Torrefaction pretreatment of sugarcane bagasse coupled with pelletization for carbonaceous fuel production

1 Introduction

With the continuous growth of global energy demand and the progressive depletion of fossil resources, biomass energy—which serves as the only renewable organic carbon source—has attracted widespread attention for fuel and chemical production (Yang et al., 2022; Yu et al., 2025; Yang et al., 2024; Yu et al., 2026; Yin et al., 2026; Yang et al., 2020; Sugiarto et al., 2025). In China, the annual biomass solid waste resources amount to over 460 million tons of standard coal equivalent. Among these, the annual output of sugarcane bagasse (SCB)—the primary by-product of the sugar industry—exceeds 20 million tons. However, its average annual value-added utilization rate remains below 5% with the majority of SCB is discarded (Wang et al., 2024), incinerated, or used solely as low-grade boiler fuel, resulting in limited high-value utilization (Alonso et al., 2017). Potential bottlenecks that hinder the high-value application of SCB biomass include the inherently low energy density (Yuksel and Karakehya, 2024), low mechanical strength and low pyrolytic char yield. SCB feedstocks commonly exhibit high moisture content, high oxygen mass fraction, and loose structural morphology (Wen et al., 2025), leading to low calorific values, which is typically only one-third to one-half that of coal (Xiong et al., 2025). Moreover, these characteristics significantly increase transportation costs and storage space requirements, thereby diminishing the overall economic viability and impeding the large-scale deployment (DeMartini et al., 2013). Consequently, pretreatment of biomass feedstocks is essential for performance improvement of SCB and its derived char. Current biomass pretreatment methods are broadly classified into three categories: physical (Ningthoujam et al., 2026), chemical (Myeni et al., 2026), and biological approaches (Du et al., 2025). Among them, torrefaction (Du et al., 2013), which is a mild thermal treatment conducted under inert or low-oxygen conditions at the low temperature range of 200 °C–300 °C, offers distinct advantages over conventional drying: it not only removes moisture more effectively but also mitigates moisture reabsorption, it improves grindability and physicochemical structure, and partially deoxygenates the biomass (Xu et al., 2024; Ma et al., 2019; Peng et al., 2025). As a result, torrefaction could enhances the energy density and mechanical properties while maintaining a high yield of char and reducing downstream transportation, storage, and energy conversion costs—making it a particularly promising pretreatment strategy for the improving the quality of abundantly available SCB feedstock (Xu et al., 2026; Yang et al., 2023).

Chen and Kuo (2011) identified three torrefaction stages at 230 °C–290 °C for biomass components: mild (230 °C, minor volatiles release), moderate (260 °C, substantial hemicellulose decomposition), and severe (290 °C, extensive hemicellulose/cellulose degradation with high mass loss and negligible interactions). van der Stelt et al. (2011) further reported that torrefaction reduces biomass volume by 30%–70%, achieves energy yields exceeding 90%, and increases energy density by approximately 30%. The aforementioned studies have clearly demonstrated the significant influence of torrefaction temperature gradients on the pyrolytic behavior and energy density of biomass components. Despite the well-recognized improvement in energy density conferred by torrefaction pretreatment, current knowledge remains largely derived from studies conducted at a limited set of discrete temperatures, with a notable absence of systematic investigation spanning a continuous temperature gradient. This knowledge gap warrants dedicated experimental efforts to elucidate the temperature-dependent evolution of biomass properties, which is essential for rational process design and optimization.

Nevertheless, the resulting torrefied char is typically granular and exhibits rapid, uncontrolled combustion when used directly—leading to poor combustion stability and low burnout efficiency (Li et al., 2026). Such characteristics fail to meet industrial requirements for solid fuels regarding particle size distribution and mechanical strength (Ivanovski et al., 2022). Therefore, cold compaction (i.e., mechanical densification at ambient temperature) is required to convert torrefied char into robust (Li and Zou, 2022), dimensionally stable pellets. It is noteworthy that existing research has empirically demonstrated that torrefaction significantly enhances the mechanical strength of cold-pressed products. Murakami et al. (2025) compared direct pelletization with torrefaction followed by pelletization. The results showed that the products obtained by the latter method exhibited significantly greater improvement in both fuel ratio and hot compressive strength, far superior to those produced by direct pelletization. The aforementioned studies have clearly demonstrated the significant influence of torrefaction temperature gradients on the pyrolytic behavior and energy density of biomass components. Nevertheless, the translation of these beneficial effects into industrial carbonaceous fuel hinges critically on the successful densification of torrefied materials. To date, most densification studies have focused on raw agricultural residues or wood chips, with little systematic inquiry into how torrefaction-induced changes in pore structure, and hydrophobicity affect the compaction behavior and binder effectiveness of torrefied products. For SCB, a feedstock with high hemicellulose content and distinctive fibrous morphology—the densification strategy across different torrefaction severities remains ambiguous, and the underlying bonding mechanisms have not been clarified.

For subsequent cold-press pelletization of the torrefied product, the necessity of a binder is determined by the torrefaction temperature. Under mild torrefaction conditions, the torrefied char often retains sufficient native binders (e.g., lignin, pectin, cellulose derivatives) to enable self-binding without added chemical binders—rendering the process simple and energy-efficient. In contrast, moderate-to-severe torrefaction substantially degrades these thermolabile binding components, necessitating the incorporation of suitable binders to ensure adequate pellet integrity (Hu et al., 2015). Common binders include organic additives (e.g., starch) and inorganic additives (e.g., hydrated lime, Ca(OH)₂) (Chin and Siddiqui, 2000). The binder type critically influences pellet impurity profile (e.g., ash composition and content) and governs key functional properties—including mechanical strength, electrical resistivity, and chemical reactivity (Liu et al., 2025). Kong et al. (2013) demonstrated that Ca(OH)₂ markedly improved the physical quality of alfalfa-based pellets and reduced compaction energy consumption; Borowski et al. (2017) reported that starch’s high calorific value enhanced both combustion performance and mechanical durability of densified fuels. In our prior work, we selected carboxymethyl cellulose (CMC) as the binder owing to its renewable origin, high water solubility, excellent film-forming capability, compatibility with green processing protocols, and absence of inorganic contaminants. Furthermore, we tested alkaline lignin and CMC as binders for cold-compacting torrefied biomass. Results revealed that alkaline lignin—owing to its excessively high ash content—impeded inter-particle adhesion and thus deteriorated pellet formation, consistent with established principles governing binder selection. Furthermore, biomass pelletization involves synergistic contributions from multiple micro-mechanisms, including mechanical interlocking (He et al., 2024), particle bridging (Kaliyan and Vance Morey, 2008), and intermolecular forces (Kaliyan and Morey, 2010), all of which collectively determine final pellet strength.

Building upon the above findings, the integrated torrefaction–cold-compaction route for producing solid biofuels—as schematically illustrated in Figure 1—represents a highly promising pathway toward sustainable energy valorization. Accordingly, this study employs a comprehensive suite of characterization techniques, including torrefied char yield analysis, thermogravimetric (TG) analysis, elemental analysis, scanning electron microscopy (SEM), proximate analysis, gasification reactivity assessment, and mechanical strength evaluation of pellets—to systematically investigate: (i) how torrefaction temperature governs key energetic parameters (e.g., energy density, higher heating value, fixed carbon content); and (ii) how binder selection modulates pelletization mechanisms and mechanical integrity. These multi-scale analytical approaches enable deep mechanistic insights into microstructural evolution and property development, not only elucidating the fundamental origins of performance changes during thermal conversion but also providing critical theoretical guidance for scaling up torrefaction technology. Ultimately, such torrefied and densified biomass fuels hold strong potential to partially substitute conventional solid fossil fuels in industrial applications—thereby enhancing resource utilization efficiency and lowering production costs.

2 Experimental

2.1 Torrefaction pretreatment

This study employed SCB as the primary feedstock. Prior to experimentation, the raw material was dried in a vacuum oven at a constant temperature of 105 °C for 24 h. Subsequently, the dried SCB was ground using a vibratory mill to achieve a uniform particle size distribution within the 60–80 mesh range. In Experiments 1∼6, torrefaction temperatures were systematically varied from 200 °C to 325 °C in 25 °C increments, yielding six distinct temperature setpoints: 200 °C, 225 °C, 250 °C, 275 °C, 300 °C, and 325 °C. We intentionally selected 200 °C as the lower boundary to capture the incipient stages of torrefaction, while extending the upper limit to 325 °C to systematically probe the transition from mild torrefaction to intensive thermal treatment—a regime in which cellulose decomposition becomes markedly pronounced. This expanded temperature gradient (200 °C–325 °C) permits a comprehensive evaluation of how progressive increases in torrefaction severity govern the physicochemical properties and fuel characteristics of biomass, and further enables a systematic assessment of escalating treatment intensity, encompassing the substantial degradation of cellulose. For each experiment, approximately 5 g of SCB was loaded to ensure experimental reproducibility and data comparability. A fixed-bed pyrolysis reactor (SY-24, Jinan Haibo Laboratory Instrument Co., Ltd.) was used. Before initiating each experiment, high-purity argon gas was purged continuously through the quartz tube inside the reactor for a sufficient duration to completely displace ambient air and establish a stable, oxygen-free inert atmosphere. Once the reactor’s internal temperature stabilized at the target torrefaction temperature, the quartz basket containing the pre-weighed SCB sample was rapidly inserted into the isothermal zone of the reactor. Xu et al. (2026) investigated the effects of torrefaction duration—set at 20, 40, and 60 min—and found that higher torrefaction temperatures combined with longer durations synergistically improved the fuel properties of biomass. Therefore, the samples were held at the designated temperature under an inert atmosphere for a representative dwell time of 1 h. Upon completion of the torrefaction reaction, the system was allowed to cool naturally to room temperature. The resulting solid residue was carefully retrieved and weighed on an analytical balance. The solid yield was subsequently calculated according to Equation 1:

y i e l d w t . % = m a s s o f p r o d u c t g m a s s o f s a m p l e g × 100 (1)

The holding time of 60 min and the temperature range of 200 °C–325 °C were selected not only for experimental suitability but also because they fall within the typical residence time (30–60 min) and operating temperature envelope of industrial torrefaction systems, thereby ensuring that our findings are directly relevant to industrial scale-up and process optimization.

2.2 TG analysis and gasification reactivity analysis

Thermogravimetric experiments and Carbon dioxide gasification were performed simultaneously using a thermogravimetric analyzer (Q50, TA, United States). (i) TG analysis: for each run, approximately 20 mg of sample was loaded into an alumina crucible. The sample was first heated from room temperature to 900 °C at a heating rate of 20 °C/min under a nitrogen atmosphere (flow rate: 100 mL/min), and then held isothermally at 900 °C for 5 min. (ii) Gasification reactivity analysis: subsequently, the nitrogen flow was switched to pure CO₂ (flow rate: 100 mL/min), and the temperature was maintained at 900 °C for an additional 30 min. The carbon conversion (x), gasification rate (r), and gasification activity index (R _*0.5*) were calculated according to the following equations (Hu et al., 2022):

x = m 0 − m j m 0 − m ∞ (2)

r = d x d t (3)

R 0.5 = 0.5 τ 0.5 (4)

Among them, m _*o*, m _*j*, and m ∞ are the mass in the initial state, time j, and final state respectively, and τ _*0.5* is the time when the gasification conversion rate (x) reaches 0.5.

2.3 Elemental analysis

The elemental compositions (C, H, O, N, and S) of the raw material subjected to different torrefaction temperatures were determined using an organic elemental analyzer (Unicube, Elementar, Germany). The oxygen removal ratio (O/C) and hydrogen removal ratio (H/C) of the roasted biomass were subsequently calculated, as defined in Equations 5, 6:

O / C = O % C % (5)

H / C = H % C % (6)

2.4 SEM analysis

Scanning electron microscopy (ZEISS Sigma 360, Carl Zeiss AG, Germany) was employed to characterize the surface morphology and pore structure of the raw material at different torrefaction temperatures.

2.5 XRD analysis

The crystal structure of the carbon material was characterized by X-ray diffraction (XRD, X’Pert PRO, PANalytical B.V., Netherlands). Prior to analysis, the sample was ground and sieved to a particle size <0.2 mm. It was then packed into an aluminum sample holder and scanned over a 2θ range of 5°–90° using Cu K radiation at an operating voltage of 40 kV and current of 40 mA. Data were collected with a step size of 0.0191°. Smoothed diffraction patterns were subjected to peak deconvolution in the 2θ range of 15°–35°, yielding two Gaussian peaks: the γ peak and the 002 peak (Yen et al., 1961). As reported in the literature, the integrated intensities (areas) of the γ peak (A _*γ*) and the 002 peak (A _*002* ) are proportional to the relative abundances of aliphatic and aromatic carbon atoms, respectively (Xie et al., 2019).

2.6 Proximate analysis

The proximate analysis of SCB char obtained at different torrefaction temperatures was conducted in accordance with GB/T 28731-2012 《Industrial Analysis Methods for Solid Biomass Fuels》.

The moisture content of biochar (M _*ad*) was determined according to Equation 7:

M a d = W e i g h t o f l o s s g W e i g h t o f i n i t i a l s a m p l e g × 100 (7)

The volatile matter content of biochar (V _*ad*) was determined according to Equation 8:

V a d = W e i g h t o f l o s s g W e i g h t o f i n i t i a l s a m p l e g × 100 − M a d (8)

The ash content of biochar (A _*sd*) was determined according to Equation 9:

A s d = W e i g h t o f r e m a i n i n g s o l i d s g W e i g h t o f i n i t i a l s a m p l e g × 100 (9)

The fixed carbon content of biochar (FC _*ad*) was calculated using Equation 10:

F C a d = 100 − M a d + A a d + V a d (10)

2.7 HHV analysis

The higher heating value (HHV) of a solid biomass fuel sample can be calculated based on the volatile matter, ash, and fixed carbon contents of the biomass feedstock pyrolyzed at different temperatures, as shown in Equation 11:

H H V = 0.3536 × F C a d + 0.1559 × V a d − 0.0078 × A s d (11)

2.8 Energy density and energy yield

Based on the experimentally determined higher heating value (HHV) and mass loss data, the energy density (W _*M*) and energy yield (HHV _*yield*) of the char material can be calculated using Equations 12, 13:

W M = H H V f H H V i × 100 (12)

H H V y i e l d % = 100 − w e i g h t l o s s × H H V f H H V i (13)

Among them, HHV_f represents the High Heat Value of lignocellulose biomass after pyrolysis, and HHV_i represents the High Heat Value of biomass before pyrolysis.

2.9 Pelletization

Upon completion of the torrefaction process, the resulting roasted product undergoes pelletization. The specific procedure is as follows: First, a total of 2 g of the mixed feedstock is accurately weighed, comprising 75 wt.% roasted material, 20 wt.% deionized water, and 5 wt% sodium carboxymethyl cellulose (CMC, analytical grade, DS: 6.5%–8.5%, PH: 6.5–8.5) as a binder. The three components are placed into a clean ableting die cavity and thoroughly homogenized via manual or mechanical stirring to ensure uniform distribution and effective binding. Subsequently, the homogenized mixture is transferred into the die cavity of a tablet press (YP-12, Shanghai Lijin Scientific Instrument Co., Ltd.) and uniaxially compressed at a pressure of 6 Mpa to yield mechanically robust, disk-shaped pellets with defined geometry. After compression, the formed pellets are carefully demolded, concluding the pelletization procedure.

2.10 Mechanical property analysis

The pelletization behavior and mechanical strength of raw materials subjected to different torrefaction temperatures—with and without the addition of CMC as a binder—were evaluated using an electronic universal testing machine (UTM5300-H, Sansi Vertical Technology Co., Ltd).

2.11 Pellet formation analysis

The microstructural morphology of the crushed pelletized raw material was analyzed using a scanning electron microscope (ZEISS Sigma 360, Carl Zeiss AG, Germany), and a theoretical pelletization diagram was constructed to investigate the pelletization mechanism.

All experiments involving yield, elemental analysis, proximate analysis, higher heating value (HHV), and mechanical strength were conducted in triplicate (n = 3) using independently prepared samples. Results are reported as the mean ± standard deviation (SD) of the three parallel measurements, or error bars are included in the corresponding figures.

3 Results and discussion

3.1 Analysis of char recovery rate

As shown in Figure 2, the variation in solid product mass yield of SCB after 1-h isothermal torrefaction at different temperatures is presented. It can be observed that the mass yield decreases monotonically with increasing torrefaction temperature. This trend arises because elevated temperatures promote progressive thermal decomposition and volatilization of organic components—particularly hemicellulose, cellulose, and partially lignin—thereby reducing the mass of the residual solid (Cen et al., 2026). Notably, a pronounced inflection point in mass yield occurs when the torrefaction temperature increases from 275 °C to 300 °C, where the rate of decline is significantly greater than in other temperature intervals. This behavior suggests a marked intensification of devolatilization reactions near 275 °C, primarily driven by the accelerated thermal degradation of hemicellulose—the least thermally stable component—which releases substantial volatile matter and leads to a sharp reduction in solid yield.

3.2 TG analysis

As shown in Figures 3A,B, thermogravimetric (TG) and differential thermogravimetric (DTG) analyses reveal pronounced and systematic changes in the subsequent torrefaction behavior of SCB with increasing torrefaction temperature. Under a final pyrolysis temperature of 800 °C, the residual char yield consistently increases with rising torrefaction temperature, indicating that high-temperature pretreatment induces extensive removal of organic components. Concurrently, the characteristic peak temperature corresponding to the maximum mass loss rate in the DTG curves shifts markedly toward lower temperatures—advancing substantially to approximately 100 °C—suggesting an earlier onset of the most intense torrefaction stage. Moreover, the shape of the mass loss curve undergoes a distinct transformation: at lower torrefaction temperatures, torrefaction exhibits a narrow and sharp single peak, reflecting rapid decomposition of hemicellulose and cellulose within a relatively narrow temperature range; following high-temperature torrefaction, however, mass loss evolves into a broad and gradual profile spanning a wide temperature interval, implying that pyrolytic reactions are no longer confined to a narrow thermal window but instead occur progressively across the entire heating process. Collectively, these observations indicate that elevated torrefaction temperatures significantly promote the premature cleavage of volatile components—particularly hemicellulose and cellulose—in SCB, accelerate condensation and densification of the carbonaceous skeleton, reduce overall thermal stability, and lower the activation energy of the primary pyrolysis reaction—thereby facilitating both the initiation and sustained progression of pyrolysis at comparatively lower temperatures (Sezgin et al., 2026).

3.3 Elemental analysis

Table 1 presents the elemental composition analysis of raw and torrefied SCB at different temperatures. Firstly, the carbon content of the resulting char increases significantly compared to that of the raw, while the oxygen content correspondingly decreases. As the torrefaction temperature increases from 200 °C to 325 °C, the carbon € mass fraction rises significantly—from 48.89% to 66.64%—exhibiting a monotonic upward trend. The most pronounced increase in carbon content occurs within the temperature range of 275 °C–300 °C. This behavior is attributed to the preferential decomposition of hemicellulose and partial cellulose during torrefaction, which involves decarboxylation, dehydration, and decarbonylation reactions, resulting in the release of oxygen-rich volatile species (e.g., CO₂, CO, and H₂O) (Park et al., 2023), while the carbon-rich backbone remains largely intact and becomes progressively enriched.

TABLE 1. Organic elemental contents of raw and torrefied SCB at different temperatures.

To further elucidate the influence of torrefaction temperature on the elemental evolution of biomass, the hydrogen-to-carbon (H/C) and oxygen-to-carbon (O/C) atomic ratios were calculated across the tested temperature range, yielding the paradigm diagram shown in Figure 4. Linear regression analysis of the H/C and O/C ratios versus torrefaction temperature reveals strong linear correlations: both ratios decrease linearly with increasing temperature. Specifically, the H/C ratio declines from 0.11 to 0.063, representing a reduction of approximately 43%, while the O/C ratio drops from 0.80 to 0.32—a decrease of about 60%. Notably, the steepest reductions in both H/C and O/C occur between 275 °C and 300 °C, indicating intensified deoxygenation (primarily as H₂O and CO₂) and concurrent aromatization. The O/C ratio decreases from 0.59 to 0.34 over this interval, signifying substantial improvements in energy density and hydrophobicity. These findings show that torrefaction SCB, especially after high-temperature treatment, has lower H/C and O/C ratios than raw, indicating enhanced dehydrogenation and deoxygenation, and is therefore a promising biofuel feedstock.

3.4 SEM analysis

Figure 5 presents the scanning electron microscopy (SEM) analysis of SCB subjected to different torrefaction temperatures, revealing a pronounced and stage-dependent structural evolution in its micro-morphology with increasing thermal treatment temperature. Within the relatively low temperature range of 200 °C–275 °C, the fibrous architecture of SCB retains a largely intact, dense, lamellar skeletal framework; inter-fiber boundaries remain clearly discernible, and the overall structural continuity and stability are well preserved. At this stage, only minor thermal degradation of hemicellulose occurs, resulting in the formation of a limited number of small, localized pores—insufficient to induce appreciable disruption of the macroscopic fiber network. However, upon further elevation of the torrefaction temperature to 300 °C, thermal effects intensify markedly, triggering substantial thermal decomposition of the primary biopolymeric constituents—namely cellulose and lignin. Consequently, the original fibrous structure exhibits evident thermal damage, manifested as localized fiber fracture, distortion, and partial collapse, leading to severe deterioration of the initial dense skeletal integrity. Upon continued heating to 325 °C, SCB undergoes complete deep carbonization: its micro-morphology is transformed entirely into a highly porous, loosely packed, honeycomb-like carbonaceous structure, characterized by densely distributed and mutually interconnected pore channels. The original fibrous morphology and skeletal architecture become fully destabilized and irreversibly converted into carbonized products—exhibiting hallmark features of high-temperature torrefaction of lignocellulosic biomass.

3.5 XRD analysis

Figure 6 presents XRD patterns of SCB samples subjected to different torrefaction temperatures. As clearly observed in the figure, a prominent diffraction peak appears at the 002 plane, accompanied by a secondary but distinct peak at the 100 plane; no other significant diffraction peaks are detected (Yoshizawa et al., 2000). This characteristic XRD profile closely matches that of native cellulose I, indicating that the thermal treatment applied in this study did not induce any crystalline phase transition of cellulose in the SCB. Further analysis reveals temperature-dependent variations in the diffraction peaks. Within the torrefaction temperature range of 200 °C–250 °C, the intensity of the 002 peak remains essentially unchanged. This stability is attributed to the onset of hemicellulose thermal degradation and partial removal within this interval, which increases the relative cellulose content and thereby enhances the prominence of its characteristic diffraction peaks. In contrast, when the torrefaction temperature is elevated to 250 °C–325 °C, the 002 peak intensity progressively declines and its full width at half maximum broadens, signifying substantial thermal disruption of cellulose crystallites and a consequent reduction in molecular ordering and overall crystallinity. Finally, at 325 °C, the 002 peak intensity reaches its minimum detectable level, and the peak becomes highly diffuse—indicating near-complete collapse of the ordered crystalline structure and severe deterioration of cellulose crystallinity.

3.6 Proximate analysis

Figures 7A–D presents the compositional variations of SCB under different torrefaction temperatures. As shown in the figure, the volatile matter content gradually decreases with increasing torrefaction temperature, whereas both ash and fixed carbon contents increase progressively. The compositional changes are relatively mild below 275 °C but become markedly accelerated above this temperature. In particular, the decline in volatile matter becomes especially pronounced at higher torrefaction temperatures, indicating intensified hemicellulose decomposition and substantial release of volatiles at 275 °C.

Figure 7E illustrates the higher heating value (HHV) of torrefied SCB, calculated using the standard empirical formula for solid biomass fuels. The HHV increases monotonically with rising torrefaction temperature (Channiwala and Parikh, 2002; Maksimuk et al., 2021). A notable enhancement in HHV is observed within the temperature range of 275 °C–300 °C, suggesting a significant thermochemical transformation occurring in this interval. This is consistent with the above analysis.

To provide a more intuitive assessment of the energy-related properties of torrefied SCB, Figure 7F plots the relative energy density and energy yield as functions of torrefaction temperature. A sharp transition occurs between 275 °C and 300 °C: the relative energy density increases dramatically from 112.03% to 125.86% (Elhenawy et al., 2024), representing the most pronounced change across all tested temperatures. In contrast, the energy yield decreases continuously with increasing torrefaction temperature; its reduction is most substantial within the same 275 °C–300 °C range, dropping from 66.57% to 51.40%. A clear trade-off was observed between energy yield and energy density. As the torrefaction temperature increased from 275 °C to 300 °C, the energy yield decreased from 66.57% to 51.40%, whereas the higher heating value (HHV) rose from 22.40 to 22.92 MJ/kg. This indicates that elevated temperatures enhance fuel quality, primarily through improved energy density, but at the expense of a substantial reduction in recoverable chemical energy. For applications where transportation logistics and combustion efficiency are paramount, the higher energy density achieved at 300 °C may be advantageous. In contrast, when maximizing the overall energy recovery rate is the primary objective, lower torrefaction temperatures are generally more conducive to enhancing the system-wide economic efficiency. Consequently, the optimal torrefaction process parameters must be determined holistically, tailored to the specific application context and balancing both end-use requirements and the economic performance across the entire supply chain.

3.7 Gasification reactivity of chars

Figure 8 shows the carbon dioxide gasification process of SCB chars prepared at different final torrefaction temperatures characterized by TGA. Table 2 lists the corresponding gasification characteristic parameters. When the purge gas switches from N₂ to CO₂, the gasification reaction of char starts, and this moment is marked as time 0 (Hu et al., 2022). The gasification end time (t _*end*) is defined as the time when the gasification conversion (x) reaches 0.98. t _*max* corresponds to the time when the gasification rate reaches its maximum value (r _*max*). After gasification is completed, the remaining mass percentage is recorded as residue (wt%) in Table 2.

TABLE 2. Gasification characteristic parameters of biomass torrefaction char.

As shown in Figures 8A,B, r _*max* shows a downward trend with the increase of torrefaction temperature, decreasing from 0.866 min^-1 at 200 °C to 0.147 min^-1 at 325 °C. R _*0.5* also decreased from 0.667 min^-1 to 0.074 min^-1. Increasing torrefaction temperature prolongs the complete gasification time and reduces the peak gasification rate. It promotes the development of pore structure while reducing active centers, indicating that char gasification reactivity deteriorates (Tong et al., 2018). Alongside elemental data in Table 1, torrefaction from 200 °C to 325 °C triggers drastic deoxygenation, with oxygen content dropping from 38.87 wt.% to 21.40 wt.% and carbon content accumulating from 48.89 wt.% to 66.64 wt.%, removing abundant oxygenated functional groups that act as core gasification active sites while forming inert aromatic carbon skeletons. Proximate analysis (Figure 7) confirms fixed carbon rises sharply from 17.45 wt.% to 50.28 wt.% and volatile matter declines from 76.13 wt.% to 42.56 wt.%; volatile release during gasification generates fresh reactive pores, an effect largely suppressed at high torrefaction temperatures. SEM (Figure 5) observations reveal mild torrefaction (200 °C–250 °C) produces interconnected micropores facilitating CO₂ diffusion, whereas intense devolatilization above 275 °C causes micropore collapse and carbon agglomeration, increasing mass transfer resistance. XRD patterns (Figure 6) demonstrate progressive carbon graphitization, where the intensified (002) diffraction peak at ≈22° reflects ordered aromatic stacking and higher gasification activation energy. Although higher torrefaction slightly elevates ash content to concentrate AAEM catalysts (Rocha et al., 2017), their promotional effect cannot offset reactivity losses from active site elimination, carbon aromatization and pore degradation. The depletion of oxygen-containing active sites and enhanced carbon graphitization are the dominant factors suppressing gasification performance, with deteriorated pore structure serving as a secondary limiting factor.

3.8 Mechanical property analysis

Figure 9 presents the pelletized samples prepared from SCB—without any added binder—subjected to torrefaction treatments at various temperatures, along with their corresponding mechanical strength test results. As clearly observed in the figure, SCB torrefied within the temperature range of 200 °C–275 °C retains a certain degree of formability: although the pelletization quality is modest, the resulting specimens can still maintain a rudimentary disc-like morphology. This behavior is primarily attributed to the thermal stability of the major biopolymeric constituents—cellulose, hemicellulose, and lignin—under these moderate torrefaction conditions. Specifically, hemicellulose undergoes only partial degradation, while cellulose and lignin remain largely intact; consequently, the material preserves relatively high inter-particle cohesion and inherent binding capacity, enabling it to retain structural integrity under compaction pressure and sustain comparatively high and stable mechanical strength. In contrast, when the torrefaction temperature is elevated to 300 °C–325 °C, SCB exhibits near-complete loss of formability: the compacted specimens readily disintegrate and often fail to yield coherent, intact discs. This dramatic deterioration stems predominantly from near-total thermal decomposition of hemicellulose and the onset of cellulose degradation at these elevated temperatures. Such extensive depolymerization severely compromises the intrinsic binding capability and structural continuity of the biomass matrix, leading to a substantial reduction in inter-particle cohesion and a consequent sharp decline in mechanical performance—rendering the material incapable of meeting the mechanical requirements for successful pelletization.

Figure 10 presents a detailed characterization of the pellet morphology and corresponding mechanical strength test results for SCB briquettes fabricated with CMC as a binder, under varying torrefaction temperatures. As clearly observed in the figure, SCB feedstock torrefied within the temperature range of 200 °C–325 °C exhibits favorable formability, enabling successful compaction into structurally intact, disc-shaped specimens. More importantly, compared with the binder-free control group, the majority of CMC-blended samples demonstrate significantly enhanced mechanical strength, underscoring the beneficial role of sodium carboxymethyl cellulose in improving the structural integrity and stability of biomass briquettes.

Notably, however, the mechanical strength of the CMC-containing samples follows a non-linear, biphasic trend with increasing torrefaction temperature—initially decreasing and subsequently increasing. In particular, at the specific torrefaction temperature of 275 °C, the CMC-blended samples exhibit lower mechanical strength than their binder-free counterparts—a counterintuitive observation. A preliminary hypothesis attributes this anomaly to partial thermal decomposition of CMC at 275 °C and/or complex interfacial interactions between CMC and native components of SCB, which may disrupt the inherent pore architecture and interparticle bonding mechanisms. Such disruption likely compromises overall densification and mechanical coherence, thereby diminishing the mechanical performance of the resulting pellets. We emphasize that this interpretation is merely a preliminary hypothesis and requires further experimental validation, which we plan to pursue in future work.

In summary, the maximum compressive strength measured in this study was 8006.16 N, which was compared with reported values for commercial biomass pellets from the literature. Under diametral compression testing conditions, the average compressive strength of commercial wood pellets typically falls within the range of 6,000–8,000 N (Scott et al., 2025), and significantly exceeds the minimum requirement of ≥ 1,000 N specified in the industrial reference standard ISO 17831-1-2025 《Solid biofuels—Determination of mechanical durability of pellets and briquettes》. The mechanical strength of the biomass pellets prepared in this study is comparable to—and slightly exceeds—that of commercially available biomass pellets, and significantly surpasses the minimum industrial requirement. This demonstrates that the pellets developed herein possess excellent mechanical strength, thereby enhancing their suitability for practical applications, including transportation, storage, and handling.

3.9 Pellet formation analysis

To further investigate the pelletization mechanism, as shown in Figures 11, 12. The mechanical strength of CMC-bonded torrefied SCB pellets declines from 200 °C to 275 °C and hits a minimum at 275 °C. Elevated torrefaction reduces surface functional groups, weakening intermolecular interactions and particle bridging between the binder and biochar (Kaliyan and Morey, 2010); forming pressure generates new intergranular pores, and subsequent desiccation triggers cohesive failure of the CMC binder, jointly lowering pellet strength. A partial strength rebound occurs at 300 °C–325 °C as denser biochar particles form tighter mechanical interlocking (Tumuluru and Fillerup, 2020), though irreversible biomass degradation prevents full recovery of initial mechanical performance. The above analysis is consistent with the experimentally observed trend in mechanical performance of the biomass feedstock across different torrefaction temperatures, as presented in Figure 10.

4 Conclusion

From an industrial perspective, this study proposes a method for producing carbonaceous fuels with high mechanical strength via torrefaction pretreatment of biomass coupled with cold-press pelletization. The proposed torrefaction-pelletization route offers a viable pathway to upgrade low-density SCB into a high-quality carbonaceous fuel with improved energy density, hydrophobicity, and uniformity. These property enhancements translate directly into reduced transportation and storage costs, more stable combustion behaviour, and lower emissions of particulates and pollutants compared to raw biomass. At scale, the process can be readily integrated with existing biomass-handling facilities, and the combustible volatiles released during torrefaction can be captured and recycled as a supplementary heat source, thereby improving overall energy efficiency. The main conclusions are as follows:

As the torrefaction temperature increased from 200 °C to 325 °C, the char yield decreased from 76.42 wt.% to 36.42 wt.%, while the higher heating value (HHV) increased from 18.31 MJ/kg to 24.52 MJ/kg. Gasification reactivity exhibited a non-monotonic trend—first increasing and then decreasing with rising temperature. Further investigation revealed a synergistic effect between binder dosage and torrefaction temperature on pellet mechanical strength. Maximum compressive strength (8006.16 N) was achieved at a binder dosage of 5 wt.% and a torrefaction temperature of 200 °C. Mechanistic analysis indicated that the binder enhances mechanical integrity through combined contributions of mechanical interlocking, particle bridging, and intermolecular interactions—thereby effectively reinforcing the structural cohesion of SCB-derived char pellets.

Low temperature torrefaction (e.g., 200 °C) affords a high raw-material yield of 76.42 wt.% and enables binder-free pelletization, facilitating low-cost production of carbonaceous fuels. In contrast, high-temperature torrefaction significantly enriches fixed carbon content—exceeding 50 wt.%—rendering the resulting char a promising candidate for industrial reducing agents.

This study presents a process strategy for converting lignocellulosic biomass into high-mechanical-strength carbonaceous fuels via torrefaction pretreatment followed by binder-assisted pelletization. The approach contributes to reducing overall processing costs and improves the economic viability of biorefining agricultural and forestry residues.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

Edited by: Xu Zeng, Huaqiao University, China

Reviewed by: Sachin C. Hallad, Maharana Pratap University of Agriculture and Technology, India

Asmelash Mekonen, Mekelle University, Ethiopia