Connect with us

Gamma irradiation of components - results

Discussion in 'Electronic Design' started by Richard Rasker, Apr 22, 2004.

Scroll to continue with content
  1. Hi all,

    Some weeks ago, I posted a few questions about gamma sterilization of
    certain electronic components. From the reactions, the conclusion could be
    drawn that gamma sterlization of electronics is quite a tricky business.
    This also became obvious when an irradiated test circuit showed
    multiple component failures.

    Nevertheless, I decided to put a small batch of components through the
    gamma irradiation process again, and an identical batch through steam
    sterilization, for comparison purposes. Most of the selected components
    are included in the design of a medical implant. Here are the results, for
    those interested.

    Components tested (one batch):
    FDC6305N Dual N-channel MOSFET (4x)
    FDC6306P Dual P-channel MOSFET (4x)
    FDC6327C N- & P-channel MOSFET (4x)
    TL084 quad opamp (1x)
    TC1047A temperature-to-voltage converter (2x)

    Sterilization procedures:
    - Autoclave (steam), 134 degrees Centigrade at 2 bars of pressure
    - Gamma irradiation, 1 dose of 25 kGray Co60 exposure

    Steam: All components were unaffected, and behaved well within their

    Gamma: All MOSFET's were severely affected:
    FDC6305N - all 8 devices exhibited short circuit from drain to source; no
    response to Vgs at all.

    FDC6306P - all devices exhibited dramatically lower Vgs threshold values,
    with a Vgs-on between -2.1V and -2.5V (normally -1V), and a maximum drain
    current Id-max (set at 0.33 A) at Vgs = approx. -3.2V (normally -1.6V).

    FDC6327C - P-FET's: same as FDC6306P.
    N-FET's: lower Vgs values: Vgs-on between -0.2V and -0.4V
    (normally +1.1V) and Id-max at Vgs = +0.2V (normally +1.6V)

    Something rather unexpected is the difference between the N-FET's in the
    FDC6327C (which merely showed lower Vgs values), and the N-FET's FDC6305N,
    which were completely destroyed. All P-FET's behaved in the same manner.
    Apart from the changed Vgs values, other parameters seemed mostly

    TL084 - tested in unity-gain circuit (output connected to - input, driving
    voltage on + input). Two of the four opamps exhibited a massive offset
    voltage of approximately +0.5V, probably as a result of damage to the
    input JFET's; the other two devices in the same package seemed unaffected.

    TC1047A - both devices tested seemed to operate within specifications;
    however, the small number of components tested makes it difficult to say
    anything definitive about the sensitivity for gamma radiation.

    Conclusion: gamma sterilization of electronic circuits is quite
    troublesome; especially FET's (both MOSFET's and JFET's, such as
    employed in the TL084) are at risk. The most prominent effect is a
    drastically lower value for Vgs, rendering the devices near useless for
    most practical purposes.

    If any more gamma ray testing is carried out, e.g. with bipolar
    semiconductors instead of FET's, I will post the results here as well.

    Richard Rasker
  2. Joerg

    Joerg Guest

    Hi Richard,

    Thanks, that was pretty interesting. Same experience here, radiation had quite
    disastrous effects whenever someone insisted that we test it. Most everything
    remotely related to semiconductors was toast afterwards. But even some
    plastics were affected in ways such as changing their color tint.

    With steam, do you mean EtO? That's what we generally use.


  3. Nevertheless, the people I work for still want to look into radiation
    hardened semiconductors. I don't think that much will come of it - if I
    understand correctly. 'radiation hardened' means a much larger
    electronically active area, and therefore larger components as a whole. I
    selected the FDC type MOSFETs especially for their combination of
    very small size, high current handling capabilities and low drive voltage,
    and I think that even if rad-resistant types exist, these will almost
    certainly be much bigger ...
    No, I really mean steam, heated water, at 134°C (273°F, 407K), under a
    pressure of 2 bar (about 30 PSI), in an autoclave - standard hospital
    issue. As said, this didn't harm the components in any way. And as long as
    I don't use electrolytical capacitors, I expect no problems with this
    method. We didn't test EtO; based on the experiences of others, this seems
    a very well-proven method for electronics, so rigorous testing in advance
    of this method is not really necessary.


    Richard Rasker
  4. Joerg

    Joerg Guest

    Hi Richard,

    Rad hard parts are very expensive and the selection you can pick from becomes
    quite limited. Some degradation will still occur since they are hard but not

    Electrolytics aren't the only problem. Switches, buzzers, display panels could
    also be affected in an autoclave process especially if it is done repeatedly.

    I don't know whether you meant EtO to be so well proven that it wouldn't need
    advance testing. My experince is that it requires lots of validation and
    testing. EtO can have quite nasty corrosive and other effects on contacts,
    glues parts with soft materials and so on. Then there is the outgassing phase
    which needs to be extremely carefully validated.

    Regards, Joerg
  5. Hallo Joerg,
    Thank you for your much valued remarks; I figured so much already, and I'm
    not overly confident that gamma irradiation is a feasible option in this
    case - to put it mildly.
    That won't be much of a problem, since the circuit contains no
    electromechanical or otherwise "special" parts. It's all semiconductors
    and standard passive components (resistors, capacitors, inductors).
    Sorry, I wasn't quite clear here. Of course several (semi-)complete
    prototype implants will be thoroughly tested in all respects, probably
    including EtO sterilization. But I don't expect EtO-testing of a handful
    of separate off-the-shelf components to yield much in the way of valuable
    or unexpected information, especially not in relation to their electrical
    With gamma irradiation, I wanted to make sure that indeed the majority of
    components would suffer damage - which was established in a rather
    convincing way. The steam procedure was chosen as an alternative, to see
    whether prolonged exposure to temperatures close to the absolute maximum
    ratings would have any measurable effects - which was not the case.

    However, any remarks, thoughts, or other bits of information are still
    very welcome.

    Richard Rasker
  6. Joerg

    Joerg Guest

    Hi Richard,

    The autoclave process isn't so familiar to me since everything I ever designed
    for med goes through EtO. But one thing comes to mind and that is cabling. If
    your design has wires with PVC insulation the high temps might be an issue.
    But that can easily be solved by changing to another type of insulation.

    Another concern would be chips. An autoclave may cause a chip with a damaged
    package to fail which then may not be testable because you can't touch sterile
    goods. Unless it had already been autoclaved some time before final test this
    could be a problem.

    I looked up your web link. Reminds me of the good old times back in NL where
    Grolsch was my standard beer. Even visited the brewery in your town. Speaking
    Dutch got rusty but I could still read the profiles page on your web site.

    Regards, Joerg

  7. Hi there Mr. Rasker.

    Interesting post, but is 25 kGy really necessary? I don't think I would
    want to be exposed to that being that (IIRC) the LD 99/60 (lethal dose for
    99% of the population within 60 days of exposure) for mid-line absorbed dose
    (mid-line being average exposure over the abdominal region) for people is
    around 5 Gy for X-ray or gamma radiation. Of course some particularly
    robust organisms can handle allot more than people, but gee...
  8. Well, I was a bit surprised at the dosage as well - especially when
    confronted with the extent of damage done to something as relatively
    simple as a semiconductor device.
    But then again, this seems to be the standard dose for medical implants -
    my guess is that they want to make absolutely certain that nothing even
    remotely protein-like will survive the irradiation unchanged. And of
    course one doesn't just have to worry about bacteria and viruses, but also
    about prions (BSE and scrapie pathogens), which are much more robust. Come
    to think of that, I wonder if the autoclave treatment takes care of that...

    Richard Rasker
  9. OK, thanks, point duly noted :)
    Yes, I can imagine water vapour being able to creep in; plans are, however
    to encase the electronics in resin or seal them off in some other way.
    We're very much in the early stages of solving these particular problems
    with sterilization, so all information is appreciated.
    Hehe, well, Grolsch is as fine a lager as you can wich for - still my
    favourite as well. And if you can't follow the Dutch, there's always the
    English (and German) version.

    Richard Rasker
  10. Joerg

    Joerg Guest

    Hi Richard,

    Grolsch beer is even available here in the US but it doesn't taste the same as
    in NL. Maybe it doesn't like to travel much. It comes in the same flip top
    bottles (beugelfles in Dutch?). But we now have many local micro breweries
    that make great beers of all types. A lot has changed for the better here in
    the brewing scene. We can even go to a local pub for a refill into a 2 liter
    glass bottle. Like in them good ol' times. Talking about sterilization, these
    bottles that are called growlers over here need to be meticulously sterilized.

    Once in a while we treat ourselves to a nice Belgian Saison Regal. People here
    watch in amazement when you pull out the cork with a loud pop and a cloud
    comes out of the top.

    Groetjes, Joerg
  11. Joerg

    Joerg Guest

    Yes, Fritz, as Richard said this process is meant to eradicate anything that
    lives, plus a healthy dosage margin on top of that. So you can't compare it to
    danger levels for humans. There will be no people present when the dosage is

    In sterilization validation there are certain bacteria "planted' in your device
    at places where it is assumed that the process has the least effect. Later it is
    measured whether any of those are alive and at which dosage. This is quite a
    labor intensive process.

    The same goes for other processes such as EtO. The quantities administered would
    definitely not be survivable by a human or any other living being. Even during
    outgassing the products are quarantined for several days. Which is why I keep a
    very safe distance whenever I see a tanker truck with EtO on the highway.

    Regards, Joerg
  12. Yup, that's the name, and this particular bottle is still a very strong
    marketing icon. Grolsch decided to keep it, where most other breweries
    have fully switched to identical, uniform "euro" bottles. Grolsch does
    however sell its beer in these standard, lightweight bottles as well.
    Indeed, here as well, the "beerscape" has become increasingly varied over
    the past 20 years or so. In the late 1970's, one could choose from at most
    five or six brands of beer (and all lager) in the supermarket, and maybe
    double this figure in liquor stores. Nowadays, the choice is overwhelming,
    with some stores offering up to 50 or 60 brands of beer :)
    Hmm, never tried one of those, but I'm looking forward to the moment when
    I'll descend on a sunny terrace, and order my first wonderfully
    refreshing Hoegaarden Witbier ("white beer") of this summer season - after
    an electronics job well done of course, that's when they taste best :)


    Richard Rasker
  13. Ken Taylor

    Ken Taylor Guest

    Normally I'd say that wasn't possible, as the levels used aren't enough.
    However that was valid for baggage screening - container-level screening may
    involve much higher doses. Having said that, we haven't had similar problems
    and we're shipping stuff in and out of the US all the time via container.

  14. Mike

    Mike Guest

    I don't know how Maxtor, Seagate, and WD import their drives, but I'd guess
    that they're all being shipped in containers...

    -- Mike --
  15. Rob

    Rob Guest

    Has anyone experienced hard disks being corrupted by customs xray
    inspection? Equipment my employer exports often arrives with the hard disks
    unreadable. The HDDs are part of a system built into a sea container. The
    problem has been attributed to US customs xraying the sea container. It only
    seems to happen when the equipment goes into the US. There have been 3 or 4
    instances of this occuring.

    Any thoughts/similar experiences?

  16. R.Legg

    R.Legg Guest

    A permanent part of the shipping container? Re-used?

    Rad effects are cumulative.

  17. Rich Webb

    Rich Webb Guest

    Are the components operational with the exception of bad/scrambled data
    on the hard disk? If so, rather than the X-Ray flux, the culprit may be
    the focusing coils.
  18. Rob

    Rob Guest

    Thanks for all the replies.

    You make a good point Rich, all the electronics functions fine, so perhaps
    it is magnetic rather than xray based.

    To clarify a couple of the replies;

    The equipment is physically built into a modified sea container, which is
    then treated like any other container for shipping purposes. We believe the
    container itself is being xrayed to inspect its contents.

    The displays/monitors used in the product are all LCD based - perhaps if
    they were CRTs we would see some effect on the CRT display purity - if the
    power on degauss didn't clean it up first.....
    I think pan is to ship the hard drives as hand luggage with the installation

Ask a Question
Want to reply to this thread or ask your own question?
You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.
Electronics Point Logo
Continue to site
Quote of the day