semi合集-English.pdf - 第5370页

SEMI M55-0705 © SEMI 2003, 2004 4 4.1.14 total thickness variation (TTV) — the difference between the m aximum and minimum thickness val ues of a wafer encountered during a scan pattern or a series of point requ irements…

SEMI M55-0705 © SEMI 2003, 2004 3

4 Terminology

NOTE 1: Many definitions and terms not given in this section can be found in SEMI M1, the SEMI Compilation of Terms, and

ASTM F154.

4.1 Definitions

4.1.1 bow  of a semiconductor wafer, a measure of concave or convex deformation of the median surface of a

wafer, independent of any thickness variation which may be present. Bow is a bulk property of the test specimen,

not a property of an exposed surface. Generally, bow is determined with a test specimen in a free, unclamped

condition. Units of bow are generally micrometers.

4.1.2 crystallite — any part of the wafer, having an arbitrary orientation of its crystallographic axis in respect to the

monocrystalline part of the wafer.

4.1.3 dopant  a chemical element, usually from the third or fifth column of the periodic table for the case of IV-

IV compounds, incorporated in trace amounts in a semiconductor crystal to establish its conductivity type and

resistivity.

4.1.4 edge contouring  on wafers whose edges have been shaped by mechanical and/or chemical means, a

description of the profile of the boundary of the wafer joining the front and back sides.

4.1.5 edge exclusion  the width X of a narrow band of wafer surface, located just inside the wafer edge, over

which the values of the specified parameter do not apply. See definition of fixed quality area below.

4.1.6 fixed quality area (FQA) — The central area of a wafer surface, defined by a nominal edge exclusion, X, over

which the specified values of a parameter apply.

4.1.6.1 Discussion — The boundary of the FQA is at all points the distance X away from the periphery of a wafer of

nominal dimensions. (See Figure 1.) The size of the FQA is independent of wafer diameter and flat length

tolerances.

4.1.7 lot  for the purpose of this document, (a) all of the wafers of nominally identical size and characteristics

contained in a single shipment, or (b) subdivisions of large shipments consisting of wafers as above which have been

identified by the supplier as constituting a lot.

4.1.8 micropipe  small hollow tube approximately parallel to the crystallographic c-axis and extending through

the whole crystal.

4.1.9 orthogonal misorientation  in {0001} wafers cut intentionally “off-orientation”, the angle between the

projection of the vector normal to the wafer surface onto the {0001} plane and the projection on that plane of the

specified direction of tilt in the {0001} plane. (See Figure 2.)

4.1.10 planar defect

 small cavity in a SiC bulk crystal with large width-to-height ratio roughly parallel to the

{0001} lattice plane. The lateral boundaries are parallel to crystallographic directions. Often one or more

micropipes are connected to a planar defect.

4.1.11 polytype  one possible crystallographic modification of a substance which shows the phenomenon of

polytypism. All polytypes of a substance have the same lattice layers with nearly the same lattice constant in

common. However the stacking sequence of these layers differs between different polytypes. Most commonly

polytypes are named after a suggestion of Ramsdell

: A symbol like 6H gives the number of layers in one periodic

stacking sequence (2, 3, 4, ... ) and the symmetry of the resulting crystal (H = hexagonal, R = rhombohedral). The

most common polytypes of SiC are 6H, 4H, 15R.

4.1.12 surface orientation  the tilt angle between the crystallographic c-axis and the wafer surface normal. (See

Figure 2.)

4.1.13 total indicator reading (TIR) — the smallest perpendicular distance between two planes, both parallel with

the reference plane, which encloses all points on the front surface of a wafer within the FQA, the site, or the subsite,

depending on which is specified.

6 L. S. Ramsdell, J. A. Kohn: Developments in Silicon Carbide Research, Acta Cryst. 5 (1952) 215 – 224

SEMI M55-0705 © SEMI 2003, 2004 4

4.1.14 total thickness variation (TTV) — the difference between the maximum and minimum thickness values of a

wafer encountered during a scan pattern or a series of point requirements. TTV is generally expressed in

micrometers.

4.1.15 warp  of a semiconductor slice or wafer, the difference between the maximum and minimum distance of

the median surface of the wafer from a reference plane, encountered during a scan pattern. Warp is a bulk property

of the test specimen, not a property of an exposed surface. Warp is generally expressed in micrometers.

5 Ordering Information

5.1 Purchase orders for silicon carbide wafers furnished to this specification shall include the following items:

5.1.1 Polytype,

5.1.2 Nominal diameter (see applicable SEMI Standard for polished SiC wafers),

5.1.3 Thickness (see applicable SEMI Standard for polished SiC wafers),

5.1.4 Dopant (see applicable SEMI Standard for polished SiC wafers),

5.1.5 Resistivity or Carrier Concentration (see applicable SEMI Standard for polished SiC wafers),

5.1.6 Total Thickness Variation (see applicable SEMI Standard for polished SiC wafers),

5.1.7 Surface orientation (see applicable SEMI Standard for polished SiC wafers),

5.1.8 Polarity of Surfaces (see applicable SEMI Standard for polished SiC wafers),

5.1.9 Lot Acceptance Procedures (see §7),

5.1.10 Certification (see §11), and

5.1.11 Packing and Marking (see §12).

6 Dimensions and Permissible Variations

6.1 The material shall conform to the dimensions and dimensional tolerances as specified in the applicable polished

silicon carbide wafer standard.

6.2 The material shall conform to the crystallographic orientation details as specified in the applicable polished

silicon carbide wafer standard.

6.3 If edge contoured wafers are specified on the purchase order, the profile shall conform to the following

requirements at all points on the wafer periphery.

6.3.1 When the wafer is aligned with the SEMI Wafer Edge Profile Template (see Figure 3) so that the x-axis of the

template is coincident with the wafer surface and the y-axis of the template forms a tangent with the outermost radial

portion of the contour, the wafer edge profile must be contained within the clear region of the template. (See Figure

4 for example of acceptable and unacceptable contours.)

6.3.2 Cosmetic attributes of the edge contour are not covered by this specification. They shall be agreed upon

between supplier and purchaser.

6.4 Flats shall conform to the requirements of §9 and the appropriate polished silicon carbide wafer standard.

NOTE 2: For edge chips and indents see §10.

7 Sampling

7.1 Unless otherwise specified, ASTM Practice E122 shall be used. When so specified, appropriate sample sizes

shall be selected from each lot in accordance with ANSI/ASQC Z1.4. Each quality characteristic shall be assigned

an acceptable quality level (AQL) or lot total percent defective (LTPD) value in accordance with ANSI/ASQC Z1.4

definitions for critical, major and minor classifications. If desired and so specified in the contract or order, each of

these classifications may alternatively be assigned cumulative AQL or LTPD values. Inspection levels shall be

agreed upon between the supplier and the purchaser.

SEMI M55-0705 © SEMI 2003, 2004 5

8 Test Methods

NOTE 3: SiC wafers are extremely fragile. While the mechanical dimensions of a wafer can be measured by use of tools such as

micrometer calipers and other conventional techniques, the wafer may be damaged physically in ways that are not immediately

evident. Special care must therefore be used in the selection and execution of measurement methods.

8.1 Test Plan for Crystal Quality within One Crystal  Determine by a method agreed upon between the supplier

and purchaser.

NOTE 4: The assessment of the crystal quality is a problem of great practical impact as it can be very time consuming and costly

or even impossible in the case of destructive test methods to test every wafer. However in general crystal quality does not change

abruptly in a crystal. The evaluation of a subset of all wafers from a given crystal will give sufficient information about the

quality of the whole crystal.

8.2 Polytype  For nominally undoped material (n < 10

/cm

) determine by visual inspection at 77K (liquid

nitrogen) under UV excitation. For doped material (n > 10

/cm

) determine by visual inspection of the colors of the

doped material under diffuse lighting conditions.

8.3 Diameter  Determine by ASTM Test Method F2074.

8.4 Thickness, Center Point  Determine by ASTM Test Method F533.

8.5 Flat Length  Determine by ASTM Test Method F671.

8.6 Bow and Warp  Determine bow in accordance with ASTM Test Method F534 and warp in accordance with

ASTM Test Method F657.

NOTE 5: ASTM has standardized two methods for measuring warp. ASTM Test Method F1390 is an automated, non-contact

method which provides for correction of the wafer deflection due to gravitational effects. The scan pattern covers the entire fixed

quality area. ASTM Test Method F657 is a manual, non-contact method which has a continuous, prescribed scan pattern which

covers only a portion of the wafer surface. There is no provision for correction of the wafer deflection due to gravitational

effects. As noted in Appendix 2, different reference planes are used for the two methods. Because Test Method F657 employs a

back surface reference plane, the measured warp may include contributions from thickness variation of the wafer. Test Method

F1390 employs a median surface reference plane and is not susceptible to interferences from thickness variations. In general,

Test Method F1390 is preferred, especially for wafers 150 mm in diameter and larger.

8.7 Total Thickness Variation  Determine by ASTM Test Method F657.

NOTE 6: ASTM Test Method F533, DIN 50441/1 and JIS H 0611 are all 5 point methods, while Test Method F657 involves a

continuous scan pattern over a portion of the wafer surface and Test Method F1530 involves an automated continuous scan

pattern over the entire wafer surface. JIS H 0611 differs from ASTM Test Method F533 and DIN 50441/1, in that the

measurements in JIS H 0611 are taken at the center and at 5 mm from the edge on diameters parallel and perpendicular to the

primary orientation flat or notch bisector, while the measurements in ASTM Test Method F533 and DIN 50441/1 are taken at the

center and at the same radial distance (R

nominal

-6 mm) on diameters 30° and 120° counterclockwise from the bisector to the

primary orientation flat or notch (with the wafer facing front surface up).

8.8 Surface Polarity  Determine by a method agreed upon between the supplier and the purchaser.

NOTE 7: There are several destructive and non-destructive methods. Most common examples are a chemical etch of the surface

(destructive), by comparing the differently reacting carbon and silicon faces, and wet oxidation (non-destructive) by comparing

the different growth rates on both surfaces by measuring the oxide layer thickness.

8.9 Flat Orientation  Determine by ASTM Test Methods F847.

8.10 Surface Orientation  Determine by ASTM Test Methods F26.

8.11 Orthogonal Misorientation  Determined by a method agreed upon between the supplier and purchaser.

8.12 Surface Defects and Contamination

8.12.1 Visually Observable Surface Defects  Determined by ASTM Practice F523 or a method agreed upon

between the supplier and purchaser.

8.13 Edge Contour  Determine by ASTM Test Method F928.

8.14 Resistivity  For conductive wafers determine by ASTM Test Method F43 or ASTM Test Method F673. For

high-resistivity or semi-insulating material determine by DIN 50448.