Density Of Methanol + Water 2g6z57

Densities and Viscosities of Methanol-Water Mixtures S.Z. MlKHAlL and W.R. KIMEL Department of Nuclear Engineering, Kansas State University, Manhattan, Kan.

HYDROGEN

ION MOBILITY in aqueous alcoholic mixtures was investigated. Its contribution to abnormal ion mobility requires accurate values of the densities and viscosities of these media a t different temperatures between 25" and 50" C. For methanol-water mixtures, accurate density values have been determined by Clifford and Campbell (3)a t 25" C. only. Subnis, Bhagwat, and Kanugo (14) determined the densities at 30", 40", and 50" C., but their values were always very high-e.g., their densities at 30" C. were even higher than either Clifford and Cambell's or ours at 25" C. Our data agreed with those obtained by Clifford and Campbell a t 25" C., which experimental data a t 25" C. are plotted on our density-mole per cent curve in Figure 2. Density values, corresponding to their mole percentages, calculated from our empirical Equation 3 agree, within the errors shown in Table 111, with their reported experimental values. Viscosities of only a few of these mixtures are available from the International Critical Tables. The results for the densities and viscosities of these mixtures a t 25". 30", 40°, and 50" C. are reported here; also, our data are reported empirically with a power series equation of the form

In Equation 1, y denotes the density d , viscosity 7, or molal volume V, and x , the independent variable, denotes alcohol weight percentage W, alcohol mole percentage N , or temperature t" C. The constants, a,, were determined by the least squares procedure using the IBM 650 digital computer, except a. which was, in some cases, chosen to agree with accepted values of density and viscosity of water. Using Equation 1 and the computed constants included in the tables, the experimenter can calculate tables of values for density and/or viscosity for any value of the independent variable. The maximum deviation of the calculated values from Equation 1 as compared with the experimental values was always less than 0.1 and 0.6% for density and viscosity data, respectively. This deviation can be further reduced to any reasonable arbitrary minimum by adding more to Equation 1. EXPERIMENTAL

Chemically pure methanol was further purified from acetone traces, using the method recommended by Bates, Mullaly, and Hartley (1). It was then fractionated to constant boiling point through an efficient column, dried with aluminum amalgam by the McKelvey (12) method, and then boiled for 6 hours with freshly dehydrated copper sulfate while ing a current of pure air following Hartley and Raikes' method (6). The resulting absolute methanol was odorless, free of any detectable ketone, and had a density d? of 0.7869 gram per ml. compared to the International Critical Tables value of 0.78683 gram per ml., and to Clifford and Campbell's value of 0.78687 gram per ml. ( 3 ) . The water used was triply distilled from acid, alkaline, and neutral permanganate solutions and had a specific conductance of 0.8 x VOL. 6 , NO. 4, OCTOBER 1961

Results of a literature survey by Manufacturing Chemists Association (13) reveal that a majority of such measurements gave d? = 0.7865 gram per ml. However, the literature is not unanimously definite on this point. Densities obtained at higher temperatures for the same methanol sample (Table I) agree within 0.0001 with all the corresponding literature values available. Solutions were prepared by adding the calculated amount of absolute alcohol by weight to a preweighed amount of water. Correction for buoyancy was evaluated and taken into . The set of weights used was certified by the National Bureau of Standards. Density determinations were made a t each composition and a t each temperature with two pycnometers calibrated with water as a standard. Values obtained by the two pycnometers always agreed to within i.0.0001. Viscosities were determined by a modified Ostwald viscometer of the type described by Cannon and Fenske (2). The viscometer dimensions were carefully selected to give a maximum calculated kinetic energy correction of about 0.2% for the range of viscosities encountered in this work. Among several advantages of this viscometer is that it is free from surface tension corrections because practically all the efflux time is measured when the menisci of both the upper and lower levels are in tubes of equal diameter. This is very advantageous, for in the case considered, a change of over 300% in surface tension takes place between pure water and pure methanol. Loading errors never exceeded 0.002%. Errors resulting from failing to align the viscometer in the constant temperature bath in an exact vertical position were avoided by the aid of a small plumb bob made of silk and a small piece of lead, and placed in the open arm of the viscometer. The viscometer was calibrated by measuring the efflux time of pure water at 20" C. The value of 10.02 millipoises was used for the viscosity of water at 20" C. as recommended by the Bureau of Standards (15). Correction of the viscometer constants for change in temperature was eliminated entirely by preheating the samples and the viscometer to test temperature before loading (2). Both density and viscosity measurements were carried out in a constant temperature bath controlled by an electronic unit capable of keeping the temperature constant to within & 0.005" C. Temperature measurements were made by very sensitive Bureau of Standards calibrated thermometers.

*

RESULTS

Tables I and I1 summarize the experimental results obtained. Empirical equations relating the different parameters follow. Density of the mixtures, d:, as a function of weight percentage methanol, W. 5

cG ( g . / d . in vacuo) =

C a, W

(2)

I-0

In this equation, a. was chosen to agree with the density values of water given in the International Critical Tables (8). Values of a. through as evaluated a t the five different

533

Table I. Densities, db, and Molal Volumes, V,of Methanol-Water Mixtures a t Different Temperatures

30" c.

25" C. Wt. % Methanol

Mole Yo" Methanol

d, g./ml.

Vol., ml.

0 0.9971 18.068 5.88 0.9802 19.221 12.33 0.9647 20.468 19.43 0.9492 21.852 27.27 0.9315 23.447 25.293 35.99 0.9119 0.8904 27.442 45.76 29.951 56.75 0.8673 69.22 0.8426 32.905 36.468 83.50 0.8152 91.44 0.8006 38.524 40.720 100 0.7869 ' Molecular weight = 32.043. ' Reference (8). Ob 10 20 30 40 50 60 70 80 90 95 100

35" c

d, g./ml.

Vol.,

0.9957 0.9786 0.9628 0.9467 0.9285 0.9096 0.8870 0.8637 0.8381 0.8110 0.7962 0.7819

18.094 19.253 20.508 21.909 23.523 25.357 27.548 30.076 33.081 36.657 38.737 40.981

ml.

working temperatures are recorded in Table 111. The maximum percentage error given in column 8 is: max. ( d M - d d ) 100/d&. The mean percentage error given in column 9 is:

d, g./ml. 0.9941 0.9768 0.9606 0.9440 0.9254 0.9060 0.8835 0.8595 0.8333 0.8057 0.7911 0.7771

40" c. d, g./ml. 0.9923 0.9747 0.9582 0.9410 0.9220 0.9022 0.8795 0.8553 0.8291 0.8014 0.7864 0.7723

Vol., ml. 18.123 19.288 20.555 21.972 23.602 25.457 27.657 30.223 33.272 36.898 38.987 41.234

50" C.

Vol., ml. 18.156 19.330 20.607 22.042 23.689 25.565 27.783 30.371 33.440 37.096 39.220 41.490

Vol., ml.

18.233 19.424 20.728 22.174 23.876 25.787 28.057 30.687 33.795 37.498 39.669 42.013

Absolute viscosity of mixtures, 7, as a function of weight percentage methanol, W. 5

7

(Millipoise) = 4

=o

a, W

In this equation, a0 was chosen to agree with the values

, = I

where rn denotes the number of experimental values obtained between pure alcohol and pure water a t one and the same temperature. Figure 1 graphically illustrates the closeness of fit between Equation 2 and typical experimental density data. A smaller number of in Equation 2 was inadequate for precise data representation. Density of mixtures, d:, as a function of mole percentage methanol, N.

c a, N' 5

di, (g./ml. in vacuo) =

I

(3)

=a

In this equation a0 is the same as an of Equation 2. Constants an through a5 are given in Table 111. Figure 2 graphically illustrates the closeness of fit between Equation 3 and typical experimental density data. Molal volumes of mixtures, V, as functions of mole fraction methanol, N.

v (ml.) =

5

a, N'

(4)

,=O

In this equation an was chosen to agree with the molal volumes of water computed from density values of water given above. Equation 4 can also be used to derive the expressions of the partial molal volumes ( 1 1 ) as follows: Methyl partial molal volume urn = V + (1 - nr) d V / d N

(5)

Water partial volume v , = V - N d V / d N

(6)

Temperature dependence of density, d:, between f = 25O and 50' C.

for

mixtures,

4

d: =

,=o

a, t'

(7)

An equation of the fourth degree was found to give the same order of accuracy as Equations 2, 3, and 4 in this particular case. Constants a. through a4 were determined for different methanol percentages by weight and recorded in Table IV. 534

d, g./ml. 0.9881 0.9700 0.9526 0.9354 0.9148 0.8944 0.8709 0.8465 0.8204 0.7928 0.7775 0.7627

Table II. Viscosities of Methanol-Water Mixtures ai Different Temperatures

Viscosity, Millipoises Wt. 'X Mole % Methanol Methanol" 25°C. 30°C. 35°C. 40°C. 50°C. 0' 0 8.95 8.00 7.21 6.54 5.49 10 5.88 11.55 10.15 8.94 8.00 6.56 13.92 12.06 10.62 20 12.33 9.41 7.59 30 19.43 15.40 13.36 11.72 10.34 8.23 40 27.27 15.95 13.81 12.13 10.74 8.54 50 35.99 15.34 13.37 11.80 10.46 8.41 60 45.76 14.01 12.39 10.98 9.81 7.95 70 56.75 12.18 10.85 9.74 8.76 7.22 80 69.22 10.05 9.07 8.29 7.52 6.30 90 83.50 7.82 7.14 6.67 6.12 5.22 95 91.44 6.59 6.12 5.66 5.32 4.58 5.41 5.07 100 100.00 4.74 4.50 3.96 a Molecular weight = 32.043. 'Values of water viscosity are reported by Dorsey (4). Since Dorsey's value for viscosity of water at 20" C. is 10.09 millipoises, correction of all his values to the new National Bureau of Standards basis of 10.02 millipoises was given some consideration. The following are pertinent points: Dorsey's values were obtained "from a consideration of all pertinent data available" (4, p. 182). He adds, "They are the results of an entirely independent study of the recorded data, and involve many complete recomputations." This led us to believe that his values for 30", 35", 40", and 50" C. are not necessarily based on his 10.09-millipoisevalue for 20" C. Comparison between Dorsey's values and those of Bingham and Jackson (4, p. 184) reveals that Dorsey's values at 30", 35", 40", and 50" C. are lower than the corresponding Bingham and Jackson's values, although the latter values are based on 10.05 millipoises for water a t 20" C. Hardy and Cottington [ J . Research Natl. Bur. Standards 42, 573 (1949)] reported a value of 6.551 millipoises for water at 40" C. based on 10.05 millipoises for water at 20" C. Correction of this value to 10.02 basis gives 6.531 millipoises compared to Dorsey's value of 6.536 millipoises without correction. In view of this lack of agreement, it was decided to keep Dorsey's values as they are. Whether Dorsey's values will, in the light of future investigations, need to be corrected or not, the viscosity values recorded in Table I1 will not be affected because those are based on 10.02 millipoises for water at 20" C.

JOURNAL OF CHEMICAL AND ENGINEERING DATA

Table Ill. Computed Constants of Various Equations at Different Temperatures

Max.% Mean.% Error Error

Temp.,

c.

25 30 35 40 50 25 30 35 40 50

al

a0

a2

0.9971 0.9957 0.9941 0.9923 0.9881

-1.8527 -1.8577 -1.8278 -1.8409 -1.9417

x

0.9971 0.9957 0.9941 0.9923 0.9881

-2.8930 -2.9883 -3.0331 -3.0601 -3.1257

x x x x x

X

x x x

Equation 2: Density, 2.2013 x +1.9615 x 10-j 1.3727 x 1.2424 x 1.7956 x

a5

a4

a3

d: in G./ MI., us. Weight % Methanol, W -1.8012 x lo-” -6.0418 x lo-’ +5.3912 x lo-’ 4.4949 x lo-’ -1.4274 X lo-’’ -5.3092 x 2.3168 x lo-’ -4.9507 x lo-’* -3.5368 x lo-’ 2.4143 X lo-’ -6.1299 x -3.4028 x lo-’ 5.4346 x lo-’ -1.9714 x l o - ” -5.5712 x lo-’

Equation 3: Density, diin G./Ml., us. Mole % Methanol,” N 5.9438 x lo-’ 2.9907 x -6.0876 x lo-’ 7.5625 x lo-’ -7.4395 x lo-’ 3.4526 x 6.7625 x lo-’ 3.3232 x -6.8854 x lo-’ 5.6578 x lo-’ -5.8571 x lo-’ 2.9977 x 5.5450 x lo-’ -5.3337 x lo-’ 2.7332 x

x x x x x

-2.0581 -2.7269 -2.3385 -1.9397 -2.0748

-0.05 +0.05 -0.04 -0.03 -0.06

0.02 0.01 0.01 0.01 0.02

-0.08 -0.07 -0.05 -0.06 +0.09

0.04 0.03 0.02 0.03 0.04

+0.08 +0.05 -0.04 +0.05 -0.08

0.04 0.03 0.02 0.02 0.04

+0.27 -0.33 -0.19 -0.32 +0.33

0.10 0.10 0.08 0.11 0.10

+1.80 +1.41 +1.96 +1.66 +1.34

0.72 0.56 0.78 0.61 0.47

lo-” lo-’’ lo-” lo-” lo-”

Equation 4: Molal Volumes, in Ml., us. Mole Fraction of Methanol,”N

a

19.324 19.623 19.756 19.791 19.951

25 30 35 40 50

18.068 18.094 18.123 18.156 18.233

25 30 35 40 50

8.95 8.00 7.21 6.54 5.49

0.24099 0.20233 0.15417 0.13139 0.10196

25 30 35 40 50

8.95 8.00 7.21 6.54 5.49

0.54929 0.44634 0.36936 0.30680 0.22387

-7.6821 2.0494 -0.39832 9.3660 -14.873 5.0875 15.192 -2.1367 -12.394 3.7326 -1.7467 13.767 2.4280 10.244 -8.7175 -0.40699 3.2833 8.8399 -9.1047 0.81614 Equation 8: Viscosity, 7 in Millipoises, us. Weight 70Methanol, W 3.8479 x -2.1872 x 2.3409 x -8.1493 x lo-’ 1.6602 X 10 -5.6258 X lo-’ 2.6359 x -1.5929 X lo-‘ 1.8071 X 10 -6.4851 X 10 4 3.4838 x -1.6859 x lo-‘ 1.3872 x -4.8712 x lo-’ 2.7079 x -1.3226 x 1.3259 x -7.6409 x 10-j 7.8579 x lo-’ -2.7156 x lo-’ Equation 9: Viscosity, 7 , in Millipoises, us. Mole So Methanol,” N

-‘ -‘

-1.2468 -9.7573 -7.7954 -6.1310 -4.6508

4.0795 x lo-’ 2.5606 x 1.4103 x IO-’ 1.4968 X lo-‘ 1.1518 x

x x x X

x

7.2924 x 6.3434 X 5.9862 x 6.0250 X 2.8063 x

lo-’ lo-‘ lo-’ lo-’

lo-’

-4.7553 x lo-’ -3.9059 X lo-’ -3.5458 X lo-’ -3.3182 X lo-’ -1.7008 X lo-’

Molecular weight = 32.043.

ss

. I

2

9c

rn

L .

* t z

Y

0!

0c

7!

t

0 10

20 30 PERCENTAGE

40 METHYL

50 60 70 00 A L C O H O L BY W E I G H T , W

90

20

I

,

30

40

\ 1

1

1

1

50

60

70

80

4

90

100

Figure 1. Comparison of experimental and computed data for density, d i vs. percentage methanol by weight, W

VOL. 6, MO. 4, OCTOBER 1961

I

IO

MOLE % METHANOL

Figure 2. Comparison of experimental and computed data for density, dfi, vs. mole per cent methanol

535

Table IV. Computed Constants for Different Weight Percentages of Methanol

Wt. 70 Methanol 0 10 20 30 40 50 60 70 80 90 100 0 10 20 30 40 50 60 70 80 90 100

a0

1.1648 1.1117 8.8852 8.5160 9.6817 9.6206 9.9612 9.3393 8.9900 9.9865 8.6237

a1

lo-' x lo-' x lo-' x lo-' X lo-' x lo-' x lo-' x lo-' x lo-' X

25.503 29.880 19.832 24.914 27.385 23.126 30.371 16.780 41.483 X lo-' 55.288 x lo-' 16.429

d: in G./Ml., us. Temperature, to C. X lo-' -1.5558 X 1.0631 X lo-' x lo-' -1.1955 X 8.2093 X lo-' x 8.2770 x -5.7565 x lo-' X lo-' 1.0555 x lo-' -7.2290 X lo-' -2.7738 X 1.9733 x lo-' x -3.2283 X lo-' 2.2369 X lo-' x -7.6195 x 5.0549 X lo-' x lo-' -4.2553 x lo-' 2.8876 X lo-' x -2.8846 X lo-' 1.9575 X lo-: X -1.5900 X 1.1127 X l o - ' X lo-' -5.0440 X 3.4857 X lo-' x Equation 10: Viscosity, 7 , in Millipoises, us. Temperature, t" C. -1.4464 4.8228 x lo-' -8.0351 X lo-' 5.1496 X -1.4226 4.1265 x lo-* -6.5175 X lo-' 4.1783 X 3.0041 x l o - ' -4.1435 x lo-' 9.8528 X lo-' -7.4939 X -5.5901 x lo-' -2.7324 x lo-' 7.1099 X lo-' -5.5129 X -1.8143 X lo-' -2.5714 X lo-' 7.3737 x lo-' -6.0257 X 1.0769 x lo-' -3.3298 x lo-' 8.1673 X lo-' -6.2318 x -1.1161 2.6533 x lo-' -3.8134 X 2.3553 X lo-' 1.0788 x lo-' -2.2029 x lo-' 5.0683 X lo-' -3.7180 X 1.1447 -6.1256 x lo-' 1.2072 X -8.4552 X 1.8792 -8.9590 X lo-' 1.7346 x -1.2170 x -1.0987 4.1657 x lo-' -7.3310 x 4.7938 x

-1.9364 -1.4933 9.7838 1.2585 -3.4993 -4.5013 -1.0699 -6.1523 -4.7077 -1.9995 -7.1042

Equation7: Density, lo-' 8.2879 x lo-' 6.3450 x -4.3998 x lo-' -5.6522 x 1.3800 x 1.6846 x lo-' 4.1849 x 2.2865 X 1.5801 X lo-' 8.3439 x 2.6727 X

of the viscosity of water from Dorsey's selection ( 4 ) . Constants a0 through a5 are given in Table 111. Figure 3 illustrates the closeness of fit between Equation 8 and typical experimental viscosity data. Absolute viscosity of mixtures, 7, as a function of mole percentage methanol, N. 5

7

Max. % Error

a4

a3

a2

(Millipoises) = I

=o

a, N'

(9)

In this equation, a, is the same as a0 of Equation 8. Viscosity values computed from this equation had a maximum error of 1.96% and a maximum mean error of 0.78% compared to the corresponding values of 0.3 and 0.11% obtained from Equation 8 using the constants given in Table 111. The difference in maximum error in these two cases is a natural result of the least squares

Mean Yo Error

-0.05 -0.01 +0.01 +0.02 0.00 0.00 0.00 -0.01 -0.02 -0.07 -0.01

0.02 0.01 0.01 0.01 0.00 0.00 0.00 0.00 0.01 0.03 0.00

+0.25 -0.30 +0.31 +0.19 +0.19 +0.25 -0.13 +0.18 +0.54 -0.54 -0.57

0.14 0.15 0.15 0.09 0.09 0.12 0.06 0.09 0.27 0.50 0.28

analysis, since the data in the two cases are different and the number of parameters is identical. The closeness of fit in such analyses of experimental data, with a given number of parameters generally becomes better as the number of inflection points in the data decreases. Inspection of Figure 4 reveals one additional inflection point compared with Figure 2; therefore, for our six parameter least squares fits, the closeness of fit is expected to be better in the relation represented by Equation 8 than that represented by Equation 9. Readers interested in computing viscosity values for solution mixtures of known mole percentages are therefore advised to calculate the corresponding weight percentages and use Equation 8, rather than using Equation 9 directly unless i 1.96% maximum error is tolerable. Temperature dependence of viscosity, 7, between t = 25' and 50' C.

for

mixtures,

IS

13

F: w

Figure 3. Comparison of experimental and computed data for viscosity, 1,vs. percentage methanol by weight, W

L"

I

0

7

S

3~

10

20

40

PERCENTAGE

536

50 METHYL

60

ALCOHOL EY

10 WEIGHT.

W

JOURNAL OF CHEMICAL AND ENGINEERING DATA

Figure 4. Viscosities of water-methanol mixtures, 7,vs. mole per cent methanol

I

I

I

7

I

(Millipoises) =

I

,=o

I

1

a, t’

(10)

This empirical representation by a power series is more precise, within this range of temperature, than that predicted by the theoretically based equations suggested by Gutmann and Simmons ( 5 ) for viscosity of liquids in general or by Kampmeyer (IO)and by Innes (7) for water and mercury. It is also more precise than that predicted by the equation:

used in the International Critical Tables (9) for methanol. A simple theoretical temperature dependence equation for viscosity of mixtures of liquids has not yet been derived. LITERATURE CITED (1) Bates, H.H., Mullaly, J.M., Hartley, H., J.. Chem. SOC.123, 401 (1923).

Cannon, M.R., Fenske, M.R., I d . Eng. Chem., A d . Ed. 10, 297 (1938). Clifford, G., Campell, J.A., J . Am. Chem. SOC.73, 5449 (1951). Dorsey, E.N., “Properties of Ordinary Water Substances,” pp. 183-5, Reinhold, New York, 1940. Gutmann, R., Simmons, L.M., J . Appl. Phys. 23, 977 (1952). Hartley, H., Raikes, H.R., J . Chem. SOC.127, 525 (1925). Innes, K.K., J . Phys. Chem. 60.817 (1956). International Critical Tables, vol. 111, p. 25, McGraw-Hill, New York, 1928. Ibid., vol. V , p. 11, 1929. Kampmeyer, P.M., J . Appl. Phys. 23,99 (1952). Lewis, G.N., Randall, M., “Thermodynamics and the Free Energy of Chemical Substances,” p. 33, McGraw-Hill, New York, 1923. McKelvey, E.C., Natl. Bur. Standards, Bull. 9, 364 (1913). Manufacturing Chemists Association, Selected Values o f Properties of Chemical Compounds, MCA Research Project. Washington, D. C., 1960. Subnis, S.W., Bhagwat, W.V., Kanugo, R.B., J . Indian Chem. SOC.25, 575 (1948). Swindells, J.F., Coe, J.R., Jr., Godfrey, T.B., J. Research Nutl. Bur. Standards 48, 1 (1952). RECEIVED for review November 10, 1960. Accepted June 13, 1961.

CORRECTION

In “Critical Properties of Mixtures of Normal Paraffin Hydrocarbons” [Doyle 0. Etter and W.B. Kay, J. CHEM. ENG.DATA6, 409 (1961)], the legends for Figures 2 and 3 should be reversed. Equation 5 should read:

VOL. 6 , NO. 4, OCTOBER 1961

537