I can't remember who said it anymore, and I don't really want to single anyone out, but the statement that buying cheap sand that contains quartz will ruin your tank is plain-and-simple bunk. I'm not sure where the idea that silica sand is dangerous to a reef tank came from, but typically silica sand is 99.0-99.9% SiO2 (depending on the source and grade), which is about the exact same chemical composition as the glass of your aquarium. If the addition of pure quartz sand is somehow dangerous to keeping a reef tank, we'd better all get our animals out of glass aquaria...
Quartz (SiO2) is considered "totally insoluble" in water according to the US MSDS, and is also nontoxic (although inhalation of silica has many well-documented health risks for humans -- I'll explain at the bottom if anyone cares). Yes, water is a "universal solvent" and yes, everything (including the silicone) dissolves slightly into the water over time, but the amount of dissolution is so low that it is impossible for it to make a difference to your aquarium. There certainly are highly soluble forms of silica that will increase the level of dissolved silicates in the water (such as aluminosilicate) and are likely to cause problems, but quartz sand (SiO2) is not one of them.
The fact is that quartz sand (and the walls of our aquarium and even the silicone rubber which is the most soluble of the lot) do not dissolve enough in seawater to be measurable If silica sand contributed in any significant way to dissolved silica, then you would expect there to be big differences in the silica concentration around sandy beaches and on calcareous beaches (such as the red "sand" of Bermuda, which is composed mainly of calcareous foraminiferan skeletons) but there are not -- although calcium concentrations do vary significantly, the silica concentration in either location is about the same (roughly 2 ppm everywhere other than adjacent to the mouths of rivers where FW inputs increase the level). That suggests to me that quartz sand doesn't make much of a difference to the silicate concentration of seawater. Silica gets into water by being in a more soluble form than SiO2 (such as aluminosilicate), and the most common source of contaminating silicate in aquaria is the freshwater used for top-off or mixing. In fact, normal river runoff entering the sea has 2-5 times the amount of dissolved silica present in the surrounding seawater (which as I just said is higher than the norm), and researchers studying oceanic silica cycles consider quartz sediments a "dead end" for silica (so little is released it does not contribute to the global silica budgets of the ocean -- if it doesn't make a difference on a global scale with all the silica sand in the ocean, how much difference do you think it can make in our tanks?). The major input of silicate into seawater remains freshwater runoff into the sea, not the minuscule (and unmeasurable) amount of dissolution from the *enormous* amount of quartz in the sea...
In fact, that same quartz sand that people are recommending against is what was smolted and fused to form the glass walls of your aquarium... I don't know exactly what (if any) chemical changes are involved with the smolting process, but according to the glassblower for the Department of Chemistry, it's just melted and reformed into the appropriate shape -- it's not really doing anything to the sand other than burning off any organic contamination in the sand (the melting point is about 3110F). Even after being formed, glass is still SiO2, so there isn't any reason to suspect that there are important chemical changes occurring. Also, there should be no changes occurring as the sand passes through the guts of the animals in the tank -- is no noticeable degradation of the SiO2 spicules from ingested sponge tissue as that passes through the guts of animals (such as angelfish, sea stars & urchins) adapted to eating sponges, and that's where you'd expect some effect of digestion if any was going to occur. Given that, it's pretty hard to argue that using quartz sand is bad when the glass box that you're putting it into is made of the same stuff.
The standard recommendation to decrease silicates applies to your *water* not your sediments. If you look through D&S, Moe, Spotte, Adey and Ron's articles they all have a sentence buried somewhere that says something to the effect of the use of silica sand for sediments is fine, but that carbonate sands have a couple of advantages that make it more desirable (primarily the ability to absorp phosphate, and the buffering capacity, which I'll come back to later). Somewhere along the line, someone got it in their head that keeping silicates low in the aquarium meant not including quartz sand, because that is primarily silica. This "fact" has spread across the net, but it's another one of those things that "everyone knows" but that lacks any convincing evidence (or logic for that matter).
Having said that, however, some diatoms are apparently capable of pulling silica out of glass or quartz, and if when culturing phytoplankton in glass containers, additional silicates are usually not added to the culture medium because the algae seem to get what they need from the walls of the flask (hence to common observation that diatoms grow fastest on the walls of the aquarium). Even with the presence of lots of dissolved silicate in the tank, however you still need the other nutrients (nitrates and phosphates) to get a bloom of diatoms. In fact, diatoms turn out to need *more* nutrients in culture (people generally also add glutamate & glycerophosphate to culture media to really get diatoms growing well) than standard microalgae. If there are high silicates but undetectable levels of nitrates and phosphates, you're still unlikely to ever get a significant increase in diatom growth. Some people *do* report having diatom blooms with silica sand (although the cause for such blooms is unclear). Stories about "I used quartz beach sand and the diatoms exploded in my tank" may be coincidence, it may be that the diatoms were already there growing on the glass, but the addition of sand gives them more surface area to inhabit and therefore they seem to bloom, or it may be that by using the beach sand, those diatoms were *introduced* to the tank in the first place. I think that the "dreaded diatom bloom" is as likely to be attributed to the introduction of the diatoms as to the provision of nutrients...
In fact, I have a couple of tanks that are set up on washed silica sandbeds from the hardware store, and I don't see any problems with them. When I first started to get into using deep sandbeds in my tanks, there were *no* carbonate sands available through any sources, so I had no choice but to use silica sands to establish my early sandbed tanks. Those tanks are still set up and running just fine back in Canada despite the fact that my father (who has a black thumb of coral death) took them over when I left home many years ago . It makes no difference to the animals in the sandbed what the grains are made of, and silica is actually more resistant to size degradation over time than are carbonate sands (which continually get smaller due to dissolution and passage through the digestive system of animals in the tank). Of course, therein lies the benefit of using aragonitic sands -- the dissolution of these sediments allows for some increased buffering capacity and calcium release into the aquarium and the organisms which ingest them. That is a benefit in itself and a good reason to use carbonate sands if they are available. However the buffering capacity of various carbonate substrata is also overblown. Laboratory experiments comparing the buffering capacity of crushed oyster shell (composed primarily of the most soluble forms of calcium carbonate -- high magnesium calcite and aragonite), crushed coral gravel, and dolomite sediments compared to that of silica sand. Of course silica sand had no buffering capacity and the tank pH dropped below 7.0 within 90 days. The other sediments didn't work so well, either -- after two months in a tank with a reasonable bioload, the pH of seawater in the tank with dolomite was 7.4-7.5, the crushed coral/aragonite was a little better at 7.7-7.8, and the oyster shell was the best at 7.8-7.9 -- again though, this was over only 2 months. In any case, none of them buffered the pH enough to rely on the capacity of the sediments alone to provide buffering over the long-term.
I'm not saying that anyone should stop using carbonate sands, or that there aren't advantages to doing so. I'm simply trying to point out that this is another one of those great urban legends that runs around in the hobby. Once everyone "knows" it's true, no one seems to question it anymore, no matter how improbable the factoid seems. I certainly prefer to use oolitic sand and have in all my recent tanks, but I'm trying to explain that the sandbed critters couldn't care less for the most part (there are obviously some exceptions) as long as the particle size is correct, and that it is not *necessary* to do so.
Just had to get that off my chest...
PS -- As for the health risk to humans, silicosis is a respiratory disease resulting from the inhalation of crystalline silica. It typically occurs after decades of exposure (but can occur quickly in cases of massive infusion of freshly crushed silica), during which crystalline silica is trapped in the lungs, because the crystalline particles repeatedly injure the lining of the lung wall, and we mount an immune response (technically a nodular fibrosing pneumoconiosis, if anyone cares ). There is more to the disease than simple cutting, however. When crystalline silica is deposited in the lung, SiOH groups on the surface of the particle eventually form chemical bonds to membrane proteins and phospholipids, and these chemical reactions with the lung lining lead to denaturation of the membrane proteins and phosopholipids. During crushing of crystalline silica, surface free radicals are formed (half life is about 30 hours), and this makes the effects *much* worse in freshly ground silica than aged silica. That is why people who do sandblasting, mining or other rock work with quartz are at risk. Our fish, cukes and the like are generally not crushing the sand grains, they simply pass them through the gills or gut, and the particles are certainly aged by the time that you could get them (unless you're crushing your own rock to make sand ). Furthermore the surface is completely coated by a bio-organic film (primarily bacteria and their by-products) within hours to days of being introduced to seawater, and the surface of the particles are unlikely to ever come into direct contact with the gills or gut of the sand fauna for long enough to start the chemical reactions that are damaging to the tissues.
Addendum 4 Aug 2000
I fired that message off one morning without a lot of thought and never really intended it to become the reference article it has ended up being. After some pretty extensive discussions of this subject with a number of people (I especially thank Randy Holmes-Farley and Craig Bingman for their input), I wanted to post a clarification of the original comments.
First of all, I was speaking specifically of quartz sand in that post, and I must emphasize that the comments in it apply only to > 99% quartz sand that does not contain a significant amount of feldspar (or any other particularly soluble silica compound) contamination. If you plan to include a quartz sandbed in your tank, and are worried about possible diatom growth, it is advisable to get a chemical analysis from the company (many list this on their web pages), or get a good silica test kit and soak the sand for a couple of weeks in a little DI/RO water before testing it to see if there is any measurable amount of silica liberated by the sand. Randy and I discussed this at length in the Reefs.Org archives if you're interested in the nitty-gritty details.
Second, Craig Bingman has corrected me on a couple of points that I made in my original post. It turns out that the folks in the glass shop gave me 2 bad pieces of information: 1) that silicone rubber contains a lot of aluminosilicate (which is among the most soluble forms of silica) impurities from the manufacture, and 2) flat glass is simply melted quartz sand. Both of these are incorrect. The first may be true of certain cheap silicone sealants, but is apparently not true of anything that would be considered for use in aquaria, so that comment ought to be ignored. The second point turns out also to be untrue, and Craig tells me that almost all glass (which is really a supercooled liquid where the viscosity has become high enough for them to be mistaken for a solid) have other things dissolved in them to alter its physical properties. For example, most plate glass that is used by humans is soda lime glass (roughly 2/3 - 3/4 SiO2), which has sodium oxide and calcium oxide dissolved in the melt. These compounds lower the melting temperature of the glass, thereby making it easier to work with, but also make the glass much more soluble in water. So even though diatoms can pry silicic acid or silicate out of glass aquarium walls fairly easily, the same is not true of quartz sand grains. Craig's comment was that it may be possible for diatoms to get their silica requirement from quartz, but "quartz is the ultimate hard case for silica-loving animals, and it turns out to be much easier for critters to pull silicates out of the walls of the aquarium than out of a quartz sand bed." In fact, depending on the source of the aragonite added to an aquarium, a carbonate bed with a reasonable inclusion of biogenic opal (in the form of ancient diatom frustules and sponge spicules, etc.) could well provide more available silicic acid to the system than would a really clean, crystalline quartz substrate...
Craig is working on a detailed article on the chemistry of silica in aquariums, and in the next issue of Aquarium Frontiers, he will argue that silica is not the bugbear that we have all come to worry about. I would strongly suggest people ought to read that article to find out a lot more detail.
Anyhow, my personal feeling here is that the bottom line is still the same, but some of the details are a little fuzzy there (sorry folks) and I wanted to correct them for the record.
Last modified 2006-11-24 18:41