semi合集-English.pdf - 第7351页

SEMI MF1391-0704 © 2004 3 5.3.5 FWHM — acronym for full width at half maximum , the width of the a bsorbance band at half its magnitude as measured from the baseline. 5.3.6 reference spectrum , n — the spectrum of the re…

3.3 Spectrophotometer technique is critical to a

successful carbon determination. The manufacturer's

instrument instruction manuals should be studied to

familiarize the operator with the proper use of the

spectrometer. Since the transmittance at the carbon

peak can be very low, while the transmittance at the

baseline regions is about 40%, extremely good

photometric linearity is critical. Wavenumber precision

is also critical because the carbon peak lies on the

shoulder of a very intense lattice absorption band.

3.4 The FWHM of the carbon absorption band at room

temperature must be less than 6 cm

−1

for acceptable

measurements. At cryogenic temperatures (below 80

K), it must be less than 3 cm

−1

for acceptable

measurements. Excessive width may be due to

improper thickness matching, to stress, or to the use of

a low resolution setting of the instrument. In dispersive

instruments, excessive widths can also result from

incorrect instrument balance setting or too fast a scan

speed. In Fourier transform instruments, excessive

widths can result from the use of a source-defining

aperture that is too wide.

3.5 Specimens that do not exceed the instrument beam

size cause error. Use of apertures at the sample, or

preferably beam condensers to reduce the beam size at

the sample, can correct this problem.

3.6 The main two-phonon lattice band of silicon, at

about 610 cm

−1

(16 µm), is very intense; the absorption

coefficient for this band is about 9 cm

−1

at room

temperature, and about 5 cm

−1

at 78 K.

This broad

band peak is close to the wavelength of the carbon-in-

silicon band and so presents a problem in measuring the

intensity of the carbon band.

3.7 Reference and test slices must be as close as

practically possible to the same temperature to avoid

the effects of temperature on the intensity of the lattice

band.

3.8 The minimum detection level of this test method is

limited by the signal-to-noise ratio of the spectrum.

Thus attaining the highest possible sensitivity by this

test method requires long measurement times and stable

spectrophotometers.

3.9 Free carrier absorption in silicon specimens with

resistivities less than 3 Ω·cm for p-type, or 1 Ω·cm for

n-type, reduces the available energy below the level

required for satisfactory operation of most

spectrophotometers.

3.10 For samples at cryogenic temperatures (below 80

K), plane parallel, polished surfaces may cause

interference fringes on the spectrum. Increasing the

2 Johnson, F. A., Proc. Physics Society 73, 265–272 (1959).

sample thickness or inducing a wedge (non-flatness) in

the sample reduces these interference fringes.

4 Referenced Standards

4.1 SEMI Standard

SEMI MF1241 — Terminology of Silicon Technology

4.2 ASTM Standard

E 131 — Terminology Relating to Molecular

Spectroscopy

NOTICE: Unless otherwise indicated, all documents

cited shall be the latest published versions.

5 Terminology

5.1 General definitions for terms related to infrared ab-

sorption spectroscopy are found in ASTM Terminology

E 131.

5.2 Definitions for terms related to silicon materials

technology are found in SEMI MF1241.

5.3 Definitions

5.3.1 background spectrum, n — in FT-IR instruments,

the single-beam spectrum obtained without a specimen

in the infrared light path that is usually obtained with

only nitrogen, dry air, or a vacuum in the beam.

5.3.2 baseline, n — a straight line interpolation

between points on either side of the carbon peak of the

absorbance spectrum, drawn to represent the spectrum

that would have been obtained in the absence of the

impurity.

5.3.3 baseline absorbance, n — the value of the

baseline at the wavenumber corresponding to the

carbon peak that is used for evaluating the absorbance

peak height.

5.3.4 Fourier transform infrared (FT-IR) spectrometer,

n — type of infrared spectrometer in which the data are

obtained as an interferogram.

5.3.4.1 Discussion — An interferogram is a record of

the modulated component of the interference signal

measured by the detector as a function of retardation in

the interferometer. This interferogram is then subjected

to a Fourier transformation to obtain an amplitude-

wavenumber (or wavelength) spectrum. FT-IR

instruments are always used in conjunction with a

computer to control the interferometer, collect and

manipulate the data, and for spectral output.

3 Annual Book of ASTM Standards, Vol 03.06, ASTM International,

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

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

www.astm.org

5.3.5 FWHM — acronym for full width at half

maximum, the width of the absorbance band at half its

magnitude as measured from the baseline.

5.3.6 reference spectrum, n — the spectrum of the

reference specimen.

5.3.6.1 Discussion — For true double-beam

instruments, it is obtained directly with the reference

specimen in the sample beam and the reference

(second) beam empty. For FT-IR and other single-

beam instruments, it is the result of ratioing the single-

beam spectrum of the reference specimen to the

background spectrum (see Section 5.3.1).

5.3.7 sample spectrum, n — the spectrum obtained

when the sample specimen is placed in the infrared

beam.

5.3.7.1 Discussion — For true double-beam

instruments, it is obtained directly with the sample in

the sample beam and the reference (second) beam

empty. For FT-IR and other single-beam instruments, it

is the result of ratioing the single-beam spectrum of the

sample to the background spectrum (see Section 5.3.1).

6 Summary of Test Method

6.1 At room temperature, test slices are prepared that

are polished on both sides to a nominal thickness of 2.0

mm. At cryogenic temperatures, the thickness of the

test slices can range from 2.0 to 4.0 mm.

6.2 A reference slice of known low carbon content (see

Section 3.2) is prepared in the same manner.

6.3 After verifying that the instrument is suitably set up,

transmittance spectra of the sample and reference are

obtained over the range from 700 to 500 cm

−1

(14.3 to

20.0 µm) on a double-beam dispersive or single-beam

Fourier Transform Infrared (FT-IR) spectrophotometer

in accordance with manufacturer's instructions.

6.4 Absorbance spectra are derived from these

transmittance spectra and a carbon-only spectrum is

obtained as the difference between the two absorbance

spectra.

6.5 A baseline is drawn between the regions on both

sides of the carbon peak on this difference absorbance

spectrum, and absorbance values of both the peak and

baseline are recorded.

6.6 The absorbance peak height is taken as the

difference between these two values. This peak height,

corrected for sample thickness, is multiplied by a

constant to calculate the substitutional carbon concen-

tration. Two constants are used, one for measurements

at room temperature (300 K), and one for measure-

ments at cryogenic temperatures (below 80 K).

7 Apparatus

7.1 Infrared Spectrophotometer — Either a dispersive

(computerized or non-computerized) or a Fourier trans-

form (FT-IR) spectrophotometer may be used. The

resolution of the spectrophotometer must be at least 2

−1

at room temperature, or 1 cm

−1

at cryogenic

temperatures, over the range from 500 to 700 cm

−1

for

either dispersive or Fourier transform infrared spectro-

photometers. The total operating range of the

spectrophotometer shall include the range from 500 to

2000 cm

−1

7.2 Micrometer Caliper — or other instrument, capable

of measuring the thickness of the specimens to an

accuracy of 0.005 mm.

7.3 Equipment and Materials for Slicing and Polishing

Silicon — to a final thickness tolerance of 0.005 mm or

less, and a total thickness variation of 0.01 mm or less.

7.4 Thermometer — or other instrument capable of

measuring the temperature of the sample chamber (for

room temperature measurements) or the sample holder

(for cryogenic temperature [below 80 K]

measurements) to within ± 2°C.

7.5 Calcium Fluoride Crystal (CaF

), cut to nominal

thickness of 5 mm.

8 Sampling

8.1 Unless otherwise specified, a silicon slice used for

the carbon test is to be measured at the nominal slice

center.

9 Test and Reference Specimens

9.1 A single crystal slice of about 2-mm thickness must

be used at room temperature, and a slice of 2.4 to 3.5-

mm thickness is preferable at cryogenic temperatures.

9.2 Both the test and reference specimens must be

carefully shaped to the following criteria:

9.2.1 Thickness variation over the measurement area

shall be 0.005 mm or less.

9.2.2 Surface preparation shall be identical,

9.2.3 Thickness equality shall meet one of the

following:

9.2.3.1 When the measurement is made with double

beam simple dispersive infrared spectrophotometers,

the final thickness of test and reference slices shall be

equal to within ± 0.01 mm (see Section 10.4).

9.2.3.2 When the measurement is made with computer

assisted double-beam dispersive or single-beam FT-IR

spectrophotometers, the final thickness of test and

reference slices shall be equal to within ± 0.5 mm (see

Section 10.5).

9.2.4 Surface area large enough such that with respect

to the holders no incident radiation can bypass either

the test or reference specimen.

9.3 The reference specimen must be selected from float

zone silicon with a minimal substitutional carbon

concentration and a carrier concentration such that there

is no measurable free carrier absorption of infrared

radiation in the range from 500 to 2000 cm

−1

NOTE 1: A satisfactory method of selecting usable reference

specimens is to prepare polished slices of equal thickness

from many different low-carbon silicon crystals produced by

the float zone method, and then compare them to each other in

an infrared radiation (IR) spectrophotometer. The selection

must be carried out at cryogenic temperatures for cryogenic

measurements. The specimen(s) showing the highest relative

transmittance at 605 cm

−1

at room temperature, or at 607.5

−1

at cryogenic temperatures (below 80 K), can be used as

reference specimen(s).

9.4 Both test and reference specimens must have

resistivity greater than 3 Ω·cm for p-type silicon and

greater than 1 Ω·cm for n-type silicon.

10 Procedure

10.1 Instrumental Checks

10.1.1 With reference to the transmittance spectrum,

establish the 100% transmittance line to measure the

stability and the noise level by one of the following

methods.

10.1.1.1 For double-beam instruments, record the

transmittance spectrum with both the sample and

reference beams empty.

10.1.1.2 For single-beam instruments, record the trans-

mittance spectrum as the ratio of two spectra taken with

the sample beam empty. Take the two spectra at times

separated by at least the time required to load samples

and take two sequential spectra.

10.1.1.3 In both cases, plot the transmittance spectrum

over the wavenumber range covering 500 to 700 cm

−1

to obtain the 100% line.

10.1.1.4 If the 100% transmittance line is 100 ± 0.5%

over the entire range, continue to the next step.

Otherwise, adjust or repair the instrument so as to meet

this criterion.

10.1.2 Establish the 0% transmittance line, (T).

10.1.2.1 For double-beam dispersive (DIR)

instruments, record the transmittance spectrum between

700 and 500 cm

−1

with the sample beam blocked.

10.1.2.2 For FT-IR instruments, collect a background

spectrum with the beam empty. Next, collect a

spectrum with a CaF

crystal wafer, 5-mm thick, in the

beam. Obtain the spectrum that is the ratio of the CaF

spectrum to the background spectrum, and plot the

percent transmittance spectrum between 700 and 500

−1

10.1.2.3 In either of these cases, the 0% transmittance

must not exceed ± 0.5% over the range between 700

and 500 cm

−1

10.1.2.4 Alternatively, for FT-IR instruments, plot the

single-beam background spectrum (empty beam only)

from 1000 to 200 cm

−1

. The recorded energy in a

region where the detector is known to give a zero

response (below 300 cm

−1

for most detectors) must not

be greater than 1.0% of the maximum signal in this

region.

10.1.3 To determine the mid-scale linearity of the

instrument obtain a spectrum of the silicon reference

specimen over the wavenumber range from 1600 to

2000 cm

−1

. If the value of the transmittance is not 53.8

± 2% over this wavenumber range, align the sample at a

small angle to the axis of the incoming IR beam, in

order to minimize undesirable reflections between the

silicon surfaces and the spectrometer components.

NOTE 2: This angle may be determined by initially placing

the silicon slice normal to the axis of the incoming beam, and

then gradually tilting the specimen while repeatedly obtaining

the transmittance spectrum above 1600 cm

−1

. The optimum

angle is typically less than 10°.

10.2 Measure and record the thicknesses of the test and

reference specimens to within ± 0.005 mm, at their

centers.

10.3 For room temperature measurements, measure and

record the temperature of the sample chamber to ± 2°C.

At cryogenic temperatures (below 80 K), measure and

record the temperature of the sample holder.

10.4 Obtain the carbon-only absorbance spectrum by

one of the following methods (see Figure 1):

4 Chase, D. B., Applied Spectroscopy 38(4), 491–494 (1984).

5 Hoffman, P., and Knozinger, E., Applied Spectroscopy 41(8),

1303–1306, (1987).