Data
analysis sheet for step height
measurements from one step
height test structure.
a)
b)
Figure SH.1.1.For a CMOS step height test
structure: a) a design rendition and
b) a cross-section.
To obtain the
following measurements, consult SEMI
standard test method MS2 entitled
"Test Method for Step Height
Measurements of Thin Films."
date data taken (optional) =
/
/
Table 1 - Preliminary
INPUTS
Data Set Prelims
Description
1
proc
=
which process?
2
which
=
For CMOS SRM chips,
which of the six step
height measurements? For
MUMPs chips, which quad?
3
which2=
For CMOS chips, which
iteration of the test
structure where "first"
corresponds to the
topmost test structure
in the column?
4
orient =
orientation of the test
structure on the test
chip
5
×
magnification
6
align =
alignment ensured?
7
level =
data leveled?
8
μm
certified value of
physical step height
used for calibration
9
μm
certified one sigma
uncertainty of the
certified physical step
height used for
calibration
10
μm
maximum uncalibrated
range of the six
calibration measurements
taken before the data
session at the same
location on the physical
step height or after the
data session at the same
location on the physical
step height (whichever
is larger)
11
μm
the uncalibrated average
of the six calibration
measurements from which
zrepeat(shs)
was found
12
μm
uncalibrated drift in
the calibration data
(i.e., the uncalibrated
positive difference
between the average of
the six calibration
measurements taken
before the data session
at the same location on
the physical step height
and the average of the
six calibration
measurements taken after
the data session at the
same location on the
physical step height)
13
the
z-calibration
factor = the certified
value of the physical
step height divided by
the average of the
twelve calibration
measurements taken at
the same location on the
physical step height
14
%
if
applicable, over the
instrument's total scan
range, the maximum
percent deviation from
linearity, as quoted by
the instrument
manufacturer (typically
less than 3%)
15
σ
μm
the uncalibrated surface
roughness of
platNX measured
as the smallest of all
the values obtained for
σplatNXt.
(However, if the
surfaces of platNX,
platNY, and
platNr all have
identical compositions,
then it is measured as
the smallest of all the
values obtained for
σplatNXt,
σplatNYt,
and
σplatNrDt
in which case
σroughNX=σroughNY.)
16
σ
μm
the uncalibrated surface
roughness of
platNY measured
as the smallest of all
the values obtained for
σplatNYt.
(However, if the
surfaces of platNX,
platNY, and
platNr all have
identical compositions,
then it is measured as
the smallest of all the
values obtained for
σplatNXt,
σplatNYt,
and
σplatNrDt
in which case
σroughNX=σroughNY.)
Nomenclature:
"N" refers to the
test structure number ("1,"
"2," "3,"
etc.),
"X" and "Y"
refer to the platform letter
("A," "B," "C,"
etc.),
"r" indicates a
reference platform,
"D" directionally
indicates which reference
platform, and
"t" indicates which
data trace ("a," "b,"
or "c").
Note 4:
stepNXY = AVE(stepNXYa,
stepNXYb, stepNXYc)
Note 5:
uLstep =SQRT[(σplatNXave-
calz
σroughNX)2
+ (σplatNYave-
calz
σroughNY)2]
Note 6:
uWstep =
σstepNXY
=
STDEV(stepNXYa,
stepNXYb, stepNXYc)
Note 7:
ucert = |σcertstepNXY
/ cert|
Note 8:
urepeat(shs) = |zrepeat(shs)stepNXY
/ [2(1.732) z6]|
Note 9:
udrift = |(zdriftcalz) stepNXY
/ [2(1.732) cert]|
Note 10:
ulinear = |zpercstepNXY
/ (1.732)|
Note 11:
ucSH =
SQRT(uLstep2+uWstep2+ucert2+urepeat(shs)2+udrift2+ulinear2)
(Each of the standard
uncertainty components is
obtained using a Type B
analysis, .
except for uWstep,
which uses a Type A analysis.)
Report the results as follows: Since it can be assumed that the
estimated values of the
uncertainty
components are
approximately uniformly
or Gaussianly distributed with
approximate combined standard
uncertainty
ucSH, the step
height is believed to lie in the
interval stepNXY
±
ucSH (expansion
factor k=1)
representing a level of
confidence of approximately 68 %.
Modify the input data,
given the information
supplied in any flagged
statement below, if
applicable, then
recalculate:
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
σroughNX
and
σroughNY
should be greater than
0.0
μm and less than or
equal to the smallest
measured value for
σplatNXt
and
σplatNYt,
respectively.
13.
14.
15.
16.
σplatNXt
and
σplatNYt
should be between 0.00
μm and 0.025 μm,
inclusive.
