This study included 48 adult male Sprague–Dawley rats, each weighing 450 g and aged 4 months (Janvier Labs, Le Genest-Saint-Isle, France). One examiner completed all the steps of the study experiment. This investigation was conducted in accordance with the guidelines of the European Parliament and the council on the protection of animals used for scientific purposes, ARRIVE (Animal Research: Reporting of In Vivo Experiments) [7], and Directive 2010/63/EU. The study protocol received ethical approval and consent to participate from the appropriate local authority (Landesamt für Natur und Verbraucherschutz, Recklinghausen, Germany; Ref. 2019A276).

For a tempering time of 1 min, the six test groups were divided into cold and warm temperatures of 4 °C, 3 °C, 2 °C, 48 °C, 49 °C, and 50 °C. These temperatures were chosen based on the findings of a prior investigation [8]. In each group, six temperatures were randomly tested in eight animals. Each animal received one implant in each tibia (Modus 2.0 3-mm cortical screws, Medartis AG, Basel, Switzerland). Therefore, for each temperature, eight test und eight control samples were investigated, resulting in 12 groups with 96 implants. All the procedures were performed under general anesthesia. At 30 min before the start of surgery, the animals were weighed. Then, the animals received 0.03-mg buprenorphine subcutaneously (0.1-ml Temgesic/kg of body weight, Remedix GmbH, Germany). Anesthesia was induced with an inhalation narcotic in an induction box with isoflurane (4 vol.%) and oxygen (vol. 30%) air mixture (Isofluran, Piramal GmbH, Hallbergmoos, Germany). Inhalation narcosis with an isoflurane (2 vol.%) and oxygen (30%) air mixture was continued via a nasal mask. After the induction of anesthesia, the animals were positioned on an adjustable warming mat.

Before implant insertion, one tibia was randomly chosen as the test side in each animal. Without tempering, the contralateral side was used as the control group. Before temperature application and implantation, the skin was shaved, disinfected, and cut with a scalpel after sterile draping. Implantation was performed after exposure of the tibia by predrilling with a pilot drill of 1.5-mm diameter under strict cooling with sterile saline (Medartis AG, Basel, Switzerland). In the case of the test group, the thermal treatment was administered after predrilling and before implant placement. The device exhibited a congruent clamp fit in the drill studs. A screw was inserted in accordance with the manufacturer's protocol by using a screwdriver and a torque of 15–20 Ncm. After primary stable insertion of the implants, resonance frequency analysis with hand-screwed individual smart pegs (Osstell, Gothenburg, Sweden) was performed to measure the primary stability after implant insertion (Fig. 1). Primary stability was measured with the implant stability quotient (ISQ) in four directions (i.e., from left to right and from front to back [9]), resulting in a calculated mean ISQ value. The wounds were closed with suture material. By using a standardized 90° dental X-ray computed tomography device (Dentsply Detrey, Germany, Konstanz), the distance between the alveolar crest at the implant and the implant shoulder was measured directly after surgery on both implant sides [10]. The measurements were taken between the implant shoulder and the greatest coronal level of the direct bone-to-implant contact (Fig. 2). All the measurements were performed using specialized computer software (ImageJ Version 1.51, National Institutes of Health, USA) [11]. Every day postoperatively, the animals were treated once with carprofen 4 mg/kg subcutaneously (Rimadyl, Zoetis GmbH, Berlin, Germany) for analgesia. At the second follow-up, 7 days after surgery, ISQ evaluation and radiographic imaging were repeated.

A After implant insertion, the ISQ values were measured for the first time. B The mini-screws used. For the ISQ values, threads were cut on the head of the screw. C The second implant stability measurement was performed after 7 days

A In the case of the test group, the thermal treatment was administered after predrilling. The surface of the device individually tempered the bone surface B After the thermo-treatment, implantation was performed. C The radiographic evaluation was performed directly after the surgery and after 7 days. D A standardized radiographic measurement procedure was used to measure the crestal bone changes around the implant body

The EDX and TEM analyses included six drill samples, one for each of the following temperatures: 4 °C, 3 °C, 2 °C, 48 °C, 49 °C, and 50 °C. The EDX analysis was performed using a scanning electron microscope (SEM; ESEM XL 30 FEG; FEI, Eindhoven, the Netherlands) equipped with an EDX detector system (EDAX, Mahwah, NJ) in the backscatter mode, with an acceleration voltage of 15 kV. The samples for the SEM and EDX analyses were fixed in 3% glutaraldehyde in 0.1 M Sorensen’s phosphate buffer, dehydrated in an ascending ethanol series (30%–100%), and dried at 37 °C, in accordance with the procedure used in a previously published study [6]. The EDX analysis was performed with an EDAX Genesis system (EDAX), at eight measurement points using a mean value for each sample.

For TEM, the samples were fixed in 3% glutaraldehyde in 0.1 M Sorensen’s phosphate buffer and decalcified in 0.5 M EDTA. After post-fixation in 1% OsO4 (Roth, Karlsruhe, Germany) in a 25 mM sucrose buffer, the samples were dehydrated in an ascending ethanol series, incubated in propylene oxide (Serva, Heidelberg, Germany) and embedded in Epon resin (Serva). Ultrathin Sects. (70–100 nm) were cut and stained with 0.5% uranyl acetate and 1% lead citrate (both EMS, Munich, Germany) for contrast enhancement. The samples were viewed at an acceleration voltage of 60 kV by using a Zeiss Leo 906 TEM service (Carl Zeiss, Oberkochen, Germany).

Analyses were performed using the Prism 8 software for Mac OS X (GraphPad, La Jolla, CA) running on Apple OS X. The variables were analyzed using the Kolmogorov–Smirnov normality test. The Kruskal–Wallis and Dunn’s multiple comparison tests with adjustment were used to identify differences between the parameters.

Post hoc power analysis was performed with the G*Power software (Heinrich-Heine-Universität, Düsseldorf, Germany) using the post hoc analysis of variance with groups to determine a power of 100% (parameter primary study aim EDX) based on the total sample size of 30 and six groups using an effect size of 11.23 and an α of 0.05.