Surface Science 566–568 (2004) 751–754

www.elsevier.com/locate/susc

Hydrogenation of Pd capped Mg thin filmsat room temperature

Kazuki Yoshimura *, Yasusei Yamada, Masaharu Okada

National Institute of Advanced Industrial Science and Technology (AIST), 2266-98 Anagahora, Shimoshidami, Moriyama-ku,

Nagoya 463-8560, Japan

Available online 19 June 2004

Abstract

Pd capped pure Mg thin films were prepared by DC magnetron sputtering and their hydrogenation at room tem-

perature has been investigated. After exposure to 4% hydrogen gas diluted by argon, the Pd/Mg thin films on Si

substrates at 295 K show drastic optical changes in reflectance within 5 s. XPS analysis reveals that there exists an

intermixing layer of Pd and Mg near the surface, and MgH2 is formed very quickly in this intermixing layer. It is very

interesting that such a fast hydrogenation occurs at room temperature.

� 2004 Elsevier B.V. All rights reserved.

Keywords: Alkaline earth metals; Palladium; Hydrogen atom; Sputter deposition; X-ray photoelectron spectroscopy

1. Introduction

Much attention has been paid to the magne-

sium–hydrogen system from both fundamental

and practical viewpoints [1]. Especially for the

purpose of hydrogen storage, magnesium has been

extensively studied because magnesium can absorb7.65 wt% of atomic hydrogen. However, its oper-

ation temperature is over 600 K, and it strongly

restricts applications of magnesium. On the other

hand, dramatic optical change by hydrogenation

was discovered for Pd capped rare-earth metal

hydrides in 1996 [2]. After that the same phe-

nomena have been observed for some Pd capped

* Corresponding author. Tel.: +81-52-736-7306; fax: +81-52-

736-7315.

E-mail address: k.yoshimura@aist.go.jp (K. Yoshimura).

0039-6028/$ - see front matter � 2004 Elsevier B.V. All rights reserv

doi:10.1016/j.susc.2004.06.115

magnesium alloy thin films such as Gd–Mg [3],

Sm–Mg [4], Y–Mg [5] and Ni–Mg [6,7]. In these

materials, hydrogenation occurs at room temper-

ature. Moreover, Yamamoto et al. found that Pd

capped pure Mg thin films can be hydrogenated

and made to be transparent at a temperature of

373 K [8]. These findings imply that Pd/Mg thinfilms have the possibility to reduce the hydroge-

nation temperature to about room temperature. In

this paper, the Pd capped pure Mg thin films were

prepared by DC magnetron sputtering with vari-

ous conditions and their hydrogenation properties

at 295 K were characterized by monitoring optical

reflectance of the sample surface. We found that

Pd/Mg thin films prepared under specific condi-tions show drastic reflectance changes after expo-

sure to a H2 atmosphere due to hydrogenation

which occurs very quickly even at room tempera-

ture.

ed.

752 K. Yoshimura et al. / Surface Science 566–568 (2004) 751–754

2. Experimental

Pd capped Mg thin films were prepared by DC

magnetron sputtering with metal Mg and Pd tar-

gets. Using a loadlock system, the base pressure ofthe deposition chamber was kept below 2 · 10�5

Pa. After evacuation, Mg thin films were deposited

on a Si substrate, followed by the Pd deposition.

This Pd cap layer is necessary for enhancement of

the hydrogen uptake kinetics and protection of the

Mg layer [9]. The substrate was not heated during

deposition for all samples. Sputtering conditions

are summarized in Table 1.After removal from the chamber, deposited

films were exposed to 0.1 MPa 4% H2 gas diluted

by Ar. Their reflectance change was monitored by

using a diode laser (k ¼ 670 nm) and a Si photo-

diode with acquisition every one second. A silver

thin film deposited on a Si substrate was used for

the calibration of reflectance. As-deposited and

hydrogenated films were characterized by opticalmeasurements using a UV–Vis–NIR optical pho-

tometer (JASCO V570) and by X-ray photoemis-

sion spectroscopy (XPS) measurements using a

VG Sigma probe system.

90

Pd ( 4 nm) /Mg (60 nm)

3. Results and discussion

We prepared Pd capped Mg thin films under

various conditions by changing total pressure, in-

duced power and film thickness. After observa-

tions of optical reflectance change of the prepared

samples by exposure to a 4% H2 atmosphere, we

Table 1

Deposition conditions

Method DC magnetron sputtering

Targets Mg (99.9%) 50 £ · 3 mm

Pd (99.99%) 50 £· 1 mm

Base pressure 2· 10�5 Pa

Ar flow rate 200 sccm

Total pressure 0.8 Pa

Discharge power Mg: 30 W

Pd: 14 W

Deposition rate Mg: 0.8 nm/s

Pd: 0.5 nm/s

Target-substrate distance 100 mm

Sample size 10· 20 · 0.8 mm

found that the samples prepared under specific

conditions (total pressure around 0.8 Pa, and dis-

charge power for Mg and Pd sputtering around 30

and 14 W, respectively) show drastic reflectance

changes after H2 exposure.

Fig. 1 shows the optical reflectance change at670 nm with respect to exposure time for samples

with different thickness. All as-deposited films

have a silver-like metallic surface. After exposure

to the H2 atmosphere, the surface of these samples

became tinged with brown and the reflectance de-

creased. These changes were very fast and were

completed within 5 s. As shown in Fig. 1, with

increase of the thickness of Mg layer, the per-centage change of the reflectance increases. How-

ever, this dependence is not so strong. A 410%

change of Mg thickness caused only a 50% change

of reflectance. The sample with thinner Pd layer

showed a slightly slower change.

Optical reflectance spectra of a Pd (10 nm)/Mg

(250 nm) thin film in the as-deposited state and

after 20 s exposure to H2 is shown in Fig. 2. Alarge reflectance change is observed in the wave-

length range from 250 to 2000 nm. The amount of

change is maximum at 500 nm and is about 40%.

With increase of wavelength, the relative change

decreases gradually. The reflectance minimum in

the short wavelength region around 500 nm causes

the dark brown colour in the exposed sample.

These reflectance changes are caused byhydrogenation of the Mg layer in the film, because

80

70

60

50

40

Ref

lect

ance

at 6

70 n

m (%

)

86420Time (s)

Pd (10 nm) /Mg (60 nm)Pd (10 nm) /Mg (120 nm)Pd (10 nm) /Mg (250 nm)

Fig. 1. Time evolution of optical reflectance at 670 nm after 4%

H2 exposure for Pd capped Mg thin films with different thick-

ness.

100

80

60

40

20

0

Ref

lect

ance

(%)

200015001000500Wavelength (nm)

as- depositedexposure to H2

Fig. 2. Optical reflectance spectra of the Pd (10 nm)/Mg (250

nm) thin film. Solid line is for the as-deposited state. Broken

line is for the hydrogenated state.

Peak

Inte

nsity

(Arb

. Uni

ts)

8006004002000Etching Time (s)

Mg SiPd

O

MgH2

Fig. 3. Depth profile of the hydrogenated Pd (10 nm)/Mg (250

nm) thin film (d: Mg, �: Pd, N: oxygen, �: Si, respectively).

Thick solid lines with . designates the distribution of MgH2.

Inte

nsity

(Arb

. Uni

ts)

338 336 334 332Binding Energy (eV)

120 s

240 s

360 s

Inte

nsity

(Arb

. Uni

ts)

1308 1306 1304 1302 1300Binding Energy (eV)

240 s

120 s

360 s

Pd 3d 5/2 Mg 1s

Fig. 4. XPS spectra of the Pd 3d5=2 state and Mg1s state for the

hydrogenated Pd (10 nm)/Mg (250 nm) thin film after different

etching time.

K. Yoshimura et al. / Surface Science 566–568 (2004) 751–754 753

a 10 nm thick pure Pd thin film deposited on Si

shows only a 0.4% reflectance change after expo-

sure to H2. Upon hydrogenation, magnesium hy-

dride (MgH2) is believed to be formed in the film.

Because MgH2 is theoretically predicted to be a

large band gap insulator [10] and transparent, the

optical reflection conditions of Pd and Mg layersare changed, which may be the cause of the de-

crease of reflectance. Hence, such a large reflec-

tance change means that a considerable amount of

MgH2 is formed in the film within 5 s. It is a

surprising result that hydrogenation of magnesium

occurs so fast at room temperature, because the

diffusion coefficient of H in Mg at 305 K is re-

ported to be 1.1 · 10�20 m2/s [11], which means thatit takes 30 h for a hydrogen atom to diffuse 1 nm.

To investigate what happens in these films after

hydrogenation, we measured the XPS spectra of

these films before and after exposure to hydrogen.

XPS spectra of the Mg1s state, the Pd 3d5=2 state,

the oxygen 1s state and the Si 2p state were taken,

as well as their depth profile, which was measured

with the aid of argon sputter etching. The excita-tion source was monochromatized AlKa radiation

(1486.48 eV). Fig. 3 is the depth profile of each

element for the hydrogenated Pd (10 nm)/Mg (250

nm) thin film. The vertical axis is proportional to

the integrated counts of each peak. The depth

profile of the as-deposited sample is almost the

same as the hydrogenated sample. This depth

profile shows that the interface between the Pdlayer and the Mg layer is not sharp compared with

the interface between the Si substrate and the Mg

layer. Magnesium distribution spreads to the top

surface. Also some amount of oxygen exists near

the surface and its signal intensity decreases with

increase of depth.

Fig. 4 shows the XPS spectra of the Pd 3d5=2

and Mg1s states for the hydrogenated Pd (10 nm)/Mg (250 nm) thin film after the designated etching

time. The peak position of the Pd 3d peak after 120

s etch is different from the bulk value (335.1 eV

[12], which is indicated by a vertical line) and the

peak energy is shifted 0.9 eV to the higher binding

energy side. The peak position of Mg 1s after 120 s

etch is different from the bulk value (1303.5 eV

[12], which is indicated by a vertical line), either.The peak position is shifted 0.2 eV to lower

754 K. Yoshimura et al. / Surface Science 566–568 (2004) 751–754

binding energy in this case. These results indicate

that Pd and Mg form an intermixed layer near the

surface.

The Mg1s spectra after 240 s etch splits into two

peaks. The higher peak position is 1305.4 eV, which

coincides with the reported peak position of theMg1s state of MgH2 [12]. These results clearly

shows that MgH2 is formed in the hydrogenated

film. The relative peak height of the Mg1s state of

MgH2 to that of Mg, which is determined by the

deconvolution of the Mg1s spectrum, is plotted in

Fig. 3 with respect to etching time. This plot indi-

cates the magnesium hydride is centred around a

depth which corresponds to about etchingtime¼ 200 s.

Because the sputtering rate of Mg is larger than

that of Pd, making depth not proportional to

etching time, it is difficult to determine the thick-

ness of the hydride layer. However, these results

strongly imply that a hydride with considerable

thickness exists in the intermixing layer of Pd and

Mg, and in this intermixing layer, hydrogenation isfast and MgH2 is formed very quickly. There is

only a few reports on bulk intermetallic compound

of Mg and Pd; Mg6Pd [13] and Mg89Pd11 [14]. In

both reports, the formation of MgH2 and Mg5Pd2

after hydriding of these alloys is suggested. On the

other hand, the same kind of experiments have

been done for Pd/Mg thin films by some authors

[15,16]. However, such a fast hydrogenation atroom temperature has not been observed. They

prepared the thin films by evaporation under high

vacuum conditions instead of sputtering. Thus we

think that the formation of the intermixing layer is

promoted by sputtering, because sputtering is a

highly non-equilibrium process. Also there is a

possibility that oxygen plays an important role for

hydrogen diffusion in the intermixing layer, whichwas pointed out by Hjort et al. [17].

4. Conclusions

Pd capped Mg thin films were prepared by DC

magnetron sputtering and their hydrogenation

properties at room temperature have been inves-

tigated. After exposure to 4% hydrogen in Ar, Pd

(10 nm)/Mg (250 nm) a thin film on a Si substrate

shows remarkable optical reflectance change

within 5 s. XPS analysis reveals that there exists an

intermixing layer of Pd and Mg near the surfaceand in this mixing layer hydrogen may diffuse fast

and MgH2 is formed very quickly even at room

temperature. This formation of MgH2 causes a

large reflectance change.

References

[1] G. Alefeld, J. Volkl (Eds.), Hydrogen in Metals, Topics in

Applied Physics, Springer, Berlin, 1978.

[2] J.N. Huiberts, R. Griessen, J.H. Rector, R.J. Wijngaarden,

J.P. Dekker, D.G. de Groot, N.J. Koeman, Nature 380

(1996) 231.

[3] P. Van der Sluis, M. Ouwekerk, P.A. Duine, Appl. Phys.

Lett. 70 (1977) 3356.

[4] M. Ouwekerk, Solid State Ionics 113–115 (1977) 431.

[5] D.G. Nagengast, A.T.M. van Gogh, E.S. Kooij, B. Dam,

R. Griessen, Appl. Phys. Lett. 75 (1999) 2050.

[6] T.J. Richardson, J.L. Slack, R.D. Armitage, R. Kostecki,

B. Farangis, M.D. Rubin, Appl. Phys. Lett. 78 (2001)

3047.

[7] K. Yoshimura, Y. Yamada, M. Okada, Appl. Phys. Lett.

81 (2002) 4709.

[8] K. Yamamoto, K. Higuchi, H. Kajioka, H. Sumida, S.

Orimo, H. Fujii, J. Alloys Compd. 330–332 (2002)

352.

[9] S.J. van der Molen, J.W.J. Kerssemakers, J.H. Rector, N.J.

Koeman, B. Dam, R. Griessen, J. Appl. Phys. 86 (1999)

6107.

[10] R. Yu, P.K. Lam, Phys. Rev. B 37 (1988) 8730.

[11] P. Spatz, H.A. Aebischer, A. Krozer, L. Schlapbach, Z.

Phys. Chem. 181 (1993) 393.

[12] A. Fischer, H. K€ostler, L. Schlapbach, J. Less-Common

Metals 172–174 (1991) 808.

[13] Y. Kume, A. Weiss, J. Less-Common Metals 136 (1987)

51.

[14] T. Yamada, J. Yin, K. Tanaka, Mater. Trans. 136 (2001)

2415.

[15] J. Ryden, B. Hjorvarsson, T. Ericsson, E. Karlsson, A.

Krozer, B. Kasemo, J. Less-Common Metals 152 (1989)

295.

[16] A. Krozer, B. Kasemo, J. Less-Common Metals 160 (1990)

323.

[17] P. Hjort, A. Krozer, B. Kasemo, J. Alloys Compd. 237

(1996) 74.