1、Designation: F 1893 98 (Reapproved 2003)Guide forMeasurement of Ionizing Dose-Rate Burnout ofSemiconductor Devices1This standard is issued under the fixed designation F 1893; the number immediately following the designation indicates the year oforiginal adoption or, in the case of revision, the year
2、 of last revision. A number in parentheses indicates the year of last reapproval. Asuperscript epsilon (e) indicates an editorial change since the last revision or reapproval.1. Scope1.1 This guide defines the detailed requirements for testingmicrocircuits for short pulse high dose-rate ionization-i
3、nducedfailure. Large flash x-ray (FXR) machines operated in thephoton mode, or FXR e-beam facilities are required because ofthe high dose-rate levels that are necessary to cause burnout.Two modes of test are possible: (1) A survival test, and (2)Afailure level test.1.2 The values stated in Internati
4、onal System of Units (SI)are to be regarded as standard. No other units of measurementare included in this standard.2. Referenced Documents2.1 ASTM Standards:2E 666 Practice for Calculating Absorbed Dose from Gammaor X-RadiationE 668 Practice for the Application of Thermoluminescence-Dosimetry (TLD)
5、 Systems for Determining Absorbed Dosein Radiation-Hardness Testing of Electronic Devices3. Terminology3.1 Definitions:3.1.1 dose rateenergy absorbed per unit time per unitmass by a given material that is exposed to the radiation field(Gy/s, rd/s).3.1.2 high dose-rate burnoutpermanent damage to asem
6、iconductor device caused by abnormally large currentsflowing in junctions and resulting in a discontinuity in thenormal current flow in the device.3.1.2.1 DiscussionThis effect strongly depends on themode of operation and bias conditions. Temperature may alsobe a factor in damage to the device shoul
7、d latchup occur priorto failure. Latchup is known to be temperature dependent.3.1.3 failure conditiona device is considered to haveundergone burnout failure if the device experiences one of thefollowing conditions.(1) functional failurea device failure where the device undertest, (DUT) fails the pre
8、-irradiation functional tests followingexposure.(2) parametric failurea device failure where the deviceunder test, DUT fails parametric measurements after exposure.3.1.3.1 DiscussionFunctional or parameteric failures maybe caused by total ionizing dose mechanisms. See interferencesfor additional dis
9、cussion.3.1.4 survival testA “pass/fail” test performed to deter-mine the status of the device after being exposed to apredetermined dose-rate level. The survival test is usuallyconsidered a destructive test.3.1.5 burnout level testa test performed to determine theactual dose-rate level where the de
10、vice experiences burnout.3.1.5.1 DiscussionIn such a test, semiconductor devicesare exposed to a series of irradiations of differing dose-ratelevels. The maximum dose rate at which the device survives isdetermined for worst-case bias conditions. The failure leveltest is always a destructive test.4.
11、Summary of Guide4.1 Semiconductor devices are tested for burnout afterexposure to high ionizing dose-rate radiation. The measure-ment for high-dose-rate burnout may be a survival test con-sisting of a pass/fail measurement at a predetermined level; orit may be a failure level test where the actual d
12、ose-rate level forburnout is determined experimentally.4.2 The following quantities are unspecified in this guideand must be agreed upon between the parties to the test:4.2.1 The maximum ionizing (total dose to which thedevices will be exposed, and4.2.2 The maximum high dose rate to which the device
13、s willbe exposed.5. Significance and Use5.1 The use of FXR radiation sources for the determinationof high dose-rate burnout in semiconductor devices is ad-dressed in this guide. The goal of this guide is to provide asystematic approach to testing for burnout.1This guide is under the jurisdiction of
14、Committee F01on Electronics , and is thedirect responsibility of Subcommittee F01.11 on Quality and Hardness Assurance.Current edition approved May 10, 1998. Published July 1998.2For referenced ASTM standards, visit the ASTM website, www.astm.org, orcontact ASTM Customer Service at serviceastm.org.
15、For Annual Book of ASTMStandards volume information, refer to the standards Document Summary page onthe ASTM website.1Copyright ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959, United States.5.2 The different type of failure modes that are possible aredefined
16、 and discussed in this guide. Specifically, failure can bedefined by a change in device parameters, or by a catastrophicfailure of the device.5.3 This guide can be used to determine the survivability ofa device, that is, that the device survives a predetermined level;or the guide can be used to dete
17、rmine the survival dose-ratecapability of the device. However, since this latter test isdestructive, the minimum dose-rate level for failure must bedetermined statistically.6. Interferences6.1 There are several interferences that need to be consid-ered when this test procedure is applied.6.2 Ionizin
18、g Dose DamageDevices may be permanentlydamaged by the accumulation of ionizing dose. This limits thenumber of radiation pulses that can be applied during burnouttesting. The ionizing dose sensitivity depends on fabricationtechniques and device technology. Metal oxide, semiconductor(MOS) devices are
19、especially sensitive to ionizing dose dam-age, however, bipolar devices with oxide-isolated sidewallsmay also be affected by low levels of ionizing dose. Themaximum ionizing total dose exposure of the test devices mustnot exceed fifty percent (50 %) of the typical ionizing dosefailure level of the s
20、pecific part type to ensure that a devicefailure is caused by burnout, and not by an ionizing total dose.6.2.1 Radiation Level Step SizeThe size of the stepsbetween successive radiation levels limits the accuracy of thedetermination of the burnout failure level.6.3 LatchupSome types of integrated ci
21、rcuits are suscep-tible to latchup during transient radiation exposure. If latchupoccurs, the device will not function correctly until power istemporarily removed and reapplied. Permanent damage (burn-out) may also occur during latchup, primarily caused by asubstantial increase in power supply curre
22、nt that leads toincreased power dissipation, localized heating, or both.Latchup is temperature dependent and testing at elevatedtemperature is required to establish worst-case operating con-ditions for latchup. Latchup testing is addressed elsewhere.6.4 Charge Build-up DamageDamage to a device mayoc
23、cur due to direct electron irradiation of the DUT leads. Whenusing direct electron irradiation of the DUT leads. When usingdirect electron irradiations, (see Section 7), all device leadsmust be shielded from the electron beam to reduce chargepickup that could cause abnormally large voltages to begen
24、erated on internal circuitry and produce damage not relatedto ionizing dose-rate burnout.6.5 Bias and Load ConditionsThe objective of the test isto determine the dose-rate survivability of the test deviceswhen tested under worst case conditions.6.5.1 Input BiasUnless otherwise specified, the input b
25、iascondition shall be chosen to provide the worst-case operatingconditions. For example, for digital devices, input pins that arein the high state should be tied directly to the supply voltage.For analog devices, input voltages generally should be at themaximum levels expected to be used. For both d
26、igital andanalog devices, it is desirable to perform the burnout test usingat least two different input conditions, such as minimum inputlevels and maximum input levels, or alternately with half theinputs tied high and the remaining tied low.6.5.2 Output LoadingUnless otherwise specified, theDUT out
27、puts shall be chosen to provide the worst-caseconditions for device operation. For digital devices, worst caseconditions should include maximum fan-out. For analog de-vices, worst-case conditions should include maximum outputvoltage or load current. For both digital and analog devices, itmay be desi
28、rable to perform the burnout test using at least twodifferent output conditions.6.5.3 Operating VoltageUnless otherwise specified, test-ing shall be performed using maximum operating voltages. Thetest setup shall be configured such that the transient powersupply photocurrent shall not be limited by
29、the external circuitresistance or lead inductance. Power supply stiffening capaci-tors shall be included to keep the power supply voltage fromvarying more than 10 % of the specified value during and afterthe radiation pulse.6.6 Over-StressThe high dose-rate burnout test should beconsidered destructi
30、ve. Peak photocurrents in excess of 2 to 3amperes can occur during these tests. These large currents canproduce localized metalization, or semiconductor melting thatis not readily detected by electrical testing, or both, but mayadversely affect device reliability. Devices that exceed manu-facturers
31、absolute limits for current or power during burnouttesting should not be used in high-reliability applications.6.7 Test TemperaturesTesting shall be performed at am-bient temperature, or at a temperature agreed upon between theparties to the test. If testing is performed in a vacuum,overheating may
32、be an issue, and temperature control isrequired.7. Apparatus7.1 GeneralThe apparatus used for testing should includeas a minimum, the radiation source, dosimetry equipment, atest circuit board, line drivers, cables and electrical instrumen-tation to measure the transient response, provide bias, andp
33、erform functional tests. Precautions shall be observed toobtain an electrical measurement system with ample shielding,satisfactory grounding, and low noise from electrical interfer-ence or from the radiation environment.7.1.1 Radiation SourceThe most appropriate radiationsource for high dose-rate bu
34、rnout testing is a FXR machine.The required dose rate for burnout cannot usually be achievedusing an electron linear accelerator (LINAC) because LINACstypically cannot produce a sufficiently high dose rate over thecritical active area of the device under test. Linear acceleratorsshall be used only w
35、ith agreement of all parties to the test.7.1.2 Flash X-ray (Photon Mode)The choice of facilitiesdepends on the available dose rate as well as other factorsincluding photon spectrum, pulse width and end-point energy.The selection of the pulse width is affected by; (a), the doserate required, and (b),
36、 the ionizing dose accumulation perpulse. Finally, the FXR end-point energy for the photon mademust be greater than 1 MeV to ensure device penetration.7.1.3 Flash X-ray (E-beam Mode)An FXR operated inthe e-beam mode generally provides a higher dose rate thansimilar machines operated in the photon mo
37、de. However,F 1893 98 (2003)2testing in the e-beam mode requires that appropriate precau-tions be taken and special test fixtures be used to ensuremeaningful results. The beam produces a large magnetic field,which may interfere with the instrumentation, and can inducelarge circulating currents in de
38、vice leads and metals. The beamalso produces air ionization, induced charge on open leads, andunwanted cable currents and voltages. E-beam testing isgenerally performed with the DUT mounted in a vacuum toreduce air ionization effects. Special dosimetry techniques arerequired to ensure proper measure
39、ment of the dose. Finally, theFXR endpoint energy must be greater than 2 MeV to ensuredevice penetration. some necessary precautions are:7.1.3.1 The electron beam must be constrained to the regionthat is to be irradiated. Support circuits and components mustbe shielded.7.1.3.2 The electron beam must
40、 be stopped within the testchamber and returned to the FXR to prevent unwanted currentsin cables and secondary radiation in the exposure room.7.1.3.3 All cables and wires must be protected from expo-sure to prevent extraneous currents. These currents may becaused by direct deposition of the beam in
41、cables, or bymagnetic coupling of the beams into the cable.7.1.3.4 All cables and cable entries must be shielded fromelectromagnetic radiation caused by the firing of the FXRmachine.7.1.3.5 An evacuated chamber for the test is required toreduce the effects of air ionization.7.2 Dosimetry EquipmentDo
42、simetry equipment shall in-clude the following:(a) a system for measuring ionizing dose, such as athermoluminescent dosimeter (TLD) or calorimeter,(b) a pulse shape monitor, and(c) a dosimeter that allows the dose rate to be determinedfrom electronic measurements, for example, a positive intrinsicne
43、gative (PIN) detector, Faraday cup, secondary emissionmonitor, or current transformer.NOTE 1PIN represents a semiconductor structure consisting of highlyP and N regions on the two sides of an intrinsic or relatively pure region.7.2.1 Thermoluminescent Detector (TLD)Exposure ofthermoluminescent detec
44、tors to ionizing radiation creates ther-moluminescent centers that when subsequently heated, emitlight. The radiant energy is proportional to the total absorbeddose in the detector. This type of detector can cover a doserange from approximately 0.1Mrd to 1Mrd (see PracticeE 668).7.2.2 CalorimeterA s
45、ilicon calorimeter system can beconstructed by attaching a thermocouple to a small (1 by 1 by0.1 mm) block of silicon. The thermocouple-silicon blockassembly is surrounded by closed-cell polyurethane foam andmounted in an aluminum housing. The aluminum provideselectron isolation and equilibration in
46、 a medium-energy photonenvironment, and the polyurethane foam provides thermalisolation. A typical thermal decay time constant for such asystem is about 3 to 4 s and typical sensitivities are about 1000to 1500 rd(Si)/V.7.2.3 PIN DiodesA PIN diode is the solid state equivalentof an ionization chamber
47、. The magnitude of photo-chargegenerated and collected in a back-biased diode is directlyproportional to the absorbed dose. Since the generation rate forsilicon is 4.3 3 109. . . rd(Si), 4.3 3 109carrier pairs/rd (Si),these devices can be calibrated knowing only the detectorgeometry. Calibration dep
48、ends on the PIN bias and may changewith accumulated exposure. Most PIN diodes have a linearresponse up to a dose rate of approximately 1 3 1010rd(Si)/s.(WarningCare must be taken when using PIN diodes toensure that the indicated PIN dose rate is equivalent to thatabsorbed by the DUT. Factors that ca
49、n affect dosimetry includethe FXR photon spectrum, the method used to calibrate the PINdiode, and the location of the PIN diode relative to the DUT.)7.2.4 Opti-chromic DosimetersOpti-chromic dosimetershave many of the same advantages as TLDs. These devices arerelatively small, passive, inexpensive, and retain accurate doseinformation for months between irradiation and measurementof dose. The useful dose range of these devices is 400rd(Si) to20Krd(Si). The device response is nearly linear with dose.Opti-chromic dosimeters are calibrated in a Co60cell usi
