Journal of Polymer Research

, Volume 18, Issue 6, pp 1947–1953

Morphological development of polypropylene in immiscible blends with cellulose acetate butyrate

Authors

  • Chao-Hua Xue
    • Division of Textiles and ClothingUniversity of California
  • Dong Wang
    • Division of Textiles and ClothingUniversity of California
  • Bei Xiang
    • Division of Textiles and ClothingUniversity of California
  • Bor-Sen Chiou
    • Bioproduct Chemistry and EngineeringUSDA/WRRC/ARS
    • Division of Textiles and ClothingUniversity of California
Original Paper

DOI: 10.1007/s10965-011-9602-x

Cite this article as:
Xue, C., Wang, D., Xiang, B. et al. J Polym Res (2011) 18: 1947. doi:10.1007/s10965-011-9602-x

Abstract

Isotactic polypropylenes (iPP) with different melt flow indexes were melt blended with cellulose acetate butyrate (CAB) and then prepared into microspheres or nanofibers following a novel process of producing well dispersed CAB/iPP immiscible blends and subsequent removal of the CAB matrix. The morphologies of iPP microspheres were investigated by a scanning electron microscopy, and the dimensions of iPP microspheres were evaluated. The melt viscosities of iPP, CAB, and CAB/iPP blends were measured by using a capillary rheometer. The influences of the viscosity, viscosity ratio, and composition ratio of iPP/CAB on the morphology formation of iPP in CAB matrix were studied.

Keywords

Immiscible polymer blendsPolypropyleneViscosity ratioMicro-spherical beads

Introduction

Formation and morphology of immiscible polymer blends have been investigated extensively in the field of polymer science and industry due to the fact that novel properties could be resulted from the blends. Because of the occurrence of phase separation, the dispersed phases could exist in the forms of spheres, ellipsoids, fibrils and lamellas in the blends depending on the compositions, chemical, and physical properties of the polymers, as well as processing conditions [1]. It is also known that sizes and shapes of dispersed phases significantly affect the properties of the resulted polymer blends. Because of these features, many studies were focused on improving mechanical properties of polymer blends by controlling morphologies of the dispersed phase[25]. Recently, a novel process of preparing thermoplastic nanofibers by extrusion of well dispersed immiscible polymer blends through spinnerets has attracted great attention [68]. Polymer nanofibers are considered as important materials used in high efficient water and air filters, scaffolds for tissue engineering, protective clothing, reinforcement in composite materials and sensors [6, 911]. Moreover, the surface of polymer nanofibers can be functionalized, which leads to development of biocompatible, biorecognizable, biomimic, and other fibrous products for applications in high sensitivity and high selectivity sensors.

The novel nanofiber production process possesses attractive features of high productivity, versatility of thermoplastics, controllability, and environmental friendliness in manufacturing thermoplastic nanofibers. The whole process includes mixing of two polymers in a twin-screw extruder; extrusion of the blends into fibers in an islands (nanofibers)–in-sea(sacrificial polymer) form; and removal of the sacrificial polymer, cellulose acetate butyrate (CAB), from the extruded blends. The sample thermoplastic polymers including isotactic polypropylenes (iPP), poly(trimethylene terephthalate), poly[ethylene-co-(glycidyl methacrylate)], and low density polyethylene, have been successfully processed into nanofibers [1, 68, 11, 12]. Fabrications of other thermoplastic nanofibers by using this process are ongoing.

Several factors such as mixing ratio, viscosity ratio of two polymers, interfacial tension, and drawing ratio of the formed fibers, could affect morphology of the polymer blends and consequently the formation and uniformity of the resulted nanofibers. Earlier studies have revealed the critical impact of mixing ratio of two polymers and interfacial tension on dispersion of the thermoplastic polymer in CAB and sizes of the formed nanofibers [6, 7]. In order to have better understanding on the morphological development of thermoplastic polymers in the sacrificial CAB, iPPs with different melt flow indexes and viscosities were selected into the study of formation of microspheres and nanofibers via this process. The morphologies of CAB/iPP blends, iPP micropheres and iPP nanofibers were investigated by scanning electron microscopy, and the sizes of iPP micropheres and iPP nanofibers were evaluated. The melt viscosities of iPP, CAB, CAB/iPP blends were measured by using a capillary rheometer. The influences of viscosity and composition ratio on the morphology formation of iPP in CAB were studied, with respect to providing guidance for future polymer nanofiber fabrication.

Experimental

Materials

Cellulose acetate butyrate (CAB; butyryl content: 35–39%) was purchased from Acros Chemical (Pittsburg, PA). Four iPPs with different melt flow indexes and molecular weights were purchased from Aldrich Chemical (Milwaukee, WI). The characteristics and the abbreviations of the iPPs used are listed in Table 1.
Table 1

Viscosity ratios of iPP to CAB at a constant shear rate of 100 1/s, and size and size distribution of dispersed phase obtained from blends of iPP/CAB at a ratio of 80/20

iPP

Melt flow index (g/10 min, 190 °C, 2.16 kg)

Mw

Mn

Apparent shear viscosity at 100 1/s (Pa·s)

Apparent shear viscosity ratio of iPP/CAB at 100 1/s

PP-1

0.5

580,000

166,000

946.3

3.56

PP-2

4.0

340,000

97,000

456.7

1.72

PP-3

12.0

250,000

67,000

259.0

0.97

PP-4

35.0

190,000

50,000

127.5

0.48

CAB

   

265.9

/

Melt mixing

Mixtures of CAB and iPP in different weight ratios were blended in a mixer (ATR Plasti-Corder, C. W. Brabender, USA) for 10 min. Nitrogen gas was purged into the mixing chamber to reduce oxidation during extrusion. The screw speed was 100 rpm. The temperature was 240 °C. The final blends were collected and frozen in iced water immediately after mixing.

Melt extrusion

The mixtures of CAB/iPP were gravimetrically fed into a Leistritz corotating twin-screw (18 mm) extruder (Model MIC 18/GL 30D, Nurnberg, Germany). The feed rate was 12 g/min, and the screw speed was 100 rpm. Barrel temperature profiles were 150, 180, 200, 220, 235, and 240 °C. The blends were extruded through a two strand (3 mm in diameter) rod die. The extrudates were hot-drawn at the die exit by a take-up device and air cooled to room temperature.

Measurement and characterization

The melt flow behaviors of iPP, CAB, and CAB/iPP blends were investigated with a capillary rheometer LCR 8052 (Kayness, PA 19543), using a capillary round die with an L/D ratio of 30 and an entrance angle of 120 º. The barrel was preheated to the set temperature before loading. CAB/iPP blends were prepared by melt mixing, and chopped before measurement.

The blends prepared from melt mixing were immersed in acetone at room temperature for 30 min and centrifuged to remove CAB from the blends. To obtain clean iPP microspheres, this process was repeated at least 4 times. The extrudates prepared from melt extrusion were cut into small pieces and packed in a thimble, and then Soxhlet extracted in boiling acetone for 24 h to remove CAB from the blends. Thus bundles of iPP fibers were obtained.

Microspheres and fiber bundles of iPP were observed using a Philips XL30 scanning electron microscope (SEM). Number averaged diameters of iPP microspheres were calculated by equation (1).
$$ {D_n} = \frac{{\sum {{n_i}{D_i}} }}{{\sum {{n_i}} }} $$
(1)
where Dn is the number averaged diameter and ni is the number of microspheres with a diameter of Di.

Results and discussion

Effect of melt flow index on dispersion of iPP in blends

In an early study, difference in thermoplastic polymer viscosities was found not critical to size and size distribution of the formed nanofibers fibers by using this process [6]. However, due to the use of different thermoplastics, the viscosity difference could be shadowed by the difference of interfacial tension, another important factor in determining formation of nanofibers in extrusion of immiscible blends. The impact of viscosity on the morphology development and formation of nanofibers should exist and could be investigated by carefully selecting proper polymers. Thus, four isotatic polypropylene samples with different melt flow indexes were selected in this study. Due to the structural similarity, these polymers should have same interfacial tensions with CAB, and therefore, effect of the interfacial tension could be eliminated in the study.

SEM images of iPP spheres from the CAB/iPP blends at a component ratio of 80/20 are shown in Fig. 1a-d, obtained from samples with melt flow indexes varied from very low to high. It can be seen that, after removing CAB, iPPs form large elliptical spheres or small microspheres in all four samples. The surfaces of the formed spheres are very smooth, owing to the immiscibility between CAB and iPP. Different melt flow indexes of iPP resulted in obviously different sizes of the spheres in these samples. Most of the dispersed iPP forms spherical structures after vigorous mixing and relaxation, indicating the disappearance of the shear deformation of the dispersed phase even during the ice freezing process after mixing. Some iPP rods or elongated elliptical structure found in these spheres might be caused by the retained shear deformation caused by ice freezing.
https://static-content.springer.com/image/art%3A10.1007%2Fs10965-011-9602-x/MediaObjects/10965_2011_9602_Fig1_HTML.gif
Fig. 1

SEM images of iPP spheres from CAB/iPP blends. (a) CAB/PP-1 = 80/20, (b) CAB/PP-2 = 80/20, (c) CAB/PP-3 = 80/20, (d) CAB/PP-4 = 80/20, (e) CAB/PP-3 = 70/30, and (f) CAB/PP-3 = 90/10

The average diameters and distributions of iPP spheres from the blends corresponding to that of Fig. 1 are shown in Fig. 2. It can be seen that PP-1 spheres obtained have the broadest diameter distributions ranging from 1 μm to 35 μm, as well as the largest number average diameter of 13.4 μm. For polymer PP-3, the diameter distributions were narrowed to a range of 1–25 μm, mostly to the range of 1–15 μm, and the number average diameter was reduced to 4.8 μm. The diameter distributions and number average diameters of spheres for PP-2 and PP-4 were 1–35 μm, 1–30 μm, and 9.0 μm, 6.0 μm, respectively. The general trend was that the lower melt flow index of iPP, the larger size and distribution of particles resulted in the dispersed phase, except the sample PP-4.
https://static-content.springer.com/image/art%3A10.1007%2Fs10965-011-9602-x/MediaObjects/10965_2011_9602_Fig2_HTML.gif
Fig. 2

Diameter distributions of iPP spheres from CAB/iPP blends. (a) CAB/PP-1 = 80/20, (b) CAB/PP-2 = 80/20, (c) CAB/PP-3 = 80/20, (d) CAB/PP-4 = 80/20, (e) CAB/PP-3 = 70/30, and (f) CAB/PP-3 = 90/10

Relationship of viscosity ratio of polymer blends to sizes and size distribution of the dispersed phase

The melt flow behaviors of CAB and iPPs are shown in Fig. 3a, b. It can be found that when the apparent shear rate was increased, the apparent shear viscosity decreased at the shear rate above 100 1/s. This indicates that the CAB and the thermoplastic melts are non-Newtonian fluids under the mixing conditions and all samples followed shear thinning behavior under high shear rate conditions. For CAB/iPP blends under the same processing condition, the shear rate and the matrix viscosity are constant. The normal stress is dependent on the viscosity ratio of dispersed phases to matrix. The experimental viscosity ratios of dispersed thermoplastic polymers iPPs to matrix are significantly different over the shear rate range, as shown in Fig. 3b, in which viscosity ratio of PP-3 to CAB is close to the dot line of viscosity ratio 1. Since all the blends of CAB/iPP were prepared at a constant screw speed, thus we also measured the apparent viscosity iPPs and CAB under a constant shear rate of 100 1/s. The apparent viscosity ratios of iPPs to CAB were also calculated, shown in Table 1.
https://static-content.springer.com/image/art%3A10.1007%2Fs10965-011-9602-x/MediaObjects/10965_2011_9602_Fig3_HTML.gif
Fig. 3

(a) Apparent shear viscosity of CAB, iPPs versus apparent shear rate at 240 °C; (b) Apparent shear viscosity ratio of iPP to CAB versus apparent shear rate at 240 °C (Dot line in (b) shows the apparent viscosity ratio of unity); (c) Viscosity ratio versus average diameter of micro-spheres obtained from blends of iPP/CAB at a ratio of 80/20; (d) Apparent shear viscosity versus apparent shear rate; (e) Apparent shear viscosity at constant shear rate of 100 1/s at 240 °C of CAB, PP-3, and blends of CAB/PP-3 at different ratios

It can be seen that among the four iPPs, PP-1 with the lowest melt flow index (Table 1) showed the highest melt viscosity of 946.3 Pa·s, as well as the highest viscosity ratio of 3.56. Such a high viscosity might cause high elasticity and make the dispersing of high viscosity PP-1 in low viscosity matrix (CAB) more difficult [13], decreasing the trend of breaking up large drops of the disperse phase. Hence, large spheres were formed, which can be seen from the SEM image of Fig. 1a and the diameter distribution in Fig. 2a. Lowering the viscosity of the dispersed phase favors drop breakup and formation of fine spheres, which correlates with the sphere diameter changes of PP-1, PP-2, and PP-3 samples. However, sample PP-4 with the lowest viscosity ratio of 0.48 does not have the smallest average sphere diameter. It has been shown that when viscosity ratio between two immiscible components decreases in the range of 0.5 to 1, the flow in the dispersed droplet should become less sensitive to viscosity ratio [13]. However, when the viscosity of the dispersed phase is much lower than that of the matrix, such as in the case of PP-4 where the viscosity ratio was lower than 0.5, the flow of the dispersed phase might be hindered by the matrix due to their immiscibility. And propelled by the matrix, the dispersed droplets are easy to get coalesced in collision. The viscosity ratio of iPP-3 to CAB was calculated to be 0.97, as listed in Table 1, in addition, iPP-3 spheres have the smallest average diameter with the narrowest diameter distribution, as shown in Fig. 2. This corresponds well with previous reports [1416] that smallest particles were obtained when the viscosity ratio was around unity; as the viscosity ratio was increased far above or below unity, the dispersed particles become larger in size.

The dispersion of the thermoplastics in CAB should be related to viscosity ratios of the thermoplastic polymer to CAB, and interfacial tensions between the two polymers [6, 7, 14]. Since the four iPPs used in this system have the same chemical compositions, the only main difference is in the molecular weights. Therefore, the interfacial tensions between CAB and the various iPPs are practically the same. This is because in the high molecular weight ranges, the surface tensions are practically independent of molecular weight [14, 17]. Therefore, when the composition ratio of CAB/iPP is at 80/20, the formation of iPP spheres is mainly related to viscosity ratios of CAB to the thermoplastic polymer under a given shear field.

The normal stress difference (δPn) responsible for the dispersion across the drop/matrix interface can be given by [6]:
$$ \delta {P_n} = - 4(G{\eta_m}/{F_0})\sin (2\varphi ) $$
(1)
F0 is a function of the viscosity ratio, λ, G is the effective shear rate, ηm is the viscosity of matrix and ϕ is the orientation angle with respect to the direction perpendicular to the flow direction.
$$ {F_0} = \frac{{16\lambda + 16}}{{19\lambda + 16}} $$
(2)
and
$$ \lambda = \frac{{{\eta_d}}}{{{\eta_m}}} $$
(3)
ηd and ηm are the viscosities of the dispersed (target) and matrix (sacrificial) polymers, respectively. The interfacial capillary stress (δPt) which resists the deformation and tends to restore the initial equilibrium spherical shape is given by the Laplace equation:
$$ \delta {P_t} = \gamma (\frac{1}{{{C_1}}} + \frac{1}{{{C_2}}}) $$
(4)

Here C1 and C2 are the two principal radii of curvature; γ is interfacial tension of polymers. For dispersed phase, the drop is roughly spherical, C1 = C2 = d/2, \( \varphi = \frac{\pi }{4} \), d is diameter of sphere.

The normal stress difference is a function of shear rate and viscosity ratio of dispersed and matrix polymers (Eq. 1) [14]. When δPt = δPn, diameter of disperse phase could be calculated. However, according to the definition of F0 (Eq. 2) its value is in the range of 0.872 to 0.965, which does not change very much when viscosity ratio (λ) changes from 3.56 to 0.48, thus F0 would not clearly reflect relationship of the viscosity ratio (λ) to the diameter of formed spheres. But based on the experimental results, there is a relationship between (λ) and diameter of the spheres. Figure 3c shows a plot of viscosity ratio (λ) versus average diameter of micro-spheres formed by the dispersed polymer. Under the condition of constant interfacial tension, constant shear rate of 100 s−1, in the same matrix, the average diameter of the dispersed polymers formed in the matrix polymer is almost linearly related to viscosity ratio (λ) when λ is higher than one, but reversely related to λ when it is less than 1. However, when the viscosity ratio (λ) is less than unity, the dispersed polymer will become difficult to be dispersed by a more viscose and immiscible polymer, the sizes of the dispersed phase increased. These results are similar to the behavior of other immiscible blends [14].

Effect of composition ratio on the morphology of iPP in blends

To investigate the effect of blend ratio on the formation of iPP dispersed phase in CAB, PP-3 was blended with CAB at CAB/iPP ratios of 70/30, 80/20, and 90/10 in the mixer. SEM images of spheres extracted from the CAB/PP-3 blends are shown in Fig. 1c, e, f, corresponding to 80/20, 70/30, and 90/10, respectively. The average diameters and distributions of PP-3 spheres from the blends at different composition ratios were also calculated, as shown in Fig. 2c, e, f. From the SEM images of Fig. 1c, e, f, it can be seen that the blend ratio of CAB to iPP plays an important role in determining the morphology of blends and possibly in controlling the size of the formed iPP microspheres. Increasing the amount of CAB resulted in the diameter distributions and number average diameters of formed iPP microspheres becoming narrower and smaller, respectively, as shown in Fig. 2c, e, f. About 44% iPP microspheres prepared from CAB/PP-3 at a ratio of 90/10 have diameters of about 1 μm. About 92% of microspheres have diameters between 1 and 5 μm. However, for PP-3 microspheres obtained from CAB/PP-3 at a ratio of 80/20, only 21% of them have diameters of about 1 μm and 90% of them had diameters between 1 and 5 μm. The average diameter increases from 3.7 μm to 4.8 μm as the CAB/PP-3 ratio changes from 90/10 to 80/20. As the ratio of CAB/PP-3 changes to 70/30, the average diameter increases to about 7.4 μm. It should be noted that for CAB/PP-3 at the ratio of 70/30, the diameter was calculated from the dispersable microspheres, after removing the large solids which are not soluble or not dispersable in the acetone solution, as shown in Fig. 4a. For CAB/PP-3 ratios of 80/20 or 90/10, the PP-3 can form particle dispersion in acetone after removal of CAB, as shown in Fig. 4b, c. This agrees with the previous reports [7, 16, 18, 19] about influence of composition on the morphology formation in immiscible polymer blends, that increasing of composition ratio of thermoplastic polymer to CAB would increase the collision rate of particles, resulting in accelerated coalescence in flow as well as in quiescent states. Therefore, larger iPP particles were generated and even solids formed in the blend of CAB/iPP at higher composition ratios of the disperse phase. It seems that prevention of coalescence of the dispersed phase in the immiscible blends by adjustment of composition ratio is critical to control the size and size distribution of iPP in the blends.
https://static-content.springer.com/image/art%3A10.1007%2Fs10965-011-9602-x/MediaObjects/10965_2011_9602_Fig4_HTML.gif
Fig. 4

Pictures of iPP spheres dispersed in acetone after stirring. The samples were obtained from CAB/PP-3 at the ratio of (a) 70/30, (b) 80/20, and (c) 90/10

From Fig. 3d, e, it can be seen that the viscosity of CAP/PP-3 blends are much lower than that of bulk CAB or PP-3 in the main range of shear rates over 100 1/s, as shown in Fig. 3d, or at a constant shear rate of 100 1/s, as shown in Fig. 3e. The decrease in viscosity is very obvious when the composition changes from bulk to blends of 90/10 and 80/20. This is because increasing the amount of iPP in the blend could result in increasing interfacial contact areas as well as increasing interfacial slip between the two phase components in the blend due to the immiscibility of CAB and iPP. The contact area is influenced by not only the number of particles formed but also the particle size of the minor phase. When the composition ratio was changed from 70/30 to 50/50, the decrease in viscosity slows down. This is because within this ratio range, increasing the amount of iPP tends to accelerate formation of larger particles, due to the collision and/or coalescence rate increase. This rheological property explains the SEM micrographs involving the influence of composition ratio on iPP morphology of spheres formed in the CAB/PP-3 blends.

Fiber extrusion

Based on removal of the matrix from the blend of immiscible components, a method of fabrication of thermoplastic nanofibers was developed. Instead of using mixer to blend the immiscible polymers, followed by removing the matrix CAB to obtain iPP micropheres, this method uses an extruder to blend the immiscible polymers, followed by removing the matrix from the extrudates to obtain iPP nanofibers. As shown in Fig. 5a, after removing CAB, the remains were found to be composed of nano-microfibers or fiber bundles which can be easily separated. Higher magnification SEM image in Fig. 5b shows that the diameter of the fibers was about 90 nm.
https://static-content.springer.com/image/art%3A10.1007%2Fs10965-011-9602-x/MediaObjects/10965_2011_9602_Fig5_HTML.gif
Fig. 5

SEM images of remains after removing CAB from the blend of CAB/PP-3 with the composition ratio of 90/10 at a drawn ratio of 24. (b) The higher magnification of (a)

The focus of the investigation was on the factors influencing the morphology formation of iPP in CAB, which could provide guidance to the optimization of thermoplastic nanofiber fabrication as well as the understanding of nanofiber formation of thermoplastic polymers in immiscible blends. The morphology of the formed fibers in the extrudates is related not only to the melt mixing of the blend prior to entering the die, but also to the post treatment of the extrudate after exiting from the die. In fact, after the composite fibers were extruded, the shear deformation of iPP caused by cold drawing, viscosity ratio change of iPP and CAB caused by temperature decrease also have effects on the morphology evolution of iPP nanofibrils in CAB, and such impacts were discussed in another publication [20].

Conclusions

The viscosity ratio of polymer blends was identified as an important factor in determining the particle size of iPP in immiscible blends with CAB when the interfacial tensions between CAB and the various iPPs were same. The particle size of the dispersed iPP was smaller when the viscosity ratio of iPP/CAB was low, and smallest particles were obtained when the viscosity ratio was closer to unity with other factors being equal. Increasing the amount of CAB in the blend favors the formation of smaller diameter particles with narrower distribution. Varying the composition ratio between iPP and CAB could control the diameter of iPP spheres, and avoid large solid formation of the dispersed phase in the immiscible blend. This work provides guidance to the optimization of thermoplastic nanofiber fabrication as well as understanding of thermoplastic nanofiber formation.

Acknowledgements

This work was supported by a contract (HDTRA1-08-1-0005) from the Defense Threat Reduction Agency of The Defense Department.

Copyright information

© Springer Science+Business Media B.V. 2011