298
D. Ferri . C. Manfredi . E. Vasca . C. Fontanella . V. Caruso
At the time (1984) when the TDB project started, none of the existing databases
fulfilled all these criteria. The example of the TDB is a clear demonstration of two facts:
a Although the attempts, carried out by diligent authors to give order to the immense
mess of the literature on equilibrium studies are greatly appreciated, we must admit
that, at the first serious challenge to demonstrate their adequacy to model real systems, the existing compilations have badly failed. Much work is still to be done. The
OECD/NEA project has shown that the right way to do the job is through international cooperation. The amount of labour for this enterprise demands for joint ventures, since we do not know the legendary Hercules' present address (read the sarcastic story of the mythical hero cleaning the stables (databases?) of king Augeas
(Grenthe and Puigdomenech 1997, pp. 131-152)). While a deeply critical revision of
the existing literature is indispensable, missing data must also be provided on mixed
complex formation, on complexation of metals by large ligands as humic and fulvic
acids, and on thermodynamic quantities, in addition to !J.G. Spectroscopic investigations (to gain structural information on the species found by the methods of equilibrium analysis) must also corroborate, when possible, the results of the equilibrium analysis. Finally, the results obtained in the presence of an ionic medium must
be recalculated on other scales of activity or extrapolated to zero ionic strength,
possibly by means of the Pitzer (Pitzer 1979) or the SIT (Br0ensted 1922; Br0ensted
1923; Scatchard 1936; Guggenheim 1966; Biedermann 1975; Ciavatta 1980) equations,
which are the most reliable. The first is more precise (see Chapts. 1 and 8) but requests more parameters, which are rarely available. The second, which is simpler to
apply, has proved to be the most advantageous choice. At this point it is evident that
in our opinion the existing equilibrium data compilations must be considered generally inadequate as sources on which to base modelling. Somebody can suggest to
begin amending the databases by eliminating from the literature the infesting data
on applying a set of rules, by assigning, for instance, a weight to the experimental
method of investigation. Whoever makes use of potentiometry and has experimented
with its incomparable power, provided that an adequate electrode exists, would suggest the highest ranking for this method, but probably somebody would disagree.
And yet the application of this rule would be only the first step. We should also ascribe different weights to the choice of the ionic medium (if there is one) and its
concentration, to the type of electrochemical cell adopted (with or without liquid
junction), to the type of glass electrode chosen (single or combined), to the number
of free concentrations measured (if more than one is accessible) and so on. Then
comes the delicate moment of the interpretation of the data in terms of stoichiometry of the complexes and their formation constants. Recently, May and Murray (2001)
discussed the possibility of building up a database that should achieve thermodynamic consistency automatically. But let us go back to the initial statements about
the accuracy of the experimental data.
b The advent of high-speed computers and consequently of powerful programmes for
the interpretation of the data has somehow favoured a dangerous tendency to use
less effort in the acquisition of experimental data, for instance measuring a single
free concentration (one electrode, usually the H+ glass membrane), when a deeper
insight in the system could be attained by also measuring other electrodes (amalgam of the metal, second type electrodes for halogenides, oxalate, etc.). The choice
D. Ferri . C. Manfredi . E. Vasca . C. Fontanella . V. Caruso
At the time (1984) when the TDB project started, none of the existing databases
fulfilled all these criteria. The example of the TDB is a clear demonstration of two facts:
a Although the attempts, carried out by diligent authors to give order to the immense
mess of the literature on equilibrium studies are greatly appreciated, we must admit
that, at the first serious challenge to demonstrate their adequacy to model real systems, the existing compilations have badly failed. Much work is still to be done. The
OECD/NEA project has shown that the right way to do the job is through international cooperation. The amount of labour for this enterprise demands for joint ventures, since we do not know the legendary Hercules' present address (read the sarcastic story of the mythical hero cleaning the stables (databases?) of king Augeas
(Grenthe and Puigdomenech 1997, pp. 131-152)). While a deeply critical revision of
the existing literature is indispensable, missing data must also be provided on mixed
complex formation, on complexation of metals by large ligands as humic and fulvic
acids, and on thermodynamic quantities, in addition to !J.G. Spectroscopic investigations (to gain structural information on the species found by the methods of equilibrium analysis) must also corroborate, when possible, the results of the equilibrium analysis. Finally, the results obtained in the presence of an ionic medium must
be recalculated on other scales of activity or extrapolated to zero ionic strength,
possibly by means of the Pitzer (Pitzer 1979) or the SIT (Br0ensted 1922; Br0ensted
1923; Scatchard 1936; Guggenheim 1966; Biedermann 1975; Ciavatta 1980) equations,
which are the most reliable. The first is more precise (see Chapts. 1 and 8) but requests more parameters, which are rarely available. The second, which is simpler to
apply, has proved to be the most advantageous choice. At this point it is evident that
in our opinion the existing equilibrium data compilations must be considered generally inadequate as sources on which to base modelling. Somebody can suggest to
begin amending the databases by eliminating from the literature the infesting data
on applying a set of rules, by assigning, for instance, a weight to the experimental
method of investigation. Whoever makes use of potentiometry and has experimented
with its incomparable power, provided that an adequate electrode exists, would suggest the highest ranking for this method, but probably somebody would disagree.
And yet the application of this rule would be only the first step. We should also ascribe different weights to the choice of the ionic medium (if there is one) and its
concentration, to the type of electrochemical cell adopted (with or without liquid
junction), to the type of glass electrode chosen (single or combined), to the number
of free concentrations measured (if more than one is accessible) and so on. Then
comes the delicate moment of the interpretation of the data in terms of stoichiometry of the complexes and their formation constants. Recently, May and Murray (2001)
discussed the possibility of building up a database that should achieve thermodynamic consistency automatically. But let us go back to the initial statements about
the accuracy of the experimental data.
b The advent of high-speed computers and consequently of powerful programmes for
the interpretation of the data has somehow favoured a dangerous tendency to use
less effort in the acquisition of experimental data, for instance measuring a single
free concentration (one electrode, usually the H+ glass membrane), when a deeper
insight in the system could be attained by also measuring other electrodes (amalgam of the metal, second type electrodes for halogenides, oxalate, etc.). The choice