The Effect of Blending Resins on Plastisol

Performance
M. K. BANNISH and J. T. RENSHAW
Monsanto Industrial Chemicals Co.
Plasticizers Division
St. Louis, Missouri 63166

Blending resins are well known by plastisol compounders as an

economical substitute for PVC dispersion resin and as viscosity
modifier. As dispersion resin supply tightens, interest grows in blend-
ing resins and how to choose the one that best fits the compounder’s
specifications. Eleven representative blending resins have been
characterized, compounded into plastisols and screened for perform-
ance. A regression analysis and resulting contour plots have supplied a
means of correlating resin molecular weight, particle size and surface
area with plastisol, gel/fusion, and physical properties. The trends
described in the contour plots allow the prediction ofthe best blending
resin criteria for a specific application.

INTRODUCTION and three resins of similar particle size but varying

B lending resins affect plastisol properties in dif-
ferent ways. Although the literature discusses
the effects of certain blending resins on a single
plastisol property, there is a need today for an over-
view of major blending resins and their effects on a
wider range of plastisol properties. The purpose of
this study is to associate resin physical characteris-
tics with plastisol performance in terms of rheolog-
ical and physical properties, in order to identify the
resin that best suits a given application’s needs.
EXPERIMENTAL RAW MATERIALS
Nine commercial PVC blending resins were cho-
sen to represent the broadest available range of
molecular weight and particle size (Table 1 and Fig.
1). Resin performance can be affected by differ-
ences in PVC synthesis technology and production,
so resins of four different suppliers are represented
here. During the initial property screening, one
additional resin was eliminated on the basis of par-
ticle size duplication. Within this group of resins,
two series were studied: two resins of similar
molecular weights but with varying particle size,
Table 1. Blending Resin Propertles
molecular weights. Two co olymer resins were
P
also included, a 5 percent ma eic copolymer and a 5
percent vinyl acetate copolymer. Two specialt
homo olymer blending resins were also evaluated:
f
one o large particle size and one indigenous to the
flooring industry. The lasticizer used in these ex-
P
periments was a dialky C7.11 phthalate. A standard
commercially available stabilizer package was in-
corporated for processing and heat stability.
EXPERIMENTAL
A statistical experiment was designed to relate
blending resin physical properties to plastisol
properties. Physical properties were determined
on the resin itself: molecular weight (GPC), inher-
ent viscosity, particle size and particle size distri-
bution, surface area by nitrogen absorption and
plasticizer absorption by a Rub-Out Test (ASTM D
281).
For every blending resin, three plastisols were
compounded in 500-gram batches using a 4-quart
Hobart mixer. The basic formulation was 100 parts
homopolymer PVC dispersion resin with inherent
viscosity of 1.20 and 65 parts of dialkyl C7.11 phthal-

Molecular
Blending welght Inherent Particle Surface Rub-Out
resin ( x 1000) viscosity size (4 area (mYg)

A 104 .73 32 .35 49

6 108 .86 30 .36 55
C 137 1.15 23 .48 58
D 128 .83 73 .09 40
E 133 .83 35 .24 45
F 139 .89 57 .62 44
G 126 .78 62 .17 48
H 145 .97 26 .41 47
I 114 .74 40 .20 45
Standard 135 1.20 8 5.7 39

JOURNAL OF VINYL TECHNOLOGY, DECEMBER 1980, VOL. 2, NO. 4 227

 M . K . Bannish and 1. T. Renshaw

D
Thus the effect of four independent variables was
70 t G
used to determine the response of the rheological
properties of the plastisol and the physical proper-
ties of the fused vinyl composition (see Table 5 ) .
F The four independent variables were inherent vis-
cosity, resin surface area and particle size, and
8 blending resin concentration.
/
E METHOD OF DATA ANALYSIS
A
“5
..
c 30 e Each of the dependent plastisol properties was
z
CI
0)
0)
x C
C sensitive to one or more of the four blending resin
physical properties ( X ) found to be predominant.
The method used to evaluate the variable was a
10
1°1 S backstep regression analysis which produced an
equation for each plastisol property (Y), described
1 1 I I I I I I
by only those blending resin terms affecting it (X).
.6 .7 .8 .9 I 1.1 1.2
The other terms drop out.
Inherent Viscosity The equation model is:
Fig. 1 . Distribution of blending resin properties.
Y = c, + x , x j + xxcip,
ate. To describe the entire range of blending resin The regression program also gave t test values
effects, 10, 25, and 45 percent blending resin was which weighed each term according to its impor-
substituted for the dispersion resin. Table 2 lists the tance. The closeness of curve fit to the data points
formulation variations. was expressed as percent variation explained. The
From these plastisols, samples were cast with a standard deviation of residuals was a measure of
50-mil knife and fused in a forced-air 180°C oven variation in reproducibility. These equations clar-
equipped with a rotating table. Test samples were ified just what property in the blending resin pro-
prepared, aged and tested for tensile strength and duces specific changes in the plastisol. The results
60” gloss. The remaining plastisol was used to de- were also graphed as contour plots so that trends
termine gel temperature, Brookfield and Severs could be seen. This gave a visual means of predict-
viscosity, settling characteristics, and air release ing the optimum blending resin for a specific plas-
times. Actual raw data are tabulated in Tables 3 and tisol property desired.
4.
EVALUATION OF BLENDING RESIN SUMMARY OF THE RESULTS
PROPERTIES
In summary, the major effects of substituting
In the analysis of the physical properties of the blending resin in a plastisol are the reduction of
per se blending resins, the resin properties which viscosity and the diminishing of dilatancy charac-
reflected the same information were eliminated. To teristics. This viscosity reduction levels off around
describe a resin’s ability to absorb plasticizer, sur- 40 phr and is maximized by using a blending resin
face area by nitrogen adsor tion was selected for
accuracy over the ASTM Ru -Out Test. As an indi-
cator ofpolymer size, inherent viscosity was chosen
E of larger particle size and lower surface area. Con-
currently, air release is improved while tensile
strength, gloss, and gel temperature diminish. The
over molecular weight by GPC due to familiarity in degree to which these effects take place is contin-
the plastisol field with inherent viscosity numbers. gent upon the surface area, particle size, inherent
Particle size and particle size distribution was viscosity, and amount of the blending resin used.
determined by Coulter Counter (Model TA) and Thus one may maximize the desired effects and
verified visually by scanning electron microscope. minimize the loss of critical processing and end use
The validity of eliminating given per se properties properties by the careful selection of the blending
ofthe blending resins was verified in the regression resin characteristics.
analysis; i.e., duplicate information was dropped.
This program evaluated which properties in-
fluenced plastisol behavior and how closely they DISCUSSION
correlated. The regression equations along with their con-
Table 2. Formulations (phr) tour plots express the trends described by the data.
All the plots measure inherent viscosity versus per-
Standard 1 2 3 cent blending resin. To incorporate the other two
Homopolymer 100 90 75 55
variables, particle size and surface area, several
Blending resin - 10 25 45 different graphs were printed, each with different
,,
Di alkyl C, phthalate 65 65 65 65 values for particle size and surface area. The gel
Epoxidized soybean oil 3 3 3 3 temperature, for example, is shown here as isomet-
Modified barium-cadmium- ric temperature curves for a resin particle size of 48
zinc stabilizer 2 2 2 2
and a surface area of 0.3 m2/g.

228 JOURNAL OF VINYL TECHNOLOGY, DECEMBER 1980, VOL. 2, NO. 4

 The Effect of Blending Resins on Plustisol Performance

Table 3. Rheological Properties

Brookfield (poises, 50 rpm)

Gel Settling
Blending resin temperature, "C 2 hr 1 day 1 week 1 wk(ml)

A 10 phr 79 17 20 24 0
25 83 15 16 18 0
45 88 17 18 18 1
6 10 81 18 20 30 0
25 87 17 19 25 0
45 93 18 18 18 0
C 10 80 17 19 22 0
25 90 17 19 24 0
45 100 17 18 21 2
D 10 80 18 20 33 0
25 96 14 14 20 2
45 111 10 8 12 15
E 10 81 20 23 22 0
25 89 13 14 14 1
45 96 15 15 16 2
F 10 79 9 10 9 0
25 94 14 14 10 1
45 105 18 18 9 9
G 10 79 15 18 24 0
25 92 16 17 17 0
45 100 11 10 11 12
H 10 80 19 21 43 0
25 82 16 17 19 0
45 85 14 14 16 1
I 10 80 18 21 26 0
25 81 18 19 23 0
45 82 15 15 15 1
Standard 78 27 26 61 0

Table 4. Physical Propetties

Tensile
Blending Percent strength Abrasion
resln 60" gloss elongation (PSI) (grams lost)

10 phr 47 382 2533 ,0216

A 25 27 386 2430 ,0201
45 20 354 2234 .0345
10 53 390 2649 .0360
6 25 35 407 2465 ,0367
45 25 438 2576 ,0376
10 55 476 2861 .0420
C 25 32 449 2785 ,0215
45 19 423 2572 ,0318
10 57 349 231 1 ,0193
D 25 77 41 6 2065 ,0233
45 38 326 1 800 ,0263
10 43 363 2524 ,0214
E 25 25 347 2145 ,0215
45 25 340 2022 .0268
10 35 362 2294 ,0261
F 25 31 357 2073 .0279
45 31 308 1932 ,0342
10 56 377 2476 ,0259
G 25 38 375 2291 .0209
45 30 352 2042 ,0255
10 54 426 2512 ,0270
H 25 28 433 2632 .0369
45 22 429 2397 .0346
10 38 396 2221 .0342
I 25 21 390 21 25 ,0284
45 16 380 2007 ,0298
Standard 78 427 2743 ,0253

Gel Temperature shows that adding the initial amounts of blending

resin will affect the gel temperature noticeably, but
The regression equation shows that gel tempera- above 25 phr, little difference is seen unless very
ture is affected predominantly by percent blending high molecular weight blending resins are used.
resin used and its surface area. The contour plot The negative sign denotes an inverse relationship

JOURNAL OF VINYL TECHNOLOGY, DECEMBER 1980, VOL. 2, NO. 4 229

 M . K. Bannish and I. T . Renshaw

Table 5. Dependent Variables

Gel Temperature
Brookfield Viscosity
Wet Film Strength Test
Syringe Test
Settling
Tensile Strength
60 Degree Gloss

between gel temperature and surface area: as sur-

face area increases, more plasticizer is used to wet
the surface, there is less free plasticizer, and the el
temperature will decrease. The t test value fort is
term is 14, the highest value in this entire analysis,
a
which verifies the strong 97 percent correlation be-
tween resin property and el temperature (see Fig, OL
f
2). In actual practice, if ge temperature is a critical Fig. 3. Initial Brookfield.
parameter, the blending resin of choice would be
one with a high surface area and/or a lower molec-
ular weight. The choice of these types of blending
resins would have minimum negative impact on
this parameter.
Brookfield Viscosity
The viscosity is affected equally by the amount of
blending resin used, the particle size, and the sur-
face area. These negative first-order terms have a
high t test value for initial, one-day and one-week
results, which verifies the 92 percent correlation: as
more blending resin is added, the viscosity drops.
The three contour plots, Figs. 3 , 4 , and 5 , show the
drop in viscosity with the addition ofblending resin
up to approximately 40 phr. These results indicate ol
that the larger particle size, lower surface area Fig. 4 . One day Brookfield.
blending resins are most effective in reducing ap- INHERENT VISCOSITY
parent plastisol viscosity.
Settling
Figure 6 is a contour plot of the settling proper-
ties of a medium particle size blending resin as a
function of inherent viscosity and blending resin
concentration. The results here confirm the direct
correlation between increased settling with in-
creased particle size and amount of blending resin
used. I n compounding practice, settling properties

r
are quite often the limiting factor in the utilization
INHERENT VISCOSITY

Fig. 5 . One week Brookfield.

of blending resins. In applications where settling is

critical, blending resins usage is normally re-
stricted to filled plastisol formulations or formula-
Fig. 2. Gel temperature. tions with low plasticizer concentrations.
230 JOURNAL OF VINYL TECHNOLOGY, DECEMBER 1980, VOL. 2, NO. 4
 The Effect of Blending Resins on Plastisol Performance

INHERENT VISCOSITY

5
v)
W
w a

I1
0
Y
W
z-
0
z
w
-I
m
LL
0

T
Fig. 6. Settling.
I-
z
W
V
a
w
0-

Air Release
Both Huffs wet film strength test (Fig. 7 )and the
syringe test (Fig. 8) gave similar results. The initial Fig. 8. Air release (syringe test}.
25 phr blending resin brought about a two- to
threefold decrease in deaeration time. Additional
blending resin usage resulted in a diminishing ef-
fect on air release time. The Huff test gave 88 per-
cent correlation with the negative second-order
surface area term predominating. Air release is
aided by any emulsifier on the resin surface. As the
surface area increases, the air release time de-
creases. The syringe test gave an 82 percent corre-
lation of data. Ease of deaeration and release of
entrained air is an important plus contributed by
the use of blending resin in the plastisol area.
60 Degree Gloss
Gloss was lowered by any addition of blending
resin at any PHR level, shown in the equation by
two predominant terms, the first-order term of per-
cent resin used, and the second-order term of parti-
cle size times the surface area. The contour plot,
Fig. 9, reflects this relationship, and shows that a
lower molecular weight resin tends to depress gloss
more than a medium molecular weight resin.
As has been previously reported in the literature,
the data indicate that gloss is adversely affected by
the use of blending resin at any level (5). This is
illustrated in the regression analysis by the two
INHERENT VISCOSITY OL
Fig. 9. 60 Degree gloss.

predominant terms, the first-order term of percent

blending resin used and the second-order term of
particle size times surface area. In point of actual
end use effects, where gloss is a prime considera-
tion, blending resin usage is limited to about the 5
to 10 percent level and high surface area resins are
preferred.
Tensile Strength
The tensile strength of the fused plastisol is sen-
sitive to two blending resin properties, surface area
and the amount of resin used (see Fig. 10). The
ol correlation is 96 percent, a very close fit. As antici-
Fig. 7. Air release ( w e t film strength). pated, the contour plot trends show that tensile

JOURNAL OF VINYL TECHNOLOGY, DECEMBER 1980, VOL. 2, NO. 4 231

 M. K . Bannish and I. T. Renshaw

INHERENT VISCOSITY with plastisol. The amount of settled resin was

measured at intervals and recorded as milliliters of
resin settled. Method described by Friedman and
Gibby, "Blending Resins in Plastisols-How to
Evaluate Them," SPE 30th ANTEC Preprints, Vol.
XVIII (Part 2), pp. 329-340 (1972).

APPENDIX I1
Regression Equations for the Rheological and
Physical Properties as a Function of Blending
Resin Properties
A regression equation for each property, Y, is
listed below. The underlined terms carried the
heaviest influence (t value) over the dependent var-
iable, Y.
Y = Gel temperature ("C) =
89 + ml+ 22X2- 15x4 - 52x2, - 47x2,
OL + lOlXZ, + 179XlXZ - 0.79XIX3 - 331XiX4
Fig. 10. Tensile strength.
Percent variation explained = 97
Standard deviation of residuals = 2.2
strength increases with the surface area of the
blending resin and its molecular weight, and de- Y = Initial Brookfield (poises) =
creases regularly with the amount ofblending resin
substituted. The use of blending resin always re- 17 - 11x1 + 13X2+ 28x2, - 47x5
sults in a deterioration of physical properties of the - L0 57X 3 - 20X1X4+ 0.48XZX3
l - X
fused vinyl composition. This effect can be best
ameliorated by the use of low molecular weight Percent variation explained = 80
homopolymer blending resins or by the use of co- Standard deviation of residuals = 1.6
polymer blending resins.
Y = One-day Brookfield (poises) =
CONCLUSIONS
18 - 17X, + 16X2+ 35x2, - 66x5 - 0.57XIX3
As a general conclusion based on our data, it can
be stated that the use of blending resin will im- + 0.56X2X3
prove the rheological properties of a plastisol but Percent variation explained = 87
will negatively impact the physical properties of Standard deviation of residuals = 1.6
the fused plastisol. It is, therefore, mandatory that
the compounder balance these effects against the y = One-week Brookfield (poises) =
performance requirements of his application. And
with the results stated here, the fine tuning of this 22 - -
24X, + 13X2+ 28X4+ 5lX; - 145X;
balance is now possible. --
165x2, - 0.28X1X3+ 85X1X4- 2Xg4
APPENDIX I Percent variation explained = 88
Test Methods Standard deviation of residuals = 2.6
The test methods used throughout this study are
described below. Y = Settling (ml of sediment) =
Gel Temperature: Gradient Bar Method. A film 0.98 + -
15x1 - 8x4 + 63x2, + 65x2,
ofplastisol is spread on a foil covered aluminum bar
with a temperature gradient of 40 to 130°C in 10" + 28X1X2+ 0.98X1X3
increments. After 3 min, the foil is removedand the Percent variation explained = 93
line between flowho flow plastisol is determined Standard deviation of residuals = 1.4
by laying a strip of 1-mil plastic on top, then peeling
it off, starting from the gelled end. This point (the Y = Air Release (Wet Film Strength)""" =
gel point) on the film is correlated to the tempera-
ture of that point on the gradient bar. 16 - 15x1 + 39x2 + 7x4 + 78X: - 64x2,
Brookfield Viscosity: ASTM D 1824
Air Release: SPI-VD-t20 (1969) Wet Film + 4ox1x2- 0.53X1X3- 4ox1x4
Strength and Syringe methods. + 2.0X2X3- 0.91X&4
60" Gloss: ASTM D 523-67
Tensile Strength: ASTM D 638-77a Percent variation explained = 88
Settling: A 50-ml graduate cylinder was filled Standard deviation of residuals = 2.1

232 JOURNAL OF VINYL TECHNOLOGY, DECEMBER 1980,VOL. 2, NO. 4

 The Effect of Blending Resins on Plastisol Performance

Y = Air Release (Syringe)”““= Percent variation explained = 81

Standard deviation of residuals = 0.003
1.5 - m X , + 9.6X: + 9.OX; - 53x2,
+ 7.6X1X2 - 0.1X,X3 + 0.06X2X3 Where X, = Concentration of blending resin
X2 = Inherent viscosity
Percent variation explained = 82 X3 = Particle size
Standard deviation of residuals = 0.26 X4 = Surface area
Y = 60” Gloss (Percent Reflected Light) =
65X1 + 36X4 - 254x2, + l.lX1X3
36 - -
f 94xlx4 - 4 x J 4 ACKNOWLEDGMENTS
Percent variation explained .= 79 The authors wish to thank their co-workers in the
Standard deviation of residuals = 8.1 Plasticizers Division of Monsanto Industrial
Chemicals Company for their help in this program.
Y = Percent Elongation = Special thanks to Jack Hinchen for statistical assist-
409 - 60X1 + 297x4 - 480x2, - 921X; ance, and to the Physical Sciences Group at
Monsanto Corporate Laboratories. We wish to
- 1093x2,- llX2X3- 13x&~, thank Monsanto Company for permission to pub-
Percent variation explained = 80 lish the work, and the Society of Plastics Engineers
Standard deviation of residuals = 23 for including it in this meeting.
Y = Tensile Strength (psi) =

2480 --734x1 + 1582x4

-+ 1353x2,- 3683x2, REFERENCES
- 7907x2, - 1619X1Xz + 15xlx3+ 5431X,X4 1. L. J. Friedman and W. Gibby, “Blending Resins in
Plastisols-How to Evaluate Them,” SPE 30th ANTEC Pre-
- 53X2X3 - 4 8 X a 4 prints, Vol. XVIII (Part 2), pp. 329-340 (1972).
Percent variation explained = 96 2. L. J. Friedman, “Vinyl Dispersion Coatings for Foam Floor-
ing and Fabrics,” SPE 28th ANTEC Preprints, Vol. XVI, pp.
Standard deviation of residuals = 77 226-232 (1970).
Y = Abrasion (gr lost/lOO cycles) = 3. P. M. Peltz, Mod. Plast., 43, 154, 156, 158 (Sept. 1965).
4. J. T. Renshaw, J. A. Cannon, and J. D. Hinchen,PoZym.-Plost.
0.03 + 0.07x4 + 0.16X2, - 0.36X2, - 0.19X2, Technol. Eng., 6, 137 (1976).
5. J. T. Renshaw and J. D. Gabbard, Polym.-Plast. Technol.
- 0.12X,X2 - 0.06X1X4 - 0.002XJ4 Eng., 10, 131 (1978).

JOURNAL OF VINYL TECHNOLOGY, DECEMBER 1980, VOL. 2, NO. 4 233