semi合集-English.pdf - 第7333页

SEMI MF1389-0704 © 2004 1 SEMI MF1389-0704 TEST METHODS FOR PH OTOLUMINESCENCE AN ALYSI S OF SINGLE CRYSTAL SILICON FOR III-V IMPURITIES This guide was technically approved b y the Global Silicon Wafer Com mittee and is …

SEMI MF1366-0305 © SEMI 2003, 2005 7

APPENDIX 1

ANALYSIS OF DATA FROM THE MULTI-INSTRUMENT EXPERIMENT

NOTICE: The material in this appendix is an official part of SEMI MF1366. Approval was by full letter ballot

procedures with publication authorized by the NA Regional Standards Committee on December 10, 2004.

A1-1 The precision estimate was taken from data generated in one laboratory using three instruments and ten

instrument operators. All the instruments were CAMECA IMS 3f or 4f SIMS instruments. The test specimens were

all taken from one silicon wafer that was lightly doped and verified by infrared absorption spectroscopy to have

uniform interstitial oxygen levels across the central region where the test specimens were taken. The test specimens

were chemically mechanically polished on one side. The standards had oxygen levels that bracketed the expected

level of oxygen of the test specimen.

A1-2 The measurements were made in 191 loads over a two-month period.

A1-3 Both the load line calibration and the average load factor calibration methods were used to convert the SIMS

data to infrared absorption equivalent oxygen. The load line calibration method gave an average oxygen level of

18.64 ppma (IOC-88), a one standard deviation of 0.39 ppma, and a relative standard deviation of 2.09%. A

frequency distribution is shown in Figure A1-1. The average load factor calibration method gave an average oxygen

level of 18.61 ppma (IOC-88), a one standard deviation of 0.365 ppma, and a relative standard deviation of 1.97%.

A frequency distribution is shown in Figure A1-2.

NOTE 1: The oxygen concentration scale in Figures A1-1 and A1-2 is given in Old ASTM units. These must be multiplied by

0.652 to get the values of oxygen concentration in IOC-88 units. This difference in scale arises from an historical basis and does

not affect the conclusions reached. Note also that the oxygen values in §A1-3 have been adjusted to obtain the results in IOC-88

units.

Figure A1-1

Frequency Distribution of SIMS Measured Oxygen

Using the Load Line Calibration Method

Figure A1-2

Frequency Distribution of SIMS Measured Oxygen

Using the Average Load Factor Calibration Method

NOTICE: SEMI makes no warranties or representations as to the suitability of the standards set forth herein for any

particular application. The determination of the suitability of the standard is solely the responsibility of the user.

Users are cautioned to refer to manufacturer's instructions, product labels, product data sheets, and other relevant

literature, respecting any materials or equipment mentioned herein. These standards are subject to change without

notice.

By publication of this standard, Semiconductor Equipment and Materials International (SEMI) takes no position

respecting the validity of any patent rights or copyrights asserted in connection with any items mentioned in this

standard. Users of this standard are expressly advised that determination of any such patent rights or copyrights, and

the risk of infringement of such rights are entirely their own responsibility.

International), 3081 Zanker Road, San Jose, CA 95134. Reproduction of

the contents in whole or in part is forbidden without express written

consent of SEMI.

SEMI MF1389-0704

TEST METHODS FOR PHOTOLUMINESCENCE ANALYSIS OF SINGLE

CRYSTAL SILICON FOR III-V IMPURITIES

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 for publication by the North

American Regional Standards Committee on March 14, 2004. Initially available at

www.semi.org

May

2004; to be published July 2004. Original edition published by ASTM International as ASTM F 1389-92.

Last previous edition SEMI MF1389-00.

1 Purpose

1.1 Electronic-grade polycrystalline silicon producers

and users require information regarding impurities for

quality assurance as well as for research and

development purposes. Polysilicon is float-zoned and a

sample from the zoned rod is analyzed following these

test methods to obtain impurity densities that can be

related to the impurity content of the starting material

(see SEMI MF1723).

1.2 Photoluminescence analysis identifies and

quantifies the electrically active dopant impurities in

monocrystalline silicon. These test methods address

boron, phosphorus, arsenic, and aluminum, found as

impurities in electronic grade silicon.

1.3 These test methods can be applied to doped and

undoped float-zoned or Czochralski material.

2 Scope

2.1 These test methods cover the simultaneous

determination of electrically active boron, phosphorus,

arsenic, and aluminum content in low-dislocation

monocrystalline silicon.

NOTE 1: These chemical species can also be determined by

the low temperature infrared analysis procedure of SEMI

MF1630.

2.2 These test methods can be used for samples that

have dopant densities between approximately 1 × 10

and approximately 5 × 10

atoms/cm

2.3 The concentrations obtained using these test

methods are based on an empirically determined

relationship of the logarithm of the concentration to the

logarithm of specific luminescence line-intensity ratios.

2.4 The empirical relationship established assumes a

constant sample excitation level for all measurements

on a given instrument.

2.5 To accommodate differences in instrumentation,

two methods are included. Test Method A refers to

procedures appropriate for dispersive infrared

spectrophotometers operating under the high sample

excitation conditions and Test Method B refers to

procedures appropriate for Fourier transform

instruments operating under low excitation conditions.

2.5.1 Typical calibration curves for each test method

are provided. These curves are modified for each

instrument using the analysis of standard samples as

reference data. Once modified, the curves for a given

instrument should produce sample dopant density

values that agree with other similarly operated

instruments using the same test method. Data obtained

using Test Method A may not agree with data obtained

using Test Method B, hence values must be reported

with reference to the test method used.

NOTE 2: Several different methods of photoluminescence

analysis are currently in practice worldwide. These test

methods address two of these, one (Test Method A) in use

primarily in Japan

and the other (Test Method B) primarily in

the United States. Recently published works

2,3

describe other

approaches.

2.6 Many laboratories use photoluminescence to

analyze epitaxial layers. However this application

encounters many variables and the underlying physics

is not fully understood; hence these test methods do not

attempt to outline standard practices regarding such

analysis.

NOTICE: This standard does not purport to address

safety issues, if any, associated with its use. It is the

responsibility of the user of this standard to establish

appropriate safety and health guides and determine the

applicability of regulatory or other limitations prior to

use.

1 Test Method A is essentially equivalent to JIS H 0615, Test

Method for Determination of Impurity Concentrations in Silicon

Crystal by Photoluminescence Spectroscopy. This standard is

available from Japanese Standard Association 1-24, Akasaka 4

Chome, Minato-ku, Tokyo 107-0000, Japan. Telephone: 81-(0)3-

3583-8005; Fax: 81-(0)3-3586-2014; Website:

www.jsa.or.jp

2 Colley, P. McL., and Lightowlers, E. C., “Calibration of the

Photoluminescence Technique for Measuring B, P, and Al Concen-

trations in Silicon in the Range 1e12 to 1e15 at/cm 3 Using Fourier

Transform Spectroscopy,” Semiconductor Science and Technology 2,

157–166 (1987).

3 Schumacher, K. L., and Whitney, R. L., J. Electron. Materials

18(6), 681–687 (1989).

3 Limitations

3.1 Variations in Excitation Intensity — The extrinsic

bound exciton (BE) and intrinsic free exciton (FE)

luminescence features do not vary at the same rate with

excitation intensity.

The FE features increase

proportionally with excitation intensity, while the BE

features increase proportionally at low excitation levels

but become more slowly increasing at higher excitation

levels generally above the electron hole droplet (EHD)

onset point. Since the calculated concentrations are

derived from the ratio of extrinsic features to the

intrinsic feature, the ratio will be decreasing as the

excitation intensity is increasing. Thus, if a sample is

measured at a higher excitation level than the

instrument is calibrated for, the calculated

concentration will be artificially low, and vice versa.

3.2 Sample surface damage, bulk defects, or other

lifetime reducers also affect the line-intensity ratios and

overall luminescence intensity owing to their tendency

to reduce the steady state population of excitons, thus

mimicking the effects of a lower excitation intensity.

Other mechanisms related to the presence of these

defects may also introduce decalibrating effects. Some

luminescence features are generated by defects; for

example, those at 6510 and 7050 cm

−1

are typical of

thermally stressed samples. Such features can provide

qualitative information about the presence of defects.

5,6

3.3 The ratios and line widths of the silicon

luminescence features vary strongly with temperature,

hence variations in sample temperature must be

avoided. Corrections for the effects of temperature

variations are included in these test methods (see

Section 7.1).

3.4 Overlapping Spectral Features

3.4.1 The boron B

(BE) feature at 8812.6 cm

−1

overlaps the P

(b′

) feature at 8812.7 cm

−1

causing a

direct error when calculating boron concentration. The

beta-series P

(b′

) line is approximately one-tenth the

intensity of the alpha series P

(BE) line at 8806.6

−1

, so a subtraction based on the amount of

phosphorus present can be made in the boron feature

measurement. An alternative approach is to use the

) boron line, which is free from a coincident

4 Nakayama, H., Nishino, T., and Hamakawa, Y.,“Analysis of the

Excitation Luminescence of Silicon for Characterization of the

Content of Impurities,” Jap. J. Appl. Phys. 19(3), 501–511, (1980).

5 Drozdov, N. A., Patrin, A. A., and Tkachev, V. D.,

“Recombination Radiation on Dislocations in Silicon,” Soviet

Physics, J.Experimental and Theoretical Physics 23(11), 597 1976).

6 Suezawa, M., Sasaki, Y., Nishina, Y., and Sumino, K., “Radiative

Recombination on Dislocations in Silicon Crystals,” Jap. J. Appl.

Phys. 20(7), L537 (1981).

phosphorus line, provided the instrument has been

calibrated for this feature.

3.4.2 The antimony Sb

(BE) line falls between the

boron and phosphorus transverse optical (TO) features.

Since the line widths are broad relative to the line

positions of these features, the presence of antimony

affects the apparent intensities of boron and phosphorus

TO features. The no-phonon, (NP) features for these

elements could be used providing the instrument is

calibrated for them.

3.4.3 In general, the broad TO phonon region features

can interfere, particularly when a sample contains

widely differing levels of impurities. For example,

when boron greatly exceeds phosphorus, the P

(BE)

line must be used, and when phosphorus greatly

exceeds boron, one of the approaches described in

Section 3.4.1 must be used.

3.5 Stress in samples can cause NP region feature

splitting and thus affect the established TO/NP line

ratio for a given impurity. The concentrations

calculated from such spectra would be artificially low.

Peak area calculation methods reduce the

measurement' s sensitivity to line splitting effects.

3.6 The calibration curves presented in these test

methods are traceable to resistivity measurements (see

SEMI MF723), neutron transmutation doped, (NTD)

reference material, and low-temperature infrared

absorption spectroscopy (FT-IR) measurements (as a

secondary standard) (see SEMI MF1630). Hence

inaccuracies in the work that produced the calibration

curves impact the accuracy of the PL results.

4 Referenced Standards

4.1 SEMI Standards

SEMI C28 — Specifications and Guidelines for

Hydrofluoric Acid

SEMI C30 — Specifications and Guidelines for

Hydrogen Peroxide

SEMI C35 — Specifications and Guideline for Nitric

Acid

SEMI MF723 — Practice for Conversion Between

Resistivity and Dopant Density for Boron-Doped,

Phosphorus-Doped, and Arsenic-Doped Silicon

SEMI MF1630 — Test Method for Low Temperature

FT-IR Analysis of Single Crystal Silicon for III-V

Impurities

SEMI MF1723 — Practice for Evaluation of

Polycrystalline Silicon Rods by Float-Zone Crystal

Growth and Spectroscopy