*3.1. Materials*

For catalyst preparation, hydrolysis lignin was obtained through a biomass hydrolysis process from Sekab Oy, Sweden. Other chemicals used in catalyst preparation, conversion reactions, partitioning experiments and analyses were used as received, without any purification. For high pressure liquid chromatography (HPLC) sample preparation, regenerated cellulose (RC) syringe filters (0.45 μm, 25 mm, Phenomenex) were used for the organic samples and polytetrafluoroethylene (PTFE) syringe filters (0.45 μm, 4 mm, Phenomenex) were used for the water samples.

#### *3.2. Furfural Partitioning in Biphasic Reactor System*

Furfural partitioning experiments were carried out using a 4.7 wt% furfural solution in water as feed and toluene or MIBK as an organic solvent. The furfural solution and organic solvent were measured with alternative ratios (1:1, 1:2 and 1:3) into a microwave reactor tube with a magnetic stirring bar. The tube was closed, heated five minutes at 160 ◦C and cooled to room temperature. Samples from both layers were analyzed by HPLC to calculate the partitioning for furfural in different solvent systems.

#### *3.3. Catalyst Preparation and Characterization*

Activated carbon (AC) supports were prepared from hydrolysis lignin, dried in oven at 105 ◦C and crushed to a particle size of <425 μm. Activation was performed using either a chemical or physical activation method. Chemical activation was done by impregnation of zinc chloride into the dried lignin using a 2:1 mass ratio of ZnCl2:biomass. ZnCl2 dissolved into H2O was mixed with the biomass for 3 h at 85 ◦C and then dried in the oven at 105 ◦C until achieving a constant weight. The carbonization and activation of the dried ZnCl2-impregnated lignin was done in a stainless-steel tube in a tube furnace (Nabertherm RT200/13) (Nabertherm GmbH, Lilienthal, Germany) at 600 ◦C for 2 h using a heating ramp of 10 ◦C/min. During the thermal heating process, the reactor was flushed continuously with N2 (flow 10 mL/min). Alternatively, carbonization followed by physical activation was performed in in one-step process in a stainless-steel tube in a tube furnace using a heating ramp of 10 ◦C/min to 800 ◦C. At the target temperature, steam was added by feeding water at 0.5 mL/min into the reactor for 2 h. During the thermal heating process, the reactor was flushed continuously with N2 (flow 10 mL/min). Both resulting activated carbons were washed with hot water, dried overnight at 105 ◦C, crushed and sieved to a fraction size of <100 μm. The supports were named ACz (AC zinc chloride-activated and water washed) and ACS (AC steam-activated and water washed). In addition, a support with chemical activation and HNO3 treatment was prepared (ACzN). This was performed in a round bottom flask with a 10:1 mass ratio of 3 M HNO3 per support and heated for 4 h at 85 ◦C. After the acid treatment, the support was filtrated and washed with hot distilled water until neutral pH was obtained and finally it was dried in the oven at 105 ◦C.

In order to modify the carbon supports with iron, metal salts (FeCl3·6H2O or Fe(NO3)3·9H2O) were added by incipient wetness impregnation on the support, aiming that the targeted concentration of iron in the catalyst was 5 or 10 wt% of the total catalyst mass. The metal salts were dissolved in distilled water equal to the pore volume of the support and mixed with the support, matured for 5 h at room temperature and finally dried in an oven at 105 ◦C for 16 h. Finally, the catalysts were calcined at 400 ◦C for 2 h with a continuous flush of N2 (flow 10 mL/min). The iron-impregnated catalysts were named 5Fe-ACs, 10Fe-ACs, 5FeNO3-ACz, 5Fe-ACz and 5Fe-ACz N according to the targeted iron concentration, type of support and type of iron precursor (FeNO3 if mentioned, otherwise FeCl3).

Specific SAs and pore size distributions were determined from the physisorption adsorption isotherms using nitrogen as the adsorbate. Determinations were performed with a Micromeritics ASAP 2020 instrument (Micromeritics Instrument, Norcross, GA, USA). Portions of each sample (100–200 mg) were degassed at low pressure (0.27 kPa) at a temperature of 140 ◦C for 3 h in order to remove adsorbed gas. Adsorption isotherms were obtained by immersing sample tubes in liquid nitrogen (-196 ◦C) to achieve constant temperature conditions. Gaseous nitrogen was added to the samples in small doses and the resulting isotherms were obtained. SAs were calculated from adsorption isotherms according to the BET (Brunauer–Emmett–Teller) method [56]. The percentual distribution of pore volumes (vol%) was calculated from the individual volumes of the micropores (pore diameter <2 nm), mesopores (pore diameter 2–50 nm) and macro-pores (diameter >50 nm) using the DFT (Density Functional Theory) model [57]. The instrumental setup enabled the measurement of micropores down to 1.5 nm in diameter, even if there might have been some contribution from smaller pores. The SAs were measured with a precision of ~5%.

The metal contents of the catalysts and supports were measured by ICP-OES using the 5110 VDV instrument (Agilent Technologies, Santa Clara, CA, USA). Zn, Fe, Ca, K, Mn, S, Na and Mg were measured from the ACs support; Zn and Fe contents were measured from ACz, 5FeNO3-ACz, 5Fe-ACz, ACz N and 5Fe-ACz N; while only the Fe content was determined from 5Fe-ACs and 10Fe-ACs. For determination, samples of 0.1–0.2 g were first digested in a microwave oven (MARS, CEM Corporation) using the EPA 3051A method with 9 mL of HNO3 and 3 mL of HCl [58]. Subsequently, the solution was diluted to 50 mL with water and the former elements were analyzed with the ICP-OES.

XPS analyses were performed using the ESCALAB 250Xi XPS System (Thermo Fisher Scientific, Waltham, MA, USA). With a pass energy of 20 eV and a spot size of 900 μm, the accuracy of the reported binding energies (BEs) was ±0.3 eV. Fe, C, O and Cl were measured for all samples. The measurement data were analyzed using Avantage software. The monochromatic AlK α radiation (1486.6 eV) was operated at 20 mA and 15 kV. Charge compensation of the BEs was performed by applying the C1s line at 284.8 eV as a reference.

XRD was used to study the phases of 5Fe-ACs and 10Fe-ACs utilizing PANalytical X'Pert Pro X-ray di ffraction equipment (Malvern Panalytical, Almelo, Netherlands). The di ffractograms were collected in the 2θ range of 5–90◦, with a step size of 0.017◦ and a scan speed of 1.06◦/min using monochromatic CuK α1 radiation (λ = 1.5406 Å) at 45 kV and 40 mA. The crystalline phases and structures were analyzed with HighScore Plus software and the peaks were identified using International Centre for Di ffraction Data ICDD (PDF-4 + 2020).

The morphology of the catalyst particles was studied using SEM and STEM. A JEOL JEM-2200FS energy-filtered transmission electron microscope equipped with a scan generator (EFTEM/STEM) (JEOL Ltd., Tokyo, Japan) was used for STEM analysis. The catalyst samples were dispersed in pure ethanol and pretreated in an ultrasonic bath for several minutes to create a microemulsion. A small drop of the microemulsion was deposited on a copper grid pre-coated with carbon (Lacey/Carbon 200 Mesh Copper) and evaporated in air at room temperature. The accelerating voltage in the measurements was 200 kV, while the resolution of the STEM image was 0.2 nm. The metal particle sizes were estimated visually from the STEM high-angle annular dark field (HAADF) images. The SEM was performed

with a Zeiss Sigma Field emission scanning electron microscope (FESEM). In the sample preparation, a powder sample was placed on a conductive glue tape. The SEM images were taken at a voltage of 5 kV and a working distance around 5 mm.

Catalyst surface acidity was characterized by applying the Boehm titration method [59–63]. A total of 100 mg of catalyst was weighed and mixed with 50 mL 0.01 M NaOH. Samples were shaken (300 rpm) in sealed tubes for 72 h at room temperature and then filtered using a syringe and syringe filter (0.45 μm, regenerated cellulose). Titration was carried out using a back-titration method by taking 10 mL of filtrate, mixing it with 20 mL of 0.01 M HCl and finally back-titrating with 0.01 M NaOH. Acidic groups were calculated using Equation (2), based on the theory that NaOH neutralizes all acidic oxygen groups (including phenols, lactonic groups and carboxylic acids) present on carbon. Nonconsumed base content was neutralized with acid and then nonconsumed acid was quantified through simple acid-base titration.

#### *3.4. Furfural Production from Xylose*

In a conversion reaction, 0.25 mmol (37.6 mg) of xylose and 0.0036/0.050 mmol of homogeneous metal salt (AlCl3·6H2O, ZnCl2, CrCl3·6H2O, SnCl2·2H2O or FeCl3·6H2O) or 5 mg heterogeneous carbon-based catalyst were placed into a 5 mL reaction tube. A magnetic stirring bar, water (1 mL) and MIBK (3 mL) were added and the tube was sealed. The reaction was carried out in a Biotage Initiator microwave reactor (Biotage, Uppsala, Sweden) at 160/170/180 ◦C for 30 min to 3.5 h. After the reaction, approximate 1 mL samples from both layers were filtered with a syringe filter (an RC filter for the organic layer and a PTFE filter for the water layer) and then analyzed with HPLC.
