WATER  POLLUTION CONTROL RESEARCH SERIES
                                     17040 DFC 10/70
          FEASIBILITY STUDY
                 OF
         REGENERATIVE  FIBERS
                 FOR
       WATER POLLUTION CONTROL
U.S. ENVIRONMENTAL PROTECTECTION AGENCY

-------
          WATER POLLUTION CONTROL RESEARCH SERIES
The Water Pollution Control Research Series describes the
results and progress in the control and abatement of pollution
in our Nation's waters.  They provide a central source of
information on the research, development and demonstration
activities in the Environmental Protection Agency, through
inhouse research and grants and contracts with Federal,  State,
and local agencies, research institutions, and industrial
organizations.

Inquiries pertaining to Water Pollution Control Research
Reports should be directed to the Chief, Publications Branch
(Water), Research Information Division, R&M, Environmental
Protection Agency, Washington, D.C. 20U60.

-------
            FEASIBILITY STUDY OF REGENERATIVE FIBERS
                FOR WATER POLLUTION CONTROL
                                  by
                          Uniroyal, Inc.
                      Wayne, New Jersey 07470
                              for the
                ENVIRONMENTAL PROTECTION AGENCY
                       Program #17040  DFC
                       Contract #14-12-815
                         October, 1970
For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. 20402 - Price 75 cents

-------
                  EPA Review Notice
This report has been reviewed by the Environmental Protection
Agency and approved for publication.  Approval does not
signify that the contents necessarily reflect the views and
policies of the Environmental Protection Agency nor does
mention of trade names or commercial products constitute
endorsement or recommendation for use.
                          ii

-------
                               ABSTRACT

The feasibility of making fibers which have the ability to sorb minerals
from water has been investigated.  High molecular weight organic polymers
cospun with fiber-forming polymers and crosslinking agents can be spun
into fiber and chemically treated to give ion exchanging materials.  In
some cases the ion exchange active material can be spun directly (e.g.,
polyamines and polyvinylpyridines); in other cases the activity must be
introduced after spinning (e.g., the sulfonation of fibers containing
polystyrene).

Styrene sulfonic acid cation exchange fibers having capacities up to
4.2 meq/g dry fiber have been prepared.  Polypropylene and polybutadiene
served as fiber-forming polymer and crosslinking agent, respectively.

By cutting the fiber to very short lengths, it was possible to test these
materials by column elutions in much the same way as conventional ion
exchange beads are evaluated.  Accordingly, the breakthrough capacities
for ion exchange fibers were found to be higher than those for beads,
probably due in part to a faster rate of ion exchange in fibers.

Weak acid cation exchangers were produced from fibers containing poly-
methylmethacrylate by hydrolysis of the ester to polymethacrylic acid.
Capacities up to 1 meq/g were obtained.

Vinylpyridine and aliphatic amine ion exchange active materials were spun
directly and then crosslinked with dihaloalkanes.  Quaternization was
achieved with alkyl halides.  Vinylpyridine exchangersgave capacities up
to 3.8 meq/g, and fibers containing both Vinylpyridine and polyethyl-
eneimine show promise of yielding higher capacities. ,,,

Selectivities of ion exchange fibers were similar to those of conven-
tional beads, but different selectivities may be possible because of
the anisotropic swelling of the fibers.

This report was submitted in fulfillment of Project Number 17040 DFC,
Contract 14-12-815, under the sponsorship of the Water Quality Office,
Environmental Protection Agency.
                                  iii

-------
                             CONTENTS


Section                                                             Page

 I         Conclusions 	 	  1

 II        Recommendations	3

 III       Introduction 	   5

 IV        Experimental  	  9

              Preparation of Ion Exchange Fibers  	  9

              Evaluation of Ion Exchange Fibers -- Physical
              Properties	10

              Evaluation of Ion Exchange Fibers -- Chemical
              Properties	11

 V         Results and Discussion	13

              Fiber Spinning	13

                 Polypropylene-Polystyrene Mixtures   ...  	  13

                 Mixtures Containing Butadiene  	  13

                 Mixtures Containing Epoxy Resins   	  24

                 Mixtures for Weak  Acid Cation  Exchangers	24

                 Anion Exchange Fibers Containing  SDM  	  25

                 Vinylpyridine Anion Exchange Fibers  .  	  25

                 Mixtures Containing Polyethyleneimine  	.  .  26

                 Miscellaneous Fiber Mixtures and  Treatments  ....  27

              Evaluation of Cation  Exchange  Fibers 	  27

                 Sulfonic Acid Exchangers  	  27

                 Column Elutions of Sulfonic Acid  Exchangers  ....  29

                 Selectivity Coefficients  for Sulfonic  Acid
                 Exchangers	39

                 Methacrylic Acid Exchangers 	    44
                                      iv

-------
                             CONTENTS (contd.)




Section                                                             Page




             Evaluation of Anion Exchange Fibers  	  44




                SDM Fibers	44




                Vinylpyridine Fibers  	  45




                Selectivities for Vinylpyridine Fibers  	  47




                Polyethyleneimine Fibers  	  55




                Selectivities for Polyethyleneimine Fibers  	  61




 VI       Acknowledgments	69




 VII      References	71




VIII      Glossary	  .  .  73

-------
                                 FIGURES




No.                                                                   Page
1

2
3
4
5
6
7
8
9
10

11

12
13
14

15

16

17

18

19
Effect of the Number of Titration- Regeneration Cycles
on the Capacity 	 	
Total Capacity of Sulfonated PP-PS-ABS Fiber 	
Breakthrough Curve for Sulfonated PP-ABS Fiber (1-141) . .
Breakthrough Curves for Sulfonated ABS Fiber (1-177) . . .
Breakthrough Curve for Rohm & Haas IR-120 Resin .  . . . .
Breakthrough Curves for Cation Exchange Fiber II- 9 ....
Total Capacity versus Swelling for MS-WP-40B 	
Sulfonated MS-WP-40B, First Elution 	
Sulfonated MS-WP-40B, Second Elution .... 	
Calcium Ion Breakthrough for Ammonium- Calcium Elution
of Sulfonated MS-WP-40B 	 	
Ammonium Ion Breakthrough for Ammonium-Calcium Elution
of Sulfonated MS-WP-40B 	
Breakthrough Curve for Anion Exchange Fiber (1-159) . . . .
Breakthrough Curves for Anion Exchange Fiber II -2 	
Breakthrough Curve for Anion Exchange Fiber 11-52
(MS-WP-12E) 	
Chloride Ion Breakthrough for Second Elution of
Sample 11-52 	
Phosphate Ion Breakthrough for Second Elution of
Sample 11-52 	 ; 	
Chloride Ion Breakthrough Curve for Third Elution of
Sample 11-52 	
Phosphate Ion Breakthrough Curve for Third Elution of
Sample 11-52 	
Fourth Elution of Sample 11-52 	

30
31
32
34
35
37
38
40
41

42

43
46
48

50

51

52

53

54
56
                                     vi

-------
                             FIGURES (Contd.)

No.                                                                   Page

20     Fifth Elution of Sample 11-52 ......	   57

21     Sixth Elution of Sample 11-52 . 	 ..........   58

22     Chloride Ion Breakthrough Curve for Nitrate-Chloride
       Elution of Sample 11-52, Seventh Elution  	 .59

23     Chloride Ion Breakthrough Curve for Nitrate-Chloride
       Elution of Sample 11-52, Eighth Elution  	   60

24     Breakthrough Curve for Anion Exchange Fiber 11-37
       (MS-WP-44A)	   62

25     Breakthrough Curve for Anion Exchange Fiber 11-55
       (MS-WP-44A)	   63

26     Second Elution of Sample 11-55  	 .....   64

27     Third Elution of Sample 11-55 (Chloride + Phosphate)  ....   65

28     Phosphate-Chloride Elution of Sample 11-63  	   68
                                    vi x

-------
                                  TABLES




No,




 1     Polypropylene-Polystyrene Mixtures  	   13




 2     Styrene-Butadiene Mixtures . .	   14




 3     Fiber Compositions in Terms of Major Components 	   16




 4     Mixtures with Epoxy Resin	   17




 5     Fibers for Weak Acid Exchange .	   18




 6     Vinylpyridine Mixtures ....  	  ......   19




 7     Polyethyleneimine and Miscellaneous Mixtures   . 	   20




 8     Physical Properties -- Cation Exchange Fibers  	   21




 9     Physical.Properties -- Anion Exchange Fibers   	   22




10     Irradiation of Fibers  .  . .	   27




11   .  Effect of Titration-Regeneration Cycles on the Capacity  ...   28




12     Breakthrough Capacities  for CaCl2 Elutions 	   33




13     Sulfonation of Ms-WP-40B	    36




14     Elutions for Anion Exchange Fiber 11-52	  .  .  .   49




15     Phosphate-Chloride Elutions of Anion Exchange  Fibers   ....   66
                                    viii

-------
                              SECTION I

                             CONCLUSIONS
1.  The feasibility of using high molecular weight organic polymers
cospun with fiber forming polymers and crosslinking agents to sorb
minerals from water has been established.

2.  Ion exchange fibers can be made strong and durable while maintaining
high percentages of active material in the fibers.

3.  Selectivities of ion exchange fibers parallel those of conventional
beads; however, different selectivities may be possible because of the
anisotropic swelling of the fibers.  Based on theoretical considerations
and indirect experimental evidence, the rates of ion exchange for fibers
are concluded to be much greater than corresponding rates in beads.

4.  Styrene sulfonic acid cation exchange fibers having capacities
(4.2 meq/g dry fiber) of comparable magnitude to those of conventional
ion exchange beads can be produced.  Polymerized butadiene serves as a
crosslinking agent for the styrene sulfinic acid exchangers.

5.  Weak acid cation exchangers may be produced from fibers containing
polymethylmethacrylate by hydrolysis of the ester to polymethacrylic
acid.

6.  Vinylpyridine and aliphatic amine ion exchange active materials may
be crosslinked with dihaloalkanes.  Quaternization may be achieved with
alkyl halides.  Vinylpyridine exchangers having capacities of 3.8 meq/g
dry fiber can be made.  Fibers containing both vinylpyridine and poly-
ethyleneimine show promise of yielding higher capacities.

-------
                              SECTION II

                           RECOMMENDATIONS

This research was restricted to determining the feasibility of making
ion exchange fibers of reasonable capacity and strength.  Pursuant to a
fuller understanding of the properties and potential applications of
these ion-sorbing fibers, the following program is recommended:

1.  Existing fibers should be thoroughly tested for rate of attrition by
following the capacity over a large number of exhaustion-regeneration
cycles.

2.  The effect of organic foulants on ion exchange fibers should be
determined.  Also the effect of fiber porosity on organic fouling should
be determined.  Some of these tests should be made using actual secondary
effluents.

3.  Physico-chemical studies of rates and selectivity should be performed
on ion exchange fibers under conditions simulating actual operations, and
these results should be compared with those for ion exchange beads.

4.  A bench scale model of a continuous ion exchange system utilizing a
moving belt fabricated from ion exchange fiber should be constructed.
The belts could be tested for rates of exchange, selectivity, attrition,
and fouling.

5.  Improved ion exchange fibers should be sought -- fibers having maximum
ion exchange capacity while keeping maximum strength.

-------
                             SECTION III

                             INTRODUCTION
As our population grows and the per capita water requirements likewise
grow, it has become apparent that within the next two decades the daily
need for water will exceed our fresh water resources.  Long before this
high demand for water is reached, it will not be practical to discharge
secondary effluents into streams; "purification by dilution" will no
longer be effective.  Not only will tertiary treatment be necessary
before discharge into rivers and streams but in some cases water will
have to be conditioned for immediate reuse.  In order to promote wide-
apread application of tertiary treatment, a highly advanced technology
must be developed -- especially toward reducing the cost of waste water
renovation for reuse.

Four distinct processes have evolved as the most feasible for demineral-
ization of waste water.  These are distillation, electrodialysis, reverse
osmosis, and ion exchange.  A number of variations and combinations of
these basic processes have been proposed.  Clearly, all of the above
methods of water treatment have both technological and cost deficiencies.

Ion exchange has been most successful in, applications requiring the
demineralization of water having reasonably low levels of organic
pollutants and has been used widely in the treatment of industrial
wastes.  Irreversible'fouling by organic contaminants has been greatly
reduced by improved ion exchange resins, and ion exchange processes
have been successfully demonstrated for the treatment of secondary
effluents (1).  Although ion exchange processes are generally regarded
to be most economical for low solids water, improved processes have been
developed for water with high total dissolved solids including desalin-
ization of sea water  (2).

Ion exchange as a unit operation requires periodic regeneration of the
exchanger.  Thus, two ion exchange systems are necessary in order to
maintain continuous demineralization.  On the other hand, continuous ion
exchange systems are  rather complex and require costly plumbing for
transport of the resin from the exhaustion column to a regeneration com-
partment and back.  Attrition of the ion exchange resin is higher than
in conventional fixed-bed equipment (3).  Another difficulty in continuous
ion exchange systems  is maintaining column integrity while moving the
resin.  It is desirable, then, to consider forms for ion exchange
materials which differ from the usual spherical bead and are more amen-
able to continuous operations.  Three fundamental shapes can be dis-
tinguished -- the bead  (usually spherical), the fiber  (one dimension much
larger than the other two), and the film  (two dimensions much larger
than the third).  Ion exchange films or membranes are employed in elec-

-------
trodialysis and various other applications.  The films are generally
weak and are not suitable in applications involving mechanical move-
ment.  Ion exchange fibers could be practical for column operations
since the fibers (in the form of a bundle of cylindrical lengths) pack
more efficiently than spherical beads.  Furthermore, ion exchange fibers
can be woven into tough, durable materials capable of withstanding the
duress of continuous mechanical movement.  A strong belt can be fabri-
cated from very thin fibers, thus exposing a very large surface area of
ion exchange material.

The requisite physical properties for an ion exchange fiber differ
greatly from those for the conventional ion exchange bead.  Whereas the
beads are hard and brittle, the fiber must be flexible and elastic.
For this reason the chemical composition of an ion exchange fiber will
likely differ from that of spherically shaped ion exchange materials.

An ion exchange fiber is composed of three basic elements:  (1) an
ion exchange active material,  (2) a fiber forming resin, and  (3) a
crosslinking agent.  A given chemical constituent in the fiber may
serve in one or more of the three categories.  Certain ion exchange
active material can be melt spun with a fiber forming polymer.  How-
over, highly ionic materials usually cannot be melt spun, and for these
cases the ion exchange activity must be introduced by chemical reaction
after spinning.  The chemically active groups suitable for ion exchange
fibers are identical to those  available in conventional ion exchange
beads.  Moreover, the chemical framework to which the exchange active
groups are attached is often similar in fiber and beads.  For example,
the skeletal structure for a sulfonic acid cation exchanger might be
polystyrene in both fiber and  beads.

In general, neither the organic polymers with ion exchange capacity nor
those polymers which are converted to ion exchangers after spinning will
form strong fibers.  Therefore, it is usually necessary that  part of the
ion exchange fiber be composed of a strong fiber forming material  (for
example, polyolefin, polyamide, polyester, etc.).  As little  as ten
percent of this inert fiber forming polymer in the ion exchange fiber
can improve its strength by more than a factor of ten.  The amount of
inert filler in the ion exchange fiber is a compromise between fiber
strength and loss of total capacity since these fiber forming resins
have no sorption capacity.

The third major ingredient of  the ion exchange fiber  is the crosslinking
agent.  It is not possible to  spin crosslinked material, and  the  actual
crosslinking must be effected  after spinning.  In conventional ion
exchange beads, crosslinking is often achieved at little expense  to
total capacity.  For example,  in a styrene polymerization the divinyl-
benzene added for crosslinking may be sulfonated with the styrene to
produce additional cation exchange sites.  Thus, any  loss in  theoretical
capacity is due only to the small difference in the equivalent weight
of styrene and divinylbenzene. In contrast to the above efficient

-------
crosslinking system, crosslinking in ion exchange fibers often requires
a material which has no capability for conversion to ion exchange
activity, and in such cases the crosslinking agent significantly
reduces the maximum theoretical total capacity.  However, the presence
of an inert fiber forming material and the usually high degree of
crystallinity in the fiber allow for the use of less crosslinking agent
than would otherwise be required.

Some early work in the use of fibrous materials for ion exchange was
presented by Muendel and Selke in 1955 (4).  The ion exchange fibers
reported were exclusively chemically treated natural fibers.  Since
this preliminary work, very little has appeared in the literature with
the exception of several patents (5-7).  The purpose of the present
study was to determine the feasibility of using melt-spun polymers for
the removal of ions from water and waste water.  Melt spun fibers have
the potential advantage of greater capacity and lower cost.  The study
included the preparation of multicomponent fibers which function as ion
exchangers and the evaluation of these fibers in column operation for
comparison with conventional bead ion exchangers.

-------
                              SECTION IV

                             EXPERIMENTAL
PREPARATION OF ION EXCHANGE FIBERS

All fibers were spun from the melt in a 1-inch Modern Plastics
Machinery Co. extruder feeding a size 1/2 gear pump.  The melt emerged
from an eight hole die (0.020 in. hole diameter) at temperatures ranging
from 230 to 260C and at a rate of about 5 g/min.  The fiber was taken
up at 200 to 300 ft/min.  Prior to spinning, the ingredients were
thoroughly mixed by one of three methods depending on their physical
state and stability.  If all the ingredients had approximately the same
small particle size, mixing was readily achieved at room temperature on
a ball mill.  Those ingredients having disparate particle size or requir-
ing mixing of solids and liquids were banded on a steam heated two-roll
mill and ground in a Cumberland grinder.  Alternatively, for mixtures
lacking the oxidative stability for mill mixing at elevated temperatures,
the ingredients were extruded through a single hole (1/8 in. diameter)
die at 230 to 260C and chopped to produce pellets about 1/4 in. long.
For materials which were easily degradable, the extrusion operations
were carried out under an argon atmosphere.  Whenever possible the fiber
strength was increased by hot drawing.  The fiber, entering a 130-150C
oven at a fixed rate, was drawn from the oven at a faster rate, the
ratio of speeds being the draw ratio.  Because most of the fibers prior
to drawing were very weak, the speed at which the undrawn fiber could be
fed into the oven was usually less than 50 ft/min.

After spinning, the next step in the preparation of ion exchange fibers
is to introduce sufficient crosslinking in the fiber to render it water
insoluble and prevent excessive swelling.  Two crosslinking methods were
employed for most of the ion exchange fibers - one for cation exchangers
and one for anion exchangers.  The use of these two methods lay partly
in their simplicity and partly because the crosslinking agents did not
add excessive inert weight to the fiber.

Cation exchange fibers were crosslinked by tying together polybutadiene
units which were included in the original fiber mixture.  That is,
styrene-butadiene copolymers (SBR), acrylonitrile-butadiene (NBR)
copolymers, acrylonitrile-butadiene-styrene  (ABS), or some combination
of the above was spun with the original fiber.  Points .of unsaturation
were then linked by free radicals initiated by cold sulfuric acid or
tertiary-butyl peroxide.  In the sulfuric acid cure the fiber was placed
in cone. H  SO, at 0C for several hours after which the temperature was
allowed to  rise to that of the room.  The fiber remained in the sulfuric
acid at ambient temperature for  periods varying from a  few hours to
several days.  This procedure resulted  in a small amount of sulfonation
as well as  crosslinking.  Several hours exposure  to a methanol or
ethanol solution of t-butyl peroxide at room temperature also effectively

-------
linked points of unsaturation in fibers containing polybutadiene.

All of the anion exchange materials were of either aliphatic amine or
pyridine type.  Crosslinking these materials was effected by quatern-
ization with dibromoalkanes , typified by the reactions:
             Br(CH2)6Br
       [R3N+(CH2)6Br]Br~    R3N - > [R3N+(CH2)6N+R3] 2Br"


Several days exposure to a decane solution of dibromoalkanes at 50C
was sufficient for corsslinking polyvinylpyridines but the method was
not as successful in the crossl inking of polyethyleneimine - some of
which was always extracted from the fiber during evaluation,

For the weak base ion exchange fibers, crosslinking was the last step
in the preparation since mixtures containing the exchange active ingre-
dient could be spun directly.  To make a strong base ion exchange fiber,
the weak base fiber was further quaternized with a bromoalkane (usually
n-bromobutane) .

Another crosslinking method attempted for both anion and cation fibers
was exposure to high energy electrons.  Several samples were irradiated
with a Van de Graaff generator electron source (about 1.5 Mev) ; exposures
were between 8 and 32 watt-hr.

Cation exchange materials could not be spun directly, and the activity
had to be introduced after spinning and crosslinking by chemical
reaction.  Strong acid exchangers were made by sulfonation of fibers
containing polymerized styrene.  Fuming sulf uric acid and chlorosulfonic
acid were the sulfonating agents of choice.  Rinsing of the fiber after
sulfonation must be carried out very carefully in order to avoid
charring or even melting of the fiber by the large heats of mixing of
the acids with water.

Weak acid cation exchangers were prepared by the hydrolysis of a poly-
ester such as polymethylmethacrylate.  Hydrolysis of the. ester and
crosslinking of butadiene in the fiber were accomplished simultaneously
with sulf uric acid.

EVALUATION OF ION EXCHANGE FIBERS - PHYSICAL PROPERTIES

Several parameters relating to physical properties of the fibers were
measured after spinning and also after drawing.  The properties measured
were (a eight -filament denier, (b) percent elongation at break, and
(c) tenacity (8).  The eight -filament denier is defined as the weight in
grams of a 9000 meter length of eight filament yarn.  The eight -filament
denier is related to the monofilament diameter by

                  diameter (in.) = 1.65 X 10"  V denier,

                                     10

-------
                                       3
assuming the fiber density to be 1 g/cm .   For example, a denier of
400 corresponds to a monofilament diameter of about 0.0033 in.  Denier
is measured by weighing a three meter length of fiber.  The percent
elongation and tenacity are derived from Instron data in which two-inch
lengths of eight filament yarns are pulled apart at a constant rate and
the load-extension curve recorded.  Percent elongation is derived from
the length of the fiber just prior to breaking.  The relationship between
the tenacity in grams/denier and the tensile strength is:

                      1 g/denier = 12,800 psi,
                                                                    3
again with the approximation that the density of the fiber is 1 g/cm .

Another physical property which is important in the ion exchange fiber is
the degree of swelling in aqueous solution.  In many cases the swelling
was simply estimated by visual inspection.  For finely cut fiber the
swelling could be estimated by measuring the volume occupied by the fiber
in a column.  For more accurate determination of the degree of swelling
the fiber was photographed in the wet and dry states at a magnification
of about 40X.  The fiber diameters were measured to determine the degree
of swelling on the assumption that the increase in fiber length was
negligible.

EVALUATION OF ION EXCHANGE FIBERS - CHEMICAL PROPERTIES

Generally, the first test made on an ion exchange fiber was a determin-
ation of the total capacity.  For the strong acid or strong base ion
exchange fibers the total capacity was determined by simple acid-base
titration.  An excess of sodium chloride was added to the fiber in an
aqueous solution.  The acid liberated in exchange for sodium ions by the
cation exchanger is titrated with sodium hydroxide.  Similarly, the
base liberated in exchange for chloride ions is titrated with hydro-
chloric acid.  In the case of weak acid or weak base ion exchange fibers,
the total capacity was estimated by back titration after neutralization
of the exchanger.  Thus, a known excess of base is added to the weak
acid ion exchange fiber, and after a few minutes of contact the fiber
and solution are separated.  Then the excess base is back titrated to
neutralization with acid.  The meq of acid subtrated from the meq base
gives the total capacity of the fiber.

In order to determine the rate of attrition of the ion exchange fibers,
the total capacity was followed as a function of the number of cycles.
For strong acid exchangers this was accomplished by the simple titration
method; for weak base exchangers the total capacity was derived from
column elution data.

All of the column elutions were carried out in 100 ml burets having
cross-sectional area of  1.9 cm  .  Finely cut  fiber was packed in the
buret with a glass plug  below and above the fiber.  The  fiber was cut
either by hand, in a Waring blender, or in a Wiley mill.  Hand cutting,
though tedious, yielded  the best  samples because the blender  and mill
tended to crush and disintegrate  the fiber.


                                  11

-------
General performance of the exchangers was tested by elution with calcium
ion solutions for cation exchangers or chloride ion solutions for anion
exchangers.  The effluent entered a siphon cup which delivered volumes
ranging from 16 to 20 ml to a fraction collector.  Selected effluent
samples were then analyzed for the ions of interest.  The influent
solution was gravity fed to the column and the rate of elution was
adjusted by the buret stopcock.

Those elutions for determining ion selectivity required solutions with
two ions present, for example, chloride and phosphate.  In the case of
phosphate and ammonium the second ion was separately analyzed, but in
the elution of a chloride-nitrate solution the nitrate ion concentration
was known only in the influent solution.

Ion concentrations were determined as follows:

      Calcium - EDTA titration, Eriochrome Black Indicator (9)

      Chloride - AgNO- titration, CrO~ Indicator

      Phosphate - Absorption at 880 m^j, of an ant imony- phosphate -
                  molybdate (10) ascorbic acid complex

      Ammonium - absorption at 425 mjx of nesslerized ammonia solution
                                  12

-------
                              SECTION V

                        RESULTS AND DISCUSSION

FIBER SPINNING

A complete list of the mixtures which were formulated for melt spinning
is given in Tables 1 through 7.  A description of the materials corres-
ponding to the codes used in these tables is given in the Glossary.
Physical properties of the fibers are given in Tables 8 and 9.  A dis-
cussion of the principal characteristics of the fiber formulations and
physical properties follows.

Polypropylene-Polystyrene Mixtures

It is clear from Table 1 that polypropylene (PP) and polystyrene (PS)
can be spun in all proportions.  PS is a fiber forming material, but the
filaments are very weak.  Thus, a decrease in tenacity with increasing
PS content is noted for these compositions.  Sulfonation of the fibers
in Table 1 will give a sulfonic acid cation exchanger.  Only very low ion
exchange capacities are possible with these fibers because of the solu-
bility of sulfonated polystyrene.  To a small extent the polypropylene
prevents extraction of the PS so long as the degree of sulfonation is
quite low.  Signer e_t al. (11) have reported that 10% sulfonation renders
polystyrene water soluble.

                             Table 1

                Polypropylene-Polystyrene Mixtures

               Code No.     % PP     % PPP     % PS

               MS-WP-OA      90                 10

               MS-WP-OB      50                 50

               MS-WP-OC      10       10        80

               MS-WP-OD      10                 90

Mixtures Containing Butadiene

In order to prevent extraction of water soluble polystyrene sulfonic
acid, butadiene was included in the  fiber  formulations; crosslinking was
then effected via the points of unsaturation after spinning.  All of the
mixtureS'in Table 2 contain butadiene in the form of a copolymer or
polymer blend.  In Table  3 the mixtures of Table 2 which could  be melt
spun are resolved in terms of  inert, styrene, and butadiene content.

The PP-PS-ABS system  (MS-WP-8A to  -8G) yields good fibers  throughout a
wide variation in concentrations.  The mixture MS-WP-8E was prepared with
                                   13

-------
                                        Table 2

                               Styrene-Butadiene Mixtures

Code No.
MS-WP-8A
8B
8C
8D
8E
8F
8G

% PP
40
30
20
30
25
25
20

% PS
50
50
40
40
40


/o /o /o /o fo
ABS SBR NBR PMS SB-1
10
20
40
30
35
75
80
                                                                    Z       %
                                                                   SB-2    SB-3    % NW
MS-WP-21                        100
MS-WP-26A
26B*
26C
26D*
26E*
26F*
26G*
26H
261*
MS-WP-31A*
31B
31C
31D
31E*
31F*
20
20
15
15
20
30
25
30
20






70
60
70
65
60
55
55
50
64






                                       10
                                       20
                                       15
                                       20
                                       20
                                       15
                                       20
                                       20
                                       16

                                 90     10
                                 92.5   7.5
                                 95     5
                                 97.5   2.5
                                 85     15
                                 80     20
                 * Could not  be melt  spun

-------
            Table 2 (Cont.)
Code No.
MS-WP-33A
33B
MS-WP-34A
34B
34C*
34D*
MS-WP-48A
48B
48C
MS-WP-40A*
40B
MS-WP-49A*
MS-WP-7A*
MS-WP-29A*
29B*
MS-WP-30A*
MS-WP-32A*
MS-WP-45A*
% PP
15
25

15
10
5
15
25
25
30
15
30
20
25
20
% %
% PS ABS SBR
40 35 10
45 20 10
90
85
80
82
80
80
85
60
65

30


37.5

/o /o /o /o /o
NBR PMS SB-1 SB- 2 SB- 3 % NW

10
15
20
18
5
10
10
20
15
15 65
40
85
70
80
37.5
60 20
*Could not be melt spun

-------
                                 Table 3

             Fiber Compositions in Terms of Major Components


Code No.              % Inert          % Styrene          % Butadiene

MS-WP-8A               41.6              57.7                 0.7
      8B               33.2              65.4                 1.4
      8C               26.4              70.8                 2.8
      8D               34.8              63.1                 2.1
      8E               30.6              67.0                 2.4
      8F               37.0              57.8                 5.2
      8G               32.8              61.6                 5.6

MS-WP-21               16.0              77.0                 7.0

MS-WP-26A              20.0              72.5                 7.5
      26C              15.0              73.8                11.2
      26H              30.0              55.0                15.0

MS-WP-31B              14.8              73.1                12.1
      31C              15.2              74.4                10.4
      31D              15.6              75.7                 8.7

MS-WP-33A              20.6              69.4                10.0
      33B              28.2              62.9                 8.9

MS-WP-34A              15.9              69.3                14.8
      34B              15.8              65.5                18.7

MS-WP-48A              28.6              61.6                 9.8
      48B              24.3              61.6                14.1
      48C              20.1              65.5                14.4

MS-WP-40B              27.2              60.0                12.8
                                    16

-------
                                 Table 4




                        Mixtures with Epoxy Resin






Code No.           % PP       70 PS      % ER     % ABS     % SBR
MS-WP-20A
20B
20C*
MS-WP-22A
22B*
MS-WP-23A
23B*
23C*
MS-WP-28A*
28B*
30
30
15
15
10
10
10
5
20
25
65
60
65


40
40
40
60
55
5
10
20
5
30
10
25
25
10
10



80
60
40
30
30-










10
10
         *Could not be melt spun
                                    17

-------
                                  Table 5 - Fibers for Weak Acid Exchange
oo

Code No.
MS-WP-11A*
11B*
11C*
11D*
MS-WP-18A*
18B*
18C*
18D*
MS-WP-18AA*
18CC*
MS-WP-27A*
27 B*
MS-WP-35A*
35B*
MS-WP-36A*
36B*
MS-WP-37A*
37 B*
MS-WP-38A*
MS-WP-42A*
MS-WP-47A*
MS-WP-41A
MS-WP-46A*

% PP
75
25
60
90
80
70
30

80
30
90
80
40
20
20
30
20
10
16.7




% % % % % 7
EMA % PS ABS NBR PAA GMA SMA-1
25
75
40
10
10 10
20 10
30 40
80 20
10
30
10
20
60
80
10 70
15 55
10 70
30 60
16.7 8.3
45 5 50
60 20 20
45 5
30
                                                                                  SMA-2
PMM
PAAS
                                                                                                    10
                                                                                                    40
                                                                                    58.3
                                                                                             50

                                                                                             70
               *Could not  be melt  spun

-------
                                   Table 6 Vinylpyridine Mixtures

Code No.
MS-WP-1A*
IB*
1C*
ID
MS-WP-2A
2B
MS-WP-3A*
MS-WP-4A
4B
4C
4D*
MS-WP-5A*
MS-WP-6A
MS-WP-12A
12B
12C
12D
12E
MS-WP-13A*
13B*
MS-WP-14A*
L4B*
%
% PP NPA
80
60
40
20
75
25
50
30
15
27.5
37.5
50
20
25
30
25
25
15
30
30
15
15
/o /o /o /o 10 /o
EMA % PS ABS PVP PVP-HAC SVP ZnO % ER
20
40
60
80
25
75
25 25
10 60
5 80
2.5 70
2.5 60
50
80
5 70
10 60
15 60
25 50
5 80
60 10
50 20
5 75 5
5 60 20
MS-WP-15A*
15
77.5
2.5
            *Could not be melt spun

-------
                                 Table 7
Polyethyleneimine and Miscellaneous Mixtures
Code No.
MS-WP-16A*
MS-WP-17A*
MS-WP-19A*
19B*
19C*
MS-WP-44A
44B
44C
44D
44E*
RF-PB-129D
MS-WP-9A*
9B*
/o /o /o
7o PP EMA % PS ABS PVP
70 10
70 10
80 10
50 25
80
15 5 70
20 10 60
15 5 60
15 5 50
20 10 40
40
75
25
% %
PEI SDM
20
20
10
25
20
10
10
20
30
30
60

%
PAS





25
75
MS-WP-10A*
20
75
           *Could not be melt spun
                                    20

-------
                  Table 8
Physical Properties - Cation Exchange Fibers
Code No.
MS-WP-OA
OB
OC
OD
MS-WP-8A
8B
8B
8C
8D
8D
8E
8E
8F
8F
MS-WP-21
Draw 8 Filament
Ratio Denier
537
463
549
531
560
687
2:1 351
753
573
3:1 200
602
6:1 96
684
3:1 228
711
Percent
Elongation
803
_
-
-

-
53
_
.
43
-
55
_
39

Tenacity
(g/denier)
0.72
0.59
0.49
0.50
0.47
0.48
i.io
0.41
0.48
1.31
0.50
2.49

1.08
0.53
MS-WP-26C
      26C

MS-WP-31B
      31B
      31C
      31D
      33B
      33B

MS-WP-34B

MS-WP-48A
      48B
      48C
      48C
      48C

MS-WP-40B
      40B
      40B(2)

MS-WP-20A
      20A
      20A
      20A
      20B

MS-WP-41A
   1.7:1
    4:1
    2:1
    3:1
    4:1
    4:1
    3:1
   1.2:1
   3.5:1
    6:1
 639
 381

1794
 443
 664
 687

 714
 524
 270

1130

 733
 797
 972
 330
 238

 717
 168
 248

 523
 444
 150
  85
 543

 535
 52
 50
 38
 27
 54
 38
103
 87
 43
                 0.27
0.18
1.34
0.32
0.51
0.33
0.36
0.98

0.20

0.23
0.28
0.24
0.99
1.01

0.34
1.14
1.79

Oo46
0.46
1.33
2.34
0.44

0.23
                     21

-------
                               Table 9
             Physical Properties - Anion Exchange Fibers
Code No.

MS-WP-1D

MS-WP-2A
      2B

MS-WP-4A
      4B
      4C

MS-WP-6A

MS-WP-12A
      12C
      12D
      12E
      12E

MS-WP-44A
      44A
      44A
      44B
      44B
      44C
      44D
      44D
 Draw
Ratio
 2:1
1.6:1
1.8:1

1.6:1
2.3:1
8 Filament
  Denier

    484

    650
    618

    589
    723
    848

    597

    762
    624
    842
    662
    355

    869
    545
    468
    740
    447
    889
   1839
    789
 Percent
Elongation
                               298
 Tenacity
(g/denier)

   0.39

   0.69
   0.84

   0.70
   0.64
   0.43

   0.56

   0.39
   0.61
   0.36
   0.60
   0.89

   0.38
   0.48
   0.50
   0.32
   0.55
   0.25
   0.16
   0.48
                                   22

-------
the expectation of obtaining good physical properties together with high
ion exchange capacity.  It is a compromise between MS-WP-8C,  which has a
high loading of PS and ABS but forms a brittle fiber, and mixtures
MS-WP-8B and -8D which have good physical properties but lower ion
exchange potential.  MS-WP-8E with its 2.49 g/denier tenacity was the
strongest fiber obtained in this work.  Because of the effectiveness of
acrylonitrile-butadiene-styrene (ABS) in preventing loss of sulfonated
polystyrene, even without any crosslinking after spinning, attempts were
made to spin higher proportions of this material with polypropylene
(MS-WP-8F and -8G); and, in fact, it was found that ABS could be spun by
itself (MS-WP-21).

Another approach for the incorporation of butadiene in the fiber mixtures
is the addition of a copolymer which is high in butadiene.  Such is the
case with MS-WP-26A to -261 where SBR (25% styrene, 75% butadiene
copolymer) is combined with polystyrene and polypropylene.  The use of
SBR allows higher butadiene and styrene content and a minimum of inert
material.  However, the fiber obtained from SBR mixtures is usually
extremely weak.  For example, the tenacity of MS-WP-26C was too low to
be measured by the Instron test, and the tenacity of the drawn fiber
was only 0.27 g/denier.

Physical properties were greatly improved by spinning SBR with ABS
(instead of PS and PP).  Thus, MS-WP-31B is strong, 1.34 g/denier, and
contains high proportions of styrene and butadiene  (see Table 3).  Com-
binations of PP, PS, ABS, and SBR as in Ms-WP-33A and -33B resulted in
no substantial improvement in physical properties.

Further refinements in fiber properties were obtained by the use of NBR
(15% acrylonitrile, 85% butadiene copolymer).  The  NBR, having greater
high temperature viscosity than the SBR used, was easier to spin.  In
the series MS-WP034A to -34D, NBR was combined with ABS to give fibers
with high butadiene content.  MS-WP-34B had the highest butadiene content,
18.7%, and still had good tenacity, 0.98 g/denier.  The series MS-WP-48A
to -48C was an attempt to improve physical properties while maintaining
high butadiene content.  Although these mixtures had improved spinning
properties, there was  little improvement in the fiber tenacity.  On the
other hand, mixture MS-WP-40B, combining  polypropylene, polystyrene,  and
NBR yielded a much stronger  fiber,  1.79 g/denier, while maintaining
reasonably high styrene and  butadiene content.

Mixture MS-WP-49A  is  the  same as  -40B except that the polystyrene was
replaced with  polymethylstyrene  (PMS).  This substitution  has  potential
advantage in that  the  formation of  sulfone bridges  during  sulfonation
is more prevalent  in  PMS  (12).  The PMS used in this mixture was  unsuit-
able for spinning  with PP and NBR because of its  low melting  point.

As a general rule  the difficulty  in spinning fibers increased with  the
butadiene content  partly  because  of instability of  the  mixtures  toward
hot working and partly because of poor  fiber forming properties  of  the
butadiene copolymers.   Although  improvement  in spinning behavior was

                                  23

-------
seen by decreasing the mixing time prior to spinning (for example,  com-
pare 40B with 40B(2) in Table 8),  the effectiveness of crosslinking
appeared to be enhanced by thorough mixing.  It is reasonable to assume
that the most efficient crosslinking which can be achieved by a given
amount of butadiene will come from a styrene-butadiene copolymer rather
than a blend of polystyrene and polybutadiene.'  For1 this reason styrene-
butadiene copolymers were specially made for mixtures MS-WP-7A, -29A,
-29B, -30A, -32A, -45A.  These copolymers ranged from 72% styrene,
28% butadiene to 90% styrene, 10% butadiene.  None of the mixtures  could
be spun.  MS-WP-7A showed evidence of being partially crosslinked as the
mixture caused very high extruder pressures.  Neither SB-1 (90% styrene,
10% butadiene) nor SB-2 (80% styrene, 20% butadiene) could be spun
because of thermal instability.  The addition of stabilizer in MS-WP-45A
did not noticeably improve the spinning properties of SB-2.  Low
molecular weight and perhaps residual monomer contributed to the insta-
bility of SB-1 and SB-2 which decomposed below 200C.  Furthermore,
even though the polymerization was taken only to 85% conversion, there
was evidence of excessive branching.  Nevertheless, copolymers of this
type should ultimately lead to improved sulfonic acid ion exchange
fibers.

Mixtures Containing Epoxy Resins

Epoxy resins (ER) provide another promising method for crosslinking in
fibers containing sulfonated polystyrene.  The epoxy resin can be cross-
linked with amines, peroxide, or possibly sulfuric acid itself.  More-
over, the phenyl groups in the ER can be sulfonated, thus minimizing the
loss in capacity due to the crosslinking agent.  The miscibility of ER
with PP-PS or PP-ABS mixtures is limited to about 10% ER (see Table 4).
Attempts to include both epoxy resin and SBR with PP-PS (MS-WP-28A and
-28B) were unsuccessful.  However, the ability to spin such mixtures is
extremely sensitive to variations in the control of the spinning extruder
as well as thermal history of the mixture.  The principal difficulty in
spinning epoxy resin was due to its low viscosity at typical PP-PS melt
temperatures; thus, higher molecular weight ER is necessary for better
results.

Mixtures for Weak Acid Cation Exchangers

As shown in Table 5, many attempts were made to spin weak acids and
anhydrides for use as weak cation exchangers.  None of the mixtures
could be melt spun.  Polyacrylic acid (PAA) could not be spun with
either PP or EMA (sodium salt of ethylene-methacrylic acid copolymer)
nor could the sodium salt of PAA be spun with these materials.  The
problems in melt spinning mixtures containing PAA were probably the
ionic character of the acid and the possibility of crosslinking through
formation of the acid anhydride.  The mixtures containing PAA which
showed the best spinning behavior, MS-WP-18A and -18B, were prepared
for spinning by mixing the PAA (25% aqueous solution) with powdered PP
before dehydration, thus decreasing the chances of crosslinking via acid
anhydride formation.

                                  24

-------
As an alternative to the ionic PAA, the spinning of several maleic
anhydride copolymers was attempted with the aim of hydrolyzing the
anhydride after spinning.  GMA (copolymer of methyl vinyl ether and
maleic anhydride) could not be spun with PP because of its decomposition
at ordinary PP melt temperatures.  Both the 1:1 and 2:1 copolymers of
styrene and maleic anhydride (SMAT1 and SMA-2) were tested for compat-
ibility with PP or EMA.  None of these mixtures could be spun, but the
2:1 (styrene:maleic anhydride) copolymer had better extruding behavior.
Although the SMA copolymers are fiber forming, the fiber pulverizes with
a small amount of flex or compression, extruded filaments from the SMA
mixtures in Table 5 had physical properties similar to pure SMA.

The only successful approach to making a weak acid cation exchanger was
the mixture MS-WP-41A containing PMM (methyl methacrylate-styrene-
butadiene blend).  After spinning, the methyl methacrylate can be
hydrolyzed to give methacrylic acid.  The fiber (MS-WP-41A) was quite
weak with only 50% PMM, but increased PMM content and increased strength
will undoubtedly be possible.

Anion Exchange Fibers Containing SDM

The anion exchange fibers had two basic types of active ingredients -
polyvinylpyridines and polyamines.  In previous work (13) it was demon-
strated that SDM (styrene-dimethylaminopropyl maleimide copolymer) could
be cospun with PP in the ratio 60% SDM, 40% PP.  This .fiber, RF-PB-129D,
has excellent physical properties and can be crosslinked with dibromo-
alkanes; but because of the high equivalent weight of SDM (286 gequiv),
the maximum calculated ion exchange capacity, 2 meq/g, is low.  Therefore,
attempts were made to spin fiber with PVP (1:1 copolymer of 2-vinyl-
pyridine and 2-methyl-5-vinylpyridine) whose equivalent weith is
112 g/equiv, about 40% that of SDM.

Vinylpyridine Anion Exchange Fibers

PVP and PP mixtures are compatible for spinning only in the extremes of
very low concentration of PVP or high concentrations, about 80% PVP or
greater (see Table 6).  In the latter case the fibers are quite weak.
PVP neutralized with acetic acid  (MS_WP-5A) was equally incompatible
with polypropylene.  Polystyrene has a marked effect on the ability to
spin PVP with polypropylene.  Although the 4:6 PP-PVP mixture  (MS-WP-1C)
could not be spun, the mixture MS-WP-4A  (3:1:6 PP-PS-PVP) gave a  fiber
of good appearance, fair tensile strength, and slight cold draw capa-
bility.  As little as 5% polystyrene gave excellent physical properties
to the fibers.  With 2.5% PS  (MS-WP-4C and  -4D) the fibers were quite
weak and could not be drawn.  Mixtures with ABS in the place of PS
 (MS-WP-12A to -12E) were more easily spun and of comparable strength.
None of the mixtures with PS could be drawn; but one fiber with ABS,
MS-WP-12E, was drawn to improve the strength.

Nylon 6 and PVP are compatible for spinning over a wide range of  concen-
trations  (MS-WP-2A and-2B).  Fiber MS_WP-2B has good tenacity but no

                                  25

-------
appreciable elongation at break; the fiber could not be drawn.  EMA,
which is a compatibilizer for SDM and PP, does not aid in the spinning
of PVP and PP as shown by mixture MS-WP-3A.

Since polystyrene is a compatibilizer for PVP and PP, one might expect a
styrene-vinylpyridine copolymer (SVP) to be compatible with PP.  Such is
the case with MS-WP-6A.  This fiber, however, is not particularly
superior to those with PVP and has the disadvantage of a high monomer
equivalent weight compared with PVP.  If, on the other hand, SVP is used
in small amounts in an effort to make PP and PVP compatible (that is, in
the same capacity as PS or ABS in MS-WP-4A to 4D or MS-WP-12A to 12E),
the mixtures cannot be spun.  Thus, MS-WP-13B, which had the same poly-
propylene, styrene, and vinylpyridine concentrations as MS-WP-4A but in
different polymer form, could not be spun.
                             \
In the compounding of polyamines with elastomers, zinc oxide is found to
have a curing effect on the polyamines.  Thus, it is plausible that ZnO
might provide crosslinking in PVP and other polymeric amines used for
ion exchange fibers.  The mixture MS-WP-14A containing 5% ZnO behaves as
a crosslinked material when extruded (manifested by a high pressure on
the extruder die), but the extruded material swells rapidly and loses
PVP in acidic solution.  Based on these experiments it is unlikely that
ZnO can be useful in the prevention of PVP extraction.

Since it is known that PVP is a cure for epoxy resin, the effect of
adding a small amount of ER to PP-PS-PVP mixtures was investigated.  The
mixture MS-WP-15A with 2.5% ER had excessive crosslinking and could not
be spun.  It appears that in order to have a spinnable mixture the degree
of crosslinking must be so low as to be practically negligible, and there-
fore, all crosslinking must be effected after spinning.

Mixtures Containing Polyethyleneimine

Polyethyleneimine (PEI) is not compatible for spinning with PP.  In mix-
tures MS-WP-16A and -17A (Table 7) it was found that unlike the case
with PVP neither PS nor ABS improved the spinning characteristics of
PP and PEI.  Furthermore, the series MS-WP-19A to -19C showed that EMA
is not compatible with PEI nor does it compatibilize PP and PEI.  As
seen by mixtures MS-WP-44A to -44E, PEI is compatible with PVP or at
least is compatible with a PP-ABS-PVP mixture.  It would appear then
that a four component system is necessary for spinning PEI along with PP;
that is, compatibility of PEI requires PVP which is in turn compati-
bilized with PP by ABS.  The substitution of PEI for PVP should lead to
higher capacities since the PEI equivalent weight is only about 35% that
of PVP.  Up to 30% PEI has been spun with PP, ABS, and PVP; this fiber,
MS-WP-44D, has a maximum calculated amine content of 12 meq/g.
                                  26

-------
Miscellaneous Fiber Mixtures and Treatments

The polyaminostyrene (PAS) - polypropylene mixtures could not be extruded
(MS-WP-9A, -9B, -10A).  The polyaminostyrene was likely impure or cross-
linked since it caused very high pressures at the extruder die.

Fibers for which crosslinking was attempted by irradiation are listed in
Table 10 along with dosages.  The effect of the irradiation on the anion
exchange fibers was too small to be useful.  Likewise none of the fibers
containing polystyrene were appreciably crosslinked.  Crosslinking via
irradiation probably has little merit for polypropylene-polystyrene mix-
tures as in most of the fibers in Table 10.  This is due to the fact
that sufficient irradiation to crosslink the polystyrene at the same
time seriously degrades the polypropylene.

                               Table 10

                        Irradiation of Fibers

                                     Dosage (watt-hour)

           Sample No.          8        16        24        32

           MS-WP-2B            x        -         x         -
           MS-WP-4B            xx--
           MS-WP-OB            x        x
           MS-WP-OC            x        -         x         -
           MS-WP-OD            x        -         x         -
           MS-WP-8B            x        x         x         x
           MS-WP-8C            x        x
           MS-WP-8D            x        x         x         x

               x indicates irradiated samples

EVALUATION OF CATION EXCHANGE FIBERS

Sulfonic Acid Exchangers

Studies of the sulfonation of polystyrene and polypropylene-polystyrene
fibers indicated that although a crosslinking agent is definitely needed
for such fibers, polypropylene itself holds some of the sulfonated poly-
styrene in the fiber.  Samples of MS-WP-OB  (1:1 PP-PS) were sulfonated
using (1) concentrated sulfuric acid, (2) fuming sulfuric acid  (30% S0_)
and (3) chlorosulfonic acid.  Concentrated sulfuric acid was ineffective
in producing a high degree of sulfonation in the fiber; the maximum
capacity obtained was 0.5 meq/g  (dry fiber in the acid form).   Hot
sulfuric acid gave some crosslinking but the loss of polystyrene sul-
fonic acid was faster than the rate of crosslinking.  The action of
fuming sulfuric acid on the fiber permitted a higher degree of  sulfon-
ation, but with increasing sulfonation there was a corresponding loss of
weight in the fiber due to increased water solubility of the sulfonated
polystyrene.  Moreover, several regenerations brought marked decrease in

                                  27

-------
the total capacity of the fibers treated with fuming sulfuric acid.   The
highest degree of sulfonation in the fiber, 3.0 meq/g, was achieved  with
chlorosulfonic acid.  The capacity was unchanged after three regener-
ations indicating that some crosslinking had taken place.   In separate
experiments polypropylene fiber samples were treated with fuming sulfuric
and chlorosulfonic acids.  In neither case was sulfonation of the poly-
propylene fiber detected.  Thus, it is calculated that the fiber
described above having a capacity of 3.0 meq/g has 78% of its aromatic
rings sulfonated.  Since 10% sulfonation would render the polymer water
soluble, the chlorosulfonic acid treatment must have caused sufficient
crosslinking to maintain the 3 meq/g capacity.  Presumably the cross-
linking is accomplished by means of -S- or -S02- bonding.   Subsequent
experiments showed that this type of crosslinkage is not very stable to
prolonged exposure to aqueous solution.

The sulfonation of MS-WP-OC and OD (80% PS and 90% PS, respectively)
resulted in capacities up to 4 meq/g.  For these samples the crosslinking
afforded by chlorosulfonic acid sulfonation was clearly inadequate.
Swelling of the fiber was excessive; and after several hours in solution,
a significant amount of the sulfonic acid was leached from the fiber.
The presence of peroxidized polypropylene in MS-WP-OC did not result in
a noticeable increase in crosslinking.

The substitution of acrylonitrile-butadiene-styrene for part or all  of
the polystyrene reduced the loss of styrene sulfonic acid from the fiber
even without any special efforts to crosslink the butadiene.  Samples of
MS-WP-8A, -8B, -8C, and -8D were sulfonated with chlorosulfonic acid for
two minutes.  After rinsing, the ion exchange capacity was determined and
the fiber was regenerated.  Four cycles of titration and regeneration are
shown in Table 11.  The "capacities" are listed in terms of meq of acid

                               Table 11

               Effect of Titration-Regeneration Cycles
                           on the Capacity

                          MS-WP-8A     MS-WP-8B     MS-WP-8C     MS-WP-8D
No. of Cycles             meq acid     meq acid     meq acid     meq acid

     1                      5.91         6.60         9.46         5.12
     2                      4.53         5.03         7.78         4.73
     3                      4.13         4.74         6.80         4.20
     4                      4.11         4.49         6.42         4.06

Sample weight after
cycle No. 4  (g)             1.455        1.920        2.124        1.555

titrated rather than meq/g because the samples were dried and weighed
only after the fourth regeneration.  The apparent decrease in capacity is
almost certainly accompanied by a decrease in sample weight due to the
loss of highly sulfonated polystyrene.  Thus, the change in total

                                  28

-------
capacity expressed in meq/g is smaller than the change in meq acid
titrated.  However, since in the evaluation of ion exchange performance
the sample weight for the final cycle determines the maximum capacity
for succeeding cycles, an "apparent capacity" defined as the meq of acid
titrated divided by the sample weight during the final cycle may be used
to describe the change in sample performance with the number of titra-
tion-regeneration cycles.  This is shown in Figure 1 which is essentially
a plot of the data of Table 11 normalized so that the ordinate for the
final titratiori is the true'total capacity of the fiber in meq/g.  With
regard to swelling, the fibers having more ABS have less swelling.

A more extensive study of the change in capacity with cycles of exhaus-
tion and regeneration is shown in,Figure 2.  The sample, MS-WP-8D
(3:1 draw), was sulfonated with H2SO^'15% S03, and its total capacity
was followed through 25 titration-regeneration cycles.  The sample was
dried and weighed only after the twentieth cycle; all the capacities of
Figure 2 are calculated on"the basis of that weight.  The scatter in the
data (probably due mostly to variations in the degree of regeneration)
does not permit establishment of a definite trend in the total capacity.
From the second through the twentieth cycle, there is no change in total
capacity within experimental error, but the five cycles subsequent to
drying indicate a loss of capacity.  The average capacity of the sample
through 20 cycles (excluding the first) was 2.91 meq/g dry fiber.  After
the 25 cycles there was no visible deterioration of the fiber.

In other sulfonation experiments on the PP-PS-ABS fibers it was found
that the total capacity is not simply related to the time of exposure to
the sulfonating agent.  Rather, the capacity increases with time of sul-
fonation to a maximum and then decreases.  For example, a sample of
MS-WP-8D (3:1 draw) sulfonated in C1S03H for 2 minutes had a total
capacity of 2.5 meq/g whereas another sample of the same fiber exposed
to C1S03H for 8 minutes had capacity of 2.4 meq/g.  The reason for
this behavior is the loss of highly sulfonated polystyrene from the
fiber which eventually competes with the rate of sulfonation.  There-
fore, the PP-PS-ABS fibers  still have insufficient crosslinking.

Column Elutions of Sulfonic Acid Exchangers

Improved sulfonic acid cation exchanger fiber was obtained from the sul-
fonation of MS-WP-8F  (25% PP, 75% ABS) with  H2S04'157o S03.  This  fiber
was  finely cut and placed in a buret for evaluation.  By elution with
excess NaCl and titration of the liberated acid the total capacity was
determined to be 3.16 meq/g dry  fiber.  The  column was  regenerated with
H2S04 and eluted with 0.004 N NaCl at 6.5 ml/min.  The  effluent was
collected periodically by the emptying of a  siphon cup, and  the acid  con-
tent of  each  fraction of effluent was determined.  The  breakthrough curve
for  this elution is shown in  Figure  3 where  the ratio of effluent to
influent concentrations  is  plotted against the number of siphon cup
fills.   The volume of effluent emptied from  the siphon  cup had unex-
pectedly large variations  resulting  in a random error for each measure-
ment.   However,  the error  is not cumulative  so that the sense of  the

                                   29

-------
     4.5
     4.0
bO

D"
.U
H
U
o,
TO
4J
c
01
tl
(0
(X
a,
     3.5
3.0
     2.5
                                      O  MS-WP-8A
                                            -WP-8B
                                I
                                               I
                                2                  3

                                    No. of Cycles


                                       Figure 1

                   Effect  of the Number of Titration-Regeneration

                                Cycles on the Capacity
                                   30

-------
           3.4
       00

       cr
u>
4J
H
O

g,

3
       (0
       4J
       O
       H
                              O

                                                 10
                                                              15
                                                       no. of cycles
                                                          Figure 2

                                       Total  Capacity of Sulfonated PP-PS-ABS Fiber
20
                                                                                                             25

-------
1.0
0.8
0.6
0.4
0.2
  0
     -O
                            10         20          20                      30
                                Effluent Volume (X 0.0184 ml.  0.004 N NaCl)
                                           Figure  3
                    Breakthrough Curve for Sulfonated PP-ABS Fiber (1-141)
40

-------
curve is accurate; since a total of 1850 mg NaCl solution was passed
through the column, the average volume per siphon cup fill was 54 ml.
The breakthrough capacity, 1.5 meq/g, was slightly less than half the
total capacity under these operating conditions.

Fiber such as MS-WP-8F, which was used for the column evaluation of
Figure 39 has enough swelling to seriously degrade its physical proper-
ties.  In fibers with higher degrees of sulfonation the swelling was
correspondingly greater and in some cases sufficient to cause hydraulic
problems in column operation.  In fiber containing epoxy resin (for
example, MS-WP-23A) swelling was reduced by curing the epoxy with
diamines prior to sulfonation.  The sulfonation process alone did not
give appreciable crosslinking.

Further improvements in crosslinking were obtained with fiber MS-WP-26A
(20% PP, 70% PS, 10% SBR) by treatment with t-butyl peroxide.  The cross-
linking afforded by this procedure greatly reduces the swelling of the
sulfonated fiber in comparison with the uncured sulfonated fiber.  Sul-
fonation of the cured fibers with funing sulfuric acid gave total capac-
ities up to 4.0 meq/g.  However, during the sulfonations the loss of
sulfonated polystyrene was evident and the total capacities decreased by
several percent after the first cycle of titration and regeneration.

In contrast with MS-WP-26A, the peroxide curing of MS-WP-21 (ABS fiber)
was much more effective even though the butadiene content was about the
same in both fibers.  Very little sulfonated polystyrene was lost during
the sulfonation and the swelling of the ABS was less than that of
MS-WP-26A.  A finely cut sample (2 g) of cured and sulfonated ABS fiber
was placed in a column and eluted with 0.05 N CaC^.  The results of
three elutions at various flow rates are given in Figure 4.- Total capac-
ities were obtained from the area above the breakthrough curves.  By way
of comparison the same solution was passed through a sample of equal
weight of commercial resin (Rohm & Haas IR-120) in an identical column.
One such elution is shown in Figure 5.  The total capacity of the fiber
was 4.2 meq/g compared with 4.6 meq/g for the resin beads.  On the other
hand, the breakthrough capacity of the fiber at 8.4 ml/min was 3.8 meq/g
compared with 1.3 meq/g for the beads at a flow rate of 6.0 ml/min.  An
even greater difference between the fiber and beads was evident at higher
flow rates.  At 26.8 ml/min the breakthrough capacity of the fiber was
3.4 meq/g whereas at 32.6 ml/min the breakthrough capacity of the beads
was 0.3 meq/g.  The fiber and bead capacities are summarized in Table 12.

                             Table 12

            Breakthrough Capacities for CaCl0 Elutions

Flow Rate
(ml/min)
3.4
8.4
26.8
ABS Fiber
Breakthrough
Capacity
(meq/j)
3.8
3.8
3.4
"Rohm & Haas IR-120
Flow Rate
(ml/min)
2.5
6.0
32.6
Breakthrough
Capacity
(meq/g)
2.9
1.3
0.3
                                   33

-------
                                                                          10,
OJ
              1.0
              0.8
              0.6
          c/c
              0.4
              0.2
                                                  _	
                                                  //    i
                              3.4 ml/min
                                                          8.4 ral / rain
                                                                  i	//
                                     9   7/    7  ^              9    7/


                                                    Influent  CaCl,,  (meq)


                                                          Figure  4



                                    Breakthrough Curves for Sulfonated ABS  Fiber (1-177)
10

-------
co
             1.0
             0.8  _
             0.6  -
       C/C,
             0.4  -
             0.2  -
                                                        8
12
                                                  Influent CaCl- (tneq)
                                                     Figure 5
                                 Breakthrough Curve for Rohm & Haas IR-120 Resin
16

-------
In the case of volume capacities the fiber would not compare so favorably
with the commercial resin.  The swelling of the fiber was about four
times that of the beads which to a large extent accounts for the high
ratio of breakthrough to total capacity in the fiber.  Nonetheless,
these shallow bed elutions indicate the probability of a higher rate of
exchange in the fiber.

The most direct approach toward reducing the swelling of the cation
exchange fibers in aqueous solution is to increase the butadiene content
since the degree of crosslinking obtained in the peroxide or sulfuric
acid cures is proportional to the butadiene content (swelling is also
reduced by incomplete sulfonation but at the expense of decreased ion
exchange capacity).  The fiber having the greatest percentage of butadiene
was MS-WP-34B with 18.7%.  A finely cut sample of this fiber, cured  in
sulfuric acid and sulfonated with fuming sulfuric acid, occupied a volume
of 5 ml per gram of fiber in a column.  The column was eluted with 0.02 N
CaCl2-  The breakthrough curves are shown in Figure 6.  In the elution at
3.0 ml/min the breakthrough capacity was 96% of the total capacity,  3.8
meq/g.  At the higher rate of 15.3 ml/min the breakthrough capacity  was
still 92% of the total capacity.  The breakthrough capacity of this  fiber
exceeds that of sulfonated styrene-divinylbenzene resin beads (Rohm  &
Haas IR-120, see Table 12) both on a dry weight and volume basis.  How-
ever, the swelling of this fiber in aqueous solution still causes unde-
sirable loss in strength and durability.  Much of the weakness of this
cation exchange fiber can be attributed to the original fiber which  was
weak and could not be drawn.  Therefore, the need for a compromise
between high butadiene content and improved fiber strength is evident.

In fiber MS-WP-40B (25% PP, 60% PS, 15% NBR) the butadiene content of
12.8% was less than the 18.7% of MS-WP-34B but the fiber could be drawn
to give a tenacity of 1.8 g/denier compared with 0.2 g denier for
MS-WP-34B.  A sample of this fiber, hand cut to an average length of
3 mm, was cured with sulfuric acid and then divided into four parts  for
sulfonation with ^804-7.5% SOg for varying lengths of time.  The total
capacity is given for each sulfonation time in Table 13.  Also listed in

                             Table 13

                     Sulfonation of MS-WP-40B

          Sulfonation          Total Capacity          V + AV
          Time (min)               (meq/g)                V

                5                   2.75                 2.65
               10                   3.96                 4.86
               20                   4.16                 7.18
               40                   4.19                16.14

Table 13 is the volume swelling factor, (V + AV)/V, calculated from
changes in diameter on the assumption of negligible swelling in the axial
direction.  Figure 7 shows the relationship between the swelling factor

                                  36

-------
   1.0
   0.8
  0.6
C/C
  0.4
  0.2
                   3.0 tnl/min
                                                                              15.3 ml/tnin
      17      18
19       20
ri        7r
Influent CaCl,
17       18
19
20       21
                                             Figure 6

                         Breakthrough Curves  for Cation Exchange Fiber II-9

-------
U)

00
             60
            ^
             cr
             o
             n)
             a
             CD
             o
            4-1
            o
            H
                 4.0
3.0
                 2.0
                   (5 min.)
                                                                              12
                                                                    V +
                                                 Swelling Factor
                                                                            e-

                                                                             (40  min.)_
                                                                                 16
                                                     v
                                                       Figure 7


                                    Total  Capacity versus Swelling for MS-WP-40B

-------
and the total capacity.  The increase in swelling after the leveling of
the total capacity is attributed to a loss of highly sulfonated poly-
styrene while the sulfonation proceeds.  It appears that the maximum ion
exchange capacity represents a steady state balance between continuing
sulfonation and extraction of highly sulfonated polystyrene.  Thus, the
increase in swelling is related to an increase in degree of sulfonation
of the styrene remaining in the fiber.  Figure 7 points to the importance
of a strictly controlled sulfonation for this type of fiber in order to
obtain high capacities with minimum swelling.

Selectivity Coefficients for Sulfonic Acid Exchangers

In accordance with the results of Table 12 a sample of knitted MS-WP-40B
was cured with H2SO^ and sulfonated with chlorosulfonic acid to a total
capacity of 3.5 meq/g.  The fabric was then finely cut and placed in a
column for evaluation.  Breakthrough curves for calcium elutions are
shown in Figures 8 and 9.  The column was also eluted with a solution
containing ammonium ion in addition to calcium ion.  Breakthrough curves
for calcium and ammonium are given in Figures 10 and"11, respectively.
The analysis of ammonium ion in the effluent presented serious reproduc-
ibility problems.  Although it is difficult to extract quantitative
results from Figure 11, some information may be derived by comparison
with Figures 8, 9 and 10.  It is seen from Figures 8 and 9 that the
total capacity of the fiber is fairly constant, i.e., about 8.8 meq; the
same capacity is therefore expected for Figure 10.  Substracting the
calcium absorbed in Figure 10 from the total capacity (assumed to be a
maximum of 8.8 meq) of the fiber gives the maximum amount of ammonium
ion absorbed in the column, 6.525 meq.  This does not agree with the
ammonium ion sorption of Figure 11.  The difference between the two
results is represented graphically by the area enclosed by the square
in Figure 11.  Clearly, because of the scatter of the data, this area
could be accounted for by drawing a different curve - especially between
4.0 and 5.0 meq on the abscissa.  The first sample analyzed for ammonium
ion was at 4.86 meq NH^*, just prior to calcium ion breakthrough.  How-
ever, it appears that ammonium ion breakthrough may have occurred well
before detection of calcium ion breakthrough.

The selectivity coefficient Kg may be calculated from the steady state
absorptions on the column and influent concentrations.  Selectivity
coefficients are variously defined in terms of different concentration
units.  Here we use the molal selectivity coefficient defined as
follows:
                                            z
                                             A
                        B                  A
                                  39
                                               t

-------
C/C
                                       Influent Calcium (meq)
                                        Figure 8
                          Sulfonated MS-WP-40B,  First Elution

-------
           1.0
           0.8
                                                    '
c/c
           06
           0.4
           0.2
                                                o
                                                           _L
                                    Influent Calcium (meq)
                                           Figure 9
                             Sulfonated MS-WP-40B, Second Elution
                                                                       10

-------
      c/c
NJ
              1.0
              0.8
              0.6
              0.4
              0.2
              o.4 r i
                                                                                  ,
      ^78                       9
                                Influent Calcium  (meq)
                                   Figure  10
Calcium Ion Breakthrough for Ammonium-Calcium Elution of Sulfonated MS-WP-40B
                                                                                                      in

-------
3.0
  0
                    567
                     Influent Ammonium (meq)
                            Figure 11
          Ammonium Ion Breakthrough for Ammonium-Calcium
                 Elution of Sulfonated MS-WP-40B
                             43
                                                    LIBRARY U.S.

-------
where in and m are molalities in the absorbed and liquid phase respec-
tively, and z is the valence.  The selectivity coefficient as defined
above is not an equilibrium constant; the equilibrium constant has the
same form as the molal selectivity coefficient but is expressed in terms
of activities.  Therefore, the selectivity coefficient may be expected
to vary with the relative concentrations of A and B and with the total
concentration of A and B.

Using the results of Figure 10 and assuming a total capacity of 8.80 meq
for the column, the molal selectivity coefficient may be calculated:

                                          (.005846)  _
                    -H-  ~  8.275         (. 004243 )2 ~   '
                   Ca

It can be inferred from the above relationship that in the case of solu-
tions having ammonium ion in concentrations of several parts per million
and comparable calcium concentration, the fraction of ammonium ion
retained by the column at saturation would be practically negligible.
This situation is generally typical of ion exchangers whereby the
divalent ion is preferentially absorbed.

Methacrylic Acid Exchangers

In some preliminary experiments with fiber MS-WP-41A which contains poly-
methylmethacrylate (PMM) , it was demonstrated that hydrolyzed PMM, i.e.,
polymethacrylic acid, can serve as a weak acid ion exchanger.  The PMM
was hydrolyzed and the polybutadiene in the fiber was crosslinked by
exposure to concentrated sulfuric acid.  Because of the presence of
styrene in the fiber which was partially sulfonated by the sulfuric acid,
the resulting ion exchanger was a combination weak acid and strong acid.
The fiber had a total capacity of 1.0 me.q/g and a salt splitting capac-
ity of 0.3 meq/g.  Thus, methacrylic acid accounted for 0.7 meq/g.
Higher capacities should be possible with more complete hydrolysis of
the polyester and with a greater fraction of ester in the fiber.

EVALUATION OF ANION EXCHANGE FIBERS

SDM Fibers

Fiber RF-PP-129D containing 60% styrene-dimethylaminopropylmaleimide
(SDM) required 1.0 meq/g of HC1 for complete neutralization.  Based on
the starting composition of the mixture to be spun the calculated
capacity of SDM is 2.1 meq/g.  The SDM was successfully crosslinked with
1 ,12-dibromododecane, but the resulting capacity was only 0.7 meq/g.
Because of the low theoretical capacity of this SDM fiber and the still
lower capacity actually obtained, further evaluation was discontinued
in favor of working with materials having lower equivalent weights.
                                  44

-------
Vinylpyridine Fibers

Several attempts were made to crosslink fibers containing polyvinyl-
pyridine (PVP) with epoxy resin (ER).  In separate experiments with pure
PVP it was demonstrated that ER and PVP react rapidly at 80C.  Thus, a
controlled coating of ER on the fiber followed by heating to react the
epoxy with the PVP could form a shell which would prevent loss of PVP
but still allow for water permeability and a small amount of swelling.
Alternatively the epoxy resin could be allowed to diffuse throughout the
fiber prior to heating in order to obtain uniform crosslinking.  Both
the diffusion of the ER into the fiber and the attainment of a uniform
coating of ER on the fiber were difficult to achieve.  For example, a
sample of MS-WP-2B was epoxy coated and cured at 75C for an hour.  The
ion exchange capacity of the fiber decreased from 2.9 meq/g on the first
exhaustion-regeneration cycle to 1.1 meq/g on the second cycle.  Cross-
linking by the epoxy coating method appears to be less desirable than
treatment with dihaloalkanes because of the relative simplicity of the
latter.

Fibers containing PVP were adequately crosslinked with mixtures of
1,6-dibromohexane and 1,12-dibromododecane but the capacities were
generally low.  For example, fiber MS-WP-12A (25% PP, 5% ABS, 70% PVP)
treated with the dibromides gave a capacity of less than 1 meq/g.  By
first swelling the fiber with dilute HC1 and then reacting it with an
alkyl dibromide mixture, the capacity was improved to 1.9 meq/g.  How-
ever, the swelling procedure resulted in some loss of the polyvinyl-
pyridine before crosslinking was effected.  Other swelling agents were
sought in order to minimize the loss of vinylpyridine from the fiber.
For example, a sample of MS-WP-12A was exposed to a solution of dibromo-
halides to which were added several drops of 2-pentanol.  The fiber was
then reacted with a n-heptane solution of 1-bromobutane and 1,10-
dibromodecane; again several drops of 2-pentanol were added to swell the
fiber.  This treatment resulted in a total capacity of 3.2 meq/g and a
salt splitting capacity of 1 meq/g.  Further increases in capacities
should be possible since a direct titration of fiber MS-WP-12A with HC1
requires 3.6 meq HC1 per gram of dry fiber.

A sample of MS-WP-12A was crosslinked with a mixture of 1,12-dibromodo-
decane and 1,10-dibromodecane, quaternized with 1-bromobutane, finely
cut, and placed in a column.  The column was eluted with 0.025 N HC1,
and the chloride ion concentration in the effluent was determined by
silver nitrate titration.  In the first elution with HC1 the break-
through capacity was 1.1 meq/g for an average flow rate of 1.6 ml/min.
It was clear  in the course of this elution that channeling was a serious
problem - easily observed since the hydroxide and chloride forms have
different colors.  The fibers, 1 to 4 mm in length, were only loosely
packed and in some places were matted.  The column was therefore care-
fully repacked, the volume occupied by the fiber being reduced by nearly
a factor of two.  In the second elution at 1.8 ml/min the channeling
was significantly diminished although still present.  The breakthrough
curve for this elution is shown in Figure 12.  The breakthrough capacity

-------
                                                                      o
    1.0
    0.8
    0.6
C/C.
    0.4
    0.2
      Breakthrough
    Capacity
1.6 meq/g
                                                                          -e-
                               10                     15

                                     Influent HCl (meq)
                                         Figure 12

                    Breakthrough Curve for Anion Exchange Fiber (1-159)
                                                               20

-------
is 1.6 meq/g; and the total capacity, determined from the area over the
breakthrough curve, is 2.2 meq/g (dry fiber in the chloride form).   A
third elution at 0.5 ml/min resulted in the same breakthrough capacity
as for the second elution at 1.8 ml/min.  The breakthrough capacities
were probably decreased considerably due to the poor packing of the
fibers in the column.

Improvement in packing efficiency was made by decreasing the average
fiber length to about 1 mm.  Total capacity was increased by crosslink-
ing with lower molecular weight alkyldihalides and by using fiber with
a higher fraction of PVP.  Thus, a finely cut sample of ^fiber MS-WP-4B
(15% PP, 5% PS, 80% PVP) was crosslinked with a heptane solution of
1,6-dibromohexane and further quaternized with a n-butyl bromide.  A 4 g
sample of this fiber, packed in a 100 ml buret, occupied about 5.4 ml/g
fiber.  The large volume to weight ratio was mainly attributable to poor
packing efficiency rather than swelling.  The column was eluted at flow
rates from 3.5 ml/min to 24.9 ml/min with 0.02 N HC1.  A maximum capac-
ity of 2.9 meq/g was found for this sample.  The breakthrough curves are
shown in Figure 13.  In general, the breakthrough curves are normal in
that the efficiency of ion exchange increases with decreasing flow rates.
However, as seen from Table 14, the total capacity decreased with the
number of elutions or alternatively the total capacity decreased with
increasing flow rate.  The reason for this behavior could not be ascer-
tained since the column was dismantled after the third elution.

Fiber MS-WP-12E (15% PP, 5% ABS, 80% PVP) yielded the highest capacity
of all anion exchange fibers tested.  Short lengths of the fiber were
obtained by cutting a knitted sample of the drawn fiber.  The fiber was
then crosslinked in a decane solution of 1,6-dibromohexane and 1,10-
dibromodecane  (seven days at 45-55C).  After washing with methanol and
drying, the sample was placed in a column  (sample No. 11-52).  The
results of an elution with 0.0157 N HC1 are shown in Figure 14.  The
breakthrough capacity is about 86% of the total capacity, 3.81 meq/g.
The higher breakthrough capacity was a .result of improved column pack-
ing efficiency  (less than 4 ml/g fiber).

Selectivities for Vinylpyridine Fibers

Sample 11-52  (MA-WP-12E) was eluted with a solution containing both
chloride 0.0190 N and phosphate 0.0012 N E^PO^') at pH = 1.7.  The
results of this elution are shown in Figure 15  for chloride ion and
Figure 16 for  phosphate ion.  At this acidity the phosphorus is assumed
to consist almost entirely of the monovalent ion, H2PO,".  In this case
the selectivity is highly in favor of chloride  ion.  Pertinent data
for this elution is given in Table 14.

The influent solution for the third elution of  sample 11-52 also had
phosphate and chloride ion, the pH of the  solution being about 2.6.
Breakthrough curves  for chloride and phosphate  ions are  shown in
Figures 17 and  18, respectively.  Examination of the results for this
elution  (Elution No. 3 in Table 14) show that if the sorbed phosphorus

                                  47

-------
oo
                            1     i  w I    7/~T    f
                                  3.6  ml/min
8.7 ml/min
                                                   j	I     I
                                         II       10        12
                                               Influent HCl (meq)
                                                   Figure 13
                               Breakthrough Curves for Anion Exchange Fiber II-2
                              10       12        14

-------
                                             Table 14

                             Elutions for Anion Exchange Fiber 11-52

Elution No.                        1        2         3       4        5        6        7        8

Chloride Absorbed on
Column (meq)*                    13.72    13.91     6.88    13.49    13.41     13.41      3.62     6.51

Phosphate Absorbed on
Column (mm)**                              0.25     3.68

Nitrate Absorbed on
Column (meq)                                                                            9.79     6.90

Chloride Cone, (meq/liter)       15.66    18.96     3.544   15.73    15.73     15.73     8.221    3.637

Phosphate Cone, (mm/liter)                 1.25     3.196

Nitrate Cone, (meq/liter)                                                              10.28    4.548

Total Capacity (meq/g)            3.82     3.9      3.7      3.74     3.73      3.73


     K^PO                                3.62



     KSo4                                         -0033

     KN3
      C1J                                                                               2.2      0.85

Ave. Elution Rate (ml/min.)      12.4     15.6     13.6     16.7     13.5     12.3     14.1     14.7
*meq = milliequivalents
**mm  mi11imo1es

-------
Ui
o
            C/C
                  1.0
                  0.8
                  0.6
                  0.4
                  0.2
                                                                      I         I
     ave.  elution rate =  12.4 ml./min.
        sample weight = 3.60 g
                                   -e'   o o
       10                  12                  14

                            Influent HCl  (meq)

                           Figure 14

Breakthrough Curve for Anion Exchange Fiber 11-52 (MS-WP-12E)
                                                                                        16

-------
C/C
      1.0
     0.8
      0.6
     0.4
     0.2
      0
0,'n   Q
                      I
                                               \
_L
  10
                                                                          16
                    12                  14
                           .Influent  Chloride  (meq)
                         Figure 15
Chloride Ion Breakthrough for Second Elution of Sample 11-52
                                      18

-------
3.0
                                 0
2.0
1.0
                                _L
I
   0.6                0.8                1.0               1.2
                         Influent Phosphate (meq)
                          Figure 16
        Phosphate Ion Breakthrough for Second Elution of Sample 11-52
                              52

-------
  1.0
  0.8
  0.6



C/C
   o




  0.4
  0.2
    0
                               6                       8                     10

                                  Influent Chloride (meq)

                                         Figure 17

             Chloride Ion Breakthrough Curve for Third Elution of Sample 11-52

-------
   1.6
   1.2
C/C
   0.8
   0.4
        468
                    Influent  Phosphate  (raeq)
                          Figure  18
      Phosphate  Ion  Breakthrough  Curve,for  Third  Elution
                       of  Sample  11-52
10
                              54

-------
is considered to be in the monovalent ion state, I^PO^',  the total
capacity of the fiber is 2.9_meq/g.  However, if the phosphorous is in
the divalent ion state, HK>4~, the total capacity, 3.7 meq/g, conforms
more closely to the capacities obtained in the first two  elutions.
Therefore, it is assumed that the divalent ion predominates on the
exchanger, and the selectivity coefficient is accordingly calculated for
HPO^".  As expected the selectivity increasingly favors the phosphate
ion over chloride as the pH is increased since ions of higher valence
are preferentially absorbed.

Elutions 4 through 6 of sample 11-52 are shown in Figures 19 to 21.  The
purpose of these elutions with 0.0157 N Cl" was to check the stability
of the total capacity of the fiber.  A meaningful estimation of the life-
time of the ion exchange fiber is not possible from the elutions of
Table 14.  On the other hand, in elutions 4, 5, and 6 there is no indi-
cation of serious attrition.  Between elutions 1 and 4 there is a 2%
decrease in capacity, but this is not uncommon in any type of ion
exchanger 

The influent solution for the seventh elution contained chloride and
nitrate ions.  Since the nitrate ion breakthrough curve was not deter-
mined, the total capacity from the sixth elution was utilized in order
to determine the steady state absorption of nitrate ion on the column.
The breakthrough curve for chloride ion is shown in Figure 22.  Selec-
tivity is seen to be in favor of the nitrate ion.  In the eighth and
final elution both chloride and nitrate ions were again present but in
lesser concentrations.  The ratio of chloride to nitrate ion is constant
in elutions 7 and 8.  Figure 23 gives the chloride ion breakthrough
curve for the eighth elution.  As with the seventh elution, the nitrate
absorbed on the column was taken as the difference between the total
chloride absorbed in the sixth elution and the chloride absorbed in the
eighth elution.  In this case chloride ion is preferentially absorbed
by the fiber.  This preference would probably prevail at still lower
concentrations.

From the results of the above elutions of sample 11-52 (MS-WP-12E), the
properties of the vinylpyridine anion exchange fiber are seen to be
typical of common commercial ion exchangers, although the capacities are
generally somewhat lower.  No physical deterioration of the fiber was
detected after the eighth elution.

Polyethyleneimine Fibers

In the evaluation of fiber MS-WP-44A containing 1.0% polyethyleneimine
(PEI), it became apparent that PEI is not as easily rendered water
insoluble by dibromoalkanes as is polyvinylpyridine.  Initial cross-
linking attempts led to almost complete extraction of the PEI in acidic
solution.  By prolonged heating of the fiber in dibromoalkane solution
the loss of PEI in acidic medium was considerably but not completely
reduced.  The column evaluations of these samples are described below.
                                   55

-------
        1.0
        Oo8
                                                                                        O
C/C
        0.6
        0.4
        0.2
                                               O
                        O	
                             10                 12
                                     Influent Chloride (meq)
                                          Figure 19
                               Fourth Elution of Sample 11-52
14
16

-------
         1.0
         0.8
         0.6
      C/C,
Ln
 'O
0.4
         0.2
                          J--Q      L
                          10
                                               I         I         I          I
                                               J	I
                                     12
14
                                                 Influent Chloride  (nieq)
                                                      Figure 20
                                            Fifth Elution of Sample 11-52
                    I          I
                                                                                     16
18

-------
           1.0
                     i         I          I         r
   c/c
in
00
           0.8
           0.6
0.4
           0.2
                                        I
                              10
                                               i          r
                                               j	L
                                    12                 14

                                 Influent Chloride  (meq)

                                       Figure 21

                             Sixth Elution of Sample 11-52
16

-------
VO
       1.5
       1.0
       C/C
       0.5
         0
            12
    16                      20                      24
         Influent Chloride + Nitrate (meq)

                     Figure 22

Chloride Ion Breakthrough Curve for Nitrate-Chloride
      Elution of Sample 11-52, Seventh Elution
                                                                                                          28

-------
 1.5
 1.0
C/C
 0.5
                            12                       16
                           Influent Chloride + Nitrate (meq)
                                          Figure 23
20
               Chloride Ion Breakthrough Curve for Nitrate-Chloride Elution of
                                 Sample 11-52, Eighth Elution

-------
Sample 11-37 (from finely cut MS-WP-44A) was treated with a 5% dibromo-
hexane, 95% decane solution for 8 days at room temperature and an
additional 2 days at 50C.  An exemplary column elution of this fiber
with 0.016 N HC1 is shown in Figure 24.  The low breakthrough capacity
for this fiber, less than 2 meq/g, is a result of inefficient packing;
breakthrough of more than 10% occurred above 3 meq/g.  The total
capacity, 3.4 meq/g, is quite low due to the extraction of PEI during
the elutions.

Another sample (11-55) of MS-WP-44A was treated with a 10% dibromohexane,
5%'dibromodecane, 85% decane solution at 45 to 50C for a week after
which n-butyl bromide was added and the solution was maintained at 50
for an additional 5 hours.  A column elution of this sample with 0.02013
N HC1 is shown in Figure 25.  In this case the breakthrough capacity
exceeds 3 meq/g, and the total capacity, 3.6 meq/g, is an improvement
over sample 11-37.  Nevertheless, this fiber should have a greater
capacity than the 3.8 meq/g obtained from fiber MS-WP-12E (e.g., sample
11-52) since both have 80% PVP for MS-WP-12E.  A second elution of sample
11-55 with 0.02 N HC1 is shown in Figure 26.  Again the total capacity
of the sample is 3.6 meq/g.                                 '*

Selectivities for Polyethyleneimine Fibers

With the establishment of a reasonably stable total capacity  (i.e., no
further extraction of PEI), sample 11-55 was eluted a third time with a
phosphate-chloride solution.  The results of this elution are given in
Figure 27.  Because the elution was stopped before a steady ;state con-
dition had been reached, the extrapolation necessary to approximate
attainment of the steady state (dashed line of Figure 27) introduces a
sizable error in the determination of phosphate absorption on the column.
However, it is still possible to determine the selectivity coefficient
within about 25% error.  Selectivity is clearly in favor of the chloride
ion  (see Table 15).  It is predicted that the selectivity coefficient
would not vary greatly as the total ion concentration is reduced  (at con-
stant pH) since the ratio of activity coefficients would be nearly
constant for the dilute solutions being considered.  On the other hand,
a change in relative concentrations of chloride and phosphate ions will
affect the selectivity coefficient.  Generally the selectivity coeffic-
ient decreases as the fraction of ion preferentially absorbed (chloride)
increases; however, experiments to confirm this relationship have not
been performed.  It should be noted that the phosphate-chloride solutions
used in the elutions for Table 15 had a pH of 1.7; at this acidity the
monovalent anion I^PO," is probably absorbed.

A  finely cut sample  (11-63) of fiber MS-WP-44D  (15% pp, 5% ABS, 50% PVP,
30%  PEI) was exposed to a decane  solution of 1,6-dibromohexane and
1,10-dibromodecane  for 14 days at 50C.  After the second day, 1-bromo-
butane was added.  Although the prolonged treatment with dibromoalkanes
provided more extensive crosslinking than previously obtained with
fibers containing PEI, part of the fiber was still soluble in dilute


                                  61

-------
to
            1.0
            0.8
            0.6
            0.4
            0.2
                                   I
ave. elution rate =9.2 ml/min
sample weight = 3.91 g
                                   8                  10                 12
                                                Influent  HCl  (meq)
                                                    Figure  24
                         Breakthrough  Curve  for Anion  Exchange  Fiber  11-37  (MS-WP-44A)
                                                                     14

-------
CO
             1.0  |
             0.8
             0.6
         C/C
             0.4
             0.2
                 0    18
                                                                 O
                            ave. elution rate = 13.4 ml./min.
                             sample weight = 6.05 g.
20
22
24
                                                   Influent HCl (meq)
                                                   Figure 25
                        Breakthrough Curve for Anion Exchange Fiber 11-55  (MS-WP-44A)
26
                                                                                                            j

-------
     1.0 r-
     0.8
     0.6
'-C/C
     0.4
     0.2
                 18
   20                 22
       Influent HCl (meq)
          Figure 26
Second Elution of Sample 11-55
24
26

-------
                                                     D Phosphate
                                                     OChloride
                   20                 24
        Influent Chloride + Phosphate (meq)

                     Figure 27
Third Elution of Sample 11-55 (Chloride + Phosphate)
                          65

-------
                                Table 15

         Phosphate-Chloride Elutions of Anion Exchange Fibers


                                    Sample 11-55       Sample 11-63

Phosphate Absorbed on
Column (meq)                         0.23 + .06            0.25

Chloride Absorbed on
Column (meq)                           21.97              18.50

Phosphate Concentration
(meq/liter)                             1.28               1.22

Chloride Concentration
(meq/liter)                            18.96              18.96

Total Capacity (meq/g)                  3.7                3.6


K!1..-                                6.5 + 1.5             4.8
                                  66

-------
acid.  Nevertheless, even with the loss of PEI the total capacity was in
excess of 3.5 meq/g which indicates that at least part of the PEI was
rendered insoluble.  This is about the same capacity as obtained from
the fiber having 80% PVP.  Thus, if extraction of PEI can be prevented,
higher capacities should be possible.

Sample H-63 (MS-WP-44D) was also eluted with a chloride-phosphate solu-
tion.  The results of this elution are shown in Figure 28 and Table 15.
Within the experimental error the selectivity is the same as for .sample
11-55.  Since samples 11-55 and 11-63 both contain the same components
but in different proportions, the selectivity coefficiepts are expected
to be nearly equal.
                                   67

-------
   5.0
   4.0
   3.0
C/C
   2.0
   1.0
                                              D  Phosphate

                                              O  chloride
       15        17          19          21
         Influent Chloride + Phosphate (meq)

                      Figure 28

      Phosphate-Chloride Elution of Sample 11-63
23
25
                          68

-------
                              SECTION VI
                            ACKNOWLEDGMENTS

The support of the project by the Water Quality Office, Environmental
Protection Agency, and the assistance and helpful discussions with Mr.
Richard A. Dobbs, Project Officer, and Mr. Jesse M. Cohen, Chief,
Physical and Chemical Treatment Research Program, Robert A. Taft
Water Research Center are gratefully acknowledged.

The work was conducted at the Uniroyal Research Center, Wayne, New
Jersey under the administrative direction of Dr. Robert L. Bergen, Jr.
with Dr. Ronald W. Fuest as Principal Investigator.  Experimental
work was performed by Dr. Malcolm J. Smith with the assistance of
Mr. Theodore Budynski.
                                   69

-------
                               SECTION VII

                                REFERENCES

 1.   Pollio,  Frank X.  and Kunin, Robert,  Environmental  Science and Tech-
     nology,  2,  54 (1968).

 2.   Bregman, Jacob I. and Shackelford, James M.,  Environmental Science
     and Technology, 3, 336 (1969).

 3.   Applebaum,  Samuel B., Demineralization by Ion Exchange, New York
     Academic Press, 1968, p.  341.

 4.   Muendel, C. H. and Selke, W. A., Ind. & Eng.  Chem.,  47, 374 (1955).

 5.   Richter, George A., Jr.  (to Rohm & Haas Co.), U.S. Patent 2,933,460
     (April 19,  1960).

 6.   Richter, George A., Jr.  (to Rohm & Haas Co.), U.S. Patent 2,974,101
     (March 7, 1961).

 7.   Cowers,  Donald S. (to Rohm & Haas Co.), U.S.  Patent  3,213,016
     (October 19, 1965).

 8.   Moncrieff,  R. W., Man-Made Fibres, New York,  John Wiley & Sons, Inc.
     1963, p.3.

 9.   Standard Methods for the Examination of Water and Wastewater, New
     York, American Public Health Assoc., Inc., 12th ed.  (1969).

10.   FWPCA Methods for Chemical Analysis of Water  and Wastes, Cincinnati,
     Ohio, Federal Water Pollution Control Administration, November 1969.

11.   Signer,  et  al., Makromol. Chem., 19/19, 139 (1956).

12.   Roth, Harold H., Ind. & Eng. Chem., 49, 1820  (1957).

13.   Fuest, R. W., et al., Development of Regenerable Fibers for Removal
     of Sulfur Dioxide from Waste Gas, Final Report, Contract PH 86-68-74,
     National Air Pollution Control  Administration, July 1970.

14.   Helfferich, Friedrich, Ion Exchange, New York, McGraw-Hill Book Co.,
     1962, p. 152.
                                   71

-------
                              SECTION VIII

                                GLOSSARY

                    Materials Used in Fiber Mixtures

ABS      acrylonitrile-butadiene-styrene (16:7:77 polymer blend) Dow
         "Tybrene"

EMA      sodium salt of ethylene-methacrylic acid copolymer, DuPont
         "Surlyn 1559"

ER       epoxy resin, Dow "DER 662"

GMA      copolymer of methyl vinyl ether and maleic anhydride, GAF
         "Gantrez AN 169"

NBR      15:85 copolymer of acrylonitrile and butadiene, Arco "Poly
         BD, CN-15"

NPA      Nylon-6 polyamide

NW       alkylated p-cresol type antioxidant, Uniroyal "Naugawhite"

PAA      polyacrylic acid, Rohm & Haas Acrysol

PAAS     sodium salt of PAA

PAS      polyaminostyrene  (Norsk Hydro)

PEI      polyethyleneimine, Dow PEI-1000

PMM      methyl methacrylate-styrene-butadiene,  Rohm & Haas "Acryloid"

PMS      polymethylstyrene, Amoco Resin 18

PP       polypropylene

PPP      peroxidized  polypropylene

PS       polystyrene

PVP      1:1 copolymer of  2-vinylpyridine and 2-methyl-5-vinylpyridine

PVP-HAc  PVP neutralized with acetic acid

SB-1     9:1 copolymer of  styrene and  butadiene

SB-2     8:2 copolymer of  styrene and  butadiene

SB-3     72:28  copolymer of  styrene  and butadiene

                                   73

-------
                           GLOSSARY (contd.)






SBR      1:3 styrene-butadiene copolymer,  Arco "Poly BD,  CS-15




SDM      styrene-dimethylaminopropyl maleimide copolymer




SMA-1    1:1 copolymer of .styrene and maleic  acid,  Arco Chemical  Co.




SMA-2    2:1 copolymer of styrene and maleic  acid,  Arco Chemical  Co.




SVP      3:7 copolymer of styrene and 2-vinylpyridine
                                   74

-------
  SELECTED WATER
  RESOURCES ABSTRACTS
  INPUT TRANSACTION FORM
                     1. Report No.
                        3. Accession No.
                                          W
  4. Title
   FEASIBILITY STUDY OF REGENERATIVE FIBERS FOR WATER
   POLLUTION  CONTROL
  7. Autbor(s)
   Fuest,  R.  W., and Smith, M. J.
                                          5. Report Date
                                          6.
                                          8. Performing Organization
                                            Report No.
  9.  Organization
   Uniroyal,  Inc.
   Research Center
   Wayne,  New Jersey 07470
                                         10. Project No.
                                            17040 DFC
                                         11. Contract/Grant No.
                                            14-12-815

                                         /3. Type of Report and
                                            Period Covered
 12. Sponsoring Organization
  15.  Supplementary Notes
  16.  Abstract
         The feasibility of making fibers which have  the ability to sorb minerals from
   water has been investigated.   High molecular weight  organic polymers cospun with
   fiber-forming polymers and crosslinking agents  can be spun into fiber and chemically
   treated to give ion exchanging materials.  In some cases the ion exchange active
   material can be spun directly (e.g., polyamines and  polyvinylpyridines);  in other
   cases the activity must be introduced after spinning (e.g., the sulfonation of
   fibers containing polystyrene).
         Styrene sulfonic acid cation exchange fibers having capacities up  to 4.2 meq/g
   dry fiber have been prepared.  Polypropylene and  polybutadiene served as  fiber-
   forming polymer and crosslinking agent, respectively.
         Weak acid cation exchangers were produced  from fibers containing polymethyl-
   methacrylate by hydrolysis of the ester to polymethacrylic acid.   Capacities up
   to 1 meq/g were obtained.
         Vinylpyridine and  aliphatic amine ion exchange active materials were spun
   directly and then crosslinked with dihaloalkanes.  Quaternization  was achieved
   with alkyl halides.  Vinylpyridine exchangers gave capacities up to  3.8 meq/g,
   and fibers containing both Vinylpyridine  and  polyethyleneimine show  promise of
   yielding higher capacities.
  17a. Descriptors
    *Ion Exchange  Fiber
    *Capacity
    *Selectivity
    *Fiber Spinning
  17b. Identifiers
    *Ion Exchange
    *Demineralization
  1 7c. CO WRR Field & Group  05D
  18. Availability
19. Security Class.
   (Report)

20. Security Class.
   (Page)
21. No. of
   Pages

22. Price
Send To:
                                                        WATER RESOURCES SCIENTIFIC INFORMATION CENTER
                                                        U.S. DEPARTMENT OF THE INTERIOR
                                                        WASHINGTON, D. C. Z024Q
  Abstractor  M. J. Smith
              Institution
                                                            Inc.
WRSIC 102 (REV. JUNE 1971)
                                                                                   6PO 913.261

-------