Toward a New Way for the Valorization of Miscanthus Biomass Produced on Metal-Contaminated Soils Part 2: Miscanthus-Based Biosourced Catalyst: Design, Preparation, and Catalytic Efficiency in the Synthesis of Moclobemide

: The conception of two biosourced catalysts (biocatalysts) using stems of miscanthus from the first part of this study are described herein. The temperature and the process used to extract metals from plant as mixture of Lewis acids were investigated in detail and proved to be essential in the design of the biosourced catalysts and their catalytic efficiency. One part of the crude mixture of Lewis acids extracted from the aerial parts of miscanthus plants was used without further treat-ment as a homogeneous biocatalyst (M1), and the other part was supported on montmorillonite K10 to provide a heterogeneous biocatalyst (MM1). M1 and MM1 were next tested in the synthesis of moclobemide (main ingredient of a drug used to treat depression) and led to excellent yield. Additional comparative experiments with different commercial metallic salts (NaCl, KCl, CaCl 2 , MgCl 2 , CuCl 2 , ZnCl 2 , FeCl 2 , FeCl 3 , MnCl 2 , and AlCl 3 ) and their mixtures were carried out and underlined the importance of the multimetallic synergy on catalytic activity. Finally, a comparison of this new synthetic method assisted by the biosourced catalyst with the previously described procedures to access moclobemide was realized by calculating their green chemistry metrics. This study revealed that the use of the biosourced catalyst led to one of the greenest synthetic methods described today to produce moclobemide.


Introduction
Nowadays, the practice of chemistry more respectful of human and the environment has become an essential axis of research. According to the 12 principles of green chemistry [1], there are many ways to reduce the environmental impact of organic synthesis. For instance, the use of biobased raw materials, limitation of the use of solvent, or the development of synthesis under reasonable pressure and temperature conditions. In this study, we will focus on the well-known concept in organic chemistry, which fits perfectly into this perspective of green chemistry: catalysis. Indeed, the main effect of catalysts is the reduction of the energy threshold necessary for the reaction progress, which generates a higher yield, shorter reaction time, or lower activation temperature. There are many families of chemical species capable of playing the role of catalyst [2]. Lewis acids are the most common and widely used in homogeneous and heterogeneous organic syntheses when they are supported on montmorillonite K10 [3][4][5].
Still in the context of green chemistry, the development of inexpensive catalysts from renewable sources is becoming a challenge for industrials and researchers. In this light, a decade ago, Pr. Grison and her team developed for the first time a new type of catalyst [6], called eco-catalyst or biosourced catalyst. These catalysts originated from hyperaccumulative biomasses grown on metal contaminated soils. Metal-uptake by selected plants then takes place, producing biomass enriched with metals. Biomasses are then recovered and after thermal and several chemical treatments, metals are extracted in the form of Lewis acids, thus constituting plant-based biocatalysts. As summarized by Helchelski et al. [5], this technology has proved effective in many different chemical reactions such as well-known Friedel-Crafts alkylation and acylation, Biginelli reaction, or Suzuki-Miyaura coupling. Nevertheless, the use of hyperaccumulating plants presents several drawbacks. Their size and therefore the amount of biomass ultimately generated is often small (e.g., Anthyllis vulneraria L.), requesting the use of specific tools for their harvesting. The plants are often endemics (i.e., Grevillea exul Lindley, [7]) and so, there are unusual plants in some regions (e.g., Noccaea caerulescens (J. Presl and C. Presl) F.K. Mey, Bacopa monnieri (L.) Wettst.) and finally not recommended for production in large-scale contaminated soils due to the lack of seeds. The use of so-called "tolerant" plants, which can grow on metal contaminated soils while concentrating a significant amount of metals [8] may therefore constitute a sustainable alternative. Indeed, in the first part of the current study [9], our group demonstrated that Miscanthus × giganteus, with the assistance of a monocalcium phosphate (MCP) amendment, accumulated, to a certain extent, metals of interest in the aerial parts.
In the incessant search for "tolerant" plants and their requalifying into bio-ore resources, the current study deals with (i) the conception of two biocatalysts using the stems of miscanthus from the first part of this study [9]; (ii) the evaluation of the catalytic activity of the biocatalysts prepared in the synthesis of moclobemide, a main ingredient of a marketed drug used to treat major depressive episodes and anxiety (marketed under different commercial names: Amira ® , Aurorix ® , Clobemix ® , Depnil ® , or Manerix ® ); (iii) additional comparative experiments with different commercial metallic salts in order to identify the real species responsible for the catalytic activity present in the biosourced catalysts; (iv) comparison of our method with the other described synthesis pathways to access moclobemide using green chemistry metrics.

Mineralization and Analysis
The mineralization method was based on the procedure described in literature [10,11]. The concentration of heavy (Cd, Pb, Zn, Cu, Mn, and Fe), alkali (Na and K), and alkaline earth (Ca and Mg) metals in ashes and as well as the HCl-extracted metals were determined by flame atomic absorption spectrometry (AA-6800, Shimadzu, Tokyo, Japan) following the recommendations described in the literature [11,12] to avoid potential spectra interferences. Details on characteristics of light source, limits of detection and quantification were given in Waterlot and Hechelski [11].

Conception of Biocatalysts
Stems of miscanthus (57 g) from mesocosm experiment using dicalcium phosphate (DCP) as amendments from the first part of this study [9] were transformed into ashes in a muffle furnace (Nabertherm P330, Lilienthal, Germany) under air flow using the following temperature program: (i) 20-250 °C in 30 min, (ii) 250 °C for 1 h, (iii) 250-500 °C in 2 h, and (iv) 500 °C for 8 h. The concentration of metal in ashes from the stem of miscanthus are summarized in Table 1. The resulting ashes were treated with HCl aqueous solution at different concentrations ranging from 0.5 to 6 M with a weight ratio ashes/HCl: 1/10 ( Table 2). The best extraction rate of metallic species from ashes was obtained when a 2 M HCl aqueous solution was used and ranged from 80% (Cu) to 100% (Fe) ( Table 2). The mixture was stirred at 70 °C for 2 h and the final suspension was poured through a filter paper in a Buchner funnel (Cloup, champigny-sur-Marne, France). After cooling, the mixture was filtered through celite and the filtrate was evaporated to dryness. The resulting solid was dried at 80 °C for 12 h, to provide 1.9036 g of a white powder. From this, 0.5036 g were used as such and constituted the homogenous catalyst (M1) and the remaining 1.5 g were supported on montmorillonite K10 to obtain 1.22 mmol Zn per g of support in dry methanol to provide the heterogeneous catalyst (MM1) [13]. Upon concentration in vacuo, MM1 was finally activated at 120 °C overnight.

Atom Economy (AE) and Percentage of Waste
Atom economy is a theoretical value introduced by Trost et al. [14]. It represents the percentage of atoms from reactants which are involved in a reaction providing a product. It is defined as the ratio of the molar mass of the isolated product over the sum of the molar masses of all the products which appear in the stoichiometric equation (Equation (1)).
The percentage of atoms included in the waste resulting from the reaction (Equation (2)).

Environmental Factor (E-Factor)
The environmental factor is an indicator very representative of the waste generated during a reaction [15]. It is calculated by dividing the mass of waste by the mass of product (Equation (3)), giving a value representing the mass of waste generated by mass of products.

Reaction Mass Efficiency (RME)
Reaction mass efficiency allows to calculate the mass conservation of a reaction [16]. It is defined as the ratio of the weight of the isolated product over the sum of the weight of the reactants (Equation (4)).

Application of the Biosourced Catalysts in the Synthesis of Moclobemide and Additional Experiments with Different Commercial Metal Salts
The extracted Lewis acids obtained from miscanthus stems led to the conception of two biosourced catalysts (M1 and MM1) which were involved in the synthesis of moclobemide. The conversion rates generated by the biosourced catalysts were first compared with that obtained in the absence of catalyst. From Table 3, we can observe that without any catalytic system, at 100 °C for 18 h, the conversion rate was 25%, while the use of M1 and MM1 increased the conversion rate to 90% and 79%, respectively. These first results clearly proved the catalytic activity of the biosourced catalysts. Moreover, the different efficiencies between M1 and MM1 can be explained by the nature of the catalyst. Indeed, M1 was composed of crude extracted Lewis acids without further workup or purification, while in MM1, extracted crude Lewis acids were supported on MK-10 clay. The same Zn/reagents ratio was conserved during the chemical syntheses of moclobemide using M1 or MM1 catalysts. However, it is important to note that even if MM1 showed slightly reduced catalytic activity compared to M1, thanks to its heterogeneous nature, it was recovered and reused at the end of the reaction. The catalyst MM1 showed good catalytic performance for up to five cycles of utilization. We tested it only on five runs of the same chemical transformation. At the end of the 5th run, reused MM1 still induced moclobemide formation (56%), which was much higher compared to the reaction carried out without catalyst (25%) ( Table 3). From a global point of view, the use of the two biosourced catalysts led to a very clean process proceeding in solvent-free conditions, without the need of specific atmosphere conditions (e.g., argon, nitrogen), large excess of reactant, or the use of expensive and/or exotic catalyst or ligand, while preserving very good performance. As detailed in Table 1, the Cd and Pb concentrations are two of the three lowest concentrations. Taking into account the percentage of extracted metal using 2 M HCl, the percentage of Cd and Pb are 0.035% and 0.022%, respectively. Considering the general procedure, this means that 7 µg Cd and 4.5 µg Pb were introduced in the mixture. After purification of moclobemide, it was analyzed by ETAAS to highlight possible metallic traces due to the catalysts. The concentrations of Cd and Pb were below the limit of detection (LDCd = 0.02 µg/L; LDPb = 0.06 µg/L). After having proved the effectiveness of the designed miscanthus-based catalysts M1 and MM1, additional chemical efforts were undertaken to try to identify the real species responsible for the catalytic activity. In this perspective, we evaluated the catalytic activity of all commercial metal chloride versions of the metals contained in miscanthus stems. The results are presented in Table 4. A different catalytic behavior was observed for the salts tested. Some metallic salts provided better conversion rate (MnCl2, FeCl2, AlCl3 with, respectively, 56%, 53%, and 54%) compared to others (FeCl3, ZnCl2, and CuCl2, respectively, 41%, 36%, and 39%). However, none of the investigated pure metal chloride provided a conversion rate equivalent to that obtained with the biosourced catalysts M1 and MM1. These results corroborated well with the statement that the catalytic activity does not come from a single metal but is probably due to a multimetallic synergy, in accordance with the work of Professor Grison's team related by Hechelski et al. [5]. The investigation of binary equimolar mixtures of commercial metal chlorides was next envisaged. Fourteen mixtures of six metal chlorides (CaCl2, ZnCl2, AlCl3, MnCl2, FeCl2, and FeCl3) were thus tested in the same synthetic procedure to access moclobemide. The choice of the metal salts for the tests in mixture was dictated by the results obtained with the pure salts used alone (Table 4). Indeed, since CaCl2, MgCl2, NaCl, and KCl provided similar yields (42-44% yields), only CaCl2 was retained for binary mixture investigation. CuCl2 was also excluded since its efficiency proved to be limited (only 39% yield). Results obtained with selected mixtures of metal chlorides are reported in Table 5. Again, the equimolar mixtures of the metal salts tested did not allow to achieve the same yields as those obtained with the biobased catalysts M1 or MM1. The most efficient tested binary mixture was the one composed of MnCl2 and FeCl2 and provided moclobemide in 54% yield not exceeding the performance of these same species used alone. The conclusions which emerged from this study were threefold: (i) the equimolar mixture was not representative of the chemical composition of the biosourced salts; then, (ii) the counterion of the metal had an importance on the catalytic activity; finally, (iii) in investigated mixtures, the metal salts seemed to be in competition which resulted in situ in a decrease in catalytic activity in the most cases (see ZnCl2 and AlCl3 alone (36% and 54% yield, respectively) in Table 4 compared to ZnCl2/AlCl3 mixture (30% yield) in Table 5). The next question that we wanted to answer was the importance of the ratio between the metallic species in order to highlight a possible mutual poisoning of the species and the impact on the catalytic activity. For this issue, MnCl2 (56% yield) and FeCl2 (53% yield) were selected, two of the most efficient metal chlorides identified during the previous assays. As a reminder, manganese and iron were also two of the metals present in large quantities in the biosourced catalysts ( Table 1). Mixtures of 0.005, 0.01, 0.015, 0.02, 0.025 eq. of FeCl2 with 0.045, 0.04, 0.035, 0.03, and 0.025 eq. of MnCl2 were then tested (Table 6). Indeed, the ratio of metal salts proved to be important for the catalytic activity. A small excess of FeCl2 in the mixture FeCl2/MnCl2 0.03/0.02 eq. resulted in an increased yield of moclobemide (78%) compared to pure FeCl2 (53% yield) and MnCl2 (56% yield) studied alone. Interestingly, another argument that supported the importance of these ratios was the fact that the reverse ratio of 0.02 eq. of FeCl2 mixed with 0.03 eq. of MnCl2 completely inhibited the reaction and gave a modest yield of 20%. All other tested proportions were less effective than the mixture FeCl2/MnCl2 0.03/0.02 eq. ( Table 6). This corroborated well with the fact in the biosourced catalyst, iron was detected in a greater concentration than that of manganese. Therefore, iron and manganese species can be considered as important contributors to the catalytic activity induced by the biosourced catalysts M1 and MM1 in the synthesis of moclobemide. The multimetallic composition of these catalysts allowed a synergy of action, which was not reachable using monometallic sources.

Conclusions
Two biosourced catalysts (M1 and MM1) were successfully obtained from miscanthus biomass cultivated on metal-contaminated soils and were investigated in the synthesis of moclobemide.
Since many syntheses of moclobemide previously described have drawbacks, a review of all the procedures that allowed to obtain moclobemide with a yield greater than 75% was carried out (Figure 2). From this review, the test of the newly obtained biosourced catalysts demonstrated interesting catalytic activity and allowed to obtain moclobemide in good (79% for MM1) to excellent yield (90% for M1). The reaction proceeded in solvent-less conditions under heating at 100 °C for 18 h and needed 0.05 equiv of catalyst only. Of interest, even if the catalyst MM1 induced decreased catalytic activity compared to M1, it had the advantage of being regenerable at the end of the reaction and was reused for up to five runs.
Encouraged by the catalytic activity of M1 and MM1, additional studies to find the metal species responsible for their catalytic activity were carried out. Among the tested pure commercial metal chlorides, some of them resulted in better conversion rate (MnCl2, FeCl2, and AlCl3 induced, respectively, 56%, 53%, and 54% yield) compared to others (ZnCl2, CuCl2, NaCl, KCl, CaCl2, and MgCl2 providing, respectively, 36%, 39%, 43%, 43%, 42%, and 44% yield). No pure species equaled the performance of biobased catalysts M1 or MM1 and highlighted the importance of the synergy of metallic species. Binary equimolar mixtures of commercial metal chlorides were also tested in the same synthesis to obtain moclobemide. The best efficiency was obtained with a mixture of MnCl2 and FeCl2 but did not exceed the performance induced by M1 or MM1 (54% versus 90% and 79%, respectively). This allowed to conclude that the equimolar mixture is not representative of the composition of the biosourced salts. Different ratios of MnCl2 and FeCl2 were finally studied and highlighted the importance of the proportion of MnCl2 compared to that of FeCl2. While a small excess of FeCl2 compared to MnCl2 (0.03:0.02 eq.) allowed to reach 78% yield equivalent to that of MM1, the reverse ratio FeCl2:MnCl2 0.02/0.03 eq. considerably inhibited the chemical transformation and provided moclobemide in only 20% yield. This corroborated well with the composition of the biosourced catalyst, iron being detected in a greater concentration than that of manganese. Therefore, iron and manganese species proved their important role to the catalytic activity induced by the biosourced catalysts M1 and MM1 in the synthesis of moclobemide.
Finally, previously and newly described procedures to access moclobemide were all compared in terms of green chemistry metrics (AE, RME, and E-Factor) to evaluate their environmental impact. The biosourced catalysts, especially M1 (AE of 89.3%, RME of 80.4%, and E-Factor of 0.243), led to one of the greenest synthetic routes described to date to produce moclobemide. This study, therefore, opens new perspectives for the use of biobased catalysts obtained from plants as substitutes for products from the petrochemical industry. Moreover, their use can be further considered in various chemical transformations using Lewis acids and consequently contribute to the circular economy.
Author Contributions: Formal analysis, investigation, methodology, writing-original draft and editing, T.G.; conceptualization, design of experiments, A.G. and C.W.; writing-review, C.W. and A.G. All authors have read and agreed to the published version of the manuscript.

Funding:
The authors thank the Région Hauts-de-France and JUNIA for financial support (FEDER funding) of this work and technical help and facilities.

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