Molybdenum, Vanadium, and Tungsten-Based Catalysts for Sustainable (ep)Oxidation

This article gives an overview of the research activity of the LAC2 team at LCC developed at Castres in the field of sustainable chemistry with an emphasis on the collaboration with a research team from the University of Zagreb, Faculty of Science, Croatia. The work is situated within the context of sustainable chemistry for the development of catalytic processes. Those processes imply molecular complexes containing oxido-molybdenum, -vanadium, -tungsten or simple polyoxometalates (POMs) as catalysts for organic solvent-free epoxidation. The studies considered first the influence of the nature of complexes (and related ligands) on the reactivity (assessing mechanisms through DFT calculations) with model substrates. From those model processes, the work has been enlarged to the valorization of biomass resources. A part concerns the activity on vanadium chemistry and the final part concerns the use of POMs as catalysts, from molecular to grafted catalysts, (ep)oxidizing substrates from fossil and biomass resources.


Introduction
Among the several challenges that the chemical industry has to face in the future, the most visible is to diminish/erase its negative image, soiled for years by several unfortunate contaminations issues, hazards being mainly bad handling procedures and storage mistakes (ammonium nitrate explosions, chemical leaks from tankers, burning of chemicals) [1]. For a century, the market of organic molecules has mainly been based on relatively cheap and available fossil resources. The "cheap" version of those raw sources is diminishing and industries have to face some geopolitical issues, increasing their prices. Maintaining the production of such organic molecules, with a relative low price/cost, obliges academic/industrial chemists to consider the development of new sources. New fossil sources can still be discovered for several years (ca. 50-100 according to Association for the Study of Peak Oil-ASPO) but with low accessibility [2,3]. The use of fossil sources being one cause of global warming, the quest toward renewable and sustainable sources seems more than urgent. In the current time, there is high interest in the carbon footprints of all processes/products for public policies but also for energy-and matter-saving processes. This was among the reasons that academic and industrials scientists began to think about solutions for a "better chemistry". Those ideas have been gathered in 12 principles in the beginning of 1990 called "Green Chemistry" [4]. This concept points out the urgent need for new processes that are more sustainable and less energy demanding. Presented solutions recommend using cleaner and safer processes to replace the actual chemical processes where hazards might more frequently occur [5]. All of those ideas are a straight line to solutions anticipating fossil depletion. nature of the metal and the ligand in the case of fundamental studies and valorization is added in an extra sub-chapter.

Methodological Approach and Model Study with Cyclooctene
Among the metal-containing species exhibiting interesting catalytic activity, molybdenum was quite active and deeply used. Coordination complexes bearing tridentate ligands attracted our attention. Indeed, Mo is known to be at the core of several industrial processes, but conditions were not as green as they could be [24,25]. The work engaged herein started from the chemistry of [MoO2L] complexes with tridentate ligands, accessible complexes for sustainable chemistry. The studied tridentate H2L ligand (H2SAP) is obtained as the "Schiff base" formation by condensation of 1,2-aminophenol and salicylaldehyde. The related molybdenum complex is obtained starting from one precursor, [MoO2(acac)2] (acac = acetylacetonate), reacting in a stoichiometric amount with H2L ligand. [MoO2(SAP)(EtOH)] complex was previously described, and its catalytic activity toward epoxide was shown in organic media. In several articles from Sobczak and Ziółkowski [26,27], a good catalytic activity of the [MoO2(SAP)(EtOH)] was announced and placed the complex as a very promising catalyst. In the frame of a sustainable process, the reaction could be improved since the reaction was performed in the presence of a halogenated solvent (bromobenzene) and needed TBHP in decane as an oxidant, i.e., the presence of two organic solvents, one being halogenated.
The idea developed herein has been to test the reactivity of the "MoO2(SAP)" complex and derivatives as catalysts in a more sustainable way. Thus, catalytic experiments were done in absence of any organic solvent, using TBHP in water as an oxidant and no organic solvent addition than the model substrate itself within the reaction media, i.e., cyclooctene ( Figure 1). Water (from TBHP) was not a solvent in the process since it was observed that a biphasic system occurred with the presence of the catalyst within the organic phase (phase containing the substrate itself). As previously mentioned, the catalytic activity of those species was interesting and a beginning of mechanism postulation proposed an equilibrium between [MoO2L]2 and [MoO2L(ROH)] within the reaction media, being the starting base of the postulated mechanism with a potential pentaco-ordinate species [MoO2(L)] calculated by DFT and explained later.

Preliminary Studies toward the Best Backbone and Mechanistic Study
Two different ligands were synthesized using salicylaldehyde and 1,2 aminoethanol (leading to H 2 SAE ligand) or 1,2 aminophenol (leading to H 2 SAP ligand), Figure 2 [28]. Molecular structures of [MoO 2 (SAP)(EtOH)] and a polymorphic form of [MoO 2 (SAE)(SAEH 2 )] [29] were determined by X-ray diffraction. The "MoO 2 L" species have been studied in DFT according to the previously mentioned mechanism [26,27] and it was shown that both [MoO 2 (L)] 2 (L = SAP, SAE) complexes needed less energy to be converted into the pentaco-ordinate mononuclear complex [MoO 2 L] than their ethanolstabilized congeners [MoO 2 L(EtOH)] [28]. The catalytic activity of [MoO2L]2 and [MoO2L(MeOH)] complexes (L = SAE, SAMP, SAP) was studied through cyclo-octene epoxidation [30]. The nature of the ligand helped the system to be more active. Indeed, "MoO2(SAP)" based complexes were stable, but the SAE and SAMP ligands from complexes were hydrolyzed during the catalytic process. An induction period observed with the [MoO2L(MeOH)] species and not [MoO2L]2 confirmed that more energy was needed to deco-ordinate the monomer stabilized by a solvent than the dimer into the pentacoordinate active species, confirmed by the DFT calculations. A direct relationship could be established between the catalyst stability and selectivity. From the experimental work with those complexes, one DFT-calculated pathway fitted to experimental observations, showing a new type of TBHP activation, corresponding to the Bartlett postulation done in the Prizhaev reaction, i.e., when an olefin does react with a peroxyacid as m-CPBA [31]. The transition state (TS) corresponds to a loose coordination of TBHP to the Mo together with an H-bonding with an oxido moiety linked to the Mo center. From this, the oxygen atom from the TBHP linked to the hydrogen can be transferred to the olefin (as seen in Figure 3A). Although [MoO2(acac)2] is an active catalyst, the advantage of [MoO2(SAP)] derivatives lie in their strong stability.

Ligand Tuning and Optimization
From this mechanism, optimization of the process was pursued using a slight modification of the ligand. The first change has been on the coordination sphere, comparing ONO with the ONS coordination sphere [32]. The [MoO2(SATP)] complex (ONS coordination sphere) was more active than [MoO2(SAP)] one (ONO coordination sphere) and activity was confirmed with DFT calculations (Figure 4a). The catalytic activity of [MoO 2 L] 2 and [MoO 2 L(MeOH)] complexes (L = SAE, SAMP, SAP) was studied through cyclo-octene epoxidation [30]. The nature of the ligand helped the system to be more active. Indeed, "MoO 2 (SAP)" based complexes were stable, but the SAE and SAMP ligands from complexes were hydrolyzed during the catalytic process. An induction period observed with the [MoO 2 L(MeOH)] species and not [MoO 2 L] 2 confirmed that more energy was needed to deco-ordinate the monomer stabilized by a solvent than the dimer into the pentacoordinate active species, confirmed by the DFT calculations. A direct relationship could be established between the catalyst stability and selectivity. From the experimental work with those complexes, one DFT-calculated pathway fitted to experimental observations, showing a new type of TBHP activation, corresponding to the Bartlett postulation done in the Prizhaev reaction, i.e., when an olefin does react with a peroxyacid as m-CPBA [31]. The transition state (TS) corresponds to a loose coordination of TBHP to the Mo together with an H-bonding with an oxido moiety linked to the Mo center. From this, the oxygen atom from the TBHP linked to the hydrogen can be transferred to the olefin (as seen in Figure 3A). Although [MoO 2 (acac) 2 ] is an active catalyst, the advantage of [MoO 2 (SAP)] derivatives lie in their strong stability. The catalytic activity of [MoO2L]2 and [MoO2L(MeOH)] complexes (L = SAE, SAMP, SAP) was studied through cyclo-octene epoxidation [30]. The nature of the ligand helped the system to be more active. Indeed, "MoO2(SAP)" based complexes were stable, but the SAE and SAMP ligands from complexes were hydrolyzed during the catalytic process. An induction period observed with the [MoO2L(MeOH)] species and not [MoO2L]2 confirmed that more energy was needed to deco-ordinate the monomer stabilized by a solvent than the dimer into the pentacoordinate active species, confirmed by the DFT calculations. A direct relationship could be established between the catalyst stability and selectivity. From the experimental work with those complexes, one DFT-calculated pathway fitted to experimental observations, showing a new type of TBHP activation, corresponding to the Bartlett postulation done in the Prizhaev reaction, i.e., when an olefin does react with a peroxyacid as m-CPBA [31]. The transition state (TS) corresponds to a loose coordination of TBHP to the Mo together with an H-bonding with an oxido moiety linked to the Mo center. From this, the oxygen atom from the TBHP linked to the hydrogen can be transferred to the olefin (as seen in Figure 3A). Although [MoO2(acac)2] is an active catalyst, the advantage of [MoO2(SAP)] derivatives lie in their strong stability.

Ligand Tuning and Optimization
From this mechanism, optimization of the process was pursued using a slight modification of the ligand. The first change has been on the coordination sphere, comparing ONO with the ONS coordination sphere [32]. The [MoO2(SATP)] complex (ONS coordination sphere) was more active than [MoO2(SAP)] one (ONO coordination sphere) and activity was confirmed with DFT calculations (Figure 4a).

Ligand Tuning and Optimization
From this mechanism, optimization of the process was pursued using a slight modification of the ligand. The first change has been on the coordination sphere, comparing ONO with the ONS coordination sphere [32]. The [MoO 2 (SATP)] complex (ONS coordination sphere) was more active than [MoO 2 (SAP)] one (ONO coordination sphere) and activity was confirmed with DFT calculations (Figure 4a).   Table 1).   Table 1).
The modifications of the pending organic functions surrounding the H 2 SAP ligand (second co-ordination sphere) were performed to evaluate their effects (according to nature and/or position) on catalytic properties (Figure 4b). Additional DFT calculations could help to refine the proposed theoretical model. Thus, different substituents have been added and activity was compared toward cyclo-octene, as shown in Table 1.
The addition of functional groups (R 2 = diethylamino and/or R 5 = nitro groups known for push-pull effects and modifications of NLO properties) at specific positions on the SAP backbone led to the comparison of four different molecules based on the presence or absence of both substituents [33]. The electronic effect of substituents on ligand might influence the electronic density of the molybdenum atom and subsequently the catalytic  Figure 3) value was observed for highly active processes and vice-versa.
The presence of the OH group on the salicylic part of the ligand was also interesting to study. Thus, the effect of the presence of one pending OH function in ortho (R 1 ), meta (R 2 ), or para (R 3 ) position to the phenolic oxygen atom linked to Mo on catalytic activity has been studied with cyclo-octene as a model substrate and compared to [MoO 2 (SAP)] moiety. The OH position strongly influences the activity toward cyclooctene with activity vs. [MoO 2 (SAP)] very high, slightly higher, and identical when OH lie in ortho, para, and meta positions, respectively [34]. The behavior was confirmed through DFT calculations with similar trends.
Other groups (OMe on ortho position (R 1 ) as for the OH and Me on the aminoalcohol part of the ligand (R 4 and R 5 )) have been added on SAP ligand and the corresponding [MoO 2 L(MeOH)] species (obtained through mechanochemistry and a solventless protocol, LAG) have been tested on different substrates [35]. With cyclo-octene, results showed that the presence of the methoxy group was the main factor toward a better activity. The methyl substituents in R 4 and R 5 positions did not have a noticeable effect.

ONS, ONO Mo-Enlarging the Scope of Complexes through Collaboration
From the knowledge of [MoO 2 L] species to act as catalysts under organic solvent-free conditions, a collaboration has been established with Croatia in 2010. All catalytic results have been separated according to the nature of the ligands around the molybdenum. Carbazones, hydrazones, and thiosemicarbazones ligands have been developed and presented according to the nature of the hydrazide used. For all the species with such geometries, we emphasized the TON (Turn Over Number) and TOF (Turn Over Frequency) parameters of the catalytic processes. TON is defined as the number of substrates transformed per unit of catalyst at the end of the studied time, i.e., the number of cycles per catalyst. TOF corresponds to the number of substrates transformed per unit of catalyst in time intervals all along a reaction. This shows how the reaction can start fast or not, according to the catalyst.
Pyridoxal Fragment within the Ligand ONS coordination sphere-thiosemicarbazones The synthesis of the thiosemicarbazonato dioxomolybdenum complexes based on pyridoxal ligand has been developed in Zagreb [36]. Pyridoxal possesses a free CH 2 OH pending function on a pyridine ring. On the other side of the ligand, thiosemicarbazide precursors possess different possible R terminal groups (R = H, Me, Ph). ( Figure 5) Depending on the pyridoxal protonation; the formed molybdenum species can be neutral "[MoO 2 L]" or charged "[MoO 2 LH]Cl". According to the presence (resp. absence) of donor molecules (for instance methanol), species can be stabilized as the monomer (resp. polymer) with polymerization mode influenced by the nature of R.  Those species have been tested as catalysts and results ( Table 2) indicated nature of R, catalytic ratio, charge (neutral/anionic) of the complex, and monomer meric form are all factors influencing the catalytic results. The most efficient catalyst under organic solvent-free conditions is [MoO2L(M i.e., the neutral complex (with R = Ph) stabilized by a molecule of methanol, whi sponding polymer [MoO2L]n (stabilized through the CH2OH pending function of a MoO2L complex) is the less active. The studies showed that the addition of MeOH quantity inhibited the activity of the catalyst at the same level as the correspondi mer itself. This study exhibited the role of methanol and its strong donation effect equilibria within the process are in favor of the previously postulated mechanism i the active species is the pentaco-ordinate complex [MoO2L] [37].
The charged complexes could be monomers or polymers as observed for the charges. The activity was lower than the neutral charges but the lower activit charged species corresponding to the most active neutral compound was mo nounced with monomers than for polymers [38].
ONO coordination sphere-hydrazones Other pyridoxal containing Mo complexes have been studied, using ligands a hydrazonato function and a ONO coordination sphere ( Figure 6). The differenc the X, being phenyl (X = CH), pyridine (X = N) or a phenol group (X = C-OH). A ONS, monomeric [MoO2L(MeOH)] or polymeric [MoO2L]n species were obtaine Those species have been tested as catalysts and results (Table 2) indicated that the nature of R, catalytic ratio, charge (neutral/anionic) of the complex, and monomeric/polymeric form are all factors influencing the catalytic results. The most efficient catalyst under organic solvent-free conditions is [MoO 2 L(MeOH)], i.e., the neutral complex (with R = Ph) stabilized by a molecule of methanol, while corresponding polymer [MoO 2 L] n (stabilized through the CH 2 OH pending function of a second MoO 2 L complex) is the less active. The studies showed that the addition of MeOH in high quantity inhibited the activity of the catalyst at the same level as the corresponding polymer itself. This study exhibited the role of methanol and its strong donation effect. All the equilibria within the process are in favor of the previously postulated mechanism in which the active species is the pentaco-ordinate complex [MoO 2 L] [37].
The charged complexes could be monomers or polymers as observed for the neutralcharges. The activity was lower than the neutral charges but the lower activity of the charged species corresponding to the most active neutral compound was more pronounced with monomers than for polymers [38].
ONO coordination sphere-hydrazones Other pyridoxal containing Mo complexes have been studied, using ligands bearing a hydrazonato function and a ONO coordination sphere ( Figure 6). The difference lies in the X, being phenyl (X = CH), pyridine (X = N) or a phenol group (X = C-OH). As for the ONS, monomeric [MoO 2 L(MeOH)] or polymeric [MoO 2 L] n species were obtained. Interestingly, mixed hybrid compounds could be obtained, formally composed by two cationic coordination complex moieties [MoO 2 (LH)] + and one Lindqvist polyanion [Mo 6 O 19 ] 2− [39]. Those species ( Figure 6) were tested as catalysts under the same experimental conditions ( Table 3). Conversion of cyclooctene depends on the nature of the species, i.e., monomer, polymer, or hybrid. Considering monomers vs. polymers, the monomeric complex [MoO 2 L(MeOH)]with the highest conversion after 6 h (X = C-H)-has the lowest in a polymeric state. As for the ONS species, the monomeric species bear higher TOF, i.e., faster activation. The species containing POMs are more active with a higher conversion (they contain two potential active substances) but the higher selectivity is due to the [MoO 2 L] species. [MoO2L(MeOH)]-with the highest conversion after 6 h (X = C-H)-has the lowest in a polymeric state. As for the ONS species, the monomeric species bear higher TOF, i.e., faster activation. The species containing POMs are more active with a higher conversion (they contain two potential active substances) but the higher selectivity is due to the [MoO2L] species.

From Mo to W Complexes with Salicylaldehyde Part: Some Experimental Features
Replacing the pyridoxal moiety by differently substituted salicylaldehydes led to monomeric [MO2L(EtOH)] (M = Mo, W) and polymeric [WO2L]n species [40] (Figure 7). The catalytic activity of molybdenum complexes was quite high (Table 4), with conversion up to 90% and selectivity quite low compared to the complexes containing pyridoxal moieties. Tungsten-related complexes exhibited very low activity under the same experimental conditions. The oxidant TBHP in water did not seem to be the right oxidant to achieve good conversion and good selectivity.

From Mo to W Complexes with Salicylaldehyde Part: Some Experimental Features
Replacing the pyridoxal moiety by differently substituted salicylaldehydes led to monomeric [MO 2 L(EtOH)] (M = Mo, W) and polymeric [WO 2 L] n species [40] (Figure 7). The catalytic activity of molybdenum complexes was quite high (Table 4), with conversion up to 90% and selectivity quite low compared to the complexes containing pyridoxal moieties. Tungsten-related complexes exhibited very low activity under the same experimental conditions. The oxidant TBHP in water did not seem to be the right oxidant to achieve good conversion and good selectivity.  The quest toward better experimental conditions was done with new ligands keeping the salicyladehyde moiety but adding isonicotinoyl hydrazines (X = C-OH), leading to [WO2L(MeOH)] and [WO2L(EtOH)] complexes [41].
Those species were tested under several experimental conditions. The reaction was faster using acetonitrile with TBHP in water as an oxidant but selectivity was worse than without acetonitrile. Reaction performed with TBHP in decane exhibited equivalent conversion than with TBHP in water but with higher selectivity, exhibiting the role of water. The effect of the coordinating alcohol was noticed, with a better conversion with EtOH, certainly due to a looser interaction with the W and a faster activation into a pentacoordinated species. In the case of nicotinoyl hydrazides and molybdenum complexes (Figure 8), it was shown that the oligomerization rate has an influence on the reactivity. Indeed, while all  The quest toward better experimental conditions was done with new ligands keeping the salicyladehyde moiety but adding isonicotinoyl hydrazines (X = C-OH), leading to [WO 2 L(MeOH)] and [WO 2 L(EtOH)] complexes [41].
Those species were tested under several experimental conditions. The reaction was faster using acetonitrile with TBHP in water as an oxidant but selectivity was worse than without acetonitrile. Reaction performed with TBHP in decane exhibited equivalent conversion than with TBHP in water but with higher selectivity, exhibiting the role of water. The effect of the coordinating alcohol was noticed, with a better conversion with EtOH, certainly due to a looser interaction with the W and a faster activation into a pentacoordinated species.
In the case of nicotinoyl hydrazides and molybdenum complexes (Figure 8), it was shown that the oligomerization rate has an influence on the reactivity. Indeed, while all compounds existed in polymeric form (with high nuclearity) specific experimental conditions could lead to the isolation of tetranuclear scaffolds that appeared to be highly active, more than the corresponding polymer. (Table 5) It exhibited that the mechanism suggesting the formation of pentacoordinate species was relevant and easier with oligomers than for polymers [42]. In this case, it was also shown that TBHP in decane was more efficient [43].
compounds existed in polymeric form (with high nuclearity) specific experimental conditions could lead to the isolation of tetranuclear scaffolds that appeared to be highly active, more than the corresponding polymer. (Table 5) It exhibited that the mechanism suggesting the formation of pentacoordinate species was relevant and easier with oligomers than for polymers [42]. In this case, it was also shown that TBHP in decane was more efficient [43].  Oligomerization was emphasized with compounds having 4-aminobenzhydrazone ligands ( Figure 9) [44]. Those species existed as monomer [MoO2L(ROH)] and dimer [MoO2L]2 due to the ligand bearing NH2 substituent coordinating the Mo of a neighboring molecule. The dimers reacted faster than the monomers, another proof of the postulated mechanism ( Table 6). The comparison with the complexes bearing 2-aminobenzydrazones showed that the position of the NH2 is important within the stabilization of the transition state since those species were even more active. This has also been assessed through DFT calculations [45].  Oligomerization was emphasized with compounds having 4-aminobenzhydrazone ligands ( Figure 9) [44]. Those species existed as monomer [MoO 2 L(ROH)] and dimer [MoO 2 L] 2 due to the ligand bearing NH 2 substituent coordinating the Mo of a neighboring molecule. The dimers reacted faster than the monomers, another proof of the postulated mechanism ( Table 6). The comparison with the complexes bearing 2-aminobenzydrazones showed that the position of the NH 2 is important within the stabilization of the transition state since those species were even more active. This has also been assessed through DFT calculations [45].
As expected, reactivity in TBHP in decane was faster. Different induction periods were observed between monomers and polymers and the substitution on the benzaldehydic part of the ligand seemed to have an influence.
Replacing -NH 2 by -OH led to a series of monomeric and oligomeric MoO 2 L complexes ( Figure 10) tested as catalysts using TBHP as an oxidant in water (W) or in decane (D) and no other added solvent [46]. Results are compiled in Table 7. Very good results are obtained with such structures, exhibiting the role of OH on both aromatic rings.  As expected, reactivity in TBHP in decane was faster. Different induction periods were observed between monomers and polymers and the substitution on the benzaldehydic part of the ligand seemed to have an influence.
Replacing -NH2 by -OH led to a series of monomeric and oligomeric MoO2L complexes ( Figure 10) tested as catalysts using TBHP as an oxidant in water (W) or in decane (D) and no other added solvent [46]. Results are compiled in Table 7. Very good results are obtained with such structures, exhibiting the role of OH on both aromatic rings.      Very recently, another variation was done, replacing the MeO on the benzaldehydic part with OH on the ortho and para position and having the 2-or 4-NH 2 benzydrazide ( Figure 11) [47].   With those compounds, catalysts tests with cyclo-octene have been performed using three types of oxidants (Table 8), TBHP in water (W), TBHP in decane (D), and H2O2 in water ( Table 8). As seen previously seen, TBHP in decane as oxidant gives the best results but TBHP in water gives relatively good results and is greener. With H2O2, the protocol With those compounds, catalysts tests with cyclo-octene have been performed using three types of oxidants (Table 8), TBHP in water (W), TBHP in decane (D), and H 2 O 2 in water ( Table 8). As seen previously seen, TBHP in decane as oxidant gives the best results but TBHP in water gives relatively good results and is greener. With H 2 O 2 , the protocol was tested herein for the first time and the reaction was shown to be quite slow. A mechanistic study with H 2 O 2 as oxidant showed that the less energetic mechanism is similar to the mechanism with TBHP. Cyclohexene (CH) has been studied because the corresponding epoxide (CHO) can be readily opened in the presence of water and leads to the trans-cyclohexanediol (CHD). (Figure 12) Further steps can lead to the complete opening of the 6-membered ring toward a very valuable species, adipic acid. The activity of the catalyst was also assessed. With a very active catalyst, the ring opening could be quickly observed. It explained why a less active catalyst did not give the diol in big quantities. This was interesting and has been used for the other presented works with applicative purposes. was tested herein for the first time and the reaction was shown to be quite slow. A mechanistic study with H2O2 as oxidant showed that the less energetic mechanism is similar to the mechanism with TBHP.

Other Olefins
Cyclohexene (CH) has been studied because the corresponding epoxide (CHO) can be readily opened in the presence of water and leads to the trans-cyclohexanediol (CHD). (Figure 12) Further steps can lead to the complete opening of the 6-membered ring toward a very valuable species, adipic acid. The activity of the catalyst was also assessed. With a very active catalyst, the ring opening could be quickly observed. It explained why a less active catalyst did not give the diol in big quantities. This was interesting and has been used for the other presented works with applicative purposes. Figure 12. Organic solvent-free cyclohexene epoxidation and subsequent ring-opening with water. Figure 12. Organic solvent-free cyclohexene epoxidation and subsequent ring-opening with water.

Application to Biomass Substrates
The success of the process with simple liquid cyclic alkenes was also the basis of extra development toward the valorization of biomass with the epoxidation of a sesquiterpenes [48] and lignans [49], showing that the [MoO 2 (SAP)] coupled with TBHP with/without organic solvent-free conditions could replace the m-CPBA method. The species extracted from natural sources are interesting sources of chemicals [50].

Himachalenes
Himachalenes are sesquiterpenes obtained from the essential oil of an endemic Moroccan Tree, Cedrus atlantica [51]. Those species are relatively abundant and cheap interesting substrates. Those epoxidations exist with classical organic methods, i.e., m-CPBA oxidant in CH 2 Cl 2 , a solvent to avoid in industrial or academic processes. m-CPBA is known to be efficient but post-treatment is tedious, time-consuming, and contains one step with a basic phase that could be avoided by the use of a more atom-economic oxidant. Thus, in collaboration with a Moroccan research group expert within himachalene chemistry, the epoxidation reaction of three different substrate types was performed. [48]The starting mixture of himachalenes contains the different positions of double bonds present on the seven-membered ring, leading to a mixture of three isomers ( Figure 13). It was possible to see that the regioselectivity of epoxidation with MoO 2 (SAP)/TBHP is identical to that using m-CPBA, privileging epoxidation on the internal bonds of the 7-member ring (β isomer) vs. α.

Application to Biomass Substrates
The success of the process with simple liquid cyclic alkenes was also the basis of extra development toward the valorization of biomass with the epoxidation of a sesquiterpenes [48] and lignans [49], showing that the [MoO2(SAP)] coupled with TBHP with/without organic solvent-free conditions could replace the m-CPBA method. The species extracted from natural sources are interesting sources of chemicals [50].

Himachalenes
Himachalenes are sesquiterpenes obtained from the essential oil of an endemic Moroccan Tree, Cedrus atlantica [51]. Those species are relatively abundant and cheap interesting substrates. Those epoxidations exist with classical organic methods, i.e., m-CPBA oxidant in CH2Cl2, a solvent to avoid in industrial or academic processes. m-CPBA is known to be efficient but post-treatment is tedious, time-consuming, and contains one step with a basic phase that could be avoided by the use of a more atom-economic oxidant. Thus, in collaboration with a Moroccan research group expert within himachalene chemistry, the epoxidation reaction of three different substrate types was performed. [48]The starting mixture of himachalenes contains the different positions of double bonds present on the seven-membered ring, leading to a mixture of three isomers ( Figure 13). It was possible to see that the regioselectivity of epoxidation with MoO2(SAP)/TBHP is identical to that using m-CPBA, privileging epoxidation on the internal bonds of the 7-member ring (β isomer) vs. α. After suppressing one internal double bond through different pathways on the α(β) isomer, only one internal (external) double bond is present to study the diastereoselectivity of the approach and it was here seen that epoxidation could be achieved with stereoselectivity slightly different than the m-CPBA method ( Figure 14).  After suppressing one internal double bond through different pathways on the α(β) isomer, only one internal (external) double bond is present to study the diastereoselectivity of the approach and it was here seen that epoxidation could be achieved with stereoselectivity slightly different than the m-CPBA method ( Figure 14).

Application to Biomass Substrates
The success of the process with simple liquid cyclic alkenes was also the basis of extra development toward the valorization of biomass with the epoxidation of a sesquiterpenes [48] and lignans [49], showing that the [MoO2(SAP)] coupled with TBHP with/without organic solvent-free conditions could replace the m-CPBA method. The species extracted from natural sources are interesting sources of chemicals [50].

Himachalenes
Himachalenes are sesquiterpenes obtained from the essential oil of an endemic Moroccan Tree, Cedrus atlantica [51]. Those species are relatively abundant and cheap interesting substrates. Those epoxidations exist with classical organic methods, i.e., m-CPBA oxidant in CH2Cl2, a solvent to avoid in industrial or academic processes. m-CPBA is known to be efficient but post-treatment is tedious, time-consuming, and contains one step with a basic phase that could be avoided by the use of a more atom-economic oxidant. Thus, in collaboration with a Moroccan research group expert within himachalene chemistry, the epoxidation reaction of three different substrate types was performed. [48]The starting mixture of himachalenes contains the different positions of double bonds present on the seven-membered ring, leading to a mixture of three isomers ( Figure 13). It was possible to see that the regioselectivity of epoxidation with MoO2(SAP)/TBHP is identical to that using m-CPBA, privileging epoxidation on the internal bonds of the 7-member ring (β isomer) vs. α. After suppressing one internal double bond through different pathways on the α(β) isomer, only one internal (external) double bond is present to study the diastereoselectivity of the approach and it was here seen that epoxidation could be achieved with stereoselectivity slightly different than the m-CPBA method ( Figure 14).

Lignans
The MoO 2 (SAP)/TBHP catalytic system was tested on lignans [49]. The specific lignans studied-extracted from the knotwood of Norway Spruce (Picea Abies)constitute interesting renewable biphenolic material studied in collaboration with a research group from Abo Akademi in Finland [52,53]. Those compounds, derived from imperanene and imperaneic acid, possess a very interesting double bond between two aromatics. Their oxidation products-through the formation of non-isolable epoxide followed by acid-assisted epoxide ring opening and rearrangements-can lead to different heterocyclic species, such as tetrahydrofuranes, lactones or tetralin structures (Figure 15). The [MoO 2 (SAP)]/TBHP method needs the use of an organic solvent since the substrate was solid. At the difference with the sesquiterpenes, the complexity of the lignans led to oligomeric side products that diminished the yields.

Lignans
The MoO2(SAP)/TBHP catalytic system was tested on lignans [49]. The specific lignans studied-extracted from the knotwood of Norway Spruce (Picea Abies)-constitute interesting renewable biphenolic material studied in collaboration with a research group from Abo Akademi in Finland [52,53]. Those compounds, derived from imperanene and imperaneic acid, possess a very interesting double bond between two aromatics. Their oxidation products-through the formation of non-isolable epoxide followed by acid-assisted epoxide ring opening and rearrangements-can lead to different heterocyclic species, such as tetrahydrofuranes, lactones or tetralin structures (Figure 15). The [MoO2(SAP)]/TBHP method needs the use of an organic solvent since the substrate was solid. At the difference with the sesquiterpenes, the complexity of the lignans led to oligomeric side products that diminished the yields.

Limonene
Another interesting part concerns a very simple substrate, limonene. [54] A byproduct of the orange juice industry, this terpene is cheap and relatively abundant. Used in the food and perfume industries, it can be at the base of building blocks for pharmaceutical compounds. The epoxidation of the limonene (Figure 16) is preferential on the inner double bond and can create two stereoisomers, cis-and trans-limonene epoxides (LimOs). Those epoxides are useful in different applications, from a molecular point of view, as precursors of several biobased polymers. Both LimOs are relatively stable but the presence of water can open both in two different diols, named here equatorial (eq) and axial (ax) limonene diols (LimDs). The reactivity favors the stable ax-LimD. With the use of MoO2L complexes, it has been shown that organic solvent-free conditions led to the formation of both LimOs without difference in selectivity. The interesting point lies in the ring opening of both limOs that showed in major form the ax-LimD (exclusively this species starting from cis-LimO) but also the formation of the unfavored eq-LimD without real caution within the experimental conditions, which is different from all other described protocols for this specific LimD. Indeed, to compare, the existing methods to produce the "unusual" equ-LimD requires first the isolation of the trans-LimO from a cis/trans LimO mixture and the use of mercuric reagent [55] or enzymatic protocols [56] under buffered conditions. An interesting comparison between [MoO2(SAP)] and [MoO2(SATP)] exhibited that the ONS coordination sphere around the molybdenum favored the eq-LimD generation.

Limonene
Another interesting part concerns a very simple substrate, limonene. [54] A byproduct of the orange juice industry, this terpene is cheap and relatively abundant. Used in the food and perfume industries, it can be at the base of building blocks for pharmaceutical compounds. The epoxidation of the limonene (Figure 16) is preferential on the inner double bond and can create two stereoisomers, cisand translimonene epoxides (LimOs). Those epoxides are useful in different applications, from a molecular point of view, as precursors of several biobased polymers. Both LimOs are relatively stable but the presence of water can open both in two different diols, named here equatorial (eq) and axial (ax) limonene diols (LimDs). The reactivity favors the stable ax-LimD. With the use of MoO 2 L complexes, it has been shown that organic solvent-free conditions led to the formation of both LimOs without difference in selectivity. The interesting point lies in the ring opening of both limOs that showed in major form the ax-LimD (exclusively this species starting from cis-LimO) but also the formation of the unfavored eq-LimD without real caution within the experimental conditions, which is different from all other described protocols for this specific LimD. Indeed, to compare, the existing methods to produce the "unusual" equ-LimD requires first the isolation of the trans-LimO from a cis/trans LimO mixture and the use of mercuric reagent [55] or enzymatic protocols [56]

Carveol Oxidation
With Mo complexes ( Figure 17) similar to those used in Figure 8, oxidation of several alcohols was done ( Figure 18). One alcohol of biomass origin, carveol, is the most interesting. The oxidants tested have been H2O2 and TBHP (water or decane) [57]. Results compiled in Table 9 showed different phenomena. H2O2 was the best oxidant to transform into carvone (40% selectivity) vs. 4-20% with TBHP. TBHP in decane reacted faster but certainly to the corresponding epoxide. With TBHP, the presence of water seems detrimental for alcohol oxidation. The main factor influencing the activity is the OH on the ligand in position R1 that seems to more quickly activate the catalyst. Some further kinetic studies showed that cis-carveol is preferentially transformed with H2O2 and trans-carveol with TBHP.

Carveol Oxidation
With Mo complexes ( Figure 17) similar to those used in Figure 8, oxidation of several alcohols was done ( Figure 18). One alcohol of biomass origin, carveol, is the most interesting. The oxidants tested have been H 2 O 2 and TBHP (water or decane) [57]. Results compiled in Table 9 showed different phenomena. H 2 O 2 was the best oxidant to transform into carvone (40% selectivity) vs. 4-20% with TBHP. TBHP in decane reacted faster but certainly to the corresponding epoxide. With TBHP, the presence of water seems detrimental for alcohol oxidation. The main factor influencing the activity is the OH on the ligand in position R 1 that seems to more quickly activate the catalyst. Some further kinetic studies showed that cis-carveol is preferentially transformed with H 2 O 2 and trans-carveol with TBHP.

Carveol Oxidation
With Mo complexes ( Figure 17) similar to those used in Figure 8, oxidation of several alcohols was done ( Figure 18). One alcohol of biomass origin, carveol, is the most interesting. The oxidants tested have been H2O2 and TBHP (water or decane) [57]. Results compiled in Table 9 showed different phenomena. H2O2 was the best oxidant to transform into carvone (40% selectivity) vs. 4-20% with TBHP. TBHP in decane reacted faster but certainly to the corresponding epoxide. With TBHP, the presence of water seems detrimental for alcohol oxidation. The main factor influencing the activity is the OH on the ligand in position R1 that seems to more quickly activate the catalyst. Some further kinetic studies showed that cis-carveol is preferentially transformed with H2O2 and trans-carveol with TBHP.

Vanadium Species
As for molybdenum, vanadium is an element used in oxidation processes. In addition to being present in nature in some enzymatic processes, vanadium can activate smoother and cleaner oxidants, such as TBHP or H2O2 and the most convenient oxidant, O2. Several vanadium-containing compounds have been shown to be active in catalysis, for example the species used by Mimoun [58], Rehder [59], Maurya [60], or Hartung [61]. The mechanisms are numerous according to the nature of the ligands.

Vanadium Species
As for molybdenum, vanadium is an element used in oxidation processes. In addition to being present in nature in some enzymatic processes, vanadium can activate smoother and cleaner oxidants, such as TBHP or H 2 O 2 and the most convenient oxidant, O 2 . Several vanadium-containing compounds have been shown to be active in catalysis, for example the species used by Mimoun [58], Rehder [59], Maurya [60], or Hartung [61]. The mechanisms are numerous according to the nature of the ligands.

SAP
It was interesting to take advantage of the H 2 L ligands (H 2 SAE, H 2 SAMP, and H 2 SAP) presented in the molybdenum section and to study the activity of equivalent vanadium species. Through reaction with [VO(acac) 2 ] as a vanadium precursor and subsequent oxidation, the dinuclear complexes [(L)VO] 2 O complexes have been isolated ( Figure 19) and the structure of the compound (L = SAE, SAMP) was determined through X-ray crystallography. The catalytic activity of 1 mol% complex vs. substrate was tested with [VO(SAP)] 2 O as a catalyst, with both TBHP or H 2 O 2 as oxidant, and for the first time under organic solvent-free conditions. The epoxidation of cyclo-octene gave a 94% conversion (and 83% selectivity towards the epoxide) after 5.5 h with TBHP and no reaction with H 2 O 2 [62].

Vanadium Species
As for molybdenum, vanadium is an element used in oxidation processes. In addition to being present in nature in some enzymatic processes, vanadium can activate smoother and cleaner oxidants, such as TBHP or H2O2 and the most convenient oxidant, O2. Several vanadium-containing compounds have been shown to be active in catalysis, for example the species used by Mimoun [58], Rehder [59], Maurya [60], or Hartung [61]. The mechanisms are numerous according to the nature of the ligands.

SAP
It was interesting to take advantage of the H2L ligands (H2SAE, H2SAMP, and H2SAP) presented in the molybdenum section and to study the activity of equivalent vanadium species. Through reaction with [VO(acac)2] as a vanadium precursor and subsequent oxidation, the dinuclear complexes [(L)VO]2O complexes have been isolated ( Figure 19) and the structure of the compound (L = SAE, SAMP) was determined through X-ray crystallography. The catalytic activity of 1 mol% complex vs. substrate was tested with [VO(SAP)]2O as a catalyst, with both TBHP or H2O2 as oxidant, and for the first time under organic solvent-free conditions. The epoxidation of cyclo-octene gave a 94% conversion (and 83% selectivity towards the epoxide) after 5.5 h with TBHP and no reaction with H2O2 [62].

ONO, ONS, and Mechanism
Based on the pyridoxal moiety used for the Mo complexes mentioned above, different families of vanadium complexes were synthesized with an ONS coordination sphere and containing one vanadium atom with general formulas [VO2(LH)] and an ONO coordination sphere around the vanadium creating neutral [V2O3L2] or charged [V2O3(HL)]Cl2 molecules containing two vanadium atoms (Figure 20) [63]. Those species have been tested as catalysts for the epoxidation of cyclo-octene under organic solvent-free conditions (Table 10).

ONO, ONS, and Mechanism
Based on the pyridoxal moiety used for the Mo complexes mentioned above, different families of vanadium complexes were synthesized with an ONS coordination sphere and containing one vanadium atom with general formulas [VO 2 (LH)] and an ONO coordination sphere around the vanadium creating neutral [V 2 O 3 L 2 ] or charged [V 2 O 3 (HL)]Cl 2 molecules containing two vanadium atoms ( Figure 20) [63]. Those species have been tested as catalysts for the epoxidation of cyclo-octene under organic solvent-free conditions (Table 10). With TBHP in water as oxidant. Results were moderate but it was interesting to try to elucidate the mechanism through DFT. Thus, from the complexes of general formulas [VO 2 (LH)] (X = C-OH), calculations showed that the most energetically favorable pathway went through the formation of hydroxido-alkylperoxido [VO(OH)(OOMe)(HL)] ( Figure 21) [63].

Keggin-Type Polyoxometalates as (ep)Oxidation Catalyst
Keggin Polyoxometalates (POMs) is the second class of catalysts studied using the sustainable methods presented herein. Known for a very long time for fundamental research but also for their applications in biology and in catalysis, in both homogeneous and heterogeneous conditions, POMs have the advantage of being an extremely stable species, very simple to be synthesized (mostly in solution methods but recently using solvent-free methods using mechanochemical activation). Those species can very easily activate  With TBHP in water as oxidant. Results were moderate but it was interesting to try to elucidate the mechanism through DFT. Thus, from the complexes of general formulas [VO2(LH)] (X = C-OH), calculations showed that the most energetically favorable pathway went through the formation of hydroxido-alkylperoxido [VO(OH)(OOMe)(HL)] ( Figure 21) [63].

Keggin-Type Polyoxometalates as (ep)Oxidation Catalyst
Keggin Polyoxometalates (POMs) is the second class of catalysts studied using the sustainable methods presented herein. Known for a very long time for fundamental research but also for their applications in biology and in catalysis, in both homogeneous and heterogeneous conditions, POMs have the advantage of being an extremely stable species,

Keggin-Type Polyoxometalates as (ep)Oxidation Catalyst
Keggin Polyoxometalates (POMs) is the second class of catalysts studied using the sustainable methods presented herein. Known for a very long time for fundamental research but also for their applications in biology and in catalysis, in both homogeneous and heterogeneous conditions, POMs have the advantage of being an extremely stable species, very simple to be synthesized (mostly in solution methods but recently using solvent-free methods using mechanochemical activation). Those species can very easily activate smooth oxidants (H 2 O 2 , TBHP, O 2 , UHP) for several oxidation reactions. Among the active species relative to POMs; the classical Venturello-Ishii catalyst, a peroxooxomolybdenum complex, has been deeply studied and Keggin species were more explored for the sulfoxidation reaction.

Pyridinium Salts
Very simple [PMo 12 O 40 ] 3− and [PW 12 O 40 ] 3− Keggin type heteropolyanions have been tested for catalysts as organic salts using tetrabutylammonium (TBA), butyl-(BP) or cetyl-pyridinium (CP) as cation, in order to use the catalyst in the organic medium, i.e., the substrate itself [64]. The reactions were done without organic solvent and oxidants (H 2 O 2 or TBHP) in aqueous solution ( Figure 22). The results (Table 11) exhibited activity depending on all parameters, i.e., PMo 12 vs. PW 12 , nature of cation and nature of oxidant. With 0.1% POM vs. cyclo-octene, at 80 • C, TBHP gave a better selectivity toward epoxide without formation of the cyclo-octanediol. This latter compound was observed when H 2 O 2 was used as oxidant, this certainly explained the low selectivity. The catalysts could be recycled and separated easily from the reaction mixture very easily at room temperature in the case of TBA and BP salts, the CP giving an emulsion that was hard to separate (although efficient). The reaction certainly takes place in the organic phase, and it was found that homogenous and heterogeneous reactions coexist (the solubility of the POMs being strongly cation dependent), explaining the difference within the selectivity. The alkyl pyridinium being the most soluble, a different test with a low catalyst charge was performed, exhibiting activity until 2 or 5 ppm POM vs. substrate ratio but a lower selectivity. Thus, it was supposed that a heterogeneous process gave better selectivity within the reaction media. In addition, it was also shown that [PMo 12   that homogenous and heterogeneous reactions coexist (the solubility of the POMs being strongly cation dependent), explaining the difference within the selectivity. The alkyl pyridinium being the most soluble, a different test with a low catalyst charge was performed, exhibiting activity until 2 or 5 ppm POM vs. substrate ratio but a lower selectivity. Thus, it was supposed that a heterogeneous process gave better selectivity within the reaction media. In addition, it was also shown that [PMo12O40] 3− species were more efficient with TBHP and [PW12O40] 3− with H2O2.

Supported Catalysts
Within the aim of a sustainable and recyclable process, it was interesting to further study the catalytic activity of the simple Keggin polyanions once ionically grafted on solid support. Two types of supports have rightly been studied: functionalized organic polymer and silica nanoparticles.

Grafted POMs on Merrifield Resins [65]
The protocol consisted in the functionalization of a commercial Merrifield resin by quaternization of alkyl imidazoles by the chloromethyl pending functions present on the polymer. From those functionalizations, [PM 12 O 40 ] 3− (M= Mo, W) were ionically grafted on those polymers. In order to compare the activity when grafted or as a free molecule, molecular analogs with same type of imidazolium countercations have been also synthesized. (Figure 23) The species were stable and could be characterized through several methods. It was possible to load more Mo Keggin than W Keggin on the Merrifield resins with a range of 55.6-66.7 µmol/g polymer for PMo 12

Supported Catalysts
Within the aim of a sustainable and recyclable process, it was interesting to further study the catalytic activity of the simple Keggin polyanions once ionically grafted on solid support. Two types of supports have rightly been studied: functionalized organic polymer and silica nanoparticles.

Grafted POMs on Merrifield Resins [65]
The protocol consisted in the functionalization of a commercial Merrifield resin by quaternization of alkyl imidazoles by the chloromethyl pending functions present on the polymer. From those functionalizations, [PM12O40] 3− (M= Mo, W) were ionically grafted on those polymers. In order to compare the activity when grafted or as a free molecule, molecular analogs with same type of imidazolium countercations have been also synthesized. (Figure 23) The species were stable and could be characterized through several methods. It was possible to load more Mo Keggin than W Keggin on the Merrifield resins with a range of 55.6-66.7 μmol/g polymer for PMo12O40 and 12.2-18.9 μmol/g polymer for PW12O40. The organic and the grafted salts of POMs have been tested for the epoxidation of cyclohexene, an interesting precursor for the synthesis of adipic acid (AA) (Figure 24). With four equivalents of oxidant starting from cyclohexene (Table 12), AA yields from 46-61% were obtained with molecular catalysts and 33-51% with grafted catalysts. The interesting fact lies in the low POM content in general (0.025% POM vs. substrate with molecular catalysts and within 0.001-0.007% range for the grafted catalysts). The study considered each step of the postulated mechanism.   The organic and the grafted salts of POMs have been tested for the epoxidation of cyclohexene, an interesting precursor for the synthesis of adipic acid (AA) (Figure 24). With four equivalents of oxidant starting from cyclohexene (Table 12), AA yields from 46-61% were obtained with molecular catalysts and 33-51% with grafted catalysts. The interesting fact lies in the low POM content in general (0.025% POM vs. substrate with molecular catalysts and within 0.001-0.007% range for the grafted catalysts). The study considered each step of the postulated mechanism.

Supported Catalysts
Within the aim of a sustainable and recyclable process, it was interesting to further study the catalytic activity of the simple Keggin polyanions once ionically grafted on solid support. Two types of supports have rightly been studied: functionalized organic polymer and silica nanoparticles.

Grafted POMs on Merrifield Resins [65]
The protocol consisted in the functionalization of a commercial Merrifield resin by quaternization of alkyl imidazoles by the chloromethyl pending functions present on the polymer. From those functionalizations, [PM12O40] 3− (M= Mo, W) were ionically grafted on those polymers. In order to compare the activity when grafted or as a free molecule, molecular analogs with same type of imidazolium countercations have been also synthesized. (Figure 23) The species were stable and could be characterized through several methods. It was possible to load more Mo Keggin than W Keggin on the Merrifield resins with a range of 55.6-66.7 μmol/g polymer for PMo12O40 and 12.2-18.9 μmol/g polymer for PW12O40. The organic and the grafted salts of POMs have been tested for the epoxidation of cyclohexene, an interesting precursor for the synthesis of adipic acid (AA) (Figure 24). With four equivalents of oxidant starting from cyclohexene (Table 12), AA yields from 46-61% were obtained with molecular catalysts and 33-51% with grafted catalysts. The interesting fact lies in the low POM content in general (0.025% POM vs. substrate with molecular catalysts and within 0.001-0.007% range for the grafted catalysts). The study considered each step of the postulated mechanism.   eropolyacids precursors and the reaction giving other products than CHD (maybe AA). With Mo, the CHD was more visible showing that the reaction was slower. With Lim, the reaction was very fast and mainly led to the formation of LimDs, as well as a few quantities of carveol (Col) and carvone (Cone). CyOH gave interesting information, exhibiting better activity for the heteropolyacids, certainly due to the inner acidity (compared to the grafted ones).
Those objects were even reused and showed recyclability after the third run. Figure 25. Schematic representation of POMs grafted on functionalized silica. Figure 25. Schematic representation of POMs grafted on functionalized silica.  Table 14. CH epoxidation with heteropolyacids and corresponding anionic species from Figure 25. Analyzed species are in Figure 26.        Table 16. CY ol oxidation toward CY one with heteropolyacids and corresponding anionic species grafted on silica schemed in Figure 22. Studied products are those from Figure 27.  Table 16. CY ol oxidation toward CY one with heteropolyacids and corresponding anionic species grafted on silica schemed in Figure 22. Studied products are those from Figure 27.

Conclusions
It has been shown here all the different directions taken in Castres, France toward sustainable processes. Ligand engineering (with some mechanistic DFT explanations) for the coordination complexes and catalysts grafting were the strategies employed to use a very low quantity of catalysts for different (ep)oxidation processes. All has been progressively oriented recently toward the valorization of biomass substrates, in order to situate this research in the context of circular economy. The advances in this research are still in progress.  CyOH gave interesting information, exhibiting better activity for the heteropolyacids, certainly due to the inner acidity (compared to the grafted ones).
Those objects were even reused and showed recyclability after the third run.

Conclusions
It has been shown here all the different directions taken in Castres, France toward sustainable processes. Ligand engineering (with some mechanistic DFT explanations) for the coordination complexes and catalysts grafting were the strategies employed to use a very low quantity of catalysts for different (ep)oxidation processes. All has been progressively oriented recently toward the valorization of biomass substrates, in order to situate this research in the context of circular economy. The advances in this research are still in progress.
Author Contributions: Conceptualization, D.A.; methodology, D.A. and J.P.; writing-original draft preparation, D.A.; writing-review and editing, D.A. and J.P. All authors have read and agreed to the published version of the manuscript.