1

C95-4

Effect Of Molecular Structure On The Mechanical, Optical And

Electrical Properties Of High Technology Fibers
Richard V. Gregory, School of Textiles, Fibers and Polymer Science, Clemson University,
Clemson, South Carolina 29634-1307

and

Robert J. Samuels, School of Chemical Engineering, Georgia Institute of Technology,

Atlanta, Georgia 30332-0100

Abstract and Goals:

The primary focus of this research over the last several years has been to develop fibers and films
from electroactive polymers and to determine their structure/property relationships. We report here for
the first time the observance of a melt transition in the electroactive polymer polyaniline (PANI) in its
base form which may facilitate some degree of melt processability. Structural characterization of films and
fibers utilizing unique optical methods developed at Georgia Tech. has resulted in a new understanding of
the relationship between the optical properties and the polymer morphology necessary to the use of these
materials in the developing light emitting diode and photonics industries. Fiber spun from
leucoemeraldine base polyaniline (LEB/PANI) have tenacities of 5.2 g/d and after doping conductivities of
better than 150 S/cm while maintaining tenacities of 3.2g/d. The mechanical properties of the fiber are
more than adequate for a variety of commercial applications and are the highest reported to date in the
literature. In addition we report on the substantial effect of solvent types and characteristics on the
crystallinity of the formed polymer.
This work has resulted in the publication of over seventeen scientific papers, a filed patent
application, and many invited talks at national and international meetings dealing with fiber technology
and electroactive materials.

Introduction:

Over the last three years this project has primarily focused on the characterization of the
electrical and mechanical properties of electroactive polymers and of fibers formed from these materials.
The polymer of choice has been polyaniline (PANI) due to its facile processing, ease of switching, and
stability compared to other intrinsically organic electroactive polymers as previously demonstrated and
reported in the scientific literature by our laboratories and other workers in the field.[1] Fibers produced
from the reduced lecoemeraldine base (LEB) form of the polymer shown in Fig. 1(a), after draw and
doping, have significantly improved mechanical properties over those previously reported and
conductivities as high as 150 S/cm. [2] Structural characterization of films formed from the fiber forming
material by the Georgia Tech laboratories clearly demonstrate that the morphology of the formed material
is significantly dependent on the formation process and on the initial oxidation state, polymer solution
morphology, and solvent used in the fiber or film forming solutions. X-ray structure analysis of the
formed PANI powder prior to fiber or film formation also demonstrates the significant effect of solvent,
annealing , and reduction in oligomeric content on the final structure of the formed material.
Recently we have observed and reported for the first time a melting transition in PANI films and
fibers which may in future work open up possibilities for melt processing polyaniline.[3] Structural
characterizations of the materials will lead to quantification of the parameters necessary to producing
fibers and films with reproducible mechanical and electrical properties on a segment by segment basis
 2

whether they are melt, solution , or otherwise processed. Such reproducibility is not only required for
commercially viable products with well characterized mechanical and electrical properties but also for the
developing photonic device technology now being realized which extensively utilizes the unique
electronic, magnetic, and optical properties of electroactive polymers. Results obtained by the
cooperative efforts of the Clemson and Georgia Tech. laboratories has substantially advanced the
understanding of these materials and provided a much needed information base for the processing of
electroactive fibers and films. Over the last year the development of a potentially melt processable form of
PANI at Clemson and a deeper understanding of the structure and effect of differing formation
methodologies of the formed material developed at Georgia Tech. has resulted in numerous publications
in the scientific literature and several invited talks at national and international scientific meetings. This
report will detail recent work in fiber formation, melt phenomena, and structure property relationships
developed in unison at both the Clemson and Georgia Tech. laboratories.

Results and Discussion:

Continuous processing of the PANI polymer in the base form for extended periods of time has
been a problem reported by several different workers since solutions of solids in excess of 8% in N-
methyl-2-pyrrolidinone(NMP) and 16% in N,N'-dimethyl propylene urea (DMPU) are unstable due to
gelation. This precludes many applications which require formation of polymer from stable solutions at
high solids concentrations or where solution storage for some period of time is desired. This gelation
results from interactions between imine nitrogens and amine nitrogen protons on neighboring PANI
chains. Stable solutions are necessary for the production of fiber or film with segment to segment
reproducibility of mechanical and electrical properties. Such solutions cannot be obtained if the dissolved
polymer is undergoing a dynamic change such as network formation and subsequent gelation. This
dynamic change is realized in concentrations of PANI above 6% in NMP and 12% in DMPU as
increasing viscosity due to an elastic network formation. Potential uses of PANI on a large scale in fiber or
film form in electrochromic devices, rechargeable batteries, EMI shielding, LED's, etc. will depend on
formation of materials with predictable and reproducible morphologies on a segment to segment basis
resulting in reproducible polymer properties. Such reproducibility cannot be obtained if the polymer
solution is changing its viscous and elastic properties during polymer formation.
PANI can exist in three distinct oxidation states as shown in Figure 1. Of these, only the
emeraldine base (EB) form (Fig. 1b) can be doped by acids to form the electrically conductive emeraldine
salt. Both the EB and the reduced emeraldine base, called leucoemeraldine base(LEB), are soluble in
NMP and DMPU and fibers and films can be processed from these solutions with subsequent doping to
varying levels of conductivity. Since the inception of this project we have carried out and reported the
results of our studies on the rheology and viscoelasticity of fiber forming solutions of PANI/EB and
PANI/LEB. In addition we have studied the mechanical properties of fibers formed from these solutions
to better understand the effect of dynamic changes such as gelation on the mechanical and electrical
properties of the formed fiber or film. Work in the Clemson and Georgia Tech. laboratories and at I.B.M.
by Angelopoulos and coworkers has shown DMPU to be an excellent solvent for higher solids
concentrations of PANI in the EB form. Viscoelastic studies showed, however, that the solution becomes
more elastic with time, indicating an unacceptable degree of solution instability at higher solids
concentrations. This solution elasticity results from intra- or inter- chain secondary bond formation
between the imine nitrogens and the amine hydrogens on neighboring PANI chains . Reduction to the
LEB oxidation state by phenyl hydrazine or other suitable reducing agents produces a stable solution
where the increase in solution elasticity is substantially reduced. Interaction between amine moieties on
the PANI backbone is substantially weaker than between imine nitrogens and amine hydrogens and can be
easily broken up by solvents such as DMPU. We have also determined and reported that the electrical
properties of PANI fibers formed from either the insulative LEB or EB forms, and subsequently doped,
have conductivities of 140 to 350 S/cm or higher, and in the case of fibers formed from LEB solutions,
mechanical properties approaching those of some commercially available polyamides such as Nylon 6.
This initial work demonstrated that the electrical properties are dependent on the degree of orientation of
the fibers and films and on the stability of the forming solution. In an effort to characterize the fibers and
films formed from the more stable LEB/DMPU solutions, several studies to determine the morphologies
and the relationship between the polymer microstructures and observed electrical and mechanical
properties were conducted.
 3

During recent investigations of fibers and films formed from LEB/DMPU solutions we have
observed crystals which melt at elevated temperatures under a nitrogen atmosphere. The surface of this
material appears to flow at elevated temperatures, indicating a potential for melt processing of LEB PANI.
While several workers have studied the thermal transitions of polyaniline, most have
concentrated only on the emeraldine base form produced from NMP solutions.[4] Figure 2 contrasts the
DSC data of films formed by dip coatings of emeraldine base and leucoemeraldine base produced from
DMPU solutions. A broad exotherm at approximately 200 ºC was observed for the EB films, most
probably due to oxidation or crosslinking, consistent with results obtained by others [4,5]. The Tg of the
dip coated LEB films was found to be about 200 0C, and an endotherm was observed at approximately 385
0
C. Thermogravimetric analysis was carried out to confirm that the endotherm was not due to
degradation, since the films showed no significant weight loss up to a temperature of 450 ºC in nitrogen.
The specific enthalpy change associated with the endotherm observed for the LEB fiber was
measured and found to be ~36 J/g. Even larger endotherms were observed in films prepared by spin
coating LEB/DMPU solutions. Thermal analysis results for the spin coated film demonstrate an
endotherm at about 390 ºC, with a specific enthalpy change of ~70 J/g. The TGA for this film clearly
shows that the film is not decomposing at the observed endotherm temperature. The film is stable in
nitrogen to more than 450 ºC. There is some weight loss between 175 and 260 ºC associated with the
loss of residual solvent, confirmed by both TGA and DSC, compared to that of annealed film in which no
weight loss or thermal transitions were observed between 175 and 250 ºC.
Optical microscopy was used to determine the surface features of the films prior to the observance
of the endotherm in the DSC and after passing through the endotherm temperature. Prior to the
endotherm temperature the film surface was found to be somewhat rough. After passing through the
temperature where the endotherm occurs the film surface becomes much smoother, more reflective, and
begins to distort. The dramatic difference between the surface characteristics occurs at almost exactly the
same temperature as the endotherm observed from the DSC. After cooling, it was observed that a small
amount of polymer had flowed from between the microscope slides. The above evidence (DSC,TGA, and
microscopy) is consistent with a melting transition at approximately 390 ºC.
X-ray diffraction was employed as an independent method to confirm the existence of crystalline
structure in the LEB/DMPU spin coated films. Since spin coated films are thin (~3 microns), x-ray
diffraction patterns were obtained from stacked mats of both EB and LEB/DMPU films at Georgia Tech.
As can be seen from the x-ray patterns in Figure 3, the EB/DMPU spin coated films (3a) are amorphous
while the LEB/DMPU films (3b) are crystalline. In the case of the heated and then quenched LEB/DMPU
single film sample, although the x-ray data is somewhat inconclusive due to the thin nature of the film,
much less intense scattering is observed consistent with the melting of a crystalline phase. More detailed
x-ray analysis is currently under way.
Additional data indicating that the endotherm described above is, in fact, due to melting of
crystallites was obtained from dynamic mechanical thermal analysis of a bundle of fibers. Figure 4
presents the storage and loss moduli of this fiber bundle. This shows classic behavior of the moduli
through the glass transition, but an increasing storage modules from about 225 to 325 ºC, possibly due to
an increase in crystallinity. Above 350 ºC, both the storage and loss moduli abruptly decrease. Since the
sample was not broken at the conclusion of the measurement, the decrease in the dynamic moduli
indicates that the fibers experienced a dramatic softening consistent with melting.
In addition to the dynamic moduli, the change in sample length with temperature was obtained.
Below the glass transition, there is no significant shrinkage as expected for a glassy material. When
enough thermal energy was available for large scale segmental mobility (at the glass transition), an
increase in the rate of shrinkage was observed. Since the fibers were partially oriented (coagulation bath
draw ratio =1.5/1 ), shrinkage occurs in order to maximize the entropy. From the glass transition to
approximately 375 ºC, the rate of shrinkage remains nearly constant, but dramatically increased above
375 ºC, consistent with melting.
Experimental evidence obtained to date in our laboratory (x-ray, TGA, DSC, DMTA, and
microscopy) is consistent with the conclusion that the endotherm at approximately 385 to 395 ºC is due to
the melting of a crystalline phase in polyaniline films and fibers produced from solutions of
leucoemeraldine base in DMPU. A detailed discussion of this work has recently been accepted for
publication in Polymer Communications [3]. The results presented in this article are believed to be the
first report of melting in an electrically conductive homopolymer, and has significant implications for the
possibility of melt processable PANI.
 4

It should be recognized, however, that the PANI is in the reduced form and must be subsequently
re-oxidized and protonated to achieve electrical conduction. The two times post-drawn fibers produced
from the LEB/DMPU wet spinning process, oxidized with H2O2 and doped with HCl exhibit a
conductivity of 140 S/cm, somewhat lower than the 350 S/cm reported for the fibers produced from
EB/DMPU spin solutions. Optimization of the process is expected to result in improved electrical
properties surpassing those of the EB/DMPU fibers.
To fully explore the significance of the above observations and determine the optimum
fabrication conditions, studies are being conducted on PANI films fabricated by different processes and
from different solvents. Although PANI is opaque, the near infrared wave guide technique developed at
Georgia Tech [6,7,8] allows three dimensional refractive indices to be obtained from PANI films. This
information allows us to identify the molecular orientation and any crystallinity that develops during
fabrication and deformation. By combining the refractive indices with polarized infrared and x-ray
diffraction measurements further quantitative information about the molecular morphology can be
obtained.
As mentioned above the primary film fabrication procedures examined to date have been spin
coating and film casting with subsequent drawing, and the effect of casting solvent on the morphologies
produced. For detailed structure studies emphasis has concentrated on EB/PANI films, with fabrication
solvents NMP and DMPU. Figure 5 shows the refractive indices of both (a) cast and (b) spin coated
oriented EB films [8,9,10]. The spin coated films were prepared from DMPU solution at Clemson while
the cast and drawn films were prepared for us from NMP solution at the Los Alamos National Laboratory
by Drs. B. Mattes and H. Wang.
Figure 5 illustrates important differences between the spin coating and casting processes. In this
figure, the Z-direction is defined as the optical symmetry axis (draw) direction in the film, the Y-direction
is perpendicular to the symmetry axis in the plane of the film, while the X-direction is normal to the film
plane. For random orientation in the film, Nz=Ny=Nx, For uniaxial orientation along the draw direction
(fiber symmetry),Nz does not equal Ny and Ny =Nx For planar orientation,Nx does not equal Ny and Ny =
Nz. Further the average refractive index is proportional to the density which is a measure of the
crystallinity. If Ave.N is constant then there is no change in the density of the sample.
Figure 5 (a) shows the three dimensional refractive indices of the film cast [~1 mil] from NMP
solution both before and after drawing to a draw ratio of 3.5. The first thing to notice is that the cast film
before drawing is isotropic. That is, all three principle refractive indices Nz, Ny, and Nx are equal. After
drawing the film becomes anisotropic and uniaxial. The molecules align along the stretching direction, as
shown by the increase in Nz, while the orientation transverse to the stretch direction is symmetrical (Ny =
Nx) and the refractive indices decrease. The average refractive index increases with extension showing
that there is an increase in density and, as will be discussed below, a consequent increase in crystallinity.
In an earlier examination of EB/PANI film cast from DMPU solution at Clemson [8] the cast
film before drawing was also isotropic. After drawing the film to a draw ratio of 2.8 it also became
anisotropic and uniaxial. However, in contrast to the NMP solution films, these DMPU cast films showed
no change in average refractive index with extension and therefor no change in density.
In contrast to film casting from NMP solution the effect on the three dimensional refractive
indices of fabricating film by spin coating from DMPU solution is also shown in Figure 5. In spin coating
(5b) a polymer droplet is deposited onto a spinning surface and is spread out uniformly from the center.
The figure shows that the resulting undrawn film orientation is different from that of the cast film. The
film is randomly oriented in the film plane, Nz = Ny, but the principle refractive index in the thickness
direction, Nx, is different. The spin coated film has a more planar-like structure.
With extension of the freestanding spin coated film, changes in orientation occur in all three film
planes. Nz increases with increasing draw ratio as the random molecules align increasingly in the draw
direction. Ny decreases with increasing draw as does Nx. The rate of decrease in Ny is greater than that of
Nx so that the two values approach each other at the highest draw ratio. Thus the spin coated film goes
from a planar to almost a uniaxial structure with increasing extension. Again, as with DMPU solution cast
films, the average refractive index remains constant indicating there is no phase change with extension.
The above observations indicate that not only do significant molecular orientation differences
occur in films formed from film casting processes as compared to those formed from spin coatings, but
also significant morphological differences occur depending on the polymer solvent used. Whether spin
coated or cast the average refractive index of films produced from DMPU solution showed no change with
film deformation, while films cast from NMP showed increasing average refractive index with extension.
We turn to x-ray diffraction to clarify this observation. X-ray diffraction of the DMPU spin coated and
 5

DMPU cast film both show an amorphous structure (see for example Figure 3a). The film cast from NMP
on the other hand has a crystalline diffraction pattern that increases in degree of crystallinity with
extension [9,10]. Thus our X-ray diffraction measurements are completely consistent with the refractive
index measurements. This should come as no surprise since polymer theory predicts crystallization is
strongly effected by the solvent used.
Polarized infrared measurements have also been used to study the EB PANI films. Polarized
infrared measures the absorption distribution of specific chemical groups within the polymer structure.
Different groups respond at different frequencies and thus there are many absorption bands available. The
transition moment angle is a measure of the angle of oscillation of the group with respect to the molecular
axis. By examining the relation between the infrared dichroic function for different frequencies and the
observed birefringence from the same sample it is possible to identify if the molecular structure remains
unchanged during fabrication. Our work [8] has shown that the EB PANI molecule has an intrinsic
structure with a specific intrinsic birefringence and specific infrared transition moment angles.

Conclusion:

Thus the picture emerging from this work is that EB/PANI is a single material that
morphologically manifests itself differently depending on the fabrication environment it undergoes. This
is an important observation, for it means that rather than naming each morphological manifestation as a
new or different EB/PANI, EB/PANI can now be treated in a consistent unified manner with defined
intrinsic molecular properties for process-structure-property studies. The success of this work to date
suggests a similar treatment of the LEB/PANI system will be equally rewarding and work is continuing on
these systems.
Additionally a thorough knowledge base has been constructed for the processing of electroactive
polymers and particuarly polyaniline which continues, as evidenced by its world wide attention, to be the
polymer of choice for many applications. The work sponsored under this project by the National Textile
Center is recognized internationally as evidenced by an entire chapter devoted to the work in the new
“Handbook of Conductive Polymers”.
 6

REFERENCES:

[1] R. Gregory; Handbook of Conductive Polymers, Chpt. 18; 2nd ed. T. Skotheim, R. Elsenbaumer,
and J. Reynolds, eds. Marcel Dekker. In press 1997
[2] A.P. Chacko, S. S. Hardaker, R. J. Samuels, and R. V. Gregory; Synthetic Metals, 84, (1997) 41
[3] A. P. Chacko, S. S. Hardaker, T. Hanks, and R. V. Gregory; Polymer Communications, in press
1997
[4] Y. Wei, F. K. Hsueh, and G.W. Jang; Macromolecules, V. 27 pp. 518 1994
[5] M. Angelopoulus, Y. H. Liao, B. Fruman, and T Graham; Macromolecules, V. 29 1996 pp.
3046: A. P. Chacko, S. S. Hardaker, B. Huang, and R. V. Gregory; SPE ANTEC Proceedings,
V.1 1996 pp. 1358: A. P. Chacko, S. S. Hardaker,B. Huang, and R. V. Gregory; J. Material
Research Soc., V. 413, 1996 p. 503
[6] S. S. Hardaker, C.Y. Cha, S. Moghazy and R. J. Samuels , Advances in Polyimide Science and
Technology; C. Feger, M. Khojasteh and M. Htoo, eds. Lancaster, Pa., Technomic Pub. Co.,
(1993) 571.
[7] S. S. Hardaker, Moghazy, S., Cha, C. and R.J. Samuels J. Polymer Sci., Part B: Polymer
Physics, 31, (1993),1951.
[8] C. Cha, S. S. Hardaker, R.V. Gregory, and R. J. Samuels, Synthetic Metals, 84, (1997) 743.
[9] R. Ou, T. Liu, H. Wang, B. Mattes, S. S. Hardaker, L. Ding, R. V. Gregory and R. Samuels, SPE
ANTEC 98 Proceedings, April(1998), submitted.
[10] R. Ou, T. Liu, H. Wang, B. Mattes, S. Hardaker, L. Ding, R. Gregory and R. Samuels, Synthetic
Metals, submitted.

Graduate Students: Georgia. Tech.: Runqing Ou, Tau Liu

Clemson University: Steve Hardaker,
Greg Wood, Lili Ding, Jim Yon