Influence of Natural Mordenite Activation Mode on Its Efficiency as Support of Nickel Catalysts for Biodiesel Upgrading to Renewable Diesel

In the present work, natural mordenite originated from volcanic soils in Greek islands, activated using HCl solution and HCl solution followed by NaOH solution, was used as support for preparing two metallic nickel catalysts (30 wt.% Ni). The catalysts were thoroughly characterized (XRF, N2 adsorption–desorption, SEM, XRD, TEM, H2-TPR, NH3-TPD) and evaluated for biodiesel upgrading to green (renewable) diesel. Double activation of natural mordenite optimized its supporting characteristics, finally resulting in a supported nickel catalyst with (i) enhanced specific surface area (124 m2 g−1) and enhanced mean pore diameter (14 nm) facilitating mass transfer; (ii) easier nickel phase reduction; (iii) enhanced Ni0 dispersion and thus high active surface; (iv) balanced population of moderate and strong acid sites; (v) resistance to sintering; and (vi) low coke formation. Over the corresponding catalyst, the production of a liquid consisting of 94 wt.% renewable diesel was achieved, after 9 h of reaction at 350 °C and 40 bar H2 pressure, in a semi-batch reactor under solvent-free conditions.


Introduction
The alarming rate of climate change in combination with the diminishing of petroleum reserves have spurred to the urgent need to develop sustainable alternative fuels suitable for replacing fossil fuels. Although biodiesel (FAMEs) produced using transesterification of triglyceride biomass with a light alcohol [1,2] has already entered in the fuel market as a very promising alternative to petro-diesel, its utilization without mixing with the latter is still problematic. Biodiesel is very unstable (at low temperature it is prone gelling and at high temperature it grows mold), it may damage filters and pipes of the existing vehicles and has lower fuel efficiency than petro-diesel [2]. Considering these drawbacks, it is of great interest to design catalytic systems to upgrade biodiesel to the next generation, higher value fuel consisting of hydrocarbons in the diesel range, so-called renewable or green diesel.
Hydrotreatment has proven to be a very promising process for upgrading FAMEs to renewable diesel [3][4][5]. This process aims to their selective deoxygenation (SDO), avoiding extended cracking to hydrocarbons with carbon atoms much less than those existing in the initial fatty acids. Several catalytic systems have been studied for SDO. Noble metal catalysts have proven very effective [6][7][8] but too expensive for industrial applications. Sulfided NiMo [9] and CoMo [10,11] catalysts, which are traditionally used for oil refining, have also been studied. However, the need for continuous sulfidation upon SDO process makes their use undesired. In the last decades, base metals have attracted researchers' interest for the development of SDO monometallic, bimetallic or multi-metallic catalysts

Acid and Acid-Base Treatment of Natural Mordenite
Acid activation of natural mordenite was performed as described previously [5] using treatment with hydrochloric acid aqueous solution (2 M, mass of solid to solution volume ratio: 1 g/20 mL) followed by drying (110 • C overnight), calcination (500 • C for 2 h) and grinding. The support obtained is symbolized as MO A .
Acid-base treatment of natural mordenite was performed by treating a MO A sample with NaOH aqueous solution (2 M) in a solid mass-to-solution volume ratio equal to 1 g/15 mL. The suspension was stirred at 65 • C for 30 min, filtered, washed several times with distilled water and dried at 110 • C overnight. The solid obtained and the NH 4 NO 3 aqueous solution (1 M) were mixed in a round-bottomed flask, retained at 80 • C for 24 h under stirring, followed by cooling at 25 • C. The suspension was filtered and washed several times with distilled water until the pH of the filtrate reached 7. The obtained solid was dried at 110 • C for 12 h, calcined at 500 • C for 2 h and grinded to become fine grained. The support obtained is symbolized as MO AB .

Nickel Catalysts Preparation
The deposition-precipitation method was used for the catalysts' preparation with 30 wt.% Ni. They are symbolized as NiMO x , where x represents the activation mode of natural mordenite (A or AB). Details of the preparation procedure are provided in Supplementary Materials.

Catalysts' Characterization
The supports and catalysts' compositions were determined using X-ray Fluorescence spectroscopy (XRF). Textural characteristics of the samples were determined using nitrogen physisorption. X-ray powder diffraction (XRD) was used to identify crystal phases of the samples. The morphology of the samples was observed using scanning electron microscopy (SEM). Metallic nickel crystal size distribution was calculated using transmittance electron microscopy (TEM) results. Samples' acidity was determined using temperatureprogrammed desorption of ammonia (NH 3 -TPD). Reducibility of the catalyst precursors was studied using temperature-programmed reduction with H 2 (H 2 -TPR). Combustion elemental analysis was used to determine the coke deposited on the spent catalyst samples. Experimental details and setups used are provided in Supplementary Materials.

Catalysts' Evaluation
Catalyst performance was evaluated in a high-pressure (40 bar) semi-batch reactor fed with H 2 (100 mL/min) at reaction temperatures 310, 330 and 350 • C under solvent-free conditions [28]. For more details, see Supplementary Materials.

Catalysts' Characterization
In the present work, nickel catalysts supported on (singly-acid and doubly-acid/base) activated natural mordenite were prepared using the deposition-precipitation method. Ni loading (30 wt.%) and Si/Al atomic ratio in the natural mordenite as well as the activated supports (Table 1) were determined using XRF.

Textural Properties
The porosity and the surface area of the supports and catalysts were determined using the nitrogen adsorption-desorption method. The isotherms obtained are shown in Figure S1. The MO nat. isotherm, according to IUPAC classification, is of type I; the activated supports and the corresponding nickel catalysts isotherms are of type IV. They present a steep curvature at low relative pressure, which is characteristic of microporous materials (pore diameters ≤ 2.0 nm) and a hysteresis loop at higher relative pressures indicating the development of a mesoporous network of pores (2.0 ≤ pore diameters ≤ 50.0 nm). The activated supports present as an H1 hysteresis loop, which is characteristic of porous materials exhibiting well-defined cylindrical-like pores. The NiMO A and NiMO AB catalysts show an H4-type hysteresis loop, indicating the formation of narrow slit pores [29]. Table 1. Physicochemical characteristics of the catalysts. (Si/Al Ratio obtained using XRF analysis of the materials, total specific surface area, S BET ; specific surface area of micropores, S micro ; specific pore volume in meso-and macropores, V BJH ; mean pore diameter of meso-and macropores, d BJH ; mean crystal size of Ni 0 nanocrystals determined using XRD, MCS Ni 0 and Ni 0 surface area, S Ni 0 ). The textural characteristics of the materials are illustrated in Table 1. The acid treatment of MO nat. seems to drastically enhance the S BET , mainly the surface of the micropores. This is due to the removal of Na + , K + , Mg 2+ , Ca 2+ and Fe 3+ from the zeolite channels. Upon acid treatment of MO nat. , Al 3+ ions are removed from its framework (dealumination process), resulting in an increase in Si/Al ratio. This process leads to the hydrolyzation of Si-O-Al bonds, producing -Si-OH and -Al-OH, and forming vacant sites in the zeolite framework. Further treatment of this material with a base leads to Si/Al ratio diminution of MO A (desilication process) and to a further enhancement of S BET . Acid and acid-alkaline treatment of MO nat. also lead to an increase in both pore volume and pore diameter (Table 1) [30].

Sample
The addition of Ni to the activated supports seems to decrease the corresponding specific surface area, mainly by blocking the micropores. This is reflected by the substantial decrease in S micro values (Table 1). On the other hand, the values of the specific pore volume are higher on the nickel catalysts than those of the corresponding supports. This indicates that new pores are created after nickel phase deposition. This becomes more obvious in Figure 1, which presents the pore size distribution curves of the materials. Indeed, the NiMO A and NiMO AB catalysts' curves appear maximal in the mesoporous range 20-30 nm. This is a positive fact, reducing mass transfer limitations upon reactions over such catalysts.    (Figure 2a), which seems to be created by the adhesion of fibrous structures. On the other hand, acid-base treatment made the fibrous structure in MOAB more visible (Figure 2b). Deposition of the nickel phase on the above supports created flower-like structures that covered the corresponding supports (Figure 2c,d). Combining the new structures with the N2 physisorption results discussed in Section 3.1.1, one can conclude that these are responsible for the creation of new mesopores.   Figure 3 presents the XRD patterns of the activated mordenite supports (MO A and MO AB ) and the corresponding catalysts (NiMO A and NiMO AB ). Inspection of this figure shows that zeolite framework remains almost intact after either acid or acid-alkali treatment of the parent material or after the deposition of the nickel phase. However, a slight shift towards low 2 theta angles of mordenite diffraction peaks was observed after acid-alkali treatment. This could be attributed to the desilication mentioned above and is in accordance with the increase in S BET value ( Table 1). The diffraction peaks at 2θ 44.58, 51.80 and 76.31 • are attributed to the metallic nickel crystals (Ni 0 ) with (111), (200) and (220) planes (PDF 87-0712). The diffraction peaks at 2θ 37.23, 43.29 and 62.92 • are attributed to the NiO (111), (200) and (220) planes, respectively (PDF 65-6920). Metallic nickel is the main crystal phase detected in the NiMO A and NiMO AB catalysts. In addition, NiO diffraction peaks are rather obvious in the NiMO A catalyst. This indicates that the NiO species formed upon thermal treatment (at 400 • C under Ar) before reduction interact more strongly with the MO A than the MO AB support. The mean crystal size values of the Ni 0 (MCS Ni 0 ) were estimated from the XRD data at 2θ equal to 51.8 • when using Scherrer's equation (Table 1). In both catalysts, these values were found to be almost the same and equal to 10 nm. This small value is evidence of very good dispersion of the nickel phase. However, deconvoluting the overlapped peaks corresponding to NiO (2θ: 43.29 • ) and Ni 0 (2θ: 44.58 • ), we made a rough estimation of the (Ni 0 /NiO) ratio and found a value of about two in the NiMO A catalyst, while this ratio became five in the NiMO AB catalyst. This is in good agreement with H 2 -TPR results discussed in the next subsection.

Structural Properties of the Materials
fraction peaks are rather obvious in the NiMOA catalyst. This indicates that the NiO species formed upon thermal treatment (at 400 °C under Ar) before reduction interact more strongly with the MOA than the MOAB support. The mean crystal size values of the Ni 0 (MCSNi 0 ) were estimated from the XRD data at 2θ equal to 51.8° when using Scherrer's equation (Table 1). In both catalysts, these values were found to be almost the same and equal to 10 nm. This small value is evidence of very good dispersion of the nickel phase. However, deconvoluting the overlapped peaks corresponding to NiO (2θ: 43.29°) and Ni 0 (2θ: 44.58°), we made a rough estimation of the (Ni 0 /NiO) ratio and found a value of about two in the NiMOA catalyst, while this ratio became five in the NiMOAB catalyst. This is in good agreement with H2-TPR results discussed in the next subsection.  Figure 4 shows representative TEM images of the catalysts, which prove that the nickel phase is evenly distributed on the supports' surfaces. The nickel particle size distributions of the NiMOA and NiMOAB catalysts were also determined (Figure 4 right) using statistical analysis of about 250 particles. As can be seen, the mean particle size of Ni 0 is ~7 nm for both catalysts, in good agreement with the XRD results (Table 1).  Figure 4 shows representative TEM images of the catalysts, which prove that the nickel phase is evenly distributed on the supports' surfaces. The nickel particle size distributions of the NiMO A and NiMO AB catalysts were also determined ( Figure 4 right) using statistical analysis of about 250 particles. As can be seen, the mean particle size of Ni 0 is~7 nm for both catalysts, in good agreement with the XRD results (Table 1).

Reducibility and Acidity Characteristics
In order to determine the suitable reduction temperature for the catalysts' activation and to further investigate the strength of the metal−support interactions and thus reducibility of the NiO to Ni 0 , H2-TPR experiments were performed using catalyst precursor samples (after Ar treatment at 400 °C and before reduction). Figure 5a shows the H2-TPR curves obtained. Inspection of the curve of the NiMOA

Reducibility and Acidity Characteristics
In order to determine the suitable reduction temperature for the catalysts' activation and to further investigate the strength of the metal−support interactions and thus reducibility of the NiO to Ni 0 , H 2 -TPR experiments were performed using catalyst precursor samples (after Ar treatment at 400 • C and before reduction). Figure 5a shows the H 2 -TPR curves obtained. Inspection of the curve of the NiMO A sample reveals that its NiO is reduced in a wide temperature range (235-542 • C). This means that well-dispersed NiO with various interaction strengths with the MO A support was formed upon preparation [5]. A shoulder appearing in the range 542-625 • C could be attributed to the reduction in Ni 2+ -species incorporated in the support surface layers [31]. As for the NiMO AB sample, the corresponding reduction curve shows two peaks. The first one, with a maximum at~200 • C, indicates the existence of NiO weakly interacting with the MO AB surface. The second and more intense peak, appearing at the 235-542 • C temperature range, indicates that the main part of NiO interacts moderately with the support surface. It should also be stressed that the NiMO AB reduction curve does not present the aforementioned shoulder in the range 542-625 • C, indicating that incorporation of Ni 2+ species in the support surface layers did not take place in this sample. The H 2 -TPR results confirm our previous conclusion, drawn using XRD analysis, that there is a stronger interaction between NiO species and the MO A support. This is also in good agreement with the fact that the total amount of H 2 consumed for the reduction in NiO supported on MO AB is higher than that consumed for the NiO supported on MO A ; however, the two samples have the same nickel loading. NH3-TPD experiments were performed to investigate the acidity of the supports and the nickel catalysts [5,32]. Figure 5b illustrates the corresponding curves. Based on these curves, we calculated the total acid site populations and their distribution according to their strength ( Table 2). All these curves were deconvoluted into four peaks. A low temperature desorption peak, with a maximum at ~106 °C, has been attributed to desorption of NH3 physisorbed on the sample's surface. The corresponding amount was not taken into account for the calculation of the surface acid sites' population. A second peak, appearing at ~190 °C, is attributed to weak acid sites. The third peak, with a maximum at ~300 °C, is attributed to surface acid sites with moderate acidity. The forth peak, at ~400 °C, is attributed to the strong acid sites. Table 2 involves the total acidity values as well as the percentages of acid sites according to their strength. Inspection of Figure 5b and Table  2 shows that the number of acid sites of MOAB is higher than that of MOA. The acidity of the supports is mainly attributed to Brønsted acid sites corresponding to the hydroxyl group formed on the oxygen atom that bridges an aluminum atom with a silicon one, so a negative charge occurs, which can be recompensed by the proton (H + ). Both supports exhibit enhanced population of weak acid sites ( Table 2). The alkali treatment of MOA leads to its desilication (Table 1) and to the enhancement of the SSA. As a result, the population of the aforementioned weak acid sites increases on this support. The addition of nickel decreases the weak acid sites indicating that probably the deposition precipitation mechanism of Ni 2+ ions involves a first ion exchange step. On the other hand, nickel deposition creates a significant population of intermediate and strong acid sites (Figure 5b and Table 2). The latter could be due to the empty d orbitals of Ni 0 (Lewis acid sites), which is better dispersed on the MOA surface.  NH 3 -TPD experiments were performed to investigate the acidity of the supports and the nickel catalysts [5,32]. Figure 5b illustrates the corresponding curves. Based on these curves, we calculated the total acid site populations and their distribution according to their strength ( Table 2). All these curves were deconvoluted into four peaks. A low temperature desorption peak, with a maximum at~106 • C, has been attributed to desorption of NH 3 physisorbed on the sample's surface. The corresponding amount was not taken into account for the calculation of the surface acid sites' population. A second peak, appearing at~190 • C, is attributed to weak acid sites. The third peak, with a maximum at~300 • C, is attributed to surface acid sites with moderate acidity. The forth peak, at~400 • C, is attributed to the strong acid sites. Table 2 involves the total acidity values as well as the percentages of acid sites according to their strength. Inspection of Figure 5b and Table 2 shows that the number of acid sites of MO AB is higher than that of MO A . The acidity of the supports is mainly attributed to Brønsted acid sites corresponding to the hydroxyl group formed on the oxygen atom that bridges an aluminum atom with a silicon one, so a negative charge occurs, which can be recompensed by the proton (H + ). Both supports exhibit enhanced population of weak acid sites ( Table 2). The alkali treatment of MO A leads to its desilication (Table 1) and to the enhancement of the SSA. As a result, the population of the aforementioned weak acid sites increases on this support. The addition of nickel decreases the weak acid sites indicating that probably the deposition precipitation mechanism of Ni 2+ ions involves a first ion exchange step. On the other hand, nickel  Table 2). The latter could be due to the empty d orbitals of Ni 0 (Lewis acid sites), which is better dispersed on the MO A surface.

Catalysts' Evaluation
The evaluation of the catalysts' performance for biodiesel upgrading to renewable diesel was studied in a high-pressure semi-batch reactor without solvent, in a volume of biodiesel-to-catalyst mass ratio equal to 1 g/100 mL and hydrogen pressure of 40 bar. The GC analysis of the liquid product after 9 h reaction time is shown in Figure 6. Inspection of this figure shows that both catalysts (NiMO A and NiMO AB ) reached almost total conversion of the feed even at 310 • C. The double activation mode (acid-alkali) of natural mordenite seems to almost double the hydrocarbon production in the green diesel range from 27 wt.% to 52 wt.%. This improvement seems to take place at the expense of the intermediate high molecular weight esters (HMWE), the hydrodeoxygenation of which is considered to be the slowest reaction step under solvent-free conditions [33]. The improved catalytic performance of the NiMO AB catalyst could be attributed to (i) its enhanced specific surface area and mean pore diameter (S BET and d BJH in Table 1) facilitating mass transfer; (ii) its easier reduction (Figure 5a) resulting in enhanced Ni 0 dispersion ( Figure 4) and thus high active surface; (iii) its balanced population of moderate and strong acid sites ( Figure 5b and Table 2), which is expected to result in low coke formation (see below). Nanomaterials 2023, 13, x FOR PEER REVIEW 9 of 14

Catalysts' Evaluation
The evaluation of the catalysts' performance for biodiesel upgrading to renewable diesel was studied in a high-pressure semi-batch reactor without solvent, in a volume of biodiesel-to-catalyst mass ratio equal to 1 g/100 mL and hydrogen pressure of 40 bar. The GC analysis of the liquid product after 9 h reaction time is shown in Figure 6. Inspection of this figure shows that both catalysts (NiMOA and NiMOAB) reached almost total conversion of the feed even at 310 °C. The double activation mode (acid-alkali) of natural mordenite seems to almost double the hydrocarbon production in the green diesel range from 27 wt.% to 52 wt.%. This improvement seems to take place at the expense of the intermediate high molecular weight esters (HMWE), the hydrodeoxygenation of which is considered to be the slowest reaction step under solvent-free conditions [33]. The improved catalytic performance of the NiMOAB catalyst could be attributed to (i) its enhanced specific surface area and mean pore diameter (SBET and dBJH in Table 1) facilitating mass transfer; (ii) its easier reduction (Figure 5a) resulting in enhanced Ni 0 dispersion ( Figure  4) and thus high active surface; (iii) its balanced population of moderate and strong acid sites (Figure 5b and Table 2), which is expected to result in low coke formation (see below). Based on the above findings concerning the influence of the mordenite activation mode on the performance of the final catalysts, we tested the most active one (NiMOAB) at higher reaction temperatures (330 and 350 °C) in order to further improve the yield of the process. The results presented in Figure 6 show that the intermediate products (acids and esters) are converted to final products to a greater extent (hydrocarbons in the diesel range) as the reaction temperature increases. Indeed, as high as 94 wt.% of the liquid products consisted of renewable diesel after 9 h of reaction at 350 °C. Based on the above findings concerning the influence of the mordenite activation mode on the performance of the final catalysts, we tested the most active one (NiMO AB ) at higher reaction temperatures (330 and 350 • C) in order to further improve the yield of the process. The results presented in Figure 6 show that the intermediate products (acids and esters) are converted to final products to a greater extent (hydrocarbons in the diesel range) as the reaction temperature increases. Indeed, as high as 94 wt.% of the liquid products consisted of renewable diesel after 9 h of reaction at 350 • C.
It is well known that nickel catalysts used for SDO of triglycerides and relative compounds favor the DeCOx (DeCO and DeCO 2 ) pathway instead of HDO (Scheme S1), as hydrocarbons with odd carbon atoms are the main fraction of the hydrocarbons produced [24,34]. Figure 7 confirms that this is the case for all catalysts studied in the present work as well.   Table S1. Based on these data, we attempted a kinetic analysis of the process.    Table S1. Based on these data, we attempted a kinetic analysis of the process.
Previously reported kinetic studies use the Langmuir-Hinshelwood or Eley-Rideal mechanism to describe the SDO process of FAME and triglyceride biomass for renewable diesel production. These models describe the experimental results excellently in several cases but are quite complex and require a lot of kinetic data in order for the reaction rate constants of the various reactions involved to be determined [7,35,36]. However, some recent studies [33,[37][38][39], attempting to simplify the perplexing mechanism of triglyceride biomass SDO, adopt pseudo-first-order reaction steps. Following a similar approach, we considered that the transformation from biodiesel to green diesel obeys a pseudo-firstorder reaction rate law. Based on the kinetic data obtained from the NiMO AB catalyst for hydrocarbon production (Table S1), and using Equation (1), we calculated the apparent reaction constant k at the three reaction temperatures (310, 330 and 350 • C).
where Y is the hydrocarbon yield and t denotes reaction time.
reached after 4 h of reaction, whereas the yields of acids and esters passed through a maximum and then decreased, confirming that these are intermediate products. The yield of hydrocarbons increased monotonically with time, indicating that these are final products. Similar kinetics were obtained at all temperatures (310, 330 and 350 °C) over the NiMOAB catalyst, and the results are presented in Table S1. Based on these data, we attempted a kinetic analysis of the process.   Figure 9 shows that introduction of the calculated k values in an Arrhenius plot results in a very good linear correlation (R 2 = 0.99999), confirming the assumption made for the reaction rate law. Based on the straight-line slope, we calculated the corresponding activation energy Eα = 64.4 ± 2.7 kJ mol −1 . This value corresponds to the rate-determining steps, namely the SDO of FFAs and HMWE [33]. An Ea value higher than 40 kJ mol −1 also ensures that our results were obtained under kinetic regime, and the mass transfer phenomena's influence on the activity results is negligible. Previously reported kinetic studies use the Langmuir-Hinshelwood or Eley-Rideal mechanism to describe the SDO process of FAME and triglyceride biomass for renewable diesel production. These models describe the experimental results excellently in several cases but are quite complex and require a lot of kinetic data in order for the reaction rate constants of the various reactions involved to be determined [7,35,36]. However, some recent studies [33,[37][38][39], attempting to simplify the perplexing mechanism of triglyceride biomass SDO, adopt pseudo-first-order reaction steps. Following a similar approach, we considered that the transformation from biodiesel to green diesel obeys a pseudo-firstorder reaction rate law. Based on the kinetic data obtained from the NiMOAB catalyst for hydrocarbon production (Table S1), and using Equation (1), we calculated the apparent reaction constant k at the three reaction temperatures (310, 330 and 350 °C).
where Y is the hydrocarbon yield and t denotes reaction time. Figure 9 shows that introduction of the calculated k values in an Arrhenius plot results in a very good linear correlation (R 2 = 0.99999), confirming the assumption made for the reaction rate law. Based on the straight-line slope, we calculated the corresponding activation energy Eα = 64.4 ± 2.7 kJ mol −1 . This value corresponds to the rate-determining steps, namely the SDO of FFAs and HMWE [33]. An Ea value higher than 40 kJ mol −1 also ensures that our results were obtained under kinetic regime, and the mass transfer phenomena's influence on the activity results is negligible.

Spent Catalysts' Characteristics
The characterization results of the spent catalysts are presented in Table 3. They show a dramatic decrease in SBET in all cases in comparison with the values of the corresponding fresh catalysts (Table 1). XRD and TEM analysis of the spent catalysts (Table 3, Figures S2  and S3) show that sintering of nickel phase took place. This phenomenon could be one of the reasons for the aforementioned decrease in SBET. The sintering is more intense in the case of the NiMOA catalyst, as according to TEM results, the mean size of the Ni 0 nano-

Spent Catalysts' Characteristics
The characterization results of the spent catalysts are presented in Table 3. They show a dramatic decrease in S BET in all cases in comparison with the values of the corresponding fresh catalysts (Table 1). XRD and TEM analysis of the spent catalysts (Table 3, Figures S2 and S3) show that sintering of nickel phase took place. This phenomenon could be one of the reasons for the aforementioned decrease in S BET . The sintering is more intense in the case of the NiMO A catalyst, as according to TEM results, the mean size of the Ni 0 nanoparticles increases 5.5 times (from 7.35 nm to 40.45 nm) and according to XRD results, the mean size of the Ni 0 nanocrystals increases 1.7 times (from 11 nm to 19 nm). In contrast, the relevant changes observed in the case of the NiMO AB catalyst correspond to an increase in the mean size of the Ni 0 nanoparticles by 3.2 times (from 6.83 nm to 21.71 nm, TEM) and the mean size of the Ni 0 nanocrystals by 1.5 times (from 10 nm to 15 nm, XRD). This effect is also combined with another negative effect, the coke deposited on catalysts' surface. Coke formation is generally low in the studied catalysts. However, it is slightly higher in the case of the NiMO A catalyst and seems to decrease with the reaction temperature in the case of the most active NiMO AB catalyst (Table 3). This behavior has previously been mentioned several times for nickel catalysts [33,40].

Conclusions
The main conclusions drawn from the present study are summarized as follows: • Natural mordenite is a promising support for Ni catalysts used for the SDO of biodiesel to green diesel; • Double activation of natural mordenite using acid (HCl) solution followed by alkaline (NaOH) solution optimized its supporting characteristics, finally resulting in a supported nickel catalyst with (i) enhanced specific surface area and mean pore diameter facilitating mass transfer; (ii) easier nickel phase reduction (iii) enhanced Ni0 dispersion and thus high active surface; (iv) balanced population of moderate and strong acid sites; (v) resistance to sintering; and (vi) low coke formation.
Supplementary Materials: The following supporting information can be downloaded at: https:// www.mdpi.com/article/10.3390/nano13101603/s1, Experimental details; Scheme S1: Reaction pathways upon biodiesel transformation to renewable diesel over Ni supported catalysts; Figure S1: N 2 adsorption-desorption isotherms of natural mordenite, acid activated mordenite, acid-base-activated mordenite and the corresponding nickel/mordenite catalysts; Figure S2: Nickel particle size distributions for the spNiMO A and spNiMO AB catalysts after 9 h of reaction at 310 • C; Figure S3: XRD patterns of the spent NiMO AB catalyst after reaction at various temperatures;