On the Hive, I remember reading about the speculative use of dimethylaminoethanol and indole to prepare DMT directly, in a bomb. I don't know if that is even possible, but it occurred to me recently that a dimethylaziridinium salt, which would certainly condense with indole, could be prepared from DMAE.

I have some papers which are not with me at the moment (I'll post them when I can) but from memory, higher temperatures caused relatively more condensation at the 1 position, but then even at the boiling point of indole the yield was some 40% I think. 2-methyl indole had a dramatically higher yield for what its worth.

There are industrial processes for the gas phase dehydration of aminoethanol to aziridine and they all seem to involve the oxide of wolfram and sometimes silica and other "stuff". Perhaps WO3 (and/or silica, or some catalyst or another... alumina itself dehydrates ethanol at such temperatures in the gas phase) works just well enough in the liquid phase to allow some reaction of indole and DMAE, yielding DMT and H2O? A toluene solvent would both lower boiling point to possibly more desirable temperatures (depending on catalytic activity) and help remove H2O as formed.

'Cause that would be just spectacular (except for removing the impurities).

Alternately an acid might do the trick provided the anion is non-nucleophilic (and on this I have another most fascinating paper, sorry to tantalize!) the bisulphate salt which is the standard method to aziridinium ions from aminoalcohols I forget the name of the rxn.

Very very sorry to keep you waiting!

2-Aminoethylation of Indole and of 2-Methylindole briefly describes aziridinium BF4- and indole to tryptamine: 40% yield at 70 deg C and 20% at 130 deg.

Nucleophilic ring opening of aziridines describes on p. 2706 the "InCl3-catalyzed ring opening of aziridines with heteroaromatics" in fairly good yield. The aziridines in question are N-tosyl rather than aziridinium salts.

 (note the lewis acid; some kind of which might be an important catalyst to add for a hypothetical one pot dmt)

Reaction of Indole with Aziridines has some interesting little hints: "It is shown that the presence of electron-donor substituents attached to the carbon atom or of bulky substituents attached to the nitrogen atom in aziridines favors cleavage of the aziridine ring by indole to give derivatives of tryptamine and tryptophan... However, 2-phenylaziridine (I) reacted with indole under the influence of boron trifluoride etherate and particularly readily when it was heated with tetracetyl diborate... A bulky substituent, by hindering polymerization, favors reaction with indole."

The importance of a non nucleophilic anion is for the stability of the aziridinium ion itself, to which nucleophiles will readily add, but like mustard even the addition product (for halogens at least, right?) will recyclise in some equilibrium and go on to react with the indole (though possibly at the cost of yield by extending rxn time). So that's presumably a lesser stumbling block.

In addition to the attached, the patent Continuous process for the preparation of alkyleneimines might be relevant as it summarizes known preparations of aziridine from aminoethanol, including through the hydrogen sulphate, which is known as the Wenker process/reaction   -  http://www.freepatentsonline.com/7683216.html

Aziridines are of course toxic and mutagenic, and luckily relatively unstable or at least impersistent. While keeping the temperature lower increases yield, even calling it a 10% yield of DMT from indole and DMAE in one shot would be a godsend.