semi合集-English.pdf - 第4112页

SEMI F60-0301 © SEMI 2001 8 5 Lambda 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 Angstroms 3 Lambda 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 1 2 3 …

100%1 / 7923
SEMI F60-0301 © SEMI 20017
Figure A1-2
Example of ESCA spectrum of the oxide passive
surface of stainless steel
A1-1.5 Illuminated Volume and Surface Sensitivity —
The incident x-ray beam penetrates well below the
surface. Photoelectrons and Auger electrons are
generated all through this illuminated volume. The
inelastic mean free path (IMFP) of an electron in a solid
is the mean distance through the solid that an electron
can travel before losing some of its energy (suffering an
inelastic interaction). The IMFP is a function of the
electron’s energy and the matrix through which it is
traveling. Electrons with an energy of approximately 1
keV have an IMFP of about 10 Å. That means that a 1
keV electron will travel, on average, 10 Å before
interacting with the matrix and giving up some of its
energy. In other words it will be scattered to lower
energy and will not be part of the photoelectron or
Auger electron line for its element. The IMFP is a
statistical parameter and an electron may travel several
IMFPs before having a collision. The “several IMFPs”
is called the escape depth of the electron, or the depth
of analysis, as shown in Figure A1-3. Although ESCA
and Auger are considered surface analytical tools, their
depth of analysis is on the order of 50 Å, or about one
to two times the depth of the oxide film found on a
typical passivated stainless steel surface.
Figure A1-3
The electron escape depth of low energy electrons in
some metals.
A1-1.6 Depth of Analysis Function — There has been
some discussion with respect to the “Depth of Analysis
Function”. There is one school that believes that it is 5
IMFPs and another that uses 3 IMFPs. These two
functions will be referred to as 5λ and 3λ respectively.
It will be seen that it makes little difference which
function is used even though the depth of the oxide
films is of the order of the functions. These functions
are seen in Figure A1-4, which shows the proportion of
the total signal generated from each depth below the
surface. The y-axis is in decimal and the x-axis is in
angstroms from the surface. The first point of the 5λ
curve is 0.16 at the 1
st
angstrom level. That means that
16% of the signal comes from the 1
st
angstrom, 14%
from the 2
nd
angstrom level and so on down to 1% from
the 25
th
angstrom level. Thus elements that do not
appear on the surface or in the first 24 angstroms in
from the surface, but do exist at the 25 Å level will be
detected.
t
rum Skip Auto by 1
010020030040050060070080090010001100
0
1
2
3
4
5
6
7
8
9
x 10
4
FCEPSSe1.spe
Binding Energy (eV)
c/s
-O KLL
-O KL
L
-Fe LMM
-Fe2p3
-Cr2p3
-O1s
-N1s
-C1s
-P2s
-P2p
-Fe3s
-Fe3p
-Cr3p
-Mo3d
-Ni LMM
-Ni3p
-Ni2p3
SEMI F60-0301 © SEMI 2001 8
5 Lambda
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
0.18
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
Angstroms
3 Lambda
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
0.18
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
Angstroms
Figure A1-4
These two graphs represent the “Depth of Analysis”
that will be seen when analyzing 1 keV electrons.
The top curve is the 5λ
λλ
λ curve and bottom the 3λ
λλ
λ.
A1-1.7 Modeling — Figure A1-5 is the composition
profile of a “perfect” oxide. It represents 25 Å of pure
Cr
2
O
3
on an atomically flat 316L stainless steel surface,
with no surface or interfacial contamination present.
Figures A1-6a and A1-6b are the theoretical depth
profiles derived by assuming a depth of analysis of 5λ
and 3λ respectively. The first factor to note is that Fe,
Ni and Mo appear most or all of the way through the
depth profile of the oxide due to the contribution of the
metal substrate to the detected signal, the “depth of
analysis” effect as introduced above. 1 keV electrons
can escape from as deep as 25 Å. This must be
appreciated when interpreting these depth profiles.
Theoretical
0
10
20
30
40
50
60
70
80
0 5 10 15 20 25 30 35 40 45
Angstroms
Atomic Concentration
Cr Fe Ni Mo O C
Figure A1-5
Composition versus depth of a pure Cr
2
O
3
layer on
stainless steel.
Convoluted 5 Lambda
0
10
20
30
40
50
60
70
80
90
100
0 5 10 15 20 25 30 35 40 45
Atomic Concentration
Cr Fe Ni Mo O C Cr:Fe
Figure A1-6a
Theoretical depth profile of Figure A1-5 assuming
5λ
λλ
λ depth of analysis.
Convoluted 3 Lambda
0
10
20
30
40
50
60
70
80
90
100
0 5 10 15 20 25 30 35 40 45
Angstroms
Atomic Concentration
0
100
200
300
400
500
600
700
Cr Fe Ni Mo O C Cr:Fe
Figure A1-6b
Theoretical depth profile of Figure A1-5 assuming
3λ
λλ
λ depth of analysis.
A1-1.8 The second factor to notice is the thickness of
the oxide. Historically, the FWHM of the oxygen
profile has been used as the measure of oxide thickness.
SEMI F60-0301 © SEMI 20019
It is apparent in Figures A1-5 and A1-6 that this
underestimates the oxide thickness by 15 to 20%.
A1-1.8.1 It should also be noted that the Cr:Fe ratio
measured from Figures A1-5 and A1-6 has its
maximum value at the initial surface, decreasing as the
depth profile progresses, whereas the actual Cr:Fe ratio
in Figure A1-5 is infinite down to 25 Å. Most actual
depth profiles exhibit a maximum of the Cr:Fe ratio at
some depth below the initial surface, referred to as the
depth of maximum enrichment. This is a consequence
of variation in the actual composition of the oxide,
possibly having a higher concentration of Fe near the
surface, a phenomenon commonly termed a “detached
iron oxide layer”, or of contamination on the surface,
generally hydrocarbons. These are discussed in the
next section.
A1-1.9 Effects of Detached Iron Oxide Layer and
Surface Contamination — The model composition
profile of Figure A1-7 shows 3 Å of pure Fe
2
O
3
over 22
Å of pure Cr
2
O
3
on 316L stainless steel. This is a
model of a detached iron oxide layer. Figure A1-8 is
the theoretical depth profile of this model derived
assuming a depth of analysis of 5 lambda. Note that the
initial Cr value in the profile is higher than the Fe value,
even though the surface is pure Fe
2
O
3
. This is a
consequence of the depth of analysis detecting the Cr
from levels beneath the surface. The Cr:Fe ratio for
this model has its maximum value at 3 Å below the
initial surface. The theoretical depth profile using a
depth of analysis of 3λ is very similar.
Theoretical
0
10
20
30
40
50
60
70
80
90
100
0 5 10 15 20 25 30 35 40 45
Angstroms
Atomic Concentration
Cr Fe Ni Mo O C
Figure A1-7
Composition versus depth of a Fe
2
O
3
layer over a
Cr
2
O
3
layer on stainless steel.
Convoluted 5 Lambda
0
10
20
30
40
50
60
70
80
90
100
0 5 10 15 20 25 30 35 40 45
Angstroms
Atomic Concentration
0
5
10
15
20
25
Cr Fe Ni Mo O C Cr:Fe
Figure A1-8
Theoretical depth profile of Figure A1-7 assuming
5λ
λλ
λ depth of analysis.
A1-1.10 The model composition profile of Figure A1-9
shows 3 Å of pure carbon over 22 Å of pure Cr
2
O
3
on
316L stainless steel, representing an idealized model of
the adsorbed hydrocarbon contamination generally
found on stainless steel surfaces exposed to the
atmosphere. Figure A1-10 is the theoretical depth
profile of this model assuming a depth of analysis of
5λ. In this case the O, Cr and Fe atomic concentration
values are reduced by the presence of the carbon until
the depth profiling proceeds past the carbon, but they
have the same relative values (ie: same Cr:Fe ratio)
versus depth as derived in the uncontaminated model.
The maximum of the Cr:Fe ratio is seen to be at the
initial surface. Although the actual oxide thickness is
less in this model, the FWHM measure of the oxide
thickness from the depth profile is the same due to the
presence of the carbon layer. Note that the Oxygen
concentration profile initially increases to a maximum,
then decreases. This is typical of Oxygen concentration
profiles seen on actual samples, which will generally
have some adsorbed hydrocarbon contamination on the
surface.