Diffusion dynamics of ionic liquids during the coagulation of solution spinning for acrylic fibers.
Acrylic fiber is a very versatile fiber, which has emerged not only as a major substitute for wool but also as the important precursor for carbon fiber. So far, the spinning process most commonly used for acrylic fiber is solution (wet or dry) spinning as the melt spinning method is not suitable for the reason that acrylic polymer decomposes before melting although plasticization melt spinning has been invented. Acrylic polymers are only dissolved in a few polar solvents, which are often non-green solvents such as dimethyl sulphoxide (DMSO), dimethylacetamide (DMAc), dimethyl formaide, and sodium thiocyanate (NaSCN)/[H.sub.2]O .
Ionic liquids (ILs) have been known as prospective environmentally friendly media for green physical and chemical processes . ILs are composed of organic cations and inorganic anions while there are no molecules in ILs. The special chemical structure makes ILs to have different properties than conventional solvents, such as nonvolatility, nonflaming, stability, and recycling capability. Therefore, ILs have been widely applied in electrochemical processes, separations, extractions, chemical and biochemical reactions and functional materials [3, 4].
In 2002, Swatloski et al.  found that ILs could be used as nonderivatizing solvents for cellulose and cellulose could be precipitated from the ILs solution by addition of water or other precipitating solvents. Another group  also studied the dissolution of cellulose in 1-allyl-3-methylimidazolium chloride ([AMIM]Cl) and the results were similar to that of Swatloski's. Tu et al. [7, 8] have found that ILs are also good solvents for some aromatic polyamides and polyarylsulfone.
In our work, we have found that ILs are good solvents for acrylic polymers [9, 10]. The purpose of this work is to investigate the mass transfer process occurring during the formation of acrylic fiber used [BMIM]Cl as solvent and [H.sub.2]O as nonsolvent.
The formation of fibers by solution spinning proceeds is very complex, which involves a combination of rheology phenomena, mass transfer and super molecular structuralization that are difficult to be understood well together. Some attempts have been made to study these processes under ideal situations to learn more about fiber formation. As early as 1968. Paul [11, 12] discussed the spinnability and diffusion dynamics in the wet-spinning of acrylonitrile-vinyl acetate copolymer with DMAc as solvent and water as the nonsolvent. To simplify the experiments. Paul performed his experiments on gelled rods of concentrated solutions of acrylic copolymer in DMAc. Oh et al.  studied the coagulation process of polyacrylonitrile wet-spinning, which was simulated by the way of line-procedure based on the numerical analysis of polyacrylonitrile/nitric acid solution coagulated from water. Law et al.  investigated the compositional changes and the filament diameter reduction during the wet-spinning of acrylic fibers from aqueous sodium thiocyanate (NaSCN/[H.sub.2]O) solution. The diffusion coefficient of [H.sub.2]O in the protofibers prepared by acrylic homopolymers was determined by Chen et al. , using [H.sub.2]O/DMSO mixture as the coagulation bath during wet-spinning process.
As ILs are composed of organic cations and inorganic anions without any molecule, the relative large molecular dimension of [BMIM]Cl and the high viscosity of the polymer/ILs solutions as well as the specific interactions between the macromolecule and ILs make the coagulation process and diffusion dynamics of PAN/ILs/[H.sub.2]O different from that of conventional solvents.
The coagulation dynamics of acrylic copolymer (PAN) was investigated with l-butyl-3-methylimidazolium chloride ([BMIM]Cl) as solvent for PAN and [H.sub.2]O as nonsolvent. On the basis of the content of [BMIM]Cl in the coagulated filament at different coagulation time, a mass-transfer model of [BMIM]Cl from concentrated PAN/[BMIM]Cl solution was established. And then, the diffusion rate of [BMIM]Cl was discussed according to the factors that affect the coagulation mechanism, including the concentration of PAN solution, coagulation bath temperature and the concentration of coagulation bath.
[BMIM]Cl was synthesized according to the procedures described in the literature  in our laboratory. Dimethylacetamide (DMAc) was purchased from Shanghai Chemical Agent Company. PAN composed of acrylonitrile (91 wt%), methyl acrylate (7.5 wt%), and itaconic acid (1.5 wt%) was supplied by SINOPEC Shanghai Petrochemical ([bar.M.sub.[eta]] = 6.0 x [10.sup.4]).
Dissolution of PAN in [BMIM]Cl
PAN was dried at 60[degrees]C under vacuum for 10 h, and then mixed with [BMIM]Cl under 80[degrees]C at different concentration.
Model Spinning Experimental
The spinning experiment was carried out on the model dry-wet spinning apparatus in our laboratory. The concentrated solution was extruded from a one-hole spinneret by compressed nitrogen ([N.sub.2]) with the pressure of 3.5 kg at the temperature of 90[degrees]C. The emerging filaments passed through the air first and then the coagulation bath. The distance from the spinneret to the surface of the coagulation bath is 1 cm. The coagulation bath consisted of various mixtures of [BMIM]Cl and water. The coagulated filaments were taken up at the low velocity of 10 cm/min. The concentration of solution the temperature and concentration of the coagulation were adjusted to investigate the diffusion rate. The larger diameter of the filament lengthened the diffusion time scale to a range convenient for measurement.
Coagulation of Concentrated PAN/[BMIM]Cl Solutions
The extruded freshly formed fluid filament was immersed into the coagulation bath and taken out after different coagulation time. The coagulated filaments consisted PAN/[BMIM]Cl/[H.sub.2]O were weighted as Ml. And then the filaments dried at 80[degrees]C under vacuum to remove water were weighed as M2. Finally, the filaments washed with hot water to remove solvent and dried were weighed as M3. The composite of the coagulated filament were calculated as following.
[H.sub.2]O content in filament (wt%) = (M1 -M2)/M1 x 100% (1)
[BMIM]Cl content in filament (wt%) = (M2 - M3)/M1 x 100% (2)
Morphology of the Concentrated Filament
The coagulated filaments were dried at -83[degrees]C for 24 h to remove the water. The cross sections of the dried samples were examined by scanning electron microscopy (SEM, JSM-5600LY).
RESULTS AND DISCUSSIONS
Coagulation Process of PAN/[BMIM]Cl Solutions
During the coagulation process of acrylic polymer solution, binary diffusion occurs including the solvent diffusing out of the concentrated fluid while nonsolvent diffusing into it. Diffusive interchange between the solvent and nonsolvent actually crosses the bath side interface, solidified polymer skin and inner ternary fluid core respectively. During the initial period, solidified skin is very thin and the filament is a ternary gelled concentrated solution. The difference of the concentration between solution and coagulation bath provides a driving force to promote the diffusion in this period. Because the solvent concentration difference between solution and coagulation bath varies a lot and gelled filament contains loose structure with macroscopic voids, solvent and nonsolvent can interchange easily, which causes solvent/nonsolvent diffuse at a high rate. However, as the diffusion continues, the skin of the filament becomes thicker and denser, which causes the difficulty of solvent and nonsolvent diffusion. Meanwhile the diffusion rate becomes lower gradually down to the equilibrium condition with the concentration difference between solution and coagulation bath decreasing. The structure variation of the filament during coagulation is shown in Fig. 1. The thickness of the skin increased with the coagulation time.
The binary diffusion can be considered as two independent counter unilateral diffusion processes. In this article, the diffusion of [BMIM]Cl and [H.sub.2]O is considered respectively, shown in Fig. 2. At the beginning, both [BMIM]Cl and water diffuse rapidly and then diffuse slowly with the increase in diffusion time until the diffusion equilibrium is reached. To simplify the simulation, only the diffusion dynamics of [BMIM]Cl is focused on in the following.
[FIGURE 1 OMITTED]
Establishment of Diffusion Dynamics Model
The actual spinning procedure would be very difficult to study with the idea of obtaining fundamental information. Only model experiments consisted of coagulating large specimens of the spinning solution in static experiments were performed. And the diffusion model was established to give information regarding the diffusion rate associated with coagulation. Such knowledge can be useful both for practical applications and for elucidating the mechanisms of coagulation.
[FIGURE 2 OMITTED]
The hypothesis is pointed out as follows to simplify the model:
1. The diffusion of [BMIM]Cl and [H.sub.2]O can be considered as two independent uni-diffusion processes, then the ternary system can be regarded as two independent binary systems.
2. The diffusion along the radial direction is considered, while that along the axial direction is ignored.
3. The cross section of the filament is uniform and the diameter of the filament does not change during the process of diffusion.
4. The boundaries are fixed and thermal effects during the diffusion process are ignored.
It is assumed that the diffusion of [BMIM]Cl is conformity with Fick's second law of diffusion according to the assumption and elucidation. In cylindrical coordinates, assuming axial symmetry, the rate of concentration change of [BMIM]Cl at any radial point can be expressed by equation
[FIGURE 3 OMITTED]
[[partial derivative]C(r, t, T)]/[partial derivative]t = [1/r][[partial derivative]/[partial derivative]r](rD[[partial derivative]C(r, t, T)/[partial derivative]r]) (3)
where C is the concentration, r is the radial distance, t is the time of diffusion, T is the temperature of the system, and D is the diffusion coefficient of [BMIM]Cl. Here, it is assumed that D is constant at given temperature, which does not change during the diffusion process, so Eq. 3 can be simplified as:
[partial derivative]C/[partial derivative]t = [1/r][[[partial derivative]/[partial derivative]r](rD[[partial derivative]C(r, t]/[partial derivative]r])] = D([[[[partial derivative].sup.2]C]/[[partial derivative][r.sup.2]]] + [1/r][[partial derivative]C/[partial derivative]r]). (4)
The concentrated PAN/[BMIM]Cl solution is uniform, and [BMIM]Cl in solution is initially at a concentration [C.sub.0], that is,
C = [C.sub.0], t = 0, 0 < r < R (5)
where R is the filament radial.
Under experiment condition, the volume of coagulation bath is much larger than that of concentrated PAN/[BMIM]Cl solution so it is assumed that the volume of coagulation bath is infinite. If the concentration of the coagulation bath does not vary with diffusion time as well as there is no difference of concentration between solution and coagulation bath at the interface, the boundary condition is
C = [C.sub.[infinity]], r = R, t [greater than or equal to] 0 (6)
where [C.sub.[infinity]] is the concentration of coagulation bath.
On the basis of hypothesis of the initial and boundary condition, the partial differential equation (4) can be solved by separating the variables. Finally, the solution satisfying Eqs. 5 and 6 is
C(r, t) = [C.sub.[infinity]] + 2([C.sub.0] - [C.sub.[infinity]]) [[infinity].summation over (n=1)] [e.sup.-D[a.sub.n.sup.2]t][[[J.sub.0]([a.sub.n]r)]/[R[a.sub.n][J.sub.1]([a.sub.n]R)]] (7)
where [J.sub.0]([a.sub.n]r) and [J.sub.1]([a.sub.n]r) are the Bessel functions of the first kind of order 0 and order 1, respectively; the as are the roots of [J.sub.0] ([a.sub.n]r) = 0 (n = 1, 2, 3, ...).
Since the diameter of the filament is quite small and it is difficult to obtain the concentration distribute of [BMIM]Cl in fiber during the diffusion process, the concentration of the [BMIM]Cl at diffusion time t can be calculated according to the average concentration of [BMIM]Cl in the filament:
[bar.C](t) = [[[integral].sub.0.sup.R]2[pi]rC(r, t)dr]/[[[integral].sub.0.sup.R]2[pi]rdr] = [[pi][R.sup.2][C.sub.[infinity]] + 4[pi][R.sup.2]([C.sub.0] - [C.sub.[infinity]])[[infinity].summation over [n=1]][[-[e.sup.-D[a.sub.n.sup.2]t]]/[(R[a.sub.n])[.sup.2]]]]/[[pi][R.sup.2]] (8)
[bar.C](t) = [C.sub.[infinity]] + 4([C.sub.0] - [C.sub.[infinity]])[[infinity].summation over (n=1)][[e.sup.-D[a.sub.n.sup.2]t]/[[R.sup.2][a.sub.n.sup.2]]]. (9)
The parameter D in the model can be adjusted with different system. Parameter values are determined using standard nonlinear least-squares procedures based on the experimental results.
Diffusion Dynamics of [BMIM]Cl
At static conditions, a few factors that affect the diffusion dynamics are taken into consideration.
Polymer Concentration. From Fig. 3, it is clear that the polymer concentration has a little effect on the diffusion rate of [BMIM]Cl at the beginning but no distinct effect on the [BMIM]Cl content in filament when the system reach diffusive equilibrium. It indicates that the hypothesis of the boundary condition is reasonable. Thus, the model has a good fit with the experimental data (shown in Fig. 3). The concentration of [BMIM]Cl decreases rapidly at the beginning and then slowly with coagulation time increasing till it reaches the diffusion equilibrium. The effect of polymer concentration on the diffusion rate from the diffusion coefficient of [BMIM]Cl can be seen in Table 1. The diffusion coefficient decreases continuously with an increase of polymer concentration as a whole. Despite the larger drive force from the concentration difference with increasing PAN concentration, the larger viscosity of dope increases the boundary and diffusion resistance for [BMIM]Cl.
[FIGURE 4 OMITTED]
Coagulation Bath Temperature. The coagulation bath temperature has a very complex effect on the coagulation process. On one hand, an increase of temperature makes the molecules move quicker, which would increase the diffusion rate. On the other hand, the morphology structure of the filament would change with the variation of temperature. As shown in Fig. 4, the skin is formed slowly in the cross section of the coagulated filaments when the coagulation bath temperature is lower than 30[degrees]C. It is an excited result for us because it indicates that the uniform fiber can be prepared by this technology even if the coagulation bath is water due to the slow diffusion of IL from the concentrated polymer fluid. When the coagulation temperature is up to 50[degrees]C. the thickness of the skin increases rapidly at the same coagulation time and the voids are clear. As the coagulation temperature is up to 70[degrees]C, skin-core structure is very clear and the circular cross section is deformed.
[FIGURE 5 OMITTED]
From the diffusion dynamics results, the diffusion model has a good agreement with the experimental data when the coagulation bath temperature varies as shown in Fig. 5. It is shown from Table 2 that the diffusion coefficient of [BMIM]Cl increases with an increase of temperature especially when the temperature is up to 70[degrees]C. As the temperature increases, the molecular movement increases that accelerates the diffusion of [BMIM]Cl. Meanwhile, PAN/[BMIM]Cl solution is of great flow activity energy. Therefore, the viscosity decreases rapidly as the temperature increases and the resistance of diffusion decreases remarkably.
Coagulation Bath Concentration. The coagulation bath concentration is the key factor that affects the final content of [BMIM]Cl in filament. The established model can illustrate the diffusion trend well that the content of [BMIM]Cl increases and finally tends to the initial concentration of coagulation bath, which is coordinated with the boundary conditions in the model (Fig. 6). It is shown that the diffusion coefficient of [BMIM]Cl increases at first and then decreases as the bath concentration increase (Table 3). There are two factors that devote to the maximum value of the diffusion coefficient. Firstly, as the concentration gradient is the main drive force to help [BMIM]Cl diffuse from solution to coagulation bath, the increase of bath concentration will weaken this drive force and make the diffusion rate lower. Secondly, when the concentration of coagulation bath is low, the skin is easy to form which results in the decrease of diffusion rate.
[FIGURE 6 OMITTED]
ILs are excellent solvents for acrylic fiber because acrylic polymer contains cyano-group(-CN) as electron acceptor, and [BMIM]Cl contains [.sup.-.C]l as great electron donor, the interaction between -CN and [.sup.-.C]l will weaken the interaction of -CN between the acrylic polymer in the dissolution process. From the coagulation process, it is shown that the diffusion rate of [BMIM]Cl is slow, which is helpful to keep the fiber structure uniform even if the coagulation condition is strong. On the basis of the experimental data, a diffusion dynamics model was established to simulate the diffusion rate of [BMIM]Cl from concentrated PAN filament during coagulation, which had a good fit with the experimental data. The diffusion coefficient is affected by the polymer concentration, concentration and temperature of coagulation bath. The diffusion rate of [BMIM]Cl is more sensitive to the coagulation concentration and temperature other than the dope concentration. The diffusion coefficient D decreases a little with the dope concentration due to the increase of viscosity. The diffusion coefficient D decreases with the coagulation bath temperature. As the coagulation bath concentration increasing, the diffusion coefficient D exhibits a maximum value. The results are helpful to understand the coagulation process for polymer/ILs solutions although some other important factors such as spinning velocity and flow rate of the coagulation bath must be taken into consideration. The further work is in the process.
1. X. Wang, R. Ni, Q. Liu, and Z. Wu, Synth. Fiber, 29, 23 (2004).
2. R.D. Rogers and K.R. Seddon, Science, 302, 792 (2003).
3. C. Chiappe and D. Pieraccini, J. Phys. Org. Chem., 18, 275 (2005).
4. P. Kubisa, J. Polym. Sci. Part A: Polym. Chem., 43, 4675 (2005).
5. R.P. Swatloski, S.K. Spear, J.D. Holbrey, and R.D. Rogers, J. Am. Chem. Soc., 124, 4974 (2002).
6. H. Zhang, J. Wu, J. Zhang, and J. He, Macromolecules, 38, 8272 (2005).
7. T. Zhao, H. Wang, B. Wang, X. Tu, Y. Zhang, and J. Jiang, Polym. Bull., 57, 369 (2006).
8. X.P. Tu, Y.M. Zhang, T.T. Zhao, and H.P. Wang. J. Macromol. Sci. Part B: Phys., 45, 665 (2006).
9. D. Li, Y. Zhang, H. Wang, J. Tang, and B. Wang, J. Appl. Polym. Sci., 102, 4254 (2006).
10. W. Liu, L. Cheng, H. Zhang, Y. Zhang, H. Wang, and M. Yu, Int. J. Mol. Sci., 8, 180 (2007).
11. D.R. Paul, J. Appl. Polym. Sci., 12, 2273 (1968).
12. D.P. Paul, J. Appl. Polym. Sci., 12, 383 (1968).
13. S.C. Oh, Y.S. Wang, and Y. Kooyeo, Ind. Eng. Chem. Res., 35, 4796 (1996).
14. S.J. Law and S.K. Mukhopadhyay, J. Appl. Polym. Sci., 69, 1459 (1998).
15. H. Chen, Y. Liang, and C. Wang, J. Polym. Res., 12, 49 (2005).
16. S. Csihony, C. Fischmeister, C. Bruneau, I.T. Horvath, and P.H. Dixneuf, New J. Chem., 26, 1667 (2002).
Yumei Zhang, Xiaoping Tu, Weiwei Liu, Huaping Wang
State Key Laboratory for Modification of Chemical Fibers and Polymer Materials, Donghua University, Shanghai 200051, People's Republic of China
Correspondence to: Y. Zhang: e-mail: firstname.lastname@example.org
Contract grant sponsor: Shanghai Municipal Education Commission (DAWN Project).
TABLE 1. Effect of PAN concentration on diffusion coefficient of [BMIM]Cl. PAN concentration in PAN/[BMIM]Cl concentrated solution (wt%) 14 16 18 20 D ([cm.sup.2]/ 3.90 x 2.93 x 1.96 x 1.42 x Sec) [10.sup.-8] [10.sup.-8] [10.sup.-8] [10.sup.-8] TABLE 2. Effect of coagulation bath temperature on diffusion coefficient of [BMIM]Cl. Coagulation bath temperature ([degrees]C) 30 40 50 D ([cm.sup.2]/sec) 2.93 x 3.57 x 8.37 x [10.sup.-8] [10.sup.-8] [10.sup.-8] 60 70 D ([cm.sup.2]/sec) 1.24 x [10.sup.-7] 9.61 x [10.sup.-7] TABLE 3. Effect of concentration of [BMIM]Cl in coagulation bath on diffusion coefficient of [BMIM]Cl. Concentration of [BMIM]Cl in coagulation bath (wt%) [H.sub.2]O 5 10 D ([cm.sup.2]/sec) 2.93 x 5.20 X 3.25 x [10.sup.-8] [10.sup.-8] [10.sup.-8] 15 20 D ([cm.sup.2]/sec) 2.15 x 1.76 x [10.sup.-8] [10.sup.-8]
|Printer friendly Cite/link Email Feedback|
|Author:||Zhang, Yumei; Tu, Xiaoping; Liu, Weiwei; Wang, Huaping|
|Publication:||Polymer Engineering and Science|
|Article Type:||Technical report|
|Date:||Jan 1, 2008|
|Previous Article:||Effects of crystallization on dispersion of carbon nanofibers and electrical properties of polymer nanocomposites.|
|Next Article:||Uniaxial ratchetting behavior of vulcanized natural rubber.|