[OLD] REE data

[webmaster]

From: Greg Anderson

Subject: REE data

There are quite a few problems with the REE data in thermo.com.v8.r6+.dat.

1. There are many redundant species. Two methods are used to write REE hydrolysis reactions, differing only in the number of H2O. The number of metal ions and H+ are the same, so log K should be the same. However, the two methods come from different references, so there are some (generally small) differences. Using Dy as an example, the two methods are

Dy+++ + 4H2O = Dy(OH)4- + 4H+; log K(25C) = 33.4803

Dy+++ + 2H2O = DyO2- + 4H+; log K(25C) = 33.4804

Reactions in the DyO2- form are from Haas, Shock and Sassani (1995) and have data for higher temperatures; data in the Dy(OH)4- form are from Spahiu and Bruno (1995), and are for 25C only. I have no idea as to accuracy, but as there are small differences at 25C, using the Haas et al. data at least gives access to someone's idea of high T behavior. I suggest commenting out the (OH)4- forms. The REE with redundant forms are:

log K(25C)

============================================================

Nd Nd(OH)4- and NdO2- 37.0803 and 37.0721

Sm Sm(OH)4- and SmO2- 36.8803 and 35.0197

Eu Eu(OH)2+ and EuO+ 14.8609 and 16.3370

Eu(OH)3(aq) and EuO2H(aq) 24.1253 and 25.4173

Eu(OH)4- and EuO2- 36.5958 and 34.5066

Gd Gd(OH)4- and GeO2- 33.8803 and 34.4333

Dy Dy(OH)4- and DyO2- 33.4803 and 33.4804

Er Er(OH)4- and ErO2- 32.5803 and 32.6008

Yb Yb(OH)4- and YbO2- 32.6803 and 32.6741

=============================================================

 

2. Many of the REE have a sign error on the oxidation reaction,resulting in the divalent form being stable at all oxidizing conditions. For example, the reaction

Ce++ + H+ + 0.25 O2(aq) = Ce+++ + 0.5 H2O; log K = -83.6754

The log K is actually +83.6754. The log K values are:

Element Current logK Correct logK status

=======================================================

La -72.4754 72.4754 error

Ce -83.6754 83.6754 error

Pr -79.9754 79.9754 error

Nd -64.3754 64.3754 error

Pm -65.2754 65.2754 error

Sm 47.9624 47.9624 OK

Eu 27.5115 27.5115 OK

Gd -84.6754 84.6754 error

Tb -78.7754 78.7754 error

Dy -61.0754 61.0754 error

Ho -67.3754 67.3754 error

Er -70.1754 70.1754 error

Tm -58.3754 58.3754 error

Yb 39.4595 39.4595 OK

========================================================

 

The ones with the wrong sign are from Spahiu and Bruno (1995), and this is an error in their compilation, not in converting the data to GWB. The ones with the right sign are from Shock and Helgeson (1988), and can be checked using supcrt92. In fact all of them are in supcrt. If you are using REE, it might be a good idea to change all the data to supcrt data, because 1. they then have a certain consistency, and 2. they then all have logKs at high temperatures. If you do this don't forget the O2(g) = O2(aq) swap. I trust no one is fooled by the 4 decimal places. S&B give only one figure past the decimal. If you want to have a copy of Spahiu and Bruno, it is available free from SKB in Sweden. I saw some data there that are not in GWB, but I did not do a thorough comparison.

 

3. There is a typo in the data for YbPO4(aq). The reaction in thermo.com.v8.r6+.dat is

YbPO4(aq) + H+ = HPO4-- + Yb+++; log K = -9.5782

This is a typo. log K should be -0.5782. The reaction in Spahiu and Bruno (1995) p.51 is

YbPO4 = Yb+3 + PO4-3; log K = -12.9

Add to this the reaction from the database

PO4--- + H+ = HPO4-- log K = 12.3218

and you get -0.5782 for the above reaction.

 

These errors came to light when I translated thermo.com.v8.230 from GWB format to Phreeqc format.