1. Introduction
Meltblown nonwoven fabrics are widely used as filtration media, medical fabrics, sound-absorbing and cushioning materials, and so on. Fine fibers are produced by pumping molten resin through a die with tiny holes whose size is of the order of 0.1 mm, thus low viscosity grades of resin with quite a high value of melt flow rate (MFR) are chosen as the raw materials [
1]. For polypropylene (PP), the value of MFR around a thousand is required for the meltblown process. For example, low-viscosity PP with MFR of 1550 g/10 min at 230
C was used for investigating unstable behavior of melt-blowing process [
2]. However, such special grades of resin are not dealt with so much on market and they cost quite high compared to the standard grades. It could be useful and economical that those special grades of resin are controllably produced from the commercial ones which are always available anywhere at low costs.
One of the methods, which is well known, is to degradate resin by using organic peroxides. In this method, a resin and an organic peroxide are compounded by a twin-screw extruder and the peroxide radicals chemically decompose the polymers during the mixing process. Numerical simulation of the peroxide-induced degradation was also developed so far. The molecular weight loss and change in viscosity of PP in a modular self-wiping corotating twin-screw extruder was studied by coupling a global one-dimensional flow model of the twin-screw extrusion process with kinetic equations of the evolution of molecular mass in the peroxide-induced degradation of PP [
3]. To evaluate the viscosity reduction, the relationship between the molecular weight and the rheological parameters, which were given in [
4], were used. However, it is generally hard to treat such highly reactive additives and to control the degradation. The molecular weight distribution of PP during the peroxide-induced degradation in a continuous stirred tank reactor was calculated and it was argued that imperfect mixing of the peroxide with PP could lead to a broadening of the molecular weight distribution [
5]. The broadening of the molecular weight distribution was explained by spatially inhomogeneous degradation which occurred around the peroxide particles because of the fast dissociation reaction and relatively slow diffusion of the peroxide [
6]. Recent developments of the modeling of the reactive extrusion were summarized in [
7]. In addition, due to residues that are caused by the decomposition with peroxides, the degradated resins by this method are not suitable for medical and hygiene products. Moreover, organic peroxides are specified as hazardous materials which must be carefully treated and stored.
Here, we propose another method without using peroxides, which we call a high-shear rate processing method. To date, the high-shear rate processing was mainly applied to the compounding of immiscible polymers and the dispersion of inorganic fillers in a polymer matrix. A high-shear processed poly(vinylidene fluoride) (PVDF)/polyamide11 (PA11) blend was studied and it was shown that PA11 could be dispersed in the PVDF phase with a domain size of several tens of nanometers by using a high-shear extruder [
8]. The rotational speed of the extruder was 1200 min
. Unmodified multiwalled carbon nanotubes (CNT) were compounded with poly(styrene-
b-butadiene-
co-butylene-
b-styrene) (SBBS) using a high-shear extruder and the dispersion of CNTs was greatly improved upon increasing the exerted shear rate and a homogeneous dispersion was successfully achieved with the screw rotation speed of 2000 min
[
9]. The structures and properties of polycarbonate (PC)/poly(methyl methacrylate) (PMMA) blends fabricated using their high-shear extruder were investigated and it was shown that nanostructured PC/PMMA blends with high transparency and improved mechanical properties were obtained at the screw rotation speed of 2250 min
[
10]. On the other hand, a reduction of mechanical properties of compounds with the high-shear rate processing was also reported. The effect of high-shear rate processing on the compatibilization of high impact PP/ethylene propylene rubber/high density polyethylene (PP/EPR)/PE blends was investigated by using a co-rotating twin-screw high shear extruder and a decrease of the size of PE core encapsulated by EPR shell was observed when increasing the screw rotation speed from 300 to 600 min
[
11]. However, as the screw speed was increased to 800 min
and 1200 min
, a noticeable reduction of the mechanical properties of (PP/EPR)/PE blends was found due to the thermal degradation. Polyamide 6(PA6)/PP blends were produced by a high-shear rotational processing machine and PA6/PP alloy fabricated under the screw rotation speed of 3000 min
with a mixing time of 30 s showed considerably lower breaking strain and yield strength than that of neat PA6, which was caused by the reduction of the molecular weight due to the high-shear rotation [
12].
In this paper, we positively use the ability of the degradation with the high-shear rate processing to fabricate the low-viscosity polymers without using peroxides. Our high-shear extruder is a type of single screw extruder with a maximum rotational speed of 3600 . Similar to a twin-screw extruder, the screw is composed of multiple screw elements so that we can realize different screw configurations. We apply high-shear stress to molten resin by using a single screw extruder with high screw rotation speed. The molten resin is thermally decomposed of its own shear-induced heat. Of course, we could decompose resin only by heat conduction from the barrel with high temperature, but it costs much time to heat or cool the barrels. Furthermore, there is usually an upper limit of barrel temperature which is not so high. Our method utilizes the shear-induced heat which is quickly generated in the bulk of resin and makes the resin at a higher temperature than the barrel temperature. In addition, as the major operating condition to adjust the degradation is only the screw rotational speed, different grades of resin are quickly and easily available. With the screw rotational speed of 3600 min and the barrel temperature over 300 C, we have produced low-viscosity PP with a value of MFR over a thousand from standard grades of PP which are commercially available. The meltblown nonwoven fabric sheet which is made of the degradated PP has been also produced. To our best knowledge, this is the first example to fabricate the low-viscosity PP for the meltblown process with the high-shear rate processing without using any peroxides.
We have also developed a numerical simulator which can predict the zero-shear viscosity of the degradated resin during the high-shear rate processing. The simulation of thermal degradation in a modular corotating twin-screw extruder was firstly developed in [
13]. They calculated the molecular weight loss of PP along the screw axis by coupling a two dimensional flow model with the kinetic equations of the thermal degradation. Since the maximum screw rotation speed of their study was 300 min
, the effect of the shear-induced heat was considered to be small. Thermal and mechanical degradation of polystylene in an ultra-high speed twin-screw extruder and the predictions of the molecular weight loss were studied [
14]. However, the viscosity reduction due to the molecular weight loss during the extrusion process and the back reactions to the physical quantities such as pressure and temperature were not considered. In our simulation, we used the simplest kinetic equation of the viscosity reduction and coupled it with a one dimensional non-Newtonian flow model in a single screw extruder. Our kinetic model of viscosity reduction could be easily derived by assuming that a resin was composed of linear polymers and thermal decomposition of polymers followed the random scission process. The model took the similar form to the kinetic equation of MFR described in [
15], however we determined the order of reaction of the kinetic equation for the zero-shear viscosity theoretically. Residence time distributions (RTD) of PP in a high-shear twin-screw extruder was measured by an on-line UV fluorescence device, and they found that deviations between the experimental results and the simulation results of RTD were observed at a high screw speed more than 800 min
and a small flow rate less than 4 kg/h [
16]. They also found that at a high screw speed of 1200 min
and a barrel temperature of 200
C, the calculated exit temperature overestimated the experimental result by around 30
C. They concluded that this difference occurred because the dependence of the heat transfer coefficient on the screw speed was not treated and the resin was not cooled by the barrel in the simulation [
16]. Therefore, we decided to evaluate the global Nusselt number by solving a one dimensional transient heat conduction equation in the case of the simple shear flow. The calculated Nusselt number was dependent on the operational conditions such as the screw rotation speed and the barrel temperature. We found that our simulation results agreed well with the experimental results for the degradation using the high-shear extruder. To our best knowledge, no one simulates and investigates the viscosity reduction in the high-shear extruder over the screw rotation speed of 2000 min
.
5. Conclusions
In general, to manufacture melt-blown nonwoven fabric sheets of PP, we need a low viscosity grade of PP with the value of MFR around 1000. We demonstrated that we could produce the degradated PP with the value of MFR over 1000 g/10 min from the commercial PP (MFR = 30 g/10 min) by the high-shear extruder without using the peroxides. To fabricate the PP with the value of MFR over 1000 g/10 min, the maximum screw rotation speed of 3600 min
and the barrel temperature over 300
C were used. From the TGA analysis shown in
Figure 7, the thermal decomposition of PP started at around 250
C in the air. The temperature of PP in the high-shear extruder exceeded the thermal decomposition temperature by 100
C. By varying the screw rotation speed from 2000 min
to 3600 min
with the fixed barrel temperature of 195
C, the outlet zero-shear viscosity decreased by around 100 Pa·s. This shows that the viscosity of polymers can be adjusted easily and quickly only by the screw rotation speed of the high-shear extruder.
Using the degradated PP, a meltblown nonwoven fabric sheet was really produced, which is shown in
Figure 15. The value of MFR of the degradated PP was 1148 g/10 min, and the yellow index was 5.68, which was a little bit higher than that before high-shear rate processing, 3.71. As can be seen in
Figure 15, the appearance of the sheet was white enough. The feeling of touch was also good. In fact, no shot (agglomerates of polymers that are larger than fibers) was confirmed in the image of the scanning electron microscope (SEM) which is shown in
Figure 16. The average fiber diameter was 1.6 µm.
It is interesting to apply the high-shear rate processing to fabricate low-viscosity engineering plastics for the meltblown process because those raw materials are less commonly available in market than low-viscosity PP. The meltblown nonwoven fabrics made of the engineering plastics are applied to the heat-resistant air filters, etc. However, it is challenging to degradate any engineering plastics by the high-shear extruder sufficiently because those materials are obviously hard to decompose thermally. In the high-shear extruder, the shear-induced heat becomes gradually small as the resin is thermally decomposed because the shear stress is proportional to the viscosity. Then, the heat conduction to the barrel becomes dominant compared to the heat generation. In fact, in the simulation of the experiment 1, the resin was cooled by the barrels as we saw in
Figure 13. This will suppress the thermal decomposition of resin with a high thermal decomposition temperature. This issue might be solved by replacing the barrels to adiabatic walls.
The viscosity of the degradated PP in our experiments was too low to extrude in a strand shape. Thus, small lumps of the extruded PP were crushed and the particles, whose size was of the order of 1 mm, were obtained to be fed into the meltblown equipment. To pelletize such low-viscosity PP, we have to use an under water cutter, which is hard to manage. A low-viscosity grade of resin for the meltblown process is sometimes only available in powder forms, and feeding the powders to the meltblown equipment is awkward because powders generally tend to be cohesive and also tend to entrap air. However, if the high-shear extruder is combined tandemly with the meltblown equipment, we can manufacture the meltblown nonwoven fabric sheets from the standard grades of resin, which are supplied in pellet form and always available anywhere at low cost. Furthermore, various meltblown nonwoven fabric sheets with different physical properties will be easily produced by changing the screw rotational speed of the high-shear extruder. We are now developing the integrated manufacturing process of meltblown nonwoven fabrics.
We also developed the simulation of the shear-induced thermal decomposition of resin in the high-shear extruder. By the nature of the high screw rotation speed and the wide range of the operational conditions, the simulation of the high-shear rate processing was not so easy. In fact, as described in [
16], the heat transfer coefficient should be modulated by the screw rotation speed to predict the resin temperature at high-shear rates. Of course, we need to take into account the effect of the viscosity reduction by the thermal decomposition. In our one dimensional flow model, the global Nusselt number in the screw was estimated by solving the heat conduction equation Equation (
8) in the case of the simple shear flow. We found that the Nusselt number did not change so much as the barrel temperature was increased from 195
C to 300
C, but the large deviations occurred by changing the screw rotation speed from 100 min
to more than 1000 min
. By coupling the flow model to the kinetic equation of the viscosity reduction, which was derived in
Section 2.2, we found that our simulation results agreed well with the experimental results even when the operational conditions were widely changed. The simulation predicted that the large viscosity reduction occurred in front of the dams such as the through-holes and the die because the shear stress and the residence time were expected to be large in the fully filled regions of screw. Therefore, we installed many dams in the screw to obtain the low-viscosity PP with MFR over 1000 g/10 min in the experiment 2.
The numerical simulation to predict the number-average molecular weight and weight-average molecular weight under the thermal degradation in a twin-screw extruder were developed [
13]. Using their kinetic equations of the molecular weights, we will be able to predict the polydispersity index during the high-shear rate processing. In the meltblown process, the polydispersity index of the raw material will affect the variance of the fiber diameters. Since the high-shear rate processing utilizes the thermal decomposition of polymers, which is the random scission process, the polydispersity index tends to be small. However, the quantitative prediction during the high-shear rate processing is missing. It will be interesting to see how to change the polydispersity index with the screw configurations and the operational conditions of the high-shear extruder.
In our simulation, the resin temperatures at the exit of the high-shear extruder were a little lower than the experimental results. The possible reason is that the dependence of the Nusselt number on the distance in the system was not considered in our model. However, obtaining the full local Nusselt number is difficult. This is because the heat conduction equation is usually solved forwards with a homogeneous initial condition, but we have to proceed the calculations backwards to specify the unfilled regions. If the barrels of the high-shear extruder were adiabatic, consideration of the heat conduction would not be appeared.