Study on Molecular Structure Model and Reactivity of Spent Mushroom Substrate: Experiment and Simulation

The infrastructure features of the spent mushroom substrate (SMS) were detected and explored based on experimental data acquired from Fourier Transform Infrared Spectroscopy, Pyrolysis–Gas Chromatography/Mass Spectrometry as well as Thermogravimetric Analysis. Further, molecular simulation methods were combined with diverse testing technologies to build the 2D and associated 3D models from cellulose, hemicellulose, and lignin. Eventually, a rational 3D model was constructed for SMS using geometry optimization calculations and annealing dynamics simulations. According to the Mulliken population analysis and electron density analysis, the π-π conjugation effect of the aromatic ring structures was influenced by the oxygen atoms, thus, resulting in the electron accumulation on these oxygen atoms. At the same time, oxygen exhibited a greater electronic charge density and reactivity in the Furan ring and pyran ring structures in contrast to other atoms. Furthermore, the molecular orbitals for HOMO and LUMO were computed to check the chemical reaction characteristics of the SMS. According to the HOMO–LUMO energy gap, the structure of the polycyclic aromatic hydrocarbons provided reaction sites for the reaction and played a key role in chemical bond breaking.


Introduction
The spent mushroom substrate (SMS) refers to an agricultural and forestry waste that remains after the fungus Auricularia auricula is produced. The main raw materials contain cereal straws, cottonseed shells, and wood sawdust, mainly cellulose, hemicellulose, and lignin [1]. Accumulation of a large number of cellulose-rich substances makes the material recalcitrant. After the decomposition and enzymatic hydrolysis of the mycelium, the waste fungus rod contains several structural polysaccharides and lignins that can be utilized as biomass energy. Due to the complex structure of the SMS, early research mainly discussed the apparent pyrolysis characteristics, and reports on the pyrolysis intrinsic characteristics and microchemical reaction mechanism are scarce. The current study aimed to understand the SMS pyrolysis mechanism and the law of biomass in pyrolysis for developing the biomass pyrolysis technology and effectively using biomass energy.
The advancement and integrated utilization of multiple advanced technologies like the Fourier Transform Infrared (FTIR) spectroscopy, thermogravimetric analysis, and pyrolysis-gas chromatography/mass spectrometry (Py-GC/MS) over the last decades have enabled the availability of a large amount of information on the chemical structure of the SMS [2][3][4][5]. According to the experimental results, the statistical characteristics of the SMS were computed for identification and quantification.
For the complete characterization of SMS, FTIR analysis was performed according to Eliescu et al. [3]. After increasing the composting time, aliphatic and polysaccharidelike structures in the SMS decreased, while oxygenation, polycondensation, and polymerization processes increased. Li Feng et al. determined the molecular structure model of the Shengli lignite via TG-GC/MS for the analysis of the pyrolysis products structure and obtained contents and the structural parameters by peak-fitting of FTIR data [6]. Chefetz et al. characterized the chemical properties and transformations of organic matter (OM) by Py-GC/MS [7]. They described explicit characterization outcomes for SMS in functional group locations and carbon structural parameters. The understanding of the physicochemical properties of SMS was improved to a large extent by these experimental analyses.
SMS is a lignocellulosic material with the potential to generate fermentable sugars (like glucose and xylose). A water-soluble polysaccharide (PL) was purified from the SMS by Zhu et al. to develop and efficiently use SMS [8]. Besides, the monosaccharide composition and the total sugar content were measured using the phenol-sulfuric acid approach. Infrared spectroscopy was conducted to characterize the structure. The whole sugar content of crude polysaccharides in SMS approached 25.8%. Besides, the polysaccharides comprising PL1 and PL2 contained glucose, rhamnose, and mannose in the molar ratio of 1:3.13:1. 16. Recent studies on lignocellulose have adopted molecular modeling and simulation techniques for developing the model of cellulose, hemicellulose, and lignin molecules. The hierarchical structure of the SMS complicates the modelbuilding task. Generally, a representative unit like the plant cell is studied to understand the physicochemical structure and properties of the plant biomass [9,10]. The prime cells in the softwood are the tracheids, whereas those in the hardwood are fibers; this distinction simplifies the computational simulation problem. The gap of cellulose fibrils is filled with amorphous hemicellulose and lignin [11,12]. Recently, a model for cellulose-hemicellulose-lignin matrix has been raised using NMR data. In the proposed model, every cellulose microfibril comprised 18 cellulose chains that were categorized into two domains featuring discrepant hydroxymethyl conformations. Galactoglucomannan and xylan were attached to the surface of the cellulose microfibril. Hence, the microfibril is bound to the larger fiber of a lateral size of around 28 nm, while its outside is surrounded by lignin [13]. Generally, there is no direct connection between cellulose and lignin,they are linked by hemicellulose molecules [14,15]. The cellulose-hemicellulose-lignin matrix, as well as the associated Macro-molecular assembly, requires further study. Since force fields exist for both lignin and carbohydrates, MD simulations for lignocellulose can be of great use. However, the density-functional theory (DFT) methods cannot be applied. Based on experimental findings, cellulose fibrils are thought to be embedded into the mixed hemicellulose and lignin matrix. Yet, the exact interactions of these 1 3 components are unclear. For plant cell walls, considering the chemical similarity and interaction affinity between hemicellulose and cellulose, a model was built and adopted for MD modeling. This model featured hemicellulose in close proximity to the cellulose surface, as well as lignin molecules, filling up the remaining space [16,17]. Although the model chemical reactivity was not discussed in that particular study, lignocellulose biomass modeling was interpreted completely using a practical approach. Thus, the SMS carbon skeleton microstructure was better understood. Besides, only a few MD studies reported using simulated plant cell walls using cellulose, lignin, and hemicellulose because of the complexity of the material and the scarcity of information. Buehler et al. established an easy, coarse-grained model of the wood cell walls with cellulose fibrils and hemicellulose wherein only weak interactions between hemicellulose and cellulose were considered [18]. Later, they created a comprehensive model of the wood cell wall with cellulose, lignin, and hemicellulose [9]. Hemicellulose covalently connects cellulose and lignin that were not in direct contact. The model featured a layered structure of about 8 nm thick wherein cellulose sheets formed the bottom and top layers, and hemicellulose and lignin filled the gaps. Comparative research on such natural structures is highly beneficial, and corresponding comprehensive models should be established, computational modeling plays an important role here. Furthermore, their research provided details of internal wood cell wall microstructure as well as fundamental knowledge concerning its transport, elastic properties, and pore distribution. Based on Faulon et al., a model denoting the plant secondary cell walls was constructed [16]. A combined model integrating lignin, cellulose chains, and water molecules with xylans was simulated for exploring the impact of lignin containing different proportions of the basic monomers (S, H, and G units), chain lengths, as well as linkage types. Few studies have described the lignocellulose pyrolysis behavior by adopting the DFT method. A novel perspective of describing the physicochemical behavior of the SMS model was provided by lignocellulose biomass modeling. Reports investigating the lignocellulose biomass reactivity under the perspective of chemical reactions are few. Moreover, the electronic properties in the lignocellulose biomass model have been studied previously. These properties were significant for determining the cleavage or formation of chemical bonds and the pyrolysis reaction characteristics in the lignocellulose biomass molecule.
In the current study, a novel model was proposed by combining molecular simulation methods based on many testing techniques. Multiple molecular dynamics (MD) and quantum chemistry approaches were used to construct a reasonable 3D SMS model. Further, based on the density functional (DFT) approach, electron density, molecular orbitals, Mulliken population analyses through quantum chemistry, and MD simulations, the chemical properties possessed by the lignocellulose biomass were studied. Finally, the results obtained from the study offered an elaborated insight into SMS-related reaction mechanisms and reactivity. Thus, they explained the chemical behavior of the SMS molecules at the microscopic level and enriched the theoretical explorations for lignocellulose biomass molecule modeling.

Sample Preparation
SMS samples were collected from Huangsongdian town located in Jilin Province, China. The mushroom substrate comprised bean curd (6%), wheat bran (9%), sawdust (80%), and N and P fertilizers applied to satisfy the mushroom cultivation requirements. After cultivation for a few cycles, the SMS was gathered but discarded later because of insufficient nutrients. The SMS was preserved in a sealed bag at the factory before being delivered to the laboratory. It was pulverized after natural drying at room temperature, followed by ball milling to ensure a participle size below 0.45 mm. The SMS samples underwent oven drying treatment at 40 °C for 24 h and were preserved using a desiccator for experimental uses.
Proximate analyses on SMS based on air drying treatment were performed following the standard in China (GB/T 30728-2014). The content of carbon, hydrogen (GB/T 28734n2012), nitrogen (GB/T307284n2012, and sulfur (GB/T28732rn2012), carbon, S based on air drying treatment using the EuroVector elemental analyzer (EuroEA3000 CHNS-O Analyzer, EuroVector SpA), whereas oxygen was measured from the difference. The proximate/ultimate experiments were carried out thrice to check the reliability of the results with those obtained in earlier studies [19]. The SMS components were measured with the paradigm method for the three replicates. Table 1 presents the proximate and ultimate analyses and components of the SMS samples. The proportion of each element in the Macro-molecular structure was predicted, and the simplest molecular formula of the constructed average molecular structure model was determined. The elemental composition of cellulose, hemicellulose, and lignin was CHO. Assuming that the molecular weight of sample 1 of SMS was about 3000, the simplest molecular formula (C 900 H 1395 O 531 ) of the average molecular structure model constructed in this study was determined.

Fourier Transform-Infrared Analysis
The chemical structure of the SMS biomass is complex. The spatial configuration of the internal molecules, chemical bond force constant, and other parameters were obtained through solid-phase infrared detection. The internal structure of the SMS was investigated with the Spectrum Two FT-IR (Fourier transform-infrared) spectrometer (PerkinElmer). The main technical parameters were as follows: Precision: 0.1 cm −1 , Wavelength accuracy: 0.01 cm −1 , Wavelength range: 350-8300 cm −1 , and Spectral resolution: 0.5 cm −1 .

Pyrolysis-Gas Chromatography-Mass Spectrometry (Py-GC/ MS)
The SMS sample (5 mg, < 0.2 mm) was put into the Multi-Shot pyrolyzer (EGA/PY-3030D, Frontier Laboratories) with a double-shot sampler for performing pyrolysis. Next, the SMS pyrolysis products were directly applied for performing GC/MS analysis (Clarus SQ 8 GC/MS, Perkin Elmer) using the pyrolyzer with He as the carrier gas. The pyrolysis conditions were as follows: pyrolysis time: 10 s; pyrolyzer furnace temperature: 600 °C. The GC-MS acquisition parameters were: initial oven temperature of 50 °C for 5 min; ramp rate of 10 °C/m for up to 250 °C; holding time of 10 min; injection port temperature of 300 °C; transfer temperature of 280 °C; source temperature of 280 °C; scan range of 50-450 Da; and column dimensions of 30.0 m × 250 µm.

Computational Methods and Simulation Details
SMS refers to a material consisting of organic chemical compounds with uncertain chemical formulae or weights. Accordingly, the 3D molecular model for SMS required reasonable optimization and adjustment before the follow-up analysis of structural properties.
Firstly, the 3D molecular models for cellulose, hemicellulose, and lignin with the SMS model were performed using annealing dynamics calculations through an annealing process with the Forcite Plus module of the Materials Studio 2019 software (Dassault Systèmes BIOVIA). Besides, the details in the calculation are presented as: Condensed-phase Optimized Molecular Potentials for Atomistic Simulation Studies (COMPASSII) served as the force field for energy calculations [20]. Annealing cycles amounted to 25. The initial temperature at the beginning and the end of every annealing cycle was 298 K, the temperature achieved its maximum of 800 K until reaching the midpoint in every annealing cycle. There existed 100 temperature steps between the initial/final temperature and the mid-cycle temperature. To ensure sufficient time for minimization and getting rid of energy barriers, the total number of steps in the model was predetermined as 25,000; this number was sufficient for searching and calculating the lowest-energy conformation. Besides, the thermodynamic ensemble was set as NVT in the dynamics calculation. Moreover, the Nose algorithm was adopted as a thermostat for controlling the temperature to a time step of 1 fs. The general quality level of the annealing dynamics calculation was fine.
The lowest-energy conformation was retrieved following annealing. Subsequently, the total energy in the chosen conformation was continually minimized in the Forcite geometry optimization task under the Forcite Plus module. Details in the calculation were: COMPASSII and Smart served as the force field and algorithm, respectively. The general quality level of the geometry optimizations was fine.
Following geometry optimization, the energy and dynamics parameters in the optimized molecular models were computed in the Forcite dynamics task under the Forcite Plus module. Details of the calculation included: The NVT ensemble was chosen. Besides, the target temperature for calculating the dynamics was 298 K. The time step (1 fs) and the total simulated time (5 ps) were properly predetermined, maintaining the dynamics simulation system at a steady temperature during the ultimate conformation output. In addition, the dynamics calculations were of the desired quality.
The electron density, molecular orbitals, and population were calculated with the quantum mechanics method in the DFT geometry optimization task under the DFT module. Meanwhile, the details of the calculation included: The Smart algorithm was applied during optimization, and the corresponding convergence thresholds were predetermined. The convergence tolerance of energy and force were 0.02 kcal/mol and 0.1 kcal/mol/Å, separately. Self-consistent charges were chosen to improve accuracy. To ensure a sufficient number of calculation cycles that satisfied the convergence criteria, the maximum iterations were set as 100,000. The general quality level in the DFT calculation was fine. The associated parameters included the convergence threshold value of 2 × 10 -5 Hartree (Ha) for energy, 0.005 Ha/Å for force, 0.05 GPa for stress, and 0.006 Å for displacement.

Lignocellulosic Materials and Their Structures
Multiple experimental technologies were employed along with FTIR and Py-GC/MS analysis to explore the structural properties of SMS. Based on the above experimental results, naphthenic hydrocarbon and aromatic compounds were identified as the main components in the SMS, with determined structural units linked through short-chain alkanes. The chemical model of the SMS was constructed with the following procedures. The initial microstructures were formed using the Materials Studio software from BIO-VIA [21]. First, the 2D models of cellulose, hemicellulose, and lignin were built using the experimental characterization data. Cellulose is generally represented by the chemical formula (C 6 H 10 O 5 ) n [22]. In recent research, Zheng et al. built a cellulose chain polymer at the polymerization degree of 60 using MD pyrolysis simulation [23]. A linear chain of Macro-molecular polysaccharides connected through β-1,4-glycosidic bonds was dominant in the SMS and, thus, selected as the typical cellulose model. Cellulose chains are linear instead of straight because pyranose ring puckering is straight [24]. After determining the degree of polymerization of cellulose, the molecular weight of hemicellulose and lignin was determined according to the ratio measured previously (cellulose:hemicellulose:lignin = 5:3:2). Notably, galactose, furan glucose, and furan (determined based on the Py-GC-MS data of 3.2) were chosen for building the structure of aromatic clusters of the model. Xylan was selected to represent the hemicellulose structure. The hemicellulose structure consists of two segments, namely, O-acetyl-4-Omethylglucurono-xylan (the molar ratio of methylglucuronic acid, xylose, and acetyl structures was 1:8:4) and arabinose-4-O-methylglucuronate-xylan (the molar ratio of glucuronic acid, xylose, and arabinose was 1:6:1) [25,26]. The weight ratio of these two segments was approximately 1. During lignin modeling, the number of aromatic carbons, the aromatic carbon rate, methyl ratio, and the aromatic hydrogen rate in the model matched the corresponding values measured and calculated by FTIR. The lignin molecule contains three elementary structural sub-units, i.e., p-hydroxyphenyl, guaiacyl, and syringyl units that vary significantly in proportion in different SMS samples. The percentages and sequences of inter-unit linkages also vary in different samples. To lower the complexity and identify the chemical properties for lignin in SMS, the amorphous lignin molecule was generated by polymerizing three units linked through the β-O-4 and α-o-4 bonds in SMS [27]. The type H lignin fraction was barely detectable in the infrared spectrum of the SMS lignin because the molar ratio of guaiacyl, syringyl, and p-hydroxyphenyl lignin in the SMS was 14:11:1. Moreover, the three intermonomer linkages of the type β-o-4, β-5, β-β, 5-5, α-o-γ, α-o-4, and β-1 (31%, 3%, 8%, 4%, 4%, 34%, and 16% per 100 monomers) formed the aromatic heteropolymer with a three-dimensional network structure. The units and modeled structure of cellulose, hemicellulose, and lignin are presented in Fig. 1.
Cellulose, hemicellulose, and lignin were bound by hydrogen bonds, and covalent bonds existed between hemicellulose and lignin in addition to hydrogen bonds: ferulic acid and lignin monomer were connected by β-O-4 or ester bonds [28]. In this study, three O-acetyl-4-O-methyl glucuronic xylans and one arabinose 4-O-methyl glucuronic xylan were linked to the lignin Macro-molecular model by ester and ether bonds, respectively. The two connection methods and the modeled structure of hemicellulose-lignin are given in Fig. 2.
As can be seen in Fig. 2, panel a represents the free radical coupling of the lignin monomer and ferulic xylanan. The main xylose residue (grey) is replaced by an arabinose residue (pink), which is modified with ferulic acid ester (brown), ferulic acid and lignin monomers are linked by β-O-4. Panel b represents the rearomatization of the quinone methyl intermediate of lignin by hemicellulose nucleophilic water. The methyl intermediates of quinoone of two lignin monomers are linked by a β-O-4 bond and are rearomatized by the C6 hydroxyl group of the methyl glucuronic acid residue of hemicellulose. The methyl glucuronic acid residue (blue) ether is attached to the diene group. In addition, other sugar residues can also form bonds, with the curly arrows indicating the movement of electrons, and further polymerization of lignin can occur on the dotted line labeled as carbon. Panel c represents three O-acetyl-4-O-methylglucuronic xylans and one arabinos-4-O-methylglucuronic xylan that are related to the Macro-molecular model of lignin in the form of b and a, respectively. Furthermore, an optimal and reliable model was chosen from these models by performing dynamics computations (Fig. 3).

FTIR Analysis
In the current work, FTIR analysis was performed in the infrared band of 500 cm −1 -4000 cm −1 to quantify the distribution of the oxygen-containing functional groups in the SMS (Fig. 4). Four peaks were observed at binding energies of 1240, 1626, 2970, and 3350 cm −n , which corresponded to C-O-C/COOH, C=C/C=O, C-H, and C-O-H structures, respectively. The findings acknowledged the hydroxyl, carbonyl, and carboxyl groups as the major oxygen-containing functional groups of the SMS. Functional groups in complex organic Macro-molecular mixtures are rich in composition and complexity. Moreover, the absorption bands of characteristic spectra overlap partially or completely. Hence, accurate analysis and estimation of the relative contents of various functional groups from the infrared spectrum are impossible. The Gaussian peak fitting on the superpositioned spectrum peaks (the number and position of the interval absorption peaks were obtained by the second derivative of the FTIR data in the interval) and curve simulations were performed using the Origin software (Fig. 5). Peak Fit v4.12 was used for the peak-fitting of the SMS FTIR data. The infrared spectrum comprised four segments. The proportion of different benzene substitutions in the model was determined from the aromatic structure between 700 and 900 cm −1 with a correlation coefficient of R 2 = 0.9887; the types and proportions of oxygen-containing functional groups and hydrocarbon structures were determined between 1000 and 1800 cm −1 ; the fatty acyl functional group structure was derived from 2700 to 3000 cm −1 ; hydroxyl functional group structure was determined from 3000 to 3600 cm −1 . The overlapping peaks were separated effectively to accomplish the individual absorption peak area, and then the relative contents of the corresponding functional groups were characterized.
Some structural parameters like the aromatic hydrogen ratio of the hydrogen carbon atom to methyl ratio and aromatic carbon ratio were obtained. (1) The aromatic hydrogen ratio refers to the ratio of hydrogen atoms from aromatic compounds to total hydrogen atoms in the complex organic Macro-molecular structure. The absorption area in the 700-900 cm −1 region in the infrared spectrum represents the aromatic hydrogen content H ar . The absorption area within the region of 2800-3000 cm −1 represents the adipose hydrogen content H al ; Here, A ar represents the area of aromatic hydrogen absorption peak in the region of 700-900 cm −1 of the infrared spectrum, A al represents the area of adipose hydrogen absorption peak in the region of 2800-3000 cm −1 in the infrared spectrum; (1) H ar H = A ar A ar + A al (2) The absorption area in the region of 2800 to 3000 cm −1 represents the hydrogen content in fat; H al H ar = Aromatic hydrogen rate; (3) The hydrogen/carbon mole ratio (H/C) is the mole ratio of the hydrogen atom to the carbon atom in the complex organic Macro-molecular structure Here, H m is the mass of the hydrogen atoms; C m is the mass of the carbon atoms; H∕C = 1.55 (Table 1). (4) Methyl methylene ratio is a parameter derived from the ratio of carbon atoms of methyl to methylene Here, A CH3 represents the area of the methyl absorption peak in the infrared spectrum; ACH 2 represents the area of the methylene absorption peak in the infrared spec- Aromatic carbon ratio is the ratio of all aromatic carbon atoms to total carbon Here, C al /C represents the ratio of aliphatic carbon to total carbon; H al /C al stands for hydrogen/carbon mole ratio in the aliphatic group.C al = (12∕x) ⋅ H al , x is the stoichiometric description of fatty substances, and its value is 1.8 ( Table 2).
The infrared Gaussian peak fitting curve describing the peak classification of the aromatic structure indicated that the SMS contained three substituted modes of the benzene ring; the main benzene ring was tetrasubstituted, followed by the trisubstituted benzene ring, and the disubstituted benzene ring, as shown as Table 3. In the infrared band of 1000 cm −1 ~ 1800 cm −1 , the main corresponding oxygencontaining functional groups were the carboxyl, hydroxyl, carbonyl groups and the ether oxygen bond. Further, the C=C vibration of the aromatic hydrocarbons in this spectral peak was maximum at 18.18%, indicating that the benzene ring structure constituted the main part of the Auricularia furl. The stretching vibrations of C-O included the alkyl ether C-O vibration in the aryl ether and phenolic hydroxyl group, wherein the aryl ether content formed up to 11.20% of the fatty acyl spectrum peak. The CH 2 stretching vibration accounted for 89.24% and was the main existence mode of fatty acyl functional groups. A hydroxyl functional group is an important group affecting the reactivity of the SMS. The fitting curve of SMS indicated the presence of a greater number of self-associating hydroxyl hydrogen bonds and cyclic hydrogen bonds but a limited number of hydroxyl π hydrogen bonds.

Py-GC/MS Analysis
The total ion chromatogram (TIC) for the pyrolyzed SMS and the corresponding pyrolysis products are described separately in Fig. 6 and Table 4. The primary volatile products observed in the current study included aromatics, saccharides, phenols, acids, carbonyl compounds, alkanes, and alkenes. The primary pyrolytic products included acids, saccharides, and aromatics. The distribution of alkanes and (4) alkenes was sophisticated, and their carbon atoms ranged between 4 and 32, suggesting that the SMS sample mainly comprised short-chain alkanes. The distribution analysis on pyrolysis products showed that straight or branched chains could be used as the main linkers of aromatic units.
Benzene and relevant derivatives (ARHs) were obtained by decomposing lignin and (hemi-) cellulose. The primary bio-oil compounds of lignin pyrolysis were phenols subject to higher temperatures [29]. The phenolic compounds of SMS observed in the current study were phenol, 2-methylphenol, and pyrocatechol, revealing the contribution of these aromatic rings to SMS chemical composition. Therefore, benzene, heterocyclic compounds, and cycloalkanes were chosen to build the ring units for the SMS model. Py-GC-MS provided more rational details concerning the chemical structure of the SMS in comparison with the previous analysis methods, particularly in the measurement of hydrogenation for the aromatic ring, alkanes or alkenes, and oxygen-containing compounds [4]. Thus, according to the results derived from Py-GC/MS, the specific structural fragments and connections among the fragments were determined to generate a real SMS model.   structures were organized into cubes in a simulation of cells.

Establishment of Molecular Models of the SMS
The initial system density was set with a low value of 0.5 g/ cm 3 to avoid the overlapping of atoms. Later, the energy in these initial systems was minimized with the Forcite Tools program of the Materials Studio software. To gain a similar density, these systems were optimized to reach 200 ps at 0.1 MPa and 298 K by the NPT ensemble that maintained constant pressure and temperature. The annealing dynamics process was adopted for the search for low-energy structures, since the model system temperature increased and decreased periodically. Thus, to avoid being trapped in the structure in case of minimum local energy, the 3D conformations were fully relaxed with global energy minimization. In addition, the lowest-energy conformation formed in the annealing dynamics process was chosen. The second geometry optimization was performed to ensure the accuracy of the model structure. Dynamics simulation details: global calculation efficiency and accuracy are arranged fine, electrostatic and Van der Waals forces strive for interaction and methods for Ewald, Ewald precision = 418,585 kJ/mol (1.0 × 10 5 kcal/ mol), Simulated temperature = 300 K, ensemble is NVT, Time step = 1 fs, Total simulation time = 30 ps; Step number of dynamics calculation = 30,000; The temperature control algorithm is Nosé. The molecular structure of SMS was optimized through molecular dynamics simulation so as to minimize the energy. As Fig. 7 shown, the energy change undergoes about three stages: significant descent at 0-3 ps; slow decrease at 3-17 ps; and keeping nearly unchanged at 17-30 ps. So 30 ps of total simulated arranged is proved successful to obtain the molecular structure model with minimized energy.
Eventually, the 3D conformation was formed as the ultimate SMS model. Figure 8 exhibits the 3D molecular conformations for the SMS following annealing and geometry optimization. According to Sects. 3.1 and 3.2, various functional groups such as the furan ring, pyranoid ring, aldehyde group, aromatic group, alcohol, hydroxyphenol, hydroxyl carboxyl group were reflected in the Macro-molecular structure model. Besides, the chemical properties of a substance are determined by the functional groups in the molecular system so the Macro-molecular structure model of SMS exhibited the chemical properties of SMS. Table 5 summarizes the partial structural parameters and experimental data on SMS. The value of Root Mean Square Error (RMSE) equals 0.151, which indicates that the difference between the two values reaches small enough to testify the reasonability for optimized configuration of the SMS.

Model Verification
Thermogravimetric (TG) analysis is essential for an improved understanding of the pyrolysis behavior and products as a function of time and temperature under a controlled atmosphere [5]. Yang et al. suggested that hemicellulose, cellulose, and lignin degradation occurred under varying temperatures and released different gas products in pyrolysis because of the differences in the functional groups and chemical structures [30]. The rationality of the SMS model was verified by proving the importance of the mixed sample of cellulose, hemicellulose, lignin in the SMS pyrolysis. The content ratio for the three components in the SMS was cellulose: hemicellulose: lignin = 5:3:2, and the model was constructed according to this ratio. To check the consistency in the general trend of pyrolysis of the SMS and the model, the samples mixed in the ratio of 5:3:2 were used as the replacement samples of the model. Avicel PH Microcrystalline cellulose (FMC BioPolymer), xylan from beechwood (Sigma-Aldrich), and Lignin (dealkaline) (Sigma-Aldrich) were selected as the commercial samples. Xylan was selected on behalf of hemicellulose. Before the experiment, these  Fig. 6 Distribution of pyrolysis products of SMS (pyrolysis temperature: 600 °C) 1 3 three biomass components were dried at 105 ℃ for a day to remove the free water and reduce the influence of other factors on the experiment. The process of pressing and mixing was undertaken. A very small amount (about 0.5 mg) of the sample was required for the experiment. Hence, the samples (2 g) were weighed on a balance according to the ratio (5:3:2) and pressed by the tablet technology. The operation of grinding-pressing-re-grinding-re-pressing was repeated at least thrice for each sample to achieve uniform mixing of the components. The final mixed sample was utilized for loading and pyrolysis experiments. A thermogravimetric analyzer (Mettler-Toledo TGA/ DSC 1 thermobalance) was adopted to investigate the thermogravimetric characteristics of SMS samples and the evolution of the gaseous product from the cellulose-hemicellulose-lignin mixture. The two samples were heated at 50-600 °C with a rate of 10, 20, 30, and 40 °C/min in an open Al 2 O 3 crucible under a nitrogen gas flow; the carrier gas was nitrogen with over 99.99% purity. To keep an inert atmosphere during decomposition, the flow rate of nitrogen was maintained at 100 mL/min. Sample masses were limited to 5 mg to prevent any potential temperature gradient and guarantee kinetic control in the process.
The thermal stability of the three components varied because of the differences in their chemical structures. The hemicellulose initially was pyrolyzed at 197-257 °C, cellulose at 237-357 °C, and lignin at 277-497 °C [31].
As can be seen in Fig. 9, at four pyrolysis rates, the TG curves of the auricularia bran and model samples showed the same thermogravimetric trend, with a crossover point at 239 °C. When the pyrolysis temperature was 239-346 °C, the mass loss of the mixed sample was faster than that of the fungus bran. This was because the experimental samples with three structural components used in this paper were not directly extracted and separated from the biomass source of fungus bran, but were commercially obtained. Second, although the samples were evenly mixed, there was no covalent bond between the three structural components, so, compared to the fungus bran, the thermal decomposition reaction was more likely to occur. In the pyrolysis process of biomass, the existence of covalent bonds between cellulose, hemicellulose, and lignin will lead to corresponding interactions. Therefore, the molecular model of the crosslinking of the three structural components is studied in the following simulation work. It can be seen from the pyrolysis curve that the pyrolysis trends of the two are roughly the same except for the above areas. In conclusion, the model samples can reflect the pyrolysis trend of the auricularia bran, and the model established in this paper takes into account the connective bond between the three structural components mentioned above, so the model is reasonable.

The DFT (Density-Functional Theory) Method
The three-dimensional Macro-molecular structure model of SMS was established according to the structural characteristics described in Sect. 2.1. The DFT and MD techniques have their respective strengths and weaknesses; the drawbacks can be partially overcome by narrowing down the simulation. By utilizing the strengths of the two methods, information can be transferred across varied scales. Thus, these methods exhibit a great potential for exploration [32]. DFT is applied to systems with limited molecules because of its ability to cope with electrons. Representative systems, including as many as ten glucopyranose units or four aromatic units, were examined for cellulose and lignin [33],34,   chains in width, and 4 layers in thickness, with methyl group ends) with varying C6 hydroxymethyl group confirmations in energy optimization as well as 13 C NMR signal prediction [36]. Because MD methods have low requirements for calculation accuracy, DFT simulations provide the ideal option that balances accuracy and computational efficiency [37]. Therefore, population analysis, electron density, and molecular orbitals were computed using DFT to evaluate the reactivity of the SMS model. However, in earlier studies, DFT was merely applied to small-scale systems (scores of atoms). To reduce the time and cost of calculation, pyrolytic properties of the cellulose-hemicellulose-lignin model compounds were studied to substitute the SMS structures (Fig. 10). It is a molecular system of cellulose (

Electron Density Analysis
The total electronic charge density in the SMS molecule was calculated. In the meantime, the isosurfaces for electronic charge density were produced on a succession of isovalues. Slices of structures were made to describe electron density distribution in the smallest structural unit of cellulose, hemicellulose, and lignin from the SMS (Fig. 11). The oxygen heteroatom in the heterocyclic structure showed a stronger electronic charge density. Thus, demonstrating the stronger reactivity of these atoms compared with that of the other atoms in the heteroaromatic structures that determined the chemical changes in the SMS as reaction sites. Besides, the π-π interaction between the atoms in the aromatic structures was greatly influenced by heteroatoms (oxygen), and such interactions caused the electrons to accumulate on the heteroatoms. According to Fig. 10b Fig. 9 The pyrolysis behavior of synthetic biomass samples containing three components and SMS 1 3 ring in cellulose and Furan ring of hemicellulose structure were substituted by the oxygen atom (b and c). This resulted in an uneven electron density distribution on the aromatic ring associated with the oxygen atom. For the β-O-4 chain carbon skeleton structure of lignin, the electron density distribution mainly concentrated on the carbon atom, and the distribution was uneven, as shown in (d).

Molecular Orbital Analysis
The highest occupied molecular orbital (HOMO) and the lowest unoccupied molecular orbital (LUMO) contribute greatly to comprehending the molecular reactivity of chemical reactions. Under the frontier molecular orbital theory, chemical reactions occur due to the maximum overlap matching between the HOMO in a molecule and the LUMO in another molecule. The HOMO-LUMO gap is measured for assessing molecule chemical reactivity, particularly for the π-conjugated system [38].

Bond Length and Mulliken Population Analysis
Regarding molecular configuration, bond length, in association with the bond strength and bond dissociation energy, can be used as a key parameter to estimate the stability and reactivity of the SMS. The bonds turn shorter and stronger as more electrons are involved in bond formation. Simultaneously, Mulliken charges under the SMS model were computed and allocated to associated atoms to analyze the chemical reactivity of the SMS. Table 5    contrast, those of alkanes and aromatic structures carried a negative charge. The terminal methyl and secondary carbon C atoms in the SMS molecular structure carried a negative charge as electron donors in chemical reactions. During SMS pyrolysis, the O atoms were bound to oxygen due to the weak C-O bonds. Therefore, carbon dioxide was produced and considered a pyrolysis product of Py-GC/MS. The corresponding calculation results are given in Table 6. On the basis of the different chemical bond lengths and Mulliken population analysis between the monomers of the three structural components of the mold, the breaking sequence of each chemical bond in the molecular structure of auricularia bran can be obtained:

Pyrolysis Prediction of SMS Molecular Structure Models
Because stable compounds or chemical bonds form at all times during the pyrolysis process, the chemical bonds in the SMS molecular structure model cannot be completely and completely broken. As a result of the bond length calculation and the above SMS molecular structure analysis, the chemical bond fracture mechanisms in different chemical environments in the molecular structure model can be reasonably predicted and classified. SMS macromolecules are first decomposed into different structural fragments based on the strength of the bonds during pyrolysis. Chemical bonds with long bond lengths are more likely to preferentially break as reaction sites in the SMS molecular model. The breaking sites in the SMS molecular model can thus be predicted. During the pyrolysis process, the auricularia furfural modeler molecular model decomposes into three component monomers based on the strength of the bonds. As the reactive sites break, the benzene ring of lignin and the pyran ring of ester cellulose, hydrogen atoms are added to the broken molecular segment structure to saturate the fragment products. Figure 13 depicts the SMS molecular structure's thermal decomposition mechanism model. Although this mechanism model does not include the secondary reaction stages of SMS pyrolysis (such as polymerization, condensation, and aromatization), the fracture properties of some weak chemical bonds play an important role in predicting the evolution process of SMS thermal decomposition. According to the simplified mechanism model proposed in this paper, the hydroxyl functional group (-OH) is easily broken and then reacts with the hydrogen radical to form H 2 O. A large number of short chain carbon compounds are formed during the carbon skeleton decomposition process, which is accompanied by the formation of CH 4 products. Furthermore, as pyrolysis gas products, carbon-containing compounds and alkane products such as carbon monoxide (CO), ethane (C 2 H 6 ), and propane (C 3 H 8 ) are released. During thermal decomposition, aliphatic and aromatic structures will be further cracked into biomass oil components and converted into semi coke products as the pyrolysis reaction progresses [39].

Conclusions
In conclusion, the chemical structural parameters of the SMS were examined using FTIR, Py-GC/MS, and TG. Additionally, a new lignocellulose modeling method was developed by integrating molecular simulation methods and diverse testing technologies. Two-dimensional models and associated 3D models were established at varying ratios (5:3:2) of cellulose, hemicellulose, and lignin. A minimum-energy 3D SMS model was rationally chosen after complicated annealing dynamics and geometry optimization calculations with molecular dynamics and quantum chemistry methods. Besides, the DFT method was adopted in atomistic modeling and chemical reactivity assessment to better understand SMS reactivity. In exact simulations, the DFT method fulfilled the requirements for the dissociation or formation of chemical bonds and electronic properties of the SMS model. The TG curvilinear trend in the SMS and the mixture containing the three components indicated the reasonability of the model. For small-scale simulations, the DFT method satisfied the requirement for the dissociation or formation of chemical bonds and electronic properties for the SMS model. Three components (cellulose, hemicellulose, and lignin) were selected as simulation objects to accurately calculate the properties of the model and reduce the amount of calculation. The electron density analysis indicated that the π-π conjugation effect of the aromatic ring structures was subject to heteroatoms, and it caused the electrons to accumulate on the heteroatoms. Thus, the electronic charge density and Fig. 13 Molecular structure pyrolysis mechanism model of SMS reactivity of oxygen were higher in heterocyclic structures than those observed in the other atoms. These properties inversely determined the chemical changes in the SMS as reaction sites. According to HOMO-LUMO energy gap (Egap = − 3.387 eV), the higher reactivity of SMS. The Mulliken population analysis of some C and O atoms in the model was analyzed by quantum chemistry. The electron density distributions of aromatic ring units, pyran ring in cellulose, furan ring in hemicellulose and β-O-4 carbon chain structure of lignin were obtained by calculating the electron density characteristics of the molecular structure model of furan compound in SMS. Finally, a model for predicting the pyrolysis fracture mechanism of SMS compounds was proposed. The quantum chemistry calculation results of the current study were satisfactory, and the presented molecular structure supported further development and extension in a wide scope. However, a larger model combining the SMS with extracts (such as starch and protein) should be developed to measure the kinetic properties of the SMS for more reasonable results. Though the proposed SMS molecular model did not involve temperature and pressure, the methods provided basic knowledge concerning electron transfer for future research and other biomass structure-related domains, including the chemical pathways for products produced from SMS pyrolysis/combustion as thermal conversion reactions in lignin pyrolysis. Further studies should be performed to simulate the thermal cracking and thermodynamic properties of the SMS in quantum chemistry and dynamics.