*4.2. Geochemistry*

Whole rock major element composition as well as multication parameters (R1, R2 [21]) and rock classification based on major element signatures (TAS for plutonic rocks [22], R1-R2) are listed in Tables 1 and 2 for both sample series KB G and KB S.

**Table 1.** Major element composition of KB G samples from western part metasomatites. TAS classification for plutonic rocks according to [22]. QAPF classification according to [20] on the basis of volumetric mineral composition as shown in Figure 2. Multication parameters R1 and R2 were calculated according to [21] with: R1 = 4Si − 11(Na + K) − 2(Fe + Ti), and R2 = 6Ca + 2Mg + Al. Abbreviations for TAS and QAPF classification: Gran = granite, Mon = quartz-monzonite, GranDio = granodiorite; abbreviations for de la Roche: Qtz rich = outside the classification fields, Alk Gran = alkali granite, Syeno Gran = syenogranite, Monzo Gran = monzogranite; LOI = loss on ignition.


As is indicated by the SiO2 values, all rocks are silica oversaturated. The samples of KB S, especially, show an increased concentration of SiO2 compared to KB G, which is in accordance to the increased quartz content indicated by the modal composition Figure 2. Other major elements are more variably distributed in the KBG sample suite, in which KBG 6 and 7 show elevated amounts of fluorite, but are in contrast to all other samples containing no albite in Figure 2. This leads to significantly decreased Na2O and Al2O3 concentrations, but enhanced values of CaO and LOI (Figure 2, Tables 1 and 2). In addition to fluorite, gittinsite, which is present among the Zr-silicates in some samples (e.g., KB G1), is a carrier of CaO, too. Variations in major element composition are mirrored in element-based rock classification. Lower SiO2 content in KB G samples but variable sums of Na2O + K2O result in TAS classifications of monzonite, granodiorite, and granite (Table 1). In contrast, the de la Roche classification, which, in addition to Si and the alkalis, includes Fe, Ti, Ca, Mg, and Al, reflects element variability by assigning the classes syenogranite, alkali granite, granodiorite, and monzogranite to the KB G samples. While all KB S samples are entirely classified as granites in the TAS scheme, the de la

Roche scheme leads to classification as alkali granite and syeno granite in some cases or does not allow classification at all. The latter can be observed in samples with highest SiO2 concentrations and R1 multication parameter > 3000. Therefore, these samples are classified as "Qtz rich" in this study.


**Table 2.** Major element composition of KB S. TAS classification for plutonic rocks according to [22]. QAPF classification according to [20]. Parameters R1 and R2 and resulting classification calculated according to [21]. Abbreviations for TAS, de la Roche, and QAPF classifications as defined in Table 1.

The results of the REE analyses of the bulk rock samples are given in Tables 3 and 4. Chondrite normalized distribution patterns are given in Figure 7. All samples show REE concentrations strongly enriched in comparison to chondritic composition [23] and they have negative Eu anomalies, which are more pronounced in KB G samples (Eu/Eu\* < 0.3) in contrast to KB S (Eu/Eu\* < 0.2), see Tables 3 and 4. The HREE (heavy rare earth elements Gd-Lu) elements show a significantly higher enrichment in the KB G samples in comparison to KB S. The latter samples, on the other hand, show a steeper pattern, with elevated enrichment of LREE (light rare earth-elements La-Eu) ratios up to 7000 but a generally lower enrichment of HREE ≤ 1000 (Figure 7). KB G samples, instead, are less enriched in LREE < 3000, which leads to flat patterns.


**Table 3.** Rare earth elements (REEs) + Zr and Nb composition of KB G. Zirconium data from handheld X-ray fluorescence (XRF) measurement. Europium anomaly Eu/Eu\* calculated based on the geometric mean of normalized Sm and Gd values.

**Table 4.** REE + Zr and Nb composition of KB S. Zirconium data from handheld XRF measurement.


Zirconium has variable concentrations across the data set. KB G samples contain Zr up to 3.3 wt. % (Figure 7 orange range) whereas only two samples have concentrations < 1.9 wt. %. Within both sample suites, those of low Zr concentrations show decreasing enrichment of the HREE. In contrast, samples with high Zr concentration are characterized by increased HREE enrichment.

Niobium concentrations show a distribution similar to the Zr. maximum values given by the method's upper limit of quantification at 2500 mg/kg. Minimum concentrations of Nb in KB G samples generally are two times higher than the concentrations of KB S samples ranging between 490 and 790 mg/kg. Concentrations of Nb are generally higher in KB G and show a variation between 1000 mg/kg and 1900 mg/kg. In comparison, the concentration of Nb in KB S samples, excluding the samples exceeding the upper limit of quantification, cover the range from 490 mg/kg to 1945 mg/kg.

**Figure 7.** REE pattern for KB G (**upper**) and KB S (**lower**). Coloured range represents maximum and minimum values of individual sample patterns (orange and red) of parallel evolution for high content of Zr in KB G and low Zr content in KB S. The samples KB G3 and G5 were identified as having the lowest enrichment of Zr in the KB G samples and showed a decreasing tendency within the HREE. The same behaviour is shown for the KB S dataset where, in contrast, the samples KB S6 and S8 contain significantly higher amounts of Zr than the other samples (red range) representing the increasing tendency of the HREE. REE concentrations were normalized against chondritic composition according to [24].
