Note: Descriptions are shown in the official language in which they were submitted.
This invention concerns an improved process for the
preparation of crosslinked vinyl copolymers as discrete copolymer
beads in aqueous dispersions using oxygen in the polymerization
in a novel way. The invention also concerns the ion exchange
; resins having improved physical characteristics obtained by ap-
pending conventional ion exchange functional groups to said co-
polymers.
The techniques of preparing crosslinked vinyl copolymers
in bead form (as precursors for conversion into ion exchange re-
sins) by free-radical catalyzed polymerization of the monomer
mixture in aqueous dispersion are well known. The term "cross-
linked vinyl copolymer" and the like is used for the sake of
brevity herein to signify copolymers of a major proportion, e.g.,
from 50 upwards to about 99.5 mole percent, normally 80 to 99~,
of a monovinyl monomer, preferably, monovinyl aromatic monomers,
e.g., styrene, vinyl toluene, vinyl naphthalene, ethyl vinyl
benzene, vinyl chlorobenzene, chloromethyl styrene, and the like,
with a minor proportion, e.g., of from about 0.5 up to 50 mole
percent, preferably 1 to 20~, of polyvinyl compounds having at
least two active vinyl groups polymerizable with the aforesaid
monovinyl monomer to form a cross-linked, insoluble, infusible
copolymer, for example, divinyl benzene, trimethylolpropane tri-
methacrylate, ethylene glycol dimethacrylate, divinyl toluene,
trivinyl benzene, divinyl chlorobenzene, diallyl phthalate, di-
vinylpyridine, divinyltoluene, divinylnaphthalene, ethylene glycol
diacrylate, neopentyl glycol dimethacrylate, diethylene glycol
divinylether, bisphenol-A-dimethacrylate, pentaerythritol tetra-
and tri-methylacrylates~ divinylxylene, divinylethylbenzene, di-
vinyl sulfone, divinyl ketone, divinyl sulfide, allyl acrylate,
diallyl maleate! diallyl fumarate, diallyl succinate, diallyl
carbonate, diallyl malonate, diallyl oxalate, diallyl adipate,
diallyl sebacate, divinyl sebacate, diallyl tartrate~ diallyl
~ $~
, . .
silicate, triallyl txicaxballylate, triallyl aconitr~te! txiallyl
citrate, triallyl phosphate, N~N~methylenediacrylamide, N,NI-
methylene dimethacrylamide~ N,N'-ethylenediacrylamide, trivinyl
naphthalene, polyvinyl anthracenes and the polyallyl and poly-
vinyl ethers of glycol glycerol, pentaerythritol, resorcinol and
the monothio and dithio derivatives of glycols. The copolymer may
also have incorporated therein polymerized units of up to about 5
mole % of non-aromatic vinyl monomers which do not effect the
basic nature of the resin matrix, for example, acrylonitrile,
methyl acrylate, butadiene and others known in the art.
The conventional conditions of polymerization used
heretofore lea~ to crosslinked vinyl copolymers, which, when con-
verted to ion exchange resins by attachment of functional groups
thereto, have certain operational deficiencies resulting from
physical weaknesses.
The practice of the present invention yields ion ex-
change resins in which the polymer beads have greater mechanical
; strength and increased resistance to swelling pressures which are
produced within a bead during acid/base cycling (i.e., osmotic
shock). The greater mechanical strength of the beads manifests
itself in improved resistance to physical breakdown from external
forces such as weight of the resin column bed, high fluid flows
and backwashingr Thus, the physically stronger ion exchange
resins embodied herein are especially useful in treating fluid
streams of high flow rates, for example, condensate polishing ap-
plications in which resins of lesser quality are prone to mechan-
ical breakdown and short life spans.
In the past, it has been the practice not to include
oxygen during the preparation of crosslinked vinyl polymers used
as the base matrix copolymer for ion exchange resins since oxygen
presents a safety hazard and has been generally regarded as detri-
mental to the properties of said polymer obtained by free-radical
polymerization.
In accordance with this invention, the yinyl monomer~
cross-linking monomert and other optional monomer or monomers,
are polymerized in an aqueous dispersion in the presence of a
free-radical initiator and (1) in contact with an oxygen-contain-
ing gaseous mixture, (2) using a pre-oxygenated monomer mixture
or (3) both (1) and (2), advantageously within a range of reaction
temperatures of from about 30 to about 95C., preferably 50 to
70C. Thus, in order to improve absorption of oxygen by the mono-
mer mixture, it is generally preferred to employ polymerization
temperatures somewhat below, e.g.,5-25~C., those normally used hereto-
fore in suspension polymerization for similar products. Accor-
dingly, the free-radical initiator used herein is one suitable
for catalyzing polymerization at such temperatures, for example,
such initiators as di(`4-t-butylcyclohexyl) peroxydicarbonate, di-
cyclohexyl peroxydicarbonate, di-(sec-butyl) peroxydicarbonate,
- di-(2-ethylhexyl) peroxydicarbonate, dibenzyl peroxydicarbonate,
diisopropryl peroxydicarbonate, azobistisobutyronitrile), azobis-
(2,4-dimethylvaleronitrile), t-butyl peroxypivalate, lauroyl
peroxide, benzoyl peroxide, t-butyl peroctoate, t-butyl peroxy-
isobutyrate, and the like. The amount of initiator employed is
normally from about 0.1 to about 2 percent, based on monomer
weight, preferably 0.3 to 1%. It also may be advantageous when
using catalysts which are active at relatively low temperatures,
such as 30-60C., to employ a second so-called "chaser catalyst"
which is active at higher temperatures in order to achieve higher
yields of crosslinked vinyl polymer, for example, from about 0.05
to 0.1~, based on monomer weight of such initiators as benzoyl
peroxide, t-butyl eeroctoate, t-butyl peroxyisobutyrate, and the
like.
As mentioned above, the process of this invention in-
volves contacting the monomer mixture with oxygen such that oxy-
gen is absorbed by the monomer mixture at least until the poly-
; - 4 -
r
. `'`~
~.~.2~ 6
merization reaches the gel point~ i~e., the point at which an
infinite polymeric network occurs (see, for example, Fundamental
Principles of Polymerization by G. F. D'Alelio, John Wiley & Sons,
Inc., 1952, page 93). Known procedures for involving gaseous
reactants in polymerization systems are used to incorporate the
oxygen in the monomer mix. For example, the head space above the
reaction medium is purged with an oxygen-nitrogen gaseous mixture
(prior to initiation of reaction by raising the temperature) and
then a gaseous sweep of the appropriate O2-N2 mixture is passed
thxough the head space during the reaction period. The gaseous
mixture may contain as much as 20~ oxygen, however, for purposes
of safety in the avoidance of explosion-prone conditions, lower
levels may be required depending on the explosive range of the
` mixtures of the specific vinyl monomer or monomers with oxygen
in the vapor phase, e.g., less than 9% oxygen in the case of a
styrene and divinylbenzene mixture. Since the absorption of
oxygen by the monomer droplets depends not only on temperature
and the partial pressure thereof in the gas in the head space,
but also on the area of reaction medium exposed to said head
space, the configuration of the kettle will determine whether it
is advantageous to operate at atmospheric pressure or under in-
creased pressures, for instance up to five or more atmospheres,
; inasmuch as increased pressure causes greater oxygen absorption.
An alternative method of introducing oxygen into the monomer mix-
ture is to sparge the gaseous mixture into the monomer mixture
before and/or during polymerization.
The aqueous medium in which the polymerization is con-
ducted in dispersion form will contain minor amounts of the con-
ventional suspension additives, that is, dispersants such as
xanthan gum (biosynthetic polysaccharide) r poly (dially dimethyl
ammonium chloride), poly acrylic acid (and salts), poly acryl-
amide, magnesium silicate, and hydrolyzed poly (styrene-maleic
~S7
.
anhydride); ~xotective colloids such as carboxymethyl cellulose,
hydroxyalkyl cellulose, methyl cellulose, polyvinyl alcohol~
gelatin, and alginates; buffering aids such as phosphate and
borate salts; and pH control chemicals such as sodium hydroxide
and sodium carbonate.
The crosslinked, high~molecular weight copolymers are
recovered from the reactor as hard, discrete beads of particle
size within the range of about 0.02 to 2 mm, average particle
size being on the order of 0.2 to 1 mm. These copolymers are
converted to ion exchange resins by functionalization according
to known means, such functional groups ineluding sulfonamide,
trialkylamino, tetraalkyl ammonium, earboxyl, earboxylate, sul-
fonie, sulfonate, hydroxyalkyl ammonium, iminodiaeetate, amine
oxide, phosphonate, and others known in the art. Funetionalizing
reaetions whieh may be performed on vinyl aromatie eopolymers to
produce ion exchange resins are exemplified by sulfonation with
coneentrated sulfurie acid, chlorosulfonation with chlorosulfonic
aeid followed by amination, reaetion with sulfuryl chloride or
thionyl ehloride followed by amination, and ehloromethylation
followed by amination. Ion exehange resins may be further de-
lineated by the types: strong aeid eation, i.e., eontaining the
; groupings sulfonie (~S03H) or sulfonate (-SO3M, where M is usually
an alkali metal ion); weak aeid eation, i.e., containing the car-
boxyl (-CO2H) or carboxylate (-CO2M, where M is usually an alkali
metal ion) groupings; strong base anion, i.e., containing the
tetraalkyl ammonium grouping: -NR3X, where R is an alkyl or hy-
droxy alkyl group and X is usually chloride or hydroxide; and
weak base anion, i.e., containing a trialkylamino group: -NR2,
where R is an alkyl or hydroxyalkyl group.
The uni~ue properties of the copolymers produced accor-
ding to this invention are reflected in their different charac-
teristics under thermal analysis and solvent swelling and when
- 6 -
.
converted to ion exch~nge resins by the attachment of the afore-
said functional groups. The enhanced physical strength of these
latter resins is apparent from their resistance to crushing which
is conveniently measured on the Chatillon instrument, as well as
by visual inspection before and after use in ion exchange appli-
cations. For exampler strongly acidic styrene-type resins fre-
quently exhibit Chatillon values in the range of about 900 to
about 5000 gm. force per bead, preerably 1200~5000, in contrast
to resins derived from copolymers prepared by prior art polymeri-
zation methods which have Chatillon values in the range of about
50 to 550 gm./head. Similarly, anion styrene-type resins of the
invention exhibit Chatillon values of about 500 to 2500 prefer-
ably 600-2500 and often in the 900-1500 range in contrast to
resins derived from copolymers prepared by prior art methods
which have typical Chatillon values of 25 to 400.
The improved gel ion exchange resins of the present
invention~ particularly the most common commercial resins pro-
duced from aromatic copolymers, can be easily distinguished from
the prior art resins by one or more of various physical para-
meters including (1) perfect bead count (fewer cracked and frag-
mented beads~, (2) resin friability (Chatillon test), (3) resis-
tance to fracture upon repeated cycles of exhaustion/regeneration
~`:
(Microcycling Test)~ and (4) the birefringence patterns of the
beads. Test methods and observations of these distinguishing
characteristics follow.
IN THE DRAWINGS
.
Figure 1 illustrates the birefringence pattern of
cation exchange resins made in accordance with the present
invention,
~ 30 Figure 2 represents photomicrographs of birefringence
; pattern cation exchange resins made in accordance with several
methods existing in the prior art;
-- 7 --
~:
.2~
~ igure 3 represents photomicro~raphs of birefringence
patterns of anion exchange resins prepared in accordance with the
present invention except for ~igure 3D which is a comparison be-
tween present resins and the prior art resins; and
Figure 4 represents photomicrographs of anion exchange
resins available from competitive prior art sources.
In some instances, prior art resins have exhibited high
physical stability by one or more of tests (1)-(3) above, but
have failed to achieve excellence in all three criteria. In the
cation resins, about 75% to 90~ of the resins used commercially
have intermediate levels of crosslinker, that is about 4-12%
usually 7-10% crosslinker (preferably divinylbenzene, DVB). The
most common anion resins, from a commercial standpoint, are those
containing relatively low levels of crosslinker, that is, about
1~10% usually 2-5% crosslinker. The improved products of the
invention result with all levels of crosslinker, although illus-
trated herein principally with the most common types. The dif-
ferences between the birefringence patterns of the novel resins
disclosed herein and similar resins of the prior art may be less
pronounced at lower crosslinker levels where internal residual
resin stress is a less significant factor. The improvement can
; nevertheless ~e ascertained when comparing the novel resins with
the same type (and crosslinker content) resins of the prior art
in either a relaxed or artificially stressed state (e.g., in
swelling solvents~.
; PERFECT BEAD COUNT
., ~
Perfect bead count is determined microscopically after
functionalization of thR copolymer such as by sulfonation or
chloromethylation and amination of the copolymer. Perfect beads
are those which contain no visible flaws, that is, beads which are
perfectly spherical with no cracks, fragments, pits or surface
defects. Products of this invention contain at least 90% or more
- 8 -
~'
of perfect beads, typically 23~99% perfect beads, by visual ob-
servation and count. Prior art resins typically contain about
40-99~ perfect beads. However, many grades of commercial resins
typically have perfect bead counts of only 40-50~ (e.g., see~
Figure II, C and D described below).
ACID/BASE C~CLING (~MICROCYCLING) TEST
Microcycling is designed to simulate on an accelerated
time scale the conditions under which the resin will be used.
These studies are conduct~d over a period of a few days rather
than months or years typical of field conditions. Repeated ex-
haustion-regeneration cycles are performed on the resin at pre-
determined intervals in a fully automated apparatus.
The resin to be tested is screened to a -20 30 U.S.
mesh cut size and examined under a microscope for appearance
before microcycling: four different fields of view of a mono-
layer of beads are observed and the average result for each of
'~ the following is recorded:
(a) % perfect beads
(b) % cracked beads
2Q (c~ ~ fragmented/broken beads
A small portion of the screened resin (0.5 ml) is
placed in a sintered glass filter tube such that a monolayer of
beads is formed in the tube. This small quantity of resin beads
' assures good contact between solution and resin and total conver-
sion of the resin during each step. The solutions used for ex-
haustion and regeneration are made up in advance and stored in
50 liter tanks. The solutions used for anion and cation resins
are described below:
Resin Type Exhaustion Solution Regeneration Solution
.
Anion 0.25 N H2SO4 1.0 _ NaOH
Cation 0.5 _aOH 1.0 N HCL
_ g _
~ ~Y ~ '3 6
; During a typical experiment, approximately 200 ml of
- exhaustion solution is added dropwise to the resin sample over
10 minutes, followed by removal of bulk exhaustant by mild vacuum,
a deionized water rinse followed by mild vacuum, and dropwise
addition of regenerant solution over 10 minutes followed by re-
moval of bulk regenerant by mild vacuum and a water rinse; com-
pletion of the aforementioned process represents an exhaustion-
regeneration cycle and requires approximately 30 minutes. Com-
plete automation allows 100 cycles to be completed in about 48
hours. After completion of 100 cycles, the resin is recovered
and inspected microscopically for appearance. The reduction in
% perfect bead content is recorded as the breakdown.
The product of the invention generally show a reduction
of perfect bead count of less than about 30%, normally not more
than about 15% after 100 cycles by the Microcycling Test. The
cation resins exhibit less reduction of perfect beads, not more
than 10%r and usually 0-5%. Anion resins may show reductions of
up to 30%t normally 0-15%. By comparison, prior art cation resins
" are known to exhibit reductions of from 15-80% most typically
30-50%. Anion resins available heretofore show perfect bead re-
ductions of 15-80% after 100 cycles with 15-50 being most typical.
CHATILLON TEST FOR RESIN FRIABILITY
The Chatillon test is named for an apparatus manufac-
tured by John Chatillon and Sons, New York, N. Y. and designed to
measure resin friability. This instrument (MODEL LTCM, Gauge
DPP-2.5KG) measures the force (grams) required to crack or frac-
ture a resin bead when it is placed between two parallel plates.
The plates are gradually brought together at a uniform rate until
the resin "breakpoint" is reached. The purpose of this test is
to simulate the frictional and pressure forces exerted on indi-
vidual resin beads under actual use conditions.
Specifications for testing include converting the
resin into the proper form (hydrogen or sodium for cation resins
-- 10 --
.~ .~,.
tested hexein and chloride form for anion resins tested hexein)
by well known standard procedures. The converted resin is
screened to a -20 + 30 U.S. mesh cut size and then allowed to
fully hydrate in deionized water for at least 15 minutes prior
to testing. Actual testing is done on a single resin bead
(covered by a small drop of water) in the Chatillon instrument
using the lowest practical speed of descent of the crushing
plate. The individual fragmentation forces are recorded from the
instrument in grams per bead and the results are presented as an
- 10 average (20 beads minimum, typically 30 beads), a standard devia-
tion, a 95% confidence interval, and the percentage of beads
which meet a minimum friability standard.
BIREFRINGENCE TEST
An analytical test which aids in identifying gel resins
of the invention and generally distinguishing the same from prior
art counterparts is the birefringence test. The technique for
obtaining birefringence patterns involves the use of an optical
microscope (e.g., Carl Zeiss Photomicroscope) set up for bright
field illumination at law mganification (e.g., 34X). Polarized
lenses are inserted above and below the microscope stage and
oriented perpendicular to one another. A piece of frosted glass
is mounted on the stage to provide diffuse illumination of the
samples. Approximately 30~50 beads of the sample resin to be
analyzed are then placed in the concave well of a deep-dish micro-
scope slide. The well is filled with water and then covered with
a large coverslip. The slide so prepared is placed on the frosted
glass mounted on the stage, the focus adjusted to optimize defi-
nition of the outer edge of the beads, and a photomicrograph is
made to illustrate the birefringence pattern.
Observations of a large number of birefringence patterns
taken of ion exchange resin samples produced by the invention and
comparison of the same with patterns of contemporary commercial
-- 11
.~.
- resins produced by the various manufacturers have revealed clear
distinctions between the patterns. Both cation and anion resins
are distinguishable from prior art counterpart resins, but on a
somewhat different basis and therefore the cation and anion re-
sins shall be descr~ed separately.
CATION RESINS
The use of birefringence strain patterns to identify the
stresses in cation resins is not new to the art of ion exchange
(see, for example Wheaten R. M., et al., Industrial and Engineering
10 Chemistry, Vol. 44, No. 8, August 1952, pp. 1796-1800). We have
now further identified a number of characteristic patterns and
empirically correlated such with the physical properties of the
resin including residual internal bead stress so as to obtain a
qualitative identification of resin origin and properties. Migh
internal residual stress in the resin beads has been found to
correspond directly with low physical stability. The cation, low
stress resins of this invention are generally illustrated by
patterns A-D in Figure I appended hereto while some of the most
- widely used resins presently available from various manufacturers
are illustrated by patterns A-D in Figure II. The photographs
showing the birefringence patterns were made at essentially the
same film exposure and sample illumination.
In general, the patterns showing the highly stressed
products (and therefore those more susceptable to fracture or
breakage) can be identified by the brightness and sharpness of
the pattern as well as the pattern type. Referring to the appen-
ded Figures, the more highly stressed beads are found in Figure II
` which contains sharper, brighter individual bead patterns on an
overall basis. Note that this observation and other observations
described herein are often made on overall or gross appearance
of a sample owing to differences between the individual beads
in a single batch or sample. Further, it is postulated that
:~ - 12 ~
some of the commercial samples observed consist of composites or
physical mixtures of materials produced under different condi-
tions and therefore the patterns may illustrate the sensitivity
of the product quality to variations in the process of prepara-
tion.
A Maltese cross, or some variation thereof, is typic-
ally observed in resin strain birefringence patterns and is in-
dicative of spherically symmetric stress or orientation. Any
application of physical stress on an ion exchange resin bead
produces a strain pattern, typically a Maltese cross. This
phenomenon may be observed when compressing a relatively unstressed
bead between parallel planes and when inducing stress through
osmotic pressure such as when swelling a bead in solvent. The
width and sharpness of the arms of the Maltese cross furnish a
qualitative (sometimes quantitative) indication of strain. The
sharper, narrower arms indicating higher stress, especially when
accompanied by bright areas between the arms.
Applying the foregoing general considerations, the
strain birefringence patterns in Figures I and II can be distin-
guished. The low stress cation resin products illustrated in
Figure Ir A-D resins in hydrogen form have at least one of three
identifying patterns, namely:
(1) a broad Maltese cross enclosed by an extinction
(dark~ ring around the periphery of the bead (pattern
predominating in Figure I, A and B).,
(2) a broad Maltese cross enclosed by an extinction
ring ~hat is distinctly inside and separated from the
periphery of the bead (Figure I, A, B and C - perhaps
best observed in lower half of C), and
(3) an irregular pattern, sometimes resembling a ran-
domly oriented chain and sometimes recognizable as a
distorted version of (2), above (Figure I, patterns C
and D, but best observed in D).
- 13 -
Each of the patterns of Figure I r A thru C containing a ~altese
cross are relatively dull, with broad, and somewhat blurred arms
comprising the Maltese cross. While the pattern of Figure I, D
is less distinct~ it too is somewhat dull with more random stress
patterns, probably indicative of random stress orientation. All
of the patterns in Figure I are atypical of the prior art cation
gel resins which are illustrated in Figure II. All cation resins
in Figure I- were produced from a styrene/divinylbenzene (8~) co-
polymer backbone using oxygen moderated copolymerization tech-
niques described in the examples contained herein. The material
of Figure I, C is a composite of three laboratory-prepared samples.
All were sulfonated to produce strongly acidic resins.
Typical strain birefringence patterns for prior art gel
resins are illustrated in Figure II, A thru D, which patterns have
- at least one of three identifying criteria:
(a) A square superimposed upon a Maltese cross (see
~-~ Figure II, A),
(~b~ a sharp Maltese cross having narrow arms and
bright regions between the arms, with or without an
outer extinction ring (see Figure II, C and some beads
in B~, and
(c) an irregular pattern, sometimes resembling a
distorted cross (or swastika) and sometimes a square
superimposed upon a cross resembling (a), above
(Figure II, D.)
The pattern in Figure II, B represents a sample of relatively
high quality prior art sy$rene/DvB gel resin containing about 8%
DVB in the copolymer backbone (sample of manufacturers regular
product line). The pattern in Figure II, C represents a sample
of relatively poor quality prior art styrene/DVB gel resin con-
; taining about 8~ DBV, and exhibiting many surface defects and
poor physical siability by both the microcycling and Chatillon
- 14 _
~ ~L~
Tests described herein (sample obtained from m~nufacturerls regu-
lar product line). Another poor styrene/DVB resin containing
surface defects and bubbles and having low physical stability is
illustrated in Figure II, D (manufacturer's commercial product).
The sample from which the pattern of Figure II, A was produced
was a styrene/8% DVB gel resin of intermediate prior art quality
(manufacturer~s normal commercial product). All resins illus-
trated in Figure II were strongly acidic and in the sulfonic acid
form.
While individual beads in a given pattern in Figure I
may have strain patterns nearly the same as patterns in Figure
II, one may easily distin~uish the products on an overall basis.
To illustrate, a similarity may be seen between some beads in
Figure I, A or B and Figure II, B but a substantial number of
beads are dissimilar. The resins of Figure I are highly superior
to the res-ins of Figure II (even the best samples thereof) from
; a standpoint of the Chatillon test, perfect bead count and accel-
erated use testing (Microcycling Test).
On the basis of the above, and other studies of strain
birefringence patterns, it is postulated that the differences in
patterns between the new resins of the invention and those of
prior art resins reflect different levels of residual stress
within the resin beads. Although it is not intended that the
invention is dependent upon any theory expressed herein, the
patterns associated with the new resins are believed to represent
conditions of low internal stress, whereas those patterns asso-
ciated with the resins of Figure II, A-D, reflect higher levels
of internal resin stress. Since the stress which is responsible
for the birefringence pattern is believed to be the residual
stress within the bead, it follows that higher levels of stress
would be expected to correspond to poorer physical quality. Bire-
fringence patterns therefore offer a simple quantitative method of
identifying and distinguishing the products of this invention.
- 15 -
Copolymer precursoxs for the Cation and ~nion resinsmay also be distinguished from the prior art copolymers on the
basis of thermal analyses ana solvent swelling characteristics.
Since these copolymers result in improved resins, it is clear
that the copolymers are improved in composition over the prior
art copolymers.
-15 ~ 17-
z ~L~
ANIOr~ RESIriS
Anion exchange resins produced by the improved
copolymerization techniques described herein may also be
distinguished from prior art anion resins by strain birefrin-
gence patterns which correlate with improved physical
properties. In general it has been discovered that the anion
resins are distinguished principally on the basis of differ-
ences in the intensity of the birefringence patterns rather
than differences in the shape or nature of the patterns
themselves. Consequently, the experimental conditions must
be standardized as much as possible and a sample used as an
lntensity reference, in order to allow direct comparison of
; birefringence patterns from one day to the next. Normally,
it is preferable to focus the microscope on the outer edge
of the beads. Factors such as the intensity of the light
source, radiation lossesin the microscope, the positicn of
condensing lenses, the sensitivity of the film, and the
exposure time greatly influence the overall intensity of
the recorded image. However, for a given microscope, all
of these factors are adequately reproduced and given a sample
as an intensity reference, conditions from one day to anotner
can be matched satisfactorily to allow direct comparison
using photomicrographs.
The microscope and asgociated optics for obtaining
birefringence patterns of anion resi~s were the same as had
been used for the cation resins. However, the swelling
solvent in which the anion resins were examined was ethanol
rather than the water used for cation resins. Each anion
resin was oven dried at 90-100C for ca. 4 hours in vacuo,
equilibrated overnight under ambient conditions, and then
lmmersed in ethanol until swelling equilibrium had been
--18--
,: IB
Z~
achieved. All birefringence patterns of anion resins
presented were obtained from samples in the chlorlde form
- which had been swollen in ethanol at least 7 days.
. All of the anion resins, including those produced
by the novel oxygen moderated process described herein,
exhibit patterns which may be described qualitatively as a
broad Maltese cross having little or no extinction ring at
the periphery of the bead. However, when compared to the
prior art resins, the resins of this invention, swollen to
equilibrium in ethanol, exhibit patterns that are signifi-
cantly more intense (brighter). Intensity differences inwater are more difficult to characterize.
Figure III, A-C illustrate the birefringence
; patterns of different samples of the improved resins at 34X
magnification. The uniformity of pattern intensity and
configuration is typical of anion resins by this invention.
Figure III, D illustrates a composite sample of the new
resin (top of photo/bright) together with beads produced by
prior art methods (bottom of photo/dull) without oxygen
in the copolymerization process.
Figure IV, A-D illustrate prior art products pro-
duced by four different manufacturers. All of the photo-
micrographs of Figure III and IV were made at the same
exposure under conditions controlled, as explained hereto-
fore, to allow comparisons of the pattern intensities. Eventhe low-intensity patterns of Figure IV generally exhibit
broad Maltese crosses, with a few beads showing strong stress
patterns characteristic of the prior art cation resins.
The dominating characteristic of anion resins does
3 not appear to be the residual internal stress in the resin
'
--19--
. , .
.
beads as is the case with cation resins. However, the
intensity of the broad r~altese cross birefringence patterns
correlates directly and consistently with physical stability
of the resin, the brighter patterns serving as "fingerprints"
of the most physically stable resins. Whereas the dominating
feature of cation resins was postulated to be the internal
bead stress, applicant believes the greater swelling pressure
that sustains in superior anion resins is evidence of a
greater elastic component associated with the crosslinked
gel network, which makes it possible for the network to better
acoo~date an externally applied stress without craze or
crack formation. The more intense patterns of the new
resins swollen to equilibrium in ethanol is indicative of a
higher swelling pressure for the new resins.
It is recognized, of course, that certain selected
samples representing the extremes in pattern intensity for
new and prior art resins may be difficult to distinguish.
Also, a given bead in a pattern may deviate substantially
from the overall pattern of a prior art sample. However,
based upon data for a substantial number of products exam-
ined, ambiguities in determining the quality of a particular
product can be resolved by multiple analyses, preferably of
different lots of the same product. In some cases, both
for cation and anion resins of the p~ior ar~, the overall
patterns showing wide divergence in patterns and/or intensity
- are indicative of the quality of the overall product rather
than the presence of both good and poor quality beads within
a single sample. To illustrate,the birefringence pattern~ in
Figure IV, B and D, should be interpreted as "fingerprinting"
a typical prior art product which overall correlates with
':
B -20_
: .
,2r~L.~
physical stability inferior to the products of this inven-
tion, rather than as comprising both good (high brightness)
and bad (low intensity) beads. The bright patterns for a
few beads in those samples are indicative of high internal
stress and are to be excluded from a comparison of the
patterns.
The need for different interpretations for the
birefringence patterns between cation and anion resins is
thought to be a consequence of different inherent properties
between the cation and anion resins under study, owing
principal]y to compositional factors such as crosslinking
uniformity and the level of primary crosslinking which lead
to differences in the relative contributions of swelling
pressure and residual stress to the overall level of stress
in the cation resins swollen in water vs. the anion resins
swollen in ethanol. The importance of backbone elasticity
in the anion resins has, at least in part, been substantiated
by thermal mechanical analysis (TMA) above the copolymer
glass transition point where a secondary yield point has
been detected for both cation and anion copolyrners prepared
by the novel oxygen copolymerization method. Residual
stresses appear to be substantially less important in the
anion resins since the broad Maltese cross patterns that
typify both the resins of the improved technology and those
of the prior art are not suggestive of highly stressed beads.
Although it may be theorized that elasticity plays
an important part in the physical stability of both cation
and anion resin stability it has not been independently
-21-
4 ~
characterized with cation resins whose main "fingerprints"
are differences in pattern types.
Reactlon rate kinetic studies have indicated some
moderation of the crosslinker (DVB) reaction rate when
using oxygen-copolymerization compared to conventional
me~thods leading to a premise that the copolymer matrlx may
be more homogeneously crosslinked by the method of the inven-
tion, and explaining the improved elasticity of the resins.
All of the resins represented by Figures III and
IV contained low crosslinker levels typical of the most
widely used styrene/divinyl benzene type anion exchange
resins, that is, between about 2% and ~bout 5~ crosslinker
(DVB).
In the cation resins (Figures I and II, above),
about 75% to 90% of the resins used car ~ cially have inter-
; mediate levels of crosslinker, that is 7-10% crosslinker
(usually divinylbenzene, DVB). The most common anion resins,
from a commercial standpoint, are those containing relatively
low levels of crosslinker, that is, about 2-5~ crosslinker.
The improved products of the invention resuit with all levels
; of crosslinker, although illustrated herein principally
with the most common types. The differences between the
novel resins disclosed herein and similar resins of the
prior art may be less pronounced at~lower cross-
linker levels where internal residual resin stress is a less
significant factor. The improvement can nevertheless be as-
certained when comparing the novel resins with the same type
(and crosslinker content) resins of the prior art in either
a relaxed or artificially stressed state (e.g., in swelling
solvents).
-22-
LA`: ~
;
The process of the inventiPn is clarified by the
following illustrative examples which are not to be construed
as limitative thereof.
Polymerization Procedure
The polymerization reactor is a two-liter, three
neck, round bottom flask equipped with a two-blade paddle
stirrer, thermometer, condenser, heating mantle with
temperature controller, and provision for sweeping in a
blanket of a blend of oxygen and nitrogen. (Oxygen concen-
tration in t;he gas stream is monitored by gas-liquid chro-
matographic (GLC) techniques, and in the monomer mix is
checked by a Beckman oxygen analyzer).
The monomer phase containing initiator is charged
to the reactor, and the head space is swept with an appro-
priate gas mixture (e.g., 2% oxygen in nitrogen) until
equilibrium is reached at 25C.~ Then the aqueous phase is
charged and the stirrer is set at about 210 rpm to produce the
droplets of monomer in aqueous dispersion, while the gas
sweep is maintained. The following is representative poly-
merization reaction materials charge in grams.
Styrene 489.4
Divinylbenzene (54.7% conc.) 85.3
~ Methyl acrylate 8.8
; "Percadox-16" initiator ~ 2.04
(di(4-t-butycyclohexyl)
peroxydicarbonate)
Water 510.3
**
"Padmac A" dispersant20.1
poly(diallyl dimethyl
3 a.nmonium chloride)
.~ ***
"Pharmagel" protective1.6
c~lloid (gelatin)
Boric acid o.88
.
. * 'rradOEk
** Trademark
*** Trademark -23-
.~ .
r~
Sodium nitrite 0.59
Sodium hydroxide solution (50
conc.) added to pH lO - 10.5
The oxygen-nitrogen gas sweep is passed at 140 cc/
min. over the dispersion as it is heated from 25C to 57C.
in 45 minutes, then maintained at 57 + 2C. for 7 hours. The
batch is then heated to 75C. over a 30 minute period and
held at 75C. for one hour. The copolymer beads are washed
and excess water is removed by vacuum filtration on a Buchner
funnel.
Sulfonation-of Copolymer
A portion of the wet polymer beads prepared above
(llO gms.) is added to 600 grams of 95~ H2S04 in a one liter
flask equipped with stirrer, condenser, dropping funnel,
thermometer, caustic scrubber and heating means. Thirty grams
of ethylene dichloride (bead swelling agent) are added, and
the suspension is heated from 30C to 120C over a 3 hour
period. This is followed by a hydration procedure in which
water is added to quench the product. The polymer beads are
` 20 transferred to a backwash tower and backwashed to remove
residual acid. The resulting ion exchange resin product is
characterized by the following properties:
; Whole beads 99%
Cracked beads 2%
Fragmented beads 1%
Perfect beads 97
Friability: Chatillon
;~ value, g/bead 2139
~i Solids, H+ form44.7~
Solids, Na+ form51.5%
;~ Salt Splitting Cation
Capacity, meq./g dry 5.21
-24-
At ~ _ ~6
Additional crosslinked sytrene copolymers are
prepared as above with variations in the oxygen concentra-
tion in the reactor head spaceS then sulfonated as above
to yield ion exchange resins, the properties of which are
compared to commercial sulfonated resins made from copolymers
prepared without oxygen addition during polymerization. In
the following table, the resins of this invention are desig-
nated as A, B and C.
Oxygen Chatillon, Microcycling Stability*
Resin Level % g/bead Before** After
8 2150 97/3/0 96/4/o
B 8 2360 98.5/1.5/0 98.5/1.5/0
C 4 2300 100/0/0 98/2/o
Commercial
Resin A - 300 72/26/2 49/46/5
15 Commercial
Resin B - 510 98.5/1.5/0 55/42/3
Notes:
*100 cycles with lN HCl and 0.5N NaOH solutions.
**Perfect/Cracked/Fragmented.
Other crosslinked styrene copolymers are prepared
in accordance with this invention using oxygen incorporation,
then chloromethylated and aminated in a conventional manner
to form strong base anion exchange resins, the properties of
which are compared to commercial resins having the same
functional groups and made from copolymers prepared without
oxygen addition during polymerization. In the following
table, the resins of this invention are designted as D, E
and F.
~. :
-25-
~p ~
:1: r .
a~
c~ ~
hc: I I I
h o ~ H
h o N
r~L. ~ C~ O~
r~ ~ rl ~ a~
¢
~ ~a
r~l td 00 )~J O O
o~ ~ ~ J o
J~ D ~ o~1 :~
~' :
U~
~ o~ a Lr~
:. ~
O ~ CO ~D ~D N
~ ::t ~ 3 3
.~, 1~
'`.,:; ,1~
- ": a~
a)
: ~ ~ . bD
~1
~d ~ ~ ~1 o o ~1
rl F
H o~ G~
h
~ .
bC
.~ ~ a,
, bD~ ~
,," F~ 6 ..
.` S N J.=to o
,. ~1 ~) ~ N S
~:- X ~ . . . . .
-; ~3 ~ 3 ~~r ~ 3 ~:s
~ O
s~ c) c~
'' O ~d
~:
. Cl C~
:.~ m
., I
.. .
,~
h
h h ~v
s: ~ S~ ~v r~ r~
.. ~ ~; .rl E
~q ~ ~ ~ E ~n ~ u~ ¢
.,~ ~ O (V O Q)
.. ~L; V P~ C~ L *
26
~ 2r~
Additional Specific Examples
In a manner similar to the process described above
in the "Polymerization Procedure" additional copolymers were
prepared and functionalized to produce strong acid cation
and strong base anion resins. Using the same reactor set-up
as described above a monomer phase (represented by "A" below)
containing an initiator was charged to the reactor and
either the monomer was previously saturated with oxygen or
the reactor bead space swept with an oxygen-containing gas,
e.g., 8% 2 in nitrogen, until equilibrium was reached at
25C. (typically 30 minutes). The aqueous phase ("B" below)
was then charged (monomer : aqueous ratio = 1.1 : 1.0) and
the stirrer was set at about 210 rpm to produce droplets of
monomer in aqueous dispersion. The oxygen-nitrogen gas
` 15 sweep, if any, was passed over the dispersion at 140 cc/min
;~ for the remainder of the reaction (alternatively a pressure
- of about 5-15 psig was used).
The following were representative reaction material
charges in parts per hundred of each solution.
Solution A (Monomer Phase)
(a) Styrene 83.6
(b) Divinylbenzene (54.7% conc.) 14.6 (8.o active)
(c) Methylacrylate 1.5
(d) Di(4-t-butylcyclohexyl)-
peroxydicarbonate:Percadox-16
(initiator) 0.35
(e) t-But;yl peroctoate (chaser)
~ .
Solution B (Aqueou`s Phase)
(a) Water 95.3
3o (b) Poly~tetraall~yl ammonium
chloride)(dispersant) 3.75
(c) Gelatin (protective colloid) 0.30
r~L~
(d) Boric acid 0.16
- (e) Sodium nitrite 0.11
(f) Sodium hydroxide solution
(50% conc.) added to pH
10.5 - 11.0 0.2 - 0.4
- The reaction mixture was heated from 25C to
57C in 45 minutes and maintained at 57 + 2C for 7 hours.
The batch was then heated to 75C over a 30 minute period
and held at 75C for one hour (chase step); if a coinitiator
was used as a chaser, e.g., t-butyl peroctoate (tBP), the
batch was then heated to 95C over a 30 minute period and
held at 95C for one hour (final chase step). The batch
was then cooled and the copolymer beads were washed and
-; excess water was removed by vacuum filtration on a Buchner
funnel. The specific reaction conditions and final product
properties of styrene/DVB resins are summarized in the
following table, where crcsslinker content is expressed
as the percent "active" crosslinker ingredient, and other
~; monomer components of the commercial grade crosslinker
(principally ethyl vinyl benzene) are calculated as part
of the monovlnyl monomer. Crosslinker content given else-
where herein and in the claims is also calculated on an
"active" basis. Further, all ion exchange resin test
values given herein and in the claims, relate to fully
. 25 functionalized copolymers, that is to resins of high capa-
city and reasonable commercial quality.
-28_
.. . . .. . .
~ ~ 2
o
o o o o Lr~
h ~
~ a~ Lf\ Lr~ U~ O ~1
r I ~ O O O (~ ~1 ~ O O O
_~ ~ ~ \ ~ ~ ~ ~ ~ ~ ~D ~ ~1
C~ Lr~
~; O ~ . ~ ~ . . ~ ~ 3
E, O h ~ ~1 ~ o~ D co 0 o~ a~ c~
C~ ~ C) ~n C) ~ O~ ~ ..
~ + ~ ~
g X ~ ~
~0-` O
~I H H td .
a
~ X J~ ,D C~ O oC)t-- O`\ O ~D o ~ ~) o ~1 o ~1 )
H L~ ~t) ~ ~ co~D ~ H 0 ~ o ~ CO ~1 ~D
4 ~ O ~ (r)~-- O~1
~r; ~ ~ \J
P~ O h
td h o o Co O o :~ o o O ~1 o
h C
~1 ~ O O ~U ~ ~ O ~ O ~
G~ C~
P. h o o C~C~0~ 0 ~ o ~ c o~ Lr\ co
~ aJ o o o~ o ~ o ~ o ~ o~ o~
c~ ~ ~ ~1
u~
h
a~ oooooo
h ~e
., O O O O O O O O I O N
J~ oæ
o ~
I h^
e Lr~
~ O o oLr~ 1~ o o O ~ cr) cO Ll~
o ~ a~
h-rl ~oo ~ ~ooo~) O co 3 t~ t~ ~ tr~
. C) ~1 ~ ~1 ~1
~ C
O O ' bD bO ~ bC
~ '
,1 c) a~ O o o o o o O ~, ~Q O ~0 O
r~ td ~>
~a ~ ~~e ~e ~ ~e ~ ~ ~ ~ L~ ~~e ~ ~ ~e ~ ~e ~ ~e
c r; u~ ~ ) æ c~ 3 ~I Z C~ æ co ,~
h
o a~ a~
bO ~ ba
X O h h h h h
O O ~ ~ ~e~e ~ ~ ~ ~ ~ ~e ~lPe ~e ~e ~e
M ~ ~ O L~ O
c) a~ h . . .
:, ~L)-rl O
P ~
--`
O ~)
O
a~ o ~
~D ' a,
~1 ~1
h I rl
O X h
~ O^ J~
rl ~d ~ 3 t~3 ~ 3 3 ~1 ~r) 3 3 3 3 3 3
~C~' ........ ...... o
h O O O O O C O o O O O O O o H
H P4 h
~ .. ..
r-l ~ M M
~ O
E~ ~,1 ,1 o rl
~ ~ M rl M Z
x ~ a~ _ ~ a) ¢
~:; ~ ~ H)~X ~ ~~ æ o ~ c~ ~r; c~ *
29
Lq ~
ADDITIONAL PROCEDURE (I~ACROPOROUS RESINS)
- In a manner similar to the process described above in
the "Polymerization" procedure, macroreticular copolymers were
prepared and functionalized to produce strong acidation
exchange resins.
Example I
,
. The monomer phase consisted of styrene (387 g.),
divinyl benzene (97.4 g.), methylisobutylcarbinol (215.6 g),
Percadox-16 (2.18 g.) and t-butyl peroctoate (0.49 g). The
; 10 aqueous phase eonsisted of water (770 g.), gelatin (3.0 g.),
borie acid (3.17 g.), Padmac A (27.9 g.), and sodium
ehloride (23.1 g.).
,. .
. EXAMPLE II
The monomer phase consisted of styrene (369 g.),
, divinylbenzene (92.9 g.), methylisobutylearbinol (238 g.),
'r Percadox-l6 (2.08 g.), and t-butylperoetoate (0.51 g.).
Upon eompletion of the copolymerizations using the
aforementioned conditions, the copolymerization mixtures were
heated slowly to 100C. to remove methyl isobutyl carbinol,
and the eopolymer beads were then washed and dried prior to
sulfonation.
SULFONATION PROCEDURE
In a manner similar to that deseribed in the
"Sulfonation of Copolymer" the two macroreticular copolymers
prepared above were also sulfonated. Dried eopolymer (100 g.)
is added to 98~ H2SO4 (615 g.) followed by ethylene diehloride
(35 g.). The stirred mixture is heated to 122C. in 65
minutes and held at 122C. for one hour. This is followed by
a hydration proeedure in whieh water is added to queneh the
produet and the quenched product is treated with 80 g. of 50%
NaOH to convert to the sodium salt.
-30-
~.J ~
MACRORETICULAR STRONG ACID CATION EXCHANGE RESINS
:,
EXAMPLES CHATILLON CATION EXCHANGE
(llPCT. DVB) (G/BEAD PCT. SOLIDS CAPACITY (MEQ/GM)
Control (Typical 36S-412 50-55 4.4 - 4.5
values)
I 508 55.4 4.47
~ II 651 48.9 4.50
.` 10
`:
