QIN Shiqian, YANG Zhiyi, ZHANG Shuai, ZHANG Zhiye, WANG Xinlong,YANG Xiushan, LUO Tao, YANG Lin

(School of Chemical Engineering, Sichuan University, Chengdu 610065, China)

Abstract: A halogen-free flame-retardant (hydroquinone bis (N, N’-diarylphosphoramidate), 4N-HDP)containing phosphorus-nitrogen was synthesized. Its structure was characterized by infrared spectroscopy(IR), nuclear magnetic resonance (1H-NMR and 31P-NMR). Thermogravimetric analysis (TG), limiting oxygen index (LOI), UL-94 vertical burning test (UL-94), thermogravimetric-infrared instrument (TG-IR) and scanning electron microscopy (SEM) were used to compare the flame-retarding performance and mechanism of hydroquinone bis (diphenyl phosphate) (HDP) and 4N-HDP. TG, IR and TG-IR were used for comparative analysis, indicating that both HDP and 4N-HDP are flame-retardants, and the gas phase and condensed phase act synergistically. In the pyrolysis process, it is divided into two steps: the first step is the breakage of large molecules to small molecules; the second step is the gasification and carbonization of small molecules, and eventually produces phosphate ester and non-flammable gases. Through the comparison of various results, it could be found that 4N-HDP has better flame-retarding performance compared to HDP in the composite with polycarbonate (PC).

Key words: flame-retardant; phosphoramine; thermogravimetric analysis; polycarbonates; TG-IR

1 Introduction

Polycarbonate (PC) mixed with halogen-free flame-retardants (FRs) has been widely used in many fields, because of its outstanding features as rigidity,excellent impact strength, eminent transparency and thermal stability[1,2]. Despite of high flame-retardancy,PC still can’t satisfy the requirements in many fields such as electronic and electric applications[3]. At present, the simplest way to improve the flame-retardancy is adding flame-retardants to PC. Considering the compatibility between the flame-retardants and PC, organic flame-retardants would be a better choice.

Halogen-containing flame-retardants, especially the bromine and chlorine systems, has become the most important class of flame-retardants in the past few decades. However, the disadvantages of halogen-containing flame-retardants are particularly obvious: during combustion a lot of corrosive gases and especially some strong carcinogens such as dioxins can be produced, which is definitely unfavorable to people and to the environment[4]. Halogen-containing flame-retardants are a good choice if environmental pollution was not considered[5,6]. Therefore, halogen-free flame-retardants are getting more attention for being environmentally friendly[7].

Organic phosphorus flame-retardants have not only the advantages of low toxicity and low smoke,but also can generate carbon residues with high stability, which would greatly improve the flame-retarding properties of materials[8]. Among various organic phosphorus flame-retardants, the phosphorus-nitrogen one has become the focus of research in recent years[9]. The main advantages of N-containing compounds are low toxicity and good stability. In case of fire, the materials would not produce dioxins, halogen acids, and a lot of smoke[10,11].

Scheme 1 The synthetic route of the target flame-retardant (4N-HDP)

Table 1 Structure and characteristics of FRs employed in this study

In this study, hydroquinone bis (N, N’-diarylphosphoramidate) (4N-HDP), a new kind of phosphorus-containing flame-retardant, is synthesized successfully and its chemical structure is characterized comprehensively. Besides, the flame-retardant performances of 4N-HDP and a model compound, HDP, are compared. Eventually, the flame-retardant mechanism is studied by TG, IR, and TG-IR.

2 Experimental

2.1 Materials

Oxalic acid, aniline, triethylamine and hydroquinone were supplied by Chengdu Jinshan Chemical Reagent Co. Ltd. (China). Aluminum chloride (AlCl3)and sodium hydroxide were supplied by Tianjin Fuchen Chemical Reagent Co. Ltd. (China). Phosphorus oxytrichloride was supplied by Tianjin Guangfu Fine Chemical Research Institute. All the chemicals mentioned above were of analytic grade. PC was supplied by Bayer (Germany). HDP was synthesized according to previously reported methods[12].

2.2 Synthesis of 4N-HDP

Scheme 1 shows the synthetic route of the flame-retardant 4N-HDP. At first the assembled device was purged with nitrogen for 5 minutes, then, 36 mL of phosphorus oxychloride was added to a three-neck flask. The temperature was set at 80 ℃. Then 11 g of hydroquinone was added to the flask, followed by 0.5 g of AlCl3[13]. Afterwards, excess phosphorus oxychloride was removed by vacuum distillation, and an intermediate product - hydroquinone phosphoryl chloride was yielded. The yield of the first synthetic product was 99.33%. The obtained intermediate product was dried under vacuum until constant weight. 13.84 g of the intermediate product was dissolved in 10 mL of tetrahydrofuran (THF), added to a three-neck flask, and then 2 g of triethylamine (Et3N) was added, and finally 16.76 g of aniline was added. The control temperature was 66 ℃. The end point of the reaction was detected by thin-layer chromatography. The reaction mixture was washed three times with oxalic acid (1 wt%) and hot water at 60 ℃, respectively. Then a solid crystallized and was filtered, washed with cold water to obtain the final product 4N-HDP. The yield of the final product was 95.04%. The structures and names of 4N-HDP and HDP are listed in Table 1.

2.3 Characterization of 4N-HDP

Infrared spectroscopy (IR) analysis was conducted with a spectrophotometer (Nicolet 6700, USA). The mixture of FRs and KBr was pressed to a tablet for the analysis. The wavenumber range was from 4 000 to 500 cm-1.

The HDP and 4N-HDP samples were placed in a muffle furnace, and they were heated from room temperature to 650 °C at a heating rate of 10 ℃/min. Samples from the furnace were taken every 50 ℃ from 100 to 650 ℃, and the obtained samples were analyzed by IR with a spectrophotometer.

Nuclear magnetic resonance (1H-NMR and31P-NMR) measurements of 4N-HDP were performed with an NMR spectrometer (DMX-300, Bruker) with deuterium dimethyl sulfoxide (d6-DMSO) as the solvent.

2.4 Polycarbonate composites preparation

PC and different amounts of 4N-HDP or HDP which had been dried were prepared into PC composites by a Huck mixer at about 210 ℃, then the composite was cut into particles with a pelletizer. The samples were pressed in a vulcanizing machine (PG-63 Plate) for 10 min at 210 ℃ and under a pressure of 10 MPa, and then compressed for 5 min at 10 MPa and 20 ℃. Finally, a 3 mm thick sample plate was obtained.The weight ratio of PC to flame-retardant is shown in Table 3.

2.5 Polycarbonate composites characterization

2.5.1 Measurement of limiting oxygen index (LOI)

An instrument (XZT-100A, China) was used to measure the LOI value. The material was made into a specimen of size 127 mm×10.5 mm×3 mm according to a national standard (GB 2406-93, China). In the measurement, the mixed gas flow rate was 14 L/min.

2.5.2 Measurement of UL-94 vertical burning test(UL-94)

The UL-94 was tested with a CFZ-3 instrument(China). According to the UL-94 standard, observe the dripping phenomenon while recording the burning time, and the sample size is 127×12.7×3 mm.

2.5.3 Thermogravimetric analysis (TG)

TG was performed on a thermal analyzer (TG/DSC 1 Mettler/Toledo). The analyses were performed for PC, HDP and 4N-HDP, as well as HDP/PC (7/93 in weight ratio) and 4N-HDP/PC (7/93). The heating rate was 10 ℃/min, nitrogen flow rate was 60 mL/min,the flow rate of protecting gas was 20 mL/min, and the temperature range was 30 to 780 ℃.

2.5.4 Thermogravimetric-infrared instrument (TG-IR)

TG-IR analysis was performed with a TG 851 thermogravimetric analyzer interfacing with a Nicolet iS10 IR spectrometer. The sample was heated at a rate of 10 ℃/min under nitrogen in separate experiments(flow rate = 30 mL/min).

2.5.5 Elementary and morphological analysis of the carbon layers

The residual carbon after combustion in the LOI measurement was taken for scanning electron microscopy (SEM) analysis with an S-4800 microscope(Hitachi, Japan). The accelerating voltage was 20 kV.The distribution of P elements in the carbon layer was observed with the energy dispersive spectrometer (EDS)of Thermo Scientific ESCALAB250XI (VG, Britain).

3 Results and discussion

3.1 Structural characterization of 4N-HDP

Fig.1 shows the IR spectrum of 4N-HDP. The characteristic absorption peaks of 4N-HDP are as follows: the adsorption band of the P-N stretching is observed at 943 cm-1; the P-O-CArstretching vibration band is observed at 998 cm-1; observed at 1 180 cm-1is the stretching vibration band of C-O; the P=O stretching vibration band is observed at 1 285 cm-1;the N-CArstretching vibration band appeared at 1 384 cm-1; the aromatic C=C stretching vibration band is observed at 1 597 cm-1and weak bands are observed at 3 187 cm-1corresponding to the N-H, respectively[14-23].Compared with HDP, 4N-HDP has an N-CArabsorption peak, without absorption peak of CAr-O, which can be broken when heated, and finally produces a substance containing N to act as a dilution air. The IR characterization shows that the structure is consistent with the structure of the target product.

Fig.1 IR Spectrum of 4N-HDP

1H NMR and31P NMR is also used to further identify the structure of 4N-HDP. Fig.2 shows the1H-NMR spectrum of 4N-HDP. The identification of chemical shifts in the1H-NMR spectrum (400 MHz, DMSO-d6)are as follows: 8.44 (2H, broad signal, N-H), 7.12-7.20(20H, broad signal, CAr-H, benzene ring attached to N),6.86 (4H, broad signal, CAr-H, benzene ring attached to O). For the31P-NMR (400 MHz, DMSO-d6) spectrum as shown in Fig.3, there is only one peak atδ=-12.15 ppm, which indicates that the target flame-retardant contains only one type of phosphorus atom. That is to say, the synthesis of 4N-HDP is successful.

Fig.2 The 1H NMR spectrum of the 4N-HDP

Fig.3 The 31P spectrum of the 4N-HDP

3.2 Thermal behavior of PC, HDP and 4N-HDP

Fig.4 TG for PC, HDP and 4N-HDP under N2

Fig.4 and Fig.5 shows the TG and differential thermogravimetric (DTG) curves of PC, HDP and 4N-HDP. Table 2 shows the main thermal characteristics of PC, HDP and 4N-HDP. According to Fig.4,Fig.5 and Table 2, the initial decomposition temperature (Tinitial, at 5% weight loss) of PC is 482 ℃, and the thermal degradation process of PC only has one step with the main peak of weight loss (Tmax) at 517 ℃ and its residual char is 24.6% at 780 ℃. In the contrary,HDP and 4N-HDP are both degraded in two steps: theTinitialof HDP is 277 ℃, and that of 4N-HDP is 236 ℃;HDP decomposes between 277 ℃ and 780 ℃ in a main decomposition step and a small shoulder. For HDP,Tmax1of the main step is 338 °C andTmax2of the shoulder is 537 °C, and the residual char at 780 °C is 14.2%.The degradation process of 4N-HDP is similar to HDP,Tmax1of the main step is 311 °C andTmax2of the shoulder is 542 °C, but its residual char content is as high as 34.4%. Based on these results, it can be seen that the thermal stability of 4N-HDP is better than that of HDP,and there are more carbonized residues.

Fig.5 DTG for PC, HDP and 4N-HDP under N2

Table 2 Thermal degradation data under pure nitrogen by thermogravimetric (TG) analysis

Fig.6 IR of 4N-HDP at different temperature

In addition, in order to better understand the formation of char in the thermal degradation process of HDP and 4N-HDP, IR characterization of FRs after the treatment at different temperatures is carried out. The results are shown in Figs.6 and Fig.7. In Fig.6, before 250 ℃, the main characteristic functional groups are CAr-H (3 075 cm-1, 3 019 cm-1), N-H (3 197 cm-1), C-N(1 384 cm-1), C-O (1 175 cm-1), P-N-CAr(946 cm-1).With the increase of temperature, the absorption peaks of CAr-H, C-N, P-N-CArdisappear, which indicates that the degradation process of the four benzene rings in the 4N-HDP molecule are mainly involved in this stage. In addition, the mass proportion of the four benzene rings is 54%, and the total mass loss during this degradation should be 54%, which is consistent with the result of 56.4%, from TG. A new absorption peak that indicates O=P-OH appears at 450 ℃, and the absorption peak of O=P-O appears at 650 ℃. These indicate that the second pyrolysis of 4N-HDP produces phosphoric acid,which is then dehydrated into phosphate ester. However, the proportion of element P is small, so the weight lost in this stage is relatively less than that in the first stage, which could be mutually confirmed with TG analysis. The results of IR analysis of carbon residues at different temperatures are consistent with the weight loss analysis at different temperatures from TG. This shows that the pyrolysis of 4N-HDP mainly have two steps: the first step is the degradation of the macromolecule into small molecules, mainly in the form of the benzene ring detachment from the main macromolecule at 250 ℃; the second step is mainly the decomposition of small molecules at 450 ℃. At 650 ℃, there are also obvious bonds: N-H (3 415 cm-1), P-O (1 284 cm-1), P-N (1 008 cm-1) and C=C (1 800 cm-1). It is implied that the decomposition products of 4N-HDP are phosphate ester, phosphorus and nitrogen compounds,benzene rings and small amounts of alkenes. In Fig.7,there is nearly no degradation for HDP before 400 ℃.The degradation occurs mainly between 400 and 500℃. This may be different from the thermogravimetric analysis result, because at this initial stage the muffle furnace may contain some air. The absorption peaks of HDP (CAr-H, C=C, P-O-CAr, the CAr-H and P-N-CAr)disappear at 500 ℃. This stage is similar to the pyrolysis stage of 4N-HDP, mainly because the benzene rings are detached from the macromolecule, resulting in small molecules. At 650 ℃, there are the O=P-O and C=C absorption peaks, indicating that phosphate ester and alkenes are formed, and the pyrolysis process is similar to corresponding process of 4N-HDP, and the result is the same as TG. By comparing the IR spectra of HDP and 4N-HDP solids at different temperatures,it is obvious that the pyrolysis process of the two flame-retardants can be divided into two steps: the first step has most of the weight loss, and the second step has only a small amount of weight loss. This result is consistent with the TG analysis.

Fig.7 IR of HDP at different temperature

3.3 Flame-retarding performance

The measurement of LOI value and the UL-94 vertical test could give important indications of the flame-retarding properties[24]. The flame-retarding properties of FRs are tested by mixing them into PC. The addition amount of FRs is 1 wt%-7 wt% of the mass of PC. The composition of samples used for the LOI and UL-94 tests is shown in Table 3.

Table 3 LOI and UL-94 results for the mixtures of PC with various FRs

The LOI values increase with the increase in addition amount of FRs. The LOI value of PC is 28.0 and those of mixtures containing 1 wt%-7 wt% HDP are in the range of 29.0-30.4. On the other hand, those of mixtures containing 1 wt%-7 wt% 4N-HDP are 29.5-32.0.When the HDP is added at 7 wt%, the UL-94 vertical test reache V-0 without dripping, while the 4N-HDP is added at 5 wt%, the UL-94 vertical test reache V-0 without dripping. It is clear that the flame-retarding effect of 4N-HDP with nitrogen is better than that of HDP.

Fig.8 TG for PC, PC/HDP and PC/4N-HDP under N2

Table 4 Thermal degradation data under pure nitrogen by TG

In order to better explore the relative flame-retarding efficiency, thermogravimetric analyses are performed. Fig.8 shows the TG curves of neat PC and PC with 7 wt% flame-retardants (PC/FRs). Table 4 shows the main thermal characteristics of PC, HDP/PC and 4N-HDP/PC. Both mixtures with 4N-HDP and HDP have only a one-step degradation, which illustrates that both FRs are compatible with PC. PC/4N-HDP has 2.2 % more carbon residue than PC/HDP. 4N-HDP decomposes at a lower temperature as compared to HDP(Figs.6, 7), while the decomposition temperature of PC/4N-HDP is higher than PC/HDP. This indicates that the flame-retarding performance not only depends on the performance of flame-retardants, but also depends on the type of polymer. The amount of carbon residue from HDP alone is very low, but after adding it into PC, the amount of carbon residue is higher than that of pure PC. This indicates that the flame-retarding mechanism is the same as the condensed phase flame-retardant. The addition of 4N-HDP into PC increases the residual carbon amount by 2.3% as compared to that of pure PC, even though the amount of carbon residue in PC/4N-HDP is lower than that of pure 4N-HDP, because 4N/HDP is added as an additive to the PC to alter the flame-retarding performance of PC. The addition of 4N-HDP could significantly improve the residual carbon amount of PC, indicating that the flame-retarding mechanism is condensed phase flame-retardant. From the results of TG, it is easy to see that 4N-HDP has more residual carbon than HDP. When the composite is burned, 4N-HDP would form a thicker protective layer to protect the inner layer from burning, and it is less prone to dripping. This is also the reason why 4N-HDP’s LOI and UL-94 test results are better than those of HDP.

3.4 Discussion on possible flame-retarding mechanism

3.4.1 TG-IR analysis of 4N-HDP

In order to better understand the flame-retarding mechanism, TG-IR analysis was employed. Fig.9 shows the IR spectra of the gas phase product from the 4N-HDP pyrolysis process under nitrogen at different temperatures. No significant absorption peaks can be detected before 200 ℃, indicating that there is no thermal degradation before 200 ℃. At 200 ℃,there appears obvious peaks of the functional groups including mainly N-H (3 479 cm-1, 3 414 cm-1), CAr-H(3 075 cm-1, 3 019 cm-1), C=C (1 615 cm-1, alkene),C=C (1 498 cm-1, ring), O-P=O (1 275 cm-1) and C-O(1 175 cm-1). The N-H bond probably belongs to hydronitrogen. O-P=O could be attributed to phosphate ester. CAr-H, C-O and C=C (1 498 cm-1) are probably from the phenol compounds. C=C (1 615 cm-1, alkene)is ascribed to the formation of carbide by dehydrogenation. 200-400 ℃, the absorption peaks of CAr-H, C=C(alkene), C=C (ring) and O-P=O are very prominent,indicating that there are a large number of small alkene molecules, benzene ring molecules and phosphate ester in this stage. This is consistent with the IR analysis of solid residual carbon at different temperatures. At the 500-700 ℃ stage, the four absorption peaks have very low absorbance and the peak area is small, which indicates that only a small amount of material has evaporated into the gas phase at this stage. In addition,the N-H and C-O absorption peaks are all weak in the range of 100-700 ℃, indicating that little carbon monoxide, carbon dioxide and hydronitrogen are emitted throughout this process. Based on the results above,it is clear that the first stage is from 200 ℃ to 400 ℃,a large amount of materials degrades from the solid phase and is released into the gas phase in which they can be detected by the infrared detector. The degradation products in the gas phase mainly contain benzene,alkene and PO2·, and only a small amount of carbon monoxide, carbon dioxide and hydronitrogen are produced. The second stage is from 500 to 700 ℃. At this stage, the absorption peaks are weak, implying that only a small amount of the degradation products mentioned above are formed. According to the literature the amount of a material is proportional to the peak area in the infrared spectrum and inversely proportional to the temperature[25,26]. Therefore, from Fig.9, the amount of substances degraded in the first stage is five times that of the second stage, which is the same as the result of the TG analysis. From TG, IR analyses of solid residues at different temperatures, and TG analyses of the composites, it could be drawn that HDP has the same degradation mechanism as 4N-HDP. HDP degradation eventually produces phosphate ester, alkene, carbon,and some nonflammable gases.

Fig.9 IR spectra of pyrolysis products for 4N-HDP at different temperature under N2

3.4.2 Semi-quantitative analysis of 4N-HDP

The release of volatile gases (hydronitrogen) and phosphate ester during the combustion of FRs under N2make contributes to the flame-retardant properties. The release of hydronitrogen and phosphate ester could act to dilute the combustion-supporting substances in the air. So semi-quantitative analysis on hydronitrogen and phosphate ester would be informative. The obtained IR spectra are subjected to area integration by the OMNIC program and divided by the corresponding temperature, and the results are shown in Figs.10 and 11. It shows the peak area-temperature curves obtained for the hydronitrogen and phosphate ester under nitrogen.The two curves both have two peaks, which is consistent with the observation above that the decomposition has two steps. The release trends of hydronitrogen and phosphate ester are synchronous. The release amount starts to increase sharply at a temperature of about 150℃, and the maximum peak is centered around 285 ℃,which could be the degradation peak of 4N-HDP in the composite. The area of the second peak (around 517℃) is much smaller than that of the first peak. From the differential weight loss of corresponding FRs during TG analysis in Fig. 10 and Fig. 11 (the line above), it is clear that the weight loss and the release amount of the gases are synchronous. This indicates that the degradation of these two substances is consistent with the trend of weight loss, and the more weight loss, the more substances are released.

Fig.10 Release amounts of hydronitrogen under N2

Fig.11 Release amounts of phosphate ester under N2

3.5 Microstructure of the carbon layers

In order to better understand the mechanism of the flame-retarding property, the morphology and microstructure of the surface residues of the compounds after combustion are investigated and the results are shown in Fig.12.

Fig.12 shows the residual carbon of the two composites after the LOI test, and the surfacemorphology is observed with a scanning electron microscope, and the distribution of element P is analyzed by energy-dispersive spectroscopy (EDS). The carbon residue resulted from the flame-retarding PC composite displays a lot of holes, which can be ascribed to the generated gases in the combustion process. During combustion, phosphate-based flame-retardants produce a large amount of H2O, and phosphorus-containing radicals like PO·and PO2·[27]. The decomposition products of PC include CO, CO2and also a lot of H2O. These small molecules could pose a significant inhibitory effect on the combustion of the material, when a large number of such small molecules aggregate at the surface of the composite material. They can play a certain flame-retarding effect. As could be seen from the SEM images and the EDS mapping graphs, compared to the carbon residue resulted from PC/HDP, the carbon residue from PC/4NHDP has more P elements in the surface layer, and its surface pores are larger. These larger pores could accommodate more small molecular substances, could have more small molecules to overflow to dilute the air,and therefore have a good flame-retarding effect. The above results may be caused by the difference in the content of phosphorus in the composite material due to the compatibility between the flame-retardant and the matrix material, thus affecting the distribution of phosphorus in the carbon layer after combustion. The above two tests show that 4N-HDP have better flame-retarding effect than HDP, which is consistent with the test results of LOI test and TG.

Fig.12 EDS before LOI test and SEM after LOI test (×500)

3.6 Summary of mechanism of the flameretardant

By comparing the TG, IR analysis results of solid residues at different temperatures, SEM and EDS analysis of the carbon residues from the composites with HDP or 4N-HDP, and the TG-IR and semi-quantitative analysis of the combustion process of 4N-HDP, we can draw a conclusion that 4N-HDP has better flame-retarding performance than HDP, as a flame-retardant in PC. For the discussion of the flame-retarding mechanism, it is found that 4N-HDP has the same degradation mechanism as HDP: the flame-retarding mechanism of these two flame-retardants is combined with gas phase flame-retardancy and condensed phase flame-retardancy. When heated, it would be divided into two steps of pyrolysis. The first step breaks the macromolecule into smaller molecules, this step is mainly realized by the detachment of benzene ring from the main body to form phenols, and also the formation of phosphoric acid, alkene and some hydronitrogen. In this step, the solid will lose a large portion of molecular weight.The second step is mainly the carbonization of small molecules. In this step, the phosphoric acid would continue to lose water and yield phosphate ester, and the phenol would be further carbonized, accompanied by a small amount of hydronitrogen. In the flame-retarding process, hydronitrogen acts to foam and dilute the flame-retardants gas in the air, and the residual carbon acts to block the air outside and prevents flames from entering the interior of the material.

4 Conclusions

The LOI value of PC/4N-HDP (a weight ratio of 95:5) reaches 30.8, which is higher than 30.0 for PC/HDP (95:5) and 28.0 for neat PC. The UL-94 tests suggest that PC/4N-HDP (95:5) reaches the V-0 grade while PC/HDP (95:5) only reaches the V-2 grade. By IR of the gas and solid phases at different temperatures,the pyrolysis mechanisms of HDP and 4N-HDP are discussed, and it is found that they have the same degradation mechanism. And the degradation process can be divided into two steps: the first step is mainly the detachment of small molecules from the host molecule;The second step is the carbonization and gasification of the small molecules. Hydronitrogen and residual carbon are generated during the whole pyrolysis process. The hydronitrogen dilutes air and acts as a gasphase flame-retardant. Residual carbon is an insulation material and acts as a condensed phase flame-retardant.From the TG analysis, 4N-HDP has a lower decomposition temperature and more residual carbon than HDP.From the perspective of pyrolysis mechanism, 4N-HDP would release more hydronitrogen to dilute the combustion-supporting gas in the air, and which would form a thicker carbon layer to block air and materials.Although HDP also produces some gases to dilute the combustion gases, the formation of the carbon layer is relatively late. So, 4N-HDP has better flame-retarding performance. In addition, through SEM and EDS analyses, it is easy to see that the carbon layers resulted from the PC/4N-HDP composite have more holes and more uniform pore sizes than those of PC/HDP, and the distribution of P elements is more uniform and denser,which proves the previous viewpoint: 4N-HDP has superior flame-retarding properties than HDP in the composite with PC.