FIELD OF THE INVENTION

The invention relates to a process for producing a monohydric alcohol by hydrogenolysis of a polyhydric alcohol, a catalyst for use in the process, and a process for producing the catalyst.

BACKGROUND TO THE INVENTION

Lower alcohols such as methanol and ethanol are strategically important to many countries as they can be used as transportation fuels and as platform chemicals for synthesis. Currently, however, they are synthesized by inefficient processes. More than 80% of the world's energy consumption and production of chemicals originates from the use of fossil resources. This non-renewable route is unlikely to be viable for a long period of time with an ever increasing world population, reducing natural reserves and prevalent green house gas problems (D. E. Bloom, Science, 333, 562 (201 1); J. Tollefson, Nature, 473, 134 (2011); R. D. Cortright, R. R. Davda, J. A. Dumesic, Nature 418, 964 (2002); G. A. Olah, Angew Chem Int. Edit, 44, 2636, 2005). One trend of the global economy is to depart from the dependence on fossil fuels to renewable resources (J. Tollefson, Nature, 473, 134, 2011). Thus, there is an urgent need for research in the chemistry of biomass to unlock its energy content and to open up a wider usage of renewable biomass as a greener energy resource in the future (R. D. Cortright, R. R. Davda, J. A. Dumesic, Nature 418, 964, 2002). Catalytic production of lower alcohols such as methanol and ethanol from biomass is very important as these high energy density liquids can be used as transportation fuels with good volatility (for combustion engines or fuel cell devices) and key platform chemicals in synthesis (G. A. Olah, Angew Chem Int. Edit, 44, 2636, 2005). Currently, methanol is synthesised at large scale by an indirect, energy-inefficient process which requires an initial, thermally intensive step to break down biomass completely into CO and ¾, before these components are then re-assembled to produce the methanol molecule (ex- ICI process). Similarly, ethanol is manufactured by fermentation. Although the

fermentation process is selective, being a bioprocess it has a limited temperature regime and is associated with low productivity.

There is therefore an ongoing need for more energy-efficient processes for producing lower alcohols such as methanol, ethanol and other monohydric alcohols, from biomass and biomass-derived polyols like ethylene glycol and glycerol. From an energetic consideration it would be ideal to synthesize such alcohols directly from such polyols without the need for a stepwise process involving re-assembling CO and H ₂ to produce the alcohol in question. However, this would require highly cooperative catalysis. A key challenge would be to identify an appropriate catalytic surface for concerted C-C cleavage followed by C-H formation to form methanol from the polyol, preferably without conversion of the adsorbed carbonaceous intermediate moieties into thermodynamically more stable compounds such as methane and carbon oxides. This is counterintuitive, though, as such highly reactive carbonaceous intermediates are expected to be prone to further reaction to form the thermodynamically more stable products. Similarly, concerted C-0 bond breakage and C-H formation would be required to synthesize ethanol and higher monohydric alcohols from such polyols. A key challenge is therefore to identify a compound capable of catalysing such reactions with a view to providing energy-efficient processes for producing lower alcohols from biomass.

SUMMARY OF THE INVENTION

It is a finding of the present invention that, by using a particular catalyst, monohydric alcohols such as methanol and ethanol can be formed selectively by direct catalytic hydrogenolysis of polyols. The heterogeneous catalyst, which can be prepared by co-precipitation from solution, contains a highly-dispersed metal phase and an iron oxide support, which can achieve the desired concerted C-C and/or C-0 bond breakage and C-H bond formation. The catalyst can be used to form monohydric alcohols such as methanol and ethanol directly from biomass and biomass-derived polyols. It can also catalyse formation of monohydric alcohols from polyols made via oil or coal related technology. For instance, ethylene glycol made on a petrochemical site, e.g. by steam cracking naphtha, can subsequently be converted into methanol and/or ethanol by direct catalytic

hydrogenolysis using the catalyst of the invention. Thus, the catalyst of the invention not only provides a route for producing lower alcohols from biomass, but it also provides petrochemical sites with more versatility in terms of the products that they can produce.

Accordingly, the invention provides a process for producing a monohydric alcohol by hydrogenolysis of a polyhydric alcohol, which process comprises treating a polyhydric alcohol with hydrogen in the presence of a solid catalyst, which solid catalyst comprises (a) iron oxide and (b) a metal which is a noble metal or nickel. Typically, the metal is a noble metal. In preferred embodiments, the noble metal is palladium. Due to a unique metal-support interaction, the palladium in the preferred catalyst is highly dispersed on the iron oxide support and remains so even after heat treatments and after use of the catalyst. The high dispersion, which can ranges from small Pd-containing nanoparticles (i.e. clusters) down to individual Pd atoms, is thought to be responsible for a highly cooperative catalysis, which selectively promotes C-C or C-0 cleavage with simultaneous C-H formation, via spillover hydrogen from the active palladium-containing nanoparticles. Also, the close proximity of the metal nanoparticles and atoms to the iron oxide surface is thought to assist polyol adsorption on the catalyst surface for the cooperative catalysis. Polyol can be converted to methanol and ethanol over this surface with a combined selectivity of greater than 80%.

Accordingly, the invention further provides a catalyst which comprises (a) iron oxide and (b) nanoparticles, which nanoparticles comprise palladium. The mean particle size of said nanoparticles is usually less than 5 nm, and more typically less than 3 nm, and usually the particle size distribution of the nanoparticles is such that at least 90 % of the nanoparticles have a particle size of less than 5 nm. Generally the catalyst also comprises individual atoms of the palladium metal dispersed on the surface of the iron oxide. These have been observed by high angle annular dark field (HAADF) scanning transmission electron microscopy (STEM).

The catalyst of the invention may be prepared by a co-precipitation process, and particularly useful catalysts have been obtained in this way.

Accordingly, the invention further provides a process for producing a catalyst, which catalyst comprises iron oxide and nanoparticles, which nanoparticles comprise palladium, which process comprises:

(1) a co-precipitation step, comprising contacting (a) a solution, which solution comprises a palladium salt and an iron salt dissolved in a solvent, with (b) a base, to produce a precipitate, which precipitate comprises one or more compounds comprising said iron and said palladium;

(2) a separation step, comprising separating the precipitate from the solvent; and

(3) a calcination step, comprising calcining the precipitate by heating the precipitate in air.

Typically, the process further comprises: (4) a reduction step, comprising heating the calcined precipitate in the presence of ¾. The solvent used in step (1) is typically water. The precipitate produced in step (1) generally comprises one or more hydroxide or oxide-hydroxide compounds comprising Pd and Fe. The Pd atoms are usually atomically dispersed and bonded within a sol-gel like matrix, formed during the co-precipitation step, comprising an Fe-O-Fe network. In the matrix, iron and palladium atoms are generally connected to one other via bridging oxygen atoms. Typically, therefore, the precipitate produced in step (1) comprises a polymeric network comprising Pd atoms and Fe atoms bonded together via bridging oxygen atoms. A schematic representation of an example of such a network is shown in Fig. 1. The formation of such a polymeric network, throughout which the Pd atoms are dispersed, facilitates the production of final catalysts in which both individual Pd atoms and small Pd- containing nanoparticles are finely dispersed on the iron oxide surface.

The invention further provides a catalyst which is obtainable by the process of the invention for producing a catalyst.

Further provided is the use of a catalyst which is obtainable by the process of the invention for producing a catalyst, as a catalyst for the hydrogenolysis of a polyhydric alcohol.

BRIEF DESCRIPTION OF THE FIGURES

Fig. 1 is a schematic representation of the type of polymeric network which could be formed during the co-precipitation step (1) of the process of the invention for producing a catalyst. The co-precipitated network comprises Pd and Fe atoms bonded together via bridging oxygen atoms. The envisaged networks of -Pd-O-Fe- are thought to be created by hydrolysis and cross-condensation between -PdO- and -FeO- species.

Fig. 2 shows the performances of catalysts containing 5 wt % Pd supported on different oxides (prepared by co-precipitation) for the hydrogenolysis of ethylene glycol. For each catalyst, the % conversion of ethylene glycol is shown, as is the % selectivity for the various products obtained.

Fig. 3 shows a comparison of 5 wt % Rh and 5 wt % Pd respectively on iron oxide, for the hydrogenolysis of ethylene glycol.

Fig. 4 shows the change of gross selectivity over 5 wt % Pd on Fe ₂ 03 versus time in the batch- wise hydrogenolysis of ethylene glycol.

Fig. 5 shows time-fraction selectivity over 5 wt % Pd on Fe ₂ 0 ₃ at different periods in the batch-wise hydrogenolysis of ethylene glycol. Fig. 6 shows the catalytic yields of methanol and ethanol obtained over fresh and reused 5 wt % Pd/Fe ₂ 0 ₃ catalysts in the hydrogenolysis of ethylene glycol under different temperatures.

Fig. 7 is a histogram of the particle size distribution of the metal nanoparticles on the surface of one example of a Pd/Fe ₂ 0 ₃ catalyst of the invention, as measured by high- resolution transmission electron microscopy (HRTEM). As can be seen from the histogram, the dark field HRTEM images showed a large number of particles in the range ca. 1.0-2.5 nm.

Fig. 8a shows XRD patterns of the 5 wt% Pd/Fe ₂ 0 ₃ catalyst before (lower trace) and after (upper trace) hydrogenolysis of ethylene glycol at 195°C, 20 bar H ₂ (RT).

Fig. 8b (insert b in Fig. 8) shows an enlarged region of the XRD, showing two small peaks in the trace for the catalyst after the hydrogenolysis reaction, corresponding to PdFe alloy rather than Pd.

Fig. 9 shows core level Fe 2p ₃ / ₂ spectra of Pd/Fe ₂ 0 ₃ catalysts with different alcohol selectivities due to different hydrogen pre-reduction times (lh, 5h, 48h) after reaction at 195°C for 24h. The right-hand dashed line corresponds to the binding energy of Fe 2p _3/2 and the left-hand dashed line corresponds to Fe 2p _3/2 in PdFe alloy (literature values).

Fig. 10 shows core level Pd 3d _5/2 spectra of Pd/Fe ₂ 0 ₃ catalysts with different alcohols selectivities due to different hydrogen pre-reduction times (lh, 5h, 48h) after reaction at 195°C for 24h. The dashed line on the right hand side corresponds to the binding energy of Pd 3d sn and the left-hand darker dashed line corresponds to Pd 3d / ₂ in PdFe alloy (literature values).

Fig. 1 1 shows H ₂ -TPR profiles of Fe ₂ 0 ₃ (peaks at ~350°C and ~600°C) and Pd/Fe ₂ 0 ₃ (peaks at ~120°C and ~600°C).

Fig. 12 shows 1 ^st and 2 ^nd repeated TPR profiles of Pd/Fe ₂ 0 ₃ in H ₂ after the temperature programmed heating terminated at 200°C, as compared to the 2 ^nd TPR profile of Pd/Al ₂ 0 ₃ .

Fig. 13 shows the rate of methanol production as a function of ethylene glycol concentration, over the 5 wt% Pd/Fe ₂ 0 ₃ catalyst of the invention.

Fig. 14 shows the rate of methanol production as a function of H ₂ pressure, over the 5 wt% Pd/Fe ₂ 0 ₃ catalyst of the invention. Fig. 15 shows a plot of ln(k / L mol ^"1 hr ^"1 ) (y axis) versus 1/(T/K) (x axis) for the hydrogenolysis of ethylene glycol at 20 bar ¾, for 24 hrs, at temperatures of 428, 448, 468 and 488 K, over the 5 wt% Pd/Fe ₂ 0 ₃ catalyst of the invention.

DETAILED DESCRIPTION OF THE INVENTION

In the process of the invention for producing a monohydric alcohol by

hydrogenolysis of a polyhydric alcohol, the polyhydric alcohol is treated with hydrogen in the presence of a solid catalyst, which solid catalyst comprises iron oxide and a metal which is a noble metal or nickel.

The catalyst is a hydrogenolysis catalyst. The term "hydrogenolysis catalyst", as used herein, means a catalyst which is capable of catalysing a hydrogenolysis reaction.

The term "monohydric alcohol" as used herein, takes its normal meaning in the art, i.e. an alcohol with a single -OH group. Examples of monohydric alcohols are methanol, ethanol, propanol, butanol and pentanol. Monohydric alcohols which can be produced by the process of the invention include compounds of formula R-OH, wherein R is a C _{1 -1 0} alkyl group. More typically, R is a C _{1 -6} alkyl group. R may for instance be a C alkyl group. In a preferred embodiment, for instance, R is ethyl or methyl. Thus, in a preferred embodiment, the process of the invention is for producing a monohydric alcohol of formula R-OH, wherein R is methyl or ethyl. The monohydric alcohol is typically therefore methanol or ethanol. Of course, more than one monohydric alcohol may be produced by the process of the invention, i.e. a mixture of products may be produced, which may comprise more than one monohydric alcohol. A mixture of ethanol and methanol may for instance be produced.

As used herein, a C O alkyl group is an unsubstituted, straight or branched chain saturated hydrocarbon radical having from 1 to 10 carbon atoms. For example, a CMO alkyl group may be methyl, ethyl, propyl, butyl, pentyl, hexyl, heptyl, octyl, nonyl or decyl. CMO alkyl groups include C _{1 -6} alkyl groups, for example methyl, ethyl, propyl, butyl, pentyl or hexyl. C _1-10 alkyl groups also include C _{1 -4} alkyl groups, for example methyl, ethyl, i-propyl, n-propyl, t-butyl, s-butyl or n-butyl.

The term "polyhydric alcohol" as used herein, also takes its normal meaning in the art, i.e. an alcohol with two or more -OH groups. A polyhydric alcohol may also be referred to as a "polyol". The terms "polyhydric alcohol" and "polyol" are used herein interchangeably. Examples of polyhydric alcohols include sugar alcohols, for instance ethylene glycol, glycerol and longer-chain su ar alcohols of formula (I) below:

in which n is 0 or an integer equal to or greater than 1.

Sugars are also polyhydric alcohols. For instance, polyhydric alcohols include monosaccharides, such as glucose, and other com ounds of formula (II) below:

in which m is an integer equal to or greater than 1.

Polyhydric alcohols also incude disaccharides, for instance sucrose, as well as oligosaccharides and polysaccharides. Examples of polyhydric alcohols which are polysaccharides include, for instance, starch, amylose, amylopectin, glycogen, cellulose, pectin, chitin, laminarin, chrysolaminarin, xylan, arabinoxylan, mannan, fucoidan or galactomannan.

Other examples of polyhydric alcohols which are polymers include polyethers and polyesters. Polyethers and polyesters often comprise multiple terminal OH groups.

As the skilled person will appreciate, the term "noble metal" is used in chemistry to refer to a particular group of transition metals which have outstanding resistance to corrosion, namely the second and third row transition metals of groups 8 to 11 of the periodic table. Thus, the term "noble metal", as used herein, means a metal selected from the group consisting of ruthenium, rhodium, palladium, silver, osmium, iridium, platinum and gold.

The metal, which is a noble metal or nickel, in the solid catalyst used in the process of the invention is generally supported on the iron oxide. Thus, the solid catalyst typically comprises a support material which comprises said iron oxide, wherein the metal is supported on said support material. In other words, the catalyst usually comprises a metal, which is a noble metal or nickel, supported on iron oxide.

It is thought that the metal, which is a noble metal or nickel, is predominantly present in the catalyst in metallic form, i.e. in the oxidation state 0. Typically, therefore, the solid catalyst comprises: (a) iron oxide; and (b) a metal, which is a noble metal or nickel, in the oxidation state 0. The solid catalyst may, in this embodiment, further comprise said metal in an oxidation state other than 0, for instance in an oxidation state greater than 0, for instance an oxidation state of from +1 to +4.

The metal, which is a noble metal or nickel, is typically present in the form of nanoparticles, supported on the iron oxide.

Accordingly, the catalyst used in the process of the invention for producing a monohydric alcohol typically comprises: (a) said iron oxide; and (b) nanoparticles which comprise said metal, which metal is a noble metal or nickel.

As used herein the term "nanoparticle" means a microscopic particle whose size is measured in nanometres (nm). Typically, a nanoparticle has a particle size of from 0.2 nm to 1000 nm, for instance from 0.5 nm to 1000 nm. A nanoparticle may be crystalline or amorphous. A nanoparticle may be spherical or non-spherical. Non-spherical nanoparticles may for instance be plate-shaped, needle-shaped or tubular. The term "particle size" as used herein means the diameter of the particle if the particle is spherical or, if the particle is non-spherical, the volume-based particle size. The volume-based particle size is the diameter of the sphere that has the same volume as the non-spherical particle in question.

The nanoparticles which comprise said metal are also referred to herein as clusters.

Generally the iron oxide acts as a support material. Typically, therefore, the nanoparticles which comprise said metal are supported on the iron oxide. The nanoparticles are typically present on the surface of the iron oxide.

It is thought that the metal in the nanoparticles, which is a noble metal or nickel, is predominantly in metallic form, i.e. in the oxidation state 0. Typically, therefore, the nanoparticles comprise said metal in the oxidation state 0. The nanoparticles may however further comprise said metal in an oxidation state other than 0, for instance in an oxidation state greater than 0, for instance an oxidation state of from +1 to +4.

In some embodiments, the nanoparticles consist of said metal, which metal is a noble metal or nickel.

In other embodiments, however, some or all of said nanoparticles which comprise said metal further comprise iron. In one embodiment, only some of the nanoparticles further comprise iron. Thus, the catalyst may comprise nanoparticles which consist of said metal (which is a noble metal or nickel), and nanoparticles which comprise iron and said metal. In another embodiment, all, or substantially all, of the nanoparticles further comprise iron. The nanoparticles may for instance consist of said metal (which is a noble metal or nickel) and iron.

The iron present in the nanoparticles may be present in the nanoparticles in metallic form, i.e. in the oxidation state 0.

The iron present in the nanoparticles is usually present in the form of an alloy with the noble metal. Thus, typically, said nanoparticles which further comprise iron comprise an alloy of said noble metal and iron.

Accordingly, in one embodiment, some or all of said nanoparticles comprise an alloy of the noble metal and iron. In one embodiment, only some of the nanoparticles comprise an alloy of the noble metal and iron. Thus, the catalyst may comprise

nanoparticles which consist of said noble metal, and nanoparticles which comprise an alloy of the noble metal and iron. In another embodiment, all, or substantially all, of the nanoparticles comprise an alloy of the noble metal and iron. In some embodiments, the nanoparticles consist of said alloy of the noble metal and iron.

The nanoparticles which further comprise iron are typically bimetallic

nanoparticles.

Thus, typically, the nanoparticles are bimetallic nanoparticles which comprise (i) said metal which is a noble metal or nickel, and (ii) iron.

Generally, the mean particle size of the nanoparticles in the solid catalyst which is used in the process of the invention is less than 50 nm. More typically, however, the mean particle size of the nanoparticles is less than 20 nm; even more typically, it is less than 10 nm. Thus, for instance, the mean particle size of the nanoparticles in the solid catalyst may be from 1 nm to 50 nm, or for instance from 1 nm to 20 nm. It could for instance be from 1 nm to 10 nm. In some embodiments, however, the mean particle size of the nanoparticles is equal to or less than 1 nm.

Usually, the particle size distribution of said nanoparticles is such that at least 90 % of the nanoparticles have a particle size of less than 50 nm. More typically, at least 90 % of the nanoparticles have a particle size of less than 20 nm, or for instance less than 10 nm.

In a preferred embodiment, the mean particle size of the nanoparticles in the solid catalyst used in the process of the invention is less than 5 nm. The mean particle size of said nanoparticles may for instance be less than 3 nm.

The mean particle size of the nanoparticles in the solid catalyst may for instance be from 1 nm to 5 nm, or for instance from 1 nm to 3 nm. In a particularly preferred embodiment, the mean particle size of the nanoparticles in the solid catalyst used in the process of the invention is from 1.0 nm to 2.5 nm. For instance, the mean particle size of said nanoparticles may be about 1.5 nm.

Preferably, the particle size distribution of the nanoparticles in the solid catalyst used in the process of the invention is such that at least 90 % of the nanoparticles have a particle size of less than 5 nm. For instance, the particle size distribution of said nanoparticles may be such that at least 90 % of the nanoparticles have a particle size of less than 3.5 nm. Such a distribution is shown in Fig. 7.

In addition to nanoparticles, the metal, which is a noble metal or nickel, may also be present in the solid catalyst in the form of individual metal atoms. Typically, the individual metal atoms are on the surface of the iron oxide. Thus, the metal, which is a noble metal or nickel, may be present in the solid catalyst which is used in the process of the invention both in the form of nanoparticles and in the form of single metal atoms dispersed on the iron oxide surface.

Typically, therefore, the solid catalyst further comprises individual atoms of said metal. Usually, said individual atoms are dispersed on the surface of the iron oxide.

In one embodiment of the solid catalyst which is used in the process of the invention, the mean particle size of said nanoparticles which comprise said metal, which is a noble metal or nickel, is less than 4 nm, the particle size distribution of said nanoparticles is such that at least 90 % of the nanoparticles have a particle size of less than 5 nm, and the catalyst further comprises individual atoms of said metal dispersed on the surface of the iron oxide.

Usually, the metal is a noble metal. Typically, therefore, in the process of the invention for producing a monohydric alcohol by hydrogenolysis of a polyhydric alcohol, the polyhydric alcohol is treated with hydrogen in the presence of a solid catalyst, which solid catalyst comprises iron oxide and a noble metal. Usually, the noble metal is selected from palladium and rhodium.

In other embodiments however the metal may be selected from palladium, rhodium and nickel.

Usually, the noble metal is palladium or rhodium.

Typically, therefore, in the process of the invention for producing a monohydric alcohol by hydrogenolysis of a polyhydric alcohol, the polyhydric alcohol is treated with hydrogen in the presence of a solid catalyst, which solid catalyst comprises iron oxide and a noble metal selected from palladium and rhodium.

In a particularly preferred embodiment, the noble metal is palladium.

Preferably, therefore, in the process of the invention for producing a monohydric alcohol by hydrogenolysis of a polyhydric alcohol, the polyhydric alcohol is treated with hydrogen in the presence of a solid catalyst, which solid catalyst comprises iron oxide and palladium.

Typically, the metal is palladium and the catalyst comprises (a) said iron oxide and (b) nanoparticles comprising said palladium.

As mentioned above, the iron oxide can act as a support material. Typically, therefore, the nanoparticles which comprise said palladium are supported on the iron oxide. The palladium nanoparticles are typically present on the surface of the iron oxide.

It is thought that the palladium in the nanoparticles is predominantly in metallic form, i.e. in the oxidation state 0. Typically, therefore, the nanoparticles comprise said palladium in the oxidation state 0. The nanoparticles may however further comprise palladium in an oxidation state other than 0, for instance in an oxidation state greater than 0, for instance an oxidation state of from +1 to +4. Oxidation states greater than 0 for palladium include for instance +1, +2 and +4.

In some embodiments, the nanoparticles consist of said palladium.

In other embodiments, however, some or all of said nanoparticles which comprise said palladium further comprise iron. In one embodiment, only some of the nanoparticles further comprise iron. Thus, the catalyst may comprise nanoparticles which consist of palladium, and nanoparticles which comprise iron and palladium. In another embodiment, all, or substantially all, of the nanoparticles further comprise iron. The nanoparticles may for instance consist of palladium and iron.

The iron present in the nanoparticles may be present in the nanoparticles in metallic form, i.e. in the oxidation state 0.

The iron present in the nanoparticles is usually present in the form of an alloy with the palladium. Thus, typically, said nanoparticles which further comprise iron comprise an alloy of palladium and iron.

Accordingly, in one embodiment, some or all of said nanoparticles comprise an alloy of palladium and iron. In one embodiment, only some of the nanoparticles comprise an alloy of palladium and iron. Thus, the catalyst may comprise nanoparticles which consist of palladium, and nanoparticles which comprise an alloy of palladium and iron. In another embodiment, all, or substantially all, of the nanoparticles comprise an alloy of palladium and iron. In some embodiments, the nanoparticles consist of said alloy of palladium and iron.

The nanoparticles which further comprise iron are typically bimetallic

nanoparticles.

Thus, typically, the nanoparticles in the catalyst used in the process of the invention are bimetallic nanoparticles which comprise (i) said palladium, and (ii) iron.

Typically, the mean particle size of the nanoparticles which comprise palladium is less than 20 nm; even more typically, it is less than 10 nm. Thus, for instance, the mean particle size of the nanoparticles in the solid catalyst which comprise palladium may be within the range of from 1 nm to 20 nm, or for instance within the range of from 1 nm to 10 nm. It could for instance be from 1 nm to 7 nm. In some embodiments, however, the mean particle size of the nanoparticles is equal to or less than 1 nm.

Usually, the particle size distribution of said nanoparticles which comprise palladium is such that at least 90 % of the nanoparticles have a particle size of less than 20 nm. More typically, at least 90 % of the nanoparticles have a particle size of less than 10 nm, or for instance less than 7 nm.

In a preferred embodiment, the mean particle size of the nanoparticles which comprise palladium is less than 5 nm. The mean particle size of said nanoparticles may for instance be less than 3 nm, or for instance less than 1 nm.

The mean particle size of the nanoparticles which comprise palladium may for instance be from 1 nm to 5 nm, or for instance from 1 nm to 3 nm.

In a particularly preferred embodiment, the mean particle size of the nanoparticles which comprise palladium is from 1.0 nm to 2.5 nm. For instance, the mean particle size of said nanoparticles may be about 1.5 nm.

Preferably, the particle size distribution of the nanoparticles which comprise palladium in the solid catalyst used in the process of the invention is such that at least 90 % of the nanoparticles have a particle size of less than 5 nm. For instance, the particle size distribution of said nanoparticles may be such that at least 90 % of the nanoparticles have a particle size of less than 3.5 nm. Such a distribution is shown in Fig. 7. In some

embodiments, however, the particle size distribution of said nanoparticles may be such that at least 90 % of the nanoparticles have a particle size of equal to or less than 1 nm. In addition to nanoparticles, the palladium may also be present in the solid catalyst in the form of individual palladium atoms. Typically, the individual palladium atoms are on the surface of the iron oxide. Thus, the solid catalyst which is used in the process of the invention may comprise palladium both in the form of said nanoparticles which comprise palladium, and in the form of single palladium atoms dispersed on the iron oxide surface.

In one embodiment, the mean particle size of said nanoparticles which comprise palladium is less than 5 nm, the particle size distribution of said nanoparticles is such that at least 90 % of the nanoparticles have a particle size of less than 10 nm, and the catalyst further comprises individual atoms of palladium dispersed on the surface of the iron oxide.

In one embodiment, the mean particle size of said nanoparticles which comprise palladium is less than 4 nm, the particle size distribution of said nanoparticles is such that at least 90 % of the nanoparticles have a particle size of less than 5 nm, and the catalyst further comprises individual atoms of palladium dispersed on the surface of the iron oxide.

Typically, the iron oxide in the solid catalyst used in the process of the invention for producing a monohydric alcohol, comprises Fe ₂ 0 ₃ . Fe ₂ 0 ₃ is typically formed during a calcination step in the preparation of the catalyst. However, the iron oxide in the solid catalyst may further comprise other, more reduced forms of iron oxide, particularly if the catalyst has undergone a pre-reduction step during the preparation of the catalyst, or for instance if the catalyst is being re-used after having been exposed to hydrogen already in a hydrogenolysis process (such as the process of the invention for producing a monohydric alcohol by hydrogenolysis of a polyol). It is known that the reduction of bulk iron oxide by hydrogen proceeds through the following steps:

Fe ₂ 0 ₃ -» Fe ₃ 0 ₄ - FeO -» Fe

Accordingly, the iron oxide in the solid catalyst used in the process of the invention typically comprises Fe ₂ 0 ₃ . It may however further comprise Fe ₃ 0 ₄ or FeO (which may be present due to partial reduction of Fe ₂ 0 ₃ ) or a mixture of Fe ₃ 0 ₄ and FeO, in addition to the Fe ₂ 0 ₃ .

The solid catalyst used in the process of the invention may further comprise Fe, i.e. iron metal. Such Fe may for instance be present in the form of an alloy with the noble metal or nickel in the nanoparticles, as explained above. Alternatively, for instance, iron metal may be present in the catalyst in the form of nanoparticles consisting of iron metal.

Typically, the iron oxide in the catalyst used in the process of the invention has a surface which has undergone reduction. Such surface reduction of iron oxide may for instance have occurred during a pre -reduction step during the preparation of the catalyst, or for instance if the catalyst is being re-used after having been exposed to hydrogen already in a hydrogeno lysis process (such as the process of the invention for producing a monohydric alcohol). It is thought that the reduction of the surface of the iron oxide (with the noble metal or nickel in close proximity) will give nanoparticles comprising (a) Fe and (b) the noble metal or nickel (for instance Fe and palladium). These small but catalytically active clusters are thought to offer much enhanced catalyst activity and selectivity due to electronic modification of band structure. In addition, oxygen vacancies will be formed in close vicinity to the Fe/noble metal or Fe/Ni nanoparticles on the defective iron oxide surface due to the maintenance of electrical neutrality upon reduction (their presence may stabilize the clusters), which may further assist adsorption of the ethylene glycol molecule

Accordingly, the iron oxide in the solid catalyst which is used in the process of the invention for producing a monohydric alcohol typically comprises oxygen vacancies. Typically, the oxygen vacancies are at or near the surface of the iron oxide.

Typically, the metal (which is a noble metal or nickel) is present in the solid catalyst in an amount which is equal to or less than 50 weight %, based on the weight of the iron oxide. More typically, the metal is present in an amount which is equal to or less than 30 weight %, based on the weight of the iron oxide. The metal may for instance be present in an amount which is equal to or less than 20 weight %, or for instance which is equal to or less than 10 weight %, based on the weight of the iron oxide. In some embodiments, for instance, the metal is present in an amount which is equal to or less than 8 weight %, based on the weight of the iron oxide.

Typically, the metal (which is a noble metal or nickel) is present in the solid catalyst in an amount of from 0.1 weight % to 50 weight %, based on the weight of the iron oxide. More typically, the metal is present in an amount of from 0.5 weight % to 30 weight %, based on the weight of the iron oxide. The metal may for instance be present in an amount of from 1 weight % to 20 weight %, or for instance from 2 weight % to 10 weight %. In some embodiments, for instance, the metal is present in an amount of from 2 weight % to 8 weight %, based on the weight of the iron oxide, for instance from 3 weight % to 7 weight %.

Typically, the metal is a noble metal.

Typically, the noble metal is rhodium or palladium.

More typically, it is palladium. Preferably, for instance, the catalyst used in the process of the invention for producing a monohydric alcohol is a catalyst which comprises iron oxide and

nanoparticles, which nanoparticles comprise palladium, and which catalyst is obtainable by a process of the invention as defined herein for producing such a catalyst.

The iron oxide in the catalyst used in the process of the invention may be doped with a dopant element. It is a finding of the invention that doping the iron oxide support of the catalyst with a dopant element can lead to further improvements in the activity and/or selectivity of the catalyst, when used in the process of the invention. The dopant element is a metal other than iron, for instance a transition metal other than iron. The dopant element, when present, is present in the catalyst in addition to said noble metal or nickel. Thus, the dopant element, when present, is typically other than said metal which is a noble metal or nickel. (That is to say, when the metal is a noble metal, the dopant is typically a metal other than iron and other than said noble metal, and when the metal is nickel, the dopant is typically a metal other than iron and other than nickel.)

The dopant element may for instance be a first row transition metal other than iron. Thus, the dopant element may be selected from Sc, Ti, V, Cr, Mn, Co, Ni and Cu. Usually, the dopant is selected from Sc, Ti, V, Cr, Mn, Co and Cu. More typically, it is selected from V, Cr, Mn, Co and Cu. Most preferably, the dopant element is cobalt (Co). Such dopant elements, and especially cobalt, may provide further improvements in the activity and selectivity of the catalyst.

Accordingly, the iron oxide in the catalyst used in the process of the invention may or may not be doped with a dopant element. When the iron oxide is doped, the dopant element is preferably cobalt. Thus, in some embodiments, the catalyst used in the process of the invention comprises (a) cobalt-doped iron oxide and (b) said metal which is a noble metal or nickel. The iron oxide may alternatively, of course, not be doped with a dopant element, i.e. the iron oxide may be undoped. Catalysts in which the iron oxide is not doped with a dopant element (i.e. in which the iron oxide is "undoped") are described in further detail hereinbelow, in the Example.

Advantageously, the catalysts of the invention can be used to produce monohydric alcohols from of a wide range of polyols, ranging from sugar alcohols such as ethylene glycol and glycerol, sugars, including polysaccharides such as starch, and other polymeric polyols, for instance polyester polyols and polyether polyols.

Typically, therefore, the polyhydric alcohol is a sugar alcohol, a sugar or a polymer. When the polyhydric alcohol is a sugar alcohol, it is usually a sugar alcohol of formula (I)

in which n is 0 or an integer equal to or greater than 1. The variable n may for instance be 0 or an integer of from 1 to 5,000. For example, n may for instance be 0 or an integer of from 1 to 1 ,000, or for example n may be 0 or an integer of from 1 to 100. More typically, n is 0 or an integer from 1 to 8.

In one embodiment, the sugar alcohol of formula (I), from which a monohydric alcohol may be produced in accordance with the process of the invention has from 2 to 10 carbon atoms. Thus, typically n is 0 or an integer from 1 to 8. Examples of sugar alcohols of formula I) include, but are not limited to, the following compounds:

sorbitol mannitol

In one embodiment, n is 0 or an integer of 1 to 6. In this embodiment, the sugar alcohol is selected from ethylene glycol (C2), glycerol (C3), butane- 1 ,2, 3, 4-tetrol (C4), pentane-l ,2,3,4,5-pentol (C5), hexane-l,2,3,4,5,6-hexol (C6), heptane-l ,2,3,4,5,6,7-heptol (C7), octane-l,2,3,4,5,6,7,8-octol (C8). More typically, n is 0, 1, 2, 3 or 4. In this embodiment, the sugar alcohol is selected from ethylene glycol (C2), glycerol (C3), butane- 1,2,3, 4-tetrol (C4), pentane- 1,2, 3,4,5- pentol (C5) and hexane- 1,2,3, 4,5, 6-hexol (C6).

Even more typically, n is 0, 1 or 2 and the sugar alcohol is selected from ethylene glycol (C2), glycerol (C3) and butane- 1,2,3, 4-tetrol (C4).

Usually, however, n is 0 or 1, and the sugar alcohol is ethylene glycol or glycerol.

In some preferred embodiments, n is 0 and the sugar alcohol is ethylene glycol.

When the polyhydric alcohol is a sugar, it may for instance be a sugar which comprises a compound of formula II)

in which m is an integer equal to or greater than 1.

In one embodiment, the sugar of formula (II), from which a monohydric alcohol may be produced in accordance with the process of the invention, has from 3 to 10 carbon atoms. Thus, typically m is an integer from 1 to 8. Examples of sugars of formula (II) include the following compounds:

D-Erythrose D-Threose

D-Ribose D-Arabinose D-Xylose D-Lyxose

D-Allose D-Altrose D-Glucose D-Mannose D-Gulose D-Iodose D-Galactose D-Talose

The polyhydric alcohol may alternatively be a sugar which is a disaccharide, for instance sucrose, or an oligosaccharide, for instance a sugar with from 3 to 10 saccharide units.

Alternatively, the polyhydric alcohol may be a sugar which is also a polymer, i.e. a polysaccharide. The polysaccharide may for instance comprise starch, amylose, amylopectin, glycogen, cellulose, pectin, chitin, laminarin, chrysolaminarin, xylan, arabinoxylan, mannan, fucoidan or galactomannan, or a mixture or one or more thereof.

Polyhydric alcohols also include other kinds of polymers for instance polyethers and polyesters.

Accordingly, in one embodiment, the polyhydric alcohol is a polymer which is a polysaccharide, a polyether or a polyester.

Most typically, however, the polyhydric alcohol is ethylene glycol or glycerol. In one embodiment, the polyhydric alcohol is ethylene glycol.

In another embodiment, the polyhydric alcohol is glycerol. Monohydric alcohols which can be produced by the process of the invention include compounds of formula ROH, wherein R is a C _1-10 alkyl group. Typically, therefore, the process of the invention is a process for producing a monohydric alcohol of formula ROH, wherein R is a CMO alkyl group. More typically, though, R is a Ci _-6 alkyl group. Even more typically, R is methyl or ethyl.

In a preferred embodiment, therefore, the process of the invention is a process for producing a monohydric alcohol of formula ROH, wherein R is selected from the group consisting of methyl and ethyl. Thus, preferably, the monohydric alcohol is methanol or ethanol.

Of course, more than one monohydric alcohol may be produced by the process of the invention, i.e. a mixture of products may be produced by the process of the invention which mixture may comprise more than one monohydric alcohol.

In one embodiment, therefore, the process is for producing a mixture comprising said monohydric alcohol (which may be termed a first monohydric alcohol) and a second monohydric alcohol, by hydrogenolysis of said polyhydric alcohol. The first monohydric alcohol may for instance be methanol and the second monohydric alcohol may for example be ethanol. The process of the invention for producing a monohydric alcohol may therefore be a process for producing methanol and ethanol.

In one embodiment, the process is for producing methanol. In this embodiment, any other monohydric alcohol that is produced by the process may be considered to be a byproduct, and the process may further comprise a purification step to separate that byproduct from the desired monohydric alcohol, methanol.

In another embodiment, the process is for producing ethanol. Again, in this embodiment, any other monohydric alcohol produced by the process may be considered to be a by-product, and the process may further comprise a purification step to separate the by-product from the desired monohydric alcohol, ethanol.

Preferably, the polyhydric alcohol is glycerol or ethylene glycol, and the

monohydric alcohol is methanol or ethanol.

In one preferred embodiment, the polyhydric alcohol is ethylene glycol, and the monohydric alcohol is methanol or ethanol.

In another preferred embodiment, the polyhydric alcohol is glycerol, and the monohydric alcohol is methanol or ethanol. In the process of the invention, the step of treating the polyhydric alcohol with hydrogen is carried out in the presence or absence of a solvent, typically in the presence of a solvent. The solvent, when present, should be an inert solvent. The term "inert solvent", as used herein, means a solvent which does not itself undergo hydrogenolysis,

hydrogenation or any other chemical conversion under the reaction conditions of the process of the present invention. Thus, the solvent cannot be an organic compound which would undergo C-C bond cleavage and hydrogenolysis during the present process.

Accordingly, the solvent, when present, is typically an inorganic solvent. Usually, the solvent is water.

The source of the hydrogen used in the process of the invention may be a solid or liquid hydrogen storage material. The solid or liquid hydrogen storage material may be any suitable hydrogen storage material that is capable of providing or releasing hydrogen in a form which is suitable for hydrogenolysis. Typically, the hydrogen is provided in the form of molecular hydrogen. Thus, the hydrogen storage material may be a material which is capable of releasing molecular hydrogen, for instance hydrogen gas. Additionally or alternatively however the hydrogen may be provided by the hydrogen storage material in another reactive form, for instance in the form of single hydrogen atoms (hydrogen atom radicals) or hydride anions.

The hydrogen used in the process of the invention may be provided by such a hydrogen storage material in situ.

Accordingly, the step of treating said polyhydric alcohol with hydrogen may comprise treating the polyhydric alcohol with a solid or liquid hydrogen storage material.

The step of treating said polyhydric alcohol with hydrogen may for instance comprise treating the polyhydric alcohol with a solid or liquid hydrogen storage material and generating the hydrogen in situ from said hydrogen storage material.

The process of the invention may comprise generating said hydrogen from a solid or liquid hydrogen storage material.

In some embodiments, the process of the invention comprises generating said hydrogen in situ from a solid or liquid hydrogen storage material.

The hydrogen storage material may be a compound which comprises hydrogen, i.e. a material in which hydrogen is stored chemically. Alternatively, it may be a material in which hydrogen is physically stored, for instance a solid onto which hydrogen is adsorbed or a liquid in which hydrogen is dissolved, and from which hydrogen can be released. The hydrogen storage material may for instance comprise a chemical hydride, such as for instance a metal hydride; a chemical borohydride, for instance lithium borohydride; a protic solvent, for instance an alcohol such as isopropyl alcohol; a carbohydrate; a hydrocarbon; ammonia; an amine borane complex; formic acid; an imidazolium ionic liquid;

phosphonium borate; nanoporous carbon; graphene; fullerene; carbon nanotubes; a porous metal-organic framework (MOF). Such hydrogen storage materials are known in the art.

Usually, however, the hydrogen used in the process of the invention for producing a monohydric alcohol is molecular hydrogen. Typically, it is hydrogen gas.

Thus, the step of treating said polyhydric alcohol with hydrogen usually comprises treating said polyhydric alcohol with ¾.

The step of treating said polyhydric alcohol with hydrogen usually comprises treating said polyhydric alcohol with hydrogen gas.

Thus, usually, the process of the invention for producing a monohydric alcohol comprises treating said polyhydric alcohol with hydrogen gas.

Typically, the step of treating said polyhydric alcohol with hydrogen is carried out at a temperature of at least 50 °C.

The hydrogenolysis reaction is typically however carried out at relatively mild temperatures and pressures.

Accordingly, the step of treating the polyhydric alcohol with hydrogen is typically performed at a temperature of 250 °C (523 K) or less than 250 °C (473 K), and more typically at a temperature of 220 °C or less. In one embodiment, the step of treating the polyhydric alcohol with hydrogen is carried out at a temperature of 200 °C or less.

In one embodiment, the step of treating the polyhydric alcohol with hydrogen is carried out at a temperature of from 50 °C to 250 °C, or for instance from 100 °C to 250 °C, from 100 °C to 220 °C, or for example from 150 °C to 200 °C.

The hydrogen pressure was found typically to affect the selectivity of the hydrogenolysis reaction for methanol, and was found to affect the conversion of the polyhydric alcohol to methanol.

Accordingly, the step of treating the polyhydric alcohol with hydrogen is usually performed at a hydrogen pressure of at least 1 bar, typically at least 12 bar, more typically at a hydrogen pressure of at least 15 bar, and even more typically at a hydrogen pressure of at least 18 bar. Typically, the hydrogen pressure employed is from 1 bar to 250 bar, more typically from 12 bar to 250 bar, and even more typically from 15 bar to 250 bar, for instance from 18 bar to 250 bar. In one embodiment, the hydrogen pressure employed is about 20 bar.

The process of the invention may be carried out as continuous process or a batch process. In one embodiment, the process is a continuous process in which hydrogen is cofed in a recycled catalyst bed, either in the presence or absence of a solvent.

In the case of a batch process, the step of treating the polyhydric alcohol with hydrogen is usually performed for longer than 5 hours, and more typically for at least 10 hours, in order to ensure that a relatively high percent conversion of the reactant is achieved. Even more typically, the reaction time is at least 15 hours, for instance for about 15 hours or for about 24 hours. In some embodiments, the step of treating the polyhydric alcohol with hydrogen is performed for longer than 24 hours, for instance for at least 72 hours. As shown in Fig. 5, the selectivity for methanol and ethanol was found to increase when the catalyst was used for such longer durations.

Typically, the monohydric alcohol produced by the process of the invention is recovered from the reaction mixture. Typically, the reaction mixture comprises the monohydric alcohol product, the catalyst and, when the reaction is carried out in the presence of a solvent, the solvent. The reaction mixture usually also comprises unreacted starting material, i.e. unreacted polyhydric alcohol. The reaction mixture may also contain by-products (i.e. reaction products other than the monohydric alcohol of interest, which may for instance include other monohydric alcohols and other polyhydric alcohols) and/or impurities.

Thus, the step of recovering the monohydric alcohol from the reaction mixture typically involves separation of the monohydric alcohol from the catalyst, from any unreacted starting material and, when solvent is present, from the solvent. Typically, when a by-product or impurity is present in the reaction mixture, the monohydric alcohol is also separated from that by-product or impurity.

Accordingly, in one embodiment, the process of the invention further comprises the step of recovering said monohydric alcohol.

The monohydric alcohol produced by the process of the invention may be used as a fuel, typically an automotive fuel, or as a building block, reactant or feedstock in the production of other chemicals. The monohydric alcohol may for instance be used to esterify unwanted free fatty acids present in the production of biodiesel. Thus, the process of the invention can be used to convert the glycerol by-product of a biodiesel production process into methanol and/or ethanol, and that methanol and/or ethanol can in turn be used to esterify free fatty acids present in the biodiesel production process in order to produce further biodiesel.

Accordingly, in one embodiment, the process of the invention further comprises esterifying a fatty acid with the monohydric alcohol thus produced. Typically, the fatty acid is from a biodiesel production process and the esterification of the fatty acid with the monohydric alcohol produces further biodiesel. Typically, in this embodiment, the monohydric alcohol is methanol or ethanol and the compound which is converted to the monohydric alcohol in the first place is glycerol, which glycerol is a by-product of the same biodiesel production process.

Certain catalysts used in the process of the invention for producing a monohydric alcohol are themselves novel. Accordingly, the invention further provides a catalyst which comprises (a) iron oxide and (b) nanoparticles which comprise palladium.

The iron oxide in the catalyst of the invention can act as a support material.

Typically, therefore, in the catalyst of the invention the nanoparticles which comprise said palladium are supported on the iron oxide.

The palladium nanoparticles are typically present on the surface of the iron oxide.

It is thought that the palladium in the nanoparticles of the catalyst of the invention is predominantly in metallic form, i.e. in the oxidation state 0. Typically, therefore, the nanoparticles comprise said palladium in the oxidation state 0. The nanoparticles may however further comprise palladium in an oxidation state other than 0, for instance in an oxidation state greater than 0, for instance an oxidation state of from +1 to +4. Oxidation states greater than 0 for palladium include for instance +1, +2 and +4.

In some embodiments, the nanoparticles consist of said palladium.

In other embodiments, some or all of said nanoparticles which comprise said palladium further comprise iron. Thus, the catalyst of the invention may for instance comprise nanoparticles which consist of palladium, and nanoparticles which comprise iron and palladium. In another embodiment, all, or substantially all, of the nanoparticles further comprise iron. The nanoparticles which further comprise iron may for instance consist of palladium and iron.

The iron present in the nanoparticles may be present in the nanoparticles in metallic form, i.e. in the oxidation state 0. The iron present in the nanoparticles is usually present in the form of an alloy with the palladium. Thus, typically, said nanoparticles which further comprise iron comprise an alloy of palladium and iron.

Accordingly, in one embodiment, some or all of said nanoparticles comprise an alloy of palladium and iron. In one embodiment, only some of the nanoparticles comprise an alloy of palladium and iron. Thus, the catalyst may comprise nanoparticles which consist of palladium, and nanoparticles which comprise an alloy of palladium and iron. In another embodiment, all, or substantially all, of the nanoparticles comprise an alloy of palladium and iron. The nanoparticles which comprise an alloy of palladium and iron may in some embodiments consist only of said alloy of palladium and iron.

The nanoparticles which further comprise iron are typically bimetallic

nanoparticles.

Thus, typically, the nanoparticles in the catalyst of the invention are bimetallic nanoparticles which comprise (i) said palladium, and (ii) iron.

Typically, the mean particle size of the nanoparticles in the catalyst of the invention is less than 50 nm. Usually for instance, the mean particle size of the nanoparticles in the catalyst of the invention is less than 20 nm. Even more typically, the mean particle size is less than 10 nm.

Thus, for instance, the mean particle size of the nanoparticles in the catalyst of the invention may be within the range of from 1 nm to 20 nm, or for instance within the range of from 1 nm to 10 nm. It could for instance be from 1 nm to 7 nm. In some embodiments, however, the mean particle size of the nanoparticles is equal to or less than 1 nm.

Usually, the particle size distribution of the nanoparticles in the catalyst of the invention is such that at least 90 % of the nanoparticles have a particle size of less than 20 nm. More typically, at least 90 % of the nanoparticles have a particle size of less than 10 nm, or for instance less than 7 nm.

In a preferred embodiment, the mean particle size of the nanoparticles in the catalyst of the invention is less than 5 nm. The mean particle size of said nanoparticles may for instance be less than 3 nm, or for instance less than 1 nm.

The mean particle size of the nanoparticles in the catalyst of the invention may for instance be from 1 nm to 5 nm, or for instance from 1 nm to 3 nm. In a particularly preferred embodiment, the mean particle size of the nanoparticles in the catalyst of the invention is from 1.0 nm to 2.5 nm. For instance, the mean particle size of said nanoparticles may be about 1.5 nm.

Preferably, the particle size distribution of the nanoparticles in the catalyst of the invention is such that at least 90 % of the nanoparticles have a particle size of less than 5 nm. For instance, the particle size distribution of said nanoparticles may be such that at least 90 % of the nanoparticles have a particle size of less than 3.5 nm. Such a distribution is shown in Fig. 7. In some embodiments, however, the particle size distribution of said nanoparticles may be such that at least 90 % of the nanoparticles have a particle size of equal to or less than 1 nm.

In one embodiment of the catalyst of the invention, the mean particle size of said nanoparticles is less than 3 nm and the particle size distribution of said nanoparticles is such that at least 90 % of the nanoparticles have a particle size of less than 5 nm. For instance, the mean particle size of said nanoparticles may be less than 3 nm and the particle size distribution of said nanoparticles may be such that at least 90 % of the nanoparticles have a particle size of less than 3.5 nm.

In addition to nanoparticles, the catalyst of the invention may further comprise palladium in the form of individual palladium atoms. Typically, the individual palladium atoms are on the surface of the iron oxide. Thus, the catalyst of the invention may comprise palladium both in the form of said nanoparticles which comprise palladium, and in the form of single palladium atoms dispersed on the iron oxide surface.

In one embodiment, the mean particle size of said nanoparticles which comprise palladium is less than 5 nm, the particle size distribution of said nanoparticles is such that at least 90 % of the nanoparticles have a particle size of less than 10 nm, and the catalyst of the invention further comprises individual atoms of palladium dispersed on the surface of the iron oxide.

Typically, the mean particle size of said nanoparticles which comprise palladium is less than 4 nm, the particle size distribution of said nanoparticles is such that at least 90 % of the nanoparticles have a particle size of less than 5 nm, and the catalyst of the invention further comprises individual atoms of palladium dispersed on the surface of the iron oxide.

In one embodiment of the catalyst of the invention, the mean particle size of the nanoparticles is less than 3 nm, the particle size distribution of said nanoparticles is such that at least 90 % of the nanoparticles have a particle size of less than 4 nm, and the catalyst further comprises individual atoms of palladium dispersed on the surface of the iron oxide.

The mean particle size of the nanoparticles is in some embodiments less than or equal to 1 nm, and the particle size distribution of said nanoparticles may be such that at least 90 % of the nanoparticles have a particle size of less than 4 nm, and the catalyst further comprises individual atoms of palladium dispersed on the surface of the iron oxide.Typically, the iron oxide in the catalyst of the invention comprises Fe ₂ 0 ₃ . Fe ₂ 0 ₃ is typically formed during a calcination step in the preparation of the catalyst. However, the iron oxide in the solid catalyst may further comprise other, more reduced forms of iron oxide, particularly if the catalyst has undergone a pre-reduction step during the preparation of the catalyst, or for instance if the catalyst has already been exposed to hydrogen in a hydrogenolysis process (such as the process of the invention for producing a monohydric alcohol by hydrogenolysis of a polyol).

Accordingly, the iron oxide in the catalyst of the invention typically comprises Fe ₂ 0 ₃ . It may however further comprise Fe ₃ 0 ₄ or FeO (which may be present due to partial reduction of Fe ₂ 0 ₃ ) or a mixture of Fe ₃ 0 ₄ and FeO, in addition to the Fe ₂ 0 ₃ .

The catalyst of the invention may further comprise Fe(0), i.e. iron metal. Such Fe may for instance be present in the form of an alloy with palladium in the nanoparticles, as discussed hereinbefore. Alternatively, for instance, iron metal may be present in the catalyst in the form of nanoparticles consisting only of iron metal.

Typically, the iron oxide in the catalyst of the invention has a surface which has undergone reduction. Such surface reduction of iron oxide may for instance have occurred during a pre-reduction step during the preparation of the catalyst, or for instance if the catalyst has been already exposed to hydrogen in a hydrogenolysis process (such as the process of the invention for producing a monohydric alcohol). It is thought that the reduction of the surface of the iron oxide (with the palladium in close proximity) will cause the formation of nanoparticles comprising Fe and and palladium. These small but catalytically active nanoparticles are thought to offer much enhanced catalyst activity and selectivity due to electronic modification of band structure. In addition, oxygen vacancies will be formed in the close vicinity of the Fe/Pd nanoparticles on the defective iron oxide surface due to the maintenance of electrical neutrality upon reduction. The presence of oxygen vacancies may stabilize the clusters, and may further assist adsorption of the polyhydric alcohol molecule. Accordingly, the iron oxide in the catalyst of the invention typically comprises oxygen vacancies. Typically, the oxygen vacancies are at or near the surface of the iron oxide.

Typically, the palladium is present in the catalyst of the invention in an amount which is equal to or less than 50 weight %, based on the weight of the iron oxide. More typically, the palladium is present in an amount which is equal to or less than 30 weight %, based on the weight of the iron oxide. The palladium may for instance be present in an amount which is equal to or less than 20 weight %, or for instance which is equal to or less than 10 weight %, based on the weight of the iron oxide. In some embodiments, for instance, the palladium is present in an amount which is equal to or less than 8 weight %, based on the weight of the iron oxide.

Typically, the palladium is present in the catalyst of the invention in an amount of from 0.1 weight % to 50 weight %, based on the weight of the iron oxide. More typically, the palladium is present in an amount of from 0.5 weight % to 30 weight %, based on the weight of the iron oxide. The palladium may for instance be present in an amount of from 1 weight % to 20 weight %, or for instance from 2 weight % to 10 weight %. In some embodiments, for instance, the palladium is present in an amount of from 2 weight % to 8 weight %, based on the weight of the iron oxide, for instance from 3 weight % to 7 weight %.

The iron oxide in the catalyst of the invention may be doped with a dopant element. It is a finding of the invention that doping the iron oxide support of the catalyst with a dopant element can lead to further improvements in the activity and/or selectivity of the catalyst, when used in the process of the invention. The dopant element is a metal other than iron, for instance a transition metal other than iron. The dopant element, when present, is present in the catalyst in addition to the palladium. Thus, the dopant element, when present, is typically other than palladium. (That is to say, the dopant is typically a metal other than iron and other than palladium.)

The dopant element may for instance be a first row transition metal other than iron. Thus, the dopant element may be selected from Sc, Ti, V, Cr, Mn, Co, Ni and Cu. Usually, the dopant is selected from Sc, Ti, V, Cr, Mn, Co and Cu. More typically, it is selected from V, Cr, Mn, Co and Cu. Most preferably, the dopant element is cobalt (Co). Such dopant elements, and especially cobalt, may provide further improvements in the activity and selectivity of the catalyst. Accordingly, the iron oxide in the catalyst of the invention may or may not be doped with a dopant element. When the iron oxide is doped with a dopant element, the dopant element is preferably cobalt. Thus, in some embodiments, the catalyst used in the process of the invention comprises (a) cobalt-doped iron oxide and (b) palladium. The iron oxide may alternatively, of course, not be doped with a dopant element, i.e. the iron oxide may be undoped. Catalysts in which the iron oxide is not doped with a dopant element (i.e. in which the iron oxide is "undoped") are described in further detail hereinbelow, in the Example.

The invention further provides a catalyst of the invention as defined herein which is obtainable by a process of the invention as defined herein for producing such a catalyst.

The invention further provides the use of a catalyst as defined herein as a catalyst for the hydrogenolysis of a polyhydric alcohol.

The catalysts of the invention, which comprise (a) iron oxide and (b) nanoparticles which comprise palladium, may be produced by the process of the invention for producing a catalyst, which process comprises:

(1) a co-precipitation step, comprising contacting (a) a solution, which solution comprises a palladium salt and an iron salt dissolved in a solvent, with (b) a base, to produce a precipitate which comprises one or more compounds comprising said iron and said palladium; (2) a separation step, comprising separating the precipitate from the solvent; and (3) a calcination step, comprising calcining the precipitate by heating the precipitate in air. Usually the process further comprises: (4) a reduction step, comprising heating the calcined precipitate in the presence of H ₂ .

As the skilled person will appreciate, the percentage by weight of palladium in the catalyst can be accurately controlled by varying the proportions of the palladium and iron salts employed in the solution used in step (1). Thus, any of the catalysts of the invention defined above, having any of the abovementioned percentages by weight of palladium based on the mass of iron oxide in the catalyst, can be produced by dissolving the correct amounts of palladium and iron salts in the solution used in step (1).

Typically, therefore, the ratio of palladium to iron in said solution is selected to produce a catalyst wherein the palladium is present in an amount which is equal to or less than 50 weight %, based on the weight of the iron oxide. More typically, the ratio of palladium to iron in said solution is selected to produce a catalyst wherein the palladium is present in an amount which is equal to or less than 30 weight %, based on the weight of the iron oxide. The ratio of palladium to iron in said solution may for instance be selected to produce a catalyst wherein the palladium is present in an amount which is equal to or less than 20 weight %, or for instance which is equal to or less than 10 weight %, based on the weight of the iron oxide. In some embodiments, for instance, the ratio of palladium to iron in said solution is selected to produce a catalyst wherein the palladium is present in an amount which is equal to or less than 8 weight %, based on the weight of the iron oxide.

Typically, the ratio of palladium to iron in said solution is selected to produce a catalyst wherein the palladium is present in the catalyst of the invention in an amount of from 0.1 weight % to 50 weight %, based on the weight of the iron oxide. More typically, the ratio of palladium to iron in said solution is selected to produce a catalyst in which the palladium is present in an amount of from 0.5 weight % to 30 weight %, based on the weight of the iron oxide. The ratio of palladium to iron in said solution may for instance be selected to produce a catalyst wherein the palladium is present in an amount of from 1 weight % to 20 weight %, or for instance from 2 weight % to 10 weight %, based on the weight of the iron oxide. In some embodiments, for instance, the ratio of palladium to iron in said solution is selected to produce a catalyst wherein the palladium is present in an amount of from 2 weight % to 8 weight %, based on the weight of the iron oxide, or for instance from 3 weight % to 7 weight %.

Usually, the solvent used in the co-precipitation step comprises water.

Typically, therefore, the co-precipitation step (1) comprises contacting (a) a solution, which solution comprises a palladium salt and an iron salt dissolved in a solvent which comprises water, with (b) a base, to produce a precipitate which comprises one or more compounds comprising said iron and said palladium.

The palladium salt may be any suitable salt which is soluble in the solvent. The solvent is typically water. Typically, therefore, the palladium salt is a water soluble palladium salt. It is typically a water-soluble palladium (II) salt. An example of a water soluble palladium salt is palladium nitrate. Thus, the palladium salt may for instance be palladium nitrate.

Similarly, the iron salt may be any suitable salt which is soluble in the solvent. The solvent is typically water. Typically, therefore, the iron salt is a water soluble iron salt. It is typically a water-soluble iron (III) salt. An example of a water soluble iron salt is iron nitrate. Thus, the iron salt may for instance be iron nitrate. Typically, it is iron (III) nitrate. Typically, the co-precipitation step comprises contacting: (a) said solution, which is an aqueous solution, with (b) a second aqueous solution which comprises said base. The contacting may be performed whilst stirring.

Any suitable base may be used in the process of the invention for producing a catalyst. Usually, the base is a metal carbonate. Typically, the metal carbonate is an alkali metal carbonate, for instance sodium carbonate.

The precipitate produced in step (1) generally comprises one or more hydroxide or oxide-hydroxide compounds comprising Pd and Fe. The Pd atoms are usually atomically dispersed and bonded within a sol-gel like matrix, formed during the co-precipitation step, comprising an Fe-O-Fe network. In the matrix, iron and palladium atoms are generally connected to one other via bridging oxygen atoms. Typically, therefore, the precipitate produced in step (1) comprises a polymeric network comprising Pd atoms and Fe atoms bonded together via bridging oxygen atoms. A schematic representation of an example of such a network is shown in Fig. 1. The formation of such a polymeric network, throughout which the Pd atoms are dispersed, is thought to facilitate the production of final catalysts in which both individual Pd atoms and small Pd-containing nanoparticles are finely dispersed on the iron oxide surface.

Typically, therefore, the precipitate produced in step (1) comprises one or more hydroxide or oxide-hydroxide compounds comprising said iron and said palladium. The precipitate may for instance comprise a mixed hydroxide of iron and palladium, or a mixed oxide-hydroxide of iron and palladium. Alternatively, for instance, the precipitate may comprise a mixture of palladium hydroxide and iron hydroxide, a mixture of palladium hydroxide and iron oxide-hydroxide, or any combination of these.

Thus, the precipitate produced in step (1) typically comprises one or more hydroxide or oxide-hydroxide compounds comprising said iron and said palladium.

The precipitate produced in step (1) may for instance comprise one or more hydroxide compounds comprising said iron and said palladium.

Usually, the precipitate produced in step (1) comprises a polymeric network comprising Pd atoms and Fe atoms bonded together via oxygen atoms.

Usually, in step (1) of the process, the step of contacting said solution with said base comprises increasing the pH of the solution from a first pH to a second pH, wherein the second pH is greater than the first pH. Typically, the second pH is at least 8. The first pH is typically less than 8, for instance from 5 to 7.5, or for instance equal to or less than 7. More typically the second pH is at least 8.5, for instance at least 9. The second pH is typically for instance from 8.0 to 12.0, for example from 8.5 to 10.0, for instance about 9.0.

The co-precipitation step (1) may further comprise an aging step. The aging step may comprise allowing the precipitate to remain in contact with a solvent for a period of time. Typically the solvent is the solvent from which it was precipitated (usually water). The period of time may for instance be at least 1 hour, for instance at least 5 hours, or for instance up to about 24 hours. The aging step may or may not comprise heating the co- precipitate in the presence of said solvent. For instance, the co-precipitate may be heated to a temperature of up to about 80 °C, or for instance up to about 90 °C, during said period of time. The co-precipitate may be heated at the temperature for up to about 24 hours.

Usually, however, the aging step does not comprise heating.

In the separation step (2), any suitable means can be used to separate the precipitate from solution. For instance, the separation may be performed by filtration or by

centrifugation. Typically, the separation step further comprises washing the precipitate, after separating the precipitate from solution. Typically, the precipitate is washed with deionised water. The separation step may additionally further comprise drying the precipitate. The precipitate is typically dried at a temperature of equal to or greater than 70 °C, e.g. at a temperature of from 80 to 120 °C. It is typically dried at the temperature for a number of hours, e.g. for 4 hours or more. It is typically dried at the temperature for 8 to 16 hours. The precipitate is usually dried in air.

The calcination step (3) typically comprises heating the precipitate in air to a temperature of at least 200 °C. More typically, the precipitate is heated in air to a temperature of at least 250 °C, or for instance to a temperature of at least 280 °C. Usually, the precipitate is heated in air to a temperature of about 300 °C.

Typically, in the calcination step, the precipitate is heated in air at the temperature for at least 1 hour, more typically for at least 1.5 hours, for instance for about 2 hours. The co-precipitate is typically heated to the temperature in static air.

The reduction step (4) comprises heating the calcined precipitate in the presence of H ₂ . Typically, the reduction step comprises heating the calcined precipitate in the presence of a mixture of H ₂ and an inert gas, such as N ₂ , and more typically under a flowing stream of H ₂ and the inert gas. The calcined product is typically heated in the presence of said H ₂ for at least 1 hour. The calcined precipitate may be heated in the presence of said ¾ to a temperature of at least 120 °C. More typically, the reduction step (4) comprises heating the calcined product in the presence of said ¾ to a temperature of at least 150 °C, or, for instance, to a temperature of at least 170 °C.

The calcined product is typically heated in the presence of said H ₂ at said temperature for at least 1 hour.

The catalyst produced by the process of the invention for producing a catalyst which comprises (a) iron oxide and (b) nanoparticles which comprise palladium, may be as further defined herein for the catalyst of the invention.

The process of the invention for producing a catalyst typically further comprises recovering the catalyst.

In one embodiment of the process of the invention for producing a catalyst which comprises (a) iron oxide and (b) nanoparticles which comprise palladium, the process further comprises: using the catalyst thus produced as a catalyst for the hydrogenolysis of a polyhydric alcohol.

Thus, the process of the invention for producing a catalyst which comprises (a) iron oxide and (b) nanoparticles which comprise palladium, may further comprise: producing a monohydric alcohol, by treating a polyhydric alcohol with hydrogen in the presence of said catalyst.

Typically, the step of treating said polyhydric alcohol with hydrogen in the presence of said catalyst is as further defined herein for the process of the invention for producing a monohydric alcohol.

The invention further provides a catalyst which is obtainable by a process of the invention as defined above, for producing a catalyst which comprises (a) iron oxide and (b) nanoparticles which comprise palladium. Further provided is the use of such a catalyst for the hydrogenolysis of a polyhydric alcohol.

The present invention is further illustrated in the Example which follows:

EXAMPLE

In this Example, ethylene glycol, the simplest representative of biomass-derived polyols, is converted directly to the lower alcohols, by hydrogenolysis, with high selectivities (>80%) using a noble-metal / Fe ₂ 0 ₃ catalyst. This opens up an exciting new catalytic process, namely a non-enzymatic bio-alcohol production process which depends on concerted C-C or C-0 bond breakage and C-H bond formation. A unique metal-support interaction, within the carefully prepared co-precipitated catalyst, is revealed to give a supported metal phase of extremely high dispersion ranging from small nanoparticle clusters down to individual metal atoms, on defective iron oxide. This is thought to be responsible for highly cooperative catalysis.

1. Experimental Details

Synthesis of Metal Oxide Supported Catalysts made by co-precipitation

A typical synthesis for preparing 1 g of 5 wt% Pd/Fe ₂ 0 ₃ catalyst (i.e. Pd/Fe ₂ 0 ₃ containing 5 weight % Pd, based on the mass of the iron oxide) was as follows: 0.3309 g of palladium nitrate (containing 15.1 lwt% Pd) and 5.071 g of Fe(N0 ₃ ) ₃ .9H ₂ 0 were dissolved in a 200 mL water at room temperature under stirring. Then, 1.0 M Na ₂ C0 ₃ solution was added dropwise to the mixed solution until the pH of the final solution reached ca. 9.0. After stirring and standing overnight, the resulting precipitate (denoted as Pd/Fe(OH) _x ) was filtered, washed with deionized water several times, air-dried at 1 10 °C overnight, and calcined in air at 300 °C for 2 hours. The dried calcined powder (Pd/Fe ₂ 0 ₃ ) was then pre- reduced in a flowing stream of 33% H ₂ /N ₂ at a steady rate of 60 mL/min in a tube furnace at 200 °C for 1 , 5, 48 or 72 hours, respectively. The furnace was then cooled down to room temperature and the powder was transferred into a glove box for storage before the catalytic testing. The as-reduced catalyst was denoted as Pd/Fe ₂ 0 ₃ .

Catalyst testing: hydrogenolysis of ethylene glycol in a liquid phase batch reactor

The hydrogenolysis reaction was carried out in liquid phase using a 160 mL autoclave high pressure reactor. In a typical experiment, 30 mL of ethylene glycol solution (0.85 mol/L) and 0.50 g of catalyst were loaded in a glove box. The reactor was flushed with pure hydrogen to remove any traces of oxygen. The autoclave was pressurized to 20 bars of hydrogen at room temperature. Then, it was heated up to 195°C and allowed to react for 24 hours with constant stirring

Product analysis

After reaction, the autoclave was then cooled down to -60 °C by dry ice / acetone bath. Here, the gas phase was analysed by a Perkin Elmer Autosystem XL Arnel Gas phase GC- FID-Methanator in order to determinate the concentrations of CO, CH ₄ and C0 ₂ after the reaction. The liquid phase was analysed by a Perkin Elmer 200 series HPLC equipped with a refractive index detector where a 0.60 mL of 58 mmol/L sucrose solution was added as an external standard to determinate the concentration of ethylene glycol, methanol and ethanol after the reaction. The mass balance of this typical catalytic batch reaction of over 95% was ensured.

2. Results, Characterising Data and Discussion Initial Screening

Catalysts containing 5 wt% Pd (based on the weight of the metal oxide) on various metal oxides (Fe ₂ 0 ₃ , ZnO, Ga ₂ 0 ₃ , Ce0 ₂ and A1 ₂ 0 ₃ ) were prepared by a co-precipitation technique as described above, where 1.0 M Na ₂ C0 ₃ solution was added dropwise to a mixed solution of the metal nitrates until the pH of the final solution reached ca. 9.0. Each sample was aged, dried and calcined in an identical manner. The samples were tested in ethylene glycol solution after a pre-reduction stage with hydrogen at 200°C for 1 h. As seen from Fig.2, the conversions of ethylene glycol molecules in aqueous solution are in general small over all the Pd containing catalysts under high hydrogen pressure; this is presumably because the metal surface is primarily covered with adsorbed hydrogen and water molecules. However, once the ethylene glycol is activated on surface, the main product is carbon dioxide (C-C bond cleavage followed by C-0 bond formation). Various activities for ethanol production from the hydrogenolysis of ethylene glycol (C-0 bond cleavage followed by C-H formation) were observed over the different samples. It was interesting to note that a highly dispersed Pd (no clear visible Pd peak from XRD) on iron oxide was the only sample which showed a discernible activity for methanol production (C-C cleavage followed by C-H formation) but with no methane formation. Iron oxide appeared to be an essential component for alcohol production.

Further Catalyst Testing

A number of metals supported on iron oxide (5 wt% metal based on the weight of iron oxide) were then screened; these catalysts were prepared by the same co-precipitation process, (described above for Fe/Fe ₂ 0 ₃ ). The two most promising catalysts, Rh/Fe ₂ 0 ₃ and Pd/Fe ₂ 0 ₃ , gave relatively high selectivity towards the lower alcohols (as seen from Fig. 3). However, Rh gave much higher conversion and methane as the primary product. Larger sizes of metal and metals with stronger intrinsic adsorptive properties are known to catalyse methane production from hydrocarbons/alcohols with exhaustive hydrogenolysis on their surface (R. Burch, A. R. Flambard, J Catal 78, 389,1982). The combination of small Pd particles on iron oxide clearly shows a synergistic effect in the production of lower alcohols as meta-stable products, which surprisingly gave even higher selectivity at the prolonged period of testing in our batch reactor configuration (24h).

Figure 4 shows a further improvement in accumulative selectivity of over 50% after the catalyst was tested for 72h. Time fraction analysis in Fig. 5 clearly suggests the catalyst initially gave higher ethylene glycol conversions but mainly producing C0 ₂ . This could be due to the aqueous reforming reaction of ethylene glycol to carbon dioxide and hydrogen (C ₂ H ₆ 02 + 2H ₂ 0→ 2C0 ₂ + 5H ₂ ) which is known to take place near our reaction temperature (R. D. Cortright, R. R. Davda, J. A. Dumesic, Nature 418, 964 (2002); J. W. Shabaker, G. W. Huber, R. R. Davda, R. D. Cortright, J. A. Dumesic, Catal Lett 88, 1 , 2003). Also, ethylene glycol may act as a reductant for iron oxide (3C ₂ H ₆ 0 ₂ + 5Fe ₂ 0 ₃ → 6C0 ₂ + 9H ₂ 0 + 1 OFe) in the presence of Pd. That the latter reaction has probably played an important part for the initial production of C0 ₂ . When the catalyst was pre-reduced in hydrogen at a longer duration (72h) or at a higher temperature, selectivity for the alcohols was increased accordingly. After this induction, the catalyst gave much higher selectivity towards the alcohols. At least 80% selectivity (with methanol selectivity >50%) towards the lower alcohols (methanol and ethanol) was obtained at about 4.3% conversion between the period of 24 to 72 h. Re-testing used samples confirmed that the catalyst typically required an induction period for extensive reduction to reach steady state in order to achieve good alcohol selectivity. The optimum temperature is at around 195°C before a significant destruction of the products at longer contact time (Fig. 6). Such conversion levels for ethylene glycol or related polyols to the lower alcohols at such high selectivities have not previously been reported, which clearly reflects the unusual behavior of the Pd/Fe ₂ 0 ₃ . It would be important to investigate the reason(s) for the extremely good metal dispersion (strong metal-support interaction) over this Pd/Fe ₂ 0 ₃ catalyst and the cause for methanol/ethanol formation. TEM Microscopy

A scanning transmission electron microscopy high angle annular dark field (STEM- HAADF) image of the sample after pre-reduction was taken using a JOEL 3000F electron microscope with both high-resolution transmission electron microscopy (HRTEM) and scanning transmission electron microscopy (STEM) modes. The image showed a large number of very small metal particles of high contrast in a number of thin areas, i.e. at the edge surfaces of the iron oxide support). The particles are around ca.1.0-2.5 nm in diameter with 1.5 nm being the most probable particle size. Using aberration corrected 2200MCO - HAADF STEM microscopy to focus on a thin area of iron oxide support near the edge confirmed the dense array of small metal particles. The particle size distribution of the metal nanoparticles is shown in Fig. 7. Higher magnifications clearly showed the presence of even smaller metal clusters (many of them below 1.0 nm) with occasionally single atoms found on support surface. This result is very surprising with particular regards to the small metal sizes and their stability on surface since the sample had experienced various heat treatments. EDX analysis suggested that the surface clusters contained Pd and possibly Fe but it was unable to ascertain this measurement in the background of bulk iron oxide.

XRD

Fig. 8 shows XRD patterns of the pre-reduced Pd/Fe ₂ 0 ₃ (lh), which gives a characteristic Fe ₂ 0 ₃ phase but with no Pd peak observed, indicative of extremely small Pd particles on the iron oxide surface. On the other hand, two additional broad diffraction peaks marginally differentiable from the background, as shown at 40.76 and 47.09 (2Θ degree) were observed after the sample was tested for this reaction at prolonged time (a small degree of sintering). According to the database (JCPDS 88-2335, 65-3253), the diffraction peaks of Pd (1 1 1) and Pd (2 0 0) are at 40.01 and 46.54 (2Θ degree) and PdFe (1 1 1) and FePd (2 0 0) peaks at 40.79 and 47.27 (2Θ degree), respectively. Thus, the positions of the two peaks matched more closely with those of PdFe than Pd. In addition, XRD peaks of our previously synthesized PdFe nanoparticles (ca. 4.6 nm) via a polyol process (C. H. Yu, C. C. H. Lo, K. Tarn, S. C. Tsang, J Phys Chem C 11 1, 7879, 2007) fitted well to these two peaks. XPS

As previously stated, samples pre-treated with hydrogen at different durations were tested for ethylene glycol hydrogeno lysis. It was shown that a progressive increase in the lower alcohols selectivity at longer pre-reduction periods. The XPS shown in Figs. 9 and 10 indicates that the binding energies of Pd and Fe are shifted to higher values as compared to literature values (A. D. Romig, J. I. Goldstein, Metall Trans A 9, 1599 (1978); K. Noack, H. Zbinden, R. Schlogl, Catal Lett 4, 145 (1990); S. L. Zhang, J. R. Zhang, Phys Status Solidi B 182, 421, 1994), indicative of alloy formation (A. D. Romig, J. I. Goldstein, Metall Trans A 9, 1599 (1978); K. Noack, H. Zbinden, R. Schlogl, Catal Lett 4, 145, 1990). The progressive binding energy shift of Pd 3d _5/2 of 335.16eV (pure Pd) to the converging value of 336.17eV shown in Fig. 10 (matched with the value of PdFe sample and extensively hydrogen pre-reduced sample) clearly suggested an increasing degree of alloying of PdFe (A. D. Romig, J. I. Goldstein, Metall Trans A 9, 1599, 1978) which corresponded well with the increasing alcohols selectivity. On the other hand, the correlation on the shift of binding energy of Fe 2p _3/2 (Fig.9) with respect to the selectivity was not quite linear due to the bulk iron oxide background.

TPR

It is thus evident that there is a deep reduction of Pd/Fe ₂ 0 ₃ by ethylene glycol (self- oxidized to C0 ₂ ) assisted by Pd to give PdFe during the induction period to account for the poor initial alcohols selectivity. It is noticed that extensive reduction of Fe ₂ 0 ₃ by hydrogen should not be taking place under our pre-reduction or testing conditions (G. Neri, A. M. Visco, S. Galvagno, A. Donate, M. Panzalorto, Thermochimica Acta 329, 39, 1999). Hydrogen temperature-programmed reduction (TPR) in Fig. 1 1 indeed shows no reduction of Fe ₂ 0 ₃ until the temperature is at about 340°C where some surface reduction by gaseous hydrogen occurred. The hydrogen reduction of bulk Fe ₂ 0 ₃ involving much a slower diffusion of lattice oxygens would require higher activation energy, hence takes place at higher temperature (broad peak with the maxima at 600°C). It is interesting to note that the presence of Pd appears to catalyse reduction of the surface Fe ₂ 0 ₃ more readily. The reduction temperature was shifted to around 130°C due to the incorporation of Pd. This implies that co-reductions of -PdO- and -FeO- in hydrogen can now take place under our pre-reduction/testing conditions to produce surface PdFe which may offer the

thermodynamic driving force for this reaction. Fig. 12 clearly shows the absence of typical hydrogen evolution peak of β-Pd-H (C. W. A. Chan, Y. Xie, N. Cailuo, K. M. K. Yu, J. Cookson, P. Bishop, S. C. Tsang, Chem. Commun. 47, 7971, 2011) as compared to alumina supported Pd at around 70°C (negative peak) in the second TPR profile, the fact is consistent with the formation of highly dispersive Pd atoms and PdFe clusters.

Kinetic Study

A kinetic study on the rate of methanol production as a function of ethylene glycol concentration (Fig. 13) or hydrogen pressure (Fig. 14) was carried out. The rate equation can be written as: r = d[CH ₃ OH]/dt = k [C ₂ H ₆ 0 ₂ ] ^m [H ₂ ] ⁿ

Fig. 14 clearly shows that an increase in hydrogen pressure resulted in a decrease in methanol production. This suggests that the ethylene glycol molecule and H ₂ are competing with the metal site. At elevated pressures adsorption of ethylene glycol is greatly suppressed (n = -1). On the other hand, Fig. 13 shows that there was a direct linear relationship on activity with respect to ethylene glycol concentration (m = +1). The linearity for the In k vs 1/T plot (Fig. 15) gave a slope of -Ea/R. The apparent activation energy for methanol production was determined to be 96 kJmol ^"1 .

Discussion

It is clearly demonstrated from this work that methanol and ethanol can be synthesized at high selectivity and good conversion by direct hydrogenolysis of biomass derived ethylene glycol over Pd/Fe ₂ 0 ₃ , opening up a new direction in biomass activation. From a mechanistic viewpoint, conversion of the ethylene glycol molecule to alcohols must be carried out by break a C-C or a C-0 bond with simultaneous C-H formation over a catalytically active site in a highly cooperative and concerted manner otherwise the exhaustive surface adsorption of the substrate on large metal catalyst will lead to thermodynamically more stable C0 ₂ or CH ₄ formation (M. Salciccioli, W. Yu, M. A. Barteau, J. G. Chen, D. G. Vlachos. J. Am. Chem. Soc 133, 7996, 2011). Pd/Fe ₂ 0 ₃ catalyst prepared by co-precipitation displays an unusual material interaction, which may be called a 'strong metal-support interaction'. However, our unprecedented evidence suggests the existence of extremely well dispersed metal clusters even down to individual metal atoms on iron oxide support surface. This is thought to explain the unexpected low selectivity towards the thermodynamic stable methane/C0 ₂ products but high selectively to alcohols due to the presence of catalytically extremely active metal sites with a strong support interaction for this reaction. However, atoms and small clusters are generally highly unstable against sintering. Their dispersion and surface stabilization on iron oxide are somewhat surprising given the fact that the sample had experienced various pre-treatments. It is believed a sol-gel like chemistry through hydrolysis and cross condensation during the co-precipitation of iron and palladium nitrates plays a role in dispersing Pd atomically in the Fe-O-Fe network, as shown in Fig. 1. It is interesting to note that the rates for SN ₂ hydrolysis and condensation of Pd species are more comparable to those of Fe species than the elements of the other support materials tested, with no phase segregation. The fact that the solubility product values for the Pd and Fe species are closer than those of other support precursors is evident from IUPAC-NIST Solubility Database, Version 1.0, NIST Standard Reference Database 106, 2007. Upon controlled reduction individual metal atoms and small clusters could be produced dependently on the local dispersion of the Pd. It is thought that the stable -Fe-O-Fe- rich area will constituent to the bulk support where the reduced metal atoms and clusters may still be partially connected to the iron oxide support to offer the kinetic stability against sintering after the heat treatments. The co-reduction of a small surface -Fe-O- network with Pd at close proximity will give PdFe clusters. These small but catalytically active PdFe clusters are thought to offer much enhanced catalyst activity and selectivity due to electronic modification of band structure. In addition, oxygen vacancies will be formed in close vicinity to the PdFe clusters on the defective surface due to the maintenance of electrical neutrality upon reduction (their presence may stabilize the clusters), which may further assist adsorption of the ethylene glycol molecule (T. L. Chen, D. R. Mullins, J. Phys. Chem. C, 1 15, 13725 (2011) for concerted C-C or C-0 bond breakage with simultaneous C-H formation to methanol by spillover hydrogen from the clusters. In conclusion, a new process has been found for producing non-enzymatic bio- methanol/ethanol, by catalytic hydrogeno lysis of biomass derived glycol over an 'atom dispersive surface' using the unique metal-support interaction of Pd/Fe ₂ 0 ₃ . This new capability can open up new ways to activate biomass molecules to produce strategically important fuels and chemicals. The combined selectivity for methanol/ethanol of 80% with the rest of C0 ₂ and H ₂ co-products from aqueous reforming may allow reduction or elimination of the need for expensive hydrogen gas input, rendering the catalytic process promising at this stage.