There is one “interesting” approach to providing carbon dioxide to the aquarium which bears mentioning. This is to use food as the source of CO₂. Note I do NOT recommend this method. It is challenging, to say the least.
Let’s do some simple math here. In a planted aquarium it is desirable to add VERY ROUGHLY 0.5 ppm CO₂ to the aquarium every hour. One PPM per 100 gallons is 0.36 grams. One ppm per 20 gallons is 0.09 grams. So a twenty-gallon tank needs VERY ROUGHLY 0.09×0.5×24= 1 gram CO₂ per day.
Since carbon dioxide is 27% carbon this is a need for 0.27 grams carbon. Fish food is roughly 65% carbon by dry weight. So one needs 0.43 grams of fish food per day to supply 1 gram CO₂ per day to a 20-gallon aquarium IF THE AQUARIUM HAS LITTLE OR NO AERATION.
A half gram of dry fish food is not a lot of food! The higher the protein content of the food the better this will work. 45% or greater protein is excellent. If this fish food addition is supplemented with a pinch of potassium nitrate (0.2 grams), the added nitrogen will ensure the carbon goes into carbon dioxide and not into polysaccharide rich biofloc. The added potassium is great for vascular plants.
It is beneficial here to add phosphate and iron into the substrate as root tabs. The food has some phosphate but adding potassium nitrate will take the nitrogen/phosphate ratio out of whack, with 3:1 being the ideal.
Now the caveat here is that there MUST be a lot of biofiltration with little disturbance of the surface of the water in an aquarium. This typically requires some baffling of flow. For instance the outflow of any canister would need to go into a dead-end pipe about three times the diameter of the exit flow out of the canister. This reverses the flow 180 degrees and spreads it out without creating a surface disturbance.
This outflow needs to be done as low in the aquarium as possible. Any “turbulent flow” at the surface of any planted aquarium will remove the CO₂ from the aquarium, with some degree of exception if one has a lot of fish (see below).
A sponge filter or an in-tank cartridge type filter will not provide enough biofiltration to safely do this type of CO₂ addition. One needs a canister full of good media like plastic pot scrubber, foam or K1 media. If one does not have a very large amount of biofiltration, the water will become filled with infusoria and very cloudy. This will make for a very poor looking aquarium. But then some like this “natural” look.
Now this approach is GENERALLY NOT compatible with sizable amounts of sizable fish. The fish will require aeration. And aeration will remove the CO₂. BUT and this is a BIG BUT, some aeration will actually work. Father Fish has some gorgeous, planted tanks with air stone aeration and heavy fish loads. But Father Fish is careful with the aeration. He often only puts an airstone a few inches deep into the aquarium. And Father Fish feeds VERY lightly.
The key here is that carbon dioxide is held in the water in a loose quadripolar association with four water molecules. Oxygen is simply a gas. So oxygen will permeate water significantly faster than CO₂ will gas off. So some limited aeration will work. But it is not easy to get this modicum. One is basically creating a controlled cesspool.
Meer informatie over CO₂ systemen
General Information on CO₂ in the aquarium can be found at this link:
15.6. CO₂ in the Planted Aquarium
Measurement of the CO₂ level in an aquarium is covered in this link:
More information on setting up a high tech CO₂ injection system can be found at this link:
There is a low cost, low tech system which adds considerable CO₂ to an aquarium. The details of this system can be found at this link:
15.6.1 Low Tech CO₂ Aquarium System
The intricate relationship between Plants, KH, pH and CO₂ is covered in this link:
15.6.2. KH, pH, CO₂ relaties in een beplant aquarium
And some additional data on the very complex Bermuda Triangle of pH, KH and CO₂ can be found in this link:
English 
Dutch