(43) Vol.38, No. 2 (1982) T-61

(Received July 1, 1981)

MELT SPINNING OF POLY(VINYL ALCOHOL) PLASTICIZED WITH GLYCERIN

By Piao Dong-shwi* and Toshio Kitao

(Kyoto University of Industrial Arts and Textile Science, Matsugasaki, Kyoto 606)

PVA compounds plasticized by various amounts of glycerin were spun into fibers by melt process

and followed by the extraction with methanol and the drawing in a hot air. The structure and

properties of the fibers were studied by a x-ray instrument, a polarizing light microscope, a density

gradient column, a differential thermal analyser and a tensile tester. The best spinnability was found

for the compound containing 27 wt %glycerin. However, the exact glycerin content of the as-spun

fiber prepared from this compound was 11 wt%. The drastic decrease in glycerin content might be

attributed to the vaporization during the melt spinning and the bleeding after the melt spinning.

The extracted fiber could be drawn by around 7 times at 220•KC, while the plasticized as-spun fiber

was drawable up to 6 times. The drawn extracted fiber had the higher anisotropy evaluated by

polarized microscopy but the lower crystalline orientation by x-ray analysis than those of the

extracted fiber. This may be responsible to the mobility and the orientation of amorphous chains

in both plasticized and extracted fibers.

INTRODUCTION

The three techniques for producing fibers of synthetic linear polymers have been established:

melt, dry and wet spinnings. If a polymer can be fusible under reasonable conditions, the melt spin

ning is preferred because this is most convenient and simplest.

Poly(vinyl alcohol) (PVA) is currently spun into fibers by either wet or dry process since the de

composition temperature of PVA is very close to

the melting point. These solution techniques seem to be much complicated than the melt spinning. For example, the wet process is made up of many

sub-processes; i. e., the. preparation of spinning dope, the control of coagulant, the washing of

resultant fibers, etc. Many researchers, therefore, have challenged to

the melt spinning of PVA applying a variety of

techniques. Uzumaki et al. 1) studied the melt spinning of a fully saponified PVA of which degree of polymerization ranged from 1,000 to

2,000 and obtained fine fibers in a small scale. According to their patent, however, the spinning speed was limited to 30 m/min. Many attempts

have been made on the improvement of the spinnability of PVA. Kawakami et al.2) melt-spun some partially saponified PVA and then com

pleted the saponification after melt spinning. Matsumoto et al.3) studied the melt spinning of

poly(vinyl alcohol-ethylene) copolymer and found that the copolymer had to contain at least 50 mole%

of ethylene for smooth spinning. Kato et al.4)

made a fiber from a polymer blend of 30 wt% nylon 6 with 70 wt% PVA. However, it is obvious

that the chemical and/or the physical properties of the spun fibers must differ from the fiber of pure PVA.

Another catelogy of investigations is characterized by the use of additives. Sakamoto et al.5)

employed water vapor as a plasticizer of PVA to reduce the melt viscosity and to spin fine fibers

at higher take-up speeds. Yamada et al.6) also used water together with chemicals such as ethylene glycol, glycerin, and phthalic esters. These ideas have been extensively developed by Mashio et al.7) They used plasticized commercially avail

able PVA's with either ethylene glycol, ethylene urea, or glycerin and succeeded in the high speed melt spinning up to 500 m/min for a compound

comprising 10 parts of PVA and 5 parts of ethylene

urea. They also described that another mixture

* On leave from The Institute of Chemical Fibers, Changchun City, Jilin Province, China

T-62 SEN-I GAKKAISHI (報 文) (44)

containing 60 parts of PVA and 40 parts of

glycerin could be spun at 300m/min. These fibers could be hot-drawn, annealed, and acetalized

in the conventional manner. In their patent, how

ever, no attempt was made on the elimination of

plasticizer from the resultant fibers. They have also mentioned that the tensile strength of the

fiber plasticized with glycerin was 8g/d. It implies

that the plasticizer might play an important role

on the molecular orientation in the course of

hot-drawing.

Unfortunately, the details of the procedure and

the properties have not been revealed in the

patents. In the present paper PVA's plasticized with varying amount of glycerin were melt spun

into fibers and then glycerin was extracted from

the fibers. The hot drawing of the fibers was made

on both the plasticized and the extracted samples.

The orientation behavior was estimated by using

x-ray diffraction analysis and polarized microscopy.

EXPERIMENTAL

MATERIALS

A PVA was supplied by UNICHIKA Co. Ltd.

According to UNICHIKA, the average degree of

polymerization was 1,700, the saponification de

gree was 99.91mole%, and the residual sodium

acetate was around 0.5 mole%.

Prior to the fabrication, the PVA powder was

washed with large amounts of water and methanol

to eliminate the residual sodium acetate and then

dried in an air oven. The 70 parts of glycerin

(reagent grade) was mixed with 30 parts of distilled

water, as a swelling agent to PVA and the carrier

of glycerin. A predetermined amount of this

glycerin-water mixture was sprayed onto the PVA

powder. After being aged at about 95•KC for 24

hrs, the compound was dried in a vacuum oven at

105•KC for 24 hrs to eliminate the water. The

residual water results in the formation of bubbles

during the melt spinning at elevated temperatures.

The dry compounds prepared were allowed to

stand for two more days in a desiccator with silica

gels. The glycerin content of the final compounds

was varied from zero to 45 wt%.

MELT SPINNING

Melting point of various glycerin plasticized

PVA's was first determined by using a differential

thermal analyser (DTA), Model DT-20B, Shimadzu

Co. Ltd. The heating rate was 20•KC/min.

The spinnability of these plasticized PVA's was

qualitatively estimated by using a plunger type

extruder equipped with a spinneret having a 1.0

mm orifice, which was designed in our laboratory.

The actual melt-spinning experiments were per

formed with an another laboratory spinning instru

ment equipped with a 25 mm single screw extruder,

a spinneret with twelve 0.5 mm nozzles, and a

conventional take-up device.

The glycerin in the as-spun fibers was extracted

with boiling methanol in a Shoxlet apparatus for

more than 14 hrs. The drawing was made in a

Bistron instrument, Iwamoto Machinary Ltd.,

which was basically designed for the biaxial draw

ing of plastic films.

MICROSCOPIC OBSERVATIONS

The surface structures of the resulting fibers

were observed with a scanning electron microscope.

Gold-sputtered samples were subjected to the

observations and a standard technique was em

ployed.

O RIENTATION

The overall degree of orientation of fibers was

estimated with a polarizing microscope equipped

with a Berek compensator. The measurement was

made under the sodium D-line, of which wave

length is 589 nm. The crystalline orientation was

estimated from wide angle X-ray scattering (WAXS)

studies radiated by the nickel filtered Cu-ku beam

through a tubular collmetor with 0.5 mm slit.

DENSITY AND CRYSTALLINITY

Assuming the simple two phase model, the

degree of crystallinity was estimated by the densi

ty, measured in a n-heptane-carbon tetrachloride

density gradient column. The density D is related

to the crystallinity X using a common equation:

1/D=(X/Dc)+(1-X)/Da where suffices c and a

refer to the crystalline and amorphous, respective

ly. The crystalline and the amorphous densities

have been reported by many researchers. Among

them, in the present study, we chose the values

proposed by Sakurada et al.8), namely Dc=1.345

g/cm3 and Da=1.269g/cm3

TENSILE PROPERTIES

Tensile experiments were run at room tempera

ture by using an Instron type tensile tester,

Shinkoh Model TOM 200D. Guage length was

30mm and cross head speed was 30 mm/min.

(45) Vol.38, No.2 (1982) T-63

RESULT AND DISCUSSION

Figure 1 shows the variation of the melting

point of PVA as a function of the weight fraction

of glycerin, indicating that the melting point is a

linear decreasing function of the glicerin content.

The spinnability of these PVA-glycerin compounds

was qualitatively examined by using a small plunger

type extruder. Table 1 summarizes the spinnability

of various compounds. The best spinnability was

found for two samples which contained 27 and

35wt% glycerin. For all samples the optimum

spinning temperature was approximated to be

thirty degrees higher than their melting point, i. e.,

(Tm+30)•KC. When the spinning was conducted at

temperatures lower than (Tm+30)•KC, the polymer

stream emerged through the spinneret orifice be

haved as an elastomer and was hardly melt-drawn

into fine fibers. Whereas when the spinning tem

perature raised to higher, the extrudates colored

brownish and decomposed just below the spinneret.

Although the compound containing 35wt% glycer

in was almost equally spinnable to that of 27wt%

glycerin, the latter was subjected to the further

experiments in order to minimize the plasticizer

Fig. 1. The variation of melting point as

a function of glycerin content.

content.

In the melt spinning with a screw extruder, the

temperature of the spinneret was fixed at 232•KC

which was just 30•KC higher than the melting point

of this compound. Under such conditions of the

spinning, the compound was very good at the

spinnability and hence the fiber having the diame

ter less than 60 microns could be obtained at the

take-up speed of 300m/min or faster. For the

convenience of handling, however, the rates of

extrusion and taking-up were controlled to obtain

an as-spun fiber of 150 microns.

Since the bleeding of glycerin was apparently

observed on the surface of the as-spun fibers, the

fibers thus prepared were stored in a desiccator

with silica gel for more than three weeks to equi

librate the glycerin content. The amount of

glycerin in the as-spun fiber was cross-checked by

both Shoxlet extraction and DTA techniques. The

actual glycerin content determined from the ex

traction experiment was about 11wt%. This

amount was in good agreement with the result of

DTA, i. e. the melting point of 220•KC measured

with DTA. (c. f. Fig. 1)

This drastic decrease of glycerin from 27 to 11

wt% must be caused not only by the bleeding from

the as-spun fibers but also by the evaporation in

the course of melt-spinning. The bleeding of

glycerin from the as-spun fibers was confirmed

from its moisture regain behavior. When the as

spun fiber was allowed to stand in a desiccator

conditioned at 20•KC and 65%RH, the sample

weight increased significantly as represented in

Curve 1 of Figure 2. In the earlier stage, the

increase is rapid and fairly linear with respect to

the time stored. But the rate of increase gradually

slows down and finally levels off on prolonged

storage. This suggests that the sorption of water

must be governed by two different mechanisms:

Table I. Spinnability of PVA plasticized by various amount of glycerin.

Note; -indicates the skip of the examination.

T-64 SEN-I GAKKAISHI (報 文) (46)

Fig. 2. The plot of moisture regain against the

time stored at 20•KC and 65%RH for

as-spun (_??_) and extracted (_??_) fibers.

in the earlier stage of the rapid increase, water

must be sorbed into the glycerin which bled on

the surface of the fibers. After the saturation of

water in this glycerin layer, the diffusion of water

from glycerin to the fiber may be attributed to

the apparent rate of the weight increase.

Some samples of the as-spun fibers were ex

tracted in a Shoxlet apparatus to get extracted

fibers and the rest was subjected to drawing

without extraction. Both the plasticized and

extracted fibers were drawn in the Bistron instru

ment at various temperatures. The maximum draw

ratio varied markedly with temperature as shown

in Figure 3. It is shown that the drawability of the

extracted fiber was always greater than that of the

plasticized one. This is possibly attributed to the

molecular relaxation during the course of extrac

tion.

Figure 4 shows the variation of the fiber density

with draw ratio. The density of glycerin at 30•KC

is 1,26g/cm3 which is very close to the amorphous

density of PVA 1,269g/cm3, given by Sakurada

et al.8). Assuming that glycerin locates only in

amorphous region, we can adopt the conventional

equation: (1/D)_??_(X/Dc)+(1-X)/Da. The density

(and hence the crystallinity) of the extracted fiber

was independent to the draw ratio, whereas that

of the plasticized fiber increased monotonically

with increasing draw ratio.

In usual melt spinning the cross section of the

resulting fiber is almost circular, as the volumetric

retraction due to the solidification is rather small.

It is interesting to see whether the fibers prepared

Fig. 3. The effect of draw temperature on the drawability of plasticized (_??_) and extracted (_??_) fibers.

Fig. 4. The effect of drawing at 220•KC on the

density and the crystallinity of plasti

cized (_??_) and extracted (_??_) fibers.

in the present study have circular cross section or

not, even after the elimination of glycerin. The

SEM photographs shown in Figure 5 indicate that

all the fibers are circular and no detectable defect

can be seen on their surfaces. So, the anisotropy

of the fibers was determined by using a conven

tional polarizing microscope.

Figure 6 shows the relation between birefrin

gence and draw ratio for both the plasticized and

extracted fibers. Before drawing, the birefringence

of the plasticized fiber is slightly greater than that

of the extracted one. The difference may be

attributable to the molecular relaxation during the

extraction of glycerin, as was mentioned above.

The birefringence of the extracted fiber increased

with drawing and reached to 38•~10_??_3 at draw

ratio of 7. The birefringence of the plasticized

fiber tends to increase only in the early stage of

(47) Vol. 38, No. 2 (1982) T-65

(A)

(B)

(C)

(A)

(B)

(C)

(D)

(E)

(F)

(G)

(H)

Fig. 5. The SEM photographs of various PVA fibers: (A) plasticized and undrawn

(60x), (B) extracted and undrawn (60x), (C) extracted and drawn by 7 times (120x).

Fig. 6. The effect of drawing on the birefringence of plasticized (_??_) and extracted (_??_) fibers.

Fig. 7. The WAXS patterns of various drawn fibers: (A) plasticized and undrawn, (B) plasticized and drawn by 2.1, (C) plasticized and drawn by 4.0, (D) plasticized and drawn by 5.0, (E) extracted and undrawn, (F) extracted and drawn by 2.1, (G) extracted and drawn by 4.0, and (H) extracted and drawn by 5.0.

drawing and soon levels off to 22•~10-3.

Figure 7 shows the wide angle X-ray scattering

patterns of various PVA fibers. It is clear from

Figure 7(A) and 7(E) that the use of the plasticizer

does not modify the crystalline form but reduces

the crystal size. In other words, the plasticizer

does not induce any unfavourable effect on the

T-66 SEN-I GAKKAISHI (•ñ•¶) (48)

crystallization of PVA. On drawing, the crystallites in the both fibers tend to orient along the fiber axis. The WAXS patterns obtained for the drawn extracted fibers seem to be more diffused than those of the drawn plasticized ones. Indeed, in the patterns taken for the extracted fibers the

(101) reflection comes in contact with the broad (200) arc. On the other hand, those of the drawn plasticized fibers are obviously distinguished each other. In addition, the azimuthal intensity distributions of the (101) and (200) reflection are sharper for the plasticized fibers than for the extracted ones.

The situation is just opposite to the result obtained by the polarizing microscopy. The difference in the orientation behavior must be explained in terms of the mobility of their amorphous chains. The plasticizer may facilitate the reorganization of crystallites during the drawing but may enhance the relaxation of amorphous chains after the fibers were released from the drawing force.

Figure 8 shows the typical stress-strain curves of drawn fibers. As was expected, the drawn extracted fiber has the greater tensile strength, the

Fig. 8. The stress-strain relation for plasticized

(1) and extracted (2) fibers.

higher initial modulus and the smaller elongation

at break. Numerical values of the tensile properties

of the fibers are given in Table II.

Finally the extracted drawn fibers were acetalized

in the ordinary manner. The softening temperature

in water for the unacetalized fiber was 81.4•KC.

ACKNOWLEDGEMENT

The authors indebted to Drs. T. Yasui and A.

Kubotsu of Kurare Co. Ltd. for obtaining the SEM

photographs.

REFERENCES

1) M. Uzumali, E. Shimoda, and M. Takamura; Jpn. Patent s36-12559 (1961)

2) H. Kawakami, H. Fujii, and H. Takachi; Jpn. Patent s47-22099 (1972)

3) T. Matsumoto et al.; Kobunshi Kagaku, 23, 610 (1971), Sen-i Gakkaishi, 30, T391 (1974), ibid, 30, T398 (1974), and ibid, 31, T152 (1975)

4) H. Kato, K. Fujiwara, and U. Anzai; Jpn. Patent s48-22833 (1973)

5) T. Sakamoto, H. Kizu, Y. Yokomaku, S. Hibara, I. Otsubo, and S. Kitagawa; Jpn.

Patent s23-1140 (1948) 6) M. Yamada, T. Kinoshita, and T. Inoue; Jpn.

Patent s25-356 (1950) 7) F. Mashio, K. Yamaoka, H. Kawakami, and

E. Sato; Jpn. Patent s37-9768 (1962) 8) I. Sakurada, K. Nukushina, and Y. Sone;

Kobunshi Kagaku, 12, 506 (1955) 9) H. Kawase; "Sen-i Binran", Maruzen, Tokyo

(1968), p. 659 10) E. Nagai; ibid, p. 659

Table II. Tensile properties of fibers.

(49) Vol. 38, No.2 (1982) T-67

グ リセ リンで可塑 化 した ポ リビニル アル コール の溶融 紡 糸

京都工芸繊維大学繊維化学科 朴 東旭,北 尾敏男

グ リセ リンを 種 々の割合 で混 合 したPVA配 合物 を溶

融紡 糸 した。 この繊 維 を メタノール紬 出 しグ リセ リンを

除去 した後,加 熱 空気 中で延伸 した。繊 維の 構造 およ び

性 質を,X線回 折,偏 光顕微鏡,密 度 勾配管,DTA,

および引張試験機 に よ り評価 した。 グ リセ リンを27wt

%含 む試料 が最 も良 好 な紡糸性 を示 した。 この配合 物か

ら作 った未延 伸繊維 は,紡 糸お よび保存 過程 におい て グ

リセ リンの一部 を失 ってお り,延 伸 に供 した 試料中 には

11wt%の グ リセ リンが残 って いた。 メ タノール抽 出に

よ り完全 に グ リセ リンを 除いた繊 維 は,約7倍 延伸す る

ことがで きた 。グ リセ リンを除 いた繊維 は グ リセ リンを

含 む繊維 よ りも大きい 複屈折 と小 さい結 晶勾 配度を示 し

た 。 この原 因 は,非 晶質 の配向度 の相達 によ って説明 さ

れ た。