Desorption mechanism of hydrogen isotope from metal oxides

Contents1. Background2. Experimental system and Mechanism3. Results and discussion4. Conclusions

Yasuhisa Oya, Ryushi Jinzenji, Takuji Oda and Satoru TanakaThe University of Tokyo

Background（ tritium decontamination)

Methods:surface washing, thermal desorption, light irradiation etc.

•For the safety of a fusion reactor， effective decontamination methods from construction materials of a reactor（ such as piping materials, vacuum chamber and so on） are required.

Advantage of light irradiationEffective for strongly bonded OH ， non-contact， selectivity of reaction products.

• To understand a desorption behavior of hydrogen isotope from metal oxides and construct desorption models of desorbates from those surfaces.

• To estimate the effect of electron transition between components（ Fe2O3-TiO2) on the behavior of desorbates．

Purpose

Current situation：Desorption process by light irradiation has not understood yet.

Different species of H2O adsorption on metal oxidesThere are several chemical form of H2O adsorbed on metal oxides.

MM

H

O

M

H

O

M

H

O

M

H

OM

H

O

M

H

O

M

H

O

M

H

O M

H

O

M

H

O

M

H

O

M

H

OM

H

O

M

H

O

M

H

O

M

H

OM

H

O

M

H

O

M

H

O

M

H

O

H OH

HO HHO H

HO H

HO H

H

O H

HO H

HO H

HO HH

O H

HO H

HO HH

O H

Molecular adsorption [Hydrogen bonding]

M

H

O

M

H

O

M

H

O

M

H

OM

H

O

M

H

O

M

H

O

M

H

OMMMM

HO H

Dissociative adsorption （ chemical form:OH）

MMM

H

O H

HO H

OO

Molecular adsorption（ intermolecular force）

Light irradiation

HO

Thin metal oxide film

Desorption

Characteristic of the two main desorption process

Thermal desorption Photon-stimulated desorption (PSD)

M

H

O

M

H

O

H

O H

Negative charging

M

H

O

M

H

O

Vertical

vibration

e-e-

HO H

OH-TiO2

Energy4σ

1π

3σ

occupied

Unoccupied

• Vertical vibration• 1-photon process• Higher translational energy

than that via thermal process

• Random vibration• Thermal equilibrium

• To understand a desorption behavior of hydrogen isotopes which are adsorbed on metal oxides in several chemical forms.

Methods In order to construct the desorption model of hydrogen isotope,

Analysis of desorption behavior•Velocity distribution•Desorption amount

•Laser power dependency•Wave length dependency•Adsorption state

By using TOF technique, we can obtain information about

•Velocity distribution of desorbed species from the surface

•Analysis of chemical form and desorption amount.

•Desorption behavior

•Adsorption state etc.

Time of Flight Mass Spectroscopy(TOF-MS)

Correspondence of experiments to adsorption states

Experimental conditions

２４℃

２４℃

Sample temperature

H2O exposure in the chamber

Estimated adsorption state （ chemical form： H2O）

○physical （ multilayer）

△dissociative （ chemical form:OH）

◎ physical（ a few layers）

△ dissociative (chemical form:OH）

◎dissociative (chemical form:OH）

１５０℃ TiO2

Fe2O3 ◎ physical（ a few

layers）

△ dissociative (chemical form:OH）

Exposure

（ 2x10-5Pa)

No exposure

(4x10-6Pa）

Detector and MCP

Sample and Anode

Flight tube

Laser for desorptionProbe laser

Time of Flight Mass Spectroscopy

Ionization laser:266nm,43mJLaser for desorption:355and 430nm,0.8-4.8mJTotal pressure in the chamber : 6~8x10-6PaPartial pressure of H2O : 3~4x10-6Pa

Repulsion between electrons by space-charge effect in MCP

⇒Causing the width of TOF spectrum

v [m/s]

Desorption amount at each delay time is estimated as follows;

MCP Detector

Diffusion(Thermal and concentration gradient）

Velocity error at the ionization point is supposed to be zero

t

Area： Q　 (charge）I [A]

v

Q

v

vQrIdvdf

)()()(/)(

2

fluxjacobian

3000 4000 5000 6000 7000 8000

- 1.0

- 0.8

- 0.6

- 0.4

- 0.2

0.0

0.2

0.4

0.6

Typical TOF spectrum of desorbates from metal oxidesurface

e/ m =18 (H2O)

e/ m =17 (OH)

e/ m =16

e/ m =2Inte

nsit

y /

mV

Time of Flight / ns

Principle of measurement of velocity distributions

acceleration

( v=s/τ)

( v=s/τ)

Velocity distribution

0 500 1000 1500 20000.000

0.005

0.010

0.015

0.020

355nm 0.8- 3.2mJ

Fe- Ti 5% MB fit

f(v)

velocity / m s- 1

A08 B12 C16 D20 E24 F28 G32

0.0 0.8 1.6 2.4 3.20

5

10

15

20

25

y=9.7x- 4.7

D

esorp

tion

amoun

t /

arb.

unit

Laser intensity / mJ

355nm without H2O exposure

Fe- Ti (5%)

Desorption amount of desorbates

Desorption amount can be estimated as the area of velocity distribution.

By calculating the desorption amount , we can estimate the laser intensity dependence for desorption amount obtained in various experimental conditions.

Sample preparation of TiO2, Fe2O3, Fe-Ti(5at%) oxide.

Ti Fe Fe-Ti(5at%)

Purity [%] 99.5 99.5 99.9

Thickness [mm] 0.3 0.25 1

Oxidation time [hour] 2 2 2

Partial pressure of O2

[Pa]10 10 10

Ti, Fe and Fe-Ti alloy(5at%) were used as samples. These samples were oxidized in the vacuum chamber and thin oxide film was formed on the surface of these metals.

0.8 1.6 2.4 3.2 4.0 4.8

0

2

4

6

8

10

12

y=0.9exp(x/ 1.9) y=0.2exp(x/ 1.4)

H2O exposure at 2 x 10- 5Pa

Heated at 150℃:Desorbed H

2Os

were not observed

Ref at 4 x 10- 6Pa

Sample : TiO2, λ :430 nm

Des

orp

tion

amoun

t /

arb.

unit

Laser intensity / mJ

0.0 0.8 1.6 2.4 3.2 4.0 4.80

10

20

30

40

50

60

y=2.1exp(x/ 1.5) y=0.3exp(x/ 1.2)

H2O exposure at 2 x 10- 5Pa

Ref at 4 x 10- 6Pa

Sample : TiO2, λ :355 nm

Deso

rpti

on a

mount

/ a

rb.u

nit

Laser intensity / mJ

Result 1 TiO2 rutile (Band:3.0eV)~Laser power dependency of desorption amount of H2O

430 nm(2.89eV) < Bandgap(3.0eV)

355 nm(3.49eV) > Bandgap

In the case of 355 nm(>bandgap), •Exponentially increasing

→Hot electron does not transit efficiently from conduction band to unoccupied orbital of OH.

Desorption via both thermal and PSD process occurs.

Bandgap

OH-TiO2

E4σ

EBG3.0eV

0 500 1000 1500 2000 2500 30000.0

0.2

0.4

0.6

0.8

1.0

MB fit

f(v)=Av2exp(- Bv2)

TiO2, λ :355nm,4.0mJ

Under H2O expo.

f(v)

/ a

rb.u

nit

Velocity / m s- 1

0 500 1000 1500 2000 2500 30000.0

0.2

0.4

0.6

0.8

1.0 Fe2O

3,λ :355nm,4.8mJ

Under H2O exposure

MMB fit

y=Av2exp(- B(v- v0)2)

f(v)

/ a

rb.u

nit

Velocity / m s- 1

0.0 0.8 1.6 2.4 3.2 4.0 4.80

5

10

15

20

25

PSD+Thermal

y=5.8x- 4.6

Y=0.8x- 0.7 PSD+Thermal

PSD

Ref at 4 x 10- 6Pa

Laser intensity / mJ

Deso

rpti

on a

mount

/ a

rb.u

nit

Heated at 150℃

H2O exposure at 2 x 10- 5Pa

Fe2O

3, λ :430 nm

0.0 0.8 1.6 2.4 3.20

5

10

15

20

25

PSD+Thermal

PSD+Thermal

PSD+Thermal

y=Aexp(x/ 0.8)

y=Aexp(x/ 0.9)

H2O exposure at 2 x 10- 5Pa

Heated at 150℃

Ref at 4 x 10- 6Pa

Fe2O

3, λ :355 nm

Des

orpt

ion

amou

nt /

arb

.uni

t

Laser intensity / mJ

Result 2 Fe2O3 (Band:2.2eV)~Laser power dependency of desorption amount of H2O

In the case of 430 nm, because of easy transition of hot electron, most H2O desorbed from surface via 1-photon process so that the desorption amount of H2O increases linearly with laser power.

4σ

EBG2.2eV

OH-Fe2O3

E

0 500 1000 1500 2000 2500 30000.0

0.2

0.4

0.6

0.8

1.0

Experiment

y=Av2exp(- Bv2)

Under H2O exposure

Tsample

: 24℃Laser power: 3.2 mJ

f(v)

/ a

rb.u

nit

Velocity / m s- 1

4σ

EBG2.2eV

OH-Fe2O3

E

0.0 0.8 1.6 2.4 3.20

10

20

30

40

50

60

70355 nm

Heated at 150℃Ref at 4 x 10- 6Pa

H2O exposure at 2 x 10- 5Pa

y=0.1exp(x/ 0.8)

y=22.4x- 11.2

y=9.6x- 4.4

Des

orpt

ion

amou

nt /

arb

.uni

t

Laser intensity / mJ

Result 3 Fe-Ti (5at%) oxide~Laser power dependence for the desorption amount of H2O

0 1 2 3 4 50

10

20

30

40

50

60

70

y=2.4x- 3.6

Heated at 150℃

430 nmH

2O exposure at 2 x 10- 5Pa

Ref at 4 x 10- 6Pay=15.2x- 1.1y=5.1x- 1.4

D

esor

ptio

n am

ount

/ a

rb.u

nit

Laser intensity / mJ0 500 1000 1500 2000 2500 3000

0.0

0.2

0.4

0.6

0.8

1.0λ =430 nm

Experiment

y=Av2exp(- Bv2)

f(v)

/ a

rb.u

nit

Velocity / m s- 1

0 500 1000 1500 2000 2500 30000.0

0.2

0.4

0.6

0.8

1.0λ =355 nm

f(

v) /

arb

.uni

t

Velocity / m s- 1

Experiment

y=Av2exp(- Bv2)

Sample temperature 24℃ 150℃

Sample H2O exposureNon H2O exposure

Non H2O exposure

Fe2O3

(EBG:2.2 eV)

430nm (>EBG)

PSD

(MMB, linear)

PSD

(MMB, linear)PSD

(MB+MMB,exp)

355nm (>EBG)

PSD+Thermal (MB?, exp.)

PSD+Thermal (MB?, exp.)

PSD+Thermal (MB?, exp.)

TiO2

(EBG:3.0 eV)

430nm (<EBG)

Thermal (MB, exp.)

Thermal (MB, exp.)

No or little desorption

355nm (>EBG)

PSD+Thermal (MB?, exp.)

PSD+Thermal (MB?, exp.)

PSD +Thermal (OH >> H2O)

Fe-Ti

(5at%)(EBG:??)

430nm (>EBG)

PSD (MMB?, linear)

PSD (MMB?, linear)

PSD (MMB?,linear)

355nm (>EBG)

PSD

(MMB?, linear)

PSD

(MMB?, linear)PSD

(MMB?,exp)

Desorption process (Fitting results, Laser intensity dependence)

Desorption mechanism (Now under consideration)

2.89eV3.49eV

430nm 355nm

In case of a transition of hot electron, the conduction band of Fe plays an important role. In case of 430nm, only one path can be considered. On the other hand, in case of 355nm, two paths in which hot electrons transit between components of bulk can be considered.

4σConduction band

EBG2.2eV

95% 5%OH-Fe2O3

E

TiO2

Valence band

λ:430 nmE

OH-Fe2O3TiO25% 95%

4σ

EBG3.0eV

λ:355 nm

Conclusions

The efficiency of the electron transfer from the substrate to the adsorbed species makes a large influence on the water desorption.

To achieve high efficiency of water desorption, an overlap between the conduction band of substrate and the unoccupied orbital plays an important role in the electron transfer from the substrate to absorption species.

For Fe-Ti(5at%) oxide, the desorption process caused by the effect of electron transfer between Fe and Ti oxides would exist.