Preparation and properties of modi ﬁ ed ammonium polyphosphate synergistic with tris(2-hydroxyethyl) isocynurate for ﬂ ame-retardant LDPE

: UF-SiAPP was prepared by modifying ammonium polyphosphate with the urea formaldehyde resin (UF), tetraethylorthosilicate (TEOS), and vinyltriethoxysi-lane (A-151). Moreover, a new intumescent ﬂ ame retardant (IFR) used for low density polyethylene (LDPE) ﬂ ame retar-dant was obtained by mixing UF-SiAPP with tris(2-hydroxy-ethyl) isocynurate (THEIC). The structure, morphology, and mechanical properties of ﬂ ame retardants and LDPE composites were characterized by Fourier transform infrared spectroscopy, scanning electron microscopy, TG, and universal material testing machines. Furthermore, the ﬂ ame-retardant properties of LDPE composites were comprehensively investigated by limiting oxygen index (LOI), UL-94, and cone calori-metry tests. The results show that UF-SiAPP exhibits excellent hydrophobic properties and good compatibility with LDPE after being modi ﬁ ed by UF, TEOS, and A-151. And the ﬂ ame-retardant properties of LDPE composites are signi ﬁ cantly improved by adding IFR-containing UF-SiAPP. Especially, the LDPE composites attained the LOI value of 30.5% and reached the V-0 level after introducing 20.0 wt% UF-SiAPP and 10.0 wt% THEIC. And the tensile strength reached 13.55 MPa, which was 10.33% higher than that of LDPE/IFR with UF-APP and 13.21% higher than that of LDPE/IFR without THEIC in the same proportion. The smoke density tests demonstrate that the addition of UF-SiAPP not only e ﬀ ectively improves the residual carbon content and thermal stability of LDPE composites but also signi ﬁ cantly reduces smoke emissions.


Introduction
Low density polyethylene (LDPE) is an important polymer material, which is widely used in industry, agriculture, and people's daily life due to its low price, good physicochemical and processing properties, little harm to human beings and the environment, easy reuse, and other advantages (1,2). However, LDPE is flammable and can produce a large number of molten droplets during combustion, which severely limits its application in cables and other products with high flame-retardant requirements (3)(4)(5).
The effective way to improve the flame retardancy of LDPE is to add flame retardants (6), among which halogenbased flame retardants, inorganic flame retardants, and intumescent flame retardants are the most common (7). Halogen-based flame retardants have excellent flame retardancy, but they will produce toxic gases when burning, which will cause great harm to the environment (8)(9)(10)(11). Inorganic flame retardant has the characteristics of good stability, low toxicity, no corrosive gas, and long flame retardancy prescription, but the shortcoming is that it generally requires a large amount of additives, which greatly deteriorates the mechanical properties of LDPE (12,13). Therefore, it is crucial to seek flame retardants having both fire retardancy and excellent mechanical properties.
Intumescent flame-retardant system (IFR) is usually composed of three components: carbon source, acid source, and gas source. Among them, ammonium polyphosphate (APP)/pentitol (PER)/melamine (MH) system is widely used as a polyolefin flame retardant (14)(15)(16). As one of the most important components, APP simultaneously serves as an acid source and a gas source. And it has been the research hotspot of inorganic phosphorus flame retardant due to the advantages of non-toxicity, high phosphorus and nitrogen content, good thermal stability, and low smoke. In the previous work, APP was often used in the form of encapsulation combined with carbon forming agent to flame-retardant polyethylene. For example, Xu et al. (17) modified APP with tetraethyl orthosilicate (TEOS) and 3-(trimethoxysilyl) propyl methacrylate (KH570) and mixed it with dipentaerythritol (DPER) to prepare flame-retardant LDPE. The results revealed that when 30.0 wt% IFR containing Si-MAPP/DPER was added, the limiting oxygen index (LOI) reached 26.8% and the tensile strength increased to 3.30 MPa. Kang et al. (18) implemented the modification by mixing APP coated with melamine (MF-APP), silane (GW-APP), and epoxy resin (MC-APP) with Novolac and ME. The results showed that GW-APP had the best synergistic effect with Novolac and ME, and the POM/ GW-APP composites had an LOI value of 34.0% and reached the V-0 level. Huang et al. (19) encapsulated APP with sodium silicate and melamine formaldehyde resin (MCSF) and used it with polyamide 6 (PA-6) to prepare flame-retardant polypropylene. The results showed that when MCSF and PA-6 were combined with 20.0 and 10.0 wt% respectively, the LOI value of PA-6 composites was 28.6% and the UL-94 reached the V-0 level. Gao et al. (20) encapsulated APP in situ with boron-modified phenolic resin. And the relevant results showed that APP microcapsules had good thermal stability and high carbon residues, which can be used as an inherent flame retardant. Liu and Lv (21) synthesized a halogen-free flame retardant (ODOPM-CYC) based on DOPO (9,10-dihydro-9-oxa-10-phosphino-anthracene-10-oxide) and doped it into rigid polyurethane foam (RPUF). The results showed that the addition of ODOPM-CYC significantly improved the flame-retardant properties of RPUF. The flame-retardant RPUF obtained an ultimate oxygen index (LOI) value of 26% and achieved a UL-94 V-0 rating at a phosphorus content of 3 wt%. Li et al. (22) synthesized a phosphorus-nitrogen flame-retardant curing agent poly(p-phenylene terephthalamide spirocyclic pentaerythritol bisphosphonate) (PPXSPB).
The results showed that with the increase in phosphorus content, the oxygen index and residual carbon of the system increased significantly, and the exothermic rate decreased gradually, which is important for retarding the occurrence of fire. Xia and Wang (23) introduced expanded vermiculite (EV) into rigid polyurethane (RPU) foam and synthesized melamine phenyl phosphate (MPP) and introduced it into RPU/EV composite foam resulting in a significant increase in the oxygen index and residual carbon of the system and a gradual decrease in the exothermic rate. Yu et al. (24) synthesized a reactive flame-retardant hexa(ethylene oxide)-cyclotriphosphonitrile (HCCP-EP) and used it to improve the flame retardancy of PLA. The results showed that the ultimate oxygen index of PLA increased from 19.5% to 27.3% with the addition of 3 wt% HCCP-EP, and the PLA/HCCP-EP blend achieved Underwriters Laboratories (UL)-94 V-0 rating. Yang and Xiao (25) synthesized diphenyl allylaminophosphate (DPCA) and N-allyl-P,p-diphenyl hypophosphoramidate (DCA) using the chloride reaction and introduced them into EP for the manufacture of EP composites. Combustion tests showed that the addition of 5 wt% DPCA or 5 wt% DCA to EP resulted in abnormal limited oxygen index (LOI) values (27.1% or 31.6%).
Vinyl triethoxysilane (A-151) and tetraethyl orthosilicate (TEOS) are composed of inorganic skeleton formed by two-dimensional Si-O short chain and organic substituent, which can enhance the interface compatibility between inorganic filler and organic material, thus strengthening the adhesion between the two (26)(27)(28)(29). Urea formaldehyde resin (UF) is widely used in wood adhesives due to its advantages of low cost, insulation, weak acid resistance, and alkali resistance (30). Especially, UF is a kind of hydroxyl polymer with high carbon and nitrogen content, which is an important guarantee of flame retardancy. Based on this, we proposed a novel idea to prepare an IFR flame retardant system that was composed of micro-encapsulated core-shell flame retardant (UF-SiAPP) alone or mixed with tris (2-hydroxyethyl) isocynurate (THEIC). Thereinto, UF-SiAPP was synthesized by encapsulating UF resin and TEOS through in situ polymerization technology and then modified by A-151. Finally, a new halogen-free IFR LDPE was prepared by melt blending using UF-SiAPP as the acid source and gas source, and THEIC as the carbon source in the IFR system. Fourier transform infrared spectroscopy (FTIR), X-ray photoelectron spectroscopy (XPS), and scanning electron microscopy (SEM) were employed to characterize the structure, flame retardancy, and thermal stability of the polymer composites. This work provides an effective method for preparing LDPE composites with high intumescent flame retardancy and strong mechanical strength.

Synthesis of urea-formaldehyde resin prepolymers
About 5.0 g of urea and 10.0 g of formaldehyde solution were first placed in a 100 mL three-necked flask and then 50 mL of water was added. Next, the pH of the mixture was adjusted to 8.5 with a 10% solution of triethanolamine. The mixture was then immediately mechanically stirred in an oil bath at 80°C until it became clear. Finally, the clear prepolymer solution was cooled to room temperature to obtain the urea-formaldehyde resin prepolymer (UF).

Synthesis of UF-SiAPP
A continuous sol-gel method was employed to prepare APP encapsulated by A-151 and SiO 2 (Si-MAPP). The procedure was as follows: 50.0 g APP, 100 mL ethanol, and 50 mL deionized water were added to a 500 mL three-necked round-bottom flask, and the pH was adjusted to 10.0 with triethanolamine. Then, 1.0 g OP-10 was added, and the temperature was increased to 40°C. The reaction was stirred at 500 rpm for 10 min, and 12.5 g TEOS was added to keep stirring for 4 h. Then, 2.5 g A-151 was added, and the reaction was carried out at 60°C for 1 h. Next, the UF prepolymer solution was added dropwise to the suspension, and the pH was adjusted to 5.0 with acetic acid, and then the reaction was carried out at 80°C for 2 h. Finally, the mixture was cooled to room temperature, filtered, and washed three times with ethanol, and dried under vacuum at 100°C to obtain a white solid powder named UF-SiAPP. As a control, APP encapsulated in urea-formaldehyde resin (UF-APP) was also prepared ( Figure 1).

Preparation of flame-retardant LDPE composites
According to the formulation of IFR LDPE in Table 1, different samples of IFR LDPE were prepared by mixing APP, UF-APP, UF-SiAPP, THEICR, and LDPE in a mixer at 20 rpm for 15 min at 165°C. And then, the samples were hot pressed on a plate vulcanizer at 150°C for 10 min to obtain standard strips for testing.

Characterization
Field emission scanning electron microscopy analysis: A SIGMA-500 Field Emission Scanning Electron Microscope (Zeiss, Germany) was used to observe the surface morphology of the samples. Prior to analysis, the samples were dried and placed on a sample stage with conductive adhesive and gold sputtering treated under a vacuum. FTIR analysis: The characterization was carried out using a WQF-520A Fourier Transform Infrared Spectrometer (Riley, North Branch, China). The tests were conducted at room temperature, mixing 50 mg KBr powder and an appropriate amount of sample by grinding and pressing method, with a scanning wave number range of 400-4,000 cm −1 , 32 scans, and a resolution of 4 cm −1 .
XPS analysis: The elemental composition of the flame-retardant particles was determined using an Escalab Xi + XPS (Thermo Fisher, USA).
Thermogravimetric analysis (TGA): The samples were tested using a TA Q-50 thermogravimetric analyzer (USA TA Instruments). Characterization was performed under a nitrogen atmosphere with a test temperature from 30°C to 800°C and a heating rate of 10°C·min −1 .
Contact angle test: The samples were tested using a JC2000D2 contact angle measuring instrument (Shanghai Zhongchen Digital Technology Equipment Co., Ltd., China).
The tensile strength and elongation at break were tested according to GB/T1040.2-2006, with a sample size of 2.0 mm × 4.0 mm (dumbbell shape) and a running speed of 5.0 mm·min −1 .
The LOI test was conducted using a 5801A fully automatic oxygen index tester (Suzhou Yangyi Volch Testing Technology Co., Ltd., China) according to the American National Standard ASTM D2863-97, with a sample size of 120.0 mm × 6.5 mm × 3.2 mm; the UL-94 vertical combustion test was conducted using a JT-UL94 model (Dongguan Jinte Instruments Co., Ltd., China), according to the American National Standard ASTM D635-77 with a sample size of 125.0 mm × 12.5 mm × 3.2 mm; the conical calorimetry test was carried out by using an FTT-iCone mini conical calorimeter (Occidental Instruments China Co., Ltd.) with a sample size of 100.0 mm × 100.0 mm × 4.0 mm and the sample was exposed to a radiation cone (45 kW·m −2 ), and the heat release rate (HRR), total heat release (THR), total smoke production (TSP), and smoke temperature (ST) were recorded.

Characterization of modified APP
The APP, UF-APP, UF-SiAPP, before and after modification were characterized by SEM, FTIR, XRD, and TGA, respectively.
It can be seen from Figure 2 that the surface of the unmodified APP is slightly rough and has obviously granular protrusions. However, the surface of the UF-APP becomes significantly smoother after being encapsulated by UF or co-encapsulated by UF and TEOS. In particular, a small number of silicon spheres can be found attached to the surface of UF-SiMAPP after being enclosed by TEOS. Figure 3 shows the hydrophobicity of APP, UF-APP, and UF-SiAPP. The results imply that the unmodified APP and UF microencapsulated UF-APP are significantly hydrophilic, while the contact angle of UF-SiAPP is 135°, indicating that the hydrophobicity of UF-SiAPP after encapsulation with UF resin and TEOS and modified with A-151 is significantly increased. This is due to the fact that the surfaces of APP and UF-SiAPP contain a large number of + NH 4 and OH − hydrophilic groups, whereas after being modified with TEOS and A-151, the surfaces exhibit a lot of hydrophobic silane groups and C]C double bonds; thus, the hydrophobicity is significantly enhanced, and the flame retardant has better compatibility with the LDPE matrix. Figure 4 shows the FTIR spectra of APP, UF-APP, and UF-SiAPP. Among them, the characteristic peak at 3,106 cm −1 is associated with the stretching vibration of N-H, the peak at 1,020 cm −1 is attributed to the stretching vibration of PO 2 and PO 3 , and the peak at 1,253 and 1,079 cm −1 correspond to the symmetric stretching vibration of P]O and P-O, respectively. Furthermore, the asymmetric stretching vibration of P-O and P-O-P is illustrated by the peak at 886 and 802 cm −1 .  The thermal degradation behavior of APP, UF-APP, and UF-SiAPP in the N 2 atmosphere was analyzed by TGA. It can be seen in Figure 5 that the residual weight of UF-APP and UF-SiAPP increased compared with that of APP. The reason is that APP modified by TEOS and A-151 is encapsulated by silica gel, while SiO 2 cannot decompose during heating,   thus contributing to the increase of residual weight. According to the TGA results, the introduction of TEOS and A-151 can effectively inhibit the thermal decomposition of APP. Figure 6 shows the surface elemental analysis results of APP, UF-APP, and UF-SiAPP. The peaks located at 130.51 and 187.14 eV are attributed to P2p and P2s of APP, respectively. In UF-APP and UF-SiAPP, the intensity of these peaks is noticeably weaker, proving that APP is properly encapsulated. In contrast, the new peak at 102.05 eV is ascribed to the Si2p of TEOS and A-151. These changes in the above spectral peaks are mainly caused by the outer shell layer formed on the APP surface by UF, TEOS, and A-151. Furthermore, the elemental analysis of the three samples in Table 2 reveals that UF-SiAPP has large levels of C, N, and O as well as 1.67% Si, all of which are necessary to improve the flame retardancy of the composite material.

Flame-retardant properties of the LDPE composites
To evaluate the flammability of LDPE composites, the LOI and UL-94 vertical combustion tests were carried out. As can be seen from Table 3, pure LDPE is flammable, with many molten droplets during combustion, an LOI of 17.0% and no ranking in the UL-94 test. However, IFR can promote the formation of a carbon foam layer on the LDPE surface during combustion, which plays the role of heat insulation, oxygen barrier, smoke suppression, and antidripping, with a substantially higher LOI value. When the APP content reaches 30.0 wt%, the LDPE composites have a LOI value of 19.8, but still cannot achieve a V-0 level in the UL-94 test. This should be attributed to the poor carbon formation capacity due to the lack of carbonforming agents in the IFR system. At the same content, the LOI of LDPE with UF-APP or UF-SiAPP is higher than that of APP and increases from 19.8% to 21.3% and 21.5%, respectively. Especially, the LDPE composites with UF-SiAPP attain a V-1 level in the UL-94 test, which is because the UF resin and silica layer cover the APP surface and play important roles: the former produces NH 3 was released at high temperatures to dilute the combustible gases and the oxygen in the air, and the latter promotes the formation of the carbon layer acting as an oxygen inhibitor and temperature barrier. Therefore, UF-APP and UF-SiAPP effectively improve the flame-retardant properties of the LDPE composites.
To further enhance the flame retardancy of the LDPE composites, APP and THEIC were used in combination to    prepare the flame-retardant LDPE. As can be seen in Table 3, the LOI of the LDPE composites is further increased after adding THEIC. When the content of UF-SiAPP and THEIC is 20.0 and 10.0 wt%, respectively, the LDPE composites eventually have an LOI of 30.5 and reach a V-0 level in the UL-94 test. The reason is that the phosphoric acid decomposed by APP during the combustion process can promote the rapid formation of carbon layer by THEIC, thus improving the LOI of the LDPE composites, showing the synergistic effect between APP and THEIC, which plays a significant role in improving the flame-retardant properties of the LDPE composites. Figure 7 shows the tensile strength and elongation at the break of the LDPE composites. Compared to pure LDPE, the mechanical properties of the LDPE composites have variedly diminished. This is due to the poor compatibility of IFR with the polymer, which causes a deterioration in the mechanical properties of the LDPE composites. Compared to APP, the tensile strength of UF-APP and UF-SiMAPP modified LDPE shows a slightly enhanced, increasing from 11.45 to 12.65 MPa, and 11.76 MPa, respectively. And the elongation at break increases from 19.0% to 25.0%, and 24.0%, respectively. This is because UF-APP and UF-SiAPP modified by UF, TEOS, and A-151 are more compatible with the polymer and their dispersion in the LDPE matrix is better than that of APP. With the introduction of THEIC, the tensile strength of the LDPE composites is further improved. Compared with the APP/THEIC/LDPE composites, the tensile strengths of both the UF-APP/THEIC/LDPE and UF-SiAPP/THEIC/LDPE composites are decreased from the previous 14.28 to 12.15 MPa, and 13.55 MPa, respectively. And the elongation at break is decreased from 270% to 200%, and 250%, respectively, which is due to the antagonistic effect of UF and THEIC.t Figure 8 shows the cross-sectional morphology of the LDPE composite. It can be noticed that the size of the dispersed phase (APP) in LDPE is larger, and the two-phase interface is obvious, which is due to the poor compatibility between LDPE and APP. Compared to LDPE1, the components of LDPE2 and LDPE3 show better compatibility with each other, and the two-phase interface becomes more blurred, indicating a better modification by UF and A-151. In addition, the compatibility between the components of LDPE5 and LDPE6 becomes slightly worse again compared to LDPE4, with the two-phase interface becoming clearer, indicating that the introduction of THEIC reduces the interaction between LDPE and APP. These phenomena are consistent with the results of flame retardancy and mechanical properties.

Thermal degradation behavior of the LDPE composites
For LDPE0, it is almost completely degraded at high temperatures, just as shown in Figure 9a. However, when flame retardant is added, the carbon residue of the LDPE composites all increases. LDPE1 is an APP/LDPE composite, which formed significantly more carbon residues than LDPE at high temperatures. LDPE2 is a UF-APP/LDPE composite, in which APP has been modified by UF and microencapsulated, and it has significantly fewer carbon residues than LDPE1, which is due to the large amount of NH 3 produced by UF at high temperatures. LDPE3 is a UF-SiAPP/LDPE composite and has higher carbon residues formed at high temperatures compared to LDPE2 because SiO 2 is not decomposed during combustion. LDPE4, LDPE5, and LDPE6 are the APP/THEIC/LDPE, UF-APP/THEIC/LDPE, and UF-SiAPP/THEIC/LDPE composites, respectively. The carbon residues of LDPE4 and LDPE5 further decrease following the addition of THEIC. However, LDPE6 contains the highest amount of residual carbon formed at high temperatures when compared to LDPE4 and LDPE5. As can be seen from Figure 9b, the thermal degradation of LDPE0 undergoes only one stage in the N 2 environment, whereas the LDPE composites with flame-retardant addition all experience a three-step degradation process. Compared to LDPE0, the matrix decomposition temperatures of LDPE1, LDPE2, and LDPE3 are all pushed back by 30°C, and the maximum thermal decomposition rates are all delayed by 40°C. In particular, the decomposition temperatures of LDPE5 and LDPE6 are pushed back more  significantly than that of LDPE4. The reason is that APP is encapsulated by a silica gel shell layer containing UF resin and vinyl silicon gel, and SiO 2 remains stable and less prone to decomposition at high temperatures.

Cone calorimetric testing of the LDPE composites
The combustion performance of the LDPE composites was investigated using a cone calorimeter. Figure 10a-d show the HRR, THR, TSP, and ST values versus time for LDPE0, LDPE1, LDPE2, LDPE3, LDPE4, LDPE5, and LDPE6 at a heat flow rate of 50 kW·m −2 . As can be seen in Figure 10a, pure LDPE burns rapidly and its HRR quickly reaches a peak of 892 kW·m −2 . Compared to LDPE0, the PHRR (peak of HRR) of the LDPE composites is decreased when 30.0 wt% of APP, UF-APP, or UF-SiAPP is introduced. To be specific, the PHRR will decrease from 395 to 327 kW·m −2 for both LDPE2 and LDPE3 compared to LDPE1. This is because the LDPE composites can form a silica layer on the surface during combustion when it is mixed with modified APP, thus reducing the HRR of the composites. After the introduction of THEIC, the PHRR of LDPE5 and LDPE6 even decreases from 216 to 195 kW·m −2 , and 129 kW·m −2 , respectively, compared to LDPE4. The reason is that the phosphoric acid decomposed by APP during combustion can promote the rapid formation of the carbon layer by THEIC. A similar trend of THR is observed as shown in Figure 10b.  LDPE1, the TSP of LDPE3 decreases from 14.7 to 9.7 m 2 . And when THEIC is introduced, that of LDPE4 decreases to 12.6 m 2 . Further, by introducing UF, TEOS, and A-151 modified APP, that of LDPE5 and LDPE6 can decrease by 5.2 and 5.9 m 2 , respectively. Figure 10d gives the ST value of the LDPE composites as a function of time during combustion. Compared to pure LDPE, all ST values of the composites are decreased. It should be noted that the introduction of UF, TEOS, and A-151 will further reduce the ST value. And in terms of the ST alone, the flame-retardant capacity of UF-SiAPP is more effective than that of UF-APP.
The residual carbon base of LDPE6 after conical calorimetric testing was explored using SEM. The results are shown in Figure 11. When UF-SiAPP and THEIC were added to the LDPE, the char layer obtained by combustion was very compact. There is a clear thin layer and no holes on the surface. The main reason for this phenomenon is that the flame-retardant releases NH 3 , CO 2 , N 2 , and other nonflammable gases at high temperatures to expand the surface of the LDPE substrate, and the phosphoric acid and hypophosphoric acid produced by the decomposition of APP promote the dehydration into char and the silica gel covered by APP surface acts as a physical barrier to protect the substrate. It effectively prevents the exchange of oxygen and heat, restricts the diffusion of organic substances, and achieves a good flame-retardant effect. Table 4 and Figure 12 show the elemental content data, elemental diagrams, and EDS results of the carbon residue of LDPE6 after cone calorimetric testing. As shown in the figure, the char residue of LDPE6 after combustion contains C, O, Si, and P with 13.08, 51.13, 8.04, and 27.75 wt%, respectively. These elements were uniformly distributed on the surface of the LDPE6 carbon residue, and it is possible that UF-SiAPP released non-combustible gases such as NH 3 and CO 2 at high temperatures, so the content of C elements in the LDPE6 carbon residue decreased sharply and the content of N elements was 0 wt%. The presence of elements such as P and Si in the carbon residue once again proved that the flame retardant has a cohesive phase flame-retardant effect in LDOPE, and the phosphoric acid

Flame-retardant mechanism
Based on the above experimental results, a possible flameretardant mechanism was proposed, as shown in Figure 13. APP can produce NH 3 during the combustion of the LDPE composites and form a highly condensed polyphosphoric acid, which provides an acid source for the flame-retardant system and induces the carbonization of THEIC and partially decomposed LDPE to form foamed char, so as to achieve the flame retardancy of condensed phase. During the process, the volatile gases (NH 3 , N 2 ) are acted as expansive agents for the formed char, which can play a gas phase flame-retardant role. The introduction of UF, TEOS, and A-151 into APP will not only supplement APP with a large amount of N 2 and produce enough NH 3 to dilute external combustible gases and O 2 in the air during combustion but will also cover the surface with a dense silica layer. The silica layer not only prevents or delays small molecules of combustible matter from entering the outside of the composites and coming into contact with the air to play a flame-retardant role but also promotes the formation of a char layer and increases its thickness.

Conclusions
The intumescent flame-retardant LDPE composites were prepared by designing a new IFR based on UF-SiAPP and THEIC. And the synergistic effects of UF-SiAPP on flame retardancy, mechanical properties, and thermal properties were investigated. The highly efficient encapsulated ammonium polyphosphate (UF-SiAPP) containing "silicon sourceair source-acid source" was prepared by continuous sol-gel method, and a thermally stable, water-resistant, and moisture-absorbent flame retardant was prepared. The decomposition of APP was delayed by the coating of SiAPP with TEOS and A-151. By introducing UF resin coated with SiAPP, the APP has both condensed phase and gas phase flame-retardant mechanism. By compounding UF-SiAPP with THEIC, the phosphoric acid and hypophosphoric acid produced by the decomposition of UF-SiAPP at a high temperature can catalyze the production of THEIC carbon layer, thus increasing the decomposition temperature of the LDPE composite matrix. The results show that APP is successfully encapsulated by the UF resin and silica, and the modified UF-SiAPP is more compatible with LDPE and can be ideally dispersed in the matrix, thus improving the mechanical properties of the flame-retardant LDPE composites. Compared with the conventional IFR (APP and THEIC), the new IFR (UF-SiAPP and THEIC) has excellent flame-retardant properties. To be specific, when the amount of IFR is 30.0 wt% (20.0 wt% UF-SiAPP and 10.0 wt% THEIC), the LOI, tensile strength, and elongation at break of the LDPE composites can reach 30.5%, 13.55 MPa, and 250%, respectively. Furthermore, the introduction of UF resin and silica significantly improves char production and densification, thus effectively reducing the TSP and HRR values of the flame-retardant composites. Author contributions: Tingxuan Dong: methodology, data curation, writingoriginal draft; Guxia Wang: supervision, project administration, writingreview and editing; Zhaoshuai Li: formal analysis; Dan Li: writingreview and editing; Yuan Liu: resources; Peng Zhou: resources; Shengwei Guo: methodology, conceptualization, funding acquisition.

Conflict of interest:
The authors state no conflict of interest.
Data availability statement: All data included in this study are available upon request by contact with the corresponding author.