semi合集-English.pdf - 第7467页

SEMI MF1617-0304 © SEMI 2003, 2004 3 secondary ion coun t rate or ion count r ates of known relative intensity such as natural 28 Si/ 30 Si) on each detector. 6.7 The net integrated 23 Na, 27 Al, 39 K, and 54 Fe signals …

SEMI MF1617-0304 © SEMI 2003, 2004 2

from plastic cassette containers. The relative

importance of the organic interference is dependent

upon the surface organics of the test wafer.

3.1.2 Another significant interference occurs from

ubiquitous BO

when the aluminum is in the range of

to 10

atoms/cm

, since surface boron at the 10

atoms/cm

range is common for all wafers, both n-type

as well as p-type.

3.1.3 If the surface contains high levels of sodium,

there may be a NaO

molecular interference for

In principle,

and

can be a molecular

interferences for

and

K respectively.

3.1.4 The

+ 1

signal can have interferences from

+ 1

3.1.5 Discrimination of molecular ion interferences can

be achieved using magnetic mass spectrometers

operated under high mass resolution or in some cases

using quadrupole mass spectrometers via energy

filters.

3.2 Bias in the relative sensitivity factors (RSFs)

derived from reference materials can introduce bias into

the SIMS measured areal densities.

3.3 Mass interferences can introduce bias if the

instrument mass resolution, or subsequent detection

scheme, is not sufficient to exclude the interference.

3.4 The SIMS sodium, aluminum, potassium, and iron

instrumental backgrounds can limit the detection of low

levels of surface metals.

3.5 The accuracy and precision of the measurement can

be significantly degraded by analysis of specimens

whose surfaces are not all at the same inclination with

respect to the ion collection optics of the SIMS

instrument. The specimen holder must be constructed

and maintained such that after specimen(s) are loaded

into the holder, the inclination of the surface of each

specimen is constant from specimen to specimen.

3.6 The accuracy and precision of the measurement

significantly degrade as the roughness of the specimen

surface increases. This degradation can be avoided by

using chem-mechanical polished surfaces.

3.7 If an oxygen leak is not used in conjunction with

the measurement, there may be a bias due to the effect

1 Mollenkopf, H., “Chemicals and Cleanroom Filtered Air Effects on

Boron Contamination and Its Near Surface Detection in Silicon

Wafers,” Extended Abstracts, Vol 93-2, Abstract No. 170, The

Electrochemical Society, Pennington, NJ, 1993, pp. 273–274.

2 Frost, M. R., “On the Use of Quadrupole SIMS for the

Measurement of Surface Metallic Contamination,” Contamination

Control and Defect Reduction in Semiconductor Manufacturing III,

ECS Proceedings Volume 94-9, edited by D. N. Schmidt (The

Electrochemical Society, Pennington, NJ, 1994) pp. 339–350.

of different chemical native-oxide thicknesses upon ion

yields. This effect has not been studied.

4 Referenced Standards

4.1 ASTM Standards

E 122 — Practice for Choice of Sample Size to

Estimate a Measure of Quality for a Lot or Process

E 673 — Terminology Relating to Surface Analysis

NOTICE: Unless otherwise indicated, all documents

cited shall be the latest published versions.

5 Terminology

5.1 All terms in this test method are in conformance

with those given in ASTM Terminology E 673.

6 Summary of Test Method

6.1 Specimens of mirror-polished single crystal silicon

are loaded into a sample holder. The holder is

transferred into the analysis chamber of the SIMS

instrument.

6.2 A primary ion beam, typically O

, is used to

bombard each specimen with a sputter rate less than

0.015 nm/s (0.9 nm/min).

6.3 The area of analysis may be different for different

instruments and may range from 100 µm × 100 µm to 1

mm × 1 mm.

6.4 Depending upon instrumentation, a molecular

oxygen jet or leak may be focused on the analysis area.

6.5 The positive secondary ions

Na,

Al,

K, and

Fe are mass analyzed by a mass spectrometer, and

detected by an electron multiplier (EM) or equivalent

high-sensitivity ion detector as a function of time until

the signals reach background levels or 1% of the initial

signal rates of each element. The instrumentation must

be able to discriminate the elemental ion signals from

molecular interferences.

6.6 A BLANK silicon sample is used to evaluate

whether the lower limit of detection arises from

molecular ion interferences, elemental instrumental

backgrounds, or count rate limitations. The matrix

positive secondary ion count rate for silicon (

Si,

Si) is measured by a faraday cup (FC) or

appropriate detector during, or at the end of, the profile.

If multiple detectors are used during the test, the

relative sensitivities of the detectors are determined by

measuring standard ion signals (either the same positive

3 Annual Book of ASTM Standards, Vol 14.02, ASTM International,

100 Barr Harbor Drive, West Conshohocken, PA 19428. Telephone:

610-832-9500, Fax: 610-832-9555, Website:

www.astm.org

4 Annual Book of ASTM Standards, Vol 03.06.

SEMI MF1617-0304 © SEMI 2003, 2004 3

secondary ion count rate or ion count rates of known

relative intensity such as natural

Si/

Si) on each

detector.

6.7 The net integrated

Na,

Al,

K, and

Fe signals

are converted to quantitative areal densities using the

detector efficiency ratios (if multiple detectors are used)

and RSFs measured from reference materials.

7 Apparatus

7.1 SIMS Instrument — Equipped with a primary ion

beam, preferably O

, a mass spectrometer with some

method of discriminating molecular ion interferences

from elemental ions of interest, an electron multiplier

detector, or faraday cup detector, or similar detector

system capable of measuring secondary ion count rates,

or any combination thereof, and the ability to sputter

the surface at less than 0.015 nm/s. An molecular

oxygen jet may be used to stabilize the surface ion

yield; the local oxygen gas flux to the specimen surface

must be stable enough that the secondary ion yield does

not vary during the analysis. The stability of the

oxygen jet effect can be checked by monitoring a

silicon matrix signal during a profile; and if the matrix

signal is not monitored during a profile, the vacuum

chamber pressure can be monitored for fluctuations in

the oxygen gas pressure. The SIMS instrument should

be adequately prepared and maintained so as to provide

the lowest possible instrumental backgrounds.

7.2 Test Specimen Holder — Reserved for the SIMS

measurement. In some instruments the holder can

support multiple 5 by 5-mm samples that are held face

down against metal (tantalum) windows. In other

instruments the holder can support one or more 15 by

15-mm samples by spring clips on the edge of the

specimen or silver paste on the sample back surface.

Some instruments can hold full silicon wafers. In all

cases, the sample holder should be adequately prepared

and maintained so as to provide the lowest possible

contribution to instrumental backgrounds.

7.3 Stylus Profilometer — Or equivalent device (for

example, atomic force microscope) to measure SIMS

crater depths. This is required to calibrate depth scales

for concentration profiles of calibration standard

samples. For routine depth scale calibration of test

specimens, the instrument must be capable of crater

depth measurements of 10 nm with an accuracy and

precision of 10%. In the absence of such an instrument,

a secondary depth calibration standard (such as a silicon

wafer with a known impurity depth profile) must be

prepared and can be used to measure the SIMS sputter

rate during the measurement.

NOTE 1: The accurate measurement of sputter crater depths

on the order of and below 10 nm is still an area of research,

because the sputtered silicon is very reactive to oxygen in the

air when removed from the SIMS chamber for crater depth

measurement. For example, sputter craters expected to be

only 1 nm in depth can actually be thicker than the original

native oxide due to this rapid oxide growth on sputtered

silicon, that is, one observes a bump on the surface where the

erosion was, rather than a crater. If a reference sample is

similar to the unknown in type, that is, intentional surface

contamination rather than shallow ion implant, it may be

possible to avoid the need for a direct shallow crater depth

measurement on each sample.

8 Sampling

8.1 Since this procedure is destructive in nature, a

sampling procedure must be used to evaluate the

characteristics of a group of silicon wafers. No general

sampling procedure is included as part of this test

method, because the most suitable sampling plan will

vary considerably depending upon individual

conditions. For referee purposes, a sampling plan shall

be agreed upon before conducting the test. See ASTM

Practice E 122 for suggested choices of sampling plans.

9 Specimen Requirements

9.1 Sample specimen surfaces must be flat and smooth

(see Sections 3.5 and 3.6).

9.2 Sample specimens must be cleaved or diced to fit

within the sample specimen holder. Specimens need

not be all the same size, but typical specimen sizes are

approximately square with one side of dimension 5 to

15 mm.

9.3 Prepare a BLANK specimen with a level of surface

metal below that expected to be detected by this test

method, and which has surface organics of a level and

type expected for the unknown specimens.

NOTE 2: If measurements are to be made on specimens

taken soon after a cleaning process, the organics on the

surface from clean room air and cassette carriers are very low

and negligible. However, if the specimens are stored for one

day or more, the level of organics may be significant for a

molecular ion interference in the measurement

; this

interference must be removed by the detection scheme.

10 Calibration

10.1 The calibration is by RSFs which are determined

empirically from reference materials.

10.1.1 Reference materials may be either:

10.1.1.1 Spin Coat Contaminated Samples —

Calibrated by Vapor Phase Decomposition/Atomic

5 Muller, A. J., Sinclair, J. D., Psota-Kelty, L. A., and Krauttler, H.

W., “Measurement of Airbourne Concentrations and Surface Arrival

Rates of Organic Contaminants in Clean Rooms,” 1993 Proceedings,

Vol 1 (Institute of Environmental Sciences, Mount Prospect, IL,

1993) pp. 373–379.

SEMI MF1617-0304 © SEMI 2003, 2004 4

Absorption Spectroscopy (VPD/AAS) or Vapor Phase

Decomposition/Inductively-Coupled-Plasma Mass

Spectrometry (VPD/ICP-MS) and shown by qualitative

SIMS to be spatially uniform, since spatial non-

uniformity can introduce variability or bias to the SIMS

quantitative measurement. With a spin coat

contaminated reference sample, it may be possible to

avoid the need for the crater depth measurement and its

bias/variability. For this reference sample, the areal

density D to be used in Equation 1 (see Section 12.1) is

the elemental areal density determined by the

VPD/AAs or VPD/ICP-MS, corrected for isotopic

abundance since the SIMS measurement of SI

isotopic.

10.1.1.2 Silicon Wafers — Dipped in an intentionally

contaminated SC-1 (NH

OH:H

O) bath where

the calibration of the metal deposition is by VPD/AAS

or VPD/ICP-MS and shown by qualitative SIMS, or by

TXRF for iron, to be spatially uniform, since spatial

non-uniformity can introduce variability or bias to the

SIMS quantitative measurement. With an SC-1 dipped

reference sample, it may be possible to avoid the need

for the crater depth measurement and its

bias/variability. For this reference sample the areal

density D to be used in Equation 1 (see Section 12.1) is

the elemental areal density determined by the

VPD/AAS or VPD/ICP-MS, corrected for isotopic

abundance since the SIMS measurement of Si

isotopic.

10.1.1.3 Ion Implants — Such as

Na,

Al,

K, and

Fe, (or

Fe) implanted into silicon and shown by

qualitative SIMS, or by TXRF for iron to be spatially

uniform, since spatial non-uniformity can introduce

variability or bias to the SIMS quantitative

measurement. It is necessary to demonstrate that RSFs

obtained from these reference materials and the

instrumentation and operation used for the

measurement are consistent with RSFs from spin coated

reference samples.

For this reference sample, the areal

density D to be used in Equation 1 (see Section 12.1) is

the implanted isotopic dose.

11 Procedure

11.1 Specimen Loading and SIMS Instrument Set-up

11.1.1 Cleave or dice individual specimens (unknowns,

reference, BLANK) to fit within the sample holder.

The reference sample must contain

Na,

Al,

K, and

Fe (or

Fe) or there must be multiple reference

samples, each with one or more of these elements. Do

this preparation in such a way as to minimize

contamination of the specimens with surface metals.

11.1.2 Load the specimen(s) into the SIMS sample

holder.

11.1.3 Transfer the loaded holder into the SIMS

instrument sample chamber.

11.1.4 Turn on the instrument in accordance with the

manufacturer' s instructions.

11.1.5 Set the appropriate analytical conditions

including the method for molecular ion mass

interference discrimination.

11.1.5.1 Select a primary ion current, primary beam

raster size, and secondary spectrometer transmission,

which will result in an appropriate sputter rate (less

than 0.015 nm/s).

11.1.5.2 Select the secondary spectrometer conditions

so that the maximum secondary ion-count rates result in

dead time losses below 10%.

11.1.6 Confirm that analytical conditions are

appropriate to the measurement goals by measurement

of known reference standard(s) and BLANK sample.

11.1.6.1 Confirm that the analytical sputter rate is such

that each element monitored in a single profile is

sampled at a frequency greater than or equal to once

every 0.2 nm of sputtered depth.

11.1.6.2 If an oxygen jet is used, confirm that 0-leak

pressure is suitable by measuring a matrix ion depth

profile and demonstrating that matrix secondary ion

yield is constant (within 20%) during the analysis of the

first 10 nm of a specimen. Perform this confirmation

measurement on a typical specimen at the same sputter

rate to be used for the measurement of the surface metal

impurities. If the ion yield shows significant variation,

increase the 0-leak pressure by a factor of 2× and

reconfirm stability.

11.1.6.3 Depending upon the instrumentation used and

if needed, measure the detector efficiency ratio(s) for

any detectors used during the measurement (for

example, the electron multiplier and the faraday cup

detectors). Make this measurement by intercomparing

standard secondary ion signals of the appropriate

intensity (to minimize dead-time losses). The

secondary ion count rates used here may be different

for this measurement than for the analysis, and may be

generated under different sputter rate conditions than

the analysis.

11.2 Analysis of Specimen

11.2.1 Center the primary ion beam, initialize the

SIMS instrument control software, and begin a SIMS

profile. For sodium, aluminum, potassium, and iron,

the secondary ion signals generated by sputtering these

6 Liu, Y. A., and Fleming, R. H., “Reduced Electron Multiplier Dead

Time in Ion Counting Mass Spectrometry,” Rev. Sci. Instr, 64, 1661

(1993).