Notes on the Dzombak and Morel plots

. the E series of plots (% sorption vs pH) match up well with the
  published ones in most cases.

. the first two published curves for Ag (1EAG1 and 1EAG2, p 156) look
  wrong in that the higher HFO in the first plot does not lead to higher
  sorption in the second, observed or calculated, which it should do.
  Without going back to the original data, I suspect the captions may be
  the wrong way round. I've actually tried to refit the published data
  based on this assumption. It does not work that well and clearly the
  original data need to be checked more thoroughly and the literature
  searched for more recent data. This fitting is done more for an
  example of how it might be done rather than as a serious exercise.

. using the minteq.v4.dat database (#  $Id: minteq.v4.dat 794 2006-02-27
  21:06:22Z dlpark) for CrO4 does not reproduce the observed curves
  exactly. The published database includes log K's for two reactions
  with the log K for the second reaction estimated by a LFER and taken
  from Table 10.10 (though the other log K has changed from 10.85 for
  one site to 10.82 in Table 10.10). It is not clear whether Table 10.10
  was produced by refitting with some type of constrained optimisation
  incorporating the LFER (possible in PhreePlot), or whether it merely
  reports the interpolated log K without refitting. In any case, my
  advice would be to not use the second reaction at this stage and
  revert to the original reaction scheme (Table 7.48) with one log K =
  10.85. Do this by either removing or commenting out the second
  reaction in the database or redefine the log K's for the second
  reactions with some very small values to make their contributions
  insignificant. With this, the fits are as published on p. 232.

. some of the Hg fits are not too good, especially 1EHG11, though others
  are close (1EHG10). This is probably due to different aqueous models
  since virtually all of the Hg is complexed with OH- or Cl- and so is
  highly dependent on the assumed aqueous reactions and their log Ks.
  Even with the reportedly-used values, I could not get a good
  correspondence in all cases. Needs closer investigation so use the
  published values with care.

. I had trouble reproducing the isotherm (I) set of plots, eg, for Zn,
  even though I managed to reproduce the example, Fig. 2.3 (anion),
  satisfactorily (see demos). My other calculated isotherms have the
  correct shape but the calculated amounts bound do not match very well
  with D&M. I suspect I have made an error somewhere, possibly with the
  units. More work to be done here.


David Kinniburgh
12 November 2008

