You should be careful with choice of buffer. Phosphate will precipate the Al3+ and several other buffers will create complexes with Al(III).

Since you are adding known amounts of AlCl3 and you know the amount of acidity it creates, I recommend that you add the matching amount of alkalinity (base) of the AlCl3 at each dosing level: AlCl3 +3 H2O -> Al(OH)3 + 3H+

Considering the speciation of Al(II) at pH 7 (see link) and the solubility of Al(OH)3 I think the best guess is that each mole of AlCl3 added produces 3 mole of acid. Thus add 3x the molar concentration of NaHCO3 to the molar concentration of AlCl3 at each testing level.

https://en.wikipedia.org/wiki/File:Al_hydrolysis_speciation_diagram.png

If the pH goes from 7.2 to 5.4, that means you are forming Al(OH)x. SO, the final solution does not have the bio-available concentration you think. That could be an issue. 

If that isn't a concern, you can add a buffer to your solution. You'll have to go through some books and find a buffer that works around a pH of 7.2 and will not ruin your experiment.  

Dear pedram, the answer of Andrew is right,  and your Al probably complex with OH. You can add NaOH to control the pH, but it is sure that part of Al will form hydroxyde and will not have the same influence in terms of toxicity. But it is also representative of the environment.

 You should be careful with choice of buffer. Phosphate will precipate the Al3+ and several other buffers will create complexes with Al(III).

Since you are adding known amounts of AlCl3 and you know the amount of acidity it creates, I recommend that you add the matching amount of alkalinity (base) of the AlCl3 at each dosing level: AlCl3 +3 H2O -> Al(OH)3 + 3H+

Considering the speciation of Al(II) at pH 7 (see link) and the solubility of Al(OH)3 I think the best guess is that each mole of AlCl3 added produces 3 mole of acid. Thus add 3x the molar concentration of NaHCO3 to the molar concentration of AlCl3 at each testing level.

https://en.wikipedia.org/wiki/File:Al_hydrolysis_speciation_diagram.png

I totally agree with HRA in addition to the suggestions from AE and OR. As can be seen from the diagram (Al hydrolysis speciation diagram from Wiki Ref.as above) almost more than 90% Aluminium hydroxide will be available in the pH range of 6.2-7.2 in dissociated form. Considering the stoichiometric relationship of hydrolysed AlCI3;  if three molar base (either sodium hydroxide or sodium bi-carbonate) per mole of AlCI3 is  added, then it would be sufficient to provide the required buffering action and maintain the expected pH.

Pedram,

I've used a hydrogen ion buffer called HEPES for toxicity assays.  It can maintain a neutral pH and doesn't appreciably complex with metals.   If you want to maintain circum-neutral pH and not affect metal speciation this is a good choice.  0.5 mM should be sufficient.

A more mechanically intensive option, but one that actually may be most suitable, is to use pH controlling hardware.  These are used often in studies of metal toxicity where pH, hardness and alkalinity need to be controlled.  Of course, this also turns the entire study into a flow through exposure simulation.  As opposed to direct flow through and pH control, Mount-Brungs proportional dilutors are another option.  Would agree with Andrew and suggestion of James as a low intensity option.  HEPES buffer is not unknown to metal ecotoxicologists.   

HEPES is commonly used in metal toxicity tests, I have also used before. Just recommend you, if possible, to use an extra control without the contaminant (Al) but with the same concentration of HEPES, in order to prove that the amount of buffer you used is not making any interference with your test.  

Environment Canada has developed a procedure to stabilize pH in acute toxicity tests. 

Procedure for pH Stabilization During the Testing of Acute Lethality of Wastewater Effluent to Rainbow Trout - Report EPS 1/RM/50

https://www.ec.gc.ca/faunescience-wildlifescience/default.asp?lang=En&n=74E8CDAC-1

As mentioned above the solution pH determines the preponderance of the Al+ species exposed.  Also the ionic composition and strength of the dilution water will affect Al+ species.  The dissolved organic carbon concentration in the dilution water will also affect toxicity - think of biotic ligand effects. 

The addition of buffers will change the the test conditions and the type of Al exposure. 

Aluminum is a very complex metal as illustrated by Anderson above.  Be sure to read the latest water quality quality reviews on aluminum toxicity.  Controlling the test conditions to address the hypothesis will be challenging. 

At pH 7,2 you don`t have Al 3+ at all.  After pH=5,0 different  ions are fiormed

Al(HO)2+, {Al(HO)2}+  before Al(OH)3 is precipitated.

AlCl3 solution is always acid ( HCl) to avoid the formation of that ions an the

precipitation of the hydrated oxide.

Hi,

You have received many answers now. The toxicity of Al compounds is complicated.

Its generally accepted that Al3+ is the most toxic fraction followed by Al(OH)2+ and so on. The equilibrium between these and all other Al-compounds depends on the pH-value as Luis Bertollo pointed out.

So either decide to use a lower pH-value for all your tests with only acid as a reference and additions of AlCl3 for testing. I would suggest a pH-value of 5.5. That is likely to give a low pH-effect on your test organism and be relevant (and even high for some waters in Sweden) for real conditions.

As pointed out earlier; don't use any buffer!

Good luck

Anders Wilander

Other solution is to add an acetate buffer pH 7,0  in all tests .