development of robust hydrogen separation membranes

May 2011

Development of Robust Hydrogen Separation MembranesNational Energy Technology Laboratory-Regional University AllianceNETL-RUA

2011 DOE Hydrogen Program Review

This presentation does not contain any proprietary, confidential, or otherwise restricted information.

PD008

2

H2 Separation Membrane Team Members (Collaborators)

NETL Research FacultyDr. Andrew Gellman, CMU (ChemE)

Dr. James Miller, CMU (ChemE)Dr. Petro Kondratyuk, CMU (Chem)

Dr Michael Widom, CMU (Phys)Dr. Brian Gleeson, Pitt (MatSci)Dr. Ted Oyama, VT (ChemE)

U.S. DOE - NETLDr. Bryan Morreale, PIT (ChemE)Dr. Bret Howard, PIT (InorgChem)

Dr. David Alman, ALB (MatSci)Dr Omer Dogan, ALB (MatSci)Dr Michael Gao, ALB (MatSci)

NETL Site Support ContractorsDr. Mike Ciocco, PIT

3

H2 Separation Membrane Team Structure (Approach)

4

Overview

Timeline• Project start date: 10/1/2010• Project end date: 9/30/2011• Percent complete: 58%

Budget

• FY11 funding: $1,361k• FY10 Funding: $681k• FY09 Funding: $746k• FY08 Funding: $1,000k• FY07 Funding: $1,230k

Barriers(1)

• (G) H2 Embrittlement• (H) Thermal cycling• (I) Poisoning of catalytic surface• (J) Loss of structural integrity and

performance

Partners

• Carnegie Mellon University• University of Pittsburgh• Virginia Tech• NCCC, Wilsonville

(1) 2008 Hydrogen from Coal Program: Research, Development and Demonstration Plan

5

Background(Relevance)

• Overall goal– Development of robust

hydrogen separation membranes for integration into coal conversion processes, including integrated WGSMR

Studies suggest that incorporating separation membranes into coal conversion processes can reduce costs by 8%.

6

Hydrogen separation performance targets(Relevance)

Hydrogen from Coal Program, RD&D Plan, External Draft, U.S. Department of Energy, Office of Fossil Energy, NETL, September 2008

7

Robust Metal Membrane Development(Approach)

• Develop an advanced membrane system for hydrogen separation

– High activity for hydrogen dissociation– High H-atom permeability– Resistance to S-poisoning– Mechanically robust

• Apply computation and experiment

– Characterization of H2 dissociation kinetics

– Identification of third component that broadens the high-permeability region

– Characterization of corrosion kinetics/products

– Demonstrate performance of promising candidates

– Coupon testing of promising candidate materials at NCCC

- layer Membrane Concept

Corrosion Resistant

Barrier

Catalytic

CoatingHighly Permeable

Substrate

Catalytic Layer

Multi- layer Membrane Concept

Corrosion Resistant

Barrier

CoatingHighly Permeable

Substrate

Catalytic Layer

8

Facilities & Capabilities(Approach)

• 3 Membrane Test Rigs– Continuous, bench-scale units– T to 1000°C, P to 1000 psi

• 2 Laboratory Membrane Screening Rigs

– Continuous, lab-scale units– T to 1000oC, P to 30 psi

• Materials Lab– Deposition chamber(s)– High-T box and annealing ovens– XRD w/hot-stage– SEM w/EDS, EBSD– TGA for use with H2S– Imaging XPS– He+ ion scattering

• High Throughput Materials Science– Deposition tools– Spatially resolved characterization

• Computation– DFT, Kinetic Monte-Carlo, COMSOL

CFD

TI

GC

FCV FCV

PCV PCV

H2 Ar

Heater

Membrane

Heater

Heater

H /He2 Ar

Ar/H2

Ar/H2

9

Robust Metal Membrane Development(Approach)

Alloying Pd with minor component(s) can:• improve mechanical robustness and sulfur tolerance

• maintain Pd’s surface activity and high permeance,

PdCu

Ag Au Ce Cu Y

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

Ag (20

%)

Ag (23

%)

Ag (27

%)

Ag (30

%)

Ag (52

%)

Au (5%

)

Au (20

%)

Au (40

%)

Au (55

%)

B (0.5%

)

Ce (7.7

%)

Ce (12

.7%)

Cu (10

%)

Cu (40

%)

Cu (45

%)

Cu (55

%)

Cu (70

%)

Ni (10%

)

Ru (5%

)

Y (6.6%

)

Y (10%

)

Alloy Composition, wt% [Balance Pd]

Norm

aliz

ed P

erfo

rman

ce (k

' Allo

y / k

' Pd)

10

Robust Metal Membrane Development(Background: previous results)

Two mechanisms of performance deterioration caused by exposure to H2S: (L) growth of low permeabilty sulfide scale on Pd and (R) catalytic poisoning of alloy surface.

0 2 4 6 8 10 120

4

8

12

16 H2 + H2SH2

H 2 flu

x (c

m3 /c

m2 /m

in)

Test duration (hours)

Pd4 S thickness (µm

)

0

1

2

3

4

5

6

7

0 2 4 6 8 10 120

4

8

12

16

H2 + H2SH2

H 2 Flu

x (c

m3 /c

m2 /m

in)

Test Duration (hrs)

Pd: 25µ, 350oC

Pd47Cu53: 25µ, 350oC

11

1000ppm H2S in H2, steady state flux typically obtained after 5 days on stream

Robust Metal Membrane Development(Background: previous results)

True S-tolerance exists in the PdCu binary, but at conditions of low “base permeance”: High Cu content + elevated temperature

Illustrates potential of Pd alloys and teaches how to frame the problem

Current activities:• basic understanding of PdCu• expansion to ternary alloys

12

Material Thermodynamics(Technical Accomplishments – Stability & Scale Prediction)

log pO2 (atm)0-5-10-15-20-25-30

0

log

pS2

(atm

)

-5

-10

-15

-20

-25

-30

Cu

Cu2S

CuS

Cu2O CuOPd

Pd4S

Pd16S7

PdO

PdS

aCu = 1; aPd = 1H2-10%He-0.1%H2S

log(p S2)=-15.8

gas

Pd16S7

PdPd4S

Predicted Stabilitygas

Cu2S

Cu

gas

Pd16S7

Pd4S

Pd70Cu30

22 HSM SH M bba ba +↔+

13

Robust Metal Membrane Development(Technical Accomplishments – Stability, Scale Growth & Transport)

• Direct measurement of Pd4S permeability– In the presence of H2, appears

to follow Sievert’s law– Permeability of Pd4S is ~10x

less than Pd

• H2S causes incremental flux decline– Likely mechanism is site

blocking (link to H-D exchange work)

0 50 100 150 2000.00

0.02

0.04

0.06

0.08

~14 µm Pd4S/Pd~6 µm Pd4S/Pd

~2 µm Pd4S/Pd

~1 µm Pd4S/Pd

Pd

H 2 Flu

x (m

ol/m

2 /s)

∆PH21/2 (Pa1/2)

0 50 100 150 2000.00

0.02

0.04

0.06

0.08

~14 µm Pd4S/Pd~6 µm Pd4S/Pd

~2 µm Pd4S/Pd

~1 µm Pd4S/Pd

Pd

H 2 Flu

x (m

ol/m

2 /s)

∆PH21/2 (Pa1/2)

0.008

0.006

0.004

0.002

0.000

H 2 Flu

x (m

ol/m

2 /s)

Pure H2

H2 + H2S

14

Robust Metal Membrane Development(Technical Accomplishments – Catalytic Activity, Gas-Scale Interface)

H2-D2 exchange provides fundamental insight into hydrogen dissociation on model membrane surfaces—across alloy composition and in the presence of H2S

H2 + D2 2HD

15


H2 + D2 2HDH2S adsorption onto Pd4S reduces effective surface area for H2 adsorption

Pd4S

400 500 600 700 800 9000

2

4

6

8

10

2000 ppm H2S1000 ppm H2S500 ppm H2S200 ppm H2S100 ppm H2S50 ppm H2S

HD fl

ow ra

te (m

L/m

in)

Temperature (K)0 500 1000 1500 2000

0.0

0.2

0.4

0.6

0.8

1.0

Norm

aliz

ed e

ffect

ive

Pd4S

sur

face

are

aH2S concentration (ppm)

16


Interpret H2-D2 exchange data via a micro-kinetic model to estimate activation barriers and pre-exponentials—a rational basis for comparing membrane surface activities

H2 + D2 2HD

0.0012 0.0016 0.0020 0.0024-24

-22

-20

-18

-16

-14

CuAdsorption-limited

mol

ln(k

ads)

1/T (1/K)

Eads = 0.58 ± 0.01 eV

νads = 10 -3.1 ± 0.1

m2 s Pa

0.0020 0.0024 0.0028 0.0032-12

-10

-8

-6

-4

-2

0

PdDesorption-limited

m2smol

β-Pd-hydride

ln(k

des)

1/T (1/K)

Edes = 0.54 ± 0.01 eV

νdes = 10 4.2 ± 0.1

Edes = 0.38 ± 0.02 eV

νdes = 10 2.0 ± 0.3

α-Pd-hydride

m2smol

17


Barriers to dissociative adsorption (upstream membrane surface) and recombinative desorption (downstream membrane surface) both depend on surface composition

α-Pd

β-Pd

Pd4S

0.0

Ene

rgy

(eV

)

BCC Pd47Cu53

H2(g)

H2*

2H(ad)

FCC Pd47Cu53

-1.0

1.0

0.5

-0.5

Cu

H2 + D2 2HD

18

Coupon exposure test conditions (average)• Fuel: Coal •Duration: 5 weeks• Temperature: 400ºC•Hydrogen concentration: 7.5% (dry basis) •Hydrogen sulfide concentration: 250 ppm

Coupon types exposed•Membrane (NETL Pittsburgh)

• 19 samples (coupons approximately 100 µm thick)• Pd, PdCu (5 samples), PdCuX (4 samples), PdX (7 samples), 2 misc. alloys

• Supported membrane (WPI)• 6 samples (composite Pd and PdCu on porous SS)

• Structural alloy (NETL Albany)• 4 samples

Metal coupons consisting of materials of interest to NETL were exposed to an actual gasifier syngas at the National Carbon Capture Center in Wilsonville, Alabama to evaluate real-world corrosion effects versus those observed under laboratory conditions.

Robust Metal Membrane Development(Approach – Stability, Scale Growth & Transport)

19

Pd 80wt%Pd-Cu 60wt%Pd-Cu

XRD: Pd4S only phase detected(Pd coupon completely sulfidedand fractured into pieces)

XRD: Pd13Cu3S with trace of Pd4S (thick sulfide corrosion layer - metal not detected)

XRD: Only B2 PdCu detected (sulfides visible by SEM are below detection limit of standard XRD scan conditions used)

Robust Metal Membrane Development(Technical Accomplishments – Stability, Scale Growth & Transport)

NCCC Coupon Testing: preliminary results• B2-structured alloys suffered much less corrosion• Particulates observed to impact surface• Trace component effects appear important

• Arsenic detected in several coupons so far• Surface morphology for some alloys very different than observed in lab tests

20

Robust Metal Membrane Development(Approach – New Materials Development)

Add a third component to PdCu to:•Broaden B2 phase field•Enhance H2 permeability• Improve sulfur poisoning resistance

PdCu

21

Cu50Pd(50-x)Mx

4d TM

CuPd CuM

Calculation of enthalpy of formation using ab initio density functional theory (DFT) methods identifies most stable ternary alloys

Robust Metal Membrane Development(Technical Accomplishments – New Materials Development)

22

Robust Metal Membrane Development(Approach – New Materials Development)

Strategy:• Replace Pd with M at

“BCC boundaries”• Measure BCC region

broadening w/HT XRD• Evaluate for corrosion

resistance• Measure flux (future)• Characterize for

dissociation activity (future)

23

Cu50Pd44M6 alloys that displayed a B2 phase at room temperature (L) were studied at elevated temperatures

Inte

nsity

Cu50Pd44Ti6Cu50Pd44Y6Cu50Pd44La6Cu50Pd44Al6Cu50Pd44Mg6


24


Addition of third component (M) expands range of B2 stability

Alloy Ts ( C) Tf ( C)

Cu56Pd44 ~530 ~580Cu-Pd-Mg ~640 >860

Cu-Pd-Y 575-600 675-700Cu-Pd-Al 650-675 825-850Cu-Pd-Ti <400 775-800

Cu-Pd-La <400 625-650

25

• Temperature = 500°C• Process gas composition

– 50%H2, 30%CO2, 1%CO, 19%H2O, no H2S– H2S will be added in future trials


Except for La, no significant oxidation was observed

26

Computational study of surface segregation of Y on

CuPd (011) alloy


Y in Pd bulk position

DFT ( )bulkYE ,

Y in Pd surface position

DFT( )surfYE , ( ) ( )bulkYEsurfYEE surf

segr ,, −=

Y in Pd sub-surface position

DFT ( )surfsubYE −, ( ) ( )bulkYEsurfsubYEE surfsub

segr ,, −−=−

eVE surfsubsegr 33.0−=−eVE surf

segr 23.0−=

Y atoms migrate to the sub-surface

>

27

• Screening all possible compositions is prohibitive– Prepare Composition Spread Alloy Films (CSAFs) with all

possible compositions on a single substrate

AxB1-x

AxByC1-x-y

BA

C• Scientific/technical challenges

– Tools for preparation and characterization of CSAFs are not commercially available and must be developed internally.

Robust Metal Membrane Development(Technical Accomplishments – High Throughput Methods Development)

28

− Physical vapor deposition by evaporation.

− Flux distribution across surface determined by precise placement of filaments.

Mo

Cu

0.0035 nm/s3 min 8 min

16 min 0.0035 nm/s


29

• Prepare 100 nm-thick PdxCu1-x CSAF on Mo substrate

• Anneal at 800 K, then quench

• Characterize structure by Electron Backscatter Diffraction

• Phase diagram matches those reported in the literature

• Exciting capability for screening new membrane alloys and their responses to contaminants!


30

HD

Au

Cu

Pd

• PdCuAu CSAF• H2-D2 exchange at ~130 oC• XPS used to analyze the composition of CSAF• Au and Cu suppress HD exchange.


Rapid characterization of H2 dissociation kinetics over broad composition space of relevance to the membrane application

31


H2-D2 exchange activity (from previous slide)

Characterization of surface segregation

HDactivity

high

low

H2 + D2 2HD

Cu

PdAu

Au(LEIS)/Au(XPS)

32

• Computational evaluation of PdCuX alloys’ surface compositions and interactions with H2 and important contaminant molecules

• Fabrication of PdCuX alloys for experimental characterization of− Corrosion resistance (in process)− H2 permeance (according to protocol, just started)− Coupon exposure tests at NCCC (in process)

• Application of HT methods to measure PdCuX alloy properties across complex composition space− H2 dissociation activity in a contaminant environment− Phase stability in a contaminant environment

• Development of HT permeation reactor (started)

• Fabrication and demonstration of membrane systems

Robust Metal Membrane Development(Proposed future work)

33

Summary

• Diverse, cross-functional team…– Materials scientists, surface scientist, engineers students– NETL, CMU, Pitt, Virginia Tech

• …with a wide range of capabilities– Computational chemistry– Alloy fabrication– Materials characterization– Surface analysis– Membrane screening– Membrane performance testing

• Development of design basis for robust metal membrane– Complete, fundamental understanding of PdCu

• Identification of dual S-deactivation mechanisms• Characterization of corrosion products’ activity, permeability• Measurement of hydrogen dissociation activity across alloys, and in presence of H2S

– Extension to PdCuM• Computational evaluation of stability, potential for interactions for S• Preparation, corrosion characterization of alloys designed to expand “high permeabiity window”

• New capability in high-throughput materials and surface science– Rapid screening/understanding of PdCuM properties across composition space

development of robust hydrogen separation membranes

Documents