Fluid Phase Equilibria 340 (2013) 42–45

Contents lists available at SciVerse ScienceDirect

Fluid Phase Equilibria

journal homepage: www.elsevier.com/locate/fluid

Isobaric vapor–liquid equilibrium for binary system of methanol and acetonitrile

Yumei Li a,b , Peng Bai a,∗ , Qionghong Zhuang a
a
School of Chemical Engineering and Technology, Tianjin University, Tianjin 300072, PR China
b
Department of Chemistry, Dezhou University, Shandong 253023, PR China

a r t i c l e i n f o a b s t r a c t

Article history: Isobaric vapor–liquid equilibrium (VLE) data for methanol and acetonitrile system was measured at
Received 18 June 2012 101.3 kPa by using equilibrium still. The results show that methanol and acetonitrile system can form a
Received in revised form 2 December 2012 minimum temperature azeotrope. Thermodynamic consistency of the experimental data was confirmed
Accepted 6 December 2012
by means of Herington semi-empirical method. The experimental values obtained were correlated by
Available online 17 December 2012
using the Margules, Van Laar, Wilson and NRTL activity-coefficient models, respectively. The binary inter-
action parameters of the activity-coefficient models were determined. The average absolute deviation in
Keywords:
boiling point and vapor-phase composition were determined by comparing the experimental values with
Methanol
Acetonitrile
those calculated by the activity-coefficient models. The calculated maximum average absolute deviations
Vapor–liquid equilibrium were 0.25 K and 0.0056 for the boiling point and vapor-phase composition, respectively. Therefore, it was
shown that the four activity-coefficient models used satisfactorily correlate the experimental results of
methanol and acetonitrile system.
© 2012 Elsevier B.V. All rights reserved.

1. Introduction 2. Experimental

Methanol and acetonitrile have been used extensively as 2.1. Chemicals

solvents and important fine chemicals in the chemical and phar-
maceutical production process [1–4]. Several processes in chemical Chemicals used: All the reagents were analytical purity grade.
industries produce mixtures of methanol and acetonitrile, so sep- Methanol and acetonitrile were purchased from Tianjin University
aration of them is indispensable. Availability of accurate and Ke Wei Company. The purities of all the chemicals were checked
quantitative information on the phase equilibrium of binary or by a gas chromatograph (GC) equipped with a flame ionization
multicomponent mixtures is essential for designing distillation detector. The water content in chemicals was determined using
processes of liquid mixtures [5,6]. To our knowledge, no isobaric a Karl Fischer moisture titrator. The purity and water content are
VLE data are available for methanol and acetonitrile mixture. In listed in Table 1. The chemicals were degassed and subsequently
this paper, we present isobaric vapor–liquid equilibrium measure- dried over molecular sieves (Nankai University 3A) for 24 h prior
ments for the mixture of methanol and acetonitrile at 101.3 kPa to use. The purity of chemicals was further confirmed by com-
to provide a beneficial reference for separation of methanol and paring the experimental densities, the refractive indexes and the
acetonitrile. normal boiling points with those reported in [11,12], as they are
The experimental VLE data of the binary systems investigated listed in Table 2. The densities of the pure components were mea-
were found to be thermodynamically consistent and the activity sured at 293.15 K using an DA-510 densimeter, and the refractive
coefficients have been calculated and correlated with the following indexes were measured at 293.15 K using an ATAGO RX-5000.
equations: Margules and Akad [7], Van Laar [8], Wilson [9] and The accuracy in density, and refractive index measurements are
NRTL [10]. ±2.0 × 10−5 g/cm3 and ±0.00001, respectively. Temperature was
controlled to ±0.01 K with a thermostated bath.

2.2. Apparatus and procedure

The equilibrium still, the same as in the previous work [13],

was employed to measure isobaric VLE data at 101.3 kPa, and the
∗ Corresponding author. Tel.: +86 022 27406369. favorable performance of the apparatus was verified in the previous
E-mail address: [email protected] (P. Bai). measurement.

0378-3812/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.fluid.2012.12.003
 Y. Li et al. / Fluid Phase Equilibria 340 (2013) 42–45 43

Table 1 1.0
Purity and water content of pure components.

Component Purity (mass%) Water (mass%)

0.8
Suppliers’ GC

Methanol >99.7 99.75 0.03

0.6
Acetonitrile >99.5 99.61 0.03

y1
0.4
360

355 0.2

350 0.0
0.0 0.2 0.4 0.6 0.8 1.0
T/ K

345 x1

340 Fig. 2. y–x Diagram for the methanol–acetonitrile system at 101.3 kPa.

335 Table 3
Experimental VLE data for the methanol–acetonitrile binary system at 101.3 kPa.
330 T (K) x1 y1 1 2
0.0 0.2 0.4 0.6 0.8 1.0
338.41 1.000 1.000 1.000
x1 y1 338.04 0.962 0.945 0.998 2.535
336.96 0.866 0.860 1.030 1.863
Fig. 1. Vapor–liquid equilibrium phase diagram for the methanol–acetonitrile sys- 337.02 0.805 0.814 1.051 1.705
tem at 101.3 kPa. 337.09 0.766 0.785 1.064 1.640
337.71 0.646 0.706 1.119 1.464
338.09 0.602 0.687 1.158 1.376
The experiments were accomplished at 101.3 kPa. A mercury 339.38 0.437 0.600 1.324 1.188
thermometer calibrated by Shanghai Metrology Institution was 340.55 0.310 0.529 1.535 1.074
used for temperature measurement within accuracy ±0.01 ◦ C. The 344.6 0.171 0.380 1.750 1.047
350.19 0.054 0.189 2.225 0.994
actual atmospheric pressure was measured by a mercury barom- 355.66 0.000 0.000 1.000
eter whose accuracy was ±50 Pa with a control system whose
uncertainty is 50 Pa. Equilibrium conditions were assumed when
constant temperature and pressure were obtained for 25 min
could be obtained by using y1 − x1 versus x1 diagrams in order
or longer, and to verify the equilibrium state, condensate was
to determine x1az at y1 − x1 = 0 together with T versus x1 diagrams
analyzed every 10 min until the variation of the vapor phase com-
where Taz should be minimum at x1az . The composition and boiling
position of the last three samples is less than 0.001 (mole fraction).
temperature of the azeotrope obtained were: 0.84 and 63.85 ◦ C,
Then the samples of liquid and condensate were taken simulta-
respectively.
neously for analysis. The sampling was carried out with special
syringes that allowed withdrawal of small volume samples (0.5 ␮L).
At least two analyses were made for each sample. 3.2. Thermodynamics consistency verification

3. Results and discussion To confirm thermodynamic consistency of experimental data,

all experimental data were verified by means of the Herington
3.1. Experimental results semi-empirical method [12], in which the value of (D − J) must be
less than 10.

The vapor–liquid equilibrium data, T, xi and yi , along with the I  
calculated activity coefficients,  i , are gathered in Table 3. The D =  × 100, J = 150 ×
corresponding phase diagram was shown in Figs. 1 and 2. As Tmin
can be seen from Fig. 2, there is intercrossing point between the 1  1 
equilibrium curve and the diagonal line, which reveals that there where I = 0 ln(1 /2 ) dx1 , = 0 ln(1 /2 ) dx1 ,  =
exists azeotrope in the methanol + acetonitrile binary system at Tmax − Tmin , Tmax and Tmin were the highest and the lowest
101.3 kPa. It can be also seen that the binary system exhibited boiling point in the system, respectively (K).
positive deviation from Raoult’s law and a minimum temperature The results of thermodynamic consistency test are demon-
azeotrope was found (Fig. 1). The location of the azeotropic points strated in Fig. 3 and Table 4. The calculation result was that

Table 2
Density , normal boiling point Tb , and refractive index nD of pure components.

Component  (293.15 K) (g cm−3 ) Tb (101.3 KPa) (◦ C) nD (293.15 K)

a b
Experimental Literature Experimental Literature Experimental Literaturea

Methanol 0.79182 0.7913 65.26 64.70 1.32835 1.3286

Acetonitrile 0.78338 0.7822 82.51 81.6 1.34358 1.34411
a
Taken from Ref. [11].
b
Taken from Ref. [12].
44 Y. Li et al. / Fluid Phase Equilibria 340 (2013) 42–45

1.0 Table 5
The Antoine constants of pure substances.

Component A B C
0.5 Methanol 7.4182 1710.2 −22.25
Acetonitrile 6.34522 1388.446 −34.856
ln( 1/ 2)

0.0

355
-0.5

350
-1.0
0.0 0.2 0.4 0.6 0.8 1.0 345

/
x1
340
Fig. 3. Diagram of ln( 1 / 2 ) to x1 .

Table 4 335
Thermodynamic consistency test of VLE data. 0.0 0.2 0.4 0.6 0.8 1.0
Tmax Tmin  (K) I  D J x1 y1
355.66 338.5 17.16 0.0086 0.4432 1.9404 7.6393
Fig. 4. T–y–x diagram for the methanol–acetonitrile system at 101.3 kPa: (, )
experimental data and (—) Wilson equation.
|D − J| = 5.6989 < 10, which indicated that all experimental data
were in accordance with thermodynamic consistency. where A, B, and C are Antoine constants and T is the temperature
in Kelvin. The Antoine constants were obtained from the literature
3.3. Correlation of binary vapor–liquid equilibrium [12], and were presented in Table 5.
Because the experiment was carried out at 101.3 kPa, the vapor
The general equilibrium relationship between the vapor phase phase could be regarded as ideal gas, the parameters could be
and liquid phase can be expressed as: regarded as constant values. In order to find parameters of the equa-
  tions, activity coefficients were correlated with the Margules, Van
Vil (p − psi )
ˆ V p = xi i s ps exp
yi  (1) Laar, Wilson and NRTL equations. Estimation of the parameters of
i i i RT
the equations was based on minimization of the following objective
The Poynting factor, the exponential term in Eq. (1), is close to function (OF):
unity at low pressure and the vapor phase is assumed to be an ideal  exp 2
gas and the pressure dependence of the liquid phase fugacity is OF = (ln ical − ln i )k (4)
neglected. Eq. (1) can be simplified to k i

pyi = psi i xi (2) where k is the number of data points and i the number of compo-
nents in the mixture. The experimental and calculated values are
where P and T are the total pressure and the temperature in the exp
denoted by the superscripts ‘exp’ and ‘cal’, respectively. i is cal-
equilibrium system. xi and yi are the liquid and vapor phase mole
ˆ V is the fugacity coefficients
fractions in equilibrium, respectively.  culated with Eq. (2) and ical is calculated with Margules, Van Laar,
i
ˆ s is the fugacity coefficient Wilson and NRTL models. The correlated interaction parameters,
of component i in the vapor mixture,  i together with the average deviations of temperature and the vapor-
of pure vapor i at equilibrium system temperature T and saturation
phase composition for the methanol and acetonitrile binary system
vapor pressure psi , Vil is the mole volume of pure liquid i, R is the
at 101.3 kPa between the experimental and the calculated values
universal gas constant,  i is the activity coefficient of component i.
obtained from Margules, Van Laar, Wilson and NRTL models, were
psi is the saturated vapor pressure of pure component i at the system
shown in Table 6. The correlated interaction parameters obtained
temperature, calculated with the Antoine equation
through simplex search were, then, used to calculate y1cal , T cal , 1cal
B and 2cal under corresponding liquid compositions and pressure
lg psi (kPa) = A − (3)
T (K) + C (101.3 kPa) and the results calculated were given in Table 7. The

Table 6
Margules, Van Laar, Wilson and NRTL parameters and mean absolute deviations of the binary system.

Model Parameters |T|a |y|b

Methanol (1) + acetonitrile (2) at 101.3 KPa

Margules A12 = 0.884 A21 = 0.889 0.25 0.0048
Van Laar A12 = 0.834 A12 = 0.940 0.21 0.0056
Wilson 12 = 0.688 21 = 0.525 0.22 0.0054
NRTL g12 − g22 = 1993.75 g12 − g11 = 1755.63 ˛12 = 1.09 0.22 0.0044

  of data
 points. 
c

T  =  Texp − Tcal  /kc .

k: number
a
   
b y = yexp − ycal  /kc .
 Y. Li et al. / Fluid Phase Equilibria 340 (2013) 42–45 45

Table 7
Correlation results for methanol + acetonitrile system at 101.3 kPa.

x1 Margules Van Laar Wilson NRTL

y1cal Tcal (K) 1cal 2cal y1cal Tcal (K) 1cal 2cal y1cal Tcal (K) 1cal 2cal y1cal Tcal (K) 1cal 2cal

1.000 1.000 338.22 1.000 2.433 1.000 338.22 1.000 2.560 1.000 338.22 1.000 2.602 1.000 338.22 1.000 2.788
0.962 0.950 337.88 1.001 2.277 0.949 337.82 1.002 2.367 0.948 337.81 1.002 2.389 0.946 337.75 1.002 2.475
0.866 0.853 337.42 1.016 1.950 0.851 337.32 1.019 1.977 0.852 337.30 1.020 1.975 0.853 337.24 1.024 1.956
0.805 0.805 337.35 1.035 1.781 0.805 337.25 1.039 1.786 0.806 337.23 1.041 1.780 0.810 337.20 1.047 1.745
0.766 0.778 337.38 1.050 1.687 0.779 337.27 1.056 1.682 0.780 337.26 1.058 1.675 0.785 337.24 1.065 1.640
0.646 0.708 337.70 1.118 1.451 0.712 337.62 1.129 1.432 0.713 337.62 1.131 1.428 0.717 337.66 1.134 1.408
0.602 0.685 337.91 1.151 1.382 0.690 337.83 1.164 1.362 0.691 337.84 1.165 1.359 0.693 337.91 1.165 1.346
0.437 0.601 339.14 1.325 1.186 0.607 339.12 1.340 1.169 0.606 339.15 1.336 1.171 0.601 339.33 1.316 1.179
0.310 0.524 340.82 1.526 1.090 0.527 340.89 1.532 1.079 0.526 340.92 1.526 1.082 0.517 341.11 1.489 1.094
0.171 0.396 344.19 1.838 1.027 0.394 344.39 1.815 1.023 0.394 344.38 1.813 1.024 0.389 344.39 1.791 1.032
0.054 0.186 349.97 2.207 1.003 0.181 350.18 2.129 1.002 0.182 350.13 2.146 1.002 0.187 349.91 2.220 1.004
0.000 0.000 354.80 2.421 1.000 0.000 354.80 2.302 1.000 0.000 354.80 2.337 1.000 0.000 354.80 2.506 1.000

comparisons of the experimental results with those calculated by activity-coefficient models. The calculated maximum average
Wilson activity-coefficient models were shown in Fig. 4. absolute deviations were 0.25 K and 0.0056 for the boiling point and
It can be seen from Table 7 that the differences among the vapor vapor-phase composition, respectively. Therefore, it was shown
phase mole fraction and boiling temperature calculated by the Mar- that the four activity-coefficient models used satisfactorily corre-
gules, Van Laar, Wilson and NRTL models are indistinct. As can be late the experimental results of methanol and acetonitrile system.
seen from Table 6, the average values of |T| and |y| were less than
0.25 K and 0.0056, respectively. Thus, all of the four models can cor- References
relate the VLE data for the binary systems methanol + acetonitrile
at 101.3 kPa well. [1] S.M. Mane, B.N. Thorat, M.R. Sawant, J. Disper. Sci. Technol. 33 (2012) 357–361.
[2] M. Abdulla-Al-Mamun, Y. Kusumoto, M. Muruganandham, Mater. Lett. 63
(2009) 2007–2009.
4. Conclusions [3] P.S. Niphadkar, P.N. Joshi, H.R. Gurav, Catal. Lett. 133 (2009) 175–184.
[4] H. Tamiaki, A. Shinkai, Y. Kataoka, J. Photoch. Photobio. A 207 (2009)
115–125.
Isobaric VLE values were determined experimentally for [5] A. Aucejo, S. Loras, V. Martinez-Soria, J. Chem. Eng. Data 51 (2006) 1051–1055.
methanol and acetonitrile binary system at 101.3 kPa. The results [6] G. Ovejero, M.D. Romero, E. Díez, T. Lopes, I. Díaz, J. Chem. Eng. Data 52 (2007)
show that methanol and acetonitrile system can form a mini- 1984–1987.
[7] M. Margules, S.-B. Akad, Wiss. Wien, Math-Naturwiss. KI. II 104 (1895)
mum temperature azeotrope. Thermodynamic consistency of the 1234–1239.
experimental data was confirmed by means of Herington semi- [8] J.J. van Laar, Z. Phys. Chem. 72 (1910) 723–751.
empirical method. The VLE values were correlated by the four [9] G.M. Wilson, J. Am. Chem. Soc. 86 (1964) 127–135.
[10] H. Renon, J.M. Prausnitz, AIChE J. 14 (1968) 135–142.
activity-coefficient models, namely, Margules, Van Laar, Wilson
[11] N.L. Cheng, Solvent Handbook, Chemical Industry Publishing House, Beijing,
and NRTL. The corresponding binary interaction parameters of 2007.
the models were obtained. The average absolute deviation in [12] P.S. Ma, Chemical Engineering Thermodynamics, Chemical Industry Publishing
boiling point and vapor-phase composition were determined by House, Beijing, 2005.
[13] Y.M. Li, P. Bai, Asian J. Chem. 24 (2012) 1861–1863.
comparing the experimental values with those calculated by the