Rheology of PVC Dispersions

I. Effect of Particle Size and Particle Size Distribution

E D W A R D A. C O L L I N S AND D. J. H O F F M A N N
Diamond Shamrock Corporation, T. R. Evans Research Center, P.O. Box 348,
Painesville, Ohio 44077

AND

P R A V I N L. S O N I
Case Western Reserve University, Cleveland, Ohio 44106
Received February 9, 1979; accepted April 9, 1979

PVC particle fractions prepared by air classification were used to study the effect of
particle size and particle size distribution on viscosity, viscosity-shear rate, and viscosity-
time behavior. The results show that a broadening of the distribution leads to elimination of
the shear instability and a decrease in the shear thickening (dilatant) observed at high shear
rates. This change is found to be more pronounced when the size distribution is broadened by
the addition of coarse particles. Addition of coarse particles is also shown to result in a reduc-
tion in the viscosity at low shear rates. The results are interpreted in terms of particle
packing and electrokinetic effects.

INTRODUCTION ability of the smaller particles to fit the

The rheology of vinyl plastisols is k n o w n
to be affected significantly by the particle
size and size distribution of the PVC resin.
Broadening of the size distribution results
in a decrease in viscosity of the paste. It
has been suggested that this is achieved
mainly by the alteration of void spaces. In
suspensions of spherical particles in liquids
there are void spaces b e t w e e n the particles
which h a v e to be filled by the plasticizer.
If all the spheres are of the same size, it
has been shown that flow will o c c u r only at
plasticizer levels of - 4 0 vol% or greater.
This is a p p r o x i m a t e l y equal to the void
volume of closely layered planes of hexag-
onally p a c k e d spheres (13). If, h o w e v e r , a
polydisperse s y s t e m were available, the
required plasticizer v o l u m e p e r c e n t a g e
would tend to decrease, b e c a u s e o f the
21
Journal o f Colloid and lnterjhce Science, Vol. 71, No. 1, August 1979
interstices b e t w e e n the larger hexagonally
p a c k e d particles, thus filling up some o f the
void space. Using the same level of plasti-
cizer, a polydisperse s y s t e m would thus
h a v e more plasticizer available as the dis-
persing m e d i u m than would a m o n o d i s p e r s e
system. This would, therefore, a c c o u n t for
the lower viscosity of a paste with a wider
distribution in particle sizes. E x p e r i m e n t s
with different size distributions ( 1 - 6 ) have
b o r n e out these observations.
If m o n o d i s p e r s e d suspensions are used it
is found that the viscosity increases with a
decrease in particle size, and that this effect
is more p r o n o u n c e d at low shear rates
(1, 4, 7-11). H o w e v e r , if reported in t e r m s
of reduced viscosity and dimensionless
shear rate (15), the viscosity for suspensions
of noninteracting spheres should be inde-
pendent of particle size. The a s s u m p t i o n of
0021-9797/79/100021-09502.00/0
Copyright © 1979 by Academic Press, Inc.
All rights of reproduction in any form reserved.
22 COLLINS, HOFFMANN, AND SONI

I00 IO0

80 80

t/)
-- 60
<
< 60

t.d
I.d
>
> 40 40
_J
.J
u.l
nr 2O ZO
u

A-I
I
I00 IO i O.

DIAMETER (MICRONS)

I00 io0

8O -- eo~
u~
c~
~')
< 60
w
W <
~o

/
_J

W
11~ z o zo U

A-2 \
tO0 IO o.
DIAMETER ( M I C R O N S )

FIG. 1. Particle size distribution curves of samples A-1 through A-4 in dibutyl sebacate, as
measured by a Micromeritics Sedigraph 5000D.

noninteracting spheres is not always valid
and some influence on viscosity should be
observed.
In most of the studies reported thus far
in the literature, the effect of particle size
and size distribution on viscosity has been
treated in a narrow range of shear rates.
Thus, although the general nature of the
effect on viscosity is understood, it is not
clear how the size distribution affects
viscosity over a broad range of shear rates.
In particular, the effect of size distribution
on the instability point (13, 14) is not under-
stood. The effect of particle size on aging
has also received very little attention. It
Journal of Colloid and Interface Science, Vol. 71, N o . I, A u g u s t 1979
is the purpose of this work to study these
effects in detail.
EXPERIMENTAL
Pastes were prepared by mixing 60 parts
o f dioctyl phthalate (DOP) per 100 parts of
resin (by weight) in a small container, using
a laboratory, stirrer to achieve a uniform
dispersion.
Viscosity measurements were conducted
on two viscometers with different ranges
in order to c o v e r a broad range o f shear
rates. The low-shear viscosity (10 -2 to 10
see -1) was measured on a Rheometrics
cone and plate viscometer equipped with a
 RHEOLOGY OF PVC DISPERSIONS 23

,°°L I00

8O 8o

co
to
cO <
60 60 5:
<
w
>
40 -- 40
_.1
D
...1
w D
of g0 20 U

A-3
0
I00 IO 0.1

DIAMETER (MICRONS)

I00 10o

80 -- 80

to
v9 to
<
to
< 60 6O 5;

W
>
w
>
--
I-
4.0 40 ~
._1
< D
,._1
or" 2:0 D
20 U

A-4
I00 I0 o.I
DIAMETER (MICRONS)

Fig. 1--Continued.

sensitive load cell (1 x 10 -6 to 1 x 10 -4 Stokes' equation is used to calculate the

kg-m). High-shear viscosity measurements
were made on a modified Ferranti-Shirley
cone and plate viscometer. The overlap in
data points from the two instruments was
found to be quite satisfactory. All measure-
ments were made at 79°F (26°C). Aging
was studied by making measurements at 2
hr, 1 day, and 7 days after the preparation
of the paste.
Particle size measurements were made
using a Micromeritics Sedigraph 5000D
particle size analyzer. This instrument
operates on the principle of gravity settling
of the particles in the suspending medium.
diameter of the particles.
Suspensions required for the particle
size analysis were prepared as follows:
0.9 g of the powdered polymer was first
dispersed in 25 ml of the suspending
medium (dibutyl sebacate) by means of a
magnetic stirrer. This suspension was then
sonicated in an ultrasonic bath for 10 min
to break up the loose aggregates and form a
uniform suspension. The sonification time
was chosen through experimentation to
yield a particle size distribution equivalent
to a lab-mixed DOP plastisol.
The PVC resin was divided into different
Journal of Colloid and Interface Science, Vol. 71, No. 1, August 1979
24 COLLINS, HOFFMANN, AND SONI

.I00 I00

80 -- 80~

U3
u3
< 60 -- 60~;

bJ
>

> 40 -- 40
_J
< D
_1
Ld
20 m --20 U

B-I
I00 I0 0.1
DIAMETER ( MICRONS)

I00 I00

8O -eo~
<
60 80
kl.l
>
hi <
> 40 40 ~
I-- _J
<

S
,_1
laJ
r~ 20 ~ 20t)

B-2
I00 IO 0. I
DIAMETER (MICRONS)

I00 100

80 --80 1~
t~
t/)
<
<[ 60 --60 ~;
:E
b.l
t.i >

4O --40 ~
I- _J
< D
_1
W
20 ~ 20 U

B-3
I00 I0 I 0.1
DIAMETER (MICRONS)

FIG. 2. Particle size distribution curves of samples B-1 through B-3 in dibutyl sebacate, as
measured by a Micromeritics Sedigraph 5000D.

Journal of Colloid and Interface Science, Vol. 71, No. 1, August 1979
 RHEOLOGY OF PVC DISPERSIONS
size-range fractions by means of air classi-
fication.
RESULTS AND DISCUSSION
Two types of resins (A and B) were used
in the preparation of the plastisols. The dif-
ference between these resins was the form
of emulsifier used during polymerization,
resin A with a branched sodium alkyl
sulfate and resin B with a linear sodium
alkyl sulfate. Each resin was divided into
fractions with various size distributions by
the air classification technique. Due to the
intensity of the shearing forces present
during air classification it is felt that the
particles are primary particles or well-fused
agglomerates. A total of seven fractions
were used, four fractions of resin A and
three of resin B. The size distributions of
the various fractions used are shown in
Figs. 1 and 2.
Flow curves of sample A-1 are shown in
Fig. 3. An abrupt increase in viscosity is
observed at ca. 100 sec -1. Similar jumps in
viscosity have been reported by Hoffman
(14) and Williams (9). It has been suggested
that this jump in viscosity results from the
breakdown of an ordered, layer-like particle
structure. The size distribution of sample
I03
. . . . . 2 HOURS
. . . . . I DAY
I WEEK
102
u3
~01 _
,o ° I I
io-2 ~o-I
FIG. 3. Viscosity vs shear rate for sample A-l, viscosity instability point at -100 sec -1.
25
A-1 (Fig. 1) was found to be quite narrow,
with average diameter of 1 /zm and with
10% of the particles having a diameter less
than 0.5/xm. This shows that the presence
of particles finer than the average diameter
in quantities up to 10% of the total weight
does not eliminate the onset of the in-
stability point. However, further increase in
the fines' content (15%) results in a shift
to higher shear rates or an elimination of
the instability point, presumably caused by
a broadening of the size distribution due to
the higher content of fine particles.
The size distribution of sample A-2
(Fig. 1) is essentially identical to that of
sample A-l, the only difference being that
the former has 5% more fines than the latter.
However, their viscosity curves are quite
different at the high shear rates, as can
be seen by comparing Figs. 3 and 4. In
particular, the sudden jump in viscosity at
ca. 100 sec -1 is absent in Fig. 4.
When the amount of coarse fraction is
increased, a similar effect is observed.
Thus, sample A-3, Fig. 1, which has a broad
distribution of particles centered around
9/xm in addition to the narrow distribution
centered around 1 ~m, exhibits a gradual
dilatancy in contrast to the abrupt increase
g o ffp 1.00
I 0 ~i, O.SO,~.
I I
to o to ~ io 2 103
~I SEC - I
Journal of Colloid and lnterJace Science, Vol. 71, No. 1, August 1979
 26 COLLINS, HOFFMANN, AND SONI

10 3

HOURS 8 5 '~b 1 . 0 0 x4
DAY I5 ~ O.50zz.
WEEK

10 2

t/)

i01

,oo
I I I
10-2 I0"1 I00 I01 I02 tO3
~sEc -t
FIG. 4. Viscosity vs shear rate for sample A-2.

seen in sample A-2. Furthermore, at still
higher shear rates, the flow in the case of
sample A-3 becomes pseudoplastic (Fig. 5).
Increasing the total amount of coarse frac-
tion from 25% (in sample A-3) to 30% (in
sample A-4) is found to have little effect on
the flow behavior as seen from Figs. 5 and 6.
These results suggest that an increase in
size distribution acts to suppress the sudden
jump in viscosity seen in pastes with a
narrow particle size distribution.
The viscosity curves of sample B-1 are
shown in Fig. 7. It is apparent that although
samples A-1 and B-1 have similar size
distributions, their viscosity curves are
significantly different. At low shear rates
the viscosity is seen to be much higher for
B-1 than it is for A-1. Furthermore, sample
B-1 does not exhibit the sudden increase in
viscosity which is seen in the flow curves of
A-1. Since the only difference between
these resins is the amount and type of

10 3

. . . . . 7' HOURS 25 '~'o 2 - 2 0 zz.

. . . . . I DAY 6 2 91, I . 0 5 zz
I WEEK I 3 ~ O. 5 0 / d -

10 2

t/1

O.

i0 I

0
IO I I I I
10-2 I0 -I tO 0 I0 I 102 103
II SEC -i

FiG. 5. Viscosity vs shear rate for sample A-3.

Journal of Colloid and Interface Science, Vol. 71, No. 1, August 1979
 RHEOLOGY OF PVC DISPERSIONS 27

103

..... 2 HOURS 30 ~/o 2 - 2 0 Jd~

..... I DAY 53 ~ 1.00~
I WEEK 17 ~e O.S0z~

102

u3

101

,oo I I I I
10"2 |0-1 i0 0 i01 10 2 103
SEE -~

FIG. 6. Viscosity vs shear rate for sample A-4.

surfactant on the surface of the particles, it
is suggested that this very difference is
responsible for the drastically different flow
behaviors. It is possible that interparticle
repulsion is quite significant in the case of
B-I giving rise to the high viscosity at the
low shear rates. A similar effect in poly-
styrene latex suspensions has been reported
by Eguiluz and Krieger (15). They found
that when the interparticle repulsion was
increased by varying the concentration of
the added electrolyte, the low-shear vis-
cosity increased significantly.
The higher interparticle repulsion may
also be responsible for the absence of the
viscosity jump in sample B-1. Hoffman (14)
has suggested that electrical and van der
Waals forces act to keep the particles in
layers. When the shear stress overcomes
these forces, the layer-like structure is de-
stroyed and an abrupt jump in viscosity
results. It is possible that in sample B-l,

iO3

I02

u3
b
EL

101

..... 2 HOURS 90 ~'o l.Odd~

..... I DAY 10 ~ 0. S£L
I WEEK

,oo
I I I I
10-2 i0 -I i00 i01 102 103
SEC - i

FIG. 7. V i s c o s i t y v s s h e a r r a t e for s a m p l e B-1.

Journal of Colloid and Interface Science, Vol. 71, No. 1, August 1979
28 COLLINS, HOFFMANN, AND SONI

103

.,,x
15 ~o 3- 20 z/.
" - ~ 70 * 1.00 IJ.
~'-<~ % ~ 15 % O. 50X,L

102
tO
b
13.
i='
I01 . . . . . 2 HOURS "" •~ •~ " -- "'--"~"-
. . . . . I DAY
- - I WEEK

I00
I I I I
10-2 I0-1 100 I01 102 103
lJ SEC -I
FIG. 8. Viscosity vs shear rate for sample B-2.

the higher interparticle repulsion would re- found to be suppressed. On further increas-
quire a higher shear stress to break up the ing the amount o f coarse fraction from 15 to
structure. This may cause shear thickening 45% (Fig. 9), the low-shear viscosity de-
to appear at the higher shear rates. creases further and the dilatant behavior at
Sample B-2 has a similar particle size the high shear rates is completely elimi-
distribution to that of B-1 except there are nated, resulting in a continuously shear
15% coarse particles, Fig. 2. The flow thinning flow.
curves o f sample B-2 are shown in Fig. 8. F r o m these observations, it can be con-
It is apparent that the low-shear viscosity is cluded that a broadening of the size dis-
lower than that of sample B-1. The dilatant tribution of the particles results in a de-
behavior at the high shear rates is also crease in the shear thickening effect at

103

45 9k 3 - 30
"."~.'~ 8 ~z. O.SO~

I0 2

cO

(3
EL
"~" ~ . ~ . ~ , . ~
"-..~..~_
I01 - - . . . . . 2 HOURS "~
..... | DAY
I WEEK

0
tO I I I I
10-2 I0 -I I0 0 I01 102 I03
$ S E C -a

FIG. 9. Viscosity vs shear rate for sample B-3.

Journal of Colloid and Interface Science, Vol.71, No. 1, August1979
 RHEOLOGY OF PVC DISPERSIONS
high shear rates. This change is more
prominent when the size distribution is
broadened by the addition of coarse
particles. Addition of coarse particles also
results in a reduction in viscosity at low
shear rates. Furthermore, the amount and
type of emulsifier associated with the
particles has a significant influence on the
flow behavior of the pastes.
EFFECT OF PARTICLE SIZE DISTRIBUTION
ON AGING
It is clear from Figs. 3 and 4 that a small
increase of the amount of fine fraction has
an insignificant effect on aging or the in-
crease with time. However, when the coarse
particle fraction is increased, as shown in
Figs. 4 - 6 , aging (at high shear rates) is
found to be rapid in the first 24 hr with
little aging occurring between 1 and 7 days.
It is also clear that at low shear rates,
changes in particle size distribution have
very little effect on aging.
With resin B, the reverse is found to be
true. A narrow distribution reveals rapid
aging in the first 24 hr (Fig. 7). By broaden-
ing the size distribution (Figs. 8 and 9), the
viscosity increase in the first 24 hr is found to
decrease. Thus, it becomes apparent that
the type of emulsifier associated with the
29
particles not only affects the flow behavior,
but also influences the aging characteristics.
REFERENCES
1. Kline, M., SPE VIPAG Newslet. 1(3), 14 (1964).
2. Bjerke, O., in "Advances in Polymer Science
and Technology," SCI Monogr. No. 26, pp.
370-380 Soc. Chem. Ind., London, 1967.
3. Soldatov, V. M., Kirillov, A. I., Mol'kov, A. D.,
and Sharikova, L. I., Soy. Plast. 6, 3-5 (1970).
4. Peggion, E., Testa, F., and Gatta, G., Chim. Ind.
(Milan) 46(1), 9 (1964).
5. Gatta, G., Vianello, G., and Benetta, G., Chim.
Ind. (Milan) 51(11), 1234 (1969).
6. Shtarkman, B. P., Voronkova, I. A., and Soldatov,
V. M., Kolloid. Zh. 34(4), 594 (1972).
7. Khanna, R. K., J. Oil Colour Chem. Assoc.
57, 161 (1974).
8. Willey, S. J., Ph.D. dissertation, University of
Minnesota, 1977.
9. Williams, G. E., Ph.D. dissertation, Lehigh
University, 1977.
10. Mathews, G., in "Vinyl and Allied Polymers"
(P. D. Ritchie, Ed.), Vol. 2, pp. 242-274 Iliffe,
London, 1972.
11. Palmgren, O., "Proceedings, International Sym-
posium on Macromolecules, Macromolecular
Conference of IUPAC, Helsinki, 1972."
12. Gillespie, T., J. Colloid Interface Sci. 22, 554
(1966).
13. Hoffman, R. L., Trans. Soc. Rheol. 16(1), 155
(1972).
14. Hoffman, R. L., J. Colloid Interface Sci. 46,
491 (1974).
15. Krieger, I. M., and Eguiluz, M., Trans. Soc.
Rheol. 20(1), 29 (1976).
Journal ofColloid and lnterjace Science, Vol. 71, No. 1, August I979