**7. The chlorites in both methods**

Due to the great variability of the chlorites, 3 versions (I, II, and "Grundversion" = GV) were used as a basis for the calculation of slatenorm. Later only the "Grundversion" was used as a basis for slatenorm and slatecalculation. Jung & Wagner (1996–2000) [24] determined for chlorite porphyroblasts in roof slate relatively high iron and aluminum contents with Fe/(Fe + Mg) ratios between 0.6 and 0.7 and a replacement of silicon by aluminum [IV] of 35–40 (Weight-%). According to the classification of the chlorites, these are aphrosiderites in this case (**Figure 1**). The three versions concern the inherently variable proportion of aluminum that replaces tetravalent silicon (Si). The Mg-Al-chlorite (mac for short) and Fe-Al-chlorite (fac for short) are taken as the basis as standard minerals.

They are given here as simplified formulas without taking trivalent Fe into account:

Mg-Chlorite:

Serpentine 6MgO 4SiO2 4H2O in all versions slatenorm and slatecalculation for mc. Clinochlore 5MgO 1Al2O3 3SiO2 4H2O in slatenorm version II for mac. X 4,5MgO 1,5Al2O3 3SiO2 2,5H2O in slatenorm version I for mac.

### **Figure 1.**

*Composition of chlorite porphyroblasts in Spanish roof slates, light—Field of variation of natural chlorites, further explanations in the text [33].*

Amesite 4MgO 2Al2O3 2SiO2 4H2O in slatenorm "Grundversion" (= GV) and slatecalculation for mac.

Fe-Chlorite:

Greenalite 6FeO 4SiO2 4H2O in all versions slatenorm and slatecalculation for fc. Chamosite 5FeO 1Al2O3 3SiO2 4H2O in slatenorm version II for fac. Y 4,5FeO 1,5Al2O3 3SiO2 2,5H2O in slatenorm version I for fac.

Daphnite 4FeO 2Al2O3 2SiO2 4H2O in slatenorm "Grundversion"(= GV) and slatecalculation for fac.

As an example, full chemical analyses of five different sample (**Tables 3** and **4**) were selected, including a shale, a phyllite, two low-carbonate roofing slates from Germany and Spain, and roofing slate with carbonate (Magog). The biggest difference between slatecalculation and the three versions of slatenorm (I, II, and "Grundversion" = GV) is when hydro-micas (illite and brammallite) are calculated. In this case, more SiO2 and Al2O3 are "consumed" than with micas and thus fewer chlorite minerals are calculated.

In the case of shales (**Table 3**), there are only meaningful results with slatecalculation, i.e., when considering and calculating hydro-micas. The old slatenorm, on the other hand, does not lead to meaningful, sometimes negative results.


### **Table 3.**

*Seven exemplary full chemical analyzes [5, 7, 33–35].*



**Table 4.**

*Results of a comparison of all versions of slatenorm and slatecalculation (analysis from Table 3) [33].*

In the German roof slates, too, slatecalculation results in the hydro-micas illite and brammallite, which also lead to lower chlorite values. The appearance of chloritoid in slatenorm version II in the roof slates from Altlay/Germany and San Pedro de T./Spain, which actually do not contain this mineral, shows that this version is incorrect. There are differences in ore minerals including titanium minerals between the results of slatecalculation and all versions of slatenorm. Slatecalculation determines additional minerals, such as pyrrhotite pn and titanomagnetite tm, with the additional consideration of Fe2O3. This leads to higher ore mineral values. At the same time, the Fe chlorite values are lower (fc and fac).

Due to the natural variability of chlorites and the use of standard minerals, such as fc and mc, which do not actually occur as end links in nature (**Figure 1**), the results of both standard mineral calculations should only be given as sums (chlorite = mac +

mc + fac + fc), possibly divided into Fe-chlorite (= fac + fc) or Mg-chlorite (= mac + mc). The "Grundversion" of slatenorm works better than the other two versions I and II and is therefore preferable. The more complex calculation of slatecalculation leads to better values than slatenorm, especially when hydro-micas are calculated.
