Development of a Finishing Process for Imbuing Flame Retardancy into Materials Using Biohybrid Anchor Peptides

Abstract

Flame retardants are commonly used to reduce fire risk in various products and environments, including textiles. While many of these additives contain harmful substances, efforts are underway to reduce their usage. Current research aims to minimize flame-retardant quantities and enhance durability against external factors. This involves utilizing anchor peptides or material-binding peptides (MBPs), which are versatile molecules that bind strongly to surfaces like textiles. MBPs can be equipped with functional molecules, e.g., flame-retardant additives, by chemical or enzymatic bioconjugation. In this research, biohybrid flame retardants and an adapted finishing process are developed. Specifically, biobased adhesion promoters, the so-called MBPs, are used to finish textiles with flame-retardant additives. To date, there is no finishing process for treating textiles with MBPs and so a laboratory-scale finishing process based on foulard was developed. Necessary parameters, such as the take-off speed or the contact pressure of the squeezing rollers, are determined experimentally. In order to develop an adapted finishing process, various trials are designed and carried out. Part of the trials is the testing and comparison of different textiles (e.g., glass woven fabrics and aramid woven fabrics) under different conditions (e.g., different ratios of MBPs and flame retardants). The finished textiles are then analysed and validated regarding their flammability and the amount of adhered flame retardants.

1. Introduction

Protection against flames is a key issue in the industry [1,2]. Over 2 million tonnes of flame-retardant chemicals were used in 2020 [3]. Flame retardants cover a highly diverse range of substances, materials and areas of application [4,5,6]. Areas of application include personal protective equipment (PPE) and fire curtains, used, for example, in airports for the purpose of separating fire compartments [7,8,9,10]. In particular, these textiles are impregnated with flame retardants [8,11].

Many of the flame retardants commonly used today are applied to textiles with adhesion promoters [12,13]. The flame retardants as well as the adhesion promoters are harmful to health and the environment [13,14]. The European Union’s REACH regulation (Registration, Evaluation, Authorization and Restriction of Chemicals) will ban around 1400 chemicals over the next few years [15,16]. Around 220 chemicals are already banned. The banned chemicals are used for flame retardancy, among other things [5,17,18,19,20,21,22].

The banned substances have properties that are harmful to health and the environment [2,14]. Brominated flame retardants are one of the substances that are harmful to health and the environment and had a market share of around 18 % in 2020 [3,23]. The use of brominated flame retardants is still subject to a time limit until the use of these substances is banned [24,25]. It is therefore necessary to develop alternatives to conventional flame retardant/adhesion promoter combinations that are less harmful to the environment than conventional flame retardants and do not fall under the REACH bans.

In addition to the REACH regulation, the sustainability concept is also crucial [14,26]. The use of innovative bio-based adhesion promoters, such as anchor peptides, extends the period until a new finish is required for the textiles [27,28,29,30,31]. As the flame retardant remains on the textiles for longer, less flame retardant is required. Extending the time interval for reimpregnation is particularly interesting in the context of maintaining PPE, e.g., in blast furnace protection [32].

Material-binding peptides (MBPs) are highly diverse in terms of their small amino acid sequence, ranging from 30 to 100 amino acids in length. MBPs have a specific material-binding ability to a broad portfolio of materials including polymers, metals, ceramics, glass, textiles, and other natural surfaces [33]. The affinity and efficient binding of MBPs to material surfaces rely on non-covalent multiple-site interactions (e.g., electrostatic, hydrophobic, π–π interactions, and hydrogen bonds), enabling highly selective and intensive binding under mild conditions [34]. For example, a mere 1 g of MBPs covering up to 250 m², at a cost of less than one cent per square meter, further solidifies their application [28]. Moreover, the binding properties can be tailored by directed evolution (e.g., KnowVolution) to meet varied formulation and application requirements [35]. The diverse binding affinities of MBPs bring a wide range of applications in plant/human health, the immobilization of catalysts, bioactive coatings, accelerated polymer degradation and analytics for micro/nanoplastic quantification [36].

To date, no industrial process is known for the sustainable finishing of textiles with material-binding peptides (MBPs) [37,38]. Due to the high specificity of MBPs, the bifunctional peptide, constructed by fusing two MBPs, brings new strategies for surface functionalization. The bifunctional peptide comprises two MBPs connected by a linker to prevent intermolecular interactions [39]. One MBP binds specifically to a material surface (e.g., textiles), while the other MBP selectively interacts with the desired functional molecule (e.g., flame-retardant additives). Applying bifunctional peptides to treat textiles is both cost-effective and resource-efficient. This process, carried out in aqueous solution, forms a dense monolayer with occupancy densities typically exceeding 90%, thereby enabling efficient endowment of material surfaces with specific functionalities [37].

In order to apply the newly developed substances, namely, flame retardants linked with bifunctional peptides, to textiles and thus to make them industrially viable, existing processes are restructured or new processes are developed [37]. In terms of redesigning existing processes, the already established padding process is integrated with the potential benefits of bifunctional peptides, which perform as bio adhesion promoters. By adapting the padding process by redefining the key process parameters, the flame-retardant additives, which are not banned by the REACH regulation, can be applied on textiles.

In this work, a parameter study for the process development of flame-retardant impregnations based on bifunctional peptides is carried out. For this, flame-retardant coating tests on a textile surface mediated by a bifunctional peptide are first carried out. Based on the results, a test plan is developed and carried out. The test plan includes the influence and interactions of various parameters, such as the dwell time of the textile, the dipping frequency and the squeezing force of the rollers. Afterwards, microscopy images are taken to determine the occupancy density of flame retardants and flame tests are carried out in accordance with DIN EN ISO 15025 to assess the flame resistance of the textiles [40]. The results are then statistically analysed and processed. The parameter study is carried out iteratively. Based on the results of previous test series, changes are made to the process and further tests are carried out. The textiles are analysed with regard to the maximum impregnability with water, the flame-retardant properties of the novel flame retardant, and the amount of flame retardant applied.

2. Materials and Methods

Preliminary tests were first carried out in order to identify influenceable and non-influenceable input variables in the finishing process. In addition, the target variables of the test series were described. On the basis of these preliminary tests, a factor test plan and the test procedure were drawn up.

2.1. Preliminary Tests for Determing Paramters

The preliminary experiments were designed to determine the most important parameters for the finishing of the textiles. The textiles were impregnated using a padder; see Figure 1. The padder, designed and constructed at Institut für Textiltechnik (ITA) of RWTH Aachen University, Aachen, consists of a dip bath, a guide roller in the dip bath, a guide roller on the draw-in side and two rollers for squeezing on the draw-off side. The lower squeezing roller is mounted in such a way that the centre distance can be flexibly adjusted.

In the first series of tests, the textiles were drawn off manually to determine the parameters for the design of experiments. In this series of tests, knowledge was gained about the behaviour of the different materials and the significant influencing variables (factors and parameters).

Initially, three different textiles were used for the tests to define the parameters. The materials have inherent flame-retardant properties and are widely used in the field of flame protection. In a further step, these textiles were treated with a coating based on polyurethane (PU). Due to the PU treatment, it was necessary to apply a top coating that protects the PU finish against fire. A side effect of the flame retardant impregnation is that the flame-retardant properties of the textiles are further enhanced. The following materials were selected for the tests; see

Table 1. Tested woven fabrics.

Woven Fabric	Material	Warp Density [1/cm]	Weft Density [1/cm]	Surface Mass [g/m2]
Optisyn	Aramid	-	-	220
KlevoGlass 660	Glass	6.8	5.8	640
KlevoGlass 550	Glass	5.0	5.0	430

:

The basin of the padder was first filled with one litre of a concentrated solution of flame retardant functionalised with bifunctional peptide CgDef-MH (FR@CgDef-MH). The concentration of the FR@CgDef-MH was c_FR = 40 g/L. The design of bifunctional peptides can refer to the previous literature, in which MH was selected as the binding peptide for Klevogalss, and CgDef was selected as the binding peptide for Addiflam Pow MCA, and then the bifunctional peptide was constructed by fusing the two parts [37]. The final finishing agent was prepared from the concentrated solution by adding three litres of 50 mM Tris-HCl buffer solution (tris(hydroxymethyl)aminomethane hydro-chloride) with a pH value of 8. The total amount of the finishing agent was four litres. The test settings are summarised in

Table 2. Experimental settings.

Characteristic	Value
Anchor peptide	CgDef-MH
Flame retardant	Addiflam Pow MCA
Volume of the flame retardant VFR [L]	1
Concentration of the flame retardant cFR [g/L]	40
Volume of the buffer solution VB [L]	3
pH (buffer) [-]	8
Total volume VTot [L]	4

. Magnetic stirrers were used in the padder basin to prevent the anchor peptides from settling in the basin and to keep the anchor peptides well dispersed.

The three different fabrics (see Table 1) were passed through the padding mangle and then air-dried. The size of the tested fabrics was 20 cm × 300 cm. After each pull-off process, the fabrics were visually checked for uniform impregnation. Following the impregnation and drying process, microscopy pictures were taken to check the impregnation coverage. In addition, flame tests were carried out in accordance with DIN EN ISO 15025 [40]. Relevant factors, parameters and target values were derived with the help of the flame tests and the light microscope images.

2.2. Design of Experiments

A 2ⁿ-factorial experimental design was selected as the experimental design. The schematic structure of a 2³-factorial test plan is shown in

Table 3. 2³-factorial test plan.

	A	B	C
(1)	-	-	-
a	+	-	-
b	-	+	-
ab	+	+	-
c	-	-	+
ac	+	-	+
bc	-	+	+
abc	+	+	+

. The higher setting of the factors A, B and C was labelled “+”; the lower setting, analogously, was labelled “-”.

The tests were analysed using statistical methods. To determine the effect A* of A on the target variable, the mean value of the differences between the settings of A+ and A- was calculated using the following formula:

$$\mathrm{A}*=\frac{1}{4\left[\left(\mathrm{a}-\left(1\right)\right)+\left(\mathrm{ab}-\mathrm{b}\right)+\left(\mathrm{ac}-\mathrm{c}\right)+\left(\mathrm{abc}-\mathrm{bc}\right)\right]}$$

(1)

The effect A is often defined as half the value of A*. With this definition, the target value deviates by the positive value of the effect when the high level of a factor is set and by the negative value when the low level of a factor is set.

$$\mathrm{A}=\frac{1}{8\left[\left(\mathrm{a}-\left(1\right)\right)+\left(\mathrm{ab}-\mathrm{b}\right)+\left(\mathrm{ac}-\mathrm{c}\right)+\left(\mathrm{abc}-\mathrm{bc}\right)\right]}$$

(2)

The procedure for determining the effects of the other factors is analogous. In addition to the effect on the target variable, the factors can also influence each other. In this case, we speak of interactions. Interactions work parallel to the actual effect. The interaction AB is calculated from the difference between the A effect on B+ and the A effect on B-. AB is defined as half of the difference.

$$\mathrm{AB}=\frac{\left[\left(\left(\mathrm{ab}-\mathrm{b}\right)+\left(\mathrm{abc}-\mathrm{bc}\right)\right) -\left(\left(\mathrm{a}-\left(1\right)\right)+\left(\mathrm{ac}-\mathrm{c}\right)\right)\right]}{8}$$

(3)

Five factors were selected that significantly influenced the finishing process. Among other factors, the temperature of the finish bath, the dwell time of the textile in the padder, the immersion frequency of the textile in the finish, the concentration of the functionalised anchor peptides in the solution, the contact pressure of the squeeze rollers and the textile itself all had an influence on the system. Based on the results, design changes were made to the padder and further test plans were developed.

2.2.1. Factors

The following factors were selected from the influencing variables:

• Dwell time;
• Contact pressure of the squeezing rollers;
• Frequency of immersion of the textile in the finish bath;
• Concentration of the anchor peptide;
• Textile.

The dwell time of the textiles can be set via the take-off speed of the padder. As an alternative to the take-off speed, it is also possible to vary the number of passes or the number of guiding rollers in the padder basin. Repeated passage of the padder realises long dwell times of the textile in the process. In the further course of the work, the dwell time was selected via the repeated passage.

The impregnation of the textile was set via the contact pressure of the squeeze rollers. This factor was set via the gap width between the two squeezing rollers. The dipping frequency was set via the distance between the deflection rollers and the take-off speed.

The fourth factor is the concentration of flame-retardant chemicals in the finish bath. The concentration of the anchor peptides cannot be precisely determined as these are extracted from cells and are not purified in the production process. It can be assumed that a sufficient quantity of anchor peptides was present in the finish. The aim of varying the amount of flame-retardant chemicals was to investigate whether better flame retardancy can be guaranteed.

The fifth factor is the textile itself. However, the preliminary tests have shown that KlevoGlass 550) is best suited for impregnation due to the high binding affinity of selected anchor peptides (CgDef-MH). Therefore, the “textile” factor was also considered as a parameter in the further experiments.

2.2.2. Parameters

Parameters are those variables that cannot be influenced. The parameters in this work include:

• Temperature;
• Air pressure;
• Air humidity;
• Textile.

One property of anchor peptides is their ability to bind completely under physiological conditions. Physiological conditions are conditions at temperatures, pressures and pH values at which cells or cell components and their products do not denature and, thus, do not change their properties. As the padder is an open system at room temperature, it is not necessary to control or regulate the temperature of the finish externally (heating or cooling). As no chemical reaction takes place, no heat input into the system is to be expected. Therefore, the temperature is approximately constant at room temperature during the tests.

2.2.3. Target Values

The target value is the value that is significantly influenced by the factors. One target variable is flame retardancy. The process was, therefore, optimised with regard to flame retardancy. Flame retardancy was analysed in accordance with DIN EN ISO 15025 [40]. In the tests, the damage caused to the finished textile by a defined flame was tested. In addition, the textile was examined under a microscope. Thus, it was possible to assess the condition of the finish and the effect of the flames on the textile.

2.3. Design of Experiments Regarding Take-Off Speed and Total Impregnation Time

To ensure a continuous impregnation process, the textile was guided in a circle. The textile was sewn together at both ends for circular guidance. The textile was continuously guided through the process at a variable take-off speed. Parameter “A” describes the circular frequency ω of the motor. The motor was connected to one of the two squeezing rollers via a gear. The circular frequency of the motor was converted by the gear with a transmission ratio of i = 35.91, so that the rotation speed n of the rollers was calculated using the following Equation (1):

$$\mathrm{n}=\frac{\mathsf{\omega }}{2\mathsf{\pi }\times }$$

(4)

The gap distance between the squeezing rollers was initially set so that the squeezing rollers touch each other. In order to vary parameter “B” (squeezing force), a further series of tests was carried out. The test plan of the test series with regard to the take-off speed and total duration of the process is summarised in

Table 4. Design of experiments regarding take-off speed and total duration.

Trial	(A) Immersion Frequency [Hz]/Speed [1/min]	(C) Total Duration of the Process [min]
1	0.14/1.33	(single take off)
2	0.14/1.33	30
3	0.14/1.33	60
4	0.56/5.32	30
5	0.56/5.32	60

.

2.4. Design of Experiments Regarding Squeezing Force

To carry out the series of tests to determine the squeezing force, the design of the padder was modified. The lower of the two squeezing rollers of the padder was tensioned using springs so that an adjustable squeezing force acted on the textile. The spring force was freely adjustable in a range of 0–170 N per spring, i.e., from 0–340 N in total. The test plan for the spring force is summarised in

Table 5. Design of experiments regarding squeezing force.

Trial	(A) Immersion Frequency [Hz]/Speed [1/min]	(B) Spring Length [cm]/Squeezing Force [N/cm]	(C) Total Duration of the Process [min]	Mass m0 [g]
1	0.14/1.33	25/0.35	30	340
2	0.14/1.33	30/2.9	30	328
3	0.56/5.32	30/2.9	30	342
4	0.56/5.32	30/2.9	10	342
5	0.14/1.33	30/2.9	10	346
6	0.56/5.32	25/0.35	30	338
7	0.56/5.32	25/0.35	10	342
8	0.14/1.33	25/0.35	10	328

.

The squeezing force was calculated by a moment equilibrium (see Equation (6)) around the bearing of the suspension of the squeezing rollers. The free section of the forces for the moment equilibrium is shown in Figure 2.

The torques around the centre of rotation were calculated from the cross product of the position vector $\overrightarrow{\mathrm{r}}$ and the acting force vector $\overrightarrow{\mathrm{F}}$. The torques rotated around the bearing.

$$\overrightarrow{\mathrm{M}}=\sum \overrightarrow{\mathrm{r}} \times \overrightarrow{\mathrm{F}}$$

(5)

The spring force F_Spring was calculated using Hook’s model. The spring force was calculated from the product of the spring constant c_S and the elongation Δl; see Equation (3). Δl describes the deflection of the spring from its rest position. The initial length of the spring was 20 cm. The spring constant c_F was c_F = 1.2 N/mm.

$${\mathrm{F}}_{\mathrm{Spring}}={\mathrm{c}}_{\mathrm{S}} \ast \mathsf{\Delta }\mathrm{l}$$

(6)

The weight force was calculated from the product of mass m and gravity acceleration $\overrightarrow{\mathrm{g}}$ (see Equation (4)).

$${\mathrm{F}}_{\mathrm{Weight}}=\mathrm{m} \ast \overrightarrow{\mathrm{g}}=56.3 \mathrm{N}$$

(7)

Considering the geometry of the padder and the angle functions, the following moment equilibrium results (Equation (5)) for a padder with two springs connected in parallel:

$$2 \ast {\mathrm{F}}_{\mathrm{Spring}} \ast 0.08\mathrm{m}=({\mathrm{F}}_{\mathrm{Squeezing}} \ast \sin (80°)+{\mathrm{F}}_{\mathrm{Weight}} \ast \cos (10°\left)\right) \ast 0.15 \mathrm{m}$$

(8)

Rearranging the moment equilibrium results in the squeezing force (Equation (6)).

$${\mathrm{F}}_{\mathrm{Squeezing}}=(2 \ast {\mathrm{F}}_{\mathrm{Spring}} \ast 0.533 - {\mathrm{F}}_{\mathrm{Weight}} \ast \cos (10°\left)\right)/\sin (80°)$$

(9)

The textiles used for the tests to determine the squeezing force were first stored in a climatic chamber at a constant temperature (20 °C) and humidity (65%) for at least 24 h. The textiles were equally loaded with water during storage. The textiles were first weighed and then fed into the padding process (with tap water) for a predefined dwell time. The respective water load of the textile was calculated from the difference between the masses of the dry textile and the soaked textile. The evaluation of the test series to determine the squeezing force was carried out by calculating the loading of the textiles. The evaluation shows that a squeezing force of 2.9 N/cm leads to a stable process.

In the series of tests using the flame-retardant finish, the FR@CgDef-MH is observed to sink due to the weight of FR. Based on the findings from all previous test series, the squeezing force was set to 2.9 N/cm and a circulation pump was installed in the dip tank. The circulation pump prevented the FR@CgDef-MH from sinking to the bottom of the tank by continuously mixing the finish.

2.5. Design of Experiments Regarding the Concentration of Flame Retardants

In a final series of tests to determine the process duration and the flame retardant concentration, “KlevoGlass 550” was finished with a flame-retardant finish with different concentrations of FR@CgDef-MH in the adapted process control. The textiles finished with different concentrations of finishes were coated with a flammable adhesive based on polyurethane (PU) for the test. The textiles treated with flame-retardant finish and PU were then subjected to a flame test in accordance with DIN EN ISO 15025 [40]. The boundary conditions for this test are summarised in

Table 6. Boundary conditions for final tests.

Characteristics	Value
Anchor peptide	CgDef-MH
Flame retardant	Addiflam Pow MCA
Concentration of the flame retardant cFR [g/L]	Factor
Volume of the buffer solution VB [L]	Variable
pH (buffer) [-]	8
Total volume VTot [L]	Variable
Take-off speed [Hz]/rotation speed [1/min]	5/1.33
Spring length [cm]/FSqueezing [N/cm]	30 (Δl = 10 cm)/2.9

.

After finishing the textiles (see

Table 7. Test plan for finishing the textiles at different concentrations of the flame retardant.

Trial	(D) Concentration of Flame Retardant cFR [g/L]	Total Time of Impregnation [min]	m0 of Textile [g]
1	40	30	336
2	40	60	326
3	28.6	30	334
4	28.6	60	338

), the basis weight of the dried textiles was measured to determine the loading of the textiles with flame retardants. The dried textiles were also examined microscopically with regard to the visual appearance of the finish and in the flame test according to DIN EN ISO 15025 with regard to the flame behaviour [40].

The PU-based adhesive was applied to the impregnated textiles after they had dried completely. The adhesive was mixed in two batches with the same formulation. The textiles treated with adhesive were flame tested in accordance with DIN EN ISO 15025 [40].

Table 8. Surface weights of the samples of the final test series with adhesive on the anchor peptides (samples 4.1–5.3 are different batches of the adhesive with the same formulation).

Trial	Surface Weight [g/m2]	Average Surface Weight [g/m2] (Incl. Flame Retardant and PU-Based Adhesive)
1.1	644	639
1.2	633
1.3	641
2.1	636	649
2.2	644
2.3	666
3.1	636	643
3.2	637
3.3	655
4.1	647	653
4.2	654
4.3	659
5.1	649	646 (only adhesive)
5.2	644
5.3	644

lists the basis weights and the average basis weights of the individual test specimens.

3. Results and Discussion

The series of tests to determine the dwell time, immersion frequency of the textile, the squeezing force and the concentration of the flame retardants in the flame retardant finish were examined using an optical microscope, weight measurement and standardised flame tests in accordance with DIN EN ISO 15025and evaluated using statistical methods [40]. Light microscopy was used to evaluate whether the flame retardants adhere to the fabrics. If the flame retardants adhere to the fabric, the MBPs are functional. In addition, the loading was evaluated using the basis weight before and after finishing. The flame-retardant behaviour was also tested. If the fabrics do not burn, the MBPs are also flame-resistant. The effects of the factors (immersion frequency, dwell time in the process, squeeze-off force and concentration of the flame retardant in the flame-retardant finish) and interactions were analysed. Three tests were carried out for each sample. Measured values were averaged to analyse the factors and their interactions.

3.1. Evaluation of the Test Series to Determine the Immersion Frequency and Dwell Time

In the following sections, the evaluation of the test series to determine the dwell time and immersion frequency is described. The samples were first analysed regarding their uniformity of impregnation. Afterwards, the flame resistance was determined. The following abbreviations are used for the samples:

• 1-5: pulled off once;
• 5-30: 5 Hz motor frequency, 30 min dwell time;
• 5-60: 5 Hz motor frequency, 60 min dwell time;
• 20-30: 20 Hz motor frequency, 30 min dwell time;
• 20-60: 20 Hz motor frequency, 60 min dwell time.

3.1.1. Light Microscopic Analysis of the Textiles

The applied finish is recognisable at a magnification of 10× under an optical microscope. The individual particles of the flame-retardant finish look spherical and darker than the textile “KlevoGlass 550”. The shape of the particles of the flame-retardant finish makes it possible to compare the amount of flame retardant on the glass fibre filaments of the fabrics. The samples were analysed by optical observation and comparison of the photographs taken by optical microscopy. Figure 3 shows representative microscopy images from the series of tests to determine the dwell time. Magnifications of 10× and 20× are shown side by side. The images are arranged vertically at different parameter levels. The images are compared in terms of the quantity of flame-retardant particles and the distance between the flame-retardant particles and the glass fibre filaments.

Compared to the other samples in Figure 3, the largest amount of flame retardants can be seen on the surface of sample “5-60”. The smallest amount of functionalised flame retardants is present on sample “20-30”. The dark dots in the figures are the functionalised anchor peptides. It can be clearly seen that there are more dark dots on the glass fibre filaments in the marked areas of sample “5-60” than in the marked areas of sample “20-30”; see Figure 4.

A comparison of samples “5-60” and “5-30” as well as samples “20-60” and “20-30” shows that after a longer dwell time (60 min), more functionalised flame retardants can be seen on the surface of the textile than with a shorter dwell time (30 min). The dwell time has a positive influence on the finishing with anchor peptides.

A comparison of the microscopy images of the fabric samples “5-60” and “20-60” as well as “5-30” and “20-30” shows that the immersion frequency of the textiles has a slight negative influence on the amount of functionalised flame retardant on the surface of the fabrics. The amount of flame retardant on the surface of the textiles is greater at a lower immersion frequency (0.14 Hz) of the textile than at a higher immersion frequency (0.56 Hz).

The textile finishes are further assessed by comparing the impregnated textiles with each other. The assessment is made in three categories: strong/high (1), moderate (0.5) and low (0). The results are summarised in

Table 9. Summary of the results of the test plan with regard to dwell time and motor frequency. 1 = strong/high, 0.5 = moderate and 0 = low. 1-5: Pulled off once; 5-30: 5 Hz motor frequency, 30 min dwell time; 5-60: 5 Hz motor frequency, 60 min dwell time; 20-30: 20 Hz motor frequency, 30 min dwell time; 20-60: 20 Hz motor frequency, 60 min dwell time.

Sample	Amount of Anchor Peptide	Bonding Strength
1-5	0.5	0
5-30	0.5	0.5
5-60	0.5	1
20-30	0.5	0.5
20-60	1	1

.

3.1.2. Evaluation of the Flame Tests Based on DIN EN ISO 15025 [40]

The evaluation of the tests is based on the damage caused to the samples by flaming. The damage is categorised into two categories. Firstly, the formation of holes is assessed, and secondly, the flammability of the material after the flame has been removed from the material. The result of the test is assessed as 0 if no hole is formed or the textile does not ignite. If a hole is formed or the textile ignites, the result of the test is rated as 1. Several test specimens are tested for each factor level. None of the samples developed a hole nor did the textiles ignite. The best-performing sample of the finished “KlevoGlass 550” is shown in Figure 5, as an example.

No in-depth statistical analysis was carried out in the evaluation of this series of tests, as no effect of the factors and no interactions between the factors can be recognised in the results.

3.2. Statistical Evaluation of the Test Series with Regard to the Applied Squeezing Force

Table 10. Evaluation of the test series regarding the squeezing force.

Trial	m0 [g]	m1 [g]	Δm [g]	Load [g/g]
1	340/338	422/428	82/90	0.24/0.27
2	328/328	410/408	82/80	0.25/0.24
3	342/340	426/432	84/92	0.25/0.27
4	342/338	446/434	104/96	0.3/0.28
5	346/346	438/432	92/88	0.27/0.25
6	338/338	424/436	86/98	0.25/0.29
7	342/338	444/428	102/90	0.3/0.27
8	328/328	412/414	84/86	0.26/0.26

analyses the samples in relation to the spring force applied by the pressing springs. Two samples per factor level were tested with water. The parameters are as follows:

• Immersion frequency;
• Squeezing force (spring length);
• Immersion time.

The evaluation was carried out by determining the load of the textile samples. The mass Δm of the water in the textile is related to the mass of the dry textile.

The data are averaged in

Table 11. Average values for evaluating the squeezing force.

Trial	m0 [g]	m1 [g]	Δm [g]	Load [g/g]	Name for Calculation of Effects and Interactions
1	339	425	86	0.25	c
2	328	409	81	0.25	bc
3	341	429	88	0.26	abc
4	340	440	100	0.29	ab
5	346	435	89	0.26	b
6	338	430	92	0.27	ac
7	340	436	96	0.28	a
8	328	413	85	0.26	(1)

and then analysed with regard to the effect and interactions of the factors (compare Table 3).

The effects of the three parameters, immersion frequency, spring length and the total duration of the process, were calculated according to Equation (1). The effects of the factors are listed in

Table 12. Effect of immersion frequency, spring length and the total duration of the process on the loading of KevoGlass550.

Factor	(A) Immersion Frequency	(B) Spring Length	(C) Total Duration of the Process
Effect	7.5 × 10−3	−1.25 × 10−3	−7.5 × 10−3

.

From the evaluation of the effects, it can be seen that none of the factors set in the tests has a significant influence on the impregnation of the textile. In the next step, possible interactions between the factors are analysed according to Equation (3).

The interactions between take-off speed (A), spring length (B*) and the total duration of the process (C) are analysed in

Table 13. Interactions of immersion frequency, spring length and the total duration of the process.

Factors	AB	AC	BC	ABC
Interactions	0	−2.5 × 10−3	−2.5 × 10−3	−2.5 × 10−3

.

It can be seen that there are no significant interactions between the factors (all values are almost 0). During the tests, it can be seen that a greater spring force leads to the textile being pulled through the padder much better. If the spring force is too low, the fabric is not pulled through the padder in a controlled manner. The pressure on the take-off roller is not high enough to guide the textile through the process. The frictional force between the textile and the driven squeezing roller is also greater due to the greater spring force. As a result, the textile is guided through the process with minimal disruption. It is, therefore, beneficial for the stability of the process control to realise the highest possible squeezing force.

Table 14. Subjective evaluation of process stability with different squeeze force settings.

Setting of the squeezing force	Low (0.35 N/cm)	High (2.9 N/cm)
Subjective evaluation	Poor	Good

summarises the subjective evaluation of the process stability with different settings of the squeezing force.

3.3. Statistical Evaluation of the Final Test Series with Regard to the Loading of Flame Retardants Functionalized with Bifuanctional Peptides

Table 15. Evaluation of the final test series without adhesive, and evaluation of the grafting with anchor peptides.

Trial	Surface Weight [g/m2]	Average Surface Weight [g/m2]	Amount of Flame Retardants (Linked with Bifunctional Peptide) per Surface Weight [g/m2]
1 (d)	583/590/598	590	40
2 (dc)	600/586/591	592	42
3 (1)	590/595/587	590	40
4 (c)	602/585/580	589	39
5 (-)	Not impregnated	550	0

analyses the loading of the textiles with flame retardants functionalized with bifunctional peptides (CgDef-MH). It can be seen that the basis weight of the textiles is not significantly affected by varying the time and the concentration of flame retardant in the finish. Therefore, the concentration of the flame retardant and the process time are chosen to be low in order to keep the costs of the process low.

The averaged values are analysed in

Table 16. Effect of concentration and immersion time on the surface weight of textiles.

Factor	(A) Concentration	(B) Total duration of process
Effect	0.75	0.25

for their respective effect on the dwell time. The effects of immersion time and concentration on the basis weight are rather small but significant. A higher concentration of flame retardant has three times as strong an effect on the basis weight as a longer dwell time of the textile in the process.

The interaction between the concentration of the flame retardant in the finish and the immersion time of the textile in the process is summarised in

Table 17. Interaction of concentration and immersion time.

Factor	(DC)
Interaction	0.75

. The factors favour each other. With a high concentration of flame retardant, a long dwell time of the textile in the process also has a positive effect on the basis weight.

Furthermore, the images of differently finished textiles were compared under the microscope. Textiles are shown that have not yet been treated with the adhesive. Figure 5 shows a sample that has been finished for 60 min in the finishing agent with a flame retardant concentration of 26.8 g/L. In Figure 6, it can be seen that the flame retardant does not form any clusters and occurs rather selectively but at a high density. The functionalised flame retardants are densely deposited on the filaments of the glass fibre fabric. So, it can be interpreted that the flame retardant treatment using a bifunctional peptide as an adhesion promoter is successful.

Figure 7 also shows a large amount of flame retardant on the textile, where the amount of bifunctional peptide is highest at the factor settings of 60 min and 40 g/L. Clusters have formed from the flame retardant impregnation and the flame retardants are densely deposited on the glass fibre filaments. It can be interpreted that the flame-retardant treatment using the bifunctional peptide as an adhesion promotor is successful under the boundary conditions selected for this sample.

Additionally, the flame resistance according to DIN EN ISO 15025 of the textiles treated with the flame retardant finish and adhesive was analysed. As none of the samples were damaged in this flame test (no hole formation or flame formation), the results were not analysed statistically. In this case, no knowledge can be gained from a statistical evaluation.

4. Conclusions

In this work, the textile padding process parameters were adapted to develop a new, sustainable flame retardant finishing process, based on material-binding peptides. This investigation is inspired by the current legislation of the European Union, the so-called REACH regulation, as well as the concept of sustainability.

Finished flame-retardant textiles are used, for example, in the field of flame-retardant curtains. The aim of the work is to adapt a laboratory-scale process for the finishing of glass fibre fabrics with flame retardants based on MBPs. In order to carry out the adaptation and optimisation, various full-factorial 2ⁿ test plans were developed and implemented in iterative processes.

The textiles were tested by weighing, microscopy and standardised surface flaming. Based on the respective test results, the process was adapted and a new iteration was started.

It was found that a high concentration of the flame retardant has a positive effect on the finishing of the selected textile. Due to the high concentration of the flame retardant, the driving force of adsorption on the textile surface is greater than with lower concentrations. Glass fibre fabric is inherently flame-resistant and the bifunctional peptide (CgDef-MH) presence on flame retardants also does not burn. The squeezing force has a minimal negative effect on the impregnation of the textiles with water. The positive effect of the squeezing force on the process stability clearly outweighs this. The textile is not ignited during the surface flame treatment, and there are no damages in terms of hole formation. So, it can be concluded that the flame retardant treatment based on material-binding peptides is successful.

The best results for flame retardant impregnation were achieved with the following factor levels:

• Dwell time of the textile in the process (60 min);
• Immersion frequency of the textile in the finish (0.14 Hz);
• Squeezing force (2.9 N/cm);
• Concentration of the flame retardant in the finishing agent (40 g/L).