Novel P/Si based nanoparticles for durable flame retardant application on cotton

Cotton fabric has a wide application due to its hygroscopicity, air permeability, and large production of cotton fiber used to make the fabric. However, cotton materials are a safety hazard during its application because of flammability (limiting oxygen index is about 18%). In order to improve the flame retardancy of cotton fibers and reduce the damage of its mechanical properties, novel P/Si based flame retardant (PFR) nanoparticles were synthesized by one-step radical polymerization. Vinyl phosphoric acid and tetramethyl divinyl disiloxane were introduced into the nanoparticles. The structure, morphology and thermal stability of PFR was characterized by fourier transform infrared spectroscopy (FT-IR), field emission scanning electron microscopy (FE-SEM), thermogravimetric analysis test (TGA). Durable flame retardant cotton fibers were prepared by dip-coating and plasma induced crosslinking methods. Micro-calorimeter (MCC) characterization showed that the peak of heat release rate (pHRR) and the total heat release (THR) were reduced by 47.3% and 29.8% for modified cotton fibers compared with pure cotton fibers. Limiting oxygen index (LOI) of modified cotton fibers was increased to 27%. The residue carbon of modified cotton fibers was 19.0% at 700 °C, while the value of pure cotton fibers was 3.0%. Besides, durability of the modified cotton fibers was approved by cyclic washing test. In addition, flame retardant mechanism was revealed by collecting and analyzing condensed and gaseous pyrolysis products. The data of FE-SEM for residue carbon, FT-IR spectra of products at different pyrolysis temperatures and pyrolysis gas chromatography mass spectrometry (Py-GC–MS) showed that PFR was a synergistic flame retardant contained barrier and quenching effecting applied on cotton materials. Novel phosphorus-silicon based nanoparticles were synthesized by one-step radical polymerization and applied to improve the flame retardant of cotton materials by dip-coating and plasma induced crosslinking.


Introduction
Cotton fibers, used to make cotton fabric that is widely used because of its hygroscopicity and its air permeability (Zhang et al. 2021;Mathangadeera et al. 2020;Schumacher et al. 2020). However, the limiting oxygen index (LOI) of pure cotton materials is about 18%, making cotton fibers a safety hazard during use (Cheema et al. 2013). Therefore, it has practical significance to improve the flame-retardant properties of cotton materials.
Nowadays, the methods of endowing function to cotton materials are mainly focused on surface modification, such as coating (Li et al. 2011;Alongi et al. 2013;Pan et al. 2017), layer by layer, grafting and so on (Alongi et al. 2011;Alongi et al. 2014;Indraneel et al. 2017;Li et al. 2019a, b;Wang et al. 2020). Surface modification is an effective and convenient method to offer flame-retardant properties to combustible materials (Kim et al. 2014). Firstly, most flame retardants are concentrated on the surface of the protected textiles to achieve the highest protection. Secondly, flame-retardant layers do not alter the mechanical properties of the protected textiles. Thirdly, flame-retardant layers can be integrated with functional components purposely to achieve multiple functions (Chen et al. 2015). Compared with coating and layer-by-layer self-assembly, grafting method has stronger durability via connecting the flame retardant with the substrate through chemical bonds. Plasma has been explored widely to graft cotton fabric by inducing crosslinking. Tsafack et al. (2008) investigated the simultaneous grafting and polymerization of flame retardant monomers on cotton fabric induced by argon plasma. Edwards et al. (2012) prepared two new types of amino phosphate monomers and applied them to cotton by atmospheric pressure plasma treatment. Shahidi et al. (2014) utilized low temperature plasma technology to endow cotton fabric with anti-UV and flame retardant. Plasma is an effective path to modify surface of cotton material.
Besides, flame retardant as a functional additive to improve the flame retardancy of polymers was attracting more attention. Both industrial and academic researchers have focused their efforts on the design and development of chemicals to prevent the combustion or delay the spread of flame after ignition since the 1960s (Horrocks et al. 2012). Among the synergistic flame retardant, phosphorus-based compounds represent a suitable alternative to halogenbased flame retardants (Alongi et al. 2013). Hou et al. (2018) modified two-dimensional Co-based metal organic framework by conjugating DOPO to prepare organic and inorganic hybrid phosphorus flame retardant. Pethsangave et al. (2019) synthesized a phosphorus functionalized polymer-based graphene composite as an efficient flame retardant. Effective flame retardancy depends on the interaction between the flame retardant and the matrix strongly, as well as the structure-property relationship between them in process of thermal decomposition. The mode of action can be classified into condensed and gas-phase mechanisms generally and many successful phosphorus flame retardants include both of them (Rabe et al. 2017). Yang et al. (2019) established P and N flame retardant to reduce the fire hazard of epoxy resin. Balabanovich et al. (2009) prepared the P and Si system to achieve synergies and enhance flame retardant of polyester. Sun et al. (2021) synthesized a P/N based flame retardant to reduce the flammability of cotton. Synergetic flame retardant has become the dominant topic in the field of flame retardant. Ecofriendly and sustainability play guiding roles in novel flame retardant system (FRs) (Velencoso et al. 2018). In order to improve the compatibility between flame retardant and matrix, and to reduce damage to mechanical properties, the FRs are fabricated into nano scale. Jiang et al. (2018) explored mesoporous silica as the carrier for flame retardant modification to improve the compatibility of epoxy resin. Wicklein et al. (2015) prepared heat insulation and flame retardant light anisotropic foam based on nanoscale cellulose and graphene oxide.
In this research, novel nanoparticle flame retardant containing phosphorous and silicon elements was synthesized by one-step reaction of free radical polymerization to achieve synergistic efficiency. Vinyl phosphoric acid (VA) and tetramethyl divinyl disiloxane (THOD) were introduced during synthesis process. Plasma induced crosslinking method was utilized to connect nano flame retardant and cotton fibers to realize the intrinsic flame retardancy. Multiple flame retardant performance indicators were evaluated. In addition, durable flame retardancy was tested and flame retardant mechanism was revealed. Besides, the effect of nanoparticles on the mechanical properties of cotton fibers was also investigated.

Synthesis of flame retardant containing P and Si
The flame retardant was synthesized by radical polymerization as shown in scheme 1. Typically, TDOH (11.16 g, 0.06 mol) and VA (12.96 g, 0.12 mol) were dissolved in distilled water (10 mL) and placed in a three-neck flask, using AIBA (0.50 g, 0.0018 mol) as catalyzer. The mixture was heated up to 70°C with mechanical stirring for 2 h. Then, the nanoparticles were obtained by centrifugation and freeze-drying, designating as PFR. (yield 83.1%).

Preparation of flame retardant modified cotton fibers
The preparation of flame retardant cotton fibers involved two steps: dip-coating of PFR precursor and radio frequency capacitively-coupled plasma (RF-CCP) treatment (Malshe et al. 2012). In detail, the PFR (0.40 g) was ultrasonic dispersed in ethanol solution (50 mL). Cotton fibers (2.00 g) were placed in the solution and dipped for 1 h, dried in vacuum oven at 70°C. Coated cotton fibers were treated by plasma for 60 s in RF-CCP reactor (AP600, Nordson-March, USA) with 13.56 MHz, the RF power was 50 W and plate distance was 10.16 cm. The reactor was pumped down to a base pressure of 20 mTorr and then the argon gas was introduced into the chamber to reach 200 mTorr for plasma generation. The treated fibers were achieved and designated as pCo-PFR after ultrasonic washing for 15 min, drying in an oven at 70°C. The weight gain rate (A) was calculated by weighting method of Eq. (1). In addition, cotton fibers of plasma treated without dipping in the PFR dispersion were taken as contrast sample, and designated as pCo. Pure cotton fibers were conducted as control sample and named as Co. The weight gain rate of pCo and pCo-PFR were 0% ± 0.02% and 14% ± 0.2%, respectively.
In this equation, W f is the weight of the modified cotton fibers and W i is the weight of the pure cotton fibers.

Measurements
Field Emission Scanning Electron Microscope (FE-SEM) images were obtained with Hitachi SU8010 to observe the morphology of PFR and cotton fibers at ambient temperature. In details, the cotton fibers before and after modification were observed. All the samples were sputter-coated with a conductive gold layer before FE-SEM testing.
Element analysis (EA) and inductive coupled plasma emission spectrometer (ICP) experiments were performed by using an Elmentar Vario EL III-EA and Prodigy-ICP to determine the element composition of nanoparticles. The samples were diluted in deionized water to 20 mL suspension for testing.
Fourier Transform Infrared (FTIR) spectra of PFR was recorded in the 4000-500 cm -1 spectral region via a Thermo Fisher Nicolet 6700 under the resolution of 4 cm -1 .
Thermal gravimetric analysis (TGA) was performed to evaluate the thermal and thermo-oxidation decomposition of PFR, pure and grafted cotton fibers via German NETZSCHTG209F1 instrument. About 5 mg samples were heated from 30°C to 700°C under continue nitrogen or air atmosphere with a heating rate of 10°C/min. X-ray photoelectron spectra (XPS) were recorded on Thermo ESCALAB 250XI to test the contents of silicon and phosphorus elements on the surfaces of modified cotton fibers.
Limiting oxygen index (LOI) was tested by Qingdao zr-1 according to the testing method for flame retardant property of viscose staple fibers-oxygen index (FZ/T 50,016-2011). The cotton fibers (20 g) were spun (linear density B 60 tex) into yarn. The cotton yarn (0.9 g) was knotted and twisted in 120 twists, and folded in half to form a strand as test spline. The dimensions of specimen were 75 mm. The distance from the top to the clamping position was 70 mm. The flame application time was 10 s, with an interval of 3 s removed the igniter to observe and record the combustion behavior. The samples were pre-humidified according to GB/T 6529 standard. Five specimens were tested for each composition. The temperature and humidity of the experimental environment was 23 ± 2°C and 45 ± 5%, respectively. If the sample extinguished immediately after ignition, the smoldering Scheme 1 Synthesis route of phosphorus and silicon nanoparticles for flame retardant (PFR) time didn't exceed 180 s, or the combustion length was less than 50 mm, it passed the test.
The microscale combustion calorimeter (MCC) (Fire Testing Technology Ltd., UK) was employed to evaluate the combustion behavior. The sample (5 mg) was heated from room temperature to 700°C at a heating rate of 1°C/s according to ASTM D7309-2007. Six specimens were tested for each composition.
The horizontal combustion test was conducted in UL-94 horizontal flame chamber according to ANSI/ UL94-1985. The samples were woven into 127.0 mm 9 12.7 mm 9 3.0 mm as shown in Fig. S1, five specimens for each composition. The sample was passed the test when the combustion speed was B 38.1 mm/min or the combustion stopped before reaching the 102 mm.
Circle washing test of modified fibers was carried out according to procedure 5 N of standard ISO 6330: 2012, MOD, hang to dry mode (Han et al. 2019).
Pyrolysis Gas Chromatography mass spectrometry (Py-GC-MS) test was conducted to analysis modified cotton fiber pyrolysis behaviors and products by using GCMS-QP 2010 from Shimadzu Corporation of Japan and a Frontier single-point cracker in Japan. The pyrolysis temperature was 380°C, and the pyrolysis products were separated and identified by a coupled GCMS-QP2010 gas chromatograph mass spectrometer. Mass spectrometry conditions: EI source, mass-tocharge ratio m/z range 30 * 600, searched using Nist107 mass spectrometry library.
Raman spectroscopies of residue chars after thermal gravimetric analysis test were performed on an inVia-Reflex Raman spectrometer (UK) with a 532 nm laser line.
The mechanical properties of the samples were determined by tensile measurements with a YG001B testing machine according to standard method ASTM D5035-2011. The tensile speed was 10 mm/min. The pretension was 0.2 cN/dtex. The spacing length of the upper and lower grippers was 10 mm. Each sample was tested 50 times.

Results and discussion
Formation and morphology of PFR Radical polymerization reaction was employed as a convenient and simple synthesis method to obtain nanoparticles as shown in scheme 1. The morphology of nanoparticles PFR was obtained by FE-SEM. In Fig. 1a, PFR exhibited nano agglomeration and the particle size was about 25 * 30 nm. The element and content were determined by element analysis (EA) and inductive coupled plasma emission spectrometer (ICP) as shown in Table 1. The chemical structures of PFR were estimated via FTIR (Fig. 1b). In the FTIR spectrum of PFR, the symmetric stretching vibration peaks of=C-H at about 3051 cm -1 and 2963 cm -1 disappeared. The symmetric stretching vibration peaks at about 1593 cm -1 and 1610 cm -1 attributed to C=C also disappeared. Besides, the stretching vibration peak of Si-C or P-C appeared at 840 * 670 cm -1 (Zope et al. 2017). The stretching vibration peak of Si-O appeared in the wavenumber range of 1090 * 920 cm -1 (Guo et al. 2020;Jian et al. 2020). The data indicated the successful synthesis of PFR.

Thermal stability of PFR
The thermal stability of PFR was analyzed by thermogravimetric analysis (TGA) under continuous nitrogen or air atmosphere from 30°C to 700°C. The results ( Fig. 1c and d) showed that the maximum decomposing rate (T max ) of PFR occurred at about 480°C in nitrogen and 240°C in air. This might be due to the conversion of phosphorus into phosphate and phosphonate compounds was easier in aerobic environment (Lazar et al. 2020). The residue weight percent (CR) of PFR was 70.2% in nitrogen at 700°C and 30.1% in atmosphere environment. All the data certificated PFR had high thermal stability.
Morphology and structure of flame retardant modified cotton fibers Cotton fibers were modified with the PFR to reduce flammable property by dip-coating and radio frequency capacitively-coupled plasma (RF-CCP) treatment as shown in graphic abstract. Seen from the images of FE-SEM (Fig. 2), it could be observed that there were visible grooves and natural spirals on the surface of raw cotton fibers (Co), while plasma without dip-coating treated cotton fibers (pCo) presented obvious grooves and scratches compared with Co. The dip-coating and plasma induced crosslinking cotton fibers (pCo-PFR) apperanced attached particles on the surface compared with Co and pCo.
In order to further confirm the crosslinking reaction of PFR and cotton fibers, the FT-IR and XPS of Co and modified cotton fibers were tested. As shown in Fig. 3a, compared with the FT-IR spectrum of Co and pCo, the FT-IR spectrum of pCo-PFR exhibited obvious new absorption peak at about 1260 cm -1 and 799 cm -1 , which mainly due to the stretching vibration peak of P=O and Si-C (Zope et al. 2017;Guo et al. 2020). There was no significant difference between the FT-IR spectrum of Co and pCo, except for the peak area of hydroxyl group was reduced in FT-IR spectrum of pCo. The above results indicated that onestep of plasma induced crosslinking had no obvious effect on the fiber structure or elements. New structure and element units appeared on the fiber surface after the process of dip-coating and plasma induced crosslinking. In addition, the XPS spectra corresponded to the conclusion of the FT-IR. In Fig. 3b, compared with XPS spectra of Co, the new binding energy was found on spectrum of pCo-PFR at 150.5 eV, 130.6 eV and 99.8 eV, which could be assigned to Si2s, P2p and Si2p, respectively (Chu et al. 2018). The C 1s XPS spectra of Co and pCo-PFR were shown in Fig. 3c. The new peak was distinguished as Me 2 CO (287.6 eV) on C 1s XPS spectra of pCo-PFR, certifying the new structure was formed.

Thermal stability of nano flame retardant modified cotton fibers
The thermal stability of nanoparticles modified cotton fibers were assessed by thermogravimetric analysis (TGA) under continuous nitrogen from 30°C to 700°C. The heating rate was 10°C/min. As shown in Fig. 4 and Table 2, the initial degradation temperature (T 5% ) and fastest degradation temperature (T max ) of pCo-PFR was more sensitive than Co. This phenomenon was explained by the decomposition of phosphorus to catalyze the formation of carbon in the fibers at lower temperature (Li et al. 2021). Mass loss rate @ T max of pCo-PFR decreased from -34.6%/ o C to -28.7%/ o C, indicating that addition of PFR slowed down the degradation of cotton fibers. The residue weight percent (CR) of pCo-PFR was 19.0% at 700°C. At the same test condition, the CR of Co   and pCo were 3.0% and 3.2%, respectively. The CR of pCo-PFR was higher than the Co and pCo distinctly, and also higher than the theoretical value. The results confirmed that PFR not only endowed cotton fibers more residual carbon, but also promoted the fibers to form carbon by themselves.

Evaluation durable flame retardancy of nanoparticles modified cotton fibers
Micro calorimeter test has been recognized as one of the most acceptable fire testing apparatus for fiber materials that provides the heat release parameters information to evaluate the fire hazard (He et al. 2018). Figure 5a and b showed the heat release rate (HRR) curves and total heat release (THR) curves of Co, pCo and pCo-PFR. The peak of heat release rate (pHRR) and THR are significant parameters for evaluating the effectiveness of flame retardant (Wu et al. 2018). It could be observed clearly that pHRR of Co was 451 W/g and the THR was 12.1 kJ/g. Compared with Co, the pHRR and THR values of pCo-PFR were decreased by 47.3% and 29.8%, respectively. Compared with pCo, the pHRR and THR values of pCo-PFR were reduced by 37.4% and 28.0%, respectively. Furthermore, time to reach peak of heat release rate (t pHRR ) is also an important parameter for fire risk assessment, and longer t pHRR means more time for escaping. The data in Table 3 showed t pHRR of Co, pCo and pCo-PFR were 35 s, 34 s and 33 s, respectively. The t pHRR of pCo-PFR was shortened slightly. In order to evaluate the flame retardancy of pCo-PFR contained ignition time (TTI), pHRR and t pHRR comprehensively, the fire performance index (FPI) and fire growth index (FGI) were introduced as the Eq.
(2) and (3) (Hong et al. 2013): Generally, higher FPI value and lower FGI value present a lower fire risk (Wang et al. 2018). In Table 3, FPI of Co, pCo and pCo-PFR were 0.011, 0.015 and 0.042, respectively. FGI of Co, pCo and pCo-PFR were 12.9, 11.2 and 7.2, respectively. Above data proved pCo-PFR had better property of fire suppression and provided more time to escape from the scene of fire. The flame retardant property of pCo-PFR was attributed to the destruction of cyclic combustion through synergistic effects of PFR. In addition, the results of LOI were given in Table 3. Co and pCo were easy to ignite and LOI value were about 18%, while LOI of pCo-PFR increased to 27%. The data revealed that the ignition and combustion abilities both decreased with introduction of phosphorus and silicon contents in the cotton fibers.
The flame retardant durability of dip-coating and plasma induced crosslinked cotton fibers was investigated by washing (cycle time was set as 5, 10, 30 and 50) according to ISO 6330: 2012, MOD standard. As shown in Fig. 5c and Table 4, after 50 washing cycles, the pHRR and THR of pCo-PFR were 399 W/g and 10.3 kJ/g, still decreased by 11.6% and 14.9% compared with Co. Meanwhile, the weight gain rate of pCo-PFR decreased with the increase of washing times. As shown in Fig. 5d, the weight gain rate of pCo-PFR was still maintain at 10.2% after 50 times circle washing. Besides, the data in Table 4 showed that LOI of pCo-PFR was fluctuated slightly before 30 times washing circles. After 50 washing cycles, LOI of pCo-PFR was still maintained at 23.9%. The data proved dip-coating and plasma induced crosslinking brought stability and durability to flame retardant modified cotton fibers.
The horizontal combustion test is the necessary means to analyze the flame-retardant property of pure cotton fibers and modified fibers. The tests were conducted by an UL-94 instrument. As shown in Table 5, the combustion of pure cotton fibers was rapid and complete with no char left, and the burning rate was 43.3 ± 0.5 mm/s. However, the burning rate of modified fibers was reduced to 9.1 ± 0.5 mm/s evidently owning to flame retardants. The burning rate of pCo-PFR was increased with continue washing. When the washing times was 30, the burning rate of pCo-PFR was 16.5 ± 0.5 mm/s, and the washing times reached at 50, the burning rate was 25.3 ± 0.5 mm/s. It was undeniable that modified cotton fibers had advantageous washability via dip-coating and plasma induced crosslinking.

Mechanism analysis of flame retardant modified cotton fibers
To investigate the flame retardancy mechanism in gas phase and/or condensed phase, the microstructures of the residual char were collected after combustion test and analyzed by SEM and FT-IR as shown in Fig. 6 and 7. In Fig. 6, Co presented loose and shapeless structure after combustion, which was attributed to diffuse oxygen and flammable gas forming cyclic combustion. The residue sample of pCo showed loose and chaotic state. Different from the Co and pCo, carbon residue of pCo-PFR appeared relatively complete and kept the spiral shape of cotton fiber. The condensed phase products at different decomposition temperatures were analyzed by FT-IR spectroscopy as shown in Fig. 7. The temperatures of 330°C, 380°C, 430°C and 600°C were important decomposition periods for cotton matrix in the TGA test. There was no obvious change in chemical structure for Co, pCo or pCo-PFR at 330°C. Further degradation at 380°C, the dehydrogenation of Co, pCo and pCo-PFR occurred and produced substances of aldehydes, ketones (Pastorova et al. 1993). In the FT-IR spectrum at 430°C, the unsaturation of molecular chain in the residue carbon increased, and the conjugation of C=C and C=O made the corresponding infrared absorption peak blue shift (Zhang et al. 2021). At 600°C, the degradation was completed basically. The carbon residue with higher C/H ratio could be obtained by deoxidation and dehydrogenation. The stretching vibration peak of Si-C (799 cm -1 ) was prominent in FT-IR spectrum of pCo-PFR (Li et al. 2019a, b). The results showed that the structure of Si-C migrated to the surface of matrix and carried out condensed phase flame retardant to cut off the heat and oxygen necessary for circulating combustion with the increase of temperature (Nechyporchuk et al. 2017).
Besides, the gas phase pyrolysis of Co and pCo-PFR at 380°C were monitored by pyrolysis gas chromatography mass spectrometry (Py-GC/MS) as shown in Fig. 8. It was reported that decomposition of activated cotton materials was mainly through two competitive pathways (Cai et al. 2018;Qu et al. 2011). Dehydrating and charring produced coke and small molecule gas, meanwhile depolymerizing formed non-volatile levoglucosan, which was a crucial middle product of cotton during the pyrolysis process. Levoglucosan continuous decomposition tended to form small molecule gases and secondary coke, glycolaldehyde, tar, etc. (Zhu et al. 2016). Thus, reducing production of levoglucosan meant a decrease in combustible substances. During decomposition of Co, the peak of levoglucosan appeared at about 15 min. In contrast, the relative peak area of levoglucosan in pCo-PFR's Py-GC/MS spectra reduced by 29.4%. In addition, the peak area percentage of flammable gases produced also decreased by pCo-PFR, such as furans ( (1) and (2)), alcohols ( (2) and (3)), ketones (4) and aldehydes (5). The nonflammable small molecules contained phosphorus and silicon elements (6) were produced from the cracking process of pCo-PFR.  pCo-PFR 9.1 ± 0.5 9.9 ± 0.5 11.1 ± 0.5 16.5 ± 0.5 25.3 ± 0.5 So, P and Si elements endowed nanoparticles flame-retardant function through two aspects. In condensed phase, the PFR mediated dehydration of the polymeric structure and formation of char by inducing cyclization, crosslinking, and aromatization/graphitization. The formation of carbonaceous char reduces the release of volatiles (Camino et al. 2001;Schartel et al. 2016;). In the gas phase, the PFR slow down heat transfer via radical generators and non-flammable gases (Bourbigot et al. 2004;Lazar et al. 2020).
Raman spectroscopy is a suitable method to investigate the structure of carbonaceous materials formed during burning (Rao et al. 2021). The residual chars of Co and pCo-PFR after thermal gravimetric analysis were studied by Raman spectroscopy as shown in Fig. 9 a and b. The peak fitting mode was based on the Gaussian model. All samples displayed overlapped peaks with intensity maxima at 1350 cm -1 (called D band) and 1585 cm -1 (called G band), which stand for the typical polyaromatic species or graphitic structures (Rao et al. 2021). The D band corresponds to disordered graphite or glassy carbons and the G band represents the organized graphitic structures of the char layer. More importantly, the peak area ratio of D and G bands (I D /I G ) is related to the microcrystal size of the char layer. Smaller I D /I G value indicates a larger size of carbonaceous microstructures, which means batter flame retardance (Hou et al. 2018). As shown in Fig. 9, it was found that the I D /I G ratio followed the sequence of pCo-PFR (1.56) \ Co (2.21). The char layer of pCo-PFR became more compact than that of Co. Therefore, the synergy of

Mechanical properties
The Mechanical properties of Co, pCo and pCo-PFR were shown in Fig. 10a, b and Table 6. It was found that the breaking strength of Co, pCo and pCo-PFR were 4.22 cN, 3.93 cN and 3.79 cN, respectively. The breaking strength of pCo-PFR was decreased by 10.2% than that of Co. Linear density of pCo-PFR was slightly higher than that of Co. Considering the fiber fineness, the fracture strength of single fiber of Co, pCo and pCo-PFR were 2.78, 2.49 and 2.26 cN/ dtex. Besides, elongation at break of pCo-PFR reduced 8.4% compared with Co. The mechanical property test showed that the dip-coating and plasma induce crosslinking modification had limited damage of mechanical property to cotton fibers.

Conclusions
A novel nanoparticle flame retardant was synthesized by radical polymerization of THOD and VA in deionized water. Flame retardant activity assessment   demonstrated that the nanoparticle had excellent flame retardant for matrix by synergistic mechanism. For the dip-coating and plasma induced crosslinking cotton fibers, the pHRR and THR decreased by 47.3% and 29.8%, respectively. The durability of the modified cotton fibers was approved by cyclic washing test. The pHRR and THR of pCo-PFR were still reduced at 399 w/g and 10.3 kJ/g after 50 times of washing. Besides, nanoparticle eliminated the influence on mechanical and comfort properties of cotton, reduced the modification dosage and saved cost. To conclude, flame retardant property enriched application field of cotton materials based on the nanoparticle.