**4. Conclusions**

In this study, the conversion of xylose to furfural was studied using lignin-based activated carbon-supported iron oxide catalysts. Three di fferent activated carbon supports and five di fferent catalysts were prepared and studied in furfural production. Di fferent activation methods, metal precursors and metal concentrations were used for the catalysts and di fferent temperatures and reaction times were studied in the conversion reactions. Chemical activation resulted in a higher surface area and pore volume than physical activation but in conversion reactions, physically activated catalysts produced better reaction selectivity. FeNO3 precursor yielded higher xylose conversion than FeCl3 precursor but the furfural yield and selectivity were higher with FeCl3 precursor. The best results for xylose conversion to furfural were achieved with a 4 wt% iron-containing catalyst (5Fe-ACs), which produced a 57% yield, 92% conversion and 65% selectivity at 170 ◦C in 3 h. The results with a catalyst containing more iron (9.2 wt%) were lower (54% yield, 93% conversion and 60% selectivity) in similar conditions. The catalytic amount of Fe in 5Fe-ACs was only 3.6 μmol and using this amount of homogeneous FeCl3 as a catalyst, reduced the furfural yield, xylose conversion and selectivity. Based on catalyst characterization, iron was in the form of iron oxide on the surface of the heterogeneous catalyst, which may have a ffected to its catalytic activity positively compared to FeCl3. Moreover, hydroxyl groups were detected on the surface of 5Fe-ACs, which increases catalyst Brønsted acid

sites and therefore can increase furfural production. The recycling experiments revealed that part of the iron is easily leached out of the catalyst at a high temperature and in acidic conditions and the catalyst adsorbed some reactions products. These factors decreased the furfural yield and xylose conversion after the first round of recycling but then they remained constant. Although the activated carbon-supported iron oxide catalyst needs some improvements for better stability, it is a feasible alternative to homogeneous FeCl3.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/2073-4344/10/8/821/s1, Table S1: Metal analysis from ACs by ICP-OES, Table S2: XPS results of ACs, 5-Fe-ACs and 10Fe-ACs, Figure S1: XRD results of 5Fe-ACs and 10Fe-ACs, Figure S2: Boehm titration curves, Figure S3: EDS spectra of area shown in Figure 4, Figure S4: Graphical presentation of the results presented in Table 5, Figure S5: HPLC chromatogram of water (a) and organic (b) phase of reaction solution using 5Fe-ACz catalyst. Grams show increasing side product peak at 3.1 min and furfural shoulder at 8.5 min, when 180 ◦C was used as reaction temperature, Figure S6: STEM HAADF image of three times used 5Fe-ACs, which shows large agglomerated iron particles (diameter approx. 15¨C40 nm) as well as small single particles (diameter approx. 5 nm), Figure S7: SEM images of unused 5Fe-ACs (a,c) and used 5Fe-ACs (b,d) catalysts.

**Author Contributions:** Conceptualization, A.R.; methodology, A.R. and R.K.; formal analysis, A.R., R.K. and T.H.; investigation, A.R. and R.K.; data curation, A.R.; writing—original draft preparation, A.R.; writing—review and editing, K.L., R.K., J.K., T.H. and U.L.; visualization, A.R.; supervision, K.L., J.K. and U.L.; funding acquisition, A.R., U.L. and K.L. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by Fortum Foundation, gran<sup>t</sup> numbers 201800022 and 20190005, the EU/Interreg Botnia-Atlantica, gran<sup>t</sup> number 20201508, the Foundation of Tauno Tönning, gran<sup>t</sup> number 20190154 and Nessling Foundation, gran<sup>t</sup> number 201800070.

**Acknowledgments:** Sari Tuikkanen is acknowledged for completing part of HPLC measurements and Riina Hemmilä for AAS-measurements.

**Conflicts of Interest:** The authors declare no conflict of interest.
