Classification analysis of archaeological findings from early-iron production (Turopolje region, NW Croatia) based on multi-analytical profiling

The combination of XRF, SEM-EDS, ICP-AES, ICP-MS, and chemometric methods for the investigation of early-iron production artefacts is presented. Visually similar samples were randomly collected from stored humps of material that was excavated in the Turopolje region, NW Croatia. Elemental analysis of different kinds of slags and pit furnace remains is performed from major (Fe, Si, Al, Ca, K, Mg, P, Mn, and Ti) to trace constituents (Ba, Cu, Cr, Ni, Mo, Pb, V, Zn, Zr, and REE's). Applied statistical analyses (PCA and HCA) for enlarged set of measured data allow the similar features of heterogeneous materials to become clearly discernible. Thus, three types of slags, namely, iron-rich tap slag, bloom slag, and ceramic-rich slag are recognized. Moreover, retrospective examination of morphology data from SEM-EDS observations along with elemental signature and performed statistical treatment enables identification of an exception in a set of slags; it is identified as a non-processed bog iron ore. These results are advantageous for forthcoming multidisciplinary studies of early-iron production in the Pannonic area. The combination of presented methods facilitates recognition of inhomogeneous slag material from archaeometallurgical production sites, especially when numerous similar samples are scattered at an excavation site. A combination of multi-analytical methods and chemometric tools facilitate recognition of different types of archaeometallurgical iron slags.