novel process that achieves 10 mol/kg sorbent swing ......ar sujan, rp lively et al., ind. eng....

Ryan P. Lively Yoshiaki Kawajiri, Matthew J. Realff, David S. Sholl, Krista S. Walton

Stephen J. DeWitt, Rohan Awati, Jongwoo Park, Eli Carter, Hector Rubiera Landa

Georgia Institute of TechnologySchool of Chemical & Biomolecular Engineering

Atlanta, GA 30332

Novel Process That Achieves 10 mol/kg Sorbent Swing Capacity in a Rapidly Cycled Pressure Swing Adsorption Process

Spinneret

Quench bath

Fiber collection50 m/minHollow fiber

sorbent spinning

Module make-up

Sorbent dispersed in polymer solution

AR Sujan, RP Lively et al., Ind. Eng. Chem. Res. 2018, 57(1)SJA DeWitt, RP Lively et al., Ann. Rev. Chem. Bio. Eng. 2018

Rapid thermal swing adsorption 3

Swing capacity and cycle time are key for driving down capital costs of

adsorption-based CO2 capture systems!

Key question: Can we increase swing capacity by 10x and reduce cycle time by 5x to dramatically drive down adsorbent costs?

Y Fan, CW Jones et al., Int. J. Greenhouse Gas Control 2014, 21, 61-71

Costs dominated by costs of adsorbent

Capital cost of RTSA system for NETL 550 MWe baseline: ~$1B

RP Lively, WJ Koros et al., Int. J. Greenhouse Gas Control 2012, 10(1)RP Lively et al., U.S. Patent 8,409,332 WJ Koros, U.S. Patent 8,658,041

Rapidly cycled pressure swing adsorption using MOFs 4

Cycle times of ~20 seconds are common for industrial RCPSA (>5x faster than RTSA)

JM Simmons, T Yildirim et al., Energ. Env. Sci., 2011, 4(6), 2177-2185



0.0 0.5 1.0 1.5 2.0 2.50

10

20

30

40

50

CO

2 Upt

ake

(mol

/kg)

Pressure (bar)

213 K

228 K

243 K

258 K

273 K

Pressure Swing AdsorptionΔP = 1.9 bar

Sub-Ambient ΔNCO2~ 40 mol/kg

210 220 230 240 250 260 270 2800

5

10

15

20

25

30

35

40

45

∆NC

O2 (

mol

/kg)

Temperature (K)

Pads = 2.0 bar Pdes = 0.1 bar Pdes = 0.2 bar Pdes = 0.3 bar Pdes = 0.5 bar Pdes = 1.0 bar

J Park, RP Lively, DS Sholl, J. Mater. Chem. A. 2017, 5, 12258-12265

𝑝𝑝𝐶𝐶𝐶𝐶𝐶𝑎𝑎𝑎𝑎𝑎𝑎 = 2 𝑏𝑏𝑏𝑏𝑏𝑏𝑝𝑝𝐶𝐶𝐶𝐶𝐶𝑎𝑎𝑑𝑑𝑎𝑎 = 0.1 𝑏𝑏𝑏𝑏𝑏𝑏



J Park, RP Lively, DS Sholl, J. Mater. Chem. A. 2017, 5, 12258-12265

How to economically achieve these pressurized, sub-ambient conditions?

Enabling 10 mol/kg swing capacities via flue gas pretreatment 7

Air Liquide Sub-Ambient Membrane System

Key parameters: swing capacity & selectivity

Sub-Ambient Adsorption System

D Hasse, S Kulkarni et al., Energy Procedia, 2013, 37, 993-1003

MIL-101(Cr) emerged as a promising candidate 8

220 230 240 250 260 270 280

CO2/N2 0.14/0.86 mixture

Temperature (K)

PCO2,ads = 2.0 bar PCO2,des = 0.1 bar PCO2,des = 0.2 bar PCO2,des = 0.3 bar PCO2,des = 0.5 bar PCO2,des = 1.0 bar

220 230 240 250 260 270 2800

5

10

15

20CO2 single component

∆NC

O2 (

mol

/kg)

Temperature (K)


Low cost ligands (benzene dicarboxylate)Relatively low cost metal centers (chromium nitrate)Scale-up is straight forward (70% yield on large batches)Water stable

J Park, RP Lively, DS Sholl, J. Mater. Chem. A. 2017, 5, 12258-12265L Hamon, GD Weireled et al., J. Am. Chem. Soc. 2009, 131, 8775-8777

9

Manufacturing MIL-101(Cr) fiber sorbents

2 µm

~50 wt% MOF73-80 vol% MOF200 µm

10 µm

10

PressurizationFeed Blow-

downEvacuation

PSA separation of simulated flue gas mixtures

11

0 500 1000 1500

0.0

0.2

0.4

0.6

0.8

1.0

C/C

0

Time (Seconds)

He-273 CO2-273 CO2-263 CO2-253 CO2-243

-50 0 50 100 150 200 250 3000

20

40

60

80

100

Frac

tion

of G

as in

out

let (

%)

Time (Seconds)

N2 CO2

Simulated Evacuation (5 second Blowdown, then He sweep)

0 2 4 6-2

0

2

4

6

8

10

12

14

16

18

Bed

Pres

sure

(bar

)

Time (Hours)

PressurizationFeed Blow-

downEvacuation

PSA separation of simulated flue gas mixtures

Fiber sorbents were cycled in CO2/N2 for 4 weeks

(~4000 cycles)

12

0 200 400 600 800 1000

0

5

10

15

20

25

30

35

Pristine 99.5% Dry SO2

1000ppm Dry So2

Qua

ntity

Ads

orbe

d (m

mol

/g)

Absolute Pressure (mbar)

CO2 Isotherms on SO2 Exposed MIL-101(Cr)

Stability to acid gases

Fiber sorbents stable to aggressive acid gas

exposures

13

Tem

pera

ture

[K]

Issues with heat effects

50K increase in temperatureduring adsorption

Sorbent-loaded porous polymer

matrix

µPCM

Melt/Freeze

Sorption/desorptionenthalpy

Fiber Module

Phase Change Material

Impermeable microcapsule

SJA DeWitt, RP Lively et al., Ind. Eng. Chem. Res. 2018, 58(15)SJA DeWitt, RP Lively et al., PCT US18/48110; WO 2019/09908

14

Tem

pera

ture

[K]

Issues with heat effects

Sorbent-loaded porous polymer

matrix

µPCM

Melt/Freeze

Sorption/desorptionenthalpy

Fiber Module

Phase Change Material

Impermeable microcapsule

Tem

pera

ture

Enthalpy

ADSORPTION

DESORPTION

Concept of Phase Change Material for PSA Heat Management

SJA DeWitt, RP Lively et al., Ind. Eng. Chem. Res. 2018, 58(15)SJA DeWitt, RP Lively et al., PCT US18/48110; WO 2019/09908

15

“Passive” thermal management via microPCM capsules

SJA DeWitt, RP Lively et al., Ind. Eng. Chem. Res. 2018, 58(15)

16

0 200 400 600 800 1000

Time , [ s ]

0

0.02

0.04

0.06

0.08

0.1

0.12

0.14

0.16

Mol

ar fr

actio

n , [

- ]

CO2 breakthrough profiles , 100 sccm

“Passive” thermal management via microPCM capsules

17

“Passive” thermal management via microPCM capsulesMIL-101(Cr) at 243 K, cyclic steady state simulations

Inclusion of thermal modulation pushes Pareto front into the “attractive” zone for post-combustion CO2 capture

Process economics – from molecular models to PSA simulation to flowsheet analysis

10 -1 10 0 10 1 10 2 10 3 10 4

Concentration , [ mol / m 3 ]

10 -4

10 -3

10 -2

10 -1

10 0

10 1

10 2

Am

ount

ads

orbe

d , [

mol

/ kg

ads.

]

Nitrogen / SENWOZ , "in silico"

213 K

228 K

243 K

258 K

273 K

dual-site Langmuir fit

qsat1 = 1.705b1 = 2.451e-08B1 = 2140

qsat2 = 20.13b2 = 1.293e-05B2 = 862.7

10 -2 10 0 10 2 10 4


10 -5

10 -4

10 -3

10 -2

10 -1

10 0

10 1

10 2

Am

ount

ads

orbe

d , [

mol

/ kg

ads.

]

Carbon dioxide / SENWOZ , "in silico"

213 K

228 K

243 K

258 K

273 K

tripleSiteLF fit

10 -1 10 0 10 1 10 2 10 3 10 4


10 -4

10 -3

10 -2

10 -1

10 0

10 1

10 2

Am

ount

ads

orbe

d , [

mol

/ kg

ads.

]

Nitrogen / SENWIT , "in silico"

213 K

228 K

243 K

258 K

273 K

dual-site Langmuir fit

qsat1 = 2.041b1 = 3.538e-08B1 = 2044

qsat2 = 17.79b2 = 1.28e-05B2 = 912.8

10 -2 10 0 10 2 10 4


10 -4

10 -2

10 0

10 2

Am

ount

ads

orbe

d , [

mol

/ kg

ads.

]

Carbon dioxide / SENWIT , "in silico"

213 K

228 K

243 K

258 K

273 K

tripleSiteLF fit

qsat1 = 13.75

b1 = 4.343e-06

B1 = 0.0006933

1 = 0.09992

qsat2 = 4.597

b2 = 6.154e-06

B2 = 1906

2 = 0.5396

qsat3 = 9.802

b3 = 6.759e-08

B3 = 3025

3 = 0.7413

…Prediction of binary

isotherms from CoreMOF database

0 0.2 0.4 0.6 0.8 1

Purity

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

Rec

over

y UTEWOG_tripleSiteLF

0 0.2 0.4 0.6 0.8 1

Purity

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

Rec

over

y SENWITPareto fronts from PSA optimizer

0 0.2 0.4 0.6 0.8 1

Purity

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

Rec

over

y SERKEG_quadPlusL

0 0.2 0.4 0.6 0.8 1

Purity

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

Rec

over

y WONZOP

18

…

Flowsheet optimization

for each material

19

A B C D E F G H I0

102030405060708090

100

82.6

Process economics – from molecular models to PSA simulation to flowsheet analysis

52.254.449.3

41.8

52.5

41.1

335.7

MIL-101(Cr)UiO-66

Cos

t of C

aptu

re ($

/tonn

e C

O2)

13X

122.3

75.94%

9.63%

14.44%

MOF Sorbent µPCM Fiber Manufacture

Fiber sorbent cost:

Total capex: $216M (25% of RTSA)

20

Accomplishments and outcomes

• Developed a “template” flowsheet for any sub-ambient pressure-driven CO2 capture process

• Created multi-scale workflow for process-driven material screening and selection for adsorption processes

• Scaled-up two different MOFs (MIL-101(Cr) and UiO-66) to >1 kg scale• Fabricated MOF fiber sorbents with integrated, passive thermal management• Constructed two PSA minipilot systems (~500 grams of CO2/day productivity)• Humid acid gas stability of MOFs, PSA, and fiber sorbents demonstrated• Capital and operating cost estimation for sub-ambient PSA CO2 capture

1. J Park, RP Lively, DS Sholl, ″Establishing upper bounds on CO2 swing capacity in sub-ambient pressure swing adsorption via molecularsimulation of metal-organic frameworks″, J. Mater. Chem. A 2017, 5, 12258-12265.2. J Park, JD Howe, DS Sholl, ″How Reproducible Are Isotherm Measurements in Metal-Organic Frameworks?″, Chem. Mater. 2017, 29, 10487-104953. SJA DeWitt, HO Rubiera Landa, Y Kawajiri, MJ Realff, RP Lively. “Development of Phase-Change-Based Thermally Modulated FiberSorbents”, Ind. Eng. Chem. Res. 2019, 58, 155, 768-57764. SJA DeWitt, A Sinha, J Kalyanaraman, F Zhang, MJ Realff, RP Lively. “Critical Comparison of Structured Contactors for Adsorption-BasedGas Separations” Annu Rev Chem Biomol Eng. 2018 Jun 7;9:129-152.5. SJA DeWitt, Y Kawajiri, HOR Landa, MJ Realff, RP Lively. “Incorporation of microencapsulated phase change materials into wet-spin dry jetpolymer fibers”. PCT US18/48110; WO 2019/099086. BR Pimentel, AW Fultz, KV Presnell, RP Lively, “Synthesis of water-sensitive metal-organic frameworks within fiber sorbent modules”, Ind.Eng. Chem. Res. 2017, 56, 17, 5070-5077

Papers and Patents

1 submitted, 5 to be submitted Fall 2019

Conclusions and perspectives 21

Key question: Can we increase swing capacity by 10x and reduce cycle time by 5x to dramatically drive down adsorbent costs?

• Combining RCPSA cycles with appropriate metal-organic frameworks in sub-ambient conditions results in highly productive adsorption systems (i.e., ~30 tonne CO2/tonne adsorbent-day)

220 230 240 250 260 270 2800

5

10

15

20CO2 single component

∆NC

O2 (

mol

/kg)

Temperature (K)


• Significant “real world” complexities exist, but hollow fiber sorbent platform provides solutions to many of these (scalability, transport limitations, etc.)

• Costs in the range of $40-$50/tonne CO2 at sequestration pressures may be achievable using these materials in this process concept, but significant work remains. Advantages of small size, material stability to flue gas conditions, and modularity are important.

22

https://lively.chbe.gatech.edu

novel process that achieves 10 mol/kg sorbent swing ......ar sujan, rp lively et al., ind. eng....

Documents