The objective of this research was to investigate the effect of non-durable flame retardant (NDFR) coating of samples of polyester fabric untreated and treated with UV/Ozone for different periods. For this purpose, these samples were tested by Fourier transform infrared (FTIR) spectroscopy, thermal analysis tests as thermo-gravimetric analysis (TGA) and differential scanning calorimeter (DSC). The ignition test was applied using limiting oxygen index (LOI), flame chamber (UL/94). Results indicated that both AZ2 (dried at room temperature) and AZ8–12 (dried at 80 °C for 30 min after coating with non-durable fire retardant (NDFR) coating) polyester samples have significantly decreased the rate of burning and increased the limiting oxygen index.
Introduction
Polyester (PES) is one of the most important synthetic polymers. It can be used in two ways: as a raw material by itself or as a blend because of its wrinkle-resistant property and its ability to retain its shape [1]. The modification of polyester fiber was widely studied by several researchers. Many methods have been successfully used for modifying the physical and chemical properties of polyester fiber, such as blending with other polymers [2], coating with tetraethoxysilane for thermal property improvement [3] and modifying with inorganic materials [4]. PES is widely used in many fields such as textile, automotive industries, garments, pants, shirts, suits, and bed sheets either by itself or as a blend [1,5]. Conversely, it has various disadvantages such as highly flammable combined with dripping, smoking, shrinking effect, low dyeability, less wearing comfort, difficulties in finishing and insufficient washability associated with their hydrophobic nature [6,7]. For these reasons, it is necessary to improve the anti-dripping and fire retardant properties of PES. The ignition process depends on heat, fuel (such as Gasoline, acetone, ether, pentane), oxygen and free radical chain reaction. If a single angle in the fire triangle is missing the ignition can’t take place.
Flame-retardant additives play an important role in saving lives and protecting property from damage. It can be added to polyester by three methods: additives to the polymer melt, flame-retardant co-polymers and topical finishes – which have all been used commercially to produce flame-retardant polyester textiles [8]. Non-durable fire retardant (NDFR) treatments involve water soluble chemicals that can be washed off with plain water. These treatments can withstand non-aqueous laundering with dry cleaning solvents. Non-durable chemical treatments as borax, boric acid, organophosphorous compounds can be added as fire-retardants [9]. These treatments can withstand non-aqueous laundering with dry cleaning solvents. The main advantage of the non-durable fire retardant coating has always been their relative low cost. Di-ammonium phosphate (DAP) is an effective intumescent fire retardant for several kinds of polymer-based materials [10–12]. It is a high molecular weight chain phosphate. Its efficiency is generally attributed to an increase in the char formation through a condensed phase reaction. The flame retardants can be classified into two types; the first one can be mixed with polymer and called additive flame retardants, while reactive flame retardant referred to add flame retardants by reaction [13]. A diagram of the current model of combustion of textile fibers is given in Fig. 1[14]. Combustion is an exothermic process that requires three components, heat, oxygen and a suitable fuel. Combustion consists of the generation and emission of heat and light which makes the phenomenon visible. The emitted light color depends on the released amount of energy [15]. This work aimed to improve the ignition properties of the polyester by treating with NDFR coating. The effect of this new coating was examined by mechanical test (tensile and elongation), thermal analysis (DSC and TG), Fourier transform infrared (FTIR), and flammability tests (UL/94, LOI and small ignition tests).
Experimental
2.1. Materials
Mill bleached polyester fabric of weight 140 g/m2, plain 1/1 weave was kindly supplied by the Misr Company for Spinning and Weaving, El-Mahala El-Kobra, Egypt. The fabric was washed with distilled water at 60 °C for 30 min and dried horizontally at ambient condition before testing.
2.2. Synthesis of non-durable coating
The water based solution has been synthesised by adding di-ammonium phosphate ((NH4)2HPO4, Merck chemical company, Germany), urea (CO(NH2)2, 99%, Fluka) and boric acid (H3BO4, 96%, Aldrich) to a round flask and stirring at ambient temperature for 2 h. Sodium borate (Na2B4O7·10H2O, Aldrich) was then added slowly to the mixture, and stirred for 30 min until a white color appeared. The properties of the fire-retardant coating are tabulated in Table 1.
2.3. Prepared of the treated samples with final coating
Firstly, the polyester samples were exposed to UV/O3[16] for different irradiation times (0, 10, 20, 30, 40, 50 min) to excite the outer surface of the polyester sample then immersed in the freshly prepared non-durable coating for 10 min. Finally, the samples were removed from NDFR solution and padded under a constant pressure for 100% uptake and dried in an oven at 80 °C for 30 min [17].
2.4. Weight of polyester fabric
Five samples were cut to a specific area and weighed by Mettler Toledo of accuracy (0.1 mg). The weight was calculated for the square meter [18,19].
2.5. Mechanical properties
Tensile test is known as a basic and universal engineering test to achieve material parameters such as ultimate strength, yield strength, % elongation, % area of reduction and Young’s modulus. These important parameters obtained from the standard tensile testing are useful for the selection of engineering materials for any applications required. Fabric tensile strength test was conducted according to the ASTM method (1994), which is the standard method for breaking force and elongation of tensile fabrics [20] using H5KT Atlas instrument, USA.
2.6. Fourier transform infrared spectroscopy (FTIR)
FTIR of 100% PES fabric samples (treated and untreated) was recorded by means of Nicolet 380 Spectrometer – USA that is equipped with zinc selenide crystal, in the wavelength range of 4000–400 cm−1. To ensure reproducible contact between the crystal face and the fabric, a pressure of about 18 Kpa applies to the crystal holder. The FTIR absorbent frequencies for the treated samples are recorded with an average of 32 scans using a resolution of 4 cm−1[21].
2.7. Thermal analysis test
2.7.1. Thermogravimetric analysis (TGA)
TGA is a technique used to study the thermo-oxidative degradation of composites [22,23]. It was carried out using Thermal Analyzer (TGA-50 Shimadzu Instrument, Japan). The weight loss due to the formation of volatile products such as CO and CO2 after degradation at high temperature is monitored as a function of temperature. Tested samples ranging between 6.0 and 6.4 mg were placed in platinum crucibles and subjected to a temperature ranging from 25 to 750 °C in N2 gas with a flow rate of 30 ml/min at a heating rate of 10 °C/min [5,24].
2.7.2. Differential scanning calorimeter (DSC)
It is used to give information about thermal changes that do not involve a change in sample mass, characterize the thermophysical properties of polymers such as melting temperature (Tm), heat of melting, percent crystallinity (ΔHm), and glass transition (Tg). DSC of the untreated and treated samples was carried out from room temperature to 650 °C under N2 atmosphere using a DSC-50 Shimadzu (Japan) analyzer [25]. DSC has an important role in determining the exothermic reaction (it is a chemical reaction that generates heat) and endothermic reaction (a chemical reaction causing the absorption of heat). New substances formed by the chemical reactions contain more heat energy than prior to the reaction.
2.8. Fire test
The most popular test for textile materials are limiting oxygen index (LOI), flame chamber (UL-94), small ignition test (SI) and thermal analysis tests (TGA and DSC).
2.8.1. Limiting oxygen index
The minimum concentration of oxygen in a mixture of oxygen and nitrogen flowing upward in a test column (glass chamber) supports combustion measured under equilibrium conditions such as candle-like burning [26]. The flame source is propane gas. Samples were cut in dimensions of 50 mm by 1500 mm and then clamped in a U type holder which stood vertically in the oxygen index tester. Both nitrogen and oxygen were connected to the apparatus through pressure regulators. Before the ignition, the nitrogen/oxygen ratio should be set and pre-conditioned for 30 s. Three replicates were tested for each sample to obtain the average LOI value [27]. The oxygen concentration is adjusted until the sample supports sustained burning. It is useful to assign materials into experimentally meaningful groupings based on their LOI. The percentage of O2 in air is about 20.90% oxygen, if a material’s LOI is less than this, it will burn easily in air. For that, the value of oxygen, 20.90% can be considered as a threshold value while grouping materials. Table 2, includes the classification of the materials according to most researchers based on LOI [28,29].
2.8.2. Ignition behavior (UL/94)
Standard test method [30] was applied to determine the horizontal burning rate of materials. During testing the sample the burning behavior and molten drip and time of burning were investigated. In this standard, the sample was cut in dimensions of 35.5 × 10 cm2 (5 × 15 cm2) with a thickness not more than 1.3 cm and then held horizontal in a U holder. A burner with a flame of 3.8 cm is exposed to the free end of the sample for 15 s and then removed and the flammability properties of the samples are studied. After ignition, the time of ignition was subtracted from the total combustion duration. The remaining time was then reported as the burning time and the distance was calculated after the fire passed the first line (after 3.8 cm). Finally, the burning rate (BR) was calculated using this law:
2.8.3. Small flame test
This test was achieved on six samples (3 weft and 3 warp direction) with dimensions of 25 × 9 cm2 and thickness less than 6 cm to determine the ignitable effect of building products by direct small flame under zero impressed irradiance using samples tested in a vertical orientation. Single flame fire tests were based on EN ISO 11925-2 [31], burner was angled at 45° and propane gas with a flow rate of 10 ml/min was used during the experiments. The burner (with length 2 cm) was moved closer and placed from the free edge of the sample (while, 0.8 cm inside and 1.2 cm out of the sample). The total time of the test was 20 s divided into 15 s to expose the sample to the flame then removed and 5 s to study the effect of burning on the sample. The burning behavior of the sample is observed for flame spread, occurrence of burning particles and droplets during the total test period (20 s).
3. Results and discussion
3.1. Polyester fabric weight loss
The effect of irradiation with UV/O3 on the weight sample (g/m2) of PES is shown in Fig. 2. This figure illustrated that the sample weights were decreased with the increasing of time for all studied ranges (0–50 min), since the AZ7 sample recorded the lowest weight (126 g/m2) compared with the other. Weight loss was affected by irradiating up to 30 min followed by an insignificant decrement in extended irradiation time due to dry and photo-oxidation actions [32]. The changes could be returned to the combined effect of UV/Ozone on the PES. Ozone action is always based on the effect of direct and indirect reactions with PES fibers. This is consequent to the disintegration of ozone (O3 → O2 + O.); oxygen free radical reacts with the water content producing OH-radicals and/or primary and secondary hydroxyl groups (ROH) on the cellulose fibers. These radicals are very short-living and have an even stronger oxidation mechanism than that of ozone. The mechanism has been mentioned in a previous article [19]. AZ2 sample has the highest weight because it is exposed to dip-coating directly (without irradiated to UV/O3) and then dried at ambient temperature and weighted. This may be due to many reasons such as non-evaporation of water completely and volatile components of the coating structure and no chemical reaction occurring between the sample surface and the coating. Finally, by irradiating the sample surface with UV/O3, free radicals are formed over the surface, then reacted with OH groups from the non-durable coating by hydrogen bonds since AZ10 (after irradiating for 30 min) has a higher weight than AZ8–9 (154 and 158 g/m2) and AZ11–12 (141 and 134 g/m2) this referred to semi-burning to the surface in case of AZ11–12 samples.
3.2. 2 Tensile strength and elongation
Tensile and elongation properties were determined and results are shown in Table 3. Table 3 illustrates that the percentage of elongation decreased when the sample was irradiated with UV/O3 for a long time since recording the lowest (67.88%) and the highest (83.82%) percentages in the case of samples AZ7 and AZ1, respectively. The results of tensile strength go with those of elongation as mentioned in Table 3. This may return to the partially surface burning in the case of samples AZ7 which has the lowest tensile value and elongation percentage compared to the other. These results improved when AZ7 was irradiated and immersed in NDFR coating, then dried in oven for 50 min forming sample AZ12 (56.93 kgf and 70.15%). On comparing sample AZ5 (irradiated with UV/O3 for 30 min without coating) with AZ10 (coated with non-durable coating, then dried in oven for 30 min) it was found that the non-durable coating has a direct effect on sample AZ10 since the tensile strength and elongation percentage were increased due to the inter-chemical reaction between the free radical on the surface of PES and OH groups in the fire retardant coating. Finally, the results of the polyester sample are coated without UV/O3 irradiation (AZ2) and after irradiation for 30 min (AZ10). The tensile strength was increased from 66.31 to 73.55 kgf for AZ2 and AZ10, respectively. This explained the main role of UV/O3 to excite the surface of the polyester to form a free radical which forms a hydrogen bond with a non-durable coating. These hydrogen bonds play an important role in decreasing the tensile strength accompanied by a decrease in the elongation %, due to a decrease in crystallinity when compared to untreated sample, which indicates that the polyester filaments have become stiffer.
3.3. FTIR spectroscopic analysis
It is used to identify and study the chemical composition of the uncoated and coated samples with NDFR coating. Figs. 3 and 4 illustrate the comparison between uncoated (AZ1, AZ5 and AZ7) and coated (AZ2, AZ10 and AZ12) samples. In the case of uncoated samples of polyester, the main structure of the polyester fabric is made of ester, alcohol, anhydride, aromatic ring and heterocyclic aromatic rings. Alcohol was able to react with anhydride and produce ester groups [33]. That was the reason there was still alcohol and anhydride as residual reactants left in the polyester. The carboxyl, ester, anhydride and alcohol groups showed the polyester fabric was pure PES. It is clear from Fig. 4 that, uncoated samples, exhibit a broadband near 3428–3432 cm−1 due to the OH-stretching vibration of free and hydrogen bonded groups (most probably due to the humidity absorbed by KBr during the preparation of the pellets) [34,35]. The CH2 anti-symmetric stretching appeared at 2920–2935 cm−1. The ester functional groups appeared at 1710–1725 cm−1, 1240–1245 cm−1 and 1110–1122 cm−1, while the aromatic ring was observed at 2960–2972 cm−1, 1575–1585 cm−1, 1502–1510 cm−1, 1000–1022 cm−1 and 718–725 cm−1. Bands assigned for an ethylene CH2 group of O(CH2CH2)O moiety are also observed at 1448–1455 cm−1 and 855–870 cm−1. The peak at 1405 cm−1 corresponded to CC stretching vibration of the benzene ring which was a stable group. It was the characteristic absorption peak of PES. The presence of ester groups ranging between 1720 and 1268 cm−1 referred to a break under certain conditions. The peak at 1632 cm−1 disappeared in cases of AZ5 and AZ7. Fig. 4 illustrates that, AZ7 has the sharpest and the biggest area under the peaks between 1700 and 620 cm−1. This may be attributed to the semi burning of the AZ7 sample surface due to exposure to UV/Ozone for a long time compared to AZ5 and the untreated one (AZ1).
Fig. 4 illustrates that, the band at 3440 cm−1 is related to stretching vibrations of OH group in the case of AZ2 sample. As can be seen from the spectra of AZ10 and AZ12, the stretching vibrations of OH group were decreased to 3432 and 3436 cm−1 for AZ10 and AZ12, respectively. The bands in the range of 2900–3100 cm−1 corresponded to stretching vibrations of CH groups such as CH2 and CH3. Strong CH and CH2 stretching vibrations between 2920 and 2930 cm−1 have also been observed. This figure included 3230 (N–H), 1233 (P = O), 1082 (P–O symmetric vibration), 866 (P–O asymmetric stretching vibration), and 792 cm−1 with intensities of 87.718%, 94.273%, 91.099%, 97.016%, and 97.090%, respectively.
3.4. LOI
During the LOI test, the energy transferred from the flame to the burning surface maintains the surface temperature required for pyrolysis of the polymer and this supplies gaseous fuel to form a combustible mixture with O2/N2 stream. As the oxygen concentration is decreased, the flame temperature decreases, resulting in a reduction in the heat feedback and the supply of fuel to the flame zone. Fig. 5 shows the results of the LOI values of untreated and treated samples. In the case of AZ1 and AZ3–7 (since no fire retardant coating added) the oxygen index values were in the range of 18–20.9%, but in the case of samples AZ8–12 the oxygen index values are raised in varying degrees (13–15%). Similar results were reported in the case of sample AZ2 compared to blank since the LOI was observed to increase from 19.8% to 23.0%. In the case of sample AZ10, LOI increased to 26.0%, which can qualify this sample to be called as flame retardant. LOI values of samples AZ8–9 and AZ11–12 are slightly lower than those of sample AZ10, but higher than those of uncoated samples. The AZ1 sample of polyester is burned the entire length, so that there was no residual cloth left, only some ash.
3.5. Flame chamber
Table 4 shows the flammability of uncoated and coated polyester fabrics. In case of samples AZ1 and AZ3–7, they were burned combined with shrinking, dripping and the time of burning, decreased from 80 s for a blank sample to 71 s in the case of AZ7 (irradiated with UV/O3 for 50 min), while the burning rate increased from 101 mm/min to 114 mm/min due to exposure to UV/O3. The results obtained in the case of uncoated samples are investigated completely in the case of coated samples. Short length was burned for a long time, which recording the lowest burning rate values compared to uncoated samples. All kinds of coated samples have succeeded in classifying the polyester samples as flame retardant [class I (AZ10) and class II (AZ2, AZ8–9 and AZ11–12)] according to the ISO 3795 standard. Class I implies that the flame extinguishes before reaching the first position at 3.8 cm from the free edge (the ignited edge) of the sample. Thus, the coated sample (AZ10) used in this study has rendered flammable PES fabrics, fire resistant due to the presence of free radicals as the result of irradiation by UV/Ozone for 30 min and forming hydrogen bonds with OH groups from the coating solution, while for the other coated samples the flame has passed the first line with little area recording burning rate values of 30, 17, 12, 9 and 13 mm/min for samples AZ2, AZ8, AZ9, AZ11 and AZ12, respectively. Samples AZ11 and 12 have a semi-burning surface due to irradiation with UV/Ozone for a long time, so recording the highest rate of burning compared to AZ10.
3.6. Small flame
Tests were carried out according to the standard EN ISO 11925-2 [31]. The results, comments and observations of samples test are given in Table 5. The tabulated data showed that, the addition of the FR improved notably the fire retardant. When polyester burnt, the material started dripping (AZ1 and AZ3:7) as soon as the flame was in contact with the material and the integrity of the material was completely lost immediately. But in case of coated samples without irradiating to UV/O3 and also, the sample which was treated with UV/O3 for different periods passed this test without dripping during the ignition. AZ2 and AZ8:12 samples passed the test and were classified as class E.
3.7. Mass change by TGA
Results of thermogravimetric analysis are expressed by the TGA curves. The TGA curve of samples AZ1–3 and AZ10 is shown in Fig. 6 and tabulated in Table 6. The amount of final residue provides quantitative information about the flame-retardant activity. Fig. 6 shows TGA curves for uncoated (AZ1 and AZ3) and coated (AZ2 and AZ10) samples. It is illustrated that tested samples AZ1 and AZ3 degrade at two steps. The initial stage is observed upto 261 °C and includes small weight loss (1% weight loss) due to the volatile content as water, which vaporized out, from 261 °C upto 750 °C referred to decompose the total components of the polyester fabric rapidly forming CO, CO2 and ash residue.
In the case of coated samples (AZ2 and AZ10) the degradation was achieved in three stages: up to 179 °C (18% weight loss) the volatile components were removed then the second stage of pyrolysis started as the result of the low stability of the flame retardant coating (up to 471 °C, 20% weight residue). Finally, when the temperature is higher than 471 °C in the flame-retardant coating samples, they exhibit more thermally stable properties than untreated and uncoated samples due to degradation of phosphonates, which formed phosphoric acid. It had an important role in protection polyester from ignition return to form the aggregation layer prevent oxygen to support the fire process. The Td (the temperature at which the weight loss is 10%, 20% and 50%) of uncoated samples is higher than that of the coated samples, but in the case of AZ2 and AZ10 they have the highest temperature (471 °C and 485 °C, respectively). Tmax is the maximum rate decomposition temperature, whereas, the uncoated samples have a Tmax value of 432 °C, the AZ10 has the maximum rate of decomposition temperature (445 °C).
3.8. Differential scanning calorimetry (DSC)
The heat flow (exothermic and endothermic reaction) during the thermal conversion is measured by DSC. DSC data for uncoated and coated samples are given in Table 7 and is graphically represented in Figs. 7 and 8. Fig. 7 illustrated that, the melting temperature of PES ranges from 225 to 275 °C [36,37], this varying at Tm, referred to the effect of radiation, which is directly proportional to the time. The uncoated samples were thermally decomposed through one endothermic stage between 425 and 460 °C, with flow heat of 147:110 J/g. The area under the decomposition peak is varied since recording the highest area in the case of sample AZ1 and the lowest area in AZ3 sample. From Table 6, Tg values were changed due to the effect of UV/Ozone on the polyester fabric surface since Tg of blank sample has a value of 31.6 °C while AZ3 (irradiated for 10 min) and AZ7 (irradiated for 50 min) samples have values of 33.8 and 32.7 °C, respectively.
In the case of coated samples, they have two points of Tm, one for the melting temperature of coated samples with the non-durable coating ranging between 260 and 264 °C, and the other for the polyester fabric samples (132:136 °C). The melting temperatures changed due to the semi-burning in the case of uncoated samples and a chemical reaction occurred between the irradiated surface and the OH groups from the coating. In the case of Fig. 8, there are three endothermic peaks and one exothermic peak. All coated samples were decomposed through two stages. The first endothermic stage takes place at the temperature range of 140–170 °C with the heat of decomposition of 34–81 J/g. The second stage was exothermic (435–505 °C) and occurred with the heat of decomposition ranging between 20 and 35 J/g. In the case of the AZ10 sample, the total heat of decomposition was decreased from 37.85 (A2) to 17.25 J/g, thereby suggesting higher thermal stability of the AZ10 one. In the case of uncoated samples one stage of the endothermic reaction occurred with heat flow ranging between 105 and 147 J/g, while the AZ1 (blank) sample has the highest heat of decomposition (146.49 J/g) compared to the uncoated samples due to the effect of UV/O3 and heat on the surface of samples.
Acknowledgements
The author expresses his gratitude to Prof. Dr. N.A. El-Zaher, Textile Metrology, National Institute for Standards, Egypt and Dr. Abd El Aziz Gomaa Consultant for their cooperation and help.