² = 0.03), with q_e,cal ≈ q_e,exp, indicating both physisorption and chemisorption may occur [16,109,110]. The Langmuir model also suggests favorable adsorption (0 < R_L < 1) as explained conceptually in [118,119,120], likely as a monolayer, with minimal dye aggregation and homogeneous active sites (K_L = 3.33) [69,107,108,121]. Though it has slightly higher errors than Freundlich (RMSE = 1.133, χ² = 1.08), the fit remains good. The lower q_m than q_e,exp may stem from site competition or experimental factors. The Temkin fit (Figure 14C) was weaker (R² = 0.59), though errors were minimal (RMSE = 0.002, χ² = 0), implying some degree of applicability. The low B value (2.25 J/mol) hints at low adsorption heat and physisorption dominance at higher dye levels. As the concentration rose from 8 to 164 ppm, q_e increased linearly, but with decreasing heat of adsorption as fewer high-energy sites remained [122]. Thus, Temkin supports a weaker interaction model, ranking behind Freundlich and Langmuir in overall reliability [123].

To summarize, the Freundlich model followed by the Langmuir isotherm model indicates favorable methyl orange adsorption over B600, where both monolayer and multilayer interactions can be deduced. Despite being less contrary to the theory, the Temkin isotherm is ranked last here only due to the low model correlation factor.

(D) Effect of Adsorption Duration—Adsorption Kinetics

As the contact time increased from 0 to 240 min, complete equilibrium was not yet reached (Figure 13D). However, both q_t and removal efficiency steadily increased: from 0.64 to 0.73 mg/g (77–89%) for 8 ppm and from 11.99 to 14.49 mg/g (73–89%) for 164 ppm. Over 50% of dye removal occurred within the first 30 min, indicating a rapid initial adsorption phase [111]. Despite the higher dye concentration, the curve shape suggests removal rates were not significantly impacted, likely due to a limited mass transfer gradient and increased competition for active sites. By 24 h (used in isotherm experiments), maximum q_e (14.63 mg/g) and removal (89.39%) were recorded, which were only slightly higher than those at 240 min (by 0.14 mg/g and 0.85%, respectively), indicating near-saturation. Beyond this, minimal change suggests that most B600 sites were occupied. Any further adsorption was likely limited by electrostatic repulsion, site saturation, and slow intra-particle diffusion.

As for the kinetics involving initial methyl orange concentrations of 8 ppm and 164 ppm, pseudo-first-order, pseudo-second-order, and intra-particle diffusion models were used to calculate adsorption parameters (

Table 7. Adsorption kinetic parameters.

Dye Concentration	Pseudo-First-Order Kinetics			Pseudo-Second-Order Kinetics			Intra-Particle Diffusion Kinetics
	qe (mg/g)	k1 (min−1)	R2	qe (mg/g)	k2 (g/mg.min)	R2	k3 (mg/g.min1/2)	C (mg/g)	R2
Linear fitting with linear regression
8 ppm	0.194	0.004	0.978	0.748	0.158	1.000	0.010	0.583	0.983
164 ppm	2.752	0.012	0.951	14.872	0.010	1.000	0.214	11.473	0.772
Non-linear fitting with non-linear regression
8 ppm	0.700	0.059	0.495	0.726	0.254	0.879	0.010	0.583	0.983
164 ppm	14.107	0.056	0.517	14.716	0.011	0.890	0.214	11.473	0.772

), like in [99,124,125,126,127,128]. Both linear (Figure 15A–C) and non-linear (Figure 15D,E) fitting plus regression methods were evaluated and compared, as in [97]. All error estimations for analyzing the closeness between q_e,exp and q_e,cal for both linear and non-linear kinetic modeling are computed separately in

Table 8. RMSEs derived from comparison of q_e,exp and q_e,cal for two initial methyl orange concentrations (Co: 8 ppm and 164 ppm) during adsorption kinetic model evaluation.

Kinetic Model	Co (ppm)	Average qe,exp (mg/g)	qe,cal (mg/g)	RMSE
Linear fitting with linear regression
Pseudo-first-order	8	0.812	0.194	0.234
Pseudo-first-order	164	14.630	2.752	4.489
Pseudo-second-order	8	0.812	0.748	0.024
Pseudo-second-order	164	14.630	14.872	0.091
Intra-particle diffusion	8	0.812	0.583	0.087
Intra-particle diffusion	164	14.630	11.473	1.193
Non-linear fitting with non-linear regression
Pseudo-first-order	8	0.812	0.700	0.042
Pseudo-first-order	164	14.630	14.107	0.198
Pseudo-second-order	8	0.812	0.726	0.033
Pseudo-second-order	164	14.630	14.716	0.033
Intra-particle diffusion	8	0.812	0.583	0.087
Intra-particle diffusion	164	14.630	11.473	1.193

.

The pseudo-second-order model showed excellent linearity (R² = 1) for both 8 ppm and 164 ppm methyl orange concentrations, indicating chemisorption as the primary mechanism [129]. This suggests strong, slow-forming chemical bonds between B600 and methyl orange, consistent with systems like Zn-activated sawdust biochar [130]. Functional groups on B600, particularly oxygen- and hydrogen-containing moieties, likely interact with methyl orange’s sulfonic acid groups via electron sharing and diffusion-driven processes [131]. A high B600 dosage ensures abundant active sites, making site availability and dye concentration the key limiting factors. The q_e,cal values (0.75 mg/g for 8 ppm and 14.87 mg/g for 164 ppm) closely matched q_e,exp, with RMSEs of 0.02–0.09. Non-linear fits confirmed the trend (q_e,cal = 0.73 and 14.72 mg/g for 8ppm and 164 ppm; RMSE = 0.03), though R² decreased slightly to 0.88 and 0.89. This validates the pseudo-second-order model, particularly at lower concentrations, where k₂ > k₁, reinforcing chemisorption’s role.

For the pseudo-first-order model, linear regression yielded high R² values of 0.98 (8 ppm) and 0.95 (164 ppm), suggesting a role for physisorption, particularly influenced by methyl orange diffusion at the solid–liquid interface. At 8 ppm, dye molecules interact uniformly with available B600 sites, while at 164 ppm, site saturation and competitive interactions reduce efficiency. This is reflected in the large gap between q_e,cal (2.75 mg/g) and q_e,exp (14.63 mg/g) at 164 ppm, with a high RMSE of 4.49, compared to just 0.23 at 8 ppm. This implies that pseudo-first-order kinetics better describe adsorption at low dye concentrations. In contrast, non-linear fitting produced very low R² values (0.50 and 0.52), questioning its reliability despite close matches between q_e,cal and q_e,exp at 164 ppm (14.11 versus 14.63 mg/g; RMSE = 0.20). Across both methods, k₁ < 0.1 min⁻¹, indicating that weak forces and diffusion limitations govern the process, consistent with physisorption behavior.

For the intra-particle diffusion model, linear and non-linear fits showed a strong R² of 0.98 at 8 ppm, but only 0.77 at 164 ppm. This drop suggests that higher dye concentrations cause diffusion barriers, limited pore access due to B600 agglomeration, and the exhaustion of high-energy binding sites. The q_e,cal rose from 0.58 mg/g (8 ppm) to 11.47 mg/g (164 ppm), with RMSE increasing from 0.09 to 1.19. At 8 ppm, k₃ > k₁, highlighting intra-particle diffusion’s dominance over pseudo-first-order kinetics. At 164 ppm, k₃ exceeds both k₁ and k₂, indicating that once surface sites are saturated, the dye continues diffusing into pores until mass transfer resistance and full saturation limit further uptake. Thus, adsorbate transport, active site binding, and B600 porosity are key to the adsorption process.

To sum up, physisorption, chemisorption, and molecular diffusion within the pores of the adsorbent are the key adsorption mechanisms identified, each of which harbors one or multiple rate-limiting steps with respect to the concentration of the adsorbate, adsorption sites, binding interactions, and finally, pore diffusion.

(E) Effect of Temperature—Adsorption Thermodynamics

The theory behind adsorption thermodynamics were similar to [132,133] and the calculated thermodynamic parameters for the adsorption process are presented in

Table 9. Adsorption thermodynamic parameters and the influence of temperature.

Enthalpy	Entropy	Gibbs Free Energy
ΔH (KJ/mol)	ΔS (KJ/mol.K)	Temperature (K)	ΔG (KJ/mol) = ΔH-TΔS	ΔG (KJ/mol) = −RTlnkd
144.153	0.460	293	9.419	8.376
		303	4.823	5.437
		313	0.227	2.488
		323	−4.369	−6.207

. As the temperature increased from 20 °C to 50 °C (293–323 K), q_e rose from 0.20 to 0.81 mg/g and removal efficiency improved from 24 to 99% for an initial methyl orange concentration of 8 ppm, aligning with observations in [134]. This temperature dependence necessitates thermodynamic evaluation. From the lnK_d-versus-1/T plot, a ΔH value of 144.15 kJ/mol was calculated, indicating a strongly endothermic process characteristic of chemisorption, likely involving covalent, hydrogen, and π–π bonding. The high-energy input facilitates increased dye molecule mobility, leading to more frequent and forceful collisions with B600 surfaces [102]. A positive ΔS (0.46 kJ/mol·K) points to increased disorder as dye molecules move from the bulk solution to the adsorbent interface, allowing for interfacial diffusion and eventual attachment. However, entropy decreases once adsorption completes and molecular motion becomes restricted. The calculated ΔG values ranged from +8.38 kJ/mol at 20 °C to –6.21 kJ/mol at 50 °C, indicating that adsorption becomes increasingly spontaneous at higher temperatures, as external energy requirements decline and system favorability rises [135].

3.4.4. Adsorption Mechanism and Theory

Being a surface-driven process, methyl orange molecules migrated from the bulk solution to B600’s surface, where adsorption was influenced by factors like surface charge, functional groups, pore structure (micro- to mesopores), and energy-diverse binding sites. Ambient pH, temperature, contact time, adsorbent dosage, and dye concentration also played crucial roles. Under acidic conditions, B600’s surface became positively charged, promoting electrostatic attraction with the anionic azo dye and minimizing repulsion [136]. FTIR analysis revealed hydroxyl, carboxyl, and amide groups on B600, which facilitated covalent and hydrogen bonding with methyl orange’s sulfonate group at low pH.

As observed in Figure 9D, the adsorption of methyl orange affected the surface elemental composition of B600, which is evident from a drop in surface carbon from 95.3% to 94.1% and an increase in surface oxygen from 4.5% to 5.6% after adsorption. Such strong chemical interactions resulted in chemisorption, which was affirmed by the pseudo-second-order kinetic model, where adsorption efficiency increased and showed signs of irreversibility. This may be evident from the Langmuir isotherm parameter R_L, which was close to 0 (still > 0 and favorable). The Langmuir isotherm’s monolayer adsorption with a fixed number of active binding sites with high energies and the Freundlich isotherm’s heterogeneous multilayer interactions (0 < 1/n < 1) are both explainable. Pseudo-first-order kinetics also fit well as a supportive mechanism, with the adsorption data showing the dominance of weak physical forces via physisorption, as discussed in [137].

FTIR also indicated C=C and C≡C linkages, suggesting π–π stacking interactions with methyl orange’s aromatic rings [138]. π–π electron donor–acceptor (EDA) mechanisms may also be involved, as reported in [139]. The BET surface area (~323 m²/g) and strong intra-particle diffusion model fit at low dye concentrations point to pore filling and molecular diffusion. As shown in Figure 8J–L, dye adsorption visibly altered B600’s porosity (disappearance) and surface morphology (growth of clusters).

4. Conclusions

Sustainable management of CRD wood waste remains challenging due to its non-recyclable status and degradation from demolition and surface treatments, but this study shows that pyrolysis offers a scalable and sustainable valorization pathway with industrial and environmental benefits. Using a statistically validated DOE approach, the effects of pyrolysis temperature, BRT, and feedstock mass were assessed across lab-scale (horizontal tube furnace) and semi-pilot (rotary retort-furnace) reactors. High-temperature biochars like B800 showed peak properties—FC (87%), TSF (96%), carbon content (92%), BET surface area (300 m²/g), and micropore volume (0.146 cm³/g)—alongside declines in VC (9%), hydrogen (0.9%), oxygen (2.2%), and Van-Krevelen ratios (H/C: 0.1; O/C: 0.02), with a low yield of 21%, supporting their use in metallurgical reduction or as coke blends. The moderate-temperature biochar B600 balanced reactivity and stability—FC (81%), TSF (87%), H/C: 0.3, O/C: 0.1, surface area (324 m²/g), and micropore volume (0.145 cm³/g)—enabling >90% methyl orange removal under low pH, a high dose, and a long contact time. Its adsorption followed the Freundlich isotherm (R² = 0.97, 1/n = 0.5) and pseudo-second-order kinetics (R² = 1 for linear fitting and regression; 0.87–0.89 for non-linear fitting and regression), confirming dominant multilayer chemisorption. Notably, no chemical or physical activation was applied, avoiding environmental burdens. Cross-scale comparisons confirmed consistent B400–B800 properties, with the pyrolysis temperature being most influential, raising FC by 15% and TSF by 19% and reducing VC by 53%. The HHV ranged from 25 to 33 MJ/kg, supporting energy applications. Despite its demonstrated feasibility, challenges remain: securing CRD wood supply, managing inherent contaminants, utilizing co-products (bio-oil, syngas), and designing cost-effective, scalable systems. Future work will address pilot-scale trials with an even higher mass of CRD wood feed, post-pyrolysis thermal activation of biochar, and self-ignition risks to support safe deployment.