SOFT X-RAY EMISSION SPECTRA OF Zr IN Zr-H ALLOYS.

Article abstract of DOI:10.1016/0038-1098(79)90981-5

Zr-L//3 emission bands of ZrH//x alloys (x equals 0, 0. 54, 0. 95 and 1. 90) have been measured. With increasing x, a low-energy subband emerges and develops at 7 ev below the Fermi level at the expense of the intensity of the main band just below the Fermi edge. These changes in intensity cancel with each other, suggesting that a remarkable redistribution of energy of 4d electrons at Zr sites takes place by the influence of protons on interstitial sites.

Full text of DOI:10.1016/0038-1098(79)90981-5

Solid State Communications, Vol. 30, 173—176.
PergamonPress Ltd. 1979.Printed in Great Britain.
SOFT X-RAYEMISSION SPECTRA OF Zr IN Zr—H ALLOYS
K. Tanaka, N. Hamasaka and M. Yasuda
Departmentof Metallurgical Engineering, Nagoya Institute of Technology, Showa-ku, Nagoya 466, Japan
and
Y. Fukal
Department of Physics, Chuo University, Kasuga, Bunkyo-ku, Tokyo 112, Japan
(Received 25 December 1978 by Y. Toyozawa)
Zr—L3 emission bands of ZrH~alioys(x
=
0,0.54,0.95 and 1.90) have
beenmeasured. With increasing x, a low-energy subband emerges and
develops at 7eV belowthe Fermi level at the expense of the intensity of
the main bandjust below the Fermi edge. These changesin intensity
cancel witheach other, suggesting strongly thata remarkable redistribution
of energy of 4d electrons at Zr sites takes place by the influence of
protons on interstitial sites.
THE USEFULNESS of soft X-ray spectroscopy (SXS)
(f.c.t.) for 1.66 ~ x ~ 2.0;6-hydride (f.c.c.) for
for studying the electronic structure of metal—hydrogen 1.59 ~ x < 1.66; and a-phase (h.c.p. Zr) + & hydride
alloys has beenrecently demonstrated in VH (D)~[1]
and NbHX [2]. The measurement of V—L3 emission
spectra of WI (D)~hasgiven evidence of a gradual
for lower x, at room temperature. The presence of ~
hydride (ZrH)has also been suggested as a metastable
product in the a + 6 region. The ehydride is considered
change of host d-band with the addition of H (D) and a
formation of H(D).inducedstates at the bottom of the
to have a fluorite structure with one cube axis con-
tracted to adegree which depends on x. Based on this
d..band, in general agreementwith results of theoretical
energy-bandcalculations by Switendick [3]. Similar
results have also been obtained in Nb—L3 emissions of
diagram, our samples are expected to be composed of
following phases: ZrH1.~(e)Z; rH0.~(a+ 6);
ZrHO.M(a + 6); and Zr(a). By X-ray diffraction
NbH~I.n contrast to these Va elements,where the
hydrogen content is limited tox less than 0.8 under 1
analysis these phases were clearly identified in the
samples,except for ZrH 0~a~nd ZrH0,~where dif-
atm of H2 at room temperature, IVa elements (Ti, Zr
fraction linesof only 6-phase were able to be detected.
Since the measurement hasbeen carried out not on
powdered samplesbut on bulk crystals used for the SXS
and Hf) are known to accommodate hydrogen up to
x
—
2.0, and constitute anothermetallic group suitable
for investigating the electronic structure of metal—
hydrogensystems in awider range of the hydrogen
concentration. The present paper reports on measure-
mentsof Zr—L3 emissions (5s4d-band - ~ 2p312 level
transition) in Zr—H alloys,
measurements, the diffraction lines of coexisting a-phase
might have unexpectedly been missed in the latter
samples.
The SXS measurements have been carried out in the
same way as described in our previous report [1]. A
A reactor.grade Zr sponge (0.080wt.%0 and
curved ADP(1 10) crystal (2d = 1.064nm; R
was used as an analyzer. The instrumental resolution at
the Zr—L3 band (X 0.559 nm) is rather poor and is
estimated at 1.0 eV from a measured P—Ka1 (X
=
200mm)
0.078 wt.% Fe; purity >99.8 wt.%) was loaded with
hydrogenby holding at 600°Cfor 6 hr under purified
H2-gas flow, and slowly cooled. Next, the sample was
=
=
successively dehydrogenated by annealing at 340 —~
600°Cfor several minutes in vacuum,followed by slow
cooling with valves of the vacuum furnace shut down,
It was finally completely outgassedat 800°Cin vacuum
of 1 x iO ~Pa. This procedure enabled us to prepare
0.616 nm) line width. The total resolution including
the natural width of the Zr 2p3/2 level ( - ~ 1.5 eV) [5] is
estimated to be no less than 2.5 eV.
Figure 1 representsL~215(L3 band) and L~3
(3p3/2 - + 2s) spectra in pure Zr before hydrogen charging
three ZrH ~alloys withx = 1.90 ± 0.10,0.95 ± 0.12 and as a function of the photon energy. The L3 band is
0.54 ± 0.10, in addition to a pure Zr sample.The con-
centrationwas determined by gas-extractionanalysis.
The equilibrium phase diagram of the Zr—H system
given by Beck [4] indicates the presence of c-hydride
located on the high.energy tail (dashed curve) of the
1433 characteristic line, which is utilized as a standard of
both energy and intensity of the L3 band. The FWHM
value of this band (4.4 ± 0.1 eV) is about I eV larger
173

174
SOFT X-RAY EMISSION SPECTRA OF Zr IN Zr—H ALLOYS
Vol. 30, No.3
x 1 0 3
6
L$3
Pure Zr
( f l 4 .
0
P4)
L~2,I5
2
0
o 2
S
5
- - - - - -
0
I
•
I
•
I
2190
2200
2210
ENERGY (eV)
2220
Fig. I. L~2,15(L3 band;ad band - ~ 2p3,2) andL$3 (3P3,2 - ‘ 2s)spectra in pure Zr.
Zr Lband
(I)
. 0
I—
z
LIU-
ZrHo.,~
-
Pure Zr
— 2— 2p 25
2205
2210
2215
2220
ENERGY (eV)
Fig. 2. Zr—L3 band spectra inpure Zr and ZrH~T. he intensity of each spectrum is normalized to itsLj33 peak height.

Vol. 30, No.3
SOFTX-RAY EMISSION SPECTRA OF Zr IN Zr—H ALLOYS
Table 1. Incrementsinarea of emission bands of
175
Zr- L band
E
ZrH~correspondingto A, B and Cshown in Fig.
Zrtssetto 100
Samples
A
B
C
Total
~—
ZrH1,~
4.rHT0h~1e9,w5hole a69r.e48a of th
subtracted. T Ienicnrteemnesinttysoinf earaecah spectru in Fig. 2 is
eseape ofhythogen from ~ples during measurements
normalized to itsL~3peakheight. It must be mentioned
—
e e1m60is40sion ba5n8d ofiur41e88
ZrHo.~ 4.1
4.8 1.0 0.3
in Fig. 3 after the backgro—und contributions have been
_ _ _ _ _ _
_ _ _ _ _
here that a gradual alteration of spectral shapes due to
ZrH0~4
as observed inNbHX [2] hasnever occurred in ZrH~,
and their SXS spectra havebeen foundto be quite
reproducible. We notethe following features of theL3
band of Zr—H alloys.
7eV below the F e ~ edgeE ~(chosen at the half-height
(1) Alow-energy subband emergesand develops at
Pure Zr
2210
position) with the addition of hydrogen.
(2) The main band is depressed in height, and the
22I5
2220
2225 Fermi edge is shifted toward higher energy as much as
ENERGY (cv)
0.5eV in ZrH1 ~o.It has beenconfirmed that these
Fig. 3. Zr—L3 emission bands after the L ~ components spectral changes due to hydrogenation are entirely
have beensubtracted.
restored when the samplesare completely outgassed.
It may be helpful to make a comparison of the
Zr—L3 band of Z rH 1~w~ith that of pure Zr, as shown
than a reportedvalue [5],reflecting the differenceof
the instrumental resolution. Spectra of theL3 band
in Fig. 4. Three features are evidently seen in the figure:
observed inZrH~are shown in Fig. 2 and arereproduced an increment ofthe intensityat the low-energy subband
Zr— L band
Pure Zr
0
>-
I -
z
w
I -
z
—15
—10
—5
O(—Ef)
5
ENERGY (eV)
Fig. 4. Comparison of the Zr—L3 band of Z rH 1~w~iththat of pure Zr.

176
SOFT X-RAY EMISSION SPECTRA OF Zr IN Zr—H ALLOYS
Vol. 30,No.3
(A), a decrement at the main peak (B), and a small
increment at the high.energy edge (C). Setting the
whole area of the emission band of pure Zr to 100,
Zr—H alloys are not yet available, a rough estimationof
the number of additional electrons in the 4d-band may
be made from ourpresent data. On the assumption that
changes in area corresponding to the regionsA, B and C 4 valence electronsper Zr atom are contributingto the
are tabulated in Table 1, including results on other integrated intensity of the L3 band of pure Zr, the num-
samples. We findthat in each alloy Aand B cancel with ber of the extraelectrons in ZrH 1~g~iv, en by the area C
each other within an experimental uncertainty, while C in Fig. ~, is estimated as 4 x 5/100 0.2. However, this
=
tends to increase slightly with increasing hydrogen
would probably be an underestimate of its true value
content,
because of an appreciable self-absorption correction
The low-energy subband found in ZrH~corresponds yet to be applied at the emission edge. To make amore
evidently to those observedin V—L3 and Nb—L3
quantitative analysis of the spectra, further corrections
emission spectra in WI (D)~[1] and NbHX [2], respec- of the transitionprobability, Auger broadening and spec-
tively. In addition,our recent study on the Ti—L3 band tral resolution must be applied. Nature ofsp band in
in TiHX [6] shows that it has spectral features which
closely resemble those found in ZrH~T. hese fmdings
ZrH~may be understood by measuring Zr—M5
emissions (5p band - ~3d5, 2). The measurement is now
strongly suggest that the subband is due to a formation
of bonding states between H—ls and metal—d states,
under way.
Acknowledgement
—
This work haspartly been
whichcauses aheavy distortion of the host d-band struc- supported by the Grant-in-Aid for Scientific Research
ture of the metal. The fact that the decrement of the
from the Ministry of Education.
main band of the Zr—L3 emission in ZrH~is nearly
complemented by the increment of its low-energy sub-
band implies that the number of 4d electrons at a Zr site
remains unchanged in spite of a significant redistribution
of their energy through an interaction with surrounding
REFERENCES
1. Y. Fukai, S. Kazama,K. Tanaka & M. Matsumoto,
interstitialprotons. The slight increments of the area C
on hydrogen loading could be associated with additional
electrons supplied by hydrogen atoms filling up 4d-band
holes. According to an energy-band calculation on T1H2 ~:SolidState Commun. 19, 507 (1976).
~
(normal fluorite structure) [3], only0.85 out of 2 dcc-
trons donated by hydrogen atoms are accommodated to
fill up 3d-band holes of Ti, while the remaining electrons 4.
are supplied to an sp band pulled down belowthe Fermi
G. Alefeld& J. VOIkl), p. 101. Springer-Verlig,
Berlin (1978).
R.L. Beck, 7)un& ASM55,542(1962).
5. V.V. Nemoshk5leilko, A.P. Shpak&V.P.
level. This sp band is believed to be responsible fora
distortion of the Ti—K~5emission observed in TiH1 .83
6.
~
(to be published).
[7], but does not practically influence the L3 emission
band. Though detailed band-structure calculations on
7. S.A. Nemnonov &K.M. Kolobova, Fiz. MetaL
Metalloved. 22, 680 (1966).