chemical etching/milling

Hi James
I made some etched brass serial number plates for the dashboard in a mates Sentinel steam waggon.
I got all the stuff I needed from Maplin Electronics. spray YM62S Developer AP01B Ferric Chloride XX12N
You are supposed to expose the resist material using a UV lamp, but I stuck it in the window for 10 minutes one day when the sun was out, and it worked a treat.
This website will show you how to go about it. Hope this helps you out. Phil

Okay a bit of an update here and a couple more questions. I'm trying to chemically mill some .6mm aluminium sheet. I have a solution right now of sodium hydroxide and dissolved aluminium and that's working fine. It really eats through the aluminium especially when heated slightly.
What I'm now haveing problems with is the photoresist. The stuff I got (from maplins) is for pcb/copper and doesn't seem to hold up to the sodium hydroxide. So I have two options - maybe try ferric chloride? I'm not sure how good that would be with ali though. Or find a different photoresist - which is what I would probably preferr.
The stuff I'm doing isn't all that accurate but I was getting a sharp image with the photoresist until it pealed away. The solution I'm using is pretty strong - could it be fighting through the photoresist?
Any suggestions?

Randy Gordon-Gilmore uses dry-film resist, have a look at - My equipment - Photoetching. I have used this kind of resist myself but not with aluminium and sodium hydroxide.

Warm sodium carbonate is an effective, but slower, etch for aluminium: it might be kinder to your photoresist too.
Ferric chloride is not the answer!

Ferric Chloride should be just fine at etching Ali - I tried some (heavily) used ferric chloride on some Ali plate once (Ali of unknown parentage I'm afraid). The results were amazing - I dropped a 1" x 6" strip into a 1 gallon container and after a couple of seconds the throth from the reaction was jetting out to a height of a couple of feet - then I had the problem of getting the strip out - a bit of a chemical spill clean up operation I'm afraid.
The plate had got thinner by about 1/2mm in a couple of minutes.
The Ferric Chloride had been used (many times) for PCB etching - so a substantial amount will have turned to copper chloride (if I remember my chemistry right) - but I don't think this makes much difference.
Steve

By the way I'd be surprised if the gas being produced wasn't chlorine - so be careful as its toxic - etch outside - stand upwind -and don't breathe in the fumes.
Steve

Dire warnings about chlorine production:
You have been warned!
Steve

More likely this produces hydrogen and aluminium chloride, I'd have thought, but better safe than sorry (and hydrogen's explosive).
Scrim

[re using ferric chloride to etch aluminium]
Done right it gives ferrous chloride and Al chloride, no gases at all. It can get violent though, when gases can be produced - keep it dilute and cold.
Aluminium can be etched with: ferric chloride, sodium hydroxide, sodium carbonate, mixtures of sodium carbonate and sodium hydroxide (/and bicarb), copper sulphate (add about the same amount of sodium chloride, and a little sodium hydroxide or sodium carbonate or sodium bicarbonate), hydrogen peroxide and hydrochloric acid (4 parts 30 vol H2O2 : 1 part HCl :3 parts water is one recipe),
... the list goes on, but those are some of the common ones.
For brass, ferric chloride is normally used, or occasionally ammonium persulphate. Hydrochloric acid can be added to ferric chloride etchant to stretch it out. Ammonia will etch brass.
Stainless steel has a problem, the used etchant is highly toxic and polluting. Not really suitable for the home.
Maplins is probably the easiest place to get ferric chloride etchant and photoresist. Many of the chemicals used in etchant mixes are available elsewhere: ammonia is available in hardware stores and the like, HCl is sold as "spirits of salt" in paint stores.
You don't have to use photoresist - you can use Staedtler permanent lumocolour pens to write resist directly on metal, remove with nail varnish remover; or you can melt Tesco's value raspberry jelly (16p!) on the metal, harden in a gentle oven, and scrape it off where you want holes*. Paint, grease and the like, even ordinary screen printing ink, will also act as resists. Try on a piece of scrap first though.
Sericol used to supply special resists for screen printing, been a while since I bought any though, I imagine other screen printing suppliers will too, if needed.
*this is an old technique, once widely used. You can also add ammonium dichromate to the Tesco's value raspberry jelly (16p!) and use it as a photoresist!
It is also possible to do electrochemical etching, rather than the passive etching I have been talking about. Ordinary salt in water can be used as the electrolyte.
NaOH is used to develop many common photoresists, and developed resist should not break down in the concentrations used for Al etching (although it can be close). If you are getting breakdown, try increasing the exposure, thicker resist, or a different resist.
Or try colder NaOH, or replace some of the NaOH with sodium carbonate or bicarb.
Another cause of difficulty can be the adhesion of the photoresist to the aluminium. It is hard to get it to adhere properly. Clean the Al thoroughly, degrease, and dip in one of the etching solutions above until all the surface changes colour. This mostly removes the oxide film. Then coat immediately with resist.
Once you have etched the piece, you may want to rebuild and harden the film by anodising. Al is highly reactive, it forms an oxide film immediately on contact with air, and it's tough enough to protect the Al from more air, but not from even mild abrasion.
Anodising is done in two parts, first the oxide film is grown thicker by electrical action, after which under a microscope it looks like the surface is covered in threepenny bits: then the bits are expanded to fill the gaps.
The growth phase is done by first cleaning the Al as above, though just cleaning the resist off is enough if you have already cleaned it for etching. I'm going on about cleaning, but well over half the chemical and processing (excluding hand-wiring of pieces) cost of commercial anodising is in cleaning. If you want good results get it really clean first.
The Al is made the anode in a cell - lead is good for the cathode, and it should be bigger than the Al surface. The pieces should be wired together to the supply with Aluminium or Titanium wire/frames/whatever.
For a reasonably hard film, use a 10% sulphuric acid electrolyte with a little Glauber's Salt* added, at 20 C, about 10 amps per square foot. 7 minutes if agitated, 10 minutes without agitation, will give a ~20um film.
* or sodium hydroxide, or sodium carbonate, or sodium bicarbonate.
The piece can be dyed now if desired, then the threepenny-bits are expanded by boiling in water for 10-20 minutes.
"Hard" anodising is quite difficult and nasty to do at home, but it can give a very hard surface - the problem with that is that the base metal is still Al, and soft.

Just a quick update to say Ferric Chloride was first tried because thats what I already had in a bubble tank and it worked a treat. It did have a tendency to run away but it did work fairly quickly which is what I needed :)
I have to touch up my photoresist process - and I'll try it again.
After that I think I'll make up another solution next week and try that.
Thanks everyone !

Okay - this is what I have already so this is what I started with. Initial results where okay - then the second time (after trying to improve the masking a bit) the part seemed to 'burn up' in that it wen't all blackish - /etc. I'm not sure if the solution was too hot, too strong or what.
I have noticed that there is a thick black film on afterwards which I figure is the spent iron. Is there a better solution out there that is cleaner etching? I will anodising these afterwards to this is critical.
I was going to order the stuff for this given this recipe here:
CuSO4 (copper sulfate -- bluestone) 1 kilogram NaCl (sodium chloride -- table salt) 250 grams NaHSO4 (sodium bisulfate -- Sani Flush ) 25 grams H20 (water) - depending on bath strength 10-20 liters
From what I've read this is a salt based etchant doesn't bring with it as many of the hazards as the other methods.
The material I'm etching is 0.5 to 1mm aluminum sheet with a fairly intregate design on it.
I would really love some more help on this :)
Thanks,
James

I don't have much practical experience in etching Al with FeCl3, but I can explain the chemistry, which should give you some ideas. It's a lot more complex than this in real life btw, this is a simplification.
In the reaction between copper and hydrochloric acid
Cu + 2 HCl -> CuCl2 + H2
hydrogen gas is given off. The reaction is too slow to be useful for etching copper though, mainly because the processes that produce the gas and the bubbles need quite a lot of energy which is not readily available. The not-quite-gas builds up, and slows the reaction to a near stop.
There are two chlorides of iron, ferric chloride FeCl3 and ferrous chloride FeCl2. The normal FeCl3 etching process on copper works like this
2 FeCl3 + Cu -> 2 FeCl2 + CuCl2
The "extra" chlorine from the ferric chloride is given to the etched metal, which turns into the metal chloride, which dissolves. Other reactions happen in FeCl3 etching of copper, but they are similar in nature.
Note that no gas is given off, gas and bubble producing processes are not involved, and thus the reaction can be quite rapid.
With Aluminium there is a whole different level of available energy. Aluminium is very energetic fuel, and instead of the normal etching reaction
Al + 3 FeCl3 -> AlCl3 + 3 FeCl2
reactions like this can occur:
2 Al + 2 FeCl3 + 6 H2O -> 2 AlCl3 + 2 Fe(OH)3 + 3 H2 - gives off hydrogen
Al + FeCl3 -> AlCl3 + Fe - very energetic, a "thermite" reaction!*
so the reaction gets very hot and gives off gas.
* mixed FeCl3 and Al in powder form will produce molten iron if ignited, and can explode. It can also go off spontaneously. Extreme caution. No kidding.
Reactions are usually speeded up by increased temperature, and by increased mixing. This reaction is quite quick to begin with, the heat given off increases the temperature, and the bubbles of gas increase the mixing, both making the reaction even faster.
There is nothing to immediately stop this quickening, so it can "run away" and get very violent. That's not including the risk of the hydrogen getting mixed with air and exploding.
What can you do? Diluting the etchant has several effects - first, the reaction rate is slower, as the amount of ferric chloride in contact with the Al surface at any instant is less.
Second, the increase in temperature from the energetic reactions is lessened, as the water has to be heated up too.
Third, for reasons I won't go into gas production is slowed, favouring the normal etching reaction. A thin film of aluminium hydroxide gel forms, futher slowing the reaction.
Fourth, the mixing rate increase is lessened because the bubbles are bigger and slower to form and break away; and the slower-developing bubbles stay on the surface longer, protecting it from further fresh etchant.
In summary, dilution slows the reaction, and is good for stopping runaway.
However, dilution may make the etching process too slow. If that's the case, then heating the etchant will make it faster. This can be dangerous though. Al has a tough surface film of oxide, which etchant can take a while to remove. You may think it is going too slowly, then the film will break down and the reaction will start to go at it's "proper" rate. Boom.
I don't know that I would recommend using FeCl3 for etching Al at all, but if you insist one suggestion would be to try a cold dilute etchant, let it go for a while, and then if it is still too slow heat it up. Once you know how the "proper" rate goes, you can start with a hot dilute etchant if needed, although a cold dilute etchant may well be okay.
If dilution and temperature control aren't enough, you may have to change etchant. Let me know, I may have some suggestions.
Copper sulphate, table salt and sodium bisulphate is a fairly new one which has received good reviews. See also helpfully describes common cheap sources for the chemicals involved.
I must emphasise though that I have very little practical experience with etching Al with Ferric Chloride, and this is all a wild-assed guess. The chemistry of aluminium's mixed chlorides/oxides/hydroxides and their hydrates is particularly complex. I take no responsibility. Be careful, take precautions - suitable clothing, open air and goggles at a minimum.
Strong bicarbonate of soda solution will neutralise FeCl3 etchant spills on skin and clothing. Cheap. Keep some handy.
Oxalic acid will remove FeCl3 etchant stains, but it's poisonous, and I don't recommend even considering it unless it's your favourite shirt and you have already tried stain remover without success.
By the way, I missed something in my last post - some modern resists are designed to be stripped with NaOH. Last time I bought photoresist it used NaOH as a developer, not a stripper, but that was some time ago. It didn't work all that well then though, todays chemistry may be better.

Used etchant contains ferrous chloride, which will react with Al like this
2 Al + 3 FeCl2 -> 2 AlCl3 + 3 Fe
And if it's been used for etching copper it will have copper chloride on it:
2 Al + 3 CuCl2 -> 2 AlCl3 + 3 Cu.
I haven't tried it myself, but it's said to be good stuff. About four times more salt is now the rage, and if you are through-etching I'd use half again the amount of Sani Flush.

Snipped etc..
Digressing a bit - going back to the electrolytic cleaning of rusted parts using a sodium carbonate solution (which has worked great for me), could you do a step through the reaction there? I've tried a couple of times and my first conclusion that sodium hydroxide gets produced at one stage and as a second stage I get confused!
TIA
Steve

Not simple! But I'll do my best. You might want to skip down to the equations below, the first time you read this. ;)
It's a subject that for years has had a lot of chemically impossible nonsense talked (and published) about it. It's only in the last few years that proof of what really happens has emerged - which is mostly what sensible chemists thought all along, plus some new stuff we suspected might happen.
First, what is rust? It's commonly symbolised as ferric oxide, Fe2O3, but in fact it is a very complex mixture of simple and mixed ferric and ferrous hydroxides, simple and mixed oxides and carbonates; and hydrates of those; some appearing in more than one form.
Ferrous and ferric? There are two (main) forms of ionic iron compounds, ferrous and ferric. Ferric iron ions have three positive charges, ferrous iron ions have two. So the ferrous and ferric chlorides would be FeCl2 and FeCl3 respectively.
The simple oxides of iron are ferrous oxide FeO and ferric oxide Fe2O3; but the real life oxides are a bit (!) more complicated. There are two important ones in electrolytic rust removal; rust itself, and magnetite Fe3O4, which is a mixed ferrous and ferric oxide. It is known as a mixed oxide because some of the iron atoms are in the ferric state with three ionic charges, and some are in the ferrous state with two.
You can think of magnetite as Fe2O3 and FeO joined together. It is the black stuff used in cheaper magnetic recording tapes. More on magnetite later.
Rust has several unusual properties. It conducts oxygen, carbon dioxide, and water; oxygen and hydroxyl ions; hydrogen gas and hydrogen ions; and it conducts electricity.
Rust may not seem to conduct electricity when you try to make an electrical contact to a rusty piece of iron, but nevertheless it does conduct a little bit of electricity; and when the contact is over a large surface area and through a thin film it can conduct quite a lot. This is what allows the electrolytic rust removal process to start.
Rust has another property - it is bigger than the iron underneath it, so it swells up and cracks on a rusting surface, exposing bare metal at the bottom of the cracks. This can speed up the start of the process, but it is not needed.
Rust may not seem to conduct the other stuff, oxygen for instance, either; but if it didn't, the cracks would be the only way that oxygen could get to more metal, and the rusting process would quickly stop (much like Aluminium forming an oxide layer; although the Al2O3 layer is smaller than the base metal rather than larger, cracks still form. Al2O3 doesn't conduct anything much at all though, unlike rust).
Instead the oxygen gets through the rust layer and makes more rust underneath. Making rust needs carbon dioxide and water as catalysts (that's why iron doesn't rust in dry conditions) and the CO2 and H2O are conducted through the rust too.
In electrolytic rust removal a lot of the chemistry happens in the solid rust, not in the liquid or at the solid/liquid border. The reactants are conducted through the rust, or for instance processes like this happen: electrons travel through the rust, converting ferric iron ions in the centre of the rust to ferrous ions; and oxygen ions diffuse out through the rust and react with hydrogen ions at the liquid/rust surface.
Having said all that, I'll write rust as Fe2O3, which is mostly it's chemical composition, ignoring water. Imagine rust is a special form of Fe2O3 - the ochres (used as pigments) are some other forms of Fe2O3, and there are plenty more.
Back to magnetite. Magnetite has several interesting properties too. The most important is that it is quite stable, insoluble in water, and doesn't easily oxidise to ferric oxide (or rust) - it is the black scale on black mild steel, and is also used to protect guns (I can't remember the name of the process, anyone? You let the barrel rust a bit under controlled conditions, then boil it to convert the rust to magnetite, and repeat).
Magnetite is about the same size as the iron it forms from, so no cracks form,; and it doesn't conduct oxygen etc like rust does, so it protects a surface and won't rust underneath. It's magnetic, hence the name.
The other property of interest to us is that in the electrolytic rust removal process it forms small hard lumps, which fall off or are brushed off by the bubbles.
So what happens in electrolytic rust removal? You can write ionic/redox half reactions to describe it, but I won't bother. Overall the main process is
3Fe2O3 + H2 -> 2Fe3O4 + H2O.
The H2 comes from electrolysis of water
2H2O -> 2H2 + O2.
The sodium carbonate is unchanged.
The electrolysis of water happens a lot, due to the comparatively high voltages and currents used. This is what causes the bubbles. The bubbles are important, as they physically remove the magnetite, and any paint and other stuck on debris. Although this happens mostly at the edges of paint flecks, hydrogen (as ~ ions, kinda) can actually get through the paint, turn to gas, and form bubbles behind it!
Some people say that iron is formed from the rust as well, but the evidence for this is a bit dubious. It may happen to a small extent at the metal/rust surface, and while it will happen under some laboratory conditions when the process is unusually prolonged and vigorous, it is not usual*. Nor would it be desireable.

Sorry, that bit is wrong, magnetite does conduct electricity. Rest should be okay.

Not all that well though. It's one of the ways of insulating electrical machine laminations. Put on in a steam oven. When we did it at GEC Large Machines it was claimed to have a better bond to the steel than most varnishes could achieve. You could certainly get a good black finish on steel tools and other items by putting them through the plant :-)
Mark Rand RTFM

Mark,
That sounds an interesting process - was it under pressure and dry steam or just good old wet kettle steam?
AWEM

IIRC it was dry/saturated stream and air at atmospheric pressure with the steel parts being heated first on the conveyor with the steam being admitted further down the line.
regards Mark Rand RTFM