Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
--1--
~I~LE
Melt-Processible ~etrafluoroethylene
Copoly~ers and Pr~cesses for Preparing ~hem
F I E L D OF ~HE I NYE NT ] ON
~his invention relates to melt-processible
tetrafluoroet~ylene copolymPrs having good par~icle
flow charac~eris~ics and thermal stability.
Such melt-processible copolymers can be
extruded onto wire or extruded into film or tubing,
or used as a coating, or tan be used in rotomolding
applications to make hollow articles or linings.
BACKGROUP~D OF THE INVENTION
Tetrafluoroethylene polymers are of two
types. One is non-melt-processible polymers where
the melt viscosity is too high to process the
polymers by ordinary ~elt-extrusion processes.
Instead, the polymers are ordinarily sintered or
paste extruded depending on the type polymer made.
The other class is ~elt-protessible
tetrafluoroethylene copolymers having melt
viscosities in the mel~ extrudable range.
Melt-processible tetr3fluoroethylene (TFE)
copolymer resins directly from the polymerizer and/or
coagulator are referred to as fluff or powder. ~he
fluff is normally humid heat treated and/nr melt
extruded to stabilize it, such as described in U,S~
30 Patent 3,085,DB3, There are applications such as
rotocasting in which a free-flowing powder (herein
called "granules") is preferable to melt-extruded
pellets or where a high degree of purity of the resin
is desired. AlthDugh rotolining and rotocoatingD 5415 35 processes differ in several technital respects fro~
roto~olding, for the sake of convenience the term
"rotocasting" is used herein to refer to dll three
generically unless otherwise lndicated.
~2-
To facilitate handling of such sranules, i~
is desirable to improve particle charac~eristits.
Melt-processible copolymers that are toagula~ed from
an aqueous dispersion and ~ried are friable~ and form
fines easily which give poor handling propertiesO It
would be desirable to provide a melt-processible
copolymer that is bDth stable and easily handled in a
minimum of processing steps. It is particularly
desirable to provide a copolymer that could be used
both in con~entional melt-fabrication processes and
in rotocasting applications where particle
characteristics are important.
It is also desirable to obtain resins that
are thermally stable. A number of stabilizatior,
appr~aches are known in the art, most of which
require melting the resins. Thus resins stabili2ed
by these methods are generally available only as
pellets -- not (without tedious and expensive
regrinding steps) as ~he free-flowing granules that
are the basis of this invention.
Another desirable feature of such resins is
that the granules should be low in metal
contamination. If the granules have been melted in
traditional thermoplastic processing equipment,
contamination occurs inevitably when the corrosive
tetra~luoroethylene copolymer melts come in tontact
with the interior metal surfaces of thermoplastic
processing equipment, even when corrosion-resistant
alloys are used. Copolymers having low levels of
metal contamination are particularily desir3ble for
applications in the semiconductor industry.
SUMMARY OF THE INVE~ION
A conventional produtt form for
melt-processible ~etrafluoroethylene copolymers js
extruded pellets -~ either str~nd-cut right
cylinders, or mel~-cut ~iscs or cylinders. ~hese
pellets are used as the feed to thermoplastit
processing equipment.
An alternative product form for
melt-fabricable tetrafluoroethylene copolymers is
~ery finely divided powders. ~his produc~ for~ has
been used as the feedstock for coating operations,
which are well known in the trade.
The subject of this patent is a new product
form, namely, free-flowing~ attrition-resistant,
generally spherical9 heat-stable granules~ ~hese
granules are of high purity and thermal s~ability ~n
air, having particular utility in ~abricating
free-standing rotomolded articles and providing
defect-free polymeric coatings or linings, especially
those produced by rotolining metal process
equipment. The novel compositions have improved
thermal stability and low bubble tendency. More
specifically, the compusition is a melt-processible,
substantially nonelastomeric tetrafluoroethylene
copolymer comprising B0-99.5 mole X
tetrafluoroethylene and 0.5 to 20 role X of at least
one eopolymerizable comonomer, ~hith copolymer has
(a) a melt viscosity between 0.1 x lU4
and 100 x 104 poise at 372C,
(b) a substantially spherical particle
shape and a sphere factor less than
1 .~,
(c) an attrition factor of less than
6~,
(d) fewer than a total of 8U unstable
end groups per 10 carbon atoms, ~nd
(e) an average particle size
between 200 and 3000 mitrometers.
The process of ~his invention starts with
melt-processi~le tetrafluoroethylene copolymers that
have been polymerized in an aqueous medium and
contain unstable end groups. When prepared in an
aqueous medium, the copolymers are isolated by
solvent-aided coagulation preceded by gelation. ~he
resulting coagulated granules are spherical in shape,
which facilitates handling. The granules are then
dried and hardened by subjecting them to elevated
temperature between the differential scanning
calorimeter (DSC) peak melting point and 25C below
the melt onset temperature (i.e., the granules are
heat treated to harden them, but not so as to
completely melt or substantially deform them). The
hardening fatilitates screen sieving or methanical
screen sifting into desired particle sizes and
facilitates handling by reason of reduced
friability. ~he granules are then subjected to an
atmosphere containing fluorine to convert unstable
end groups to stable fluorinated end groups, thereby
reducing bubbling or evolution of volatiles during
further end-use heat processing.
These granules are especially well suited
for rotocasting applications because of the optimal
particle size and free-flowing characteristics
combined with low bubble tendency.
A further benefit of the stabilized
free-flowing granules is that such granules have not
been melted in conventional thermoplastit processing
3~ equipment and are low tn metal oontamination.
--5--
DESCRIPTION OF THE INV~NTION
_
Represen~ative fluorinated ethylenically
unsaturated comonomers copolymerizable with
tetrafluoroethylene are represented by thP formulas:
F\ Cl
/C=CF2 /C=CF2 or ~ =CH2
R1 F R2
wherein R1 is -Rf, -Rf-X, -O-Rf or -O-Rf-X
in which Rf is a perfluoroalkyl radical of 1-12
carbon atoms, -Rf- is a perfluoroalkylene diradical
1~ f 1-12 carbon atoms in which the attaching valences
are at each end of the chain, and X is H or Cl; and
R2 is -Rf or R~-X.
Specific copolymerizable fluorinated
ethylenically unsaturated comonamers include
hexafluoropropylene, perfluoro(methyl vinyl ether),
perfluoro(n-propyl vinyl ether), perfluoro(n-heptyl
vinyl ether)~ 3,3,3-trifluoropropylene-1,
3,3,4,4,5~5,6,6,6-nonafluorohexene-1,
3-hydroperfluoro(propyl vinyl ether), or mixtures
thereof, such as a mixture of hexafluoropropylene and
perfluoro(propyl vinyl e~her). Preferably the
comonomers are selected from perfluoro(alkyl vinyl
ethers) of the formula Rf-O-CF=CF~; or
hexafluoropropylene; or compounds of the for~ula
Rf-CH=CH2, wherein -Rf is a perfluoro3lkyl
group of 1-12 carbon atoms.
Comonomer content can range from 0.5 mole
percent up to about 20 mole percent, and more than
one comon~mer can be present.
3~
4 ~1~#~1
-6-
The comonomer content is low enough that the
copolymers are plastics rather than elastomers, i.e.,
they are partially crystalline and after ex~rusion do
not exhibit a rapid retraction tD original length
from a stretched condition of 2X at room temperature~
The aqueous polymerization of TFE with
various comonomers is well known. The reaction
medium consists of waterJ monomers9 a dispersing
agent, a free-radical polymerization initiator,
optionally, a chain-transfer agent and, op~ionally, a
water-immiscible fluorocarbon phase, as described,
for example, in UnS~ Patent 3,635,926.
Po1ymerization temperatures between
20-140~C may be employed and pressures of 1.4-7~0
MPa are ordinarily usedO Generally higher
temperatures and pressures are employed to 1ncrease
polymerization rates, especially iF a comonomer is
less reactive relative to TFE. ~he TFE and sometimes
the comonomer are fed continuously to the reaction
vessel to maintain reaction pressure, or in some
instances the ~omonomer is all added in~tially and
pressure is maintained with TFE Feed only. The
monomer(s) are fed until the desired final dispersion
solids level (15-SOX) is achieved. The agitator
speed in the reaction vessel may be held constant
during polymerization or it may be varied to control
polymerization rate.
Initiators commonly employed are
free-radical ini~iators such as ammonium or potassium
persulfate or disuc~inic ac~d peroxide. The
dispersing agent will be present in an amount between
0.01-0.5 percent based on weight of 3queous medium
and preferably between 0.05-0.1 percent,
-6-
By the term '!melt-processible" is meant that
the copolymer can be processed (i.~., fabricated into
shaped articles such as films, fibers, ~ubes, wire
coatings and the like) by conventional
melt-processing equipment. Such requires tha~ the
melt-viscosity of the copolymer at ~he processing
temperature be no more than 107 poise~ Pre~erably
it is in the range of 10 to 10 poise ak 372DC.
Melt viscosities of the melt-processible
polymers are measured according to American Society
for Testing and Materials ~ethod D-1238, modified as
follows: The cylinder, orifice and piston tip are
made of a corrosion-resistant alloy, such as Haynes
Stellite~tm) 19 or lnconel(tm) 6250 ~he 5.0 9 sample
îs charged to the 9.53 mm inside diameter cylinder
which is maintained at 372~C + l~Cu Five minutes
after the sample is charged to the cylinder, it is
extruded through a 2.10 mm diameter, 8.00 mm long
square-edge orifice under a load (piston plus weight)
of 5000 grams. ~his corresponds to a shear stress of
44.8 kPa. The melt viscosity in poises is calculated
as 53170 divided by the observed extrusion rate in
grams per minute.
The copolymers prepared by the foregoing
aqueous polymerization process are colloidally
dispersed in the polymerization medium~ The polymer
is recovered from the ~ispersion by toagul~tion.
Normal coagulation of aqueous polymer dispersiDns by
mechanical shear tends to giv~ a very finely divided
powder which has poor handling characteristics.
Several techniques might be used to obtain the
preferred larger partitle sizes. The combin~tion of
mechanical agitation and certain chemical ~dditions
can be used to Dbtain larger, spherical particles.
~8--
In the process of 2he invention, the aqueous
dispersion is gelled with a gelling agen~, a mineral
acid, ~hi1e being agitated. Preferably nitric acid
is used as the gelling agent. A water-immiscible
liquid is then added to the gel while continuing the
agitation. The gel breaks up into separate phases Gf
water and liquid-wetted polymer particles. The
particles are then dried. The granule si2e is a
function of the dispersion particle size9 the ra~io
of water-immiscible liquid to polymer, and the
agitation condi~ions. The granule size is, as
desired, much larger than that achieved if the
dispersion is coagulated by mechanical shear ac~ion
alDne. Usually, the amount of water-immiscible
liquid will be 0.25 to l.0 part per part of polymer
on a dry weight basis. About O.l to 10 parts of
HN03 per 100 parts sf polymer weight can be usedO
Hitric acid is preferred because it is not corrosive
to stainless-steel equipment and readily volatilizes
in a subsequent baking step. Coagulated particles
obtained by this process generally have a size
between 2Q0-3000 micrometers. The product is
separated, washed and dried at 80 to 150C for from 4
to 30 hours.
Preferably, the water-immiscible liquid
should have a surface tension of not more than 3~
dyne/cm at 25C and it should have a normal boiling
point in the range of 30 to 150C. Typical exa~ples
of the immiscible liquid usable in the invention are
aliphatic hydrocarbons such as hexane, heptane,
gasoline and kerosene, or mixtures thereof,
- 9-
aromatic hydrocarbons such as benzene, toluene and
xylene, halogenated derivatives such as carbon
tetrachloride, monochlorobenzene, the
trichlorotrifluoroethanes,
difluorotetrachloroethanes, and liquid oligomers of
chlorotrifluoroethyleneO
Other techniques might also be used to
obtain the particle s~zes preferred in this
invention. Nucleation agents might be added to the
aqueous dispersion before coagulation which would
result in larger particle sizesO Small polymer
particles which were obtained from mechanical
coagulation might be redispersed ~n a two-phase
liquid mixture and thus agglomerated into larger
particles. The polymerization itself might be
carried out with a water/immiscible liquid mixture so
that particles of the desired size could be obtained
directly from polymerization.
~ he dried particles are generally spherical
and have a sphere factor less than 1.~, and
preferably less than 1.2. The sphere factor is a
measure of the degree of roundness of the particles.
A sphere factor of 1 represents a geometrically
spherical particle.
The particles are then hardened by heat
treatment until the attr~tion factor, as described
herein, is less than 60 and preferably less than 25,
but before the granules agglomerate.
~ 4~d
-10-
By ~he ~erm "before ~he granules agglomerate" is
meant that the D50 as hereinafter defined does not
increase by more than 20X.
Heat hardening of the granules formed in the
toagulation step occurs relatively close to the
copolymer melting point. The ~emperature a~ which
hardening occurs depends not only on ~he copolymer
melting point but also on other characteristics such
as comonomer and molecular weight distributionsO
These characteristics influence the temperature at
which the onset of melting occurs.
~his heat-hardening phenomenon occurs when
the copolymer granules are held at a temperature
within the range between the copolymer melting point
and a temperature ~5~C below the melt onset
temperature, as measured by differential scanning
calorimetric (DSC) methods described herein. The
granules must be exposed to ~emperatures within this
range for a time sufficient to impart a useful degree
of hardness. The resulting heat-hardened gr~nules
are not completely melted and are only partially
sintered. If the melting heat ratio as hereinafter
defined is belaw 1.2, the polymer granules have been
melted and begin to fuse together9 After heat
hardening, the granules have a level of hardness
useful in preventing attrition and fines generation
during subsequent steps in the manufacturing process
and also in melt fabrication.
~ he manufacturing process for the granules
may optionally include sizing, such as screen
granulation to mechanically force all the gr~nules
through a screen of selected mesh size which breaks
up the oversize particles while maintaining the
useful particle characteristics described herein.
Some lump formation occurs during heat hardening ~nd
- 1 0 -
fluorination. Such screen granulation is ef~i ti en~
in nemoving these lumps, which are de~rimental in
rotocasti ng operations.
These particles contain unstable end
groups. The end groups found in the untreated
polymer directly from polyme,rization depend on the
initiator used and on the presence of pH and
molecular weight modifiers. For example~ if ammonium
or potassium persulfa~e is employed as the initiator,
the polymer end groups are essentially all carboxylic
acid (-C02H). The acid end groups are foun~ in
1~ both monomeric or dimeric forms. If a pH modifier
such as ammonium hydroxide is present, then a large
portion of the carboxylic acid ends may be converted
to amide ends (-CONH2). If a molecular weight
modifier such as methanol is employed~ then a portion
of the ends may be carbinol (-CH20H) as well as the
more stable difluoromethyl ends (-CF2H). ~he
presence of methanol can also lead to methyl ester
ends ( C02CH3). Vinyl ends (-OF=CF2) are
generally not a direct result of polymeri7ation but
are formed as a result of decarboxylation of the
initially formed oarboxylic acid ends~ Acid fluoride
ends (-COF) generally result from air oxidation of
the vinyl ends or ~he carbinol ends. All of the end
groups described above (except ~CF2H) are
considered to be thermally and/or hydrolyt~cally
unstable. This is what is meant by the term
"unstable end groups". They have a tendency to cause
bubbles or voids upon melt fabrication. These vo~ds
can be detrimental to the physical or electrical
properties of fabricated articles. It i~ desirable
to have less than 80 of these unstable ends per 106
carbon atoms in the polymer.
9~
-12-
The unstable end groups described above may
be reduced or eliminated by ~reatment of ~he polymer
~ith fluorine. The f 1 uorination may be carried out
with a variety of fluorine radic~l generating
compounds but preferably the polymer is contacted
with fluorine gas. Since reactions with fluorine are
very exothermic, it is preferred to dilute the
fluorine with an inert gas such as nitrogen. The
level of fluorine in the fluorine/inert gas mixture
may be 1 to 50 volume X but is preferably 10-30%.
Any reaction temperature between O~C and the polymer
melting point may be used but a temperature range
between 130 and 200C appears to be practital to
permit reasonable reaction times (1 to 5 hours under
fluorine)~ It is preferred to agitate the polymer to
expose new surfaces ~ontinuously. The gas pressure
during fluorination may range from atmospheric to 1
MPa. If an atmospheric pressure reactor is used, it
is convenient to pass the fluorine/inert gas mixture
through the reactor continuously. After exposure of
the polymer for the desired length of time, the
excess fluorine is purged from the sample, which is
then cooled.
Most of the unstable end groups are
converted to perfluoromethyl (-CF3) ends by the
fluorine. The acid fluoride ends are the most
resistant to fluorine but will react at sufficiently
high temperature or with sufficient time.
The preferred copolymers should have a
melting heat ratio greater than 1.2. By melting heat
ratio ~s meant the ratio of the heat of melting on
its first melting to the heat of melt1ng recorded on
a second melting. This is an indication that the
particles have not been melted completely.
-13-
TEST PRDCEDURES
BUBBLE INDEX
The bubble index test referred to ln the
examples is performed as follows: A 159 sample of
copolymer is weighed into a new clean aluminum pan
which is about 50mm in diameter, 16mm deep and 0.08mm
thick. The sample (with controls for comparison) is
baked at 50~5C above the melting point for 40
minutes in a high-temperature recirculating air
oven. The baking time is defined as specimen
exposure time between closing and opening the oven
door. The oven air temperature is preset and
recovers to set temperature within 5 to 10 minutes
after sample addition. After cooling to room
temperature, the polymer specimen is removed from the
pan. The degree of bubble formation 1s observable
qualitatively and is measured by the pertentage
increase in specific volume of the specimen rela~ive
to the bubble-free polymer. The Bubble Index is
defined as-
Bubble Index = ~((A - W) G)- 1~ 100
where:
G = Specific Gravity of
bubble-free copolymer as
determined by ASTM Method D-792.
A = Net weight of specimen in air.
W = Net weight of specimen in water by
displacement method.
-13-
-14-
The entire exposed Bubble Index specim~n ~ 5
~eighed in air and under water ~o ~he neares~ O~OI
gram on an elettronic balance. 70 obtain the net
weight submerged in ~ater, ~ stainless-steel wire
harness with depth mark is suspended from a small
ring stand on the balance an~ ~he t~re welgh~ of ~he
harness i5 set ~o null in water before the spec~men
is added to the harness and submerged ~o the fixed
depthO About 800 ml of demineralized water
containing one drop o~ Triton* X-100 or X-500
surfactant is used for submersion at room
temperature. Specimens are observed under water to
insure constant lmmersion depth and ~he absence of
bubbles on the specimen surfate.
A~TRITION FACTOR
____
Particle hardness ~s measured by D screening
test as follows:
Equipment:
Fritsch Pulverisette*, Model 24-0~17-000
(TeKmar Company, Cincinnat1, OW)
Sieve (U5A Standard Testing Sieves)
- 51mm high x 203mm di 3 X 30 mesh for
granules of D50 greater than 700
micrometers.
- 51mm high x 293mm dia x 80 mesh for
granules of D50 less than 700
micrometersO
Pan and dome lid, 203mm dia.
I9mm diameter st~inless-steel balls (329 each)
Procedure:
Place IOO.O g of polymer (~s) onto the
screen which has been placed on the prewe~ghed pan
(~0). Place the dome lid on top and posit~on ~n the
Fritsch Pulverisette(tm) ~pparatus. Preset a~ol~tude
to l.~mm (amplitude setting of 3).
Attach the retain~ng straps to the lid and
*denotes trade mark
-14-
-15-
tighten securely~ Set the timer for 10 min and
activateu At the end of 10 min remove ~he lid and
screen~ brushing polymer adhering to ~he ~nsid2 of
the bot~om rim of the screen into ~he pan. ~eigh the
pan twl)- Place 12 of the stainless-steel balls on
the screen and reassemble pan/screen/lid and place in
the Pulverisette(tm). Set timer for 10 min and
activate. After 10 min disassemble the screen again
brushing polymer adhering to the inside of bottom rim
into the pan. ~eigh the pan and contents again
(W2). Calculate attrition factor as follows:
1~ Attrition = r ~2 w~ O
Factor ~ ` ~ J
DETERMINATIOH OF SPHERE FAC~OR
,
A small amount of sample is placed on a
glass microscope slide, dispersed into a single layer
by shaking slightly, and then photomicrographed. On
a print, the largest and shortest diameters (a and b)
of each particle are accurately measured to within
+ ~X using more than 30 partitles selected at random.
The sphere factor is calculated ~ccording to
the method of U.S. Patent No. 3,9119072 dS follows:
(n = number of particles measured)
Sphere Factor = 1 ~ aj
in~ ~in (i - 1,2,3,..,9n)
END GROUP ANALYSIS
The end groups in a fluorocarbon polymer are
determined from the infrared spectrum of compression
molded films. This technique has been described in
previous patents such as U.SO P3tent 3,085,083.
The quantitat1ve measurement of the number
of end groups is obtained using the absorptivities
measured on model compounds containing the end groups
-16-
cf interest, The end groups of concern, the
~avelengths involved, and th~ calibration factors
determined from model compounds are shown below:
~avelength, Calibration Factor
End~roup micrometers _ (CF~ _
-COF 5.31 406
-C02H(M) ~,52 33S
1~ -C02H(D) 5.64 320
-C02CH3 5.57 368
-CONH2 2.91 914
-CF=CF~ 5.57 635
-CH20H 2.75 2220
M = Monomèric, D = Dimeric
~he calibration factor is a mathematical
conversion to give end group values in terms of ends
per 106 carbon atoms. The concentra~ion of each
type of end in a polymer film may generally be
obtained from this equation:
End Groups per = absorbance X CF _
106 carbon atoms film thickness
where film thickness is in millimeters (~0.003mm).
Some of the absorbance peaks may interfere
with one another when -C02H(D), -C02H(M), and
-CF=CF2 ends are all present. Correttions have
been developed for the absorbances of -C02H(D)
(hydrogen-bonded carboxylic acid dimer) and the
-CF=CF2 endsO ~hese are as follows (where ~ is
wavelength in micrnmeters):
absorbance 5.4~ 0.3 x absorbance 5.58~~corrected
for -C02H(D)
35 absorbance 5.57~ - 10.3 x absorbance_5~5B41 corrected
0.91 ~bsorbante
for -CF=CF2
-16-
2~a~
~17-
The presence of -CONH2 or -C02CH3 may
also interfere with the acid and -CF=CFz
~bsorbances. Since these groups are generally the
result of additives to polymeriza~ion their presente
is generally predictable. A SU5pi Ci on of -CONH~
~bsorbance in the vicinity of 5.6 micrometers can be
checked by searching for the auxili~ry -CONH2 band
~t 2.91 micrometers.
The polymer films (0.2~ to 0~30mm thick) are
scanned on a Perkin-Elmer*~83B spectrophotometer ~ith
a film of the same thickness, and known to con~ain
none of the ends under analysis, ln ~he instrument
reference beam. ~he instrument is set up with a
Response Time setting of 1, ~ Scan Time setting of 12
minutes, Ordinate Expansion of 2, ~ Slit Program of
7, and an Auto-Chek 6ain control of 20X. The films
are then scanned through the pertinent regions of the
spectrum making sure that adequate base lines are
established on each side of the pertinent absorbances.
The polymer films are generally compression
molded ~t 270-350C~ The presence of certain salts,
particularly alkali metal salts, may cause end group
degradation within th~s temperature range. If these
salts are present, the films should be molded at the
lowest possible temperature.
HEXAFLUOROP~OPYLENE (HFP) CON~EN~ DE~ERMIN_~ION
The HFP content ~n the melt-processible
TFE/HFP copolymers described herein is determined by
measurement of the ratio of the infrared absorbance
at 10.18 micrometers to the ~bsorbance ~t 4.25
micrometers. This ratio is referred to as th~ HFP
*denotes trade mark
-18-
index or ~FPI. Re~erence ~ilms of known HFP con~en~,
as determined by Fl9 NMR, are also run ~o calibrate
~he HFPI. The mole perc~nt HFP presen~ is equal to
2.1 times the HFPI. Co~pression-molded films
approximately 0.10 - 0.11mm thick are scanned under a
nitrogen atmosphere~
PERFLUOROPROPYLVINYL ETHER (PPVE~ CONTE~T
The PPYE content in ~he melt-processible
TFE/PPYE copolymers described herein is also
determined by infrared spectroscopy. The ratio of
absorbance at 10.07 micrometers to that at 4.25
micrometers is determined under a nitrogen atmosphere
using films approximately 0005 mm thick. The films
are compression molded at 350C, then immediately
quenched in ice water. This absorbance ratio is then
used to determine per~ent PPVE by means of a
calibration curve established with reference films of
known PPVE content. Fl9 NMR is used as the primary
standard for calibrating the reference films.
AVERAGE PARTICLE SIZE
, . _ . ~
U.S. Patent 3,929,721 describes a dry-sieve
analysis procedure. The "average particle si2e" is
determined by a dry-sieving procedure in ~ccordance
with ASTM Procedure D-1457-81a (12.3.3) modified as
follows. The 203mm diameter sieve set is assembled
in order, with the largest m~sh opening on top. From
the listing of U.S.A. Standard Testing Sieve si es
~ASTM E-ll Specification), four to eight ~djacent
sieves are selected with openings between the limits
of 6 mesh and 200 mesh and which bracket the range
into which the average particle size is expected to
f a 1 1 .
- 1 9
A 40 to 609 represent~tive p~r~ion o7 the
5 sample t~ be tested, preferably obtained using a
sample spli~ter and weighed to the nearest 0.~19, is
charged to the top screen. ~he screen se~ ~s shaken
in a sieve shaker (typically a .~o-Tap* shaker
obtained from Fisher Sc~entific Co., Çat. No. 4-90~)
for a~out 10 minu~es. After shaking, the net weight
of material retained on eath sieve ~s determ~ned ~o
the nearest 0.019.
The weight average particle size ls
determined based on plotting the cumulative
percentage retained vs. sieve opening on
log-probability paper as described 1n ASTM meShod
D-1921-63, or by equivalent computer ~nterpolation of
these data. The average particle s~ze in micrometets
is read from the plot at the 50th percentile (D~0)
abscissa of cumulative weight percentage retained.
DSC A~ALYSlS
_
DSC analyses are carried out w~th a Du Pon~
Model lO9D Thermal Analyzer using a ~odel 910 DSC
tell hase and the Du Pont General Analysis Program,
Yersion l.D. The )nstrument is calibrated as
recommended by the manufacturer, using a lOmg indium
standard. Polymer sample si2e 1s 6 to 10 mg, trimped
into an aluminum capsule. Different heating and
tooling cycles are used depending upon the melting
point of the sample. Samples are scanned twice
across the melting endotherm at 10C per minute from
a temperature which is 90 ~ 5C below ~o
temperature 40 ~ 5~C above the melting endotherm pea~
*denotes trade mark
- 1 9 -
-20-
temperature. Between the first and second scanning
of the endo~herm, ~he sample is cooled from the
maximum ~o ~he minimum scan temperature at a rate of
10C/min. The "melting endotherm peak temperature"
is defined dS the peak temperature of the first
melting endotherm, ~he heats of melting (Hl ana
H2) are calculated from the first and second
melting scans, respectively. The "melting heat
ratio" (Hm ratio) is defined as Hl/H2. The
melting heats Hl and H2 are determined by
instrumental integration using a base line from 80C
below to 30~C above the peak temperature. The "melt
S onset temperature" is determined graphically by
plotting the derivative of the firs~ meltin~ scan
using the Du Pont General Analysis Program, Yersion
1.0~ It is defined as the temperature where the
expanded derivative curve first increases above the
~ero base line (on the low temperature edge of the
melting curve) hy 0.2 m~/min.
Example 1
A tetrafluoroethylene/hexafluoropropylene
(TFE/HFP~ copolymer, 7.6 mole X HFP, in aqueous
dispersion form was obtained by polymerizing TFE and
HFP in an aqueous medium according to the general
procedure of U. S. Patent 4,380,618 using potassium
and ammorium persulfate initiators and ammonium
perfluorocaprylate surfactant. The copolymer was
coagulated by using 125U ml dispersion (26.4X solids)
diluted with ~00 ml of demineralized water in a
3.5-liter stainless-steel beaker (152 mm in diameter)
equipped with four equally spaced, rectangular
baffles protruding 13 mm into the beaker. The
3~ ~gitator impeller had four 34mm x 17 mm x 3,2 mm
thick blades welded onto a 17mm diameter hub at 35 to
40 pitch from hori20ntal to pump upward when rotated
-20-
-21-
clockwise. Impeller diamet~r was 85mm. Th~ contents
were agita~ed ~t 900 rpm and 30D ml of 70 weight X
nltric acid was then added to produce a thick gel,
After 3 minutes, 160ml of Freon* 113 was added to
break the gel and granula~e the polymer. Agi~ation
was stopped 5 minu~es later. The aqueous phase was
poured off, I00~ ml of ~eminerali2ed wa~er ~as added,
and the polymer agitated for 5 minutes at 500 rpm.
ThP aqueous phase was again poured off and the
poly~er ~as dried ~n 2 150C air oven for 4 hours.
This overall procedure was repeated three more ~imes
to obtain a total of 1500 9 of polymer (melt
viscosity 6.2 x I04 poise at 372~C). This
copolymer was screened on a 30-mesh sieve to remDve
~ines and yield a product with a D5U of 1210
micrometers and a sphere factor of I.33~ About IOOOg
of this polymer was divided into eight essentially
equal samples using a sample splitter. Seven of
these samples were baked in ~n air oven 3t various
conditions to harden the granules. The eighth sample
was left unbaked as a control. The attrition factors
measured on 211 eight s~mples are given below.
Bakin~ Conditions
Attrition
SampleTim~hrs TemperatureC Fattor
I 2 222 35.5
2 2 233 I3.I
3 4 ~33 4.7
4 2 239 6.
4 23~ 2.2
6 2 24S 3.9
7 4 245 I.8
Unbaked control 92.3
*denotes trade mark
.
-21-
-2~-
All the temperatures for Samples 1 through 7
are between 25C below the DSC melt onse~ temperature
and the melt endotherm peak temperature.
Two samples of this polymer (125g each after
screening to removP fines), one which had been baked
at 239C for four hours ~o harden the granules, and
the second which was not baked, were fluorinated in a
stainless-steel shaker tube for 4 hours at 160C
using a 25X fluorine in nitrogen atmosphere at
0.69 MPa gauge pressure. Total processing time was
iust over 5 hours. These samples were screened on a
30-mesh sieve to determine the amount of fines
generated in the shaker-tube treatment with the
following results:
X Fines
Sample Attrition Genera~ed
Factor_ (through 30 mesh)
Unbaked 92.3 6.1
Baked at
239C 2.2 0.5
DSC data were as follcws:
Before Baking After Bakiny
Peak Temperature 262C 263C
Melting
Heat Ratio 1.45 1.56
Melt Onset
Temperature248C 244C
The dried polymer had 440 unstable end
groups per lD6 carbon atoms. After fluorination no
unstable end groups were detected.
Example Z
A TFEtHFP copolymer (5,9 mole ~ HFP) was
polymerized at 3.1 MPa gauge pressure ~nd 95~C with
ammonium perfluorocaprylate dispersing
-22-
agent and ammonium persul~ate initiatorO ~he
resul~ing dispersion (l9.OX polymer) was coagula~ed
similarly t~ that of Example 1. Per 100 parts of
copolymer on a dry basis, 6 parts of 60 weight %
ni~ric acid and 93 parts of Freon(tm) 113 were used.
The polymer was washed several times with
deminerali~ed water to remove the nitric acid~ ~he
Freon(tm) was boiled off by a warm water 160C) wash
under slightly reduced pressure. The polymer was
separated from the aqueous phase and dried~baked in a
220C circulating air oven for 8 hours. Analysis
showed the presence of 448 unstable end groups per
carbon atoms consisting of -COF, -C02H(M),
and -C02H(D).
A 22.7-kg portiDn of the baked granules was
treated with fluorine at 190C for three hours while
being tumbled in a vessel described as follows. The
~luorination reactor was a 0.1 m3 double-cone
blender provided with gas inlet and vent connections
and an electric heating mantle. The gas inlet dipped
down into the particles and the vent pointed up into
the vapor space. Both lines were fixed and remained
stationary when the blender was rotated. The polymer
granules were placed in the reactor which was then
sealed and rotated at S rpm. The polymer ~s heated
by both the electric mantle and a preheated ~ir
stream flowing through the reactor. ~hen the polymer
reached the desired temperature, the air flow was cut
off and a vacuum was applied. The pressure was
raised to a~mospheric with a mixture of
fluorine/nitrogen (25X/75X by volume) ~nd this
mixture was fed through the reactor continuously for
three hours while maintaining the temperature with
the electric mantle heater. The gas feed was then
switched to 100~ nitrogen until no fluorine was
-23-
-2~-
detect2d in the off-gas using mois~ened starth-iodide
paper. The resin was ~hen cooled with old air
passed through the reactor. ~,e reac~or ~as ~hen
opened and the resin was collec~ed. ~he granules had
the following properties:
Melt Yiscosity . 12.6 x 104 poise
~t 372CC
Average Particle
Size (D50) 14~0 micrometers
Attrition
Factor 54.4
Sphere Factor 1.16
lS Unstable Endgroups p r
106 Carbon Atoms 21
DSC Melting Heat
Ratio 1.60
The fluorinated granules were fed to a 32-mm
diameter Waldron-Harti~ extruder ~ith a 20:1 L/D
barrel and coated onto AWG #20 19/32 stranded copper
conductor with an insulation thickness of 0.25mm. No
electrical flaws ~ere detected ~n the coating at
either of two extruder ~emperature profiles~ The
coated ~ire had a dielectr~c strength of 69 kV/mm
(ASTM D-149).
~e~
An aqueous dispersion of tetrafluoroethylene
30 (TFE) with 1.3 mole X perfluoropropyl vinyl ether
(PPVE) copolymer was prepared in ~ctordance with U.S.
Patent 3,635,926. This dispersion, cont~in~ng 26.9
weight X copolymer~ was obtalned by polymerizing the
monomers using ~mmonium persulfate lnitiator,
35 ammonium perfluorocaprylate surfactant and ethane
cha~ n-transfer agent i n the presence of ~mmoni um
hydrvxide pH modifier ~nd Freon(tm) 113 as a
water-immiscible phase.
*denotes trade mark
-24-
-Z5-
The above TFE/PPVE copolymer dispersion ~as
coagulated at 35C by a method similar to that of
Example 2. With agitation, 5.8 parts of 60 weight X
nitric acid and 85.5 parts of Freon(tm) 113 per 10~
parts by weight of copolymer (dry basis) were addedO
The resulting granules were washed , with
agi~ation, 3 times with 20-~5C demineralized water,
followed by a wash heated to 60C to remove the
treon(~m) 113, ana by a final water wash at 20-40C.
The resulting polymer was separated from the wash
water and dried at 180C for 6 hours in a circulating
5 air oven. The soft granules were characterized 2S
0 1 1 OW S .
Average Particle Size (D50) = 360 micrometers
Attrition Factor = 81.8
Sphere Factor = 1.18
Melt ~iscosity - 3.9 x 104 poise at 372C
PPYE Comonomer Content = 1.3 mole X
Melting Heat Ratio = 1.53
Melting Endotherm Peak Temperature = 311C
Melt Onset Temperature o 287DC
Bubble Index = 26
The infrared scan showed 93 amide ends per
10~ carbon atoms and a few vinyl and/or carboxylic
acid ends per 10 carbon atoms.
The resin was heat hardened at ~bout 285C
~or three hours and the gr~n~les screened through a
20-mesh screen. They were characterized as follows.
Average Particle Size (D5U) - 340 micrometers
Attrition Factor = 3.1
Melt Viscosity = 7.9 x 104 poise at 372C
3~ PPVE Comonomer Content ~ 1.3 mole X
Melting Heat Ratio = 1.59
Melting Endotherm Peak Temperature = 311C
Melt Onset Temperature ~ 2B9C
Bubble Index 66
-26-
The much reduced at~rition ~actor shows a
marked improvement in the hardness of ~he granules.
Infrared analysis showed 88 amide ends and a ~ew
vinyl or carboxylic acid ends per 1U6 carbon a~oms.
The resin was fluorinated using a high-
pressure stainless-steel cylindrical batch reactor,
equipped with gas and vacuum connections, electric
heaters and shaker-type agitation~ Polymer granules
were charged and the vessel was sealed. A vacuum was
applied followed by pressurization to 1 MPa gauge
pressure with a mixture of fluorine/nitrogen (25X/7SX
by volume) at 190C. The total processing time
including start-up, ven~ing and cool-down was just
over 5 hoursO The granules were heated in a
circulating air oven for over an hour to remove
traces of fluorineO Particle integrity was
preserved. The granules were characterized as
follows~
Average Particle Size (D50) = 285 micrometers
Attrition Factor = 6.3
Melt Viscosity = 7.5 x 104 poise at 372~C
Melting Heat Ra~io = 1.6D
Melting Endotherm Peak Temperature = 311C
Melt Onset Temperature - 291C
Bubble Index = 15
-26-
Infrared analysis sho~ed ~ha~ fewer than 50
unstable end groups per 106 carbon atoms were
present a~er fluorination.
Example 4
By a procedure similar to that of Example 3,
a heat-hardened resin was ~btained. The granules
were fluorinated as follows: An amount of polymer
granules corresponding to about one-fourth of the
reactor capacity was sealed ~nside the reactor o~
Example 2 and fluorinated for four hours at 185C to
189DC using a reactor rotation speed of 5 rpm. After
fluorination, the granules were redueed in size by
forcing them through a U.S. 30-mesh, 203-mm diameter
sieve on a Fritsch Pulverisette(tm) shaker, Twelve
stainless-steel balls l9mm in diameter were placed on
the screen and vibrated until all the material,
except for 306X of very hard partic1es which were
discarded, had been forced through the screen.
The granulated resin was characterized as
follows-
Average Particle Size (D50) = 337 micrometers
Attrition Factor - 4.1
Sphere Factor ~ 1.13
Melt Viscosity - 8.0 x 104 poise at 372C
PPVE Comonomer Content - 1~2 mole X
Melt~ng Heat Ra~io = 1.56
3 Melting Endotherm Peak Temperature = 311C.
Melt Onset Temperature = 289C
Infrared analysis indicated no detectable
unstable end groups.
-28-
The Bubble Index on ~his sample was 11~
eompared to 45 for a non-heat-hardened, unfluorinated
control sample.
A comparison was made of the rotolining
performance of heat-hardened, fluorinated granules
and unfluorina~ed control granules, using a 3-inch
flanged pipe tee as a mold. A 6479 quanti~y of the
resin was placed inside the mold. The mold was then
mounted in a fixture on one of the arms of the spider
of a McNeil-Akron rotocasting machine of the type
described in U.S. Patent No. 4,3I~,96l. The machine
was indexed to advance the arm ~nto the oven. The
mold was rotated by the fixture about mutually
perpendicular axes to cause the resin to tumble and
contact all interior surfaces of the mold. The
major/minor axis speeds were B/9 rpm respectively.
The mold and its contents were heated for 15 minutes
at 329C before the temperature was raised to 352~C
for the processing times shown below.
On tompleting the heating cytle, the spider
arm indexed to a cooling station. ~hile continuing
its rotation, the part was cooled in sequence by an
air stream for ~ minutes, by a water spray for 12
minutes, ~nd again by air for 2 minutes.
Th2 rotocasting machine was ~hen indexed to
bring the finished part to the unloading station
where it was removed. The lining of the finished
part was inspected for bubbles or other porosity.
The fluorinated resin of this 1nvention gave a
smoo~her surface than the control as shown in the
table below.
-~8-
~ ` ~z~
-29-
Heat-
Protessing Resin Rotomolded Part
Conditions ~ Observations _ _ _
110 min, 352C Fluorinated Bubble-free
110 min, 352C Unfluori- Many small lumps,
nated bubbles in side neck
1 80 min, 352C Fluorinated Bubble free
8U min9 352C Unfluori- Many small lumps
nated throughout part;
many bubbles in
middle of wall
~0
-29-
