Ab Initio Study of Hydrogen Storage on CNT Zhiyong Zhang, Henry - - PowerPoint PPT Presentation

Ab Initio Study of Hydrogen Storage on CNT

Zhiyong Zhang, Henry Liu, and KJ Cho Stanford University

Financial Support: GCEP (Global Climate and Energy Project)

Presented at the ICNT 2005, San Francisco

Rational Design of Hydrogen Storage Material

• High Storage Capacity: 6.5 wt%, 65

g/L

• Desorption Temperature: 60~120 °C
• Thermodynamically slightly stable

hydrogen storage systems are desired and can be achieved through material engineering.

• Reversibility dilemma: Reversibility

is usually associated with high- energy barrier for dissociation and adsorption.

• Solution: Engineered catalyst can

tune the energy barrier.

Multiscale Design of Nanomaterials for Hydrogen Storage Chemisorption: H2 + NM  NM-2H

• H-H bond breaking = 4.5 eV

 poor reversibility (too strong bonding) + high surface coverage ? stability of NM-2H states

Physisorption: H2 + NM  NM-H2

+ weak adsorption  good reversibility

• small surface coverage (much less than a few wt. %)

Controlled Catalytic Chemisorption

+ high surface coverage & good reversibility

Experimental results

• Dillon et al., Nature, 386, 377 (1997)

– Room temperature, 300 torr, soot with 0.1 wt% SWNT – 0.01 wt% hydrogen adsorption, equivalent to 5~10 wt% on pure SWNT.

• Ye et al., Appl. Phys. Lett. 74, 2307 (1997)

– 80K, 100 bars, pure sample. – 8.25 wt% hydrogen adsorption.

• Liu et al., Science 286, 1127 (1999)

– Room temperature, 100 MPa, pure sample. – 4.2 wt% hydrogen adsorption.

• Chen et al., Science 285, 91 (1999)

– Ambient pressure, Li or K doped carbon nanotubes. – 14 ~ 20 wt% hydrogen adsorption. FTIR show that H2 is dissociative. Other experiment suggests it is due to water adsorption. – Only 0.4 wt% hydrogen adsorption without doping.

• Nu¨tzenadel et al., Electrochem. Solid-State Lett. 2, 30 (1999)

– 0.39 wt%, electrochemical hydrogen storage.

Controlled Hydrogen Bonding on Carbon Nanotubes

H Prediction (solid curves) Ab-initio results ( and )

Stable adsorption stability of chemisorbed hydrogen (NM-2H) states relative to H2 gas can be tuned by nanotube size: (8,0) - (12,0) CNTs would have enough binding energy (up to 0.5 eV) per H2 molecule

• S. Park, D. Srivastava, and K. Cho, "Local reactivity of fullerenes

and nano-device applications," Nanotechnology 12, 245 (2001).

Optimal Configurations of Adsorbed Hydrogen Pairs

C1 C4 C3 C2 C5

LDA, US LDA, PAW GGA, US GGA, PAW C1C2: 0.263 eV 0.291 eV 0.142 eV 0.222 eV C1C3: 0.018 eV 0.046 eV

• 0.374 eV -0.264 eV

C1C4: -0.045 eV

• 0.017 eV
• 0.403 eV -0.306 eV

C1C5: -0.317 eV

• 0.289 eV
• 0.483 eV -0.402 eV

Projected LDOS H C1 C4 C3 C5 C2

Optimal Configurations of Adsorbed Hydrogen Pairs

2nd pairs: 0.709 eV 3rd pairs: 1.281 eV 4th pairs: 2.138 eV

Coverage Dependence of Binding Energy

• The maximum chemi-sorption

capacity is around 50% or less

• Binding is very strong at low

coverage.

Coverage Dependence of Binding Energy

• The maximum coverage is around 75%

with external binding.

• The binding energies are in a narrow

range of 0.4 eV ~ 0.6 eV (0.12 ~ 0.22 for GGA and PAW)

Binding Energy vs. Coverage, Most Probable Coverage Pattern

• 3.0
• 2.0
• 1.0

0.0 1.0 0.2 0.4 0.6 0.8 1 Coverage Binding Energy per Hydorgen Molecule H2 (eV) Avearge Incremental Incremental, GGA, PAW results

Intensity (arb. units)

288eV 286 284 282

Binding energy (eV)

(1) (2) (3) (4) (5) (6) (7)

C1s XPS

(1) T=40

• C

(2) T=200

• C

(3) T=300

• C

(4) T=400

• C

(5) T=470

• C

(6) T=580

• C

(7) T=750

• C

Intensity (arb. units)

288eV 286 284 282

Binding energy (eV)

C1s XPS

Absorption of Atomic Hydrogen on SWCNT

• Atomic source of H for hydrogenation.
• Diameter of SWCNT in the range of 1nm to 1,8 nm.
• 65±15% (5.1±1.2wt%) of hydrogenation.
• Stable in the temperature range of 300~600°C.

Carbon Nanotube Sorption Science External Peer Review of NREL Activities, January 19-23, 2004

• No hydrogen storage was observed in pure single-walled carbon nanotubes.
• Roughly 3 wt.% was measured in metal-doped nanotubes at room temperature.

Experimental results, Dissociative H2 Adsorption

C1 C2 C3 C4 C5 C6

H2 Dissociative Adsorption on CNT

• 0.5

0.5 1 1.5 2 2.5 3 Reaction Coordinate Energy (eV) C1C4 C2C3 C1C2

C7

• 1.5
• 1
• 0.5

0.5 1 1.5 2 2.5 3 2 4 6 8 10 C4C7+C2C3 C2C3

• 0.6
• 0.4
• 0.2

0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 2 4 6 8 10 C1C4 C2C3+C1C4

H2 Dissociation on CNT

C2 C1 C3 N C5 C6 C7 C8

• 0.500

0.000 0.500 1.000 1.500 2 4 6 8 10 12 14

C5C8/Doped CNT, Physisorption C1C4 Pure CNT C5C8/Doped CNT

H2 Dissociation on N-Doped CNT

C1 C2 C3 C4 C5 C6 C7

• 1
• 0.5

0.5 1 1.5 2 2.5 3 3.5 2 4 6 8 10 12 14 C1C4 to C2C3 C1 to C2 then C4 to C3 C1/C1, C4 to C3 C3/C3, C6 to C2 80+C2toC3

H Diffusion CNT

Conclusion

• The hydrogen biding energy strongly depends
• n the coverage on the CNT. Computational

results indicate that the binding energies fall within a narrow range of energy in the most likely coverage pattern.

• We have identified the most likely dissociative

adsorption pathway and studied the diffusion pathway of hydrogen on CNT.

• Computational results indicate that doping the

CNT can substantially lower the dissociative adsorption pathway and thus improve the hydrogen storage kinetics.