1

Development of a Novel Biphasic CO2 Absorption Process

with Multiple Stages of Liquid–Liquid Phase Separation

for Post-Combustion Carbon Capture

(DOE/NETL Agreement No. DE-FE0026434)

Yongqi Lu

Illinois State Geological Survey, Prairie Research Institute

University of Illinois at Urbana-Champaign

2016 NETL CO2 Capture Technology Meeting

Pittsburgh PA • August 11, 2016

Project Overview

Project objectives

Developing new biphasic solvents

Demonstrate phase separation-coupled CO2 absorption

process

Generate and assess engineering and scale-up data

Project duration

10/1/15 – 9/30/18 (36 months for two BPs)

Funding profile

2

DOE funding 1,999,996

BP1 1,079,663

BP2 920,333

Cost share (Cash & In-kind) 501,052

BP1 269,920

BP2 231,132

Total 2,501,048

Project Participants

University of Illinois

Illinois State Geological Survey

• Solvent screening & development

• Solvent equilibria, kinetics & properties measurements

• Absorption and desorption column testing

Illinois Sustainable technology Center

• Evaluation of solvent stabilities and corrosion impact

Applied Research Institute

• Molecular dynamics simulation study for solvent screening

Trimeric Corporation

• Process feasibility and TEA analysis

3

Conventional Monophasic Absorption Approach

4

Conventional CO2 Absorption Process (e.g., MEA)

Flue gas

Cleaned gas

Absorber

(40-50 C)

CO2-rich solvent

Cross-heat

exchanger

Stripper

(90-140 C)

Aqueous CO2

lean solvent

Combined

CO2-lean solvent

CO2

compressor

Reboiler

Steam

Reflux

Cooler

5

Advantages of Biphasic CO2 Absorption Processes

Conceptual of Biphasic Absorption Processes by Other Developers

Reduced equipment size due to reduced mass of solvent to be regenerated in

stripper

Reduced energy use and compression cost due to increased CO2 loading

capacity (concentrated feed), reduced mass, and increased stripping pressure

Impacts on stripper:

CO2-rich

phase

CO2-lean

phaseFlue gas

Cleaned gas

Absorber

(40-50 C)

CO2-rich solvent

Cross-heat

exchanger

Stripper

(90-140 C)

Aqueous CO2

lean solvent

Combined

CO2-lean solvent

LLPS

Reboiler

Steam

Reflux

Cooler

CO2

compressor

Biphasic CO2 Absorption Process with

Multi-Stages of Liquid-Liquid Phase Separation (BiCAP)

6

LP

steam

Flue gas

Cleaned

flue gas

Absorption

column

High-

pressure

stripper

Reboiler

CO2 (1)

Cooler

Cross heat

exchanger

Lean

solvent

CO2-rich

phase

CO2-lean

phase

Condensate

to power

plant cooling

tower

Inter-stage

cooler

Rich

solventWater

separator

Rich

solvent

tank

Solvent

makeup

Solvent

tankCondenser

Phase

separator

LP

steam

Flash

CondenserCO2 (2) to

compressor

Condensate

to cooling

tower

Reduced viscosity with separation of rich, viscous phase improves mass

transfer rate

Lean phase to next packed bed improves kinetics

Reduced mass of solvent to next packed bed

Adds impacts on absorber to the impacts on stripper:

Advantages of Multi-Stage Phase Separation (LLPS)

during CO2 Absorption

Modified operating curve allows for a higher mass transfer driving

force and thus a faster absorption rate7

0 0.2 0.4 0.6 0.8 1 1.2

CO

2p

art

ial

pre

ssu

re,

kP

a

CO2 loading, mol/mol solvent alkalinity

Flue gas inlet

Flue gas outlet

Equilibirum

curve

D'

BB'

CC'

D

A

Operating

curve

12 kPa

1.2 kPa

A

B

B’

C

C’

D

LLPS

LLPS

LLPS

Illustration of operating & equilibrium curves with 3 stages of CO2 absorption-LLPS

(the equilibrium curve that may change after each LLPS is not displayed in this illustration)

BiCAP vs MEA and Other Biphasic Processes

8

Biphasic processes vs MEA

Biphasic solvents have

larger loading capacity for

CO2 stripping due to

absorbed CO2 concentrated

in one phase as feed

solution to the stripper

Reduced mass and

elevated P for CO2 stripping

Reduced heat duty (low

sensible heat and

stripping heat)

Reduced compression

work requirement

BiCAP vs other biphasic processes

Absorption process:

Multi-LLPS in BiCAP allows for low CO2

loading and low viscosity throughout the

absorber, resulting in a faster absorption

rate and reduced absorber size

Solvent:

Phase transition behavior of BiCAP

solvents are tunable, facilitated with the

use of a unique solubilizer(s), allowing

for a wide range of solvent component

selection

Desorption process:

Desorption with a flash step to obtain

high-pressure CO2 and reduce

compression requirements

Project On-Track and Initial Milestones Achieved

9

WBS Lead Description Start End Q1 Q2 Q3 Q4 Q5 Q6 Q7 Q8 Q9 Q10 Q11 Q12

1.0 Project management and planning 10/01/15 09/30/18

1.1 ISGS/All Project management and planning 10/01/15 09/30/18

1.2 ISGS/All Briefings and reports 10/01/15 09/30/18

2.0 Screening and characterization of biphasic solvents 10/01/15 06/30/16

2.1 ISGS Solvent screening tests on CO2 absorption and phase transition behavior 10/01/15 06/30/16

2.2 ISGS Solvent screening tests on CO2 desorption performance 10/01/15 06/30/16 a

2.3 ARI Molecular simulation study for solvent screening 10/01/15 06/30/16 b

3.0 Measuring phase equilibria, absorption kinetics, & solvent properties 01/01/16 09/30/16 A

3.1 ISGS Measurement of VLE data under absorption & desorption conditions 01/01/16 09/30/16

3.2 ISGS Measurement of CO2 absorption kinetics 04/01/16 09/30/16 c

3.3 ISGS Measurement of solvent properties 07/01/16 09/30/16

4.0 Determining thermal and oxidation stabilities of the selected solvents 04/01/16 12/31/16

4.1 ISTC Oxidation stability of solvents under simulated absorption conditions 04/01/16 12/31/16

4.2 ISTC Thermal stability of solvents under simulated desorption conditions 04/01/16 12/31/16 e

5.0 Testing CO2 absorption and phase separation in a packed-bed column 04/01/16 03/31/17 B

5.1 ISGS Modification of absorption column to incorporate multi-LLPS operation 04/01/16 09/30/16 d

5.2 ISGS Parametric testing of CO2 absorptionand LLPS in the packed-bed column 07/01/16 03/31/17 f

5.3 ISGS Rate-based modeling of CO2 absorption in the packed-bed column 10/01/17 03/31/17

6.0 Development of a process sheet and preliminary process analysis 04/01/16 03/31/17

6.1 Trimeric Development of a conceptual process flow sheet 04/01/16 12/31/16

6.2 Trimeric Preliminary process analysis 07/01/16 03/31/17 g

7.0 Testing CO2 desorption in a high-pressure flash and stripping column 04/01/17 03/31/18 C

7.1 ISGS Modification of an existing packed-bed column by incorporating a flash unit04/01/17 09/30/17 h

7.2 ISGS Parametric testing of high-pressure flash and stripping 07/01/17 03/31/18 j

7.3 ISGS Design modeling of CO2 desorption in the flash and stripping column 10/01/17 03/31/18

8.0 Assessing the impact of solvent corrosion on the equipment 04/01/17 12/31/17

8.1 ISTC Assessing the impact of solvent corrosion on the equipment 04/01/17 12/31/17 i

9.0 Technical and economic feasibility study 10/01/17 09/30/18 D

9.1 Trimeric Process simulation and mass & energy balance calculations 10/01/17 06/30/18

9.2 Trimeric Technical and economic feasibility study 01/01/18 09/30/18 k

START/ENDSOPO BREAKOUT SCHEDULE BUDGET PERIOD 1 BUDGET PERIOD 2

Aug 11, 2016

COMPLETE

Project Scope and Technical Approach

10

Biphasic solvent

screening tests

(~50 solvents)

Molecular dynamics

simulations for

solvent screening

Preliminary selection

of solvent

components

Downsize to 5-10

solvents

VLE, kinetics, and

properties

measurements

Downsize to 2-3

solvents

Assessing solvent

oxidation and thermal

degradation

Parametric testing of

CO2 absorption &

LLPS

Process flow sheet

and techno-economic

analysis

Parametric testing of

high-pressure flash

and stripping

Assessing solvent

corrosion impact on

equipment

Biphasic

solvent

screening

(Q3)

Biphasic

solvent

characterization

(Q5 & Q9)

Biphasic

solvent-enabled

process testing

(Q6 & Q10)

Process

analysis & TEA

(Q12)

Key Milestones and Success Criteria

11

BP1 (by Q6):

Identify 2-3 top-performing solvents

(based on phase transition & CO2 enrichment behavior, CO2 loading capacity,

absorption kinetics, and viscosity)

Complete lab testing of 2-3 solvents in an absorption column with multi-LLPS

(CO2 capacity and kinetics 5 M MEA; each LLPS stage ≤ 5 min; 80% CO2

enrichment in the rich liquid phase)

Demonstrates reliable operability of the multi-stage absorption & LLPS

configuration during lab-scale testing

BP2: (by Q12)

Complete lab testing of 2-3 solvents in a flash / stripping system

(≥5 bar stripping pressure; working capacity ≥2 times that of 5M MEA)

Initial techno-economic feasibility study shows significant progress toward

achievement of DOE performance goals

Task 2. Solvent Screening Experiments: Absorption Capacity

12

0.0

1.0

2.0

3.0

4.0

5.0

6.0s1 s2 s3 s4 s5 s6 s7 s8 s9

s10

s11

s12

s13

s14

s15

s16

s17

s18

s19

s20

s21

s22

s23

s24

s25

s26

s27

s28

s29

s30

s31

s32

s33

s34

s35

s36

s37

s38

s39

s40

s41

s42

s43

s44

s45

s46

s47

s48

s49

s50

s51

s52

Tota

l CO

2 lo

adin

g in

so

lven

t, m

ol/

L

Solvents

Total loading, mol/L

--- 5M MEA (benchmark)

Overall CO2 capacity tested in gas impingers for 60 min of absorption

under atmospheric CO2 at 40C:

Most solvents achieved a comparable or slightly higher CO2 loading

than 5M MEA for absorption

Phase Transition Behavior

Formation of dual phases and their volumes are tunable

CO2 loading is highly concentrated in rich phase (91-99% of total loading)

Loading capacity of CO2 desorption for most solvents is improved by 37-

234% over 5M MEA (as only rich phase solution is used for regeneration)13

M

EA s1

s2

s3

s4

s5

s6

s7

s8

s9

s1

0s

11

s1

2s

13

s1

4s

15

s1

6s

17

s1

8s

19

s2

0s

21

s2

2s

23

s2

4s

25

s2

6s

27

s2

8s

29

s3

0s

31

s3

2s

33

s3

4s

35

s3

6s

37

s3

8s

39

s4

0s

41

s4

2s

43

s4

4s

45

s4

6s

47

s4

8s

49

s5

0s

51

s5

20

2

4

6

8

10CO

2 loading in rich phase

% of rich phase CO2 in total CO

2 uptake

CO

2 lo

ad

ing

in

ric

h p

hase, m

ol/L

90

92

94

96

98

100

No

PS

No

PS

No

PS

No

PS

No

PS

% o

f ri

ch

ph

ase C

O2 in

to

tal u

pd

ate

No

PS

Desorption Pressure

PCO2 = 0.710 bar at 120 C at lean CO2 loadings of ~0.30.5 mol/mol

(determined to achieve 90% CO2 removal), which is much higher than

that of lean MEA (pressure at ~0.25 mol/mol)14

10

100

1000

0 0.2 0.4 0.6 0.8 1

CO

2 p

art

ial p

ressu

re, b

ar

CO2 loading, mol/mol of total amine in rich phase

120C, De-S1

120C, De-S2

120C, De-S3

120C, De-S4

120C, De-S5

120C, De-S6

120C, De-S7

120C, De-S8

120C, De-S9

120C, De-S10

120C, De-S11

120C, 5M MEA

MEA op range

BiCAP desorption op range

(lean conditions to be determined)

10

1

0.1

Task 2. Molecular Dynamics Modeling for Solvent

Screening: Methodology Flowchart

15

Predicted efficiency based

on Gibbs free energy of

carbamate formation

Efficiency of phase separation• Implemented steered molecular dynamics simulation to

characterize phase separation process

• Computed free energy change for phase separate

system for selected solvent compositions including MEA

• Developed a method to characterize chemical potential

differences for species in rich an lean phases

Efficiency of carbon capture• Calculated enthalpy of formation for carbamate,

zwitterions, and protonated amines

• Performed stability analysis for selected ion pairs via

association energy in order to construct initial

configurations for phase separation simulations

• Computed Gibbs free energy change of selected reactions

Characterization of reaction pathways and barriers• Identified multiple reaction schemes as a function of solvent composition

• Characterized efficiency of bicarbonate protonation

• Computed energy barriers between transition states and carbonic acid via DFT

• Validated reactive molecular dynamics force field for proton transfer

Experimental Input• NMR, MS-GS, water content in each phase, pH

Transport and mixing properties• Computed self-diffusion coefficient of reactants and

products

• Computed viscosity for selected amines

Characterized transport

including diffusivity of species

Predicted efficiency based

on free energy change for

phase separation

Speciation

completed

Molecular Dynamics Modeling for Efficiencies of

Carbon Capture and Phase Separation

Carbon capture efficiency screening is performed via thermodynamic

calculations using semi-empirical molecular orbital theory (18 reactions

considered). Below is one example for the carbamate formation:

16

• G < 0: spontaneous

reactions; H < 0:

exothermic reactions

• Approach is general to

screen any

stoichiometry &

reactions of interest.

𝐴𝑚 + 𝐶𝑂2 + 𝐴𝑚 → 𝐴𝑚𝐶𝑂2− + 𝐴𝑚𝐻+

System F111

after SMD run

Zwitterions and

carbamates

“steered” to

separate inside

the simulation

domain

System’s energy

change during SMD

Favors phase

separationPhase separation

not favored

• Work done by the

system or constraint is a

measure of the driving

force behind the phase

separation process

Phase separation efficiency screening is performed via steered molecular

dynamics simulation

Task 3. Absorption Rate

Rate tests using both a stirred tank reactor (exemplary results displayed

in plot above) and a wetted wall column reactor (work is ongoing)

Rates tunable by addition of a promoter or selection of a different

solubilizer

* S18 (S17+promoter) and S31 (solubilizer different from S17)17

0.0E+00

5.0E-04

1.0E-03

1.5E-03

2.0E-03

2.5E-03

3.0E-03

0.2 2

CO

2ab

so

rpti

on

rate

, m

ol/

m2.s

CO2 partial pressure, psia

MEA S17 S18

S23 S26 S31

S40 S28 S53*

0.22-0.25 mol CO2/

mol amine

~0.5 mol CO2/

mol amine

0.29-0.42

mol/mol

Task 3. Viscosity Measurement and Optimization

Lean phase viscosity < 9 cP (data not displayed)

Rich phase viscosity was successfully decreased from ~400 cP to ~30

cP by optimizing total solvent concentration and selecting suitable amine

structures

18

0

50

100

150

200

250

300

350

400

450

500

#1 #2 #3 #4 #5 #6 #7 #8 #9 #10#11#12#13#14#15#16#17

Ric

h p

ha

se

vis

co

sit

y, c

P

Rich phase solution

60-70wt% amines

50wt% amines

40wt% amines

Task 5. A Lab Absorption System with 3-Stages of Packed

Beds and LLPS Vessels Designed and Under Fabrication

3 stages (4-in ID, 7-ft packed-bed for each stage) arranged horizontally

to accommodate lab ceiling limit

3 stages in one vertical column envisioned for practical use 19

CO2

cylinder

Aircompressor

~ Elec steam generator

T4

T3

T2

T1

P1

P2

T8

T7

T6

T5

P3

P4

T12

T11

T10

T9

P5

P6

Other gases (Optional)

MFC

MFC

MFC

FM

T0

Water cooler

Water cooler

Water cooler

Elec heater ~

Gas exhaust

Gas 3 to analysis

Gas 4 to analysis

Gas 1 to analysis

Gas (CO2)analyzer

Gas stream drying

Gas 1 Gas 2 Gas 3 Gas 4

T, P, Flow ratedisplay/control

Ts Ps Fs

Pump

Pump

Pump

LLPS

Regenerator

Solvent tank

Liquid sampling

T13

T14

Rotameter

Gas 2 to analysis

Rotameter

Rotameter

Vent

Liquid samplingLiquid sampling

Pump Pump

Liquid sampling

LLPS LLPS

Lab Prototype Phase Separation Vessel Achieved Efficient

and Stable Separation

20

Phase separation vessel design

Based on density difference

(lean phase ~0.85 vs. rich

phase ~1.1 g/cm3)

Residence time ≤ 5 min

(preferred at <1-2 min)

Actual separation performance

Separation efficiency better

than the design

Able to maintain constant

levels of both G-L and L-L

interfaces

Both interface levels adjustable

by adjusting their weir heights

Very stable operation

(Liquid volume of 10 L, total volume

of 15 L, liquid flow rate of 2 L/min)

Task 6. Preliminary Process Flow Diagram

Developed for BiCAP

Work underway to improve process/unit configuration, identify opportunities to minimize

equipment items to reduce cost, and assess integration options into a power plant 21

Future Work Plan in this Project

VLE data under absorption &

desorption conditions

Absorption kinetics

Solvent properties (Hr; viscosity,

diffusivity, Cp, density)

Thermal & oxidation stabilities

Solvent corrosion tendency

For 2-3 selected solvents:

Testing CO2 absorption and LLPS

in a 3-stage packed-bed column

Testing CO2 desorption in a flash

+ stripping column

Preliminary process analysis and

TEA feasibility study

22

Biphasic

solvent

screening

(Q3)

Biphasic

solvent

characterization

(Q5 & Q9)

Biphasic

solvent-enabled

process testing

(Q6 & Q10)

Process

analysis & TEA

study

(Q12)

WWC for kinetics

Low-P VLE

High-P VLE

Oxidative

stability

Thermal stabilityRx calorimeter

GC-MS

NMR

Stripping system

to be modified

CO2

cylinder

Aircompressor

~ Elec steam generator

T4

T3

T2

T1

P1

P2

T8

T7

T6

T5

P3

P4

T12

T11

T10

T9

P5

P6

Other gases (Optional)

MFC

MFC

MFC

FM

T0

Water cooler

Water cooler

Water cooler

Elec heater ~

Gas exhaust

Gas 3 to analysis

Gas 4 to analysis

Gas 1 to analysis

Gas (CO2)analyzer

Gas stream drying

Gas 1 Gas 2 Gas 3 Gas 4

T, P, Flow ratedisplay/control

Ts Ps Fs

Pump

Pump

Pump

LLPS

Regenerator

Solvent tank

Liquid sampling

T13

T14

Rotameter

Gas 2 to analysis

Rotameter

Rotameter

Vent

Liquid samplingLiquid sampling

Pump Pump

Liquid sampling

LLPS LLPSAbsorption column

under constructionStripping system

to be modified

Next-Stage Technology Development

Current project is a laboratory development

If process and TEA feasibility proven in the current project,

next stage would be a close-loop bench or small pilot demo

with simulated or actual flue gas

Rigorous process design and optimization modeling to

enhance performance

Analysis of technical risks and mitigation for scale-up

Identify industrial partners (design, construction, and

testing)

23

Acknowledgements

DOE/NETL Project Manager: Andrew Jones

University of Illinois:

Kevin O’Brien (Co-PI; PhD, Director)

Hong Lu (PhD, Chemical/Environmental Engineer)

David Ruhter (MS, Lab Manager)

Yang Du (PhD, Chemical/Environmental Engineer)

Qing Ye (PhD Student)

Wei Zheng (PhD, Senior Chemist)

Brajendra K Sharma (PhD, Senior Chemical Engineer)

Viktoriya Gomilko (MS, Assistant Research Chemist)

Joe Pickowitz (Environmental Engineer)

Santanu Chaudhuri (PhD, Principal Research Scientist)

Naida Lacevic (PhD, Lead Simulation Specialist)

Trimeric Corporation:

Ray McKaskle (Subaward PI; P.E., Senior Chemical engineer)

Andrew Sexton (PhD, P.E., Senior Chemical Engineer)

Kevin Fisher (VP, P.E., Senior Chemical Engineer)

24