Format And Type Fonts


 

 CCHHEEMMIICCAALL  EENNGGIINNEEEERRIINNGG  TTRRAANNSSAACCTTIIOONNSS  
 

VOL. 39, 2014 

A publication of 

 
The Italian Association 

of Chemical Engineering 

www.aidic.it/cet 
Guest Editors: Petar Sabev Varbanov, Jiří Jaromír Klemeš, Peng Yen Liew, Jun Yow Yong  

Copyright © 2014, AIDIC Servizi S.r.l., 

ISBN 978-88-95608-30-3; ISSN 2283-9216 DOI: 10.3303/CET1439032 

 

Please cite this article as: Steltenpohl P., Graczová E., 2014, Aromatics extraction. 2. economic aspects, Chemical 

Engineering Transactions, 39, 187-192  DOI:10.3303/CET1439032 

187 

Aromatics Extraction. 2. Economic Aspects 

Pavol Steltenpohl, Elena Graczová* 

Institute of Chemical and Environmental Engineering, Faculty of Chemical and Food Technology, Slovak University of 

Technology in Bratislava, Radlinského 9, 812 37 Bratislava, Slovakia 

elena.graczova@stuba.sk 

Besides the core-equipment design parameters, investment and operational costs of a separation unit are 

influenced by the down-stream separation apparatuses design. In case of aromatics liquid-phase 

extraction, down-stream processing is represented by the extraction solvent regeneration carried out in 

order to obtain pure products, i.e. components of the original mixture, and the solvent that is repeatedly 

used in the extractor. Solvent regeneration is typically done in a distillation column as the extraction 

solvent boiling point differs considerably from those of the original mixture components. The presence of a 

distillation/stripping column and the conditions applied in this equipment show a significant impact on the 

overall production line profitability. Moreover, purity of regenerated extraction solvents influences the 

separation efficiency of an extractor. Here, the impact of the extraction solvent nature on the production 

costs is discussed assuming ionic liquids 1-ethyl-3-methylimidazolium ethylsulfate ([EMim][ESO4]), 1-ethyl-

3-methylimidazolium bis(trifluoromethylsulphonyl)imide ([EMim][NTf2]), 3-methyl-N-butylpyridinium 

tetracyanoborate ([3-mebupy][B(CN)4)], and 1-ethyl-3-methylpyridinium bis(trifluoromethylsulphonyl)imide 

([EMpy][NTf2]), or sulphotane as an extraction solvent used for heptane–toluene mixture separation. 

1. Introduction 

Separation of aromatics from complex hydrocarbon mixtures is usually carried out using liquid-phase 

extraction, extractive distillation, or azeotropic distillation. The separation method selection depends on the 

aromatics content in the hydrocarbons feed (Weissermel and Arpe, 2003). 

Concerning liquid-phase extraction, several extractive solvents were proposed and used for the separation 

of aromatics from their mixtures with less polar hydrocarbons. Due to its high selectivity and fair extraction 

capacity, sulpholane became a benchmark of the extraction solvents for aromatics separation by the 

liquid-phase extraction (Rawat and Gulati, 1976). Pure aromatics stream is then obtained from the extract 

during the solvent regeneration in a distillation column. Non-aromatics are obtained as a raffinate from 

which the solvent is separated in order to prevent its loss. Simplified flow sheet of a liquid extraction-based 

aromatics separation with the extraction solvent regeneration is presented in Figure 1. 

Recently, imidazolium- and pyridinium-based ionic liquids (ILs) have been shown to possess comparable 

or even better extractive solvent separation characteristics (aromatics to aliphatic hydrocarbons selectivity 

and aromatics distribution coefficient) compared to sulphotane (Meindersma, 2005). Moreover, ILs are 

considered as non-volatile liquids with very low values of saturated vapor pressure even at elevated 

temperatures (Rocha et al., 2011); thus, they are easy to separate from the mixture with hydrocarbons by 

simple distillation or stripping. On the other hand, very low solubility of ILs in non-aromatics makes the 

solvent recovery from a raffinate stream unnecessary (Meindersma, 2005). 

According to Meindersma and de Haan (2008), major part of investment and operational costs of an 

aromatics separation unit are connected with the solvent regeneration and prevention of its loss in the 

raffinate stream. In case of traditional extraction solvents, e.g. sulphotane, the extract is distilled in a 

column during the solvent regeneration. Heating and cooling duties of the distillation column reboiler and 

condenser depend on the amount of extract processed, the value of the reflux ratio and the components 

vaporization heat. To the regeneration costs, also costs of the solvent recovery from the raffinate should 

be included. When ionic liquid is used as the extractive solvent for the aromatics separation, IL 

regeneration by stripping is less energy demanding than distillation. 



 

 

188 

 

 

Figure 1: Simplified scheme of the liquid-phase extraction of aromatics from a hydrocarbons mixture with 

extractive solvent regeneration and recovery 

Consequently, overall costs of the aromatics separation using sulphotane are substantially higher than 

those connected with the use of ionic liquids as the solvent for the liquid-phase extraction. Here, the 

influence of separation characteristics of the selected ILs on some economic parameters of aromatics 

separation is presented. 

2. Aromatics separation simulation 

The model mixture to be separated consisted of heptane and toluene representing aliphatic and aromatic 

hydrocarbons, . As extractive solvents, 1-ethyl-3-methylimidazolium ethylsulfate ([EMim][ESO4]), 1-ethyl-3-

methylimidazolium bis(trifluoromethylsulphonyl)imide ([EMim][NTf2]), 3-methyl-N-butylpyridinium 

tetracyanoborate ([3-mebupy][B(CN)4]), 1-ethyl-3-methylpyridinium bis(trifluoromethylsulpho-nyl)imide 

([EMpy][NTf2]), and sulphotane were chosen. Liquid–liquid equilibria of individual ternary systems were 

modeled using the NRTL excess Gibbs energy model (Renon and Prausnitz, 1968). Binary parameters of 

the NRTL equation were obtained from literature. Data for sulphotane were taken from Meindersma 

(2005), for [EMim][NTf2] and [EMpy][NTf2] from Corderí et al. (2012), for [EMim][ESO4] from González et 

al. (2011), and for [3-mebupy][B(CN)4)] from Meindersma et al. (2011). Values of the NRTL binary 

parameters for the chosen systems are summarized in Graczová et al. (2014). 

Simulation of the model hydrocarbon mixture separation in the presence of extraction solvents was carried 

out using the equilibrium model of a counter-current extraction column (Figure 2). Simultaneous solution of 

material balances of the equilibrium stages was combined with summation equations of mole fractions in 

equilibrium phases and the ternary liquid–liquid equilibrium description: 

          R E R E 1 1 R 1 F F A,B,C 1j j ij ij j ij ij in n D x n D x n x i j  (1) 

            R 1 R 1 R E 2 R2 E 1 1 R 1 0 A,B,C 2,3,..., 2, 1j ij j j ij i j ij ijn x n n D x n D x i j N N  (2) 

        R 1 R 1 R E R S S A,B,Cj ij j j ij ij in x n n D x n x i j N  (3) 

where xi is the i-th component mole fraction in the extract, feed, raffinate, and solvent (E, F, R, and S) from 

the j-th equilibrium stage and n  is molar flow. Dij is the i-th component distribution coefficient  

 

 

Figure 2: Schematic representation of an extraction column composed of N equilibrium stages 

1 

F F
, 

i
n x  

R1 R 1
, 

i
n x  

E1 E 1
, 

i
n x  

E2 E 2
, 

i
n x  

2 

R2 R 2
, 

i
n x  

E3 E 3
, 

i
n x  

N – 1 

 R 2 R 2
, 

N iN
n x  

 R 1 R 1
, 

N iN
n x  

 E 1 E 1
, 

N iN
n x  

E E
, 

N iN
n x  

N 

R R
, 

N iN
n x  

S S
, 

i
n x  

··· 

Extractor 

Hydrocarbons 

feed 

Raffinate 

Solvent 

Make up 

solvent 
Separator 

Non-aromatics product 

Distillation/

stripping 

column 

Aromatics product 

Extract 

Regenerated solvent 



 

 

189 

Table 1: Input data for the extractor simulations 

F
n /(kmol h

–1
) xFA xFB xFC N xRBN 

10 0.85 0.15 0.00 10 0.005 

Solvent composition xSA xSB xSC 

Computation set 

1 0 0 0 

2 0.0005 0.0005 0.999 

3 0.001 0.001 0.998 

 
at the j-th equilibrium stage representing the isoactivity condition: 




    

E R

R E

A,B,C 1,2,..., 1,
ij ij

ij

ij ij

x
D i j N N

x
 (4) 

γRij is the i-th component activity coefficient in the heptane-rich phase from the j-th equilibrium stage and γEij 

is the i-th component activity coefficient in the extraction solvent-rich phase from the j-th equilibrium stage. 

3. Results and discussion 

Three sets of computations were carried out for every extraction solvent using the above-mentioned 

extractor model assuming different purities of the extraction solvent used. In all simulations, the number of 

extractor theoretical stages, feed amount and composition, as well as the mole fraction of toluene in the 

final raffinate were fixed. The input values are presented in Table 1. The output information was the 

consumption of the extraction solvent relative to the amount of feed and the complete composition of all 

streams appearing in Figure 2 and also in the set of material balances Eqs (1)–(3). 

In Figure 3, extraction solvent selectivity (SBA) versus toluene distribution coefficient (DB) is presented. The 

values of both parameters were calculated for the composition of the final raffinate (heptane-rich phase) 

and its conjugated extract (extraction solvent-rich phase) leaving the last (N-th) extractor equilibrium stage. 

For comparison, also the value corresponding to sulphotane is shown in Figure 3. Solvent selectivity is 

given by the following equation: 

 B EB RA
BA

A RB EA

N N N

N N N

D x x
S

D x x
 (5) 

From Figure 3 it can be deduced that, besides [EMim][ESO4], the chosen ILs posses different extraction 

selectivity and capacity in the toluene–heptane liquid phase extraction compared to sulphotane. 

15

20

25

30

35

40

0 0.5 1 1.5 2

D B

S
B

A

5

2

4

1

3

 

Figure 3: Selectivity of toluene–heptane extraction separation versus the toluene distribution coefficient 

diagram for the chosen extraction solvents: [EMim][ESO4] (1), [EMim][NTf2] (2), [3-mebupy][B(CN)4] (3), 

[EMpy][NTf2] (4), and sulphotane (5). The values were computed for fixed value of toluene content in the 

final raffinate xRBN = 0.005 



 

 

190 

 
Figure 4 presents the comparison of extraction solvents consumption relative to that of sulphotane 

assuming different purity of solvents. It can be seen that the ILs consumption is lower compared to that of 

sulphotane (values of relative consumption lower than 0.4) with the exception of [EMim][ESO4], which is 

slightly higher (relative consumption of about 1.2). Moreover, considering solvents which contain different 

levels of the impurities content, a decrease in the consumption of ILs relative to that of sulphotane was 

observed. The only exception was the increase in the relative consumption of [EMim][ESO4] when a higher 

level (0.002 mole %) of impurities content in the extractive solvent was taken into account. 

Relative solvent consumption (RSC) values were obtained by comparing the amount of IL used as the 

extraction solvent to the amount of sulphotane at the same simulation conditions: 

   


S IL S Sulfolane
RSC n n  (6) 

Generally, the use of extraction solvent containing impurities causes an increase in the solvent 

consumption, the excess of which increases exponentially with the decreasing value of the toluene 

distribution coefficient (Figure 5). In case of [3-mebupy][B(CN)4] (DB = 1.52), the increase was only about 

0.8 % and 1.7 % for the impurities content of 0.001 mole % and 0.002 mole % compared to the 

consumption of pure solvent. On the other hand, the increase in solvent consumption for [EMim][ESO4] (DB 

= 0.26) was about 10.2 % and 22.5 % for the two levels of impurities content considered in this study 

(Figure 5). The value of the increase in solvent consumption (ISC/%) compares the consumption of 

regenerated and pure extraction solvent according to the following equation: 

     S with impurities S pureISC 1 100%n n  (7) 

The increase of the solvent amount used in aromatics liquid-phase extraction can cause an increase in the 

dimensions both of the extractor and of the down-stream processing equipment. Moreover, energetic 

equipment (heaters, coolers) dimensions and the consumption of heating and cooling media can be 

influenced. 

In case of ILs used as extraction solvents for aromatics separation from complex hydrocarbons mixtures, 

extracted aromatics and an unavoidable amount of non-aromatic hydrocarbons are removed from the 

extract via stripping with permanent gas. The consumption of stripping medium in stripping column can be 

assumed to be proportional to the amount of hydrocarbons present in the extract. Afterwards, the 

extracted aromatics with a portion of other hydrocarbons are recollected from the stripping medium in a 

cooler as the aromatics product. Consumption of the cooling medium is proportional to the amount of the 

strippant used. 

1
2

3
4

0.002

0.001
0

0

0.2

0.4

0.6

0.8

1

1.2

1.4

 

0

5

10

15

20

25

0 0.5 1 1.5 2

D B

In
c
re

a
s
e
 i
n
 s

o
lv

e
n
t 
c
o
n
s
u
m

p
ti
o
n
/%

 

Figure 4: Variation in the solvent consumption 

relative to that of sulphotane (RSC) versus the 

extraction solvent purity for: [EMim][ESO4] (1), 

[EMim][NTf2] (2), [3-mebupy][B(CN)4] (3), and 

[EMpy][NTf2] (4) 

Figure 5: Variation of the extraction solvent 

consumption increase (ISC) versus the toluene 

distribution coefficient (DB) for: pure solvent () 

and solvent containing 0.1 mole % () and 0.2 

mole % () of impurities. Data for sulphotane 

are given in grey color 

R
e
la

ti
v
e
 s

o
lv

e
n
t 
c
o
n
s
u
m

p
ti
o

n
  

Impurities 

content/mole % 

Ionic liquid 



 

 

191 

0

2

4

6

8

0 0.5 1 1.5 2

D B

In
c
re

a
s
e

 i
n

 h
y
d

ro
c
a

rb
o

n
s
 c

o
n

te
n

t/
%

 i

 

Figure 6: Increase of the amount of hydrocarbons (heptane and toluene) in extract (ICH) versus DB for: 

pure solvents () and solvents containing 0.1 mole % () and 0.2 mole % () of impurities 

Thus, the use of solvent with higher content of impurities should be avoided and, therefore, thorough 

purification of the extraction solvent is recommended during the regeneration step. 

Figure 6 shows the influence of the solvents purity on the cumulative amount of heptane and toluene in 

extracts. An exponential increase of the excess hydrocarbons amount in the extract is observed when 

plotted against DB. The increase in the hydrocarbons content (IHC/%) was computed by comparing the 

amount of hydrocarbons accumulated in the extract when using regenerated and pure extraction solvent: 

   

   
EA1 EB1E with impurities

EA1 EB1E pure

IHC 1 100%
n x x

n x x

 
  

 
 

 (8) 

However, from the separation unit economics point of view, the most relevant information is the overall 

amount of hydrocarbons present in the extract. The values of hydrocarbons yield in the extract for the 

chosen ionic liquids are shown in Figure 7 with respect to the toluene distribution coefficient. No direct 

relation between the separation characteristics (SBA and DB) and the hydrocarbons content in the extract 

was found, although this chart resembles a mirror reflection of the SBA versus DB diagram (Figure 3). 

16

18

20

22

24

0 0.5 1 1.5 2

D B

H
y
d

ro
c
a

rb
o

n
s
 y

ie
ld

/%
 i

1

2

3

4

 

0.6

0.7

0.8

0.9

1

0 0.5 1 1.5 2

D B

x
B

1

2

3

4

 

Figure 7: Hydrocarbons (heptane + toluene) yield 

in extracts obtained by liquid-phase extraction of 

the model mixture versus DB using: [EMim][ESO4] 

(1), [EMim][NTf2] (2), [3-mebupy][B(CN)4] (3), and 

[EMpy][NTf2] (4) as extraction solvents 

Figure 8: Toluene mole fraction in aromatics 

product from the model mixture extraction 

versus DB using: [EMim][ESO4] (1), 

[EMim][NTf2] (2), [3-mebupy][B(CN)4] (3), and 

[EMpy][NTf2] (4) as extraction solvents 



 

 

192 

 
The highest yield hydrocarbons in the extract was observed in case of [EMpy][NTf2], the ionic liquid with 

the second highest value of the toluene distribution coefficient (DB = 0.95) among the ILs chosen as 

extraction solvents for aromatics separation. On the other hand, selectivity of this IL computed for the 

preset raffinate purity value, xRBN = 0.005, was SBA = 15.6, the lowest of all ILs considered in this study. 

Also, composition of the aromatics product obtained from the extract by stripping of the whole amount of 

hydrocarbons from ILs was investigated. Large differences in the composition of this product were 

observed as shown in Figure 8. Mole fraction of toluene in the aromatics product (xB) did not exceed the 

value of 0.90 irrespective of the extraction solvent used. Thus, another purification step (extraction or 

azeotropic distillation, Weissermel and Arpe (2003)) is required to provide a pure aromatics stream. 

4. Conclusions 

Simulation results confirmed that high value of DB is a key parameter influencing the aromatics extraction 

process (Meindersma and de Haan, 2008). Figures 4 and 5 show that the solvent purity has a significant 

impact on the solvent consumption (2-22 % increase for the solvent impurities content of 0.2 mole %) and 

the extraction column size/operation conditions. The excess amount of solvent necessary for the extraction 

process increases exponentially with the decrease of DB. The presence of impurities causes an increase of 

the hydrocarbons amount in the ILs-rich phase, thus increasing the costs of the extract down-stream 

processing. An exponential increase of the extra amount of hydrocarbons (0.2-2.5 % at the impurity level 

of 0.1 mole % and 0.5-7 % at 0.2 mole %) with the decreasing value of DB was observed (Figure 6). 

Similar extra costs connected with separation of hydrocarbons during ILs regeneration (strippant and 

cooling medium consumption as well as the equipment size) are to be expected. In order to get a pure 

aromatics stream, hydrocarbons mixture resulting from the ILs regeneration (toluene mole fraction on the 

level of 62-87 mole %, Figure 8) requires further purification step irrespective of the IL used as the solvent. 

Acknowledgement 

The authors acknowledge the Research and Development Assistance Agency (APVV-0858-12) for 

financial support. 

References 

Corderí S., Calvar N., Goméz E., Domínguez A., 2012, Capacity of ionic liquids [EMim][NTf2] and 

[EMpy][NTf2] for extraction of toluene from mixtures with alkanes: Comparative study of the effect of 

the cation, Fluid Phase Equilib., 315, 46-52. DOI: 10.1016/j.fluid.2011.11.016. 

González E.J., Calvar N., Domínguez I., Domínguez Á., 2011, Extraction of toluene from aliphatic 

compounds using an ionic liquid as solvent: Influence of the alkane on the (liquid + liquid) equilibrium, 

J. Chem. Thermodyn., 43, 562-568. DOI: 10.1016/j.jct.2010.11.013. 

Graczová E., Thomas L., Steltenpohl P., 2014, Aromatics extraction. 1. Separation characteristics of ILs 

for the toluene/heptane separation and equipment design aspects, PRES 2014, August 23–27, 2014, 

Prague, Czech Republic. 

Meindersma G.W., 2005, From solvent development to pilot RDC evaluation: Extraction of aromatics from 

naphtha with ionic liquids. PhD thesis, University of Twente, Enschede, the Netherlands. 

Meindersma G.W., de Haan A.B., 2008, Conceptual process design for aromatic/aliphatic separation with 

ionic liquids, Chem. Eng. Res. Des., 86, 745-752. DOI: 10.1016/j.cherd.2008.02.016. 

Meindersma G.W., Simons B.T.J., de Haan A.B., 2011, Physical properties of 3-methyl-N-butylpyridinium 

tetracyanoborate and 1-butyl-1-methylpyrrolidinium tetracyanoborate and ternary LLE data of [3-

mebupy]B(CN)4 with an aromatic and an aliphatic hydrocarbon at T = 303.2 K and 328.2 K and p = 0.1 

MPa, J. Chem Thernodyn., 43, 1628-1640. DOI: 10.1016/j.jct.2011.05.022. 

Rawat B.S., Gulati I.B., 1976, Liquid–liquid equilibrium studies for separation of aromatics. J Appl. Chem. 

Biotechnol., 26, 425-435. DOI: 10.1002/jctb.5020260163. 

Renon H., Prausnitz J.M., 1968, Local compositions in thermodynamic excess functions for liquid mixtures. 

AIChE J., 14, 135–144. DOI: 10.1002/aic.690140124. 

Rocha M.A.A., Lima C.F.R.A.C., Gomes L.R., Schroeder B., Coutinho J.A.P., Marrucho I.M., Esperanca 

J.S.S., Rebelo L.P.N., Shimizu K., Canongia-Lopes J.N., Santos L.M.N.B.F., 2011, High-accuracy 

vapor pressure data of the extended [CnC1im][Ntf2] ionic liquid series: Trend changes and structural 

shifts, J. Phys. Chem. B, 115, 10919-10926. DOI: 10.1021/jp2049316. 

Weissermel K., Arpe H.J., 2003, Industrial Organic Chemistry, 313-336. Wiley-VCH, Weinheim, Germany.