Part 2: Chemical Filtration

By: Robert T. Ricketts





This is my weakest area in filtration, so this section of the series is not going to offer as much immediately useful information as many might want. But I will try to provide a few hints on some of the chemical filtration processes, and cheat a bit by edging over into water modification at the end (the stretch isn’t that great, these are either chemical or mechanical filtration functions after all).


Granular activated carbon (GAC) or activated carbon (AC) is, for some strange reason, somewhat controversial. People argue heatedly for or against its use. Manufacturers supplying the aquarium trade do tend to rather exaggerated claims on what it will do, while sliding past what it will not do. Occasionally, they outright lie, but mostly it is just normal and standard marketing hype. GAC, or AC, is a very specially processed form of charcoal. Just about any high-carbon material could be made into charcoal, with varying degrees of practicality. In the aquarium trade, we are most likely to see GAC made from coal, specifically bituminous coal, or from various form of wood. The bituminous coal source material is the best for water use in general. The wood based material, especially coconut shell carbon, may be significantly better for gas (air) filtration applications, but is less desirable for aquatic uses. The exact process used and the resulting physical structure and final available surface area of the carbon pellet is critical as well. If carbon is thought of a chemical “sponge”, it is a special type of sponge – materials are picked up, adsorbed, only through direct contact between the water surface and the surface of the carbon itself.

Scanning electron microscope image of activated carbon pores., from NIOSH



There is no wicking of the material through the carbon mass as there would be with water through a sponge; everything is pure surface contact and adsorption phenomenon. The surface area per unit of weight for good pelleted GAC is astronomic. The final extractions and rinses done on the product are equally important. We are not likely to get any of this detail from a jar of GAC in the LFS, so we are at the mercy of suggestions from our peers and the local fish store or web outlet staff. Some carbons sold for tank use are not much more than crushed charcoal briquettes or crushed wood charcoal, and are of marginal if any use in our tanks. They could in fact import undesirable material (especially phosphates). High-porosity pelleted or granular materials made from bituminous coal are what we hope to get. We further hope that they will not phosphate-load our water. That phosphate issue you can test by a tablespoon to a quarter cup of GAC in a glass or jar of tap water, which is left out overnight on the counter alongside a control glass of water. After several hours to overnight, if it has not done radical change to the pH or the phosphate levels of the water, it is okay on that point. Most brands will require some rinsing to remove the fines (tiny dust-like particles), but none are likely to rinse clear without special processing. Extended rinsing is wasting both your time and the material by allowing uptake of materials from your tap water; extended stirring will only abrade more fines from the pellets/granules, so moderation is required. The material is best used in mesh bags, arranged so that water must pass through the bag, not around it or over it.


Okay, we have relatively clean, low-phosphate GAC in our filter with water flowing through it. What do we get? Several dissolved materials will be extracted from the aquarium water by adsorption onto the carbon. First (because this gives visible effect, we can see the difference in the water) is removal of dyes and colorants. Most tanks will build up more or less yellow in the water with the passage of time without large-scale water changes. You could consider this dilute urine if you don’t object to a grotesque image. Carbon will remove it quite well. Ditto for tannins from driftwood, bogwood, or peat, or for dyes after a medication use. The difference in the appearance of the water can be quite striking, even if the yellow/tan tint was not obvious or unattractive. The difference in water clarity is such that live plants will respond, may even burn a bit in shallower plantings if the water color was noticeable prior to treatment with GAC.

Many non-colored or too dilute to be seen dissolved organic compounds will also be taken up by carbon, and this includes a number of aromatic (smelly) compounds we are glad to have removed. Overall this is highly beneficial to the water quality, even if we cannot see the difference, or smell it. Surface pickup is still key here, and with time in the filter might, only might, work against us to a very limited and tightly defined way. There are many warnings against carbon on the boards, saying that after saturation, it can release material back into the water column. This is not true. Repeat: GAC does not and cannot release material it has captured under tank or household conditions. In the lab, carbon can be extracted to some extent by use of specialized solvents, frequently very strong bases (alkalis). This is very far outside the range of any biological or household situation. But there could be considered to be a tiny exception – certain very long-chain or cyclic organic molecules may be captured only by a specific part of the molecule, leaving some part of the attached chains hanging out in the water. In this case, it is theoretically possible that the part of that organic molecule chain not adsorbed could be attacked as food by microbes. But this would be a vanishingly small quantity in relation to the material adsorbed, and undetectable other than by highly sophisticated laboratory experiment, if the effect even exists in working, real life situations.

The much more common situation would be that the carbon is fully saturated and acting as a mechanical filter, picking up fine particulates. (Remember the massive surface area? Lots of crevasses and nooks and crannies are available here). This can be exactly comparable to the case of conventional undergravel filters, which should be biofilters, and which in use act all too well as mechanical filters. GAC, which should be a chemical filter, will also act as a mechanical filter. As the captured particulates are largely organic, this provides an available food supply for heterotrophic and other bacteria. And the carbon itself serves quite nicely as attachment for both heterotrophs and lithotrophs (the nitrification bacteria). The problem presented by using this spent or saturated material for this use is that it is anything but self-cleaning without potential or actual bacterial colony damage due to the poor abrasion resistance of GAC. So in use in tanks, GAC should be at least as well protected from smothering particulates as biomedia should be. Keeping it clean and clear will allow it to perform its design task (“chemical” filtration) as completely as possible. After it is “spent” or fully loaded, it is discarded. Despite comments to the contrary sometimes seen of the web, activated carbon cannot be burned out and reactivated under household conditions. That is an industrial process involving very high and controlled temperatures under controlled oxygen level, it is not DIY.

Some desired materials are likely to be lost to carbon, most especially the chelated metals such as iron from plant supplements, or copper if you are a user of such – the chelating agents are readily adsorbed by GAC. The adsorption of this material is specific to the material in question. For the non-chelated minerals, the adsorption is real and measurable (in a laboratory at least), but is not really significant in terms of functionality in a freshwater tank. The total capacity of normal-use quantities of carbon is really quite small for most non-chelated material, and the change in concentration would not be readily detectable in our tanks. Should it be marginally detectable, compensation by routine micronutrient supplement addition would be easy to trivial. This is one myth presented on the boards routinely –“never use carbon in planted tanks, it will rob the nutrients your plants require” – in a word, horsefeathers. It will adsorb some minerals and more chelated metals, but if you need and desire to use carbon, do so and add a small percentage more micronutrients and chelated iron (if you can measure that small a change) and you will be fine so long as you don’t overdose. The increase in light from colorant removal will be a bigger boost to plants than loss of a few micrograms or fractions thereof from the trace elements. Medications which are dyes themselves should not be used in tanks with carbon in the filters – remember using carbon to adsorb colorants? Remove carbon from the system before such dye usage.

Another point of heavy debate is how long carbon will last in your tank. The always infuriating but correct response is that it all depends. Carbon is not general or non-specific. Each and every adsorbed material will have its own uptake by both rate and total quantity taken up by a given mass/surface area of carbon. Naturally this means that the percentage removed from the tank water will depend on the material, the total quantity in the tank, and the possible uptake of that particular material by that particular carbon. If you have a heavy tannin load, a small quantity of carbon may be saturated in hours, maybe minutes depending on flow (turnover) rates. If the thing to be removed is a material that is present in the water in small to minute quantities, it could take days to weeks to saturate the carbon. If it is specifically the colorants produced by normal metabolic processes by a moderate fish load, days to weeks is quite likely the life span in routine use. In periodic use, after a long period of small water changes, the saturation of the carbon will come much faster. By the way, carbon works best at low water flow. Our filters are generally not that low in flow, but we compromise again. GAC still works at higher flow, just not as efficiently. But we want effect and end result, not efficiency here – yet another compromise.

If carbon is so great, I must use lots of it, right? In fact, no, I seldom to rarely use it, and only for specific aim (usually to get tannins out). I do not use it as a tank routine maintenance item. Why? Simple water changes will perform the majority (but not all) of the functions covered by carbon, and I do perform water changes regularly. Also, I have multiple tanks. Water is cheaper than carbon, and dilution by partial water change on a regular basis is also removing things that carbon effectively does not touch (one example: nitrate). But that is personal choice. There is no right or wrong here. If I ran only one or a small number of tanks, I would be more likely to use it routinely.

Specific absorbents:

Phosphate: There are two chemistries by which phosphates may be removed from tank water. The preferred process uses aluminum oxide. Beads or finer granules of this material may be purchased pre-bagged or in bulk containers for use in owner-supplied fine mesh bags. Again as for carbon, this material should be arranged for water flow through the material, to maximize surface contact. Also again as for carbon, the lifespan cannot be easily predicted. Instead, the water is tested until the concentration of phosphate in tank water is minimized. The expected minimum achievable concentration with a given material should be available from the manufacturer, and sometimes is even printed on the bottle.

This same material will remove silicates from tank water as well. So if diatoms (brown algae) are an issue for you, you might find this material useful even without having high phosphates in your water.

If your tap water is high in phosphates (many utilities add phosphate buffers for network and home plumbing protection) and you wish to extract such for use in planted tanks, the material should be extracted before the water is added to the tank(s). Then the proper, lower, dosage of phosphate required by the plants may be added to the tank.

Others: Materials are also available for copper and several other specifics, but I have no experience with these.

Exchange Resins:

Nitrogen: The commonest material used for this purpose is a mineral, zeolite. This material exchanges salt ions for nitrogenous ions in solution. There are also synthetic resins with similar action. In emergency situations, such as nitrification failure by the biofilter, the material may well help save your fish. But it is generally an exchange, not a simple absorption or adsorption. The downside is that the zeolite or resin is adding NaCl to your water, increasing total dissolved solids (TDS) silently in a way you cannot detect without having a TDS meter. The give-away is that these materials advertise that they can be recharged in brine solutions. In a situation such as the nitrification failure mentioned before, this can have benefits, as salt helps protect against nitrite toxicity, but this is not easily measured. If used in this manner, extra water changes should be done when the tank’s biological filtration is back in proper function and stable again. These agents can also easily starve a biofilter by absorbing the materials that the nitrification bacteria require. There are conflicts there between starving the biofilter and getting rid of toxic materials. For my own use again, water changes are simpler and safer.

Hardness: (calcium and magnesium) The term “hardness” refers to water in which it is “hard” to make a lather with soap (or detergent), and which tends to leave a hard film after a washing is done. The primary ions responsible for this function are calcium (Ca++) and magnesium (Mg++), so these ions are responsible for “hard” water. In most natural waters, calcium is present in higher to much higher concentrations than magnesium, but the relative proportions are not critical for laundry. In fish tanks, “salt” creep from the carbonates and other compounds of calcium and magnesium is common, and unattractive, along with sometimes damaging such accessories as biowheels and other filter equipment with exposed water-level components. Many of my sumps, which have water levels well below the plastic trim on the tank, develop grubby lines of mineral buildup from this.


  • “Water Softener Pillows” are frequently sold and used to “soften” the water, supposedly to make it more suitable for blackwater fish such as many Tetras, Amazonian Cichlids, etc. What is not explained is that these “pillows” (bags of ion exchange resins, again brine rechargeable), are trading two sodium ions (Na+) for each calcium or magnesium ion (both ++), as the net charge in such an exchange must match. Thus the TDS of the water is ordinarily increased by the exchange. True, blackwater is poor in Ca++ and Mg++, but it is equally poor in Na+ as well. The desired water would be low TDS, which should also be relatively “soft”. This sort of resin is best referred to as a “salt exchange” resin. They are not desirable for use for aquarium water. Household water softeners are commonly of this type, although “acid-base exchange” resins are available for home use at significantly higher cost. If the home water softener is rechargeable by salt, it is obviously a salt exchange type. Sodium ions do not interfere with lather in doing laundry.
  • “Acid-Base Exchange Resins” are those which function by exchanging hydrogen ions (H+) for Ca++, Mg++, Na+, etc. (the cations – positively charged ions), and hydroxyl ions (OH-) for carbonates, bicarbonates, sulfates, etc. (the anions – negatively charged ions). The net exchange is the mineral or salt for H+ and OH- (HOH or H20), plain and pure water. These resins are obviously fully suitable for aquarium use. Unfortunately, they are not safely recharged at home. They require the use of strong acids and bases for the recharge process. The water from this process is referred to as “deionized water”, or DI water. The Tap Water Purifier is one such unit sold for home use. The quantity of water produced by one of these units is dependent on the amount of dissolved minerals (not limited to those measured by GH and KH) in the source water.
  • “Reverse Osmosis Systems” are commonly complex systems made up of a fine particle filter, followed by a carbon adsorption filter, and finally microporous membranes, which are in reality mechanical filters at the level of small molecules and ions, plus a pump. The mechanical filters discussed in the prior article were flow-through filters. Reverse osmosis (RO) filters are generally tangential flow, meaning that the liquid flows along parallel to the membrane, with molecules and ions small enough to pass through the filter doing so because the pressure is significantly higher of the input side. This literally forces sufficiently small material through the filter, but not dead-ending or allowing the buildup of larger ions or molecules on the filter surface. (Remember the macro-filter illustration? This is not supposed to occur in tangential flow applications.) The purity of the output water is in part dependent on the pressure used, the membrane itself (the nature of the material and its microporosity), and the TDS (total dissolved solids) of the input water. Also in significant part it is dependent on the ratio between rejection water (the part that never goes through the membrane and which is richer in minerals than when it started) and the throughput water or product (which is poorer to much poorer in minerals than the source water). The highest purity of output water requires the highest rejection rate, that is, much more water sent to “waste”. RO systems are capable of producing quite high-purity water, but only when operated at high pressures and flows, with very high rejection to retention ratios (a.k.a waste to product ratios). This can make them rather expensive to operate at home if very high purity water is required (rarely for fish tanks). A more practical practice would be to follow the RO system with an acid-base exchange resin system, giving more RODI water of higher purity than is readily achieved by either system alone. Very few home FW techniques would need such pure water. Captive cultivation of some SW inverts may be easiest with such source water.
  • Sphagnum Peat Extraction: One ion exchange process frequently overlooked in discussions such as this is peat extraction. This material is the dried, ground, acid-preserved remains of a moss, sphagnum moss. It is also available as the intact dried form of the formerly live plant, but this is much more expensive. This moss occurs in extensive bogs on this continent (North America) and many other places (Germany has significant bogs). The places where the moss grows could be characterized as wet deserts, having plenty of very soft rainwater in and on the ground, but very, very few minerals. This material is a natural acid-base exchange resin, whether live or dead. The ground material can be housed in mesh bags (I use sections of my wife’s cast-off pantyhose) to soften water for tank use. The material is available quite cheaply at garden centers in bales of varying sizes for soil conditioning outdoors. The primary warning is that you want pure ground Canadian sphagnum peat, not Michigan peat, which is a sedge peat (not a moss peat) lacking the desired exchange property. You also want pure peat without additives (no fertilizers, etc.). Different brands (in reality, different locations of the material) will have different final pH at maximum exchange, and this is commonly specified on the bag.

My use has been the stocking section bag mentioned above, charged with several cups to a few quarts of peat, hanging in a large Rubbermaid container of water with a heater and a prefiltered powerhead for circulation. The prefilter on the powerhead is to capture the fines from the peat that inevitably escape the stocking. A given quantity in your container may be allowed to go to full extraction and final pH, or depending on the hardness of your source water, may exhaust the peat before reaching final pH. The good thing is that measured quantities are fairly reproducible. That is, after use of quantity x of peat, when y time has elapsed in z gallons of water, the pH and residual hardness can be pretty well predicted. You have to monitor the process the first time, after that you need only check the finished product. The water will be stained dark from tannins. This peat-extracted water may be mixed with your tap water to give the desired levels for your tank, and used for makeup water after partials. Changes in tank water should be done slowly to allow the fish plenty of time to adjust to this DIY blackwater. If the tannin staining is not desired, it can be carbon-extracted, but it takes a significant amount of high-grade GAC to extract all or most of the tannins. That carbon extraction would not be an inexpensive step in the process. Tannin-stained water eats light, so if your tank is planted, this should be considered as well.

Multi-function materials: There are a number of commercial products that employ more than one process, such as carbon plus resin(s) plus sometimes surface-charged media. Chemipure is one such material, but there are others. As these are all proprietary formulations with closely kept component materials, it is difficult without extensive and expensive laboratory testing to say just what is being extracted in what manner. On these materials, you are entirely on your own. If they accomplish a task you need to have done, and seem to do no harm to the tank or your budget, use is a personal choice item. No defense or further justification is required.

Next: Part 3, Biofiltration


Robert T. Ricketts, a.k.a. RTR


Please enter your comment!
Please enter your name here