Changes in Void Content and Free Volume in

Fibers during Heat Setting and Their Influence

on Dye Diffusion and Mechanical Properties

N. S. MURTHY, A. C. REIMSCHUESSEL, and V. KRAMER, Corporate

Technology, Allied-Signal Inc., Morristown, New Jersey 07960

Synopsis
The contributions of the changes in porosity and free volume to the differences in the properties
of dry and wet heat set fibers are analyzed. Nylon 6 fibers heat set in wet atmosphere (autoclave
and Superba process) show a significantly higher small-angle X-ray scattering intensity between
0.5" and 1.5" (Cu K a ) relative to corresponding non-heat-set fibers, whereas the fibers heat set in
dry atmosphere ( Suessen process) do not. This scattering arises from the electron density contrast
due to pores and free volume which lower the electron density between the fibrils. The average
diameter of the crystalline fibrils, as calculated from the Guinier analysis of diffuse scattering,
increases from ca. 50 to 65 A in all fibers upon heat setting. Additional scattering a t much lower
angles is attributed to aggregates of fibrils. Transmission electron microscopic studies of fibers
stained by AgI, through I,/methanol/AgN03 treatment, suggest that as AgI crystallizes in a dendritic
network along the paths opened up by the diffusion of iodine, it brings about large structural
changes in the polymer; this process reveals defects and latent diffusion paths in the fiber. The
dendritic structures of AgI are more extensive in wet heat set fibers than in dry heat set and non-
heat-set fibers. The ordering within the crystalline domains, the interlamellar spacing, and the
density of the fiber increase during both wet and dry processes. The role of these changes in the
structure and the morphology of the fibers on the dyeing and ozone fading behavior, and on the
strength and the modulus of the fibers, are discussed.

INTRODUCTION

Fibers used in home-furnishings are heat-set to stabilize the twist and crimp
so that the yarns maintain the aesthetic appeal of unmatted texture, tuft def-
inition, and resilience during wear. Dye diffusion in fibers is sensitive to the
heat setting process. While nylon fibers heat set (annealed) in the presence of
moisture (the continuous Superba process and the batch autoclave method,
both using saturated steam) dye darker, the fibers heat set in essentially dry
atmosphere (the continuous Suessen process) usually dye lighter than the non-
heat-set yarns. Analysis by wide- and small-angle X-ray methods has shown
that changes in the crystalline and interlamellar regions in both classes of
fibers (the wet and the dry process) are similar' : Lamellar spacing, crystallite
size in the equatorial plane, crystalline and fiber density, and the chain-axis
repeat all increase with annealing temperature; this is accompanied by loss of
crystalline orientation and a decrease in the van der Waals separation of the
hydrogen-bonded sheets. If dye diffusion were to be determined by these pa-
rameters alone, then there should be a consistent decrease in the dyeability in
both the wet and the dry heat-set fibers.' This is contrary to the observations

Journal of Applied Polymer Science, Vol. 40, 249-262 (1990)

0 1990 John Wiley & Sons, Inc. CCC 0021-8995/90/1-20249-14$04.00
250 MURTHY, REIMSCHUESSEL, AND KRAMER

cited above. Therefore, the differences in the dyeability of fibers heat-set by

different processes should be due to the changes in the amorphous domains, or
the porosity of the fibers.2 Irrespective of which morphological feature affects
the diffusion of dyes, wet-heat setting enhances the dye uptake, and therefore
yields fibers with a more “open” structure.
In this report, using small-angle X-ray scattering (SAXS) and transmission
electron microscopy ( T E M ) , we show that this “open” structure can be at-
tributed to a decrease in the average density of the amorphous matrix between
the fibrils caused by an increase in the volume fraction of pores (voids) or by
an increase in the free volume. We will use our data to understand the contri-
bution of voids and free volume to the differences in the dye diffusion, ozone
fading, and the mechanical properties in the non-heat-set. and the heat-set
fibers.

EXPERIMENTAL

SAXS data were collected from Suessen, Superba, and autoclave fibers listed
in Table I. Control, non-heat set fibers were also analyzed in each case. The
data were obtained using a linear proportional counter mounted on a Franks
am era.^ The intensity data were collected along the equator to eliminate the
interference from the lamellar peak in the analysis of the equatorial diffuse
scattering. The background scattering, obtained with no sample in the beam,
was subtracted from the sample curves. The curves were not smoothed. The
data were not desmeared since our scattering geometry could be approximated
by point collimation. The difference curves were analyzed using the Guinier
approximation. No corrections were made for multiple scattering.
The fibers used for electron microscopy were stained with AgI by soaking
the fibers in a solution of iodine in methanol, followed by conversion of iodine
to AgI by immersing the fibers in an aqueous solution of silver nitrate. The
fibers were then washed with distilled water, dried, embedded in epoxy, and
sectioned at room temperature for examination in a Hitachi 800 electron mi-
croscope operated at 200 kV.

TABLE I

Approximate duration
Fiber Processing mode Typical heat-setting conditions (min)

Suessen Continuous Dry heat (180-220°C) at 1

atmospheric pressure
Superba Continuous Saturated steam (110-150°C) 1
at high pressure (75 psig)
Autoclave Batch Alternate cycles of high 30
pressure (30-35 psig)
saturated steam ( 110-
150°C) and vacuum
 CHANGES IN VOID CONTENT 251

RESULTS

The background subtracted SAXS curves from the Suessen, Superba, and
the Autoclave processes are shown in Figure 1 (full lines) along with the scat-
tering curves from the corresponding non-heat-set or control fibers (points).
The scattering from the three control fibers are different, and the scattering
from heat-set fibers should be compared with the scattering from the corre-
sponding non-heat-set fibers. The left panel (curves a, c, and e ) shows the
scattering along the fiber axis, and the right panel (curves b, d, and f), shows
the scattering perpendicular to the fiber axis. The data in the left panel show

: a
40 - Suessen

Superba

50

40 Autoclave

30

20

10

n
"0 1 2 3 0 1 2
2 t) (Degrees, CuKa) 20 (Degrees, CuKa)
Fig. 1. Small-angle X-ray scattering curves from control (dotted) and heat set (full-line)
fibers. Curves in the left panel provide information about lamellae, and the curves in the right
panel contain information about fibrils: (a, b ) Suessen process; (c, d ) Superba process; (e, f )
Autoclave process.
252 MURTHY, REIMSCHUESSEL, AND KRAMER

that the separation between the lamellae, and the intensity of this lamellar
peak, increase upon annealing in all the three classes of fibers (Fig. 2). These
increases in lamellar spacing and SAXS intensity can be attributed to the
growth of the lamellae and an increase in their crystalline order (see Ref. 1
and references therein). Since the position and intensity of the SAXS peak
are sensitive to heat-setting temperature, they can be effectively used to monitor
all the three heat-setting processes. In the remainder of the paper, we will
concentrate only on the data in the right panel in Figure 1 (curves b, d, and
f ) . This equatorial scattering will be simply referred to as scattering.
The scattering falls off more rapidly in non-heat-set and Suessen-set fibers
than in Superba-set and autoclaved fibers (Fig. 1, right panel). This rate of
decrease in small-angle diffuse scattering can be analyzed using the Guinier
plots (In I(h ) vs. h2)shown in Figure 3. Although the Guinier approximation
[ eq. ( 1) ] is derived for low concentrations of scattering particles, it has been
shown4 that eq. (1)can be used even in concentrated systems, such as dense
solids, when the particles are not of uniform size and shape. It can be seen,
especially in the non-heat-set and the dry-heat-set fibers, that the Guinier plots

11c

1oc
s0)
.-C0
m
n
u)

--z
9)

J
5
9c

8C

0. 110 120 130 140 Wet

Heat Setting Temperature (“C)
Fig. 2. A plot of the lamellar spacing and SAXS peak intensity as a function of annealing
temperature in dry heat set (Suessen) and wet heat set (Autoclave) fibers.
 CHANGES IN VOID CONTENT 253

I
I 0.01 0 12
h2 (A-2)

Fig. 3. Guinier plots of the scatteringcurves shown in Figure 1 (right panel). The curves have
been offset vertically for clarity. The curves a, c, and e are from control fibers, and curves b, d and
fare from correspondingfibers which are heat set by respectively, Suessen, Superba, and autoclave
process.

can be resolved into two straight lines corresponding to scattering from entities
of two different sizes. These sizes can be estimated from the relation

where h = (47rsin 8) /A, X is the wavelength ( 1.542 A ) , 0 is half the scattering

angle, and r is the radius of the circular cross section of an elongated particle
whose long axis is normal to the incident X-ray beam.5
The results of the Guinier analysis for each of the fibers analyzed are shown
in Table 11. The slope of the Guinier plots at h2 > 0.0075A-' corresponds to a
diameter d of ca. 50 A in control fibers, and increases to ca. 65 A for heat-set
fibers. Similar values were obtained in two other sets of experiments. Note that
the scattering intensity in this angular range is higher in the wet-heat-set fibers
than in the non-heat-set and the dry-heat-set fibers. We will later argue (see
Discussion) that this increased scattering is due to higher void content and
increased free volume in wet-heat-set fibers, but the size calculated from this
scattering is related to the diameter of the fibrils.
We studied a phenol treated fiber to determine the source of the scattering
at 28 > 1.2O.A fiber (the non-heat-set fiber in the autoclave series) was immersed
in an aqueous 3.5%phenol for 16 h, washed in water for 2 h, and dried. A SAXS
scan from this partially dried fiber (Fig. 4) shows a resolvable peak at 1.4'
corresponding to a Bragg spacing D of 65 A. In other phenol-treated fibers,
254 MURTHY, REIMSCHUESSEL, AND KRAMER

TABLE I1
Estimated Sizes from the Guinier Plots of Control and Heat Set Fibers

Diameter (A)
Control Heat-set

20 > 1.2",h2 > 0.0075 A-' a

Suessen 44 57
Superba 57 70
Autoclave 47 65
20 < lo, h2 < 0.005A-*
Suessen 145 139
Superba 137 -

Autoclave 139 114

aThese values correspond to the diameter of the individual fibrils.

These sizes are tentatively assigned to fibrillar aggregates, but may in fact have contributions
from other sources such as large pores. These values are probably only lower limits since data in
Figure 2 may not extend to sufficiently low angles.

spacings as high as 100 8,were observed. This interference peak can be attributed
to periodic fluctuations in electron density between the crystalline fibrils and
interfibrillar amorphous regions. The distance D of 65-100 8, is larger than the
diameter d of 45-70 A determined from the Guinier analysis of the slope at h 2
> 0.0058,-2 from control and heat set fibers, respectively. This suggests that
the diameter d refers to the size of the fibrils, and the distance D might cor-
respond to the interfibrillar separation. These results are illustrated with a
model in Figure 5.
As already noted, besides the scattering at 20 > 1.2", which we have used to
calculate the sizes of the fibrils, an additional scattering is present at 28 < 1.2"
in all fibers, suggesting the presence of a second scattering species (Fig. 1).
From the limited range of the Guinier plots ( h 2 < 0.005A-'), we estimate the
lower limit for the diameter of these species to be ca. 150 8, (Fig. 3). This

1.51' " ' 1 ' I ' " ' ' " " " I
0 1 2 3 4 0 1 2 3
2 8 (oeg, cu K ~ ) 28 (Deg, Cu Ka)
Fig. 4. Comparison of the observed scattering curves for fibers which are: ( a ) non-heat-set;
( b ) partially dried after treating with phenol.
 CHANGES IN VOID CONTENT 255

1ooa

Fig. 5. A model of a filament of nylon 6 to interpret the SAXS and WAXD data in terms of
fibrillar crystalline and interfibrillar amorphous regions: ( 1 ) fibrils; ( 2 ) lamellae; ( 3 ) partially
extended chains in the interfibrillar regions; ( 4 ) tie molecules in the interlamellar region; ( 5 ) free
chain ends; ( 6 ) amorphous segments with large free volume and which may give rise to voids; ( 7 )
fusion of adjacent microfibrih6 Shaded areas represent the interfibrillar amorphous regions.

scattering essentially remains unchanged during heat setting, and we tentatively

attribute this scattering to aggregates of fibrils. Such aggregates have in fact
been observed in transmission electron micrographs.6
Figure 6 shows transmission electron micrographs of the non-heat-set fibers
and the fibers heat-set by Suessen [Figs. 6 ( a ) and ( b )3 , Superba [Figs. 6 ( c )
and ( d ) 1, and the autoclave [Figs. 6 ( e ) and ( f ) ] processes. The micrographs
show AgI stains formed during the I2/methanol/AgNO3 treatment. A thin
surface layer (0.5 pm) on the fibers remains unstained due to the presence of
a dense, oriented skin on the surface of the fiber. The core of the fiber also
remains unstained, and this is most likely because the stain cannot diffuse
through the already stained outer layers in the cross section of the fiber. The
stained regions in the cross sections of the fibers show AgI crystallizing in the
form of dendrites. We estimate the size of AgI crystals to be ca. 100 A (30-150
A, Fig. 7 ) , and this is in agreement with the value calculated from the line
widths in the wide-angle X-ray diffraction (WAXD) scans of the AgI stained
fibers. While the degree of staining is lower in Suessen set fibers than in the
corresponding control, the autoclave fibers stain more heavily than the corre-
sponding controls. The staining characteristics of the Superba set fibers were
somewhat similar to that of the autoclave fibers, except that the staining in
256 MURTHY, REIMSCHUESSEL, AND KRAMER

Fig. 6. Transmission electron micrographs of cross sections of AgI stained fibers. Control
fibers are shown in the left panel and heat set fibers are shown in the right panel (a, b ) Suessen
process; (c, d ) Superba process; (e, f ) autoclave process.

the Superba set fibers was not consistently heavier than in the corresponding
controls. Analysis of 15 sections of control fibers and 16 sections of Superba-
set fibers showed that the fractions of fully stained, partially stained, and un-
 CHANGES IN VOID CONTENT 257

Fig. 7. Transmission electron micrographs of a cross-section of a heat set fiber (autoclave

process) after AgI staining and a t a magnification higher than in Figure 6 ( f ) . In this photograph,
smallest dots represent - -
30 8, size AgI crystals, slightly larger dots are 150 8, AgI crystals, big
dark clusters (arrow a t the top left-hand corner) are aggregates of AgI crystals, and the small
unstained circular areas (arrow a t the lower-half of the picture) are aggregates of fibrils.

stained sections were, respectively, 25, 35, and 40% in control fibers, and 70,
20, and 10% in heat-set fibers. These results show that, in general, the degree
of staining is higher in Superba- and autoclave-set fibers relative to non-heat-
set and Suessen-set fibers. Phenol-treated fibers showed a similar increase in
AgI stains relative to the untreated fibers.
258 MURTHY, REIMSCHUESSEL, AND KRAMER

DISCUSSION

We will attempt here to interpret the changes in SAXS curves and the dif-
ferent degree of AgI staining in terms of a model which can be used in under-
standing diffusion and mechanical behavior of fibers. Two models are commonly
used, the pore and the free volume models.' The pore model describes amorphous
regions as a two-phase system made of polymer chains and the voids. The free
volume model describes amorphous regions as a single phase system in which
the segmental mobility of the chains, rather than voids, influences the dye
diffusion. Although diffusion is sometimes interpreted in terms of either a pore
model or a free volume model, it is likely that the distinction is in part due to
the probe used for measurement, and that diffusion, depending upon whether
it occurs above or below the glass transition temperature, is influenced by free
volume as well as porosity. SAXS, being sensitive only to average electron
density over distances of nanometers, cannot easily distinguish between these
two models. The average electron density of the amorphous regions can decrease
both due to pores and free volume, and we will use the two terms interchange-
ably. In the literature, this SAXS scattering along the equator is referred to as
void scattering.

Effect of Pores and Free Volume on Diffuse Scattering

S t a t t ~ nand
~ ' ~Hermans et al.' have done extensive work on the changes in
the SAXS diffuse scattering that occur during processing of polymer fibers,
and attribute the scattering in the equatorial plane to voids. In contrast, in
agreement with the work of Heyn, lo our data suggest that the equatorial small-
angle scattering at h 2 0.1 &' is due to fibrils. This distinction may not be
important, except when determining the sizes, since the scattering according
to either interpretation is essentially due to the contrast between the fibrils
and the amorphous chain segments laterally separating the fibrils. We suggest
that although the intensity of the scattered X-ray can be influenced by the void
content, the distribution of this intensity as a function of scattering angle, i.e.,
the shape of the scattering curve, is determined by the size of the fibrils.
The size of the fibrils, determined by the Guinier analysis of SAXS data,
increases from -50 to -65 A upon annealing in both dry and wet-heat-set
fibers (Table 11).We have seen earlier' that the crystallite size, as determined
from wide-angle X-ray equatorial reflections, also increases from 50 to 75 A
along the a-axis upon annealing (the increase along the c-axis could not be
reliably determined because of the overlapping peaks). This supports our con-
tention that the size determined from the diffuse equatorial scattering using
the Guinier analysis refers to the diameter of the fibrils. The size we have
calculated for the fibrils is perhaps an average of the crystallite sizes in the
equatorial plane.
It has been shown that the increase in the density of the crystalline regions
is about the same in the wet- and the dry-heat-set fibers.' Therefore, the increase
in the contrast and the accompanying higher small-angle intensity in wet-heat-
set fibers should be due to a decrease in the density in the interfibrillar amor-
phous regions. These low density regions are most likely to be regions in which
 CHANGES IN VOID CONTENT 259

a significant fraction of the volume is not occupied by polymer chain segments.

In a non-heat-set fiber, these regions may be occupied by gases entrapped during
spinning. Furthermore, as the moisture introduced during the wet-heat-setting
process leaves the fiber, it can leave behind voids and thus increase the free
volume available for mobile chain segments. Our results and analyses are con-
sistent with both pore and free volume models, as both pores and enhanced
segmental mobility lower the average amorphous density and thus give rise to
excess SAXS intensity.
The AgI stains in the electron micrographs show enhancement of any defects
already present in the fibers. It is known that iodine forms chemical complexes
with nylon 6.” Although iodine could change the morphology of the polymer,
the diffusion path of iodine could still be regarded as regions most accessible
to iodine and methanol, and hence least dense. Unlike I/KI treatment, the 121
methanol/AgNO, treatment does not produce the gamma form of nylon 6, and
the fiber remains in the alpha form. However, as AgI crystallizes along the path
opened up by the diffusion of iodine and methanol, it can bring about large
structural changes in the morphology of nylon 6. Thus, AgI dendrites indicate
regions which are least dense, and which are most susceptible to crack propa-
gation. A higher concentration of AgI stained areas in wet-heat-set fibers reflects
a more “open” structure than in dry-heat-set fibers. The lower limit of 30 A
for the size of the AgI crystals measured in electron micrographs (Fig. 7 ) , is
the same as that derived from SAXS and WAXD data (Fig. 5). This is consistent
with the proposition that iodine in methanol, as well as dye molecules, diffuse
through the interfibrillar regions. An increase in the porosity or free volume in
these regions contributes to higher rates of diffusion.
Gas adsorption measurements showed that the surface area increased only
in autoclaved fibers (0.41-0.60 m 2 / g ) , and was essentially the same in Suessen
and Superba fibers (0.43 and 0.30 m2/g, respectively). Since these values rep-
resent the contribution to the surface area from only those pores which are
open to the surface, the higher surface area found in autoclaved fibers suggests
an increase in latent cracks, defects, or voids near the surface of the filament.
This is consistent with the considerable increase in AgI stains in autoclaved
fibers. The more open structure in autoclaved fibers is seen in SAXS as lower
average density of the amorphous material in the lateral space between the
fibrils as compared with that in non-heat-set and dry-heat-set fibers.

Influence of Heat Setting on Porosity and Free Volume

The increase in small-angle diffuse scattering is observed only in wet-heat-
set fibers and not in fibers heat-set in dry atmosphere. Statton’ reported that
in dry and dispersion spun fibers, dry, high temperature heat treatment reduced
the amount of diffuse scattering. Wet processing of these fibers, on the other
hand, did not show any collapse of the voids. Statton also reported that the
effect of after-treatments was negligible in melt-spun fibers. Our data from
Suessen-set fibers, which show little change in the diffuse scattering, is con-
sistent with Statton’s conclusion. However, in contrast with the findings of
Statton, we show that after-treatments, such as wet heat setting, can have a
significant effect on the SAXS curves even in melt-spun fibers. More impor-
260 MURTHY, REIMSCHUESSEL, AND KRAMER

tantly, our SAXS data show that wet processing of melt-spun fibers leads to
increased electron density contrast, and higher SAXS diffuse scattering. We
suggest that lower amorphous density resulting from increases in voids and
free volume contribute to the higher electron density contrasts.
Both the Superba and the autoclave processes use saturated steam, and are
usually referred to as wet processes; in contrast, the Suessen process is a dry
heat setting process (Table I ) . The crystalline regions become more ordered
after heat setting in both the wet and the dry methods, but the volume fraction
of pores increases only in the wet-heat-set fibers. The marginal decrease in the
volume fraction of the pores in dry heat set fibers is likely due to the partial
melting and recrystallization of nylon 6 at the high heat setting temperatures.
It is possible that in wet-heat-set fibers the moisture diffusing into the fibers
occupies a finite volume, thereby swelling the fiber, and when the moisture
leaves the fiber upon drying, more free volume becomes available for the mobile
chain segments, i.e., the moisture leaves behind microvoids or pores in the
amorphous matrix by a process similar to the action of a foaming agent. Partial
hydrolysis of the nylon 6 chains in the presence of moisture could also contribute
to the increase in the void content. It is quite possible that the pores enhance
the mobility of the amorphous chain segments. The pores are probably occupied
by neighboring amorphous chain segments for only a fraction of the time during
the random thermal fluctuations from the equilibrium positions of these chain
segments. Therefore, the voids could decrease the density in the interfibrillar
region, and thus enhancing the SAXS intensity due to fibrils.

Influence of Heat Setting on Diffusion and Mechanical Properties

Some of the physical properties of a fiber can be understood in terms of its
morphology, i.e., by the distribution of crystalline fibrils, the amorphous regions,
and the void content. At low levels of crystallinity, the crystalline regions can
undergo large changes (such as increase in crystallinity), and thus influence
the properties of the fiber. As the crystallinity approaches a limiting value,
other factors such as density and orientation of the amorphous domains, and
defects such as voids, significantly affect the properties of the fiber.' Although
these various factors, as well as other parameters such as glass transition tem-
perature, crystalline perfection, crystallite size, and amorphous and crystalline
orientation, can directly or indirectly affect the physical properties of a fiber,
we here focus only on the role of the voids and free volume.
Dye diffusion decreases in dry-heat-set fibers probably not so much by the
increase in the crystalline perfection as it is by the decrease in the void content.
The observed increase in the volume fraction of the voids in Superba and au-
toclove fibers might directly contribute to an increase in the dye diffusion.
Thus, wet processing yields fibers which dye darker and faster. Our data suggest
that the increase in the dye uptake in fibers treated with aqueous solutions of
benzyl alcohol, phenol, and formic acid is due to an increase in the void content
as proposed by Subramanian et a1.12 In contrast, Moore and Weigmann13 at-
tribute increased dye uptake to greater segmental mobility. As discussed earlier,
the presence of pores might increase the free volume and enhance the segmental
mobility of the chains in the amorphous regions.
 CHANGES IN VOID CONTENT 261

The changes in the void content can be used to interpret some of the data
on ozone fading in nylon fibers. Ozone fading is a phenomenon in which, a t
high humidities, atmospheric ozone appears to oxidize the dyes, thus fading
the fibers. Haylock and Rusch14 noted that dyes which have the least mobility
in the fiber are most resistant to ozone fading: that acid dyes (which react with
the amine end-groups) desorb from nylon 6 at a much slower rate, and are
more ozone resistant than disperse dyes; and that premetallized dyes, which
do not desorb significantly into water a t 35"C, offer the most resistance to
ozone fading. Although such differences in resistance to ozone fading might
arise in part from the inherent differences in the resistance of the dyes to
degradation by ozone, to explain the generally observed correlation between
dye diffusion and the fading of the dyes in humid atmosphere in the presence
of ozone, Haylock and Rusch suggested that dye molecules are destroyed a t the
surface of the fibers once they diffuse out from the interior of the fiber. But
microspectrophotometric studies of dye diffusion by Kamath, Ruetsch, and
Weigmann l5 suggest that ozone penetration into the fiber substantially con-
tributes to the fading process. One cannot, however, completely discount the
hypothesis of Haylock and Rusch, since it is possible that the reactivity of
ozone may decrease rapidly as it diffuses into the fiber. Whatever may be the
diffusing entities, dyestuff or ozone, fading is higher in humid atmospheres and
in wet-heat-set fibers. In view of our data on the void content, the increase in
the rate of fading of dyes due to atmospheric ozone in a humid atmosphere,
and the independently observed increased rate of dye diffusion in steam-heat-
set fibers, can now be attributed to the increase in free volume or the volume
fraction of the voids which offer least resistance to the diffusion of the dyes
and ozone.
The observed differences in the free volume or the void content can also be
used to understand the differences in the mechanical properties of the fibers
after dry and steam heat setting.16 It was noted that, although the crystallinity
and the density were higher in steam-heat-set fibers, the increase in the Young's
modulus in steam-heat-set fibers was not as large as in dry-heat-set fibers. It
is possible to explain these apparently incompatible results by invoking the
role of the increased void content in the steam-heat-set fibers, and by the in-
crease in the number of latent cracks, defects or voids near the surface of the
filament. These changes can also account for the decrease in the tensile strength
of the fibers in steam-heat-set fibers, and a corresponding increase in dry-heat-
set fibers, relative to non-heat-set fibers.

CONCLUSION

Heat-setting (annealing) nylon 6 fibers in a moist atmosphere results in

fibers with higher void content or free volume. This contributes to increase in
the dye uptake and decrease in the tensile strength of the fibers.

We thank Dr. A. J. Signorelli for valuable suggestions throughout this work, Mr. J. L. Rush
for discussions on heat setting and dyeing, and for providing some of the fibers, Drs. K. Zero and
M. E. McDonnell for porosity measurements and for reviewing the manuscript, Drs. W. Hammond
and R. A. F. Moore for their valuable comments, and Professor S. Krimm for frequent discussions.
262 MURTHY, REIMSCHUESSEL, AND KRAMER

References
1. N. S. Murthy, H. Minor, and R. A. Latif, J . Macromol. Sci.Phys., B 2 6 , 427-446 (1987).
2. R. M. Rohner and H. Zollinger, Text. Res. J.,56, 1-13 (1986).
3. N. S. Murthy, Norelco Rep., 30,35-39 (1983).
4. A. Guinier and G. Fournet, Small-Angle X-Ray Scattering, Wiley, New York, 1955 pp. 70,
187.
5. A. Guinier and G. Fournet, Small-Angle X-Ray Scattering, Wiley, New York, 1955, p. 134.
6. A. C. Reimschuessel and D. C. Prevorsek, J. Polym. Sci. Polym. Phys. Ed., 14, 485-498
(1976).
7. W. 0. Statton, J. Polym. Sci., 22, 387-397 (1956).
8. W. 0. Statton, J. Polym. Sci., 56, 205-220 (1962).
9. P. H. Hermans, D. Heikens, and A. Weidinger, J. Polym. Sci., 35, 145-165 (1959).
10. A. N. J. Heyn, Text. Res. J., 23,782-787 (1953).
11. N. S. Murthy, Macromolecules, 20, 309-316 (1987).
12. D. R. Subramanian, A. Venkatataman, and N. V. Bhat, J. Appl. Polym. Sci., 27, 4149-
4159 (1982).
13. R. A. F. Moore and H.-D. Weigmann, Text. Res. J.,56, 180-190 (1986).
14. J. C. Haylock and J. L. Rush, Tent. Res. J., 46, 1-8 (1976).
15. Y. K. Kamath, S. B. Ruetsch, and H.-D. Weigmann, Text. Res. J., 53,391-402 (1983).
16. M. Tsuruta and A. Koshimo, J.Appl. Polym. Sci., 9, 3-9 ( 1965).

Received December 27, 1988

Accepted March 20,1989