Studies on Regenerated Protein Fibers. 111.

Production
of Regenerated Silk Fibroin Fiber by the
Self-Dialyzing Wet Spinning Method

KlYOlCHl MATSUMOTO,' HlROYUKl UEJIMA,',* TATSUO IWASAKI? YOSHlYUKl SANO?

and HlROSHl SUMINO'

'Faculty of Textile Science, Kyoto Institute of Technology, Matsugasaki Sakyo-ku, Kyoto 606, Japan; 'Shiga Prefectural
Junior College, Hassaka Hikone, Shiga 529, Japan; and 3 K a ~ h Junior
o College,
Chioin Higashiyarna-ku, Kyoto 605, Japan

SYNOPSIS

The spinnable dope containing 20 wt % of silk fibroin dissolved in the MU solvent (e.g.,
LiBr * H 2 0 = 40 wt %, and EtOH : H 2 0 = 100 : 0 in volume) was spun into a coagulant and
subsequently drawn, to give the following results.

1. Methanol, ethanol, and isopropanol were a predominant coagulant for the dope,
and the coagulant containing 10% LiBr H20 is best.-
2. Spinnability depended upon the ratio of nozzle hole length to diameter (LID) and
was improved by increasing LID from 4 to 10. Coagulativity was not sufficient at
the SF concentration of 16%, but improved by raising it up to 20%. Drawability of
the spun filaments in water was improved at a temperature above 60°C.
3. The drawn fiber showed excellent mechanical properties, especially tensile toughness
when as-spun fiber was drawn in water at 61OC. An X-ray diffraction diagram of
the drawn fibers indicated their crystalline axes were partially oriented along the
fiber axis. 0 1996 John Wiley & Sons, Inc.

INTRODUCTION or sodium sulfate, to produce some regenerated ma-

terials of various shapes. However, they did not suc-
In 1930 Yazawa et al. dissolved silk fibroin (SF) in ceed in producing any regenerated SF fiber. The
a concentrated magnesium nitrate solution to pro- reason for such a failure in the past resides in the
duce a raw dope, which was then dialyzed and con- fact that any suitable system of solvent/nonsolvent
centrated before spun into a saturated ammonium was not established by them.' The later proposals
sulfate solution. They reported that the regenerated that relied on the dialysis of the dope not only in-
fiber had a breaking strength of 2.5 g/den a t an volved very intricate operations but gellation of SF
elongation of 20-25%.' or a like phenomena of the dope. Consequently, the
Yazawa's report was followed by many similar regenerated fibers were not free from undesirable
proposals in which SF was dissolved in various in- variation in their mechanical and/or other proper-
organic salt aqueous solutions, and the salts were ties. To overcome such drawbacks inherent in prior
removed by dialysis to prepare spinnable dopes. methods, a novel system of solvent/nonsolvent was
Those dopes were also coagulated in a saturated so- established by us.
lution of inorganic salt such as ammonium sulfate We found that the Matsumoto/Uejima (MU)
solvent as previously reported3 is an excellent sol-
vent for a spinning dope, and that methanol and
* To whom correspondence should be addressed.
Journal of Applied Polymer Science, Vol. 60,503-511 (1996)
others are appropriate nonsolvents that can coag-
0 1996 John Wiley & Sons, Inc. CCC 0021-8995/96/040503-09 ulate the spinning dope even if undialyzed.

503
504 MATSUMOTO ET AL.

The results of spinning experiments carried out Measurement

on our basic concept mentioned above are reported
Viscosity of the Diluted Solutions
as follows.
Viscosity of the solution of specific amounts of SF
dissolved in MU solvents (each comprising 40% by
-
weight of LiBr H 2 0 in ethan01)~ was measured us-
EXPERIMENTAL ing an Ubberohde viscometer at 30°C. The data were
used to calculate the values qsp/C,and the relation-
Row Material ship between qsp/Cand C was plotted to give straight
lines. Those lines were then extrapolated to C = 0
Silkworm cocoons produced in the farm at the Kyoto so as to determine intrinsic viscosity. [o].
Institute of Technology were used. After removing
sericin by the conventional method, the cocoons X-Ray Diffraction Analysis
were dried to provide SF. Some cocoons were divided
X-Ray equipment made by Rigaku Denki Co. Ltd.
into their strata, each composed of a continuous fil-
was used under the following conditions: Ni-filtered
ament, before removing sericin by the conventional
Cu-K, line was used as the X-ray source, voltage
method so as to prepare a discrete SF sample for
of 40 kV, current of 20 mA, exposure for 3 h, and
each stratum. Further, initial filaments that the co-
camera length of 32.31 mm.
coons had spun a t first for sticking their succeeding
length of filament to were also collected separately
Infrared Absorption Spectrum
from the succeeding normal filaments, and their
sericin was removed by this method. An infrared spectrometer IRA-1 made by Nihon
Bunko Co. Ltd. was used under the following con-
ditions: wavelength range of 4000 to -650 cm-',
Preparation of SF solution resolving power of 4 cm-', scan speed of 4 cm, and
temperature of 20°C.
Given amounts of the dried SF were mixed well with
the MU solvents of varied compositions, and heated Density
to 70°C while being stirred. The SF thus dissolved
in the solvents provided the SF solutions. Density gradient columns with mixtures of n-hep-
tane and carbon tetrachloride were used at 30°C.

Rating of Coagulativity
Preparation of Coagulant
Visual Inspection. The dope was coagulated in a se-
Ion-exchanged water was used, together with ries of coagulants in the manner described above.
ethanol, methanol, or isopropanol (Wako Junyaku The coagulated films were classified by inspection
Co., Ltd.), which were of the highest reagent grade. of their overall coagulativity as shown in Table I.
Lithium bromide was the industry grade (Wako A droplet of 10% SF solution (dope) on a prep-
Junyaku Co., Ltd.). These reagents were used alone aration glass was sandwiched with a cover glass, as
or blended a t varied ratios to provide the coagulant. shown in Figure 1. The dope was then gently pressed
to spread it around, and to make a thickness of 0.5
mm using the pulp spacer. The dope was dipped in
Formation of Sample Films the coagulant at 20°C for a specific time, then the
length of a coagulated portion was measured under
An appropriate amount of the concentrated SF so-
a microscope without removal from the coagulant.
lution was poured on a glass plate, and a knife was
used to spread the dope over the plate with the cau-
Mechanical Properties
tion not to involve any bubbles. After keeping the
plate horizontal to make a smooth surface of the A uniaxial tensile tester made by Iwamoto Seisak-
poured dope, the glass plate was immersed in a co- usho Co. Ltd. was used under the following condi-
agulant bath at 20°C to form a film. The fully co- tions: gauge length of specimens of 20 mm and close
agulated film was desalted and kept in a methanol head speed of 25 mm/min. The moisture content of
aqueous solution to prevent hardening and decaying. samples was regulated under standard conditions.
 REGENERATED PROTEIN FIBERS. 111 505

Table IA Results of Film Forming Test Coagulation boundary length Cover glass

Composition of Solvent:
EtOH/H20" ( V O ~%)
Coagulation
Agent 100/0 80/20 65/35 45/55

MeOH o/c o/c OA,C A/A,B,C

EtOH A/C A/C A/A,C X
i-PrOH e/C A/A,B A/A,C x
HZ0 X X X X
MeOH/H,O (vol %)
97/3 O/C A/C O/A,C X
95/5 O/B O/C A/A X
90/10 A/C A/C A/C X Figure 1 Measurement of the coagulation length.
85/15 A/C A/C A/C X
MeOH/LiBr * HzO (wt %)
ulation bath, stretched between the first and second
95/5 o/c o/c A/C A/C
roll, and then wound on a take-up roll. Important
89/11 @/C o/c o/c A/C
spinning conditions were as follows: the nozzle had
85/15

a
o/c o/c A/C
LiBr H20/EtOH,H20= 50/50 (wt %) constant. Bath tem-
A/C
six spin holes having a diameter (D)of 0.3
-
mm and a length ( L )of 0.6 4 mm. The throughput
0.5 -
perature, 20°C. Key: @, better; 0, good; A, bad; X, did not co- of the dope was 1 cm3/min, with the coagulation
agulate; A, contains small bubbles; B, a little swelling; C, effusion.
Open marks signify translucent film. Solid marks signify white bath of methanol and the stretching bath of water
film. or methanol. The spinnable dopes were prepared by
dissolving 16 or 20% by weight of SF in MU solvents
Spinning Conditions
prepared by dissolving 40% by weight of LiBr H 2 0 -
in a mixture of EtOH and water, with the ratio of
A test apparatus as shown in Figure 2 was used for EtOH to H 2 0 being 100 : 0.
wet spinning for the dope samples. The dope ex-
truded through a nozzle was coagulated in a coag-
RESULT AND DISCUSSION
Table IB Results of Film Forming Test Inter- and lntracocoon DP of SF
Composition of Solvent: It is an established but questionable theory that
EtOH/H,O" (vol %) SF molecules are of varied molecular weight falling
Coagulation within a range of lo4to lo6, and the thickness of the
Agent 100/0 80/20 65/35 45/55 SF filament varies among the portions of each co-
coon4Note that SF is inevitably decomposed in the
MeOH O/A O/B OA,B A/A,B
course of preparation to determine the viscosity of
EtOH ./A ./A e/A,B A/A,B
i-PrOH @/A @/A,B A/A,B A/A,B diluted SF solutions in the solvents used conven-
HZO X X X X tionally. However, the MU solvent keeps the SF
MeOH/HzO (vol %) molecule in the solutions without degradation. We
97/3 O/A O/A O/A O/A employed the MU solvent to determine intrinsic vis-
95/5 O/A O/A O/A,C A/A,B,C cosity [ 711 of SF and to examine the variation of inter-
90/10 O/A O/C A/A,C A/A,C
85/15 A/C A/C a/c A/C
-
MeOH/LiBr HzO ( w t %) EX 1 : Nozzle
2 : Coagulation Bath
3 : First Roll (Double)
95/5 O/C O/B,C O/B O/B 4 : Drawing Balh
89/11 @/C @/C @/C @/C 5 : Second Roll (For Winding)
85/15 o/c o/c o/c o/c
P
'EtOH/H,O = 90/10 (vol %) constant. Bath temperature, d
2OOC. Key: Q, better; 0, good; A, bad; X, did not coagulate, A, 4 5
contains small bubbles; B, a little swelling; C, effusion. Open
marks signify translucent film. Solid marks signify white film. Figure 2 Spinning apparatus.
506 MATSUMOTO ET AL.

Table I1 Intra- and Intercocoon DP Variation of SF

Standard
Inner Intermediate Outer Average Deviation

[vl (dL/g) 0.62 0.58 0.65 0.62 0.03

0.84 0.68 0.72 0.75 0.07
0.63 0.66 0.69 0.66 0.02
0.74 0.68 0.83 0.75 0.06
0.71 0.76 0.71 0.73 0.02
Average 0.71 0.67 0.72 - -
Standard
deviation 0.08 0.06 0.06 - -

and intracocoons. Table I1 summarizes results of our aqueous solutions of inorganic salt. Both the alcohol
measurement. There is no or almost no difference in and lithium bromide in each MU solvent behave as
the viscosity-average degree of polymerization (DP) a good solvent in that SF molecules can fully ex-
between or in each of the cocoons that were produced tend.2,6On the contrary, water acts as a poor solvent
at the same time and the same area. in which SF molecules shrink when the solution of
SF dissolved in MU solvent is dialyzed by the tra-
ditional method: and the inorganic salt (i.e., lithium
Production of Regenerated Silk Fibroin Fiber
bromide) and alcohol are removed, thereby unde-
Coagulativity of the Concentrated Silk sirably rendering spherical the SF molecules that
Fibroin Solution had been in their extended state. Dialysis of such a
solution will thus produce an unsatisfactory dope
Coagulating Ability. Coagulation of SF aqueous so- for the spinning process. From this viewpoint, we
lution has been studied for many years, and, gen-
have sought a novel coagulation system from which
erally, saturated aqueous solutions of inorganic salts
SF molecules can coagulate in the extended state.
have been used in the so-called salting-out m e t h ~ d . ~
We had observed in our previous research that less
However, the solubility in water of SF is not high,
so that an aqueous solution of SF separates SF in a
-
than 30 wt ?6 of LiBr H 2 0 can never dissolve the
SF, at any ratio of alcohol/water in MU solvents.
gel state easily; SF molecules in aqueous solution We expected that such an MU solvent with low con-
cannot be presumed to be stable. Thus, it is unlikely tent of LiBr * H 2 0 could act as a favorable poor sol-
that the SF molecules would be salted out in their vent, which ensures the extended state of SF mol-
well-extended state. The MU solvents' are alcoholic-

R
C I
xzn

I
50 100 150
Dipping Tim (sec)
Dip@ng Tim?(sec)
Figure 4 Effect of the ratio of C,H50H to H 2 0 (V loo/
Figure 3 The initial coagulation rate for the SF solution 0, A 95/5, 0 90/10, W 87/13) in the MU solvent
-
of MU solvents (LiBr H,O/ROH = 50/50) R, CH3H, CzH5 .
(LiBr H20/CZH50H= 50/50) on the initial coagulation
0, i-C3H70HA, at 27°C. rate in CH30H at 27°C.
 REGENERATED PROTEIN FIBERS. 111 507

ure 3 shows the initial coagulation rate for the dif-

ferent alcohols. i-PrOH gives the largest coagulation
length, MeOH gives the smallest length, and EtOH
gives an intermediate one. However, the alcohols,
except MeOH, failed to produce a film whose round
contour was distinct and clear. This fact means that
a certain quantity of the SF dope drained out a t the
initial stage of coagulation with the alcoholic co-
agulants composed of i-PrOH or EtOH. This result
also shows that MeOH is the best alcohol to for-
I mulate the coagulant.
Ob 50 100 150 Figure 4 shows the effect of ratios of EtOH to
D i e ’Ilme [sec] water in the MU solvent for dissolving SF, on the
coagulation in MeOH-based coagulant. Coagulation
Figure 5 The initial coagulation rate of SF solution
dissolved in MU solvents (LiBr * H20/C2H50H= 50/50) length is greatest at a ratio of alcohol to water a t
-
for the concentration of LiBr H,O; 0, 0 5, A 10, V 13 95 : 5. The increase of water content in coagulant
-
wt % in CH30H (LiBr H20/C2H,0H) at 27°C. made the coagulativity poorer as previously sum-
marized. It is concluded that the richer water content
in coagulant allowed a greater amount of dope to
ecules being coagulated. Following this scheme, we drain out into the coagulant.
prepared a series of specified MU solvents in which Figures 5 and 6 show the effects of the blend ratio
-
LiBr H 2 0 content was varied, with the blend ratio -
of MeOH to LiBr H20.The initial coagulation rate,
of alcohol/water kept constant to evaluate their co- defined as the value at 10 s dipping, was decelerated
agulating ability. -
as the LiBr H 2 0concentration exceeded 10%in the
bath, in a case wherein SF dissolved in an MU solvent
Film Formation in Coagulant. Methanol (MeOH), whose LiBr * H 2 0 content was 50% and the ratio of
ethanol (EtOH), or iso-propanol (i-PrOH)was used EtOH to water was 1OO:O. With an increase of
and the alcohol component of the coagulant. The LiBr.H20 concentration of more than 13%, the
coagulant ability was inspected in terms of the abil- boundary region of coagulation became so indefinite
ity be form a film; results are summarized in Table that coagulation length could not precisely be mea-
I. Among the alcohols, MeOH could produce a better sured. In the wet spinning process, fiber formation
film than the other alcohols, for the SF solution with (i.e., dope coagulation) is regarded as a “phase sep-
MU solvent comprising EtOH : H 2 0 > 65 : 35. Nei- aration process”’ of a certain kind. One of the factors
ther bubbles nor swelled portions were observed in determining the rate of this process is the ratio of
this case. An increased ratio of the water in MU diffusion rate of a solvent to that of a nonsolvent. In
solvent did cause bubbles in the regenerated film,
nevertheless alcohol was used to formulate the co- 5.0
agulant. MeOH would be regarded as one of the best
candidate alcohols in the coagulant. As far as the
content of water in the coagulant is concerned, the
ratio of MeOH to H 2 0 > 95 : 5 gave a better film.
The LiBr-H20 content in coagulant was also al-
tered, and MeOH : LiBr.H20 at 89 : 11 made co-
agulation satisfactory. These results suggest the best
candidates for the solution and the coagulant for SF
regeneration are as follow: the solvent is MU solvent
those water content is 20%or less and the coagulant
is composed of another MU solvent containing 11%
-
or less of LiBr H 2 0 dissolved in MeOH as the al-
LiBr Y O in Cbag Age& (W!)
cohol blended with 5% or less of water.
Figure 6 Relation between the initial rate of coagulation
Rate of Initial Coagulation. The initial coagulation -
and concentration of LiBr H 2 0 in CH30H (LiBr H 2 0
rates in the coagulant are shown in Figures 3-6. Fig- = 50%).
508 MATSUMOTO ET AL.

Table I11 Mechanical Properties of Fibers

Tensile Young's
Sample Strength Fineness Density Strength Elongation Modulus
No. Strength Bath Ratio (den) (g/mL) (d/g) (760) (g/d)

1 Air 1.0 x 391/6f 1.345 0.70 1.7 46

2 Air 2.2 x 202/6f 1.347 0.96 2.4 49
3 Water (18.5"C) 2.8 X 151/6f 1.347 1.17 2.5 59
4 Water (61.0"C)
Reference data (Silk)
Reference data (Wool)
3.2 X
-
-
134/6f
-
-
1.33-
1.347

1.32
1.45 3.0
1.0
--
1.09
4.0
1.7
--
11.0
15 25
25 35
50
11
56
- 100
- 25

the present case, this factor depends on the difference tent, the dope's adhesion to the nozzle surface was so
.
in LiBr H 2 0 concentration between the two MU good that some extruded filaments could not be spun.
-
solvents. Dilution of LiBr H 2 0 in the dope becomes We visually inspected the running state of those dope
more difficult with a decrease in the difference of con- streams near the nozzle holes, and found a phenom-
centration, and the coagulation is rendered less ex- enon that resembled the so-called "Barus" effect?
peditious. The indefinite contour of the film coagulated Our countermeasures were as follows: an increased
-
in the bath containing 13%or more of LiBr H 2 0con- ratio of L/D of nozzle holes, a reduced as-spun rate
tent is probably a result of such a phenomenon. for the dope, and a reduced dope viscosity? The re-
duction of as-spun rate did not improve the problems.
Spinnability Increasing the L / D ratio from 0.6/0.3 (=2) to 2.0/0.5
The spinnability of the solution of SF dissolved in (=4)seemed to be better. Therefore, we further raised
MU solvent is excellent. Therefore, we presumed that the L/D ratio up to 4.0/0.4 (=lo). As a result, the
a proper coagulant could make the regenerated SF sticking problem was remarkably improved.
i n t the
~ shape of fiber. From the result of a coagulation
test, an optimum coagulant is an MeOH solution of Drawability
LiBr H20and the concentration is below 11%.When Coagulated filament was drawn to make the filament
-
the concentration of LiBr H 2 0 is small, coagulation mechanically stronger by giving orientation to the
was early, and the fiber that was obtained is easy to SF molecules. The coagulated SF filament has duc-
take up. From these result, if methanol was employed tile properties, but becomes somewhat brittle after
as a coagulant, then SF dissolved in MU solvent could drying from this state. The brittleness after drying
be regenerated to form a good fiber, without dialyzing may be caused by poor orientation of SF molecules
the SF solution in advance as previously mentioned. along the fiber axis. We therefore drew the as-co-
We also applied methanol as a coagulant for the spin- agulated (as-spun) yarn under varied conditions to
ning of the dopes, which were prepared by dissolving produce filaments that have sufficient strength. The
SF at varied concentrations in the MU solvent, which drawing ratios given in Table I11 were maximum
-
was composed of 40% LiBr H 2 0dissolved in EtOH/ ratios for the respective conditions. The hotter the
H2n0 (EtOH/H20 = 100/0). water in the drawing bath, the more extensively the
as-coagulated (as-spun) yarns could be drawn.
Coagulativity of Extruded Dope Streams. A dope
containing 16 wt % SF did not show good coagula-
tivity. Presumably, retardation in phase separation Micro Structures and Physical Properties of the
of SF was caused in this case by a retarded mutual Regenerated Silk Fibroin Fiber
exchange of the solvent components between the dope X-Ray Diffraction
and the coagulant. To resolve this problem, we in-
creased the SF content in dope up to the SF content Figure 7 (a-d) shows wide angle X-ray diffraction dia-
of 20% by weight, and found that the coagulativity grams of the fibers produced in the described manner."
and spinnability of the extruded dope were improved. Deby-Sherrer rings, which appear in the photograph
of sample 1at 2 B = 20.7" and 24.8", are inherent in
Dope Sticking Problem to Nozzle Surface. Although undrawn fibers. The more intensive Deby-Sherrer ring
the coagulativity was improved by raising the SF con- at 2 6 = 20.7' indicates the so-called ''P structure."'o
 REGENERATED PROTEIN FIBERS. I11 509

(c) (a
Figure 7 The X-ray diffraction photography of the regenerated fiber of (a) SF (11,
(b) SF (2), (c) SF (31, and (d) SF (4).

Samples 2-4 also show similar Deby-Sherrer rings, significant difference was found between the drawing
but some intensive diffraction at 20.7' appears on the conditions.
equatorial. This means that crystallization oriented to
the fiber axis proceeds by drawing of as-spun fiber. Infrared Spectrum
Figure 8 shows the IR spectrum of Samples 1 and
Density
4. As seen in Figure 8, broad absorption bands were
Table I11 shows the measured density. The drawing detected a t 1630-1660, 1530, and 1540 em-', which
of fibers slightly increased their density, though no were obviously caused by the amide-I group and
510 MATSUMOTO ET AL.

Magoshi12 in hot water, and it was found that its

crystalline structure had changed at a temperature
above 6OoC. It was also reported13 that a similar
phenomenon had occurred when stretching the film.
Our sample 4 was stretched at 61"C, significantly
higher than the temperature used for sample 3, and
also at a somewhat higher stretch ratio than for
sample 3. Such a stretching temperature for sample
4 is presumably near the critical temperature, which
allows the deformation of crystals. Therefore, the
drawing at this temperature in our research would
have rendered the twined SF molecules more ready
to be loosened and to convert themselves into p
structure, i.e., the predominant crystalline structure
of SF, whereby polymer chains were oriented along
the fiber axis.
The fresh spun filaments can be stretched ad-
vantageously under proper conditions, especially at
higher temperature, such that the chain molecules
are oriented along the fiber axis and crystallize so
as to produce a regenerated silk fibroin fiber of im-
proved mechanical properties.
I ' I
1800 1500 1200
W a v c 11u 111be r (ciii- 1)
CONCLUSION
Figure 8 Infrared spectrum of fiber: SF (1) and SF (4).
1. Methanol, ethanol, and isopropanol were
evaluated as the predominant ingredient of
amide-I1 group, respectively. The spectrum of sam- coagulant for the concentrated solution of SF
ple 1shows bands a t 1660 and 1540 cm-', which are dissolved in MU solvents. A methanol solution
inherent in the random-coiled polymer molecules, -
of LiBr HzO makes an excellent film as a co-
and the spectrum of sample 4 shows bands at 1630 agulant, and the coagulant containing 10%by
and 1530 cm-', which are inherent in the ,6 struc- -
weight of LiBr H 2 0 is best for the SF solu-
ture." These data indicate that SF molecules in tions in the MU solvent with the ratio of EtOH
sample 4 aligned along the fiber axis by drawing in to HzO 65 : 3 or more.
water at 61°C.
I .2
Mechanical Properties
Table I11 shows the mechanical properties of sam-
ples 1-4. Tensile strength as well as Young's mod-
ulus of each fiber generally increased with increasing
stretch ratio, however, the strength of sample 4 is
.
3
M
v
inferior to sample 3. The same relationship was 2 0.6
found also in their Young's modulus. However, the 2
v3
elongation a t breakage of sample 4 was greater than
the others and the shape of its stress-strain curve
greatly differed from the others, as seen in Figure
9. This difference indicates that sample 4 has a
higher tensile toughness, afforded by the following 0 6.0 12
mechanism. The stretching temperature and ratio
Strain (%)
for sample 4 were higher than those for sample 3. A
noncrystalline film of silk fibroin was treated by Figure 9 Stress-strain curve of fiber.
 REGENERATED PROTEIN FIBERS. 111 51 1

2. Coagulation length of the dope in a coagulant (Toyobo Co., Ltd.) for making many valuable suggestions
was measured in terms of coagulativity for the and for their help and cooperation.
concentrated SF solutions in MU solvent. It
was found that the leveled off coagulation
length was smallest when using methanol, and
the contour of the coagulated region was more REFERENCES
distinct than with other alcohols.
3. The spinnable dope containing 20 wt % of silk 1. S. Yazawa, J . Chem. SOC.Japan (Ind. Chem.), 63,
fibroin dissolved in the MU solvent (e.g., 1428 ( 1960).
-
LiBr HzO = 4076, and EtOH : HzO = 100 : 0 2. K. Matsumoto and H. Uejima, (to appear).
in volume) was spun into methanol as a co- 3. K. Matsumoto and H. Uejima, Japanese Unexam, Pat.
agulant and subsequently drawn to give the Publ. Hei. 3-89636 (28, 3/ 1992).
following results. Spinnability depended on the 4. K. Hayashi, “Kenshi no kouzou” Chikuma, p 340
ratio of nozzle hole length to diameter ( L / D ) ( 1949).
5. K. Otoi, Japanese Pat. Publ. Sho. 57-4723.
and was improved by increasing L I D from 4
6. K. Matsumoto and H. Uejima (to appear).
to 10. Coagulativity was not sufficient at a SF
7. A. Ziabicki, Seni Gakkaishi, 26,156 ( 1970).
concentration of 16%,but improved by raising 8. P. S. Spencer and R. E. Dillon, J . Colloid Sci., 4, 241
it up to 20%.Drawability of the spun filaments (1949).
in water was improved a t a temperature above 9. J. P. Todella, J . Appl. Phys., 2 7 ( 5 ) , 454 (1956).
60°C. The drawn fiber showed excellent me- 10. R. E. Marsh, et al., Biochim. Biophys. Acta, 16, 1
chanical properties, especially tensile tough- (1955).
ness, when an as-spun fiber was drawn in water 11. J. Magoshi, et al., J . Polym. Sci., Polym. Phys., 17,
at 61’C. An X-ray diffraction diagram of the 515 (1979).
drawn fibers indicated their crystalline axis 12. J. Magoshi, Koubunshi, 34,98 (1985).
was partially oriented along the fiber axis. 13. J. Magoshi, Seni-gakkai Reprints, 49, s-141 (1993).

We wish to express our gratitude to Dr. Prof. Tsuyoshi Received September 23, 1994
Kiyotsukuri, Dr. Naoto Tsutsumi, and for Dr. K. Yabuki Accepted August 5, 1995