Case in point-- how exactly does dilithium crystal somehow manage to survive direct exposure to antimatter? (the technobabble suggests that it somehow is able to redirect all the enormous quantities of photons and hard neutrons produced by the reaction so that the energy can be harnessed, but really does not explain how. Considering that it would just be a pretty innocuous looking metallic prism, (and a highly reactive one at that, keep away from water and oxygen kids!) with a pretty low melting point.. this just does not make much sense at all.)
Noting that dilithium isn't Li
2, but something else.
Lithium has atomic number 3 (and mass typically of 7, sometimes 6) but Dt reportedly has Z=87. In our table, that's Francium, with estimated melting point of 'warm to the touch', though the most stable
223Fr has a half-life of slightly over 20 minutes. The difference between the "lithium"ness and "francium"ness was always explained by a
natural component of the atomic structure being displaced beyond the normal three dimensions of space (and, in certain circumstances, the four of space-time!). It seems to be the only (or most stable) element in this complex-plane "island of stability" so there aren't other di'unobtanium' elements, though given its inability to be directly replicated or atomically fabricated into existence by the tech of the time is shared with Latinum, maybe it's not so alone after all.
A dilithium
crystal is famously indistinguishable to pre-warp civs from quartz, SiO
2, so that mineral museums of Earth were 'raided' for it once the truth of its existence was determined. Crystaline dilithium is a mix of Dt with Si and Fe (IIRC, though 'technobabbly', presumably representing the extradimensional component of the 4D matrix) and this presumably creates the compound solid that acts quartzlike enough for normal operation.
When subjected to a certain field frequency (EM?), though, the resonances are such that in delivering a matter/anti-matter mix the crystalline structure is kept 'out of the way' of the streams that it otherwise moderates (without the crystal it'd just be too much instantaneous undirected energy from the annihilation, so it must regulate and disyribute the ±deuterium gases' introduction like a pachinko machine) and allows the energy produced by the mutual annihilation of equal parts of the
2H,
only, to create the plasma (from the carefully tuned over-ratio component of introduced positive-deuterium) that is tapped out to be sent off directly to the warp nacelles, or slightly less directly (via the GNDN ducts) to power the ship's other systems (transporters, replicators, lights/doors, the famously exploding command-consoles, etc; but sometimes
not the holodeck, because Plot!).
This all requires a carefully (re)configured crystaline structure, obviously, for optimal power, and extended/excessive use can distort the structure beyond usability, so it's not a
trivial thing to fire a
-1D stream into a (DtSiFe...) matrix. We've barely produced enough
-1H for research purposes. No wonder contemporary science hasn't yet
quite yet confirmed the undoubted technical accuracy of the currently most forward-looking descriptions of the process, trapped as we all currently are beneath the transparent alum(in)ium ceiling in this field.