Systematic Method for Screening Ionic Liquids as Extraction

Solvents Exemplified by Extractive Desulfurization Process

REFERENCES

ACKNOWLEDGEMENT

1. Eckert, F., Klamt, A. Fast solvent screening via quantum chemistry: COSMO‐RS approach. AIChE

Journal. 2002, 48(2): 369-385.

2. Song, Z., Zhou, T., Zhang, J., Cheng, H., Chen, L., Qi, Z. Screening of ionic liquids for solvent-sensitive

extraction–with deep desulfurization as an example. Chemical Engineering Science. 2015,129: 69-77.

3. Song, Z., Zhou, T., Qi, Z., Sundmacher, K. Systematic method for screening ionic liquids as extraction

solvents exemplified by an extractive desulfurization process. ACS Sustainable Chemistry &

Engineering, 2017, 5(4): 3382-3389.

National Natural Science Foundation of China (NSFC U1462123), Major State Basic Research

Development Program of China (973 Program 2012CB720502).

Deutsche Forschungsgemeinschaft (DFG) for the Collaborative Research Center SFB/TRR 63

"Integrated Chemical Processes in Liquid Multiphase Systems„.

MOTIVATION

IL SCREENING: STATE OF THE ART

x expensive and time-consuming

x limited to simple laboratory experiments

Ionic liquids (ILs) are highly promising alternatives for volatile organic solvents in

liquid-liquid extraction, gas absorption, extractive distillation, etc.

Application challenges

• huge number of ILs, various separation processes

• complex effects of IL molecular decision variables at different levels

Goal: develop systematic methods for screening practically attractive IL solvents

for separation processes.

Ab initio calculation x computationally expensive

NRTL, UNIQUAC, EoS (PC-SAFT) x require experimental data, molecule-specific,

x limited predictive ability for novel systems

UNIFAC-IL

x GC-based, limited group parameters available

COSMO-RS model independent of experiment, molecule/group σ-profile

virtually applicable to any system

good qualitative & acceptable quantitative prediction

1 i

1 1

E Rm m

1 2

1 2

E E

R R

m mS

m m

3

RSL m

γ∞ mole-based LLE mass-based LLE

β∞/β 28089 27584 3144

S∞/S 10349 11293 834

−/SL not considered 11265 831

Modified thermodynamic criteria

Physical property estimation by GC models [ref]

31 36

1 1

( ) 288.7m i ci j aj

i j

T K n t n t

20 67 20 67

1 1 1 1

ln 6.982 i i j j i i j j

i j i j

n a n a n b n b T

Step 1: Pre-screening by modified thermodynamic criteria

thermodynamic properties

LLE(mass-based β, S, SL)

physical properties

(Tm, η)

process performance

(operating conditions,

solvent/energy consumption)

COSMO-RS

GC methods

Aspen plus

conventional solvent

as benchmark

Tm < 298.15 K

η < 100 cP

acti

vit

y c

oeff

icie

nt m

odel

(NR

TL

)

Criteria derived from mass-based LLE are more reasonable and efficient.

Step 2: Further screening by Tm and η constraints

831 163 15 Tm < 298.15 K η < 100 cP

Top-ranked 4 IL candidates

Solvent β S SL Tm (K) η (cP)

IL1 2.66 104.03 2.14×10-6 281.14 88.55

IL2 2.31 119.58 2.56×10-5 253.01 64.73

IL3 2.27 77.34 2.85×10-5 286.08 71.31

IL4 3.02 134.94 3.02×10-5 268.80 53.95

sulfolane 2.10 71.15 9.09×10-3

All ILs have notably higher β and S, as well as lower SL than sulfolane.

Step 3: Final selection by process simulation

ILs defined as pseudo-component in Aspen Plus [ref]

MW, ρ, Tnb, critical properties

NRTL model as the thermodynamic model

parameters regressed from COSMO-RS data

B2

(Evap)

S1

Fresh IL

S2

Model fuel feed

10,000 kg/hr

S3

S4

Low-sulfur fuel product

S5

Sulfur-load IL

S6

Residual liquids

S7

Recycled IL

B1

(EC)

IL process

j iS

Process simulation of IL-based processes [ref]

Systematic IL screening

3 steps

decomposition

ILs: much lower energy and solvent consumption, higher fuel product recovery ratio

EDS case study: removal of trace aromatic sulfur compounds from fuel oils

m2

m3

m1

Specific global

composition

APPLICATION – EXTRACTIVE DESULFURIZATION

Current Methods

Experimental

Computational

Main Limitations of Previous Screening

Improved IL screening methods are highly desirable!

SYTEMATIC METHOD FOR IL SCREENING

β∞, S∞

Infinite dilution,

molar properties

Effect of molecular weights (MWs) of ILs, effect of practical condition

Melting point

Viscosity

β∞, S∞

Mixtures

to be separated

MW of ILs

Physical properties

Process performance

?

simulated by thiophene/n-octane mixture (sulfur content 100 ppm)

Sulfolane is employed as the benchmark solvent

(overall 36260 ILs combined from 370 cations and 98 anions in COSMOthermX database)

IL1 IL2 IL3 IL4

Sulfolane process requires 2 distillation columns.

more capital cost

Required S/F of all ILs is lower than sulfolane.

less solvent need

IL regeneration: different dependencies of the

operating pressure on the tempeature.

different energy cost

2 3 4 5 6 7 8 9 100.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

Ma

ss r

ati

o o

f S

/F

Number of stages

IL1

IL2

IL3

IL4

Sulfolane

(a)

380 400 420 440 460 480 500 5200.0

0.2

0.4

0.6

0.8

1.0

Pre

ssu

re (

bar)

Temperature (K)

IL1 IL2 IL3 IL4

Extraction column

Evaporator

Z. Song1, T. Zhou1, Z. Qi2, K. Sundmacher1,3

1 Max Planck Institute for Dynamics of Complex Technical Systems, Sandtorstrasse 1, 39106 Magdeburg 2 Max Planck Partner Group at East China University of Science and Technology, Meilong Road 130, 200237 Shanghai 3 Otto-von-Guericke University Magdeburg, Universitätsplatz 2, 39106 Magdeburg

a list of top candidates