This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

Failure rate vs Humidity

Hi guys

My question is a general one and related to humidity influencing on integrated circuits. I have read in one scientific article that since silicon nitride passivation is used at manufacturing integrated circuits humidity hasn't been influencing factor for failure rate anymore. Exception is of course bond pad corrosion where humidity is still very influencing. I'm asking this as we have two basic models for accelerating tests - Hallberg-Peck and Arrhenius one. Which one should be preferably used for QC of integrated circuits? If the first why you don't have humidity involved in your temperature failure rate estimator's equation then? Thanks for your answer in advance!

Kr,

Primoz

  • Hello Primoz, 

    Thank you for posting this question on E2E.  While we are still researching available data to provide a reply, I invite you to review this publication as it has some basic information on this topic: 

    https://www.ti.com/lit/wp/slpa019/slpa019.pdf 

    We will get back to you with more information within the week.  

    Regards,

    ~Leonard 

  • Primoz: 

    Internal corrosion in the presence of moisture is evaluated as part of it's qualification where both unbiased and biased accelerated humidity qualification tests are run. You are right that this has a role in assessing durability of the IC and package interconnection. The package material set (For example, Mold compound, wire type, Bond pad for example) has a role in surviving those tests. The corrosion it looks for are bulk corrosion, galvanic corrosion and ionic corrosion.  

    The semiconductor industry uses the Halberg-peck model for their modelling of humidity. See Q100 Rev. J (aecouncil.com) page 42 as an example of how they apply it. It does have an Arrhenius factor to it as well as a Relative humidity acceleration factor. 

    The FIT values on ti.com is derating HTOL results and it method per JESD85 "Methods for Calculating Failure Rates in Units of FITs".  That is a simple Arrhenius model where the industry practice had assumed an Eaa of 0.7eV. 

    As to your question about moisture and failure rates, natural moisture ingress into an IC at normal ambient temperatures is not a specific intrinsic reliability concern for the electronics industry that needs a specific FIT budget.  

    I would also add that many electronic manufacturers add protections such as coating over the PCB electronics, for systems intended to be used in a harsh moisture environment. That is there since high humidity can result in condensation which can result in electrical shorts. TI has no recommendations on the use of such external coatings and the customer would be responsible for evaluating any interaction of the coating with the IC. 

  • I should have also added the humidity stress tests during a semiconductor qualification uses DI water.  

  • Allan, many thanks for the info. I've got no further questions.