Inverse Emulsion Polymerization - American Chemical Society

2 Inverse Emulsion Polymerization J. W. VANDERHOFF, Ε. B. BRADFORD, H. L. TARKOWSKI, J . B. SHAFFER, and R. M. WILEY Physical Research Laboratory, The Dow Chemical Co., Midland, Mich.

Downloaded by UNIV OF MISSOURI COLUMBIA on August 5, 2013 | http://pubs.acs.org Publication Date: January 1, 1962 | doi: 10.1021/ba-1962-0034.ch002

In an inverse emulsion polymerization, a hydro philic monomer, frequently in aqueous solution, is emulsified in a continuous oil phase using a water -in-oil emulsifier and polymerized using either an oil-soluble or water-soluble initiator; the products are viscous latices comprised of submicroscopic, water-swollen, hydrophilic polymer particles col loidally suspended in the continuous oil phase. The average particle sizes of these latices are as small as 0.05 microns.

The technique is applicable

to a wide variety of hydrophilic monomers and oil media.

The inverse emulsion polymerization of

sodium p-vinylbenzene sulfonate initiated by both benzoyl peroxide and potassium persulfate was compared to the persulfate-initiated polymeriza tion in aqueous solution.

Hypotheses for the

mechanism and kinetics of polymerization were developed and used to calculate the various kinetic parameters of this monomer.

n a conventional emulsion polymerization, a hydrophobic monomer is e m u l sified i n water using an oil-in-water emulsifier a n d polymerized using a watersoluble initiator. A schematic representation according to Harkins (.19) is shown i n Figure 1. T h e hydrophobic monomer exists i n three loci—i.e., in the emulsion droplets, i n the aqueous phase as solute molecules, a n d i n the emulsifier micelles as solubilized molecules; most of the monomer is i n the emulsion droplets. T h e emulsifier also exists i n three loci—i.e., i n the oil-water interface, i n the aqueous phase as solute molecules, and i n the micelles; most of the emulsifier is i n the micelles. T h e primary free radicals are generated i n the aqueous phase and m i grate to the monomer-water interface. Since the total surface area of the micelles is large relative to that of the monomer droplets, the probability is great that the diffusing radical w i l l enter a micelle rather than a droplet. T h e initiation of poly merization i n a monomer-containing micelle transforms it into a monomer-swollen polymer particle before the initial polymer radical is terminated. T h e rapid chain propagation is sustained b y monomer diffusing from reservoir droplets a n d neigh boring micelles w h i c h have not captured a radical. This initiation process con-

I

32 In POLYMERIZATION AND POLYCONDENSATION PROCESSES; PLATZER, N.; Advances in Chemistry; American Chemical Society: Washington, DC, 1962.

VANDERHOFF ET A i .

Inverse Emulsion Polymerization

33

Background Literature Mechanism and Kinetics of Emulsion Polymerization


Subject

Ref.

General qualitative theory of the mechanism and kinetics of emulsion polymerization General quantitative treatment (Smith-Ewart) of the foregoing theory Experimental verification of the Smith-Ewart theory for the styrenepersulfate system Emulsion polymerization of butadiene in the "mutual recipe" Emulsion polymerization of butadiene in the hydroperoxide—polyami ne system Emulsion polymerization of styrene in the persulfate and hydroperoxide-poly amine systems, and isoprene in the hydroperoxidepolyamine system Emulsion polymerization of styrene in the persulfate-fatty acid soap system Emulsion polymerization of styrene in the persulfate-Amphoseife system Emulsion polymerization of styrene in the persulfate-triethanolamineAmphoseife system The gel effect at high conversions in styrene emulsion polymerization Emulsion polymerization of styrene, vinyltoluene, and dimethylstyrene in the persulfate-Amphoseife system Emulsion polymerization of styrene in the cumene hydroperoxidefatty acid soap system The gel effect at low conversions in styrene emulsion polymerization Emulsion polymerization of styrene and styrene-acrylonitrile mixtures above and below the critical micelle concentration Emulsion polymerization of styrene using oil-soluble initiators Quantitative theoretical treatment of emulsion polymerization kinetics ( more general than the Smith-Ewart theory ) * Emulsion polymerization of methyl methacrylate in the persulfateTergitol 7 system The mechanism and kinetics of emulsion polymerization as inferred from particle size distributions determined by ultracentrifugation Emulsion polymerization of vinylidene chloride in the persulfate-metabisulfite-sodium lauryl sulfate system Emulsion polymerization of styrene using the competitive growth technique Emulsion polymerization of styrene-divinylbenzene and styrene-acrylonitrile mixtures using the competitive growth technique Emulsion polymerization of styrene initiated by 7-rays

MONOME R - SWOLLEN POLYMER PARTICLE

(19) (35) (33,34) (30) (31) (32) (4) (7) (6) (17) (IS) (22) ( 23 ) (24) ( 25 ) ( 20, 36 ) (43) (11-13) (16,21, 27, 28 ) (10,37-39)

CONTINUOUS AQUEOUS PHASE

Figure 1. Schematic representation of a conventional emulsion polymerization

In POLYMERIZATION AND POLYCONDENSATION PROCESSES; PLATZER, N.; Advances in Chemistry; American Chemical Society: Washington, DC, 1962.

( 40 ) ( 41 )


34

ADVANCES IN CHEMISTRY SERIES

tiniies until a l l of the micelles either become monomer-swollen polymer particles, or relinquish their monomer and emulsifier to a radical-containing neighbor. T h e disappearance of the micelles marks the end of stage 1, the particle initiation stage. In stage 2, the growth stage, no new particles are formed; those formed i n stage 1 continue to grow until the supply of monomer or free radicals is exhausted. This hypothesis was treated quantitatively b y Smith and Ε wart ( 3 5 ) and applied successfully to the styrene system (33, 34). Since then, it has been extended b y many investigators to other monomers and initiators over a wide range of condi tions. In an inverse emulsion polymerization, an aqueous solution of a hydrophilic monomer is emulsified i n a continuous hydrophobic oil phase using a water-in-oil emulsifier. T h e polymerization is initiated w i t h either oil-soluble or water-soluble initiators. F i g u r e 2 shows a schematic representation of this system. T h e for mation of micelles is uncertain, but is portrayed speculatively. T h e hydrophilic part of the emulsifier molecule is oriented toward the hydrophilic dispersed phase and the hydrophobic part toward the hydrophobic continuous phase. T h e initiation of polymerization proceeds b y a mechanism analogous to that of the conventional system and submicroscopic particles of water-swollen hydrophilic polymer are generated in the continuous oil phase.

AQUEOUS MONOMER POLYMER PARTICLE

CONTINUOUS HYDROPHOBIC PHASE

Figure 2. Schematic representation of an inverse emulsion polymerization In conventional latices, the colloidal stability of the particles arises from the predominance of the electrostatic forces of repulsion over the London-van der Waal's forces of attraction. These electrostatic forces o f repulsion result from the electric double layer formed b y the emulsifier ions adsorbed on the hydrophobic polymer particle surface and the counterions from the aqueous phase. T h e London-van der Waal's forces of attraction are strongest when the particleparticle distance is very small. Therefore, i n most particle-particle collisions, the particles repel one another until the particle-particle distance is decreased to the point where the London-van der Waal's forces of attraction are predominant over the electrostatic forces of repulsion. Thus, many conventional latices remain stable indefinitely without significant stratification or flocculation of the particles. In an inverse latex, these electrostatic forces are quite different. F o r waterin-oil emulsions, Albers and Overbeek ( J , 2, 3 ) found that, w i t h ionic emulsifiers, flocculation was promoted by gravity because of the diffuse nature of the electric double layer relative to that of oil-in-water systems and, w i t h nonionic emulsifiers, In POLYMERIZATION AND POLYCONDENSATION PROCESSES; PLATZER, N.; Advances in Chemistry; American Chemical Society: Washington, DC, 1962.

VANDERHOFF ET AL.


35

flocculation was not prevented b y adsorbed oleophilic chains u p to 2 0 A . i n length. Thus i t might be expected that inverse latices w i l l stratify or flocculate more readily than their conventional counterparts.


Experimental Methods, Remits, and Discussion General. Aqueous solutions of hydrophilic monomers were emulsified i n xylene using water-in-oil emulsifiers, a n d polymerized using oil-soluble initiators. T y p i c a l hydrophilic monomers were sodium p-vinylbenzene sulfonate, sodium vinylbenzyl sulfonate, 2-sulfoethyl acrylate, acrylic acid, acrylamide, vinylbenzyltrimethylammonium chloride, a n d 2-aminoethyl methacrylate hydrochloride. T y p i c a l oil-soluble initiators were benzoyl a n d lauroyl peroxides. I n some cases, water-soluble potassium persulfate was used, both separately and i n mixtures w i t h oil-soluble peroxides. O f the water-in-oil emulsifiers, one of the most effective was Span 60 (technical sorbitan monostearate, Atlas C h e m i c a l Industries, Inc.). T h e emulsions were formed b y dissolving the emulsifier i n xylene a n d adding the aqueous monomer solution w i t h stirring. Frequently the crude emulsions were homogenized (Cenco hand homogenizer) to decrease the average droplet size a n d increase the emulsion stability. T h e emulsions were heated w i t h stirring at 4 0 ° to 70° C . to effect polymerization. T h e time required for essentially com plete conversion varied from a few minutes to several hours. T h e viscosities of the final latices were high—e.g., ca. 400 centipoises, as compared to ca. 10 centipoises for a conventional polystyrene latex of the same concentration. These latices consist of submicroscopic, water-swollen, hydrophilic polymer spheres colloidally suspended i n the continuous xylene phase. A typical electron micrograph of a diluted dispersion of a sodium poly (p-vinylbenzene sulfonate) latex w h i c h h a d been treated to remove water is shown i n F i g u r e 3. T h e inverse

Figure 3. Electron micrograph of sodium poly (p-vinylbenzene sulfonate) latex latex particles are spherical and as small as 300A. i n diameter. Thus, the two criteria for a n emulsion polymerization system—i.e., a segregated free radical system and a number of loci for polymerization within a few orders of magnitude of the number of free radicals existent at a given time, appear to be met b y these systems. These inverse latices are less stable than conventional latices; upon standing, In POLYMERIZATION AND POLYCONDENSATION PROCESSES; PLATZER, N.; Advances in Chemistry; American Chemical Society: Washington, DC, 1962.


36


their particles w i l l settle out i n a few hours to a few days. I n some cases, the stratified latices may be redispersed b y gentle agitation. Continuous, gentle agitation w i l l preserve the latex i n its colloidal state indefinitely. Purification of Sodium p-Vinylbenzene Sulfonate Monomer. T h e crude monomer was purified b y recrystallization. A saturated solution of monomer was prepared i n water containing sodium nitrite polymerization inhibitor at p H 11. This solution was cooled and the monomer crystals were filtered, washed w i t h absolute alcohol, a n d dried under vacuum at 2 5 ° to 35° C . Three batches of purified monomer were prepared. Batch N o . 1 was used for the benzoyl peroxideinitiated polymerizations, batch N o . 2 for persulfate-initiated solution polymeriza tions, w h i c h w i l l be described elsewhere, a n d batch N o . 3 for persulfate-initiated emulsion and solution polymerizations. In batch N o . 1, the monomer solution containing 2 0 % monomer and 0.4% sodium nitrite was prepared at 2 5 ° C . and cooled to 0 ° C . to crystallize the monomer. T h e monomer crystals (one part) were treated twice b y shaking w i t h six parts of absolute alcohol, filtering i n a fritted glass funnel, washing with five parts of absolute alcohol, and drying under vacuum at room temperature. F o r batch N o . 2, the monomer was recrystallized three times. In the first crystallization, a saturated solution of monomer was prepared without sodium nitrite at 4 0 ° C . and cooled to 0 ° C . to effect crystallization. T h e product was dissolved i n water under nitrogen and recrystallized i n the same manner. This product was dissolved i n water containing ca. 0 . 5 % sodium nitrite at 55° C . a n d cooled to effect the crystallization. T h e crystals were filtered out at 32° a n d 22° C , squeezed d r y , shaken in absolute alcohol for one hour, filtered, washed, dried under vacuum at 3 5 ° C . for three days, a n d stored under vacuum. Various analytical data are summarized i n Table I. Batch N o . 3 was recrystal lized from batch N o . 2; the saturated solution containing ca. 3 0 % monomer and ca. 0 . 3 % sodium nitrite at p H 11 was prepared at 50° C . T h e crystals were filtered out at ca. 35° a n d ca. 2 3 ° C . and were treated i n the same manner at batch N o . 2.

Table I.

Analyses of Monomer

Compound

Batch No. 1 Original Final

Batch No. 2 Original Final

Per Cent Monomer NaBr Na S0 NaNO H 0 Ethanol Polymer Sodium bromoethylbenzene sulfonate 2

4

z

2

93 .62 1.,62 2 .37 2.,S2

98.48 None None 0.02 0.23 0.87 0.3

71, ,49 0. 12 18,,28

i .40

98.32 0.36 0.25 0.02 0.16 ϋόο 0.89

T h e consistency of the three monomer batches is demonstrated b y the rates of polymerization under equivalent conditions, 2 0 % aqueous solution, 0 . 2 0 % K S O based on water (Figure 4, Table I I ) . T h e polymerization rates were practically the same for batches Nos. 2 and 3, but were significantly lower for batch N o , 1. 2

2

s


VANDERHOFF ET AL.


37

Table II. Polymerization Rates of Various Monomer Batches Initial R Χ 70 , Moles/Liter Second 50° C. 60° C. 70° C. p

Batch No. 1 2

1.34 1.87

3

1.98

4

3.28 5.16° 3.96

11.9

9.13

4.44

11.3

4.21·


° Not shown in Figure 4.

TIME.

MINUTES

Figure 4. Variation of per cent conversion with time for various monomer batches polymerized in aqueous solution with persulfate initiator Ο Batch No. 1

Φ Batch No. 2

• Batch No. 3

Benzoyl Peroxide-Initiated Polymerizations. A series of experiments were carried out using sodium p-vinylbenzene sulfonate as the hydrophilic monomer, benzoyl peroxide as the initiator, Span 60 as the emulsifier, and o-xylene as the continuous o i l phase. In all polymerizations, the aqueous monomer solution con tained 2 0 % monomer, the xylene/monomer solution ratio was 7 0 / 3 0 , and the nominal benzoyl peroxide concentration was 0.20% based on xylene. T h e poly merization rates were determined dilatometrically at 5 0 ° , 6 0 ° , and 70° C . T h e glass dilatometers were agitated by a magnetic stirring bar and held i n a bath con trolled w i t h i n ± 0.05° C . T h e emulsions were prepared without benzoyl peroxide initiator, homogenized, degassed b y heating, and charged into the dilatometers. T h e benzoyl peroxide was added i n xylene solution at the polymerization tem perature and the rate of volume contraction with time was measured. After the polymerization had reached a high conversion, the latex was cooled, shortstopped, and analyzed for residual monomer by coulometric bromination. These analyses were used to determine the cubic centimeters of contraction per gram of poly merized monomer; the average of these values, 0.07639 ( σ = 0.00179), was used to calculate the conversion-time curves. The values for the final per cent con version are given i n Table III. In POLYMERIZATION AND POLYCONDENSATION PROCESSES; PLATZER, N.; Advances in Chemistry; American Chemical Society: Washington, DC, 1962.

38

ADVANCES IN CHEMISTRY SERIES Table III.


Emulsifier, G. Xylene, Cu. Cm. 0.033 0.042 0.050 0.067 0.083 0.125 0.033 0.042 0.050 0.067 0.083 0.125 0.033 0.042 0.050 0.067 0.083 0.125

Final Conversion,

%

Benzoyl Peroxide-

Av. Particle Diameter (σ), A.

93.5 92.2 92.7 93.2 94.7 91.9 82.4 95.6 95.6 93.9 95.2 92.7 76.1 92.8 96.9 94.2 98.8 99.4

1190(460) 1100(440) 930(260) 770(250) 710(210) 660(180) 1480(370) 1060(360) 920(320) 840(340) 720(260) 670(190) 1780(210) 1460(370) 1330(350) 770(230) 750(230) 670(190)

Av. Particle Volume X 70™, Liters 1.27 1.05 0.528 0.320 0.238 0.183 2.00 0.842 0.560 0.484 0.278 0.198 3.07 1.92 1.49 0.309 0.287 0.200

T o recover the polymer, latex samples were coagulated i n absolute alcohol, filtered, washed w i t h absolute alcohol, a n d dried under vacuum at 3 5 ° C . T h e polymer molecular weights were estimated from the reduced specific viscosities (0.400 gram of polymer per 100 c u . c m . of aqueous 0 . 5 N N a C l solution, at 30° C . ) according to the relationship of Vanderkooi, Schultz, and Sieglaff (42). W C

= 3.5 X 1 0 - * [ M ] M 7

T y p i c a l conversion-time curves are shown i n Figure 5. These curves, for 0.125 gram of emulsifier per c u . c m . of xylene at 5 0 ° , 6 0 ° , and 70° C , were normal ized to the origin to eliminate the induction period w h i c h was essentially non existent at 7 0 ° C , but as great as 4 0 minutes at 5 0 ° C . After a short initial period, the per cent conversion increased almost linearly w i t h time to 50 to 6 0 % a n d then levelled off to approach complete conversion asymptotically. T h e shape of these curves is very similar to those of conventional emulsion polymerizations. T h e values for the over-all rate of polymerization calculated from the linear portions of the conversion-time curves are listed i n Table I I I . T h e variation of these over-all rates w i t h emulsifier concentration is shown i n F i g u r e 6. A t 5 0 ° C , the over-all rate was independent of emulsifier concentration; the least squares slope is - 0 . 0 0 6 ( σ = 0.13). A t 60° a n d 70° C , the over-all rate increased w i t h increasing emulsifier concentration; the least squares slopes of these lines are 0.58 (σ = 0.11) and 0.89 ( = 0.14) respectively. T h e final latices were prepared for electron microscopy b y dilution w i t h xylene to ca. 0 . 1 % polymer followed b y distillation to remove the water. T h e distilled latices were diluted further and dried on the specimen substrates. T h e electron microscope specimens could not be calibrated w i t h monodisperse poly styrene particles because of the sensitivity of the sodium poly (p-vinylbenzene sulfonate) particles to water. Instead, the magnification was calibrated from one separate exposure of monodisperse spheres w h i c h was made on each photo graphic plate of five exposures. σ


39


VANDERHOFF IT AL.

Initiated Polymerizations


Number of Particles/Cu. Cm. Xylene X 70-* 0.253 0.304 0.597 0.970 1.28 1.46 0.160 0.377 0.563 0.637 1.09 1.34 0.104 0.164 0.210 0.996 1.05 1.32

k [/] X 70», Moles/Liter Second

Rate/Particle X 70», G./Particle/ Second

R X 70\ Moles/Liter Second

d

P

DP X 10 4.03 3.72 4.61 4.85 4.80 4.54 2.94 3.10 4.85 3.66 4.46 5.21 1.99 2.13 2.36 3.68 4.54 3.88

5.13 3.34 2.46 1.16 0.985 0.771 14.4 6.84 6.82 5.70 3.86 3.53 48.5 36.3 31.9 12.2 11.4 10.7

1.94 1.53 2.24 1.74 1.99 2.03 3.45 3.91 5.86 5.66 6.36 8.57 7.60 9.08 10.3. 19.0 19.0 25.6

1.91 1.85 1.87 1.87 1.87 1.98 7.42 7.48 7.51 7.52 7.45 7.92 29.5 29.3 29.8 29.6 29.6 31.5

T h e average particle volumes, [(w/β) ( 2 % Α / 2 % ) ] > were calculated a n d corrected to compensate for the water removed b y distillation. T h e correction factors based on a polymer density of 1.53 grams per c u . c m . were 7.12, 7.16, and 7.20 at 5 0 ° , 6 0 ° , a n d 7 0 ° C , respectively. T h e values for numbers of particles per c u . c m . of xylene (N) were calculated (Table I I I ) . T h e average particle diameters listed i n this table are the number averages corrected for the water removed b y distillation; the values of σ are given i n parentheses following the average diameter. Since the particle size distributions were fairly broad, the statistical t- a n d F-tests were applied to determine possible variations w i t h poly merization temperature; these results are given i n Table I V . A t the three highest 3

0

20

40

60

80

100

120

TIME. MINUTES

Figure 5. Variation of per cent conversion with time, benzoyl peroxide initiator, and 0.125 gram of emulsifier per cu. cm. of xylene m 50° C.

·

60° C .

Ο

70°

C.


40

ADVANCES IN CHEMISTRY SERIES 0.01

ο α.


u. Ο

0.0001 I

I

0.01

I

1 I

I I 111

I

I

I

II

I

III 1.0

0.1 GRAM E M U L S I F I E R / C U . C M .

XYLENE

Figure 6. Variation of over-all rate of polymerization with emulsifier concentra tion m 50° C.

· 60° C. ο 70° C.

emulsifier concentrations, 0.067, 0.083, a n d 0.125 gram per c u . c m . of xylene, Ν was independent of temperature and depended only upon the emulsifier concen tration. A t 0.042 and 0.050 gram of emulsifier per c u . c m . of xylene, the values of Ν at 5 0 ° and 60° C . were equivalent, but were smaller at 7 0 ° C . A t the lowest concentration, 0.033 gram per c u . c m . of xylene, Ν decreased w i t h increasing polymerization temperature. I n conventional emulsion polymerization systems initiated b y peroxide compounds, Ν usually increases w i t h increasing tempera ture; however, i n the γ-ray-initiated emulsion polymerization of styrene (41), Ν was observed to decrease w i t h increasing temperature. Since the free radical concentration of γ-ray-initiated systems is independent of temperature, it was sug gested that, w i t h increasing temperature, the temperature-dependent propagation rate produces larger initial polymer molecules i n the relatively few micelles w h i c h capture radicals. Therefore, a larger proportion of the micelles w h i c h have not captured radicals are forced to relinquish their emulsifier a n d monomer to support the growth of the radical-containing micelles. I n systems initiated b y peroxide compounds, the same competitive situation exists, but the activation energy for initiator decomposition is sufficiently greater than that of chain propa gation to override this effect. However, this explanation is not applicable to the foregoing inverse emulsion polymerization data. T h e variation of Ν w i t h emulsifier concentration is shown on a log-log plot (Figure 7 ) ; Ν increased strongly w i t h increasing emulsifier concentration (initial slope = 2.5 to 4.5) a n d levelled off at high concentrations. These results differ Table IV. Variation of Ν With Polymerization Temperature Emulsifier, G./Xylene, Cu. Cm. 0.033 0.042 0.050 0.067 0.083 0.125

Condition Decreases with increasing temperature Equivalent at 50 ° and 60 ° C. ; decreases at 70 ° C. Equivalent at 50 ° and 60 ° C. ; decreases at 70 ° G. Equivalent at 50 °, 60 °, and 70 ° G. Equivalent at 50 °, 60 °, and 70 ° C. Equivalent at 50 ° 60 ° and 70 ° C. 5

}


VANDERHOFF ET AL.

Inverse Emulsion Poly mer hat ion

41

considerably from the Smith-Ewart case where a straight line w i t h a slope of 0.6 was observed ( 3 3 ) . 5.0 ι


4.0

— -

3.0

-

2.0

h

0.01

αϊ

.15 0.03

0.1

.15 0.03

0.1 .15

GRAM EMULSIFIER/CU.CM. X Y L E N E

Figure 7. Variation of the number of particles per cubic centimeter of xylene with emulsifier concentration m 50° C.

· 60° C.

ο 70° C.

These unusual variations of Ν w i t h emulsifier concentration and temperature suggest that the aqueous monomer droplets m a y furnish loci for polymerization initiation. I n conventional emulsion polymerizations, the monomer droplets usually range i n diameter from 1 to 10 microns; therefore their total surface area is small relative to that of the micelles, and they serve only as reservoirs i n the particle initiation process. However, if these monomer droplets were smaller b y one or two orders of magnitude, they w o u l d compete effectively w i t h the monomerswollen micelles for the available radicals and therefore w o u l d serve as loci for particle initiation. In order to determine h o w small an emulsion droplet can be formed i n this system, a 2 0 % solution of sodium poly (p-vinylbenzene sulfonate) was prepared and emulsified i n xylene containing 0.083 gram of emulsifier per c u . c m . i n the same manner as the monomer solutions. The homogenized emulsion was then pre pared for electron microscopy i n the same manner as the final latex samples. F i g ure 8,B shows an electron micrograph of this dispersion, while Figure 8,A shows, for comparison, an electron micrograph of an inverse latex prepared using the same concentration of emulsifier. It is evident that this dispersion of polymer contained emulsion droplets as small as 200A. i n diameter—i.e., i n the same size range as the smallest particles of the analogous inverse latex. This is supported b y the values for the interfacial tension ( d u N u o y ring method), 0 ± 0.01 dyne per cm., of a 2 0 % aqueous monomer solution i n contact with xylene solutions of emulsifier (0.033, 0.083, and 0.125 gram per c u . c m . ) . This very l o w interfacial tension is consistent w i t h the formation of very fine emulsion droplets. T h e critical micelle concentration of Span 60 i n benzene has been determined as 0.0022 gram per c u . cm., the micellar molecular weight as 52,800, and the aggregation number as 94 (8, 9 ) . Thus, the emulsifier concentrations of the fore going polymerization experiments were considerably i n excess of the critical micelle concentration. T h e shape a n d actual size of the micelles have not been In POLYMERIZATION AND POLYCONDENSATION PROCESSES; PLATZER, N.; Advances in Chemistry; American Chemical Society: Washington, DC, 1962.


42


Figure 8.

Electron micrographs

A. Sodium poly(p-vinylbenzene sulfonate) latex B. Emulsified sodium poly(p-vinylbenzene sul fonate) solution

100

o.i '

mil

ι

ι ι ι nul

0.1

ι

ι ι 11 n i l

1.0

AVERAGE

10

PARTICLE VOLUME ι Ι 0 , LITERS , β

Figure 9. Variation of rate of polymerization per particle with average particle volume m 50° C.

m 60° C.

Ο 70° C.

determined, b u t it is reasonable to assume a diameter of the order of 100A. for a spherical model. Therefore, it is concluded that, although the monomer-swollen micelles furnish loci for particle initiation, many monomer droplets are small enough to serve also i n this capacity a n d , as a result, particle initiation occurs i n both the micellar and droplet phase. T h e reverse dependence of Ν on temperature at l o w emulsifier concentrations is still unexplained. T h e variation of the rate of polymerization per particle w i t h the average particle volume is shown i n F i g u r e 9. T h e increase o f the polymerization rate is about the same for the 60° and 70° C . samples, but is greater for the 50° C . samples. T h e least squares slopes of these lines are 0.94 ( σ = 0.078) for 50° C , 0.60 ( = 0,056) for 60° C , and 0.58 ( = 0.022) for 70° C . A t 6 0 ° and σ

σ


VANDERHOFF f Τ A l .


43

70° C , there is a subdivision effect typical of emulsion polymerization systems, while at 5 0 ° C , this effect is less marked a n d the kinetics resemble those of a solution polymerization.


90

TIME, MINUTES

Figure 10. Variation of per cent conversion with time for persulfate-initiated polymerizations at 50° C. •

Solution Ο Emulsion

Potassium Persulfate-initiated Polymerizations. A similar series of experi ments was carried out using potassium persulfate as the initiator; the concentra tion was 0.20% based o n water. A series of solution polymerizations was also carried out for comparison. F o r these polymerizations, a n unstirred glass d i l a tometer i n the shape of a cylindrical coil, 0.5 c m . i n I . D . a n d 120 c m . i n length, was used. Thus, portions of the same aqueous phase containing 2 0 % monomer and 0.20 gram of initiator per 100 grams of water were polymerized i n both emulsion a n d solution at 4 0 ° , 5 0 ° , 6 0 ° , a n d 7 0 ° C . F i g u r e 10 shows the c o m parative conversion-time curves at 5 0 ° C . T h e polymerization rate was consider ably more rapid i n emulsion than i n solution, but the shapes of the curves were similar. This disparity i n rate decreased w i t h increasing temperature (Figure 11) ; extrapolation o f the polymerization rate-temperature curves indicated that the rates should be equivalent at about 9 0 ° C . F i g u r e 12 shows the conversion-time curves of the emulsion polymerizations at various temperatures; the rates increased w i t h increasing temperature as expected. T h e pertinent kinetic data from these experiments are listed i n T a b l e V . Hypotheses Concerning Kinetic Parameters T h e foregoing data were analyzed using the following equations suggested b y van der Hoff (23) a n d taken from the theoretical analyses of Stockmayer (36) a n d H a w a r d (20). ζ = ή/(α/4) « h(a)/h(a)

(le)

ζ = η / ( β / 4 ) = tanh (Λ/4)

(lb)

β / 4 = (fk [I]/k ) N V d

t

m

A


(2)


44

Polymerizations

Table V. Persulfate-I

Temp., °C. 40 40 50 50 60 Downloaded by UNIV OF MISSOURI COLUMBIA on August 5, 2013 | http://pubs.acs.org Publication Date: January 1, 1962 | doi: 10.1021/ba-1962-0034.ch002

60 70 70

Initial R X 10\ Moles/Liter Second

[I] X 70 , Moles/Liter Second

k

P

Type Inverse emulsion Solution Inverse emulsion Solution Inverse emulsion Solution Inverse emulsion Solution

8

d

Av. Particle Volume

Av. Particle Diameter (*), A.

X 70«,

Liters

6.24

0.116

1780(550)

0.935 12.2

0.116 0.626

1490(550)

2!50

1.98 14.4

0.626 3.11

1560(660)

3U6

1590(780)

3!75

4.44 20.6

3.11 14.4

11.3

14.4

Γ

ι

ι

2.6

ι

ι

(l/T)

ι

3.0 DEGREE

2.8

«ΙΟ , 3

4.35

1

ι 32

Figure 11. Variation of rate of poly merization with temperature •

Solution Ο Emulsion

Rv = z\{kpVk )fk [I]yi*[M\ t

1/DP -

(\/z)\{kJkp*)Jk [I]\u\\/[M)) d

(3)

d

+ (k r/k ) t

(4)

p

where ft is the average number of free radicals per particle, / is the average of efficiencies of both initiator fragments i n initiating polymer chains, k is the rate constant for the decomposition of initiator, [I] is the concentration of initiator based on the dispersed phase, [ M ] is the monomer concentration i n the particles, N is Avogadro's number, i and I are Bessel functions of the first k i n d of zeroand first-order respectively, and V is the average particle volume i n liters. Equation la represents the case where the radicals are generated i n , or enter, the particles singly, while E q u a t i o n lb represents the case where radicals are gene rated i n , or enter, the particle i n pairs. T h e variations of ζ a n d az w i t h a are shown i n F i g u r e 13 [taken i n part from reference (23)] where the upper curves correspond to E q u a t i o n la and the lower curves to E q u a t i o n lb. d

A

0

t


VANDERHOFF ET AL.

45



βο

Ο

ΙΟ

20

30

40

50

TIME, MINUTES

Figure 12. Variation of per cent conversion with time for persulfate-initiated emulsion polymeriza tions + 40° C. • 50° C.

Φ 60° C. Ο 70° C.

The values of k used for benzoyl peroxide decomposition at 5 0 ° , 6 0 ° , a n d 70° C . were 0.92, 3.7, and 15 X l O ^ per second, respectively ( 1 5 ) . T h e values of k used for potassium persulfate decomposition at 4 0 ° , 5 0 ° , 6 0 ° , a n d 7 0 ° C . were 0.20, 1.0, 5.0, a n d 2 3 X 10~ per second, respectively (26). T h e calculated values of k [I] are listed i n Tables I I I and V . Except for the factor z, Equation 3 is a standard kinetic equation w h i c h should apply to the persulfate-initiated solution polymerization. Thus, i f it is assumed that f,k > and k are equivalent for both solution and emulsion polymeriza tions at the same temperature, the ratio of the initial values for R for emulsion a n d d

6

d

6

d

p

t

p

ο

Figure 13.

Variation of ζ and az with a


46


solution polymerizations should equal z. Once ζ is known, a and η may be calcu lated from the appropriate form of E q u a t i o n 1. T h e values used for R were the initial rates given (Table V ) . T o determine z, the slopes of the conversion-time curves (Figure 10) were determined at every 5% of conversion. T h e values of ζ were found to vary only insignificantly up to 40 to 8 0 % conversion. Therefore the values of ζ (Table V I ) are these averages; the corresponding values of σ are indicative, at least i n part, of the precision of measurement of the slopes. T o determine k , Equations 2 and 3 may be combined to give p

p

k = AR N V/az{M] p

p

Table V I shows the values for k as w e l l as for k (calculated i n terms of / from Equation 2 ) . A l l values of ζ are greater than 1. Since the initiation of radicals by persulfate ion has been proposed (26) as p


(5)

A

t

S Os~ 2

%

2SOÎ

it might be expected that the radicals w o u l d be generated i n pairs i n the particles and therefore the lower curve (Figure 13) w o u l d apply. However, the experi mental values of ζ are a l l significantly greater than 1, and therefore it must be assumed that some mechanism for the generation of single radicals i n an emulsion particle is operative. Possibly, as suggested for the styrene-cumene hydro peroxide system (25) even a very low rate of radical transfer i n or out of the particle w o u l d explain the results. Thus, the values of a and η (Table V I ) corre spond to those of the upper curve (Figure 13). T h e values of ft, ranging from 0.50 to 0.61, correspond to the Smith-Ewart case at 40° and 50° C , and increase w i t h increasing concentration of free radicals. T h e values of k are approximately 30 times greater than those of styrene (23, 2 9 ) . T h e Arrhenius plot of log k w i t h reciprocal temperature is shown i n Figure 14. T h e activation energy calcu lated from the least squares slope was 7.07 ( σ = 0.65) kcal. per mole. This is i n good agreement w i t h previously published values for styrene (5,14, 29, 32, 33). p

p

Table VI.

Persulfate-initiated Polymerizations k X 70-6, k χ w-\ Liters / Mole Liters/Mole Second Second η a 2.77 l.lOf 0.50 0.302 3.54 2.05f 0.50 0.332 3.53f 5.04 0.53 0.715 6.37f 7.47 0.61 1.36 t

p

Temp., °C. 40 50 60 70

ζ(σ) 6.67(0.26) 6.05(0.24) 2.96(0.25) 1.79(0.11)

T o estimate values for k /k , tr

p

Equations 3 and 4 were combined and re

a r r a n g e d to give (X/DP) - (fk [I]/R ) + (ktr/kp) d

(6)

P

If it is assumed that / is constant for a given series, a plot of ( 1 / D P ) vs. (k [l]/ Rp) should have a slope of / and an intercept of k /k at (k [l]/R ) = 0. T h e benzoyl peroxide-initiated polymerization data were analyzed i n this manner. There was considerable experimental scatter i n these plots and therefore, the data are given (Table V I I ) as the least squares slopes and intercepts as w e l l as their 9 5 % confidence limits. T h e average values of k /k , ranging from 0.8 to 1.1 X d

tr

tr

p

d

p

p



VANDERHOFF ET AL.

47


Ι Ο , are reasonable, a n d i n the same range as the values for styrene, 0.35 a n d 0.65 X 1 0 - at 5 0 ° a n d 7 0 ° C , respectively (25). However, the experimental scatter is such that these values can be regarded only as estimates. T h e values of / ranged from 1.1 to 1.3; these values are reasonable considering the experimen tal scatter and the possible errors. -4

4

Table VII. Estimated Values for k and f X 70* ktr, Liter/Mole Second 95% 95% Confidence Confidence Av. limits limits Av. 1.2 0.060-0.69 0.37 0.17-1.9 1.3 0.20-0.57 0.40-1.1 0.38 1.1 0.45-1.0 0.60-1.4 0.73 tr

ktr/k

p

Temp., °C. 50 60 70

Av. 1.1 0.76 0.98

/

95% Confidence limits 0.34-2.1 1.0-1.5 0.94-1.2

T h e other parameters for the benzoyl peroxide-initiated polymerizations were calculated i n the following manner. U s i n g the subscripts Β a n d F to denote the benzoyl peroxide- a n d persulfate-initiated polymerizations respectively, E q u a tions 2 and 3 give «B = M(J kd [I]B)/kt } N VB

(7a)

a

(7b)

B

P

B

B

ll2

A

= *\(fpk [I]p)/k )uW Vp DP

tp

A

RPB = ZB[(k ykt )fBk [I)B} [M}B p

B

(8a)

m

DB

RPP = zp\{kpyk )fpk \I]p\^[M}p tp

(Sb)

DP

These equations are combined to give CBZB = apzp(V /Vp)(R /R ) B

PB

(9)

Pp

Thus values of az can be calculated for the benzoyl peroxide-initiated polymeriza tions from the known values of az (persulfate-initiated polymerizations), and the ratios of average particle volumes a n d polymerization rates. These values of az m a y then be substituted i n Equations 3 a n d 4 to give values for f/k and fk w h i c h may then be combined to give values for / a n d k . T h e results of these calculations are shown (Table V I I I - A ) . T h e values of az are all considerably less than 2. F r o m F i g u r e 13» for the case where the radicals are generated i n , or enter, t

t

American Chemical Society. In POLYMERIZATION AND POLYCONDENSATION PROCESSES; PLATZER, N.; Advances in Chemistry; American Chemical LibrarySociety: Washington, DC, 1962.

t


48


the particle singly, az should equal 2 at values of a u p to 0.2 and then increase w i t h increasing values of a. T h e benzoyl peroxide-initiated polymerizations w o u l d be expected to follow this case, if each initiator molecule decomposed i n the con tinuous o i l phase to form a pair of radicals w h i c h then enter the particles singly. However, the calculated values of az range from 0.02 to 0.7, a l l of them consider ably smaller than 2. There are many possible experimental errors w h i c h might affect the value of az, but since this quantity is based on experimentally deter mined ratios of polymerization rates and average particle volumes, it is difficult to imagine a 100-fold error. Therefore, the case where the radicals are generated i n , or enter, the particle i n pairs must be considered. In order for this mecha nism to be operative, the polymerization must be initiated only b y those few peroxide molecules w h i c h enter the particles and decompose there, while most of the peroxide decomposes i n the continuous oil phase to form radicals w h i c h are ineffective in initiating polymerization. This hypothesis is admittedly tenuous, but at the present time, there is no other explanation w h i c h fits the observed results. Therefore the values for a, z,

Table VIII. Benzoyl PeroxideA. Temp. °C.

y

50

60

70

Emulsifier, G. Xylene, Cu. Cm. 0.033 0.042 0.050 0.067 0.083 0.125 0.033 0.042 0.050 0.067 0.083 0.125 0.033 0.042 0.050 0.067 0.083 0.125

az

a 0 .815 0 .655 0 .561 0 .383 0 .353 0 .312 1 .15 0 .788 0 .785 0 .717 0 .576 0 .563 1 .77 1 .52 1 .41 0 .868 0 .836 0 .811

0 .163 0 .106 0 .0781 0 .0367 0 .0312 0 .0244 0 .321 0 .153 0 .152 0 .127 0 .0822 0 .0788 0 .736 0 .550 0,.484 0..185 0.,172 0..162 B.

50

60

70

0.033 0.042 0.050 0.067 0.083 0.125 0.033 0.042 0.050 0.067 0.083 0.125 0.033 0.042 0.050 0.067 0.083 0.125

0. 234 0. 152 0. 112 0. 0527 0. 0448 0. 0351 0. 435 0. 207 0. 206 0. 172 0. 111 o: 107 0. 935 0. 699 0. 615 0. 235 0. 219 0. 206

Calculated from Equations

ζ 0 .201 0 .163 0 .140 0 .0955 0 .0880 0 .0779 0 .280 0 .195 0 .194 0 .177 0 .143 0 .140 0 .415 0 .363 0 .339 0..214 0 .206 0..201

Recalculated using normal 0. 240 0,.975 0..785 0. 194 0..672 0. 167 0..460 0. 115 0..423 0. 106 0..374 0. 0933 1..35 0. 325 0. 918 0. 226 0..915 0. 225 0. 836 0. 206 0..670 0. 166 0..658 0. 163 2. 01 0. 466 1..72 0. 405 1. 61 0. 382 0. 979 0. 241 0. 944 0. 234 0. 915 0. 226


VANDERHOFF Et AL.

49


and η were determined from the lower curve (Figure 1 3 ) . A s expected for this mechanism, the values of n , 0.006 to 0.2, are m u c h lower than the values of 0.5 to 0.6 obtained for the persulfate-initiated system. T h e values of / a n d k are given as the average and the range, respectively. These designations refer to the average values and 9 5 % confidence limits of k /k listed (Table V I I ) . T h e aver age values of / range from 0.8 to 1.5 a n d appear to be independent of tempera ture ( i n calculating k /k , it was assumed that f was constant for the series at a given temperature). T h e values corresponding to the 9 5 % confidence limits range from 0.1 to 2.4. t

tr

tr

p

p

T h e average values of k range from 0.5 to 54 Χ 1 0 liters per mole second and increase w i t h increasing temperature, increasing particle size, and decreasing polymer molecular weight. It is expected that k w o u l d increase w i t h temperature. The variation of k w i t h average particle volume is shown (Figure 1 5 ) . T h e lines are almost parallel; the least squares slopes at 5 0 ° , 6 0 ° , and 7 0 ° C . are 1.20 ( σ = 0.071), 1.34 ( = 0.11), a n d 1.33 ( = 0.093), respectively. T h e distance be tween these lines corresponds to an activation energy of about 15 kcal. per mole. 5


t

t

t

σ

σ

Initiated Polymerizations 3, 4, and 9 k X 70-·, Liters/Mole Second Av. Range t

ή

Av.

0.041 0.027 0.020 0.0092 0.0078 0.0061 0.080 0.038 0.038 0.032 0.021 0.020 0.18 0.14 0.12 0.046 0.043 0.041

1.5 1.4 1.3

0.93 1.1 1.2 1.2

1.3 1.0 1.5

1.3 1.3 1.0 1.2

1.1 1.1

values for R 0.059 0.038 0.028 0.013 0.011 0.0088 0.11 0.052 0.052 0.043 0.028 0.027 0.23 0.18 0.15 0.059 0.055 0.052

0.79 1.3

/

Range 0.55-2.4 0.62-2.1 0.28-2.4 0.11-1.8 0.15-2.0 0.27-2.1 1.1-1.4 1.1-1.5 0.73-1.3 1.2-1.7 0.95-1.6 0.86-1.6 0.94-1.1 1.0-1.3 0.99-1.2 0.88-1.4 0.55-1.0 1.0-1.6

3.9 3.7 1.3 0.70 0.53 0.46 16. 6.4 2.2 2.9 1.3 0.71 54. 31. 22. 2.4 1.6 1.5

1.5-6.4 1.7-5.8 0.27-2.3 0.084-1.3 0.073-1.0 0.10-0.82 14.-18. 5.4-7.4 1.6-2.9 2.4-3.5 0.96-1.6 0.49-0.94 49.-59. 28.-35. 19.-24. 1.9-3.0 1.1-2.1 1.1-1.8

0.79-3.4 0.89-3.0 0.40-3.5 0.16-2.5 0.21-2.9 0.38-3.0 1.4-1.9 1.5-2.0 0.98-1.8 1.6-2.4 1.3-2.1 1.2-2.2 1.2-1.4 1.3-1.6 1.3-1.6 1.1-1.7 0.69-1.3 1.3-2.1

3.9 3.7 1.3 0.70 0.54 0.46 16. 6.4 2.2 2.9 1.3 0.71 53. 31. 21. 2.4 1.6 1.5

1.5-6.4 1.7-5.8 0.27-2.3 0.084-1.3 0.073-1.0 0.10-0.82 14.-18. 5.4-7.3 1.6-2.9 2.4-3.5 0.96-1.6 0.48-0.93 48.-58. 28.-34. 19.-24. 1.9-3.0 1.1-2.1 1.1-1.8

p

2.1 1.9 1.9 1.3 1.6 1.7 1.7

1.7 1.4 2.0 1.7 1.7

1.3

1.5

1.4 1.4 1.0

1.7


50


However, when the l o g k is plotted against DP, (Figure 1 6 ) , a l l of the points are fitted best b y a single line. T h e vertical lines through each point correspond to the ranges of k calculated from the 9 5 % confidence limits of k /k (Table V I I I - A ) . T h e single line shown was calculated b y least squares using the average values of k . These data suggest that, even at a monomer concentration of only 2 0 % , the termination process is diffusion-controlled a n d the effect of temperature is obscured. It may be argued that the variation i n polymerization rate (Table I I , F i g ure 4) w i t h the various monomer batches may introduce gross errors i n the fore going calculations. Therefore, the polymerization rates of batch N o 1. were normalized to the average of the rates for batches Nos. 2 a n d 3; the factors used for 5 0 ° , 6 0 ° , and 7 0 ° C . were 1.44, 1.35, and 1.27, respectively. T h e recalculated values for k /k were almost identical to those of Table V I I . T h e values for / were somewhat larger, ranging from 1.4 to 1.8. T h e corrected values of az a n d the correspondingly higher values of a, z, a n d η are listed (Table V I I I - B ) . A l though the values of / are somewhat larger, the calculated values of k are virtually identical of those of Table V I I I - A . t

t

tr

p


t

tr

p

t

Acknowledgment T h e authors acknowledge gratefully the assistance of the East M a i n A n a l y t i cal Laboratory, T h e D o w C h e m i c a l C o . , w h i c h carried out the coulometric bromination and viscosity measurements.

100

J

I

1 I I I

0.1

111

I

I

1.0

AVERAGE

PARTICLE

VOLUME

JO %I0

, e

,

o.i 0

1

2

Figure 15. Variation of k with average particle volume t

m 50° C.

3 5P

LITERS

Figure 16.

4

5

6

7

χ 10"*

Variation of log k with DP

Φ 60° C. Ο 70° C.


t

VANDERHOFF ET AL.


51

Literature Cited


(1) Albers, W., Overbeek, J. Th. G., J. Colloid Sci. 14, 501 ( 1959 ). (2) Ibid., p. 510. (3) Ibid., 15, 489 ( 1960 ). (4) Bakker, J., "Kinetics of the Emulsion Polymerization of Styrene," Ph.D. thesis, University of Utrecht, The Netherlands, 1952. (5) Bamford, C. H., Dewar, M. J. S., Proc. Roy. Soc. London Al92, 309 ( 1948 ). (6) Bartholomé, Ε., Gerrens, Η., Ζ. Elektrochem. 61, 522 ( 1957 ). (7) Bartholomé, Ε., Gerrens, Η., Herbeck, R., Weitz, H. M., Ibid., 60, 334 ( 1956 ). (8) Becher, P., private communication, Atlas Chemical Industries, Inc., Wilmington, Del., September 1961. (9) Becher, P., Clifton, Ν. Κ., J. Colloid Sci. 14, 519 ( 1959 ). (10) Bradford, Ε. B., Vanderhoff, J. W . , Alfrey, T., Jr., Ibid., 11, 135 ( 1956 ). (11) Brodnyan, J. G., Ibid., 15, 563 ( 1960 ). (12) Ibid., p. 573. (13) Brodnyan, J. G., Brown, G. L., Ibid., 15, 76 ( 1960 ). (14) Burnett, G. M., Trans. Faraday Soc. 46, 772 ( 1950 ). (15) Doehnert, D. F., Mageli, O. L., 13th Annual Meeting, Reinforced Plastics Division, Soc. Plastics Ind., Inc., Feb. 4-6, 1958, Chicago, Ill. (16) Evans, C. P., Hay, P. M., Marker, L., Murray, R. W., Sweeting, O. J.,J.Appl. Polymer Sci. 5, 39 ( 1961 ). (17) Gerrens, H., Z. Elektrochem. 60, 400 ( 1956 ). (18) Gerrens, H., Köhnlein, E., Ibid., 64, 1199 ( 1960 ). (19) Harkins, W. D., "The Physical Chemistry of Surface Films," chap. 5, Reinhold, New York, 1952. (20) Haward, R. N., J. Polymer Sci. 4, 273 ( 1949 ). (21) Hay, P. M., Light, J. C., Marker, L., Murray, R . W., Santonicola, A. T., Sweeting, O. J., Wepsic, J. G., J. Appl. Polymer Sci. 5, 23 ( 1961 ). (22) van der Hoff, Β. M. E., J. Phys. Chem. 60, 1250 ( 1956 ). (23) van der Hoff, Β. M. E . , J. Polymer Sci. 33, 487 ( 1958 ). (24) Ibid., 44, 241 ( 1960 ). (25) Ibid., 48, 175 ( 1960 ). (26) Kolthoff, I. M., Miller, I. K., J. Am. Chem. Soc. 73, 3055 ( 1951 ). (27) Light, J. C., Santonicola, A. T., J. Polymer Sci. 36, 549 ( 1959 ). (28) Light, J. C., Marker, L., Santonicola, A. T., Sweeting, O. J., J. Appl. Polymer Sci. 5, 31 (1961). (29) Matheson, M. S., Auer, E . E., Bevilacqua, Ε. B., Hart, E . J., J. Am. Chem. Soc. 73, 1700 ( 1951 ). (30) Morton, M., Salatiello, P. P., Landfield, H., J. Polymer Sci. 8, 111 ( 1952 ). (31) Ibid., p. 215. (32) Ibid., p. 279. (33) Smith, W. V., J. Am. Chem. Soc. 70, 3695 ( 1948 ). (34) Ibid, 71, 4077 ( 1949 ). (35) Smith, W. V., Ewart, R. H., J. Chem. Phys. 16, 592 ( 1948 ). (36) Stockmayer, W. H., J. Polymer Sci. 24, 314 ( 1957 ). (37) Vanderhoff, J. W., Vitkuske, J. F., Bradford, Ε. B., Alfrey, T., Jr., J. Polymer Sci. 20, 225 ( 1956 ). (38) Vanderhoff, J. W., Bradford, Ε. B., Tappi 39, 650 ( 1956 ). (39) Vanderhoff, J. W., Bradford, Ε. B., Abstracts, p. 28S, 130th Meeting, ACS, Atlantic City, N. J., September 1956. (40) Ibid., p. 20-I, 131st Meeting, ACS, Miami, Fla., April 1957. (41) Vanderhoff, J. W., Bradford, E . B., Tarkowski, H. L., Wilkinson, B. W., J. Polymer Sci. 50, 265 ( 1961 ). (42) Vanderkooi, W. N., Schultz, J. Α., Sieglaff, C. L., private communication, The Dow Chemical Co., Midland, Mich., November 1958. (43) Zimmt, W. S., J. Appl. Polymer Sci. 1, 323 ( 1959 ). RECEIVED

September 6, 1961.


Inverse Emulsion Polymerization - American Chemical Society

Recommend Documents