III.
PHYSICAL ELECTRONICS AND SURFACE PHYSICS
Academic and Research Staff
Prof. Robert E. Stickney
Dr. Mark J. Cardillo
Graduate Students
Ronald L. Levin
Mehmet K. Tulga
A.
JS
MOLECULAR BEAM STUDY OF THE APPARENT ACTIVATION
BARRIER
ASSOCIATED WITH ADSORPTION
OF HYDROGEN
ON COPPER
Joint Services Electronics Program (Contract DAAB07-71-C-0300)
Medhi Balooch, Mark J. Cardillo, Robert E. Stickney
Molecular beam techniques have been employed to study the adsorption and desorption of H2 on the (100),
(110), and stepped (310) crystal faces of copper.
Each crystal
was exposed simultaneously to a supersonic molecular beam of H 2 (energy variable from
1. 6-10.7 kcal/mole) and a highly dissociated beam of deuterium. The majority of H 2 molecules are scattered while a portion adsorb
dissociatively and react catalytically with
1.0
I
I
0.9 -
adsorbed deuterium atoms to form HD mol-
I
I
o
ecules.
Cu (310)
A Cu(I00)
0.8-
Cu (110)
\
by a rotatable mass spectrometer.
S\
0.7
N 6,
\cos
So.s -
0\
DIFFUSE
EMISSION
-
o 0.5-
o
0.4
0
meation study (Fig. III-1).
B
0.2
From the dependence of the HD signal
o
0
0.1 -\10
three crystal faces the angular distributions
ment with the results of our previous per-
\
\
20
30
40 50
8r (deg)
For all
of desorbed HD deviate significantly from
diffuse emission and are in excellent agree-
\ \-cu(I0o)
)
S0O\o
0
These HD molecules desorb, and
their angular distributions are measured
60
70
80
Fig. III-1.
on the energy and incident angle of the H
2
beam, it appears that there are substantial
energy barriers to adsorption.
These bar-
riers depend on crystallographic orientation
Angular distributions of desorbed HD
Cu(310).
Cu(110),
from Cu(100),
from Cu(l0),
Cu(00),
and Cu(310).
The dashed lines represent the angular distributions obtained from a pre-
and can be correlated by the parameter E =
E. cos 2 6.,
i where 6. is the angle of the inci-
vious permeation study. 1
mal, and Ei is the average incident energy;
QPR No. 114
1
1
1
dent beam with respect to the surface norJS
is
1.0 -()
0
Cu (110)
0.14
0
0.13
0.9
0.12
0.8
0.11
0.7
0.10
0.09
0.6
0.08
0.5
0.07
0.06
0.4
0.05
0.3
0.04
0.2
0.03
0.02
0.1
I
I
0.8
to
I
I
I
I
C(b)
Cu(310)
0.7
0.01
•
- 0.10
- 0.09
o
0
0.08
A
0.07
0.5
-O
S
0.06
0.4
0.05
LL 0.3
0.04
--
0.03
0.2
0.02
0.1
a
(C)
0.8
0.01
0.10
0
Cu (100)
0.09
0.7
0.08
0.6
-
-
0.07
0.5
0.06
0.4
0.05
0.03
act
0.2
- Cu
o (100)
1 111
0.1
0
0.04
A
]
0.3
0
I
[
1
2
o
I
I
I
I
I
3
4
5
6
7
-I
8
0.02
0.01
9
E.L=Ei cos2 i (kcal/mole)
Fig. III-2.
Relative HD fluxes (left ordinate) and the dissociative
adsorption probabilities for hydrogen, H (right ordiThey are plotted
2
against the correlating parameter El = Ei cos 0. for
nate),
for Cu(110),
(310) and (100).
a range of incident angles (0i) from Z5-50" with respect
to the surface normal.
QPR No. 114
(III.
PHYSICAL ELECTRONICS AND SURFACE PHYSICS)
this suggests that a one-dimensional potential to the point of dissociation is sufficient to
explain our results.
The estimated adsorption probabilities vs E l are S-shaped curves
that appear to level off at values considerably less than unity.
Fig. III-2.
JS
They are summarized in
Both the energy dependence and the shape of the HD angular distributions
are nearly identical for the stepped (310) and (100) surfaces (the stepped (310) may be
described as a sequence of (100) terraces, 3 atoms wide, separated by steps 1 atom
high), thereby suggesting that ledge sites are not the principal regions responsible for
adsorption of hydrogen on copper.
Switching the roles of deuterium and hydrogen in the
two incident beams yields a dissociative adsorption probability for D 2 which is higher
than that for H 2 by approximately the square root of the mass ratio.
We have compared our results with a simple model Z with a single energy barrier to
adsorption and obtain qualitative but not quantitative agreement for the angular distributions of desorbed HD (Fig. 111-3).
Consideration of the adsorption probabilities when
1.0
=
--
z 0.9
D0.8
V
ss-- 800 *K
calculated
Ts =I100 °K)
a Cu ( 00)
molecular beam results
0 Cu ( 10)
1,
0.7 -
0.
<
S0.4
\\
S\,\.
distributions calculated from
van Willigan for the (110) and (100) surfaces at two temperatures using the point
of inflections from Fig. III-2 to obtain
\
<o.3
\
S0.2
mole)
\
the activation energies E
A A
. o
i0\V\mental
.
10
20
30
40
50
from 800-1100"K,
60
a
.a The experi-
molecular beam results, which
are surface temperature-independent
,\E\\\5
0.1
0
SAng ul ar
3(kcal/mole)
Ea 3(kcal/mole)
<0.6 -\
Fig. 111- 3.
70
80
are also shown.
90
Or (deg)
the effect of the nonzero energy spread of the incident beam is removed, together with
the results for isotopic switching, suggests that the simplest model must include (i) a
distribution of barrier heights associated with the oscillatory nature of the potentials
near the surface and (ii) a dynamic factor such as a characteristic collision time to
account for the change in dissociation probability with incident particle mass.
References
1.
M. Balooch and R. E. Stickney (submitted to Surface Sci.).
2.
J. E. Lennard-Jones, Trans. Faraday Soc. 28, 333 (1932); W. van Willigan, Phys.
Letters 28A, 80 (1968).
QPR No. 114
JS