Erratum to: Clinopyroxene–melt trace element partitioning and the development of a predictive model for HFSE and Sc

Erratum to: Contrib Mineral Petrol (2011) 161:423–438 DOI 10.1007/s00410-010-0540-0

Unfortunately, an error in the spreadsheet used during development of the model has resulted in publication of incorrect equations. The correct equations with original equation numbers are given below:

$$ E_{M1}^{4 + } = 11228 - 5.74T + 15204X_{Al}^{T} ( \pm 298\,{\text{GPa}}) $$

(3)

$$ r_{{_{0,M1} }}^{4 + } = 0.659 - 0.008P + 0.028X_{M1}^{Al} ( \pm 0.004\,{\text{\AA}}) $$

(7)

$$ \Updelta G_{\text{elec}}^{\text{cpx}} = 14\,{\text{kJ}} $$

$$ - {\text{RT}}\ln \left( {\frac{{D_{\text{Ti}} }}{\O }} \right) = - 35730 + 2183P + 1457P^{2} ,{\text{ with standard deviation of 3}}. 8 8 {\text{ kJ}} $$

(15)

$$ - {\text{RT}}\ln \left( {\frac{{D_{\text{Ti}} }}{\O }} \right) = - 30583 + 2984P + 1253P^{2} - 1280\Updelta Z_{m}^{2} ,{\text{ with standard deviation of 3}}. 8 2 {\text{ kJ}} $$

(16)

$$ \begin{aligned} - {\text{RT}}\ln \left( {\frac{{D_{\text{Sc}} }}{\O }} \right) - \Updelta G_{\text{elec}}^{\text{melt}} \Updelta Z_{m}^{2} = - 24232 + 143T - 4117P^{2} , \\ & {\text{with standard deviation of 3}}. 9 6 {\text{ kJ and}}\,\,\Updelta G_{\text{elec}}^{\text{melt}} = 1280 \\ \end{aligned} $$

(18)

As a result of refitting of the data, Figs. 6 and 7 have been modified slightly and are shown below.

Fig. 6

Comparison of predicted versus measured $D_{\text{M1}}^{4 + }$ and $D_{\text{M1}}^{3 + }$. Figure a reveals the good correlation between D _Ti obtained from Eq. (16) and measured D _Ti used in the regressions. The same treatment is afforded to values obtained using an expression for D_Ti from Langmuir et al. (1992). b and c experimentally obtained values of D _Hf and D _Zr (not used in regressions) are plotted against predicted D_Hf and D_Zr values from Eq. (17). E _M1 and r_0,M1 used in the predictions are obtained from Eqs. (3) and (7), using the T, P and composition of the experiments. There is a close correlation between fits using DTi calculated using Eq. (16) and fits using measured D _Ti. d Comparison of calculated and experimentally determined D _Sc values used in the regressions. Due to lack of $D_{\text{M1}}^{3 + }$ data, development of $E_{M1}^{3 + }$ and $r_{0,M1}^{3 + }$ expressions has not been possible, thus impeding the determination of $D_{\text{M1}}^{3 + }$ using Eq. (18)

Fig. 7

Plots of predicted partition coefficients along the lherzolite solidus, using P–T-X data from Walter and Presnall (1994). a Plot of D _Zr/D _Hf ratios against P. Predicted ratios along the lherzolite solidus show systematic variations with P but lie within the range observed in experimental data. b Calculated D _Sc along the lherzolite solidus predict Sc remains compatible in cpx even at high-P. c Predicted D _Ti/D _Eu indicate variable fractionation along the lherzolite solidus with varying P

Other than for the fit to $E_{M1}^{4 + }$ no longer being dependent on a pressure term, the conclusions drawn in the paper are unaffected by this erratum.

We would like to thank Chenguang Sun at Brown University for drawing attention to the error.