67

I initially posted this on chemistry.stackexchange but didn't get any answers, so I'm reposting it here.

Long story short - we have an electronic product that is submerged in fuels (kerosene being one of them) and uses an RGB LED (click here for datasheet). Due to a sealing problem in the enclosure, kerosene has managed to get in and cover the PCBs. What's interesting is the effect that has had on the PCB. The PCB's functionality has been completely unaffected, apart from the fact that the red LED in the RGB LED module has completely stopped illuminating. We've replicated this ourselves manually by submerging 2 new PCBs in kerosene for a day and then taking them out and powering them up and seeing that the red LED stops illuminating entirely. The green and blue LEDs continue to illuminate just fine.

Examination of the failed boards shows that there are no other electrical faults. It is just the red LED that completely stops illuminating. We measured the forward voltage across each of the LEDs in the failure condition, but didn't notice any significant difference that would explain the fault.

After leaving the PCBs to dry, the red LED starts working again. So the problem is not permanent.

Looking at the last page on the datasheet, the LED material is listed as AlGaInP / GaAs. Is there any obvious reaction between kerosene and these materials that would explain why just the red LED stops working?

Update 1: I've carried out the following experiments:

  • Dripping kerosene on to the LED.
  • Submerging the PCB+LED in kerosene while running.

(Videos to follow up later on today, hopefully)

In both cases, there was no perceived effect on the LED - it continued to operate just fine. This would seem to indicate that the problem is not purely an optical problem between the kerosene and the LED. So far, the problem has only occurred after soaking the LED in kerosene for some time.

Update 2: I've taken a fresh PCB with LED on it (haven't done any tests yet with just the LED) and soaked it in kerosene. I've taken some close up photographs of the LED before soaking, after soaking while it's not working and after it resumes working after it's been left to dry.

What the photos show is that there is a very obvious bulge in the LED lens during the period when it's not working. Once the bulge recedes, the LED illuminates again.

Unfortunately, I don't have a camera set up on the PCB to see the exact moment that it stops working. I'd let it soak for about an hour before it stopped working. I checked on the LED every now and then and noticed no change in the LED brightness. I came to check it once and it was just off. My suspicion is that the change is sudden.

Judging by the swelling, I'm going to guess that there is some mechanical damage internally that's moving something and once the swelling recedes it springs back to position.

Left: Kerosene-soaked LED; Right: Normal LED Left: Kerosene-soaked LED in failed state; Right: Normal LED

LED in failed state after soaking LED in failed state after soaking

Normal LED Normal LED

Left: Kerosene-soaked LED after being left to dry and in the working condition; Right: Normal LED Left: Kerosene-soaked LED after being left to dry and in the working condition; Right: Normal LED

Kerosene-soaked LED after being left to dry and in the working condition Kerosene-soaked LED after being left to dry and in the working condition

Community
  • 1
Amr Bekhit
  • 1,278
  • 10
  • 24
  • 6
    What does "doesn't work" mean? Does it only mean you can't *see any light*, does it mean the diode gets a very high or a very low DC resistance? I.e. does the forward voltage change, or the current at a fixed voltage? – Marcus Müller Aug 16 '16 at 09:44
  • You haven't addressed the question from the chemistry SE: when looking at the contacting area after the LED has been exposed to kerosene, is it visible that the individual diodes are still properly connected? – Marcus Müller Aug 16 '16 at 09:46
  • 2
    @MarcusMüller it just doesn't illuminate at all. We measured the forward voltage and didn't notice any change. Regarding the question on chemistry SE, I mentioned at the end that the PCB starts working perfectly fine after leaving it to dry. So that eliminates the possibility of any permanent physical damage. – Amr Bekhit Aug 16 '16 at 09:47
  • 1
    It would probably have been better if you asked this on physics.se since it seems that kerosine is interacting with the LED material which is more a physics thing than a chemistry thing. If you are further interested you could submerge more LEDs (no need for the whole PCB) in kerosine and test more. – PlasmaHH Aug 16 '16 at 09:47
  • Not asking about permanent damage! – Marcus Müller Aug 16 '16 at 09:47
  • 1
    @MarcusMüller, the theory on Chemisty SE was that the kerosene might be reacting with the packaging and/or dissolving the bond wires, thus disconnecting the LED inside. But since the LED started working perfectly fine after being left to dry, I ruled that possibility out. I could be wrong, of course. Unfortunately, I don't have access to a microscope to examine the LED up close. – Amr Bekhit Aug 16 '16 at 09:52
  • 9
    Wild guessing: maybe the kerosene is optically opaque to red light (especially the wawelength of your red LED models). Try to see if a red LED light can be seen through a glass full of kerosene, but keeping the LED *outside the glass*. – LorenzoDonati4Ukraine-OnStrike Aug 16 '16 at 09:57
  • @LorenzoDonati kerosene, in my experience, is "OK" on red light, but that's not the point – solving kerosene components in a plastic lens changes the (optical) structure of that lens, and that change has next to nothing to do with the optical properties of kerosene itself (see my answer).ö – Marcus Müller Aug 16 '16 at 10:01
  • 6
    Does this happen with a red LED in a minimal circuit on long leads, just immersing the LED? Does it stop working immediately or after a period of time? – pjc50 Aug 16 '16 at 10:05
  • 1
    See JRE's answer re TIR. Add kero drop to stop working and then add eg water over outside of kero. What happens ?-). Hey - you have a kero detector :-). | Have a look at [**these pages**](https://www.google.co.nz/search?q=brix+meter&num=100&tbm=isch&tbo=u&source=univ&sa=X&ved=0ahUKEwi60oKj4cXOAhUBxpQKHcGSDVIQsAQIOg&biw=1745&bih=977) for "Brix meter" - IF it's TIR then this shows how it works. Click on a few (all are linked to pages) and read. Also see [**Wikipedia - Brix meter](https://en.wikipedia.org/wiki/Brix) and [search - Brix Meter](http://bit.ly/search_brix_meter) – Russell McMahon Aug 16 '16 at 10:55
  • 6
    To those possibly thinking this is off-topic and it belongs to Physics.SE, IMO it is not. Although I concede physicists could know better about the phenomenon at hand, it is completely on topic here, since it is about the materials with which an actual part is made of and its interaction with the external environment. I think it is a very good question for this site (EE design applied to harsh/extreme environment). IMO this is one of those edge cases where cross-posting *is* useful/needed. – LorenzoDonati4Ukraine-OnStrike Aug 16 '16 at 12:22
  • 3
    And, just to nitpick and reinforce my claim that this question is on-topic, the OP's is a rather oddball case of optical coupling between an emitter and the transmission medium, which is the same problem EE designers face with optical fibers. – LorenzoDonati4Ukraine-OnStrike Aug 16 '16 at 12:23
  • This is getting very opinion-based and noisy, with zero credible sources cited in the answers. – Dmitry Grigoryev Aug 16 '16 at 13:06
  • 1
    @LorenzoDonati if kerosene were opaque to red light, it would be strongly green/blue in colour... – nekomatic Aug 16 '16 at 13:57
  • 4
    Just to make sure... You're not using red-dyed Kerosene, right? Because that's a very common thing: http://www.agriculture.ny.gov/WM/Kerosene.html – JPhi1618 Aug 16 '16 at 14:27
  • 2
    @JPhi1618, no, this is standard yellow kerosene. – Amr Bekhit Aug 16 '16 at 14:31
  • @AmrBekhit Have you had a chance to try some of the suggestions? Does the red LED stop working immediately when the LED goes into the kerosene, or does it take some time before it goes out? – JRE Aug 16 '16 at 14:43
  • @JRE Not yet, but I do intend to set them up, try and video them and then post some links. – Amr Bekhit Aug 16 '16 at 14:45
  • 2
    Some pictures would have been better – n00dles Aug 16 '16 at 15:01
  • 1
    I've found a handful of red LEDs with clear lenses in my junk box. Experiment with LEDs an kerosene (well, lamp oil) after supper. – JRE Aug 16 '16 at 16:50
  • 2
    Well, so much for that plan. Those were white LEDs, not red. Tried it any way, no effect. I thought the white might turn blue/green if the red were being blocked or whatever. – JRE Aug 16 '16 at 18:12
  • 2
    OK. With my idea proven wrong, I've deleted my answer. – JRE Aug 17 '16 at 08:51
  • Does the LED manufacturer have any contact info published that could be used to see if this is a known issue/limitation? – Todd Wilcox Aug 17 '16 at 12:13
  • Another question to OP: Can you measure the forward voltage in working and faulty state at a given forward current? The result could be interesting. At least you could rule out if something happens to the die. – Ariser Aug 18 '16 at 12:07
  • @JRE can I suggest you reconsider that deletion? Yours was a good idea, and at least until the real answer is found, an experiment showing a conclusive negative result is still a useful data point. – Oleksandr R. Aug 18 '16 at 12:24
  • @OleksandrR.: Amr Bekhit repeated the experiment, with the same negative results. I think it is better to leave my answer deleted, since it is proven wrong and the OP has posted the negative results in the question. – JRE Aug 18 '16 at 12:40
  • I've added some photos showing the physical effects of soaking the LED in kerosene. – Amr Bekhit Aug 19 '16 at 13:40
  • 1
    I checked your question after a while, to see updates. Man, this turns out to be one of the most interesting question of the whole EE.SE knowledge base, IMO! I wish I could upvote you again! Please, keep us informed of your further findings. – LorenzoDonati4Ukraine-OnStrike Sep 10 '16 at 15:30

3 Answers3

14

We measured the forward voltage and didn't notice any change.

Physically, I'm pretty sure this means the semiconductor interface is still producing photons at the same rate and wavelength as before.

So, something happens to those photons.

What you should do is get a working source of red light of the same wavelength (e.g. another one of your LEDs), extract the "lens" material from a "donor" LED:

LED picture

e.g. by cutting it off with a razor blade, testing transmission of red light before and after having soaked that material in kerosene.

Since that lens is tiny, you should probably use something like a piece of cardboard with a hole punched through it with some kind of needle (don't let the hole get to small, lest you want to have much diffraction...) and put the lens in front of that hole.

My guess is that soaking the material in kerosene leads to a drastic change in optic properties, and that might very well mean that either

  1. your lens is now absorbing red light or
  2. your lens is now not focussing red light, but spreading it.

To rule out 2., you'd need a very very dark room and some way to guesstimeasure the distribution of light. So in effect, without optical design lab equipment, either way, Kerosene contains a mixture of different hydrocarbons, and those are soluble in other hydrocarbons, such as the transparent material used to protect the actual LEDs and act as a lens.

Community
  • 1
Marcus Müller
  • 88,280
  • 5
  • 131
  • 237
  • My guess is that soaking the material in kerosene does absolutely nothing, since kerosene has nearly the same refractive index as most plastics. – Dmitry Grigoryev Aug 16 '16 at 12:58
  • 4
    well, most plastics aren't used for LED lenses! – Marcus Müller Aug 16 '16 at 13:03
  • Do you have a credible source saying that red LEDs use different plastic than LEDs of other colors? – Dmitry Grigoryev Aug 16 '16 at 13:07
  • @DmitryGrigoryev I'm assuming the LEDs in that package are under the same plastic, but soaking that in kerosene adds periodicity on a level that absorbs red light, currently. The alternative explanation of thin film of Kerosene changing emission on the air/lens boundary: backed by [Zeiss](http://www.zeiss.com/vision-care/en_de/eye-care-professionals/optical-knowledge/optical-basics/lens-materials/plastic.html) lists optical plastics with \$n\approx1.6\$, which is about 0.2 higher than [what wikipedia says about kerosene](https://en.wikipedia.org/wiki/List_of_refractive_indices). – Marcus Müller Aug 16 '16 at 13:09
  • 1
    What's *periodicity on a level that absorbs red light*? Is there an article about it? – Dmitry Grigoryev Aug 16 '16 at 13:18
  • 1
    so, going back to my solid state physics lectures: Imagine a crystal. A crystal has periodic distribution of potential. This implies that there are only discrete non-zero points in impulse space (i.e. essentially Fourier transform of space). This means that photons of a specific impulse, hence energy, hence wavelength, can interact with the material, and others can't. – Marcus Müller Aug 16 '16 at 13:38
  • 1
    @DmitryGrigoryev http://www.reading.ac.uk/ir-absorptiontheory-latticeabsorption.aspx – Marcus Müller Aug 16 '16 at 13:39
  • 3
    FYI, plastics used for optical devices are amorphous, since crystalline plastics are typically [opaque](https://en.wikipedia.org/wiki/Crystallization_of_polymers#Optical_properties). – Dmitry Grigoryev Aug 16 '16 at 13:50
  • 2
    @DmitryGrigoryev that's *exactly* what confuses me here, and why I'd love to see OP do these experiments. We both agree that if forward voltage doesn't drop, the same photons should be emitted, and the LED as is operating "normally", so this pretty much *has* to be an optical effect. – Marcus Müller Aug 16 '16 at 14:01
  • So if the red light is being absorbed, the lens should give of more heat, right? – n00dles Aug 16 '16 at 15:17
  • 1
    @n00dles there's positively no way you can measure this outside a very very well equiped lab, under very well-controlled conditions – Marcus Müller Aug 16 '16 at 15:42
  • 2
    I've tried dripping kerosene on to the LED and submerging it, but there is no immediate effect on the LED. I'll try soaking it while it's running and see if there are any gradual changes. – Amr Bekhit Aug 17 '16 at 08:49
  • @MarcusMüller there's one tiny problem with crystalline absorbtion and LEDs: the bands are quite narrow (or so I suspect), but LEDs are not that monochromatic: they actually have a pretty wide intensity-vs-wavelength distribution, as seen, for example, in [this](http://l.flashcact.us/pimg/screenshots/20160817-115710.png) graph (grabbed from a random datasheet). – FlashCactus Aug 17 '16 at 09:01
  • 2
    @AmrBekhit make sure that you're submerging *just* the led (i.e. desolder it from the board and hang it off of some wires), to simultaneously make sure that it's actually the LED itself acting up and not some obscure effect in some other part of the board. – FlashCactus Aug 17 '16 at 09:14
  • I've added some photos to the question showing how the kerosene causes the LED lens to bulge. – Amr Bekhit Aug 19 '16 at 13:41
14

My 5 cents:

Most LEDs are potted with silicone today. Silicone has a good permeability for VOCs (volatile organic compounds, e.g. alkanes and their isomeres), which are part of kerosene.

VOCs entering the silicone can interact with the silicone matrix, changing its optical properties. Often observed damage: the potting/lens can get milky or diffuse, and yellowing can be observed.

Certain VOCs will be broken up by the blue light of an LED which usually leads to a blackening of LED potting/lenses.

Those effects are known to be (partially) reversible. I.e. the discoloration of the lenses will disappear if the VOCs are able to gas out again. This happens faster if heated under the LED's operating conditions.

So my explanation is: Edit:Highly speculative Large amounts of kerosene may contain also aromatic compounds, which are known to be optically active (e.g. See azo dyes pigments). Van der Waals forces can change the resonant behaviour of aromatic compounds, which is possible when the VOCs enter a silicone rubber matrix. This could explain why fractions of the kerosene achieve a red filtering behaviour when entering the potting.

Edit: I cannot rule out interaction of VOC with the semiconductor itself, but I habe difficulties imagining how this could work. The crystal is nearly impermeable for anything at room temperature hence the interaction can only happen at the surface of the dice. Because light emission happens everywhere near the pn-boundary I doubt that kerosene components can prevent the generation of photons. IMO only absorption and filtering are the effects to look after again.

Another culprit in LED deterioration is hydrogen sulfide, which can be found among other sulfur compounds in kerosene, too. But sulphur corrosion in LED isn't reversible AFAIK, so this can be excluded IMO.

Ariser
  • 3,846
  • 3
  • 23
  • 43
  • Could these changes also include start of fluorescense? Absorb visible light reemitting it somewhere in the IR? Would you -if so- be able to see these photons with a webcam? – peter Aug 16 '16 at 21:04
  • While I really can't judge if this stuff about VOCs is true (_volatile organic compounds_, I suppose that is), it certainly sounds much more plausible than JRE's total internal reflection explanation. – leftaroundabout Aug 16 '16 at 21:42
  • 2
    There are several excellent resources describing the issue with VOC absorption in LEDs from some manufacturers eg. [Cree](http://www.cree.com/~/media/Files/Cree/LED%20Components%20and%20Modules/XLamp/XLamp%20Application%20Notes/XLamp_Chemical_Comp.pdf), [OSRAM](http://www.osram-os.com/Graphics/XPic0/00161697_0.pdf/Chemical%20Compatibility%20of%20LEDs.pdf). I am not sure if this would be the issue with the OP's LEDs though - they are relatively low power and I guess the absorbed VOC are less likely to decay in the way they do with bright/hot high power LEDs. – Matt B Aug 17 '16 at 03:50
  • 1
    The recovery effect as the VOC evaporates out of the silicone packaging definitely does occur - I've observed it myself in LEDs that had absorbed VOC produced by glue used elsewhere in the device. – Matt B Aug 17 '16 at 03:50
  • @peter it could easily be too far into the IR to see with a webcam -- *if* it flouresces. The webcam could also bbe confused by residual red getting through. – Chris H Aug 17 '16 at 08:30
  • Azo compounds are dyes, not pigments. "Van der Waals forces can change the resonant behaviour of aromatic compounds" -- agreed. But: "... which is possible when the VOCs enter a silicone rubber matrix" does not follow. Are you suggesting there are azo dyes in kerosene? Most aromatic compounds that could be found in kerosene absorb in the UV and emit in the blue, with poor fluorescence quantum yield. This is not going to be altered that much by their dissolution in silicones. It is a fair guess, but IMO it does not work as an explanation. – Oleksandr R. Aug 17 '16 at 11:43
  • My feeling is that rather than absorbing the emission, some component of the kerosene interacts with the semiconductor die to quench the electroluminescence (e.g. by providing other, kinetically or energetically preferable pathways for electron-hole recombination apart from the usual radiative one). But I haven't found any literature on this phenomenon so far, so have no solid proof that this occurs. – Oleksandr R. Aug 17 '16 at 11:54
  • Related: http://www.lrc.rpi.edu/resources/newsroom/pdf/2012-2013/LED-VOCs8511.pdf – Todd Wilcox Aug 17 '16 at 12:18
  • 2
    Thanks for your answer. After examining one of the LEDs when it was in its failure state, I couldn't notice anything visibly wrong with it. There was no sign of any clouding or occlusion - the lens seemed perfectly clear. I'm going to try and repeat the test and get some photographs so that I can post them up. – Amr Bekhit Aug 17 '16 at 12:27
  • @AmrBekhit I guess you need a good microscope to see the crucial details if there are any visible at all. – Ariser Aug 18 '16 at 11:51
  • 2
    Follow up on this, I had a look at the LEDs on two PCBs that I had tested a week or so back and there is very noticeable yellowing of the LED lens. So this reaction with the kerosene you mention seems to be happening. However, I've just updated the question with some photographs showing bulging of the LED lens in the failed state after being soaked. I have a feeling that the problem is mechanical in nature, i.e. the bulging is disconnecting something. – Amr Bekhit Aug 19 '16 at 13:44
1

My guess is that the kerosene absorbs the red photons and heats up the plastic lense causing it to bulge, which in turn causes dispersion of the photons. So you have the double effect of absorption and dispersion of the red photons. There is also the possibility that at some point, the heat produced swelling of the plastic, creates a high resistance connection, which goes back to "normal" after the LED dries.

Guill
  • 2,430
  • 10
  • 6
  • 4
    Although I didn't mention this detail in the original question, I know that this is not the case. This is because if you leave the PCB in the kerosene **unpowered** and then take it out and power it up, it won't work. Therefore, the bulging has got nothing to do with the light being produced and is purely due to exposure to the kerosene. – Amr Bekhit Aug 22 '16 at 07:49