S95-7 Page 1

Project Title: Fundamentals of Dye Diffusion Variance in Polyamide and Polyester

Fibers

Project Number: S95-7

PI(s): R McGregor (Leader), H.A. Davis, K. Hamada,A. Tonelli, R. Aspland (Clemson),

Graduate Students: R. E. Kyles, Y. Song, H. Xue (Clemson),
Other contributors: Mohan Srinivasarao (NCSU), Paul Russo (Louisiana State University)

Goals:

1. To create an understanding of the factors that control the temperature dependence of dye
diffusion into synthetic fibers in sufficient detail to provide leads for developing dyeing
methods and chemicals to completely eliminate dye streaks.

2. To develop chemical or physico-chemical approaches to eliminating dye variability in

continuous filament synthetic yarns.

The first year of this project will be spent 1) improving our experimental capability for
observing the properties we need to measure, 2) synthesizing appropriate molecular probes and
potential chemical modifiers, and 3) implementing a state of the art theoretical and computational
background. The second and third years of this project will use the technology developed during
year 1, as described in the “Technical Approach” section.

Abstract:

We are in the early stages of reviewing the current literature, and assembling and testing
new equipment and methods for dye diffusion, dynamic mechanical, molecular modeling, and
molecular probe studies. We are also strengthening our links with both industry and university
collaborators. The literature reviews have revealed important and useful advances in ESR and
NMR probe method, as well as in optical approaches to the study of diffusion . Our contacts with
other potential collaborators have been very encouraging.
Twelve novel 19F fluorinated molecular probes have already been synthesized, as have
three new 13C labeled acid azo dyes. Selected commercial dyes are being purified, and the
synthesis of other types of molecular probes is in progress.
The polymer physics literature, some of which stems from current members of our group,
has revealed new ideas about molecular motions in fiber-forming polymers which will be very
helpful in our work. We are now in a position to carry out both optical and molecular simulation
studies of polymer-penetrant interactions.

National Textile Center Annual Report: November 1996 213

 S95-7 Page 2

Relevance to NTC Goals:

Streaks, and in particular dye streaks, have been a nagging and expensive problem in the
textile industry for 50 years. Sufficient information is now available that foreign competitors
could produce yarns which give no streaks, albeit with increased ( but low cost) labor. The most
desirable approach to eliminating streaks would be a chemical one, but none is known. By
understanding the fundamental causes of dye variability, and developing chemical routes to
eliminating it, this project will help maintain a level playing field for USA industries in the
critical-dyed apparel and carpet markets.

Technical Approach:

Past work (mostly ours) and currently available physical chemical approaches to reducing
dye variability tend to point in the same direction. This seems to support the general validity of
the approaches but emphasizes the need for further confirmatory experimental data.

Fibers have a crystalline fraction which is hardly influenced by temperature or dyebath

chemicals, and a noncrystalline fraction (A), which is strongly influenced by these factors. With
increasing temperature, an increasing part of region A develops enough liquidlike mobility to
permit dye diffusion. We call this process mobilization. The main difference between two like
yarns which dye differently is the temperature dependence of mobilization, which is not well
understood. One way to eliminate streakiness would be to mobilize 100% of region A in all the
competing yarns before permitting any significant dye uptake. Physical-chemical methods to do
this have been reported, e.g. Bayer’s Telon SCR) process, and DuPont’s Infinity (R) dyeing process.

Initial streakiness arising from different initial dye sorption rates will be made worse by
the concentration dependence of the diffusion constant. In addition, dyes with high distribution
coefficients will be unable to migrate from yarn to yarn in order to reduce these initial
inequalities in dye uptake. The concentration dependence of the diffusion coefficient in ionic
dyeing systems is not fully understood, but when the dye ion is a counter-ion, this concentration
dependence can be reduced by ionic pre-treatments of the fiber.

We plan to improve the understanding of this mobilization process by combining

computer modeling of diffusion in the noncrystalline regions, molecular probe diffusion studies,
precise dye diffusion and dynamic mechanical measurements, and their comparison with the best
available theories, including the newer theories of thermoreversible gels. Based on the extension
and improvement of the understanding thus generated, we will develop novel chemical
approaches to control the polymer chain mobilization, and the rate of dyeing. This preferred
approach differs from past methods that have focussed on preventing dye from reaching the fiber
until region A is 100% mobilized.

214 National Textile Center Annual Report: November 1996

 S95-7 Page 3

Industry Outreach:

We have spent considerable effort organizing the industry connections for this project. We
examined our status and capabilities and concluded that the main functions of collaboration with
an industrial partner are to provide the project with 1) special nylon fiber samples, 2) assistance
in the design and execution of specialized nuclear magnetic resonance (NMR) testing, 3) a place
to test the ideas which come from our work and 4) a mechanism to make sure that the course
we are following is going somewhere useful.

We have discussed our project with several potential industrial partners and selected one
as the most likely to be involved initially, based on our understanding of their needs. In April
we presented a seminar covering our most recent dye diffusion work at their main research
facility and discussed future collaboration. This company wants to be involved, and will provide
fiber, contacts and NMR experiments. However, the appropriate NMR expertise appears not to
be readily available there. We are working with two of their R&D people and expect a fruitful
collaboration, but we will need to get the NMR expertise elsewhere. We are now exploring
collaborative NMR and ESR work with the Chemistry Departments at NCSU and UNC-Chapel
Hill..

Current Progress:

Experimental resources:

We have assembled a dyeing system of the type used by Grieder ( Grieaer, K.,
J.S.D.C., 92, g-13(1976)) for direct measurements of the dye diffusion coefficient, based on the
rate of dye sorption by small fiber samples. This required parallel sorption equilibrium studies
using the thermostated shaker-bath procedure described by Arora and McGregor (Textile
Research Journal, Accepted for publication). Ye Song’s measurements by this technique have
revealed a very interesting dependence of the activation energy for diffusion of CI Acid Blue 80
in nylon on the pH of dyeing.
We have begun to assemble the components for a dyebath monitoring system based on
the latest technology developed by the related NTC Project S95-4. This system can be used for
continuous monitoring and feedback control of dyeing experiments (McGregor, Arora & Jasper,
submitted for publication ). In addition, spectrophotometric microdensitometry of fiber cross
sections has been carried out in the laboratories of a fiber manufacturer.

Study of the Diffusion of CI Acid Blue 80 in Nylon 66

In our study we measured the diffusion coefficient as described by Grieder, in an infinite

dyebath, but at different pH values and temperatures. The activation energy was determined at
different pH values from the temperature dependence of the diffusion coefficient. We also found
a linear relationship between the diffusion coefficient and the dyebath pH under our
experimental conditions. The Nylon 66 yarn was obtained from Monsanto Chemical Company,
and had been heat set by autoclaving. The amine end group content was 0.042 meq/kg of fiber.

National Textile Center Annual Report: November 1996 215

 S95-7 Page 4

The finish was removed from the yarn before dyeing by scouring the fiber with a lg/l solution
of sodium carbonate and T&on-X-100@) at 45 “C for 30 minutes. The dye was purified by
reprecipitating the dye three times from 2N hydrochloric acid and 10% acetic acid solution. The
empirical formula of the purified dye was C32H2&08S2. 2.6H,O, and the Formula Weight was
681.5. The following data were obtained:

Table 1. Equilibrium concentrations for dyeings at different pH and temperatures

Temperature PH4 PH5 PH6

(“Cl
(C/kg) (%) (g/kg) (%I (g/kg) (%)
75 15.8 110.4 18.3 127.9 13.81 96.5
85 17.64 123.3 18.32 128 15.15 105.9
95 18.4 128.6 17.24 120.5 16.17 113

Table 2. Diffusion coefficient and activation energy under different dyeing conditions

Temperature
Diffusion Coefficient (cm’/min)
(“C)
PH4 PH5 PH6
75 I 1.90E-9 I 1.47E-9 I l.l6E-9
85 4.40E-9 3.18E-9 2.51E-9
95 l.l2E-8 6.76E-9 4.15E-9

Ea(kcal/mole) 22.43 19.35 16.16

The diffusion coefficients were the statistical averages from 8-10 experimental values.
They are larger at higher temperature and lower pH. The activation energy of diffusion
increases when pH decreases, and our experimental results show that there is also a linear
relationship between activation energy and pH. Since the activation energy is a measure of the
energetic barriers to diffusion, these data suggest that the energetic barriers are higher at lower
pH values, which raises interesting mechanistic questions.

216 National Textile Center Annual Report: November 1996

 S95-7 Page 5

UV-Visible Spectroscopic Study of the Spectral Characteristics of CI Acid Blue 80

A UV-Vis spectroscopic study was undertaken to determine the effects of ionic bonding
on the spectral characteristics of purified CI Acid Blue 80. The first step in this experiment was
to determine the spectral behavior in aqueous solutions. Quantitative solutions of purified CI
Acid Blue 80 were prepared using sodium acetate/acetic acid buffer solution, pH=4.0. This pH
was selected to model typical acid dyeing conditions for nylon materials. Three electronic
transitions are observed in the visible region, at wavelengths of 542, 582, and 626 nm. The
transition of maximum absorbance is 626 nm and has an absorptivity of 3 1.9 Ug-cm (correlation
coefficient of 0.9999).

The next step of this experiment was to measure and compare the spectral behavior of
purified CI Acid Blue 80 when combined with n-propylamine or n-methylacetamide in pH=4.0
buffered aqueous solutions. These compounds were chosen to model the amine end groups and
the amide linkages of nylon polymers, while maintaining aqueous miscibility. No change in the
absorbance profile was observed for the amine or the amide solutions.

The final step of this experiment was the dyeing and measurement of the spectral
characteristics of nylon 6,6 films dyed with purified CI Acid Blue 80. The films were prepared
from 14%(w/v) solutions of nylon 6,6 in 9597% formic acid. The viscous solutions were cast
onto glass plates and the solvent was allowed to evaporate under ambient conditions. The
resulting thin films were washed in distilled water and thoroughly dried. The films were then
dyed to a 1% (0.w.f.) shade depth with purified CI Acid Blue 80 from aqueous solution, buffered
to pH=4.0. The absorbance profile was then measured using transmission UV-Vis spectroscopy.
As seen in aqueous solution, three electronic transitions occur in the dyed film. However, a
bathochromic shift of 5-6 nm is observed for all three transitions. Also, in one of our dyed films,
there was a change in the transition of maximum absorbance. The transition at 587 nm absorbed
slightly more energy than the transition at 633 nm. The origin of this behavior is still unknown
and merits further investigation. Modifications of the physical properties of the films, such as
percent crystallinity, will be made to determine the effect on the spectral behavior of the dye.

Probe synthesis and application:

Kunihiro Hamada synthesized a total of 12 fluorine (19F) labeled probes, with which
he will can-y out further work at Shinshu University in Japan, and we also will use these dyes
at NCSU for NMR studies. Richard Aspland and Huachan Xue prepared l3 C labeled probes for
NMR studies, and Richard Aspland will also synthesize stable free-radical probes for ESR
studies, and fluorescent probes for fluorescence recovery after photobleaching studies (FRAP),
to provide information about molecular probe mobilities in polyamide fibers.

Richard Aspland and Huachan Xue synthesized the following seven azo dyes using
anthranilic acid as the diazo component: 1-(2-carboxyphenylazo)-2-hydroxynaphthalene;
1-(2-carboxyphenylazo)-2-hydroxy-8-acetylaminonaphthalene; sodium- 1-(2-carboxyphenylazo)-2-
hydroxynaphthalene-6-sulfonate; sodium-l-(2-carboxyphenylazo)-2-hydroxynaphthalene-8-

National Textile Center Annual Report: November 1996 217

 S95-7 Page 6

sulfonate; sodium-2-(2-carboxyphenylazo)- l-hydroxynaphthalene-3-sulfonate; sodium-2-(2-

carboxyphenylazo)- 1-hydroxy-6-aminophenylnaphthalene-3-sulfonate a n d sodium-2-(2-
carboxyphenylazo)-l-hydroxynaphthalene-5-sulfonate. The following dye:metal complexes were
then synthesized from all seven of these dyes: 2:l chromium, cobalt, and iron complexes, and
1:lcopper complexes. When the parent dyes are included, 35 different acid dyes were
synthesized from the anthranilic acid diazo component.

The above dyes were dyed at 0.2% and 0.5% owg under conventional neutral and acidic
dyeing conditions, on a Testfabrics nylon fabric with a known amino end group content of 40
meq/kg. The dyeings were then examined for their coverage of Barre . None of the dyeings
showed particularly poor barre coverage.

The parent dye, sodium-2-(2-carboxyphenylazo)-l-hydroxynaphthalene-5-sulfonate and

its 1: 1 copper and 2: 1 chromium complexes were selected for further study on the basis of their
shades. This included the synthesis of these theree products from anthranilic acid labeled with
carbon 13 at the carbonyl carbon. As a result of this work and other dyeing studies, Huachun Xue
was awarded a Masters Degree in Textile Chemistry from Clemson university in August 1996.
His Thesis is entitled “Synthesis and Application to Nylon of Selected Acid Dyes”.

P r o f e s s o r K u n i h i r o Hamada’s 19F fluorinated acid dyes include sodium

1-(2-trifluoromethylphenylazo)-2-hydroxy-6-naphthalenesulfonate, sodium
1-(3-trifluoromethylphenylazo)-2-hydroxy-6-naphthalenesulfonate, and sodium
1-(4-trifluoromethylphenylazo)-2-hydroxy-6-naph~~enesulfonate. They were synthesized by the
c o u p l i n g o f the corresponding diazotized a n i l i n e d e r i v a t i v e s w i t h s o d i u m
2-hydroxy-6_naphthalenesulfonate, p u r i f i e d , a n d d r i e d . Sodium
l-phenylazo-2-hydroxy-6-naphthalenesulfonate (Crocein Orange G, Aldrich) is being purified.
In addition, the corresponding acid dyes containing a methyl group will be synthesized.

NMR Dye Studies:

We have observed a red acid dye synthsized by Richard Aspland and his student, which
has a C-13 enriched carboxyl carbon, by means of high resolution, solid-state C- 13 NMR. This
was achieved by combining the techniques of cross-polarization (CP), rapid sample spinning at
the magic angle (MAS), and high power dipolar decoupling of H-l spins (DD). A mixture of
l-3 wt. % of dye in the C- 13 NMR transparent boron nitride powder gave a CPMAS/DD C-13
NMR spectrum with a strong carboxyl carbon signal from the dye after only a few scans.
However, when nylon fibers were dyed with this C-13 enriched dye, we observed the carboxyl
carbon from the dye and the amide carbonyl carbons of the nylon to resonate at identical
frequencies. This unfortunate overlap could not be resolved by techniques that discriminate
between differences in the mobilities of the dye carboxyl and nylon carbonyl carbons, because
a CP contact time (TCH) study revealed that both carbon nuclei CP with H-l nuclei with equal
efficiencies. It is hoped that Dick Aspland can slightly modify the structure of this red acid dye,
possibly by attaching a substituent on the phenyl ring ortho to the C-13 enriched carboxyl carbon.
This should shift the carboxyl carbon resonance on the order of 5 ppm upfield from the

218 National Textile Center Annual Report: November 1996

 S95-7 Page 7

resonance frequency observed for the unsubstituted dye and relieve the overlap of dye carboxyl
and nylon carbonyl resonances, thereby permitting separate observation of dye and nylon
mobilities by C-13 NMR. We plan to place nylon fibers dyed with the substituted dye into a
NMR tube filled with water and heated to ca. 95’ C in hopes of generating sufficient mobility
of the dye and the amorphous, swollen regions of the nylon to permit C- 13 NMR observations
by the usual solution techniques, i.e. without CPMAZYDD. If this proves feasible, then pulsed
field gradient C-13 NMR can be used to determine the diffusion of dye in the swollen, dyed
nylon fibers.
We are now beginning to attempt similar NMR studies with the -CF, (19F) labeled dyes
synthesized by Prof. Kunihiro Hamada during the first year of this NTC project.

Molecular Modeling

We have purchased and installed an SGl High Impact work station with the latest high-
speed RlOfKKl processor chip and have loaded the Biosym-MS1 suite of polymer modeling
software. Our graduate students are currently able to polymerize monomers, construct dye
molecules, create amorphous cells of polymers containing small numbers of dye and water
molecules, and soon we will attempt to conduct some molecular dynamic simulations on these
complex systems. we hope this will lead to new insights into dye diffusion in polymers.

Dye-Inclusion Complexes

We are currently attempting to form crystalline inclusion compounds (ICs) between dyes
and cyclodextrins (CD, beta and/or gamma). The purposes of creating dye-CD-ICs is to permit
their embedding into nylon fibers and films. CD-ICs formed with other small molecules and with
polymers generally have high melting points (usually above 270 C). As a consequence, melt
processing of nylon-6 (Tm + 225 C) films and fibers containing dye-CD-ICs should be possible
at T’s<260 C. Once the dye-CD-IC crystals are embedded into the nylon-6 matrix, we anticipate
that exposure to hot water will swell the nylon-6 matrix, disrupt the dye-CD-IC crystals, and
permit the preferential removal of the dissolved CD from the nylon-6 matrix, leaving behind the
dye which strongly interacts with the amorphous portions of the nylon-6. By this means we hope
to achieve the internal dyeing of nylon-6 and to study directly the in situ diffusion of dye in
the fiber without any of the complications attendant to dye diffusion studies performed on undyed
fibers placed in dye baths. We have some experience with the formation of polymer-CD-ICs
which we expect to utilize by forming CD-ICs with polymers or oligomers containing chemically
bound dyes.. Embedding of thes CD-ICs into difficult to dye polymers like PET might facilitate
their dyeing if we can disrupt the CD-ICs and preferentially remove CD leaving behind the dye
chemically bound to the included polymer or oligomer in the PET matrix.

New methods for studying dye diffusion

In this report we also introduce new methods that are capable of providing a wealth of
data that will prove useful in understanding the mechanism of dye diffusion, and have the

National Textile Center Annual Report: November 1996 219

 S957 Page 8

potential of providing that data in real-time as opposed to other methods which are time
consuming and perhaps tedious. We introduce two such methods, one based on “Fluorescence
Recovery after photobleaching” (FRAP), and one based on the “Evanescent Waves” created by
total internal reflection within the fiber being dyed. This can be termed “Evanescent Wave
Spectroscopy (EWS)“.

In the first of these methods the essential idea is to impart a known spatial pattern of dye
concentration to a fiber which is being dyed, i.e., we bleach the dye locally, thus imprinting
a pattern of varying dye concentration. Due to diffusion of the dye the bleached pattern will
decay, in other words, the diffusion process will slowly eliminate the imprinted pattern. By
observing this “recovery” process in a microscope, and measuring its time dependence, we
can obtain an estimate of the diffusion constant. Such measurements can be made in both
directions, parallel and perpendicular to the fiber axis. Information about both the axial and the
radial diffusion constants are much more difficult and time consuming to obtain using the current
conventional methods of studying dye diffusion in fibers.

We have already made preliminary measurements of dye diffusion in fibers using FRAP
with Prof. Paul Russo at Lousiana State University, giving us a diffusion constant in the range
of lo-” cm2s--1 at 25 ‘C, consistent with the data obtained from conventional methods of
estimating dye diffusion. We plan to continue this work in a more systematic fashion and are
planning to construct a FRAP instrument at NC State that would enable us to do these
experiments.

Plans are also underway to build the instrumentation necessary to start work on EWS.
We plan to shine laser light down a nylon fiber, a portion of which is immersed in a dye bath.
At the other end of the nylon fiber a detector will record both the intensity and spectral
distribution of the emitted light. As the dye diffuses radially into the fiber both the concentration
gradient of dye and the refractive index gradient across the fiber will change and should be
detectable as the temporal behavior of the spectral distribution of the emitted light. This
approach may permit an in situ, time-of-flight observation of dye diffusion into a fiber from the
bath, and also yield details concerning the temporal distribution of dye across the fiber diameter.

Conventional determinations of the dye diffusion coefficient from the time dependence
of the macroscopic uptake of dye by the fiber, and from the time dependence of the dye bath
exhaustion, provide a necessary link to the behavior of the dyes in practical situations, but they
inevitably require specific assumptions concerning the instantaneous dye concentration at the fiber
surface, and about the shape of the dye concentration profiles across the fiber, in order to extract
a dye diffusion coefficient. By using the fiber instead as an optical fiber, and observing how its
light transmission characteristics change during the course of dyeing, we hope to be able to
make direct observations of the molecular details involved in fiber dyeing,

We plan to use the fluorinated dyes in order to investigate their mobility in nylon fiber
by using NMR spectroscopy. Also classical diffusion measurements for the dyes will be carried
out to elucidate the effects of the position and kind of the substiuent. High-resolution solid-state

220 National Textile Center Annual Report: November 1996

 S95-7 Page 9

19F-NMR spectroscopy of the sorption and diffusion of fluorine-containing aromatic molecules

in polymeric media was reported recently (P. J. Toscano et al., J. Polym. Sci.: Part A: Polym.
Chem., 31, 859 (1993)). H NMR spectra-scopy and pulsed-gradient NMR techniques also seem
to hold promise as a means to probe the mobility of dyes in fibers and possibly also to monitor
their intrafiber molecular environments.

Searches have been made of the recent literature on the ESR spin probe technique. New
developments include the determination of translational diffusion coefficients using ESR imaging
(e.g. J. Freed, Ann. Rev. Biophys. Biomol. Struct., 23, 1 (1994)), and monitoring of the bimodal
distribution of the the ESR spectra of nitroxides to explore the penetration of solvents into
polymeric materials (Pekcan and Demir, J. Appl. Polym. Sci., 43, 2169 (1991) and Pekcan and
Demir, J. Appl. Polym. Sci., 49, 1877 (1993)). These papers indicate that modem ESR spin
probe techniques provide powerful methods for investigating not only rotational diffusion but
also translational diffusion in synrthetic polymers and natural membranes. Freed and his co-
workers have revealed interesting differences between the macroscopic and the microscopic
ESR-based diffusion coefficients for molecular probes in different media.

The motional freedom of the polymer chains in the accessible regions of a polyamide
fiber is known to be highly sensitive to the presence of water molecules. A preliminary study
of the molecular motions of D, 0 in dyed and undyed nylon-6,6 fibers has been carried out by
H NMR spectroscopy, via a collaboration with Dr. Edward Samulski and his graduate student
Julie Hutchison at the Chemistry Department of the University of North Carolina at Chapel Hill.
The results were not encouraging, in that significant differences were not observed.

Polymer physics:

We have also been reviewing literature in the areas of polymer physics which seem likely
to help us to understand the liquidlike/solidlike transition which appears to control dye diffusion
in fibers. In addition to the gel-like solidification process discussed in our original proposal,
another credible mechanism has been uncovered. This is the “orientational crystallization”
mechanism of Elyashevich, which is based on Flory’s liquid-crystal theory, rather than on
thermoreversible gelation. We are studying these theories in detail to determine if the two
mechanisms, which are based on fundamentally different assumptions, can be differentiated by
experiments.

Molecular modeling:

We intend to model the structures and motions of nylon and PET chain segments in the
non-crystalline regions of their fibers, where the dye molecules must reside. Hawthorne Davis’
work, in which he observed the dynamical mechanical properties (DMP) of fibers in aqueous
glycerol baths, showed that the activation energy of their DMP is very similar to the activation
energy of dye diffusion in the fiber, and that the non-crystalline fraction of the fiber consists of
a mobile, liquid-like phase, through which the dye could diffuse, together with a glassy, and
presumably inpenetrable, rigid phase. The relative ratio of the two non-crystalline environments

National Textile Center Annual Report: November 1996 221

 S957 Page 10

was found to be temperature dependent. Recent X-ray diffraction observations and analyses of
PET fibers by Wunderlich et al. also suggest the existence of two distinct non-crystalline phases,
one completely disordered and amorphous and the other partially ordered.

We have in prior unrelated work modeled the behavior of single nylon and PET chain
segments as they are confined to occupy cylinders with ever decreasing radii. Hopefully the
information generated there will serve as a starting point for modeling a small non-crystalline
portion of nylon and PET fibers containing at least 2 chain segments, with and without the
presence of a dye molecule. A molecular dynamics simulation, as well as a brute force search
of configurational space, may be utilized in the modeling.

222 National Textile Center Annual Report: November 1996