The application of a phosphorus nitrogen flame retardant curing agent in epoxy resin

2.1 Materials

P-xylylenediamine (PXDA, chemically pure) were purchased from Shandong Xiya Chemical Reagent Co., Ltd. (Shandong, China), phosphorus oxychloride (POCl₃, analytical reagent (AR)), 4-dimethylaminopyridine (DMAP, AR), Pentaerythritol (PER, AR), acetonitrile (AR), chlorobenzene (AR), and dichloromethane (CH₂Cl₂, AR) were all purchased from Sinopharm Chemical Reagent Co., Ltd (Shanghai, China). The EP resin (bisphenol A diglycidyl ether, trade name E-44, EP equivalent = 210-240 g/eq) was purchased from Jinan Zhuopu Chemical Technology Co., Ltd (Jinan, China).

2.2 Synthesis

2.3 Preparation of flame-retarded EP resin

PXDA was used as a pure sample curing agent which was used to compare with the flame retardant curing agent PPXSPB. The curing process of all flame retardant epoxy resin composites was under the same conditions. The mass ratios of all components are listed in Table 1. The mixture was stirred continuously by high speed disperser for 20 min at 90°C. When the mixture was mixed well, degassed it under vacuum for 30 min, then the mixture was poured into a preheated mold at 119°C. The samples were cured for 2 h at 119°C, 3 h at 146°C and then post-cured for 2 h at 178°C.

Scheme 1

Synthesis route of SPDPC (a) and PPXSPB (b).

2.4 Characterization

Phosphorus content was determined by 721 digitally visible spectrophotometer (Sanghai Jinghua, China) according to the GB/T11893-89, potassium persulfate was used as the oxidant to digest the solid sample.

FTIR analysis was performed using a Thermo Nicolet NEXUS-470 (Ramsey, Minnesota, USA) spectrometer by an attenuated total reflectance method. Each sample was scanned from 4000 to 400 cm⁻¹.

¹HNMR spectra and ³¹PNMR were recorded on an AVANCE III 500MHz spectrometer (Bruker, Switzerland). The chemical shifts relative to that of DMSO-d₆ were recorded.

The molecular weight was determined by micrOTOF-Q 125 (Bruker Customer, Germany), with the sample dissolved in methanol.

Thermogravimetric analysis (TGA) were performed on a STA449C/41G thermal analyzer (NETZSCH, Germany). The mass of PPXSPB was 5 mg, and EP-0, EP-1, EP-2, EP-3 were the same, 4.5 mg. Each sample was tested from 30 to 800°C at a heating rate of 10°C/min under N₂ or air atmosphere.

The limiting oxygen index (LOI) is used to indicate the concentration (volume fraction) of oxygen in a mixture of oxygen and nitrogen when it is just capable of supporting its combustion. LOI test was carried out according to the GB/T2406-1993. The LOI was measured using a JF-3 LOI chamber (Jiangning Analytical Instrument Factory, Nanjing, China), and according to the ASTM D2863 standard, the sample dimensions were 130 × 6.5 × 3.0 mm³.

The vertical burning tests (UL94) were measured by a CZF-3 instrument (Jiangning Analytical Instrument Factory, Nanjing, China), and according to the ASTM D3801 testing procedure, the sample dimensions were 130 × 13.0 × 3.0 mm³. The vertical burning categories are shown in Table 2 according to EN 60695-11-10:1999.

For each set of five test specimens from the two conditional treatments, calculate the total afterflame time for the set t_f in seconds, using the following equation:

where:

t_f – the total afterflame time, in seconds;

t_1,i – the first afterflame, in seconds, of the i^th test specimen;

t_2,i – the second afterflame, in seconds, of the i^th test specimen.

The standard sample for LOI and vertical burning was dried to constant weight at 70°C, weighed and recorded as W₀; it was placed in distilled water at 70°C for 168 h, and dried it at 70°C to constant weight, recorded as W₁. The immersed samples were subjected to LOI and vertical burning tests. The mobility of the flame retardant in EP/PPXSPB flame retardant systems is calculated according to the formula:

Cone calorimeter test was carried out according to the ISO 5660-1 standard on a FTT cone calorimeter (Fire Testing Technology Co. Ltd. UK). Specimens with sheet dimensions of 100 × 100 × 3.0 mm³ were irradiated at a heat flux of 35 kW/ m². Each sample was tested at least twice.

The tensile strength tests were performed according to the International Organization for Standardization ISO 527-2 standards using a TCS-2000 electric tensile tester (Gotech testing machines Inc., Taiwan), and the tests were performed at a speed of 10 mm‧min^-1 at room temperature. Charpy impact tests were performed according to ISO 179-1 standard using a GT-7045-MDL Charpy impact tester (Taiwan). All of the listed results are the mean of five samples. The dumbbell sample dimension of tensile tests was 200 mm × 20 mm × 4 mm, and the size of the stretched part is 60 mm × 10 mm × 4 mm. The sample dimension of charpy impact tests was 80 mm × 10 mm × 4 mm.

Scanning electron microscopy (SEM) testing by an S-3400N instrument (Hitachi, Japan) was used to investigate the combustion residues of burned samples from cone calorimeter tests.

The adhesive strength tests were performed according to GB/T 7124-2008 using electronic universal testing machine RGD-5 (Shenzhen, China). The stretching rate was 10 mm/min, Iron substrate 100 mm × 25 mm × 5 mm, sizing area of 12.5 mm × 25 mm, curing conditions of 119°C/2 h + 146°C/2 h + 178°C/2 h. The data was taken as the mean of five samples.

3.1 Structural characterization

The structure of PPXSPB was characterized by FTIR, ¹HNMR, and mass spectroscopy. Figure 1 was the FTIR spectra of SPDPC and PPXSPB, as is shown in spectra of PPXSPB, the absorption peak of P=O appears at 1245 cm⁻¹, 1022 cm⁻¹ and 1074 cm⁻¹ are the absorption peaks of P–O–C and P–N (15), respectively. Compared with SPDPC, the spectrum of PPXSPB shows the absorption peaks at 3380 cm⁻¹ is attributed to the stretching vibration absorptions of –NH₂. The absorption peak at 838 cm⁻¹ is attributed to the bending vibration absorptions of –Ph. The absorption peaks at 1598 cm⁻¹ is assigned to the variable angle vibration of –NH (16), and the absorption peak of P=Cl at 547 cm⁻¹ disappeared.

Figure 1

FTIR spectra of SPDPC and PPXSPB.

Figure 2 was the ¹HNMR (a) and ³¹PNMR (c) spectra of PPXSPB and the ³¹PNMR (b) spectrum of SPDPC. As is shown in Figure 2a, the signal at 2.50 ppm is attributed to the DMSO-d₆ protons. The chemical shift of protons (b), (g) of methylene between –NH₂ and the aromatic ring is at 3.9 ppm, and methylene proton adjacent to P–O can be observed at 3.34 ppm. The chemical shifts of –NH₂ adjacent to the aromatic ring is at 4.22 ppm, the chemical shifts of the protons (c), (d), (e) and (f) of the aromatic ring were observed at 7.15-7.56 ppm. The protons (h) of –NH between –P=O and the aromatic ring is shown at 8.07 ppm. The ³¹PNMR is shown in Figure 2c, the signal at 6.45 ppm is attributed to the P proton, compared with Figure 2b, the signal at 6.45 ppm disappeared, which proves that SPDPC has completely reacted.

Figure 2

¹HNMR (a) and ³¹PNMR (c) spectra of PPXSPB and ³¹PNMR (b) spectra of SPDPC.

Figure 3

Mass spectrum of PPXSPB.

The mass spectrum of PPXSPB is shown in Figure 3. The m/z values 232.1, 429.1, 497.2, 609.1, 789.2 are correspond to 1/13 M₈H⁺, 1/2M₂H⁺, MH⁺, 1/2 M₃H⁺, 1/2M₄H⁺, respectively, while M is the repeating unit of PPXSPB.

3.2 Thermal stability of PPXSPB and EP/PPXSPB composite

Figure 4a is the TG and DTG curves of PPXSPB in air atmosphere. Epoxy resin will decompose at about 200°C in air (17), while PPXSPB starts to decompose at 250~300°C, which proves that it can adapt to the processing temperature of epoxy resin and may delay the decomposition of epoxy resin to some extent. The residual carbon content is 56.5% at 600°C, this indicates that it has good char formation, which may have potential promotion effect on the flame retardancy of the material (18, 19, 20, 21).

In order to study the thermal properties of the material degradation process, the thermal decomposition (TGA under nitrogen) and thermo-oxidate decompositon (TGA under air) of EP/PPXSPB composited are investigated. The TG and DTG curves of EP/PPXSPB composited with different contents of PPXSPB in nitrogen and in air are shown in Figures 4b and 4c. It can be seen that the initial decomposition temperature of the EP/PPXSPB system is slightly lower than that of the pure epoxy resin in the atmosphere of both air and nitrogen. This is perhaps due to the larger steric hindrance of PPXSPB which leads to low crosslink density of the EP/PPXSPB systems (22,23). As is shown in Figures 4b and 4c, the char yields around 480°C under air atmosphere are higher than which under nitrogen atmosphere. However, at 600°C and 800°C, the result is reversed where the char yield is lower in N₂ atmosphere. So we can infer that an oxidative environment promotes the formation of a char layer earlier, it helps the substrate to fully oxidize and degrade at high temperatures.

However, it is noted that the residual carbon yield of the cured epoxy resin increases as phosphorus increases. The residual char yield of EP-1, EP-2, and EP-3 are 23%, 25%, and 27% respectively, while the residual char yield of EP-0 is only 5%. It means that the PPXSPB can improve the carbonization characteristics of the epoxy resin during combustion. As a result, it can be concluded that the phosphorus and nitrogen in the EP/PPXSPB system have a synergistic effect in the condensed phase to promote charring of epoxy resin (24,25).

3.3 Water resistance and flame resistance of EP/PPXSPB composite

The flame retardant and water resistance of the cured epoxy resins are listed in Table 3. As shown in Table 4, the LOI value of pure EP is only 18.0%, with the increase of phosphorus content, the LOI value of EP-1, EP-2 and EP-3 are increased to 28.8%, 31.4%, and 32.2%, respectively, which is due to the overall flame retardancy of the N-P synergistic flame retardant system (26,27). After immersion in water at 70°C for 168 h, the

Figure 4

The TG and DTG curves of PPXSPB (a) and EP/PPXSPB composites (b,c): (a) in air atmosphere, (b) in nitrogen atmosphere, (c) in air atmosphere.

system shows a lower flame retardant migration rate. This is mainly due to the fact that PPXSPB forms a large number of chemical bond connections with epoxy, and the compatibility is greatly improved, thereby preventing the migration of the flame retardant to the surface of the material.

3.4 The cone calorimeter test

The fire hazard of materials is essentially a combination of thermal hazard and the smoke hazard. Therefore, it is necessary to analyze the combustion characteristics of materials that cause thermal hazards and smoke hazards under fire conditions. Using the cone calorimeter is an effective way to measure these indicators. The heat release rate (HRR) is one of the most important effective evaluation parameters for thermal hazards in fire behavior, and the smoke produce rate (SPR) reflects the smoke hazard when the material is burning.

As is shown in Figures 5a-c, the peak heat release rate (PHRR), total heat release (THR) and smoke produce rate (SPR) of EP/PPXSPB composites are all lower than that of EP-0. This means that it can provide longer time for human escapes in the event of a fire. From the trend of HRR curve we can infer that after the flame retardant system begins to burn, the surface of the system rapidly forms a layer of carbon under the catalysis of the produced phosphoric acid, thereby causing a small decrease of HRR at the beginning of combustion.

Figure 5

HRR (a), THR (b), and SPR (c) curves of EP/PPXSPB composites.

SEM and digital images of the char residues after cone calorimeter test are revealed in Figure 6, combined with the residue which is showed in Table 5, it is obvious that after burning, EP-0 has only a small amount of residue, while the char of EP-2 was fully expanded and covered with small bubbles containing non-combustible gas. The surface of the system is heated to form an expanded carbon layer. On the one hand, external radiant heat is difficult to transfer to the inside, reducing the heat absorption value of the internal material and reducing the pyrolysis rate (28). On the other hand, it increases the time required for thermal degradation to release flammable volatiles. At the same time, the non-combustible gas released by the ruptured bubbles, such as NH₃, H₂O, etc., takes away the heat required for combustion, and the carbon layer formed on the surface can retard the contact speed of the material interior with oxygen to some extent, thereby delaying the burning of the materials (29).

3.5 Tensile strength, impact strength and adhesive strength of EP/PPXSPB composite

The tensile strength, impact strength and adhesive strength properties of different simples are shown in Table 6. The tensile strength and impact strength properties of EP/PPXSPB composite are reduced than EP-0. On the one hand, compared with the liquid curing agent used in EP-0, the solid curing agent used in this EP/PPXSPB composite is less likely to be evenly dispersed. During the mixing of the flame retardant system, some of the PPXSPB did not uniformly disperse, resulting in a slight agglomeration phenomenon, so that some PPXSPB cannot bond well with the epoxy resin and cause a slight gap in the system (30). Eventually, if there is external force contact, the system is unevenly stressed, and it is more prone to brittle fracture. On the other hand, steric hindrance is also one of the causes of the deterioration of the mechanical properties of the material. Larger steric hindrance of PPXSPB reduces the reactivity of PPXSPB, and some unreacted PPXSPB acts as a filler, thereby reducing the crosslink density of the system.

Figure 6

SEM and digital images of the char residues after cone calorimeter test: (a) EP-0, (b) EP-0, (c) EP-0 magnification: 200 ×, (d) EP-2, (e) EP-2, (f) EP-2 magnification: 200 ×.

As for the adhesive strength of different simples, the adhesive strength of EP-2 increase 6% than that of EP-0. The main reason for this phenomenon is that the two curing agents have different internal stresses in the curing process. It is known that the use of a liquid curing agent will cause a certain degree of volume shrinkage of the curing system. When the internal stress is large, the adhesive strength will be significantly reduced. In addition, the stress distribution around the adhesive end or adhesive gap is not uniform, the resultant stress also increased the possibility of cracks appear. While PPXSPB is a solid curing agent, PPXSPB can partially act as filler during curing process, so the volume shrinkage of the cured product was reduced (31). At the same time, when the cured material containing the filler subject to the stress, the stress will be uniformly transferred to the surface of the filler particles. The filler takes most of the stress and played a role in the uniform distribution of stress. So there is a higher adhesive strength. However, too much excessive PPXSPB is equivalent to add an excessive amount of filler into the system. Excessive filler will reduce the degree of crosslinking of the system, and result in a decrease in adhesive strength.