In this study, we assessed two different implant materials in the form of round specimens (each measuring 5.0 mm in diameter and 1.0 mm in thickness, see Table 1). Half of the specimens were made of grade 1 pure titanium (Mechanische Werkstatt Biologie, University of Regensburg, Germany) and the other half of zirconia ceramic (IPS e.max ZirCAD; Ivoclar Vivadent, Ellwangen, Germany). The grade of the titanium used is the purest commercially available alloy. In comparison to other titanium grades, it is ductile and soft; however, there are very low amounts of impurities (≤1625%) and thus the lowest interferences caused by contained trace elements. The zirconia ceramic is a high-strength yttrium-stabilized zirconium oxide ceramic and as such a metal oxide ceramic. Due to its excellent mechanical properties, this ceramic is used in a wide range of indications.

Twenty specimens of each experimental implant material were subjected to one of the following surface treatments to modify surface roughness and surface free energy. The surface of some specimens was polished to high gloss with a polishing machine (Motopol 8; Buehler, Düsseldorf, Germany) and wet abrasive paper discs (Buehler, Lake Bluff, IL) with a grit of 1000, 2000, and 4000. Other specimens were sandblasted either with 50 or 250 μm aluminum trioxide at 2.5 bar for 20 s (both; Korox, Bego, Bremen, Germany). In the second part of the investigation, we additionally modified surface free energy values on the material surfaces of the rough and smooth substrata by applying n-propylsilane; hydrophilic conditions were altered by the application of aminosilane. As a result of various surface finishes (roughness and surface free energy) and the two starting materials (titanium and ceramic), there were finally ten different groups of test specimen with unique properties.

Surface roughness values of three specimens of each of the ten material groups were determined at three different sites with a stylus instrument (Perthometer S6P; Perthen, Göttingen, Germany) and shown as the arithmetic average peak-to-valley value (R _a ). Water contact angles (hydrophobicities) were calculated from automated contact angle measurements (OCA 15 plus; Dataphysics Instruments, Filderstadt, Germany) with deionized water. Nine drops of the liquid (one drop 1 μl) were examined on each substratum, and the contact angle was measured exactly 15 s after the positioning of the drop.