Many in the 3-D IC arena have been calling for some standardization over the last few years to help speed things along. [ see PFTLE “Quotes from the Summit“, 08/09/2009 ] . Standardization is a very important activity for a new technology to become accepted and “usable” in any given infrastructure. We have previously commented on the activity of the 3D Alliance [ see PFTLE, “Recent Activity on 3-D IC Integration”, 07/27/2008 ]
There are many societies and associations involved with standardization activities around the world. One such organization, that we all look to for standards, is JEDEC. JEDEC is an independent association which facilitates standardization within the solid state technology industry. The mission of JEDEC is to “.. create, publish, and promote the global acceptance of standards, and provide a forum for technical exchange on leading industry topics”. JEDEC standards and publications are “…designed to serve the public interest through eliminating misunderstandings between manufacturers and purchasers, facilitating interchangeability and improvement of products, and assisting the purchaser in selecting and obtaining, with minimum delay, the proper product …”
Any company, organization, or individual conducting business in electronic equipment or electronics-related products is eligible for membership in JEDEC and the JEDEC committees.
JEDEC begins to tackle 3D IC
When I became aware that JEDEC had in fact issued a “document” on 3D IC [ see “3D Interconnect Shaping Future Solutions“, Semi Int, 02/16/2010 ] , I decided to delve further, to make you the reader aware of both the content and the authorship of this document. I stress the point of authorship because, to PFTLE, it is important to know exactly who (company and person ) was involved with preparing the document, especially if it is to become a standard. JEDEC, to their credit, suggests that ” The standards development process should have a balance of interests. Participants from diverse interest categories shall be sought with the objective of achieving balance.” PFTLE believes that we should know exactly who was behind writing all documents – it’s called transparency.
PFTLE was able to make contact with JEDEC and the Chairman of the responsible committee (JC -14.3 see below) but unfortunately the JEDEC position is to not release information about who specifically was involved with drawing up any of their documents (companies or individuals). PFTLE is aware of at least three of the individuals / companies involved but will respect JEDEC’s position and not reveal them here. PFTLE strongly disagrees with this position because this lack of transparency can only weaken the authority of any document, standard, rule or regulation.
JEDEC Committees and Publications
JC-14 is responsible for standardizing “quality and reliability methodologies for solid state products”. The committee is comprised of both suppliers and users. Sub committee JC-14.3 handles “Silicon Devices Reliability Qualification and Monitoring”. There are also JEDEC committees on mechanical package outlines, electrical and thermal characterization, digital logic, RAM memory, memory modules, flash memory and many other topics. 3D IC will certainly require standardization activity in all of these areas and more. In fact PFTLE has heard unsubstantiated rumors that JEDEC is in fact tackling the stacked memory area as we speak.
There are several classifications of documents at JEDEC . JEP 158 is actually a JEDEC “publication” not a JEDEC standard (which would have a “JESD” label). Documents are defined as “containing general engineering information on products, procedures, or services, that are not necessarily appropriate for standardization”
So with that long introduction, lets take a look at what is covered here.
This document can be downloaded free of charge from the JEDEC web page:
[ http://www.jedec.org/standards-documents/results/jep-158 ] and I recommend that you do that.
JEP-158 3D Chip Stack with Through-Silicon Vias (TSVS): Identifying, Evaluating and Understanding Reliability Interactions
This publication is intended as “..a guideline to describe the extension of standard tests to three-dimensional (3D) chip structures that contain stacks of two or more chips that use through-silicon vias to connect from the front side to the back side of each chip.” The main element of this extension is the addition of appropriate test structures to evaluate the reliability of the TSVs and other new features introduced in the fabrication of 3D products. This publication applies to vias-first process (TSV formation before completion of the silicon device fabrication), vias-middle (TSV formed in the BEOL or prior to the BEOL process), and a vias-last process (TSV formation after completion of the silicon device fabrication).
Terms and definitions
Good to see that JEDEC has accepted the “vias middle” terminology:
”via-middle: A TSV formed after FEOL processing and prior to or during the BEOL process. NOTE A “via-middle” process is sometimes considered to be part of a “via-first” process.”
Overview of TSV Chip Stack Manufacturing and Reliability
TSV related backside processing
Listed precautions include:
“thinning may adversely affect certain specific transistor designs that need to be addressed at the product level….. Thinning down to 25-50 μm may lead to additional warpage related stress to devices….backside processing may lead to leakage to substrate as well as mobile ion diffusion…Mechanical stresses, which can scale with size, and defects introduced by the far back end of the wafer line and bond and assembly processes, can only be addressed through representative test structure design and processing, followed by stress testing and appropriate electrical and physical characterization”
“Stresses are generated in the Si by the presence of the TSV, due to the differences in CTE between the conductive material of the TSV and that of the Si, and due to the geometry of the TSV……. Some field-effect transistors (FETs) are sensitive to stress, such that the FET electrical characteristics can vary with distance to a nearby TSV…… The connection schemes between the TSVs and chip wiring provide additional opportunities for failure…… the processing needed to fabricate [ the TSV] , the fabrication steps.. thermal excursions can have deleterious effects on small wiring features and on features in low-k dielectrics. Thus an additional set of reliability structures are needed…”
3D TSV Failure Modes
“TSVs can also be a source of a reliability failure……Two types of defects can cause reliability fails in this system. .. if the insulator surrounding the TSV is not continuous or has a defect, it can break down and allow leakage between the TSV conductor and the bulk silicon of the die. …. A second potential TSV problem is a void in the conductive material that makes up a via, or that connects to it. The void can grow over time and cause an open.”
TSV failure concerns………..
“TSVs connect both to other strata and to elements within a stratum. All of these interfaces may be somewhat different from standard packaging or chip fabrication processes and should be tested for reliability. All bonds between strata must be evaluated, especially those located at large distances-to-neutral points (DNP)…. The same interfaces are also vulnerable to electromigration failure, if not properly designed… The integrity of all of these interfaces can be affected by … whether the TSV is fabricated before the rest of the chip (stratum) is built, during the sequence of steps used to build the rest of the chip, or after the chip is built……. Individual test structures should be placed at multiple locations across the chip, and… at varying distances from TSVs”
3-D chip stack TSV test guidelines
Failure modes and detection methods are summarized in the table shown below.
Things to be documented and examined in a failure analysis are also discussed.
…the interesting footnote
Before we conclude, lets get back to the authorship issue one more time. An interesting comment that is footnoted on all the pages of this document is “Compliance with this section of the document may require requires U.S. Patent Applications No. 11/351,418 and 11/593,788. Users are advised to assess exposure to patent rights in applying this publication.” Certainly anyone who has looked at 3-D IC from a patent standpoint knows that there is a mine field out there especially due to nomenclature (everyone has had their own nomenclature for 3-D IC through the years). I’ll let you take a look at these specific patent applications yourself, if you choose to. Let me just say that of all the 3-D IC related patents I have personally come across, these were certainly not two of the ones I’d be most concerned about (NOTE – this is not a legal opinion, just my personal opinion). I then noticed who the patents were assigned to and found that that assigned company was indeed a JEDEC member. Sure would be nice to know whether they had a representative on the 14.3 committee drafting this document…..know what I mean ?
Conclusions on JEDEC JEP-158
PFTLE is very pleased with the content of this JEDEC document. This is a very good read for all those involved with or considering becoming involved with 3-D IC. It is hoped that JEDEC continues this work and helps develop the necessary standards for 3-D IC. It is also hoped that they reconsider their position on divulging authorship !
For all the latest on 3-D IC and advanced packaging technology stay linked toPFTLE……………………..