Вы находитесь на странице: 1из 3

Research: Science and Education

Regression Methods To Extract Partial Molar Volume Values in the Method of Intercepts
Leon F. Loucks* Chemistry Department, University of Prince Edward Island, Charlottetown, PE C1A 4P3, Canada

Partial molar quantities, while fundamental to the thermodynamics of solutions, are conceptually difficult for many students. In the laboratory, only volume is studied in most physical chemistry course offerings. Volume has the advantage of concrete visualization, which assists students and may permit them to extrapolate to other thermodynamic properties, once partial molar volume has been mastered. For the evaluation of the partial molar volumes of the components of binary solutions, either the slope method (14) or the method of intercepts (3, 58) is used. The method of intercepts is a very satisfying approach in mathematical and graphical theory; however, in practice it is imprecise because the graphical assignment of tangents is too arbitrary for many binary systems that show only minor nonideality in the plots of molar volume against mole fraction. This paper presents the use of a regression technique to generate an equation that accommodates the curvature in the plot of molar volume vs mole fraction for systems exhibiting a regular pattern of nonideality, in which the deviation is zero at the mole fractions of zero and unity and increases monotonically to a maximum at 0.5 mole fraction. This equation is then used to generate the first derivative so that the slope of the tangent line can be evaluated at all points along the curve; in turn, the intercepts of the tangent line are evaluated. This method provides precision far superior to manual graphical methods. In the following presentation, the notation follows that of Alberty and Silbey (5). The total volume of an ideal bi nary solution is given by V = V *1n1 + V *2n2, where V *1 and V *2 are the molar volumes of the pure components 1 and 2 and n1 and n2 are the number of moles of 1 and 2. In practice, however, the total volume of a binary solution is usually less than predicted by ideal mixing (although in some cases it is more). Such volume changes arise from the relative strengths of the intermolecular forces between the like and unlike molecules of the binary solution and from differences in the packing of unlike molecules, due to size and shape. In cases where the volume of a solution is less than the sum of the volumes mixed, the presence of the second substance provides for more efficient packing of the molecules or generates stronger interactions between unlike molecules than between like molecules. In some cases, dislocations in the hydrogenbonded structure by the insertion of foreign molecules cause collapse of structure and volume decreases. Similarly, the solvation of ions to create a tightly packed solvation shell may lead to a net volume decrease. In cases where the solution volume exceeds the sum of volumes mixed, the second substance disrupts a rather strong intermolecular interaction of a pure material and substitutes a weaker interaction between the unlike molecules; the result is a larger volume per molecule, on average.
*Email: Loucks@UPEI.ca

The differential of binary solution volume change caused by component additions to a solution being kept at constant temperature and pressure is:

dV = V n 1

n2

dn 1 + V n 2

dn 2
n1

(1)

The partial derivative coefficients in this equation arecalled the partial molar volumes and are given the symbols V1 and V2. Thus: dV = V1dn1 + V2dn2 (2) The partial molar volumes depend on the solution composition. Their values can be appreciably different from the molar volumes of the pure components. For a solution of particular composition, eq 2 may be integrated to yield a total volume given by V = V1n 1 + V2n2 (3) Division by the total number of moles (n1 + n2) results in V = x 1V1 + x 2V2 (4) where V is the molar volume of any binary solution; that is, the volume of one mole of solution. Upon insertion of x1 = (1 x2), this latter equation rearranges to V = V1 + ( V2 V1)x2 (5) The tangent line at a particular concentration is given by this equation. The tangent line has an intercept of V1 at x2 = 0; similarly, the right-side intercept at x2 = 1 must be V2, as the the slope V2 V1 can only thus be accommodated. At any different composition, the tangent line differs and V1 and V2 take on different values. Figure 1 shows an example of one tangent drawn to a plot of V vs x 2. The molar volumes of binary solutions are readily calculated from density measurements. In our studies we have found that densities measured at 25 C with a 25-mL Weldstyle density bottle for about 10 different mole fractions provide a sufficient data base for the ensuing analysis. It is helpful to select compounds of similar molar volumes in choosing the binary set, so the plot of molar volume vs mole fraction can easily reveal the curvature indicative of deviation from ideality. We have found the ethanolethylene glycol and t-amyl alcoholtoluene systems to be excellent selections for the negative and positive deviations from ideality. Table 1 shows typical student data for the two systems. The filled circles of Figure 1 show the dependence of the molar volume on the mole fraction of ethanol in the ethanol ethylene glycol system. The regular pattern shows a clear downward curvature between the intercepts, which are the molar volumes of the pure compounds. If there existed no deviation from ideality, the intercepts would be joined by a straight line.
425

JChemEd.chem.wisc.edu Vol. 76 No. 3 March 1999 Journal of Chemical Education

Research: Science and Education

To fit the data well, an appropriate regression should allow for the curvature as an effect superimposed on what would otherwise be a straight line. A parabolic term centered on the midpoint of the mole fraction axis provides a way to introduce the correction symmetrically on each side of x 2 = 0.5; thus we assume that V is given by a + bx2 as a first approximation, which can be improved by adding a parabolic term. The use of such a parabolic term has been reported in the literature (9) as the simplest term of a general method. Thus the regression needs to be conducted to satisfy V = a + bx2 + c(x 2 0.5)2 (6) The coefficients a, b, and are evaluated in a spreadsheet c multiple regression of the V values as a function of both x2 and (x2 0.5) 2. In the case of ethanolethylene glycol, the equation becomes V = 55.287 + 2.793x2 + 2.08(x 2 0.5)2 (7) The data for t-amyl alcoholtoluene give V = 108.16 + 2.416x2 2.20(x 2 0.5)2 (8)

59

Molar Volume / mL

58

57

56

55 0.0

0.2

0.4

0.6

0.8

1.0

Mole Fraction Ethanol


Figure 1. A plot of the molar volumes of ethanolethylene glycol solutions as a function of mole fraction ethanol. Circles are experimental data; the curve is calculated from the regression equation; and the straight line is the tangent calculated from the regression coefficients at a mole fraction of 0.5.

Notice the change of sign for the coefficient c as the nonideality changes character between the two systems. The R 2 values are .9995 and .9982 in these two respective regressions. The solid curve of Figure 1 has been constructed from the V values calculated from eq 7; the fit is obviously very good. Once equipped with an equation for the dependence of V on x2, one can extract the equation for the slope of the tangent line as the first derivative. Thus dV /dx 2 = b c + 2cx2 (9) As expected, the slope of the tangent varies with the mole fraction. The slope of the tangent at a particular value of V can be expressed as ( V V1)/x 2 = slope (10) or as: ( V2 V )/(1 x 2) = slope (11) The expressions for V given by eq 6 and for the slope by eq 9 can then be used to obtain: V1 = a + 0.25c cx22 (12) 2 V2 = a + b 0.75c + 2cx 2 cx 2 (13) These relationships permit us to calculate V1 and V2 values at every mole fraction value that we choose. Table 2 shows a number of values calculatedfor the ethanolethylene glycol system. The data reveal how V2 approaches the molar volume of pure ethanol (58.60 mL) as x2 approaches unity; in contrast, as x2 approaches zero the molar volume of ethanol has reached a quite different limit of 56.52 mL, representing the volume of one mole of ethanol dissolved in an infinite amount of ethylene glycol. Conversely, as x2 approaches zero, the partial molar volume of ethylene glycol becomes the molar volume of that pure substance (55.81 mL). Meanwhile, at x 2 equal to unity, the partial molar volume of ethylene glycol has taken a limiting value of 53.73 mL for the volume of one mole dissolved in an infinite amount of ethanol. Figure 2 shows a plot of the calculated values of V1 and V2 as a function of the mole
426

Table 1. Molar Volumes of Binary Solutions at 25 C Ethanol Ethylene Glycol

X ethanol
0 0.1092 0.2244 0.3321 0.4393 0.5499 0.6529 0.7818 0.8686 1

V /mL 55.828 55.902 56.060 56.245 56.513 56.866 57.178 57.621 58.004 58.591

t - Amyl Alcohol Toluene X amyl alcohol V /mL


0 0.1130 0.2232 0.3307 0.4407 0.5505 0.6621 0.7702 0.8741 1 107.554 108.155 108.549 108.915 109.240 109.452 109.650 109.859 109.976 110.041

Table 2. Partial Molar Volumes in EthanolEthylene Glycol System X ethanol V ethanol /mL V ethylene glycol /mL 0 0.05 0.15 0.25 0.35 0.45 0.55 0.65 0.75 0.85 0.95 1 56.519 56.722 57.097 57.430 57.721 57.971 58.179 58.345 58.470 58.553 58.595 58.600 55.807 55.802 55.760 55.677 55.552 55.386 55.178 54.928 54.637 54.304 53.929 53.726

Journal of Chemical Education Vol. 76 No. 3 March 1999 JChemEd.chem.wisc.edu

Research: Science and Education


59

0.0 -0.1
Excess Molar Volume / mL

Partial Molar Volume / mL

58

57

-0.2 -0.3 -0.4 -0.5 -0.6

56

55

54

53 0.0

0.2

0.4

0.6

0.8

1.0

0.0

0.2

Mole Fraction Ethanol


Figure 2. A plot of the partial molar volumes of ethanol and ethylene glycol as a function of the mole fraction of ethanol in ethanol ethylene glycol solutions. All values are intercepts of tangents of Fig. 1, calculated from the regression coefficients and the mole fraction. The upper curve is for ethanol; the lower, for ethylene glycol.

0.4 0.6 Mole Fraction Ethanol

0.8

1.0

Figure 3. A plot of the excess molar volumes of ethanolethylene glycol solutions as a function of the mole fraction of ethanol. The curve shows the values calculated from a regression fit of the experimental data to a parabola centered on a mole fraction of 0.5.

fraction of ethanol and we see that the partial molar volumes are continuously variable with the composition of the solution. Many thermodynamicists prefer to monitor the molar volume behavior of binary solutions in terms of excess molar volume. This quantity, VE, is the volume deviation per mole from the expected volume if no contraction or expansion occurs upon mixing the two components. Thus VE = V x1V *1 x2V *2 (14) For systemsthat exhibit contraction upon mixing, V will be less than x1V *1 + x2V *2; hence the excess molar volumes are negative values. Such is the case for the ethanolethylene glycol system, whereas the t-amyl alcoholtoluene system yields positive excess molar volumes. Figure 3 shows a plot of the excess molar volumes of the ethanolethylene glycol system with the maximum in the negative value occurring at about x2 = 0.5. The functional behavior is approximately parabolic centered on x2 = 0.5. The data can be fit to VE = m + c(x 2 0.5) 2 (15) Regression of the VE experimental values as a function of the square of (x2 0.5) gives the constants: m = 0.5255 and c= 2.079. These constants can be used to calculate values of VE

for any x 2; the solid curve shown on Figure 3 displays the values calculated from the regression equation. Once again a regression method using a parabolic term centered on x2 = 0.5 has been found very useful in the data analysis of a wellbehaved binary system. Literature Cited
1. Levine, I. N. Physical Chemistry, 3rd ed.; McGraw-Hill: New York, 1988; p 234. 2. Moore, W. J. Basic Physical Chemistry; Prentice Hall: Englewood Cliffs, NJ, 1982; p 179. 3. Barrow, G. M. Physical Chemistry, 6th ed.; McGraw-Hill: New York, 1996; p 282. 4. Reid, C. E. Chemical Thermodynamics; McGraw-Hill: New York, 1990; p 150. 5. Alberty, R. A.; Silbey, R. J. Physical Chemistry, 1st ed.; Wiley: New York, 1992, p 126. 6. Noggle, J. H. Physical Chemistry, 2nd ed.; Scott-Forseman: Glenview, IL, 1989; p 334. 7. Atkins, P. W. Physical Chemistry, 4th ed.; Freeman: New York, 1990; pp 182183. 8. Salzberg, H. W.; Morrow, J. I.; Cohen, S. R.; Green, M. E. Physical Chemistry; Academic: New York, 1969; p 375. 9. Redlich, O.; Kister, A. T. Ind. Eng. Chem. 1948, 40, 345348.

JChemEd.chem.wisc.edu Vol. 76 No. 3 March 1999 Journal of Chemical Education

427

Вам также может понравиться