semi合集-English.pdf - 第5175页
SEMI M33-0998 © SE MI 1998 2 SEMI M1 — Specifications f or Polished Monocrystalline Silicon Wafers SEMI M2 0 — Specification for Establishing a Wafer Coordin ate System 3.2 AST M Specifications 1 ASTM D 5127 — Stan dard …
SEMI M33-0998 © SEMI 19981
SEMI M33-0998
TEST METHOD FOR THE DETERMINATION OF RESIDUAL SURFACE
CONTAMINATION ON SILICON WAFERS BY MEANS OF TOTAL
REFLECTION X-RAY FLUORESCENCE SPECTROSCOPY (TXRF)
1 Purpose
1.1 The test provides the analytica l procedure to
determine the trace level of contaminating elements of
an atomic number higher than 15 on polished or
epitaxial silicon wafer surfaces in native or thermally
grown or tetraethylorthosilicate (TEOS) oxide or in
residues of microdroplets of process chemicals or
media as analyzed with TXRF on silicon wafer surfaces
as described in Sections 15.1 and 15.2.
2 Scope
2.1 This document specifies a VPD -TXRF (Vapor
Phase Decomposition Total Reflection X-Ray
Fluorescence Spectroscopy) method to analyze the
elemental composition and areal density of impurities,
that include cations and anions with atomic numbers
between 16 (S) and 92 (U) independent of their
chemical state, with the exception of the X-ray source
material, on polished or epitaxial silicon wafer surfaces
in native or thermally grown oxide or in residues of
microdroplets of process chemicals or media as
analyzed with TXRF on silicon wafer surfaces.
2.2 This test is especially useful fo r analyzing metallic
elements such as K, Ca, Ti, V, Cr, Mn, Fe, Co, Ni, Cu,
Zn, (Mo), Pd, Ag, Sn, Sb, Ta, (W), Pt, (Au), Hg, and Pb
and non-metallic elements such as S, Cl, As, Br, and I
through their characteristic K and L lines. (Elements in
brackets are usual X-ray sources.) For limitations in the
nature of analytes refer to the note in Section 14.7.
2.3 This test method can be used t o analyze areal
surface contamination that can be collected in a
microdroplet during the specified VPD preparation and
the collection of the digested surface contamination in
the range of 5 × 10
8
through 5 × 10
12
atoms/cm
2
.
2.4 Theoretically, the detection limit (LOD) of each
analyte depends upon its atomic number. As defined by
DIN 32645 “Limit of detection, determination and
quantification” the LOD of TXRF is also depending
upon many parameters, such as:
— excitation energy,
— intensity of incident X-ray,
— instrumental background,
— crystallographic interferences, such as Bragg
diffraction conditions,
— impurities in the beam path,
— contamination of the blank scanning solution
(see Section 4.5),
— contamination level in the analytical ambient,
— surface microroughness of wafer at the
microdroplet (see Section 6.7), and
— integration time.
2.5 Concerning the surface conditions to be analyzed,
the VPD-TXRF method is invasive. Nevertheless, the
TXRF analysis of the microdroplet residue can be
repeated many times provided that the prepared
specimen is stored in a clean environment. The
substrate and/or surrogate wafers can be recycled for
monitoring purposes.
2.6 The user of this test method must assure that the
metrology equipment is under control by the procedures
commonly utilized in the performing laboratory. In the
absence of established control procedures the use of
4.11.2 EN-ISO 9001 is recommended.
NOTICE SAFETY PRECAUTIONS — This standard
does not purport to address the safety concerns,
associated with its use. It is the responsibility of the
user of this standard to establish and maintain
appropriate safety and health practices and comply with
the local regulatory ordinance. X-ray irradiation and
handling of HNO
3
, HF and H
2
O
2
are dangerous.
Operators must comply with X-ray safety regulations
and be trained to wear protective garments and glasses
when handling HNO
3
, HF and H
2
O
2
. These chemicals
should be handled in a ventilated area (under exhaust.)
3 Referenced Documents
3.1 SEMI Standards
SEMI C7.3 — Standard for Hydrofluoric Acid, Grade 2
SEMI C7.5 — Standard for Hydrogen Peroxide, Grade
2
SEMI C7.6 — Standard for Nitric Acid, Grade 2
SEMI C10.1 — Guide for Determination of Method
Detection Limits for Trace Metal Analysis by Plasma
Spectroscopy
SEMI E45 — Test Method for the Determination of
Inorganic Contamination from Minienvironments
SEMI M33-0998 © SEMI 1998 2
SEMI M1 — Specifications for Polished
Monocrystalline Silicon Wafers
SEMI M20 — Specification for Establishing a Wafer
Coordinate System
3.2 ASTM Specifications
1
ASTM D 5127 — Standard Guide for Electronic Grade
Water (Type E-1)
ASTM E 691 — Practice for Conducting an
Interlaboratory Study to Determine the Precision of a
Test Method
ASTM F 1526 — Test Method for Surface
Metals/TXRF
3.3 DIN Specifications
2
DIN-ISO 5725 — Precision of Test Methods,
Evaluation of Round-robins
DIN 12650 Part 6 — Gravimetrical Test for Piston
Operated Volumetric Apparatus
DIN 12650 Part 6 (Apr. 1983) — Gravimetrical Test
for Piston Operated Volumetric Apparatus
DIN 32645 — Limit of Detection and of Quantification
3.4 ISO Specifications
3
EN-ISO-DIN 9001 — Quality Systems; Quality
Assurance
3.5 Other Specification
4
U.S. Federal Standard 209 — Airborne Particulate
Cleanliness Classes in Cleanrooms and Clean Zones
4 Terminology
4.1 anglescan — A measurement of the emitted
fluorescene signal as a function of the glancing angle of
incident X-ray beam.
4.2 areal density — Amount of im purities in a unit
area of native or thermally grown silicon oxide as
converted from the detected amount of analytes into the
whole analyzed (i.e., scanned surface area).
1 American Society for Testing and Materials, 100 Barr Harbor
Drive, West Conshohoken, PA 19428-2959
2 Deutches Institut fur Normung e.V., Beuth Verlag GmbH,
Burggrafenstrasse 4-10, D-10787 Berlin 30, Germany
3 ISO Central Secretariat, C2.P2 56 Clf-1211 Geneve 20,
Switzerland, available in the U.S. from American National Standards
Institute, 11 West 42nd Street, 13th Floor, New York, NY 10036
4 Available from Standardization Documents Order Desk, Bldg. 4,
Section D, 700 Robbins Avenue, Philadelphia, PA 19111-5094, Attn:
NPODS
4.3 as-polished wafer — Mirror-fi nished wafer
planarized by chemi-mechanical polishing.
4.4 azimuthal position — Orientat ion around the z-
crystal/ingot axis as specified in SEMI M20.
4.5 contamination collection — M icroanalytical
method to collect the VPD decomposition products
from the silicon surface by rolling a scanning
microdroplet on the hydrophobic silicon wafer surface
after VPD preparation as originally described in
Sections 15.2, 15.4, and 15.5 (also see Section 8.3 of
this document).
4.6 critical angle — The incident X-ray glancing
angle below which total reflection of the incident X-ray
occurs. At the critical angle the X-ray reflection equals
0.5.
4.7 detection spot area — The sur face area where
above the fluorescence counts are integrated.
4.8 epitaxial wafer — As-polished wafer covered with
a layer of monocrystalline silicon deposited from a
heterogeneous phase.
4.9 glancing angle — Incidence a ngle of X-ray
excitation.
4.10 hydrophobic surface — Contact angle of wafer >
60° (e.g., virgin epitaxial or HF- or HMDS
(hexamethyldisilazane) -treated surface as described in
Section 15.3).
4.11 impurities — Elements in/on t he specimen other
than silicon or elements in ultra pure process media as
listed in Section 2.2.
4.12 native oxide — Compound of silicon, oxygen,
and water on as-polished or epitaxial wafer, grown in
air or in cleaning solutions.
4.13 recovery rate — The ratio of a nalytes found after
the first VPD and contamination collection procedure to
the sum of the analytes found after two or more
repeated scanning with unused scanning droplet of
unchanged chemical composition.
NOTE: Recent efforts of the Statistical Task Force of the
SEMI Chemicals and Gases Committee may result in a new
definition (see Section 14.2 of this document).
4.14 spurious peaks — Peaks that are detected but not
originated from impurities of the silicon wafer (c.f.,
6.1).
4.15 thermally grown oxide — SiO
2
up to 1000 nm
thickness deposited or grown in thermal processing in
oxygen containing atmosphere.
4.16 vapor phase decomposition (VPD) — Vapor
phase decomposition of silicon oxides using HF vapor
at room temperature as a surface preparation method for
SEMI M33-0998 © SEMI 19983
microanalysis, as originally described in Sections 15.2,
15.4, or 15.5 (also see Sections 5.1 and 11 of this
document).
5 Summary of Method
5.1 The native or thermally grown oxide layer of the
silicon surface is converted with HF vapor into fluid
droplets that contain the impurities of the oxide layer.
With a scanning droplet the fluid reaction products are
collected in one microdroplet. That microdroplet is
dried on the wafer under controlled conditions and
analyzed with TXRF.
5.2 Similarly, a microdroplet of process chemicals or
media can be dried on hydrophobic polished or
epitaxial silicon wafer under controlled conditions and
analyzed with TXRF.
5.3 Preferably, monochromatic and collimated X-rays
irradiate a planarized and chemi-mechanically polished,
monocrystalline silicon wafer surface. The X-rays
impinge the surface at a glancing angle that is below the
angle for total reflection of the X-rays, preferably, at an
incident angle 70% of the angle of total reflection [1.3
mrad (or 0.07 degrees) for Mo target and 2.0 mrad (or
0.11 degrees) for W target].
5.4 The evanescent waves excite the fluorescence
energy levels of the surface atoms, which then emit
fluorescence X-rays characteristic of their atomic
number. The emitted X-rays are detected by a solid
state detector that is an energy dispersive spectrometer.
In the range of specified areal density (compare with
Section 2.3 of this document) the integrated count rate
(cps) is linearly proportional to the elemental areal
density.
5.5 For quantification, the linear re gression must be
established (c.f., Section 15.6) or a linear
proportionality is anticipated between the cps data that
are measured above the certified reference
microdroplet(s) and the cps data that are measured
above the microdroplet of the unknown analytes
without changing the anglescan conditions, according
to Sections 9.1, 13.6, and 15.7.
6 Interferences
6.1 The known interferences in X-ray fluorescence
spectroscopy also affect TXRF. Thus, overlapping
fluorescence lines, escape peak, energy gain drift, X-ray
source stability, beam path background contamination
must be evaluated according to Section 15.7 of this
document.
6.2 Baseline corrections due to va rying background
contamination must be controlled by the rules of
statistical analysis (e.g., as described in Section 15.8 of
this document).
6.3 Under the specified conditions, no corrections are
required for secondary fluorescence or for oscillations
or for matrix absorption as described in Sections 15.9–
15.11 of this document.
6.4 Accuracy of the standard refer ence specimen and
positioning accuracy and precision of the detector
define the bias in the assigned areal density.
6.5 Mechanical vibration may degrade the detector
resolution and it can also decrease the selectivity.
6.6 Multielement contamination d egrades the LOD
compared with monoelement contamination.
6.7 Increased surface microroughness and/or high-
total signal count rates result in high deadtime and can
lead to non-linearity of detected fluorescence signal
versus areal density (i.e., to degradation of LOD).
6.8 Under optical conditions satisf ying the Bragg
reflections the background noise depends upon the
azimuthal orientation of the sample. Before
quantification, a determination of the azimuthal angular
range, that shows minimum Bragg reflection
background, is recommended. Otherwise the LOD may
degrade due to high background and spurious peaks.
6.9 During handling and measurem ent particles or
volatile contamination (e.g., NH
3
) from the analytical
environment must be controlled and avoided.
6.10 During measurement Ar must be excluded from
the analytical ambient (e.g., by evacuating the chamber
or flushing it with He).
6.11 Curve smoothing and evaluatio n algorithms with
controlled Fourier parameters or Digital Filtering are
preferred to direct count rate evaluation because these
algorithms provide a higher level of statistical
confidence than a software that directly quantifies cps
as described in Section 15.12 of this document.
6.12 Recovery rates as defined in Section 4.13 of this
document depend upon the distribution of the analytes
between the scanning solution (solubility) and silicon
surface (adsorption and plating). Therefore, recovery
rates depend on the:
— applied scanning solution (Sections 4.5, 8.3),
— chemical nature of the different
analytes/elements, and
— physical and physicochemical state of the silicon
surface.
6.13 Automated contamination coll ection procedure
(scanning) increases the wafer-to-wafer reproducibility