Flame Retardancy and Mechanical Properties of Chitosan and DOPO with Different Mass Ratios in Epoxy Resin


 This work focused on the effects of chitosan (CS) and 9,10-dihydro-9-oxo-10-phosphaphenanthrene-10-oxide (DOPO) on the flammable propertied of epoxy resin matrix. The EP composites were fabricated by direct mixing method through a general curing method. The influence of CS, DOPO and CS / DOPO on the resin was investigated through cone calorimetry tests (CC), UL-94 vertical burning, limiting oxygen index (LOI), thermal gravimetric analyzer (TGA), differential scanning calorimeter (DSC) and thermogravimetric analyzer-Fourier infrared combined system (TG-FTIR). The char residues of modified EPs after CC tests were investigated by FTIR, EDX and XPS. Under the 10% addition of CS / DOPO in EP, with the mass ratio of CS and DOPO of 1 : 1, 1 : 2, 1 : 3, 2 : 1 and 3 : 1, the flame retardancy properties of EPs all increased, but only if EP/10% CS1/DOPO2 and EP/10% CS2/DOPO1 achieved a V-0 rating and their values of LOI were 33.7% and 32.5%, respectively. Compared with EP, the peak heat release rate, peak smoke produce rate and total heat release of EP/10% CS1/DOPO2 and EP/10% CS2/DOPO1 decreased, especially, total smoke release decreased by 61.9% and 71.0%, the char residuals amount increased by 84.3% and 41.6%, and the average CO2 yield decreased by 55.4% and 55.0%, respectively. It is worth nothing that the mechanical properties increased, especially the flexural strength increased by 36.0% and 38.4%, respectively. The results indicated that DOPO and CS had important synergistic effects for simultaneous increase both the flame retardancy and mechanical properties of EP composites.


Introduction
Epoxy resin is a kind of polymer material that is widely used and has obvious merits and demerits Kandola et al. 2010;Ai et al. 2020). Applications in more elds were limited due to its signi cant ammable defects (Wang et al. 2006). At present, the re resistancy of epoxy resins has been extensively studied and the direct addition of ame retardants is considered to be the most direct, e cient and economical usage mode (Tang et al. 2019;Netkueakul et al. 2020). However, the addition of some ame retardants will generally reduce the mechanical properties of epoxy resin (Zou et al. 2020).
The DOPO derivatives were widely used as phosphorous ame retardants in EP. Xie et al. synthesized a ame retardant named MBF-DOPO, when 4% MBF-DOPO was added, the samples obtained the V-0 rating in UL-94 test, and the limiting oxygen index (LOI) reached to 32.9%. Compared to EP, the peak heat release rate and total smoke production values were decreased by 29.3% and 33.6%, respectively (Xie et al. 2020). A ame retardant TAD was synthesized by Tang et al., when the mass fraction of TAD reached 12 wt.%, the samples acquired the highest LOI value of 33.5% and passed the UL94 V-0 rating (Tang et al. 2016). However, the addition of all above mentioned ame retardants in EP resulted in the decrease of the mechanical properties of EP composites.
Intumescent ame retardant (IFR) is a type of ame retardant composites, mainly composed of nitrogen and phosphorus, which is foamed and expanded when heated. It is halogen-free and does not use antimony oxide as a synergist (Lai et al. 2011). Hereby, it is a kind of environmentally ame retardant with high e ciency and low toxicity Li et al. 2009). The ame retardancy of IFR mainly depends on the density, quality and strength of the carbon layer, and the matching between the components is essential Huang et al. 2020). Previously, PER (pentaerythritol)-based carbon source had been paid much attention by researchers, which was basically petroleum derivatives (Xu et al. 2018). The main mechanism was that the hydroxyl group in its molecular structure could crosslink with the inorganic acid produced by the decomposition of acid source during combustion (Xie 2013).
Chitosan is a kind of natural biomass with abundant content in nature, and the molecular structure contains both hydroxyl and amine groups (Kim et al. 2017). It is a kind of alkaline polysaccharide with rich carbon source (Hoqani et al. 2020). Not only that, but non ammable gases can be supplied during combustion. However, when single chitosan was used as ame retardant, its poor thermal stability and low charring e ciency at high temperature limited its application in the eld where ame retardancy is required. In order to increase the carbonization e ciency of chitosan, the main research direction was to carry out chemical modi cation, using the active amine and hydroxyl carried by chitosan itself, so as to increase the thermal stability of chitosan and improve the carbonization e ciency Pan et al. 2015;Goda et al. 2021).
In order to improve e ciency and reduce energy consumption and cost, in this paper, CS and DOPO were directly compounded in a certain mass ratio, and the halogen-free ame retardancy EP composites were prepared by melt blending. The in uence of the addition of CS / DOPO ame retardant system on the ame retardancy of epoxy resin was analyzed, and on the basis, the in uence mechanism of different mass ratios of CS and DOPO on the ame retardancy properties of epoxy resin composites was also investigated, which provided basic experimental data and theoretical support for the wide application of chitosan in polymers. When the addition amount of CS / DOPO was 10% and the mass ratio of CS to DOPO was 1:2 and 2:1, the LOI value of EP/10% CS1/DOPO2 and EP/10% CS2/DOPO2 increased to 33.7% and 32.5%, respectively, and passed UL-94 V-0 rating both.
Meanwhile, the thermal properties of modi ed EPs were tested by thermogravimetric analyzer (TGA) and differential scanning calorimeter (DSC). The ame retardancy properties of modi ed EPs were analyzed by LOI (limiting oxygen index), UL-94 vertical burning test and cone calorimeter (CC). The morphology and composition of the char residual layer of modi ed thermosetting polymer were studied by SEM (Scanning Electron Microscope), Raman spectra and EDX (Energy Dispersive Spectxrmleter), and the ame retardancy mechanism was analyzed.

Preparation of modi ed EPs
In order to investigate the ame retardancy of CS / DOPO in EP, CS and DOPO were added to EP at different total additions and mass ratios. The preparation of modi ed EPs can be implemented according to Fig. 1 as followings: Firstly, CS, DOPO and DGEBA were mixed uniformly in a two-necked ask at 100 ℃ under vacuum for about 30 min, and stirred vigorously to obtain a homogeneous liquid. Then, the temperature dropped to 95°C, DDM was added, and stirred under vacuum condition. Finally, the curing process was carried out on the plate vulcanization machine. The mixture was poured into the prepared mold and put into the plate vulcanization machine. The whole curing process included two stages: curing at 100°C for two hours and at 150°C for another two hours. The various modi ed EP samples labeled on the basis of their compositions, were listed in Table 1. For example, the label EP/10% CS1/DOPO2 indicated that the total amount of CS and DOPO in the samples was 10%, and the mass ratio of CS and DOPO was 1 : 2. Scanning Electron Microscopy (SEM) and Energy Dispersive X-ray spectroscopy (EDX) were used to determine the sample morphologies and elemental distribution throughout the sample granules, respectively. The JEOL model JSM-5900LV SEM was applied to determine morphologies of char residuals. EDX was performed using Link Isis series 300 program to determine various elements distribution as a density mapping of the speci ed element.
FTIR spectra of the char residues for EPs were recorded on a Thermo Nicolet 5700 FTIR spectrophotometer using the KBr disk in a wavenumber range of 4000 -400 cm −1 .
X-ray photoelectron spectra (ESCALAB 250xi, US) of the char residues after cone calorimeter tests were recorded on using AlKα radiation (1486.6 eV) in ultrahigh vacuum conditions.
Raman spectroscopy (LabRAMHR800) was conducted to characterize the types of the carbon under 514 nm helium-neon laser line with wavelength range of 200-4000 cm −1 .
TG-FTIR was used to analyze the FTIR spectra of the gas components from the TGA, and the real time FTIR analysis of the pyrolysis gases was conducted on a thermogravimeter (TG219 F3, Netzsch Instruments, Germany) and a Fourier transforminfrared spectrometer (Nicolet IS50, Thermo Fisher Scienti c, USA.), the transfer line of gases from TGA to FTIR was heated at 280 ℃. The investigation was carried out at temperatures ranging from 40 ℃ to 800 ℃.

Thermal stability
The glass transition temperature (T g ) of modi ed EPs was characterized by DSC tests. In Fig. 2, EP exhibited a high T g value of 154.7 o C. With CS / DOPO supplementation, the T g values of EP/ CS/DOPO samples were all decreased. When the mass ratios of CS to DOPO were 1 : 1, 1 : 2 and 1 : 3, respectively, equivalent to the gradual increase of DOPO content, the T g of EPs gradually decreased. Two aspects for the decrease of T g values were concluded: 1) stereo-hindrance of DOPO structure decreased the crosslinking density of EPs (Liu et al. 2014); 2) CS with active N-H and -OH bonds could act as chain extender and reduce the crosslinking density of EPs. Meanwhile, when the mass ratio of CS / DOPO was 1: 1, 2 : 1 and 3 : 1, the T g of EPs were 132.8 o C, 146.8 o C and 143.7 o C, respectively. It can be seen that the T g of EPs increased with the addition of CS. It is particularly noteworthy that compared with the other two samples, when the mass ratio of CS to DOPO was 2 : 1, the T g value of the sample was the highest in CS / DOPO modi ed EPs. It was conjectured that with the addition of CS and DOPO, the crosslinking density of EPs decreased. However, when the mass ratio of CS / DOPO was 2 : 1, the ame retardant composite system formed a relatively dense cross-linking structure with epoxy resin. This is consistent with the mechanical properties of the materials.
TGA tests were further carried out to study the thermal stability and decomposition behaviors of EP and CS / DOPO modi ed EP samples. Fig.3 showed the TGA curve of EP, EP / CS, EP / DOPO, and EP / CS / DOPO samples, and the corresponding results were listed in Table 2. It can be seen from Fig. 3 that under N 2 atmosphere, the initial decomposition temperature (T 5 % ) and the maximum weight loss temperature (T max % ) of EP were 366.6 o C and 383.8 o C, respectively. The T 5 % and T max % of EP/10% DOPO and EP/10% CS samples were 285.2 o C, 385.6 o C and 366.3 o C, 385.3 o C, respectively. The T 5% of all modi ed EPs is higher than the curing temperature of EP ranged from 100 -150 o C, indicating that DOPO, CS or CS / DOPO can meet the solidi cation condition. The T 5% of the other samples decreased to varying degrees, but the T max % of the other samples did not change much. It can be inferred that the decrease of initial decomposition temperature T g of modi ed materials was attributed to the addition of DOPO. It can be seen from Fig. 3 that, unlike EP, the pyrolysis process of modi ed materials had two rapid weight loss stages, 240 ± 10 o C and 285 ± 5 o C, respectively. The main reasons were as follows: on one hand, the thermal decomposition temperature of CS and DOPO is lower than that of EP, it can be assigned to the preferential decompose of CS and DOPO that did not react with EP over EP/CS/DOPO materials during the heating process. On the other hand, with the loading of CS and DOPO, the epoxy groups of the system were correspondingly reduced, thereby reducing the crosslinking density of EP more or less. Except EP/10% DOPO, the char residues at 750 °C were all higher than that of EP, indicating that these EPs had superior high temperature stability and better carbonization capacity than EP. UL-94 is measured by the length of combustion times and burning dripping at self-extinguishing (Wang et al. 2009). Limiting oxygen index (LOI) refers to the minimum oxygen concentration (volume percentage) that supports the combustion of materials. The test results of ame retardancy of EPs studied by LOI and UL-94 vertical combustion tests were summarized in Table 3. As can be seen from Table 3, all the modi ed EPs did not obtain V-0 rating when the total addition amount was 8%. However, when the total addition was 10 %, EP/10% DOPO modi ed EPs passed the V-0 rating. Interestingly, when the mass ratio of CS / DOPO was 1 : 2 or 2 : 1, the modi ed EPs passed the V-0 rating, but the other mass ratio of CS / DOPO modi ed EPs only obtained the V-1 or V-2 rating. It indicated that under the premise of a certain total amount of addition with the mass ratio of the two of 1 : 2 or 2 : 1, the synergistic effect of CS and DOPO on EPs was prominent. For LOI tests, the LOI value of EP sample was 22.6 %, however, with the incorporation of CS or/and DOPO, there was a improvement on LOI values, especially with the incorporation of CS and DOPO both, the increase of LOI value was considerable. The value of EP/10% CS1/DOPO2 and EP/10% CS2/DOPO1 samples reached up to 35.1 % and 34.3 %, respectively, and simultaneously obtained V-0 rating. According to the analyses above, it was concluded that when the total addition amount was 10 % and the mass ratios of CS and DOPO were 1 : 2 and 2 : 1, EP could be endowed with good ame retardancy. , as an important parameter for evaluating re hazard, is calculated by the maximum value of HRR/t and always equal to PHRR/t p , which can estimate both the re spread rate and re scale. The lower the value of FIGRA can be considered the slower the ame spread and ame growth. It can be seen from Table 4 that the FIGRA values of EP/10% CS1/DOPO2 and EP/10% CS2/DOPO1 were signi cantly lower than that of EP, indicating that the CS / DOPO could reduce re hazard. In addition, it is worth noting that the PHRR, THRP and SPR values of the two modi ed EPs were smaller than those of EP, especially, TSP decreased by 61.9% and 71.0%, indicating that the CS / DOPO can not only reduce heat release, but also greatly enhance the smoke inhibition of EP. The possible reason was that the CS / DOPO formed a stable and dense carbon layer during combustion, which hindered the heat transfer inside the samples during ignition and reduced the ue gas release channel, thus achieving the quenching effect. In terms of reducing heat release, EP/10% CS1/DOPO2 had better performance. The av-EHC value of EP was 29.3 MJ kg -1 , while the av-EHC values of EP/10% CS1/DOPO2 and EP/10% CS2/DOPO1 were decreased by 27.6 % and 0.7 %, respectively, indicating that the gas phase ame retardancy was more obvious when the mass ratio of CS and DOPO was 1:2. Moreover, compared with EP, the average CO 2 yield decreased by 55.4% for EP/10% CS1/DOPO2 and 55.0% for EP/10% CS2/DOPO1, while the relative average CO yield only increased by 6% and 9%, respectively. The results indicated that the complete combustion degree of the modi ed EPs signi cantly reduced.
Furthermore, the ame retardancy related to ame inhibition, charring, and formation of a protective barrier layer were quantify according to the calculation method reported in the literature (Tang et al. 2017). The values of the three main modes of action can be obtained from the following equations (1)-(3) and were summarized in Table S1. A signi cant improvement in the ame-inhibition effect 27.60% was obtained when the addition amount of CS and DOPO was 10% with the mass ratio of CS and DOPO of 1:2. The reason can be contributed to the obvious quenching effect caused by more PO free radicals released during thermal decomposition by adding more DOPO. Both EP/10%CS1/DOPO2 and EP/10% CS2/DOPO1 exhibited higher barrier and protection effects, which were 20.36% and 36.52%. The better barrier and protection effects of the latter was due to the higher content of CS in it, it was also the reason why EP/10% CS2/DOPO1 can achieve V-0 rating in the UL-94 tests.

Char residues analyses after CC tests
To disclose the ame-retardant mechanism in the condensed phase, element composition, chemical structure and morphology of char residues surface were investigated by EDX, FTIR, Raman and SEM. Firstly, the elemental composition of the surface of the char residuals without crushing was studied, the results were shown in Fig. S1. For EP/10% CS1/DOPO2, the content of carbon decreased, but the content of oxygen and phosphorus increased signi cantly in comparison with EP, indicating that there were a large number of phosphorus-containing oxides on the surface of residual carbon. For EP/10% CS2/DOPO1 samples, the carbon content increased, and the phosphorus content of oxygen and phosphorus was lower than EP, indicating that the formed char residual was mainly graphitized carbon, which played a ame retardancy effect, which was consistent with the results of Raman. It was worth noting that no nitrogen was found on the char residues surface of all samples, indicating that nitrogen elements on the surface were completely converted into gaseous in the form of small molecules, such as N 2 and NO 2 , etc.
Further EDX elemental mapping analysis was conducted on the cross-section of char residual, and the detailed distribution of elements was shown in Fig. 5. As seen in Fig. 5, all elements except carbon were increased and distributed evenly. That proved the presence of carbon, nitrogen and oxygen in the chars of EPs. In addition, phosphorus was uniformly distributed in the char residuals of EP/10% CS1/DOPO2 and EP/10% CS2/DOPO1, which improved the heat resistance and oxidation resistance of the char residuals. Fig. 6 showed digital and SEM images of char residues after CC test. As shown in Fig. 6 (a1-a4), the char residuals of EP showed that there were only a small amount of char residuals left on the margin of the base plate, along with a large amount holes in its microstructure, which did not have the ability to protect the matrix from combustion. While the char residuals of EP/10% CS1/DOPO2 and EP/10% CS2/DOPO1 increased by 84.3% and 41.6% when compared with EP, respectively, and exhibited more uniform carbon layer with relatively smaller holes in the microstructure. In addition, the carbon layer of EP was discontinuous and loose, while the carbon layer of the modi ed EPs showed a more continuous and dense structure. The dense carbon layer can be used as a protective layer to prevent the continuous combustion of the matrix. At the same time, the char residuals particles of EP were a dispersion state, while the char residuals particles of modi ed EPs had good agglomeration. The agglomeration of char residuals particles of EP/10% CS2/DOPO1 was more obvious among them.
It is known that the barrier effect of condensed phase is not only affected by its quantity, but also closely related to its structure. Raman spectroscopy is an effective technique for analyzing the internal structure of carbonaceous materials after combustion. As shown in Fig. 7, two remarkable peaks at 1360 and 1590 cm -1 in all spectra are belong to D band (vibrations of amorphous carbon atoms) and G band (vibration of sp 2hybridized carbon atoms), respectively. Generally, the ratio of the integrated intensities of D to G band (I D /I G ) is usually used to measure the degree of graphitization. The lower the I D /I G value, the higher the graphitization degree and the better the ame retardancy. As seen, the values of I D /I G follow the sequence of EP (3.93) > EP/10% CS1/DOPO2 (2.88) > EP/10% CS2/DOPO1 (2.65), indicating that the graphitization degree of modi ed EPs had signi cantly increased. This denser char layer can better hinder the escape of ammable gases and entrance of oxygen in the condensed phase during combustion, thus protecting the inner materials to reduce decomposition.
The chemical structure of char residual was studied by FTIR. As shown in Fig. 8, the main absorption peaks were observed at 2927 cm -1 , 1607 cm -1 and 1080 cm -1 , which belonged to the vibration of aromatic rings (Li and Yang 2014). By comparison with EP, the new peaks at 1236 and 1040 cm -1 were attributed to the stretching vibration of P=O and P-O-C, and the absorption peaks of C-P and phenol were observed at 1510 and 754 cm -1 , respectively (Zhang et al. 2012a;Yang et al. 2016). The above results showed that the char residue was mainly composed of graphite-like compounds and organic phosphorus compounds.
Simultaneously, the chemical compositions of char residues for EP, EP/10 %CS1/DOPO2 and EP/10 %CS2/DOPO1 after CC test were further analyzed by XPS. The XPS spectra and the atom percent of C 1s , N 1s , O 1s and P 2s for the char residues of EPs were shown in Fig. 9 and Table S2, respectively. In the C 1s spectrum, there were three kinds of carbon binding states in the three samples, among which the peak at 287.7 eV belongs to C=C and C=O bonds, the peak at 285.8 eV belongs to C-O and C-N bonds, and the peak at 284.7 eV belongs to C-H and C-C bonds in aliphatic and aromatic components (Li et al. 2019). Moreover, it can be clearly seen that the peak at 285.8 eV of the two modi ed EPs increased compared with EP, suggesting that more cross-linked carbon composed of C-O and C-N was formed in the residual carbon. In the N 1s spectrum, all three samples presented two peaks at 400.3 and 398.7 eV, wherein the former was the contribution of N-H bond, and the latter was assigned to C-N bond (Wang et al. 2010). In the O 1s spectrum, similarly, all three samples had peaks at 533.3 eV and 532.3 eV, the former was attributed to the C-OH and C-O-C groups and the latter belonged to C=O (Zhang et al. 2012b). However, the modi ed EPs both had a larger peak at 533.3 eV than that of EP, which was due to the formation of C-O-P or C-O-C, P-O-P groups in the char residues of the modi ed EPs. In the P 2s spectrum, there were two new peaks, 134.4 and 133.3 eV, in the char residues of modi ed EPs (Huang et al. 2018;Sun et al. 2016), which assigned to P-O-P, P-O-C and/or PO 3group in phosphate. This further indicated the formation of phosphate-containing compounds in the char residuals, which were covered on the surface of the materials and prevented the further combustion.

TG-FTIR analysis of EP and modi ed EPs samples
In order to evaluate comprehensively the in uence of the CS / DOPO on the thermal stability of EP composites, TG-FTIR spectra were applied to further disclose the decomposition behavior of the modi ed EPs under programmed heating process. The detailed characteristics of each TG-FTIR spectrum for EP, EP/10% CS1/DOPO2 and EP/10% CS2/DOPO1 obtained at 100 ℃ , 200 ℃ , 300 ℃ , 400 ℃ , 500 ℃ , 600 ℃ , 700 ℃ and 800 ℃ were shown in Fig. 10. The 3D TG-FTIR images of EP, EP/10% CS1/DOPO2 and EP/10% CS2/DOPO1 at various temperatures were shown in Fig. S2. It can be seen that EP produced more CO 2 than other three sample and the absorption peak of phosphorus-containing substances can be clearly observed in EP/10% CS1/DOPO2 and EP/10% CS2/DOPO1.
As shown in Fig. 10, all EPs exhibited the characteristic absorption peaks of non ammable gases, such as H 2 O and CO 2 , which could be demonstrated by the broad bands at 3500-4000 and 2300-2390 cm -1 , respectively (Wang et al. 2011a;Wen et al. 2018). It can be seen that in terms of carbon dioxide release, EP released the most. Compared with EP/10% CS1/DOPO2, EP/10% CS2/DOPO1 released more CO 2 , which was coincided with the results of CC tests. Furthermore, hydrocarbons can be clearly identi ed at the broad bands at 600-700, 1100-1500, and 2900-3100 cm -1 between 400 °C and 800 °C, as the peak at about 661 cm -1 attributed to =CH-bonds, peaks at 1180 cm -1 , 1460 cm -1 and 2972 cm -1 belonged to the -CH 3 bonds, peak at 2927 cm -1 assigned to -CH 2 -bonds (Wang et al. 2011b;Wu et al. 2011). In addition, some sharp peaks appeared in the range of 1550-1800 cm -1 , which might be carbonyl-containing compounds such as carboxylic acids (1706 cm -1 ), ketones (1720 cm -1 ) and aldehydes (1738 cm -1 ) ).
Signi cant difference was observed compared to EP and EP/10% CS, as shown in Fig. 10, after forti ed with CS / DOPO for EP, the peak at about 1129 cm -1 for EP/10% CS1/DOPO2 and EP/10% CS2/DOPO1 was attributed to the Ph-P bond (Xu et al. 2016). Moreover, the peak at about 1260 cm -1 assigned to the stretching vibration of P=O, which was probably overlapped with the characteristic absorbance of ether bonds. A weak absorption peak at 1642 cm -1 was the stretching vibration of P-OH (Yuan et al. 2018). On the basis of the analysis above, a variety of phosphorus-containing compounds were generated during the thermal decomposition process for modi ed EPs, which can increase the ame retardancy e ciency of EP signi cantly. It is worth noting that compared with EP/10% CS1/DOPO2, most of the absorption peaks of non ammable gas for EP/10% CS2/DOPO1 were larger, so it was speculated that EP/10% CS2/DOPO1 was mainly gas-phase ame retardancy effect while EP/10% CS1/DOPO2 was mainly condensed-phase ame retardancy effect, which was consistent with the results of CC tests.
In order to explore the time when these EPs began to decompose into gas in the process of thermal decomposition, the curve was made with time as abscissa and infrared absorption intensity as ordinate, simultaneously the rst derivative curve was drawn. seen in the Fig. S3, the pyrolysis products of EP began to release at approximately 31.2 min. Surprisingly, the time of EP/10% CS1/DOPO2 and EP/10% CS2/DOPO1 began to release pyrolysis products almost identical, which were 30.3 min and 30.2 min, respectively, a little less than that of EP. The results showed that the thermal stability of the modi ed EPs slightly decreased after adding CS / DOPO, which was consistent with the test results of T g and TGA.

Mechanical properties
The mechanical properties of materials are important in some areas, so it is necessary to study the mechanical properties of modi ed EPs and make corresponding analysis. The mechanical properties were analyzed from three indexes: tensile strength, exural strength and izod impact strength. The results were summarized in Fig. 11 On the other hand, these active groups would react with epoxy groups in EP to undergo ring-opening reaction, which enhanced the formation of cross-linking network in the curing process. What's more, the π-π stacking among the aromatic rings were also conducive to improving its mechanical strength. The possible chemical reactions were shown in Scheme 1.

Conclusions
In this work, the direct compound of biomass and DOPO was used to modify the EP by a solution blending method. The results showed that the ame retardancy of EP was improved by adding CS / DOPO, especially when the total amount of CS / DOPO was 10% and the mass ratio of CS and DOPO was 1:2 and 2:1, both modi ed EPs obtained V-0 rating, and LOI increased from 22.6% for EP to 33.7% and 32.5%, respectively. At the same time, the PHRR, THR, PSPR and TSP values of EP/10% CS1/DOPO2 and EP/10% CS2/DOPO1 decreased. It showed that under a certain mass ratio, the relative properties of the composite can be greatly ampli ed by simply mixing the functional materials. In comparison with each other, the ame retardancy mechanism of EP/10% CS1/DOPO2 and EP/10% CS2/DOPO1 tended to be in gas phase and in condensed phase, respectively. The general ame retardancy mechanism was summarized as follows according to the results of CC, FTIR, SEM, EDX, TG-FTIR, etc. 1) part of DOPO and CS that didn't involve in curing process was preferentially decomposed into some non-combustible gases such as CO 2 and NH 3 during in the early stage, diluting the combustible gases; 2) DOPO released Pcontaining radicals during the pyrolysis process, which helped to capture free radicals of chain reaction; 3) the obstructing function on heat transfer of the dense char residue. In the meantime, the mechanical properties of EPs did not decrease but increased, especially the exural strength. Therefore, under the 10% total addition amount with the mass ratio of CS and DOPO of 1 : 2 or 2 : 1, the EPs were endowed with good ame retardancy and mechanical properties.

Declarations
Funding: This study was funded by Hubei provincial science and technology department (2018ACA158).
Con icts of interests: All authors certify that they have no a liations with or involvement in any organization or entity with any nancial interest or non-nancial interest in the subject matter or materials discussed in this manuscript.
Availability of data and material: All data generated or analyzed during this study were included in this published article (and its supplementary information les).