This experiment started out as an attempt to make a mineral acid. Mineral acids are made from minerals such as green vitriol (Iron sulphate), saltpetre (potassium nitrate) and others. They were used by medieval alchemists to dissolve metals.
The origins of them are still a little unclear – I thought it likely, based on what I have read, that they were invented by Arabic alchemists some time in the 12/13th centuries AD, following in natural progression from the distillation of solutions of the minerals mentioned already. Others think they were invented by European Alchemists in the 13th century, especially in Spain.
One way to prove the argument one way or the other is to find an Arabic text which tells you how to make acids, from the right period. But the earliest definite recipe we have for nitric acid comes from a work of pseudo-geber, writing circa 1300AD in Europe, and thus not Arabic at all. It uses iron vitriol, saltpetre and alum, distilling them together to make nitric acid with some sulphuric acid from the alum.
One researcher, Ahmad Y. Al-Hassan, reckoned he had found an Arabic recipe. In a Liber Luminem Luminum, attributed to Michael Scot by historians in the 19th century, but probably not by him, there is a recipe which says (in the 19th century printed version):
M. cum sossile et nitro salso ana in aqua resolutis ac coagulatis es ad naturam lune reduxi. R. vitrioli romani Libra 1. salis nitri libra 1 . salis armoniaci 3 . 3 . hec omnia comisce in unum terendo et pone in curcubita cum alembico et quod distillaverit serva et pone cum m. crudo ita quod in 3 aque fundatur super mediam libram m. in una ampulla et pone in cineribus bene clausam et da lentum ignem per unam diem et postea invenies m. in aquam purissmam
Translated it becomes
“This text indicates that ‘M’ (usually a contraction for mercury in alchemical texts) must first be purified by being placed with ‘sossile’ and spirit of Nitre. [‘Sossile’ I do not recognise].
When you perform a recipe, grinding together 1 pound of vitriol with 1 pound of nitre and 3 pounds of sal ammoniac, which you then heat in a flask and distill off a water.
Then you are to place the purified ‘M’ (mercury) from your first stage and place this in a flask with three more parts of this acid distillate. The flask should be well sealed and heated gently for a day. After this you should find mercury in this most pure water.“
Okay, I thought, that should be do-able.
So I mixed up the substances as required, although without so much care paid to the weights, but if anything with less sal ammoniac than was written.
It looked rather yellow as the ingredients began to react with each other.
This is the setup with the luted cucurbit to help support it at high temperatures. This technique goes back at least 600 years. You cover the glass with a mixture of clay sand and horse dung and it fires and supports the glass when it gets hot and starts getting more flexible. At which point it looks like this:
I then heated my alembic for nearly an hour, and what came over first was an orange liquid. The water part being of course the water within the minerals, since they trap water in their structure when crystallising. I thought it was orange because of nitric oxides, probably nitric oxide, which makes orange solutions.
Here’s the first vapours coming into the receiver:
Then when the temperature was up over 600°C as measured by my thermocouple, a lot more white fume came over and there was a smell of chlorine. Obviously that was from the sal ammoniac.
When bottled it looked like this:
Not having anything else to add to it, I put it in the garage for a week or two since was quite busy.
Finally I got around to checking the pH of this allegedly acidic liquid.
It was 9.8. Rather alkaline, not acidic at all. (If it was nitric acid it would be around 1 to 2, normal water being neutral at 7) I had calibrated my pH meter a week earlier, and the alkalinity was confirmed by pH paper.
Something was wrong!
As it happened I had also been working on some recipes of Rhazi, the famous 9/10th century arab alchemist. One required the distillation of a mixture of salts that included sal-ammoniac. It was called “The water of salts”, from the Kitab al-Asrar, the “Book of Secrets”.
Oddly enough the water distilled from this, despite what well known early researcher Stapleton thought, produced an alkaline liquid too, not an acid.
The explanation is pretty clear – the sal ammoniac provides ammonia, which can be smelt very strongly from both alkaline liquids when they are heated.
In the first experiment, heating the minerals causes the dissociation of the ammonia from the chlorine, and their solution in the water that came over. The potassium nitrate also breaks down at the same time, and for some reason is either neutralised by the ammonia or sufficiently oxidised that it just forms the orange nitric oxides and goes into solution.
Or in other words, al-Hassan was wrong about the recipe being to make acid.
Of course there are quibbles to be had. Maybe if I had stopped the distillation early enough, i.e. when a water was first formed, the liquid would not have been so alkaline, but then given that sal ammoniac and saltpetre break down at very similar temperatures, it seems unlikely that the result would have been different.
When I heated some of the orange liquid with a piece of silver, all that happened was that the silver got cleaner, probably the outer layer was stripped off by the ammonia or sal ammoniac. On the other hand I should try it with talc, since that is what rhazi suggests. I’ll try making some more liquid and testing it on a lump of talc (When I get one) and a calx of metal to see how it softens it as the recipes say. As an alkaline substance it may well soften sulphur well which is also referred to.
The chemistry is a little unclear – I’ve not needed to use basic chemistry since school, insofar as in ‘the good old days’ chemists would memorise the reactions of lots of elements and compounds and in theory in the modern period you can work it out using the electron configurations. An interesting thing though is that Ammonium nitrate is formed by neutralising nitric acid with ammonia, which may well have occurred in the distillation of the first recipe. It is a white crystalline solid, so I should try crystallising it out.
So I’m having trouble finding anything about mineral acid recipes that is definitely Arabic in origin. At the least, this experiment suggests that any distillation recipe which includes sal-ammoniac is not going to make an acid. Given that many such Arabic recipes use sal-ammoniac, it seems more likely that mineral acids were an innovation of Christian alchemists working in Spain or somewhere like Sicily, where there was a great deal of cultural cross fertilisation. Of course the best way to find out would be to read more of the original translated texts. But they are locked up in modern copies of Latin texts or manuscripts which nobody has looked at properly for years.
I’m picking up Latin as I go along but it’s taking a while.
The moral of this tale is that a chemist is only as good as their books and their experience of specific practical procedures is.
It also shows that we have a rather poor understanding of the chemical work of the medieval alchemists. More research is required of course, and more experimentation.