The present invention relates to a preparation of a compound of formula I with R1 and R2 independently from each other representing hydrogen or a C1 - to C10 alkyl group by a two-step process wherein the first step is an amination of the compound of formula II with an amino compound HNR ¹ R ² hydrogen in the presence of hydrogen and resulting in the compound of formula III and the second step is a ring-cleaving reaction of the compound of formula III giving the compound of formula I. 3-amino-1 ,2-propandiol (shortly APD) and 3-(methylamino)propane-1 ,2-diol (shortly MAPD) are intermediates for the pharmaceutical industry and is used in particular for the synthesis of contrast agents.

According to US 4358615, US 4360697 and EP-A 364708 APD is produced by a process using alkylene oxides, in particular epichlorohydrin or glycidol (usually obtained from epichlorohydrin) as starting materials.

WO 2012/108777 describes a continuous process for the production of APD wherein 1 -chloro- 2,3-propandiol is used as starting material.

In CN 103739501 a process for the synthesis of 2-amino-1 ,3-propylene glycol (Serinol) is disclosed which comprises a reductive amination and a hydrolysis of the cyclic ketal used as protecting group. The cyclic ketal used as protecting groups is a six-membered ring system.

Compounds comprising chlorine are considered to be toxic and may cause corrosion. If they are used efforts must be made to keep waste streams free of chlorinated compounds to avoid any contamination of the environment. It is the object of the present invention to provide a process for the production of APD and its derivatives which does not involve the use of any compounds comprising chlorine such as epichlorohydrin or glycidol which is usually obtained from epichlorohydrin. The process should be very effective and easy to perform. The yield of APD or its derivatives should be high. The formation of any by-products that cannot easily be separated from the desired product should be reduced or avoided.

Accordingly, the process as defined above has been found. To the compound of formula I

The compound of formula I is the product of the process of this invention.

Preferably, R ¹ and R ² independently from each other represent hydrogen or a C1 - to C4 alkyl group, which is specifically a methyl or ethyl group.

In a particularly preferred embodiment at least one of R1 and R2 in formula I is hydrogen.

In a most preferred embodiment R ¹ is hydrogen and R ² is methyl thus corresponding to the compound 3-(methyl amino) propane-1 ,2-diol (MAPD) or both R ¹ and R ² are hydrogen thus corresponding to 3-aminopropane-1 ,2-diol (APD)

APD is the most preferred compound of formula I. To the first step

The first step of the two-step synthesis is the amination of the compound of formula II. The chemical name of compound of formula II is (2,2-dimethyl-1 ,3-dioxolan-4-yl) methanol or iso- propylidene glycerol. The compound of formula II is known as solketal and is commercially available under this name.

The compound of formula II is reacted with an amino compound in the presence of hydrogen. The amino compound is HNR ¹ R ² wherein R ¹ and R ² have the same meaning as in formula I. Preferably, the amino compound is liquid or gaseous at 21 °C, 1 bar. In the most preferred embodiments the amino compound is ammonia Nll- or methyl amine H3C- NH ₂ .

The amination can be performed with or without a solvent. Suitable solvents are hydrophilic aprotic organic solvents, for example tetrahydrofuran, diethyl ether, methyl tertiary butyl ether or 1 ,4-dioxane.

Preferably, no solvent is used in the 1 . step. The first step may be performed continuously, semi-continuously or batch-wise. In a preferred embodiment, the first step is performed continuously by continuously supplying the starting materials to the reactor and continuously removing the products from the reactor.

The reactor may, for example, be a batch reactor, a cascade of batch reactors or a tubular reac- tor; preferably it is a tubular reactor.

In the continuous process the compound of formula II, the amino compound HNR ¹ R ² and hydrogen are continuously supplied to the reactor separately or as combined feeds, if appropriate. Hydrogen may be used as such or as a mixture of hydrogen with inert gases such as nitrogen or noble gases. In a preferred embodiment hydrogen is used as such.

The amino compound is preferably used in at least equimolar amounts, more particularly it is used in a molar excess based on the compound of formula II.

More particularly, the molar ratio of the amino compound to the compound of formula II is from 1 :1 to 50:1 , more preferably from 2:1 to 20:1.

The hydrogen is supplied in gaseous form in sufficient amounts, generally in a molar excess based on the compound of formula II.

Preferably, the first step is performed in presence of a catalyst. Suitable catalysts are any hy- drogenation catalysts. The amination may be catalyzed by a homogeneous catalysis or by a heterogenous catalysis.

In case of a homogeneous catalysis compound of formula II and the catalyst may be solved in a solvent, for example toluene, tetrahydrofuran or diethylene glycol dibutyl ether. Suitable catalysts for the homogeneous catalysis of the amination are, for example rutheni- um(lll) acetylacetonate, bis(2-methylallyl)(1 ,5-cyclooctadiene)ruthenium(ll), carbonylchlorohy- dridotris(triphenylphosphine)ruthenium(ll) or chlorocarbonylhydrido[4,5- bis((dicyclohexylphosphino)methyl)acridine]ruthenium(ll) with or without the addition of a ligand such as 1 ,1 ,1 -tris(diphenylphosphinomethyl)ethane, or 4,5- bis((dicyclohexylphosphino)methyl)acridine. Preferably, the amination is catalyzed by a heterogenous catalysis. In a heterogeneous catalysis, the catalyst used is solid. In a preferred embodiment of the invention, a heterogenous catalyst is used in form of powder or pellets, notably pellets.

Heterogeneous catalysts for hydrogenation comprise catalytically active elements or com- pounds, notably oxides. The heterogeneous catalysts may in form of particles without a support (unsupported catalysts) or may be supported catalysts, wherein the catalytically active elements or compounds have been applied to a support, for example calcium carbonate, silicon oxide, zirconium dioxide or aluminum oxide. In the following, the term metal as used in this patent application encompasses elemental metals and also chemical compounds comprising metals either in ionic form or in covalently bonded form.

Examples of catalysts comprising metals in elemental form are Raney nickel or Raney cobalt.

Metals used in form of oxides are usually at least partially reduced to the metals at higher temperatures and, preferably, in the presence of hydrogen. This reduction is often performed at the beginning of the reaction under the conditions of the reaction or it may be carried out beforehand in a separate step.

Suitable hydrogenation catalysts comprise, for example, metals of the groups IVb, Vb, Vlb, Vllb, Vlllb, lb or lib.

In a preferred embodiment, the hydrogenation catalyst is a supported catalyst; the support is preferably zirconium oxide (Zr02) or aluminum oxide (AI2O3).

Hydrogenation catalysts comprising at least one metal of the cobalt, nickel or copper group of the periodic table are preferred, particularly hydrogenation catalysts comprising at least one metal selected from cobalt, rhodium, iridium, nickel, palladium, platinum or copper are preferred.

Preferably, the content of the aforementioned metals of the cobalt, nickel or copper group in the catalyst amounts to a total of at least 5 wt%, more preferably at least 20 wt% and most preferably at least 50 wt%, based on the total weight of all active metals of the catalyst (for metal compounds, e.g., oxides, only the metal fraction is considered). Metals that are part of the chemical compounds that form the support, such as aluminum in aluminum oxide or zirconium in zirconium oxide are not considered as active metals. Further active metals that may be used along with the aforementioned metals include, for example, manganese, tin, ruthenium or else alkali metals and alkaline earth metals, whereby tin is a preferred one. In a particularly preferred embodiment, the hydrogenation catalysts comprise cobalt, nickel, copper or mixtures of these three metals in total amounts of 5 % by weight, more preferably at least 20 % by weight and most preferably at least 50 % by weight , based on the total weight of all active metals of the hydrogenation catalyst. In a more preferred embodiment, the hydrogenation catalysts comprise both, nickel and copper. Specifically, the hydrogenation catalysts comprise at least 1 %, in particular at least 5 % by weight of nickel and at least 1 %, in particular at least 5 % by weight of copper based on the total weight of all active metals of the hydrogenation catalyst. In a most preferred embodiment, the hydrogenation catalysts comprise nickel and copper and cobalt. Specifically, the hydrogenation catalysts comprise at least 1 %, in particular at least 5 % by weight of nickel and at least 1 %, in particular at least 5 % by weight of copper and at least 1 %, in particular at least 5 % by weight of cobalt based on the total weight of all active metals of the hydrogenation catalyst.

The hydrogenation catalyst may be installed in the reactor as fixed-bed. The fixed-bed can be positioned in the reactor as desired. It is generally positioned such that the total reactant stream passes through the fixed-bed. The fixed-bed therefore generally fills the total flow cross-section of the reactor. The volume of the fixed bed is preferably selected such that the mass flow of the amino compound HNR ¹ R ² over the hydrogenation fixed-bed is 0.005 to 50 kg/hour (h), more particularly 0.05 to 50 kg/h, per liter volume of the fixed-bed.

The reaction of the amino compound with the compound of formula II in the presence of hydrogen may be performed, for example, at temperatures from 50 to 250°C, preferably from 100 to 250°C, more preferably from 180 to 220°C. The reaction can be carried out, for example, at normal pressure or at elevated pressure. In a preferred embodiment, it is carried out at elevated pressure, for example, at pressure from 1.1 to 500 bar, particularly from 20 to 300 bar; more preferred is a pressure of 50 to 200 and most preferred a pressure of 80 to 150 bar. The pressure is preferably established by supplying hydrogen or the gas mixture comprising hydrogen with such pressure.

The reaction product is continuously removed from the reaction zone once it has flowed through the fixed-bed. The obtained product mixture comprises the desired product of the first step, which is compound of formula III, any unconverted starting materials and any by-products. Any volatile compounds such as ammonia may be easily removed from the product mixture by degasification. The product mixture may optionally be further worked up/purified by distillation or rectification.

As result of the first step, the compound of formula III is obtained. Yield and selectivity of the compound of formula III is high. A yield of compound of formula III above 80 %, notably above 90% is usually obtainable. The obtainable selectivity is usually above 80%, notably above 90%.

To the second step

The second step is the ring-cleaving reaction of the compound of formula III. In the reaction of the second step acetone is split off from the compound of formula III and removed.

Preferably, the ring-cleaving reaction is a hydrolysis or an alcoholysis.

In case of a hydrolysis the ring-cleaving reaction is performed in the presence of water and an acid. The compound of formula I is formed and acetone as byproduct.

In case of an alcoholysis the ring-cleaving reaction is performed in the presence of an alcohol and acid. The alcohol is preferably an alkanol or alkandiol with 1 to 4 carbon atoms, notably methanol or ethanol or ethylene glycol. The compound of formula I is formed and the corresponding ketal of acetone as byproduct.

Preferably, the ring-cleaving reaction is a hydrolysis.

The acid used in the hydrolysis or alcoholysis may be an inorganic acid or an organic acid. The term "acid" shall include ion exchange resins in the acid form.

In a preferred embodiment, the acid is selected from inorganic acids.

Preferred inorganic acids are, for example, sulfuric acid, phosphoric acid or nitric acid. Preferably, the acid is solved in water (in case of a hydrolysis), respectively in alcohol (in case of an alcoholysis).

In one embodiment of the invention, the reactor is charged with a solution of the acid in alcohol or water. Preferably, the solution is an aqueous solution. The solution may comprise, for exam- pie, 10 to 60 %, notably 30 to 60 % by weight of the acid. The compound of formula III is added to the solution. Preferably it is added slowly, respectively dropwise.

During the addition of compound III the solution is preferably kept at a temperature of 30 to 100°C, notably of 50 to 90°C.

Acetone (in case of the hydrolysis) or the ketal of acetone (in case of the alcoholysis) are removed from the reaction mixture during and after the addition of compound of formula III. The reaction has come to an end when the development of acetone, respectively the ketal has stopped and all acetone respectively ketal has been removed.

The obtained product solution comprises the protonated compound of formula I. A proton of the acid has added to the nitrogen atom turning the amino group into an ammonium group and the compound of formula I into an ammonium compound. The anion of the acid is the anion of the ammonium compound.

In order to get the compound of formula I a basic compound has to be added to the product solution or the product solution has to be brought into contact with an ion exchange resin in the hydroxide form. Preferably, the basic compound is a water-soluble base, such as an alkali hy- droxide, for example sodium hydroxide or potassium hydroxide. Preferably, an aqueous solution of the base is added to the product solution above. A salt precipitates usually from the product solution. The salt is formed from the cation of the base and the anion of the acid, respectively the anion of the ammonium compound. The precipitate may be removed by filtration. The product solution now comprises the compound of formula I.

The product mixture may optionally be further worked up/purified by distillation or rectification.

To the use of carbon dioxide for the ring-cleaving reaction In a preferred embodiment carbon dioxide is used in the ring-cleaving reaction. Preferably, the compound of formula III is brought into contact with carbon dioxide and, in addition, water (hydrolysis) or alcohol (alcoholysis).

Preferably, the ring-cleaving reaction with carbon dioxide is a hydrolysis.

The ring-cleaving reaction with carbon dioxide may be performed in a batch process, semi- continuously or continuously.

In a batch process, the compound of formula III, water, respectively alcohol, are filled into a re- actor. Carbon dioxide is fed as gas liquid or solid (dry ice). Preferably, it is fed as gas in an amount to get the desired pressure. Carbon dioxide may be used as such or in combination with inert gases, such as nitrogen or noble gases. In a semi-continuous process at least one of the starting materials is supplied continuously. In one embodiment of a semi-continuous operation of the process, the compound of formula III is filled completely into the reactor whereas gaseous carbon dioxide or a mixture of carbon dioxide and an inert gas and water (in case of a hydrolysis) or alcohol (in case of an alcoholysis) are fed continuously to the reactor; gaseous products, such as acetone are continuously withdrawn from the reactor.

Water, respectively the alcohol, is preferably used in an amount of 10 to 1000 parts by weight, in particular of 100 to 500 parts by weight and most preferably in an amount of 200 to 400 parts by weight per 100 parts by weight of compound of formula III.

The ring-cleaving reaction with carbon dioxide is preferably performed at a pressure above normal pressure. Preferably the pressure is at least 1 .5 bar, in particular at least 5 bar, more preferably at least 10 bars and most preferred at least 20 bars. Usually the pressure is at maximum 500 bars, respectively at maximum 200 bars, as higher pressures are not required. The temperature is preferably kept at 50 to 200, in particular at 90 to 170°C.

When the reaction is completed in case of a batch and semi-continuous process the product mixture may be removed from the reactor.

The product mixture comprises the compound of formula I. The addition of a basic compound is not necessary and any precipitates are avoided due to the use of carbon dioxide. The compound of formula I is obtained in high yields and selectivity.

The process of the invention is a very easy and effective process to produce compounds of formula I, in particular APD and MAPD. The process does not involve the use of any compounds comprising chlorine. High yields are obtained.

Examples

Example 1 : Heterogeneous amination of solketal

Chemical Formula: C ₆ H ₂ C>3 Chemical Formula: C ₆ H ₃ N0 ₂ Molecular Weight: 132,16 Molecular Weight: 131 ,18

A tubular reactor was filled with 1000 mL of a fixed-bed catalyst containing Ni, Co and Cu. After catalyst activation in the presence of hydrogen, the reactor was fed continuously with 400 g/h (2,2-dimethyl-1 ,3-dioxolan-4-yl)methanol (solketal), 1600 g/h NH ₃ and 1 15 NL/h H ₂ at 200 °C and 120 bar. The product stream was depressurized to atmospheric pressure and its composition was analyzed by GC. The product stream contained 86% (area % determined from gas chromatography) of (2,2-dimethyl-1 ,3-dioxolan-4-yl)methanamine (Conversion: 95%, Selectivity: 91 %), also known as solketalamine. Example 2: Homogeneous amination of solketal

Chemical Formula: C ₆ H ₂ 0 ₃ Chemical Formula: C ₆ H ₃ N0 ₂ Molecular Weight: 132,16 Molecular Weight: 131 ,18 An autoclave was charged in an argon-filled glovebox with 13.9 g solketal, 80.4 mg Chlorocar- bonylhydrido[4,5-bis((dicyclohexylphosphino)methyl)acridine] ruthenium(ll) and 30 mL tetrahy- drofuran. The autoclave was pressurized by addition of 10.7 g of liquid NH ₃ and the vessel was sealed. The reaction mixture was heated to 170 °C for 5 h. After depressurization, the composition of the mixture was analyzed by GC. The mixture contained 82% (area % determined from gas chromatography) of (2,2-dimethyl-1 ,3-dioxolan-4-yl) methanamine (solketalamine) (Conversion: 86%, Selectivity: 95%).

Example 3: ring-cleaving hydrolysis with sulfuric acid

Chemical Formula: C ₆ H ₃ N0 ₂ Chemical Formula: C ₃ H _g N0 ₂

Molecular Weight: 131 ,18 Molecular Weight: 91 ,1 1 A flask was charged with 539 g of a 50% aqueous solution of H2SO4 and 655.5 g (2,2-dimethyl- 1 ,3-dioxolan-4-yl)methanamine obtained in example 1 were added dropwise within 2 h at 70 °C. After complete addition, stirring was continued at 70 °C for 1 h. During addition and stirring, acetone was removed continuously by distillation. Acetone removal was completed under reduced pressure and the solution was then treated at ambient temperature with 444.4 g of a 50% aqueous solution of NaOH. The precipitate was removed by filtration and the filter cake was washed with ca. 300 ml. of Ethanol. The combined filtrates were concentrated under reduced pressure and the residue was distilled. 406 g (89% yield) of 3-aminopropane-1 ,2-diol (APD) were collected.

Example 4: ring-cleaving hydrolysis with carbon dioxide

Chemical Formula: C ₆ H ₃ NC>2 Chemical Formula: C ₃ HgN0 ₂

Molecular Weight: 131 ,18 Molecular Weight: 91 ,1 1 An autoclave was charged with 13.1 g of (2,2-dimethyl-1 ,3-dioxolan-4-yl)methanamine (sol- ketalamine) obtained according to example 1 and 39.3 g of H2O. The autoclave was pressurized with 30 bar C02 and the mixture was heated to 130 °C for 12 h. After depressurization, the composition of the mixture was analyzed by GC. The mixture contained 95% (area % determined from gas chromatography) of 3-aminopropane-1 ,2-diol (APD) (Conversion: >99%, Selec- tivity: 95%).