The invention relates to a method for preparing a chemical compound from a levulinic acid source.

Levulinic acid (LA) has been identified as a promising, renewable platform molecule, that can be converted by a catalytic reaction, such as a hydrogenation or deoxygenation reaction, to another valuable chemical, such as gamma-valerolactone (GVL), methyltetrahydrofuran (MTHF), pentanoic acid (PA), pentanediol (PD) or a pentanoic acid ester (PE). These products can be used as renewable fuels, additives, solvents and chemical building blocks.

Various catalysts have been evaluated for their use in the conversion of levulinic acid, including supported noble metal and base metal catalysts (Wright, W.R.H. et al. ChemSusChem 5 (2012) 1657- 1667).

WO -A- 2006/067171 relates to a process for the hydrogenation of a lactone, a carboxylic acid ester or a carboxylic acid, for example LA, in the presence of a bifunctional heterogeneous catalyst comprising (i) a zeolite or another strongly acidic component and (ii) a hydrogenating metal component. The Examples describe a method of preparing the catalyst using an incipient wetness

impregnation procedure, wherein zeolite or silica (as comparison) is impregnated with unspecified metal solutions, dried, and then calcined. In one of the examples ethyl levulinate is hydrogenated in the presence of various catalysts. The major reaction products in the experiments with the metal-zeolite catalysts (plus silica binder) are ethyl pentanoate and pentanoic acid. The experiments with the catalysts without the zeolite have GVL as major reaction product, but show considerably less catalytic activity.

Limited examples are available on beneficial effects of alloying in LA hydrogenation reactions. Wettstein, S.G. et al. showed that alloying Ru with Sn is beneficial for catalyst stability (Appl. Catal. B. 117 (2012) 321-329). Van den Brink, P.J. et al., on the other hand, described in WO-A-2006/067171 that a catalyst comprising an alloy of Pt with Ru or Re on a ΤΪΟ2 and Zr02 support did not improve the performance, but rather seemed to reduce the initial activity and enhanced deactivation for instance for ruthenium on a titania support.

There remains a need for a catalyst useful in catalysing a conversion of a levulinic acid source into another useful chemical compound, such as GVL, in particular a catalyst that is advantageous in one or more of the following aspects: catalytic activity, catalytic selectivity towards a specific useful compound of interest, catalytic productivity, robustness under reaction conditions (resistance to leaching or sintering).

In particular the inventors have found that a specific group of catalysts display a satisfactory activity, selectivity and/or productivity for a prolonged period of time in a method for converting a levulinic acid source into another useful compound, such as GVL. More in particular the inventors found that such catalysts are preparable by a specific method.

Accordingly, the invention relates to a method for preparing a chemical compound, comprising subjecting a levulinic acid source, preferably a compound selected from the group of levulinic acid, levulinic acid anhydride, levulinic acid salts, levulinic acid esters and levulinic acid amides, to a reduction reaction catalysed by a metal catalyst on an oxide support, wherein the metal catalyst on a support is obtainable by a metal-ion wet-impregnation method, preferably an anion-excess wet-impregnation method and wherein the metal comprises ruthenium.

Further, the invention relates to a method for preparing a chemical compound, comprising subjecting a levulinic acid source, preferably a compound selected from the group of levulinic acid, levulinic acid anhydrides, levulinic acid salts, levulinc acid esters and levulinic acid amides to a reduction reaction catalysed by a metal catalyst on an oxide support, wherein the metal is a metal alloy of ruthenium and at least one metal selected from the group of platinum and palladium, preferably a metal alloy of ruthenium and palladium.

Further, the invention relates to a metal catalyst on an oxide support, in particular a metal oxide support or a silica support, wherein the metal comprises ruthenium and wherein the catalyst is obtainable by an anion excess wet- impregnation method. Further, the invention relates to a metal catalyst on an oxide support, wherein the metal is a metal alloy of ruthenium and at least one metal selected from the group of platinum and palladium, preferably a metal alloy of ruthenium and palladium. In particular good results have been achieved with a metal alloy of ruthenium and palladium on an titanium oxide support.

Further, the invention relates to a method for preparing a metal catalyst on an oxide support (for use in a method) according to the invention, comprising

- providing a impregnation solution comprising a precursor for the metal catalyst - i.e. metal ions, including ruthenium ions, plus counter- anions for the metal ions (in an amount sufficient to maintain the electrochemical balance of the precursor) and an additional source of anions, i.e. the excess anions (together with counter-cations in an amount to maintain the electrochemical balance of the additional source of anions);

- providing an oxide support, in particular a metal oxide support or a silica support;

impregnating the oxide support with the impregnation solution;

- drying the oxide support impregnated with the impregnation solution; and,

- reducing said metal ions present in or on the impregnated oxide support, thereby forming the metal catalyst on the oxide support.

As illustrated in the Examples, a catalyst prepared according to the method of the invention has properties that are different from known catalysts, also if the known catalyst is made of the same catalytic metal and the same type of support material. A catalyst according to the invention is particularly

advantageous in that it has an improved catalytic activity, catalytic selectivity towards a specific useful compound of interest, especially GVL, catalytic

productivity or robustness (resistance to leaching or sintering).

Typically, the catalyst has a productivity of at least 10 molcvL . gmetai ^{" 1} . hr ¹ , preferably of at least 15 molcvL . gmetai ¹ . hr ¹ , more preferably of at least 25 molcvL . gmetai ¹ . hr ¹ , and even more preferably of at least 35 molcvL . gmetai ¹ . hr ¹ . In practice the productivity usually is less than 250 molcvL . gmetai ¹ . hr ¹ , preferably 150 molcvL . gmetai ¹ . hr ¹ or less, more preferably 75 molcvL . gmetai ¹ . hr ¹ or less, even more preferably 55 molcvL . gmetai ¹ . hr ¹ or less, in particular 25 molcvL . gmetai ^{" 1} . hr ¹ or less, more in particular 20 molcvL . gmetai ^"1 . hr ¹ or less . The productivity can be calculated by determining the moles of GVL produced (by means of GC analysis of samples taken from the reactor and quantification with an internal standard) as a function of time and divided by the grams of catalyst added to the reaction under the following conditions: reaction in liquid phase (dioxane), initially 10 wt.% LA (based on the weight of LA+dioxane), 1 wt.% (total) metal based on the weight of the metal catalyst including support, 473 K, 40 bar Lb.

The invention is in particular advantageous in that it allows the essentially complete (e.g. > 99 wt.%) conversion of an LA source, in particular LA, to a desired product, in particular GVL, without noticeable degradation of LA source, in particular without noticeable decarboxylation of LA or further hydrogenation of GVL.

Figure 1 shows a STEM - EDX mapping of a RuPd catalyst prepared according to the invention.

Figure 2 shows the production of GVL during the hydrogenation of LA using monometallic 1% Ru/Ti02 (Ru, ·) made with a metal ion impregnation method with anion excess and a bimetallic 1% Ru-Pd/Ti02 (RuPd, A) made with a metal ion impregnation method with anion excess.

Figure 3 shows a comparison of particle sizes for the fresh and spent Ru, Pd, Ru-Pd, catalysts made with anion excess using TEM. (Blank bar = fresh catalyst; Grey bar = spent catalyst after 1 run; gray, striped bar = spent catalyst after 3 consecutive runs)

Figure 4 shows the recyclability of bimetallic RuPd/Ti02 catalyst. Recyclability tests were performed at different LA conversion levels and GVL selectivities.

The term "or" as used herein is defined as "and/or" unless specified otherwise.

The term "a" or "an" as used herein is defined as "at least one" unless specified otherwise.

The phrase "the metal catalyst on the support" is abbreviated herein after as "the metal catalyst".

The term "metal" is generally used herein in a strict sense, namely to refer to the metallic form of one or more elements, unless specified or evident otherwise (e.g. when referring to metal ions or to a metal salt). Thus, the term "metal catalyst" is used for a catalyst that comprises at least one catalytically active metallic element (characteristically ruthenium). A specific form of a metal is an alloy.

When referring to a noun (e.g. a compound, an additive, etc.) in the singular, the plural is meant to be included.

For the purpose of clarity and a concise description features are described herein as part of the same or separate embodiments, however, it will be appreciated that the scope of the invention may include embodiments having combinations of all or some of the features described.

The term "substantial(ly)" or "essential(ly)" is generally used herein to indicate that it has the general character, appearance or function of that which is specified. When referring to a quantifiable feature, these terms are in particular used to indicate that it is for more than 50 %, in particular at least 75 %, more in particular at least 90 %, even more in particular at least 95 % of the maximum that feature.

The phrase "essentially consisting" of a substance (e.g. a metal) as used herein, in general means that other components than said substance are not detectible or present at a level that is generally considered an impurity. In particular, "essentially consisting of means for more that 98 wt.%, more in particular for more than 99 wt.%, more in particular for more than 99.5 wt.%.

As used herein, the term "catalytic activity" is defined (calculated) as the initial reaction rate per unit mass of catalyst. In practice this is determined by determining the (slope of concentration of LA-source vs time graph at short reaction times/time derivative of the LA-source concentration at short reaction times) divided by the grams of metal added to the reaction. A short reaction time depends on the reaction conditions and typically reflects the time wherein the first 10 mol.% or less of the initially present LA-source are converted.

As used herein, the phrase "catalytic selectivity towards a specific useful compound" is defined (calculated) as ratio of the molar amount of the reactant (LA-source) converted to said compound (e.g. GVL) relative to the total molar amount of converted reactant (LA-source). As used herein the term "catalytic productivity" is defined (calculated) as moles of desired product (e.g. GVL) per gram of metal per unit of time.

The term "impregnation" is also referred herein as "wet-impregnation".

Resistance to leaching can be determined by determining the concentration of metal in the liquid phase after set reaction time, e.g. an hour, a day or a week. The metal concentration can be determined by inductively coupled plasma mass spectrometry (ICP-MS) or inductively coupled plasma optical emission spectroscopy (ICP-OES).

The reaction conditions of the method for preparing the chemical compound from a levulinic acid source can be chosen dependent on the compound of interest, the levulinic acid source, the information disclosed herein, the cited prior art and references cited therein, common general knowledge and optionally a limited amount of testing.

Usually, the LA source is selected from the group of levulinic acid, levulinic acid anhydrides, levulinic acid salts, levulinc acid esters and levulinic acid amides. Angelica lactone is a preferred LA lactone. In particular, good results have been achieved with levulinic acid.

The reduction reaction for converting the LA source preferably is a hydrogenation reaction. Suitable reducing agents are preferably selected from the group of hydrogen, formic acid and formate salts, in particular a formate salt of formate and a monovalent cation, such as ammonium formate or sodium formate. Preferably GVL is formed in the hydrogenation reaction.

The reaction conditions for the hydrogenation reaction can be based on common general knowledge and the information provided in the present disclosure.

In particular, GVL is usually prepared at a temperature in the range of

25-250 °C, preferably at a temperature in the range of 120-230 °C, and more preferably at a temperature in the range of 180-220 °C.

In particular, GVL is usually prepared at a hydrogen pressure of 1-100 bar, preferably of 20-70 bar, in particular of 40-50 bar.

In case formic acid ("hydrogen formate") or a formate salt is used, the molar ratio of formate to the LA-source is typically at least about 1. The formate can be applied in any excess, but usually the ratio of formate to the LA-source is 1000 or less, in particular 100 or less, and more in particular 10 or less. The amount of catalyst can be chosen based on common general knowledge, depending on the type of reactor system.

The GVL prepared in accordance with the invention can be used in the preparation of another compound.

In a specific embodiment, the GVL is subjected to a an acid or base catalysed ring-opening reaction to produce a mixture of pentenoic acids.

In a specific embodiment, the GVL is subjected to an acid or base- catalysed ring-opening reaction in the presence of an alcohol to produce a mixture of alkyl pentenoates, preferably methyl pentenoates.

In a specific embodiment, the GVL is subjected to an acid or base- catalysed ring-opening in the presence of ammonia to produce a mixture of pentenenitriles. In these mixtures the double bond can be in the 2-, the 3- or the 4- position. Where applicable, the double bond may be cis or trans. These ring opening reactions can be advantageously performed in the gas phase or in the liquid phase. In the latter case the ring-opening reaction can advantageously be executed as a reactive distillation.

In a specific embodiment, the GVL is subjected to an acid or base- catalyzed reaction with an aldehyde to form an alpha-alkylidene-gamma- valerolactone. Preferably, the aldehyde used is formaldehyde which when used in the reaction forms alpha-methylene-gamma-valerolactone.

In a specific embodiment, the GVL is subjected to a reaction with ammonia or an N-alkylamine to form 5-methyl-pyrrolidinone or N- alkyl- 5-methyl- pyrrolidinone. Preferably, the N-alkylamine used is methylamine, which when used in the reaction forms N-methyl-5-methyl-pyrrolidinone.

In a specific embodiment, the GVL is hydrogenated to form 1,5- pentanediol.

In a specific embodiment, the GVL is hydrogenated to form 2- methyltetrahydrofuran.

In a specific embodiment, the GVL is hydrogenated to form pentanoic acid.

In a specific embodiment, the LA-source is converted into a salt of 4- hydroxypentanoic acid or 4-hydroxypentanamide. In a preferred embodiment, the invention relates to the use of gamma- valerolactone prepared in accordance with the invention in the production of a fuel; a monomer, in particular a monomer for the production of a poly amide; or a solvent. For instance GVL can be converted into methyltetrahydrofuran which is suitable for use as a solvent or a fuel.

Typically the catalyst of the invention comprises ruthenium and optionally one or more other metals.

In a specific embodiment, the metal essentially consists of ruthenium. Such a catalyst, in particular when obtained by an anion excess impregnation method, has been found to be advantageously suitable for converting an LA source to GVL due to its high catalytic activity and high selectivity towards GVL.

In another specific embodiment the metal catalyst comprises ruthenium and at least one other metal, preferably palladium or platinum. Ruthenium and the other metal(s) preferably form an alloy. A preferred alloy essentially consists of ruthenium and at least one of palladium and platinum. An alloy of ruthenium and palladium, in particular an alloy essentially consisting of ruthenium and palladium, has been found to be particularly advantageous due to its high catalytic activity and high selectivity toward GVL for a prolonged time.

The molar ratio of the total of the metals other than ruthenium (such as palladium or platinum) to ruthenium in this embodiment can be chosen within wide limits, usually in the range of 1:99 to 90: 10, in particular in the range of 5:95 to 80:20, and more in particular in the range of 20:80 to 70:30. If palladium is present, the ratio of palladium to ruthenium is preferably at least 10:90, in particular at least 30:70, and more in particular at least 40:60. In general, the higher the palladium content, the longer the catalyst maintains a satisfactory selectivity towards GVL.

The total metal concentration usually is in the range of 0.1-10 wt.%, in particular in the range of 0.5- 5 wt.%.

The metal catalyst usually contains at least 0.1 wt.% ruthenium, based on the total weight of metal(s) and support, preferably at least 0.3 wt.% ruthenium, in particular at least 0.5 wt.% ruthenium. The ruthenium content of the metal catalyst, based on the total weight of metal(s) and support usually is less than 10 wt.%, preferably 5 wt.% or less, in particular 2 wt.% or less. The support preferably is a non-acidic or weakly acidic oxide. A weakly acidic support is in particular a support having an isoelectric point at a

temperature of 25 °C (hereafter "IEP") of more than 1.5, in particular of more than 3.5, and more in particular 3.9 or more. The IEP of the support usually is 14 or less. For example, the highly acidic metal oxide WO3 typically has an IEP of 0.2- 0.5. S1O2 has a lower acidity, its IEP typically being in the range of 1.7-3.5. For Zr0 ₂ , the IEP typically is 4- 11, for Ti0 ₂ it typically is 3.9-8.2. Typically, the IEP of MgO is 12- 13. IEPs of many oxides are readily available in the art (Marek

Kosmulski, "Chemical Properties of Material Surfaces", Marcel Dekker, 2001).

Silica (S1O2) is a preferred non-metal oxide support for a metal catalyst of the invention. It has advantageous processing properties, such as pelletizing properties. In a specific embodiment, the supported metal catalyst is in the form of a pellet, and preferably comprises a metal catalyst on a silica support.

In particular, good results have been achieved with a metal oxide support. Titanium oxide and zirconium oxide are preferred metal oxide support materials. The oxide may be synthetic or a natural mineral. In an advantageous embodiment, the titanium oxide comprises anatase or rutile. In particular, good results have been achieved with a support comprising both anatase and rutile (70- 80 wt.% anatase and 30-20 wt.% rutile). Such a support is commercially available from Evonik (P25-Ti0 ₂ )

An example of a non-acidic support material is magnesium oxide.

The size of the support material is not critical. Usually, when preparing the catalyst on support a particulate support is provided, such as a powder. In an advantageous embodiment for the preparation of the catalyst, the particles are microp articles (particles typically having a size < 1 mm, in particular in the range of 1-200 micrometer). After the catalyst has been prepared, it may be used as such or may be shaped in a desired form, e.g. pelletized.

In an embodiment, the metal catalyst on the support (i.e. including support) has a BET surface, as determined by N2 physisorption of at least 25 m ² /g, in particular of at least 30 m ² /g. Preferably the BET surface is at least 40 m ² /g.

Usually, the BET surface is 800 m ² /g or less, in particular 400 m ² /g or less, more in particular 200 m ² /g or less. In an embodiment, the BET surface is 75 m ² /g or less, and in particular 60 m ² /g or less. In particular for a Ti02 support, BET advantageously is in the range of 20-75 m ² /g, in particular in the range of 40-60 m ² /g. In particular for a Zr02 support, BET advantageously is in the range of 50- 150 m ² /g, in particular in the range of 75- 125 m ² /g. In particular for a silica support, BET usually is in the range of 50-800 m ² /g, in particular in the range of 50-400 m ² /g.

In an advantageous embodiment, the metal on the support is present on the support surface in the form of nanop articles (particles with a size of typically less than 100 nm), and in particular in the form of metal clusters, deposited on the support. These metal clusters preferably have a particle size, as determined by TEM in the range of 0.5 to 10 nm.

It is further contemplated that - in an advantageous embodiment - a metal catalyst obtainable in accordance with the invention is characterisable by a relatively high abundance of metal clusters present on the support, more preferably clusters having a size of about 0.5 to 5 nm and/or a relatively low polydispersity with respect to the particle size distribution of the metal nano- particles. Without being bound by theory, it is contemplated that the excess anions - in particular chloride ions - stabilise relatively small ruthenium or ruthenium- palladium alloy nanop articles or clusters. Chloride is in particular considered favourable for realising a random alloy formation.

With respect to the metal catalyst, wherein the metal is an alloy, it is further contemplated that - in an advantageous embodiment - the metal-alloy is characterisable by an essentially homogeneous distribution of ruthenium and the other metal(s), in particular a random alloy structure, rather than e.g. a core-shell alloy which one would expect to perform differently. A homogeneous distribution and random alloy structure is determinable by XAFS or STEM, as illustrated by Figure 1. It has further been found that a metal catalyst wherein the metal is a metal alloy, in particular an alloy of ruthenium and palladium, the particle size stability of the metal particles is improved, compared to particles of a single metal, such as only ruthenium or only palladium.

In a method for preparing the metal catalyst on an oxide an a impregnation solution is provided comprising a precursor for the metal catalyst. The liquid phase is usually a highly polar solvent, such as an aqueous liquid. As used herein "highly polar means" having about the same polarity as water or a higher polarity than water.

The metal precursor, such as the ruthenium precursor respectively the palladium precursor, usually is a metal salt dissolved in water or an aqueous liquid. Examples of metal salts are halogen salts (chloride, fluoride, iodide and bromide) and organic acid salts. Good results have been achieved with a chloride salt. Chloride has been found in particular to contribute to providing a catalyst with good catalytic properties. Preferably the impregnation solution is essentially free (less than 1 wt.%) of nitrates and/or sulphates. The total concentration of the metal ions is usually in the range of 1 mg metal ions/mL to 100 mg metal ions 1/mL, preferably in the range of 5 mg metal ions/mL to 10 mg metal ions /mL.

For preparing a catalyst with an improved catalytic property with respect to the reduction of an LA source, the impregnation solution preferably comprises an additional source of anions, in addition to the anions from the salt serving as the source for the metal ions which serve as the precursor for the catalytic metal and may be selected from the group of acids and salts of anions with volatile bases, in particular from the group of HC1, organic acids, ammonium chloride and salts of ammonium and an organic acid. In particular good results have been achieved with a chloride, in particular hydrochloric acid. A method wherein such an impregnation solution is used is called an anion excess

impregnation method. The concentration of the acid or salt of the anions and a volatile base, preferably is 10- 100 mL of a 0.1-10 M solution, in particular 15-30 mL of a 0.1 to 5 M solution, per 0.5 to 5 gram support.

The impregnation may be carried out in a manner known per se.

The support is advantageously mixed with the impregnation solution by agitation, for instance by vigorous stirring, e.g. using a magnetic bar/stirrer set-up above 900 rpm, whereby the mixing of support and impregnation liquid is effected. The impregnation can be carried out at any temperature between the melting point and the boiling point of the liquid phase, and is advantageously carried out at ambient temperature, e.g. at about 25 °C. During or after mixing of the support with the impregnation solution, the temperature is preferably raised to about 50-90 °C. This will allow evaporation of the solvent at a desired evaporation rate. The drying may be carried out in a manner known per se. In particular good results have been achieved with a method wherein the slurry of support in the impregnation solution is dried in hot open air atmosphere typically at a

temperature in the range of 50-95 °C. The drying is preferably carried out whilst agitating the slurry, for instance by stirring or drying in a fluidized bed. The agitation is preferably vigorous enough to achieve intensive mixing of the contents of the slurry. The drying is usually continued until a visually dry product is obtained. The drying usually takes 24 hours or less, in particular 1-20 hours, more in particular 5-20 hours.

If desired, the dried product is ground, prior to reduction.

Advantageously, the dried impregnated oxide support is reduced without first having been subjected to a calcination step. Thus, in a preferred embodiment the catalyst is a non-calcined catalyst. As is generally known,

'reducing' means that the metal ions that are the precursor for the metal catalysts are reduced from the ionic state to the metallic state (oxidation state = 0). In the art, calcination is generally known as a heat treatment of a catalyst precursor in an oxidizing atmosphere (of variable composition) for variable amount of time and at a temperature sufficient to convert essentially all precursors into oxides, i.e. to remove essentially all anions except oxygen as well as any solvent molecules.

Reducing is preferably carried out using hydrogen, for instance 1-10 vol.% hydrogen in an inert gas, such as nitrogen, argon and/or helium. The reduction is preferably carried out at a temperature in the range of 250 °C to 600 °C. Preferably, the temperature is gradually raised until the desired maximum temperature is reached, in particular at a ramp-rate in the range of 1-5 °C/min. In particular within this range, an advantageous balance is reached between removing the anions, such as chloride ions, increasing the metal-support interaction and preventing the metal particles from sintering. A slower rate of ramping typically increases the dispersion of the metal particles within the catalyst.

The reduction is preferably carried out for a period of at least 1 h, in particular for a period of 2 h to 10 h.

The resultant metal catalyst on the metal oxide support is then allowed to cool. The invention is illustrated by the following examples. Example 1 Ru catalyst Preparation of the catalyst

RuC (Acros Chemicals) was used as the ruthenium precursor and was dissolved in deionized water to form an aqueous precursor solution with a ruthenium concentration of 5.28 mg / mL.

The requisite amount of the precursor solution (1.89 mL for preparing a 1 wt.% Ru catalyst on support) was charged in to a clean 50 mL round bottomed flask fitted with a magnetic stirrer, after which the requisite amount of

concentrated HC1 (37.5%) was added and the volume finally adjusted to 25 mL to obtain a final HC1 concentration of 0.5 M in deionized water. The round bottom flask was submerged in a temperature controlled oil bath and the mixture was then agitated at 298 K using a hot plate stirrer (1200 rpm). After 15 min of stirring, 0.99 gram of the support (P25-Ti02, obtained from Evonik) was added slowly with constant stirring at 298 K for about 30 min.

After the completion of the support addition, the slurry was stirred vigorously at 298 K for about 30 min and then the oil bath temperature was raised to 333 K, stirred for 1 h and then finally heated to 358 K. The slurry was stirred at this temperature overnight until all the water had evaporated. The solid powder, denoted as the "dried sample" was ground thoroughly with a mortar and pestle and approximately 350 mg of this material was then reduced in a furnace at 723 K (approx. 2 K / min ramp rate) under a flow of 5 vol.% Lb / N2 (total flow: 420 mL/min ) for 4 h, without having been subjected to a calcination step. Finally the furnace was cooled rapidly to room temperature and the catalyst sample was used without any further modification.

The resultant catalyst contained a 1 wt.% metal (Ru) concentration on a 1 g production scale.

Method of testing the catalyst

Reactions were performed with 10 wt.% levulinic acid (6.0 g, 51.7 mmol) in dioxane (54 g) with 1 wt.% catalyst (0.6 g). The reactions were run in a 100 mL Parr batch autoclave at a temperature of 473 K using a hydrogen pressure of 40 bar and a stirring speed of 1600 rpm. Before starting the reaction, the batch autoclave reactor was loaded with catalyst, substrate and solvent, purged three times with argon after which the reaction mixture was heated to reaction temperature and charged with Lb to 40 bar. 1 mL of solution was sampled at various intervals during the reaction. After the reaction, the autoclave was cooled to room temperature, the Lb was released and 2 wt.% anisole was added as an internal standard. The catalyst was separated by filtration and washed with acetone.

The reaction products were analyzed using a Shimadzu GC-2010A gas chromatograph equipped with a CP-WAX 57-CB column (25 m x 0.2 mm x 0.2 μιη) and FID detector. Products were identified with a GC-MS from Shimadzu with a CP-WAX 57CB column (30 m x 0.2 mm x 0.2 μπι). Results

The activity and selectivity of supported monometal Ru catalyst was found to be strongly influenced by both its preparation method and by alloying it with a second metal, such as Pd.

The 1 wt.% Ru/Ti02 catalyst prepared via the impregnation method according to the invention, using anion excess (as described herein above in these examples) proved to be very active when compared to a comparable 1 wt.% Ru/Ti02 catalysts prepared by standard wet impregnation (WI), as described in 'Luo, W. et al. Journal of Catalysis 301 (2013) 175-186'.

The Ru catalyst according to the invention gave full conversion after 40 min with 99.0 mol.% selectivity to GVL (0.1 mol.% MTHF and 0.1 mol.% PD as a byproduct); cf. 43.3 mol.% conversion after 40 min for the comparative Ru catalyst (WI), and 99.5 mol.% selectivity to GVL.

The productivity of the Ru/Ti02 catalyst according to the invention was 16.4 molcvL . g metai ¹ . hr ¹ , thus outperforming the catalysts listed in a recent LA hydrogenation review paper (Wright, W.R.H. et al. ChemSusChem 5 (2012) 1657- 1667). This Ru/Ti02 catalyst according to the invention was very active and gave excellent selectivity at short reaction times of 40 min. At longer reaction times, however, consecutive reactions took place causing a drop in selectivity and mass balance after 2 h (selectivity: 93.0 mol.% GVL, 1.2 mol.% PD, 0.7 mol.% MTHF).

The catalyst according to the invention showed very limited leaching (i.e. loss of 0.5 wt.% loss of the total amount of ruthenium originally present after 10 h in neat LA) and very limited sintering was observed.

Example 2

PdC salt (Sigma Aldrich) was dissolved in deionized water to form an aqueous precursor solution with a resultant palladium concentration of 3.02 mg / mL. This solution was slowly cooled and used as the palladium precursor solution. This solution was used together with the Ru precursor solution of Example 1 in amounts to provide a molar ratio of Ru:Pd of 1:1. The requisite amount of the precursor solution was charged in to a clean 50 mL round bottomed flask fitted with a magnetic stirrer, after which the requisite amount of concentrated HCl

(37.5%) was added and the volume finally adjusted to 25 mL to obtain a final HCl concentration of 0.5 M in deionized water. Method conditions were further the same as in Example 1. A RuPd on Ti02 catalyst was obtained, comprising 1 wt.% of the Ru-Pd alloy (molar ratio Ru:Pd 1: 1).

Results

The Ru-Pd metal alloy catalyst was found to have a high activity (i.e. full conversion after 30 min, 99.6 mol.% selectivity to GVL, 0.1 mol.% PD, 0.3 mol.% MTHF; note that this is at only 0.49 wt.% Ru metal loading), and stayed completely selective also at longer reaction times (1 h, sel. 99.2 mol.% GVL, 0.2 mol.% PD, 0.4 mol.% MTHF, 0.1 mol.% PA) The productivity of the RuPd/Ti0 ₂ catalyst was 17.2 molcvL . gmetai ¹ . hr ¹ .

This catalyst was also compared with a catalyst, prepared with the same method as described in Example 1, except that all Ru was replaced by Pd. This Pd/Ti02 catalyst shows negligible activity with respect to the conversion of levulinic acid. Characterization of the catalyst

Figure 1 shows a STEM - EDX mapping of a RuPd catalyst prepared according to Example 2 (molar ratio of Ru:Pd of 1: 1). Details about the mapping can be found in "Aberration Corrected Analytical Electron Microscopy Studies Of Bimetallic Nanopar tides", A.A. Herzing, M.Watanabe, C.J.Kiely, J.Edwards,

M.Conte, Z.R.Tang and G.J.Hutchings, Faraday Discussions 138 (2008) 337-351 or "Nanostructural and Chemical Characterization of Supported Metal Oxide

Catalysts by Aberration Corrected Analytical Electron Microscopy", W. Zhou, I. E. Wachs, C. J. Kiely, Current Opinion in Solid State and Materials Science 16 (2012) 10-22. The images are STEM high angle annular dark field (HAADF) images of the metallic particles obtained using an aberration corrected JEM ARM-200F STEM operating at 200 kV. X-ray energy dispersive (XEDS) spectra were acquired from individual nanoparticles larger than 1 nm in size by rastering the beam over the entire particle, while using a JEOL Centurio 0.9sr silicon drift detector.

Example 3

The performance of the RuPd catalyst of Example 2 and the Ru catalyst of Example 1 (both prepared with an metal ion impregnation method with anion excess of the invention) were compared.

Selectivity

Figure-2 shows that the GVL yield (A), was stable for several hours even after reaching quantitative conversion when the bimetallic catalyst was used. However, in the case of monometallic metallic 1 wt%Ru/Ti02 catalyst (·), the GVL yield started to drop after reaching a maximum of 99.0%. This is because of the further hydrogenation of the product GVL to 1,4-pentanediol (PD) and

methyltetrahydrofuran (MTHF).

This clearly demonstrates the advantages of catalysts preparable by the impregnation method with anion excess, and in particular of the bimetallic RuPd catalyst for suppressing the un-wanted consecutive reactions and thereby increasing the yield of the most wanted product (GVL). Particle Size, Stability and Reusability

Figure -3 shows the results from the TEM measurements of the metal particle sizes for the fresh (white bars) and spent (grey bars) monometallic (Ru or Pd) and bimetallic (Ru-Pd) catalysts, made with a method of the invention. From the figure it is evident that the metal particle sizes increased when the monometallic catalysts were used once. However, the particle sizes remained the same for the bimetallic catalysts even after 3 runs (grey, hatched bar). This shows that alloying Pd with Ru substantially helps in stabilizing the small particle for several cycles. This stability of the small particles, in the case of bimetallic catalysts, resulted in the catalytic activity of the recovered catalysts. Figure-4 shows the catalytic activity of the fresh and recovered bimetallic catalysts. The activity and selectivity (thus the yield of GVL) was stable for at least 3 catalytic runs. Thus, it is concluded that alloying Pd with Ru substantially influences the selectivity, stability and reusability of the catalyst for the hydrogenation of LA to GVL.