Many compounds will produce dark, sometimes brown colored reaction products with Marquis. However, the orange result definitely sounds like amphetamine or meth. An important question; is your Marquis reagent very old? If the liquid before reaction is a medium-dark brown color then the reagent has passed it's effective use by date. It usually still works with larger samples, but for some substances, sensitivity is affected, and exact interpretation of reaction colors can be difficult.
Although it seems from the purple color you observed, that some MDMA was present in the tablets you tested, it's worth noting that a few years ago in Australia, reports surfaced concerning pills containing a mixture of safrole (or sassafras) and meth. The pills smelled of safrole so people thought they were MDMA (note; odor is never a reliable indicator - pure MDMA should have no smell) and the darker color produced by the safrole with Marquis dominated the color observed, so it appeared as though MDMA was present. Further to this, the methamphetamine present gave a positive result with Simons reagent, indicating a secondary amine.
Marquis reaction with safrole/sass produces a dirty red-->brown color that slowly turns darker. Without looking carefully, it can be easy to mistake the red as purple and the dark end product as being black. It was found that by doing the test on a piece of glass with a light behind, the color difference between MDMA and sassafras oil reactions was more apparent.
Alkenes such as safrole can be tested for using a permanganate test, but this can also react with meth/amphetamines so they need to first be removed.
A 1/4 of the tablet is crushed and dissolved in a test tube containing 5-8 mLs of hot distilled water. This is then stirred/ shaken for 3 minutes. The water is filtered off using a bleached paper coffee filter, and the residue dried.
A few small crystals of potassium permanganate (condy's crystals) are dissolved in a 5 mLs of ethanol (only a tiny pinch of crystals - enough to produce color). It takes a bit of shaking to fully dissolve the crystals. 2-3 drops of lemon juice is then added and the pink solution shaken again. A couple of drops of this is then transferred to the dried residue. Loss of colour indicates the presence of an alkene.
