VLE/batch distillation

set of calculated data have been filed with the ACS Mi- crofilm Depository Service.

Nomemclature

GE = excess Gibbs free energy, cal/g-mol x1 = mole fraction of the more volatile component of a

y1 = mole fraction of the more volatile component of a

y1,yz = activity coefficients for components 1 and 2 in

q h , $ ~ = coefficient of correction for nonideality of the

T = total pressure, m m Hg

binary mixture in the liquid phase

binary mixture in the vapor phase

the liquid phase

vapor phase for components 1 and 2

Literature Cited

(1) Minh, D. C.. MS thesis, Sherbrooke University, Sherbrooke. Que., Canada (1970).

(2) Minh, D. C., Ruel, M., Can. J . Chem. Eng.. 4 8 , 501 (1970). (3) Minh, D. C.. Ruel, M.,ibid., 49, 159 (1971). ( 4 ) Prengle. H . W., Palm, G. F., ind. Eng. Chem., 49, 1769 (1957). (5) Ramalho, R . , Delmas, J.. Can. J . Chem. Eng., 46, 32 (1968). (6) Ramalho, R., Delmas, J., J . Chem. Eng. Data, 1 3 , 161 (1968). (7) Selected Values of Properties of Chemical Compounds. Thermody-

namics Research Center, Texas ABM University, College Station, Tex., 1970.

Received for review January 19. 1971. Resubmitted February 10, 1972. Accepted October 26, 1972. Grants from the National Research Council of Canada are gratefully acknowledged. Additional data will appear fol- lowing these pages in the microfilm edition of this volume of the Journal. Single copies may be obtained from the Business Operations Office, Books and Journals Division. American Chemical Society, 1155 Sixteenth St., N.W., Washington, D.C. 20036. Refer to the following code number: JCED-73-41. Remit by check or money order $7.00 for photocopy or $2.00 for microfiche.

Vapor-Liquid Equilibria in Mixtures of Water, n-Propanol, and n-Butanol

Richard A. Dawe,’ David M. T. Newsham,2 and Soon Bee Ng Deoartment of Chemical Enoineerino, Universitv of Manchester Institute of Science and Technology. Manchester, M60 lob, U . K .

Measurements of vapor-liquid equilibria in the systems water-n-propanol and water-n-propanol-n-butanol at atmospheric pressure are reported. The results for the ternary system have been compared witb those predicted from the binary mixtures using the equation of Renon and Prausnitz. Agreement is very satisfactory. Densities and refractive indices of the binary and ternary mixtures at 25°C are also reported and the excess volumes of binary mixing have been calculated.

This paper reports the results of measurements of vapor-liquid equilibria for the systems water-n-propanol and water-n-propanol-n-butanol at atmospheric pres- sure. Enthalpies of mixing for these systems have pre- viously been reported (5, 6) and liquid-liquid equilibrium studies have also been published ( 7 0 ) .

Vapor-liquid equilibria in the system water-n-propanol have been extensively investigated ( 7 ) and this system is therefore suitable for testing the performance of the equi- librium still used in the present work. No measurements of vapor-liquid equilibria in the system water-n-propanol- n-butanol have previously been reported.

Experimental Materials used in this investigation were purified as de-

scribed in ref. 5 and 6. For the measurements of vapor-liquid equilibria, an

equilibrium flow still similar to that described by Vilim et at. (74) was constructed. This instrument has the advan- tage that it avoids recirculation of the condensed vapor, which enables reasonably precise results to be obtained in a short time (15 min). The still used in this work dif-

fered from that of Vilim et al. in that the thermometer well was designed to accommodate a 50-ohm capsule- type platinum resistance thermometer (Rosemount Engi- neering Co. Ltd.) and the vacuum jacket that insulated the equilibrium chamber was extended to include the droplet separator (Figure 1 ) in order to minimize thermal

I

‘I 1 I

’Present address, Department of Chemical Engineering, University of Leeds, Leeds, U.K.

To whom correspondence should be addressed.

Figure 1. Equilibrium chamber of the flow still A, thermometer weli; B, vacuum jacket; C, droplet separator; D, to vapor condenser; E, 3-mm capillary; F, to liquid cooler; G, Cottrell pump; H, boiler; J. float-valve; K , reservoir

44 Journal of Chemical a n d Engineering D a t a , Vol. 18, No. 1, 1973

losses. The platinum thermometer was calibrated by intercomparison with a 25-Ohm platinum thermometer that had been calibrated at the National Physical Labora- tory (U.K.) on the International Practical Temperature Scale of 1948. Resistances were measured by potentio- metric comparison with a 100-ohm standard resistor on a vernier potentiometer (H. Tinsley and Co. Ltd., U.K.). Temperatures could be measured to within f0.02"C with the platinum thermometer. No attempt was made to con- trol the pressure in the equilibrium still, but the pressure in the laboratory was measured by a mercury manometer and cathetometer with an accuracy of f 0 . 0 5 mm Hg. All pressures were corrected to give the equivalent height of a mercury column at 0°C and standard gravity. For this purpose a value of the local acceleration due to gravity of 981.370 f 0.001 c m s e c - 2 was used.

Preliminary measurements in which pure water was boiled in the still showed that thermal equilibrium could be attained in it to within 0.02"C. Subsequent experi- ments with water and n-propanol mixtures indicated also that transfer of material between vapor and liquid phases was sufficiently rapid for a reasonable approach to equi- librium conditions to be made. The ratio of liquid to vapor flow rates was, typically, about 10. The droplet separator functioned satisfactorily throughout the measurements and no indications of entrainment in either vapor or liquid phases were observed. The liquid samples were always used in the freshly distilled state since they were contin- ually recycled through the still.

Compositions of the binary liquid mixtures were deter- mined by measurement of their refractive indices, and those of the ternary mixtures were deduced from density and refractive index measurements. A dipping refractom- eter that was equipped with thermoprisms (Carl Zeiss, Jena) and which had a measuring precision of f 2 X

was used. The temperature of the prisms was con- trolled at 25.00" f 0.05"C by circulating water from a thermostat. Illumination was provided by a sodium lamp. Densities were measured at 25°C using 5-cm3 Lipkin pycnometers that had been calibrated with distilled water. The density of water at 25°C was taken to be 0.99705 gram ~ m - ~ . The precision of the density mea- surements was estimated to be f0.0002 gram ~ m - ~ . Calibration curves for composition as a function of densi-

l o o \P,.,* /

6 0

w*

4 0

2 0

0 20 40 6 0 80

Wl

Figure 2. Curves of constant density and refractive index for water (1)-n-propanol (2)-n-butanol (3) at 25°C

ty and refractive index were prepared by measuring these properties for samples of known composition. For esti- mating ternary compositions it was necessary to prepare curves of constant density and refractive index as a func- tion of composition. This was achieved by interpolating the primary results graphically to give the curves of Fig- ure 2.

The estimated errors in the measured mole fractions are different for each component. They are f 0 . 0 0 5 for water and fO.O1-0.05 for n-propanol and n-butanol. The reason for this is apparent on examination of Figure 2 which shows that the curves of constant density are al- most vertical straight lines. This means that the percent- age of water ( Wl) can always be accurately estimated because it is virtually independent of the angle at which the curves of constant density and refractive index inter- sect. The percentage of propanol ( WZ) is more difficult to determine particularly in the region 40 < W t < 60% where the density and refractive index curves are almost parallel to each other. The butanol composition (W3) is also subject to considerable uncertainty in this region. For values of W1 < 40%, however, W1 and W3 can be determined to within about 1%. Th'ese estimates of the uncertainties are substantiated in the comparison of ex- perimental and calculated vapor compositions made later.

Results and Discussion

The densities and refractive indices for the three binary systems and the ternary system at 25°C are recorded in Table I. Table I I gives densities and refractive indices measured at 25.9"C for mixtures with compositions lying on the liquid-liquid phase boundary at 25°C.

The excess volumes of mixing for the three binary sys- tems have been computed from the densities and are re- corded in Figure 3. The vertical error bars for the n-pro- panol-n-butanol system represent an uncertainty in the density measurements of f 0 . 0 0 0 2 gram ~ m - ~ . The ap- parent asymmetry is probably not real. The other excess thermodynamic functions are not asymmetric (6).

Vapor-liquid equilibrium results for the systems water- n-propanol and water-n-propanol-n-butanol are present- ed in Tables I l l and I V , respectively. References to vapor-liquid equilibrium data for water-n-propanol may be

/ / -0 6 I

x

Figure 3. Excess volumes of mixing of water-n-propanol (0), water-n-butanol ( A ) , and n-propanol-n-butanol ( 0 ) at 25°C; x is the mole fraction of the second component

Journal of Chemical and Engineering Data, Vol. 18, No. 1, 1973 45

Table I. Densities and Refractive Indices of Solutions of Water, Table 11. Refractive Indices and Densities of Mixtures of Water, n-Propanol, and n-Butanol at 25°C n-Propanol, and n-Butanol at 25.9'C and at

Compositions Lying on the Binodal Curve at 25°C p l g r a y

7 W l W2 7 P I S cm-3 w1 wz w3 cm - 100

0 0 2.51 6.31

10.75 16.51 22.69 30.96 41.08 55.05 72.93 0 0 0 0 0 0 0 0 0 2.32 5.27 9.06

13.55 18.89 93.85 94.64 96.96 98.37

5.5 7.2 7.4 7.4 7.4 7.9 9.0 9.9

15.2 17.1 19.4 20.3 22.5 29.2 31.1 33.0 42.9 45.6 57.9 60.0 77.8

0 100

0 97.49 93.69 89.25 83.49 77.31 69.04 58.92 44.95 27.07 90.84 80.38 70.10 59.99 49.64 39.84 30.29 19.90 9.64 0 0 0 0 0 0 0 0 0

84.4 21.9 41 .O 51.9 89.3 71.5

7.6 29.9 12.6 74.0 58.1 44.6 25.9 37.0 49.5 59.8 35.6 36.5 30.1 35.9 19.5

0 0

100 0 0 0 0 0 0 0 0 0 9.16

19.62 29.90 40.01 50.36 60.16 69.71 80.10 90.36 97.68 94.73 90.94 86.45 81.11

6.15 5.36 3.04 1.63

10.1 70.9 51.6 40.7

3.3 20.6 83.4 60.2 72.2 8.9

22.5 35.1 51.6 33.8 19.4

7.2 21.5 17.8 12.0 4.1 2.7

0.99705 0.80034 0.80594 0.80737 0.81599 0.82537 0.83743 0.85012 0.86652 0.88751 0.91 688 0.95394 0.8008 1 0.80163 0.80217 0.80266 0.80347 0.80464 0.80514 0.80590 0.80696 0.81084 0.81657 0.82368 0.83202 0.8421 1 0.98788 0.98903 0.99210 0.99443 0.81373 0.81933 0.81906 0.81886 0.81780 0.81914 0.82309 0.82447 0.83496 0.83959 0.84291 0.84422 0.84881 0.86167 0.86586 0.87023 0.88883 0.89464 0.91975 0.92556 0.96196

1.33252 1.38325 1.39727 1.3831 1 1.38264 1.38182 1.38022 1.37835 1.37533 1.371 19 1.36460 1.35466 1.38446 1.38596 1.38741 1.38883 1.39024 1.39158 1.39293 1.39442 1.39573 1.39683 1.39581 1.39429 1.39227 1.38959 1.33851 1.33782 1.33539 1.33405 1.38424 1.39216 1.38950

1.38300 1.3851 7 1.39342 1.39004 1.38995 1.38129 1.38225 1.38352 1.38476 1.38000 1.37758 1.37546 1.37251 1.37090 1.36398 1.36230 1.35149

i 3 3 8 0 1

21.6 7.7 1.38609 0.84643 31.4 23.4 1.37931 0.86490 40.5 27.2 1.37395 . . . 55.6 25.2 1.36465 . . . 72.5 16.4 1.35290 . . . 77.8 13.6 1.35085 0.95925 80.8 11.9 1.34870 0.96558

Table Ill. Vapor-Liquid Equilibrium Data for Water (1)- n-Propanol (2)

P, m m H g t , O C x 2 Y2 71 Y2 '

756.33 756.73 756.73 756.73 756.07 755.86 757.97 757.97 757.97 758.08 758.08 758.08 756.33 759.13 759.13 759.13

98.18 96.47 93.91 91.92 88.57 88.04 87.72 87.62 87.54 87.53 87.62 88.97 89.40 91.69 93.07 94.87

0.0031 0.0068 0.01 30 0.0218 0.072 0.139 0.275 0.327 0.380 0.396 0.515 0.721 0.766 0.867 0.91 3 0.955

0.071 0.131 0.215 0.288 0.375 0.393 0.407 0.415 0.424 0.421 0.456 0.561 0.596 0.710 0.785 0.878

0.989 0.988 0.985 0.970 1.015 1.083 1.276 1.361 1.459 1.507 1.756 2.344 2.525 2.939 3.1 66 3.252

21.8 19.6 18.5 15.9

7.126 3.948 2.098 1.806 1.593 1.519 1.260 1.051 1.032 0.998 0.994 0.994

where difficulties are encountered in qperating an equilibri- um still because of the rapid change of boiling point with composition in this region. The boiling points of pure water and n-propanol obtained by extrapolation of the t - x diagram, after correction to 7 6 0 m m Hg, were 100.03" and 9 7 . 1 3 " C compared to values of 100.00" and 9 7 . 0 7 " C calculated from the Antoine equations of ref. 8. The esti- mated error on the extrapolated values of the boiling points is 0.1 "C.

The validity of the results may also be checked by applying thermodynamic consistency tests to the liquid phase activity coefficients. The latter were calculated from the following equation which includes corrections for vapor phase nonideality:

y , = ( y , P / x , P P j exp [ ( B a t - V d ( P - P , " ) / R T ] (1) This equation assumes that the vapor phase behaves as an ideal solution. The reference state for the activity coefficients is the pure liauid at the temperature and total pressure of the soiution. The second virial coefficients for found in the compilation Of Hala et ( 7 ) ' For the pur- water and n-propanol were taken from the work of Collins pose of evaluating the performance of the equilibrium still

used in the present investigation we have compared the data selected by Hala et al. (8) with our results in Figure

and Keyes (') and ( 2 ) 7 respectively~ and the pure component vapor pressures were calculated from the An- toine equations given in ref. 8. The maximum value of the vapor phase correction was 1 . 5 % . The consistency test was applied by writing the Gibbs-Duhem equation in the form

4 which is a plot of vapor phase composition against liq- uid phase composition. The average deviation of the vapor compositions determined in this work from the smooth curve of Figure 4 is 0.001. We find, for the azeotropic composition, y2 = x z = 0.433 compared to values of 0.432 and 0.426 given by Doroshevsky and Po- lansky (3) and Murti and Van Winkle ( 9 ) . At other com- positions the agreement is generally good; the largest de- viations (up to 0.04) occur at low alcohol concentrations

1n(YI/Y2)dxl - l:: j$ ( g ) , d x : - 0 ( 2 ) Figure 5 is a plot of In 73/72 against xl. The area ratio I A , / A 2 1 obtained by Simpson's rule integration is 0 . 9 7 5 .

t - 0

46 Journal of Chemical and Engineering Data, Vol. 18, No. 1, 1973

Table IV. Vapor-Liquid Equillbrium Data for Wator (1) and n-Propanol (2)-n-Butanol (3) ~~ ~~

Exptl Calcd Eq 7

X 1 x 2 Y1 Yz Y1 Yz AYi AYZ AYJ P, rnrntig t , " C

760.58 760.56 760.47 760.56 760.64 765.92 761.67 761.84 762.90 765.71 765.68 762.92 761.84 765.72 765.59 762.56 765.47

100.21 96.37 94.88 93.44 92.20 91.73 91.82 92.76 90.22 88.73 90.06 88.87 89.93 89.74 88.43 89.27 88.91

0.141 0.153 0.155 0.166 0.168 0.176 0.372 0.374 0.377 0.406 0.490 0.572 0.589 0.623 0.689 0.749 0.857

0.197 0.409 0.517 0.61 1 0.697 0.748 0.233 0.144 0.503 0.539 0.303 0.338 0.226 0.239 0.307 0.210 0.129

0.432 0.382 0.367 0.352 0.344 0.354 0.602 0.633 0.542 0.535 0.616 0.607 0.649 0.646 0.607 0.639 0.635

0.170 0.421 0.492 0.571 0.61 5 0.618 0.247 0.142 0.401 0.436 0.288 0.338 0.243 0.261 0.359 0.307 0.318

0.439 0.398 0.371 0.362 0.348 0.345 0.613 0.646 0.530 0.528 0.622 0.609 0.660 0.651 0.591 0.630 0.625

0.216 0.408 0.489 0.547 0.604 0.623 0.207 0.133 0.422 0.454 0.279 0.343 0.235 0.265 0.405 0.330 0.340

-0.007 -0.016 -0.004 -0.010 -0.004 +0.009 -0.01 1 -0.013 +0.012 +0.007 -0.006 -0.002 -0.011 -0.005 4-0.016 +0.009 +0.010

-0.046 4-0.013 +0.003 4-0.024 +0.011 -0.005 +0.040 +0.009 -0.021 -0.018 +0.009 -0.005 4-0.008 -0.004 -0.046 -0.023 -0.022

Average 0.009 0.019

+0.053 +0.003 +0.001 -0.014 -0.007 -0.004 -0.029 +0.004 +0.009 +0.011 -0.003 +0.007 +0.003 +0.009 +0.030 4-0.014 +0.012

0.013

1 ---

1

0 0.2 0.4 0.6 OE 1.0

x2

Flgure 4. Equilibrium vapor and liquid compositions for water- n-propanol at atmospheric pressure 0 this work, A Doroshevsky and Polansky ( 3 ) . Murti and Van Winkle (9).

This is satisfactory since the second term of Equation 2 has been neglected. However* this integral may be evalu- ated using the enthalpies of mixing of Plewes et al. ( 1 1 )

Figure 5. Plot of In ( y ~ / y * ) against XI for water (1)-n-propanol (2) at atmospheric DreSSure . .

and from values of ( a T / b x l ) p obtained by graphical dif- ferentiation of the t - x curve. When this is done the area ratio is increased to a value of 0.995. The corrected Prediction of Ternary Phase Equilibria curve cannot be shown with clarity in Figure 5 but the main differences occur for values of X I 0.2 and > 0.9 because of the high values of ( d T / a x l ) p in these regions. The closeness of the area ratio to unity confirms the thermodynamic consistency of our results.

Figure 5 also indicates the high precision of the results for the water-n-propanol system, as does Figure 6 which is a plot of the excess Gibbs energy against x 2 .

The results of the measurements on the ternary system are collected together in Table I V , and are discussed below.

Jouri w

Recently, Renon and Prausnitz (72) have proposed a method of evaluating the thermodynamic properties of multicomponent mixtures from a knowledge only of the properties of the appropriate binary pairs. The method is particularly suitable for partially miscible systems. We have therefore applied it to the system water-n-propanol- n-butanol. According to Renon and Prausnitz the excess Gibbs free energy of a binary liquid mixture is given by the following Equation:

G"RT - x l x ~ [ r 2 1 G 2 1 / ( x I + x&)+ rl2GI2/(x2 + x , ~ , , ) ] ( 3 ) 11 of Chemical and Engineering Data, Vol. 18, No. 1, 1973 47

Table V. Parameters of the Equation of Renon and Prausnitr for Binary Systems of Water ( l ) , n-Propanol (2), and n-Butanol (3)

Figure 6. Plot of the excess Gibbs free energy against x 2 for water (1 )-n-propanol (2) at atmospheric pressure

A

I .o 0 0.1 0.2 0.3 0

x3 - Figure 7. Projection of the phase diagram for water-n-propa- nol-n-butanol at 760 mm Hg -- vapor 3-phase curve, .... liquid 3-phase curves, - tie triangles, C critical point

where 721 = ( 9 2 1 - g11)/RT, 7 1 2 = (912 - 922)/RT G21 =exp(-a21721), GIZ = e x p ( - a 1 ~ 7 1 ~ ) , a n d a 1 2 =(YZI.

The quantities gi, are interaction energies and cy12 is the so-called "nonrandomness parameter. " The three in- dependent parameters of Equation 3 were obtained for the three binary systems by making a nonlinear least- squares fit to the excess Gibbs energies, assuming the parameters to be independent of temperature. The data for n-propanol-n-butanol and water-n-butanol were those of Gay ( 4 ) and Smith and Bonner ( 7 3 ) , respectively. The excess Gibbs energies for the system n-propanol-n-buta- no1 are much smaller than those of the other two systems and, within the experimental error, GE is a symmetrical function of composition. For this system, therefore, the parameter was taken to be zero when Equation 3 reduces to the second-order Margules equation:

G' = 2 ~ 1 x 2 (gzi - gii) ( 4 )

The binary parameters are given in Table V . All of the

48 Journal of Chemical and Engineering Data, Vol. 18, No. 1,

0.4 A

0 ai 0.2 0.3 a4 a5 x3 -

Figure 8. Isothermal sections (with tie-lines) of the phase di- agram for water-n-propanol-n-butanol at 89" and 91 OC --vapor curves, - liquid curves

systems could be fitted to Equation 3 with root-mean- square deviations of less than 0.003 in G E / R T .

The binary parameters were then used in Equation 5, which is the ternary form of the Renon-Prausnitz equa- tion, to calculate the activity coefficients of the ternary system and hence bubble point temperatures and vapor phase compositions at 760 m m Hg:

where Gkk = 1 and 7kk = 0. The results of the calculations are given in Table I V

where the vapor phase compositions are compared with the experimental values. The agreement is very satisfac- tory, bearing in mind the limitations of the analytical method used for obtaining the compositions of n-propanol and n-butanol.

Equation 5 has also been used to examine the phase diagram in the vicinity of the liquid-liquid region. I n an at- tempt to establish the liquid-liquid phase boundary and tie-lines by searching (graphically) for the compositions at which the activities of each component were uniform, it was found that the equilibrium compositions were too sensitive to small changes in the activity for a reliable estimate of the binodal curve to be made. Instead, esti- mates of the binodal curves at temperatures close to the bubble point were made by extrapolating the liquid-liquid equilibrium data previously reported ( 70). Liquid-liquid

1973

tie-lines were then obtained by finding the compositions at which curves of constant activity of n-propanol inter- sected the binodal curve. The resultant liquid and vapor three-phase curves are shown, in projection, in Figure 7. The three-phase curves cover a range of boiling points of only about 3°C.

The system does not form a ternary azeotrope, homo- geneous or heterogeneous. The vapor and liquid surfaces are rather flat, however, in the region between the homo- geneous water-propanol azeotrope and the heterogene- ous water-butanol azeotrope. This is evident on inspec- tion of two isothermal sections of the phase diagram at 89" and 91 "C, as shown in Figure 8.

Acknowledgment

We are grateful to G. L. Standart for his encourage- ment. We are also indebted to F. P. Stainthorp and H. M. Rash for their guidance in the preparation of the neces- sary computer programs.

Nomenclature B i j = second virial coefficient, cm3 m o l - ' GE = excess Gibbs free energy of mixing, J mol G i j = parameter of Renon-Prausnitz equation g i j = parameter of Renon-Prausnitz equation H E = excess enthalpy of mixing, J m o l - ' P = total pressure, m m Hg Pio = pure component vapor pressure, mm Hg R = gas constant, J K - ' mol-' T = Kelvin temperature, K t = Celsius temperature, "C V E = excess volume of mixing, cm3 mol - 1

x i = liquid phase mole fraction y t = vapor phase mole fraction

W i = w t %

- 1

Greek Letters

aij = parameter of Renon-Prausnitz equation T i = liquid phase activity coefficient 7 = refractive index (sodium D-line) p = density, g ~ m - ~ T i j = parameter of Renon-Prausnitz equation

Subscripts

1 = water 2 = n-propanol 3 = n-butanol i , j , k , l , r = running variables

Literature Cited

Collins, S. C., Keyes, F. G., Proc. Amer. Acad. Sci., 72, 283 (1938). Cox, J . D., Trans. FaradaySoc., 57, 1674 (1961). Doroshevsky, A.. Polansky. E., Z. Phys. Chem.. 73,192 (1910) Gay, L., Chim. Ind. 18, 187 (1927). Goodwin. S. R . , Newsham, D. M. T., J. Chem. Thermodyn., 3 , 325 (1971). Goodwin, S. R . . Newsham, D. M. T., ibid., 4, 31 (1972). Hala, E., Pick, J., Fried, V., Vilim, O., "Vapour-Liquid Equilibrium," 2nd ed., p 404, Pergamon, London, 1967. Hala, E., Wichterle. I . , Polak, J., Boublik, T., "Vapour-Liquid Equi- librium Data at Normal Pressures," Pergamon, London, 1968. Murti. P. S., Van Winkle, M . , Ind. Eng. Chem., 3, 72 (1958). Newsham, D. M. T., Ng, S. E., J. Chem. Eng. Data, 17 (2), 205

Plewes. A. C., Jardine. D. A., Butler, R. M., Can. J. Technoi.. 32,

Renon, H., Prausnitz, J. M., AlChEJ., 14, 135 (1968). Smith, T. E., Bonner, R. F., Ind. Eng. Chem., 41, 2867 (1949): Vilim, O., Hala, E., Pick, J., Fried, V.. Collect. Czech. Chem. Com- mun., 19, 1330 (1954).

(1972).

133 (1954).

Received for review March 30, 1972. Accepted August 21, 1972. S. B. Ng received financial support from the British Council.

Integral Isobaric Heat of Vaporization of Benzene-I ,2=Dichloroethane System

Yaddanapudi Jagannadha Rao and Dabir S. Viswanathl Department of Chemical Engineering, Indian Institute of Science, Bangalore- 72, India

Integral isobaric heats of vaporization of benzene-l,2- dichloroethane mixtures were measured at pressures of 684 and 760 mm of Hg using a modified Dana's apparatus. The results were found to be linear with composition.

Latent heat of vaporization is a very important property needed in the design and operation of chemical plants. Several investigators have therefore devised methods for the determination of this property. Very little data ( 7 , 2, 6-70, 72, 73, 7 5 ) are available on heat of vaporization of mixtures.

The first published work on latent heats of mixtures was by Dana (2) who worked at atmospheric pressure and cryogenic temperatures. Apart from other sources of error, the main source of error in his experimentation was

' To whom correspondence should be addressed

due to heat leak because of the considerable tempera- ture gradient between the system and the surroundings. Subsequently, attempts have been made (6-70, 72, 73) to minimize the heat leak and reduce other sources of error involved. The best modification was due to Shettigar et al. (9) who introduced a liquid meter also to avoid changes in the equilibrium condition of the experiment.

Experimental

Materials. Benzene used was of Pro analysi grade pro- duced by Sarabhai Merck Ltd., India, with a reported boiling range of 80-81°C. This material was subjected to the thiophene test. Thiophene was removed by treating it with concentrated sulfuric acid and distilling i t after sepa- ration and washing it with distilled water. Benzene then was dried over calcium chloride, filtered, and further puri- fied in a distillation column. Only the middle fractions of the distillate were collected. Table I gives a comparison of the experimentally determined physical properties with the literature values.

Journal of Chemical and Engineering Data, Vol. 18, No. 1, 1973 49