Oxidative Pyrolysis of Typical Volatile Model Compounds Under Low Oxygen Equivalence Ratios During Oxidative Pyrolysis of Biomass

Abstract

This study aims to investigate the oxidative pyrolysis of biomass volatiles with a particular focus on the formation of liquid products. Furfural, hydroxyacetone, and 3,4-dimethoxybenzaldehyde were chosen as volatile model compounds. The impacts of the oxygen equivalence ratio (ER, 0–15%) and temperature (400–500 °C) on the product composition and distribution were examined using a two-stage quartz-tube reactor. The results showed that volatile pyrolysis was limited at the lower temperature of 400 °C even with oxygen introduction, while it could be significantly promoted at 500 °C as illustrated by the observed great decrease in the GC-MS peak areas of the volatile compounds especially under an oxidative atmosphere. For instance, the peak area of 3,4-dimethoxybenzaldehyde at 500 °C under an ER of 4% was only ~9% of that at 400 °C. Oxygen introduction enhanced the volatile decomposition with the formation of mainly permanent gases (although not given in the study) rather than liquid products, but distinct impacts were obtained for varied volatile compounds possibly due to their different chemical structures and autoignition temperatures. From the perspective of liquid product formation, furfural would undergo the cleavage of C-C/C-O bonds to form linear intermediates and subsequent aromatization to generate aromatics (benzene and benzofuran). The presence of oxygen could enhance the oxidative destruction of the C-C/C-O bonds and the removal of O from the molecules to form simple aromatics such as benzene, phenol, and toluene. Hydroxyacetone mainly underwent C-C/C-O cleavage that was further enhanced in the presence of oxygen; the resultant intermediates would recombine to generate acetoin and 2,3-pentanedione. A higher ER would directly oxidize the alcoholic hydroxyl group (-OH) into an aldehyde group (-CHO) to form methyl glyoxal, while 3,4-dimethoxybenzaldehyde mainly underwent cleavage and recombination of bonds connected with the benzene ring including aldehyde group (-CHO), C_Ar-O, C_Methoxy-O bonds, thus forming 1,2-dimethoxybenzene, toluene, and 3-hydroxybenzadehyde. This study provides more fundamental insights into the homogeneous oxidation of volatiles during the oxidative fast pyrolysis of biomass, facilitating the deployment of this technology.

1. Introduction

Fast pyrolysis of biomass is an efficient thermochemical conversion technology, which enables the direct transformation of the green and net-zero-carbon biomass into more value-added products such as bio-oil and biochar [1,2,3]. Bio-oil, also called bio-crude, contains hundreds of chemicals, such as phenols, aromatics, anhydrosugars, furans, linear acids/carbonyls/aldehydes, etc. [4]. It can be used as the feedstock for a centralized biorefinery to produce all kinds of green chemicals and materials to partially substitute fossils [5,6], which is of great importance for the decarbonized future.

At present, the research on the fast pyrolysis of biomass mainly focuses on the chemical reaction aspect; how to selectively generate value-added fuels, chemicals, or materials is of the first priority [7], including optimization of the pyrolysis conditions, in situ catalytic pyrolysis, ex situ catalytic pyrolysis, and other upgrading techniques [8,9,10,11,12]. Some novel pyrolysis technologies as well as catalysts have been successfully developed and demonstrated in the laboratory [3,13,14]. However, the practical/scaled application of the fast pyrolysis of biomass still faces engineering issues, one of which is the rapid heat supply and transfer during the process. Conventional fast pyrolysis of biomass is normally undertaken in an oxygen-free atmosphere at medium temperatures (e.g., 400–600 °C); this inherently endothermic process requires an external heat source to sustain the thermal decomposition reactions [15,16]. Particularly at the commercial scale, supplying external heat typically through gas/solid heat carriers or conduction from reactor walls to the biomass particles would be increasingly difficult due to the complexity and instability of heat exchange systems and the inefficiency of heat transfer [17]. It becomes a challenging issue during the scaling up and commercialization of fast pyrolysis technology.

The oxidative fast pyrolysis of biomass has been recently developed to address the heat supply/transfer bottleneck issue associated with conventional pyrolysis processes [18]. A small amount of oxygen or air is introduced into the pyrolysis process, facilitating the in situ oxidation of certain pyrolysis intermediates to generate the heat demanded for the endothermic reactions; a self-sustained/autothermal pyrolysis process is thus achieved [19,20,21]. Polin et al. [22] reported that the oxygen equivalence ratios (ERs) for the autothermal pyrolysis of corn stover and red oak in a fluidized-bed reactor at 500 °C were only 0.06 and 0.10, respectively. It should be pointed out that the oxidative fast pyrolysis of biomass is very different from combustion and gasification; it is normally carried out at a medium temperature of 400–600 °C, which is much lower than that of gasification and combustion. The ER used in the oxidative fast pyrolysis process (0–0.15 or 15%) is also much lower than that in gasification (0.2–0.35) [23,24] and combustion (more than 1) [25,26], and the main products from the oxidative fast pyrolysis of biomass are still bio-oil, biochar, and non-condensable gas. Compared with conventional non-oxidative/inert fast pyrolysis of biomass, the oxidative fast pyrolysis of biomass shows a lot of advantages not only in terms of the heat supply and transfer toward scaling up but also in terms of the elimination of the complicated external heating, the ease of scaling up by volume (not surface area) due to the realization of bulk heating instead of surface heating, and the significant process intensification due to the in situ oxidation heating [17]. Moreover, no significant reduction in bio-oil yield is observed for fluidized-bed oxidative fast pyrolysis of biomass at 500 °C under low ERs (0.033–0.104), as found by Kim et al. [27]. The heat demand for pyrolysis is mainly satisfied by the oxidation of biochar with only a very low fraction of light bio-oil being oxidatively consumed [28]. It further confirms the advantage of oxidative fast pyrolysis of biomass in terms of bio-oil production.

As discussed previously, the introduction of oxygen in the oxidative fast pyrolysis of biomass would inevitably induce oxidation reactions of pyrolysis intermediates or products, including the heterogeneous oxidation of biochar and the homogeneous oxidation of volatiles [17,28]. The manner of oxidization of the biochar and volatiles/bio-oil during the process mainly depends on the contact between oxygen and biochar or volatile vapors, as well as on their oxidation reactivities. However, Brown [19] regarded that the autoignition temperatures of biochar and volatile/bio-oil components determined their oxidation characteristics, and the biochar with the lowest autoignition temperature was more oxidized in the fluidized-bed reactor to provide the largest contribution to the heat demanded for biomass pyrolysis. Li et al. [28] also made a similar observation, while the bio-oil composition was also found to be altered with the introduction of oxygen. Considering that the oxidation of the biochar and the volatiles occurs concurrently, although the tendency of the biochar and volatiles toward being oxidized is known, it would be still challenging to quantitatively distinguish their individual contributions during the process. The role of oxygen in oxidative fast pyrolysis and its influence on the formation and consumption of pyrolysis products remain poorly understood. Jiang et al. [29] proposed a method to decouple the oxidation reactions in oxidative fast pyrolysis by varying the oxygen injection location. A purely “homogeneous” oxidation and the combined “homogeneous + heterogeneous” oxidations for product formation in a fluidized-bed pyrolysis system were obtained and compared in detail. However, not only would the biochar and volatiles compete for oxygen for hetero-/homogeneous oxidation [30], but the different components in the volatiles would also compete for oxygen due to their different reactivities with oxygen; changes in the operating conditions can substantially influence the oxidation process as well. Despite the importance of this topic, research specifically addressing the homogeneous oxidation of volatile components in the oxidative fast pyrolysis process remains limited.

Therefore, in this study, the oxidative pyrolysis of three representative biomass volatiles from the pyrolysis of holocellulose and lignin was carried out under varying temperatures (400–500 °C) and ERs (0–15%) in a two-stage quartz-tube reactor. The characteristic products were analyzed, and the possible conversion pathways during the oxidative pyrolysis process were proposed. Through this study, we expect to gain a more fundamental understanding of the homogeneous oxidation of volatiles/bio-oil and of the selective formation of valuable products from the oxidative fast pyrolysis of biomass.

2. Experimental Section

2.1. Volatile Model Compounds

Furfural (AR, ≥99.5%, Xilong Scientific Co. Ltd., Shantou, China) and hydroxyacetone (AR, 97%, Bidepharm Co. Ltd., Shanghai, China), as the representative furans and linear carbonyls from holocellulose (cellulose and hemicellulose) pyrolysis volatiles, and 3,4-dimethoxybenzaldehyde (AR, ≥98%, Xiya Reagent Co. Ltd., Zaozhuang, China), as the representative aromatic from lignin pyrolysis volatiles, were chosen as the volatile model compounds from the fast pyrolysis of biomass [31]. They were all subjected to homogeneous oxidative pyrolysis experiments.

2.2. Oxidative Pyrolysis of Volatiles

The oxidative pyrolysis of biomass volatile model compounds was conducted in a two-stage quartz-tube reaction system, as shown in Figure 1. It consisted of a gas supply system, a two-stage quartz-tube reactor heated by an electric furnace, a condensation system, and a gas collection bag. The gas supply system mainly provided two gas streams into the reactor, including the carrier gas N₂ and the oxygen-containing oxidation gas. The carrier gas N₂ was injected into the reactor from the top with a flow rate of 250 mL/min. The oxidation gas was a mixture of N₂ and air with a total flow rate of 50 mL/min; it was injected into the reactor from the middle (between the upper and lower sections). The amount of oxygen introduced was controlled by the mass flow controller (MFC). It should be noted that the gas flows did not require preheating due to their low flow rate compared with the large power of the two electric furnaces (both 1.5 kW); insignificant temperature fluctuations were observed during the experiments as the gases were easily heated to the reaction temperature. The quartz-tube reactor featured an inner diameter of 38 mm, with the lengths of the upper and lower sections being 330 mm and 122 mm, respectively. These two sections were heated separately by two independent electric furnaces (both 1.5 kW) and maintained at different temperatures. The upper section was used to evaporate the volatile model compounds; the temperature was set at 200 °C for furfural and hydroxyacetone and 250 °C for 3,4-dimethoxybenzaldehyde, according to their volatilization temperatures. The lower section was used to oxidize the volatilized model compounds at a lower oxygen ER, and its temperature was set at 400 or 500 °C, which was identical to the oxidative pyrolysis temperature. To ensure the precise control of the temperatures in each section and minimize the mutual influence of both sections due to the large temperature gradient, Kaowool was used to isolate the furnace/reactor from the environment (reducing heat loss) and the upper section from the lower section of the reactor (minimizing mutual influence). The top, middle, and bottom of the furnace/reactor were all well insulated. A K-type thermocouple penetrating through the top rubber plug of the reactor was used to measure the temperatures of both the sections. It should be pointed out that the temperatures given above were the actual measured temperatures, which were normally 8 °C lower than the set temperatures of the electric furnaces. Pre-experiments proved that the temperatures of the two sections could be controlled independently and kept steady over the experimental timescale. The condensation system was made of an acetone (AR) washing bottle immersed in an ice bath to capture the condensable products.

For a typical experiment, 0.5 g of the abovementioned volatile model compound was accurately weighed and loaded into a quartz basket; it was put into the upper section of the reactor and initially hung out of the heating zone prior to the experiment. The reactor was heated to the desired temperatures and kept isothermally with the continuous flowing of the carrier gas N₂ and the oxidation gas. The quartz basket was then quickly inserted into the center of the upper heating zone; the volatile model compound was heated and evaporated; it was carried by the N₂ gas, mixed with the oxygen-containing gas, and went into the lower section for homogenous oxidation. The resultant gas vapors coming out of the reactor were condensed and captured inside the acetone washing bottle, and the non-condensable gas was collected in the gas bag. Each experiment lasted for 15 min according to the volatilization rates of the model compounds. Different ERs were tested during the oxidation process, with 0–6% for 400 °C and 0–15% for 500 °C. It should be noted that the ER was calculated based on the amount of the volatilized model compound (not the initial weight used) during the process. The volatilization rates of furfural, hydroxyacetone, and 3,4-dimethoxybenzaldehyde were determined by pre-experiments, with average results of 100.00 wt.%, 97.83 wt.%, and 22.33 wt.%, respectively. The ER was defined as the ratio of real oxygen input to the theoretical oxygen demand for compete combustion (full oxidation to form CO₂ and H₂O). The real oxygen concentration in the total gas ranged around 0–2 vol.% as it was largely diluted by N₂ gas at an ER of 0–15%. No significant temperature changes induced by the oxidation reactions were observed during the experiments due to the small experiment scale (0.5 g model compound, 15 min evaporation under a low ER). After the experiment finished, the gas bag was switched off and disconnected from the system. The air supply was stopped with N₂ continuously flowing; the quartz basket was lifted out of the heating zone for rapid quenching. The condensation system was also disconnected from the reactor; the liquid captured in the acetone solvent and the condensates in the pipeline were also collected for further analysis.

2.3. Product Characterization

The solid residue remaining in the quartz basket was weighed to determine the volatilization rate of the model compound during the process, as mentioned in the above section.

The gas composition in the gas bag was analyzed using a gas chromatograph equipped with TCD and FID detectors (GC-9790II, Zhejiang Fuli Analytical Instrument Co., Ltd., Taizhou, China). Unfortunately, no feasible results were obtained due to the low oxidation rate of volatile model compounds and the corresponding excessive dilution by N₂. As a result, the gas composition is not reported in this study.

The composition of the liquid products was analyzed using an Agilent gas chromatography–mass spectrometry (GC-MS) system (6890 series GC coupled with a 5975 MS). An HP-5MS capillary column (30 m × 0.25 mm × 0.25 μm) was used. The detailed operating conditions can be found elsewhere [32]. Compound identification was performed using the NIST11 spectral library, with peak areas representing the relative abundance of specific compounds. Considering that the bio-oil is diluted by acetone in different extents due to the varied mass of acetone used for capture, the different acetone volatilization rate, and the different mass of acetone used for washing and collecting after experiment, the peak area of an individual compound detected by GC-MS cannot represent the yield of that compound in the bio-oil but only represents the relative content of the compound in the bio-oil solution. To compare the relative yield of a specific compound generated, the peak area should be generalized by multiplying the peak area and the mass of the liquid product mixed solution (including both bio-oil and acetone). It can thus mitigate the impact of differences in acetone volatilization and acetone solvent dilution and be used for semi-quantitative analysis.

3. Results and Discussion

3.1. Oxidative Pyrolysis of Furfural Under Low ERs at 400 and 500 °C

Figure 2 shows the composition of products detected by GC-MS from the oxidative pyrolysis of furfural under oxygen ERs of 0–15% at 400 and 500 °C. From Figure 2a, it can be seen that, at 400 °C, only furfural (the reactant) was found in the product, with its peak area slightly decreasing with increasing ER. It indicated that a small amount of furfural might be consumed by oxygen under low oxygen ERs at 400 °C, possibly only generating small-molecular permanent gases (e.g., CO, CO₂, or H₂O) rather than condensable liquid products.

When the temperature increased to 500 °C, as shown in Figure 2b, more products were observed both without and with oxygen addition, suggesting that the pyrolysis temperature was an important parameter to improve the reactivity of furfural during the pyrolysis process. The peak area of furfural decreased significantly with increasing ER; even without oxygen (ER of 0), its peak area was much lower than that at 400 °C (Figure 2a), indicating the large thermal decomposition and oxidation consumption of furfural at a higher temperature and under an oxidative atmosphere. But, even under an ER of 15% at 500 °C, the content of furfural (in terms of peak area percentage, same thereafter) in the product still accounted for ~90%, demonstrating that the formation of new condensable liquid products was quite low; most of the furfural was converted into permanent gases.

From the perspective of liquid product formation, under an ER of 0 and at 500 °C, benzene and benzofuran (dominant) were formed from furfural. As we know, furfural could undergo thermal decomposition to form furan and CO (C₅H₄O₂→C₄H₄O+CO) [33], while benzofuran could be directly formed from two furan molecules’ recombination with subsequent dehydration (2C₄H₄O→C₈H₆O+H₂O). In addition, furan would further decompose through C-C/C-O cleavage to form linear molecules or free radicals, such as HC≡CH, CH₂CO, CH₃C≡CH, HCCCH₂, CH₂CHCHCO, CO, and H radical [33], which might be re-aromatized to form benzene or benzofuran. Under an oxidative atmosphere, with the increasing ER, the yield of benzofuran (in terms of peak area) decreased with that of benzene increasing, and other simpler aromatics such as phenol and toluene also appeared in the products. The total detected peak areas (except that for furfural) also increased. It indicated that the introduction of oxygen promoted the decomposition of furfural and the formation of more aromatic products; the presence of oxygen would lead to the oxidative destruction of C-C/C-O bonds and the O removal of the furan ring (possibly forming CO and CO₂). More linear fractions or radicals apart from the abovementioned ones like CH₃, C≡CH, H₂C=CH, etc., might be generated under an oxidative atmosphere, thus resulting in a decrease in benzofuran and an increase in benzene, phenol, and toluene due to re-aromatization, especially at higher ERs. In addition, the furan might directly react with linear intermediates such as H₂C=CH and H₂C=CHCH₃ to form benzene and toluene through Diels–Alder condensation and dehydration reactions [34]. But, under too high an ER of 15%, the formation of these aromatics would also be inhibited.

3.2. Oxidative Pyrolysis of Hydroxyacetone Under Low ERs at 400 and 500 °C

Figure 3 shows the composition of products detected by GC-MS from the oxidative pyrolysis of hydroxyacetone under oxygen ERs of 0–15% at 400 and 500 °C. From Figure 3a, it can be seen that, at 400 °C, with the increasing ER, the peak area of hydroxyacetone decreased, indicating the promoted thermal decomposition/oxidation of hydroxyacetone in the presence of oxygen. But the peak areas of acetoin and acetylacetone also decreased, demonstrating that the decomposition of hydroxyacetone mainly generated permanent gases rather than condensable liquid products. The hydroxyacetone in the liquid product still accounted for ~96%, which was largely higher than the other products, showing the limited decomposition rate at 400 °C. From the perspective of liquid product composition, acetoin and acetylacetone were the major products from both non-oxidative and oxidative pyrolysis of hydroxyacetone. It might be inferred that hydroxyacetone (CH₃COCH₂OH) might undergo C-C and C-O cleavages to form CH₃CO, CH₃COCH₂, and CH₂OH intermediates, while (1) two CH₃CO would be coupled to form CH₃COCOCH₃, with one of the C=O bonds hydrogenated (with H radical) to form acetoin (CH₃COCHOHCH₃), and (2) CH₃CO and CH₃COCH₂ might also be coupled to form acetylacetone (CH₃COCH₂COCH₃). Although the presence of oxygen enhanced the cleavages of C-C and C-O bonds (illustrated by the decreased hydroxyacetone peak area), the coupling of the intermediates was inhibited with the decreased acetoin and acetylacetone under the oxidative atmosphere. It means that the further decomposition of these intermediates under an oxidative atmosphere might be strengthened with CO, CO₂, and hydrocarbon gases such as CH₄ releasing.

When the pyrolysis temperature increased to 500 °C (Figure 3b), the peak area of hydroxyacetone was further decreased compared to that at 400 °C, indicating that a higher temperature promoted the thermal decomposition of hydroxyacetone. But the presence of oxygen did not decrease the peak of hydroxyacetone, instead increasing the peak areas of other products, which means more condensable products formed under an oxidative atmosphere, especially acetoin. The presence of oxygen might promote the cleavage of the C-C bond connected with C=O in the hydroxyacetone molecule and the subsequent coupling of two CH₃CO, with more acetoin formation. While acetylacetone was not found in the product at 500 °C, instead, 2,3-pentanedione (CH₃COCOCH₂CH₃) and methyl glyoxal (CH₃COCHO) were observed with the ER at 6–15%. The formation of methyl glyoxal at a higher ER of 10% and 15% might be due to the oxidation of alcoholic hydroxyl in hydroxyacetone (CH₃COCH₂OH+0.5O₂→CH₃COCHO+H₂O), which was enhanced with more oxygen. Meanwhile, the formation of 2,3-pentanedione might be due to the intermolecular dehydration of CH₃COCOCH₃ and CH₂OH (aforementioned intermediates, CH₃COCOCH₃+CH₃OH→CH₃COCOCH₂CH₂+H₂O), subsequently stabilized by the H radical (CH₃COCOCH₂CH₂+H→CH₃COCOCH₂CH₃).

These findings confirm that pyrolysis temperature and oxygen introduction both significantly affect the thermal decomposition of hydroxyacetone. The presence of oxygen at a lower temperature (400 °C) promoted the decomposition and oxidation of hydroxyacetone to form more permanent gases, while the presence of oxygen at a higher temperature (500 °C) enhanced condensable liquid product formation, with their yields even higher than those at 400 °C, such as acetoin at an ER of 6%, although the yield of acetoin under a non-oxidative atmosphere was lower. The decomposition pathway of hydroxyacetone also changed at a higher temperature, with different products formed.

3.3. Oxidative Pyrolysis of 3,4-Dimethoxybenzaldehyde Under Low ERs at 400 and 500 °C

Figure 4 shows the composition of products detected by GC-MS from the oxidative pyrolysis of 3,4-dimethoxybenzaldehyde under oxygen ERs of 0–15% at 400 and 500 °C. From Figure 4a, it can be observed that the oxidative pyrolysis of 3,4-dimethoxybenzaldehyde only detected one product, i.e., 1,2-dimethoxybenzene, with the majority of the compound still being 3,4-dimethoxybenzaldehyde (~99.8%) at 400 °C. Although the presence of oxygen promoted the decomposition of 3,4-dimethoxybenzaldehyde with the observed decrease in the GC-MS peak area, the decomposition rate was not high, with a very low amount of 1,2-dimethoxybenzene formation. The content of 1,2-dimethoxybenzene even decreased at an ER of 6%, indicating that a higher ER might enhance the oxidation of these compounds to form permanent gases. The formation of 1,2-dimethoxybenzene (C₆H₄(OCH₃)₂) might be mainly due to the decarbonylation of the aldehyde (-CHO) group connected to the benzene ring (CHOC₆H₃(OCH₃)₂→C₆H₄(OCH₃)₂+CO), which might be enhanced in the presence of oxygen at a lower ER of 4%. But a higher ER might further oxidize it into permanent gases.

When the pyrolysis temperature increased to 500 °C (Figure 4b), under a non-oxidative atmosphere (ER of 0), it can be clearly seen that the peak area of 3,4-dimethoxybenzaldehyde was largely lower than that at 400 °C, indicating the significantly promoted thermal decomposition of 3,4-dimethoxybenzaldehyde at a higher temperature. The peak area of 1,2-dimethoxybenzene was nearly 16 times bigger than that at 400 °C, with more products such as toluene, 3-hydroxybenzadehyde, and 3,5-dimethyl-4-hydroxybenzaldehyde formed at a higher temperature. 1,2-Dimethoxybenzene was the major product followed by toluene and 3-hydroxybenzadehyde. The significantly enhanced direct decarbonylation of 3,4-dimethoxybenzaldehyde (forming 1,2-dimethoxybenzene) was the major reason. Furthermore, the 1,2-dimethoxybenzene (C₆H₄(OCH₃)₂) molecule could further remove the methoxy (OCH₃) group and combine with a methyl radical to form toluene (C₇H₈). The formation of 3-hydroxybenzadehyde (CHOC₆H₄OH) might be mainly due to the cleavage of the C_Ar-O and C_Methoxy-O bonds of 3,4-dimethoxybenzaldehyde (CHOC₆H₃(OCH₃)₂) to form CHOC₆H₃O, CH₃, and CH₃O intermediates, then stabilized by the H radical (CHOC₆H₃(OCH₃)₂→CHOC₆H₃O+CH₃+CH₃O, CHOC₆H₃O+2H→CHOC₆H₄OH). Despite its low content in the product, 3,5-dimethyl-4-hydroxybenzaldehyde might be also formed through similar bond (connected to aromatic ring) cleavage and recombination with radicals (H and CH₃). However, an oxidative atmosphere would inhibit the formation of these compounds; not only was the peak area of 3,4-dimethoxybenzaldehyde largely decreased due to enhanced oxidative decomposition but also fewer compounds were detected in the products. Only 1,2-dimethoxybenzene and toluene were observed at 500 °C under an oxidative atmosphere. A higher temperature and an oxidative atmosphere were beneficial for the removal of the aldehyde group, possibly due to the enhanced decarbonylation reaction (forming CO) or the oxidation of the aldehyde group (CHO) to the carboxyl group (COOH) with a subsequent decarboxylation reaction (forming CO₂). A higher ER (e.g., 15%) would significantly enhance these reactions.

3.4. Further Discussion on Oxidative Pyrolysis of Typical Volatile Model Compounds

Figure 5 shows the peak areas of the three volatile model compounds in the products after oxidative pyrolysis under low oxygen ERs of 0–15% at 400–500 °C. The data were gathered from the above sections, with a focus on the comparison of the conversion rate of different compounds during oxidative pyrolysis under different operation conditions. It should be pointed out that the peak areas of different compounds do not represent the amount of such compounds, as the response factors (peak area to weight percentage) for different compounds vary largely. For instance, the volatilization of 3,4-dimethoxybenzaldehyde (22.33 wt.%) was largely lower than that of furfural (100 wt.%), but the peak area of 3,4-dimethoxybenzaldehyde in the product was greatly higher than that of furfural with less of both being decomposed under a non-oxidative atmosphere at 400 °C. But, for the same compound, the peak area can be used to compare the amounts of the compound obtained under different conditions.

Temperature indeed plays an important role during both non-oxidative and oxidative pyrolysis of volatile compounds. For instance, the peak area of 3,4-dimethoxybenzaldehyde at 500 °C under an inert atmosphere (ER of 0) was only ~25% of that at 400 °C, and it further decreased to ~9% at an ER of 4% compared with that at 500 and 400 °C. At 400 °C, with the increasing ER, the peak areas of all three compounds decreased slightly, indicating that the oxidative decomposition of the volatile compounds at a lower temperature is limited. At 500 °C, with the increasing ER, the peak area of furfural showed a clear decreasing trend, and the peak area of hydroxyacetone just changed slightly, while the peak area of 3,4-dimethoxybenzaldehyde first decreased greatly and then was relatively stable. These results demonstrate that the oxidative pyrolysis characteristics of different volatile compounds vary greatly. One possible reason would be the chemical structure differences between different compounds, resulting in distinct reactivities. For instance, hydroxyacetone, as a linear compound with unsaturated bonds, is quite reactive. It can be decomposed quite sufficiently at a higher temperature, thus reducing the impact of oxygen introduction. Another possible reason would be their differences in autoignition temperatures. For instance, 3,4-dimethoxybenzaldehyde with a lower autoignition temperature (192 °C) would be more likely to be oxidatively decomposed [19], as illustrated by the larger decrease ratio in peak area with and without oxygen introduction (500 °C). Meanwhile, furfural with a higher autoignition temperature (392 °C) showed a lower decrease ratio but a continuous decrease in peak area. For real volatiles, competitive oxidative pyrolysis is expected among different compounds, which necessitates further investigation.

4. Conclusions

Oxidative pyrolysis of three typical volatile model compounds under low ERs at 400 and 500 °C was conducted in a two-stage quartz-tube reactor. It was found that, at a lower temperature of 400 °C, the oxidative pyrolysis of volatiles was limited even with oxygen introduction, while a higher temperature of 500 °C significantly promoted the decomposition of volatile compounds with the observed great decrease in their peak areas. The presence of oxygen could enhance the decomposition of volatile compounds but showed very distinct impacts at different temperatures on varied volatile compounds possibly due to their different chemical structures and autoignition temperatures. During the pyrolysis process, the decomposition of volatile compounds mainly generated permanent gases rather than condensable liquid products. In terms of liquid product formation, furfural would undergo the cleavage of C-C/C-O bonds to form linear intermediates and subsequent aromatization to generate aromatics (benzene and benzofuran). The presence of oxygen could enhance the oxidative destruction of the C-C/C-O bonds and the removal of O from the molecules to form simple aromatics such as benzene, phenol, and toluene. Hydroxyacetone mainly underwent C-C/C-O cleavage that was further enhanced in the presence of oxygen; the resultant intermediates would recombine to generate acetoin and 2,3-pentanedione. A higher ER would directly oxidize the alcoholic hydroxyl group (-OH) into the aldehyde group (-CHO) to form methyl glyoxal, while 3,4-dimethoxybenzaldehyde mainly underwent cleavage and recombination of bonds connected with the benzene ring including aldehyde group (-CHO), C_Ar-O, C_Methoxy-O bonds, thus forming 1,2-dimethoxybenzene, toluene, and 3-hydroxybenzadehyde. This study further deepens the fundamental understanding of the homogeneous oxidation of volatiles during oxidative fast pyrolysis of biomass, offering possibility to selectively producing value-added products through biomass autothermal pyrolysis in the future.