One way to maintain the constant concentration is to bubble into the solution a gaseous mixture of air and carbon dioxide prepared under pressure. You can see the device I used long ago in the publication (in French!):

BUSSIERES P., GIROU A., CALLOT G., ANDRE L. Dissolution of particles of limestone amendment. Soil Science, 1982, 4, 263-273. It can be loaded from http://www.afes.fr/afes/sds/SDS_20_4_p263_BUSSIERES.PDF.

A pressure gauge and flow meter are used to adjust the flow rate of the mixture. If the flow rate is sufficient, the pH of the solution will depend on the CO2 content of the gas mixture and remain constant.

The liquid air industry can prepare mixtures with a desired CO2 content.

Obviously, the container containing the solution must be opened to allow the gas to exit.

Dear Bruno, not being a chemist I have no imediate solution to your question. But I doubt that the alkalinity change comes from CO2 exchange with the atmosphere: CO2 is a non-charged species and its concentration does not appear in the definition of

TAlk = [HCO3-] + 2 [CO32-] + ....

So, if CO2 enters the water, it does not change TAlk; and when it reacts with water, it forms both a HCO3- and a H+, which cancel out in the alkalinity.

My guess is that your problem has to do with starting from de-ionised water: In seawater (which is where I know most about) the high alkalinity is a consequence of weathering, which bring in ions such as Ca2+, Cl- or Na+.

Maybe you should have a look at the recipes for artificial seawater?

Cheers, christoph

Dear Bruno,

Like Christoph I cannot see any good reason why the alkalinity of your solution should change, unless precipitation/dissolution reactions involving the carbonate salts are taking place somehow.

Chemical equilibrium calculation programmes such as MINEQ+ or MINTEQ will allow you to easily calculate [HCO3-] from a knowledge of Total Dissolved Inorganic Carbon (Total DIC = your analytical concentration of carbonate salt) and pH. It turns out that over 90% of the DIC is contributed by the bicarbonate ion at pH 8 (over 95% at low ionic strength) which provides a quick estimate of [HCO3-].

The question remains, how to maintain a pH value of 8.0. There I would suggest adding an appropriate buffer solution or controlling the pH by bubbling the appropriate air/CO2 mixture, as suggested by Philippe.

Dear Philippe, Christoph and Francois,

I would like to thank each of you for your valuable answer and am certainly directing myself towards gas bubbling in the solution to reach the appropriate HCO3- concentration. I have been as well surprised by the pH change over time, but it has been measured with different pH electrodes nonetheless. It remains quite a mystery!

To be complete, I prepared a 0.01 mol/L solution of NaHCO3 by simply adding 840mg NaHCO3 to a liter of water. I do not know if this change of pH over time might be the result of a low ionic content in the water, or the absence of other salts, but the pH was indeed changing over time.

I thank you again for your insight!

Bruno

Kindly check..

http://dnr.wi.gov/regulations/labcert/documents/training/basics-genchem.pdf

More to explore..

http://www.who.int/water_sanitation_health/resourcesquality/wqmchap7.pdf?ua=1

See here too..

 http://www.fao.org/docrep/t0551e/t0551e03.htm

Also, you can see a relationship between CO2 content of the gaseous mixture and pH of a solution NaHCO3 and KCL on the figure 14 page 43 in my thesis which can be uploaded from my profile:

Maitrise_du_pH_des_sols_acides_par_les_amendements_calcaires (Oct. 1981).

 It can be usefull to make the gaseous mixture.

Philippe

Hi Bruno,

1. Which question are you ultimately addressing?  What do you want to use the water for?  Chemical or biological experiments?

2. How did the pH change? Was it fluctuating or directional and if the latter, in which way?

3. My experience is mostly re. seawater TA (total alk) and from about 20 years ago: (in line with Christoph's answer)  the classic paper for TA is still the one by Andrew Dickson 1981 for the definition of total alkalinity in seawater (http://scrippsscholars.ucsd.edu/adickson/content/exact-definition-total-alkalinity-and-procedure-estimation-alkalinity-and-total-inorganic-ca-    DOI: 10.1016/0198-0149(81)90121-7 according to which the entry and exit of CO2 does not alter TA but it affects pH.  I remember however, that geologists had their own definition of alkalintiy, for they asked different questions.

4. I would have thought that the 'easiest' way, in principle, is to keep the solution in a closed system at all times, so as not to allow any exchange of CO2 across the air/liquid interface.  If of use: In alkalinity titrations this system was used by e.g. Bradshaw and Brewer (1988) DOI: 10.1016/0304-4203(88)90023-0 

5.  Following on from what Francois has written, the more additional acid-base systems you have in there which do not involve gaseous components, the better (hence Francois mentioned buffers.  However, if you want to keep it as natural as possible, e.g. for biologigal experiments, this is not always favoured).

6. I just remember: when I titrated blanks (just H2O and NaCl, and thus not pH-buffered at all) my precision was the worst when compared to NaCO3 standards and natural seawater - presumably because of gas exchange - for I had a SEMI-closed system only.

7.  I was advised to use Na2CO3 (and dry it and keeping it in a desicator before weighing it) instead of NaHCO3 to increase the accuracy when weighing it.

Kerstin

Corrigendum: I did not express myself correctly in point 5 above:  The buffer capacity is not determined by the number of acid-base systems but by their concentration(s) and the suitability of their pK- value(s).  And if you have acid-base systems without a gaseous phase, you don't have to worry about exchange across the air/liquid interface. (What a mine field!)  Kerstin

Dear Kerstin,

Thank you very much for your thorough answer. It is quite a mine field indeed! Regarding the issue with biological components, I actually prefer water as free from them as possible. The waters are meant for modeling in flotation, and biological components are susceptible to disturb the repeatability of the test just due to bacteria colony growth or so.

Bruno

Hi Bruno,

if no interest in biology, have you considered adding mercuric chloride - or maybe BAC- to your water?  

On our JGOFS research cruises the TCO2 scientist (Carol Robinson) and my TA-me as the subsequent sampler from the Niskin bottles used mercuric chloride to stop any biol activity.  This was more important for TCO2, less relevant for TA under those circumstances.  As I am just finding out, Carol has co-authored a paper on HgCl2 and possible alternatives, i.e. BAC, and the paper may also supply you with useful updated refs on preservation - so a quick glance seems to suggest. See link.

Bon week-end,

Kerstin

 Not sure the link got through:

http://www.ocean-sci.net/11/947/2015/os-11-947-2015.pdf

Cheers

Kerstin