MIL- STD-883F 2004 TEST METHOD STANDARD MICROCIRCUITS - 第588页

MIL-STD-883F METHOD 5003 20 November 1969 6 c. Di ffus ion i mperfec tions . d. Junc tion geometries . e. Inte rmetallic phase f ormati on. f. Voids at t he bond/meta llization in terface. g. Diff usion of contac t metal…

MIL-STD-883F

METHOD 5003

20 November 1969

3.3.1 Total device cross section. This procedure shall be used where there are indications of defects in the package, die

or substrate, bonds, seals, or structural elements. The following steps shall be performed:

a. Mount the device in the appropriate orientation for cross sectioning procedures.

b. Section to reveal desired feature(s) and stain where applicable.

c. Employ bright field, dark field, or polarized light photomicrography at suitable magnification.

d. Make photographic record of defective regions or features pertinent to the mode or mechanism of failure.

3.3.2 Oxide defect analysis

. This procedure shall be used where there are indication of oxide (or other dielectric)

structural anomalies or contamination within or under the oxide or where it is necessary to determine the specific location

and structure of such defects. The following steps shall be performed:

a. Remove bonds to die or substrate and remove metallized interconnection layer(s).

b. Observe the oxide using interferometric or phase contrast photomicrography at suitable magnification and make

appropriate photographic record.

c. Observe and probe semiconductor contact (window or cut) areas as applicable, recording appropriate electrical

characteristics.

d. Mount the die or substrate in the appropriate orientation for sectioning (angle or cross) procedures, cut or lap to

reveal desired features and stain where applicable.

e. Make photographic record at suitable magnification.

3.3.3 Diffusion defect analysis

. This procedure shall be used where there are indications of diffusion imperfections,

diffusion of contact metal into the semiconductor, structural defects in the semiconductor or anomalies in junction

geometries. The following steps shall be performed:

a. Remove bonds to die or substrate and remove metallized interconnection layer(s).

b. Remove oxide or other dielectric passivation layer.

c. Probe contact regions recording appropriate electrical characteristics.

d. Stain surface to delineate junctions.

e. Mount the die or substrate in the appropriate orientation for cross sectioning or angle lapping, as applicable.

f. Cut or lap as required to expose significant features and stain junctions (may involve successive lap and stain

operations to approach specific defect).

g. Make photographic record at suitable magnification of significant features and record pertinent electrical probing

results.

3.3.4 Information obtainable

. Failure analysis in accordance with test condition C provides additional capability for

detecting or defining the following types of defects:

a. Oxide or dielectric imperfections.

b. Oxide or dielectric thicknesses.

MIL-STD-883F

METHOD 5003

20 November 1969

c. Diffusion imperfections.

d. Junction geometries.

e. Intermetallic phase formation.

f. Voids at the bond/metallization interface.

g. Diffusion of contact metal into the semiconductor or substrate.

h. Migration of metals across, through, or under the oxide or dielectric.

i. Voids in die or substrate mount.

3.4 Optional measurements

. The purpose of failure analysis is to obtain sufficient information to initiate corrective action

in device design, production, test, or application. It may be necessary to obtain more detailed information than can be

acquired in test conditions A, B, or C on the nature of contaminants or phases observed, concentrations, dimensions of

submicroscopic features, etc. The selection and use of a number or less conventional analytical techniques by highly

qualified personnel can provide this more extensive or fundamental knowledge of the precise chemical, physical, or

electrical mechanisms of failure. The decision as to which techniques are appropriate and the point in the analytical

sequence of test conditions A, B, or C at which they should be employed is contingent on the nature of information desired

and previous results obtained from the specified analytical procedures, and must be left to the discretion of the analyst. Any

of the following techniques may therefore be introduced into a failure analysis sequence at the appropriate point provided

precautions are taken to avoid destruction of the evidence of failure which may be observed in subsequent procedures.

Where multiple samples of the same type of device or failure exist, it shall be permissible to subdivide the quantity of

devices and employ destructive techniques in parallel with the specified test condition provided all samples have been

exposed to electrical verification tests and internal examination (see 3.1.1 through 3.1.3 and 3.2.1 through 3.2.5) prior to any

of the optional measurements. When any of these optional measurements are employed, they shall be listed in the failure

analysis report including the details of the method applied, conditions of test and results.

a. Residual gas analysis. When device surface contamination is indicated as a possible cause of failure, the lid of an

unopened device shall be punctured and the internal gaseous ambient analyzed for the type and concentration of

volatile products. This information then supplements electrical leakage current measurements and hermeticity

tests.

b. Surface profilometer measurement. A mechanical determination of surface topography variations can be made

using this type of instrument. This records the vertical motion of a stylus moved across the surface of the device.

This information can be used to quantitatively determine oxide, dielectric, or metal thicknesses.

c. Photoscanning. A device, with leads and interconnections intact, after being opened, can be scanned with a small

diameter beam of light which generates photovoltages in active p-n junctions. This generated photovoltage which

is dependent on many physical junction properties indicates the presence of surface channels or inversion layers

or both, caused by contamination on, in, or under the passivating oxide layer. It is also possible to locate certain

regions of enhanced high field multiplication, mask misregistration, imperfect diffusions, as well as other device

imperfections involving junction properties.

d. Infrared scanning. An IR detector, sampling infrared radiation from various points of the surface of an operating

microcircuit, can detect the location of hot spots and other thermal abnormalities.

MIL-STD-883F

METHOD 5003

20 November 1969

e. Scanning electron microscopy and electron beam microanalysis. The scanning electron microscope, employing an

electron beam with a diameter on the order of a few hundred angstroms, is the most effective means of attaining

device structural information without the need for special sample preparation procedures. The scanning electron

microscope can perform chemical analysis, such as the microanalyzer, by incorporating a nondispersive x-ray

detector. An electron beam microanalyzer can be used for x-ray spectrochemical analysis of micron sized

volumes of material. Several other device structural properties are determinable through detection and display of

back-scattered primary electrons and secondary electrons. These instruments are most generally used for:

(1) Determination of surface potential variations using secondary electron scanning microscopy. The small size

of the electron beam coupled with the properties of secondary electrons result in the ability to examine

physical defects with much higher resolution and depth of field than light microscopy.

(2) Analysis of micron sized defects such as oxide pin-holed, metallization grain structure.

(3) Determination of products of solid state reactions, such as diffusion, precipitation, and intermetallic

formation.

(4) Corrosion product identification.

f. Electron microscopy. An examination at extremely high magnification of the structure of failed metallization and

bulk materials is best accomplished using electron microscopy.

g. Special test structures. Often the amount of reacted material on a failed circuit is too small to allow definitive

determination of chemical and structural properties. In addition, it is often necessary to reproduce the failure in a

controlled experimental manner for verification of the mechanism of failure. Special test structures may be

fabricated with variations in geometry and materials permitting study of the mechanism without extraneous

influences. This is most advantageous when information is desired concerning the basic failure mechanism(s).

4. SUMMARY

. The following details must be specified in the applicable acquisition document:

a. Test condition letter (see 3.) for test conditions A, B, or C and where applicable, optional measurements (see 3.4),

identifying the specific procedures to be applied and details as to their option application.

b. Any special measurements not described in the applicable test condition.

c. Requirements for data recording and reporting including instructions as to disposition of original data, photographs,

radiographs, etc.

d. Physical and electrical specifications and limits for the device being analyzed.