semi合集-English.pdf - 第5220页
SEMI M40-0200 © SE MI 2000 2 4.2 Historically a v ariet y o f roug h n e ss parameters for profiling techniques have evolved and have been standardized, in cludi ng rms roughness (R q ) and average rough ness ( R a ). Pr…
SEMI M40-0200 © SEMI 20001
SEMI M40-0200
GUIDE FOR MEASUREMENT OF SURFACE ROUGHNESS OF PLANAR
SURFACES ON SILICON WAFER
This guide was technically approved by the Global Silicon Wafer Committee and is the direct responsibility
of the North American Silicon Wafer Committee. Current edition approved by the North American Regional
Standards Committee on December 15, 1999. Initially available at www.semi.org February 2000; to be
published February 2000.
1 Purpose
1.1 This guide provides procedures for specifying the
measurements to be used in characterizing and
reporting roughness of the planar surfaces of silicon
wafers. It may also be applicable to other types of
planar wafer materials.
1.2 This guide provides nomenclature and procedures
for roughness determination that employ three key
methodologies:
1.2.1 Standardized scan site patterns ,
1.2.2 Roughness abbreviations, and
1.2.3 Reference test methodologies with respect to
identifying specific roughness measurements.
2 Scope
2.1 This guide incorporates the following
methodologies:
2.1.1 Standardized scan patterns for both local and
full-area surface characterization,
2.1.2 A set of roughness abbreviatio ns that describe
measurement conditions in a short-hand code, and
2.1.3 Reference test methodologies for three generic
types of roughness measuring instruments. These
general categories may include, but are not limited to:
• Profilometers — AFM and other scanning probe
microscopes; optical profilometers; high-resolution
mechanical stylus systems,
• Interferometers — interference microscopes, and
• Scatterometers — Total integrating scatterometers
(TIS), angle-resolved light scatterometers (ARLS),
scanning surface inspection systems (SSIS).
2.2 Procedures to obtain a represe ntative value of
roughness for a surface are specified.
2.3 Roughness nomenclature is int ended to remove
ambiguities with respect to identifying the roughness
measurements used and the results achieved.
2.4 This guide does not purport to address safety
issues, if any, associated with its use. It is the
responsibility of the users of this guide to establish
appropriate safety and health practices and determine
the applicability of regulatory limitations prior to use.
3 Limitations
3.1 This guide does not apply to measurements in the
edge region of wafers.
3.2 This guide does not apply to sp atial wavelengths <
10 nm.
3.3 This guide is not intended to d efine specific
roughness parameters; these can be found in other
documents.
3.4 The bandwidth of the measure ment tool and the
bandwidth used can severely influence the result of
roughness measurements.
3.4.1 Differences in either long or short bandwidths
used between two instruments can produce significantly
different results.
3.4.2 The codes listed in Table 1 of this guide do not
identify instrument transfer functions.
3.5 The presence of films may affe ct light scattering
measurements.
4 Discussion
4.1 Roughness of silicon wafer su rfaces is becoming
frequently specified for bare silicon wafers (cf. the SIA
International Technology Roadmap for Semiconduc-
tors). These specifications refer to the roughness of
both the final polished front and the back surfaces of a
wafer. Various techniques are currently used to
measure the surface roughness of silicon wafers. These
techniques include AFM scanning probes, mechanical
and optical profiling, interferometric microscopes, and
light (electromagnetic radiation) scattering. Mechanical
profiling techniques are widely used in other industries
and are well standardized. Profiling techniques are
generally limited to line scans and therefore provide
information of only a very small part of a surface.
Interferometric techniques for roughness measurements
are similarly limited, whereas light scattering
techniques can scan the entire wafer surface.
SEMI M40-0200 © SEMI 2000 2
4.2 Historically a variety of rough ness parameters for
profiling techniques have evolved and have been
standardized, including rms roughness (R
q
) and average
roughness (R
a
). Profiling instruments measure the
surface topography and derive roughness statistics, such
as R
q
, from a series of height data. Light scattering
techniques measure R
q
from the angular dependence of
scattering; some light scattering systems derive other
parameters through interpretation and estimation.
4.3 A roughness value representative of an entire
wafer surface, or a large portion of the surface, cannot
be based on data obtained at a single point. Yet,
specifications often describe a single value for a wafer.
Therefore, this guideline suggests and defines
standardized patterns of scan sites that can be described
and used unambiguously. A model describing the
relationship between several types of roughness
variation, scan patterns, and reported results is included
in the attached Related Information to assist users in
specifying and interpreting these variables.
4.4 A common feature of all roughness measurements
is their dependence on the bandwidth and transfer
function of the tool used. In addition, high or low
spatial frequency software filters are common and
affect reported results. Widely different numbers can
be reported for the same surface by two measurement
instruments. This guideline includes suggestions on
specifying and reporting the instrument bandwidth and
transfer function.
5 Referenced Standards
5.1 SEMI Standards
SEMI M1 — Specifications for Polished
Monocrystalline Silicon Wafers
SEMI M20 — Specification for Establishing a Wafer
Coordinate System
5.2 ASME Standard
1
ASME B46.1 — Surface Texture (Surface Roughness,
Waviness, and Lay)
5.3 ASTM Standards
2
E 1392 — Practice for Angle Resolved Optical Scatter
Measurements on Specular or Diffuse Surfaces
F 1048 — Test Method for Measuring the Effective
Surface Roughness of Optical Components by Total
Integrated Scattering
1 American Society of Mechanical Engineers, United Engineering
Center, 345 East 47th Street, New York, NY 10017
2
American Society for Testing and Materials, 100 Barr Harbor
Drive, West Conshohoken, PA 19428-2959
F 1620 — Practice for Calibrating a Scanning Surface
Inspection System Using Monodisperse Polystyrene
Latex Spheres Deposited on Polished or Epitaxial
Surfaces
F 1811 — Estimating the Power Spectral Density
Function and Related Finish Parameters from Surface
Profile Data
5.4 DIN Standards
3
DIN 4760 — Form Deviation, Waviness, Surface
Roughness; System of Order, Terms and Definitions
DIN 4768 — Determination of Values of Surface
Roughness Parameters Ra, Rz, Rmax by Means of
Electrical Contact (Stylus) Instruments; Terminology,
Measuring Conditions
DIN 4777 — Metrology of Surfaces; Profile Filters for
Electrical Contact Stylus Instruments; Phase-Corrected
Filters
5.5 ISO Standards
4
ISO 468 — Surface Roughness – Parameters, Their
Values and General Rules for Specifying Requirements
ISO 1879 — Instrument for the Measurement of
Surface Roughness by the Profile Method - Vocabulary
ISO 1880 — Instruments for the Measurement of
Surface Roughness by the Profile Method - Contact
(Stylus) Instruments of Progressive Profile
Transformation - Profile Recording Instruments
ISO 3274 — Instruments for the Measurement of
Surface Roughness by the Profile Method – Contact
(Stylus) Instrument of Consecutive Profile
Transformation – Contact Profile Meters, System M
ISO 4287/1 — Surface roughness – Terminology – Part
1: surface and its parameters
ISO 4288 — Rules and Procedures for the
Measurement of Surface Roughness Using Stylus
Instruments
5.6 JIS Standards
5
JIS B 0601 — Surface Roughness - Definitions And
Designation
JIS B 0652 — Instruments For The Measurement Of
Surface Roughness By The Interferometric Method
JIS B 0659 — Roughness Comparison Specimens
3 Deutches Institut für Normung e.V., Beuth Verlag GmbH,
Burggrafenstrasse 4-10, D-10787 Berlin, Germany
4 ISO Central Secretariat, 1, rue de Varembé, Case postale 56, CH-
1211 Genève 20, Switzerland
5
Japanese Standards Association, 1-24, Akasaka, 4-Chome, Minato-
ku, Tokyo 107 Japan
SEMI M40-0200 © SEMI 20003
5.7 Other Documents
International Technology Roadmap for
Semiconductors
6
Optical Scattering in the Optics, Semiconductor, and
Computer Disk Industries, Second Edition 1995, J C
Stover, Editor
7
NOTE 1: As listed or revised, all documents cited shall be the
latest publications of adopted standards.
6 Terminology
6.1 autocorrelation function — th e Fourier transform
of the Power Spectral Density function. It expresses the
similarity between a surface profile and the same
profile that is slipped, or moved laterally, with respect
to itself.
6.2 autocorrelation length — the l ateral slip required
to reduce the Autocorrelation function to a value equal
to e-1 times its zero slip value. Sometimes 10% or even
0 value definitions are used instead of e-1.
6.3 average roughness (R
a
) — the average of the
surface profile height deviations Z(x) from the mean
line taken within the evaluation length (ASME B46.1).
6.4 bi-directional reflectance distribution function,
BRDF — a description of the distribution of light
scattered by a surface, it is the differential radiance
normalized by the differential irradiance, and is
approximated by the scattered power per unit projected
solid angle divided by the incident power (Stover).
6.5 fixed quality area (FQA) — the central area of a
wafer surface, defined by a nominal edge exclusion, X,
over which the specified values of a parameter apply
(SEMI M1). See also the discussion immediately
following this definition in SEMI M1.
6.6 haze — non-localized light sca ttering resulting
from surface topography (microroughness) or from
dense concentrations of surface or near-surface
imperfections (SEMI M1).
DISCUSSION — Haze due to the existence of a
collection of imperfections is a mass effect; individual
imperfections of the type which result in haze cannot be
readily distinguished by the eye or other optical
detection systems without magnification. In a particle
counter (SSIS), haze results in a background signal and
laser light-scattering events together comprise the
signal due to light-scattering from a wafer surface.
(SEMI M1). It is the total scattered optical flux
6 Semiconductor Industry Association (SIA), 181 Metro Drive, Suite
450, San Jose, CA
7 SPIE, P.O. Box 10, Bellingham, WA 98227-0010
collected by an optical system normalized by the
incident flux.
NOTE 2: Different SSIS instrument types may give
significantly different haze values on a given sample.
6.7 illumination source incidence angle the angle
of the incoming beam, measured from surface normal.
6.8 kurtosis (Rku) — a measure of the sharpness of the
histogram of surface profile height deviations Z(x) from
the mean line within the evaluation length. A complete
random surface will have a Gaussian histogram and Rku
= 3 (ASME B46.1).
6.9 laser light-scattering event — a signal pulse that
exceeds a preset threshold, generated by the interaction
of a laser beam with a discrete scatterer at a wafer
surface as sensed by a detector; see also haze (SEMI
M1). See also the discussion that follows this definition
in SEMI M1.
6.10 lay – the predominant directio n of the surface
pattern, ordinarily determined by the production
method used (ASME B46.1).
6.11 microroughness — surface roughness
components with spacing between irregularities (spatial
wavelength) less than about 100 µm (SEMI M1).
6.12 Nyquist Criterion — the shorte st spatial
wavelength detected. It is twice the sample spacing.
6.13 one-dimensional grating equa tion — in its most
common form, it is an expression that gives the
positions of diffracted orders from a one-dimensional
sinusoidal grating (Stover).
6.14 peak to valley (R
t
) — the highe st to lowest value
of the surface profile height deviations Z(x) from the
mean line taken within the evaluation Length L (ASME
B46.1).
6.15 power spectral density (PSD) function — a
surface characterization function that is proportional to
the square of the modulus of the Fourier transform of
the surface and may be considered as a roughness
power per unit of spatial frequency (ASTM F 1811).
6.16 Rayleigh Criterion (of resolvin g power) — a
condition for distinguishing a pair of diffraction
patterns whereby the maximum of one pattern overlaps
with the minimum of the other.
DISCUSSION — When a lens is free from aberrations,
the images of point objects appear as diffraction
patterns. When the principle maximum of one pattern
strikes the first minimum of another, the images are
described as being resolved. With respect to circular
optics, this criterion applies when the distance between
resolvable point objects, viewed from the objective lens
of the instrument, is