Printer Friendly

Direct drawing of gel fibers enabled by twist-gel spinning process.


Gel-spun UHMWPE (ultrahigh molecular weight polyethylene) fiber exhibits remarkable tensile properties [1, 2]; however, the relatively complex processing involved in its production leads to costs almost ten times more than conventional meltspun fiber [3, 4], Commercial UHMWPE gel-spinning processes are inherently inefficient, requiring spin-solvent concentrations in excess of 80% [5], Efficient extraction of large volumes of spin solvent is a major limiting factor using conventional techniques since the rate of spin-solvent extraction is generally limited by diffusion related variables such as concentration gradient, fiber diameter, and temperature. The present paper introduces a new twist-gel-spinning process where a majority of nonvolatile spin solvent is removed from the gel fiber by applying mechanical twisting. The twisted gel fiber can be directly hot drawn, allowing conventional solvent extraction to proceed significantly faster after drawing.

In the authors' previous study, it was demonstrated that volatile spin solvent contained in the gel fiber can be effectively removed by mechanical twisting which liberates the spin-solvent evaporation from many diffusion related restrictions [6]. Decahydronaphthalene was used as spin solvent in the previous study to demonstrate the feasibility of the mechanical solvent removal process. Decahydronaphthalene is a common volatile spin solvent in commercial UHMWPE gel-spinning processes, and it is typically evaporated from the gel fiber at elevated temperature; however, it is flammable and toxic to humans [7, 8].

A more environmentally friendly process uses the nonvolatile spin solvent, paraffin oil [9]. Nonvolatile gel-spinning introduces new processing challenges since the spin solvent cannot be directly evaporated from the gel filament. Paraffin oil is a multimer containing multiple ethylene repeating units. It is relatively nonvolatile at typical processing temperatures in gel spinning, mandating the use of an extraction solvent for spin-solvent removal. The introduction of extraction solvent to the spinning process increases solvent byproducts by two or more orders of magnitude; however, with judicious selection of the extraction solvent, the nonvolatile UHMWPE gel-spinning process can remain nontoxic and nonflammable relative to the dry spinning process using decahydronaphthalene [9].

In commercial nonvolatile gel-spinning processes, the extraction of the spin solvent occurs after quenching but before extensive drawing of the filament [9], Drawing of the gel filament at relatively low temperatures was shown to prematurely induce preferential c axis orientation associated with the generation of defects, leading to reduced drawability, and poor tensile strength [10]. Drawing the gel filament at relatively high temperatures before extracting the spin solvent is difficult or even impossible because the gel fiber is relatively weak containing at least 80% spin solvent and will return to a solution state at a temperature well below the typical UHMWPE melting temperature due to the significant presence of spin solvent. Because of these limitations, extraction of the spin solvent preferentially occurs on the undrawn gel fiber. For typical spin-dope concentrations, the initial gel fiber must have a diameter on the order of 1 mm so that the final fiber after solvent extraction and ultradrawing does not enter the subdenier regime because processing would then become even more complex. This large diameter represents another major limiting factor for solvent extraction.

The present study introduces a nonvolatile twist-gel spinning process that is relatively nontoxic and nonflammable with minimal generation of solvent by-product while simultaneously increasing the process efficiency. The process is enabled by removing a majority of spin solvent from the gel filament by mechanical twisting which permits extensive drawing of the gel filament before the relatively slow diffusion based spin-solvent extraction is applied. Using mechanical twisting, it will be shown that more than 75% of the nonvolatile spin solvent, paraffin oil, can be efficiently removed from the relatively large, undrawn gel filament. Mechanically removing a majority of the spin solvent imparts the gel fiber with sufficient strength and thermal stability to be directly hot drawn to 20X in a first drawing stage. Consequently, the initial gel fiber diameter is reduced by a factor of about ~10 from the combined draw ratio and fiber volume reduction due to mechanical solvent removal. Processing of the reduced diameter gel fiber proceeds using conventional extraction, but two significant benefits arise from the twist-gel-spinning process. Since the pre-extraction fiber diameter is 10X smaller, the resulting diffusion based extraction on the twisted gel fiber should occur 100X faster compared to the large diameter filament associated with conventional gel spinning. Furthermore, since 75% of the spin solvent is removed mechanically, the conventional extraction during twist-gel-spinning consumes 75% less extraction solvent. It will be further shown that this twist-gel spinning process does not significantly affect the ultimate fiber tensile strength, Young's modulus, surface morphology, or geometry when compared with conventional gel-spinning processes. As a result of the significantly improved efficiency, the twist-gel spinning process should reduce the production cost of gel-spun high-strength fibers, promoting the broad expansion of high performance fibers into applications that were previously cost prohibitive.



Ticona GUR UHMWPE resin was kindly supplied by Celanese with a viscosity average molar mass of approximately 4 million Da. Hydrobrite 550 PO white paraffin oil with average molecular weight of 541 Da was kindly supplied by Sonneborn. The Hydrobrite 550 paraffin oil has viscosity of about 100 cSt at 40[degrees]C and a vapor pressure less than 0.01 mm HG at 20[degrees]C. Spin dope was prepared by combining 6 wt% UHMWPE powder with paraffin oil at 20[degrees]C to form a slurry. The slurry was poured into a preheated batch mixer (C.W. Brabender Prep-Center fitted with twin roller blades) and mixed for 30 min at 150[degrees]C to obtain a homogenized solution.

Gel Spinning

Fiber spinning was performed through an Alex James and Associates piston extruder with a 2.54 cm bore diameter and 150 mL capacity. The homogenized UHMWPE/paraffin oil solution was quickly transferred into the bore (preheated at 250[degrees]C) of the extruder and allowed to equilibrate for 1 h. The solution was extruded through a 1 mm die orifice with an aspect ratio of ~15:1 maintained at a temperature of 250[degrees]C. The solution was extruded at a speed of approximately 1 mm/min. The fiber solution was quenched into a 20[degrees]C glycerol bath 2 cm from the die exit. The quenched gel fiber was collected onto spools at 2 mm/min so as to apply two times jet-stretch to the fiber solution as it exited the die. The spools of gel fiber were rinsed with water and stored on under ambient conditions (40-60% relative humidity at 20-22[degrees]C) until needed for experiments.

Spin-Solvent Removal

Paraffin oil in the gel filament was removed by two methods: conventional extraction in a chemical solvent and mechanical solvent removal. Conventional extraction was performed using n-hexane. Control fibers were prepared by winding the gel fiber on to a 1 in. diameter polytetrafluoroethylene (PTFE) rod. Both ends of the gel fiber were fixed to maintain constant fiber length. To simulate a counter-flow type extraction process, the rod as prepared was submerged into an agitated bath of n-hexane at ~20[degrees]C for 0.5 h. The volume of hexane was ~5000 larger than the volume of the gel fiber in order to maintain a maximum concentration gradient throughout the extraction process. The PTFE rod containing the extracted gel fiber was removed from the n-hexane bath and dried under forced air convection while maintaining fixed fiber ends. Mechanical solvent removal was performed by fixing one end of the gel filament and attaching the free end to a motor with adjustable rotations per minute (rpm). A control sample was attached to the device for an equivalent amount of time without applying twist. After twisting, a tissue was gently passed along the bottom of the fiber to remove any residual solvent droplets. The control sample without twist was treated similarly. The amount of mechanical solvent removal applied to the gel fibers is defined in terms of turns per mm (TPmm) given by

TPmm = t x rpm/L, (1)

where L is the fixed length between fiber ends in mm, t is the twisting time in minutes, and rpm is the rotations per minute. A TPmm from 0 to 8 was applied to the gel fibers by varying the removal time, and the resultant mass loss from the gel fiber was measured. The mass loss is assumed to be entirely from removal of the spin solvent.

Hot Drawing

Hot drawing was performed through heated polyethylene glycol (PEG). The total path length through the hot bath was 0.6 m. The bath temperature was maintained within [+ or -] 1[degrees]C of the set point. The first stage of hot drawing was performed at 120[degrees]C with a feeding speed of 0.9 mm/min and a collection speed of 18 mm/min to obtain a draw ratio of 20 X. The second stage of hot drawing was performed at 130[degrees]C with a feeding speed of 0.5 mm/min and a collection speed increased incrementally to obtain the maximum draw ratio. Draw ratio as used in this report is defined as the ratio of the collection roller speed to the feed roller speed. Fibers were drawn at the maximum ratio to obtain samples at least 15 mm long for testing.


Diameter measurements were obtained by weighing a known length of fiber and calculating the cross-sectional area assuming a density of 1.0 g/[cm.sup.3]. Before weighing, the hot drawn fibers were briefly rinsed with water to remove residual PEG from the hot drawing stage and dried.

Tensile properties for single filaments were measured using an Instron 5566 universal testing machine. Fiber samples were wound on to wooden rods approximately 2 mm in diameter and super-glued over the wound fiber ends. The prepared single filament samples were clamped using Instron 2711 Series Lever Action Grips rated for 5 N. Crosshead speed was 100 mm/min with a gauge length of ~10 cm. All tensile tests were performed under ambient conditions (40-60% relative humidity at 20-22[degrees]C). Six samples from each of three fiber spinning experiments were tested and averaged. Experimental error was estimated using the standard error of the mean, defined as the standard deviation divided by the square root of the sample number.

Wide angle X-ray diffraction (WAXD) data were collected on a Rigaku Micro Max 002 (Cu K[alpha] radiation, [lambda] = 0.154 nm) operating at 45 kV and 0.65 mA using an R axis IV++ detector. Exposure time was 30 min for all samples. The crystalline orientation factor was computed using the method developed by Wilchinsky [11], The 110 and 200 equatorial diffractions were used to determine the orientation factor based on the orthorhombic UHMWPE unit cell [12].

Differential scanning calorimetry (DSC) data were collected on a TA Q200 DSC unit (TA Instruments). Samples were crimped in hermetic aluminum pans. Nitrogen atmosphere and a scan speed of 10[degrees]C/min were used for all samples. Thermogravimetric analysis (TGA) experiments were performed using a TA TGA5000 (TA Instruments). Samples were heated to 250[degrees]C in nitrogen atmosphere and held at this temperature until the sample weight approached a steady state. Scanning electron microscopy (SEM) images were collected on a LEO 1550. SEM samples were mounted onto carbon tape and gold sputtered.


The as-quenched gel fibers are not in an equilibrium state: a small fraction of spin solvent naturally phase separates from UHMWPE gel fiber until an equilibrium concentration is reached [13, 14], The phase separation that occurs in the gel fibers in the time period following quenching has been described in terms of the free energy of mixing. The large temperature change during quenching from solution (~150[degrees]C) to gel (~20[degrees]C) state reduces the entropie contribution to mixing which induces phase separation between the solvent and the polyethylene until a new equilibrium is reached [14]. The extent of phase separation has been shown to depend on the spin-dope concentration as well as the quenching conditions [13, 14]. Figure 1 shows the change in gel fiber linear density as a function of time. The linear density of the gel fibers approaches a constant value after approximately 24 h; therefore, it assumed that the gel fibers have reached an equilibrium state with respect to the natural phase separation. All solvent extraction and mechanical removal experiments were performed one day after fiber gel extrusion so that the initial gel fiber concentration remained constant during testing.

Conventional extraction of nonvolatile spin solvent requires an extraction solvent [9], With sufficient exposure time and volume of extraction solvent, the fiber weight change as a function of extraction time approaches a steady state (Fig. 2). For the present study, complete extraction (>99.9% solvent removal) of the spin solvent takes about a whole day in the extraction solvent based on the data shown in Fig. 2. Using the weight reduction of the fiber after 24 h in the extraction solvent, the initial concentration of the gel fiber is calculated to be 12% UHMWPE/88% paraffin oil. The initial gel fiber concentration is expected to be higher than the 6% spin-dope concentration because of the phase separation described previously.

Mechanical spin solvent removal by twisting does not require an extraction solvent [6]. Detailed descriptions of the mechanical solvent removal procedure were given in a previous report [6]. The extent of nonvolatile spin solvent mechanically removed is plotted in Fig. 3. A TPmm of 0 represents the control sample without twist, and the solvent removal in this case can be attributed to mass loss from surface contact by handling the fiber between measurements. The percentages of spin solvent removed in Fig. 3 were calculated according to

% Solvent Removal = [M.sub.before - [M.sub.after]/[M.sub.before x c] X 100. (2)

The values [M.sub.before] and [A.sub.after] are the gel fiber mass before and after twisting, respectively. The concentration of spin solvent in the gel fiber, c, is defined to be 88% in this study based on the initial gel fiber concentration. The majority of solvent (~73%) can be removed by twisting to 4 TPmm. Further increase in TPmm does not significantly increase solvent removal; however, excessive deformation of the gel fiber occurs at higher TPmm leading to instability of the process, and reduced tensile properties in the final hot drawn fiber. The effect of increasing TPmm on the tensile properties of gel-spun UHMWPE fibers was reported in a previous publication [6]. With decahydronaphthalene as a spin solvent, no significant change in the tensile properties of the hot drawn fibers was observed with increasing TPmm; however, a maximum TPmm was also observed above which the fiber becomes nonuniformly deformed and frequently breaks [6].

SEM images of the fibers after mechanical spin-solvent removal are shown in Fig. 4. The helix-like features are most visible on the 4 and 8 TPmm fibers, and the distance between repeating features in both fibers is consistent with the amount of twist applied during the mechanical spin-solvent removal process. The residual paraffin oil content in the gel fibers after the mechanical spin-solvent removal process was also measured by TGA. The neat paraffin oil almost completely evaporates with about 5% residue remaining, probably consisting of the relatively high molecular weight components in the paraffin oil (Fig. 5). The fiber conventionally extracted in hexane shows no significant weight loss, indicating that nearly all of the paraffin oil was extracted. Fiber mechanically treated to 4 TPmm shows a weight reduction of about 40%.

For hot drawing to proceed effectively, a sufficient amount of spin solvent should be removed from the gel fiber before hot drawing. Too much spin solvent remaining in the gel fiber reduces the melt temperature, thereby reducing the maximum drawing temperature that the fiber can sustain. Orienting the gel fiber at relatively low temperatures has been shown to significantly reduce the tensile strength of the drawn fiber, likely due to the creation of defects [10]. Figure 6 shows that the peak melting temperature of gel fibers can be increased solely by mechanical solvent removal without the use of an extraction solvent. The peak melting temperature of the gel fiber increases from about 123[degrees]C in the un-twisted fiber to about 130[degrees]C for the fiber with 4 TPmm. The increase in melting temperature likely results from the increased polyethylene concentration in gel after removing a portion of the spin solvent. The conventionally extracted gel fiber exhibits a peak melting temperature around 137[degrees]C, which is higher than the mechanical process, but a longer extraction time is needed and a relatively large volume of extraction solvent is consumed.

The mechanical solvent removal process imposes deformation to the gel fiber; however, this deformation does not impart significant orientation to the gel fiber. As previously noted, orientation at relatively low temperatures can impart significant defects in the fiber. No significant fiber axis crystal orientation can be detected from the 2-D patterns of the mechanically treated gel fibers in Fig. 7, regardless of the amount of mechanical solvent removal applied. The conventionally extracted fiber in Fig. 7 shows diffractions along the meridian indicating crystalline orientation perpendicular to the fiber axis. This perpendicular orientation is typical for polyethylene gel fibers spun and extracted with similar conditions [10].

Since the mechanical spin-solvent removal process sufficiently increases the gel fiber melting temperature and does not impart harmful orientation at low temperature, the mechanical solvent removal can be used exclusively (without the need for an extraction solvent) to gel-spun UHMWPE fiber. This new process represents an efficient, safe, and environmentally preferable method of spinning high-strength fibers. Since paraffin oil spin solvent is the only solvent required, the process is relatively nontoxic and nonflammable. To demonstrate the effectiveness of this new process, gel fibers were mechanically extracted to 4 TPmm corresponding to approximately 75% removal of paraffin oil spin solvent. Direct hot drawing to a total draw ratio of 60X yields fibers with a tensile strength of 2.25 [+ or -] 0.2 GPa and a Young's modulus of 114 [+ or -] 6 GPa. A representative stress-strain curve is shown in Fig. 8. WAXD shows intense equatorial diffractions in the 2-D patterns and azimuthal integrations of the 110 and 200 diffractions (Fig. 9) are consistent with crystalline orientation in the fiber direction. The corresponding orientation factor was calculated to be 0.864 suggesting good crystalline alignment along the fiber direction.

Combining mechanical solvent removal with conventional solvent extraction significantly increases the extraction rate while simultaneously reducing the amount of extraction solvent required to produce ultrahigh strength UHMWPE fiber. To demonstrate the mechanical solvent removal process, gel fibers twisted to a TPmm of 4 were continuously passed through an n-hexane bath (residence time of 30 s), followed by direct hot drawing to a maximum draw ratio of 80X. For comparison, a control sample prepared solely using conventional extraction in n-hexane for 30 min was hot drawn to a maximum draw ratio of 80X. Figure 8 shows that nearly identical tensile behavior results from the two separate processes. The control fiber and the twisted fiber reach tensile strengths of 3.93 [+ or -] 0.1 GPa and 3.73 [+ or -] 0.1 GPa, respectively. Table 1 summarizes the average tensile strength, Young's modulus, strain to break, and diameter of fibers from both processes; all these data are statistically similar, even though the twist-gel-spun fibers were produced using ~75% less extraction solvent. The WAXD 2-D patterns and azimuthal integrations of the (110) and (200) diffractions are shown in Fig. 9 and do not exhibit significant differences. Correspondingly, the orientation factors for the control and twist-gel-spun fibers were calculated to be 0.946 and 0.945, respectively.

The surface morphology and the geometry of the ultimate fibers can be seen from the SEM images in Fig. 10. Despite the relatively large amount of twist applied to the precursor fiber, both the control fiber and the twisted fiber after hot drawing exhibit similar surface appearance and shape on the length scale of the images. This result would be expected since the twist occurs before hot drawing, and the original four rotations per mm of fiber (TPmm 4) will be reduced by a factor 80 since the fiber is extended 80 times during drawing. Therefore, the twisted fiber will have an apparent TPmm of 0.05 or about 5 rotations per meter after hot drawing. This small amount of twist cannot be easily characterized by SEM or optical microscopy because the field of view of these instruments is normally less than 20 cm: not sufficient to capture a complete rotation.


In this study, a mechanical spin-solvent removal process for the production of efficient and environmentally preferable, high-strength, UHMWPE fiber was developed and investigated. More than 75% of the nonvolatile spin solvent, paraffin oil, was mechanically removed from gel fibers without using an extraction solvent. An extraction solvent free process was demonstrated by directly drawing the mechanically extracted fiber to a draw ratio of 60X achieving a tensile strength of 2.25 [+ or -] 0.2 GPa. A twist-gel-spinning process was demonstrated by combining the mechanical spin-solvent removal process with the conventional chemical based spin-solvent extraction process. The twist-gel-spinning process consumes 75% less extraction solvent than the conventional extraction process, and the resulting fiber achieved a tensile strength of 3.73 [+ or -] 0.1 GPa which was statistically similar to the control fiber processed using conventional extraction (3.93 [+ or -]0.1 GPa). All of these results indicated that the twist-gel-spinning process is compatible with the production of high-strength fibers through efficient UHMWPE fiber gel-spinning processes.


[1.] J.P. Penning, H. Vanderwerff, M. Roukema, and A.J. Pennings, Polym. Bull., 23, 347 (1990).

[2.] J.P. Penning, A.A. Devries, and A.J. Pennings, Polym. Bull., 31, 243 (1993).

[3.] J.A.P.M. Simmelink, U.S. Patent 0207907 (2011).

[4.] T.Y. Tam, J.A. Young, Q. Zhou, C.J. Twomey, and C.R. Arnett, U.S. Patent 0268967 A1 (2011).

[5.] V. Mittal, High Performance Polymers and Engineering Plastics, Wiley, New York (2011).

[6.] T. Wyatt, Y. Deng, and D. Yao, Polym. Eng. Sci. 55, 745 (2015).

[7.] S. Kavesh, U.S. Patent RE41268 El (2010).

[8.] N.H. Proctor, J.P. Hughes, and G.J. Hathaway, Proctor and Hughes' Chemical Hazards of the Workplace, Wiley, New York (2004).

[9.] S. Kavesh and D.C. Prevorsek, U.S. Patent 4551296 A (1985).

[10.] W. Hoogsteen, G. Tenbrinke, and A.J. Pennings, Polymer, 28, 923 (1987).

[11.] Z.W. Wilchinsky, J. Appl. Phys., 30, 792 (1959).

[12.] Z.W. Wilchinsky, J. Polym. Sci. Part 4: Polym. Phys., 6, 281 (1968).

[13.] M.M. Xiao, J.R. Yu, J.J. Zhu, L. Chen, J. Zhu, and Z.M. Hu, J. Mater. Sci., 46, 5690 (2011).

[14.] Z. Yufeng, X. Changfa, J. Guangxia, and A. Shulin, J. Appl. Polym. Sci., 74, 670 (1999).

Tom Wyatt, Teyana Gainey, Xudong Fang, Donggang Yao

School of Materials Science and Engineering, Georgia Institute of Technology, Atlanta, Georgia

Correspondence to: D. Yao; e-mail:

DOI 10.1002/pen.24082

Published online in Wiley Online Library (

TABLE 1. Tensile data of fibers hot drawn to the maximum draw ratio.

                          draw         Tensile            Young's
Processing condition      ratio    strength (GPa)      modulus (GPa)

Conventional extraction    80     3.93 [+ or -] 0.1   154 [+ or -] 4
Mechanically assisted      80     3.73 [+ or -] 0.1   146 [+ or -] 11
No extraction solvent      60     2.25 [+ or -] 0.2   114 [+ or -] 6

                              Strain at             Fiber
Processing condition          break (%)       diameter ([micro]m)

Conventional extraction   3.87 [+ or -] 0.2          17
Mechanically assisted     3.67 [+ or -] 0.2          17
No extraction solvent     2.95 [+ or -] 0.2          20

The solvent in the fibers were removed by three different methods:
conventional spin-solvent extraction, mechanically assisted spin-
solvent extraction, and purely mechanical spin-solvent removal where
no extraction solvent is used.
COPYRIGHT 2015 Society of Plastics Engineers, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2015 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Wyatt, Tom; Gainey, Teyana; Fang, Xudong; Yao, Donggang
Publication:Polymer Engineering and Science
Article Type:Report
Date:Jun 1, 2015
Previous Article:Effect of co-solvents on the photovoltaic performance of an inverted organic solar cell.
Next Article:Enhancement of dielectric and energy density properties in the PVDF-based copolymer/terpolymer blends.

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters