Nanomaterials Engineering for

Hydrogen Storage

Bruce Clemens’ Group

Material Science

Ranadeep Bhowmick

Cara Beasley

Hongjie Dai’s Group

Chemistry

Dr. Liying Jiao

Anders Nilsson’s Group

Photon Science/SLAC

Srivats Rajasekaran

Dr. Daniel Friebel

Dr. Hirohito Ogasawara

DANIEL FRIEBEL

Hydrogenation of Nanotubes

H-H bond energy 4.5 eV

Adsorption Energy Decreases with diameterPark et.al Nano Lett. 3, 1273 (2003)

chemical shift of C1s peak due to C-H bond formation

Clean

H-treated

Graphite

0.65 eV

Single wall Carbon Nanotube

Nikitin et.al Surf. Sci. 602, 2575 (2008) Nikitin et.al Phys. Rev. Lett. 95, 225507 (2005)

Atomic Hydrogen Treatment

Maximum Hydrogenation

H

C C

H

C C

Decomposition of C1s XPS for hydrogenated SWCN film

Nikitin et.al Nano lett. 8, 162 (2008)

Ipeak 1:(Ipeak 2+Ipeak3) = 1:10

which corresponds to ~7 wt % of SWCN hydrogen capacity

Catalyst to Produce Atomic H

Motives

Molecular hydrogenation of SWNTs with

catalyst to split the hydrogen molecule.

Detection of C-H bonds formed in the

process and prove “spillover” mechanism.

Possible electrochemical hydrogenation

pathway?

Motives

Molecular hydrogenation of SWNTs with

catalyst to split the hydrogen molecule.

Detection of C-H bonds formed in the

process and prove “spillover” mechanism.

Possible electrochemical hydrogenation

pathway?

Sample studied and Preparation

methods As grown CVD mat samples

◦ 2Å thick cobalt metal deposited on silicon oxide wafers as catalyst, isopropanol as carbon source, growth temperature was varied from 700°C to 800°C.

Commercially obtained HiPCO samples◦ SWNTs dispersed in isopropanol (1mg/10ml),

sonicated for 15 minutes, and then spin cast on a quartz slide, annealed to get rid of solvent.

◦ Langmuir Blodgett (LB) method of centrifugation to unbundle the nanotubes so as to have maximum catalyst coverage (Platinum (Pt)) on SWNTs.

LB density gradient centrifugation

methodology Done to unbundle the nanotubes so that when Pt is

sputtered we would have maximum Pt coverage on SWNTs.

HiPCO single-walled carbon nanotubes (SWNTs) suspended

by sodium cholate were layered onto a 5/10/15/20/60%

Iodixanol step gradient

◦ Iodixanol is a commonly used density gradient medium

Centrifugation at ~300,000g for one hour yields a

distribution of SWNTs along the length of the gradient,

stopping at the boundary of the 60% Iodixanol

Iodixanol – 60% ρ=1.32 g/mL

5%

10%

15%

20%

60%

SWNTs

Results of separation process

After separation, SWNTs were distributed throughout the density gradient and stopped by the high density 60% iodixanol layer

Fractions of 100 µL were carefully extracted from top to bottom (fractions 1-24)

SWNT concentration in each fraction was normalized by adjusting the same to the same optical density

50 kRPM

1 hr

f1

f24

Results of separation process Photoluminiscence spectra (PL) indicate

f6 and f7 are the least bundled.

The increase in NIR PL and Raman

scattering intensity in fractions 2-6 can be

explained by length separation.

Beyond 7, bundling of nanotubes occur.

f4f3 f8f7f6f5 f12f11f10f9 f16f15f14f13 f18f17f4f3 f8f7f6f5 f12f11f10f9 f16f15f14f13 f18f17

NIR PL image of separated SWNTs

at normalized OD

0

500

1 000

1 500

2 000

2 500

3 000

3 500

4 000

4 500

5 000

5 500

6 000

6 500

Inte

nsi

ty (

cn

t/sec)

150 200 250 300

Raman Shift (cm-1)

0

500

1 000

1 500

2 000

2 500

3 000

3 500

4 000

4 500

5 000

5 500

6 000

6 500

Inte

nsi

ty (

cn

t/sec)

1 550 1 600

Raman Shift (cm-1)

f2

f24

Starting material

Stokes shift (cm-1)

Radial

Breathing ModeGraphitic

Mode

5 10 15 200

200

400

600

800

1000

1200

1400

1600

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

2.2

Ra

ma

n In

ten

sity [a

.u.]

Fraction #

RBM

G Band

Re

lative

QY

[a

. u

.]

Relative QY

Raman Scattering spectra of SWNT

fractions 1-24 at normalized OD

Sample specifications for

measurements The SWNTs were sputter deposited with Pt nanoparticles.

The diameter range of SWNTs are 0.8-1.6nm

SEM images of SWNTs (a, b) Dense and sparse distribution of as grown SWNTs (c) Spincast HiPCO SWNTs

(d) LB Film (e) AFM of LB Film (f) Representative Raman spectrum of HiPCO SWNT

Measurements performed

In-situ 4-probe conductivity measurements in the presence of hydrogen exposure ~700 torr.

Ex-situ X-Ray Photoelectron Spectroscopy (XPS) measurements before and after exposing Pt-SWNT composites to hydrogen ~ 8.25bar (120psi).

◦ XPS was done at Beam line 13-2, elliptical undulator beam line at Stanford Synchrotron Radiation Lab (SSRL) with Scienta R3000 electron analyzer with a sensitivity of 250meV.

Conductivity measurements

(Top-Left) Change in resistance with exposure time for hydrogen pressure of 700torr for

HiPCO samples with 0.6nm deposited Pt and without Pt. (Bottom-Left) Change in

resistance with exposure time for hydrogen pressure of 700torr for pure Pt. (Top-Right)

Change in resistance for Pt-HiPCO composite with different pumping cycles. (Bottom-

Right) Change in resistance for 0.6nm pure Pt film with different pumping cycles with

cycle indicated.

• Conductance (resistance) of Pt-SWNT decreases (increases)

4x times that of SWNT.

• Pure Pt film resistance on

exposure to hydrogen increases

and reaches a constant value.

• Cycling hydrogen exposure,

pumping and exposure to air

indicates little change in

resistance of Pt-SWNT while for

Pure Pt film, there are major

changes.

OBSERVATIONS

Conductivity measurements for Pt-SWNT

for different Pt thickness and different

SWNT samples

(left) Change in resistance of Pt-SWNT for different deposited Pt thickness, (Right) Change in

resistance of Pt-SWNT for different SWNT films with deposited Pt thickness = 0.6nm

Conductance (Resistance), an intrinsic property of the Pt-SWNT

composites decrease (increase) on exposure to hydrogen gas in a

way different than pure Pt or pure SWNTs with maximal change

observed for 0.6nm Pt deposited on them and for LB film samples.

XPS measurements• XPS measurements were performed on LB films and as grown CVD films with 0.6nm deposited Pt (maximal resistance change samples.

• Measurements performed at Stanford Synchroton Radiation Lab, at elliptical

undulator 13-2 beam line with Scienta R3000 electron analyzer with energy

resolution of 250meV.

C1s XPS (photon energy – 400eV) peak before and after hydrogen exposure for Pt-SWNTs;

Shirley Background subtracted and Normalized spectra wrt peak height to enhance the peak

Shape difference before and after the process of hydrogen exposure

(Left) for CVD grown films sputtered with Pt (Right) LB films sputtered with Pt

Deconvolution of XPS peak

Blue open circles indicate raw data, red indicates total fit, Grey, green and orange indicate

sp2, sp3 and the 3rd peak resply. (The individual peak fit components are offset

downwards from the raw data and the total fit); C1s could be peak fitted with one sp2

peak. (Left) CVD grown film; (Right) LB film

Interpretations and Conclusions of XPS

Peak fitting XPS peak before hydrogen exposure has FWHM 1.4eV for the as

grown CNTs while it is 1.15eV for the LB films. The broader nature of the as grown CVD samples is because the CVD samples had more defects than the LB films – could be fitted with one peak.

After exposure to hydrogen at a pressure of 8.27 bars (120 psi), the FWHM of C1s peak for as grown CNTs increased to 1.5eV while that for LB films increased to 1.35eV.

The relative weights of the sp2 (sp3) peaks are 0.84 (0.17) for the LB film and 0.87 (0.13) for the as-grown film after hydrogenation, hence hydrogen uptake is 1.2wt% for LB films and 1 wt% for the as-grown films.

The 3rd peak arises from a metal-to-semiconductor transition of the nanotubes that is induced by the hydrogenation (which could be seen in conductivity measurements). The accompanying decrease of the electric conductivity can cause a reduction of the core hole screening, resulting in a ~3.3eV higher final state energy. Alternatively, the creation of a band gap can give rise to a shake-up line which leads to less screened final state.

Interpretation and Conclusion

In-situ conductivity

measurements

Ex-situ XPS

measurements

Resistance increase on hydrogen treatment for Pt-SWNTs and not for

pure SWNTs

Hybridization change for some

carbon atoms from sp2 to

sp3Hybridization change explains resistance

increase

Insight into kinetics of spillover

mechanismFormation of C-H bonds on SWNTs

through catalyst Pt nanoparticles!

Motives

Molecular hydrogenation of SWNTs with

catalyst to split the hydrogen molecule.

Detection of C-H bonds formed in the

process and prove “spillover” mechanism.

Possible electrochemical hydrogenation

pathway?

Electrochemical pathway

exploration Cyclic voltamograms (CV) were performed on Pt-SWNTs with Ag-AgCl

reference electrode, Pt wire counter electrode and 0.05M H2SO4electrolyte solution (dE/dt = 10mV/s).

H2 evolution

O2 evolution

C-H or Pt-H formationH2 evolution

O2 evolution

Hydrogen reduction happes at -0.8V for Pt-free sample while it happens at -0.4V

for sample with Pt.

➢A positive current peak(hydrogen oxidation) is seen only in sample with Pt and this peak increases with the decreasing cathodic potential limits. Could possibly

mean formation of C-H bonds.

➢Hydrogen oxidation currenpeak appears in a potential region where we would also expect the oxidation of H adsorbed at the Pt surface, the interpretation of

the current-voltage curves alone will not be unambiguous – need for spectroscopy

Spectroscopic studies of

electrochemical treatment Measure ex-situ C1s XPS by dipping sample into electrolyte before

and after application of potential.

With 2 electrode set up apply -1V wrt to Ag/AgCl reference electrode for 10 min.

Peak broadens when

sample immersed in

electrolyte –inconclusive

as to whether C-H

bond formed

Other measurement plans

X Ray Raman Spectroscopy (XRS) – A droplet cell would be used

to measure K-edge XRS (inelastic scattering) during application of

electrochemical potential. Hard X rays of 6-7KeV would be used

for measurement.

High-pressure Inelastic X ray Scattering spectra for

graphite in horizontal and vertical directions

plotted as normalized scattered intensity versus

energy loss at different applied pressures.

Ref -W. Mao et al., Science, Vol. 302. no. 5644, pp. 425

- 427

Schematic droplet cell

Summary

We have demonstrated using conductivity and XPS measurements that a spillover mechanism is feasible to hydrogenate Pt covered carbon nanotubes with molecular hydrogen

Using non bundle nanotubes we have demonstrated 1.2wt%

From electrochemical measurements there are indications for potential hydrogenation but needs to be confirmed