During this study, a P, S and Si-containing flame retardant was used for modification of cotton fabric (Fig. 2).Fig. 2Structure of flame retardant.The compound was obtained as a result of the thiol-ene click reaction via the addition of SH group-containing phosphate to a HC=CH2 bond of triethoxyvinylsilane in the presence of a photoinitiator and UV irradiation. The method of synthesis was simple and efficient, and the product obtained with almost 90% isolation yield (NMR and FT-IR spectra are included in Supplementary Information). The obtained derivative was used for modification of cotton fabric via sol–gel process. During this process triethoxysilyl groups of synthesized silane were firstly hydrolyzed by water in acidic condition to silanol groups and next condensed to form a crosslinked network on the cotton fabric surface. A schematic illustration of modified fabric is shown in Fig. 3.Fig. 3Proposed structure of modified cotton fabric (cellulose) by silane via sol–gel process.During the study mercerized and non-mercerized cotton fabrics were treated. Table 1 shows the method of sample modification and their add-on values.Table 1 Explanation of sample symbols and add-on values.Analyzing the data presented in Table 1, it can be seen that all modified fabrics have relatively low add-on values. The add-on value increases with the silane concentration in the modifying compositions. Mercerized fabrics are characterized by slightly higher add-on values compared to analogous non-mercerized samples. The mercerization process activates the surface of cotton fibers and creates –O–Na + active centers, which promotes the formation of bonds between the modifier and cellulose. Relatively low add-on values are very advantageous from both an economic and functional point of view.The cotton samples (2.5, 5 and 10 wt.% silane concentration) were analyzed with microcalorimeter PCFC. Heat release rate (HRR) curves of samples before washing process are presented in Fig. 4.Fig. 4Heat release rate (HRR) of samples before washing process.Figure 5 shows the remaining data obtained during the microcalorimetric analysis, including: the maximum rate of heat release (Qmax), the temperature of reaching the maximum rate of heat release (Tmax) and the time of achieving the maximum rate of heat release (t at Qmax).Fig. 5PCFC analyses of modified samples before washing process.The results presented in Figs. 4 and 5 show that the applied modifications significantly reduced the HRR value compared to the unmodified fabric. Samples treated with a 2.5% silane solution are characterized by a HRR reduction of over 40% compared to the raw fabric. This is a very large difference in the heat release rate at an extremely low concentration of the modifying composition and the add-on value of the sample. The curves presented in Fig. 4 illustrate that as the concentration of the composition increased, the modifying heat release rate values decreased. Mercerized fabric impregnated with a 10% silane solution (S10M) shows a HRR reduction of almost 75% compared to the raw fabric. Moreover, the obtained results showed that mercerized samples after the impregnation process had slightly lower heat release rate values compared to analogous, non-mercerized samples. The mercerization process activated the hydroxyl groups of cellulose and created –O–Na + active centers, which promoted the formation of bonds between the modifier and the fiber surface. This process had a positive effect on reducing the HRR of modified fabrics.The findings depicted in Figs. 4 and 5 unequivocally demonstrate a notable decrease in both the temperature of maximum heat release rate (Tmax) and the time required to attain it (t at Qmax) for the modified samples in comparison to raw cotton. This reduction represents a significant alteration in the thermal behavior of the samples. It is noteworthy that the temperature at which the maximum release rate occurs aligns with the temperature of maximum pyrolysis rate. It is crucial to emphasize the disparity between the conditions of a microcalorimeter and a cone calorimeter. Unlike the latter, where combustion of the sample takes place, a microcalorimeter operates with the sample situated in an inert gas atmosphere (N2). The temperature gradually escalates until reaching Tmax, at which point pyrolysis commences. Subsequently, the resultant pyrolysis products, primarily flammable gases, are introduced into the chamber where they mix with oxygen, leading to combustion. The determination of heat release rate hinges on the measurement of oxygen loss. As documented in the literature, one effective strategy to mitigate carbon dioxide emissions involves maximizing carbon residue at the expense of reducing flammable gases. It is a well-established principle that lower temperature pyrolysis fosters greater carbon residue formation, consequently diminishing the production of flammable gases. In essence, the observed reduction in Tmax and time to reach Qmax signifies a mitigation of flammability, a highly advantageous outcome. Moreover, the values of Tmax and t at Qmax decrease with increasing silane concentration in the modifying composition.Samples were washed 10 times to confirm durability of modification. After 10 washing cycles, samples were analyzed on a PCFC pyrolysis microcalorimeter (Figs. 6 and 7).Fig. 6Heat release rate (HRR) of samples after 10 washing cycles (W).Fig. 7Data from PCFC analyses of samples after 10 washing cycles (W).One of the most challenging aspects at in fire-retardant modification of fabrics is the durability of the created coatings. A huge number of developed modifications are not resistant to washing or are only partially durable. However, due to the aspects of use, the durability of the produced coatings is extremely important. The results presented in Figs. 6 and 7 clearly confirm the durability of the developed modifications. Both the shape and course of the curves from washed fabrics are almost identical to those from unwashed samples. Putting the sample through 10 washing cycles slightly changes in the heat release rate. Also, the remaining data obtained during the PCFC analysis barely changed as a result of washing. The alkoxy groups present in the structure of the silane modifier (Si(OC2H5)3 were transformed into hydroxyl groups as a result of the hydrolysis process. This process enabled the formation of covalent bonds on the fiber surface. The obtained results allow us to conclude that the bonds formed are durable and the silane remains on the fiber surface even after repeated washing. The creation of durable fire-retardant coatings on the fiber surface is a spectacular success.Both before washing and after undergoing ten, twenty, and thirty washing cycles, the modified fabric samples were subjected to analysis of the measurement of the limited oxygen index (LOI) (Fig. 8).Fig. 8LOI of modified samples before washing and after being ten, twenty, and thirty washing cycles.The results presented in Fig. 8 allow us to conclude that all the applied modifications increased the LOI values; however, the differences between the modified and unmodified cotton are not spectacular. The limited oxygen index values increase with the concentration of silane in the modifying composition. Figure 8 shows that mercerized samples, after impregnation with the silane solution, show slightly higher LOI values than non-mercerized. It is worth noting that although subsequent washing cycles caused a slight decrease in LOI values, the differences are not significant.Next, the thermal stability of the modified and unmodified fabrics was examined. For this purpose, thermogravimetric analyses (TGA) were performed in a nitrogen atmosphere (Table 2). Thermogravimetric curves of unwashed and washed samples are shown in Figs. 9 and 10.Table 2 Temperatures of mass loss of samples in the nitrogen atmosphere.Fig. 9TG curves of modified samples before washing.Fig. 10TG curves of modified samples after washing.The TGA analysis shows a considerably improved thermal stability for modified cotton fabrics compared to pure cotton fabrics. Pure cotton begins to degrade at 290 °C and has 0.38% residue at 700 °C. On the other hand, the modified samples present much higher values of residues, pointing to their higher thermal stability. For example, the mercerized cotton sample modified with a 10% modifier (S10M) presents the highest residue, which is 36.8%, and the washed sample (S10MW) demonstrates a residue of 36.9%. These large residue values point to the efficiency of the modification in enhancing the thermal resistance of the fabric.As can be seen from the comparison between the washed and the unwashed samples, the thermal degradation is almost the same, which means that washing has minimal effect on the modification’s efficiency. Specifically, for the S5M mass loss occurs at 241 °C, 258 °C, and 277 °C for 5%, 10%, and 20% weight loss, and a residue of 21.4%. For the S5MW washed sample, the corresponding mass loss temperatures are 243 °C, 262 °C, and 281 °C, with a residue of 14.2%; the slight differences in these values suggest that the modifier is still primarily intact after washing, thus keeping the thermal properties in the fabric. The 2.5% modified sample (S2.5) has mass loss temperatures at 245 °C, 266 °C, and 283 °C, with a residue of 11.8%, and the washed version (S2.5W) shows mass loss temperatures at 249 °C, 270 °C, and 288 °C, with a residue of 8.4%. Similarly, the 10% modified sample (S10) shows mass loss temperatures at 237 °C, 253 °C, and 271 °C, with a residue of 29.5%, while the washed version (S10W) shows temperatures at 235 °C, 251 °C, and 269 °C, with a residue of 27%. These results confirm minor changes due to washing for the 2.5% and 10% modified samples.Mercerization also affects thermal stability, generally resulting in slightly lower mass loss temperatures but higher residues. For example, the mercerized sample with a 10% modifier (S10M) showed mass loss temperatures at 234 °C, 250 °C, and 269 °C, with a residue of 36.8%, while the non-mercerized sample (S10) showed slightly higher mass loss temperatures at 237 °C, 253 °C, and 271 °C, but a lower residue of 29.5%. This indicates that mercerization enhances the binding or incorporation of the modifier into the cotton fibers, resulting in improved thermal stability. Differences in results are minimal for the 5% modified samples, where the unwashed sample (S5) shows mass loss temperatures at 248 °C, 264 °C, and 279 °C, with a residue of 20.6%. In contrast, the washed sample (S5W) shows mass loss temperatures at 244 °C, 265 °C, and 283 °C, with a residue of 15%.Based on TG analyses, DTG derivatives were also determined. Table 3 lists the main DTG peaks and the mass loss rate they achieve. Figures 11 and 12 show the DTG curves for samples before and after washing.Table 3 DTG onset and the temperatures at which it occurred in nitrogen atmosphere.Fig. 11DTG curves of modified samples before washing.Fig. 12DTG curves of modified samples after washing.The DTG results also show significant reductions in the degradation rate of the modified fabrics compared to the pure cotton. Pure cotton has a DTG peak at 361 °C coupled with a degradation rate of 2.6%/°C. The modified samples have a much lower rate of degradation. For example, the peak of the S10M sample is at 252 °C, with a degradation rate of 0.60%/°C, and that of the washed sample (S10MW) is 292 °C, with a rate of 0.71%/°C. The degradation rate is reduced significantly with the modification treatment, enhancing thermal stability. The differences between the washed and unwashed samples are minor, a property representing modifying durability. For example, the 5% modified sample (S5M) peaks at 295 °C with a rate of 0.89%/°C; in contrast, the washed sample (S5MW) peaks at 302 °C with a rate of 1.19%/°C. The difference is, however, slight, and it may be inferred that the modifier is not heavily washed out, hence the thermal stability of the fabric is preserved. In general, the mercerized samples show lower rates of degradation compared to the non-mercerized, indicating that there is a better incorporation of the modifier. For instance, the 2.5% mercerized sample (S2.5M) has its peak at 308 °C at a rate of 1.48%/°C, while the non-mercerized sample (S2.5) has its peak at 303 °C at a rate of 1.31%/°C. In the curves derived from fabrics modified with 10% solutions, two peaks are visible. This is due to the decomposition of the modifier occurring first, followed by the decomposition of the cotton fabric. It is important to note that the decomposition of the modifier occurs at lower temperatures than the fabric, which is advantageous for flame retardants. Furthermore, these results correlate with the HRR curves.By comparing these DTG results with TGA data, an integral concept of the thermal behavior of the fabrics is given. Considering the decreased degradation rates in the DTG, the high residues obtained by TGA at 700 °C with the modified samples are 36.8% for S10M and 36.9% for S10MW. This indicates that the existence of the modifier does not only reduce the degradation rate but also leads to higher thermal stability and a lower extent of mass loss in general. The beginning of thermal degradation in TGA lies in the data for lower temperatures of modified samples—for example, 234 °C for the S10M sample. It corresponds to DTG data whereby the modified samples, in turn, show lower maximum temperatures—for instance, 252 °C for S10M. It does improve the validity of the conclusion about the decisive impact of modification on the thermal properties of such fabrics.TGA and DTG results show that washing had a minimum influence on thermal properties. For example, TGA degradation of 21.4% and 14.2% for S5M and S5MW, respectively, with the same degradation trend in the DTG (0.89%/°C for S5M and 1.19%/°C for S5MW), the results are proofs that the modification is stable to washing. It is possible to sum up that the results obtained from thermogravimetric and derivative thermogravimetric analyses underline the positive influences of modifying cotton fabrics. The modification led to considerably less degradation and higher residues, proving higher thermal stability. The slight differences between washed and unwashed samples further underline the durability of the modification. Lower onset temperatures are found only in mercerized samples but are thus associated with better incorporation of the modifier and therefore better thermal features. This further underlines the efficiency of the chemical method in increasing the thermal performance and durability of cotton fabrics.Subsequently, vertical burning tests were performed on the modified samples. The test results are presented in Fig. 13 and Table 4.Fig. 13Vertical burning test results of modified samples.Table 4 Vertical burning test results of modified samples.The modifications resulted in significant changes in the burning process of the samples compared to the raw cotton fabric. It is evident that the modified fabrics exhibit a much larger char residue after the burning process, whereas the raw fabric burned completely. Moreover, for samples modified with the highest concentration of silane, unburned fragments of the fabric are visible. It is noteworthy that modified fabrics have a zero after-glow time. Additionally, the burning time was extended compared to the unmodified fabric.The samples were subjected to FT-IR analyses to confirm the effectiveness of the modification (Fig. 14).Fig. 14FT-IR spectra of a raw cotton fabric and modified cotton fabrics (S2.5M, S2.5MW, S5M, S5MW, S10M and S10MW).The FT-IR spectra of pure cotton fabric and silane-modified cotton fabrics (S2.5M, S5M and S10M washed and non-washed) are shown in Fig. 14. The spectra of modified samples show the bands at about 2976 cm−1 coming from CH stretching vibration of CH3. The spectra show bands coming from POCH2CH3, P=S and P–S of organophosphorus group at about 1009 cm−1, 976 cm−1, 824 cm−1, 795 cm−1 and 658 cm−1. These bands are the most intense for samples modified with the highest silane concentration of 010%, both unwashed and washed (S10M and S10MW) and invisible in the spectra for samples modified with a concentration of 2.5%. The bands in the range 950–1100 cm−1 are attributed to the characteristic peaks of cellulose and overlap the Si–O–Si bands.The cotton samples, including the sample of unmodified cotton as a reference sample, and sample S10M after burning were subjected to SEM microscopic analysis (Fig. 15).Fig. 15SEM images of pure cotton, modified cotton fabrics and fabric S10M after burning.In the first photo of pure cotton the varied fiber morphology typical of raw fabric is visible. In the subsequent photos clear differences resulting from the modifications can be seen. The photos of the modified samples show a layer of silane on the surface of the fibers. The photos show a homogeneous siloxane layer clearly surrounding the fibers. Analyzing the SEM photos, it can be assumed that the silane not only formed bonds with the fibers but also condensed between them, producing a siloxane layer. Moreover, it can be seen that as the concentration of silane in the modifying composition increases, the more distinct layer formed on the fiber surface. When using a 10% solution, single agglomerates also appear. It is important to note that the silane layer is clearly visible in the photos of washed samples and the photos look similar to unwashed samples. This proves the durability of the modifications after repeated washing. The last photo shows the S10M sample after burning. We can observe preserved cotton fibers. Modification promoted the formation of a compact char layer which can prevent further combustion by cutting off the heat and oxygen. SEM photos of the remaining samples are included in the Supplementary Information.To confirm the presence of silane on the fiber surface, SEM–EDS elemental analysis was performed (Table 5).Table 5 SEM–EDS results of modified samples before and after washing.The results presented in Table 5 clearly confirm that the modification was successful. There are clearly visible differences in the elemental composition of modified fabrics compared to unmodified fabrics. Impregnated samples are characterized by the presence of silicon, phosphorus and sulfur in their composition. These are elements characteristic of silane used to reduce the flammability of fabrics. Therefore, the presence of the mentioned elements in the modified fabrics clearly confirms that the modification has occurred. The content of elements characteristic of the silane used increases with its percentage concentration in the modifying compositions. Moreover, mercerized samples have slightly higher contents of elements characteristic of the modifier compared to non-mercerized ones. These results correlate perfectly with previous analyses. Moreover, what is most noteworthy is that the elemental composition of unwashed and washed samples remains almost unchanged. The SEM–EDS analysis once again confirmed the durability of the modifications made. The process of washing 10 times did not cause any significant changes in the elemental composition of the samples. All element values of washed and analogous unwashed samples remain at a very similar level. A slight decrease in the content of elements after washing may be caused by the washing out of particles that did not form bonds with the fiber surface and only accumulated them. However, washing away this small amount of modifier did not result in any deterioration of the remaining parameters as a result of washing. The element ratio of cotton before and after combustion changed significantly. The weight percentage of C increases after burning from 33.9 to 45.2% indicating that the fibers were carbonized during combustion. The weight percentage of P and Si increases considerably after burning. Based on the obtained results, conclusions can be drawn regarding the mechanism of flame retardancy of the applied modification, that P and Si remain in condensed phase and promoted the formation of a char layer. Silicon-containing part form siliceous barrier and protect from heat flux and volatile formation37. The weight percentage of S decreases which may be related to the gas-phase flame retardant mechanism of sulfur.The bending length (stiffness) and flexural rigidity of the pure cotton and the modified cotton samples are shown in Table 6.Table 6 Bending length and flexural rigidity of pure cotton and modified cotton fabrics.From the results given in Table 6, one can realize that the applied modifications do not, to a great extent, influence the flexibility of the fabric. The fabric bending length was slightly increased with the increase of silane concentration and the process of mercerization has not influenced on additional increase in fabrics stiffness. The bending length for the mercerized fabric modified with a 10% solution (S10M) increases by only 2 mm as compared with the raw fabric. Further, the flexural rigidity of the modified fabrics is very near to that of the unmodified fabric. These results would indicate that the process of mercerization has not influenced the stiffness of the fabrics and modification using both 2.5%, 5% and 10% silane solutions does not significantly affect the stiffness of fabrics. The elasticity of the chemically modified fabrics is very important with regard to their usability.
