semi合集-English.pdf - 第5354页
SEMI M53-1103 © SEMI 2003 5 RC depends on the light source used, is in gene ral non- linear and m ay contain regions wi th dips t hat make the response/diameter relationship multi-valued. (See Section 3.3). 5.3.18 scanni…
SEMI M53-1103 © SEMI 2003 4
signal due to haze together comprise the signal due to
light scattering from a wafer surface.
5.3.10 latex sphere equivalent (LSE) — the diameter of
a monodisperse polystyrene latex sphere that, under
identical test conditions, produces the same detected
scattering intensity as the LLS under investigation.
5.3.10.1 Discussion — If the LLS is assumed to be due
to a particle (or pit), the LSE size (diameter) of the
particle (or pit) is given in units of length followed by
LSE; for example, 0.12 µm, LSE. This unit varies in
different ways for different materials from instrument
to instrument because of differences in the optical
systems and signal processing procedures of different
instruments. Therefore a particular LLS generally does
not have the same LSE size when measured on different
model instruments or on different channels of the same
instrument. If elements of the optical system, such as
incidence angle, collection solid angle, or polarization,
of an SSIS can be varied, the LSE size of a particular
LLS will not necessarily be the same for each
configuration of the optical system.
5.3.11 localized light scatterer (LLS) — an isolated
feature, such as a particle or a pit, on or in a wafer
surface, resulting in increased light scattering intensity
relative to that of the surrounding wafer surface;
historically called light point defect because under high
intensity optical illumination features of sufficient size
appear as an isolated point of light.
5.3.11.1 Discussion — Localized light scatterers are
observed by automated inspection techniques as laser-
light scattering events. Automated inspection
techniques are quantitative in the sense that scatterers
with different scattering intensities can be segregated.
However, the amplitude of the scattered light intensity,
or “laser-light scattering event”, as measured by any
combination of incident beam direction and collection
optics, does not by itself convey topographical
information about the LLS; particles and pits cannot be
distinguished solely on the basis of single-channel
amplitude data. Also, the observer cannot deduce the
size, shape, or composition of the LLS from single-
channel amplitude alone. The presence of LLSs does
not necessarily decrease the utility of the wafer.
5.3.12 missing count — the case in which an LLS fails
to produce a laser-light scattering event; also called
false negative.
5.3.13 multipoint calibration — a procedure for
calibrating the size response of an SSIS using a set of
accurately sized polystyrene latex spheres, deposited on
a wafer surface of the type to be inspected by the SSIS.
5.3.13.1 Discussion — The purpose of a calibration is
to relate the amount of light captured by an SSIS to the
physical size of the light scatterer. The amount of light
scattered from a localized light scatterer (LLS) that is
captured by the SSIS is a function of both the scattering
characteristics of the LLS (including the directional
dependence of the scattering) and the geometry of the
collection optics of the SSIS. For a given wavelength
of incident radiation, regular objects of certain sizes
exhibit non-linearities, usually in the form of a dip, so
that the curve of scattering amplitude as a function of
physical size of the scatterer may not be monotonic.
Thus, similar objects with modest variations in physical
size can scatter the same amount of light. Different
materials may exhibit non-linearities at different sizes.
Because these effects tend to be less pronounced for
irregular particles such as may be found on polished
wafer surfaces, instrument calibration is usually carried
out by avoiding the sizes of spheres that correspond to a
dip. Sometimes, however, it is desired to fully
characterize the response of the SSIS to a particular
type of scatterer; in this case, the calibration can be
carried out with a set of particles with sizes that cover
the desired range with adequate density to ensure that
any non-linearities in the scattering amplitude-size
curve are detected.
5.3.14 nominal sphere size distribution — the stated
diameter distribution of a suspension of PSL spheres of
a certified diameter used in the calibration of SSISs.
5.3.14.1 Discussion — The nominal sphere size
distribution is usually expressed as the standard
deviation or the coefficient of variation of the
distribution about the certified diameter. Depending on
how the deposition is made, the size distribution of the
deposited spheres may be equal to the distribution in
the original suspension or it may be significantly
narrowed by the deposition process.
5.3.15 nuisance count — a signal pulse that arises from
discrete or area surface or near-surface features other
than the localized light scatterers being investigated;
compare false count.
5.3.15.1 Discussion — The presence of nuisance
counts is dependent on the threshold and gain settings
and may be a function of the optical configuration of
the SSIS, the orientation of the wafer surface, or both.
5.3.16 reference wafer — for calibrating an SSIS in
accordance with this practice, an unpatterned wafer
with the same surface films and finish as the wafers to
be examined by the calibrated SSIS and upon which
one or more PSL sphere depositions have been certified
to specified uncertainties for peak diameter and
diameter distribution.
5.3.17 response curve (RC) — the modeled relation
between measured scattered light intensity (or SSIS
signal response) and PSL sphere peak diameters. The
SEMI M53-1103 © SEMI 2003 5
RC depends on the light source used, is in general non-
linear and may contain regions with dips that make the
response/diameter relationship multi-valued. (See
Section 3.3).
5.3.18 scanning surface inspection system (SSIS) — an
instrument for rapid examination of the entire quality
area of a wafer to detect the presence of localized light
scatterers or haze or both; also called particle counter
and laser surface scanner.
5.3.19 threshold — the level set on a scanning surface
inspection system (SSIS) to discriminate between signal
pulses of different size.
5.3.19.1 Discussion — Thresholds may be set to
discriminate between true counts and surface or
electrical noise (nuisance or false counts, respectively)
or between different sizes of light scatterers. Because
of spatial non-uniformity of the intensity of the
scanning beam and the general use of overlapping scans
in an SSIS, a localized light scatterer with equivalent
size near the threshold may generate a signal greater
than or less than the threshold depending on its location
with respect to the path of the scanning beam. The
former is identified as a true count and the latter is
identified as a missing count.
5.3.20 true count — a laser-light scattering event that
arises from the localized light scatterers being
investigated.
5.3.21 unimodal distribution — a distribution
represented by a histogram with constant bin size that
has a single bell-shaped peak.
6 Summary of Practice
6.1 The range of LSE values to be used for calibration
is defined for each dark channel that is to be calibrated.
6.2 The number of calibration points required is
defined for each defined range.
6.3 Suitable reference wafers with certified PSL
distributions are obtained.
6.4 The reference wafers are scanned by the SSIS
being calibrated under machine conditions identical
with those to be used in examining wafers with the
calibrated SSIS.
6.5 The peak diameter of the PSL spheres deposited on
each reference wafer is assigned to the peak value of
the SSIS signal units.
6.6 An RC, the curve of SSIS channel response as a
function of SSIS calibration PSL sphere diameter, is
constructed through the data points obtained. A
separate calibration curve is developed for each
channel.
7 Apparatus
7.1 Scanning Surface Inspection System — designed to
detect, size, and map localized light scatterers (LLSs)
on unpatterned semiconductor wafers, that has the
following capabilities:
7.1.1 Scans the entire fixed quality area of the surface
of a wafer with a laser beam,
7.1.2 Detects localized light scatterers as laser-light
scattering events,
7.1.3 Has a user definable sensitivity threshold,
7.1.4 Can generate a data set file of the distribution of
the detected LLSs as a function of reported size (LSE),
7.1.5 Can generate a histogram from the data set file,
or can output the data set file in a form that can be
imported to a spreadsheet or other application program
that can generate the histogram,
7.1.6 Is sufficiently repeatable for the intended
application, and
7.1.7 Handles wafers in a Class 4 or better clean
environment as defined in ISO 14644-1.
8 Reference Wafers
8.1 Substrates — Use bare semiconductor wafers with
a native oxide (or other filmed) surface of the type
intended to be tested with the SSIS to be calibrated as
substrates for the certified depositions of the PSL
spheres. This is particularly important because SSIS
response is affected by the optical properties of the
substrate. Semiconductor wafer surfaces, such as
unpatterned polished, epitaxial, or film layers, have
different optical properties. The wafers must meet the
dimensional requirements of SEMI M1 for the
appropriate nominal wafer diameter and must be laser
marked in a manner agreed upon between supplier and
user.
8.2 Range of Calibration Diameters — Choose the
diameters of the PSL spheres so that the measurement
range for the intended application is covered. Use
spheres of size ranging from the largest measurable size
down to a size with an estimated capture rate less than
50%. Do not exceed the dynamic range of the SSIS
channel being calibrated. Use sufficient sphere
diameters to achieve the required 5% PSL sphere sizing
accuracy of the response curve between calibration
points using a calibration curve that is produced by
fitting the calibration points. In the absence of other
criteria, choose spheres with diameter ratios of
approximately 1.7.
8.2.1 To accomplish the accuracy check (see Section
10.4.2), choose an additional set of PSL spheres of size
SEMI M53-1103 © SEMI 2003 6
approximately midway between adjacent pairs of the
original set.
8.3 Background Contamination
8.3.1 Handle and store reference wafers with great care
to avoid contamination and damage.
8.3.2 Establish that the bell-shaped peaks in the SSIS
LLS histogram of each reference wafer to be used,
which are generated from the PSL sphere depositions,
are well defined and well above the background level
over all of the response curve except near the threshold.
Also verify that each unimodal bell-shaped curve
extends to less than 50% of its peak value on both sides
of the peak within a diameter range of ± 15% of the
PSL sphere diameter at the peak of the distribution. If
the deposition on any reference wafer fails either of
these criteria, obtain a new reference wafer.
NOTE 3: In most cases, the LLS histogram peaks are wider
than the actual deposition diameter distributions. See Related
Information 1 for more details.
8.4 Data to Accompany Reference Wafers — A
certificate with the following information must
accompany each reference wafer.
8.4.1 For each CRM on the reference wafer provide:
8.4.1.1 The deposition peak diameter and the
uncertainty in accordance with the requirements of Row
3.3 of Table 3 of SEMI M52.
8.4.1.2 The maximum possible value of the deposition
diameter distribution full width at half max (FWHM)
expressed as a percent of peak diameter. (See Related
Information 1.)
8.4.1.3 The particle count of each CRM and the
associated sample coefficient of variation.
8.4.1.4 The approximate location of each CRM on the
reference wafer by the x- and y-coordinates (as
specified in SEMI M20) of the center of the deposition
area or by a map or drawing of the wafer.
8.4.2 Identification of the deposition system used for
the deposition by model and serial number.
8.4.3 The date of production.
8.4.4 Wafer identification by laser mark on 200 mm
and smaller wafers per SEMI M12 or 300 mm wafers
per SEMI M1.15 including alpha numeric message.
8.4.5 Name and address of the reference wafer
manufacturer.
8.4.6 Identification of the deposited PSL spheres by
manufacturer, lot number and model.
9 Procedure
9.1 Set up the SSIS in accordance with the
manufacturer's instructions for the wafer diameter, sizes
of PSL spheres, and other machine conditions to be
used during the calibration procedure. Ensure that
machine conditions are identical with those to be used
in examining wafers with the calibrated SSIS.
9.2 Ensure that the SSIS is operating properly for the
selected machine conditions.
9.3 Load the first reference wafer into the SSIS.
9.4 Scan the wafer.
9.5 Generate a data set file of the distribution of
localized light scatterers as a function of reported SSIS
signal.
9.6 Repeat Sections 9.3 through 9.5 for each of the
depositions on the remaining reference wafers and for
all of the SSIS channels.
10 Interpretation of Data
10.1 Construct a histogram for the data set from each
of the PSL sphere depositions used for the calibration.
10.2 Determine the standard deviation and peak value
from curve fits to the histograms from each deposition.
10.3 Associate each peak value of reported SSIS
signals determined in Section 10.2 to the certified value
of PSL sphere diameter for that CRM deposited on that
reference wafer.
10.4 Create and check the SSIS response curve.
10.4.1 To obtain the SSIS response curve make a fit to
the set of peak values in such a manner that the
response to PSL sphere deposition CRMs that size
between calibration points meets the uncertainty
requirements of SEMI M52.
10.4.2 Check the accuracy of the resulting SSIS
response curve by either measuring PSL sphere
deposition CRMs with locations mid-way between
calibration points, or by comparison to a modeled
result, or both.
10.4.3 If necessary repeat the curve fitting procedure
until the required accuracy is met.
10.5 Use this fitted response curve to establish the LSE
size of localized light scatterers with any particular
machine response value.
11 Report
11.1 Report the following information:
11.1.1 Operator identification;