ASTM F1892-12(2018)

This international standard was developed in accordance with internationally recognized principles on standardization established in the Decision on Principles for the
Development of International Standards, Guides and Recommendations issued by the World Trade Organization Technical Barriers to Trade (TBT) Committee.
Designation: F1892 − 12 (Reapproved 2018)
Standard Guide for
Ionizing Radiation (Total Dose) Effects Testing of
Semiconductor Devices
This standard is issued under the fixed designation F1892; the number immediately following the designation indicates the year of
original adoption or, in the case of revision, the year of last revision. A number in parentheses indicates the year of last reapproval. A
superscript epsilon (´) indicates an editorial change since the last revision or reapproval.
INTRODUCTION
This guide is designed to assist investigators in performing ionizing radiation effects testing of
semiconductor devices, commonly termed total dose testing. When actual use conditions, which
include dose, dose rate, temperature, and bias conditions and the time sequence of application of these
conditions, are the same as those used in the test procedure, the results obtained using this guide
applies without qualification. For some part types, results obtained when following this guide are
much more broadly applicable. There are many part types, however, where care must be used in
extrapolating test results to situations that do not duplicate all aspects of the test conditions in which
the response data were obtained. For example, some linear bipolar devices and devices containing
metal oxide semiconductor (MOS) structures require special treatment. This guide provides direction
for appropriate testing of such devices.
1. Scope 1.5 Determination of an effective and economical hardness
test typically will require several kinds of decisions. A partial
1.1 This guide presents background and guidelines for
enumeration of the decisions that typically must be made is as
establishing an appropriate sequence of tests and data analysis
follows:
procedures for determining the ionizing radiation (total dose)
1.5.1 Determination of the Need to Perform Device
hardness of microelectronic devices for dose rates below 300
Characterization—For some cases it may be more appropriate
rd(SiO )/s. These tests and analysis will be appropriate to assist
to adopt some kind of worst case testing scheme that does not
in the determination of the ability of the devices under test to
require device characterization. For other cases it may be most
meet specific hardness requirements or to evaluate the parts for
effective to determine the effect of dose-rate on the radiation
use in a range of radiation environments.
sensitivity of a device. As necessary, the appropriate level of
1.2 The methods and guidelines presented will be applicable
detail of such a characterization also must be determined.
to characterization, qualification, and lot acceptance of silicon-
1.5.2 Determination of an Effective Strategy for Minimizing
based MOS and bipolar discrete devices and integrated cir-
the Effects of Irradiation Dose Rate on the Test Result—The
cuits. They will be appropriate for treatment of the effects of
results of radiation testing on some types of devices are
electron and photon irradiation.
relatively insensitive to the dose rate of the radiation applied in
the test. In contrast, many MOS devices and some bipolar
1.3 This guide provides a framework for choosing a test
devices have a significant sensitivity to dose rate. Several
sequence based on general characteristics of the parts to be
different strategies for managing the dose rate sensitivity of test
tested and the radiation hardness requirements or goals for
results will be discussed.
these parts.
1.5.3 Choice of an Effective Test Methodology—The selec-
1.4 This guide provides for tradeoffs between minimizing
tion of effective test methodologies will be discussed.
the conservative nature of the testing method and minimizing
1.6 Low Dose Requirements—Hardness testing of MOS and
the required testing effort.
bipolar microelectronic devices for the purpose of qualification
or lot acceptance is not necessary when the required hardness
is 100 rd(SiO ) or lower.
This guide is under the jurisdiction of ASTM Committee E10 on Nuclear
Technology and Applications and is the direct responsibility of Subcommittee
1.7 Sources—This guide will cover effects due to device
E10.07 on Radiation Dosimetry for Radiation Effects on Materials and Devices. 60
testing using irradiation from photon sources, such as Co γ
Current edition approved March 1, 2018. Published April 2018. Originally
irradiators, Cs γ irradiators, and low energy (approximately
approved in 1998. Last previous edition approved in 2012 as F1892 – 12. DOI:
10.1520/F1892-12R18. 10 keV) X-ray sources. Other sources of test radiation such as
Copyright © ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959. United States
F1892 − 12 (2018)
linacs, Van de Graaff sources, Dymnamitrons, SEMs, and flash 3.2 Definitions of Terms Specific to This Standard:
X-ray sources occasionally are used but are outside the scope 3.2.1 accelerated annealing test, n—procedure utilizing el-
of this guide.
evated temperature to accelerate time-dependent growth and
annealing of trapped charge.
1.8 Displacement damage effects are outside the scope of
this guide, as well. 3.2.2 category A, n—used to refer to a part containing
bipolar structures that is not low dose rate sensitive.
1.9 The values stated in SI units are to be regarded as the
standard. 3.2.3 category B, n—used to refer to a part containing
bipolar structures that is low dose rate sensitive.
1.10 This international standard was developed in accor-
dance with internationally recognized principles on standard- 3.2.4 characterization, n—testing to determine the effect of
ization established in the Decision on Principles for the dose, dose-rate, bias, temperature, etc. on the radiation induced
Development of International Standards, Guides and Recom- degradation of a part.
mendations issued by the World Trade Organization Technical
3.2.5 delayed reaction rate effect (DRRE), n—a time and
Barriers to Trade (TBT) Committee.
temperature dependent effect where the rate of degradation for
a second irradiation is much greater than the rate of degrada-
2. Referenced Documents
tion for the first irradiation after a delay time that is dependent
2.1 ASTM Standards:
on the temperature of the part during the time between the two
E170 Terminology Relating to Radiation Measurements and
irradiations.
Dosimetry
3.2.6 enhanced low dose rate sensitivity (ELDRS), n—used
E666 Practice for Calculating Absorbed Dose From Gamma
to refer to a bipolar part that shows enhanced (greater)
or X Radiation
radiation induced damage for a fixed dose at dose rates below
E668 Practice for Application of Thermoluminescence-
about 50 rd(SiO )/s compared to damage at the same dose for
Dosimetry (TLD) Systems for Determining Absorbed
dose rates of >50 rd(SiO )/s. The enhancement may be a result
Dose in Radiation-Hardness Testing of Electronic Devices
of true dose rate effects or time dependent effects, or both.
E1249 Practice for Minimizing Dosimetry Errors in Radia-
3.2.7 gray, n—the gray (Gy) symbol, is the SI unit of
tion Hardness Testing of Silicon Electronic Devices Using
absorbed dose, defined as 1 Gy = 1 J/kg (1 Gy = 100 rd).
Co-60 Sources
E1250 Test Method for Application of Ionization Chambers
3.2.8 in-flux tests, n—measurements made in-situ while the
to Assess the Low Energy Gamma Component of
test device is in the radiation field.
Cobalt-60 Irradiators Used in Radiation-Hardness Testing
3.2.9 in-situ tests, n—electrical measurements made on
of Silicon Electronic Devices
devices during, or before-and-after, irradiation while they
F996 Test Method for Separating an Ionizing Radiation-
remain in the irradiation location.
Induced MOSFET Threshold Voltage Shift Into Compo-
3.2.10 in-source tests, n—an in-flux test.
nents Due to Oxide Trapped Holes and Interface States
Using the Subthreshold Current–Voltage Characteristics
3.2.11 ionizing radiation effects, n—the changes in the
F1467 Guide for Use of an X-Ray Tester (≈10 keV Photons) electrical parameters of a microelectronic device resulting from
in Ionizing Radiation Effects Testing of Semiconductor
radiation-induced trapped charge.
Devices and Microcircuits
3.2.11.1 Discussion—Ionizing radiation effects are some-
ISO/ASTM 51275 Practice for Use of a Radiochromic Film
times referred to as“ total dose effects.”
Dosimetry System
3.2.11.2 Discussion—In this guide, doses and dose rates are
2.2 Military Specifications:
specified in rd(SiO ) as contrasted with the use of rd(Si) in
MIL-STD-883 , Method 1019, Ionizing Radiation (Total other related standards. The reason is that for ionizing radiation
Dose) Test Method
effects in silicon based microelectronic components, it is the
MIL-STD-750 , Method 1019, Steady-State Total Dose Ir- energy deposited in the SiO gate, field, and spacer oxides that
radiation Procedure
is responsible for the radiation-induced degradation effects. For
MIL-HDBK-814 Ionizing Dose and Neutron Hardness As- high energy irradiation, for example, Co photons, the differ-
surance Guidelines for Microcircuits and Semiconductor
ence between dose deposited in Si and SiO typically is
Devices negligible. For X-ray irradiation, approximately 10 keV photon
energy, the energy deposited in Si under some circumstances
3. Terminology
may be approximately 1.8 times the energy deposited in SiO .
For additional details, see Guide F1467.
3.1 For terms relating to radiation measurements and
dosimetry, see Terminology E170.
3.2.12 not in-flux test, n—electrical measurements made on
devices at any time other than during irradiation.
3.2.13 overtest, n—a factor that is applied to the specifica-
For referenced ASTM standards, visit the ASTM website, www.astm.org, or
contact ASTM Customer Service at service@astm.org. For Annual Book of ASTM
tion dose to determine the test dose level that the samples must
Standards volume information, refer to the standard’s Document Summary page on
pass to be acceptable at the specification level. An overtest
the ASTM website.
factor of 1.5 means that the parts must be tested at 1.5 times the
Available from the Standardization Documents Order Desk, Building 4, Section
D, 700 Robbins Ave., Philadelphia, PA 19111–5094. specification dose.
F1892 − 12 (2018)
3.2.14 parameter delta design margin (PDDM), n—a design (a) Radiation Source—The type of radiation source ( Co,
margin that is applied to the radiation induced change in an X-ray, etc.) that is to be used.
electrical parameter.
NOTE 1—The ionizing dose response of many device types has been
3.2.14.1 Discussion—For example, for a PDDM of 3 the
shown to depend on the type of ionizing radiation to which the device is
change in a parameter at a specified dose from the pre-
subjected. The selection of a suitable radiation source for use in such a test
irradiation value is multiplied by three and added to the must be based on the understanding that the gamma or electron radiation
source will induce a device response that then should be correlated to the
pre-irradiation value to see if the sample exceeds the post-
response anticipated in the device application.
irradiation parameter limit. For example, if the pre-irradiation
(b) Dose Rate Range—The range of dose rates within
value of I is 30 nA and the post-irradiation value at 20
b
which the radiation exposures must take place (see 6.4).
krd(SiO ) is 70 nA (change in I is 40 nA), then for a PDDM
2 b
NOTE 2—The response of many devices has been shown to be highly
of 3 the post-irradiation value would be 150 nA (30 nA + 3 X
dependent on the rate at which the dose is accumulated. There must be a
40 nA). If the allowable post-irradiation limit is 100 nA the part
demonstrated correlation between the response of the device under the
would fail.
selected test conditions and the rate at which the device would be expected
to accumulate dose in its intended application.
3.2.15 qualification, n—testing to determine the adequacy
(c) Operating Conditions—The test circuit, electrical bi-
of a part to meet the requirements of a specific application.
ases to be applied, and the electrical operating sequence, if
3.2.16 rad, n—the rad symbol, rd, is a commonly used unit
applicable, for the part during irradiation (see 6.3). This
for absorbed dose, defined in terms of the SI unit of absorbed
includes the use of in-flux or not in-flux testing.
dose as 1 rd = 0.01 Gy.
(d) Electrical Parameters—The measurements that are to
3.2.17 remote tests, n—electrical measurements made on
be made on the test devices before, during (if appropriate), and
devices that are removed physically from the irradiation after (if appropriate) irradiation.
location for the measurements.
(e) Time Sequence—The exposure time, the elapsed time
between exposure and post-exposure measurements, and the
3.2.18 time dependent effects (TDE), n—the time dependent
time between irradiations (see 6.5).
growth and annealing of ionizing radiation induced trapped
(f) Irradiation Levels—The dose(s) to which the test device
charge and interface states and the resulting transistor or IC
is to be exposed between measurements (see Practice E666).
parameter changes caused by these effects.
(g) Dosimetry—The dosimetry technique (TLDs,
3.2.18.1 Discussion—Similar effects also take place during
calorimeters, diodes, etc.) to be used. This depends to some
irradiation. Because of the complexity of time dependent
extent on the radiation source selection.
effects, alternative, but not inconsistent, definitions may prove
(h) Temperature—Exposure, measurement, and storage
useful. Two of these are: the complex of time-dependent
temperature ranges (see 6.5 and 6.6).
processes that alter trapped oxide change (ΔN ) and interface
ot
(i) Experimental Configuration—The physical arrange-
trap density (ΔN ) in an MOS or bipolar structure during and
it
ment of the radiation source, test unit, radiation shielding, and
after irradiation; and, the effects of these processes upon device
any other mechanical or electrical elements of the test.
or circuit characteristics or performance, or both.
(j) Accelerated Annealing Testing for MOS—The acceler-
3.2.19 true dose rate effect, n—a response that occurs during
ated annealing tests called for in 8.2.2.3 (a) through (e) should
low dose rate irradiation that cannot be reproduced with a high
be performed for hardness assurance testing of any device that
dose rate irradiation followed by an equivalent time anneal.
contains MOS elements by design. Further requirements and
exceptions to such accelerated annealing testing may be made
4. Summary of Guide
based on the factors discussed in Appendix X1.
4.1 This guide is designed to provide an introduction and
(k) Special Testing for Linear Bipolar—The special testing
direction to the purposes, methods, and strategies of total
procedures called for in 8.1.2.1 through 8.1.2.5 and 8.2.3.1
ionizing dose testing.
through 8.2.3.4 should be performed for hardness assurance
4.1.1 Purposes—Device or system hardness may be mea-
testing of linear bipolar devices. Further requirements and
sured for several different purposes. These may include device
suggestions for the testing of linear bipolar devices will be
characterization, device qualification, lot acceptance, line
found in Appendix X2.
qualification, and studies of device physics.
4.1.3 Strategies—Several kinds of strategies may prove
4.1.2 Methods:
useful for device testing. The strategy used will depend on the
4.1.2.1 An ionizing radiation effects test consists of per-
key impediments to accurate, repeatable, and inexpensive
forming a set of electrical measurements on a device, exposing
testing. For example, it may be useful to measure device
the device to ionizing radiation while appropriately biased, and
properties at several different dose rates and then to extrapolate
then performing a set of electrical measurements either during
to the results expected at the actual dose rate anticipated in use.
or after irradiation.
Then again, it may be more efficient to devise a method that
4.1.2.2 Because several factors enter into the effects of the
will place an upper or lower bound on the excursions that may
radiation on the device, parties to the test must establish and
be anticipated for a given device parameter.
agree to a variety of conditions before the validity of the test
can be established or before the results of any one test can be 4.2 The choice of optimal procedures for the performance of
compared with those of another. Conditions that must be total ionizing dose testing typically involves resolution of the
established and agreed to include the following: conflicts between the following four competing requirements:
F1892 − 12 (2018)
4.2.1 Test Fidelity—It is necessary that a test reproduce the higher than the dose measured by a monitoring dosimeter,
results to be expected in the projected application environment typically the average dose deposited in the dosimeter material.
to an acceptable degree of precision. The test methodology The severity of the effects is very dependent on the radiation
chosen has a strong effect on the precision of the result. source being used and the geometry of the test configuration.
Typically, however, greater test fidelity must be balanced
6.3 Bias—Most ionizing radiation effects are related to the
against greater cost. In addition, many environments cannot be
post irradiation net trapped charge in the device dielectric
reproduced in the laboratory. Often it may be necessary to have
layers, usually oxides, and to the interface traps at the
an adequate command of device physics in order to devise
dielectric-semiconductor interface. These effects often are
laboratory tests that adequately match or bound the perfor-
dependent strongly on the electrical field in the dielectric
mance to be expected in actual use.
during and after exposure (see Test Method E1250). In general,
4.2.2 Reproducibility—It is important to have test proce-
the largest effect for the net trapped charge occurs for a large
dures that can be depended upon to give approximately the
positive electric field in the dielectric during irradiation. For
same result each time when used by different laboratories.
the interface trap build-up, the worst case condition most often
Failure to achieve this goal may have significant contract
is a small electric field during irradiation and a large positive
implications. Obtaining this goal typically requires careful
field after irradiation. Radiation testing typically is performed
attention to the control of experimental variables and to the
under worst-case bias conditions. For many circuits, the
development of accurate dosimetry methods.
worst-case bias is a static dc bias with the supply voltages at
4.2.3 Single-Valued Result—For some purposes, it is desir-
their maximum rated voltage. The determination of the worst
able to have a test that can be used to simply categorize parts
case bias for the input/output lines and internal nodes of any
and that gives one answer for each part. For example, labeling
given circuit often is a complex process of circuit analysis or
of parts for the military parts system is facilitated if such a
characterization tests, or both, under many bias conditions.
characterization is available. On the other hand, the search for
Some guidance is given in the appendices for methods to
a simple characterization scheme must not be allowed to
determine the worst case irradiation and anneal bias. For
obscure real dependencies on dose rate, temperature, bias, etc.,
complementary metal-oxide semiconductor transistor (CMOS)
which may have a significant effect on operational hardness.
components, see Appendix X1; for bipolar components, see
Care must be taken to extrapolate appropriately from the
Appendix X2; and, for application-specific integrated circuits,
conditions that lead to the test rating to those conditions to be
(ASIC) see Appendix X3. The irradiation bias conditions
expected in use.
selected for any component should not exceed the manufac-
4.2.4 Testability—It is, of course, desirable to obtain a test
turer’s maximum ratings or place the component in a configu-
that is economical in its use of time, equipment, and personnel.
ration that is unrealistic for a system application.
The perfect test typically will be too expensive to perform. The
goal is to determine an optimal balance between expense and NOTE 3—Lacking information on worst-case application conditions,
preliminary analysis and characterization tests should be performed to
reliability of results.
determine worst-case conditions. In performing step-wise irradiations, it is
important to minimize the changes taking place between exposures so that
5. Significance and Use
measurements at each level accurately reflect the effects of the cumulative
5.1 Electronic circuits used in space, military, and nuclear dose to which the device was exposed. Minimum parameter changes
generally take place between exposures if the device pins are kept shorted.
power systems may be exposed to various levels of ionizing
Bias should not be changed from one level to another in a step stress
radiation. It is essential for the design and fabrication of such
sequence, in order to avoid charge neutralization effects.
circuits that test methods be available that can determine the
NOTE 4—Some space applications involve devices used at very low
vulnerability or hardness (measure of nonvulnerability) of
repetition rates; for example, electrically programmable read-only
components to be used in such systems.
memory (EPROMs.) Another example is redundant devices and circuits
that ride along in an unbiased condition until they are switched on. Still
5.2 Some manufacturers currently are selling semiconductor
another example is sensor circuits that only are biased on when a
parts with guaranteed hardness ratings. Use of this guide
measurement is to be taken. Thus, it may be desirable to characterize and
provides a basis for standardized qualification and acceptance test these devices in an unbiased condition. Ionizing dose survival levels
may be three to ten times higher in the unbiased condition than under
testing.
typical bias conditions.
6. Interferences
6.4 Dose Rate:
6.1 There are many factors that can affect the results of
6.4.1 The concentration of excess carriers depends on the
ionizing radiation tests. Care must be taken to control these
dose rate. High densities of excess carriers can affect the charge
factors to obtain consistent and reproducible results. Several of
state of trapping levels, as well as the mobilities and lifetimes
these factors are discussed as follows:
of these carriers resulting in altered post-radiation densities and
distributions of trapped charge.
6.2 Energy Spectrum—Many gamma-ray sources have as-
6.4.2 Photocurrents produced by the excess carriers gener-
sociated low-energy electron and photon components that
ated by ionization can alter internal bias levels of a semicon-
result from interaction of the gamma radiation with shielding
ductor chip, thereby causing a variation in the response of the
surrounding the source (see Practice E1249). These low-energy
device or circuit.
components can deposit their energy in a shallow layer near the
surface of the device chip. This places an absorbed dose in the 6.4.3 Because of the counteracting effects of charge anneal-
most susceptible region of a test device that can be much ing and interface state growth in some MOS device oxides, the
F1892 − 12 (2018)
dose rate at which a test is carried out can have a strong effect 6.5.2.1 The crux of the bipolar TDE issue concerns the
on the apparent device hardness (see 6.5 for further detail). properties of field oxides used to isolate the base and emitter
contacts. These oxides typically are of poor quality. The effects
6.4.4 For the reasons noted in 6.4.1 – 6.4.3, the dose rate to
of radiation on such oxides determine the radiation response of
be used in an ionizing radiation test must be established and
many bipolar transistors. A characteristic failure mechanism in
agreed upon between the parties to the test and controlled
such bipolar transistors is radiation-induced increase in the
during the test. Selection of appropriate dose rate ranges should
base current, and resulting decrease in transistor gain. This
be based on the radiation environments anticipated for the parts
excess base current largely is caused by enhanced surface
while in actual system operation.
recombination current in the emitter-base diode.
6.4.5 The use of thick absorbers in order to produce a low
6.5.2.2 For the bipolar technologies mentioned above, fail-
dose-rate Co test source must be used with caution. The
ures occur at lower doses for irradiations at low dose rates than
absorbers may cause softening of the spectrum (through
at higher rates. For example, the devices may show higher
Compton scattering). This may cause dose deposition and dose
excess base currents below 1 rd(SiO )/s than at 100 rd/(SiO )/s,
2 2
enhancement problems (see 6.2).
for the same level of accumulated total ionizing dose. Such
6.5 Time Dependent Effects:
enhanced failure at low dose rates has been observed both in
modern bipolar technologies and in relatively old designs.
6.5.1 Time Dependent Effects for MOS Devices:
These effects have been observed both in transistors and ICs.
6.5.1.1 Ionizing irradiation of MOS devices results in two
6.5.2.3 There are at least two types of enhanced low-dose-
major species of defects: trapped holes in gate (and field)
rate effects that have been characterized extensively, true dose
oxides and interface states at Si-SiO interfaces. Hole trapping
rate effects and time dependent effects. Many low-dose-rate
occurs rapidly (typically less than ;1 s) and often anneals
sensitive bipolar linear circuits have shown both types of
significantly in hours or days. Interface state density builds up
enhanced low-dose-rate effects. In addition there is a delayed
slowly (in seconds to days) and does not usually anneal
reaction rate effect described in the work of Freitag and Brown
significantly at room temperature. The relative magnitudes of
(see Refs (1, 2)) that results in an increased rate of degradation
these defects determine the effects on operation of the device
if the circuit is being irradiated at the time the interface state
and its post-irradiation time dependence. The quality of the
“precursors” arrive at the Si-SiO interface. This arrival time is
oxide determines the relative densities and saturation levels of 2
temperature dependent and for some circuits is on the order of
the defects.
several hundred thousand seconds at room temperature and
6.5.1.2 Trapped holes in the silicon oxide result in a
about ten thousand seconds at 100°C. This mechanism has only
negative shift in the gate threshold voltage for both n- and
been characterized on two circuit types to date.
p-channel devices. Interface states maintain a net negative
6.5.2.4 The true dose rate effects cannot be simulated by
charge in n-channel devices (positive gate threshold shift) and
accelerated anneal procedures, such as that recommended for
a net positive charge in p-channel devices (negative gate
MOS devices in 8.2.2.3 (a) through (e) and Appendix X1.
threshold shift). See Test Method F996.
Currently, there is no proven single universal method for
6.5.1.3 With increasing time, trapped holes are removed or
accelerating the testing of low dose-rate irradiation for all types
compensated while interface state concentrations increase.
of dose-rate sensitive bipolar devices. Some promising test
Because hole trapping occurs rapidly, initial gate threshold
methods, however, are described in Appendix X2.
shifts in both p- and n-channel devices are negative under
6.6 Temperature:
irradiation at moderate to high dose rates. As time passes, the
6.6.1 Because time-dependent effects (see 6.5) may be
gate threshold shift of n-channel devices becomes less
thermally-activated processes, the temperatures at which
negative, and, if interface states build up sufficiently, can
radiation, measurements, and storage take place can affect
eventually become positive. Whether p-channel gate shifts
parameter values. It is recommended that all radiation
become more or less negative with time depends on the relative
exposures, measurements, and storage be done at 24 6 6°C
rates of formation of interface states and the removal of trapped
unless another temperature range is called out specifically in
holes, but the shift always remains negative.
the test or is agreed upon between the parties to the test. If
6.5.1.4 The interaction of these competing effects that shift
devices are to be transported to and from a remote electrical
with time cause the sometimes complex time dependent
measurement site, the temperature of the devices shall not be
behavior of MOS parts following irradiation. This complex
allowed to increase by more than 10°C from the radiation-
behavior explains observed effects once thought anomalous:
environment temperature.
reverse annealing, in which parts continue to degrade with time
6.6.2 When the post irradiation electrical measurements are
following cessation of irradiation; the rebound effect, in which
made at a location remote from the radiation source the
n-channel devices super-recover past their preirradiation gate
irradiated parts may be stored at a temperature ≤60°C (using
threshold values and can fail due to a positive gate threshold
dry ice) to increase the time between the end of irradiation and
shift; dose rate effects where parts show little change at a
the beginning of electrical testing. The requirements for using
particular dose rate but show a significant response at either
this option are detailed in Section 8.
higher or lower dose rates (because at the intermediate dose
rate the net oxide-trapped charge buildup is balanced by
interface buildup); etc.
The boldface numbers in parentheses refer to the list of references at the end of
6.5.2 Time Dependent Effects for Bipolar Devices: this standard.
F1892 − 12 (2018)
6.6.3 Many device parameters are temperature sensitive. To 7. Apparatus
obtain accurate measures of the radiation-induced parameter
7.1 Radiation Sources Used for Ionizing Radiation (Total
changes, the temperature must be controlled.
Dose) Effects Testing:
6.6.4 Temperature effects also must be considered in estab-
7.1.1 Sources typically used for characterization, qualifica-
60 137
lishing the sequence of post-irradiation testing. The sequence
tion and lot acceptance testing include Co and Cs isotopes
of parameter measurements should be chosen to allow lowest
(mounted in pool sources, pop-up sources, and fully shielded
power dissipation measurements to be made first. Power
irradiators), and low energy (approximately 10 keV photon
dissipation may increase with each subsequent measurement.
energy) X-ray sources.
When high power is to be dissipated in the test devices, pulsed
7.1.1.1 Each source can be used satisfactorily for such tests,
measurements are required to minimize the temperature excur-
and the differences in the results from using different sources or
sions.
kinds of sources should be negligible provided that dose rates
can be matched or deemed to have no significant impact on the
6.7 Handling—As in any other type of testing, care must be
devices being tested.
taken in handling the parts. This applies especially to parts that
7.1.1.2 The radiation environment impinging on the tested
are susceptible to electrostatic discharge damage.
device must be characterized in terms of photon energy
6.8 Delidding—For some testing, it is necessary to de-lid
spectrum and dose rate. In situations where the photon energy
the devices prior to irradiation and testing. Care must be taken
spectrum impinging on the device is not or cannot be well
to make proper allowance for the effects of such a process.
defined, but is suspected to contain low energy components
that promote absorbed dose enhancement, a filter box such as
6.9 Radiation Damage:
the lead-aluminum structure (see 7.1.2.1 and Practice E1249)
6.9.1 If a test fixture is used over a long period of time,
can be incorporated into the radiation test environment to
components of the fixture can be damaged by exposure to the
harden the photon spectrum.
ionizing radiation, causing an impact on the test results. Such
7.1.2 The following radiation sources may be used to
fixtures should be checked regularly for socket or printed
support ionizing radiation effects testing:
circuit board leakage and for degradation of any peripheral 60
7.1.2.1 Co—The most commonly used source for ioniza-
components used in the test. Current leakage between pins or 60
tion radiation (total dose) effects testing is Co. Gamma rays
wires shall not be allowed to approach levels that interfere with
with energies of 1.17 and 1.33 MeV are the primary ionizing
60 60
accurate parameter measurements.
radiation emitted by Co (see 6.2). In exposures using Co
6.9.2 Ionizing radiation causes the introduction of color
sources, test specimens must be enclosed in a lead-aluminum
centers in optical materials, seriously degrading light transmis-
container to minimize dose enhancement effects caused by
sion properties. Much of the radiation damage to devices
low-energy scattered radiation. A minimum of 1.5 mm of lead
containing optical elements may be due to this effect rather
surrounding an inner shield of 0.7 to 1.0 mm of aluminum is
than to damage of the semiconductor elements. Such damage
required. This lead-aluminum container produces an approxi-
to the device under test or to test circuitry is outside the scope
mate charged particle equilibrium for silicon devices with
of this guide.
some attenuation of the gamma rays. Because of this
attenuation, the gamma ray intensity inside the container shall
6.10 Burn-In—Burn-in is a set of elevated-temperature bi-
be calibrated initially, whenever sources are changed, and each
ased anneals required by reliability testing and the system
time the source, container, or test fixture orientation or con-
application. For some devices, there is a significant difference
figurations are changed. This measurement shall be performed
in the radiation response before and after burn-in. Unless it has
by placing a dosimeter, for example a TLD, in the device
been shown by characterization testing that burn-in has no
irradiation container at the approximate position of the test
effect on radiation response, then either characterization and
device (see Practice E1249).
qualification testing must be performed on devices that have
137 137
7.1.2.2 Cs—Radiation sources based on Cs can be
been exposed to all elevated-temperature biased (or unbiased)
used for characterization testing in much the same way as Co
anneals required by reliability testing and the system
sources. The lead-aluminum box used for Cs testing will
application, or the results of characterization and qualification
require adjustment of the lead and aluminum thickness because
testing must be corrected for the changes in radiation response
of the lower energy of the gamma rays.
that would have been caused by elevated temperature anneals
7.1.2.3 A special case of radioactive source testing, for
(such as burn-in). This correction shall be performed in a
60 137
example, Co sources and Cs sources, is to support very
manner acceptable to the parties to the test.
low dose rate testing, that is, <1 rd/s. The use of attenuation to
6.11 Test Sample Size—There is a difficult trade-off in obtain a low dose rate, for example the use of lead bricks or
sheet, can add a significant low energy component to the
deciding the number of devices to use for a particular test.
Using a large number may in some cases be prohibitively radiation due to Compton scattering. The radiation effects of
such a softened beam may be significantly different than those
expensive. Then again, the reliability of a test result may be
unacceptably low if too small a sample size is used. This of the unattenuated beam. See Practice E1249 for additional
discussion. Special care is required to support such testing.
outcome results from part-to-part variability within a given test
lot. The sample sizes specified in this guide are accepted
7.1.2.4 Low Energy X-Ray Source—Low energy (approxi-
generally in the industry. mately 10 keV photon energy) X-ray sources commonly are
F1892 − 12 (2018)
used for transistor characterization. Because of the low pen- under test at up to 150°C while it is being irradiated. The
etration of such photons, devices must be tested prior to chamber should be capable of raising the temperature of the
packaging or be delidded for testing. For additional detail, see circuit under test from room temperature to the irradiation
Guide F1467. temperature within a reasonable time prior to irradiation and
cooling the circuit under test from the irradiation temperature
7.2 Bias Circuit—The bias circuit may be simple or
to room temperature in less than 20 minutes following irradia-
complex, depending on the part type and testing requirements.
tion. The irradiation bias shall be maintained during the heating
Good commercial design and fabrication practices should be
and cooling. The method for raising, maintaining and lowering
used to prevent oscillations, minimize leakage currents, pre-
the temperature of the circuit under test may be by conduction
vent device damage, and support accurate and repeatable
through a heat sink using heating and cooling fluids, by
measurements. For test fixtures holding several devices, isola-
convection using forced hot and cool air, or other means that
tion should be used between devices so that a failure of one
will achieve the proper results. Elevated temperature irradia-
device will not impact the other test units. For in-situ
tion is intended for use in characterizing bipolar circuits and
measurements, provision must be made for switching indi-
devices for low-dose-rate sensitivity (see 8.1.2.5).
vidual devices between the radiation bias circuit and the test
instrumentation used for pre- and post-irradiation parameter
8. Procedure
measurements. For remote measurements, MOS and bipolar
parts should be maintained with shorted leads during transport.
INTRODUCTION
7.3 Test Instrumentation—Various instruments for device
This section provides guidance for characterization testing
parameter measurement may be required. Depending upon the and for hardness assurance acceptance testing.
device to be tested, these can range from simple broadboard NOTE 5—Hardness assurance refers to part qualification and lot/process
quality conformance.
circuits to complex IC test systems. All equipment is to be in
NOTE 6—Semiconductor Devices and Integrated Circuits with Intended
calibration and of suitable stability and accuracy.
Use at Dose Rates above 300 rd (SiO )/s—For some strategic and possibly
7.4 Dosimetry System: some tactical military applications, the ionizing dose response of many
semiconductor devices can be non-monotonic with the severity of
7.4.1 Determination of Absorbed Dose—Determining the
non-monotonic behavior depending strongly on both ionizing dose and
absorbed dose in a semiconductor device requires a knowledge
dose rate. This problem can occur for ionizing dose in the prompt pulse
of the elemental composition and geometrical structure of the
resulting from a nuclear explosion. Parameters, such as leakage currents
materials involved, the appropriate tabulated mass energy-
and current gain, may reach failure levels during the pulse and return to
passing levels shortly after the pulse. The time during which the
absorption coefficients (μ /ρ), the energy spectrum of the
en
parameters are above failure level may cause system failure even though
radiation field (not merely that of the unperturbed radiation
they return to passing levels after a short period of time. Hardness
source, in which the exposure is conducted), and a related
assurance testing for these parts is discussed in Appendix X1.
measurement based on a dosimeter whose response is well
8.1 Characterization Testing—Characterization testing is
defined in the particular radiation field of interest.
performed for the purpose of part selection, determination of
7.4.2 For Co irradiation systems, dosimetry most often is
sensitivity to dose rate or time dependent effects, categoriza-
performed using thermoluminescent dosimeters (TLDs) to
tion for hardness assurance, or to determine the specific
measure the dose inside the lead-aluminum container delivered
nominal worst case test conditions for hardness assurance
in a fixed time period. Other dosimeters, such as cobalt glass,
testing.
radiochromic dye dosimeters (see ISO/ASTM 51275), or ion
8.1.1 MOS Devices and Integrated Circuits with Intended
chambers, however, can be used. This measurement is used to
Use At Dose Rates At or Below 300 rd(SiO )/s—Parts in this
establish the dose rate for the geometry used. Once the dose 2
category are those intended for use in, for example, space
rate is established, preselected radiation levels are attained by
systems, some tactical military systems, some nuclear power
irradiating for the proper time period. TLDs also may be used
plant electronics or associated robotics, and high energy
with any of the other radiation sources. Dosimeter systems can
particle accelerator detectors.
be calibrated through a service of the NIST. Proper use of
8.1.1.1 Parties to the test must first establish the conditions
TLD systems is described in Practice E668.
of the test. These conditions should be stated in a test plan as
7.5 Irradiation Temperature Chamber—Ionizing radiation
follows:
effects testing may require the use of an elevated temperature
(a) Development of the Test Plan—As a minimum, the
irradiation chamber if determined through characterization
following conditions should be specified: test approach (step-
testing. The chamber should be capable of maintaining a circuit
stress or continuous), test type (in-flux, in-situ, or remote),
irradiation source, total dose levels for electrical measurements
(for step-stress), dose rate(s), irradiation bias(es), irradiation
See, for example, Hubbell, J.H. and Seltzer, S.M. “Tables of X-Ray Mass
temperature(s), anneal bias(es), anneal temperature(s), anneal
Attenuation Coefficients and Mass Energy-Absorption Coefficients, 1 keV to 20
times, and use of test structures (where appropriate). In
MeV for Elements Z = 1 to 92 and 48 Additional Substances of Dosimetric Interest,”
addition, it may be appropriate to specify date code informa-
NISTIR 5632, May 1995. Available from Ionizing Radiation Division, Physics
Laboratory, National Institute of Standards and Technology, Technology
tion for the test devices (that is, limitations on the number of
Administration, U.S. Department of Commerce, Gaithersburg, MD 20899.
diffusion furnace lots or time to assemble date code lot, or
To schedule calibration services, contact Center for Radiation Research,
both). All of the possible interferences listed in Section 6 must
Radiation Physics Building, National Institute of Standards and Technology (NIST),
Gaithersburg, MD 20899. be considered when making these decisions.
F1892 − 12 (2018)
(b) Dose Rate—The dose rate for the test shall be selected 8.1.1.4 If the devices are being tested in-flux using the
from one of the following possibilities: continuous irradiation approach, place the devices in the
(1) Standard Dose Rate, Condition A—Unless otherwise irradiation test circuit inside the lead-aluminum shield box, if
used, and initiate the test circuit. Record the preirradiation
specified, the dose-rate range shall be between 50 and 300
parameter, or functional measurements, or both. Begin irradi-
rd(SiO )/s. The dose rates may be different for each radiation
ating the parts at the prescribed dose rate and continue to
dose level in a series; however, the dose rate shall not vary by
monitor the electrical parameters/functionality of the devices,
more than 610 % during each irradiation.
either continuously or at the prescribed time intervals, until the
(2) Condition B—As an alternative, the test may be
final dose level is reached or the parts become nonfunctional.
performed at the dose rate of the int
...