Fluid Phase Equilibria 224 (2004) 31–37

Liquid–liquid equilibrium data for binary alcohol + n-alkane

(C10 –C16) systems: methanol + decane, ethanol + tetradecane,
and ethanol + hexadecane
Hiroyuki Matsuda, Kenji Ochi∗
Department of Materials and Applied Chemistry, Nihon University, 1-8 Kanda Surugadai, Chiyoda-ku, Tokyo 101-8308, Japan

Received 1 February 2004; accepted 24 May 2004

Available online 28 July 2004

Abstract

Liquid–liquid equilibrium (LLE) data for three binary alcohol + n-alkane (C10 –C16 ) systems—methanol + decane, ethanol + tetradecane,
and ethanol + hexadecane—were measured using a laser scattering technique. The experimentally determined cloud points were satisfactorily
correlated by three local composition models (NRTL, Tsuboka–Katayama’s modification of the Wilson equation, and the modified complete
local composition model suggested by Nagata and Tamura). Prediction of vapor–liquid equilibria by means of these models with parameters
obtained from the LLE data was also tested.
© 2004 Elsevier B.V. All rights reserved.

Keywords: Experimental; Liquid–liquid equilibria; Laser scattering technique; Vapor–liquid equilibria; Activity coefficient models

1. Introduction (LC) models: NRTL [10], Tsuboka–Katayama’s modifica-

tion of the Wilson equation (T–K–Wilson) [11], and the
Liquid–Liquid equilibrium (LLE) data play an important modified complete local composition (CLC) model sug-
role in the design and development of separation processes. gested by Nagata and Tamura [12]. We also attempted to
In our laboratory, we have measured LLE data by means of predict vapor–liquid equilibria (VLE) using these LC mod-
the cloud point method using a laser scattering technique. els, with parameters obtained from the experimental LLE
This technique is advantageous for easy and exact determi- data for these systems.
nation of cloud points. In previous studies, we determined
the mutual solubilities for 18 binary systems [1–9], espe-
cially methanol + hydrocarbon systems in which the carbon 2. Experimental
number of the hydrocarbons is 6–8: methanol + hexane [7],
methanol + heptane [7], methanol + octane [6], methanol 2.1. Apparatus and procedure
+ cyclohexane [8], and methanol + 2,2,4-trimethylpentane
[6]. However, there is not much literature data on LLE for The cloud points of the binary mixtures were measured
hydrocarbons with higher carbon numbers. using a laser scattering technique. Details of the appara-
In this study, we made detailed measurements of LLE tus and its operation have been described elsewhere [1–4].
of three binary systems, methanol + decane, ethanol A light sensor (NP type silicon capsule module, Model
+ tetradecane, and ethanol + hexadecane, at moderate pres- 3-400S, Solar Tech, Japan) for the determination of the
sures up to the upper critical solution temperature (UCST), cloud points is placed away from the incident path of the
as a continuation of our previous studies. The experimental light beam so that the low intensity scattered light, which
solubility data were correlated by three local composition arises at the beginning of turbidity formation, can be de-
tected effectively. A thermo-regulator temperature control
∗ Corresponding author. Tel.: +81 3 3259 0793; fax: +81 3 3259 7572. system (Model UT/350, Yokokawa Electric Co. Ltd.) was
E-mail address: [email protected] (K. Ochi). adopted for the purpose of shortening the measurement time

0378-3812/$ – see front matter © 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.fluid.2004.05.006
32 H. Matsuda, K. Ochi / Fluid Phase Equilibria 224 (2004) 31–37

and improving the measurement accuracy. The equilibrium

cell was covered with a film heater for heating.
The liquid mole fraction was determined by mass with a
Mettler digital balance (Model H315 having a sensitivity of
0.1 mg and a maximum load of 1 kg), and the reproducibility
of the temperature of the cloud point was within ±0.1 K
over the whole mole fraction range.

2.2. Materials

Special grade methanol, ethanol, decane, tetradecane, and

hexadecane were supplied by Wako Pure Chem. Ind. Ltd.,
Japan. All chemicals were stored over molecular sieves in
order to reduce the water content. The purity and water
content were checked by gas chromatography. The purity
was established to be better than 99.9 mol%, and the water
content was less than maximum 5 × 10−2 mol% for all five
chemicals.
Fig. 1. Liquid–liquid equilibrium data for the system methanol (1)
+ decane (2).
3. Results and discussion

3.1. Methanol (1) + decane (2) et al. [14] for temperatures from 303 to 333 K, and by Casás
et al. [15] for 278–308 K. On the methanol-rich side, our
The experimentally determined cloud points for the sys- data agree reasonably well with the three literature data sets
tem methanol (1) + decane (2) are listed in Table 1, and a at temperatures lower than 320 K. On the other hand, on the
comparison of these data with the literature values is shown decane-rich side, some differences between our data and the
graphically in Fig. 1. The UCST determined in this study literature data are seen. Furthermore, no data near the UCST
is marked with the symbol ⊗ in this figure. This UCST are available in the literature. As the behavior of the sys-
was interpolated from the experimental cloud points near tem in the vicinity of the UCST still requires experimental
the UCST. In this system, Kiser et al. [13] have measured verification, the cloud points were measured at small com-
the cloud points for temperatures from 278 to 313 K by vi- position increments.
sual determination, and mutual solubility data determined by
means of gas chromatography were reported by Higashiuchi 3.2. Ethanol (1) + tetradecane (2)

The experimental solubility data for the ethanol (1)

Table 1 + tetradecane (2) system are summarized in Table 2, and
Experimental cloud point data for the system methanol (1) + decane (2) Fig. 2 illustrates these data compared with the values re-
ported by Dahlmann and Schneider [16]. It can be seen that
T (K) x1 T (K) x1 T (K) x1

290.24 0.0560 360.68 0.5418 348.28 0.9258

299.98 0.0834 361.55 0.5645 342.45 0.9372 Table 2
306.17 0.0969 362.97 0.6184 341.80 0.9380 Experimental cloud point data for the system ethanol (1) + tetradecane
310.00 0.1103 363.16 0.6391 338.46 0.9428 (2)
312.72 0.1226 363.70 0.6635 333.67 0.9487 T (K) x1 T (K) x1 T (K) x1
315.60 0.1310 363.56 0.6968 330.98 0.9519
321.94 0.1628 363.76 0.7265 323.92 0.9583 279.96 0.2544 304.54 0.5308 306.13 0.8360
328.61 0.2032 363.91 0.7475 316.10 0.9655 280.98 0.2596 305.85 0.5635 304.83 0.8550
331.04 0.2228 363.92a 0.7510a 312.96 0.9666 283.63 0.2735 306.94 0.6136 303.49 0.8676
334.95 0.2544 363.69 0.7717 309.53 0.9691 286.71 0.2906 307.36 0.6400 300.77 0.8856
339.12 0.2810 363.58 0.7846 303.94 0.9704 288.23 0.3061 307.48 0.6644 299.81 0.8917
341.87 0.3024 363.54 0.8068 299.13 0.9742 290.69 0.3310 307.67 0.6800 296.99 0.9027
345.06 0.3328 362.62 0.8435 295.42 0.9762 292.67 0.3541 307.71 0.6994 293.14 0.9153
348.01 0.3602 360.39 0.8725 290.98 0.9781 295.06 0.3820 307.80 0.7188 291.65 0.9193
352.49 0.4109 357.00 0.8976 285.69 0.9800 296.50 0.4010 307.81a 0.7365a 288.84 0.9258
355.09 0.4438 352.04 0.9149 282.40 0.9812 298.70 0.4290 307.70 0.7556 284.92 0.9337
358.30 0.4816 350.33 0.9202 281.52 0.9816 301.01 0.4645 307.66 0.7748 282.20 0.9382
359.60 0.5089 348.56 0.9253 277.74 0.9829 302.93 0.4933 307.14 0.8060 279.89 0.9414
a Upper critical solution point. a Upper critical solution point.
 H. Matsuda, K. Ochi / Fluid Phase Equilibria 224 (2004) 31–37 33

Table 3
Experimental cloud point data for the system ethanol (1) + hexadecane
(2)
T (K) x1 T (K) x1

298.66 0.2946 326.92 0.7628

304.70 0.3440 326.72 0.7881
306.96 0.3613 326.56 0.8167
308.66 0.3756 326.21 0.8414
312.24 0.4082 325.97 0.8523
313.66 0.4228 325.26 0.8665
318.99 0.4936 324.61 0.8749
321.91 0.5442 323.02 0.8904
322.50 0.5571 322.90 0.8914
323.80 0.5859 320.21 0.9070
324.47 0.6089 319.04 0.9170
325.59 0.6425 314.96 0.9293
326.06 0.6643 310.60 0.9437
326.40 0.6833 304.11 0.9528
326.67 0.7007 295.37 0.9595
326.87 0.7307 293.88 0.9613
326.98a 0.7465a 290.12 0.9657
Fig. 2. Liquid–liquid equilibrium data for the system ethanol (1) a
+ tetradecane (2). Upper critical solution point.

our experimentally determined values near the UCST are non-randomness parameter α12 , so that the relationship be-
somewhat lower than the literature values. The apparatus tween the binary interaction parameter gij − gjj and tem-
by Dahlmann and Schneider focuses on the measurement perature is linear in nature [9]. On the other hand, in the
LLE over a large pressure range (0.1–120 MPa), and the modified CLC equation, the constant value α12 = 1.0 was
measurement is performed by visual determination. There- used for the three systems, according to the recommendation
fore, our experimental LLE data at atmospheric pressure by Nagata and Tamura [12]. Furthermore, in T–K–Wilson
are likely to be more precise than the data by Dahlmann and modified CLC equations, the value of the liquid molar
and Schneider. Furthermore, we measured the cloud points volume,vLi , was calculated by means of the modified Rack-
at temperatures from 280 K up to the UCST (307.81 K) in ett equation [20]:
more detail, because data at temperatures lower than 298 K
RTc,i [1+(1−Tr,i )2/7]
is not available for this system. vLi = Z (1)
Pc,i RA,i
3.3. Ethanol (1) + hexadecane (2)

Table 3 and Fig. 3 report our experimental cloud point

values for this system and a comparison of these data with
literature values. The LLE data in the vicinity of the UCST
was measured by Dahlmann and Schneider [16], as well as
that for the ethanol + tetradecane system. French et al. [17]
have also reported mutual solubility data, by visual deter-
mination at 325.25 K, and from the excess molar enthalpy
data in the two-phase region measured by them from 298 to
318 K. Serious differences are apparent between the two lit-
erature data sets at temperatures above 315 K. As shown by
our previous studies, accurate determination of cloud points
is possible by combining the light sensor and delicate tem-
perature control using the thermo regulator, especially in the
vicinity of the UCST.

3.4. Data reduction

The experimental solubility data were correlated with

NRTL, T–K–Wilson, and modified CLC equations. In the Fig. 3. Liquid–liquid equilibrium data for the system ethanol (1)
NRTL equation, a constant value 0.4 was adopted for the + hexadecane (2).
34 H. Matsuda, K. Ochi / Fluid Phase Equilibria 224 (2004) 31–37

Table 4
Pure component parameters used in the calculation of LLE and VLE
Tc a (K) Pc a (MPa) Vc a (m3 mol−1 ) ωa ZRA

Methanol 512.64 8.097 118 × 10−6 0.565 0.23230b

Ethanol 513.92 6.148 167 × 10−6 0.649 0.25041b
Decane 617.70 2.110 624 × 10−6 0.490 0.25074b
Tetradecane 693.00 1.570 894 × 10−6 0.644 0.24322b
Hexadecane 723.00 1.400 1034 × 10−6 0.718 0.22545c
Tsonopoulos ad Tsonopoulos bd Antoinee Aa Antoinee Ba Antoinee Ca

Methanol – – – – –
Ethanol 0.0878 0.0572 7.33675 1648.220 −42.232
Decane – – – – –
Tetradecane – – – – –
Hexadecane 0 0 6.15357 1830.510 −118.7
a
Poling et al. [19].
b
Spencer and Adler [20].
c The value of Z
RA was not available in the table of Spencer and Adler. Therefore, it was estimated using the following equation proposed by Yamada
and Gunn [21]: ZRA = 0.29056 − 0.08775ω.
d Tsonopoulos et al. [22].
e log P S (kPa) = A − B/(T (K) + C).
i

The pure-component critical constants required in Eq. (1) obtained using the NRTL equation are shown graphically in
were obtained from Poling et al. [19], and the value of ZRA,i Figs. 1–3.
was taken from Spencer and Adler [20]. These parameters
are listed in Table 4.
The following polynomial expression of Tc − T were 4. Prediction of vapor–liquid equilibria
adopted for the expression of the temperature dependencies
of the binary interaction parameters gij − gjj for NRTL and In this study, an attempt was made to examine the capa-
modified CLC, and λij − λii for T–K–Wilson: bility for predicting the isothermal VLE by using the pa-
NRTL and modified CLC equations: rameters determined from the LLE data. The system ethanol
+ hexadecane was tested for the prediction of the VLE,
gij − gjj = Aij + Bij (Tc − T) + Cij (Tc − T)2 because it was only this system that the isothermal VLE
+ Dij (Tc − T)3 (2) data with the partially miscible were available (reported by
French et al. [17], and Pierotti et al. [18]), in three binary
T–K–Wilson equation:

λij − λii = Aij + Bij (Tc − T) + Cij (Tc − T)2

+ Dij (Tc − T)3 (3)

where Tc is the UCST. The degree of this polynomial equa-

tion was set to three after considering the correlation ac-
curacy of the solubility data. The coefficients of Eqs. (2)
and (3) were estimated by a previously reported method [1].
The LLE compositions necessary for the reduction of the
data were obtained by smoothing of the experimental cloud
points.
The estimated parameters for Eqs. (2) and (3) and the
average deviations between the calculated and experimental
mutual solubility data are given in Table 5. In the methanol
+ decane system, the results yielded by the NRTL equa-
tion showed a better correlation with the experimental
data than those yielded by the other two equations. On
the other hand, in the ethanol + tetradecane system, the
T–K–Wilson equation was more accurate, while the corre-
lation accuracy by means of this equation was inferior in Fig. 4. Predicted results for vapor–liquid equilibria by three LC models
the ethanol + hexadecane system. The calculated results for the system ethanol (1) + hexadecane (2).
Table 5
Determined parameters and deviations between calculated and experimental cloud points for the three binary systems using the three local composition models
NRTL T–K–Wilson Modified CLC

H. Matsuda, K. Ochi / Fluid Phase Equilibria 224 (2004) 31–37

ij = 12 ij = 21 ij = 12 ij = 21 ij = 12 ij = 21

Methanol (1) + decane (2) system

Aij (J mol−1 ) 7.055479 × 103 1.815871 × 103 6.724470 × 103 −3.685977 × 103 2.128869 × 103 4.237940 × 102
Bij (J mol−1 K−1 ) 2.640328 × 101 1.452246 × 102 1.815153 × 101 5.283316 × 101 −5.656397 3.590130 × 101
−1
Cij (J mol K ) −2 −2.781241 × 10 −1 −1.878796 3.564183 × 10−1 −7.313433 × 10−1 1.185100 × 10−2 −4.236640 × 10−1
Dij (J mol−1 K−3 ) 1.957136 × 10−3 1.122525 × 10−2 −2.017083 × 10−3 5.335632 × 10−3 −7.097849 × 10−7 2.699270 × 10−3
|x1 |av. a 0.0095 0.0099 0.0104
Ethanol (1) + tetradecane (2) system
Aij (J mol−1 ) 5.789005 × 103 1.769883 × 103 5.759423 × 103 −2.758062 × 103 1.660716 × 103 5.385580 × 102
Bij (J mol−1 K−1 ) 1.711979 × 101 1.707391 × 102 −9.587510 × 101 1.080870 × 102 7.647269 1.308216 × 101
Cij (J mol−1 K−2 ) −3.233318 × 10−1 −6.227185 9.962493 −6.356456 −9.948678 × 10−1 6.726169 × 10−1
Dij (J mol−1 K−3 ) 8.925731 × 10−3 1.003282 × 10−1 −2.302148 × 10−1 1.370958 × 10−1 2.306137 × 10−3 −2.306939 × 10−2
|x1 |av. a 0.0072 0.0043 0.0056
Ethanol (1) + hexadecane (2) system
Aij (J mol−1 ) 6.278592 × 103 1.710038 × 103 5.257378 × 103 −2.618803 × 103 1.877185 × 103 3.705450 × 102
−1
Bij (J mol K ) −1 1.145605 × 10 2 5.523656 × 101 −1.679662 × 102 2.044818 × 102 2.831864 × 101 −1.655825 × 101
Cij (J mol−1 K−2 ) −5.282300 1.917443 1.128167 × 101 −9.883384 −1.835401 2.120960
Dij (J mol−1 K−3 ) 7.984521 × 10−2 −4.764711 × 10−2 −1.845418 × 10−1 1.529282 × 10−1 2.934095 × 10−2 −3.805731 × 10−2
|x1 |av. a 0.0056 0.0082 0.0057


NDP 

a |x |
1 av. =

k=1 x1,exptl − x1,calct k /NDP, where NDP is the number of data points.

35
36 H. Matsuda, K. Ochi / Fluid Phase Equilibria 224 (2004) 31–37

mixtures examined in this study. Thus, we calculated the List of symbols

VLE of this binary system at temperatures from 298.15 to Aij parameters of Eqs. (2)–(3) (J mol−1 )
325.25 K using the NRTL, T–K–Wilson, and modified CLC Bij parameters of Eqs. (2)–(3) (J mol−1 K−1 )
equations with the parameters estimated from the LLE data. Cij parameters of Eqs. (2)–(3) (J mol−1 K−2 )
Liquid phase activity coefficients γi were calculated from Dij parameters of Eqs. (2)–(3) (J mol−1 K−3 )
the following expression. gij − gjj binary interaction parameters of NRTL and
  modified CLC equations (J mol−1 )
vLi (P − PiS )
Pyi ϕ̂i = γi xi Pi ϕi exp
S S
(4) NDP number of data points per system
RT P pressure
R gas constant = 8.314 (J mol−1 K−1 )
In this study, the fugacity coefficient of pure components
T absolute temperature (K)
in the vapor phase ϕ̂i , and that of pure vapor i, ϕiS , were
v molar volume (m3 mol−1 )
determined from the second virial coefficient correlation of
x liquid mole fraction
Tsonopoulos [22]. The vapor pressures of the pure compo-
|x1 |av. average absolute deviation in the liquid mole
nents, PiS , were calculated by the Antoine equation. The An-
fraction
toine constants were taken from Poling et al. [19]. The liquid
y vapor mole fraction
molar volume, vLi , was calculated by the modified Rackett
ZRA modified Rackett equation parameter
equation. Pure component parameters for the calculation of
the VLE are listed in Table 4.
Greek letters
The predicted VLE for the system ethanol + hexadecane
α12 non-randomness parameter of NRTL and
by three LC models based on the LLE data are com-
modified CLC equations
pared with the experimental values in Fig. 4. The
γ activity coefficient
vapor–liquid–liquid equilibria (VLLE) were also predicted
λij − λii binary interaction parameter of T–K–Wilson
using the binary parameters obtained from the LLE data.
equation (J mol−1 )
The calculated results are graphically shown with the sym-
ϕ fugacity coefficient
bols (䊊, , and 䊐) in Fig. 4. At 298.15 and 313.15 K, all
ω acentric factor
three LC models were capable of reproducing the VLE
qualitatively. However, at 325.25 K, which is in the vicinity
Subscripts
of the UCST (326.98 K), the calculated results were poor,
1, 2, i, j components 1, 2, i, and j
especially for ethanol mole fractions of 0.0–0.6. This lack
c critical
of predictive accuracy in this mole fraction range is con-
calct calculated value
sidered as depending on a reason that the LLE data of this
exptl experimental value
system, which is the basis for the binary parameters of LC
r reduced value
models, are shift to the ethanol-rich side. The calculated
values obtained using the NRTL equation were better than
Superscripts
those obtained using the other two equations.
L liquid phase
S saturated condition
ˆ partial molar property
5. Conclusion

Liquid–liquid equilibria were measured for the systems

methanol + decane, ethanol + tetradecane, and ethanol References
+ hexadecane, using a laser scattering technique. LLE data
were taken for temperatures up to the UCST (the determined [1] K. Ochi, M. Tada, K. Kojima, Fluid Phase Equilib. 56 (1990) 341–
UCSTs were Tc = 363.92 K and x1,UCST = 0.7510 for 359.
methanol + decane, Tc = 307.81 K and x1,UCST = 0.7365 [2] K. Ochi, M. Tada, K. Kojima, B.C.-Y. Lu, Can. J. Chem. Eng. 71
for ethanol + tetradecane, and Tc = 326.98 K and x1,UCST (1993) 982–985.
[3] K. Ochi, T. Saito, K. Kojima, J. Chem. Eng. Data 41 (1996) 361–364.
= 0.7465 for ethanol + hexadecane). LLE data were corre-
[4] K. Ochi, Y. Kato, T. Saito, K. Kurihara, K. Kojima, Korean J. Chem.
lated satisfactorily with the NRTL, T–K–Wilson, and modi- Eng. 14 (1997) 365–368.
fied CLC equations. Finally, the VLE for the system ethanol [5] K. Ochi, T. Saito, K. Kojima, Fluid Phase Equilib. 158–160 (1999)
+ hexadecane at temperatures from 298.15 to 325.25 K were 847–855.
predicted using three equations, with the adjusted parame- [6] K. Kurihara, T. Midorikawa, T. Hashimoto, K. Kojima, K. Ochi, J.
Chem. Eng. Jpn. 35 (2002) 360–364.
ters obtained from the LLE data. This prediction was rela-
[7] H. Matsuda, K. Kurihara, K. Ochi, K. Kojima, Fluid Phase Equilib.
tively successful at temperatures from 298.15 to 313.15 K. 203 (2002) 269–284.
However, in the vicinity of the UCST (325.25 K), the cal- [8] H. Matsuda, K. Ochi, K. Kojima, J. Chem. Eng. Data 48 (2003)
culated values differed from the experimental VLE data. 184–189.
 H. Matsuda, K. Ochi / Fluid Phase Equilibria 224 (2004) 31–37 37

[9] H. Matsuda, M. Fujita, K. Ochi, J. Chem. Eng. Data 48 (2003) [16] U. Dahlmann, G. Schneider, J. Chem. Thermodyn. 21 (1989) 997–
1076–1080. 1004.
[10] H. Renon, J.M. Prausnitz, AIChE J. 14 (1968) 135–144. [17] H.T. French, A. Richards, R.H. Stokes, J. Chem. Thermodyn. 11
[11] T. Tsuboka, T. Katayama, J. Chem. Eng. Jpn. 8 (1975) 181–187. (1979) 95–102.
[12] I. Nagata, K. Tamura, Fluid Phase Equilib. 135 (1997) 209–226. [18] G.J. Pierotti, C.H. Deal, E.L. Derr, J. Ind. Eng. Chem. 51 (1959)
[13] R.W. Kiser, G.D. Johnson, M.D. Sheltar, J. Chem. Eng. Data 6 95–102.
(1961) 338–341. [19] B.E. Poling, J.M. Prausnitz, J.P. O’Connell, The Properties of Gases
[14] H. Higashiuchi, Y. Sakuragi, Y. Iwai, Y. Arai, M. Nagatani, Fluid and Liquids, fifth ed., McGraw-Hill, New York, 2001.
Phase Equilib. 36 (1987) 35–47. [20] C.F. Spencer, S.B. Adler, J. Chem. Eng. Data 23 (1978) 82–89.
[15] L.M. Casás, A. Tourino, B. Orge, G. Marino, M. Iglesias, J. Tojo, [21] T. Yamada, R.D. Gunn, J. Chem. Eng. Data 18 (1973) 234–236.
J. Chem. Eng. Data 47 (2002) 887–893. [22] C. Tsonopoulos, AIChE J. 20 (1974) 263–272.