semi合集-English.pdf - 第7423页

SEMI MF1529-1104 © SEMI 2004 4 SEMI MF81 — Test M ethod for M easuring Radi al Resistivity Variation on Silicon Wafers SEMI MF84 — Test Method for Measuring Resistivity of Silicon Wafers with an In-line Four-Po int Probe…

SEMI MF1529-1104 © SEMI 2004 3

value should result in a change of sheet resistance that

is less than 0.5%.

Table 1 Nominal Current Values for Measurement of

Sheet Resistance

The current used should be from one-half to twice the nominal

value and should be chosen to give a measured voltage on the

specimen that is between 7 and 15 mV when using Configuration A.

Once the current is selected for forward direction measurements at a

given site, it must be kept constant to 0.01% for the remaining

measurements at that site.

3.4 Semiconductors have a significant temperature

coefficient of resistivity. Consequently, the

measurement current used should be small to avoid

resistive heating. The current levels recommended

should reduce the chances of this problem. If resistive

heating is suspected, it can be detected by a change in

readings starting immediately after the current is

applied. If such a change is observed, repeat the

electrical measurements at a lower current. In the

absence of Joule heating, the temperature of the water

should be uniform if the wafer is mounted on a chuck

having good thermal conductivity and large thermal

mass. Sheet resistance maps should not be distorted by

temperature non-uniformities in this case. Coefficients

for the temperature variation of the sheet resistance of a

particular layer type will depend upon the specific

dopant, or resistivity, profile of that layer type, and

must be evaluated empirically for the layer fabrication

process being used if correction of data to a fixed

reference temperature is desired.

3.5 Vibration of the probe may cause variations in

contact resistance, which is often manifested as

unstable readings. If difficulty is encountered, the

apparatus should be vibration isolated.

3.6 Penetration of either the current or voltage probes

through the layer being measured to the substrate

results in erroneous readings. This can usually be

checked by mounting the specimen directly on a metal

support that is grounded to the current supply and by

then looking for a reduction in measured specimen

voltage in at least one polarity as the ground connection

is removed and replaced. If this condition occurs,

examine the probe tips microscopically for sharp

asperities and remove these by polishing or otherwise

conditioning the probe tip, or else reduce the probe

force or use a probe with blunter probe tips.

3.7 Use of the data from the two electrical

configurations to calculate a factor for water diameter

and for position of the measurement site on the wafer is

accurate to within 0.1% as long as the measurement site

is away from the perimeter of the wafer. To meet this

requirement the site should be at least five-probe

separations from the perimeter for the case of probe

alignment perpendicular to a wafer diameter, and at

least three-probe separations from the perimeter for the

case of probe alignment parallel to a wafer diameter.

For certain processes, such as ion implantation, use of a

wafer clamp during layer formation causes a p-n layer-

to-substrate junction on the top surface of the wafer

interior to the mechanical edge of the wafer. In these

cases, the probe separation values above refer to the

location of the measurement site with respect to such

junctions.

3.8 In shallow or lightly-doped layers, an effect known

as carrier redistribution causes the number of free

carriers in the layer to be different from the number of

dopant atoms. As a result, sheet resistance values

measured by this test method may be noticeably

different from values calculated from models that imply

the dopant and free-carrier depth profiles to be

equivalent.

3.9 Surface and near-surface effects, such as the

formation of hydrogen complexes with acceptors, may

occur during or immediately after the fabrication of

many thin films, particularly if lightly doped. They

may also occur slowly with storage. These effects may

be uniformly or non-uniformly distributed across the

wafer surface. The result is to modify, generally by

way of increasing, the measured sheet resistance. Two

of the most prominent impacts of these effects are to

make the results of a given process step appear to be

more non-uniform than is actually the case, and to shift

the absolute level of a reference wafer used to monitor

the performance of the mapping tool so as to make the

tool appear to be out of control.

4 Referenced Standards

4.1 SEMI Standards

SEMI C19 — Specification for Acetone

SEMI C23 — Specifications for Buffered Oxide

Etchants

SEMI C41 — Specifications and Guidelines for 2-

Propanol

SEMI M1 — Specifications for Polished

Monocrystalline Silicon Wafers

SEMI MF42 — Test Methods for Conductivity Type of

Extrinsic Semiconducting Materials

Sheet Resistance,



Current

2–25 10 mA

20–250 1 mA

200–2,500

100 A

2,000–25,000

10 A

SEMI MF1529-1104 © SEMI 2004 4

SEMI MF81 — Test Method for Measuring Radial

Resistivity Variation on Silicon Wafers

SEMI MF84 — Test Method for Measuring Resistivity

of Silicon Wafers with an In-line Four-Point Probe

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 MF1241 — Terminology of Silicon Technology

SEMI MF1618 — Practice for Determination of

Uniformity of Thin Films on Silicon Wafers

4.2 ASTM Standard

D 5127 — Guide for Ultra Pure Water Used in the

Electronics and Semiconductor Industry

NOTICE: Unless otherwise indicated, all documents

cited shall be the latest published versions.

5 Terminology

5.1 For definitions of terms used in silicon wafer

technology refer to SEMI M1 and SEMI MF1241.

6 Summary of Test Method

6.1 An in-line four-point probe is used to determine the

specimen sheet resistance at each desired measurement

location.

6.1.1 The number and positioning of measurement

locations is determined by end-use needs, or by the

parties to the test in the case of referee measurements.

6.1.2 At each location, a direct current is passed into the

specimen, using two of the probes, as specified, and the

potential difference is measured using the other two

probes.

6.1.3 Current polarity is reversed and the potential

difference is re-measured to allow elimination of

thermoelectric effects.

6.1.4 Before the probe is raised, the process is repeated

using a different combination of probes, as specified.

At each location, the sheet resistance is obtained from

the four ratios of potential difference to current.

6.2 The adequacy of the probe is determined both by

optical examination of probe indentations made in a

polished silicon surface, and by a performance test on a

wafer of the type whose uniformity is to be checked.

6.3 The accuracy of the electronics is tested by means

of an analog circuit emphasizing the performance and

6 Annual Book of ASTM Standards, Vol 11.01, ASTM International,

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

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

www.astm.org

noise immunity of the electronics in the presence of

large contact resistances of the probe tips to the

semiconductor surface.

7 Apparatus

7.1 Specimen Preparation

7.1.1 Chemical Laboratory Apparatus — Such as

plastic beakers, graduated cylinders, and plastic coated

tweezers for use both with acids and with solvents.

Proper facilities for handling and disposing of acids and

their vapors are essential.

7.1.2 Hot Plate — Large enough to heat wafer of

interest to 200°C.

7.2 Probe Assembly

7.2.1 Four-point Probe — Having conical probe tips of

a durable material such as tungsten-carbide. The

included angle of the tips shall be in the nominal range

45 to 150°. The probe tips shall be in a straight line

with nominally equal separations in the range 250 m

(0.010 in.) to 1.59 mm (0.0625 in.). An isolation

resistance between adjacent probes, a factor of 10

larger than the sheet resistance of the film is required; a

value of 10

, or greater, is recommended for the

widest general application. Recommended tip radii and

probe force values are slightly different depending upon

whether the probe force is applied by springs, or

deadweight, as follows:

7.2.1.1 Spring-loaded Probes — Having tips that

terminate in a radius in the nominal range 25 to 250 m

or in a flat circular truncation of the cone with a circle

diameter in the range 50 to 125 m. A probe force per

pin of 0.25 to 2.0 N (approximately 25 to 200 gf) may

be needed to cover the variety of films that are covered

under the scope of this test method.

7.2.1.2 Deadweight-loaded Probes — Having probe

tips that terminate in a radius of at least 19 m. A

probe force per pin of 0.1 to 1 N (approximately 10 to

100 gf) may be needed to cover the variety of films that

are covered under the scope of this test method.

NOTE 4: In general, the blunter the probe tip, the higher the

probe force that is used to make good electrical contact to the

layer. Conversely, the sharper the probe tip, or the thinner the

layer, the lighter the probe force that should be used. The

upper end of the allowed probe force range is generally used

only for buried-peak conducting layers, such as MeV

implants, or for high-resistivity, thick epitaxial layers.

General experience indicates that there is not a simple

specification of one or two combinations of probe radius and

probe force that will cover all layers of interest. Experience

with conditioning probe tips against materials such as a

sapphire or non-polished ceramic substrate, or even on a piece

of lapped silicon, indicates that the microroughness of the

probe tip is a very important, but not-readily specifiable

SEMI MF1529-1104 © SEMI 2004 5

parameter for the proper probe type in a given application.

The combination of probe radius and force that is chosen

affects both the likelihood of probe penetration and the

quality of electrical contact, which in turn affects the

measurement noise and accuracy. A use test is given to aid in

verifying appropriateness of a given probe for a specific layer

type. Controlled lowering of the probe pins so that contact is

made without lateral scrubbing of the probes against the wafer

surface has been found to be very important.

7.3 Microscope

7.3.1 The microscope for inspecting probe damage shall

have a magnification of at least 600×, and an eyepiece

magnification no greater than 15×. The microscope

shall be capable of dark-field, interference contrast, or

oblique illumination.

7.3.2 The microscope shall have a stage capable of

moving the specimen in order to examine a number of

adjacent damage marks made by each of the four probe

points.

7.4 Measurement Stage

7.4.1 Wafer Prober Stage — With a vacuum chuck or

comparable means of holding the wafer securely during

measurement. This vacuum chuck should be of

sufficient thermal mass to keep the wafer at a constant

temperature, within 1°C, during the time required for

all measurements. The stage should be provided with

stops, pins, engraved circles, or other means for

accurately and repeatedly positioning wafers. For

measurements on wafers where the deposited or

fabricated film may extend over the edges of the

substrate and make contact to the backside, a thin layer

of mica, or other electrical insulator must be used

between the wafer and the chuck.

7.4.2 Probe Assembly Support that allows the probes to

be lowered onto the wafer surface with no evidence of

lateral movement (probe skidding). This requirement

can be verified by lowering and raising the probes a

number of times onto a polished silicon surface with

steps of 50 to 100 m between these locations, and then

observing the probe damage marks for each of the

probe points with the required microscope.

NOTE 5: For a probe with blunt tips or well-conditioned

probe points, it is generally very difficult to view the probe

damage with bright field illumination; use of dark-field,

Nomarski, or oblique illumination is recommended. To aid in

locating the probe damage, the formation of a grid of

rectangles, by scribing, etching or other suitable process on a

polished wafer surface, has been found helpful. The

rectangles should be large enough to allow all four probe

points to be readily located within the boundaries and a

number of probe impressions to be made within the confines

of a single rectangle.

7.4.3 Wafer Probe Stage with a sufficient range of

motion to allow probing all desired locations on the

largest wafers to be measured. Except for restrictions

on the exclusion of three probe spacings at the

perimeter of the layer being measured, the accuracy of

sheet resistance measurements using this test method do

not require any particular accuracy on the position

coordinates.

NOTE 6: If this test method is used for referee

measurements, uncertainty in position coordinates may

produce accurate measurements at the locations measured but

may make comparison of data more difficult. It is

recommended that the wafer be centered on the stage with an

accuracy of 1 mm, or better, and that all positions measured,

have coordinates controlled with an accuracy of 10 m, or

better, with respect to the center of the stage.

7.4.4 The wafer stage shall be instrumented with a

temperature monitor to be used for any application

where the average sheet resistance of the layer is a

parameter to be reported. The temperature monitor

may be of any convenient type, but must be accurate to

0.3°C, or better.

7.5 Electrical Measuring Apparatus

7.5.1 The conceptual layout of the electronic circuitry is

shown in Figure 1 for the case where a standard resistor

is used to monitor the applied current. The standard

resistor can be omitted if the current value is set or

known directly.

7.5.1.1 Constant DC-Current Source — Having

sufficient compliance voltage to supply a constant

current that results in a measured voltage drop on the

specimen that is between 5 and 20 mV. Currents

between 10

6

and 10

2

A are required if the sheet

resistance range 1 to 20,000  is to be covered. The

output current must be stable to 0.01%, or better, during

the time required to take all data at each location; ripple

and other noise must be less than 0.1% of the dc-current

level. A compliance voltage in excess of 10 V is

generally not needed unless measurements must be

made through a significant layer of oxide or other

dielectric. A standard resistor (Section 7.5.1.2) is

needed to determine measurement current unless the

current supply is in calibration and known to output a

dc-current that is within 0.1% or better of the set-point

value. A wet or dry battery may be used for the current

source providing there is means for regulating the

output current.