Compositions comprising manganese, an organic ligand, and hydrogen peroxide are widely used for oxidizing organic compounds, not only in the production of organic chemicals, but also in processes related to the destruction of unwanted compounds (e.g. stain removal from textiles, dirt removal during

dishwashing, waste water treatment, paper bleaching) [Bouwman et al. in Adv. Inorg. Chem. 2006, 58, 29; Lienke et al. Angew. Chem. Int. Ed. 2006, 45, 206]. Typically, manganese and the ligand are used in catalytic amounts relative to the oxidant (i.e. hydrogen peroxide). Manganese catalysts based on nitrogen containing ligands stand out in terms of catalyst activity [Feringa et al. in Modern Oxidation Methods 2004, 295], and for this reason combined with the environmental compatibility of manganese, oxidation catalysts based on this element are preferred over those based on other transition metals. Unfortunately, with many manganese catalysts, decomposition of hydrogen peroxide prevails over the desired organic compound oxidation. As a result, a large excess of hydrogen peroxide is required in order to achieve appreciable organic compound conversion. Obviously, strong hydrogen peroxide decomposition activity has a negative impact on the economics of the oxidative process. Suppression of the hydrogen peroxide decomposition activity in favor of the desired organic compound oxidation requires careful manganese catalyst design, including design of proper organic ligands [De Boer et al. in C. R. Chimie 2007, 10, 341 ]. Among the many ligands that have been developed to that end, 1 ,4,7-trimethyl-1 ,4,7-triazacyclononane (tmtacn) stands out in terms of its ability to impart a high oxidation activity on the manganese catalyst, while hydrogen peroxide decomposition activity can be minimized by a proper choice of additives and anions [Feringa et al. in Dalton Trans. 2008, 6283]. In addition, manganese + tmtacn based oxidation catalysts stand out in terms of catalyzing the desired oxidative transformations under very mild conditions with respect to pH and temperature, the latter often being sub-ambient (around 0 °C). Furthermore, manganese + tmtacn based oxidation catalysts do not only operate under aqueous conditions, but also enable oxidative transformations in organic media. The foregoing aspects of mildness of process conditions and compatibility with organic solvents are especially important in the area of oxidation catalysis for the purpose of organic synthesis, where large amounts of typically non-water soluble compounds need to be selectively transformed into the desired oxidation products. Mild conditions are not only required to achieve a high selectivity, but also to enable safe processing of a reaction that combines large amounts of an oxidizing agent with flammable organics (including the solvent). Unfortunately, tmtacn and other organic ligands for manganese oxidation catalysts that can be used to fulfill the foregoing requirements are only accessible via costly synthetic manufacturing steps, thus decreasing the general industrial applicability of these catalysts. Accordingly, there is a need for readily available, low cost manganese based oxidation catalyst systems that allow the selective oxidation of organic substrates in organic media under mild conditions and with a high efficiency with respect to usage of hydrogen peroxide for the desired transformation instead of unproductive hydrogen peroxide decomposition.

A simple water soluble manganese salt based oxidation catalyst system employing an aminocarboxylic acid as a ligand (e.g. picolinic acid) has been described recently for bleaching of waste paper with hydrogen peroxide

[JP2009235587]. This system operates only efficiently at elevated temperatures in water, and as will be shown in examples below, it is not suitable for selectively oxidizing organic substrates at sub-ambient temperature in organic solvent medium.

The invention now provides a composition comprising a soluble source of manganese, a ligand, a base, hydrogen peroxide and a ketone or an aldehyde, wherein the ligand is a pyridine heterocycle containing carboxylic acid or a precursor thereof, wherein the nitrogen atom of the pyridine ring is capable of coordinating to the carboxylate bonded manganese center, wherein the 2-position relative to the nitrogen atom is part of the N(pyridine)-Mn-0(carboxylate) containing chelate ring and the second 2-position relative to the nitrogen atom in the ring is not a carboxylic acid group.

The composition according to the invention is suitable for the oxidation of various organic substrates such as alkenes, alcohols, aldehydes, alkynes, alkanes, aralkanes, acetals and hemiacetals with only minor hydrogen peroxide decomposition. Furthermore, the composition according to the invention provides satisfactory results when applied in catalytic amounts, and can be applied in organic solvents even at sub-ambient temperature. Within the context of this invention with a soluble source of manganese is meant manganese compounds, typically manganese salts, which will dissolve in the medium of the composition.

Suitable examples for the ligand of the composition according to the invention are picolinic acid, picolinaldehyde, 2-pyridyl acetic acid, 2-acetyl-4-methyl- pyridine, 3-hydroxy picolinic acid, pyridine-2,5-dicarboxylic acid, quinoline-8- carbaldehyde, 2-pyridylacetic acid, pyridine-2-carbaldehyde and hydrolysable esters of picolinic acid. The pyridine heterocycle in the ligand of the invention can be a benzofused or a non-benzofused pyridine ring. Preferably, the aromatic heterocycle in the ligand of the invention is a non-benzofused pyridine ring. More preferably, picolinic acid is applied as the ligand in the composition according to the invention.

Suitable precursors for the chelating pyridine heterocycle containing carboxylic acid contain a nitrogen atom in the pyridine ring that is capable of coordinating to the carboxylate bonded manganese center with substituents on the 2- position that may be converted to carboxylate groups via thermal, hydrolytic, per- hydrolytic and/or oxidative reactions. Examples of such 2-substituents include, but are not limited to ester groups, aldehyde groups, acetal groups, acetyl groups, aminal groups and benzylic methylene groups.

The composition according to the invention holds a carbonyl compound. This carbonyl compound can be a ketone or an aldehyde. In case a ketone is applied, suitable examples include, but are not limited to acetone, 2-butanone, trifluoroacetone, 2,2,2-trifluoroacetophenone, 1 ,3-acetonedicarboxylic acid esters, diacetyl, pyruvic acid ester, alkylbenzoylformates and isatin derivatives. Preferably, the ketone is trifluoroacetone, acetone, 2-butanone, diacetyl, or a pyruvic acid ester. More preferably, the ketone is acetone, 2-butanone, diacetyl, or 1 ,1 ,1 -trifluoroacetone. In case an aldehyde is applied, suitable examples include, but are not limited to trifluoroacetaldehyde, chloral, or glyoxylic acid derivatives. The ketone or aldehyde may also be used in the form of the corresponding hydrates or (hemi-)acetals.

The composition according to the invention also holds a base.

Suitable examples of such a base include, but are not limited to carboxylate salts, in particular sodium acetate, carbonates, such as sodium carbonate, or hydroxides, such as sodium hydroxide. To the alternative, the required amount of base can be present or partly present in the soluble source of manganese or in the ligand, for example manganese can be added in the form of manganese acetate or the ligand can be added as sodium picolinate. The base can also be an organic compound, and as such it can also be added through the organic solvent or substrate. The amount of base to be added depends amongst others on the type of oxidation to be performed. In case of oxidation of an organic substrate, the amount of base is preferably 0.01 mol% to 20 mol% relative to the substrate. More preferably, the amount of base is 0.1 mol% to 10 mol% relative to the substrate, most preferably 0.3 mol% to 5 mol%.

The invention also relates to a process for the oxidation of an organic substrate, wherein a composition according to the invention is applied. Preferably, the composition according to the invention is applied in such a way that the soluble source of manganese and the ligand are applied in 0.00001 to 1 mol equivalent relative to the substrate, more preferably 0.00001 to 0.5 mol equivalents, even more preferably 0.0001 to 0.5 mol equivalents, most preferably 0.0001 to 0.1 mol equivalents.

In an embodiment of the process according to the invention, the temperature of the process varies between -20°C to 100°C, preferably between -10°C to 60°C, more preferably between 0°C to 40°C.

In another embodiment of the invention, the ketone : substrate ratio or the aldehyde : substrate ratio is 1 mol equivalent or lower, preferably 0.5 mol equivalent or lower, more preferably 0.3 mol equivalent or lower. This embodiment with (sub-)stoichiometric amounts of ketone relative to the substrate is particularly effective in case of ketones activated by electron-withdrawing groups, such as 1 ,1 ,1 - trifluoroacetone, 2,2,2-trifluoroacetophenone, diacetyl, 1 ,3-acetonedicarboxylic acid esters, pyruvic acid ester, alkylbenzoylformates or isatin derivatives.

In yet another embodiment of the invention, the ketone is used as a solvent or co-solvent. This embodiment with large amounts of ketone relative to the substrate is particularly effective in case of non-activated ketones, such as acetone or 2-butanone.

In another embodiment of the invention, the hydrogen peroxide : substrate ratio is between 0.5 and 10 mol equivalent, preferably between 1 and 5 mol equivalent, more preferably between 1.2 and 4 mol equivalent.

Examples of suitable substrates for the oxidation process according to the invention include, but are not limited to alkenes, alcohols, aldehydes, alkynes, alkanes, aralkanes, acetals or hemiacetals, sulfides, sulfoxides, amines. Preferably, the substrate contains an oxidizable C-H bond, oxidizable multiple bonds, or oxidizable hetero-atoms. More preferably the substrate is an alkene. Oxidation of alkenes with hydrogen peroxide to epoxides or cis-diols with manganese + tmtacn based catalysts has been disclosed before in several publications [J.W. de Boer et al. in Inorg. Chem. 2007, 46, 6353 and references therein]. The process of the invention now makes use of an inexpensive and readily available catalyst system. More specifically, the oxidation process of the invention is a cis-dihydroxylation or an epoxidation.

In a further embodiment, the substrate in the oxidation process of the invention is an electron-deficient alkene with an electron-poor double bond caused by the presence of electron-withdrawing groups. Electron-withdrawing groups are known to the person skilled in the art, and include but are not limited to -CO(0)R

(carboxylate), -C(0)R (carbonyl), -N0 ₂ (nitro), -CN (cyano), -CF ₃ (trifluoromethyl), -Ar (aryl) or -CX _n R3- _n wherein X is an electro-negative element such as a halogen, for example fluorine, chlorine, bromine or iodine, or X is an electro-negative group such as -OR (alkoxy), -N0 ₂ (nitro), -CN (cyano) or -NRC(0)R, n is≥ 1 and R is a hydrogen or an optionally substituted alkyl, aryl, alkylaryl or arylalkyl group. The electron- withdrawing group is preferably directly attached to the double bond, or attached to a carbon at an allylic position with respect to the double bond.

Examples of electron-deficient alkenes are aryl alkenes, vinyl halides, nitro alkenes, cyano alkenes, a, β-un saturated acid derivatives (including amides and imides), α,β- unsaturated ketones or aldehydes, trifluoromethyl alkenes, allylic halides, allylic alcohols or esters thereof, allylic ethers or allylic amides. Preferably, the electron deficient alkenes are α,β-unsaturated carboxylic acid derivatives. More preferably the electron deficient alkenes are maleates, fumerates, cinnamates, optionally substituted styrenes or optionally alkyl substituted acrylates such as methylcrotonate, maleimides. Oxidation of these electron-deficient alkene substrates surprisingly leads to cis- dihydroxylation. To the contrary, oxidation of non-electron-deficient alkene substrates typically leads to epoxidation.

In yet another embodiment of the invention, the oxidation product obtained in the process of the invention is (3R,4S)-3,4-dihydroxypyrrolidine-2,5-dione or an N-substituted derivative thereof.

In another aspect of the invention, the oxidation product obtained in the process of the invention is a mesotartaric acid or a mesotartaric acid derivative.

In a further embodiment, the substrate in the oxidation process of the invention is a compound containing oxidizable C-H bonds. Alkane oxidation with hydrogen peroxide by manganese + tmtacn based catalyst systems has been disclosed before [G.B. Shul'pin et al. in Tetrahedron 2007, 63, 7997 and references cited therein]. It has also been demonstrated that C-H bond oxidation catalyzed by manganese + tmtacn systems can be highly stereospecific, as shown in the hydroxylation of cis-1 ,2-dimethylcyclohexane with retention of stereochemistry [J. Kim et al. in Bull. Korean Chem. Soc. 2003, 24, 1835]. The process of the invention now makes use of an inexpensive and readily available catalyst system for achieving C-H bond oxidations, including stereospecific sp ³ C-H bond oxidations (Table 2, entry 6 in Example 4).

The invention further relates to all possible combinations of different embodiments and/or preferred features according to the composition and method according to the invention as described herein.

The invention will be elucidated with reference to the following examples, without however being restricted by these:

EXAMPLES

Example 1 : Alkene oxidation catalyzed by the Mn/picolinic acid system in acetone

The following table provides examples that demonstrate the usefulness of the Mn/picolinic acid catalyst system for the oxidation of alkenes with H ₂ 0 ₂ in acetone as the solvent, providing epoxides or (cis-)diols as the major products:

Added at 0°C, then bath allowed to warm to room temperature overnight General procedure (amounts as specified per substrate in Table 1 ): Aqueous NaOAc was added to a mixture of substrate, 1 ,2-dichlorobenzene (internal standard), Mn(CI0 ₄ ).6H ₂ 0, and picolinic acid in acetone. At 0°C, 50 wt% H ₂ 0 ₂ was subsequently added gradually. The mixture was stirred additionally overnight, with gradual warming to room temperature. Workup typically consisted of adding solid NaHS0 ₃ to reduce residual peroxide, removal of solids by filtration, evaporation, and extraction with CDCI ₃ . Analysis was done by NMR, Raman spectroscopy, and/or GC. Example 2: Alkene oxidation catalyzed by the Mn/picolinic acid system in 2-butanone

Aqueous NaOAc (0.009 mmol as a 0.6 mol/L stock solution) was added to a mixture of ethyl crotonate (0.75 mmol), Mn(CI0 ₄ ).6H ₂ 0 (0.00075 mmol), and picolinic acid (0.0023 mmol) in 2-butanone (1.0 ml_). At 0°C, 50 wt% H ₂ 0 ₂ (6 mmol) was subsequently added gradually. The mixture was allowed to warm to room temperature overnight. A 30% conversion was measured by Raman spectroscopy. Via an analogous procedure, 50% conversion was measured for styrene, and 80% for cyclooctene.

Examples 1 and 2 demonstrate efficient substrate oxidation by hydrogen peroxide with the Mn/picolinic acid catalyst system in the presence of base and a non-activated ketone such as acetone or 2-butanone as solvent.

Example 3: Alkene oxidation catalyzed by the Mn/picolinic acid system in acetonitrile using sub-stoichiometric 1 ,1 ,1 -trifluoroacetone

Aqueous NaOAc (0.03 mmol as a 0.6 mol/L stock solution) was added to a mixture of diethylfumarate (3 mmol), 1 ,1 ,1 -trifluoroacetone (1 mmol), Mn(CI0 ₄ ).6H ₂ 0 (0.003 mmol), and picolinic acid (0.009 mmol) in MeCN (6.0 mL). At 0°C, 50 wt% H ₂ 0 ₂ (6 mmol) was subsequently added gradually. The mixture was allowed to warm to room temperature overnight. After workup, the cis-diol, i.e. (2S,3S)- diethyl 2,3-dihydroxysuccinate, was obtained in >90% yield.

Example 4: Alkene oxidation catalyzed by the Mn/picolinic acid system in acetonitrile using sub-stoichiometric diacetyl

A solution of Mn(CI0 ₄ ) ₂ -6H ₂ 0 (0.0007 mmol), picolinic acid (0.006 mmol) and aqueous NaOAc (0.01 mmol from a 0.6 mol/L stock solution) in MeCN (0.5 mL) was added to a solution of diethylfumarate (1.0 mmol) and diacetyl (0.2 mmol) in MeCN (1 .8 mL). At 0°C, 50 wt% H ₂ 0 ₂ (1.5 mmol) was subsequently added in one portion. The solution was allowed to warm to room temperature and was stirred for 3 h.

A 70% conversion was measured by Raman spectroscopy.

Examples 3 and 4 demonstrate efficient substrate oxidation by hydrogen peroxide with the Mn/picolinic acid catalyst system in the presence of base and an activated ketone such as 1 ,1 ,1 -trifluoroacetone or diacetyl in sub-stoichiometric amount using acetonitrile as the solvent.

Example 5: C-H bond and heteroatom oxidation catalyzed by the Mn/picolinic acid system in acetone

The following table provides examples that demonstrate the usefulness of the Mn/picolinic acid catalyst system for the oxidation of C-H bonds with H ₂ 0 ₂ in acetone as the solvent, typically providing ketones or alcohols as the major products. The oxidation of pyridine-2-carbaldehyde (Entry 7) also demonstrates N- heteroatom oxidation with the Mn/picolinic acid catalyst system, with the picolinic acid being formed in situ from the pyridine-2-carbaldehyde and the required base being present in the form of the pyridine functionalities in the substrate. Entry 6 demonstrates the stereoselective C-H bond oxidation of cis-1 ,2-dimethylcyclohexane, which is transformed with retention of configuration into a racemic mixture of (1 R,2R)-1 ,2- dimethylcyclohexanol and (1 S,2S)-1 ,2-dimethylcyclohexanol.

Table 2

c: As 50 wt% aqueous solution

d ^: Added at 0°C, then bath allowed to warm to room temperature overnight e: Added at room temperature, then stirred additionally overnight

: No NaOAc added; pyridine unit in the substrate functions as organic base

General procedure (amounts as specified per substrate in Table 2): Aqueous NaOAc was added to a mixture of substrate, 1 ,2-dichlorobenzene (internal standard), Mn(CI0 ₄ ).6H ₂ 0, and picolinic acid in acetone. At the temperature indicated in Table 2, 50 wt% H ₂ 0 ₂ was subsequently added gradually. The mixture was stirred additionally overnight, with gradual warming to room temperature when relevant. Workup typically consisted of quenching with excess saturated aqueous NaHC0 ₃ , extraction with dichloromethane, drying on Na ₂ S0 ₄ , and evaporation. Analysis was done by NMR and/or GC.

Example 6: Oxidation catalyzed by various Mn/ligand systems in acetone

Conversions of diethylfumarate and/or cyclooctane were measured when the oxidation was carried out in acetone according to the general procedure in Example 5, typically at room temperature with 0.1 mol% Mn(CI0 ₄ ) ₂ .6H ₂ 0, 0.3 mol% ligand, 0.5, 1 , or 2 mol% NaOAc, and 2 or 8 mol-eq H ₂ 0 ₂ , with "ligand" also including compounds that are converted in situ to the actual metal-chelating compounds responsible for catalytic activity. Mostly, Raman spectroscopy was used to determine substrate conversions. No conversion was observed in case of the following ligands: glycine, N-Boc-glycine, phenylalanine, a -phenylalanine, a -phenylalanine amide, a - hexylalanine, a -nonylalanine, proline, histidine, tryptophan, methylpicolinate, nicotinic acid, pyridine-2,6-dicarboxylic acid, indole-2-carboxylic acid, pyrazine-2-carboxylic acid, 2-(1 ,3-dimethylimidazolidin-2-yl)-1 H-imidazole, or 5-(1 ,3-dimethylimidazolidin-2- yl)-1 H-imidazole. Amount of base optimization allowed a high conversion of diethylfumarate with quinoline-8-carbaldehyde as a ligand, but essentially no conversion of cyclooctene. Pyridine-2,5-dicarboxylic acid, 2-acetyl-4-methylpyridine, 3- hydroxypicolinic acid, and 2-pyridylacetic acid gave a substantial conversion for both diethylfumarate and cyclooctene under optimum base conditions. Pyridine-2- carbaldehyde as a ligand allowed a high conversion of diethylfumarate.

These experiments show that a requirement for catalytic activity is the presence of a pyridine heterocycle containing carboxylic acid or a precursor thereof, wherein the nitrogen atom of the pyridine ring is capable of coordinating to the carboxylate bonded manganese center, wherein the 2-position relative to the nitrogen atom is part of the N(pyridine)-Mn-0(carboxylate) containing chelate ring and the second 2-position relative to the nitrogen atom in the ring is not a carboxylic acid group. Ligands of this type that impart catalytic activity on the manganese center in the experiments listed above are: quinoline-8-carbaldehyde, pyridine-2,5-dicarboxylic acid, 2-acetyl-4-methylpyridine, 3-hydroxypicolinic acid, 2-pyridylacetic acid, pyridine-2- carbaldehyde (Table 3).

Example I demonstrating the requirement of a carbonyl compound in the composition: C-H bond oxidation of cyclooctane catalyzed by the Mn/picolinic acid system in acetonitrile versus in acetone

No conversion of cyclooctane was observed when the oxidation was carried out at room temperature in acetonitrile instead of acetone according to the general procedure in Example 5, with 0.1 mol% Mn(CI0 ₄ )2.6H ₂ 0, 0.3 mol% picolinic acid, 0.3 mol% NaOAc, and 8 mol-eq H ₂ 0 ₂ . In contrast, cyclooctane conversion reached 63% when carried out similarly in acetone instead of acetonitrile.

Example II demonstrating the requirement of a carbonyl compound in the composition: C-H bond oxidation of tetraline catalyzed by the Mn/picolinic acid system in various non-ketonic solvents versus in acetonitrile with added 1 ,1 ,1 -trifluoroacetone

No conversion of tetraline was observed when the oxidation was carried out at 0°C to room temperature in acetonitrile, 2-methyl-2-butanol, or ethylacetate instead of acetone according to the general procedure in Example 5, with 0.5 mol% Mn(CI0 ₄ ) ₂ .6H ₂ 0, 6 mol% picolinic acid, 2 mol% NaOAc, and 4 mol-eq H ₂ 0 ₂ . In contrast, tetraline conversion reached 88% with 47% yield of 1 -tetralone when carried out similarly in acetonitrile containing 10 vol-% 1 ,1 ,1 -trifluoroacetone.

Example demonstrating the requirement of a base in the composition: Oxidation catalyzed by various Mn/ligand systems in acetone with and without base

Conversions of diethylfumarate, cyclooctene and cyclooctane were measured when the oxidation was carried out in acetone according to the general procedure in Example 5, typically at room temperature with 0.1 mol% Mn(CI0 ₄ )2.6H ₂ 0, 0.1 (for diethylfumarate and cyclooctene) or 0.3 (for cyclooctane) mol% picolinic acid, varying amounts of base (NaOAc or NaOH up to 2 mol%), and 2 (for diethylfumarate and cyclooctane) or 8 (for cyclooctene) mol-eq H ₂ 0 ₂ . Raman spectroscopy was used to determine substrate conversions. No conversion was observed when no base was added. Conversion reached an optimum at a certain amount of base. The optimum amount of base depends on the type of conversion being carried out.

Example demonstrating the requirement of a suitable ligand plus manganese in the composition: Oxidation of diethylfumarate catalyzed by Mn without picolinic acid or by picolinic acid without Mn versus catalysis by the mixed Mn/picolinic acid system in acetone

Conversion of diethylfumarate was measured when the oxidation was carried out in acetone according to the general procedure in Example 5 at room temperature with 0.1 mol% Mn(CI0 ₄ )2.6H ₂ 0 or 0.1 mol% picolinic acid, varying amounts of Na ₂ C0 ₃ (up to 2 mol%), and 2 mol-eq H ₂ 0 ₂ . Raman spectroscopy was used to determine substrate conversions. No conversion was observed in both cases, irrespective of the amount of base. In contrast, diethylfumarate was completely converted under similar conditions but with 0.1 mol% Mn(CI0 ₄ ) ₂ .6H ₂ 0 and 0.1 mol% picolinic acid, varying amounts of Na ₂ C0 ₃ (0.5 or 1 mol%), and 2 mol-eq H ₂ 0 ₂ . A similar requirement for Mn and picolinic acid being both present in order to achieve significant substrate conversions was observed with cyclooctene.