Guys, i think it is important to differentiate between refugia and dsb!
1.) refugium---------> Refuge from predation for microlife to proliferate and their population to spill over to your display and serve as a food source for tank inhabitants. And to secondarily serve as a area where nutrients are processed into biomass by the population of microlife..
2.)DSB----------> Deep sand bed with a primary function of denitrification through creation of low oxygen envionment with some microlife which will end up as food in the display to feed corals and fish as a secondry function.
Now to answer your question: Substrate in your refugium should be as fine and soft as possible, as there is a direct correltion between particle size/shape and sandbed infauna. The finer, softer and rounder the substrate, the greater variety and quantity of microlife it will support.
If we follow this trend then surely silty mud/clay should be best??? Like the type found in mangroves swamps.
Well yes it would, IF we were trying to establish mangrove biotopes, ie areas of low flow, major detritus settlement, sulphur producing bacteria at shallow depths and a rich biodiversity. You are however stiving for a refugium to feed a reef, not an area to trap and process masses of nutrients. Personally i feel that mangrove swamps are too far removed from reefs to directly provide food to reefs.
Oh yes, your question............(s)
A few people would advocate the use of a dsb in a refugium, guess the thinking would be that if one cycles nutrients and the other provides live food, them surely the two combined in one space would be better.........
Yes sure, a dsb in a refugium will not negatively impact a refugium...........
HOWEVER, a refugium in a dsb will reduce the efficiency of the dsb.
Remember, with a DSB you are effectively trying to create an oxygen gradient by optimizing flow and reducing settlement of detritus to achieve better diffusion.
A popular way to add surface area for critters to colonise (and some added nutrient removal) to a refugium is by using some algae, preferrably something like chaeto. By having algae in a refugiums, you are slowing flow pathways and increasing detrital settlement.
I.e. interfearing with flow pathways and diffusion.
IMHO a pure refugium (as you are planning) would consist purely of an environment free of predation, which would encourage proliferation of micro and meofauna.
The optimal substrate would be aragonite, as there is no other substrate which can compete in terms of surface area and shape. (of course budget comes in to play here) Stick to a shallow sandbed, and keep a seperate dsb, with higher flowrates than in the refugium. Some small pieces of liverock should be included to provide extra biodiversiy and surface area.
If you want to, add some chaeto, although i believe this is best used in a seperate area as not to impede flow or cause too much detrital settlement an the shallow sandbed of your fuge.
Large pieces of crushed coral is an effective detritus trap, which will have to be siphoned regularly, impeding development of microlife colonies. (lets not even look at the potential as a no3 factory)