The following refs are pretty much the same reaction over a progression of time.
JACS 85 2858 (1963)
JACS 86 5272 (1964)
JOC 34 3970 (1969)
JOC 35 261 (1970)
This reaction is Constant current i.e. easy and no special equipment because it is an indirect reduction, the lithium salt is reduced to lithium which does the work turns back into the salt which is reduced again. This is great since you only need a catalytic amount of Li salt which you can reuse.
In short:
To reduce benzene to the dihydro version, the following procedure was used:
An electrolysis cell 170 mm in length by 100 mm in diameter was fitted with two dry ice condensers with 2(2x5 cm) platinum electrodes. (Note none of these dimensions really mater just keep the current density the same and stick to platinum or platinised whatever as the electrode material.)
There is no cell divider for this reaction. (They do use a cell divider (asbestos) for some of their examples to further reduce the benzene to a single double bond. But Scooby thinks the undivided reaction reducing only one of the double bonds of benzene is the go for ephedrine. Maybe a Bee with more experience with this could clarify?)
Anyways the ratio:
To reduce 0.1 mol of your favourite substance add 0.8 mol of Lithium chloride and 900 mls of methylamine. The reaction temp is around -7 deg (refluxing methylamine). Apply 2.0 Amps for 7 hours (50 000C) the voltage they saw was 85 volts. Again this is a constant current reaction voltage doesn't matter your's will be different because your cell will different. Their electrodes were pretty far apart (170 mm) to reduce that crazy voltage bring your electrode closer together, this will also reduce the heat generated by the reaction. Then do a standard workup, the yields will be different since none of you are going to reduce benzene.
However they got 90+ % for their reductions.
N-Methylimine reduction to N-Methylamines
There are two ways, either add your LiCl (0.4 mol) and ketone (0.05 mol) and methylamine (300 mls) and let it sit for 6 hours. (The addition order is important adding the methylamine last creates heat solvating the LiCl which inturn speeds up the imine formation.) Then zap it, at 2 Amps for 2 hours and 41 mins, (19,300C) Then workup.
The second way is to produce your imine first isolate it then add it to the above reaction, this time using 0.2 mole of LiCl and 200 mls of methylamine and zapping at 1 Amp for 1 hour and 15 mins (4500C) then working up.
The yields of the ketones used by the ref were 73,70,70 for 2-Heptanone, Cyclohexanone and Cyclopentanone respectively. Yields were higher for the isolated imine method but this could have been tweaked by removing water while forming the imine instead of after this would have led to a higher yield in the imine and subsequently a higher amine yield. They added an 80% pure imine (not all converted) in which they got a 77% yield of amine.
This reaction could be modified quite easily, I doubt you have to use methylamine as the solvent, probably any solvent that won't react with Lithium and can dissolve the LiCl salt. It does matter for the Imine reduction since the solvent is also the amine sorce, Which is kinda cool whatever end product you want just have the ketone swimming around in it before reduction.
Also instead of letting the MeNH2 evap off bubble it into a dil HCl solution for reuse.
I'll get the most important refs scanned in and to Rhodium.