Introduction

Energy lays a solid foundation for the survival and development of human beings from everlasting. Still, nowadays the large amounts of depletion in traditional fossil fuels have brought about serious environmental issues1. Among diverse corresponding substitute energy, biomass resources are gaining increasing concern for their ideal renewable potentials, cleanness, and economy, with the pyrolysis conversion as a preferable utilization method for transforming which to produce high-added value products like biochar, bio-oil, and gases with high calorific value2. There are abundant inorganic components such as K, Na, Ca, and Mg existing in biomass inherent structures, namely the Alkali and Alkaline Earth Metal Ions (AAEMIs) which refer to IA and IIA groups of the periodic table, respectively. AAEMIs could facilitate coke formation, the secondary cracking of volatiles, and the release of more incondensable gas3. Furthermore, AAEMIs could make a difference in the thermodynamic behavior during biomass pyrolysis, while the possible effects vary with diverse factors such as temperature, mass ratio4,5. Different catalytic effects on pyrolysis characteristics were also observed between alkali and alkaline metals6, thus it makes sense for more attention paid to the research on the possible catalytic mechanism.

Several studies on the effect of alkali and alkaline metals on biomass pyrolysis have been reported. Aho et al.7 found that organically bound alkali metals (K and Na) decreased the yield of bio-oil and increased that of biochar. Besides, effect of which is more significant than that of organically bound Ca, Mg and Mn. Dalluge et al.8 reported that acetate salts of Li, Na, and K increased the content of aromatic compounds in bio-oil at 500–700 °C during fast pyrolysis of lignin. Jalalabadi et al.9 found that molten carbonates (Li, Na, and K) could catch volatiles by trapping volatiles inside of bulky bubbles, which caused the generation of various size of pores as well as char making reactions. For the pyrolysis distribution, Yang et al.10 claimed that compounds containing C-O-C groups were the main products in biomass pyrolysis after acid washing, but with the existence of K+ and Ca2+, compounds containing C = O groups were the main products. Li et al.11 found that organically bound Ca2+ and Mg2+ promoted the generation of LG (levoglucosan or 1,6-anhydro-β-D-glucopyranose). While Patwardhan et al.12 reported that calcium salts CaCl2, Ca(OH)2, Ca(NO3)2, CaCO3 and CaHPO4 all inhibited the generation of LG. Similarity, for the product formation pathway, Yu et al.13,14,15 found that CaCl2 and MgCl2 can inhibit the cleavage of glycosidic bonds during cellulose pyrolysis, promoting the dehydration and ring opening of pyranose, further cleaved into small molecular fragments. Arora et al.16 studied the influence of Ca2+ and Mg2+ on glycosidic bond breaking reactions during cellobiose pyrolysis through DFT calculations, and found that Ca2+ and Mg2+ exerted an inhibitive effect on dehydration. In addition, for the evolution of alkali inorganic salts, Chen et al.17 claimed that a fraction of KCl could react with carboxyl and carbonyl groups to form organically bound K, producing a catalytic effect in further pyrolysis. While Knudsen et al.18 proposed that KCl remained stable in solid form at 700–800 °C. Similar expression was also found that a significant part of unreacted KCl existed in the pyrolysis of KCl-loaded cellulose as the increasing content of the potassium salts was added at 500 °C19, which implied possible role of metal salts as catalysts.

Despite the vast amount of research that has been conducted on AAEMIs and the valuable insights they have provided into the role they play in biomass pyrolysis; it is important to note that there have been instances where contrasting results have emerged. Therefore, it is still necessary to conduct a comprehensive exploration of the catalytic mechanism of AAEMIs in biomass pyrolysis to utilize biomass resources more effectively.

To further explore the catalytic mechanism of AAEMIs on biomass pyrolysis. Some researchers have studied the effect of both metal cations and anions on pyrolysis product to some extent. Yin et al.20 demonstrated KOH and K2CO3 strongly reduced the thermal stability of the lignin-related polymers and increased the biochar yield, while effect of KCl was limited. Moreover, Patwardhan et al.12 Compared the effect on cellulose pyrolysis from five species of calcium salts (CaCl2, Ca(OH)2, Ca(NO3)2, CaCO3), and CaHPO4), identifying a decrease in the LG yield in the order: Cl− > NO3− ~ OH− > CO32− ~ PO43−. Xi et al.21 adopted different potassium salts (K2CO3, KOH, K2C2O4, and K3PO4) to activate lignin-based carbon samples during pyrolysis, with results that the influence of the four potassium salts on the specific surface area and pore volume of biochar was as follows: CO32− > OH− > C2O42− > PO43−. It could be found that the selection of pyrolysis raw materials in individual study is single and limited. What’s more, it’s typically for researchers to focus on studying the effect of metal cations or anions, and change of single-phase product from biomass pyrolysis. Biomass catalytic pyrolysis is a complex process, and the reaction mechanism is closely related to the evolutionary properties of biochar, bio-oil, and gas products. Besides, both cation and anion must be analyzed when assessing the effect of salts on pyrolysis. To summarize researchers’ effort on studying effect of AAEMLs on the pyrolysis reactions of lignocellulosic components briefly, some revelatory information has been presented in a schematic format (Fig. 1). However, the coverage is still limited and necessary to comprehensively explore the effect of both cations and anions on the evolution properties of all pyrolysis products and reveal biomass catalytic pyrolysis mechanism with AAEMIs in depth.

Fig. 1: The possible effects of AAEMLs.
figure 1

The revealed effect of AAEMLs on the pyrolysis reactions of lignocellulosic components.

Full size image

In this study, cellulose and bamboo catalytic pyrolysis with different cations and anions catalysts (KCl, K2SO4, K2CO3, NaCl, Na2SO4, and Na2CO3. Alkalinity strength relationship: CO32− > SO42− ~ Cl−, K+ > Na+) were investigated in the fixed-bed reaction system, with SiO2 additive as control group for eliminating influences on mass and heat transfer. The composition of lignocellulosic biomass is complex, among which cellulose, hemicellulose and lignin are the major components and cellulose is usually the most abundant composition3. Bamboo was chosen for it a typical case of lignocellulosic biomass. The effect of cations (K+ and Na+) and anions (Cl−, SO42− and CO32−) on the evolution properties of biochar, bio-oil, and gas products was explored under both in-situ and ex-situ catalytic pyrolysis. Characterization on pyrolysis products was also carried out as comprehensively as possible. Furthermore, the mechanism of cellulose and bamboo catalytic pyrolysis was proposed based on the distribution of all pyrolysis products, which is important to the high-value utilization of biomass.

Results

Effect of alkali metal salts on the product distribution

Figure 2 shows the product distribution from cellulose and bamboo catalytic pyrolysis with different alkali metal salts. From Fig. 2a. Cellulose pyrolysis with SiO2 generated higher yield of bio-oil (60 wt%) with some biochar and gas products. KCl showed no obvious effect on the product distribution of cellulose pyrolysis. K2SO4 promoted the generation of biochar and gas products a little, while inhibited the formation of bio-oil. Greatly different from KCl and K2SO4, K2CO3 significantly increased biochar yield, and largely decreased bio-oil yield to only 41 wt%, while gas yield increased 2 times to 29 wt%. To summarize, potassium salts facilitated the secondary fragmentation of volatiles, thereby increasing the formation of gas products and biochar, while reducing the yield of liquid products. This effect was particularly noticeable for K2CO3, with its reduction in liquid products likely due to its higher alkalinity compared to K2SO4. The catalytic effect of potassium salts on cellulose pyrolysis followed the order: K2CO3 > K2SO4 > KCl. Compared with potassium salts, NaCl inhibited the formation of bio-oil, and Na2SO4 showed a stronger inhibiting effect on bio-oil. Like K2CO3, Na2CO3 also greatly promoted the gas releasing (25 wt%), while decreased bio-oil yield (44 wt%). The catalytic effect of sodium salts on cellulose pyrolysis also followed the order: Na2CO3 > Na2SO4 > NaCl.

Fig. 2: Pyrolysis product distribution.
figure 2

Effect of alkali metal salts in in-situ catalytic pyrolysis. a Cellulose pyrolysis. b Bamboo pyrolysis.

Full size image

Compared with cellulose, from Fig. 2b. Bamboo catalytic pyrolysis with SiO2 addition generated more biochar and gas products, while less bio-oil. Similar with cellulose, KCl addition had no obvious effect on the product distribution of bamboo pyrolysis. K2SO4 inhibited the bio-oil generation a little. While K2CO3 greatly promoted the formation of biochar (26 wt%) and gas products and inhibited the formation of bio-oil (39 wt%). The catalytic effect of potassium salts on bamboo pyrolysis also followed the order: K2CO3 > K2SO4 > KCl. Sodium salts also showed a similar order: Na2CO3 > Na2SO4 > NaCl. Besides, the biochar yield from bamboo pyrolysis was higher than that from cellulose, which may be due to it being easy for the secondary refinement of hemicellulose and lignin pyrolysis volatiles22,23,24. Metal salts may also enhance the interaction between biomass components, which results in the increase of biochar yield25,26.

Effect of alkali metal salt on gas products

The effect of potassium and sodium salts on the gas compositions of cellulose and bamboo pyrolysis is shown in Fig. 3. From Fig. 3a. Gas product from cellulose pyrolysis with SiO2 was mainly CO (2.1 mmol/g) and CO2 (1.4 mmol/g) with some CH4 (0.3 mmol/g) and H2 (0.1 mmol/g). KCl increased CO2 yield while decreasing the yield of other compositions slightly. K2SO4 not only promoted the generation of CO2 but also increased the yield of CH4 and H2. Greatly different from KCl and K2SO4, K2CO3 greatly increased the yield of all gas compositions that CO yield increased to 2.9 mmol/g, and the yield of CO2, CH4, and H2 increased 3 times (4.5 mmol/g), 2 times (0.4 mmol/g), and 9 times (1.2 mmol/g), respectively. It indicated that the catalytic effect of K2CO3 was significantly stronger than that of K2SO4 and KCl, which could not only accelerate the decarboxylation, demethylation, and dehydrogenation reactions but also promote the decarbonylation reactions of volatiles. It is related to its strong alkali characteristics of K2CO327,28,29. Compared with potassium salts catalysts, sodium salts showed some different catalytic properties. NaCl increased the yield of CO2 and CH4. Na2SO4 also promoted the release of CO2, CH4 and H2. Na2CO3 also greatly increased the yield of CO (2.4 mmol/g), CO2 (3 times to 4.0 mmol/g), CH4 (2 times to 0.5 mmol/g), and H2 (5 times to 0.6 mmol/g). It could be observed that the catalytic effect of potassium salts was obviously stronger than that of sodium salts. It may be ascribed to that the metal strength of K is larger than that of Na30.

Fig. 3: Gas compositions.
figure 3

Effect of alkali metal salts on the gas products in in-situ catalytic pyrolysis. a Cellulose pyrolysis. b Bamboo pyrolysis.

Full size image

The case of bamboo pyrolysis was different from cellulose. From Fig. 3b. Bamboo pyrolysis with SiO2 generated more CO2 (2.1 mmol/g), CH4 (0.6 mmol/g), and H2 (0.2 mmol/g), while less CO (2 mmol/g), which was due to the decomposition of hemicellulose and lignin in bamboo. KCl slightly decreased the yield of all gas compositions from bamboo pyrolysis. In contrast, K2SO4 significantly increased the yield of all gas compositions. Although K2CO3 greatly promoted the formation of CO2 (2 times to 4.2 mmol/g), CH4 (0.8 mmol/g), and H2 (8 times to 1.4 mmol/g), it inhibited the generation of CO (only 1.7 mmol/g). Like potassium salts, NaCl showed little effect on gas compositions, and Na2SO4 greatly promoted the gas compositions formation, while K2CO3 only increased the yield of CO2 (1.5 times to 3.6 mmol/g), CH4 (0.7 mmol/g), and H2 (5 times to 0.9 mmol/g), and decreased CO yield to only 1.5 mmol/g. The catalytic effect of potassium salts and sodium salts on bamboo pyrolysis also showed the order: CO32− > SO42− > Cl−, and the catalytic effect of K+ and Na+ showed the order: K+ > Na+. The results were still in line with the alkalinity strength relationship. It could be pointed out that CH4 and H2 from bamboo catalytic pyrolysis (Fig. 3b) were significantly higher than that from cellulose (Fig. 3a), which was due to that metal salts promoted the cracking of -OCH3 of lignin and released more CH4 and H2. While the yield of CO2 and CO showed opposite tendency with the addition of K2CO3 and Na2CO3, which may be attributed to cellulose, with regular and ordered polysaccharide structure, the primary pyrolysis product of which was rich in pyran compounds, and K2CO3 and Na2CO3 could promote the reforming of pyran rings through releasing CO2 and CO23.

Effect of alkali metal salts on bio-oil products

The main organic compositions of bio-oil from cellulose and bamboo catalytic pyrolysis are classified into pyrans (including anhydrosugar and pyrone), phenols, furans, cyclopentanones and aliphatics (short-chain aliphatic compounds), and the content of these compositions is shown in Fig. 4. From Fig. 4a. Cellulose pyrolysis with SiO2 addition mainly generated pyrans (85%, mainly including LG, with a little of furans, cyclopentanones and aliphatics. KCl and K2SO4 showed similar catalytic effect, which largely decreased the content of pyrans, while increased that of furans (mainly including furfural (FF) and 5-hydroxymethylfurfural (5-HMF)), cyclopentanones and aliphatics accordingly, but pyrans were still the main component, accounting for 71%. The decrease in pyrans may be attributed to their lower stability of them, which easily decomposed into small molecular products with the effect of KCl and K2SO4. Greatly different from KCl and K2SO4, with the addition of K2CO3, the bio-oil components showed subversive change. There was nearly no pyrans generation, while the proportion of cyclopentanones and aliphatics increased significantly, and became the main components, which accounted for 53% (mainly including 2-cyclopenten-1-one, 2-methyl-2-cyclopenten-1-one, 3-methyl-2-cyclopenten-1-one) and 28% (mainly including 1-hydroxy-2-propanone), respectively. Interestingly, a few phenolic compounds also occurred with the addition of alkali metal carbonate, indicating that rearrangement reaction to form aryl compounds was promoted31. NaCl and Na2SO4 also largely decreased the content of pyrans (which decreased to 60% and 77%, respectively), while increased the content of other components. But the catalytic effect of NaCl was significantly stronger than that of Na2SO4. Na2CO3 showed the strongest catalytic effect, which also made cyclopentanones become the dominant component (58%) with 24% aliphatics. The order of catalytic effect was CO32− > Cl− > SO42−.

Fig. 4: Composition of liquid products.
figure 4

Effect of alkali metal salts on the bio-oil in in-situ catalytic pyrolysis. a Cellulose pyrolysis. b Bamboo pyrolysis.

Full size image

The decrease of pyrans was mainly due to the decrease of LG, which was consistent with the increase of FF and 5-HMF. It indicated that Cl- promoted the ring-opened reactions of LG to form FF and 5-HMF. K2CO3 and Na2CO3 promoted the ring-opening reactions of cellulose, and the further cracking and condensation reactions of the intermediates27,32,33. Some intermediates underwent Grob cracking, dehydration reactions, and keto-alcohol isomerization reactions to form short-chain small molecule compounds and gas products23. The case was consistent with the sharp increase of gas products from Fig. 3a. and Supplementary Table 1.

Different from cellulose, for bamboo pyrolysis liquid products in Fig. 4b. The major components were phenols (47%, mainly including 4-vinyl phenol, 4-ethyl phenol, 2,6-dimethoxy phenol, phenol, and p-cresol) and aliphatics (31%, mainly including acetic acid) with SiO2 addition. It could be observed that the content of LG was strikingly decreased, which was due to the presence of lignin and hemicellulose exerting a significant inhibitory effect on the pathway of cellulose pyrolysis. Lignin was the main source of phenols through a series of reactions like hydroxyl groups removal, cleavage of weak ether bonds and alkyl side chain removal34. While the hemicellulose will release acetic acid through the fragmentation of acetyl substituents35,36. KCl and K2SO4 showed no obvious effect on the bio-oil components, which slightly increased the content of phenols. However, K2CO3 greatly promoted the generation of phenols, the content of which increased to 81%, while inhibiting the formation of other components mainly aliphatics. Like potassium salts, there was no obvious change in bio-oil components after the addition of NaCl and Na2SO4, and Na2CO3 also sharply increased the content of phenols to 81%. It has been reported that potassium additives can additionally promote decarboxylation or decarbonylation reactions, along with the removal of unsaturated alkyl branch chains37. Moreover, as shown in Supplementary Table 1 and Table 2, acids and aldehydes were not detected with the addition of alkali metal carbonate, and proportion of ketones was lowered. For one thing, these could partially be attributed to competition among different biomass components, the remarkable catalytic effect of alkali metal carbonate on ketonization reactions was inhibited. For another, due to the stronger alkaline nature, they were more prone to react with -COOH functional groups and release CO2, which was consistent with the significant increase in CO2 in the pyrolysis gas.

Figure 5 shows the composition of the phenols after the addition of K2CO3 and Na2CO3. After the addition of K2CO3 and Na2CO3, the content of phenol changed slightly. While K2CO3 and Na2CO3 promoted the formation of alky-phenols, and the content increased more than 2 times (reaching 47.9% and 47.7%, respectively). The content of methoxy-phenols remained at 20%. This was mainly due to the that demethoxylation reaction was promoted and alkylation happened during depolymerization and cracking of lignin38. Compared with OH radical the oxygen atom in the methoxy group has a higher electron density39, thus it was more preferentially for the methoxy group accounting to be adsorbed on K2CO3 and Na2CO3, resulting in the secondary decomposition.

Fig. 5: Distribution of phenols.
figure 5

Respective Effect of K2CO3 and Na2CO3 on distribution of phenols in bamboo bio-oil was compared with control group (SiO2).

Full size image

The effect of K2CO3 and Na2CO3 presented above was mainly based on cellulose in-situ catalytic pyrolysis, during the pyrolysis process both solid-solid and gas-solid contact between raw material and alkali metal salt were included. Separation of pyrolysis and catalytic processes was necessary for studying the effect of catalysis mode. Given this, the experiments of K2CO3 and Na2CO3 as catalysts for cellulose ex-situ catalytic pyrolysis were set. The composition properties of bio-oil from cellulose ex-situ catalytic pyrolysis with K2CO3 and Na2CO3 catalysts are shown in Supplementary Fig. 1. The main products in bio-oil from cellulose ex-situ catalytic pyrolysis with SiO2 were FF and LG, and the content were 4.4% and 57.6%, respectively. After the addition of K2CO3, FF content increased over 2 times to 10.8%, while LG content decreased to 44.7%. Like K2CO3, the addition of Na2CO3 increased FF content (7.4%) and decreased LG content (47.9%). Moreover, K2CO3 and Na2CO3 also greatly increased the yields H2, CO2, CH4, and CO. For cellulose in-situ catalytic pyrolysis with K2CO3 and Na2CO3 catalysts, there were no FF and LG generation. It indicated that in-situ catalytic pyrolysis had stronger catalytic effect than ex-situ catalytic pyrolysis, as FF and LG were catalytic decomposed further. FF can be reduced to 2-furanmethanol under the action of H radical40. Furthermore, the demethoxylation and decarbonylation reactions could facilitate the supply of H radicals during in-situ pyrolysis41. Thus, the catalytic effect from solid-solid contact between raw material and alkali metal salts likely had quite an important role in assisting gas-solid contact pyrolysis.

Effect of alkali metal salts on the chemical structure of biochar

To comprehensively elucidate the effect of potassium salts and sodium salts on biomass catalytic pyrolysis, the morphology structure of biochar is shown in Fig. 6. The peak at 2θ = 23°represents the peak (002) indicating the carbon layers are stacked in parallel42, and the peak (100) located at around 2θ = 44° is used as an indicator of an orderly stacking of aromatic layers43. Both of which are used as indicators of the degree of graphitization.

Fig. 6: XRD pattern of biochar.
figure 6

a Cellulose pyrolysis. b Bamboo pyrolysis.

Full size image

For cellulose catalytic pyrolysis in Fig. 6a. KCl and K2SO4 showed similar effects on the biochar structure, with a slight decrease in intensity of the peak (002) in contrast to that of SiO2 control group. While K2CO3 made a rather broader peak (002) of biochar, which indicated a highly disordered structure. Besides, a slight left shift of the peak location was observed. These indicated that the crystallinity of carbon in the stacking direction was significantly reduced. For sodium salt catalysts, NaCl featured the highest peak (002) intensity, and the curve trend of Na2SO4 was similar with that of Na2CO3. Compared with K2CO3, the amorphous degree of carbon skeleton decreased a lot in Na2CO3 group. The peak (100) in all cellulose catalytic pyrolysis groups presented low intensity.

Similar with cellulose pyrolysis, from Fig. 6b. The obvious asymmetry of peak (002) and the slight peak (100) occurred in biochar from bamboo catalytic pyrolysis. K2SO4 showed the highest broad of peak (002) in potassium salts, followed by KCl and K2CO3. Among these the peak (002) of K2CO3 turned left a lot, indicating the large amount of potassium impregnation into the biochar matrix during pyrolysis, wrecking the regular biochar structure6. In comparison, nearly all the sodium salts exhibited a higher intensity of the peak (002) than of potassium salts. Besides, the intensity of peak (100) in bamboo catalytic pyrolysis was little higher than that of cellulose catalytic pyrolysis. However, the degree of graphitization was still low with amorphous carbon being the main component.

Figure 7 shows the FTIR spectrum of biochar from cellulose and bamboo catalytic pyrolysis. From Fig. 7a. Compared with cellulose catalytic pyrolysis with SiO2, the addition of KCl had little effect on the property of biochar. For the addition of K2SO4, the bending vibration peak of the aromatic ring C-H in the infrared spectrum of biochar was strengthened, and the maximum peak position shifted from 875 to 810 cm−1, indicating that the structure of the biochar matrix changed from complex position and poly-substituted aryl structure to simple one. The addition of K2CO3 strikingly increased the absorbance of C = C (1580 cm−1), promoting the development of aromatic structure. While absorbance of aromatic C-H stretching vibration and out of plane deformation were significantly weakened, indicating that the substitution of aromatic hydrogen atoms was strengthened and fused rings were formed44,45, This could be attributed to the promoting effect on cleavage of hydrogen bond by K2CO3, which was corresponding to the increase of H2 production (9 times that of SiO2 control group) in the gas products (Fig. 3a). Besides, the content of aryl C-O-C groups was observed to be a little higher than that of other groups. NaCl and Na2SO4 were similar with that of KCl and K2SO4, respectively. Na2CO3 showed relatively inferior effect on the aromatic skeletal vibration than that of K2CO3.

Fig. 7: FTIR spectrum of biochar.
figure 7

a Cellulose pyrolysis. b Bamboo pyrolysis.

Full size image

For bamboo catalytic pyrolysis in Fig. 7b. The SiO2 control group showed that the peak intensity approximately from 1000 to 1500 cm−1 increased, representing more oxygen functional groups on the surface of biochar. KCl slightly undermined the absorbance of aromatic skeletal vibration. The addition of K2SO4 showed similar effect. At the same time, the C-O-C absorption peak of aryl ethers became more and more obvious. As to K2CO3, aromatic C-H structure was weakened. C-O group of aromatic rings became more obvious and the absorbance of aromatic skeletal vibration was also facilitated, along with more H2 production (Fig. 3b). As to sodium salts, in comparison, NaCl showed similar effects with KCl. Na2SO4 increased the aromatic skeletal vibration a lot compared with K2SO4. While Na2CO3 showed slightly inferior absorbance of functional groups than that of K2CO3.

Discussion

Based on above results, the possible catalytic pyrolysis mechanism of alkali metal salts on biomass pyrolysis is shown as Fig. 8. In in-situ catalytic pyrolysis, alkali metal carbonate showed best catalytic effect through both solid-solid and gas-solid catalytic reaction, followed by alkali metal sulfate, and alkali metal chloride had the weakest effect. For solid-solid catalytic reactions, the basic structure unit was cracked under catalytic effect by alkali metal salts on continuous reactions such as dehydration, depolymerization, ring opening, decarboxylation, decarbonylation and demethoxylation, along with adequate free radical reaction for various product species. For gas-solid catalytic reactions, alkali metal salts promoted the cracking of volatiles, with primary products decomposed into small molecule fragments, such as ring opening reaction and side chain removal of primary products with pyran and furan ring in cellulose pyrolysis. Associating separate comparison between in-situ and ex-situ catalytic pyrolysis mode, the combination between solid-solid and gas-solid catalytic reaction of alkali metal salts were likely to make more further-converted products.

Fig. 8: Possible reaction mechanism.
figure 8

Possible catalytic pyrolysis mechanism of alkali metal salts on biomass pyrolysis was proposed based on the results obtained from this study.

Full size image

In detail, KCl changed the composition of bio-oil derived from cellulose pyrolysis, mainly because that Cl− enhanced the homolytic cleavage the yields of of intermediate towards the formation of light oxygenates like 5-HMF, owing to the further conversion to linear aldehydes by ring scission, later being converted to 5-HMF via dehydration reaction under the catalysis by KCl46, and thus the reaction pathway for anhydrosugar (mainly LG) was inhibited a lot during the process of pyrolysis. K2SO4 showed similar catalytic properties with KCl. In comparison, K2CO3 considerably shifted the distribution of bio-oil and increased gas products. The formation of macromolecule (mainly LG) was greatly inhibited and the content of hydroxyacetone and cyclopentanone compound was remarkably increased47,48,49,50. For bamboo catalytic pyrolysis, K2CO3 significantly promoted the formation of phenols. Demethoxylation was facilitated and large amounts of radicals such as CH3 radical were released under the catalytic effect and random recombination occurred between different radicals, corresponding to the high content of alkylphenols. For sodium salts, NaCl promoted the formation of furans (FF and 5-HMF). Na2SO4 and Na2CO3 showed similar catalytic properties with corresponding potassium salts.

In general, alkali metal salts promoted secondary cleavage of the volatiles to form more small molecular gas and biochar products, with catalytic effect in the order: CO32− > SO42− > Cl−. The decomposition of pyran ring was facilitated, and aliphatics and cyclopentanones gradually became the main components from cellulose pyrolysis. For bamboo pyrolysis, alkali metal carbonate showed striking catalytic effect with phenols being the major component. Alkali metal salts also promoted dehydration, removal of substituent group, and the formation of aromatic skeleton. Besides, the catalytic effect from solid-solid contact between raw material and alkali metal salts performed an important role in assisting gas-solid contact pyrolysis.

Methods

Materials

Bamboo waste was obtained locally, which was crushed and sieved to obtain a particle less than 120 μm. Cellulose was purchased from Sigma-Aldrich Co. LLC, the average particle size of which was 20μm with the polymerization degree of 3000–10000. The materials were dried at 105 °C for 24 h before use. The potassium salts (KCl (> 99.8, GR), K2SO4 ( > 99.5, AR), K2CO3 ( > 99.8, GR)) and sodium salts (NaCl (> 99.5, AR), Na2SO4 (> 99.5, AR), Na2CO3 (> 99.5, AR)) were purchased from Sinopharm Chemical Reagent Co., Ltd.

Proximate analysis of the sample was carried out in a TGA-2000 analyzer (Las Navas, Spain). Ultimate analysis of sample was conducted using a CHNS/O elementary analyzer (Vario Micro Cube, Germany). The content of hemicellulose, cellulose, and lignin in bamboo waste was determined by Van Soest51,52. Bamboo waste contained high volatile content of 85.4 wt% with only 0.7 wt% ash. Carbon and oxygen content reached 49.4 wt% and 43.8 wt%, respectively. Besides, cellulose was the main biochemical composition (46.5 wt%) with some hemicellulose and lignin. While cellulose contained higher volatile content of 95.5 wt% with 50.1 wt% oxygen.

Biomass catalytic pyrolysis experiment

Biomass catalytic pyrolysis was performed in a fixed-bed reactor system. The system consisted of a feeding unit, a fixed bed reactor (400 mm height, 30 mm inner diameter), a condensing unit, and a gas cleaning and drying unit, with temperature and gas flow controllers.

Prior to each trial, a quartz basket loaded with uniformly mixing sample (~2 g, the mass of biomass and metal salt was 1.6 g and 0.4 g, respectively33,53) was held on the top of the reactor with N2 purging (99.999%, 100 mL/min). When the reactor was heated up to the selected temperature of 550 °C (biomass pyrolysis at 550 °C could obtain higher quality of biochar, bio-oil, and gas products together54), the quartz basket loaded was promptly placed in the center of the reactor and retained for 30 min. The sample was heated up rapidly and the volatiles evolved out quickly. The condensable volatiles were trapped in the ice-water mixture condensing unit. The non-condensing gas was collected with a gas bag for further analysis.

After each experiment, the reactor was cooled to the ambient temperature under N2 purging. The biochar yield was determined with the weight difference of the quartz basket before and after pyrolysis (the weight of metal salt added to feedstock was subtracted in the calculation of biochar yield). Similarly, the bio-oil yield was determined by the initial and final weight of the ice-water mixture condensing unit. The gas yield (wt%) was calculated by combining the total gas volume and the gas density. The gas volume was determined by the mass flow controller, while gaseous mean density was determined based on the gas component from the testing results of micro-gas chromatography55. Relative formulas are listed as follows:

$${Solid\; yields}\,\left({wt}. \% \right)=100 \% -\frac{M-{M}_{{after}}}{{M}_{{raw}}}* 100 \%$$
(1)
$${Liquid\; yields}\,\left({wt}. \% \right)=\frac{{C}_{{after}}-{C}_{{before}}}{{M}_{{raw}}}* 100 \%$$
(2)

Mraw is the cellulose/bamboo raw material; M is the total mass of quartz basket, raw material and catalyst before reaction; Mafter is the mass of solid residue and quartz basket after reaction; Cbefore is the mass of the empty condenser bottle; Cafter is the mass of the condenser bottle containing bio-oil.

$${V}_{{gas}}=\frac{{Q}_{{N}_{2}}* t}{{x}_{{N}_{2}}}$$
(3)
$${Gas\; yields}\,\left({wt} \% \right)=\frac{{\sum }_{i=1}^{n}\left(\frac{{M}_{i}}{22.4}{x}_{i}\right){V}_{{gas}}}{{M}_{{raw}}}$$
(4)
$${{Gas}}_{i}{yield}\left({mmol}/g\right)=\frac{{V}_{{gas}}* {x}_{i}}{{22.4M}_{{raw}}}$$
(5)

During which Mraw is the mass of raw materials. Vgas is the total volume of the collected gases (ml); \({Q}_{{N}_{2}}\) is the volume flow (200 ml/min); t is the time for collecting gases (30 min); \({x}_{{N}_{2}}\) is the volume percentage of N2 measured by GC; \({{Gas}}_{i}\) refers to one of the gas presented as follows: CO2, CO, H2, CH4 and small alkanes; \({x}_{i}\) is the volume percentage of \({{Gas}}_{i}\) measured by GC; Mi is the molar mass of \({{Gas}}_{i}\); n represent the total number of gas species.

Moreover, the blank experiment of biomass pyrolysis with quartz sand (SiO2) addition (the mass of biomass and quartz sand was 1.6 g and 0.4 g, respectively) at 550 °C with N2 was carried out to remove the effect of mass and heat transfer. Each experiment was performed at least three times, and all the data took the average. Biochar was immersed in deionized water to remove metal salts on biochar56,57, and subsequently washed repeatedly with deionized water. They were then dried in an electrical oven at 105 °C for 24 h.

Characterization

The gas products were analyzed using gas chromatography (GC, Panna A91, China) with a three-valve four-column system and two detectors. Thermal conductivity detector (TCD) was for quantitative analysis of conventional gases, e.g., H2, O2, N2, CO, CO2, and CH4) and the hydrogen flame ionization detector (FID) was for quantitative analysis of small alkanes, e.g., C2 ~ C5, C2H6, C2H4, C2H2, C3H8, C3H6, C4H10, C4H8, C5H10).

The main compositions of N-enriched bio-oil were identified using a gas chromatography-mass spectroscopy (GC-MS, HP7890 series GC with an HP5977 MS detector) with a capillary column (Agilent: DB-5MS, 19091s-433; 30 m × 0.32 mm i.d. × 0.25 μm d.f.). The injector temperature was 280 °C; the carrier gas (He) flow rate for the column was 1 mL/min, and the split ratio was 40:1. In each case, a sample injection volume of 1 μL was employed. The GC oven was initially heated at 50 °C for 3 min, after which time it was heated to 100 °C at a rate of 10 °C/min, followed by heating to 200 °C at a rate of 5 °C/min, then it was heated to 300 °C at a rate of 10 °C/min, and finally the oven temperature was maintained at 300 °C for 3 min. The mass spectrometer was operated in a mass/charge ratio (m/z) of 35–500. The compounds were identified using a National Institute of Standards and Technology library (NIST2014).

The organic functional group was analyzed with Fourier Transform Infrared Spectroscopy (FTIR, VERTEX 70 spectrometer, Bruker, Germany). The scanning range was 4000-400 cm−1 with a resolution of 4 cm−1. For pellet preparation, the dried solid sample (0.75 mg) and KBr (75 mg, spectroscopy grade, Merck) were mixed and then pressed into a transparent sheet. KBr was used as a blank background to account for noise, e.g., from moisture and CO2. The FTIR spectra were smoothed and baseline-corrected using Bruker OPUS 7.0 software. The crystal structures of cellulose and the char were investigated using X-ray diffraction (XRD, Empyrean, PANalytical, Netherlands) in the 2θ range 5–50°.