Characterization of the coagulation process: wet-spinning tool development and void fraction evaluation

Abstract

Since very little research has been published on the wet-spinning process, this article is a contribution to understanding the coagulation phenomenon. In this work, the effects of spinning parameters on the evolution of the void fraction of coarse polyamide-imide monofilaments were investigated. First of all, an experimental wet-spinning device was designed in our laboratory. Robust and reliable, this prototyping tool allows a fully automated production, ensuring the reproducibility of the different spinning tests. A four-factor Design of Experiments was built to analyze the effects and combinations between the following parameters: extrusion diameter, concentration of the coagulation bath and drawing rates applied in the coagulation bath and in the air. The results derived from a statistical analysis coupled to a scanning electron microscope study showed a highly significant influence of the concentration of solvent in the coagulation bath, concerning the evolution of the void fraction of the obtained monofilaments.

Keywords

wet-spinning void fraction polyamide-imide spinning tool Programmable Logic Controller

This research aims to improve the performance and the comfort requirements of textile protective clothing against heat and flame. Indeed, today personal protective equipment consist of three different layers: an outer layer against heat and flame, a moisture barrier and an inner layer coupled to a thermal barrier. The main purpose is to reduce the weight of such clothing by replacing their current structure with a complex woven Spacer Fabric, composed of an assembly of only two textile layers, having different properties, and linked by a binder yarn.

This work particularly focuses on the development of this binder yarn and consists of a feasibility study to demonstrate the possibility to spin coarse monofilaments (diameter above 200–300 µm) from a polyamide-imide solution. It is important to specify that only multifilaments composed of elementary strands about 10–15 µm are currently produced. Hence, there is currently no scientific publication dealing with this topic. A dedicated experimental wet-spinning device has been designed to carry out this project. Such a fully automated machine enables one to conduct detailed studies concerning the influence of multiple spinning parameters on the physical properties of the manufactured monofilaments, therefore providing a simulation of an industrial production of such fibers. Hence, this prototyping tool has been used to develop monofilaments addressing the initially stated issue.

Previous studies^1–3 were conducted to describe the diffusion process involved in coagulation and to develop mathematical models for the wet-spinning of fibers from polymer solutions. However, due to the complexity of this issue, simplifying assumptions had to be postulated, leading to simplified models.

Based on these results, a four-factor Design of Experiments has been built to investigate the effects and combinations between the following parameters: extrusion diameter, concentration of the coagulation bath and drawing rates applied in the coagulation bath and in the air.⁴

This article particularly attempts to describe the importance of controlling the coagulation degree by highlighting significant influences of factors, especially concerning the evolution of the void fraction (VF) of the monofilaments.

The importance of this experimental work is underlined by the small number of studies published in this field.

Materials and methods

Material

Polyamide-imide is obtained from a polycondensation reaction between a diisocyanate and an acid anhydride⁵ in a 1.3-dimethyl-2-imidazolidinone (DMI) medium.⁶ It is characterized by the succession of aromatic rings that are alternately connected to amide and imide functions, as illustrated in Figure 1.

As an aromatic compound, it displays high thermal and mechanical properties, justifying its use in protective clothing manufacturing.⁷ Moreover, the amide functions enable the solubilization of the polymer and thus improve its processing.

Characterization of the rheological properties of the polymer solution used was performed using a shear rheometer (AR 2000, TA Instruments) associated to a cone/plate measuring system (R = 40 mm and α = 2°) and equipped with a Peltier effect temperature control.

The rheological study in steady state (Figure 2) of the spinning dope used for this work was conducted for a shear rate between 0.05 and 300 s^–1, at 20℃, and reflects a non-Newtonian shear-thinning behavior. Indeed, a constant viscosity is firstly noted, which will decrease with increasing the shear rate.

On the other hand, a dynamic-state study was conducted. The sample was subjected to a 10 Pa sinusoidal stress amplitude, for an angular frequency between 6 and 60 rad.s^–1. The obtained results underline the importance of the viscous behavior compared to the elastic one (Figure 3) and confirm the viscoelasticity of the polymer solution. Indeed, regardless of the value of the angular frequency applied, the shear storage modulus (G’) is higher than the shear loss modulus (G’’). Moreover, this behavior is confirmed by the value of the phase lag (δ), which is higher than 45°.

These observations will enable us to develop a better understanding of the phenomenon involved in the wet-spinning process, especially during the extrusion step.⁸

Wet-spinning process

Depending on the type of the raw material, there are three main types of production processes to form synthetic fibers:⁹

‐the melt-spinning process, where phase transformation is due to the solidification from a molten mass;

‐the dry-spinning process, where the evaporation of the solvent contained in the spinning dope leads to solidification;

‐the wet-spinning process, where solidification occurs after a countercurrent diffusion between the spinning dope and the coagulation bath.

Due to its non-flammable nature, polyamide-imide does not prevent a melt from being spun. Hence, and to stick to the current production process, monofilaments will be wet-spun (Figure 4).

The wet-spinning process is rather complex to analyze in that it involves a ternary system with possible coupling effects.^1,10 Coagulation is mainly brought about by transfers of mass and momentum. Contrary to the melt- or dry-spinning processes, temperature effects and heat transfer only play a secondary role.

The wet-spinning process is classified in the solution spinning category. Spinning polymer is firstly dissolved in the appropriate solvent and then wet-spun fibers are formed by the extrusion of the obtained solution, through the spinneret, in the adequate coagulation bath. The spinneret can be directly immersed (immersed method) or there may be a small gap between the spinneret and the surface of the coagulation bath (air-gap method). Three methods of solidification can be distinguished, which are named the liquid-crystalline, gel formation and phase separation methods. For the wet-spinning of polyamide-imide monofilaments, solidification is brought about by a diffusional interchange between the extruded polymer solution and the coagulation bath. In other words, coagulant from the coagulation bath diffuses into the nascent monofilament while solvent leaves.

The coagulation firstly occurs on the surface of the filament, creating a skin/core structure that will slow down the diffusion process as described by Fick’s second law,¹¹ which involves a notion of proportionality between the variation of concentration as a function of time and the second derivative of concentration relative to distance, as described in Equation (1):²

\frac{\partial C_{i}}{\partial t} = \frac{1}{r} \frac{\partial}{\partial r} [D_{i} (T) r \frac{\partial C_{i}}{\partial r}]

(1)

where C represents concentration, D is the diffusion coefficient, t is the time and r is the distance from the surface to the center of monofilament. For i = 0, this equation relates to the diffusion of the coagulant in the spinning dope; for i = 1, this equation relates to the diffusion of the coagulant in the forming gel; for i = 2, this equation relates to the diffusion of the solvent in the spinning dope; for i = 3, this equation relates to the diffusion of the solvent in the forming gel.

Hence, controlling the coagulation degree is a key factor for ensuring a homogeneous final product with a low VF and optimal properties.

Wet-spinning machine

Polyamide-imide monofilaments were wet-spun on an experimental automatic wet-spinning device developed in our Laboratory. Like every automated system, it consists of two basic parts: an operative part, which enables the different steps needed for the transformation of the raw material into monofilaments, and a controlling part, which ensures the coordination of those steps. To fulfill the requirements imposed by the wet-spinning process and ensure the quality of the monofilaments, the production process is controlled with a Programmable Logic Controller (PLC).The choice of equipment has been made taking into account not only the characteristics of the polymer solution, coagulation and the washing baths, but also the expected properties of the final product (Figure 5).

Firstly, the injection system has been selected according to the high viscosity of polymer solution experienced. Hence, to optimize the adaptability of the experimental device, a gear pump with magnetic drive is used. To ensure a high degree of chemical resistance, the coagulation bath is designed in polypropylene and built with a triangular shape to reduce the volume of solvent needed. The third step refers to the drawing system, which consists of two drafting rolls. Each axis is moved by a brushless servo motor through a planetary speed reducer, controlled with the PLC. Hence, the kinetic of coagulation can be controlled by adapting the injection flow and the drawing ratio applied in the coagulation bath, defined with regard to the rotational speed of the first cylinder of the drawing system.

After the washing step, a drying system is available to increase homogeneity and properties of the obtained filaments. It consists of two infrared lamps, allowing a heating power of 600 or 1200 W. Temperature and residence time in the drying chamber are adjustable parameters.

Finally, a winding system allows a regular distribution of the filament around the spool by a combination of rotary and linear motions. Like the drawing system, it consists of two brushless direct current (DC) electric motors completed by two planetary speed reducers. Figure 6 is a general panoramic view of the experimental device developed.

Scanning electron microscopy

In addition to studying their morphological characteristics, wet-spun monofilaments were also analyzed by a scanning electron microscope (SEM; Hitachi S-2360 N) to define their radius ( $r_{\exp}$ ). With this parameter, the VF can be estimated by comparing the theoretical volume ( $V_{th}$ ), using the density of the spinning dope and the volume experimentally measured ( $V_{\exp}$ ) for a 1-meter long monofilament (h):

V_{th} = \frac{Weightofsample (g)}{Volumetricmassdensityofkermel ({g / cm}^{3})}

(2)

V_{\exp} = π * r_{\exp}^{2} ({cm}^{2}) * h (cm)

(3)

Experimental details

Design of Experiments

Recall that this research aims at determining the rate of coagulation of wet-spun monofilaments under different spinning parameters¹² by evaluating their VF. A four-factor Design of Experiments has been built to complete this task and investigate the effects of the following parameters: concentration of the coagulation bath, extrusion diameter and drawing rates applied in the bath and in the air, according to Fick’s second law of diffusion and the drawn Ishikawa diagram (Figure 7).¹³

Since very little research has focused on the wet-spinning process and the control of the coagulation degree, especially in the case of coarse monofilaments, a full-factorial Design of Experiments with two levels has been selected so as not to leave out any potential interaction between the different factors. The aim is to identify an initial trend for prioritizing the influence of the selected parameters on the structure of the obtained monofilaments.

From this Design of Experiments, 16 experimental testing were carried out (Table 1). They are going to enable us to build a polynomial mathematical model (Equation (4)), linked with the evolution of the monofilament VF:

Y = b_{0} + b_{1} x_{1} + b_{2} x_{2} + b_{3} x_{3} + b_{4} x_{4} + b_{12} (x_{1} x_{2}) + b_{13} (x_{1} x_{3}) + b_{23} (x_{2} x_{3}) + b_{14} (x_{1} x_{4}) + b_{24} (x_{2} x_{4}) + b_{34} (x_{3} x_{4}) + b_{123} (x_{1} x_{2} x_{3}) + b_{124} (x_{1} x_{2} x_{4}) + b_{134} (x_{1} x_{3} x_{4}) + b_{234} (x_{2} x_{3} x_{4})

(4)

where, Y represents the measured response, that is, the VF, b₀ is a constant corresponding to the average value of the measured responses, x_i are the variables corresponding to the different factors, b_i,

b_{ij}

b_{ijk}

are the coefficients to be determined,

i, j, k = 1

is the value assigned to the concentration of the coagulation bath,

i, j, k = 2

is the value assigned to the extrusion diameter,

i, j, k = 3

is the value assigned to the drawing rate applied in the coagulation bath,

i, j, k = 4

is the value assigned to the drawing rate applied in the air.

Table 1.

Representation of the Design of Experiments

Batch n°	1^st factor	2^nd factor	3^rd factor	4^th factor
Batch n°	Concentration of coagulation bath	Extrusion diameter	Drawing rate in coagulation bath	Drawing rate in the air
B1	+	+	+	+
B2	+	+	+	–
B3	+	+	–	+
B4	+	+	–	–
B5	+	–	+	+
B6	+	–	+	–
B7	+	–	–	+
B8	+	–	–	–
B9	–	+	+	+
B10	–	+	+	–
B11	–	+	–	+
B12	–	+	–	–
B13	–	–	+	+
B14	–	–	+	–
B15	–	–	–	+
B16	–	–	–	–

Level + 1	60%	0.7 mm	40%	140%
Level –1	40%	0.5 mm	0%	100%

Experimental protocol

The spinning dope used for this study is prepared from 21% dry matter dissolved in DMI and has an apparent viscosity of 16.5 Pa. The coagulation bath consists of solvent (DMI) and non-solvent (water), for which quantities are defined in Table 1, while the washing phase only uses water.

The polymer solution is injected through a spinneret directly immersed in the coagulation bath by means of a gear pump with an extrusion rate of 0.3 mL/min. When solidified, the monofilament is drawn and, before being wound, it is washed to remove the remaining solvent. It is important to specify that spinning tests were carried out by drying the obtained monofilaments at room temperature.

Results and discussion

Results

The results described in this section give information regarding the evolution of VF of the obtained monofilaments according to the selected spinning parameters.

The different average values of the VF, obtained for each spool from a batch of 10 samples, are summarized in Table 2. Such a number of measures may appear to be low, but remember that it is a preliminary study to assess the influence of the different selected factors.

Table 2.

Average values of calculated void fraction

	B1	B2	B3	B4	B5	B6	B7	B8
Average void fraction (%)	58.12	64.02	69.01	66.53	73.73	71.70	67.08	75.45
Standard deviation	1.71	1.40	0.69	0.88	0.61	0.57	0.82	1.49
CV%	2.95	2.18	0.99	1.32	0.82	0.79	1.23	1.97
	B9	B10	B11	B12	B13	B14	B15	B16
Average void fraction (%)	80.05	75.23	72.18	73.46	75.29	75.40	73.58	74.96
Standard deviation	0.97	0.19	1.21	1.40	0.73	1.49	1.01	1.16
CV%	1.21	0.25	1.68	1.91	0.97	1.98	1.38	1.55

All other things equal, increasing the solvent concentration in the coagulation bath from 40% up to 60% seems to allow the decrease of the VF, except for batches 8 and 16. As a first assumption, this addition of DMI appears to favor a better penetration of the non-solvent into the monofilament. To validate this observation, the morphological properties of wet-spun monofilaments have been investigated by scanning electron microscopy. Figure 8 shows the scanning electron micrographs of the different batches for a 0.7-mm extrusion diameter.

Method of analysis

Statistical analysis of results was conducted with the assistance of the software program Nemrodw (LPRAI Society) to prioritize the influence of the selected factors on the evolution of the VF of the obtained monofilaments. It is divided into three different parts:

a mathematical analysis, which uses the least squares method to estimate the different coefficients and residues of the postulated model;

a statistical analysis, to judge the degree of influence of each factor on the evolution of the VF, by using measures of variances;

a graphic analysis, developed according to the Lenth method,¹⁴ which simultaneously shows size and significance of the different factors and their combinations.

Mathematical analysis

The different coefficients of the polynomial model postulated are grouped in Table 3. Firstly, we focused our study up to the second-order interactions, which are already complex and difficult to justify in terms of physical phenomena.

Table 3.

Values and significance of the different coefficients.

Term	Coefficient	Significance %
b ₀	71.6112	<0.01***
b ₁	–3.4076	<0.01***
b ₂	–1.7866	<0.01***
b ₃	0.0818	56.0
b ₄	–0.4812	0.0952***
b ₁₂	–1.9975	<0.01***
b ₁₃	–1.3944	<0.01***
b ₂₃	–0.5508	0.0216***
b ₁₄	–0.7374	<0.01***
b ₂₄	0.4964	0.0693***
b ₃₄	0.5873	<0.01***

***The factor or interaction of factors represents by the coefficient has a significant influence No. *Means that it has a negligible influence.

The constant b₀ refers to the average value of the different measured results. The preponderance of coefficients b₁ and b₁₂, which equate, respectively, to the concentration of the coagulation bath and its interaction with the extrusion diameter, is immediately noted. This observation supports the assumption made above. The following statistical analysis will help us to estimate the significance of the different terms.

Statistical analysis

With a coefficient of determination standing at 94.9%, the mathematical model used is deemed to be satisfactory to describe the influence of the different factors and their interactions on the evolution of the VF. Such a value validates our assumption to limit the study up to the second-order interactions.

Compared to the DMI concentration and the extrusion diameter, drawing rates applied in the coagulation bath and in the air only have a small influence on the evolution of the VF, estimated to be non-significant. Theoretically, it would have been assumed that drawing the forming monofilaments provides a higher exchange surface and then a better penetration of the non-solvent. However, their contributions have been estimated respectively above 0.02% and 0.84% against 41.92% for the DMI concentration.

Graphic analysis

To analyze size and significance of each factor and interaction, the graphic analysis has been developed according to the Lenth method.¹² Hence, a margin of error (ME) may be superimposed on the graphical representation of effects, calculated from the following steps:

‐the absolute values of the different coefficients (b_i) are sorted in ascending order;

‐the variable s₀ is obtained from the following equation:

s_{0} = 1.5 * median | b_{i} |

the coefficients having a value higher than $2.5 * s_{0}$ are eliminated and s₀ is again estimated from the remaining coefficients (m);

a Pseudo Standard Error (PSE) is defined from the following equation:

PSE = 1.5 * median | b_{i} |_{| b_{i} | < 2.5 s_{0}}

a two-sided 95% confidence interval is constructed using $b_{i} \pm ME$ , with:

$ME = t_{d}^{1 - α / 2} PSE$ , where t refers to the Student law, and $d = m / 3$ represents the degrees of freedom and $α = 5 %$ .

Finally, each factor or interaction having a value higher than $b_{i} \pm ME$ is considered to have a significant effect on the evolution of the VF of the obtained monofilaments, and conversely. The results are reported in Figure 9, with the parameters below:

$median | b_{i} |_{b_{i} < 2.5 s_{0}} = 0.662$ ;

$PSE = 0.994$ ;

$ME = \pm 2.758$ .

The ME allows only the factor associated to the coefficient b₁ to appear as significant.

Discussion

The reliability of the study carried out is shown on the normal plots diagram of residues (Figure 10). Indeed, with a coefficient of determination standing at about 95%, the mathematical model used is a good approximation to estimate the evolution of the VF for the obtained monofilaments, depending on the wet-spinning parameters.

According to the results that were reached in each analysis, solvent concentration seems to play a key role (above 42%) in the wet-spinning process of polyamide-imide solution. This parameter has an important influence on the morphological properties of the obtained monofilaments and, hence, on their physical properties. Moreover, its interaction with the extrusion diameter also affects the homogeneity of the final material, with an influence of about 14% (Figure 11).

At a low coagulation bath concentration (40%), the obtained monofilaments have a skin-core structure with high VF, B9–B12 (Figure 8). Indeed, as the spinning dope is extruded through the spinneret immersed in the coagulation bath, the counter-diffusion of solvent and non-solvent takes place and a layer of skin is firstly formed around the nascent monofilament. Compact and dense, it slows the diffusion of non-solvent and leads to a core region-like gel, with voids. By increasing the concentration up to 60% of DMI, the coagulation process is softened, ensuring a higher homogeneity of the final material, B1–B4 (Figure 8), even if after a certain length of time there is no more exchange.

Interaction between concentration of the coagulation bath and extrusion diameter is the second factor having a positive influence on the evolution of the VF. It is easy to understand that the lower the extrusion diameter, the lower the distance from the surface to the center of the monofilament, thus facilitating the diffusional exchange.

Conclusions

The Ishikawa diagram, illustrated in Figure 7, highlights a significant number of factors, controlled and controllable, having an influence on the wet-spinning process and hence on the final properties of the obtained monofilaments.

The study of the production process and the understanding of the coagulation degree require a robust and reliable experimental device. The wet-spinning machine developed in our laboratory meets these requirements by offering a fully automated operating mode due to a programmable processing technology, the PLC.

These works particularly focused on the evolution of the VF of coarse monofilaments manufactured from this prototyping tool, depending on four factors: the extrusion diameter, the concentration of the coagulation bath and the drawing rates applied in the coagulation bath and in the air. The results derived from the Design of Experiments mainly showed highly significant influences of both concentration of the coagulation bath and extrusion diameter factors. Indeed, increasing the solvent concentration in the coagulation bath allows a softer coagulation phenomenon by which VF decreases. Moreover, it seems to be consistent that the lower the extrusion diameter, that is, the lower the distance from the periphery to the core of the monofilament, the lower the VF.

We now have an overview of the influence of the different selected parameters as well as a first mathematical model to describe the evolution of the VF for the wet-spinning of polyamide-imide monofilaments. Thereafter, this model will have to be validated with randomized experiments and adjusted if necessary.

Thanks to this experimental device, it will be possible to test additional parameters from the Ishikawa diagram (Figure 7), such as the die geometry¹⁰ or the processing temperature, and improve our understanding of the wet-spinning process. Such knowledge can be useful both for describing the mechanism of the different processes occurring during the formation of wet-spun coarse monofilaments and for improving industrial applications.

Footnotes

Acknowledgments

We would like to thank the Kermel Society for providing the spinning dope for the experimental trials and allowing the realization of the rheological study. We are also grateful for the assistance provided to us by the Research and Development Laboratory.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Région Alsace.

References

Paul

. Diffusion during the coagulation step of wet-spinning. J Appl Polym Sci 1968; 12: 383–402.

Kalabin

Pakshver

Kukushkin

. Modeling the filament formation in the wet spinning of synthetic fiber from polymer solutions. Theor Found Chem Eng 1996; 30, 3: 295–301.

Kalabin

Pakshver

. Simulation of kinetics and heat and mass transfer for the spinning of chemical fibers from polymer solutions. Theor Found Chem Eng 1997; 31: 520–525.

Weisser P. Étude de monofilaments à hautes performances thermiques: développement d’outils de filage et caractérisations. PhD Thesis, Université de Haute-Alsace, France, 2013.

Barikani

Ataei

. Preparation and properties of polyimides and polyamide-imides from diisocyanates. J Polym Sci Pol Chem 1999; 37: 2245–2250.

Rhône-Poulenc Fibres. Fils à base de polyamide-imide etleurprocédé d’obtention. Patent EP 0 360 708 B1, France, 1989.

Preston

. Fiber science, Manchester: The Textile Institute, 1949, pp. 341–341.

Ferguson

Doulgeris

McKay

. Rheological and coagulation features in the wet-spinning process. J Non-Newton Fluid 1980; 6: 333–338.

Griskey

. Polymer process engineering, New-York: Chapman & Hall, 1995, pp. 478–478.

10.

Paul

. A study of spinnability in the wet-spinning of acrylic fibers. J Appl Polym Sci 1968; 12: 2273–2298.

11.

Heyi

Huashi

Juan

. The skin-core structure of poly(acrylonitrile-itaconic acid) precursor fibers in wet-spinning. J Appl Polym Sci 2008; 108: 947–952.

12.

Knudsen

. The influence of coagulation variables on the structure and physical properties of an acrylic fiber. Text Res J 1963; 33: 13–20.

13.

Ishikawa

. La gestion de la qualité: Outils et applications pratiques, Paris: Dunod, 2007, pp. 242–242.

14.

Lenth

. Quick and easy analysis of unreplicated factorials. Technometrics 1989; 31: 469–473.