As part of the experiments to test for the release of aluminum from Phosguard, I tested a sample of water from my aquarium as well as freshly made Instant Ocean (both at a salinity of 35 ppt).  I used a new Varian ICP-OES in an attempt to detect aluminum in these samples (none were acidified or filtered unless otherwise noted).  However, I was unable to detect anything but noise in these samples.  That is, in looking at the emission spectra themselves, I simply saw a noisy background by eye, without any significant peak at the known emission wavelengths for aluminum (I used 237.312, 308.215, 394.401, and 396.152 nm).  Spiked samples (standards made by adding soluble aluminum to the samples of aquarium water  and Instant Ocean salt mix), and those where aquarium water or Instant Ocean water were exposed to Phosguard (below) did have clearly defined peaks at these wavelengths. 

Figure 2:The ICP emission spectrum between 237.2 and 237.4 nm of a sample from my aquarium (blue) and the same sample after being spiked with 0.5 ppm aluminum (red).

Figure 2 shows some typical ICP data for these samples.  The blue line represents the emission spectrum displayed by my aquarium water sample.  It is actually the average of three testing replicates taken from a single water sample.  The red line represents the same sample with 0.5 ppm aluminum spiked into it.  Clearly, the spiked aluminum gives a clearly defined signal exactly where it is supposed to be, while the aquarium water sample shows no such peak.  This same conclusion applied at each wavelength examined, and for the Instant Ocean samples as well.  From these data, I conclude that the concentrations of aluminum in my aquarium sample and in Instant Ocean are less than 0.05 ppm aluminum.

Inputs of Aluminum in Reef Aquaria: Foods

Foods are, of course, another potential source of aluminum. In a study of the amounts of different elements in certain foods,²⁵ Shimek presented the results shown in Table 2.  The values have also been normalized to show the amount of aluminum in the foods in relation to the number of calories provided.  Clearly, if aluminum is of primary concern, brine shrimp (highlighted in red) should probably not be on the menu. 

If you fed 5 grams of it to a 100 gallon tank every day, that would amount to 5 g x 120 mg/kg = 0.6 mg/day or 219 mg/year.  Added to that 100 gallons (379 L), that gives an addition of 219 mg/378 L/y = 0.6 ppm per year.  The other listed foods would, of course, contribute much less.  Unfortunately, these sorts of raw measures of aluminum say nothing about what form it is in.  For example, it might be present as soluble aluminum, or as insoluble (particulate) forms.

Inputs of Aluminum in Reef Aquaria: Calcium and Alkalinity Supplements

Another significant source of aluminum are the calcium and alkalinity supplements that aquarists use.  In a recent paper on metals in aquaria,²⁶ I quantified some of these inputs, and the results for aluminum are summarized below.

Limewater (kalkwasser) is made by dissolving calcium oxide or calcium hydroxide in water.  The calcium oxide that I use from the Mississippi Lime Company is food grade, but still has certain impurities.  The typical analysis of this material shows it to contain 0.10 % aluminum.  It is not obvious what form this takes but since aluminum is quite soluble at pH 12.4 (total solubility = 80 ppm at pH 12.4,²⁷ if saturated limewater were made from CaO with 0.1% aluminum, it would contain 1 ppm aluminum) it is a reasonable hypothesis that it dissolves into the limewater and is delivered to the aquarium.   If one were adding 2% of the aquarium volume in saturated limewater (0.0204 moles/L CaO) every day for a year, one would have added the equivalent of 8.3 ppm aluminum.

The amounts of metals that are added to an aquarium when using a CaCO₃/CO₂ reactor can also be determined. The impurities present in such media varies with the source or brand of the media, as has been shown in different articles by Craig Bingman ²⁸ and Greg Hiller ²⁹.  If we make the assumption that we want the same total amount of calcium and alkalinity as in the limewater case described above, then we can calculate the following amounts of metals added over a year: 

As can be seen, the amount added over the course of a year can be quite substantial, but is less than is delivered via limewater (at least in these calculations).  To be honest, I’m not sure why there is less aluminum delivered by CaCO₃/CO₂ reactors than limewater. Calcium oxide is made by heating calcium carbonate until the carbon dioxide is driven off.  Such a conversion shouldn’t impact aluminum concentrations.  Perhaps the difference simply reflects artifacts in one or both of the testing methods, contaminants in production, or in the nature of the original calcium carbonate chosen for each process.

Other methods of calcium and alkalinity addition presumably also deliver some amount of aluminum, though I’ve not seen any analysis of any of them to comment further.

Inputs of Aluminum in Reef Aquaria: Phosguard

Many aquarists claim to see undesirable effects on corals when using Phosguard, made by Seachem. Many aquarists have attributed that effect to released aluminum, since it is largely composed of aluminum oxide (possibly with silicon present too).  In the first phase of testing that hypothesis, I examined whether Phosguard does indeed release any aluminum into solution.

Table 4 summarizes the results for a serious of samples in which commercial Phosguard (75 mL) was placed into contact with aquarium water or freshly made Instant Ocean artificial seawater (500 mL).  The samples we allowed to sit in closed plastic containers.  Once every 3 days or so the containers where gently shaken for a few seconds.  Aliquots were removed, in some cases filtered through a 0.45 μm filter to remove “particulates”, and the aluminum was determined by ICP (without acidification).  The concentrations were determined by comparison to standard made by spiking 0.5 ppm aluminum into aquarium water or Instant Ocean artificial seawater (which had been shown earlier in this article to have no detectable aluminum).  All of the samples had a clearly definable emission peak in the appropriate place, although the lowest sample (0.06 ppm) is close to the limit of detection.

From the results in Table 4 it is evident that Phosguard does release aluminum to the water, and that the majority of this is present in particulate form (that is, that it is removed on a 0.45 μm filter (although that does not demonstrate that it was originally released as particulates).

In order to determine if these results are caused primarily by fine particles that come with the much larger Phosguard particles (typically about 2 mm spheres), a batch was rinsed very thoroughly with RO/DI water (8 times, with each rinse lasting about 1 minute and each rinse volume comprising about 20 times the solid particle volume).  These rinsed Phosguard particles were then exposed to aquarium water as above.  The results are shown in Table 5.

Not surprisingly, the concentration is reduced in the unfiltered sample, indicating that the rinsing may well have removed some fine particles that were contributing to the results in the unfiltered samples.  However, the aluminum concentration in the filtered sample is not reduced, indicating that the “dissolved” fraction of the aluminum is not altered by rinsing the Phosguard first.

Significance of Aluminum Release from Phosguard

Is the amount of aluminum released from Phosguard significant?  Moreover, is it adequate to explain the results on corals that have been reported by aquarists?  This question is extremely difficult to answer without some biological experiments.  The tests run above show reasonably high concentrations of aluminum. Possibly high enough to cause problems for the organisms shown in Table 1 .  But these tests were carried out on a large amount of Phosguard in a small amount of water.  Tests with larger volumes of water might well result in lower aluminum concentrations.  Additionally, the exact nature of the aluminum in these tests may well be different than in the toxicity tests reported above.  That is, the nature may be particulate vs. colloidal vs. soluble vs. complexed by organics, etc.

Biological Testing of Aluminum Exposure

In order to more definitively show whether aluminum released by Phosguard might be the cause of reaction in corals, it seems prudent to test aluminum on corals.  Toward this end, I set up a 30-gallon aquarium with several corals.  These included a leather coral (Sarcophyton sp.), green star polyps, and brown mushroom corals.  In addition, the tank contained sand, live rock, and some macroalgae (Chaetomorpha sp. and Caulerpa racemosa).

Pictures were taken of the several corals, and then aluminum was added.  Figure 3 shows the leather coral before any addition.  The first aluminum addition boosted the aluminum concentration by 0.005 ppm by adding 0.15 mL of an aluminum chloride solution to the water.  A pH electrode was in the water at the time and recorded no pH change (pH = 8.32).  There was no apparent change in any corals in 2 hours.