Note: Descriptions are shown in the official language in which they were submitted.
12~l8~99 ( 2)
POLYACETYLENE PRODUCTION
The present invention relates to coherent poly(acetylene) films
and to a method of producing such polyacetylenes.
Poly(acetylene) is known to have highly desirable electrical
conducting properties. Such polymers are usually produced by direct
polymerisation of acetylene gas in the presence of a catalyst, eg as
described by Ito et al (J. Polymer. Sci. Chem., 12, pp 11, 1974). The
polymers thus produced have a relatively low density of around 0.4
g/cc and have a morphology which is an open, irregular, fibrillar
structure with random orientation of the fibrils. The conductivity of
such polymers has hitherto been improved by appropriate chemical
doping. The morphology of polymers produced hitherto offers an
advantage with respect to the speed of chemical reactions such as
doping. However, due to the high surface area which is an inherent
characteristic of such a morphology, the poly(acetylene) is also
highly susceptible to oxidative degradation. Moreover, the open and
irregular morphology of the polymer and the r~ndom orientation of the
fibrils makes doping of specific areas of the film with well-defined
edges, which is the basis of the semi-conductor industry, virtually
impossible. Such polymers are also infusible and insoluble in
conventional solvents thereby making it difficult to fabricate
isotropic and anisotropic articles therefrom. There is a continuing
need in industry for a poly(acetylene) which can be easily and
conveniently fabricated into articles of a desired shape, and which
can be fabricated to possess a degree of chain alignment. This chain
alignment increases the aniso~ropy of electrical properties.
~2187~9
More recently, Edwards and Feast (Polymer, vol. 21, June 1980,
pp 595) have described a ~ethod of producing poly(acetylene) (III) by
first polymerising a precursor 7,3 bis(trifluoromethyl)tricyclo-
[492,2,02 5]-deca3,7,9-triene (I) using a catalyst in toluene and the
precursor polymer (II) so formed is spontaneously decomposed to a
black product and 1,2-bis(trifluoromethyl)benzene. When the precursor
polymer (II) was heated to 150 under a vacuum of 0.01 mm of mercury
for 5 hours the authors obtained a product which had an infra-red and
Raman spectrum corresponding to that of trans-poly(acetylene) although
the elemental analysis showed that only 96.3% of the fluorine had been
removed. When heated for a further 3 hours at 210C, 98.9~ of the
fluorine had been removed although the polymer would probably have
been degraded by this stage. The authors stated that this type of
system is too labile for convenient generation of poly(acetylene) and
that they were investigating related structures in order to find a
more stable precursor.
It has now been found that a poly(acetylene) having a higher
density and a markedly different morphology can be produced from the
same or similar precursor polymers under appropriate conditions.
Accordingly, the present invention is a coherent poly(acetylene)
film.
According to a further embodiment, the present invention is a
process for producing a coherent poly(acetylene) (III) film comprising
solvent casting a solution of a polymer of the general formula (V)
derived from a precursor (IV), and transforming the pre-cast, polymer
(V) into the poly(acetylene) (III) film and a by-product (VI) at a
temperature between 20 and 200C under reduced pressure over a
duration of between 1 and 100 hours in an atmosphere inert to the
precursor polymer (V) and to the poly(acetylene), wherein in the
general formulae IV, V and VI each of the groups Rl and R2
either (a) represent a radical selected from H, CX3, CmH2m+1 and
COORs wherein X is a halogen atom, m has a value
betweeen 1 and 4 and R5 is an alkyl group with 1-4
carbon atoms,
35 or (b) form together with the respective carbon atoms to which
they are attached a benzene nucleus,
~L2~ 9
l~ I
+
~ '~ --Q~
.
¢ ~__. . ~ _~
D`~ Y ~
"~ ~ E~i
~D
` 3
12~37~9
and
each of R3 and R4
either (c) represent H atoms,
or (d) form together with the respective carbon atoms to which
they are attached a benzene nucleus.
Specific examples of the groups R1 and R2 are a trifluoromethyl
group, an alkyl group or an alkyl carboxylate group, especially a
methyl carboxylate group.
More specifically the precursor (IV) may be a compound in which
R1 and R2 are each a trifluoromethyl group, and R3 and R4 are each a
hydrogen atom. Such a compound is shown in formula (I) above.
The precursor polymers (V) used in the solvent casting step may
be produced by conventional means e.g. that described in the paper by
Edwards and Feast referred to above. The polymerisation of the
precursors (IV) is suitably carried out in the presence of a tungsten
hexachloride/tetralkyl or aryl tin (1:2 w/w) or titanium
tetrachloride/ trialkyl or dihaloalkyl aluminium (1:2 w/w) catalyst at
ambient temperatures and pressures. Depending upon the tendency of
the precursor polymer to transform to poly(acetylene), it must be
stored at a temperature low enough to slow down this transformation.
For example, the precursor polymer (II) is suitably stored at
relatively low temperatures suitably below -10C, preferably below
-20C and typically 26C to prevent the premature transformation
thereof into poly(acetylene). At this temperature the precursor
polymer of the formula (II) is stable for at least 14 months. Where
R1 and R2 represent a benzene ring and R3 and R4 represent hydrogen
atoms then the precursor polymer may be conveniently stored at room
temperature at which temperature it is stable for at least six months.
For solvent casting, the precursor polymer (V) is preferably
dissolved in an organic solvent to a concentration which, for a given
depth of solution, gives the desired thickness of the required shape.
This concentration is typically up to 100 g/l (approximately
equivalent to eg a film thickness of up to 100 microns). The
precursor polymer is suitably cast from an organic solvent selected
from acetone, chloroform, ethyl acetate, toluene and xylene although
., .
99
solvents such as toluene and xylene are preferred for precursor
polymers with aromatic substituents.
The precursor polymer (V) may be cast into any desired shape
although shapes with a relatively high surface area e.g. a film or a
fibre facilitate the transformation reaction. During the solvent
casting process it is most desirable to minimise moisture and/or
oxygen content of the system in order to produce a coherent film
having the desirable properties of conductance. It is most preferable
to carry out the casting in an atmosphere inert with respect to the
precursor polymer (V) and the eventual polyacetylene film formed. The
inert atmosphere is suitably provided by nitrogen or argon gas. The
casting temperature may be suitably adjusted to control the rate of
deposition of the precursor polymer from the solvent.
After casting, the temperature at which and the duration for
which the precursor polymer is heated to produce the poly(acetylene)
film will depend upon the nature of the substituents in the precursor
polymer. For example, the precursor polymer is preferably heated at a
temperature between 20 and 200C for between 1 and 100 hours to
produce the poly(acetylene) film. The Table below illustrates the
preferred ranges for various substituents in the precursor polymer.
Table 1
Substituents
25 IR1 R2 R3 R~l TempC Time (hours)
H H H H 20 _ 50 1 - 20
CF3 CF3 H H 50 ~ 120 1 - 100
CO2CH3 CO2CH3 H H 50 ~ 120 1 - 100
benzene ring H H 100 - 150 1 - 100
benzene ring benzene ring 17; - 200 1 - 100
For instance, the polymer (II) is heated at a temperature below 150C,
preferably between 50 and 120C, under vacuum or in the presence of an
inert atmosphere, e.g. nitrogen, to transform the precursor polymer
(II) into a coherent poly(acetylene) film. The heating procedure may
be carried out for a period between 1 and 100 hours suitably between
10 and 50 hours to form the poly(acetylene) polymer. The rate of
heating is suitably between 1 and 10C per minute. The lower the
heating temperature, the longer the duration of heating. The
preferred ranges specified relate to that needed to achieve a
~Z18799
6 :
substantially complete transformation of the precursor polymer to a
coherent film. For some uses partial transformation may be adequate
and hence slight variations outside the preferred ranges may be
acceptable. It should be noted, however, that effort to transform the
precursor polymer below 20C results in a creased and wrinkled film
whereas at temperatures above 200C there is substantial degradation
of the polymer and loss of coherence due to appearance of voids.
The poly(acetylene) thus formed has a substantially higher
density than the poly(acetylene) polymers produced hitherto. For
instance, the density of the poly(acetylene) (III) produced according
to the present invention is approximately 1.0 g/cc whereas that of the
poly(acetylene) produced according to the prior art methods is only
about 0.4 - 0.5g/cc. The calculated density of poly(acetylene) is 1.2
g/cc. The morphology of the poly(acetylene) produced according to the
present invention is shown by scanning electron microscopy (SEM) ~o be
that of a thin, coherent solid film with no voids and, no basic
structural units are visible even at a magnification of 10,000 times.
At such magnification poly(acetylene) produced by prior art methods
reveals a clear fibrillar structure. The polymers also have a lower
crystallinity than those produced by conventional methods. The
resonance Raman spectra show that these poly(acetylenes) have shorter
lengths of conjugated double bond in sequence than those derived by
prior art processes. Typically, the coherent poly(acetylene~ films of
the present invention have a C=C stretching frequency of 1480 cm~l
which may be inte~preted as a sequence of no more than 25 double bonds
per chain whereas the poly(acetylenes) of prior art have a C=C
stretching frequency of 1460 cm~1 which may be interpreted on the same
basis as a sequence of at least 75 double bonds per chain. X-ray
diffraction and electron diffraction spectra also show the
poly(a`cetylenes) of the present invention to be of lower crystallinity
and hence distinct.
The conductivity of the pristine poly(acetylene) produced
according to the present invention is in the range of between 10-8 and
10-5 per ohm per cm.
The electrical properties of the poly(acetylene) produced
~2~879g
according to the present invention may be altered as desired by
addition of suitable dopants known in the art. Examples of ~opants
include the halogens, fluorides of arsenic, and protonic acids. The
dopants ~ay be added either to the solution from which the precursor
polymer (V) is cast or to the pre-cast polymer by diffusion thereof
from a gas or liquid phase, electrochemical diffusion or by ion
implantation techniques. The morphology of the poly(acetylene)
produced by the process of the present invention renders it
particularly suitable for selective area doping, with a resolution
which is better than 1000 ~. In comparison the fibrils in
conventionally produced poly(acetylene) give a resolution figure which
rises to around 100 microns which is approximately 1000 times larger.
Upon doping, the conductivity of these films can be substantially
improved. For instance, by using iodine as dopant the conductivity of
the coherent film may be improved to a value of between 1 and 20 per
ohm per cm.
The process of producing poly(acetylene) according to the present
invention also enables controlled chain-alignment of the molecules in
the polymer prior to the transformation reaction. The polymer (V) may
be cast on a flexible substrate or in the form of a free-standing film
which can be stretched as necessary to provide the desired chain
alignment. Alternatively it can be cast from a solution in a shear
field to achieve the desired chain alignment.
The process for producing poly(acetylene) according to the
present invention is further illustrated with reference to the
following Examples.
Example 1
A solution of 1 g of the precursor polymer (II) dissolved in
25 ml of acetone was prepared. Approximately 4.5 ml of this solution
3~ was cast onto a 3in. diameter silicon slice and the solvent allowed to
evaporate under a reduced pressure of nitrogen (200 mm of Hg pressure)
at room temperature (about 25C). The pressure was then further
reduced to less than 1 mm of Hg pressure, the temperature raised to
80C, and maintained for 10 hours. The fawn film gave way to a black,
shiny film of poly(acetylene). Examination of the film ln the
~2~87~9
,.
scanning electron microscope showed a solid, coherent film of 20 ~ 2 r
microns thickness. Ihe film had a measured density oP about 1.0
g/cc. The conductivity was found to be approximately 10-6 (ohm cm)~l,
Example 2
A solution of 0.5 g of the precursor polymer (II) dissolved in
25 ml of acetone was prepared. Approximately 0.5 ml of this solution
was cast onto a quartz plate and the solvent allowed to evaporate
under a nitrogen atmosphere at room temperature. The pressure was
then reduced to less than 1 mm of Hg pressure and the sample
maintained at room temperature (about 25C) for 24 hours. It was then
heated up to 80C and maintained at that temperature for 15 min before
cooling back to room temperature. A film was formed, in which 90~ of
the precursor polymer had been converted to a coherent polyacetylene
film. The conductivity of the film at room temperature was then
15 measured and found to be 3.5 x 10-6 (ohm cm)~l. The density of the
film was about 1.0 g/cc. The sample was then subjected five times,
still under vacuum, to a temperature cycle of between 3C and 80C
by heating at the rate of 1C per minute to 80C and maintaining at
that temperature for 20 minutes before cooling down to start the next
cycle. It was exposed to iodine vapour and doped to a molar ratio
level of 17~4% with respect to the polymer and represented by the
empirical formula (CHIo.17)X. The conductivity was measured to be 0.1
(ohm cm)~l at 50C. On examination the thickness of the film was
found to be 20+3 microns.
Example 3
Example 2 was repeated except that the precursor polymer was
heated at ôOC for 2 hours and the conductivity of the coherent
polyacetylene film (CHIo.2s)X produced was 15(ohm cm)~l at 23C.
Example 4
A precursor polymer (V) in which R1 and R2 represent a benzene
ring and R3 and R4 each represent a hydrogen atom was dissolved in
ethyl acetate (0.75 gm/25 ml) and heated at 120C for 10 hours to give
a coherent poly(acetylene) film.
Example 5
Kinetic analyses were carried out on various precursors using a
~2~B79~
Differential Scannlng Calorimeter and software which enabled analysis
of the transformation reaction from precursor polymer to a
poly(acetylene) film. The data were related to other spectroscopic
analyses to optimise the transformation. The data obtained is
5tabulated below:
Table 2
Transformation
Substituents in the (Heating Rate
precursor polymer 2.5C/min)
Temp range
-I
R1 R2 R3 R4 From To
H H H 15 60
CF3 CF3 H H 50 100
CO2CH3 CO2CH3 H H 50 105
benzene ring H H 90 160
benzene ring benzene ring 180 220*
* Degradation of poly(acetylene) sets in above 200C
Example 6
A solution of 0.25g of the precursor polymer (II) dissolved in
25 ml of acetone was prepared. Samples for spectroscopic analysis
were prepared in two ways:
(a) Transmission Electron Microscope grids coated with a thin layer
of carbon were immersed in the solution, quickly dried and heated
at 80C for 2 hours. They were then examined in a Transmission
Electron Microscope, by electron diffraction techniques. They
showed a diffuse ring in contrast to poly(acetylene~ produced by
conventional techniques [described by (i) Shirakawa, H. and
Ikeda, S., Polymer Journal (1971), Vol 2(2), p 231 and (ii)
Luttinger, L.B., Journal of Organic Chemistry, (1962), Vol 27,
p 1591. These films were produced at -78C and then isomerised
at 150C for 5 hours~ which show a pattern of sharp rings. This
indicates substantially reduced crystallinity for the
poly(acetylene) produced according to the present invention.
~L2~87g9
(b) The precursor polymer was cast from this solution on the inside
of a flask. The precursor was trans~ormed at 80C for 4 hours.
(c) The sample was examined by Resonance Raman Spectroscopy, whilst
still in the flask under vacuum. When irradiated with a laser
line of 676.4 nm the sample gave a C=C stretching frequency at
1480 cm~l. This compares with a value of 1460 cm 1 for
poly(acetylene) prepared by conventional techniques, and
indicates markedly shorter conjugation lengths.
(d3 The film was removed from the flask under nitrogen and examined
by X-ray diffraction. The trace obtained showed a broader, less
intense peak at higher lattice spacings than those obtained from
conventionally produced poly(acetylene) [cf (i) Shirakawa et al
and (ii) Luttinger Loc. Cit.~. This confirms the electron
diffraction experiment suggesting substantially reduced
crystallinity.
Example 7
A solution of 0.5g of the precursor polymer (II~ dissolved in
25 ml of deoxygenated acetone was prepared. Approximately 10 ml of
this solution was placed inside a round bottom flask. It was spun
around the inner surface of the flask by gentle rotation o~ the
latter. This process was continued until all the solvent had
evaporated under a nitrogen atmosphere at room temperature. The
pressure was then reduced to less than 10-3 torr as it was heated up
to 80C and maintained at that temperature for 2 hours before cooling
back to room temperature. A film of about 15 microns thick was
formed. A portion of this film (5.6 mg by weight) was used for the
iodine weight uptake experiments using a quartz-fibre coil spring
(extension 1 mm/mg). The diffusion coefficient of iodine in the film
thus obtained was approximately 10~14cm2s~1, which corresponds to a
geometric resolution of less than 20 microns. In comparison the
geometric resolution of the conventionally produced poly(acetylene)
[(as described by Berniere, F. et al, in Journal of Physics and
Chemistry of Solids, Vol 42, pp 649-654f, (1981)] was higher than
5 mm.
, .
12~1l8~99
11
Example 8
A solution of 2 gm of the precursor polymer (II) dissolved in
20 ml of acetone was prepared. Approximately 1 ml of this solution
was cast onto an aluminium scanning electron microscope stub, The
5 temperature was quickly raised to 150C and maintained for 5 hours
under a vacuum of 0.01 mm of Hg. The edges of the film became
detached from the stub and curled up. One such section was fractured
and the broken edge examined. It showed significant voiding.
Example 9
A solution of 5 gm of the precursor polymer (II) dissolved in
20 ml of acetone was prepared. It was cast onto a salt-plate and kept
under a vacuum o~ 0.01 mm of E~g at room temperature (approx. 22C) for
200 hours. The film appeared lumpy and shrivelled. Infra-red
examination of the film showed less than 70% loss of hexafluoro-
15 ortho-xylene.
Example 10
Approximately 5 ml of the precursor polymer solution of Example 7
was cast on a glass slide. It was first evacuated at room temperature
under less than 10~3 torr to remove the solvent (acetone). The sample
20 slide was then sealed onto the sample holder of a Perkin Elmer UV/VIS
spectrophotometer. The latter was evacuated to a vacuum of less than
10-3 torr at 23C. The sample was yellowish in colour and an optical
spectrum (0.5-5.5 eV) was first obtained at 23C. The absorption edge
(ie optical bandgap) of this spectrum was 2.0 eV. The sample was then
25 heated at 65C for 5 hours and the optical scanning was repeated every
hour. As the appearance of the polymer changed from yellow, through
orange, red and brown to eventually black, metallic lustre, the
optical bandgap decreased from 2.0 eV to 1.5 eV.