I. Kawabe et al., Distribution coefficients of REE between Fe oxyhydroxide precipitates and NaCl solutions affected by REE-carbonate complexation, GEOCHEM J, 33(3), 1999, pp. 181-197
Distribution coefficients of REE between Fe oxyhydroxide precipitates and N
aCl solutions doped with NaHCO3 have been determined in pH = 8.1-8.6 and at
room temperature and pressure. The coefficient is defined as K-d(REE: ppt.
/sol.), where REE designates each lanthanide (Ln), Y or Sc. The NaHCO3 conc
entration was changed in the range of (0 similar to 12) x 10(-3) M under a
constant NaCl concentration (0.45 M). K-d(Sc) rapidly approaches K-d(Lu) as
[NaHCO3] increases. K-d(Y) is lower than K-d(Ho) even in solutions with hi
gh [NaHCO3]. The convex tetrad effect in logK(d)(Ln) becomes less conspicuo
us with increasing [NaHCO3]. We proposed a method to determine REE-carbonat
e complexation constants from observed variations of K-d(REE) with increasi
ng [CO32-,aq]. In solutions with [NaHCO3] greater than or equal to 1 x 10(-
2) M, the dominant dissolved REE species are REE(CO3)(2)(-)(aq) except for
La. We have tentatively determined the stability constants of beta(2) for R
EE(CO3)(2)(-)(aq) from our preliminary data set, although beta(1) for REECO
3+(aq) could not be estimated. The series variation of log beta(2) are comp
atible with literature values. On the basis of the refined spin-pairing ene
rgy theory (RSPET), we have analyzed the series variations of log beta(2) a
nd logK(d)(Ln) values with and without Ln(III)-carbonate complexation effec
t. Racah E-1 parameter is approximately the same between Ln(OH)(3). nH(2)O
as the precipitate and Ln(CO3)(2)(-)(aq), whereas Racah E-3 parameter of Ln
(CO3)(2)(-)(aq) is only slightly larger than that of Ln(OH)(3). nH(2)O. Thi
s is the reason that the convex tetrad effect of logK(d)(Ln) diminishes as
Ln(III)-carbonate complexation proceeds. Our experimental logK(d)(Ln) value
s and apparent logK(d)(Ln) ones from marine Mn-Fe deposit/seawater pairs su
ggest that reported log(beta(1)/beta(2)) values for light Ln are slightly s
maller than those ought to be.