A novel method to prepare a flame-retardant polyvinyl alcohol fiber with modified acrylonitrile coatings

Abstract

To improve the flame resistance of polyvinyl alcohol (PVA) fibers, the well-designed permanent flame-retardant coatings were introduced on the surface of PVA fibers. Acrylonitrile (AN) was firstly grafted on the PVA fiber surface and then the grafted fibers (PVA-g-AN fibers) were reacted with hydrazine hydrate and copper sulfate solution to form the coatings with flame retardant performance. The structure of the fibers was characterized by Fourier transform infrared spectroscopy and scanning electron microscopy. The flame-retardant performance of the fibers was evaluated by limiting oxygen index (LOI) and micro calorimeter (MCC) tests. It was found that the coatings were effective in improving the flame resistance of PVA fibers. SEM photos of char residues and the results of the TG-IR technique revealed that flame retardance is mainly provided through the barrier action of the coatings with the partial effect of gaseous phase.

---

1. Introduction

Polyvinyl alcohol (PVA) fibers have been widely used in many fields because of their excellent properties. However, they are readily ignitable in air and their limiting oxygen index is only about 19.5%, which handicaps their application in many cases. Therefore, it is necessary to improve their flame resistance. Halogen flame-retardants, such as polyvinyl chloride, are mostly used to improve the flame resistance of PVA fibers.^1,2 Those halogen flame-retardants work very well, but the fibers release toxic gases while burning. Therefore, there has been reported a lot of favorable halogen-free flame retardants that can be used to improve the flame resistance of PVA sheets and films.^3–7 However, most of those halogen-free flame retardants can not be used for PVA fibers because of the rigorous preparation requirement of PVA fibers. As is well known, the as-spun PVA fibers have to endure thermal stretching and heat setting under about 220 °C to improve their crystallinity and orientation degree. Unfortunately, a lot of frequently-used flame retardants, such as many organo-siloxane flame retardants, will decompose at this temperature. Moreover, the formalization reaction is usually carried out on PVA fibers to improve the hot water resistance, which should be catalyzed with sulfuric acid. The structure of many effective flame retardants, such as melamine polyphosphate (MPP), melamine cyanurate (MCA), ammonium polyphosphate (APP) and most phosphoric ester flame retardants, will be completely destroyed under the rigorous acidic condition.

Therefore, post-treatment method, which is very convenient and effective, seems to be an available alternative to obtain flame-resistant PVA fibers. It has been reported that the flame resistance of many fibers or fabrics, such as, cotton, silk, viscose rayon, polyacrylonitrile (PAN) fibers, etc., was highly improved after being grafted with flame retardants onto their surface.^8–14 However, documents about PVA fibers grafting with flame retardants were few reported. Therefore, it is interesting to attempt to improve the flame resistance of PVA fibers by grafting flame retardants onto their surface.

In addition, it has been reported that the flame resistance of PAN fibers was highly improved after being modified in hydrazine hydrate and copper sulfate solution.^15–19 These modified PAN fibers decompose with negligible output of smoke and toxic gases, and form stable chars without melting and shrinking when exposed to a naked flame. It has been proved by many researchers that the excellent flame retardancy of these modified PAN fibers was mainly attributed to the promoting effect of the metal ions on charring processing with partial effect on reducing the generation of combustible gases during burning.^17,20,21 However, the reported tensile strength of those fibers is only about 1.50 cN per dtex, which severely limits their wide application.^22–24 In our previous work, PVA as reinforcing agent was added into the spinning solution of PAN to improve the tensile strength of the modified PAN fibers.²⁵ As expected, the tensile strength of the blend PAN/PVA fiber after flame retardant modification was much higher than that of the neat PAN fiber after the same modification processes. Energy-dispersive spectroscopy revealed that metal ions were mainly distributed around the surface of those modified PAN/PVA fibers, which indicated that the flame retardant modification basically occurred on the surface layer of the blend fibers, i.e., the inner PAN was not participated in the modification and thus did not contribute to the flame resistance. Therefore, it is possible to construct a novel flame-retardant fiber with PVA substrates and modified PAN coatings, which has the favorable tensile strength and flame retardance. The tensile strength of this designed material is mainly provided by PVA fiber and the flame resistance is planed to be given by the high flame-resistant metal-contained PAN coatings. Actually, many kinds of metal complexes have been synthesized and blended into PVA films as flame retardants by a lot of researchers, and those metal complexes were proved to be quite effective to improve the flame resistance of PVA films.^3,4,26 Hu prepared a kind of nickel-contained flame retardant based on chitosan, which improved the flame resistance of PVA films significantly by promoting the dehydration processes of PVA and accelerating the formation of char residues during burning.²⁶ However, these blend methods may not be used to improve the flame resistance of PVA fibers for the rigorous preparation requirement of PVA fibers as has been mentioned above.

In this paper, PVA fibers grafted with a thin layer of PAN were obtained by using the method reported in many literatures,^27–31 then the grafted fibers (PVA-g-AN fibers) were successively reacted with hydrazine hydrate and copper sulfate solution to promote the flame retardant performance. The structure, appearance and mechanical properties of the fibers were investigated by Fourier transform infrared spectroscopy, scanning electron microscopy, and tensile strength testing, respectively. Flame retardant performance of the fibers was measured by limiting oxygen index and microcalorimetry tests.

2. Experiment section

2.1 Materials

Potassium permanganate, concentrated sulfuric acid (≥98%), hydrazine hydrate (HHA), and copper sulfate pentahydrate (all from Ke Long Co., Ltd., Chengdu, China), were analytically pure and used as received. Acrylonitrile (AN), also purchased from Ke Long Co., Ltd., was analytically pure and used after removing inhibitor. Polyvinyl alcohol fiber (PVA fiber), a commercial product, was kindly provided by Sichuan Vinylon Works, SINOPEC, Chongqing, China.

2.2 Preparation of PVA-g-AN fibers

PVA-g-AN fibers were prepared according to the method described in literatures.^32,33 Firstly, PVA fibers were immersed into 0.05 mol L⁻¹ potassium permanganate solution at 40 °C for 10 min to generate free radicals on their surface, and solid–liquid ratio of the activation reaction system was 1 : 40. Then, the activated PVA fibers were washed with de-ionized water until the washing water no longer being violet. After squeezing out extra water, the activated fibers were put into 20 vol% AN water mixture at 50 °C for different time under nitrogen atmosphere. Solid–liquid ratio of the grafting reaction system was 1 : 80. Sulfuric acid was then added into the system to catalyze the grafting reaction and the concentration of sulfuric acid in the reaction system was 0.08 mol L⁻¹. PVA-g-AN fibers were finally obtained after being washed with plenty of water to remove the adsorbed homopolymers of AN. Grafting degree was calculated according to formula (1):where M_o and M_t are the weight of the fibers before and after grafting, respectively.

2.3 Flame-retardant modification

2.3.1 Preparation of H-PVA-g-AN fibers. PVA-g-AN fibers were modified in a consecutive two-stage process to improve their flame resistance. In the first stage, PVA-g-AN fibers were treated in 40 wt% aqueous solution of HHA at 95 °C for different time. Nitrile groups on the surface of PVA-g-AN fibers were projected to be reacted with HHA during this stage. After modification in HHA solution, the fibers were denoted as H-PVA-g-AN fibers.

2.3.2 Preparation of Cu-PVA-g-AN fibers. Cu-PVA-g-AN fibers, which contained copper ions, were obtained by treating H-PVA-g-AN fibers in 15 wt% copper sulfate aqueous solution at 95 °C for 1 h. The chelating groups of H-PVA-g-AN fibers would absorb copper ions after treatment.

The solid–liquid ratio of the reaction system of each step was 1 : 20, and after each stage of treatment, the fibers were washed to neutral with distilled water and dried in vacuum oven at 80 °C.

2.4 Experimental techniques

Infrared spectra were recorded using a Nexus-560 (Nicolet, USA) Fourier transform infrared spectrophotometer (FTIR) on fibers powder by transmittance methods. The wave number resolution was 2 cm⁻¹ and the scan region was from 4000 cm⁻¹ to 400 cm⁻¹.

Mechanical properties were measured using a single fiber tensile tester (Model LLY-06, Laizhou Electron Instrument Co., Ltd., China) with a 10 cN load cell. Tests were conducted on single fiber specimens using a crosshead velocity of 20 mm min⁻¹ and a gauge length of 20 mm. All the tensile tests were carried out at the constant temperature of 25 °C and relative humidity of 70%. Twenty single fiber specimens were measured to make an average.

LOI values were measured on a LOI analyzer (Model JF-3, Jiangning Co., China) according to ISO 4589-2. And an average of at least five replicas was adopted.

Thermal gravimetric analysis (TGA) was carried on a TA instrument Q600. The fiber samples were cut into powder and dried in a vacuum oven at 80 °C for 12 h. The measured temperature was varied from 100 °C to 800 °C at a heating rate of 10 °C min⁻¹ under nitrogen atmosphere.

Thermal gravimetric analysis/infrared spectrometry (TG-FTIR) was performed to analyze the volatilized products after the pyrolysis of samples under a nitrogen flow of 35.0 mL min⁻¹. All samples were approximately 10 mg and were kept at 100 °C for 5 min first to remove the absorbed water, then were heated up to 800 °C at a heating rate of 10 °C min⁻¹.

Microcalorimetry (MCC) tests were conducted on a FTT0001 micro calorimeter instrument (FTT, UK). The dried powdered fiber samples, placed in a 40 μL alumina crucible, were heated from 100 °C to 750 °C at a heating rate of 1 °C s⁻¹ in a stream of nitrogen flowing at 80 mL min⁻¹. Then, the mixer of the volatile anaerobic thermal degradation products and nitrogen gas stream was mixed with a 20 mL min⁻¹ stream of oxygen gas before entering to a 900 °C combustor. All samples were approximately 5 mg.

The content of copper ions absorbed by the fibers was determined by an ICPS-8100 inductively coupled plasma emission spectrometer (Shimadzu, Japan). The samples were immersed into sulfuric acid solution for 48 h at 25 °C with stirring so that all the copper ions of the samples were replaced by H⁺ and released into the solution. Then, the content of copper ions in the acid solution was tested by ICP.

Scanning electron micrographs (SEM) of the samples were obtained by a JSM-5900LV scanning electron microscope (JEOL). The accelerated voltage was 20 kV.

3. Results and discussion

3.1 Structural characterization

3.1.1 FTIR spectra. Fig. 1 shows FTIR spectra of those fibers before and after modification. The absorption peaks of 2947 cm⁻¹ and 2858 cm⁻¹ are due to the stretching vibration of –CH₂ in the molecular chain of the fibers. The broad peak around 3410 cm⁻¹ can be attributed to the stretching vibration of –OH in PVA fibers and the peak at 1640 cm⁻¹ may be due to the stretching vibration of C=C, which formed during thermal setting process. After grafting reaction, there appears a strong absorption peak at 2244 cm⁻¹, which is the characteristic absorption of –CN groups and indicates the successful grafting of AN. Moreover, the peak at 2244 cm⁻¹ decreases after being modified in HHA, and there appears two obvious absorption peaks at 1629 cm⁻¹ and 1555 cm⁻¹, which were attributed to the stretching vibration of C=N of the amidrazone groups and the bending vibration of N–H of primary amine, respectively.³⁴ These changes indicates that partial nitrile groups of PVA-g-AN fibers were reacted with HHA. After chelating copper ions, the absorption peak of N–H (1555 cm⁻¹) is disappeared and the absorption peak of C=N shifts from 1629 cm⁻¹ to 1651 cm⁻¹. From the results of FTIR spectra, it may be concluded that the fibers were successfully coated with PAN and modified as expected, and copper ions were firmly fixed on the fibers.

Fig. 1 FTIR spectra of the fibers before and after modification. (a) PVA fiber; (b) PVA-g-AN fiber; (c) H-PVA-g-AN fiber; (d) Cu-PVA-g-AN fiber.

3.1.2 SEM photographs. SEM was used to observe the morphology of the fibers before and after grafting with AN, and the results are shown in Fig. 2. It can be seen from Fig. 2a and b that the surface of original PVA fiber is very smooth, and it becomes very rough after grafting reaction, indicating that AN was successfully grafted on the surface of PVA fiber. Fig. 2c shows the cross section of PVA fiber and it can be found that the original PVA fiber is very compact. Clearly, there appears a loose external layer after grafting with AN, which can be seen in Fig. 2d–g. It can be easily comprehend that the compact interior of PVA-g-AN fibers is PVA substrate and the loose outer layer is the grafted AN coatings. Moreover, the diameter of the fibers increases with the increasing of grafting degree, which indicates that the coatings become thicker with the increasing of grafting degree as PVA substrates are of the same thickness.

Fig. 2 SEM photographs of the surface and cross section of the fibers before and after grafting with AN. (a and b) Surface of PVA and PVA-g-AN-55% fibers, respectively; (c) cross section of PVA fiber; (d–g) cross section of PVA-g-AN fibers with 35%, 55%, 85% and 110% grafting degree, respectively.

Fig. 3 shows SEM photographs of PVA-g-AN-55% fiber (PVA-g-AN fiber with 55% grafting degree, similarly hereinafter) after being modified in HHA for 4 h and copper sulfate solution for 1 h. It can be seen that the surface of the fibers after modification is still very rough, which suggests the coatings are very stable and could not fall off from fibers during modification. Moreover, there appears many cracks in the interior of H-PVA-g-AN and Cu-PVA-g-AN fibers, as is shown in Fig. 3c and d, which implies that flame-resistant modification may cause the decrease of the tensile strength of fibers.

Fig. 3 SEM photographs of PVA-g-AN-55% fibers after being modified in HHA for 4 h (a and c) and further modification in copper sulfate solution for 1 h (b and d).

3.2 Mechanical properties

Table 1 gives the mechanical properties of PVA and PVA-g-AN fibers. It can be seen that the breaking force of PVA-g-AN fibers increases with grafting degree at first and decreases as grafting degree continued to increase. The breaking force of PVA-g-AN fibers is mainly provided by PVA substrates with partial effect of PAN coatings. At the beginning of grafting reaction, PVA substrates are undamaged, meanwhile, the coatings become more and more thick with the increasing of grafting degree, as a result the breaking force is slightly improved. However, long time reaction may cause some damage to PVA substrates, so the breaking force starts to decrease. This can be well proved by the SEM photographs shown in Fig. 2 that the interior substrates of PVA-g-AN-35% and PVA-g-AN-55% fibers are still very compact, but there appears some tiny defects in the interior of PVA-g-AN-110% fibers. Although, the breaking force is slightly improved after grafting with AN, the tensile strength of the fibers is still reduced as the linear density of the fibers almost linearly increases with the grafting degree, as is shown in Table 1. For example, the tensile strength is dramatically reduced from 6.26 cN per dtex to 4.49 cN per dtex after grafting with 55% AN, and almost reduced by half to about 3.20 cN per dtex when grafting degree raises to 110%. This indicates that the increase of weight during grafting reaction makes the tensile strength of the fiber decrease sharply. Therefore, the grafting degree should keep as low as possible under the precondition of insuring flame resistance of the modified PVA fibers.

Table 1 Mechanical properties of PVA and a series of PVA-g-AN fibers

Sample	Breaking force/(cN)	Linear density/(dtex)	Tensile strength/(cN per dtex)
PVA	7.19 ± 0.10	1.15 ± 0.04	6.26 ± 0.13
PVA-g-AN-35%	7.74 ± 0.08	1.55 ± 0.06	4.99 ± 0.14
PVA-g-AN-45%	7.91 ± 0.11	1.64 ± 0.05	4.82 ± 0.08
PVA-g-AN-55%	8.00 ± 0.10	1.78 ± 0.07	4.49 ± 0.12
PVA-g-AN-65%	8.10 ± 0.09	1.92 ± 0.06	4.22 ± 0.08
PVA-g-AN-85%	7.99 ± 0.12	2.13 ± 0.07	3.75 ± 0.07
PVA-g-AN-110%	7.75 ± 0.09	2.42 ± 0.08	3.20 ± 0.07

---

In order to observe the influence of HHA modification on the mechanical properties of the fibers, the mechanical properties of PVA-g-AN-55% fibers after being treated in HHA for different time was tested and the results are shown in Table 2. It can be seen that the breaking force of PVA-g-AN-55% fibers drops significantly from 8.00 cN to about 7.22 cN after being treated for 1 h, but further decreases only a little after a long time reaction, for example, the breaking force still retains about 7.20 cN after 3 h treatment. Correspondingly, the tensile strength of PVA-g-AN-55% fiber is also basically unchanged after a significant decrease at beginning. This may be because the modification of HHA mainly occurred on the coatings of PVA-g-AN-55% fibers, and the structure of the coatings was destroyed during modification but the structure of the PVA substrates was not damaged. As a result, the breaking force of PVA-g-AN-55% fiber after modification in HHA is close to that of PVA fiber, as is given in Table 1. However, further treatment in HHA may also cause some damage to PVA substrate, as is proved by the SEM photographs shown in Fig. 3, so the breaking force decreases to 7.10 cN after 4 h treatment in HHA. Importantly, the long time reaction in HHA may generate more chelating groups on the surface of the fibers, which helps to absorb more copper ions and thus improve the flame resistance of the fibers more effectively. Taking into account the decrease of mechanical properties and increase of chelating groups with treatment time, the treatment time of PVA-g-AN fibers in HHA is chosen to be 4 h in this paper.

Table 2 Mechanical properties of PVA-g-AN-55% fiber after treatment in HHA for different time

Treating time/(h)	Breaking force/(cN)	Linear density/(dtex)	Tensile strength/(cN per dtex)
0	8.00 ± 0.10	1.78 ± 0.07	4.49 ± 0.12
1	7.22 ± 0.12	1.80 ± 0.08	4.01 ± 0.11
2	7.25 ± 0.13	1.80 ± 0.07	4.03 ± 0.09
3	7.20 ± 0.11	1.81 ± 0.08	3.98 ± 0.11
4	7.10 ± 0.12	1.82 ± 0.10	3.90 ± 0.13

---

After 4 h treatment in HHA, the fibers were put into copper sulfate solution for 1 h to absorb copper ions, the mechanical properties of these Cu-PVA-g-AN fibers with different grafting degree was tested and the results are given in Table 3. As expected, the tensile strength of the fibers decreases with the increasing of grafting degree. Nonetheless, the tensile strength of Cu-PVA-g-AN fiber with grafting degree as high as 110% is about 2.79 cN per dtex, which is still much higher than that of acrylic fiber modified by similar methods reported in literatures.¹⁵ Empirically, to be used in woven form, the tensile strength of fibers should be above 2.50 cN per dtex. Otherwise, the low tensile strength will restrict the application of the fibers, thus the grafting degree should be limited to 110%.

Table 3 Mechanical properties of Cu-PVA-g-AN fibers with different grafting degree

Sample	Breaking force/(cN)	Linear density/(dtex)	Tensile strength/(cN per dtex)
Cu-PVA-g-AN-35%	6.92 ± 0.15	1.58 ± 0.08	4.38 ± 0.12
Cu-PVA-g-AN-45%	7.08 ± 0.16	1.68 ± 0.08	4.22 ± 0.10
Cu-PVA-g-AN-55%	7.10 ± 0.15	1.85 ± 0.08	3.84 ± 0.08
Cu-PVA-g-AN-65%	7.28 ± 0.13	2.01 ± 0.10	3.62 ± 0.12
Cu-PVA-g-AN-85%	7.33 ± 0.14	2.26 ± 0.12	3.24 ± 0.11
Cu-PVA-g-AN-110%	7.48 ± 0.16	2.68 ± 0.12	2.79 ± 0.06

---

3.3 Flame resistance

To clarify the influence of the specially designed coatings on the flame-retardant performance of these fibers, LOI of PVA, PVA-g-AN, H-PVA-g-AN and Cu-PVA-g-AN fibers were tested. The LOI values of PVA and PVA-g-AN fibers both ranged from 19.5% to 20.0%, indicating that they are quite easy to burn in air. After being treated in HHA for 4 h, LOI of the fibers slightly improves to about 20.5%, which means H-PVA-g-AN fibers are still very combustible. However, their LOI improves greatly after absorbing copper ions, and the fibers become self-extinguishing when they were removed from the igniting flame. It also can be seen from Fig. 4 that the LOI of H-PVA-g-AN fibers barely increases with grafting degree, but the LOI of Cu-PVA-g-AN fibers increases significantly. Moreover, the growth of LOI of Cu-PVA-g-AN fibers tends to slow as grafting degree increases. The flame resistance of acrylic fibers modified by this method was relied on the content of metal ions, which has been reported in our previous work.^19,21 Similarly, the flame resistance of Cu-PVA-g-AN fibers here may be also closely dependent on the content of copper ions. Fig. 5 gives the relationship between the content of copper ions in Cu-PVA-g-AN fibers and the grafting degree of PAN. As expected, the content of copper ions increases with grafting degree and grows less rapidly when grafting degree becomes higher, which is very consistent with the changes of the LOI of Cu-PVA-g-AN fibers. As grafting degree increases, the coatings become more and more thick, so the reaction of the inner nitrile groups is blocked at last. As a result, the content of copper ions grows slower.

Fig. 4 LOI values of H-PVA-g-AN and Cu-PVA-g-AN fibers versus grafting degree.

Fig. 5 Content of copper ions in Cu-PVA-g-AN fibers with different grafting degree.

In summary, the flame resistance of Cu-PVA-g-AN fibers increases greatly with the increasing of grafting degree and the content of copper ions. However, as has been discussed above that the increasing of grafting degree also causes the tensile strength of the fibers decrease dramatically. Therefore, the mechanical properties and flame resistance of the fibers should be controlled by the adjustment of the grafting degree.

3.4 Heat release behavior

MCC testing is a useful method for investigating the combustion properties of materials.²⁶ Results including peak heat release rate (pHRR), total heat release (THR), and temperature at the peak heat release rate (T_max) can be obtained. Low value of pHRR is an indication of low flammability and low full-scale fire hazard.^35,36 THR is characterized as the total energy released by the combustion of the gases generated in the material decomposition process, which is another important parameter for fire hazard evaluation.²⁶

Fig. 6 presents HRR curves of PVA-g-AN, H-PVA-g-AN and Cu-PVA-g-AN fibers, and the corresponding experimental data are summarized in Table 4. It can be seen that PVA-g-AN-55% fiber created much combustible gases during heating, with a THR as high as 16.6 kJ g⁻¹ and a pHRR of 162.8 W g⁻¹. After being treated in HHA for 4 h, THR and pHRR decrease slightly to 16.0 kJ g⁻¹ and 151.8 W g⁻¹, respectively, indicating that the modification of HHA almost did not reduce the generation of combustible gases and was not effective in slowing the release of combustible gases during heating. However, after absorbing copper ions, THR and pHRR of the fibers reduce dramatically, which means fire hazard of the fibers is reduced with less combustible gases generated and much slower release of these combustible gases during heating. Understandably, THR and pHRR of Cu-PVA-g-AN fibers decrease with the increasing of grafting degree, as is shown in Fig. 6 and Table 4. It also can be seen that both of PVA-g-AN and H-PVA-g-AN fibers have two heat release peaks. The first peak at relatively low temperature may be contributed to the volatile gases released by the elimination of the side chains of the fibers, and the second peak at about 390 °C is caused by the further degradation of the fibers. However, the first peak is totally disappeared and the second peak shifts to a higher temperature after absorbing copper ions, besides, T_max2 of Cu-PVA-g-AN fibers increases with the increasing of grafting degree. The reason may be that copper ions make the coatings more stable, and the decomposition products of the coatings would participate in charring process rather than form into volatile gases after absorbing copper ions.²¹ The carbon residues of the coatings may work as barriers and delay the degradation of PVA substrates. In the process of pyrolysis, the combustible gases generated by the elimination of side chains at the initial stage are obstructed by the carbon layer. Therefore, the first heat release peak of Cu-PVA-g-AN fibers at low temperature is disappeared. With the heat barrier action of carbon layer, the degradation of Cu-PVA-g-AN fibers was delayed. However, with the increasing of temperature, the carbon layer was finally broken with the releasing of combustible gases. Furthermore, the coatings become more and more thicker with the increasing of grafting degree, as a result the carbon layer left by the coatings become thicker as the grafting degree increases. Thicker carbon layer may have a better barrier effect, therefore, T_max2 of Cu-PVA-g-AN fibers increases with the increasing of grafting degree.

Fig. 6 HRR curves of PVA-g-AN, H-PVA-g-AN and Cu-PVA-g-AN fibers.

Table 4 MCC test parameters of PVA-g-AN, H-PVA-g-AN and Cu-PVA-g-AN fibers

Sample	Tmax1/(°C)	pHRR1/(W g^−1)	Tmax2/(°C)	pHRR2/(W g^−1)	THR/(kJ g^−1)
PVA-g-AN-55%	289.3	42.8	392.5	162.8	16.6
H-PVA-g-AN-55%	304.3	36.6	390.1	151.8	16.0
Cu-PVA-g-AN-35%	—	—	391.2	125.1	13.6
Cu-PVA-g-AN-55%	—	—	409.2	106.5	12.1
Cu-PVA-g-AN-85%	—	—	417.9	98.0	10.0
Cu-PVA-g-AN-110%	—	—	424.6	65.0	6.7

---

3.5 Thermal properties

To confirm the hypothesis that copper ions make the decomposition products of the coatings participate in charring process rather than form into flammable gases, the TG curves of the fibers were tested and are shown in Fig. 7 and 8. It can be seen from Fig. 7 that the thermal degradation process of PVA-g-AN-55% fiber can be divided into three stages (T = 275 °C, 371 °C, 438 °C). It is well known that PVA dehydrates at the beginning of thermal decomposition and generates unsaturated compounds.^37–39 Therefore, the first two weight-loss peaks of PVA-g-AN-55% fiber are the results of elimination of side groups.³⁷ When temperature further increases, the main chains start to break down generating a lot of volatile products and leaving a small amount of char residues finally. After modification in HHA for 4 h, the amount of char residues of the fibers at 800 °C is almost unchanged; this indicates H-PVA-g-AN-55% fibers decompose drastically with the releasing of a large amount of combustible gases. Therefore, THR of the fibers, as is given in Table 4, reduces only a little after modification in HHA.

Fig. 7 TG and DTG curves of PVA-g-AN, H-PVA-g-AN and Cu-PVA-g-AN fibers with 55% grafting degree tested under nitrogen atmosphere.

Fig. 8 TG and DTG curves of a series of Cu-PVA-g-AN fibers with different grafting degree tested under nitrogen atmosphere.

It also can be seen that the peak temperature of the first two weight-loss stages shifts to high temperature after absorbing copper ions (T = 314 °C, 378 °C). It may because the coatings delayed the decomposition of the substrates and slows the releasing of volatile gases, which is consistent with the results of MCC results. Besides, the third weight-loss stage originated from the decomposition of the main chains is not obvious after absorbing copper ions, leaving a large amount of carbon residues, which verifies that copper ions affect the degradation of the coatings and is beneficial to form carbon residues during heating.

It can be found from Fig. 8 that the amount of char residues of Cu-PVA-g-AN fibers increases with the increasing of grafting degree, but less rapidly when grafting degree becomes very high; this may due to the less rapid growth of copper ions when grafting degree becomes very high, as has been discussed above. Moreover, the first weight-loss stage shifts to high temperature with the increasing of grafting degree and overlaps with the second decomposition stages, which may because of the better barrier action of the thicker coatings.

3.6 Carbon residue morphology

SEM was used to observe the carbon residues of PVA-g-AN, H-PVA-g-AN and Cu-PVA-g-AN fibers after burning in a naked flame in order to reveal the barrier action of the coatings during combustion, and the results are shown in Fig. 9. It can be seen that PVA-g-AN fibers decompose and shrink seriously during combustion, leaving an apparently loose and random-shaped char residue. Though H-PVA-g-AN fibers also shrink during combustion, their char residues are not that loose and partially retains the fibrous structure; this is due to the cross-linked structure of their coatings. Moreover, Cu-PVA-g-AN fibers barely shrink during combustion and their char residues are considerably intact, and almost completely retain the fibrous structure, which demonstrates that copper ions promote the charring process of the coatings during heating. The carbon layer formed by the coatings works as barrier and effectively insulates heat and oxygen. As a result, the amount of char residues of PVA substrates is highly improved.

Fig. 9 SEM photographs of char residues of PVA-g-AN (a), H-PVA-g-AN (b) and Cu-PVA-g-AN (c) fibers after burning in a naked flame.

3.7 Volatilized products analysis

Thermal gravimetric analysis/infrared spectrometry (TG-FTIR) technique was applied to study the flame-retardant mechanism in the gaseous phase and the results are shown in Fig. 10. It can be seen that the absorption peaks of the volatile gases released by the three kinds of fibers during heating process are similar, which indicates that the species of their pyrolysis gases are the same. The identification of the decomposition products of the samples is summarized in Table 5. The main volatile gases released by the fibers are water (3800–3500 cm⁻¹), carbon dioxide (2400–2200 cm⁻¹), hydrocarbons (3020–2600 cm⁻¹), aldehydes (1760–1660 cm⁻¹), acids (1760–1660 cm⁻¹) and ammonia (1000–900 cm⁻¹).^40–42 Obviously, all the pyrolysis gases except carbon dioxide, water and ammonia are combustible, which are responsible to the flammability of the fibers.

Fig. 10 Volatile gases spectra measured by TG-FTIR for PVA-g-AN (a), H-PVA-g-AN (b), and Cu-PVA-g-AN (c) fibers.

Table 5 Identification of molecule configuration in the evolved products

Wavenumber (cm^−1)	Peak (cm^−1)	Group	Vibration	Product
3800–3500	3735	O–H	Stretch	H2O
1580–1480	1508	H–O–H	Bending
3020–2600	2936	C–H	Stretch	Hydrocarbon
3020–2600	3016	=C–H	Stretch
1500–1300	1456	C–H	Bending
2400–2200	2358, 2320	C=O	Stretch	CO2
680–660	669	C=O	Bending
3500–3300	3333	N–H	Stretch	NH3
1650–1580	1626	N–H	Bending
1000–900	965, 930	N–H	Bending
1760–1660	1716	C=O	Stretch	Aldehyde
3000–2600	2822, 2726	C–H	Stretch
1760–1660	1745	C=O	Stretch	Acid
3800–3500	3735	O–H	Stretch

---

Carbon dioxide (2358 cm⁻¹), aldehydes (1716 cm⁻¹), ammonia (965 cm⁻¹), acids (1745 cm⁻¹) and hydrocarbons (2934 cm⁻¹) are selected to study for revealing their changes at different temperature and the results are shown in Fig. 11. The intensities of absorbance were all normalized to the samples' weight.

Fig. 11 Peak intensities of aldehydes, acids, CO₂, NH₃ and hydrocarbons released by the three kinds of fibers at different temperature.

It can be clearly seen from Fig. 11 that aldehydes, carbon dioxide, acids and ammonia released by all the fibers began to be detected at about 200 °C, however, the hydrocarbons became detectable at relatively high temperature about 320 °C. It can be comprehended that the volatile gases containing oxygen or nitrogen were mainly contributed to the elimination of the side chains, and the hydrocarbons were mainly released by the main chains of the fibers. This indicates that the side chains of the fibers were much more easily decomposed during heating. Notably, there are two peaks in the release curves of aldehydes and acids of PVA-g-AN and H-PVA-g-AN fibers. Their release peaks of aldehydes and acids at about 300 °C can well explain the results of MCC tests shown in Fig. 6 that both of them exhibit a small heat release peak at relatively low temperature. With a little amount of aldehydes and acids released at this temperature, the first heat release peak of Cu-PAN-g-AN fibers is disappeared. Moreover, both of PVA-g-AN and H-PVA-g-AN fibers show the release peaks of hydrocarbons at about 447 °C, while Cu-PVA-g-AN fiber exhibits the peak at a higher temperature about 459 °C. It indicates that the barrier action of the coatings delayed the degradation of the PVA substrates, which is consistent with the results of MCC tests.

It also can be found from Fig. 11 that the amount of combustible gases including aldehydes, acids and hydrocarbons is largely reduced after being modified in HHA, while the amount of noncombustible carbon dioxide and ammonia is improved, simultaneously. Moreover, after absorbing copper ions, aldehydes, acids and hydrocarbons are further reduced, and carbon dioxide and ammonia are further improved. Therefore, the flame resistance of the fibers is improved with the decrease of combustible gases and the increase of noncombustible gases.

It can be concluded from the results above that the flame resistance of Cu-PVA-g-AN fibers is mainly provided through the effect of barrier action with the partial effect of gaseous phase.

4. Conclusions

In this article, PVA fibers with good flame resistance are prepared successfully by covering with flame-retardant coatings. It has been proved that the flame resistance of the modified fibers mainly depended upon the grafting degree and the content of copper ions. TG shows the carbon residues of the fibers is highly improved after being modified. Moreover, the intact char residue of the coatings work as barriers that insulate heat and oxygen effectively, as a result the PVA substrates are well protected. The nonflammable gases are largely improved, while the combustible gases are significantly reduced after being modified, which is also beneficial to improve the flame resistance. With the effect of barrier action and the partial effect of gaseous phase, the flame resistance of the fibers is highly improved.

Acknowledgements

The financial support of National Natural Science Foundation of China with grant No. 51273116 is gratefully acknowledged.

References

1. T. Ashikaga, A. Ohmori and M. Ando, GB Pat. 1 421 557-A, 1976.
2. S. Okamura, T. Tagawa, K. Yagami, K. Matsuo, M. Sakabe, T. Sogo and M. Korematsu, US Pat. 3 887 676, 1975.
3. D. L. Wang, Y. Liu, D. Y. Wang, C. X. Zhao, Y. R. Mou and Y. Z. Wang, Polym. Degrad. Stab., 2007, 92, 1555–1564.
4. H. D. Lu, C. A. Wilkie, M. Ding and L. Song, Polym. Degrad. Stab., 2011, 96, 1219–1224.
5. S. Hu, L. Song, H. F. Pan and Y. Hu, J. Therm. Anal. Calorim., 2012, 112, 859–864.
6. J. X. Feng, X. M. Zhang, S. Q. Ma, Z. Xiong, C. Z. Zhang, Y. H. Jiang and J. Zhu, Ind. Eng. Chem. Res., 2013, 52, 2784–2792.
7. W. L. Chen, X. W. Fu, W. B. Ge, J. J. Xu and M. J. Jiang, Polym. Degrad. Stab., 2014, 102, 81–87.
8. M. J. Tsafack and J. Levalois-Grützmacher, Surf. Coat. Technol., 2006, 200, 3503–3510.
9. M. J. Tsafack and J. Levalois-Grützmacher, Surf. Coat. Technol., 2007, 201, 5789–5795.
10. G. Akovali and G. Gundogan, J. Appl. Polym. Sci., 1990, 41, 2011–2019.
11. J. Alongi, F. Carosio and G. Malucelli, Polym. Degrad. Stab., 2014, 106, 138–149.
12. B. Paosawatyanyong, P. Jermsutjarit and W. Bhanthumnavin, Prog. Org. Coat., 2014, 77, 1585–1590.
13. J. S. Kim, U. S. Choi, Y. G. Ko and S. H. Kim, J. Ind. Eng. Chem., 2002, 8, 218–224.
14. M. Zhang, D. L. Zang, J. Y. Shi, Z. X. Gao, C. Y. Wang and J. Li, RSC Adv., 2015, 5, 67780–67786.
15. P. Bajaj, A. K. Agrawal, A. Dhand, N. Kasturia and Hansraj, J. Macromol. Sci., Polym. Rev., 2000, 40, 309–337.
16. Z. Jia and Y. G. Yang, Adv. Mater. Res., 2012, 502, 233–237.
17. J. Z. Xu, C. M. Tian, Z. G. Ma, M. Gao, H. Z. Guo and Z. H. Yao, J. Therm. Anal. Calorim., 2001, 63, 501–506.
18. P. Bajaj and S. Kumari, J. Macromol. Sci., Polym. Rev., 1987, 27, 181–217.
19. W. L. Zhou, X. Yan, P. Q. Liu, M. J. Jiang and J. J. Xu, J. Appl. Polym. Sci., 2015, 132, 41996–42004.
20. D. N. Lu, W. Qin, S. Y. Luo and Z. Q. Zhao, J. China Text. Univ., 1989, 15, 7–13.
21. X. Yan, W. L. Zhou, X. S. Zhao, J. J. Xu and L. P. Qing, J. Therm. Anal. Calorim., 2015 DOI:10.1007/s10973-015-5180-1.
22. P. J. Akers and R. A. Chapman, Lenzinger Ber., 1987, 63, 44–46.
23. P. Borrell and T. J. Ollerenshaw, US Pat. 5,051,110, 1991.
24. L. X. Zhao, Dyeing Finish., 2001, 27, 26–28.
25. W. L. Zhou, X. Yan, M. J. Jiang, P. Q. Liu and J. J. Xu, J. Appl. Polym. Sci., 2016, 133, 43006–43014.
26. S. Hu, L. Song, H. F. Pan and Y. Hu, Ind. Eng. Chem. Res., 2012, 51, 3663–3669.
27. N. Alghezawi, O. Şanlı, L. Aras and G. Asman, Chem. Eng. Process., 2005, 44, 51–58.
28. M. T. Taghizadeh and M. A. Darvishi, Iran. Polym. J., 2001, 10, 283–292.
29. Y. L. Ren, Q. Su and P. Y. Xin, Chem. J. Chin. Univ., 2013, 34, 1264–1269.
30. N. A. M. Nazri, W. J. Lau, M. Padaki and A. F. Ismail, J. Polym. Res., 2014, 21, 594–605.
31. S. C. Jana, S. Maiti and S. Biswas, J. Appl. Polym. Sci., 2000, 78, 1586–1590.
32. H. M. Zeng, Z. D. Xu and Y. Lu, Ion Exchange and Adsorption, 1993, 9, 216–222.
33. M. Tang, S. W. Li and Z. Tong, Journal of Textile Universities, 2002, 15, 220–228.
34. E. C. Riqueza, A. P. de Aguiar, M. R. M. P. Aguiar and L. C. de Santa Maria, Polym. Bull., 2005, 55, 31–40.
35. J. W. Gilman, Appl. Clay Sci., 1999, 15, 31–49.
36. M. Zanetti, T. Kashiwagi, L. Falqui and G. Camino, Chem. Mater., 2002, 14, 881–887.
37. Y. Cai, M. Zhao, H. Wang, Y. Li and Z. Zhao, Polym. Degrad. Stab., 2014, 99, 53–60.
38. P. Alexy, D. Kachova, M. Krsiak, D. Bakos and B. Simkova, Polym. Degrad. Stab., 2002, 78, 413–421.
39. D. M. Fernandes, A. A. W. Hechenleitner and E. A. G. Pineda, Thermochim. Acta, 2006, 441, 101–109.
40. R. Lujan-Acosta, S. Sánchez-Valdes, E. Ramírez-Vargas, L. F. Ramos-DeValle, A. B. Espinoza-Martinez, O. S. Rodriguez-Fernandez, T. Lozano-Ramirez and P. G. Lafleur, Mater. Chem. Phys., 2014, 146, 437–445.
41. H. M. Zhu, J. H. Yan, X. G. Jiang, Y. E. Lai and K. F. Cen, J. Hazard. Mater., 2008, 153, 670–676.
42. Q. Tai, R. K. K. Yuen, W. Yang, Z. Qiao, L. Song and Y. Hu, Composites, Part A, 2012, 43, 415–422.

---