J. Chem. Eng.

Data 2002, 47, 923-926 923

Liquid-Liquid Equilibria for Binary Systems Containing

N-Formylmorpholine

MinSu Ko,† Sungjin Lee,‡ Jungho Cho,§ and Hwayong Kim*,†

School of Chemical Engineering & Institute of Chemical Processes, Seoul National University, Shinlim-dong,
Kwanak-gu, Seoul 151-744, Korea, Department of Liberal Arts and Science, Semyung University, Sinwol-dong,
Jechon, Chungbuk 390-711, Korea, and Department of Chemical Engineering, Dongyang University,
1 Kyo Chon-dong, Pung-gi up, Young-ju City, Kyoungbuk 750-711, Korea

Liquid-liquid equlilibrium (LLE) data were measured for four binary systems containing N-formylmor-
pholine and alkanes (i.e. pentane, hexane, heptane, and octane) over the temperature range 298 K to
413 K using circulation type equipment with an equilibrium view cell. The compositions of both light
and heavy phases were analyzed by on-line gas chromatography. The mutual solubility increased as the
temperature increased for all these systems. The binary liquid-liquid equilibrium data were correlated
with the NRTL and UNIQUAC equations using temperature-dependent parameters. Both models correlate
the experimental data well. The solubility in N-formylmorpholine increased in the following order at the
same temperature: pentane, hexane, heptane, and octane.

Introduction Table 1. Suppliers and Purities of the Used Chemicals

chemical supplier spec %a purityb/%
There have been ever-increasing demands for high-purity
aromatic compounds as a feedstock for chemical synthesis. NFM ACROS 99+ 99.99+
Many solvents such as sulfolane,1-5 N-methylpyrrolidone,6 NFM Samsung Chemical Co., Ltd. 98+ 99.14+
pentane Kanto Chemical Co., Ltd. 99+ 99.99+
glycol,7,8 and N-formylmorpholine9-11 are used to extract hexane Junsei Chemical Co., Ltd. 99.5 99.99+
aromatics such as benzene, toluene, and xylene from heptane Aldrich 99+ 99.93+
hydrocarbon mixtures. octane ACROS 99+ 99.94+
The N-formylmorpholine extractive distillation process a The purity reported by the supplier. b The purity determined
separates aromatics from the reformates. It minimizes as an area ratio by gas chromatography with a thermal conductiv-
aromatic content in gasoline and refines the extracted ity detector.
aromatics, which are used as raw materials for petrochemi-
cal processes. This process can recover most of the benzene. Experimental Section
The solvent-to-feed ratio by mass of this process is around
3.5. The operating cost and the initial equipment cost for Chemicals. The suppliers and purities of the chemicals
the overall plant can be reduced by optimization. Com- are listed in Table 1 together with the purities determined
mercial process simulators such as PROΠ(SimSci), HYSYS using a HP 5890 gas chromatograph with a thermal
(Hyprotech), and ASPEN-PLUS (Aspentech)12 have been conductivity detector. The N-formylmorpholine supplied by
used to determine the optimum solvent rate. However, Samsung Chemical Co., Ltd., was used after distillation
liquid-liquid equilibrium (LLE) data are not available for and filtration treatment through the 0.45 µm membrane.
these mixtures. The rest of the chemicals were used without purification.
Experimental Apparatus. The sampling parts of the
In this work, a newly designed circulating apparatus recirculation type apparatus used in the previous studies
equipped with a equilibrium view cell and connected were modified and used in this study.4,5 A schematic
directly to a gas chromatograph has been constructed to diagram of the apparatus is shown in Figure 1. The system
obtain the equilibrium T, x, x data for the N-formylmor- consists of eight major parts: an equilibrium cell, a
pholine (NFM) and n-alkane (pentane, hexane, heptane, magnetic pump for circulation, one vapor and two liquid
and octane) systems in the operating temperature range sampling valves, a constant temperature air bath, a
of the NFM extractive distillation process. LLE data for magnetic stirrer, a charge-discharge device, temperature
the four binary systems were measured in the temperature and pressure measurement devices, and an analysis sys-
range from 298 K to 413 K, and these data were compared tem. The volume of the equilibrium cell was 240 cm3. The
with the existing data at different temperatures (293 K to cell was made of stainless steel (SUS. 316) and placed
333 K).10,11 Experimental data were correlated with the inside the air bath, controlled by a PID temperature
NRTL13 and UNIQUAC14 models with the function of controller to the desired temperature within (0.1 K
temperature-dependent parameters. (maximum temperature 573.2 K). The cell and the quartz
window were sealed with PTFE (Teflon) gaskets. The
* To whom correspondence should be addressed. Fax: +82-2-888-6695. mixing was promoted by the magnetic stirrer and magnetic
E-mail: [email protected]. recirculation pumps. The sampling system consists of a
† Seoul National University.
‡ Semyung University. sample valve, a sample loop, and a pneumatic actuator.
§ Dongyang University. The sample injection valve was a two-position rotary
10.1021/je010328n CCC: $22.00 © 2002 American Chemical Society
Published on Web 05/10/2002
924 Journal of Chemical and Engineering Data, Vol. 47, No. 4, 2002

Table 3. Experimental and Calculated LLE Data for the

Hexane (1) + N-Formylmorpholine (2) System
top phase, x11 bottom phase, x12
T/K exp NRTL UNIQUAC exp NRTL UNIQUAC
299.46 0.9865 0.9853 0.9850 0.0311 0.0319 0.0330
306.31 0.9826 0.9833 0.9832 0.0346 0.0337 0.0340
314.24 0.9797 0.9808 0.9809 0.0365 0.0360 0.0355
324.31 0.9756 0.9770 0.9774 0.0408 0.0395 0.0384
333.24 0.9737 0.9733 0.9737 0.0428 0.0430 0.0417
344.01 0.9689 0.9680 0.9684 0.0452 0.0481 0.0469
353.74 0.9641 0.9623 0.9626 0.0523 0.0536 0.0529
363.47 0.9536 0.9557 0.9557 0.0589 0.0603 0.0606
374.03 0.9488 0.9470 0.9464 0.0730 0.0694 0.0711
383.69 0.9363 0.9372 0.9360 0.0848 0.0801 0.0835
394.37 0.9252 0.9232 0.9215 0.0947 0.0963 0.1010
402.95 0.9060 0.9079 0.9067 0.1135 0.1154 0.1192
413.33 0.8785 0.8792 0.8835 0.1597 0.1600 0.1479

Table 4. Experimental and Calculated LLE Data for the

Heptane (1) + N-Formylmorpholine (2) System
Figure 1. Flow diagram of the apparatus: 1, vacuum line for
the removal of air with a vacuum pump; 2, bottle for charging a top phase, x11 bottom phase, x12
reagent; 3, air line (8 kgf/cm2) for the leak test; 4, equilibrium cell T/K exp NRTL UNIQUAC exp NRTL UNIQUAC
with the view window (quartz); 5, magnetic pump (chromium
plating); 6, six-way valve for vapor; 7, discharge line with vacuum; 301.12 0.9889 0.9903 0.9902 0.0244 0.0249 0.0251
8, magnetic stirrer; 9, sampling loop for the light phase; 10, 311.31 0.9871 0.9864 0.9864 0.0256 0.0263 0.0262
sampling loop for the heavy phase; 11, circulating minipump; 12, 319.37 0.9847 0.9826 0.9826 0.0288 0.0278 0.0276
temperature controller (domain of preheating carrier line); 13, 329.08 0.9801 0.9769 0.9770 0.0310 0.0301 0.0299
339.00 0.9658 0.9696 0.9698 0.0339 0.0331 0.0330
temperature controller (domain of circulating fluid line).
343.75 0.9637 0.9656 0.9657 0.0350 0.0348 0.0347
Table 2. Experimental and Calculated LLE Data for the 358.43 0.9463 0.9504 0.9506 0.0380 0.0412 0.0413
Pentane (1) + N-Formylmorpholine (2) System 368.50 0.9347 0.9373 0.9374 0.0481 0.0469 0.0472
373.48 0.9320 0.9299 0.9300 0.0512 0.0502 0.0505
top phase, x11 bottom phase, x12 378.02 0.9213 0.9226 0.9226 0.0545 0.0535 0.0539
T/K exp NRTL UNIQUAC exp NRTL UNIQUAC 387.64 0.9121 0.9051 0.9050 0.0605 0.0617 0.0620
398.85 0.8774 0.8809 0.8806 0.0701 0.0736 0.0736
304.49 0.9944 0.9925 0.9934 0.0397 0.0376 0.0380 403.46 0.8753 0.8696 0.8691 0.0786 0.0795 0.0792
308.16 0.9907 0.9917 0.9924 0.0415 0.0399 0.0400 412.16 0.8422 0.8456 0.8451 0.0969 0.0924 0.0913
313.07 0.9905 0.9904 0.9908 0.0432 0.0432 0.0430
317.88 0.9901 0.9890 0.9891 0.0455 0.0467 0.0463 Table 5. Experimental and Calculated LLE Data for the
322.86 0.9895 0.9874 0.9871 0.0497 0.0506 0.0501 Octane (1) + N-Formylmorpholine (2) System
327.57 0.9856 0.9857 0.9850 0.0555 0.0545 0.0539
top phase, x11 bottom phase, x12
333.28 0.9813 0.9833 0.9822 0.0573 0.0596 0.0592
338.01 0.9771 0.9810 0.9796 0.0606 0.0643 0.0639 T/K exp NRTL UNIQUAC exp NRTL UNIQUAC
342.83 0.9748 0.9785 0.9767 0.0650 0.0693 0.0692
297.00 0.9850 0.9842 0.0120 0.0143 0.0148
348.02 0.9711 0.9754 0.9734 0.0717 0.0752 0.0753
305.67 0.9811 0.9808 0.0171 0.0165 0.0166
352.23 0.9680 0.9726 0.9705 0.0827 0.0803 0.0807
308.20 0.9730 0.9798 0.9797 0.0214 0.0172 0.0172
357.79 0.9654 0.9684 0.9665 0.0899 0.0877 0.0883
313.10 0.9729 0.9772 0.9775 0.0197 0.0187 0.0184
363.45 0.9643 0.9635 0.9621 0.0952 0.0958 0.0967
322.80 0.9712 0.9716 0.9724 0.0205 0.0218 0.0213
368.48 0.9591 0.9586 0.9579 0.1022 0.1038 0.1048
332.68 0.9634 0.9651 0.9664 0.0233 0.0255 0.0248
373.60 0.9516 0.9529 0.9535 0.1128 0.1126 0.1136
342.48 0.9615 0.9580 0.9595 0.0326 0.0297 0.0290
378.41 0.9496 0.9469 0.9492 0.1262 0.1217 0.1224
353.22 0.9530 0.9492 0.9505 0.0342 0.0351 0.0348
383.17 0.9468 0.9401 0.9448 0.1376 0.1315 0.1317
362.83 0.9463 0.9406 0.9411 0.0403 0.0408 0.0411
388.10 0.9455 0.9323 0.9401 0.1494 0.1426 0.1419
372.50 0.9351 0.9309 0.9301 0.0457 0.0477 0.0487
switching valve with an internal sample loop of 0.5 µL. This 384.17 0.9161 0.9175 0.9143 0.0573 0.0579 0.0603
392.83 0.8990 0.9058 0.9003 0.0683 0.0678 0.0710
was connected with the pneumatic actuator that actuates 402.27 0.8806 0.8893 0.8824 0.0800 0.0820 0.0852
positions. The actuator was operated with a solenoid valve. 413.66 0.8523 0.8553 0.8560 0.1127 0.1102 0.1072
The vacuum pump was used for discharge of the mixture. 416.18 0.8452 0.8389 0.8493 0.1211 0.1203 0.1129
The temperature was measured using a sheath type 100
Ω-platinum resistance thermometer. Its uncertainty was of 25 K‚min-1. Helium was used as the carrier gas at the
estimated to be within 0.05 K. The sampling system was rate of 23 cm3‚min-1. Single-phase samples of known
connected to a gas chromatograph (Hewlett-Packard 5890 composition were used to calibrate the gas chromatograph
Series Π) which had a thermal conductivity detector (TCD) in the composition range of interest. The analysis for the
and a 1.828 8 m × 0.003 175 m column packed with calibration indicated that a precision of (3 × 10-5 area
Chromosorb WHP 100/120 coated with OV-101. The liquid- ratio was obtained for each composition. The samples of
phase compositions were determined with a maximum each phase at the same temperature were analyzed three
relative error of 0.1%. times at least, and the average values were used. The
Experimental Procedure. The mixture was fed into the average deviation of the area ratio is (3 × 10-5.
equilibrium cell that was initially evacuated. The mixture
was stirred for at least 1 h with the magnetic stirrer and Results and Discussion
then left to settle for at least 2 h. The samples were The experimental liquid-liquid equilibrium data for the
analyzed by gas chromatography. The temperatures of the systems pentane (1) + N-formylmorpholine (2), hexane (1)
injector and the detector were maintained at 523.15 K. + N-formylmorpholine (2), heptane (1) + N-formylmorpho-
After 1 min of holding at 353.15 K, the column temperature line (2), and octane (1) + N-formylmorpholine (2) are given
was raised to the final temperature of 423.15 K at the rate in Tables 2-5 and Figures 2-5. Concentrations of compo-
 Journal of Chemical and Engineering Data, Vol. 47, No. 4, 2002 925

Figure 2. Experimental and calculated liquid-liquid equilibrium Figure 5. Experimental and calculated liquid-liquid equilibrium
data for the pentane (1) + N-formylmorpholine (2) mixture: points, data for the octane (1) + N-formylmorpholine (2) mixture: points,
experimental results; solid line, NRTL with R ) 0.3; dashed line, experimental results; solid line, NRTL with R ) 0.3; dashed line,
UNIQUAC. UNIQUAC.

Table 6. Temperature Dependence of the Parameters of

the NRTL Equation for the n-Alkane (1) +
N-Formylmorpholine (2) Binary Systemsa
parameter pentane hexane heptane octane
a12 -3.462 66 -10.695 26 -62.593 35 -27.818 91
a21 -1.561 02 37.690 74 71.964 65 40.841 01
b12/K 2240.924 69 1555.790 02 5137.815 99 3089.473 48
b21/K 1244.168 79 -1440.303 61 -2902.831 30 -719.764 15
d21/K-1 0.000 34 0.000 89 0.000 50 -0.000 42
e12/(ln K)-1 0.002 79 1.553 54 8.621 37 3.607 94
e21/(ln K)-1 -0.000 55 -5.292 13 -10.415 88 -6.189 97

a c12 ) c21 ) 0.3; d12/K-1 ) 0.

equilibrium data were correlated with the NRTL and

UNIQUAC equations.
The binary parameters have the temperature depen-
dence function

Figure 3. Experimental and calculated liquid-liquid equilibrium for the NRTL model:
data for the hexane (1) + N-formylmorpholine (2) mixture: points,
experimental results; solid line, NRTL with R ) 0.3; dashed line, τij ) aij + bij/T + eij ln T (1)
UNIQUAC.
Rij ) cij + dij(T - 273.15 K) (2)

where τij is dimensionless and T is in kelvin.

for the UNIQUAC model:

τij ) exp(aij + bij/T + cij ln T) (3)

The binary parameters for the models were evaluated

by a nonlinear regression method based on the maximum-
likelihood principle. The objective function was minimized
during optimization of the parameters in each of the
equations:

{(
Tkcal - Tkexp
) ( )}
2
N 2 2 xijcal - xijexp 2

OF ) ∑
k)1 σTk
+ ∑∑
i)1 j)1 σxijk
(4)

where N is the number of experimental data points in each

Figure 4. Experimental and calculated liquid-liquid equilibrium group k.
data for the heptane (1) + N-formylmorpholine (2) mixture: points, The four resulting NRTL and UNIQUAC parameters
experimental results; solid line, NRTL with R ) 0.3; dashed line,
optimized for the liquid-liquid equilibrium data are listed
UNIQUAC.
in Tables 6 and 7. The absolute average deviations of the
nents i in phase j (j ) 1, top phase; j ) 2, bottom phase) NRTL and UNIQUAC equations are listed in Table 8. The
are given in the mole fraction xij. The binary liquid-liquid NRTL model showed slightly smaller deviations in the
926 Journal of Chemical and Engineering Data, Vol. 47, No. 4, 2002

Table 7. Temperature Dependence of the Parameters of N-formylmorpholine + n-alkane systems (i.e. pentane,
the UNIQUAC Equation for the n-Alkane (1) + hexane, heptane, and octane) from experimental binary
N-Formylmorpholine (2) Binary Systems liquid-liquid equilibrium data at 331.15 K up to 413.15
parameter pentane hexane heptane octane K. The temperature dependency of the parameters of NRTL
a12 59.653 86 8.355 16 23.909 00 14.715 76 and UNIQUAC models can be expressed by eqs 1 and 3,
a21 -33.994 93 -34.691 92 -18.803 55 -20.092 80 respectively. Thus, a quantitative description of liquid-
b12/K -3501.076 52 -768.568 26 -1910.536 76 -1084.385 20 liquid equilibria of industrial interest containing N-formyl-
b21/K 1549.879 87 1657.696 94 1105.737 43 885.630 03
morpholine and n-alkanes is available to accurately simu-
c12/(ln K)-1 -8.577 48 -1.142 97 -3.240 34 -2.077 96
c21/(ln K)-1 4.974 10 5.039 81 2.603 71 2.944 10 late and optimize the extractive distillation units where
these systems are involved.
Table 8. Absolute Average Deviations (AAD%)a of the
NRTL and UNIQUAC Equations for the n-Alkane (1) + Literature Cited
N-Formylmorpholine (2) Binary Systems
(1) Deal, C. H.; Evans, H. D.; Oliver, E. D.; Papadopoulus, M. N.
x11 x12 Extraction of Aromatics with Sulfolane. Fifth World Pet. Congr.
system NRTL UNIQUAC NRTL UNIQUAC Proc. 1959, 3, 283-297.
(2) Voetter, H.; Kosters, W. C. G. The Sulfolane Extraction Process.
pentane + NFM 0.30 0.19 3.19 3.06 Sixth World Pet. Congr. Proc. 1963, 3, 131-145.
hexane + NFM 0.14 0.17 2.72 3.85 (3) Broughton, D. B.; Asselin, G. F. Production of High Purity
heptane + NFM 0.33 0.33 2.90 3.08 Aromatics by the Sulfolane Process. Seventh World Pet. Congr.
octane + NFM 0.47 0.36 5.78 7.36 Proc. 1967, 4, 65-73.
(4) Sungjin, L.; Hwayong, K. Liquid-liquid Equiliubria for the
a AAD% ) (1/N)∑i)1N |(x1,cal - x1,exp)/x1,exp| × 100. Ternary Systems Sulfolane + Octane + Benzene, Sulfolane +
Octane + Toluene, and Sulfolane + Octane + p-Xylene. J. Chem.
Eng. Data. 1995, 40, 499-502.
(5) Sungjin, L.; Hwayong, K. Liquid-liquid Equiliubria for the
Ternary Systems Sulfolane + Octane + Benzene, Sulfolane +
Octane + Toluene, and Sulfolane + Octane + p-Xylene at
Elevated Temperatures. J. Chem. Eng. Data 1998, 43, 358-361.
(6) Muller, E. Use of N-Methylpyrrolidone for Aromatics Extraction.
Chem. Ind. 1973, 518-522.
(7) Symoniak, M. F.; Ganju, Y. N.; Vidueira, J. A. Plant Data for Tetra
Process. Hydrocarbon Process. 1981, 60, 139-142.
(8) Taher, A. A. S.; Emina, K. Measurement and Prediction of Phase
Equilibria in the Extraction of Aromatics from Naphtha Refor-
mate by Tetraethyleneglycol. Fluid Phase Equilib. 1996, 118,
271-285.
(9) Cinelli, E.; Noe, S.; Paret, G. Extract Aromatics with FM.
Hydrocarbon Process. 1972, 51, 141-144.
(10) Mohamed, A. Q.; Taher, A. A.; Mohamed, A. F. Liquid-Liquid
Equilibria in Some Binary and Ternary Mixtures with N-
Formylmorpholine. J. Chem. Eng. Data 1995, 40, 88-90.
(11) Cincotti, A.; Murru, M.; Cao, G.; Marongiu, B.; Masia, F.; Sannia,
M. Liquid-Liquid Equilibria of Hydrocarbons with N-Formyl-
morpholine. J. Chem. Eng. Data 1999, 44, 480-483.
(12) Oonk, H. A. J. Phase Theory, The thermodynamics of Heteroge-
neous Equilibria; Elsevier Scientific Publishing Company: Am-
sterdam, 1981.
Figure 6. Mutual solubility for the N-formylmorpholine and
(13) Renon, H.; Prausnitz, J. M. Local Compositions in Thermody-
nonaromatic mixtures. namic Excess Functions for Liquid Mixtures. AIChE J. 1968, 14,
bottom phase. However, in the top phase, the UNIQUAC 135-144.
model has smaller deviations. The mutual solubility in- (14) Abrams, D. S.; Prausnitz, J. M. Statistical Thermodynamics of
creased as the temperature increased in all these systems. Liquid Mixtures; A New Expression for the Excess Gibbs Energy
of Partly or Completely Miscible Systems. AIChE J. 1975, 21,
The solubility of n-paraffin in N-formylmorpholine, that is, 116-128.
the bottom phase, increased in the following order at the
same temperature: pentane, hexane, heptane, and octane.
The results are displayed in Figure 6. Received for review December 17, 2001. Accepted February 28,
2002. This work was supported by the Brain Korea 21 Program
supported by the Ministry of Education and LG-Caltex Research
Conclusions and by the National Research Laboratory (NRL) Program supported
by the Korea Institute of S&T Evaluation and Planning.
In this work, we have determined the interaction pa-
rameters of the NRTL and UNIQUAC equations for the JE010328N