Unfortunately the kits aren't that sophisticated. Basically, its the sulphuric acid (H2SO4) in Marquis and Mandelin that initiates the reactions. It's important to remember that even reactions of pure substances with these reagents will more often than not result in several compounds being formed, some of which may be coloured and some of which may not.
Marquis also has formaldehyde & methanol, the later added to slow down polymerisation of the formaldehyde.
Mandelin is a mixture of ammonium vanadate & H2SO4.
The reactions of Marquis involve the formation of Ar-CH2-Ar bonds or Ar-CH2-O-CH2-Ar (ref: 2001, Bee Osmium, Ar = aromatic ring). Concentrated Sulphuric acid will cause breakage of the methylenedioxy ring and I would think, most ring substituted methoxy groups and some aromatic rings as well, with the result being a mixture of further (still) reacting compounds, some of which undoubtedly contain sulphur. These sulphur containing compounds are likely to form the colours. Formaldehyde is quite a reactive substance, especially when in an acid catalysed environment and so would extend the number of expected reaction products. The amine of the molecule plays some part, as although not easily distinguishable, MDA, MDMA & MDEA, all produce subtle variations in the reaction colours/ time.
Mandelin is a bit different although most initial reactions are likely to be the same as with Marquis. As mentioned, Mandelin also contains a Vanadium compound. Vanadium forms what's known as complexes where bonds between Vanadium and other groups (e.g. amines, cyanide, broken -O-CH2-O- or CH3-O bonds, aromatic ring etc) can produce bright colours. The bonds formed are quite a bit more complex (pardon the pun) than normal covalent or ionic bonds, and in many ways are similar to the way drugs bind with receptors using orbital overlaps influenced by subtle Van der Waals forces. Drugs as well as groups attached in some metal complexes are termed Ligands. Molecular orbital theory gives an insight into this marvellous world.
Although these mechanisms are likely to play some role - at least in the initial reactions - it is possible that Marquis doesn't show colour with PMA because the para methoxy group of PMA doesn't react with sulphuric acid. It would tend to explain the absence of colour with Marquis. However, even if this is so with Maquis, the H2SO4 in the Mandelin would definitely play a part in the PMA-Mandelin reactions.
What I'm interested in knowing is whether pure PMA, like pure MDA will cause a violent reaction with Marquis. I know there's no colour, but is there any noticeable reaction?
The number of possible compounds formed by reaction with Mandelin is considerable. As many metal complexes have distinctive colours, it's understandable that Mandelin should test for a diverse range of compounds, producing distinctive reactions and colours for many amines and ring substituted variations. Isolating and analyzing these compounds would be tricky as several would be formed with any one substance, so attempting to isolate and work backwards from a single PMA-Vanadate compound would be pointless, even if attainable. However, there are things which can promote formation of certain complexes, so it's possible mandelin could be improved.
The problem with any of these tests is that they are not specific enough, nor can they account for the endless possible combinations of both active and non-active things which may also react. Combinations of things - as we know - can fool the test or mask another reaction colour. This is of course what happens when a mixture of PMA and MDXX are tested with Mandelin. PMA on it's own would show a distinctive colour change, but the darker colours from the MDXX reaction conceal the lighter colours of the PMA reaction.
It would be fantastic to have a methoxy only reagent test. One that would not react with anything else, particularly the methylenedioxy group. There are several compounds I'm aware of that are specific to methylenedioxy groups. i.e. methoxy won't cause a colour change. But that doesn't help things much. Trouble is, the methoxy group is more stable than the methylendioxy group, and both are similar structurally and reaction-wise.
I've spent countless hours pursuing this goal. I have some interesting leads and possibilities, but they will have to wait until I finish another part of my project and see whether the department is fed up with me or not. Smooth sailing will hopefully see me being able to test a few of these reagents I have in mind.
Another area I hope to pursue further (money being the limiting factor at present) is that of UV fluorescence. Small detectors are becoming attractively priced, but separation of compound is usually first required. Methoxy and methylenedioxy groups are easily distinguished between in this spectrum. However, it may be possible to distinguish the presence of a methoxy group in a mixture. To be sure of results, this of course needs to be researched thoroughly, using various ratios of PMA and say MDA, as well as different excipient and binder mixes. It should work, and there's always the possibility of making derivatives if it doesn't. Derivatives can often produce a very distinctive emission spectra.
Yet another approach which looks promising (if you can afford to do the research) is immunolabeling. I've spent hours scanning databases looking for the right antibodies. I believe this can be done, although many of the most suitable antibodies must be stored at -20 degC and so don't lend themselves to the flexibility required for a user based test. However, many companies offer to produce specific antibodies to order.
IMO, the most ideal test may be antibody based (employing Europium as a long decay fluorescence element) used with a simple fluorescence detector. Specific antibodies would be used against a range of substances, enabling a quantitative accuracy far greater than what's required.