semi合集-English.pdf - 第5305页
SEMI M49-0704 © SEMI 2001, 2004 4 Otherwise m atching tolerance tests are perf ormed under the conditions of σ 3 tests. NOTE 4: In the absence of certified or standard reference materials ma tching may be tested by u sin…
SEMI M49-0704 © SEMI 2001, 2004
3
50441-4 — Determination of the geometric dimensions
of semiconductor slices; diameter and flat depth of
slices
50441-5 — Determination of the geometric dimensions
of semiconductor wafers; terms of shape and flatness
deviation
50445 — Contactless determination of the electrical
resistivity of semiconductor wafers with the eddy
current method
NOTE 1: Unless otherwise indicated, all documents cited
shall be the latest published versions.
4.5 IEEE Standards
4
IEEE 754 — IEEE Standard for Binary Floating-Point
Arithmetic
IEEE 802 — IEEE Standard for Local and Metropolitan
Networks: Overview and Architecture
IEEE 854 — IEEE Standard Radix-Independent
Floating-Point Arithmetic
4.6 Other Standards
FED-STD 209E —Airborne Particulate Cleanliness
Classes in Clean Rooms and Clean Zones
5
NOTICE: Unless otherwise indicated, all documents
cited shall be the latest published versions.
5 Terminology
5.1 Abbreviations and Acronyms
5.1.1 ARAMS — Automated Reliability, Availability,
and Maintainability Standard
5.1.2 ASCII — American Standard Code for
Information Interchange
5.1.3 FTP — File Transfer Protocol
5.1.4 GEM — Generic Equipment Model
5.1.5 HSMS — High Speed SECS Messaging Service
5.1.6 IEEE — The Institute of Electrical and
Electronics Engineers, Inc.
5.1.7 JPEG — Joint Photographics Expert Group
5.1.8 SECS — SEMI Equipment Communications
Standard
5.1.9 XML — Extensible Markup Language
5.2 Definitions
4 Institute of Electrical and Electronics Engineers, IEEE Operations
Center, 445 Hoes Lane, P.O. Box 1331, Piscataway, New Jersey
08855-1331, USA. Telephone: 732.981.0060; Fax: 732.981.1721
6 General Services Administration, Federal Supply Service, FSS
Acquisition Management Center, Environmental Programs and
Engineering Policy Division (FCOE), Washington, D.C. 20406
5.2.1 bias — the difference between the average of
measurements made on the same object and its true
value. Sufficient measurements are needed to mitigate
the effects of variability. (SEMI E89)
5.2.2 calibration — calibration is a measurement
process that assigns value to the property of an artifact
or to the response of an instrument relative to reference
standards or to a designated measurement process.
NOTE 2: The purpose of calibration is to eliminate or reduce
bias in the user’s measurement system relative to the
reference base. The calibration process compares an unknown
or test item or instrument with reference standards according
to a specific algorithm, often in the form of a specific
calibration curve. (SEMI E89)
5.2.3 compatibility — the capability of measurement
equipment to emulate the measurement process of other
tools. Downward compatibility refers to former
generation(s) of the same or similar type of equipment
of an equipment supplier.
NOTE 3: Compatibility can be provided by a measurement
mode in which filtering, spatial resolution, etc. of another,
older, tool is imitated.
5.2.4 correlation — the relation of measurement
results obtained by repeated measurements with the
same set of test specimen(s) and any two measurement
tools expressed in terms of a regression curve.
5.2.5 level 1 variability (σ
1
) — the variation (standard
deviation) of measurement results obtained by repeated
measurements with the same test specimen(s) and the
same measurement tool under nominally identical
conditions without replacing the test specimen between
subsequent measurement runs. σ
tests are performed
with a single calibration in the shortest possible time
interval.
5.2.6 level 2 variability (σ
2
) — the variation (standard
deviation) of measurement results obtained by repeated
measurements with the same test specimen(s) and the
same measurement tool with replacing the test
specimen between subsequent measurement runs but
otherwise under nominally identical conditions. σ
2
tests
are performed with a single calibration in the shortest
possible time interval.
5.2.7 level 3 variability (σ
3
) — the variation (standard
deviation) of measurement results obtained by repeated
measurements with the same test specimen(s) and the
same measurement tool with replacing the test
specimen between subsequent measurement runs but
otherwise under nominally identical conditions. σ
3
tests
are performed over a time period greater than σ
2
tests
without operator induced adjustment.
5.2.8 matching tolerance (∆
m
) — the difference in bias
for any two measurement tools of the same kind.
SEMI M49-0704 © SEMI 2001, 2004 4
Otherwise matching tolerance tests are performed under
the conditions of σ
3
tests.
NOTE 4: In the absence of certified or standard reference
materials matching may be tested by using appropriate wafers
complying with 130 nm technology node specifications. It is
recommended to test for matching with a set of samples
covering the parameter range of interest.
5.2.9 precision over tolerance (P/T) ratio — the ratio
of the precision of measurement equipment and a
product’s tolerance. (SEMI M27)
5.2.10 sorting — real and virtual separation of test
specimens in different categories specified by one or
multiple parameters.
5.2.11 tolerance — the absolute magnitude of the
range of the product specification. (SEMI M27)
6 Specification for Geometry Measurement
Equipment for Silicon Wafers
6.1 The specification is structured in three sections
(Tables 1–3):
• Generic Equipment Characteristics (Table 1)
• Materials to be measured (Table 2)
• Metrology Specific Equipment Characteristics
(Table 3)
6.2 Tables 1–3 contain the specifications, referenced
documents, test methods, and comments. Additional
explanations and discussions are provided in this
section.
6.3 Generic Equipment Characteristics (Table 1)
6.3.1 The section “Generic Equipment Characteristics”
consists of five subsections:
• Wafer handling
• Reliability
• Procedural
• Documentation
• Computer/User Interface/Connectivity
6.3.2 Subsections covering “Facilities Requirements”
and “Safety/Legal/Regulatory” are not included in the
present document as these issues are highly user
specific and dependent on national regulations.
6.4 Materials to be measured (Table 2)
6.4.1 Table 2 specifies the parameters of Si wafers that
the measurement equipment must be capable to handle
and to measure.
6.5 Metrology Specific Equipment Characteristics
(Table 3)
6.5.1 This section specifies the dimensional parameters
of Si wafers to be measured and to be reported by
equipment for measuring the geometry and flatness of
wafers as well as the required spatial resolution,
precision and accuracy of the measurement equipment.
6.5.2 The ability of a metrology tool to properly
measure surface features of different spatial
wavelengths is affected by the spatial bandwidth of the
tool's response function. Spatial bandwidth can be
defined in many ways and is influenced by many
factors beyond the scope of this document. Some of
these need to be standardized.
6.5.3 Spatial resolution is defined by the high spatial
frequency limit of the bandwidth of the tool's response
function.
6.5.4 Low and high cut-off frequency f
min
and f
max
define the bandwidth of the response function of
measurement equipment. The cut-off frequencies
correspond to an attenuation of 0.5 for the amplitude of
a sinusoidal surface feature with the exception of a low
pass filter (f
min
= 0) for which the attenuation remains 1
at f
min
.
6.5.5 The rate of change of the attenuation approaching
the cut-off frequencies has to be larger than the rate of a
Gaussian filter with the corresponding cut-off
frequency.
6.5.6 The present document recommends f
max
and f
min
that must be measured by the instrument.
6.5.7 Any variations in filtering procedures applied
near the FQA boundary must be described by the
supplier of the equipment.
6.5.8 The bandwidth as specified in Table 3 is a
nominal value.
6.5.9 In the present document a hierarchy of variability
levels is used to describe the performance of
measurement equipment which is calibrated and
adjusted/aligned according to the supplier's procedures.
The various terms are defined in section 5. These
variability levels are consistent with terms defined in
SEMI E89 but not fully interchangeable. Their relation
is indicated in parentheses.
6.5.9.1 Level 1 variability: standard deviation σ
1
(SEMI E89 static repeatability)
6.5.9.2 Level 2 variability: standard deviation σ
2
(SEMI E89 dynamic repeatability)
6.5.9.3 Level 3 variability: standard deviation σ
3
(SEMI E89 reproducibility)
SEMI M49-0704 © SEMI 2001, 2004
5
6.5.10 In addition two levels of systematic off-set
between different tools are defined:
6.5.10.1 matching tolerance (difference of means ∆
m
)
6.5.10.2 correlation (regression curve)
6.5.11 Explicitly specified in the present document are
only level 3 variability σ
3
and matching tolerance ∆
m
as
they correspond to the utilization of measurement
equipment for wafer manufacturing most closely. The
supplier of a specific tool may optionally provide
specifications for level 1 and/or level 2 variability,
respectively.
6.5.12 Level 3 variability σ
3
and matching tolerance ∆
m
are specified with respect to anticipated specifications
for wafer geometry and flatness as given in Table 3 for
a reference wafer.
6.5.13 In the present document P/T-ratios are used for
specifying level 3 variability σ
3.
6.5.13.1 A precision-to-tolerance ration P/T less than
10% at 6σ is recommended in SEMI M27 for
metrology equipment. This would be an extremely
demanding specification for flatness and geometry
measurement tools. In addition, flatness characterisitics
of Si wafers are typically described by a single sided
distribution with the median approaching zero, the
lower specification boundary. Therefore two grades that
are based on 3σ criteria, instead on 6σ, are
recommended for such tools in the present document:
• grade A: P/T < 10%, 3σ
3
• grade B: P/T < 20%, 3σ
3
6.5.13.2 The individual measurement features a tool
provides may be graded differently, e.g. SFQR might
meet grade A, but SBIR only grade B. This has to be
indicated appropriately in the tools’ technical
specifications.
6.5.14 Matching tolerance is specified to be less or
equal to 1.5σ
3
of level 3 variability. This corresponds to
a greater than 99% probability that the difference of any
individual measurement results obtained with two
different tools is smaller or equal to 5σ
3
.
6.5.15 The target for bias is a range of ± 1.5σ
3
with
respect to a certified value provided appropriate
reference materials are available. This corresponds to a
greater than 99% probability that any individual
measurement is in the range of ± 4σ
3
around the
certified value when a reference material is tested.
6.5.16 Reference material with a series of surface
features with appropriate height and half width is
required to verify bandwidth of a measurement tool.
The height of the features corresponds to the wafer
specification as outlined in Table 3.
6.5.17 The specifications of the measurement
equipment are verified by using wafers the parameters
of which are in a range, the upper limit of which
corresponds to 1.5 times anticipated wafer
specification, the lower limit to 0.5 of anticipated wafer
specification. These are listed in Table 3 as Reference
Wafer Specifications.
NOTE 5: The edge region of wafers represents the most
challenging area for meeting the desired performance
characteristics. This is because of the larger surface geometry
variations in the region near the edge, e.g. from polishing roll-
off.
6.5.18 Reference Wafers — The specifications of the
measurement equipment are verified by using reference
wafers with properties covering the range given in
Section 1 of Table 3. All reference wafers shall meet
the thickness and warp requirements listed in the table
but different wafers may be used to meet the flatness
and nanotopography requirements. For site related
specifications not all sites must fall within the range,
but the appropriate sites to be tested should be
indicated. In all cases, at least three wafers in the range
of values for each property shall be employed in the
testing.
6.5.19 Verification of bias, matching tolerance and the
various levels of variability are performed with
equipment which is calibrated according to the
supplier's procedures and which is under statistical
process control.
6.5.20 Compatibility of two tools is considered to be
satisfactory when the specifications of the older tool are
met with the newer tool operating in the emulation
mode.
6.5.21 The quality of a correlation between different
measurement equipment is not specified in the present
document.
7 Related Documents
7.1 ASTM Documents
E 177 — Standard Practice for Use of the Terms
Precision and Bias in ASTM Test Methods
E 456 — Standard Terminology for Relating to Quality
and Statistics