I
A. C. ZETTLEMOYER and JACK VANDERRYN Lehigh University, Bethlehem, Pa., and Kentile Corp., Brooklyn, N. Y.
Compatibilities of Coumarone-indene Resins-Solvent Systems Information essential to intelligent use of coumarone-indene resins in large consuming fields
T H E P H Y S I C A L - C H E M I C A L BEHAVIOR of polymer solvent systems has received a great deal of attention in recent years. However, the behavior of polymer solutions of low molecular weight has not received much attention, in spite of the importance of such polymers, as evidenced by their wide use in the manufacture of varnishes, asphalt tile, coating compositions, and rubber products. This study was undertaken to investigate the nature of the interaction of a polyindene-type polymer of low molecular weight with solvents or plasticizers. Solubility relationships of this polymer,
Table I.
whose molecular weight distribution and viscosity characteristics had been investigated in this laboratory (4, 5, 72), were determined in a wide variety of organic solvents and with other low polymers. Solubility and critical composition were measured first to give an insight into the type of interactions that occur. The results were compared with the behavior of other polymer-solvent systems reported in the literature. Then the heat of solution of the polymer in the various solvents was precisely determined to reveal some of the quantitative thermodynamic characteristics of the
Properties of Polymers
Softening Iodine Aniline Mol. Wt. Method4 Pt., O C* No. PointC I 769 1 100 55 60 2 1023 765 3 I1 Modified polystyrene 914 1 115 19 2868 2 973 3 I11 Mod. methylstyrene 645 1 100 135 10 IV (Coumarone)-indene 749 1 105 41 60 1023 2 86d 1. Cryoscopic. 2. Weight average, calculated from data of Pieski ( 4 ) . 3. Number average, calculated from data of Pieski ( 4 ) . b Ring and ball. c Solution temperature of equal weights of aniline and sample. d Capillary method. Resin
Type Coumarone-indene
I . .
Q
220
INDUSTRIAL AND ENGINEERING CHEMISTRY
interaction ( 7 1 ) . The heat of solution work will be reported later. Some doubt has always existed regarding the validity of the use of “cloud point” and aniline point determinations of resins and oils to serve as an index of solubility characteristics of commercially useful polymer systems. Solubility of resins in mineral oil were thus further investigated, in order to understand more fully the significance of the cloud point type of determination and to make possible comparisons with actual resinplasticizer systems. The solubility parameter for polyindene was determined. Critical compositions and solubility of the different molecular weight fractions were measured and the results compared to those obtained with high polymer solutions. Materials
Polymers. The four polymers used in this study were s)-nthetic resins of low molecular weight (manufactured by the Pennsylvania Industrial Chemical Corp.), They included two coumarone-indene, one modified polystyrene, and one modified methylstyrene resin. Some of the pertinent properties of these polymers are given in Tables I and 11. Resins I and I1 are identical to those except that different used by Pieski (4, samples were used, which may have had
WEIGHT PERCENT RESIN
Figure 2. Precipitation temperatures of resin I-resin mineral oil system
Precipitation temperatures of resins Figure 1. I and II in mineral oil
slight variations in properties. Resins I, 11, and I V correspond to Pieski’s numbers 11, IV, and 11, respectively. Resin I V is the same sample used by Pieski and the molecular weight fractions prepared by him were used in the present study for measuring solubility and heat of solution. The molecular weight and softening points of these fractions are listed in Table 111. The Three fractions of highest molecular weigh are “solids” having a density of 1.09 grams per cc., whereas the X-5 fraction is a viscous, tacky liquid having a density of 1.15 grams per cc. The four uqfractionated resins all have densities very close to 1.09 grams per cc. Resins I and I V are similar, except that they were manufactured a t different times and thus vary somewhat in their properties. The chemical analysis shows that resin I V contains little or no coumarone, as the oxygen content is low; therefore it is referred to as a polyindene polymer. Resins I and I V consist largely of polymers of indene and of methylindene. Resin I1 is a mixture of about 75% low molecular weight polystyrene and 25% polyindene, each polymerized separately. The chief constituents of resin I11 are isomers of methylstyrene with small amounts of triolefins and straight-chain compounds. Solvents. The organic solvents used in the solubility studies were reagent grade chemicals, dried over anhydrous magnesium sulfate. The mineral oil was a standard blend of paraffin hydrocarbons adjusted to an aniline point of 102’ C., prepared by blending 65 parts of Nujol with 35 parts of No. 415 oil (Magie Brothers, Chicago, Ill.). This is
the standard aniline point for mineral oil used in many laboratories to test solubility properties of the type of resins used in this study.
Methods Solubility. Solubility of the resins and mixtures of resins in mineral oil was determined by heating samples in a 4-inch test tube in a glycerol bath until a definite separation or “cloud” was evident with stirring. The uncertainty in the precipitation temperature in most determinations was f0.2’, but at concentrations greater than 0.1 5 volume fraction it rose to as much as &l.O”. Solubility in all other solvents was determined in a similar manner, except that sample tubes were reweighed after the determination if a volatile solvent was used. Very little solvent was lost in most cases. ______
Table II.zChemicaI Analysis of Resins I and IV Material % C Resin I 87.4 91.5 Resin IV Polyindene (CsaH66) 93.1
%H
%O
8.8 7.9 6.9
3.8 0.6 0
Table 111. Molecular Weights and Softening Points of Fractions
Fraction No.
x-1 x-2 x-3 X-4
Molecular Weight (Cryoscopic) 1790 1350 842 40 1
Softening Point (Capillary),
c.
162 135 94 27
111-
Results
Mineral Oil Systems. The precipitation temperatures of resins I and I1 in mineral oil were determined at resin concentrations from about 1 to 8070, as shown in Figure 1. Resin I11 was completely soluble in mineral oil a t all concentrations. Figure 1 also indicates the usual behavior for a high polymersolvent system (9) where the critical occurs , at concentration of polymer, V Z , ~ a value much less than 0.5. The precipitation temperatures of three-component mixtures of resins I, 11, and 111 with mineral oil were determined and the results plotted on triangular coordinates. These results are plotted in Figures 2, 3, and 4 as isotherms a t various temperatures. The numbers on the isotherms indicate the temperature. The resins are all completely soluble in each other a t all concentrations. The curves indicate that resin I is much more soluble in mineral oil than resin 11. Mixtures of resins I and I11 with mineral oil show regular changes in solubility; the precipitation temperature decreases regularly as the ratio of resin I11 to resin I is increased. In contrast to this, systems containing resin I 1 maintain a constant precipitation temperature as the ratio resin 11-resin I or resin 11-resin I11 is decreased until a low concentration of resin I1 is attained. The phase diagram for the system resin I-resin 111-mineral oil, shown in Figure 2, exhibits a close similarity to the phase relationships for the system polystyrene toluene paraffin hydrocarbon studied by Powers (6, 7). The com-
-
VOL. 49, NO. 2
-
FEBRUARY 1957
221
MINERAL OIL
/\ //
\/
\ RESIN II
Figure 3. Precipitation temperatures of resin I-resin IImineral oil system
parison indicates that resin 111 exhibits a behavior similar to toluene, which is the solvent in Powers’ system. This analogy seems reasonable, as resin I11 is completely soluble in mineral oil. Further analysis shows, however. that the critical resin composition is not appreciably increased (as it is with toluene) as resin I11 is added to either resin I-mineral oil or resin 11-mineral oil mixtures. This finding is emphasized in Figure 5 where the results for the resins II-IIImineral oil system are redrawn by plotting precipitation temperatures against resin concentration at various resin IIImineral oil ratios. The numbers on the curves inaicate the percentages of resin 111 in the mineral oil-resin I11 mixtures. Figure 5 indicates that the critical resin compmition is increased only from 5 to 8% as the percentage of resin I11 is increased. This indicates that resin 111 is a relatively poor solvent for resin IImineral oil mixtures, even though it is miscible in all proportions with the components individually. Analvsis of the resins I-111-mineral oil system yields similar results. Resin I1 appears to superimpose its insolubility on mixtures of resin I or 111 and mineral oil. In other words, resin I1 controls the precipitation temperature of the system. This superposition is attributed to the fact that resin I1 contains a fraction of high molecular weight which greatly influences its solubility characteristics. Mixtures of resins I and I11 with mineral oil show a regular change in solubility, because their molecular weight distribution curves are similar. The enhanced solubility of resin I11 in mineral oil is attributed in part to increased content of unsaturates in this resin, as indicated by its high iodine number.
222
INDUSTRIAL AND ENOINEERING CHEMlSTRY
25c
Figure 4. Precipitation temperatures of resin Il-resin 111mineral oil system
-
I
2 30
w
\
I
I0=10%RESIN 111
210
190 0 0
I 70
150
130
I IO
0
20
IO
20 30 40 WEIQHT PERCENT RESIN I I
Figure 5. Precipitation temperatures of resin Il-resin 111mineral oil system
50
These preliminary studies point up several important observations. The precipitation temperature, or solubility, of a n unfractionated polymer is principally controlled by the fraction of highest molecular weight. The solubility curves for the low polymers used in this study exhibit critical compositions a t unusually low polymer concentrations. I n this respect they act like polymers of much higher molecular weight. Resin 111, although it is completely miscible in the system, does not act like a true solvent, because it does not appreciably change the critical composition. Although several polymers-Le., resins I, 11, and 111-have almost identical solubility parameters, their quantitative solubilities in a system may be different. Determination of Soulbility Parameters of Polymers
Solubilities of resins I, 111, and I V were determined a t 20 weight 7 0 resin by the methods discussed (Table IV). The polymers used in this study were completely miscible with esters, ketones, and such common solvents as pyridine, chloroform, carbon tetrachloride, and benzene. Partial miscibility was found with the paraffin hydrocarbons, alcohols, nitromethane, nitroethane, and aniline. Using the data in Table IV, the solubility parameter for resin I was estimated
(2) to be 9.4; the solubility parameter of resin I V is the same or slightly less. The solubility parameter for resin I11 is slightly less than 9.4 because of its increasedsolubilityin octadecane (a = 8.2) and decreased solubility in nitromethane ( U = 11.1). The parameter for polystyrene, 9.2, approximates the value for resin 11, as the solubility parameter of a polymer is not dependent on its molecular weight and does not appreciably differ from its monomer. The solubility parameters for resins I and I V have also been calculated from their solubility in aniline, using the relation among critical solution temperature, critical composition, and solubility parameters ( 2 ):
18
16
14
12
$ IO
8
where X I and x2 are the mole fractions of polymer and solvent a t T, and u1 and u z are the solubility parameters. Using Equation 1 and the solubility curves of resins I and I V in aniline, the parameter for both resins was found to be 8.9. The agreement of this value with the value determined above (9.4) is good, especially in view of the fact that aniline is a hydrogen-bonding solvent. The average value for the parameter for polyindene thus is 9.2. Therefore the solubility parameters of the four polymers used in this study do not differ appreciably.
6 X OCTAOECANE
0.2
0.4
k
0.6
0
I
Figure 6. Precipitation temperatures of resins I and IV with normal hydrocarbons
Critical Composition and Determination of Thermodynamic Parameters
The expected maximum in the solubility curve (critical composition) is given by Equation 2 : Table IV.
Solubility of Resins I, 111, and IV (20% by weight resin)
Solvent
Solubility Parameter
n-Hexane 7.3 n-Heptane 7.4 Ethyl ether 7.4 n-Octane 7.55 n-Octadecane 8.2 Cyclohexane 8.2 Dimethylaniline 8.4 Benzonitrile 8.4 n-Butyl chloride 8.4 Methyl chloroform 8.5 Carbon tetrachloride 8.6 Ethyl benzene 8.8 Ethyl acetate 9.1 Benzene 9.15 Chloroform 9.3 Methyl ethyl ketone 9.3 Dibutyl phthalate 9.3 Chlorobenzene 9.5 10.0 Nitrobenzene Carbon disulfide 10.0 Pyridine 10.7 1-Nitropropane 10.7 Aniline 10.73 n-Amyl alcohol 10.9 Nitroethane 11.1 n-Butyl alcohol 11.4 Cyclohexanol 11.4 Acetonitrile 11.9 Nitromethane 12.6 Methanol 14.5 s. Soluble above 5' C. i. Insoluble at boiling point of solvent Precipitation temperature, O C. Q
Resin I
Solubility Resin I11
Resin IV
i i
i i
i i
s
S
s
i 180" 78"
i 136a i
i 1 70a
S
S
S
s s s
8
S
s
S
S
S
6
s
S
s s
8
s
s
S
6
s
S
s
S
S
8
s s
s s
S S S
S
S S
s
s s
S
S
65" 106a 54a 1 lo=
s
S
8
46a 82a 84"
s
s S
S
65a 83a 50"
s
s
102" s
i i i
i i i
i i i
where V2,c is the volume fraction of polymer at the critical solution temperature, T,, and X = V2/V1 where V Zand V I are the molar volumes of polymer and solvent, respectively. I n the early stages of this work, it was noticed that the critical polymer composition for some of the polymer solutions was much less than predicted by Equation 2. Table V lists the systems studied, their experimental and theoretical critical compositions, and their critical solution temperatures. The solubility curves obtained are presented in Figures 6, 7 , and 8. Figure 6 also includes data on n-dodecane and nheptane, to indicate the influence of paraffin chain length on solubility. Except for aniline and nitroethane, all the experimental values for the critical composition are considerably lower than those theoretically predicted. Preliminary investigations show that the critical compositions of resin I are also less than the theoretical value in n-butyl alcohol, 2ethylhexylamine, and hexyl ether. Inspection of Table V I shows that the poorer the solvent-i.e., the higher the VOL. 49, NO. 2
FEBRUARY 1957
223
I
2
1
O
s
X -I:AM
I10
loo 0
-2:HEPT Y L 90
).'
\X-I:4
-
P ENT E N O L I
" V
0.02
Figure 7.
0.06
01 . 8 0.2 2 0.26 OJ0 OJ4 Precipitation temperatures of resin IV fractions in alcohols
'9,
60
X-3:H E P T Y L
55 "C
50
critical solution temperature for the polymer-solvent system-the greater the deviation from the predicted critical composition. The highest critical solution temperature is exhibited by the long-chain paraffin, and the lowest by an aromatic derivative. The relation between the structure of the polymer and that of the solvent thus seems to affect the critical composition. The molecular weight of the individual fractions in this low molecular weight range did not appear to have any welldefined effect on the critical composition, although, according to Equation 2, the fraction of lower molecular weight should exhibit a higher critical composition. This implies that the chemical nature of the solvent-solute interaction is important in determining the critical compositions of polymers of very low molecular weight and not the molecular weight of the solute as theoretically predicted. Thus it appears that Equation 2 is not valid for these low polymer-solvent systems. Richards ( 8 ) has attributed the discrepancy between the theoretical and calculated value of the critical composition in polyethylene systems to an entropy of mixing which is different from the configurational entropy. Miller ( 3 ) has shown that if clustering occurs, the entropy of dilution is much less than the calculated value. To support the view that the entropy of mixing of polyindene with amyl and heptyl alcohols is low, or lower than that obtained with high polymers, values for the entropy and heat parameters of the resin IV-alcohol systems were calculated from the solubility data of the fractions of resin IV, using the method of Shultz and Flory (70)* The entropy and heat parameters, $12 and K12, are related to the critical solution temperature, T,, by the equation :
45
30
:;=(ik)[l+(~)(&+&)]
(3)
where e12 is the critical solution temperature for infinite molecular weight (inhite
i XI12
Figure 8.
224
Precipitation temperatures of resin IV fractions in alcohols
INDUSTRIAL AND ENGINEERING CHEMISTRY
1 TC
X), Figure 9 shows a plot of - us.
1 + 2x for the fractions of resin IV in
amyl and heptyl alcohols. Following the theoretical prediction, the data give a straight-line plot with the exception of the fraction of lowest molecular weight. This deviation from the straight line is probably related to the configuration of this fraction. End effects become more and more important as the molecular weight is decreased, and the spatial configuration of a polymer with only three or four units tends to be different from that of a polymer of higher molecular weight, which is able to attain a regular structure. Thus, the fraction of
L
,’,’
0 1
P
L?, H E P T Y L A L C O H O L
Figure 9. Reciprocal critical precipitation temperatures vs. 1 for resin IV fractions in amyl and heptyl alcohols
Table V.
Solvent Aniline Nitroethane Octadecane I-Octadecane Amyl alcohol
4-Penten-1-01 Heptyl alcohol
Table VI.
+
1/ 2 X
Critical Compositions of Polymer-Solvent Systems v2,c
v2,c
Solute
Calcd.
Exgtl.
Resin I Resin IV Resin IV Resin I Resin IV Resin IV Fraction X-1 Fraction X-2 Fraction X-3 Fraction X-5 Resin IV Fraction X-1 Fraction X-3 Fraction X-1 Fraction X-2 Fraction X-3 Fraction X-5
0.265 0.265 0.242 0.41 0.41 0.41 0.204 0.228 0.272 0.352 0.284 0.204 0.272 0.227 0.253 0.300 0.383
0.15 0.34 0.24 0.08 0.12 0.12 0.116 0.095 0.135 0.116 0.118 0.116 0.15 0.131 0.097 0.135 0.113
tc,
O
c.
63 65 53 180 170 122 113.5 90 57.5 18 85 85.5 27 115 96 58.6 17
Thermodynamic Parameters from Precipitation Data on PolymerAlcohol Mixtures
Component 1
Component 2
Amyl alcohol
Resin IV Resin IV Polystyrene
Heptyl alcohol Octyl alcohol
e12,
0
548 562 4 74
K.
$1 2
K12
0.674 0.742 1.30
1.24 1.40 2.05
lowest molecular weight might deviate from a linear solubility relationship such as shown in Figure 9, because of the difference in the number of sites which are available for polymer-solvent interaction. The 012 and $ q 2 values were obtained from the intercepts and slopes of the graph. Table V I lists the interaction parameters of the resin IV-alcohol systems along with the parameters for the polystyrene-octyl alcohol system given by Shultz and Flory (70). The entropy parameter, $12, should be approximately 1/2 if we consider only the configurational entropy and disregard first neighbor interactions. Shultz and Flory (70) have attributed the high values of $12 for the polystyrene-octyl alcohol system to the high degree of local order in the pure liquid alcohol. The entropy parameter for the resin IV systems is considerably lower than that for the polystyrene, even though there is little or no difference between the solvents. Thus the low polymer systems show a lower entropy of dilution parameter than the high polymer system. This finding is even more striking when it is noted that the theoretical treatment requires greater entropy of dilution parameter, the lower the molecular weight of the polymer. The high entropy parameter of the polystyrene-octyl alcohol system may mean that the bulk polystyrene is more ordered in its solid state than the polyindene and thus has a lower entropy. AS both polymers should attain approximately equal states of disorder in solution, the change in entropy on going into solution would be greater for the polystyrene. Thus the polystyrene system should have a higher entropy parameter.
literature Cited (1) Flory, P., J . Chem. Phys. 10, 51 (1942). (2) Hildebrand, J., Scott, R., “Solubility of Nonelectrolytes,” Reinhold, New York, 1950. ( 3 ) Miller, A., Nature 163, 838 (1949). (4) Pieski, E., Ph.D. dissertation, Lehigh University, 1949. (5) Pieski, E., Zettlemoyer, A. C., IND. ENG.CHEM.45,167 (1953). ( 6 ) Powers, P. O . , I b i d . , 41, 126 (1949). ( 7 ) Ibid.. 42. 2558 (1950). Richards, R., ‘Trans. Faraday SOC.42, 10 .. 11946). ,-. Shultz, A., Flory, P., J . Am. Chem. SOC.74, 4760 (1952). Ibid., 75, 3888 (1953). Vanderrvn, J.. Ph.D. dissertation, Lehigh University, 1955. Zettlemoyer, A. C., Pieski, E., IND. END.CHEW45,165 (1953). - I
RECEIVED for review November 23, 1955 ACCEPTED July 2, 1956 Division of Paint, Plastics, and Printing Ink Chemistry, Symposium on Phase Equilibria in Polymeric Systems, 128th meeting, ACS, Minneapolis, Minn., September 1955. Work supported by Kentile, Inc., Brooklyn, N. Y. VOL. 49, NO. 2
FEBRUARY 1957
225
