Field of the invention

The invention relates to the use of liquid metal and liquid metal alloys in the field of heterogenous catalysis, to a catalyst system using such liquid metal alloys and to a catalytic process using the same.

Background of the invention

Catalysis is ubiquitous in the chemical industry of all kinds. It is estimated that around 90% of all commercially produced chemical products involve catalysts at some stage in the process of their manufacture. Catalysts are not consumed in the reaction and remain unchanged after it. If the reaction is rapid and the catalyst recycles quickly, very small amounts of catalyst often suffice.

The solvent selection plays an important role in both homogeneous and heterogeneous catalytic reactions. In the case of homogeneous catalysis, the solvent may affect the intermediate states and decrease the reaction activation energy [1], The proper solvent can also shift the reaction selectivity to the targeted product. The effect of solvents on heterogeneous catalysis is more complex because it involves interfacial effects, competitive adsorption and mass transfer [2, 2a]. Therefore, the solvent effect on activity, selectivity and stability is much less predictable.

Depending on the type of the reaction, solvent can strongly interact with the catalyst [3,3a] or be totally inert, for example in the case of extraction of the product [4], The strong interaction of the catalyst with the solvent and low interaction with reagents has been recently proposed as a way to merge the advantages of homogeneous and heterogeneous catalysis. For example, ionic liquids with very low volatility in the form of a separate phase or coated over support (SILP) with dissolved active phase have been efficiently used for a number of applications [5]. However, the limited thermal stability of liquid organic salt (<200°C) and leaching of ionic liquid during reaction significantly restrict the application of this concept in catalysis.

In comparison with organic solvents, mono-component liquid metal or their alloys (liquid metal and liquid metal alloy are referred as ‘LM’ in this invention) with a higher boiling point can be used without temperature limitations. LM is being developed for different applications [6, 6a]. Low melting point, easy recovery, high mobility and surface energy are the main advantages of LM for catalytic applications. Supported catalysts promoted by LM features a molecularly defined, catalytically active liquid film/droplet layer adsorbed on a porous solid support. Such bimetallic catalysts comprising supported Ni, Pd and Pt promoted with Ga have demonstrated high selectivity, activity and stability in hydrogenation and dehydrogenation reactions [7, 7a, 7b]. The reaction proceeds on the intermetallic surface with modification of electronic structure and isolation of metal active sites. Recent results by the inventors have demonstrated that the presence of pseudo-liquid Bi and Pb film on the surface of Fe and Co catalysts results in significant enhancement of selectivity, activity and stability in high and low temperature Fischer-Tropsch synthesis [8]. The effect has been ascribed to the high mobility of Bi and Pb promoters, which facilitates CO dissociation by scavenging O atoms and improves the stability of metal nanoparticles. However, metal-based catalytic reactions are limited and possible aggregation of liquid metal promoters and catalyst active sites during the reaction or catalyst activation are disadvantageous. A lot of industrial catalytic processes nowadays require extra organic solvents as reaction medium and commonly suffer from low selectivity with the formation of side products and fast catalyst deactivation due to the sintering of metal or deposition of carbon species. It results in environmental issues and low economic efficiency of the process.

Hence there is a need to provide an alternative catalyst system which would exhibit improved selectivity, activity and stability of the catalyst, would be easy to implement, would use low-cost material and/or energy and/or would allow easy separation of the catalyst system from the substrate(s), the product(s) and/or other constituents of the reaction medium.

Unexpectedly, it was found that excess, or a large quantity, of LM in the reacting medium introduces numerous electronic, structural and interfacial effects. It can influence the adsorption of reagents, intermediates, products, mass and heat transfers. For metallic and acidic catalytic reactions, the use of LM in excess surprisingly improves the selectivity, activity and/or stability of the catalyst.

Description of the invention

One object of the invention is a heterogenous catalyst system comprising a metal or acid catalyst in a solid state and a liquid metal or a liquid metal alloy, wherein said catalyst is in contact with a liquid medium, said liquid medium comprising said liquid metal or said liquid metal alloy, and wherein the mass ratio of said liquid metal or said liquid metal alloy to said catalyst is at least 0.1 . Preferably the mass ratio of said liquid metal or said liquid metal alloy to said catalyst is at least 0.5, more preferably at least 1 , even more preferably at least 10 and advantageously at least 40 and even more advantageously at least 100. As it can be observed from the examples a mass ratio of said liquid metal or said liquid metal alloy to said catalyst ranging from 1 to 50, preferably from 5 to 30 and even preferably from 10 to 20 provides good results. The catalyst in a solid state can be in a powder, or a nano-powder, shape. It can be supported or not supported.

The solid heterogenous catalyst according to the invention may be immersed in the liquid medium, and advantageously in the liquid metal or the liquid metal alloy.

According to an embodiment of the invention, the liquid metal is a metal having a low melting temperature such as Ga (29 °C), In (156 °C), Sn (231 °C), Pb (327 °C) and Bi (271 °C).

According to a particularly preferred embodiment of the invention the liquid metal or the liquid metal alloy is a liquid metal alloy. It preferably comprises Ga, In, Sn, Pb, Bi or a mixture thereof and has advantageously a melting point inferior to 200°C. It is further preferred that the liquid metal alloy comprises or be constituted of Bi and Sn.

For example, the liquid metal alloy may be chosen in the group consisting of PbSn, InBiSn, InBi, Wood’s metal (BiPbCrSn), BiSn and BiSnAg, preferably the metal alloy is a PbSn alloy or a InBiSn alloy.

The solid acid catalyst can be a silico-aluminate (such as a zeolite, an alumina, a silico- alumino-phosphate), sulfated zirconia, and one of many transition metal oxides (such as titania, zirconia, niobia, and more). Preferably the solid acid catalyst may be a zeolite, such as a faujasite (FAU), a pentasil (MFI), a mordenite (MOR) or a beta (BEA). It can be:

- a small-pore zeolite, with 8-membered oxygen rings and a “free” diameter of 3 - 4.5 A (FER, CHA), a medium-pore zeolite, with 10-member oxygen rings and a “free” diameter of 4.5 - 6 A (MOR, MFI); or

- a large-pore zeolite, with 12-member oxygen rings and a “free” diameter of 6 - 8 A (BEA, FAU).

The acid catalyst is preferably an acylation catalyst and in particular a zeolite, such as a MFI, a MOR or a BEA.

The solid metal catalyst may advantageously be an oxidation catalyst which comprises, or consists, of Pt, Ru, Pd, Rh, Ni, Au and mixtures thereof. It can be in a bulk state, such as in the shape of powder, and/or be supported. The supporting material may be C, SiO2, AI2O3 and/or TiO2.

The heterogenous catalyst system may advantageously comprise a substrate. The substrate is a reacting material which transformation will be catalysed by the heterogenous catalyst system. Organic molecules such as alcohols, carboxylic acids, esters, alkenes, alkynes and mixtures thereof, provide good substrate materials. There is no particular limit to the size of the molecules than can be used as a substrate, but molecules having from 1 to 45 carbon atoms, preferably from 3 to 25 carbon atoms, are considered particularly suitable. Examples of such substrates are phenol, hexanoic acid, benzoic acid, resorcinol, 1 -naphtol, 1 - dodecanol, benzyl alcohol, glucose, glycerol, toluene, octene, n-undecane, and ethyl acetanoate.

Products that can be obtained using the catalyst system of the invention are esters, alcohols, polyols, aldehydes carboxylic acids and alkenes. Examples of such products are phenyl ester, resorcinol ester, naphtol ester, cumarine and its derivatives, octyloxybenzene, benzaldehyde, benzyl alcohol, benzoic acid, phenyl benzoate, glyceric acid, formic acid, glycolic acid and oxalic acid.

Particularly suitable substrates for an acylation reaction to take place are phenolic compounds and/or carboxylic acids, such as phenol, hexanoic acid, benzoic acid, resorcinol, 1 -naphtol. In that case, an acid catalyst such as the ones described above will advantageously be selected.

Particularly suitable substrates for an oxidation reaction are alcohols and/or alkenes, such as glucose, glycerol or toluene. In that case, a metal catalyst such as the ones described above will advantageously be selected.

Another object of the invention is a process to catalyse a chemical reaction to obtain a product, said process comprising the following step: reacting a metal or acid catalyst in a solid state, a substrate and a liquid medium said liquid medium comprising a liquid metal alloy, wherein the mass ratio of said liquid metal alloy to said catalyst is at least 0.1 .

The word “reacting” is used in its common sense to describe that the various components will interact to some degree with each other. As it will be directly understood, only the substrates and other reactants will the subject of the chemical reaction and be transformed into another chemical entity.

The heterogenous catalyst system which is used in the process of the invention, including the substrate(s), is advantageously as described above.

According to a particularly advantageous aspect of the invention, the reacting step may take place at a temperature which is below 200°C, preferably below 190°C. As it can be seen from Table 1 , the reacting temperature will be chosen to be superior to the melting point of the selected liquid metal alloy. Hence, this temperature can be selected to be below 142°C, below 75°C or even below 64°C. The range of temperature can advantageously range from 25°C to 200°C, preferably from 60°C to 140°C.

The reacting step of the process of the invention can take place in aerobic conditions or inert conditions (such as an inert atmosphere) and eventually in presence of another (inert) solvent such as water. Advantageously, this step is not carried out in the presence of a solvent other than the liquid metal alloy(s) and/or an organic (i.e., carbon based) solvent. Preferably the process and/or the heterogenous system of the invention comprises only the catalyst(s), the substrate(s), the liquid metal alloy(s) and, eventually, the product(s) of the reaction.

The reacting step can comprise a mixing step where the catalyst(s), the substrate(s), the liquid metal alloy(s) and, eventually, the product(s) of the reaction are mixed together before and/or during a heating step is provided. The mixing can be carried out at a rate ranging from 50 to 1000 rpm, preferably from 200 to 900 rpm, and advantageously from 350 to 750 rpm.

The process can be carried out at elevated pressure or not. Gauge pressure ranging from 1 to 50 (g) bar, preferably from 5 to 40 (g) bar, in particular from 10 to 30 (g) bar. When a substrate is a gas (e.g., air, oxygen), such as it is often the case in oxidation reactions, the gas can be provided as a flow. The flow rate can be easily selected by the skilled person but a flow rate from 50 to 400 ml/min, preferably 75 to 300 ml/min and advantageously 100 to 250 ml/min is generally preferred. It is further preferred to select, for an octanol oxidation, a gas hourly space velocity (GHSV) ranging from 300 to 750 L h ^-1 g ^-1 ; for glycerol oxidation, a GHSV ranging from 120 to 300 L h ^-1 g ^-1 ; and for glucose oxidation, a GHSV ranging from 60 to 150 L h ¹ g ¹ .

The reacting step is carried out for a reaction time which may range from a few minutes, even seconds to several days. The reacting time is preferably selected from 5 minutes to 24 hours, and more preferably from 10 minutes to 2 hours, and even more preferably from 15 minutes to 1 . 5 hours.

According to one particular aspect of the invention, the process is easy and straightforward to carry out.

The heterogenous catalytic system of the invention and the process of the invention can be used to carry out various chemical reactions which may be chosen in the group selected from acylation, oxidation, Pechmann condensation, alkylation reaction, Knoevenagel condensation, cracking of alkanes, polyethylene pyrolysis. Acylation, oxidation and alkylation reactions are preferred as a particular increase of the selectivity is noted. The O-acylated product is favoured in these reactions. This is exemplified below and this has been noted for, for example, phenol, resorcinol or naphthol reacting with ethyl acetoacetate to obtain coumarin or coumarin derivatives as well as the alkylation of phenol with octene to obtain the O-alkylated compound and the oxidation of 1 -octanol to obtain octanal. The acylation of unsaturated cyclic (C6) phenolic compounds, such as phenol, resorcinol and 1 -naphthol, with carboxylic acids is one of the particularly preferred reactions which can be carried out according to the process of the invention.

Another object of the invention is the use of liquid metal or a liquid metal alloy as described within this specification to obtain a product, to increase and/or control the conversion rate, the selectivity and/or the stability of the catalyst in a chemical reaction which comprises an acid or metal catalyst in a solid state and at least one substrate, wherein said liquid metal or said liquid metal alloy is used in a quantity so that the mass ratio of said liquid metal or said liquid metal alloy to said catalyst is at least 0.1 . The preferred features of the heterogenous catalyst system and of the process of the invention which are described above can be applied to the particular use.

Brief description of the figures

Figure 1 : Is a schematic representation of a new reacting medium for heterogeneous catalysis according to the invention;

Figure 2: Shows various acylation pathway and products obtained in Examples 1 -4.

Figure 3: Shows various oxidation products obtained in Example 6.

Figure 4: Shows various glucose oxidation products obtained in Example 7.

Figure 5: Shows various glycerol oxidation products obtained in Example 8. Figure 6: Shows various toluene oxidation products obtained in Example 9.

Examples

Table 1. Liquid metal alloys under test in these examples. All the liquid metal alloys were provided by FCT Solder and Haines & Maassen.

Alloy Composition (wt.%) Melting point (°C)

PbSn alloy Pb37Sn ₆ 3 183

InBiSn alloy ln ₅ iBi ₃ 2.5Sni6.5 61

InBi solder ln ₆ 6.7Bi33.3 72

Wood’s metal Bi ₅ oPb25Cdi2.5Sni2.5 70-74

BiSn alloy Bi ₅ 8Sn ₄₂ 138

BiSnAg alloy Bi ₅ 7Sn ₄₂ Agi 139-140

Example 1 : Catalytic acylation of phenol using BEA zeolite with PbSn alloy in the reaction mixture and comparative example

The acid catalyst is a zeolite known as BEA purchased from Zeolyst with SiO^AfeOs ratio of 25. The PbSn alloy (PbSn: 37 % Pb-63 % Sn) is a eutectic alloy with a melting point of 183 ^e C and was purchased from FCT Solder. The reaction conditions were: 10 mmol phenol, 10 mmol hexanoic acid, 0.2 g BEA and 4 g of PbSn alloy, 190°C. The results are provided in Table 2 below.

Table 2. Catalytic results of phenol acylation over BEA zeolite with and without PbSn alloy in the reaction mixture.

>

Catalyst Substance Reaction Conversion Selectivity Structure system time (h) (%) to O- acylated product (%)

BEA Phenol Hexanoic 1 26.4 acid 3 37.5

5 66.3 40

7 62.7

8 61 .8

BEA + Phenol Hexanoic 1 25.2

PbSn acid 3 33.9

5 64.1 93.3

7 75.3

8 89.1

The presence of LM totally changes the catalytic performance of the reaction. LM leads to a significant increase of the selectivity to O-acylated products without deactivation of the catalyst in comparison with low selectivity and fast deactivation of the catalyst by multi-acylated product. Microscopy (SEM and TEM) analysis demonstrates the interaction of zeolite particles with liquid metal with the presence of small 1 -3 nm alloy particles localized at the interphase zones and in the pores of zeolite. Interaction of LM with acid sites leads to continuous refreshing of the surface of zeolite with the removal of surface species as precursors of the coke (carbon species which can deactivate the catalyst) and modification of the strength of acid sites by interaction with LM. It results in excellent activity, selectivity and stability of the catalyst.

Example 2: Catalytic acylation of phenol using BEA zeolite with other alloys in the reaction mixture

In order to verify the general applicability of the process of the invention, phenol acylation with hexanoic acid in the presence of other LM, such as wood's metal, BiSn and BiSnAg alloy was further investigated (Table 3). Reaction conditions: 10 mmol phenol, 10 mmol hexanoic acid, 0.2 g BEA, 4 g LM (wood’s metal, BiSn or BiSnAg alloy), 500 rpm, 175°C, 3 hours.

Table 3. Catalytic results of phenol acylation over BEA zeolite with various alloys in the reaction mixture.

LM Conver- Selectivity Structure Selectivity Structure Selectivity Structure sion to O- to o-C- to p-C-

(%) acylated acylated acylated product product product

With LMs, all cases showed exceptional high selectivity to the O-acylated products, phenyl ester. Some decrease in phenol conversion was observed after the addition of BiSn and BiSnAg alloy, however, Wood's metal showed a comparable activity with solo zeolite catalyst and a significant enhancement of phenyl ester selectivity.

Moreover, an O-acylated selectivity higher than 95% could be maintained, suggesting the superior stability of LM assisted catalyst system.

Example 3: Catalytic acylation of various phenols (Phenol, Resorcinol, 1 -naphthol with hexanoic or benzoic acid and comparative data.

Acylation of phenyl derivatives with both hexanoic acid and benzoic acid was studied under the same conditions (Table 4 and Table 5). The starting materials were obtained as described in the previous examples.

Table 4. Catalytic results of acylation of various substrates with hexanoic acid over BEA zeolite with wood's metal alloy in the reaction mixture.

Substance LM Conversion Selectivity Structure

(%) to 0- acylated product (%)

, Hexanoic . , „ „„ „

Phenol without 72.5 60.3 acid Hexanoic ^{without 87 3 8 2 acid} with 65.4 94.6

Reaction conditions: 10 mmol phenol, 10 mmol hexanoic acid, 0.2 g H-BEA, 500 rpm, 175°C, 3 hours, in case of LM, 4 g wood’s metal.

Table 5. Catalytic results of various phenols acylation withbenzoic acid over BEA zeolite with wood's metal alloy in the reaction mixture.

Substance LM Conversion Selectivity to Structure

(%) O-acylated product (%)

Reaction conditions: 10 mmol phenol, 10 mmol hexanoic acid, 0.2 g H-BEA, 500 rpm, 175°C,

3 hours, in case of LM, 4 g wood's metal.

Catalysed by the BEA + wood's metal association, both hexanoic acid and benzoic acid afforded a high phenol conversion, up to 70%, which is comparable and even better than the case of only the zeolite being used, with significantly enhanced selectivity for O-acylated products, phenyl esters. The catalyst system also functioned greatly in the acylation of other substances, such as resorcinol and 1-nathphol with more electron-donating hydroxyl group or larger molecular size, respectively. In comparison to phenol, without LM, resorcinol is rich in electrons and showed higher conversion in both acylation reactions, while C-acylation products, aryl ketones were the main products. In particularly, when hexanoic acid was used as the acylating agent, 1-(2,4-dihydroxyphenyl)hexane-1-one was the dominating product up to 90%. By adding LM into the reactor, the products selectivity significantly shifted to O- acylated phenyl esters. Impressively, with benzoic acid as acylating agent, in the presence of LM, the selectivity to O-acylated product, 3-hydroxyphenyl hexanoate, reached 100%. In addition, 1 -naphthol was selectively acylated to phenyl esters in the presence of Wood's metal, giving 94.5% and 98% selectivity to phenyl ester with hexanoic acid and benzoic acid as acylating agent, respectively. The moderate decrease in the conversion of 1 -naphthol was possibly due to the increased steric hindrance by LM interaction with acid sites.

Example 4: Catalytic reusability and comparative data.

The catalyst system was used several cycles after being washed with ethanol and water then dried and shows markedly sustained performance and reusability as shown in Table 6.

Table 6. Catalytic reusability of phenol acylation over BEA zeolite with PbSn alloy in the reaction mixture.

Catalyst Substance Cycles Conversion system (%)

1 37.2

BEA Phenol Hexanoic acid 2 21.0

3 12.1

1 34.1

BEA + PbSn Phenol Hexanoic acid 2 36.0

3 34.1

Reaction conditions: 1 g phenol, 1 g hexanoic acid, 0.2 g BEA and 4 g alloy, 190°C, 3 hours.

Example 5: Oxidation of 1 -dodecanol using a Pd sponge and a PbSn alloy and comparative data

A different catalyst, a Pd sponge was used with 1 -dodecanol as substrate. The reagents were as described above. The Pd sponge (99.9 % metal basis) has been provided by Sigma- Aldrich and the results are shown in Table 7 below.

Table 7. Catalytic results of 1 -dodecanol oxidation over Pd sponge with and without PbSn alloy in the reaction mixture.

Catalyst LM Reaction Conversion Selectivity Structure time (h) (%) to dodecyl dodecanoate (%)

Reaction conditions: 190°C, 2 g dodecanol, 30 mg of Pd sponge and 4 g PbSn alloy.

It was observed that the activity of Pd is much higher in the presence of LM. The selectivity distribution also has been significantly affected by the LM. The main product of the reaction in this case was the ester and the selectivity was higher than 80% with the presence of much smaller amounts of aldehyde and acid.

Example 6: Oxidation of benzyl alcohol using a Pd/ALOa catalyst with InBiSn alloy in the reaction mixture and comparative data

Pd/AI ₂ Os (5 wt. % loading, powder from Sigma-Aldrich) was used with benzyl alcohol (99%, Alfa Aesar) as substrate. The LM of InBiSn was as described above. The results are shown in Table 8 below.

Oxidation of benzyl alcohol was conducted at 150°C under air flow over Pd/AI ₂ Os with and without InBiSn alloy.

Table 8. Catalytic results of benzyl alcohol oxidation over Pd/AI ₂ Os with and without InBiSn alloy in the reaction mixture.

Reaction Conversio Selectivity time (h) n Toluene Benzaldehyde Benzaldehyde Benzoic Dibenzyl Benzyl

(%) dimethyl acetal acid ether benzoate

1 22.5 26.2 44.9 12.8 0 12.6 3.6

Pd 8 61 .9 0.3 40.4 31 .9 2.4 21 .2 2.6

24 94.3 0 26.3 0 10.4 21 .9 38.5

1 30.9 26 12.9 37.2 0 21 .5 0.6

Pd+

8 48.3 4.7 66.8 0 0.6 15.4 1 1 .9

InBiSn

24 88.2 0.4 53.3 0 2.0 7.5 35.8

Reaction conditions: 4 ml benzyl alcohol, 20 mg Pd/AI ₂ Os (5% wt. Pd), 0.2 g InBiSn alloy, 150°C, 800 rpm, air flow 170 ml/min.

It can be observed that over the whole reaction time of 24 hours, both catalytic systems exhibited close benzyl alcohol conversion rates (see Table 8), indicating that the addition LM barely affected the catalytic activity. However, without LM, several kinds of oxidation products, such as benzaldehyde, benzaldehyde dimethyl acetal and dibenzyl ether (see Figure 3) were obtained at similar degrees of selectivity. This suggests that the multiple reactions took place unselectively over the Pd catalyst. However, when LM was added, the product distribution changed significantly and benzaldehyde was produced with a 66.8% selectivity. Though an equivalent amount of benzyl benzoate was produced in the presence of LM, there was no benzaldehyde dimethyl acetal among the products after 8 hours.

Though the conversion of benzyl alcohol increased along with the reaction time in a similar manner with and without LM, the evolution of the product selectivity varied between these two reaction systems. In the absence of LM in the reactor, after 1 hour, the maximum selectivity to benzaldehyde was achieved, suggesting the fast oxidation of the hydroxyl group of the benzyl alcohol molecule. Afterward, benzaldehyde is converted into benzaldehyde dimethyl acetal, benzoic acid and benzyl benzoate. The latter two products are produced by further oxidation of benzaldehyde and esterification.

A catalyst system of the invention showed higher selectivity and slower transformation into deep-oxidized products. Finally, a selectivity of benzaldehyde higher than 50% was reached 24 hours, which was much higher than what is obtained without LM (26.3%). In addition, as the reaction was processed, side products such as benzaldehyde and dibenzyl ether were consumed much more rapidly than when using the Pd only. Thus, the use of InBiSn (LM) suppressed or diminished both the over-oxidation of benzaldehyde and the occurrence of other side reactions.

Example 7: Oxidation of glucose using over Pt/C with different amount of InBiSn alloy in the reaction mixture and comparative data.

In addition, the effect of InBiSn alloy on glucose oxidation was studied by varying alloy mass from 0.1 to 1 .0 g with the constant amount of Pt/C. The results are listed in Table 9 and Table 10 below. Pt/C (5 wt. % loading, from Sigma-Aldrich) was used with glucose (D-(+)-Glucose, 99.5%, Sigma-Aldrich) as substrate. The LM of InBiSn was as described above.

Table 9. Catalytic results of glucose oxidation over Pt/C with different amounts of InBiSn alloy in the reaction mixture.

Amount of Conversion Selectivity

InBiSn (g) (%) 2-keto- Glucaric Gluconic gluconic acid acid acid

0 76 2.2 23.7 70.4

0.1 91.3 15.7 24.6 55.8

0.2 95.4 55 21 21.2

0.5 64.2 69.8 10.6 17.1

1.0 57.4 75.7 3.9 17.6 Reaction conditions: 0.5 g glucose;10 ml H ₂ O; 0.1 g Pt/C (5% wt. Pt); (0.1 -1 g InBiSn alloy;) Air flow 170 ml/min; 80°C; 24 hours.

As shown in Table 9, the glucose conversion rate was the highest when 0.2 g InBiSn alloy were added. This shows a promoting effect of LM on catalytic activity. However, too high LM loading appears to block active sites resulting in a loss of activity. A significant decrease in glucose conversions was observed for a LM amount of up to 1 g. At the same time, the selectivity for 2-keto-gluconic acid continuously increased from 2% to 76% with increasing the amount of InBiSn alloy from 0 to 1 g. The above results demonstrate that the addition of LM in the reactor can be used as an effective means for tuning the product selectivity, such as 2- keto-gluconic acid in the present case, during glucose oxidation.

The kinetics of glucose oxidation over Pt/C, with and without the addition of InBiSn alloy (0.2 g), was studied and the results are shown in Table 10.

Table 10. Catalytic results of glucose oxidation over Pt/C with and without InBiSn alloy in the reaction mixture.

Reaction Conversion Selectivity time (h) (%) 2-keto- Glucaric Gluconic gluconic acid acid acid

Pt 1 14.1 0 0.3 94.8

2 25 0 0.6 94.6

4 32.7 0.7 0.9 91.2

6 47.6 2.5 3.7 86.5

8 56.2 2.2 10.5 81.7

24 74.9 2.2 23.7 70.4

Pt+ 1 10.1 1.3 2 91.7

InBiSn 2 34.6 13.6 15.6 65.2

4 61.3 38.9 20.5 36.5

6 82.1 45.3 19 32.5

8 96.4 49.7 22 28.3

24 100 55 21 24

Reaction conditions: 0.5 g glucose;10 ml H ₂ O; 0.1 g Pt/C (5% wt. Pt); (0.2 g InBiSn alloy) Air flow 170 ml/min; 80°C; 0.5-24 hours.

In both systems, the catalytic activity increased with the reaction time. In the presence of LM, the rate of glucose transformation was however significantly higher in the first 8 hours, indicating that LM increases the catalyst activity. In terms of product selectivity, the selectivity of gluconic acid dropped with reaction time regardless of LM (24% vs 70%). However, compared to the absence of LM in the reaction medium, the production of 2-keto- gluconic acid continuously increased and it became the main product (55% selectivity after 24 hour) when InBiSn alloy was added.

Therefore, InBiSn+Pt/C shows better performances for the selective oxidation of glucose toward 2-keto-gluconic acid. This is of particular interest as this compound acts as the key precursor for the isovitamin C synthesis and its derivatives. The improved 2-keto-gluconic acid selectivity at optimized LM amount may result from the complex interactions between LM and Pt NPs and the glucose and/or the other oxidation products (shown in Figure 4) which resulted in the preferred oxidation of the hydroxyl group adjacent to the carbonyl group of gluconic acid. With the increase in InBiSn amount, it might block the sites which are more active for the further oxidation and transformation of gluconic acid to glucaric acid. Hence, the 2-keto-gluconic acid selectivity is increased for a high LM amount, despite a diminution of the catalytic activity to some degree.

Example 8: Oxidation of glycerol using Pt/C with various alloys in the reaction mixture and comparative data.

The general phenomenon that combining heterogeneous catalysts with LMs leads to enhancement of activity, selectivity and stability of the catalytic process was extended to other reactions and the effects of LM on the catalytic properties of aerobic oxidation of glycerol were also investigated. The oxidation products of this reaction are shown in Figure 5. Pt/C (5 wt. % loading, from Sigma-Aldrich) was used with glycerol (99.5%, Sigma-Aldrich) as substrate. The LMs were as described above.

Table 11. Catalytic results of glycerol oxidation over Pt/C with various alloys in the reaction mixture.

LM Conversion Selectivity

(%) Tartronic Glyceric Formic Glycolic Oxalic Mesoxalic C3/ acid acid acid acid acid acid (C1 +C2)

None 57.2 0.4 31 .7 42.9 7.1 17.9 0 0.47

InBiSn 70.1 2.2 42.1 31 .1 21 .8 2.5 0.3 0.81

InBi 47.8 0.4 40.3 14.4 25.7 19.2 0 0.67

Wood’s 42.2 1 .6 65.6 21 .6 8.1 3.2 0 2

Reaction conditions: 10 ml 0.05 g/ml glycerol aqueous solution, 50 mg Pt/C (5% wt. Pt), 0.5 g liquid metal alloy, 80°C, 600 rpm, air flow 170 ml/min.

As show in Table 11 , the use of an InBiSn alloy resulted in a remarkable increase in the glycerol conversion over the reaction time. In the meantime, the addition of LMs led to an improved selectivity towards glyceric acid from 31 .7% to 42.1% (InBiSn), 40.3% ( InBi) and 65.6% (wood's metal), respectively.

A moderate conversion of glycerol could be reached without LM. However, 42.7% selectivity of formic acid formed by cleavage of glycerol carbon chain was obtained using LM which is much higher than the selectivity of the Pt catalyst. Furthermore, it was found that regardless of the composition or type of LM introduced to the reaction system, a similar enhancement of the amount of C3 over the cumulated amount of (C2+C1 ) can be obtained. Among the LM tried, Wood's metal offered the highest selectivity to glyceric acid.

Example 9: Oxidation of toluene using MnO _x with InBiSn alloy in the reaction mixture and comparative data.

The application of LM to oxide catalyst for oxidation reaction was also studied for toluene oxidation. The oxidation products of this reaction are shown in Figure 6. MnO _x (0.5 M Mn(NOs)2 was precipitated by 0.5 M NH ₄ OH at PH=9, aged for 1 hour, dried at 120 °C overnight and calcined at 350 °C for 4 hours) was used with toluene (99.5%, Sigma-Aldrich) as substrate. The LMs were as described above. The results are shown in Table 12 below.

Table 12. Catalytic results of toluene oxidation over MnO _x with InBiSn alloy in the reaction mixture.

Catalyst Conversion Selectivity

(%) Benzald- Benzyl Benzoic Phenyl Ben- methyl 4-methyl- ehyde alcohol acid benzoate ezene benzoate 1 ,1 '- biphenyl 8.6 28.3 15.2 38.8 12.1 0.8 1 .2 2.8 2.3 53.6 14.2 13.6 1 1 .6 2.2 1 .2 2.1 1 .5 76.1 0 0 9.7 8.1 0 3.7 3 72.3 3.7 0 17.7 2.3 0 3.6

Reaction conditions: 3 g toluene, 0,1 g MnO _x , 0,1 g InBiSn alloy, 20 bar air, 200 °C, 4-20 hour.

The toluene oxidation reaction was conducted in a high-pressure reaction (Parr series). As shown in Table 12, the addition of InBiSn into the reaction system, under identical conditions, decreased toluene conversion compared to only MnOx catalytic system., which may be caused by the restricted mass diffusion of the batch reactor. However, LM significantly promoted the production of benzaldehyde products from 28.3 to 53.6 % selectivity. And when the reaction time was prolonged to 10 hours, the aldehyde selectivity achieved as high as 76.1%. These examples show that the addition of LM in the reactor during reactions such as acylation and oxidation strongly modifies the catalyst stability, activity and reaction selectivity. The contact of LM with the catalyst results in reversible modification of the active sites, stabilization of metal nanoparticles from sintering and a significant increase in the catalyst stability and reaction rate.

The contact of acid sites with LM leads to modification of selectivity due to the weakening of acid sites in BEA zeolite. Consequently, in the presence of LM, the selectivity in the acylation of phenol over BEA zeolite result in stable and selective formation of O-acylated phenyl ester products. Without wishing to be bound by theory, the interplay between metal nanoparticles an LM may be resulting in selectivity tuning which may be due to mild poisoning of unselective sites and modification of the electronic state of active sites in Pd and Pt metal catalysts. Use of LM as reacting medium in the reactor improves the selectivity and catalyst stability of numerous catalytic reactions.

The LM will perform three functions: 1 ) regeneration of the active centres of the catalyst by "washing" its surface in the presence of LM, contributing to an increase in the activity of the catalyst and the stability of its operation over time; 2) modification of the active sites of the catalyst as a result of interaction with LM to increase the selectivity of the process; 3) isolation of catalyst particles, preventing their segregation and sintering. This approach will be actively used for both metal and acid catalysts in model reactions and in biomass conversion.

References

[1 ] C. K. Regan, S. L. Craig, J. I. Brauman, Science 2002, 295, 2245-2247.

[2] aL. Shuai, J. Luterbacher, ChemSusChem 2016, 9, 133-155;

[2a] bM. A Mellmer, C. Sener, J. M. R. Gallo, J. S. Luterbacher, D. M. Alonso, J. A. Dumesic, Angewandte chemie international edition 2014, 53, 11872-1 1875.

[3] aC. Kalidas, G. Hefter, Y. Marcus, Chemical reviews 2000, 100, 819-852;

[3a] bB. Schaffner, F. Schaffner, S. P. Verevkin, A Borner, Chemical Reviews 2010, 1 10, 4554- 4581.

[4] P. Wasserscheid, W. Keim, Angewandte Chemie International Edition 2000, 39, 3772- 3789.

[5] C. P. Mehnert, R. A. Cook, N. C. Dispenziere, M. Afeworki, Journal of the American Chemical Society 2002, 124, 12932-12933.

[6] aS.-T. Liang, H.-Z. Wang, J. Liu, Chemistry - A European Journal 2018, 24,17616-17626; [6a] bT. Daeneke, K. Khoshmanesh, N. Mahmood, I. A. de Castro, D. Esrafilzadeh, S. J.

Barrow, M.D. Dickey, K. Kalaritar-zadeh, Chemical Society Reviews 2018, 47,4073-411 1 .

[7] aN. Taccardi, M. Grabau, J. Debuschewitz, M. Distaso, M. Brandi, R. Hock, F. Maier, C.

Papp, J. Erhard, C. Neiss, W. Peukert, A. Gbrling, H. P. Steinruck, P. Wasserscheid, Nature Chemistry 2017, 9, 862; [7a] bM. Jakuttis, A. Schbnweiz, S. Werner, R. Franke, K.-D. Wiese, M. Haumann, P. Wasserscheid, Angewandte Chemie International Edition 2011 , 50, 4492-4495;

[7b] cM. Armbruster, K. Kovnir, M. Behrens, D. Teschner, Y. Grin, R. Schldgl, Journal of the American Chemical Society 2010, 132, 14745-14747. [8] B. Gu, M. Bahri, O. Ersen, A. Khodakov, V.V. Ordomsky, ACS Catalysis 2019, 9, 991 - 1000; bV. V. Ordomsky, Y. Luo, B. Gu, A. Carvalho, P. A. Chernavskii, K. Cheng, A. Y. Khodakov, ACS Catalysis 2017, 7, 6445-6452.