[Updated 2019-1-21] Complete Compendium of Modded Materials

Not just oxygen, H20. Hydroxides are lighter molecules than Per/oxides and the reaction also produces hydrogen gas. A Fluorine lithium reaction might be good as well.

Not necessarily. The Dicyanoacetylene Ozone reaction is the most energetic chemical reaction known to man but has lower exhaust velocity than LOH H2 due to the relatively large molecules it creates. The molar mass of the products is as significant as the reaction energy.

Water as an oxidiser is expensive compared to classic O2 and F2, and less dense than both. Liquid alkali metals also get more expensive than the very cheap solid versions, especially lithium; even sodium at 2.22c/kg is relatively pricey compared to hydrocarbons and much if not all of its density increase goes toward making up for the water's loss. There is however a water-fluorine reaction which was added to the Stock Reactions Extension at some point. Water and fluorine is pretty much the densest and cheapest combination of this kind and the absolute maximum velocity is 2.98km/s, about as good as normal hypergolic bipropellants (hydrazine et. al.) but with a much better overall mixture density of around 1550kg/m³. Worth playing around with, I think.

This can already be found in the section for lithium propellant. I thought you would be aware of it? Well if not, try also the fluorine lithium hydride reaction, which is pretty ridiculous (>5km/s) and does a similar thing to your idea of reacting lithium with water, except the hydrogen comes bound to the fuel instead of the oxidiser.

I built an missile using that fuel chemistry. Exquisite performance from a chemical system, combined with strictly superior volumetric density compared to Hydroflourine.

How about adding CNRP (Carbon Nantube reinforced plastic)? It's already in use in the aviation markets. Most notably on the F-35 and Dreamliner.

It Would probably make for a good armor and fuel tank material option.

Rocket Witch I noticed this compendium contains both liquid aluminium under propellants and molten aluminium under metal arclamp vapors. And some of the stats seem different.

The metal arclamp vapours are someone else's work and part of a set, while liquid aluminium is my own standalone. Not wanting to arbitrate what makes it onto the compendium, as it's meant to be a catalogue of data, I've included everything. The different stats come from molten version just being normal aluminium with a viscocity added to make it liquid and thus usable in lasers, while the liquid version is actually researched with correct density and everything.

Are you using NSWR with all the other future propulsion stuff? If so I'll remove the warning next time this is updated, as whatever issues a couple people had in the past is probably just load order or something, and this kind of thing should be a non-issue nowadays.