semi合集-English.pdf - 第7473页

SEMI MF1618-1104 © SEMI 2004 3 may understat e macroscopi c film non-uni formity or fai l to respond to microscopic film non-uniformities or both. 3.5.2 If the sampling area or spot size is small, it is generally possibl…

SEMI MF1618-1104 © SEMI 2004 2

chemical vapor deposition (CVD) oxidation, and

metallization, as well as for layer modification such as

various means of layer etching.

2.4 This practice is intended for use with all silicon

wafer sizes and types when measuring uniformity of

film properties and characteristics. This practice

describes measurement site patterns and determination

of their spatial coordinates on the wafer, as well as the

statistics to be used when reducing the measurement

data to determine uniformity. For each of the sampling

plans, the exact number of measurement sites is chosen

based on the size of the wafer being used, the desired

spatial resolution of the measurement instrument, and

whether maximal, or somewhat lesser information

density is desired. However, in all such choices, the

pattern of measurement sites, the rules for selecting

their coordinates on the wafer, and the statistical

calculations of the results should remain consistent with

the procedures of this practice.

2.5 This practice can be used with any measurement

method, procedure or instrumentation that can measure

the needed film property or characteristic with

sufficient precision and spatial resolution to reveal the

needed information on spatial non-uniformity of the

film. This practice does not itself contain details on

performing any specific measurement.

2.5.1 Not all types of measurements that may need to

be used for evaluation of the uniformity of a thin film

have formal procedural standards. SEMI MF374,

SEMI MF576, SEMI MF1392, SEMI MF1393, and

SEMI MF1529 give details of measurement procedures

that may be applied to evaluating the uniformity of thin

film properties.

2.5.2 This practice does not deal with acquisition or

analysis of uniformity data where it is desired to take

more than one measurement per specified spatial cell

such as is commonly done for wafer site flatness

measurements.

2.6 This practice makes no recommendations regarding

the interpretation of the statistics that result from

analysis of the data acquired with regard to the

goodness or badness of given values of the test statistic,

nor does it make recommendations regarding decisions

about the process cycle or equipment used to produce

the thin film that was measured.

2.7 The principles of this practice may be adapted to

determine the uniformity of bulk silicon wafer

properties such as interstitial oxygen content and

resistivity, but depending on the desired property and

the chosen measurement technique, depth-dependent

variations may be misinterpreted as lateral variations.

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 practices and determine

the applicability of regulatory or other limitations prior

to use.

3 Limitations

3.1 This practice is written for evaluation of planar or

blanket films, but it may be applied to patterned films if

the pattern size, shape, and distribution do not interfere

with the spatial resolution of the selected measurement

technique and the specified measurement site selection.

If either of these interferences occur, the user may adapt

the principles of the method to the needed application,

but the interpretation of the results may change.

3.2 The principles of this practice may be adapted to

other semiconductor wafers, such as gallium arsenide,

but particular concerns with those other materials may

not be addressed adequately in this practice.

3.3 Uniformity measurements from certain types of

measurements may be misinterpreted if the user is not

aware of the full nature of the basic measurement being

made. For example, sheet resistance is a function both

of the inherent resistivity of the film being measured

and of its thickness value. Non-uniformity in sheet

resistance values across the wafer may result from non-

uniformity in resistivity (layer composition or structure)

or layer thickness (deposition rate) alone, or it may

result from simultaneous variations of both parameters.

3.4 Changes in test equipment performance, or changes

in test procedure or conditions over time may

detrimentally affect the ability to compare test results

from this practice over time for a given film property or

fabrication step. It is the responsibility of the user to

ensure that the measurement system and process

remains in sufficient control to allow time-wise

comparison of uniformity results, if such comparisons

are needed.

3.5 Sampling area or sampling spot-size may cause

misinterpretation of the cause of thin film non-

uniformity if the magnitude of the non-uniformity and

its spatial scale, or rate of variation is not well matched

to the sampling spot size and the selected distribution of

measurement sites. While this should not be a serious

concern for most modern film deposition processes, and

process equipments, which are operating in control, the

principles of this limitation are worth elaborating.

3.5.1 If the sampling area or the spot size of the test

instrument is large compared to rates of film parameter

change (gradients) that are important to identify, it must

be recognized that such large sampling area instruments

generate some form of spatial average response that

SEMI MF1618-1104 © SEMI 2004 3

may understate macroscopic film non-uniformity or fail

to respond to microscopic film non-uniformities or

both.

3.5.2 If the sampling area or spot size is small, it is

generally possible to quantify the full scale of

macroscopic film non-uniformities. However, should

spatial fine-scale systematic film variations be present,

they may affect the individual measurements in an

inconsistent manner unless the measurement sampling

plan is tailored to the spatial size scale of the fine

variations. Such tailoring of the sampling plan is

generally incompatible with the requirements of this

practice.

3.6 For the most meaningful interpretation of

uniformity data, all sampling sites should represent

equal wafer areas, and the available area of the wafer

should be fully sampled. The existence of wafer flats

causes a failure of the first requirement with all

concentric circle sampling plans. The need of the

semiconductor industry to establish as large a quality

area as possible on wafers also causes a failure of the

first requirement for the outermost circle of concentric

circle sampling plans because of the need to measure

right up to a very small edge exclusion value. The

second requirement is not met near the perimeter of

Cartesian sampling plans.

3.7 Ideally, measurements would be made at the

specified sites in a random sequence so that instrument

drift or changes in other environmental parameters that

may affect measurement results would have a random

correlation with measurement positions. It is normally

not practical to acquire data in this fashion. Therefore,

any instrumental or environmental change that may

occur will have a correlation with the spatial ordering

of the measurement sequence. While such changes are

expected to have about the same effect on both

concentric circle and Cartesian sampling plan

measurements, providing the number of data points are

about the same, they would nevertheless have different

effects on two-dimensional contour maps that resulted

from these two different types of sampling plans.

3.8 Individual test instruments and test methods that

may be used to measure properties or characteristics of

thin films are subject to various interferences that may

affect measurement precision consistency, resolution,

and accuracy. These may include such things as

sensitivity to stray illumination, rf-fields, and variations

in temperature. It is the responsibility of the user of this

practice to consult instrument manuals and appropriate

test methods in order to identify and control potential

interferences with that instrument or method.

3.8.1 It is also the responsibility of the user to devise a

test for the inherent precision of the measurement being

made so that imprecision in the measurement is not

misinterpreted as non-uniformity of the thin film being

evaluated by this practice. This is particularly

important for application of this practice to evaluating

films from process steps that are capable of a very high

degree of uniformity, for example, 1% across a wafer.

Such tests may involve using known uniform or non-

uniform test specimens whose spatial pattern of non-

uniformity has been found to be very stable over

extended periods of time.

4 Referenced Standards

4.1 SEMI Standards

SEMI M1 — Specifications for Polished

Monocrystalline Silicon Wafers

SEMI M20 — Practice for Establishing a Wafer

Coordinate System

SEMI MF81 — Test Method for Measuring Radial

Resistivity Variation on Silicon Wafers

SEMI MF374 — Test Method for Sheet Resistance of

Silicon Epitaxial, Diffused, Polysilicon, and Ion-

Implanted Layers Using an In-Line Four-Point Probe

with the Single-Configuration Procedure

SEMI MF576 — Test Method for Measurement of

Insulator Thickness and Refractive Index on Silicon

Substrates by Ellipsometry

SEMI MF673 — Test Methods for Measuring

Resistivity of Semiconductor Slices or Sheet Resistance

of Semiconductor Films with a Non-contact Eddy-

Current Gage

SEMI MF1241 — Terminology of Silicon Technology

SEMI MF1392 — Test Method for Determining Net

Carrier Density Profiles in Silicon Wafers by

Capacitance-Voltage Measurements with a Mercury

Probe

SEMI MF1393 — Test Method for Determining Net

Carrier Density in Silicon Wafers by Miller Feedback

Profiler Measurements with a Mercury Probe

SEMI MF1529 — Test Method for Sheet Resistance

Uniformity Evaluation by In-line Four-point Probe with

the Dual-configuration Procedure

NOTICE:

Unless otherwise indicated, all documents

cited shall be the latest published versions.

5 Terminology

5.1 Definitions of terms used in this practice may be

found in SEMI M1 and SEMI MF1241.

SEMI MF1618-1104 © SEMI 2004 4

6 Summary of Practice

6.1 Measurements are made at the sites specified in the

chosen sampling plan using the appropriate

instrumentation and measurement procedure for the

film parameter of interest. Measures of the dispersion

of the values are obtained by simple statistics specified

for the sampling plans.

7 Apparatus

7.1 Four-probe Sheet Resistance, Eddy-current,

Mercury Probe CV, Spectroscopic Reflectometers,

Ellipsometers, Interferometers, and Stylus

Profilometers — Typical apparatus chosen according to

the type of material parameter being monitored. Others

may be chosen according to need. The instrumentation

should include such features as temperature monitoring

and control, EMI shielding, shielding from

illumination, etc. as is necessary for good practice of

that type of measurement. In addition, all

instrumentation must include the following:

7.1.1 Ability to center the wafer on the measurement

stage to an accuracy of 0.5 mm or better,

7.1.2 Vacuum chuck, or similar fixture, to secure the

wafer to the instrument stage during measurement,

7.1.3 Ability to locate the measurement points

specified in the chosen sampling plan to an accuracy of

0.5 mm or better, and

7.1.4 Ability to store all measurement data to perform

the required calculation of uniformity.

7.2 Capability of the instrumentation to perform one or

more of the following auxiliary data presentations is

highly recommended:

7.2.1 Generating a two-dimensional contour map of the

acquired data,

7.2.2 Generating a histogram of the raw data values,

and

7.2.3 Presenting the raw data as a time-series plot.

7.3 Capability to edit data and to blend or interpolate

data in conjunction with two-dimensional mapping may

be useful, but it should be done with caution.

8 Procedure

8.1 Calibrate the measurement system in accordance

with the manufacturer’s instructions or with the

applicable test method.

8.2 Place the wafer on the instrument stage so as to

locate the wafer center within 0.5 mm of the center of

the stage. For the purposes of this procedure, the wafer

center is considered to be defined by the intersection of

any two diameters that do not intersect a wafer flat or

orientation notch.

8.3 Select the sampling plan to be used from those

listed in Appendix 1 and proceed to take measurements

at all the specified locations. It may be necessary to

enter the coordinates of the desired sampling plan into

instrument software, or to work with the manufacturer

of the equipment to modify its software if the sampling

plans available in the instrument do not match the

requirements of Appendix 1.

8.4 Take measurements at all locations specified for

the selected sampling plan. Then quantify the

uniformity from the acquired data using the calculations

specified in Section 9.

9 Calculations

9.1

Calculate the mean,

x , and the standard deviation,

s, of the N measurement values, x

, as follows:





, and (1)

2/1

)(





















s (2)

9.2 Examine the data. If agreed upon between parties

to the test, reject values that are away from the mean by

more than three times the standard deviation calculated

in Section 9.1. Recalculate the mean and standard

deviation using Equations 1 and 2, where N is now the

reduced number of data points.

NOTE 1: It is generally advisable, through use of a time

series plot of the original data, or similar means, to inspect the

data just rejected for determination of measurement site

location and relation to adjacent measurement values.

NOTE 2: The calculation for standard deviation may be

made regardless of any systematic behavior of the underlying

distribution of measured values. However, when there is a

systematic spatial behavior to the measured values, the

standard deviation may not be used to estimate confidence

intervals or tolerance limits.

9.3 Calculate the following two statistics from the

complete data set or from the reduced data set if certain

measurements were rejected in Section 9.2. These two

statistics represent the range and the distribution of data

about the mean. The use of only one of these statistics

is at the discretion of the interested parties.

9.3.1 High-Low Variation, HLV, in percent:

200(%)

minmax









HLV

(3)