1

Pacific Northwest National LaboratoryU.S. Department of Energy

Amine-Modified Multiscale Materials for Reversible CO2 Capture

Chris AardahlPacific Northwest National Laboratory

Richland, WA USA

3rd Annual Conference on Carbon Capture and Sequestration

May 4, 2004

Acknowledgements

Project Team: Richard Zheng, Leo Fifield, Glen Fryxell, Diana Tran, Shane Addleman, Daryl Brown, Tom Zemanian

Project Support: PNNL’s Carbon Management Laboratory Directed Research and Development Initiative

Helpful Discussions: Jon Gibbons, Imperial College; Ken Humphreys & Pete McGrail, PNNL

Carbon Capture with Solid MaterialsGoal: Explore and develop functional multiscalematerials to serve as reliable & regenerablesorbents for the capture of CO2 from industrial processes such as fossil fuel power plants.

Issues:Surface Chemistry/Capacity: Can a solid-based system be made that competes with traditional wet amine systems?Thermal and Chemical Stability: Can the materials remain durable with high reversible capacity in flue gas conditions over long periods of time? Cost: Can these materials be produced and implemented in process systems affordably on a life cycle basis?

Solid vs. Liquid Capture

Pros:Since all functional groups are at the surface, there is effectively no liquid side mass transfer resistance.In ideal configuration, there is no downstream purification of CO2 required after regeneration.No liquid handling required.

Cons:Not possible to store captured CO2 and regenerate at preferred times (eg., when power demand is low).No possibility for sorbent make up unless moving bed or fluidized bed system is developed.

Project Scope

Synthesize new materials and new functional approaches Characterize materials in bench scale CO2 capture apparatus to assess total and working capacityCharacterize material structure and mechanism for adsorption (NMR, FTIR, Raman, TGA, electron microscopy)Understand durability of materials under repeated cycling and exposure to potential poisons: SO2, H2S, and NOx.Understand the impact of water vapor on CO2 captureDevelop process concepts and flowsheets for implementionPerform preliminary economic analysis

Temperature Swing AdsorptionPacked BedOff-Gas

~10% CO2

Use high surface area materials → inherently high capacity.Large pore volume and surface functionalization useful to reduce mass

transfer resistances. This will increase operating capacity.Use material classes that allow facile chemical functionalization. Thermal integration with power plant is critical.

Temperature

Loading Regeneration

To Compression & Storage Stack Out

2

Functional Multiscale Materials

ApplicationsCatalysisGas or liquid phase separationsChemical storageSensors

Ordered Mesoporous Oxides & Aerogels

Carbon NanotubeComposites,

Carbon Fibers

EDA Functionalized SilicaSynthesisCondensation of an EDA-alkoxysilane and silanols on silica surface.

Substrate: SBA-15 mesoporous silicapore diameter 5.5-7.8 nm, surface area 700 m2/g

NH2

NH

Si(OCH3)3

NH2

NH

SiOHO O

3CH3OHOH OH

Silica

OH OH OH OH

Silica

+toluene

reflux

3746

(a)

(b)

1629

3362

3432

3299

1346

1601

1410

1457

Abso

rban

ce (a

.u.)

W avenum bers (cm -1)

150020002500300035004000

2931

2881

(a) SBA-15

3476 cm-1 free silanol OH3432 cm-1 associated OH1629 cm-1 adsorbed H2O

(b) EDA-SBA-15

3362 cm-1 N-H asym. stretch3299 cm-1 N-H sym. stretch2931 cm-1 C-H asym. stretch2881 cm-1 C-H sym. stretch1601 cm-1 N-H deformation1457 cm-1 CH2 scissor1410 cm-1 (EDA-silane)1346 cm-1 CH2 wagging

FTIR spectra – Before and after grafting EDA ligands

FTIR spectra confirm presence of EDA functional groups in sorbent.

CO2 Sorption Bench

• CO2 levels of 1-50% possible

• SO2 examined using mixes

• CO2 level measured with RGA

• ∆P & Isotherms easily measured

Adsorption Cell

P P

T

Vent

H2O

T

RGA

Roughing Pump

Turbo Pump

Vent

N2

Ar

CO2 mix

Breakthrough Testing

Known concentration of CO2/N2 mix was flowed through sorbent bed, followed by a TPD to release adsorbed CO2.

Time (s)0 500 1000 1500 2000 2500 3000

CO

2 m

ass

spec

sig

nal (

a.u.

)

0

1e-7

2e-7

3e-7

4e-7

5e-7

6e-7

1st breakthrough 2nd breakthrough

1st desorption 2nd desorption

25°C 25°C

150°C 150°C Type of Measurements:Adsorption capacityAdsorption kineticsDesorption temperature

Total vs. Working Capacity

−

−=CCClnQw

QCwt 0

v

B

0

eb κ

ρ

Time (s)

0 20 40 60 80 100

CO

2 Con

cent

ratio

n (x

10-4

g/cm

3 )

0.0

0.5

1.0

1.5

2.0

2.5

3.0

Amount of CO2 adsorbed at complete breakthroughAmount of CO2 adsorbed when CO2 concentrationin the eluted gas averages 1 vol%

Amount of CO2 eluted at complete breakthroughMeasured breakthrough curve dataBest-fit breakthrough using the Wheeler equation

Capacity and Rate Coefficient Data:we = 12.1 mg/g (Wheeler equation)kv = 414.2 min-1 (Wheeler equation)we = 19.3 mg/g (by integration at complete breakthrough)we = 10.2 mg/g (by integration at 1% breakthrough)

EDA-SBA-15 sorbent 0.229 gFeed gas: 15 vol% CO2 and balance

N2 at 25°CFlow rate: 40 sccm

Wheeler Eqn (Wood, 2002)

3

Performance – Dry vs. Wet

100% indicates capacity to full breakthrough. 1% indicates working capacity for total equivalent slip of 1% CO2.

Data pertain to CO2 in N2.

Wet corresponds to 2% H2O in gas mixture.

Little difference in capacity when water is added to the stream indicates good selectivity.

0

10

20

30

40

50

60

0 10 20 30 40 50 60

Wet - 1%Wet - 100%Dry - 1%Dry - 100%

Cap

acity

, mg

CO

2/g

CO2 Concentration, %

TGA-MS Analysis: Thermal Stability

Mass loss indicates significant desorption events.Mass spectrometry indicates what comes off sorbent.In He, EDA decomposes above 400oC.In air, EDA decomposes above 200oC.

70

80

90

100

5 15 25 35 45

TIME, min

MAS

S LO

SS, %

-2.5

-2

-1.5

-1

-0.5

0

MAS

S LO

SS R

ATE,

%/m

in

CO2 desorption100°C

EDA decomposition437°C

0.E+00

1.E-11

2.E-11

3.E-11

4.E-11

5 15 25 35 45

TIME, min

MS

COUN

T RA

TE

50

150

250

350

450

TEM

PERA

TURE

, C

1215161744

70

80

90

100

5 15 25 35 45

TIME, min

MAS

S LO

SS, %

-2.5

-2

-1.5

-1

-0.5

0

MAS

S LO

SS R

ATE,

%/m

in

CO2 desorption100°C

EDA decomposition437°C

0.E+00

1.E-11

2.E-11

3.E-11

4.E-11

5 15 25 35 45

TIME, min

MS

COUN

T RA

TE

50

150

250

350

450

TEM

PERA

TURE

, C

1215161744

FTIR - Reaction with CO2

Wavenumber (cm-1)

13001400150016001700

Abs

orba

nce

(a.u

.) 1696

14911602

1576

1472

1449

Regenerated sorbentAfter CO2 adsorption at 25°CAfter reaction with CO2 at 135°C

Free amine groups are abundant in regenerated sorbent.1602 cm-1 N-H deformation vibrations1449 cm-1 CH2 deformation vibrations

Upon exposure to CO2 at room temperature, the sorbent forms an intramolecular carbamate ammonium salt.

1576 cm-1 NH2+ deformation vibrations

When exposed to CO2 at higher temperature, the carbamate partially converts to ethylene urea.1696 cm-1 C=O stretch (Amide I band)1491 cm-1 Amide II band

IR band assignments are tentative. Experiments to confirm urea formation in progress.

Mechanism (Ethylene Diamine)

SiO2

O

Si

NH

NH2

O

O

Si

NH

NH2

O

O

Si

NH

NH2

OO

+ CO2

HNH+

N C

H+

e-

H

O-

=O

Stability of Zwitterionic compound (intramolecular carbamate) governs ease of regeneration and working capacity

FTIR – Effect of SO2 exposure(a) Sample exposed to ambient air

1578 cm-1 NH2+ deformation

1484 cm-1 unassigned1416 cm-1 unassigned1320 cm-1 CH2 wagging795.6 cm-1 Si-O-Si sym. stretch

(b) Sample exposed to 500ppm SO2, 21% CO2, and balance N2

Band structure same as (a) except for one new band:619.7 cm-1 SO2 bending in

-NH·SO2 complex1

Wavenumber (cm-1)

40060080010001200140016001800

Abso

rban

ce (a

.u.)

(a)

(b)

1578

795.

6

619.

7

1484

1416

1320

1574

796.

4

1481

1415

1315

It is believed that SO2 reacts with secondary amine to form a charge-transfer complex. Determination of the fate of the SO2 adduct upon heating is likely reversible based on Diaf, Garcia, and Beckman (1994).

1Vasiljev et al. (1995) Thin Solid Film, 261, 296-298 Qout

Condenser

DryerCompressor

StackGas

StackOutlet

CO2Injection

Unit 2

Unit 1

Heat forRegen.

RegenerationSorption

Parallel Bed: Thermal Swing

4

Economics

It is essential to perform economic analysis in tandem with technical discovery and development in order to guide work in direction of commercial viability.Preliminary economics suggest approach could be attractive with future expected improvements.Preliminary model is based on adaptation of carbon adsorber model available from EPA for VOC capture.Expecting posting of a new solid-based capture model from NETL soon.

Future WorkContinue to develop additional amine-modified materials.

Fill in operating temperature gaps in order to make applicable to many streams.

Complete economic analysis and use as a basis for R&D targets.Look at steam regeneration.

T, oC~0 ~70 110

2-Acylaminopyridine Propylamine* EDA

*Published work by NETL

Trend of higher working capacity

Also trend of more difficult regeneration

Various Approaches Possible

Si OOO

HN

R2N

ONH

NH2

O

Si OO

NHO

N

O

Si OOO

Si OO

NHN

N

O

Si OO

NHN

NHN

O

Si OO

NH

NH2

O

Si OO

NHN

NH2

O

Si OO

NHN

NN

O

Si OO

N

OH

O

Si OO

NH

N

O

Si OO

NH

N

• Target ligands result in facile proton transfer during H uptake

• Multifunctional ligands alsoattractive.

• Polymeric and dendrimericfunctionality are being examined

Immediate Targets

Propylene Diamine(determine impact of carbamate ring size)

Piperazine(determine impact of ligand rigidity

and amine degree)

Aminomethyl Imidazole(examine cyclic compounds with potential multifunction)

NH

NH2

NH2

N

NH

NH

NNH2

NH

Propyl Amine(verify comparison to NETL literature)

Various Polymeric Species(multifunctional approach)