Retrobright: The Yellow Strikes Back

User avatar
Muirium
µ

04 Jul 2013, 01:51

Retrobright is awesome. We've all seen the spectacular before and after pictures where a deep yellowed plastic case or set of caps is restored to its youthful creamy glory. But does it stick?

Tez at classic-computers.org.nz brings the bad news that the answer is: maybe not. He has the details in his post, but the pictures speak for themselves. These computers were Retrobrighted then stored in the dark but the yellow still returned. There's several more examples if you click.
Picture 1.png
Picture 1.png (534.03 KiB) Viewed 24636 times
[/url]
Before and after is the familiar success story. Before, just after and long after is another matter, for him.

So, Retrobrighted ABS can go yellow again without further exposure to ultraviolet light. But there is one possible upside here: these computers were stored in warm conditions. Perhaps heat is something to look out for as well as staying out of the sun?

Tez has some theories to explain his findings:
The original damage from light causes degraded or free bromides throughout the case from the fire retardants. Retr0Bright only takes these away from the surface layer. However these bromides can migrate fairly freely through the polymer. They don't need light to do this. Migration is probably accelerated in hot conditions. These pre-existing bromines from the original damage migrate to the top and within a few years the surface is yellowed again. The rate of the regression depends on the inital extent of yellowing (the more yellowed, the more bromides down deep), heat exposure and the nature/quality of the plastics.

If this hypothesis is correct then a UV sealant will not protect the case. Or any sealant maybe. Unless pre-formed bromides from previous light damage can be stopped from migrating to the surface (and I don't know how you would do that) Retr0Bright is only ever going to be temporary.

Bear in mind this is just a hypothesis which fits the evidence but has not been proven. To prove or disprove it would require research. It seems a logical supposition though.
So keep an eye on anything you've Retrobrighted and maybe take some pictures to see if this is widespread. I haven't tried Retrobrighting anything yet myself.

Limmy

04 Jul 2013, 04:31

Interesting theory. He has neat evidences that are in line with his hypothesis.

Should we prevent the sun from touching beige/white plastics at all costs? If we believe his hypothesis, we probably should.

I personally prefer black colored keyboards/key caps, and I think I have more reason to like them more. Of course, we can Retr0Bright as many times as we want, but it seems to be too much maintenance for me.

User avatar
002
Topre Enthusiast

04 Jul 2013, 04:32

Would it be fair to say that after so many applications, it shouldn't happen again?

User avatar
matt3o
-[°_°]-

04 Jul 2013, 07:43

we are talking about 30 years old plastic, of course retrobright can't do miracles, but prolonging the life of very old keyboards for 4 more years is not really that bad. I can retrobright it again if needed.

I have some lego kits that are already yellowed after a couple of years. ABS really sucks.

Topre

04 Jul 2013, 08:57

The IBM Model M's case is PBT, so it's not just ABS plastic that suffers from yellowing. It's more of a plastic problem in general.

User avatar
matt3o
-[°_°]-

04 Jul 2013, 09:12

PBT *if* it yellows it takes really a long time and you have to store it in direct sun light. If you take ABS in direct sunlight for a couple of days it is already yellowed

User avatar
Halvar

04 Jul 2013, 10:31

I read about this effect in a commodore retrocomputing forum before, where a 1541 floppy they retrobrighted yellowed back to pretty much the old state in a few months despite it being packed up the whole time.

First keyboard I retr0brighted was my SGI Granite, which has an ABS case and an ABS spacebar. The other keys are PBT so I had a pretty good comparison for the space bar. My impression was that the spacebar re-yellowed considerably in the three months after rebr0brighting, while the case didn't much change. To me it seems you can't really predict a lot of this because the flame retardants and plastics used are just so different.

I suppose that also the permeability of the stuff for wandering bromides is variable -- on some parts UV-coating will help more than on others.

User avatar
matt3o
-[°_°]-

04 Jul 2013, 10:56

it must depend on how you retrobright or the original status of the keyboard, I have stuff retrobrighted years ago that I can't say it re-yellowed that much
Last edited by matt3o on 04 Jul 2013, 10:58, edited 1 time in total.

User avatar
Jmneuv

04 Jul 2013, 10:58

Next time in boiling retr0bright solution..

User avatar
matt3o
-[°_°]-

04 Jul 2013, 11:02

I use very high vol peroxide and UV source is very good (July-August sun in Italy). I can actually see the reaction happening before my eyes. As long as the plastic is in shadow nothing happens, as soon as it comes in direct sun light it starts to "bubble" :) In 30 minutes you can see it already whitened. In 8 hours it's like new.

Anyway I have some caps I recently treated, some of them I'm not going to use, I might try to take some RAW pictures with a reference white and see what happens in say 6 months.

User avatar
Muirium
µ

04 Jul 2013, 11:18

Great idea Matt. How about keeping a matching pair in different temperatures? One in a warm room, the other in the fridge! Then side by side pictures in 6 and 12 months, with white.

Retrobright still seems a useful rejuvenation technique, but it does get described as a bit of a miracle cure which always makes me suspicious. We only care about the colour of the surface of the plastic, but chemistry doesn't share our opinion. Wandering bromides exist underneath the micro layer, quite out of reach, but ready to yellow again.

You can Retrobright again, but the plastic will be at rising risk of harm every time you do. Especially if you're using household products to make your Retrobright, as we all do surely, which are actually blends of other chemicals besides the active ones.

I don't mean to be the downer on Retrobright, because honestly I would like to try it, but it seems we're only part of the way to a true solution. Merlin and the guys behind the stuff figured out how to get the bromides on the move but only how to leech them out of the thinnest tiny layer of the plastic. A few years seems to be all it takes for the rest of them to reemerge from below. I'm no chemist but my hunch is there's a second treatment needed. Something different to lock the brightness in place. And it ain't just a lick of varnish.

Pale ABS and sunlight: argh!

User avatar
matt3o
-[°_°]-

04 Jul 2013, 11:58

I don't think anyone thinks Retrobright is the miracle solution. If I see a nice yellowed keycap set on ebay for 10 bucks I take it right away because I know I can retrobright it. If it yellows back in 1 year I can retrobright it again. Who cares if it weakens the plastics, for 10 bucks I can take the risk :)

That being said the Amiga I retrobrighted one year ago still looks pretty good, but of course I don't have reference images.

The question is not how good retrobright is but how (and if) we can improve it. There are some UV filter clear paints. I used them to coat models to protect the paint. Maybe a very thin layer of varnish might help to protect the case, but that's not really a solution for caps.

Anyway I'll take the picture to the caps this afternoon as soon as I get decent sunlight and post it here. I don't have a pro camera but should suffice for the purpose.

User avatar
Muirium
µ

04 Jul 2013, 12:16

The miracle bit is on the project's site and in most coverage of it I see around the web. Yellowed ABS + careful application of Retrobright = Result! I don't think they're being dishonest, just over enthusiastic. (As I so often am about my own things like Bluetooth in a 60%!) You're level headed enough to understand what you're getting into without explanation, Matt. I'm the other sort who leaps in first and only then realises what he's done! So I'm reading up on it.

User avatar
matt3o
-[°_°]-

04 Jul 2013, 14:03

can't believe it... it's cloudy, I've never seen a weather like this in July...

User avatar
Jmneuv

04 Jul 2013, 14:16

Trying to mock us up north here?

User avatar
matt3o
-[°_°]-

04 Jul 2013, 14:19

just that I wanted to take the reference picture (+ I have a spacebar retrobrighting)

User avatar
Peter

04 Jul 2013, 15:30

Note that this is purely a surface effect and only penetrates through the plastic for a few microns;
the remaining plastic is unaffected, as it is not in contact with air and will not oxidize.
Once the treated part has been sealed with a coating such as a clear UV resistant lacquer, the yellowing should not return.
http://retr0bright.wikispaces.com/Ultra+Violet+Energy
I've retr0brighted a few boards and some cap-sets, following the formula as closely as I could :
The caps seem less likely to 're-yellow' than does the case (of a Cherry G80-3000 with DS-caps and win-keys )
BTW : 'Xanthan Gum' = 'E415' !!

User avatar
Muirium
µ

04 Jul 2013, 15:39

What Tez reports is that UV isn't necessary for re-yellowing. It could be the New Zealand heat he lives in, but it wasn't the light.

User avatar
Peter

04 Jul 2013, 15:49

Muirium wrote:What Tez reports is that UV isn't necessary for re-yellowing. It could be the New Zealand heat he lives in, but it wasn't the light.
Well, the chemists seem to disagree, no ?
'Bromides' and 'Iodine' are related, and we all know what happens when Iodine is exposed to oxygen, right ?

User avatar
Muirium
µ

04 Jul 2013, 15:59

I'd like to see some actually show up and discuss the counter evidence. Experiment is science, after all.

User avatar
Half-Saint

04 Jul 2013, 23:17

I have an Amiga 500 that was stored in a box under my bed for years. When I got it, it was pretty much like new. After not taking it out of the box for a while, I was in for an unpleasant surprise.

The Amiga was yellow, despite being stored in a dark place! Maybe it's because of the central heating, I don't know...

User avatar
Muirium
µ

04 Jul 2013, 23:38

Half-Saint wrote:I have an Amiga 500 that was stored in a box under my bed for years. When I got it, it was pretty much like new. After not taking it out of the box for a while, I was in for an unpleasant surprise.

The Amiga was yellow, despite being stored in a dark place! Maybe it's because of the central heating, I don't know...
Assuming it's never been Retrobrighted in the past (right?) that's a pretty interesting discovery. It sounds like the bromine fire retardant stains ABS yellow whether there's a UV source, or just everyday domestic heat. What a menace!

ABS itself is the common thread, of course. That plastic's bad news without a deep dye. Annoying, when you like your keyboards white like me.

User avatar
Kurk

04 Jul 2013, 23:43

I think that Tez's explanation absolutely makes sense. In the years that those strongly yellowed cases have seen, an appreciable amount of free bromine (Br2) has probably built up in the plastic. Retr0bright is an aqueous formulation and because ABS plastic is rather water tight only a thin top layer gets penetrated by the oxidizing solution. The remainder of the bromine is free to roam and will stick its head out over time simply by diffusing through the plastic. A higher temperature will surely enhance the rate at which this happens.

Another explanation could be that the breakdown products of the free bromine after the retr0bright action are the culprit. Most likely, hydrogen peroxide converts the bromine (Br2) to bromide (Br-). Next, still during retr0brighting, hydrogen peroxide could oxidize bromide to bromate (BrO3-) in a multi-step process. If this remains in or on the plastic then bromate, in turn, can be converted to the brown bromine (Br2) again by the action of UV light.
This chain of events, however, is unlikely to happen because a) all these compounds except Br2 are highly watersoluble and would be washed away by the retr0bright solution or afterwards by the typical washing with water, and b) Tez's case re-browned in the absence of UV light.

If it's indeed the bromine inside the plastic that diffuses out then that's the end of the story, there's not much one could do about it. Possibly, you could swell the plastic in an organic solvent and treat it with an organic oxidizing agent but that would do the integrity of a computer case no good.

User avatar
mbodrov

05 Jul 2013, 02:20

The original damage from light causes degraded or free bromides throughout the case from the fire retardants. Retr0Bright only takes these away from the surface layer. However these bromides can migrate fairly freely through the polymer. They don't need light to do this. Migration is probably accelerated in hot conditions. These pre-existing bromines from the original damage migrate to the top and within a few years the surface is yellowed again. The rate of the regression depends on the inital extent of yellowing (the more yellowed, the more bromides down deep),
ABS yellowing definitely does not occur throughout the depth of the plastic. If you look at a cross-section of a yellowed part, only a very thin surface layer is yellowed, and for most of its thickness, the plastic is the original color. In fact, when you are retrobrighting something and do not know when to stop (you are not sure what the original color was like), the solution is to take a knife and shave off the thinnest slice of plastic (from somewhere like the inner side of a keyboard case, where the damage will be hidden from view), revealing the original color. When you take off that thinnest slice and the color is identical to that on the surface, you are done.

User avatar
Muirium
µ

05 Jul 2013, 02:33

Yes. In both cases, we're talking about a microscopically thin layer. What went yellow in the first place, and what Retrobright fixed. The trouble is, what looks solid to us isn't like that as far as chemistry's concerned. Ions, radicals, or whatever this is technically, migrate through the plastic. The original yellowing creates them, and Retrobright tackles those it can reach, but the plastic beyond the immediate surface is a sponge for the buggers. There's enough of them around to cause lasting trouble.

A good experiment would be to test multiple Retrobrightings on the same samples of plastic. I wonder if it gets more or less effective every time. The answer would have a real effect on our guesswork.

User avatar
webwit
Wild Duck

05 Jul 2013, 02:53

I'm not sure if this chemical compount "leaked" up from deeper in the plastic case like that in any meaningful substance to cause this effect, after all this years in a stable environment. I.e. I believe it's the surface layer itself which re-yellows, and it doesn't "leak" from deeper within.

User avatar
mbodrov

05 Jul 2013, 05:18

If everything happens in the microscopic surface layer, it does not go along with the idea that there are deep stores of "yellow pigment" which diffuses back to the surface even in total darkness. The precursor chemicals may well do so, but for the actual yellowing to occur would still require all the normal triggers, i.e. exposure to UV light.

User avatar
Kurk

05 Jul 2013, 08:13

mbodrov wrote:If everything happens in the microscopic surface layer, it does not go along with the idea that there are deep stores of "yellow pigment" which diffuses back to the surface even in total darkness. The precursor chemicals may well do so, but for the actual yellowing to occur would still require all the normal triggers, i.e. exposure to UV light.
Good point, but what seems like a thin layer to the human eye can in fact be large enough to store sufficient bromine.
In other words, if the bromine-browned layer is for example 0.1 mm thick and the retr0brightened top layer only 0.01 mm, a diffusion-controlled re-browning is possible.
I can very well imagine that the penetration depth of UV-light into ABS is much larger than that of an aqueous solutions.

We need someone with access to a microscope and a means to prepare a microscopic slice through a retr0brightened key cap!

User avatar
matt3o
-[°_°]-

05 Jul 2013, 08:18

Muirium wrote:A good experiment would be to test multiple Retrobrightings on the same samples of plastic. I wonder if it gets more or less effective every time. The answer would have a real effect on our guesswork.
I'll try this. It will just take time...
Kurrk wrote:We need someone with access to a microscope and a means to prepare a microscopic slice through a retr0brightened key cap!
this is a good idea. Slice the caps and see the layers.

User avatar
Halvar

05 Jul 2013, 10:17

I agree with Kurrk's last post.

The effect is explained on the retr0brite site here: http://retr0bright.wikispaces.com/Co-od ... nt+Bonding

The central fact here IMO is that light travels much deeper into the ABS than the air's O2 or the Retr0bright solution can. So the Bromine free radicals are created up to relatively deep in the ABS, and they are small enough to travel to the surface over time. Only in the thin surface layer thats in contact with air can they react with O2 to become essentially some sort of brown dye (2BR.O).

Retr0brite can exchange the O in the Br.O for H on the surface, which essentially "bleaches" the brown dye that has already formed. But it can't go deeper than the air could, and it can't do anything to the Br radicals that were built up deeper in the plastic over 20 years. These are still there (I'm aware that this is a simplified model, in fact they probably somehow react with the ABS or other stuff that's in there) and part of them continues to travel to the surface and continues to form new brown dye just by contact with the air. It doesn't need any UV light for this to happen any more at that point.

Post Reply

Return to “Keyboards”