CELLULOSE ACETATE FRACTIONS Intrinsic Viscosity and Osmotic Molecular Weight ARNOLD M. SOOKNE AND MILTON HARRIS' National Bureau of Standards, Washington, D. C.
The intrinsic viscosities and osmotically estimated number-average molecular weights of a series of cellulose acetate fractions have been measured. It was found that, within the range of chain lengths investigated (number-average molecular weight up to 130,000), the numberaverage molecular weights are proportional to the intrinsic viscosities in acetone solutions, in agreement with Staudinger's rule and the results of Kraemer. An estimate is provided of the relative homogeneity with respect to molecular size of the fractions and the starting material from which they were prepared.
A [VI,
authors were wed. The measurements were made a t room temperature (approximately 28" C.). The cell was placed in a box insulated with cotton waste, and the entire assembly was then covered with a wooden hood in order to minimize drifts caused by changes in temperature. Measurements of the osmotic pressures were made by means of a cathetometer sighted through a window in the hood. It was found practical t o use the same membrane for a series of measurements by using increasing concentrations of the same sample in successive measurements, as recommended by Fuoss and Mead (4). The membrane was thcn rimed thoroughly with acetone, soaked overnight in this solvent, and rinsed thoroughly again, all without taking the membrane from the cell. This treatment sufEced t o remove adhering cellulose acetate so that the membrane could be used again. With the fraction of lowest molecular weight (No. 15), and t o a .lesser extent with the starting material, appreciable diffusion of the solute through the denitrated cellulose nitrate membranes occurred. For these two samples, cellophane membranes swollen with ammonium hydroxide, according to the directions of Florj (a),were therefore used. These measurements were made by the static method in a constant-temperature room. Equilibrium was reached overnight, after which no appreciable change in osmotic pressure occurred for several days. Since the osmotic pressure measurements were made a t slightly different temperature@,the number-average molecular weights, were calcubted as follows: The apparent value of Bna t each concentration (and corresponding temperature) was first calculated, using the van%Hoff relation, and the value of %&a t zero concentration was t h e n obtained algebraically, assuming t h a t the apparent value of M,, is a linear function of concentration.
CCORDING to Staudinger's rule, the molecular weight, M ,
of a linear high polymer is related to its intrinsic viscosity, by the expression (9): I71 =
KM
K = a constant c = concentration, grams/100 ml. of solution The validity of this relation has often been questioned, and i t has recently been demonstrated (3,6 ) that several unbranched polymers do not conform to this rule, but rather t o the more general relation: [ q ] = KMa where a = a constant with a value between 0.5 and 2 Some time ago a series of cellulose acetate fractions was prepared in this laboratory, in connection with a study of the mechanical properties of this material (8). The starting material from which the fractions were prepared had an acetyl content of 38.6%, an ash content of 0.06%, a melting point of 250' C., and a char point of 301". The intrinsic viscosities and the osmotically estimated molecular weights of these fractions and the starting material have now been determined in acetone solutions. This paper, which presents the results of these measurements, shows that for this type of acetate and within the range of chain length investigated, the system cellulose acetate in acetone does conform to Staudinger's rule; i.e., fsr this system the con$tant a has the value unity in the relation cited above.
zn,
MEASUREMENTS
RESULTS AND DISCUSSION
ViscosITY. The viscosity measurements were made in a n Ostwald viscometer a t 25" * 0.03" C. Since the time of efflux of the acetone was approximately only 30 seconds, kinetic energy corrections were applied (1). These were determined by calibrating the viscometer with oils of known viscosity, obtained a t the National Bureau of Standards. (Failure to apply kinetic energy corrections leads t o errors of several per cent in the values of the intrinsic viscosities of the fractions of higher molecular weight.) OSMOTIC PRESSURE.The osmotic pressure values for all the fractions, except the one of lowest molecular weight, were determined by the dynamic method of Fuoss and Mead (4). The denitrated cellulose nitrate membranes recommended by these
The results of the viscosity measurements are presented in Table I and shown graphically in Figure 1, where the values of q b p / care plotted on a logarithmic scale. When this procedure is followed, the results fall on straight lines suitable for extrapolation to zero concentration for obtaining the intrinsic viscosity. (When the values of q*Jc and c are both plotted on a linear scale, the results do not fall on straight lines except a t very low concentrations, where the experimental errors are largest. The use of the semilogarithmic plot gives weight t o the more precisely determined values obtained in the range of higher concentration.) The results of the viscosity measurements conform quite closely to Martin's equation ( 7 ) :
1 Present address, Milton Harris Associates. 1246 Taylor Street, N.W., Washington 11, D. C.
475
INDUSTRIAL AND ENGINEERING CHEMISTRY
476
TABLE1. VISCOSITIESOF CELLULOSE ACETATE SAMPLES IS ACETONESOLUTIONS AT 25" C. Concn, Grams, 100 M1 T3P VP/C 0.126 0.218 1.73 Starting material 0.268 0.501 1.87 2.25 0.637 1.210 0.094 0.289 3.08 Fraction 2 0.273 0,999 3.66 0.546 5.08 2.77 0.096 0.278 2.89 0.345 I 27s Fraction 3 3.71 0.690 3.78 5.48 0.114 0.286 2.52 1.101 Fraction 4 0.351 3.13 3.12 0.703 4.43 0.110 0.251 2.28 0.384 Fraction 5 1.101 2.88 0.769 3.14 4.08 0.118 0.247 2.09 2.52 Fraction 6 0.353 0.891 3.48 0.775 2.70 0.119 0,224 1.88 2.22 Fraction 7 0.358 0.796 3.21 1,81 0.877 0.138 0.239 1.74 0.275 Fraction 8 0.521 1.89 0.888 0.428 2.07 0.152 0,226 1.49 0.42s Fraction 9 0.271 1.57 0,541 0,988 1.82 0.152 0.209 1.38 0.303 0.45e Fraction 10 1.49 1.81 0,684 1.238 0.149 0.179 1.20 1.30 0.385 Fraction 11 0.297 1r50 0,625 0.940 0.147 0.159 1.08 1.15 Fraction 12 0.294 0.338 1.26 0.533 0.671 0.72 0.149 0.108 0.74 0.269 Fraction 13 0.199 0.76 0.403 0.305 0.61 0.147 0.089 0.63 0.335 0.210 Fraction 14 0.66 0.390 0.587 0.231 0.273 0.063 Fraction 1 5 0.411 0.23a 0.097 T h e aarnples are numbered in the order of molecular weight to with their designations in a previous publication ( 4 ) . The lowest represents t h e sample of highest molecular weight. Sample"
~
C? I 1.59
2.75 2.59 2.25 2.06 .91 .72 ,60
.37 1.28 1.13
1.02 0.70 0.59
where c is the concentration and k is a constant characteristic of a given polymeric series in a given solvent. It should be noted that the concentration is expressed in grams per 100 ml. of solution in the data presented here, whereas it is given as weight per cent in the reference ( 7 ) . The values of [ q ] in Table I were obtained by algebraic extrapolation, using the method of least squares. The results of the osmotic pressure measurements are given i n Table IT. Since it has been shown that a plot of the reduced osmotic pressure ( T / c ) against c gives straight lines for samples of' cellulose acetate in acetone (Z), the osmotic pressure values were measured a t only two concentrations for most of the samples. The slopes of the lines obtained by plotting r / c against c are nearly identiral for all of the samples except the two fractions of lowest molecular weight (14 and 15) which give larger slopes. Fraction 15 diffused through the denitrated cellulose nitrate membranes of Fuois and Mead (4). When measurements were made by the dynamic method with these membranes and the results extrapolated back to zero time, the apparent value of %fnfor fraction 1.5 was 16,000. The value of when the less permeable swollen cellophane membranes were used was 11,000 (Table 11). Figure 2 shows the relation between the intrinsic viscosities and number-average molecular weights of the acetate fractions. The results appear to fall on a straight line passing through the origin, although the points for the two fractions of highest molecular weight appear to deviate somewhat from a line drawn through the remaining data. This deviation may be real, which would sug gest that a t higher values of molecular weight, the value of [ q ] may increase less rapidly than 111, (i.e., the value of a is less than unity in the equation [ 7 ] = K e a ) . Results of this type have been obtained by Gralen and Svedberg for cellulose in cuprammonium (5). It should be noted, however, that these fraction,cover the usual molecular aTeight range of commercial cellulose acetates. If the data are treated as though they conform with the relation [77] = KM", a value for a of 1.03 is obtained by the method of least squares. It should be noted that the value of a
zfn
0.23 conform number
TARLE IT. OSMOTICPRESSURES OF CELLULOSE ACETATE SAMPLES 1N ACETONES O L U T I O N S
Starting material
f--
5+r---
-
-
I
e FRACTIONS
Fraction
2
Fraction
3
Fraction
I
Fraction
5
Fraction
6
Fraction
7
Fraction
8
Fraction
9
Fraction 10
I
.25
I
I
015
.50
I
.75
CONCENTRATION, GA00 ML.
b'igure 1. Viscosity-Concentration Relation in Acetone Solutions for Cellulose Acetate Samples Tho samples are numbered in the order of molecular weight to couform with their designationsin a previous publication (4). Tha lowest number represents the sample of highest molecular weight.
Concn.,
Osmotic Pressure,
100 M1. 0.240 0.449 0.837 0,899 0.259 0.547 0,345 0.690 0.351 0.703 0.384 0.769 0.388 0.775 0,439 0.877 0.428 0.855 0.136 0.182 0.271 0.342 0.684 0.312 0.625 0.266 0.533 0.135 0.269 0.168 0.335 0.146 0.291
Acetone 1.96 3 71 7.19 8.03 0.67 1.47 1.10 2.37 1.37 2.83 1.71 3.59 1.82 3.98 2.49 5.33 2.44 5.27 0.85 1.15 1.75 2.42 5.14 2.59 5.44 2.68 5.49 1.34 2.67 2.41 5.15 4.52 9.75
Gram/
8nmple
-
VOl. 37, No. 5
Fraction 11 Frsction 12 Fraction 13 Fraction 14 Fraction 15
Cm.
Temp..
Apparent
26
40,000 39,000 37,000 36,000 126,000 121,000 102,000 95,000 84,000 81,000 74,000 71,000 70,000 64,000 59,000 55,000 57,000 63,000 52,000 52,000 51,000 46,000 43,000 40,000 38,000 33,000 32,000 33,000 33,000 23,000 21,000 10,400 9,700
c.
26 21 26 26 24 28 28 27 27 30 30 30 31 32 33 24 25 28 28 30 24 25 30 31 30 31 30 30 25 26 25 25
Mn
.
Mn 42,000 130.00C 110,000 86.000
78,000 76,000 82,000 61,000 53,000 48,000 42,000 34,000 33,000 24,000 11,000
,
INDUSTRIAL AND ENGINEERING CHEMISTRY
May, 1945
4
is very sensitive to the results obtained for the fractions of lower molecular weight. Moderately small errors in the results for these fractions might change the value for a appreciably. From the best straight line drawn through results shown in Figure 2, the following average relation is obtained: where
mn= 150[i~] m,, = number-average degree of polymerization
(The weight per glucose residue of cellulose acetate of this degree of acetylation is 260.) The corresponding result obtained by Staudinger (9)’who also used osmotic pressure measurements to calibrate his viscosity method, is upn = 110 [VI. It should be noted that this relation holds only for fractions of this particular degree of homogeneity. Viscosity results should properly be related to weight-average or viscosity-average molecular weights, rather than to number-average values. No estimate of T P n can be obtained by viscosity measurements made on a n unfractionated sample, or a fractionated sample of unknown degree of homogeneity.
477
than unity is entirely consistent with Kraemer’s results. However, the heterogeneity of all the samples may be somewhat different from that indicated by these ratios, since the value of Kraemer’s constant is subject to some uncertainty (6). The resulte indicate that the samples have not all been fractionated to the same degree.
TABLE111. COMPARATIVE DP, AND INTRINSIC VIscosmr RESULTS FOR CELLULOSE ACETATESAMPLES Sample Starting material Fraction 2 Fraction 3 Fraction 4 Fractjon 5 Fraction 6 Fraction 7 Fraction 8 Fraction 9 Fraction 10 Fraction 11 Fraction 12 Fraction 13 Fraction 14 Fraction 15
DFn 160 502 422 331 298 291 240 234 204 186 160 131 127 92 43
[VI
1.69 2.75 2.69 2.25 2.06 1.91 1.72 1.60 1.37 1.28 1.13 1.02 .70 59 .23
230[nl 366 633 596 518 474 439 396 368 315 294 260 234 161 136 53
2 3 1 DTn
.
2.3 1.3 1.4 1.6 1.6
1.5 1.7 1.6 1.5 1.6 1.6 1.8 1.3 1.6 1.2
From the values in Table I11 and the yields of the fractionb, summated values of [ i ~ ]and up, can be calculated. These calculated values are [ q ] = 1.50 and = 175, compared with experimental values for the starting material of 1.59 and 160, respectively. Since the results do not include fraction 1, which was not completely soluble in acetone, and since a certain amount of loss is inevitable in a long fractionation procedure, these results represent reasonably good agreement between the calculated and observed values.
m,
I.
0’
0 .
50 NUMBER-AVERAGE
I 100 MOLECULAR WEIGHT
1
x 10;)
150
LITERATURE CITED
Figure 2. Relation between Intrinsic Viscosity and Osmotically Estimated Molecular Weight for Cellulose Acetate Fractions
Kraemer measured the molecular weights of a series of secondary cellulose acetates by sedimentsation equilibrium in the ultracentrifuge, and obtained the relation (6) :
m,
= 230[q] where T P , = weight-average degree of polynierization
t
These results suggest that the cellulose acetate fractions used in this study are more homogeneous than those employed by Staudinger, but that they are not sufficiently homogeneous so that their weight- and number-average D P values are equal. This is shown by the results in Table 111. Column 2 gives the number-average DP values of the fractions and starting material as obtained from the osmotic pressure measurements. Column 3 lists the intrinsic viscosities, and column 4 the product 230 [ i ~ ]which, according to Kraemer’s resuits, should approximate DP,. Since the ratio DP,/DP, is unity for a perfectly homogeneous polymer, and increases with the heterogeneity with respect to molecular size, the value of the ratio may be taken as a measure of this heterogeneity. Because of the uncertainty involved in a e y i n g Kraemer’s constant to these samples, the ratios n , / D P , , should not be taken as absolute values. (Measurements of for several of these samples by ultracentrifugal and light-scattering methods are in progress.) The values of this ratio for the several s a m p l z a r e shown in column 5. The starting material, with a ratio D P J DPn of 2.3, is manifestly less homogeneous than any of the fractions, the ratios for which range between 1.2 and 1.8. The fact that the lowest value obtained for this ratio is slightly higher
Bingham, E. C., “Fluidity and Plasticity”, pp. 17 et mQ., NeaYork, McGraw-Hill Book Co., 1922. Dobry, A., Bull. soc. chim., [5] 2 , 1882 (1935). Flory, P. J., J. Ana. Chem. SOC.,65, 372 (1943). Fuoss, R. M., and Mead, D. J., J. Phys. Chem., 47, 59 (1943). Gralen, N., and Svedberg, Th., Nature, 152, 625 (1943). Kraemer, E. O., IND.ENG.CHBM.,30,1200 (1938). Ott, E., “Cellulose and Cellulose Derivatives”, pp. 966 et seq.. New York, Interscience Publishers, 1943. Sookne, A. M., Rutherford, H. A,, Mark, H., and Harris, M . J. Research Natl. Bur. Standards, 29, 123 (1942). Staudinger , H., “Die hochmolekularen organischen Verhi ndu ri gen”, Berlin, J. Springer, 1932. THIS paper (and the one which follows) were presented under t h e title “Mechanical Properties of Cellulose Acetate as Related t o Molecular Chair. Length”, in the Symposium on Cellulose and Cellulose Plastics before a joint session of the Divisions of Cellulose Chemistry and of Paint, Varnish, and Plastics Chemistry a t the 108th Meeting of the AMERICAN CHEwcAk SOCIETY in New York, N. Y. The authors were Research Associates at the National Bureau of Standards representing the Textile Foundation.
_ -
m,
Illuminated JIotion Display Sign of Transparent Cellulose Acetate (Courtesy, RIonsanto Chemical Company)
