Single-faced flame resistance of cotton fabrics modified via mist copolymerization

Cotton fabrics with single-faced flame resistance are successfully fabricated through a simple mist copolymerization process using pentabromobenzyl acrylate (PBBA) as the functional monomer. The comonomers are methyl acrylate (MA), which can react with the hydroxyl groups of cellulose by transesterification, and divinyl benzene (DVB), a cross-linker. SEM images indicate that a very thin copolymer layer (the thickness is about 200 nm) was formed on the cotton fiber surface and the flame resistance tests show that the modified fabrics have an improved flammability with longer time to ignition (TTI), lower peak heat release rate (PHRR), lower total heat release (THR), and lower average mass loss rate (AMLR), when compared to the original cotton fabric. The modification also results in good wearing durability because the flame-retardant coating was covalently linked to the cotton fabric surface by many ester groups. Moreover, desired cotton characteristics such as tensile strength, water absorbency, vapor permeability and flexibility are mostly retained because the mist method gives a single-faced modification of the cotton fabrics.

In our previous works, 43,44 a "mist polymerization" technique has been used to modify cotton fabric surface. Asymmetrically superhydrophobic, 45,46 and wear-resistant 47 cotton fabrics were fabricated by feeding atomized monomer solutions to an in situ polymerization to build thin polymeric coatings on the cotton fabrics. The advantages of the mist polymerization include wider range of applicable monomers, tailorable coating thinness and surface morphology, simple operation, single faced modication, and almost no damage to the original properties of the fabric. These features of the mist polymerization method are suitable for fabricating cotton fabric with single faced function.
Pentabromobenzyl acrylate (PBBA) is a polymerizable monomer containing of approximately 71 wt% bromine element, [48][49][50] belonging to the class of organobromine FR, which appears to work even at low concentration. 50 Both homopolymerization 50-52 and copolymerization 53,54 of PBBA were widely studied for the applications in commercial polymer materials such as polystyrene, polyamide 6, and polypropylene.
In this work, a mist copolymerization process utilizing PBBA as the FR monomer is applied to modify cotton fabrics. Methyl acrylate (MA) is used as a co-monomer to enhance the adhesion of the polymer coatings via a transesterication with the hydroxyl groups of cellulose. The ame resistance properties of the resulting cotton fabrics are examined by the burning tests and cone calorimeter experiments. The abrasion resistance, mechanical stability, water absorbability, and moisture transmissibility of the modied cotton fabrics are further characterized.

Materials
Pentabromotoluene (PBB) was obtained from TCI Co., Ltd (Shanghai, China). The cotton fabrics were purchased from a local fabric store (60 ends per cm, 30 picks per cm, 0.42 mm thickness, 120 g m À2 weight, 35.2 m 2 g À1 specic surface area). Before chemical modication, the cotton samples were cleaned by ultrasonic washing in ethanol (30 min) and deionized water (30 min Â 3 times), respectively. Other chemical reagents were purchased from Aladdin Co., Ltd (Shanghai, China), and all used as received without further purication. Deionized water with a resistivity of 18.2 MU cm was used in all experiments.

Synthesis of pentabromobenzyl acrylate (PBBA)
PBBBr (21.2 g, 37.5 mmol), a mixture solution of acrylic acid (AA) and sodium acrylate (NaAC) (10.4 mmol L À1 , pH ¼ 6.5, 4.6 mL), and 2-methoxyethanol (ME, 63 mL) were mixed, heated at 100 C for 3 h. The obtained precipitate was washed with deionized water (100 mL Â 3 times), and dried in a vacuum oven at 80 C for 12 h to obtain the product. Yield 18.2 mmol L À1 ) was atomized using an air compression-type atomizer (DH-M01, DongHan, China), fed (0.36 mL min À1 ) to a side surface of a cotton fabric sample (30 Â 30 mm) for 5 min, and dried at 80 C for 10 min to obtain ACN treated cotton fabric. The monomer solution was atomized, fed (0.8 mL min À1 ) to the ACN cotton sample for 5 min, heated at 60 C overnight and at 180 C for 5 min, washed with deionized water (50 mL Â 3 times), dried at 80 C for 2 h. Other modied cotton fabric samples were prepared using the procedures shown in Table 1. Fully modied cotton fabric was prepared under a similar process but changed the mist feeding step with an immersion treatment in the monomer solution.

Characterizations
Fourier transform infrared (FTIR) measurements were performed on a Nicolet Avatar 370 spectrometer (Nicolet Company, Madison, USA) in a normal transmission mode. 1 H-NMR spectra were recorded on an Avance AV-400 (400 MHz) NMR spectrometer (Bruker, Switzerland) in CDCl 3 with TMS as an internal standard. Fabric surface was observed by a JSM-6700F eld emission scanning electron microscope (FE-SEM, JEOL, Japan) aer gold coating (thickness of approximately 10 nm). Attenuated total reectance infrared (ATR-IR) spectra were collected utilizing a Nicolet Avatar 370 spectrometer (Nicolet Company, Madison, USA) equipped with an ATR accessory. Fabric abrasion tests were performed using a Martindale abrasion instrument (HZ-8029A, Heng Zhun Instrument Co., Ltd, China) according to the method of DIN EN ISO 12947-3.
The vertical burning test was carried out directly burning the fabric for 10 s in the ame of alcohol burner, referring to the standard test method ENISO 15025. The cotton fabric samples was folded to make the modied surface outward to ame (before burning), and the burning state of the fabrics was recorded by a camera. To compare the ammability of the cotton samples with and without FR modication, a cotton fabric (60 Â 60 mm) having half-modied surface was prepared by combining the mist polymerization process (same to FR-cotton4) with a shielding over the other half to keep the original cotton surface. To check the durability of the FR coating, the modied fabric samples aer 150 abrasion cycles (abrasion conditions were same to previous works [40][41][42] ) were evaluated using the ame resistance test described above. In order to further verify the ammability, the combustion behavior of the cotton fabrics was also evaluated by cone calorimeter (FTT Company, UK) with a heat ux of 35 kW m À2 . The ammability indexes of total heat release (THR), peak of heat release (PHRR), time to ignition (TTI), and average mass loss rate (AMLR) were simultaneously measured. According to the ISO 5660 standard, the samples were cut into the size of 100 Â 100 mm 2 and wrapped with aluminum foil and placed in a frame with grid. The measurement was repeated to triplicate and the average data were reported.
Water absorption ability, water vapor permeability, tensile strength, and exibility of the modied cotton fabrics were determined using the methods reported in our pervious works. [43][44][45][46][47] Results and discussion Synthesis of the ame retardant monomer As shown in Fig. 1a, the PBBA monomer was synthesized by reacting AA with PBBBr, which was obtained by brominating PBB. Fig. 1b shows the FTIR spectra of PBBA and PBBBr. The peaks at 624 cm À1 and 1429 cm À1 in the spectrum of PBBBr are attributable to the CH 2 -Br stretching and -CH 2 bending, respectively. However, these peaks disappeared in the spectrum of PBBA, and were displaced by a strong absorption peak at 1730 cm À1 , meaning that the CH 2 -Br structure has been substituted by an acrylate group. The molecular structure of PBBA was further conrmed by the 1 H-NMR spectra shown in Fig. 1c. The peak at 5.01 ppm is attributable to the methylene groups in PBBBr, and the peak at 5.69 ppm is corresponded to the -CH 2 CO 2structure in PBBA. In addition, the peaks at 5.90 ppm and 6.45 ppm are believed due to the cis and trans structures of ]CH 2 , and the peak at 6.15 ppm is assignable to the proton in the ]CHbond.

Fabrication of the FR cotton surfaces
The FR layer on the cotton fabric was obtained through a twostep procedure (Scheme 1). The oxidation of cellulose by ACN gives radicals rst. Then the mist copolymerization, which was initiated by the resulting radicals, forms the copolymer layer on the cotton ber surface. The three monomers in the mist copolymerization play respective roles: DVB is used as a crosslinker; MA is designed to introduce ester groups to react with the hydroxyl groups of cellulose; and PBBA acts as the functional monomer to decrease the ammability of cotton fabric.
Our previous works 45,46 reported that the diameters of the mist droplets range from 150 to 500 nm, and only a small number of the droplets (about 3%) is xed on the cotton surface during the mist feeding. Therefore, mist copolymerization generally results a thin polymeric layer on a single side surface of the substrate.
Five fabric samples (Table 1) were prepared using mist polymerization, and their modied surfaces were examined using ATR-IR surface analysis technique. Fig. 2 compares the ATR-IR spectra with a comparison with pristine cotton. One characteristic peak appeared at 1725 cm À1 in the spectra of modied cotton fabrics but not in the spectra of pristine cotton fabric and the opposite surfaces (data not shown). This peak is attributable to the covalent bond of C]O in the ester structure, meaning that the mist copolymerization took place on the single side surface of the cotton fabric. The low-magnication SEM images for pristine cotton and FR-cotton4 ( Fig. 3a and b) display nothing signicantly different, meaning that the copolymer layer formed on the cotton ber surface was very thin. The high-magnication SEM images (Fig. 3c-f) further suggest that the copolymer layer possesses a thinness of approximately 200 nm.

Flame resistance of the FR cotton surfaces
First, the ame resistance of the modied surfaces was tested by burning the fabrics in a ame for 10 s (Fig. 4) as an improved method basing on the standard test method ENISO 15025. Before the burning test, the fabric sample was folded to make the modied surface outward to ame, and the whole burning process was recorded by an optical camera. As a result, the pristine cotton fabric (Fig. 4a) was almost completely burned in 10 s, whereas the modied cottons (Fig. 4b-d) were partly burned. As shown in Table 1, in contrast with the pristine cotton (burning rate was 3 mm s À1 ), all the modied fabric samples exhibited signicant ame resistance effect, especially, FR-cotton4 slowed the burning rate to 1.60 mm s À1 . To investigate wearing durability of the FR coatings, the modied   surfaces were abraded using a Martindale abrasion instrument.
The ame resistance effect of the samples aer 150 abrasion cycles was evaluated again using the improved ame resistance method. As shown in Fig. 5, FR-cotton1 (Fig. 5b) was burning fast as pristine cotton (Fig. 5a), suggesting that the FR function was lost aer the repeated abrasion tests. In contrast, the copolymer coating containing MA monomer showed a promising abrasion resistance (Fig. 5d). The reason may be assigned to the covalent ester linkages between the copolymer layer and the cellulose chains. When compared with FR-cotton3 (Table 1), FR-cotton4 exhibited more satisfactory stability against repeated abrasion tests. In spite of the increased MA concentration, FR-cotton5 showed the durability similar to FR-cotton4, meaning that 24 mmol L À1 of the MA concentration was enough to enhance the durability against abrasion. Aer 150 cycles of the abrasion tests under a pressure of 12 kPa, the fabric sample still has good ame retardance (losing 7.5% in burning rate). This result suggested that the modied fabric can be used for practical applications such as curtain. 44,55 Moreover, the addition of monomer AA (FR-cotton2) imparted the FR coating without satisfying improvement on the wearing durability (Fig. 5c), suggesting that the transesterication of the MA units occurred more easily than the esterication of the AA units. To further compare the ammability of the modied surfaces, a cotton fabric (60 Â 60 mm) with a shelter over its half area, was subjected to the mist copolymerization process (condition is same to FR-cotton4) to make the FR layer coated on the other area of the fabric surface. This special fabric was folded to align the two areas, suspended above the ame to re it, and quenched immediately aer it was ignited for 1.0 seconds. Its burned hollow shown in Fig. 6a indicates that the modied area has improved FR effect by comparing with the original part, suggesting that the copolymer layer on cotton surface can prolong the ignition time. On the other hand, the burning result shown in Fig. 6b indicates that the opposite surface of the fabric has a burning rate same to original cotton. Because the mist copolymerization gives a single-side surface modication, this result is very understandable.
To estimate the combustion properties using cone calorimetry method, cotton fabric samples were wrapped with aluminum foil to make the modied surface up on a specimen holder. Fig. 7 shows the heat release rate (HRR) curves of pristine cotton and FR-cotton4 at a heating ow of 35 kW m À2 , and the cone calorimetry data were collected in Table 2. Cotton fabric is a ammable material, its peak heat release rate (PHRR)   This journal is © The Royal Society of Chemistry 2017 reaches a value z 114 kW m À2 . In comparison, the PHRR of the FR-cotton4 was reduced by about 23%. The time to ignition (TTI) of the FR-cotton4 is 6 s, longer than that of the pristine cotton. Both the total heat release (THR) and the average mass loss rate (AMLR) were slightly reduced, indicating again that the copolymer layer has an effective FR function for the cotton textile.

Inuences on the intrinsic properties of cotton
As shown in Fig. 8i, the pristine cotton samples are of excellent water absorptivity (about 260%), but the samples prepared using immersion method (FR-cotton6) are poor at 130%. In contrast, the cotton fabrics modied by the mist copolymerization show good water absorptivity ranging from 245 to 218%, which is slightly lower than that of the original cotton fabric. These results suggest that the single sided modication keeps a large part of the excellent water absorptivity of cotton. For most clothing products, the desired water absorption can lower wetting of sweat, thereby being pleasant for the wearer.
The vapor transmission rate of pristine cotton fabrics is 1450 AE 48 g per m 2 per day (Fig. 8ii), indicating that the original cotton fabric has good permeability to water vapor. However, the vapor transmission rate was decreased to about 58% by the immersion treatment (FR-cotton6, 850 AE 43 g per m 2 per day). Comparatively, the mist copolymerization process gave acceptable vapor transmission rates above 1300 g per m 2 per day, which is very near to that of the original cotton fabric (90%).
The mechanical properties of the cotton fabrics were also studied by measuring the breaking tensile strength. As shown in Fig. 8iii, the pristine cotton fabric has a general breaking strength of 16.85 MPa, whereas the FR-cotton4 was lightly strengthened to 17.31 MPa, indicating that the mechanical damages caused by the nishing treatments are quite small. Fig. 8iv compares the exibilities of the modied cotton fabrics. The original cotton fabric exhibited a good exibility, as the height of the loop less than 11.8 mm. While the FR-cotton6 revealed a large loop height of more than 16.0 mm, meaning that the exibility damage caused by the immersion method is serious. The sample obtained by the mist copolymerization process (FR-cotton4) has a loop height of 12.2 mm, which is very near to that of the original cotton fabric, suggesting that the mist copolymerization impaired the cotton exibility insignicantly.

Conclusions
A new type of FR cotton fabric with single faced function was fabricated through a mist copolymerization technique. Unlike most other ame-retardant fabrics, this fabric exhibits asymmetric FR property on two faces: one face is of FR function but the opposite same to original cotton. The modied cotton fabrics delay the burning rate in the combustion tests and show positive FR behavior (including TTI, PHRR, THR and AMLR) in the cone calorimeter experiments. Moreover, the single face modication gives not serious damages on the desired cotton natures such as water absorption and vapor transmissibility. Considering the excellent balance of the new FR function and the intrinsic cotton natures, the method demonstrated in this work is believed to have promising potential in textile industries.

Conflicts of interest
There are no conicts of interest to declare.