Analyzing the pyrolysis mechanism of avermectin via experiments and density functional theory

ZHOU Hao ，LIU Su-xiang,* ，ZHAO Bao-feng ，WANG Jing-wei ，GUAN Hai-bin ，ZHU Di,* ，LI Huan ，SONG An-gang

（1. Key Laboratory for Biomass Gasification Technology of Shandong Province, Energy Research Institute, Qilu University of Technology (Shandong Academy of Sciences), Jinan 250014, China；2. Department of Chemical Engineering, Monash University, Clayton 3800, Australia）

Abstract: In this study, the thermal degradation mechanism of avermectin (AVM) was analyzed via experiments and density functional theory calculations (DFT). The experimental results of AVMD pyrolysis indicated that the removal rate of AVM residues reached peak value of 99.88% above 250 °C. The main product of AVM pyrolysis was alcohols. Based on the C-O bonds breaking, four potential degradation pathways were proposed. The findings of the calculations were in agreement with those of the experiments. These results provide theoretical and empirical guidance for the development of safe antibiotic disposal technology.

Key words: avermectin；avermectin mycelial dreg；density functional theory；pyrolysis；degradation mechanism

Avermectin (AVM), serving as a drug in the avermectin family, has been widely used in agriculture and animal husbandry. Approximately 8-10 t of wet AVMD is generated in the production of 1 t AVM[1],which has a high yield, high moisture content, and a small amount of antibiotic residues[2-4]. Without safe handling, the residual antibiotics can be transferred and accumulated into the environment[5-10], leading to the evolution and spread of bacterial resistance[7,11], and posing a risk to the environment[12-15]. Therefore, it is crucial to safely dispose of AVMD to mitigate environmental pollution and associated health hazards.

The treatment of antibiotic mycelial dreg is mainly processed via incineration, safe landfilling, and composting[16]. However, due to the high moisture content and high output of antimicrobial mycelial dreg,this treatment process is rather costly and always leads to resource waste and secondary contamination[15,17].Similarly, composting is a highly time-consuming and inefficient method of treating antibiotic mycelial dreg,associated with ecological risks. Chen et al.[18]studied the mixed composting of cephalosporin C fermentation residue and recorded temperatures above 55 °C for three consecutive days, indicating that the compost had reached the mature stage. After 110 d of maturity treatment, the degradation rate was only 49.1%. Lan[19]combined streptomyces AVM residue, sludge, and corn straw powder at a dry weight ratio of 4∶1∶1 and inoculated the mixture with a 5% organic fertilizer fermentation agent. After 40 d of fermentation, the AVM degradation rate could be increased to 75.36%.

Pyrolysis has proven to be an effective alternative method of treating antibiotic mycelial dreg and antibiotics[20]. It can effectively eliminate infections and organic contaminants while reducing the amount of bacterial residue. For example, Wang et al.[21]found that all penicillin residue can be removed at 600 °C for treating 30 min. Chen et al.[22]used a fixed bed to eliminate penicillin residue; after 60 min, all antibiotics were pyrolyzed at a temperature above 400 °C. The pyrolysis of bacterial residue can also concentrate and stabilize heavy metals and recover high-value products,such as gas, liquid, and biochar[23]. In contrast to other treatment methods, the similar physicochemical properties of antibiotic mycelial dregs allow them to pyrolysis independently of the antibiotic dreg type[24].Pyrolysis is thus a promising method of safely treating and utilizing resources obtained from antibiotic mycelial dreg[25].

The degradation pathways of most antibiotics and associated mechanisms are still unclear, as it takes so long to research a variety of antibiotics with complicated structures through trials[26]. Quantum chemistry calculation, based on density functional theory (DFT), can analyze reaction behaviors and identify reaction mechanism at the molecular level,which, hence, serving as a compensate for the limitations of experimental methods of studying the degradation process[27]. Dou et al.[28]studied the photocatalytic degradation of amoxicillin and cefotaxime on g-C3N4by DFT calculation and found that the degradation pathway included lipidation of the β-lactam ring and direct molecular fragmentation.However, the degradation of cefotaxime first occurred via de-esterification (3-acetoxy hydrolysis), followed by the decarboxylation. The oxidative degradation mechanism of sulfamethoxazole was studied by Song et al.[29]using DFT calculation. This reaction process involved the breaking of S-N and S-C bonds, nitration and hydroxylation of the benzene ring, carboxylation,and opening of the oxazole ring. Pelalak et al.[30]investigated the oxidative breakdown of sulfonamide with the aid of DFT calculations, and they identified 31 intermediate products and suggested the potential degradation pathways. Thus, the pyrolysis mechanisms and pathways of antibiotic mycelial dreg could be wellstudied through combined with experiments and DFT calculation.

AVMD is the solid waste after AVM production.Its main components are mycelium, unused culture medium and metabolites produced during fermentation and degradation of culture medium as well as a small amount of AVM. Nowadays, only a few studies on photolysis, hydrolysis, and strain decomposition were carried out, while that on the thermal degradation of AVM is nearly un-reported. Mushtaq et al. investigated the photolysis of AVM in three different solutions with pH 7 and obtained a variety of degradation intermediates[31]. Several other studies focused on the isolation and identification of AVM degradation products from treated crops[32-35]. Others analyzed the decomposition of AVM degradation products by different strains[33,36]. Therefore, this study mainly investigated the thermal degradation of AVM through combined experiments and DFT calculation to evaluate the effect of temperature on the elimination of AVM.

1 Materials and methods

1.1 Materials

AVMD used in this study was provided by a pharmaceutical company. It was first dried in an oven at 105 °C and then screened for 0.28 mm powder.Avermectin B1a(C48H72O14, Figure 1) with 98% purity was purchased and dried at 50 °C for 24 h in a vacuum drying oven for subsequent experiments. Primary properties of AVMD was listed in Table 1.

Table 1 Primary properties of AVMD

Figure 1 Structural formula of avermectin B1a

1.2 Experimental methods

1.2.1 Fixed-bed experiment

The AVMD and AVM pyrolysis experiments were conducted in a fixed-bed reaction system, as shown in Figure 2. First, 2 g of raw materials were placed in a quartz boat, which was hung from the top of a quartz tube with an iron wire to keep it outside of the heating area. The experimental device was connected and sealed. Then, the reaction furnace was heated under N2atmosphere (30 mL/min) at a rate of 10 °C/min to reach the pyrolysis temperature and hold it, and the top quartz boat was quickly lowered to the temperature measuring point in the center of the quartz tube. A cryostat (with a temperature maintained at 0 °C) was used to condense volatiles (H2O, CO, CO2,CH4, H2and tar) produced during the rapid pyrolysis of AVMD and AVM in the presence of N2. The reaction had a duration of 30 min. The condensable components(tar) were cooled into the collection bottle, and the noncondensable gases were collected in airbags. All pyrolysis experiments were conducted thrice to ensure the accuracy of experimental data.

Figure 2 Fixed-bed reaction system used for avermectin pyrolysis

1.2.2 Gas chromatography

The gas collected in the airbag during pyrolysis was analyzed using a Micro GC Fusion gas chromatograph (Infocom, Seewen, Switzerland).

1.2.3 High-performance liquid chromatography

The liquid (tar and H2O) and solid products of fixed-bed pyrolysis were weighed, and the AVM in this residue was analyzed using an e2695 Separations Module (Waters, Milford, MA, USA) and highperformance liquid chromatography (HPLC) column(4.6 mm× 150 mm, 5 μm; Agilent, Santa Clara, CA,USA). The column temperature was 35 °C, and the detection wavelength of all target compounds was 245 nm. The mobile phase was water-acetonitrile(30∶70,v/v), with the flow rate of 1.0 mL/min. The injection volume of the analytical solution was 20 μL,and the elution conditions were isometric.

whereYsolid,msolid,csolid,m,Yliquid,mliquid,cliquidandDAVMrepresent the residual amount of AVM in solid products, the mass of residual solid, the residual amount of AVM in residual solid, the mass of reactants, the residual amount of AVM in liquid products, the mass of residual liquid, and the residual amount of AVM in residual liquid after reaction,respectively

1.2.4 Gas chromatography and mass spectrometry

A 7890B Gas Chromatograph was coupled with a 7010 Series Triple-Quadrupole Mass Spectrometer(Agilent, Santa Clara, CA, USA) to analyze the changes in solid and liquid products of AVM pyrolysis at different temperatures. Solid products of AVM pyrolysis were placed in a mortar, ground into powder, extracted twice using dichloromethane, and then combined with the solution. This solution was then analyzed using GC and mass spectrometry(MS). Gas chromatography analysis was conducted in an HP-5 MS capillary column (30 m × 0.25 mm ×0.25 μm; Agilent, Santa Clara, CA, USA). A total of 1.0 μL of helium (99.9999%) was introduced as the carrier gas under a flow rate of 0.8 mL/min using splitless injection. The oven temperature of GC was maintained at 100 °C for 1 min, and then increased to 200 °C at a rate of 10 °C/min and finally to 280 °C at a rate of 40 °C/min. It was maintained at 280 °C for 15 min.

1.3 Calculations method

In this study, the CASTEP module of Materials Studio 2017 was used to conduct molecular modeling,molecular configuration optimization and transition state calculation through TS Search. The generalized gradient approximation method and PBE exchangecorrelation functional theory were used with the selfconsistent field (SCF) tolerance of 2.0 × 10-6eV/atom.The Monkhorst-Pack k point (1 × 1 × 1) was adopted with calculating 100 steps in surface optimization[37].

The optimized results in TS Search operation were subsequently treated by a search protocol of complete linear synchronous transformation/quadratic synchronous transformation (complete LST/QST) to obtain the corresponding reactant and product. These parameters included a 2.0 × 10-5eV/atom SCF tolerance, 9999 maximum convergence accuracy cycle steps, and a maximum QST of 5. In the process between completing the LST calculation and setting the convergence criterion, the minimum optimization of the conjugate gradient and the calculation of maximum QST were conducted cyclically until the calculation was completed[37].

2 Results and discussion

2.1 AVMD pyrolysis

2.1.1 AVMD pyrolysis product yields

Figure 3 depicts the solid, liquid, and gas product yields of AVMD pyrolysis at different temperatures.The conversion rate and product yield of AVMD pyrolysis were obviously influenced by temperature.Pyrolysis conversion rate increased from 1% to 37.1%with the increase of temperature from 150 to 350 °C.However, biochar production decreased from 96.8% to 60.9%, along with the elevation of gas and liquid product yields, where the liquid product yields are markedly raised from 1% to 31.2%. Therefore,elevation of temperature increases the AVMD pyrolysis degree that deepens the color of solid products.

Figure 3 Yield of residue products in avermectin pyrolysis at different temperatures

2.1.2 AVM in AVMD degradation analysis

Figure 4 shows the degradation rate of AVM in AVMD, which increased dramatically from only 11.31% at 150 °C to 95.68% at 200 °C. At temperature above 250 °C, the degradation rate only slightly elevated. This means that most AVM in solid product had been decomposed, with only a small amount of residual remaining in the liquid product. At 350 °C, all AVM was effectively removed and the maximum degradation rate reaches 99.88%.

Figure 4 Degradation rate of avermectin residue at different temperature

2.2 AVM pyrolysis

2.2.1 AVM degradation rate analysis

Figure 5 shows the relation of degradation rate of AVM with the temperature. The degradation rate of AVM increased sharply from only 8.10% at 150 °C to 97.11% at 200 °C, which is consistent with the result of degradation of AVM in AVMD. The degradation rate of AVM reached 99.995% at 250 °C and only a marginal amount of AVM remained in the liquid products. At this condition, AVM degradation rate was almost unchanged with the temperature, indicating that AVM had almost completely degraded.

Figure 5 Degradation rate of avermectin at different temperature

2.2.2 AVM pyrolysis product yields

In order to analyze the thermal degradation mechanism of AVM, the pyrolysis of AVM with 98%purity was then performed in the range of 150-350°C.The products distribution of AVM pyrolysis in this temperature range is shown in Figure 6. With the increase of temperature from 150 to 350 °C, the yield of solid products decreased from 96.99% to 22.40%,whereas the liquid and gas yields increased from 0 to 66.27% and 0.02% to 6.59%, respectively. The yield of gas, inducing H2, CH4, CO, CO2, C2H6and C3H6, was the smallest in AVM pyrolysis products. At 150 °C, the condensable volatile products and non-condensable products were undetectable due to the almost no cracking of AVM (with a melting point 150-155 °C).The liquid yields increased slightly at 200 °C, and raised dramatically at 250 °C and then gradually elevated with the temperature, thus, making the liquid yields much higher than the gas yields. On the contrary, solid yields continuously decreased with the temperature. Therefore, high temperature could promote the decomposition of AVM.

Figure 6 Yield of avermectin pyrolysis products at different temperatures

The main liquid products of AVM pyrolysis comprised the following six categories: alcohols,hydrocarbons, acids, aldehydes, esters and ketones.These liquid products of pyrolysis at 200-350 °C were then analyzed, as shown in Figure 7. Alcohols,hydrocarbons, and acids were the main liquid products of AVM pyrolysis, although the concentrations of alcohol and acid gradually decreased with the temperature. As the representative of alcohols and acids, 2,3-dimethyl-3-hexanol and sorbic acid accounted for 25.11%-37.59% and 11.23%-17.95% of total liquid products, respectively. In contrast, the yields hydrocarbons and aldehydes gradually raised with the temperature, where 2,5-dimethyl-2,3,4-hexatriene and furfural accounted for 13.03%-14.80%and 3.78%-6.85% of total liquid products, respectively.Esters' yield was unchanged, while that of ketones declined, with the increase of temperature.

Figure 7 Composition of liquid pyrolysis products at different temperatures

2.3 DFT calculations

The photocatalytic degradation of AVM has been reported in many literatures, giving a variety of mechanisms and pathways for the degradation progress[38-40]. According to previous reported results and the above experiments, four AVM degradation pathways with the same radicals of M2, M3, M5, M7 and M8 were proposed and studied by DFT calculation,as shown in Figure 8.

Figure 8 AVM decomposition pathways

2.3.1 Layout number calculation

Based on four degradation pathways, the chemical bonds of AVM B1amodel were analyzed as shown in Table 2. The bond strength of chemical bond can be evaluated by the Mulliken layout number. The smaller layout number, the longer the bond, and thus the easier it is to break[41]. Meanwhile, the smaller Mulliken orbital population of bonds means that the bonds are easily to break during a chemical reaction[42]. In the case of C-O/C-C bond cleavage, the Mulliken layout number can be used to reflect the difficulty of breaking the C-O or C-C bond. As shown in Table 2, the bond layout number and the energy barrier are both in the order of C4'- O＜C13-O＜C16-C17＜C8-C9, while the variation of bond length is opposite. The C-C and C-O bonds are easier to break, which is consistent with the results in literature[38-40].

Table 2 Decomposition pathways and associated Mulliken layout numbers(e)

2.3.2 Transition state analysis of AVM

The transition state calculation was based on the intermediate products determined using the calculated layout values. Calculation results of four pathway are described in detail in supplementary material Table (a),(b), (c), (d), respectively. The rate determining step of every pathway was listed in Table 3. The energy barrier of various rate determining step is increased in the order of TS-A4＜TS-C2＜TS-D5＜TS-B3. This indicated that Path-A may be the easiest, among the four AVM degradation pathways.

Table 3 Duel speed step reaction energy base under four paths

In Path-A, the C4'-O bond of AVM molecular is first broke and dissociated to radicals M1and M2. The structure of M2 is similar to that of glucose, while M1 retained most of the structural characteristics of AVM,being in line with the result of AVM photolysis study[33]. Then, M1 further dissociated to M3 and M4 with the similar structure of glucose and milbemycin,respectively. Thereafter, M4 decomposed into M5 and M6, along with the breakage of 16-membered ring,resulting in significant reduction of toxicity. The ester bond in M6 was then broken and decomposed into M7 and M8, which was the rate determining step of the whole degradation pathway with the highest energy barrier of 42.70 kcal/mol.

The structure of M2 and M3 radicals are similar to glucose. According to the study of Wang et al.[43], 1,2 enediol was the main intermediate compound in the process of glucose pyrolysis. This is consistent with previous experimental results, where the mass content of 1,2 enediol reached about 18% in alcohols products.M8 radical is like 1,3-cyclohexanedione derived cyclic ketal that decomposes to 1,3-cyclohexanedione reaching about 6%[44]. No further transition state calculation was carried out for M5 and M7 radicals, as they still have complex structure, lacking of information on cracking mechanism.

3 Conclusions

In this study, the thermal degradation mechanism of AVM was analyzed through combining the experiments with the calculations. Based on the pyrolysis experiment, the degradation rate of AVM in AVMD is higher than 99.88%, at the temperature above 250 °C. The main degradation products were identified as alcohols with the mass content of 1,2 enediol and 1,3 cyclohexanedionenol of 18% and respectively. Potential AVM degradation pathway (e.g.Path-A) was proposed with the aid of DFT calculation,which is in consistent with the experimental results.The insight shown in this work provides a theoretical and empirical guidance for a better understanding of the antibiotic degradation mechanisms.