1、 JEDEC PUBLICATION Failure Mechanisms and Models for Semiconductor Devices JEP122H (Revision of JEP122G, October 2011) SEPTEMBER 2016 JEDEC SOLID STATE TECHNOLOGY ASSOCIATION NOTICE JEDEC standards and publications contain material that has been prepared, reviewed, and approved through the JEDEC Boa
2、rd of Directors level and subsequently reviewed and approved by the JEDEC legal counsel. JEDEC standards and publications are designed to serve the public interest through eliminating misunderstandings between manufacturers and purchasers, facilitating interchangeability and improvement of products,
3、 and assisting the purchaser in selecting and obtaining with minimum delay the proper product for use by those other than JEDEC members, whether the standard is to be used either domestically or internationally. JEDEC standards and publications are adopted without regard to whether or not their adop
4、tion may involve patents or articles, materials, or processes. By such action JEDEC does not assume any liability to any patent owner, nor does it assume any obligation whatever to parties adopting the JEDEC standards or publications. The information included in JEDEC standards and publications repr
5、esents a sound approach to product specification and application, principally from the solid state device manufacturer viewpoint. Within the JEDEC organization there are procedures whereby a JEDEC standard or publication may be further processed and ultimately become an ANSI standard. No claims to b
6、e in conformance with this standard may be made unless all requirements stated in the standard are met. Inquiries, comments, and suggestions relative to the content of this JEDEC standard or publication should be addressed to JEDEC at the address below, or refer to www.jedec.org under Standards and
7、Documents for alternative contact information. Published by JEDEC Solid State Technology Association 2016 3103 North 10th Street Suite 240 South Arlington, VA 22201-2107 This document may be downloaded free of charge; however JEDEC retains the copyright on this material. By downloading this file the
8、 individual agrees not to charge for or resell the resulting material. PRICE: Contact JEDEC Printed in the U.S.A. All rights reserved PLEASE! DONT VIOLATE THE LAW! This document is copyrighted by JEDEC and may not be reproduced without permission. For information, contact: JEDEC Solid State Technolo
9、gy Association 3103 North 10th Street Suite 240 South Arlington, VA 22201-2107 or refer to www.jedec.org under Standards-Documents/Copyright Information. JEDEC Publication No. 122H -i- FAILURE MECHANISMS AND MODELS FOR SEMICONDUCTOR DEVICES Contents Page Foreword iii Introduction iii 1 Scope 1 2 Ter
10、ms and definitions 1 3 Inclusions, deliberate omissions, and resources 5 4 The basic thermal acceleration equation 9 5 Models for common failure mechanisms 9 FEoL Failure Mechanisms 5.1 Time-Dependent Dielectric Breakdown (TDDB) gate oxide 9 5.2 Hot Carrier Injection (HCI) 14 5.3 Negative Bias Tempe
11、rature Instability (NBTI) 17 5.4 Surface inversion (mobile ions) 19 5.5 Floating-Gate Nonvolatile Memory Data Retention 21 5.6 Localized Charge Trapping Nonvolatile Memory Data Retention 29 5.7 Phase Change (PCM) Nonvolatile Memory Data Retention 31 BEoL Failure Mechanisms 5.8 Time-Dependent Dielect
12、ric Breakdown (TDDB) ILD/Low-k/Mobile Cu ion 34 5.9 Aluminum Electromigration (Al EM) 43 5.10 Copper Electromigration (Cu EM) 46 5.11 Aluminum and Copper Corrosion 48 5.12 Aluminum Stress Migration (Al SM) 53 5.13 Copper Stress Migration (Cu SM) 55 Packaging/Interfacial Failure Mechanisms 5.14 Fatig
13、ue failure due to temperature cycling and thermal shock 58 5.15 Interfacial failure due to temperature cycling and thermal shock 63 5.16 Intermetallic and oxidation failure due to high temperature 66 5.17 Tin Whiskers 68 5.18 Ionic Mobility Kinetics (PCB) Component Cleanliness 72 Statistics and Mode
14、ling Parameter Determination 5.19 Reliability data/analysis 75 5.20 Design of Experiments (DOE) for determination of modeling parameters 80 6 Activation energies and modeling factors 82 Annexes Annex A List of references 87 Annex B Differences between JEP122H and JEP122G 103 JEDEC Publication No. 12
15、2H -ii- FAILURE MECHANISMS AND MODELS FOR SEMICONDUCTOR DEVICES Contents Figures Page 5.1-1 Photograph of TDDB breakdown in a gate oxide mid-gate 13 5.5-1 (a) Example of failure mechanisms scenario affecting VT during data retention (from 5.5.16), and (b) extraction of Eaa for each mechanism (from 5
16、.5.15). 22 5.5-2 (a) Spectrum of detrapping time constants immediatly after cycling (black curve) and during data retention (red curves), and (b) Resulting VT(tR) transient. 25 5.5-3 (a) Comparison of time constant spectra between uniform cycling of duration tcycand an equivalent cycling where all t
17、he delays are lumped in a single wait of duration A tcycprior to the bake phase, and (b) Resulting VT(tR) transients. A = 0.2 results in similar VTloss during data retention 27 5.5-4 Extrapolation of SILC bit error rate 28 5.8-1 Time-Dependent Dielectric Breakdown (TDDB) in various dielectrics 35 5.
18、8-2 Metal stack cross section/schematic 37 5.8-3 Normal distribution of breakdown voltage 38 5.8-4 Copper short/extrusion 39 5.9-1 Examples of Aluminum Electromigration 45 5.10-1 Examples of Copper Electromigration 48 5.11-1 Aluminum bond pad corrosion 52 5.11-2 Electrochemical reaction 52 5.11-3 Co
19、rrosion rate versus surface mobility 52 5.12-1 Examples of Aluminum Stress Migration 55 5.13-1 Examples of Copper Stress Migration 57 5.14-1 Examples of temperature cycling/thermal shock damage 62 5.15-1 Example of interfacial delamination after temperature cycling 65 5.17-1 SEM of Tin Whiskers on M
20、atte Tin plated Alloy 42 leads 5.16.4 71 5.17-2 Optical Image of a Tin Whisker growing from Relay lead to case 5.16.3 71 5.17-3 FIB - matte tin whisker structure from a temperature cycled specimen 5.16.5 71 5.17-4 Optical Image - Tin Whisker growing from SAC 305 solder over Alloy 42 - matte tin 71 5
21、.17-5 Optical image of Tin Whisker on a copper coupon with matte tin plating 5.16.4 71 5.17-6 8 mm long Tin Whisker growing from a bracket holding electronics in a frame. 71 5.17-7 Tin Whisker breaking through 10 m Uralane 5750 coating (9 yr office storage) 5.16.3 71 5.18-1 Resistor corroded open du
22、e to trapped MSA residues in epoxy surface 3 73 5.18-2 Electrochemical migration between leads on a QFP 73 5.18-3 Leakage and corrosion problems with residues on tinned leads due to aggressive flux 74 5.18-4 Leakage square brackets enclose citation numbers. All equation, citation, and figure numbers
23、 include the subclause number so that individual clauses can be modified without disturbing other clauses, except for page numbers. Thus, (5.3.2) is the 2nd equation in 5.3 and 5.11.5 is the 5th citation in 5.11. The citations can be found in Annex A. JEDEC Publication No. 122H -iv- JEDEC Publicatio
24、n No. 122G Page 1 FAILURE MECHANISMS AND MODELS FOR SEMICONDUCTOR DEVICES (From JEDEC Board Ballot JCB-01-97, JCB-03-39, JCB-08-61, JCB-09-19, JCB-10-64, JCB-11-74, and JCB-16-32, formulated under the cognizance of JC-14.1 Subcommittee on Reliability Test Methods for Packaged Devices.) 1 Scope This
25、publication provides a list of failure mechanisms and their associated activation energies or acceleration factors that may be used in making system failure rate estimations when the only available data is based on tests performed at accelerated stress test conditions. The method to be used is the S
26、um-of-the-Failure-Rates method. The models apply primarily to the following: a) Aluminum (doped with small amounts of Cu and/or Si) and copper alloy metallization b) Refractory metal barrier metals with thin anti-reflection coatings c) Doped silica or silicon nitride interlayer dielectrics, includin
27、g low-dielectric-constant materials d) Poly silicon or “salicide” gates (metal-rich silicides such as W, Ni k is Boltzmanns constant (8.62 105eV/K); T1is the absolute temperature of test 1 (K); T2is the absolute temperature of test 2 (K); T1is the observed failure rate at test temperature T1 (h-1);
28、T2is the observed failure rate at test temperature T2 (h-1). NOTE 2 The best-fit linear slope of a plot of the natural log of the time-to-failure as a function of 1/kT, the reciprocal of the product of Boltzmanns constant in electronvolts per kelvin and the absolute temperature in kelvins, is equal
29、to the apparent activation energy in electronvolts. NOTE 3 q= o . AT, where qis the quoted (predicted) system failure rate at some system temperature Ts, ois the observed failure rate at some test temperature Tt, and ATis the temperature acceleration factor from Ttto Ts. activation energy (Ea): The
30、excess free energy over the ground state that must be acquired by an atomic or molecular system in order that a particular process can occur. NOTE The activation energy is used in the Arrhenius equation for the thermal acceleration of physical reactions. The term “activation energy” is not applicabl
31、e when describing thermal acceleration of time-to-failure distributions, e.g., in the Arrhenius equation for reliability; hence the need for the term “apparent activation energy”. apparent activation energy (Eaa): An energy value, analogous to activation energy, that can be inserted in the Arrhenius
32、 equation for reliability to calculate an acceleration factor applicable to changes with temperature of time-to-failure distributions. NOTE 1 An apparent activation energy should be associated with a specific failure mechanism and an observed time-to-failure distribution to calculate the acceleratio
33、n factor for converting the observed failure rate to the quoted failure rate at a different temperature. NOTE 2 An activation energy is a measure of the heat energy needed to establish the rate of reaction for a specific failure mechanism. The reaction rate and other contributing factors, e.g., radi
34、ation, voltage, humidity, magnetic fields, determine the unique time-to-failure distribution for the modeled failure mechanism. NOTE 3 The apparent activation energy is empirically determined from the change in an observed time-to-failure distribution with temperature. bathtub curve: A plot of failu
35、re rate versus time or cycles that exhibits three phases of life: infant mortality (decreasing failure rate), intrinsic or useful life (relatively constant failure rate), and wear-out (increasing failure rate). Boltzmanns constant (k): A constant equal to 1.38 x 1023joule per kelvin or 8.62 x 105ele
36、ctronvolt per kelvin. JEDEC Publication No. 122G Page 3 2 Terms and definitions (contd) cumulative distribution function of the time-to-failure; cumulative mortality function F(t): The probability that a device will have failed by a specified time t1,or the fraction of units that have failed by that
37、 time. NOTE 1 The value of this function is given by the integral of f(t) from t = 0 to t = t1 and is generally expressed in percent (%) or in parts per million (ppm) for a defined early-life failure period. See “probability density functionof the time-to-failure” for f(t). NOTE 2 The abbreviation C
38、DF is often used; however, the symbol F(t) is preferred. cumulative hazard function H(t): The fraction of units that have failed referenced to the survivors (not to the initial number of units). NOTE The value of this function at a specified time t1 is given by the integral of h(t) from t = 0 to t =
39、 t1.See “instantaneous failure rate; hazard rate” for h(t). cumulative reliability function R(t): The probability that a device will still be functional at a specified time t1,or the fraction of units surviving to that time. NOTE R(t) = 1 F(t). See “cumulative distribution function of the time-to-fa
40、ilure” for F(t). failure mechanism: The physical, chemical, electrical, or other process that has led to a nonconformance. NOTE 1 See JESD671, Component Quality Problem Analysis and Corrective Action Requirements. NOTE 2 A failure mechanism may be characterized by how a degradation process proceeds
41、including the driving force, e.g., oxidation, diffusion, electric field, current density. When the driving force is known, a mechanism may be described by an explicit failure rate model; identifying that model with associated parameters is the main objective of this document. failure mode: (general)
42、 The way in which a failure mechanism manifests itself in a failing component. NOTE 1 Examples of failure modes are a visual blemish, a bent lead, a foreign particle or material, an incorrect dopant profile or grain size, a scratch, an electrical fault (open, short, leakage, inadequate slew rate or
43、noise margin, stuck at high or low, etc.). NOTE 2 Failure rate distributions for a given failure mode can be modeled only when the failure mechanism and the relevant independent variables (forcing functions) are known. failure rate (): The fraction of a population that fails within a specified inter
44、val, divided by that interval. NOTE 1 Standard methods of reporting failure rates of semiconductor devices include 1) percent failed per 1000 hours and 2) FITs. NOTE 2 The interval may be expressed in operating hours, storage hours, operating cycles, or other units of interval measurement. NOTE 3 Ty
45、pically, the term “failure rate” means the instantaneous failure (hazard) rate. NOTE 4 The statistical upper limit estimate of the failure rate is usually calculated using the (chi-squared) function. failures in time (FITs): The number of failures per 109device hours. JEDEC Publication No. 122G Page
46、 4 2 Terms and definitions (contd) instantaneous failure rate; hazard rate h(t): The rate at which devices are failing referenced to the survivors (not to the initial number of units). NOTE h(t) = f(t)/R(t). See “probability density function of the time-to-failure” for f(t) and “cumulative reliabili
47、ty function” for R(t). mean-time-between-failures (MTBF): The average time between failures in repairable or redundant systems. mean-time-to-failure (MTTF): The average time to failure for components or nonrepairable systems. NOTE The MTTF is often the reciprocal of the hazard rate when the hazard r
48、ate is described by the Poisson or equivalent exponential function, e.g., in the constant or flat portion of the useful life region of the bathtub curve. whisker: A spontaneous columnar or cylindrical filament, usually of monocrystalline metal, emanating from the surface of a finish. NOTE Whiskers a
49、re not to be confused with dendrites, which are fern-like growths on the surface of a material, formed as a result of electromigration of an ionic species or during solidification. observed failure rate: The failure rate determined from a product or test vehicle subjected to an accelerating stress that may produce failures attributable to one or more failure mechanisms. PCB component cleanliness: The absence of chemical residues, on or in the surfac
