WO2014131674A2 - Amine composition - Google Patents

Abstract

The present invention relates to relates to an amine composition, the use of said amine composition, a curing agent composition, a curable composition, a cured epoxy resin as well as methods for producing said amine composition, said curing agent composition, said curable composition and said cured epoxy resin. Furthermore, the present invention relates to reactive polyamide resins obtainable from said amine composition, a method for their manufacture and their use. Method for the production of the amine composition comprises following steps: A) conversion of formaldehyde (FA), hydrogen cyanide (HCN) and ethylenediamine (EDA); B) hydrogenation of the reaction mixture obtained in step A); and C) purification of the hydrogenated reaction mixture obtained in step B), wherein the purification step C) comprises 1) removal of hydrogen; 2) removal of organic solvent and water; 3) passing the reaction mixture obtained in step 2) to an evaporator operated at a pressure in the range of 5 to 500 mbar and a temperature in the range of 100 to 250 °C; 4) removal of the gaseous phase from the evaporator; and 5) condensation of the gaseous phase.

Description

Amine Composition

Description The present invention relates to an amine composition, the use of said amine composition, a curing agent composition, a curable composition, a cured epoxy resin as well as methods for producing said amine composition, said curing agent composition, said curable composition and said cured epoxy resin.

Furthermore, the present invention relates to reactive polyamide resins obtainable from said amine composition, a method for their manufacture and their use.

Curable compositions on the basis of amine curing agents and epoxy resins are used in the industry on a large scale to produce cured epoxy resins. Common applications include flooring, civil engineering, marine and industrial coatings, adhesives, tooling, composites, castings, com- posite lamination and encapsulations.

Epoxy resins are an important class of polymeric materials, characterized by the presence of more than one epoxy-ring. The epoxy resins are converted into cured epoxy resins, which are solid, infusible and insoluble 3-dimensional networks, with the help of curing agents, which can undergo chemical reactions with the epoxy rings of the epoxy resin.

Commonly amines are used as functional curing agents. These amines can be either primary or secondary amines. A primary amine group can react with two epoxy groups while a secondary amine can react only with one epoxy group. Usually, primary amine groups react much faster than secondary amine groups. Tertiary amines, which have no active hydrogen, will not react with the epoxy groups at all, but will generally act as a catalyst to accelerate the epoxy reaction.

The reactivity of amines depends in the curing reaction on their chemical nature. Aliphatic amines are generally more reactive than cycloaliphatic amines, which are in turn more reactive than aromatic amines. Aliphatic amines are therefore suitable for curing epoxy-resins at room temperature whereas aromatic amines generally require higher curing temperatures.

Aromatic amines are usually employed in applications requiring high temperature stability because they lead to final materials having a high glass transition temperature (T_g). Also aromatic amines result in materials having a good resistance to chemicals. The light stability of aromatic curing agents is on the other hand insufficient for some applications. Since many aromatic amines are solid at room temperatures and due to their lower reactivity, they usually require elevated temperature cures. In addition, the viscosity of the epoxy systems is higher than that of aliphatic or cycloaliphatic amines. Cycloaliphatic amines can result in materials having a T_g ap- proaching those of aromatic amines.

An aliphatic amine composition which is widely used as curing agent for epoxy resins is commercially available triethylenetetramine (TETA). "Commercially available TETA" has been com- monly produced by the reaction of ethylene dichloride with aqueous ammonia, which produces the hydrochloride salts of ethylenediamine and higher homologues. The reaction is usually carried out in the liquid phase without a catalyst. Treatment with caustic soda liberates the free amines. The process yields various derivatives of ethylenediamine (EDA), such as piperazine, diethylenetriamine, triethylenetetramine, tetraethylenepentamine, pentaethylenehexamine and aminoethylpiperazine, which are separated by distillation. However, distillation does not yield pure linear TETA, but a mixture of linear TETA with cyclic and branched compounds.

The composition of "commercially available TETA" is specified, e.g. in the Screening Information Data Set of the Organization for Economic Co-Operation and Development (OECD SIDS) for "triethylenetetramine" (published by UNEP Publications, July 1998, available under www.inchem.org/documents/sids/sids/1 12-24-3.pdf).

According to the above mentioned reference, the content of linear TETA of formula (I)

H

is between 60 to 70%.

Major impurities are:

N,N'-bis(2-aminoethyl)piperazine (DAEPIP): 1 1 -13%;

(piperazinoethyl)ethylenediamine (PEEDA): 10-13%;

tris(aminoethyl)amine (TAEA): 4-6%;

diethylenetriamine (DETA): <= 3%; and

water: <= 0,5%.

Minor amounts of further side products, such as aminoethylethanolamine (AEEA), N-(2- aminoethyl)piperazine (AEPIP), hydroxyethylpyrrolidon (HEP) and tetraethylenepentamine (TE- PA) may be also present. DAEPIP (II), PEEDA (III) and TAEA (IV) are amine compounds comprising a tertiary amine group having the following formulas:

H

/ \

HN (III)

The use of the designation "triethylene tetramine" or "TETA" for the above mentioned mixture gives rise to some confusion, because the designation "triethylene tetramine" or "TETA" commonly used and known in the epoxy industry does not refer to the linear compound of formula I but to the commercially available mixture. In general, the linear compound of formula I is not commercially available in a purity higher than 70%. Therefore references made to "TETA" or "triethylene tetramine" in the literature relating to epoxy resins generally refer to the commercial- ly available mixture and not to the linear compound of formula I .

"Pure TETA" or "purified TETA" having a purity higher than 98% was reported in US-A- 2006/0041 170 and in the above mentioned OECD SIDS. US-A-2006/0041 170 describes the use of essentially pure TETA in pharmaceuticals. However these publications do not refer to the use of "pure" or "purified TETA" in epoxy applications. However, the production of "pure TETA" is laborious requiring several synthesis and purification steps.

WO 201 1/107512 discloses an amine composition comprising 85-98% by weight of linear TETA and 15% by weight or less of one or more amine compounds selected from the group consisting of tertiary amines derived from the condensation of ethylenediamine and methyl-substituted compounds derived from linear triethylenetetramine. Said TETA composition can be obtained by a commercially viable process.

It was reported that the amine compositions of WO 201 1/107512 had lower cure times (gel times) and higher glass transitions temperatures making these amine compositions particularly suitable for applications requiring low cycle times and an improved temperature stability.

In the context of the present invention the term "linear TETA" designates the compound of formula I. The term "commercially available TETA" refers to the industrially and commercially available product having a content of about 60 to 70% by weight of linear TETA as described above. Further, the terms "pure TETA" or "purified TETA" depict compositions comprising linear TETA having a content of linear TETA of 98% by weight or more. "TETA 85" refers to the amine composition disclosed in WO 201 1/107512 having a content of linear TETA of 85% by weight or more.

The object of the present invention is to provide an aliphatic amine composition having improved cold curing and improved cold processing properties compared to "commercially available TETA" and "TETA 85" and which can be produced by a commercially viable process from readily and commercially available raw materials. The problem of the present invention is solved by an amine composition (amine composition I) comprising by weight of linear TETA of formula I ,

by weight of a methyl-substituted TETA,

by weight of diethylenetriamine (DETA),

by weight of methyl-substituted DETA, and

by weight of N-(2-aminoethyl)piperazine (AEPIP).

More preferably the amine composition according to the present invention composition ( composition II) comprises by weight of linear TETA of formula I ,

by weight of a methyl-substituted TETA ,

by weight of diethylenetriamine (DETA),

by weight of methyl-substituted DETA, and

by weight of N-(2-aminoethyl)piperazine (AEPIP). and even more preferably the amine composition according to the present invention composition (amine composition III) comprises

75 to 85% by weight of linear TETA of formula I,

2 to 8% by weight of a methyl-substituted TETA,

3 to 10% by weight of diethylenetriamine (DETA),

1 to 3% by weight of methyl-substituted DETA, and

2 to 10% by weight of N-(2-aminoethyl)piperazine (AEPIP).

In the context of the present invention, "methyl-substituted TETA" (or "Me-TETA") is understood to mean any derivative of linear triethylenetetramine (linear TETA) in which one, two or more of the hydrogen atoms bonded to the four amino functions of the unsubstituted linear TETA are substituted by the corresponding number of methyl groups (CH3-).

"Methyl-substituted TETA" (or "Me-TETA") is characterized by formula (V)

NR

NR, N R^ ^{N R}2 (V) wherein R is either H or CH3; with the proviso that at least one substituent R is CH3. Me-TETAs according to the present invention are shown by way of example in Scheme 1 below as compounds (2) to (13).

Scheme 1

TETA mono-Me-TETAs bis-Me-TETAs tris-Me-TETAs

The term Me-TETA comprises Me-TETA with one (mono-Me-TETA; compounds 2 and 3), two (bis-Me-TETA; compounds 4 to 8) and three methyl substituents (tris-Me-TETA; compounds 9 to 13). In addition, the term Me-TETA also comprises Me-TETAs in which four, five or all six hydrogen atoms of the unsubstituted TETA are substituted by methyl groups (not shown in Scheme 1 ).

Me-TETA is preferably TETA with one methyl substituent. More particularly, the methyl- substituted TETA compound is selected from N-2-aminoethyl-N'-(2-N"-methylaminoethyl)- 1 ,2-ethanediamine and N-2-aminoethyl-N-methyl-N'-2-aminoethyl-1 ,2-ethanediamine.

In the context of the present invention, "methyl-substituted DETA" (or "Me-DETA") is understood to mean any derivative of linear diethylenetriamine in which one, two or more of the hydrogen atoms bonded to the three amino functions of the unsubstituted linear DETA are substituted by the corresponding number of methyl groups (CH3-). In the following, these compounds will be subsumed under the term "Me-DETA".

"Methyl-substituted DETA" (or "Me-DETA") is characterized by formula (VI)

wherein R is either H or CH3; with the proviso that at least one substituent R is CH3.

Me-DETAs according to the present invention are shown by way of example in Scheme 2 below as compounds (2) to (9). Scheme 2

DETA mono-Me-DETAs bis-Me-DETAs tris-Me-DETAs

The term Me-DETA comprises Me-DETA with one (mono-Me-DETA; compounds 2 and 3), two (bis-Me-DETA; compounds 4 to 6) and three methyl substituents (tris-Me-DETA; compounds 7 to 9). In addition, the term Me-DETA also comprises Me-DETAs in which four or all five hydrogen atoms of the unsubstituted DETA are substituted by methyl groups (not shown in Scheme 2).

Me-DETA is preferably DETA with one methyl substituent. More particularly, the methyl- substituted DETA compound is selected from N'-[2-(methylamino)ethyl]ethane-1 ,2-diamine and N'-(2-aminoethyl)-N'-methyl-ethane-1 ,2-diamine. The amount of water, organic solvents and other organic side products in the amine compositions of the present invention is preferably less than 5% by weight, more preferably less than 2% by weight and most preferably less than 1 % by weight.

Preferably the amine compositions l-lll comprise less than 4% by weight, preferably less than 3% by weight, more preferably less than 2% by weight and most prerferably less than 1 % by weight of a compound selected from the group consisting of DAEPIP, PEEDA and TAEA.

In a preferred embodiment the amine compositions according to the present invention composition (amine composition IV) consists of

60 to 95% by weight of linear TETA of formula I,

1 to 10% by weight of a methyl-substituted TETA,

1 to 12% by weight of diethylenetriamine (DETA),

0.1 to 5% by weight of methyl-substituted DETA,

1 to 20% by weight of N-(2-aminoethyl)piperazine (AEPIP), and

0.1 to 5% by weight of organic side products, organic solvent and water. More preferably the amine composition according to the present invention composition (amine composition V) consists of

70 to 90% by weight of linear TETA of formula I ,

2 to 8% by weight of a methyl-substituted TETA ,

2 to 12% by weight of diethylenetriamine (DETA),

0.5 to 4% by weight of methyl-substituted DETA, and

2 to 15% by weight of N-(2-aminoethyl)piperazine (AEPIP), and

0.2 to 4% by weight of organic side products, organic solvent and water. and even more preferably the amine composition according to the present invention composition (amine composition VI )consists of

75 to 85% by weight of linear TETA of formula I ,

2 to 8% by weight of a methyl-substituted TETA,

3 to 10% by weight of diethylenetriamine (DETA),

1 to 3% by weight of methyl-substituted DETA, and

2 to 10% by weight of N-(2-aminoethyl)piperazine (AEPIP, and

0.2 to 4% by weight of organic side products, organic solvent and water.

The amine compositions of the present invention can be produced by a process comprising the steps of:

A) conversion of formaldehyde (FA), hydrogen cyanide (HCN) and ethylenediamine (EDA); B) hydrogenation of the reaction mixture obtained in step A); and

C) purification of the hydrogenated reaction mixture obtained in step B), wherein the purification step C) comprises

1 ) removal of hydrogen;

2) removal of organic solvent and water;

3) passing the reaction mixture obtained in step 2) to an evaporator operated at a pressure in the range of 5 to 500 mbar and a temperature in the range of 100 to 250 °C;

4) removal of the gaseous phase from the evaporator; and

5) condensation of the gaseous phase. Conversion of formaldehyde (FA), hydrogen cyanide (HCN) and ethylenediamine (EDA) (Step

A):

The amine composition of the present invention can be produced by conversion of FA, HCN and EDA (Step A). EDA can be prepared by the EDC (ethylene dichloride) process by reaction of ethylene dichlo- ride (EDC) with ammonia in the aqueous phase. Details of the process are given, for example, in Ullmann (article "Amines, aliphatic" in Ullmann's Encyclopedia of Industrial Chemistry, Karsten Eller, Erhard Henkes, Roland Rossbacher and Hartmut Hoke, Published Online: JUNE 15, 2000, DOI: 10.1002/14356007.a02_001 , page 33).

A further means of preparing EDA consists in the catalytic reaction of monoethanolamine (ME- OA) with ammonia (article "Amines, aliphatic" in Ullmann's Encyclopedia of Industrial Chemistry, Karsten Eller, Erhard Henkes, Roland Rossbacher and Hartmut Hoke, Published Online: JUNE 15, 2000, DOI: 10.1002/14356007.a02_001 , page 33 or Hans-Jurgen Arpe, Industrielle Organ- ische Chemie [Industrial Organic Chemistry], 6th edition (2007), Wiley VCH, 2007).

EDA can also be obtained by hydrogenation of aminoacetonitrile (AAN), AAN being preparable by reaction of hydrogen cyanide, formaldehyde (FA) and ammonia.

The hydrogenation of AAN to EDA is described, for example, in WO 2008/104583.

EDA is preferably used in the form of its free base, but it is optionally also possible to use salts such as the dihydrochloride of EDA as the reactant.

The purity of the EDA used in the process is preferably 95% by weight or more, more preferably 98% by weight or more, even more preferably 99% by weight or more and especially preferably 99.5% by weight or more. A further reactant used is formaldehyde.

Formaldehyde is a chemical widely available commercially.

Preference is given to using formaldehyde as a 30 to 50% aqueous solution.

In addition, hydrogen cyanide is used in the process according to the invention.

Hydrogen cyanide is likewise a chemical widely available commercially.

Hydrogen cyanide can be prepared on the industrial scale essentially by three different processes. In a first process, hydrogen cyanide can be obtained by ammoxidation of methane with oxygen and ammonia (Andrussow process). In a second process, hydrogen cyanide can be obtained from methane and ammonia by ammodehydrogenation in the absence of oxygen. Fi- nally, hydrogen cyanide can be prepared on the industrial scale by dehydration of formamide. In general, an acidic stabilizer is added to the hydrogen cyanide prepared by these processes, for example SO2, sulfuric acid, phosphoric acid or an organic acid such as acetic acid, in order to prevent the autocatalytic polymerization of hydrogen cyanide, which can lead to blockages in pipelines.

Hydrogen cyanide can be used in liquid or gaseous form, in pure form or as an aqueous solution.

Hydrogen cyanide is preferably used as a 50 to 95% by weight, more preferably as a 75 to 90% by weight, aqueous solution.

Hydrogen cyanide is preferably used in a purity of 90% by weight or more.

Preference is given to using stabilizer-free HCN.

If a stabilized HCN is used, it is preferable that the stabilizer is an organic acid. In a preferred embodiment, HCN is used which is with substantial freedom from cyano salts such as KCN.

The conversion of EDA, HCN and FA preferably takes place in the presence of water.

The reaction of EDA, HCN and FA generally gives rise to 1 mol of water per mole of formaldehyde used.

However, water can also be supplied additionally, for example by using the reactants in the form of aqueous solutions thereof. More particularly as described above, it is generally possible to use FA and/or HCN as an aqueous solution.

The amount of water used is generally in the range from 1 to 50 mol per mole, preferably in the range from 2 to 40 mol and more preferably in the range from 3 to 30 mol per mole of EDA used. The conversion of EDA, HCN and FA preferably takes place in the presence of an organic solvent.

The organic solvents used are preferably those selected from the group consisting of aliphatic, cycloaliphatic, araliphatic, aromatic hydrocarbons, alcohols and ethers.

It is especially preferable that the organic solvent is stable under the conditions of a subsequent hydrogenation of EDDN.

It is also preferable that the organic solvent is condensable within the range from 20 to 50°C at a pressure in the range from 50 to 500 mbar.

It is also preferable that the organic solvent boils at a sufficiently low temperature to be able to establish a bottom temperature of less than 100°C in the subsequent removal of water during the workup of the reaction effluent.

Preferred organic solvents are, for example, cyclohexane, methylcyclohexane, toluene, N- methylmorpholine, o-xylene, m-xylene or p-xylene, anisole, n-pentane, n-hexane, n-heptane, n- octane, n-nonane, diisobutyl ether, light gasoline, gasoline, benzene, diglyme, tetrahydrofuran, 2- and 3- methyltetrahydrofuran (MeTHF) and cyclohexanol, or mixtures of these compounds. Particularly preferred solvents are cyclohexane, methylcyclohexane, toluene, N- methylmorpholine, o-xylene, m-xylene or p-xylene, anisole, n-pentane, n-hexane, n-heptane, n- octane, n-nonane, diisobutyl ether, light gasoline, gasoline (benzene), diglyme and MeTHF, or mixtures of these compounds.

The amount of organic solvent is generally 0.1 to 50 kg per kg, preferably 1 to 30 kg and more preferably 3 to 25 kg per kg of EDA used.

In a particularly preferred process variant, in the conversion of FA, EDA and HCN, an organic solvent having a boiling point between water and EDDN is used, especially under the conditions of the distillative depletion of water described below. As described below, organic solvents which boil within this range enable particularly efficient removal of water from the reaction efflu- ent which is obtained in the conversion of FA, HCN and EDA. Particularly preferred solvents having a boiling point between water and EDDN are toluene, N-methylmorpholine, o-xylene, m- xylene or p-xylene, anisole, n-octane, n-nonane, diisobutyl ether or diglyme, or mixtures thereof. Some of the aforementioned organic solvents can form a low-boiling azeotrope with water. A low-boiling azeotrope corresponds, in the p, x diagram, to the substance mixture at the maximum vapor pressure. The boiling point of this mixture has a minimum in the T, x diagram and is below that of the pure substances involved.

Particularly preferred organic solvents which have a boiling point between water and EDDN and which form a low-boiling azeotrope with water are toluene, N-methylmorpholine, o-xylene, m- xylene or p-xylene, anisole, n-octane, n-nonane, diisobutyl ether and diglyme, or mixtures thereof.

If the organic solvent having a boiling point between water and EDDN forms a low-boiling azeo- trope with water, it is also preferred that the organic solvent has a miscibility gap or sparing solubility in water, more particularly under the conditions of the workup steps described hereinafter. This facilitates the later separation of water and organic solvents. The solubility of such an organic solvent is preferably 1 % by weight or less, more preferably 0.5% by weight or less and especially preferably 0.1 % by weight or less. In particular, toluene is preferred as such an or- ganic solvent.

In a further preferred embodiment, in the conversion of FA, EDA and HCN, an organic solvent which has a boiling point below the boiling point of water and which forms a low-boiling azeotrope with water, especially under the conditions of the distillative removal of water described below, is used.

Particularly preferred solvents which have a boiling point below the boiling point of water and which form a low-boiling azeotrope with water are n-pentane, n-hexane, n-heptane, tetrahydro- furan, cyclohexane, methylcyclohexane, light gasoline, gasoline (benzene) or mixtures thereof. Such a solvent under standard conditions should preferably have a boiling point of at least 50°C and more preferably of at least 60°C in order thus to attain sufficiently high condensation temperatures that the use of brine in the condenser can be avoided.

It is additionally preferred that the solvent used which has a boiling point below the boiling point of water and which forms a low-boiling azeotrope with water has a low solubility in water or a miscibility gap with water under the conditions which exist in the conversion of FA, HCN and EDA or the subsequent workup. This facilitates the later separation of water and organic solvents. The solubility of such an organic solvent in water is preferably 1 % by weight or less, more preferably 0.5% by weight or less and especially preferably 0.1 % by weight or less.

In a very particularly preferred embodiment, the conversion of EDA, FA and HCN is performed in the presence of toluene as a solvent, and the subsequent hydrogenation is performed in the presence of THF. As described below, it is thus possible to establish a particularly efficient integrated solvent system which allows the recycling of the organic solvents into the process. In addition, it has been recognized that the presence of THF during the subsequent hydrogena- tion, especially when the hydrogenation is performed in suspension mode, can reduce the ag- glomeration tendency of the suspension catalysts used.

Reaction processes for converting EDA, HCN and FA in the presence of water are described, for example in WO 2008/104579, the contents of which are explicitly incorporated by reference. According to the teaching of WO 2008/104579, the conversion of FA, HCN and EDA can be performed according to options a) to d) described therein, the reactants generally being converted to mixtures comprising EDDN.

The preparation can be effected, for example, by a) first converting HCN and FA to FACH, which is subsequently reacted with EDA, or by b) reacting an ethylenediamine-formaldehyde adduct (EDFA) with hydrogen cyanide, EDFA being obtainable by reacting EDA with FA, or by c) reacting EDA with a mixture of formaldehyde and hydrogen cyanide, or by d) reacting EDA simultaneously with formaldehyde and HCN. Options a) to d) described in WO 2008/104579 are preferably performed at a temperature of 10 to 90°C, especially at 30 to 70°C. The reaction can be performed at standard pressure or else optionally at elevated pressure (superatmospheric pressure).

Preferably, options a) to d) are performed in a tubular reactor or a stirred tank cascade. Prefer- ably, the conversion of FA, HCN and EDA can also be performed as a continuous process, especially as an industrial scale process.

Hereinafter, further details of process options a) to d) are described, as are, in some cases, preferred embodiments of the respective options.

Option a)

The amine compositions of the present invention can be prepared according to option a) from HCN, FA and EDA, by first reacting FA with HCN to give FACH and then FACH with EDA.

The preparation of FACH is described, for example, in Ullmann (article "Formaldehyde" in Ullmann's Encyclopedia of Industrial Chemistry, Gunther Reuss, Walter Disteldorf, Armin Otto Gamer and Albrecht Hilt, Published Online : JUNE 15, 2000, DOI: 10.1002/14356007.a1 1_619, p. 28). It can be prepared, for example, by reacting formaldehyde with an aqueous hydrogen cyanide. A preferred variant for preparation of FACH is described in WO 2008/104579. According to this, FACH can be effected by reaction of aqueous formaldehyde with hydrogen cyanide. Formaldehyde is preferably in the form of a 30 to 50% aqueous solution; hydrogen cyanide is preferably used in 90 to 100% purity. This reaction is effected preferably at a pH of 5.5, which is preferably established with sodium hydroxide solution or ammonia. The reaction can be effected at tem- peratures of 20 to 70°C, for example in a loop reactor and/or tubular reactor. Instead of purified hydrogen cyanide (HCN), it is also possible to chemisorb crude HCN gas into FACH in an aqueous formaldehyde solution under the conditions specified above. The crude HCN gas is preferably prepared by pyrolysis of formamide and comprises, as well as water, small proportions of ammonia in particular. The resulting aqueous FACH solution can optionally be concen- trated by gentle vacuum concentration, for example with a falling-film or thin-film evaporator. Preference is given to concentrating to a 50-80% by weight aqueous FACH solution. Before the concentration, it is advantageous to stabilize the FACH solution by lowering the pH to < 4, preferably to < 3, for example by adding acid, for example by adding phosphoric acid or preferably sulfuric acid.

Preferably, a 50 to 80% by weight aqueous solution of FACH is used in the process according to option a).

In general, the molar ratio of EDA to FACH according to option a) in the reaction of EDA with FACH is in the range from 1 :1 to 1 :2 [mol/mol].

Preferably, in option a), the molar ratio of EDA to FACH is about 1 :1 .8 to 1 :2 [mol/mol], especially approx. 1 :2 [mol/mol].

In general, the conversion of FACH and EDA can be performed according to the general pro- cess conditions described above.

Option b)

The preparation of the amine compositions according to the present invention can also be ef- fected according to option b), by reacting FA with EDA to give EDFA, which can then react further with HCN to give EDDN.

According to option b), EDA is first reacted with FA to give EDFA. In a preferred embodiment, no organic solvent is fed in before or during the reaction of EDA with FA to give EDFA . The reaction takes place preferably in the presence of water, since FA, as described above, preferably used in the form of aqueous solutions.

The reaction of EDA (VII) with FA to give EDFA (VIII)) generally proceeds sufficiently rapidly that generally no catalyst is required.

For better clarity, EDFA (VIII) is represented in the formula as a hemiaminal. The preparation of EDFA generally proceeds via the intermediate EDMFA (IX), which is formed from one mole of EDA and one mole of formaldehyde.

The molar ratio of EDA to formaldehyde is 1 :1 .8 to 1 :2.2, preferably 1 :1.9 to 1 :2.1 , more prefer- ably 1 :2 to 1 :2.1 .

The pressure maintained in the reaction of EDA with FA is uncritical and generally merely has to be sufficiently high that the reactor contents are liquid. There is no upper limit, and it is preferably 1 to 10 bar, more preferably 2 to 5 bar.

The reaction of FA with EDA is preferably continuous.

For the continuous reaction of EDA with formaldehyde, it is possible to use all reactors suitable for liquid phase reactions.

Preferably, the process according to option b) is performed in a tubular reactor or a stirred tank reactor or a loop reactor, especially a loop reactor.

A loop reactor is understood hereinafter to mean a reactor in which the reactor contents are circulated. After flowing through the reactor, the reaction input can be cooled in a cooling apparatus, for example a heat exchanger, a substream of the cooled stream can be recycled into the reactor and the remaining stream can be passed into the next process stage. The loop may be an internal or external loop. The external loop can preferably be cooled in a cooling apparatus, for example a heat exchanger, especially a plate heat exchanger, shell and tube heat exchang- er or jacketed heat exchanger.

By leading off the heat of reaction which arises, for example, in the hydration of EDA or in the reaction of FA with EDA, the temperature rise in the reactor can be controlled efficiently. The residence time in the loop reactor is preferably 5 seconds to 60 minutes, more preferably 30 seconds to 20 minutes.

When the conversion to EDFA is effected in a loop reactor in which backmixing occurs, the con- version is generally incomplete. It is generally in the range from 50 to 99%.

In a very particularly preferred embodiment, a combination of loop reactor and downstream tubular reactor is therefore used as the reactor. The temperature in the conversion of FA and EDA to EDFA is generally within the range from 0 to 100°C, preferably in the range from 20 to 50°C and more preferably in the range from 25 to 45°C.

In variant b), EDFA, after preparation thereof, is subsequently reacted further with HCN to give EDDN.

Preferably, EDFA is reacted with HCN without further workup.

The molar ratio of EDFA to hydrogen cyanide (HCN) is preferably 1 :1 .8 to 1 :2.2, more preferably 1 :1.9 to 1 :2.0.

In general, the conversion of EDFA and HCN can be performed according to the general process conditions described above.

Option c)

The preparation of the amine compositions according to the present invention can also be effected according to option c), by reacting EDA with a mixture of formaldehyde and hydrogen cyanide (GFB). In general, the reaction of EDA with a mixture of formaldehyde and hydrogen cyanide can be performed according to the general process conditions described above.

The molar ratio of FA and hydrogen cyanide in the GFB is generally in the range from 0.5:1 to 1 .5:1 .

The molar ratio of EDA to GFB is preferably 1 :1 .5 to 1 :2 [mol/mol]. The molar ratio of EDA to GFB is preferably 1 :1.8 to 1 :2 [mol/mol]. Preferably, the GFB is prepared by mixing approximately equimolar amounts of formaldehyde and hydrogen cyanide.

Preferably, the reaction mixture is cooled at the outlet of the reactor. The cooling of the reaction mixture can be performed as described above and in detail hereinafter. Option d)

A further variant for preparation the amine compositions of the present invention consists, according to option d), in reacting EDA with formaldehyde and hydrogen cyanide (HCN) simulta- neously (in parallel).

The molar ratio of EDA to formaldehyde to HCN is typically 1 :1 .5:1.5 to 1 :2:2 [mol/mol/mol]. The molar ratio of EDA to formaldehyde to HCN is preferably 1 :1.8:1 .8 to 1 :2:2 [mol/mol/mol]. Preferably, in this embodiment, the three reactant components are added to the reaction vessel simultaneously or stepwise in equal molar proportions based on the particular total amount of reactant.

In general, the simultaneous (parallel) reaction of EDA with formaldehyde and hydrogen cyanide (HCN) can be performed according to the general process conditions described above.

Preferably, the reaction effluent from conversion of FA, EDA and HCN according to the variants a) through d) is worked up by performing first i) a low boiler removal and then ii) a water depletion. The low boilers are preferably depleted by stripping. For example, the reaction effluent from the conversion of EDA, HCN and FA can be stripped with nitrogen in order to remove traces of hydrogen cyanide which may occur, for example, as a decomposition product of FACH.

However, low boilers can also be removed by distillation. If low boilers are removed by distilla- tion, it is preferable that the residence time in the distillation is kept brief, for example by performing the distillation in a falling-film evaporator or wiped-film evaporator.

The low boiler removal is preferably effected by flash evaporation, wherein low boilers are removed as the gasesous phase. The water depletion after the depletion of low boilers is preferably effected in a distillation column K1 .

The column is generally operated in such a way that an aqueous stream is drawn off at the top of the column, while an EDDN-containing stream is drawn off at the bottom of the column. The effluent from variants a) through d) is preferably supplied to a distillation column K1 togeth- er with a distilling agent (as defined hereinafter) in the upper region, preferably at the top.

Preferably, the distillation column K1 has internals for increasing the separating performance. The distillative internals may be present, for example, in the form of a structured packing, for example as a sheet metal packing such as Mellapak 250 Y or Montz Pak, B1 -250 type. It is also possible for a packing with lower or increased specific surface area to be present, or it is possible to use a fabric packing or a packing with another geometry such as Mellapak 252 Y. An advantage in the case of use of these distillative internals is the low pressure drop and the low specific liquid holdup compared, for example, to valve trays. The internals may be present in one or more beds.

The number of theoretical plates is generally in the range from 3 to 25, preferably 5 to 15.

It is preferable that the bottom temperature is 100°C or less since it has been found in the con- text of the present invention that EDDN is unstable at relatively high temperatures in the presence of water and decomposes to give unwanted by-products. Preference is given to establishing a bottom temperature in the region of less than 100°C, more preferably less than 80°C and most preferably less than 60°C. The bottom temperature is preferably in the range from 20 to 100°C, more preferably in the range from 30 to 80°C and most preferably in the range from 40 to 60°C.

The top pressure in the column K1 is preferably adjusted such that the bottom temperature is within the range specified below.

The top pressure is preferably 10 mbar to 1 bar, more preferably 30 mbar to 700 mbar and most preferably 50 to 500 mbar.

In a very particularly preferred embodiment, distillation is performed in the presence of an organic solvent which has a boiling point between water and EDDN at the distillation pressure existing in the column or which forms a low-boiling azeotrope with water, which is hereinafter referred to hereinafter as distilling agent.

Preferred distilling agents are the organic solvents cited at the outset which have a boiling point between water and EDDN or which form a low-boiling azeotrope with water.

The preferred distilling agent is toluene.

It is preferable that the distilling agent is fed in before or during the conversion of FA, HCN and EDA. The amount of distilling agent should generally be such that - as described above - preferably a bottom temperature in the region of less than 100°C, more preferably less than 80°C and most preferably less than 60°C is established in the column bottom of distillation column K1 .

When the distilling agent forms a low-boiling azeotrope with water, it is necessary that the amount of distilling agent is sufficiently great to be on the right "side" of the azeotrope, which means that the amount of distilling agent must be sufficiently great that the low-boiling aqueous azeotrope is obtained at the top of the column, and essentially no further water is obtained in the column bottoms. The amount of solvent required can be determined in a routine manner by the person skilled in the art, as a function of the distilling agent selected, from commonly known tables and reference works for azeotropes.

The condenser of distillation column K1 is generally operated at a temperature at which the predominant portion of the water or of the water azeotrope is condensed at the appropriate top pressure. In general, the operating temperature of the condenser is in the range from 20 to 70°C, preferably 25 to 50°C.

In the condenser, a condensate comprising essentially water or a low-boiling water azeotrope is generally obtained.

The condensate of column K1 can either be discharged or recycled into the process. In the bottom of column K1 , the bottom product drawn off is preferably a mixture comprising EDDN.

The EDDN-containing mixture preferably comprises the distilling agent used in the distillative depletion of water.

If the distilling agent used is toluene, the EDDN-containing mixture from the bottom of column K1 comprises preferably 5 to 30% by weight of toluene and even more preferably 10 to 20% by weight and especially preferably 12 to 18% by weight, based on the bottoms discharged.

The EDDN-containing mixture from the bottom of column K1 comprises preferably less than 3% by weight, more preferably less than 1 % by weight of water, even more preferably less than 0.5% by weight and especially preferably less than 0.3% by weight of water.

The EDDN- -containing mixture thus obtained can be hydrogenated directly in a subsequent reaction with hydrogen and in the presence of a catalyst to give the amine compositions of the present invention.

In a further particularly preferred embodiment, the EDDN-containing mixture after the water depletion is, however, purified before the hydrogenation of the EDDN by treating the EDDN- containing mixture with an adsorbent.

Preferred solid acidic adsorbents are acidic metal oxides such as silicon dioxide, titanium dioxide, aluminum oxide, boron oxide (B2O3), zirconium dioxide, silicates, aluminosilicates, borosili- cates, zeolites (especially in the H form), acidic ion exchangers, and silica gel, e.g. Sorbead WS from BASF SE, or mixtures of these substances.

Very particularly preferred solid acidic adsorbents are silicon dioxide and silica gel.

Very particular preference is given to silica gels, which can be produced, for example, by acidifying aqueous sodium waterglass solutions and drying the silica sols obtained at first, as described, for example, in Hollemann-Wiberg (Lehrbuch der Anorganischen Chemie [Inorganic Chemistry], 102nd edition, Walter de Gruyter publishers, 2007, page 962). Examples of particu- larly preferred silica gels are Sorbead WA from BASF SE and KG 60 silica gel from Merck KGaA.

In a preferred embodiment, the solid acidic adsorbent is a substance selected from the group consisting of silicon dioxide, titanium dioxide, aluminum oxide, boron oxide (B2O3), zirconium dioxide, silicates, aluminosilicates, borosilicates, zeolites (especially in the H form), acidic ion exchangers and silica gel.

The treatment can be effected batchwise, for example by contacting the adsorbent with the EDDN-comprising mixture in the presence of an organic solvent. The treatment can be effected by suspending the adsorbent in the mixture to be purified, for example by stirring in a suitable vessel. The treatment time in the batchwise treatment is generally in the range from 1 minute up to 48 hours, preferably 5 minutes to 24 hours, more preferably 1 hour to 16 hours and especially preferably 2 to 8 hours. The amount of adsorbent is preferably in the range from 0.1 to 25% by weight, more preferably in the range from 0.5 to 20% by weight and most preferably in the range from 1 to 10% by weight, based on the sum of EDDN and organic solvent.

The pressure is generally not critical. However, it is preferable to establish a pressure at which the mixture to be purified is in liquid form. The pressure is generally 1 to 10 bar.

The treatment is effected generally at temperatures of less than 150°C, preferably less than 100°C, more preferably less than 80°C and especially preferably less than 60°C.

The batchwise treatment with adsorbent can be effected under an inert gas atmosphere, for example under nitrogen or argon.

After the treatment, the adsorbent can be removed from the EDDN-comprising mixture by suita- ble processes, for example by filtration, centrifugation or sedimentation.

Hydrogenation (Step B):

In Step B) the reaction mixture obtained by conversion of formaldehyde (FA), hydrogen cyanide (HCN) and ethylenediamine (EDA) in step A) is subsequently hydrogenated.

The hydrogenation can be carried out as described below.

The hydrogenation takes place in general by reaction of EDDN with hydrogen in the presence of a catalyst and an organic solvent.

The mixture which is introduced into the hydrogenation preferably comprises EDDN. The fraction of EDDN in the mixture supplied to the hydrogenation is preferably in the range from 5 to 50% by weight, more preferably 8 to 30% by weight and very preferably 10 to 20% by weight.

The mixture which is introduced into the hydrogenation preferably comprises the organic solvent which was present at the treatment with adsorbent.

Furthermore, the mixture which is introduced into the hydrogenation comprises a distillation agent which preferably was used in the distillative depletion of water after convesion of FA, EDA and HCN.

The hydrogen used in the hydrogenation is generally used in technical grade purity. The hydrogen can also be used in the form of a hydrogen-comprising gas, i.e. with additions of other inert gases, such as nitrogen, helium, neon, argon or carbon dioxide. The hydrogen-comprising gases used may, for example, be reformer offgases, refinery gases, etc., if and provided that these gases do not comprise any catalyst poisons for the hydrogenation catalysts used, for example CO. However, preference is given to using pure hydrogen or essentially pure hydrogen in the process, for example hydrogen with a content of more than 99% by weight of hydrogen, preferably more than 99.9% by weight of hydrogen, more preferably more than 99.99% by weight of hydrogen, especially more than 99.999% by weight of hydrogen.

The hydrogenation preferably takes place in the presence of an organic solvent.

It is preferred for the organic solvent to be the same solvent that was present at the treatment with adsorbent. It is, however, also possible to add a further solvent or to separate off the sol- vent which was present during the treatment with adsorbent and to add a new solvent. As organic solvent it is possible to use all organic solvents which can be employed in the conversion of EDA, FA and HCN, especially the organic solvents stated as being preferred.

The weight ratio of organic solvent to EDDN during the hydrogenation is preferably 0.01 :1 to 99:1 , more preferably 0.05:1 to 19:1 and most preferably 0.5:1 to 9:1.

However, it is very particularly preferred that the hydrogenation is performed in the presence of THF since the agglomeration tendency of catalysts in suspension mode can be reduced in THF. More preferably, the hydrogenation takes place in the presence of a sufficient amount of THF that the content of EDDN during the hydrogenation is preferably in the range from 5 to 50% by weight, more preferably 8 to 30% by weight and most preferably 10 to 20% by weight.

It is further preferred that the preparation of EDDN is effected in the presence of toluene, as described above.

The hydrogenation can also be effected in the presence of water.

However, it is preferable not to supply any further water since both EDDN tend to decompose in the presence of water.

Preference is given to using a hydrogenation feed comprising less than 3% by weight, preferably less than 1 % by weight, more preferably less than 0.5% by weight of water and especially preferably less than 0.3% by weight, based on EDDN.

Very particular preference is given to using a hydrogenation feed comprising less than 0.1 % by weight and especially preferably less than 0.03% by weight of water, based on EDDN. In a further preferred process variant, the hydrogenation takes place in the presence of basic compounds, which are preferably added to the reaction mixture in suitable solvents, such as alkanols, such as C1 -C4 alkanols, e.g. methanol or ethanol, or ethers, such as cyclic ethers, e.g. THF or dioxane.

Particular preference is given to adding solutions of alkali metal or alkaline earth metal hydrox- ides or of hydroxides of the rare earth metals in water, more preferably solutions of LiOH, NaOH, KOH and/or CsOH. Preference is given to supplying a sufficient amount of alkali metal and/or alkaline earth metal hydroxide that the concentration of alkali metal and/or alkaline earth metal hydroxide based on the mixture to be hydrogenated is in the range from 0.005 to 1 % by weight, more preferably 0.01 to 0.5% by weight and most preferably 0.03 to 0.1 % by weight.

However, the basic compounds used may also preferably be amides and/or amines, such as ammonia and EDA.

Addition of such basic additives allows the amount of by-products formed, for example AEPIP, in the hydrogenation to be reduced.

Preferred examples of such additives are ammonia and ethylenediamine.

The amount of these additives is 0.01 to 10 mol per mole of EDDN.

The basic additives can generally be supplied batchwise or continuously, and before and/or during the hydrogenation. The catalysts used for hydrogenation of the nitrile function to the amine may be catalysts which comprise, as the active species, one or more elements of transition group 8 of the Periodic Table (Fe, Co, Ni, Ru, Rh, Pd, Os, Ir, Pt), preferably Fe, Co, Ni, Ru or Rh, more preferably Co or Ni. These include what are called oxidic catalysts, which comprise one or more active species in the form of oxygen compounds thereof, and what are called skeletal catalysts (also referred to as Raney® type; hereinafter also Raney catalyst), which are obtained by leaching (activation) of an alloy composed of hydrogenation-active metal and a further component (preferably Al). The catalysts may additionally comprise one or more promoters.

In a particularly preferred embodiment, in the hydrogenation of EDDN, Raney catalysts are used, preferably Raney cobalt or Raney nickel catalysts and more preferably Raney cobalt catalysts doped with at least one of the elements Cr, Ni or Fe, or Raney nickel catalysts doped with one of the elements Mo, Cr or Fe.

The catalysts can be used in the form of unsupported catalysts or in supported form. The supports employed preferably include metal oxides such as AI2O3, S1O2, Zr02, T1O2, mixtures of metal oxides or carbon (activated carbons, carbon blacks, graphite). Before use, the oxidic catalysts are activated at elevated temperature by reduction of the metal oxides in a hydrogen-comprising gas stream outside the reactor or within the reactor. If the catalysts are reduced outside the reactor, this may be followed by a passivation by an oxygen- comprising gas stream or embedding into an inert material in order to prevent uncontrolled oxidation under air and to enable safe handling. The inert material used may be organic solvents such as alcohols, or else water or an amine, preferably the reaction product. An exception in terms of activation is that of the skeletal catalysts, which can be activated by leaching with aqueous base, as described, for example, in EP-A 1 209 146. According to the process performed (suspension hydrogenation, fluidized bed process, fixed bed hydrogenation), the catalysts are used in the form of powder, spall or shaped bodies (preferably extrudates or tablets).

Particularly preferred fixed bed catalysts are the unsupported cobalt catalysts disclosed in EP-A1 742 045, doped with Mn, P and alkali metal (Li, Na, K, Rb, Cs). The catalytically active composition of these catalysts before the reduction with hydrogen consists of 55 to 98% by weight, especially 75 to 95% by weight, of cobalt, 0.2 to 15% by weight of phosphorus, 0.2 to 15% by weight of manganese and 0.05 to 5% by weight of alkali metal, especially sodium, calculated in each case as the oxide.

Further suitable catalysts are the catalysts disclosed in EP-A 963 975, the catalytically active composition of which before the treatment with hydrogen comprises 22 to 40% by weight of Zr02, 1 to 30% by weight of oxygen compounds of copper, calculated as CuO, 15 to 50% by weight of oxygen compounds of nickel, calculated as NiO, where the molar Ni:Cu ratio is greater than 1 , 15 to 50% by weight of oxygen compounds of cobalt, calculated as CoO, 0 to 10% by weight of oxygen compounds of aluminum and/or of manganese, calculated as AI2O3 and Mn02 respectively, and no oxygen compounds of molybdenum, for example the catalyst A disclosed in this document with the composition of 33% by weight of Zr, calculated as Zr02, 28% by weight of Ni, calculated as NiO, 1 1 % by weight of Cu, calculated as CuO, and 28% by weight of Co, calculated as CoO.

Additionally suitable are the catalysts disclosed in EP-A 696 572, the catalytically active compo- sition of which before the reduction with hydrogen comprises 20 to 85% by weight of Zr02, 1 to 30% by weight of oxygen compounds of copper, calculated as CuO, 30 to 70% by weight of oxygen compounds of nickel, calculated as NiO, 0.1 to 5% by weight of oxygen compounds of molybdenum, calculated as M0O3, and 0 to 10% by weight of oxygen compounds of aluminum and/or manganese, calculated as AI2O3 and Mn02 respectively. For example, the catalyst dis- closed specifically in this document with the composition of 31 .5% by weight of Zr02, 50% by weight of NiO, 17% by weight of CuO and 1.5% by weight of M0O3. Equally suitable are the catalysts described in WO-A-99/44984 comprising (a) iron or a compound based on iron or mixtures thereof, (b) from 0.001 to 0.3% by weight, based on (a), of a promoter based on 2, 3, 4 or 5 elements selected from the group of Al, Si, Zr, Ti, V, (c) from 0 to 0.3% by weight, based on (a), of a compound based on an alkali metal and/or alkaline earth metal and d) from 0.001 to 1 % by weight, based on (a), of manganese.

For suspension processes, preference is given to using Raney catalysts. In the Raney catalysts, the active catalyst is produced as a "metal sponge" from a binary alloy (nickel, iron, cobalt, with aluminum or silicon) by leaching out one partner with acid or alkali. Residues of the original alloy partner often act synergistically. The Raney catalysts used for hydrogenation of EDDN are preferably prepared proceeding from an alloy of cobalt or nickel, more preferably cobalt, and a further alloy component which is soluble in alkalis. In this soluble alloy component, preference is given to using aluminum, but it is also possible to use other components such as zinc and silicon or mixtures of such compo- nents.

To activate the Raney catalyst, the soluble alloy component is extracted completely or partially with alkali, for which it is possible to use aqueous sodium hydroxide solution, for example. The catalyst can then be washed, for example with water or organic solvents.

Individual or several further elements may be present in the catalyst as promoters. Examples of promoters are metals of transition groups IB, VIB and/or VIII of the Periodic Table, such as chromium, iron, molybdenum, nickel, copper, etc. The activation of the catalysts by leaching out the soluble component (typically aluminum) can be effected either in the reactor itself or before introduction into the reactor. The preactivated catalysts are air-sensitive and pyrophoric and are therefore generally stored and handled under a medium, for example water, an organic solvent or a substance present in the subsequent hydrogenation (solvent, reactant, product), or embedded into an organic compound solid at room temperature.

In a preferred embodiment, a Raney cobalt skeletal catalyst is used, which has been obtained from a Co/AI alloy by leaching with aqueous alkali metal hydroxide solution, for example sodium hydroxide solution, and subsequent washing with water, and preferably comprises at least one of the elements Fe, Ni or Cr as promoters.

Such preferred Raney Co catalysts typically comprise, as well as cobalt, also 1 -30% by weight of Al, particularly 2-12% by weight of Al, very particularly 3-6% by weight of Al, 0-10% by weight of Cr, particularly 0.1 -7% by weight of Cr, very particularly 0.5-5% by weight of Cr, especially 1 .5-3.5% by weight of Cr, 0-10% by weight of Fe, particularly 0.1 -3% by weight of Fe, very particularly 0.2-1 % by weight of Fe, and/or 0-10% by weight of Ni, particularly 0.1 -7% by weight of Ni, very particularly 0.5-5% by weight of Ni, especially 1 -4% by weight of Ni, where the weight figures are each based on the total catalyst weight. The catalysts used in the hydrogenation may, for example, advantageously be a "Raney 2724" cobalt skeletal catalyst from W. R. Grace & Co. This catalyst has the following composition: Al: 2-6% by weight, Co: > 86% by weight, Fe: 0-1 % by weight, Ni: 1 -4% by weight, Cr: 1.5-3.5% by weight. The hydrogenation temperatures are generally within a range from 60 to 150°C, preferably from 80 to 140°C, especially 100 to 130°C. The hydrogenation pressure is generally within a range from 5 to 400 bar, preferably 60 to 325 bar, more preferably 100 to 280 bar and especially preferably 170 to 240 bar.

The reaction of EDDN with hydrogen in the presence of catalysts can be performed continuous- ly, semicontinuously or batchwise in customary reaction vessels suitable for catalysis, in a fixed bed, fluidized bed or suspension mode. Suitable reaction vessels for the performance of the hydrogenation are those in which contacting of the EDDN and of the catalyst with hydrogen are possible under pressure. The hydrogenation in suspension mode can be performed in a stirred reactor, jet loop reactor, jet nozzle reactor, bubble column reactor, or in a cascade of such identical or different reactors.

The hydrogenation over a fixed bed catalyst preferably takes place in one or more tubular reactors, or else shell and tube reactors.

The hydrogenation of the nitrile groups takes place with release of heat, which generally has to be removed. The heat can be removed by installed heat transfer surfaces, cooling jackets or external heat transferers in a circuit around the reactor. The hydrogenation reactor or a hydrogenation reactor cascade can be run in straight pass. Alternatively, a circulation mode is also possible, in which a portion of the reactor effluent is recycled to the reactor inlet, preferably without preceding workup of the circulation stream.

More particularly, the circulation stream can be cooled in a simple and inexpensive manner by means of an external heat transferer, and the heat of reaction can thus be removed. The reactor can also be operated adiabatically. In the case of adiabatic operation of the reactor, the temperature rise in the reaction mixture can be limited by cooling the feeds or by supplying "cold" organic solvent.

Since the reactor itself need not be cooled in that case, a simple and inexpensive design is possible. One alternative is that of a cooled shell and tube reactor (only in the case of a fixed bed). A combination of the two modes is also conceivable. In this case, a fixed bed reactor is preferably connected downstream of a suspension reactor.

The catalyst may be arranged in a fixed bed (fixed bed mode) or suspended in the reaction mixture (suspension mode).

In a particularly preferred embodiment, the catalyst is suspended in the reaction mixture to be hydrogenated.

The settling rate of the hydrogenation catalyst in the solvent selected should be low in order that the catalyst can be kept in suspension efficiently. The particle size of the catalysts used in suspension mode is therefore preferably between 0.1 and 500 μηη, especially 1 and 100 μηη.

If the hydrogenation of EDDN is performed continuously in suspension mode, is preferably sup- plied continuously to the reactor and a stream comprising the hydrogenation product TETA is removed continuously from the reactor.

In the case of the batchwise suspension mode, EDDN, optionally together with organic solvent, is introduced as an initial charge.

The amount of catalyst in the case of the batchwise suspension mode is preferably 1 to 60% by weight, more preferably 5 to 40% by weight and very preferably 20 to 30% by weight, based on the overall reaction mixture. The residence time in the reactor in the case of the batchwise suspension mode is preferably 0.1 to 6 hours, more preferably 0.5 to 2 hours.

The residence time in the reactor in the case of the continuous suspension mode is preferably 0.1 to 6 hours, more preferably 0.5 to 2 hours.

The space velocity on the catalyst in the case of the continuous suspension mode is preferably 0.1 to 10 kg, preferably 0.5 to 5 kg of EDDN per kg of catalyst and hour.

The reaction effluenteffluent from the hydrogenation generally also comprises further higher- and lower-boiling organic substances as by-products, such as methylamine, AEPIP, PIP, Me- TETA or TEPA, or basic compounds or additives which have been supplied before or during the hydrogenation, for example alkali metal hydroxides, alkoxides, amides, amines and ammonia.

The hydrogenation effluenteffluent preferably further comprises organic solvent which was pre- sent during the hydrogenation, preferably the organic solvent which was also present in the course of treatment with adsorbent, especially THF.

The reaction effluent preferably further comprises distilling agent, especially toluene, which was preferably used in the distillative depletion of water after the EDDN preparation.

The reaction effluent generally also comprises small amounts of water.

In general, the amounts of water present in the effluent from the hydrogenation correspond to the amounts which originate from the conversion of EDA, FA and HCN and the preferred workup.

The catalyst can be removed by methods known to those skilled in the art. In general, after removal of the catalyst, the hydrogen present during the hydrogenation is removed.

Purification (Step C):

After the hydrogenation (Step B), the effluent from the hydrogenation is further purified (Step C), wherein the purification comprises

1 ) removal of hydrogen;

2) removal of organic solvent and/or water,

3) passing the reaction mixture obtained in step 2) to an evaporator operated at a pressure in the range of 5 to 500 mbar at a temperature of 10 to 250°C,

4) removal of the gaseous phase from the evaporator, and

5) condensation of the gaseous phase 1 ) Removal of hydrogen

Hydrogen is preferably removed by lowering the pressure at which the hydrogenation was performed to a value at which hydrogen is gaseous, but the other components in the reaction effluent are in liquid phase. The reaction effluent is preferably decompressed from a hydrogenation pressure of preferably 60 to 325 bar, more preferably 100 to 280 bar and most preferably 170 to 240 bar down to a pressure of 5 to 50 bar in a vessel. At the top of the vessel, hydrogen, with or without ammonia, and a small amount of evaporated low boilers such as THF, are obtained. Hydrogen and any ammonia can be recycled into the hydrogenation. For example THF can be condensed out and recovered. Alternatively, THF can be recovered by offgas scrubbing with a higher-boiling solvent, for example toluene or TETA.

2) Removal of the organic solvents and/or water

Organic solvents and/or water present in the reaction effluent are preferably likewise removed by distillation. The lighter boiling solvents and/or water are generally removed at the top of one or more distillation column whereas the reaction mixture is generally obtained as the bottom product.

The number of distillation columns generally depends on the number of organic solvents used during the reaction and their respective physical parameters.

The detailed operating conditions of the respective distillation columns may be routinely calculated and adapted by a person skilled in the arts taking into account the separating efficiency of the respective distillation column using the known vapour pressures and vapour pressure equilibria. The content of organic solvent in the reaction mixture (hydrogenation effluent) is in general reduced to a level of 3 percent by weight or less, prerably 1 percent by weight or less and more preferably 0.5 percent by weight or less.

The content of water in the reaction mixture (hydrogenation effluent) also is in general reduced to a level of 3 percent by weight, prerably 1 percent by weight or less and more preferably 0.5 percent by weight or less.

3) Evaporator After the removal of organic solvent and water the reaction mixure passing the reaction mixture is passed to an evaporator (phase-change heat exchanger).

An evaporator is a means for converting a liquid phase partially into a vapor phase. Examples for suitable evaporators can be found in Ullmann's Encyclopedia of Industrial Chemistry (Chapter 2.2.2 in Article "Heat Exchange" by Ramesh K. Shah and Alfred C. Mueller, Published Online : 15 JUN 2000, DOI: 10.1002/14356007.b03_02)

Preferably, the evaporator may be designed as a thin film evaporator, a falling film evaporator, a wiped film evaporator, a scraped film evaporator a boiler or a boiler kettle, a kettle-type reboil- er, a shell evaporator, a tube evaporator, natural circulation evaporator, a forced circulation evaporator or a plate evaporator.

The evaporator is operated at a pressure in the range of 5 to 500 mbar, preferably 5 to 100 mbar and more preferably 5 to 20 mbar.

The temperature at which the evaporator is operated is in the range of 10 to 250°C, preferably 125 bis 200 °C and more preferably 125 bis 180 °C.

If the evaporator is operated under these pressure and temperature conditions, part of the reaction mixture passes into the evaporator will be converted to the vapor phase.

4) Removal of the gaseous (vapor) phase

The gaseous or vapor phase is drawn off (removed) from the evaporator 5) Condensing of the gaseous (vapor ) phase

The gaseous or vapor phase is fed to a condenser in which the gaseous or vapor phase is condensed.

Examples for condensers can also be found in in Ullmann's Encyclopedia of Industrial Chemis- try (Chapter 9 in Article "Heat Exchange" by Ramesh K. Shah and Alfred C. Mueller, Published Online : 15 JUN 2000, DOI: 10. Condensing the vapor or gaseous phase in generally yields the amine compositions I - VI of the present invention.

Unexpectedly it has been found that cured epoxy resins in which the amine compositions I through VI of the present invention are used as amine curing agents have an improved cold processing and cold curing properties compared to "commercially available TETA" or "TETA 85".

Cold curing and cold processing conditions denote conditions in which the ambient temperature is in the range below 20°C. Such conditions are encountered in construction applications, e.g. flooring applications or adhesive applications, outside the summer season and/or in some geographical locations. Under these conditions it is particularly important that the amine composition as a low initial viscosity and remains processable for a longer time. On the other hand it is important that the curing process, in particular the gel times are reduced compared to conven- tional products. Under cold conditions, it is important that the curing process is as short as possible because normally construction work cannot resume until the epoxy resin in which the amine composition is used is fully cured.

Unexpectedly it was also found that in addition to the excellent cold curing and cold processing properties that the amine compositions of the present invention are an excellent drop-in product for "commercially available TETA" achieving similar or sometimes even better properties, e.g. mechanical properties, than "commercially available TETA" in many applications, especially in epoxy applications.

Therefore the present invention also refers to the use of amine compositions I - VI as amine curing agents.

The present inventions is therefore also directed to an amine curing agent composition comprising

10 to 100% by weight of an amine composition selected from the group consisting of amine composition I, II, III, IV, V and VI; and

0 to 90% of other amine curing agents (which hereinafter will be referred to as "amine curing agent composition I").

Preferably the amine curing agent composition comprising

50 to 100% by weight of an amine composition selected from the group consisting of amine composition I, II, III, IV, V and VI; and

0 to 50% of other amine curing agents (which hereinafter will be referred to as "amine curing agent composition II"). More preferably the amine curing agent composition comprising

75 to 100% by weight of an amine composition selected from the group consisting of amine composition I, II, III, IV, V and VI; and 0 to 25% of other amine curing agents (which hereinafter will be referred to as "amine curing agent composition III").

Even more preferably the amine curing agent composition comprising

90 to 100% by weight of an amine composition selected from the group consisting of amine composition I, II, III, IV, V and VI; and

0 to 10% of other amine curing agents (which hereinafter will be referred to as "amine curing agent composition IV"). The amount of water and other organic side products in the amine curing agent compositions is preferably less than 5% by weight, more preferably less than 2% by weight and most preferably less than 1 % by weight.

In a preferred embodiment the amine curing agent compositions (curing agent composition V) consists of an amine composition selected from the group consisting of amine composition I, II, III, IV, V and VI.

Other amine curing agents which may be present in the amine curing agent compositions l-IV of the present invention are amine compounds having at least one or more, preferably two or more reactive amine hydrogen atoms in the molecule capable of reaction with an epoxy functionality.

Preferably, the other amine curing agents which may be used in combination with the amine compositions l-VI in the amine curing agent compositions l-IV are:

heterocyclic amines, such as piperazine;

cycloaliphatic amines, such as isophoron diamine, 1 ,2- (1 ,3; 1 ,4) -diaminocyclohexane, cyclo- hexylaminopropylamine (CHAPA), tricyclododecan diamine (TCD);

aromatic amines, such as the isomeric phenylenediamines, such as o-phenylenediamine, m- phenylenediamine, p-phenylenediamine, the isomeric tolylenediamines, such as 2,4- diaminotoluene and/or 2,6 diaminotoluene, the isomeric diaminonaphthalenes, such as 1 ,5- diaminonaphthalene, bis(4-aminophenyl)methane (MDA), the isomeric xylenediamines, such meta-xylenediamine (MXDA), bis(4-amino-3-methylphenyl)methane and bis(4-amino-3,5- dimethylphenyl)-methane;

substituted aliphatic amines such as ethylene diamine, propylene diamine, hexamethylenedia- mine, 2,2,4 (2,4,4) -trimethylhexamethylene diamine, 2-methylpentamethylene diamine;

ether amines such as 1 ,7-diamino-4-oxaheptane, 1 ,10-diamino-4,7-dioxydecane, 1 ,14-diamino- 4,7,10-trioxatetradecane, 1 ,20-diamino-4,17-dioxyeicosan and in particular 1 ,12-diamino-4,9- dioxadodecane;

ether diamines based on propoxylated diols, triols and Polyols;

polyalkylene polyamines, such as dipropylene triamine, tripropylene tetramine;

as well as high molecular amines or addition or condensation products containing free amine hydrogen, in particular Mannich bases. Most preferably, the other amine curing agent is isophoron diamine, 1 ,2- (1 ,3; 1 ,4) - diaminocyclohexane, cyclohexylaminopropylamine (CHAPA), tricyclododecan diamine (TCD), xylylene diamine, ethylene diamine, propylene diamine, hexamethylenediamine, 2,2,4 (2,4,4) - trimethylhexamethylene diamine, 2-methylpentamethylene diamine, 1 ,7-diamino-4-oxaheptane, 1 ,10-diamino-4,7-dioxydecane, 1 , 14-diamino-4,7, 10-trioxatetradecane, 1 ,20-diamino-4,17- dioxyeicosan and 1 , 12-diamino-4,9-dioxadodecane

Most preferably, the other amine or the mixture of other amines does not contain DAEPI P, PEEDA and TAEA. The amine curing agent compositions l-V according to the present invention can be mixed with epoxy resins to yield curable compositions.

Epoxy resins are an important class of polymeric materials, characterized by the presence of more than one epoxy-ring.

A substantial enumeration of epoxy resins suitable for curable compositions can be found in the handbook "Epoxidverbindungen und Epoxidharze" of A. M. Paquin, Springer Verlag Berlin, 1958, Chapter IV; in Lee & Neville, "Handbook of Epoxy Resins", 1967, Chapter 2; in Ullmann's Encyclopedia of Industrial Chemistry, Wiley VCH Verlag GmbH, Electronic Edition 2005, Chap- ter "Epoxy Resins" and in "Epoxy Resins, Chemistry and Technology", of C. A. May, Marcel Dekker Inc, New York, 1988 , Chapter 2.

Commercially important epoxy resins are in particularly prepared by the coupling reaction of compounds containing at least two active hydrogen atoms with epichlorohydrin followed by de- hydrohalogenation. Compounds which contain at least two active hydrogen atoms include poly- phenolic compounds, mono and diamines, amino phenols, heterocyclic imides and amides, aliphatic diols and polyols, and dimeric fatty acids.

Epoxy resins derived from epichlorohydrin are termed glycidyl-based resins. Alternatively, epoxy resins based on epoxidized aliphatic or cycloaliphatic dienes are produced by direct epoxidation of olefins by peracids.

Epoxy resins also comprise reaction products of epichlorohydrin and bisphenol A. These products are generally termed DGEBA (Diglycidyl ether of bisphenol A) (see formula (X)).

DGEBA where the degree of polymerization, n, is very low (n =0.2) is typically referred to as liquid epoxy resin (LER) whereas high molecular weight (Mw) epoxy resins based on DGEBA characterized by a repeat unit containing a secondary hydroxyl group with degrees of polymerization, i.e., n values ranging from 2 to about 35 are generally denoted as solid epoxy resins (SER).

(X)

Epoxy resins also comprise so called epoxy novolac resins. The multifunctionality of these resins provides higher cross-linking density, leading to improved thermal and chemical resistance properties over bisphenol A epoxides. Epoxy novolacs are multifunctional epoxides based on phenolic formaldehyde novolacs. Both epoxy phenol novolac resins (EPN) and epoxy cresol novolac resins (ECN) have attained commercial importance. The former is made by epoxidation of the phenol-formaldehyde condensates (novolacs) obtained from acid-catalyzed condensation of phenol and formaldehyde.

The epoxy compounds which can be used for the curable compositions and the cured epoxy resins derived therefrom are those resins described above or mentioned in the cited literature, in particular commercial products having more than one epoxy group per molecule on average, which are derived from monovalent and/or multivalent and/or multinuclear phenols, in particular bisphenols as well as novolacs, such as bisphenol-A and bisphenol-F-diglycidylether.

The epoxy resins preferably comprise epoxy resins selected from the group of bisphenol A bis- glycidyl ether (DGEBA), bisphenol F bisglycidyl ether, bisphenol S bisglycidyl ether (DGEBS), tetraglycidylmethylenedianilines (TGMDA), epoxy novolacs (the reaction products of epichloro- hydrin and phenolic resins (novolak)), and cycloaliphatic epoxy resins such as 3,4- epoxycyclohexylmethyl 3,4-epoxycyclohexanecarboxylate and diglycidyl hexahydrophthalate.

Compositions of two or more epoxy resins may be used as well. The curable compositions may comprise further additives such as additives common in epoxy resin technique. To be noted are, for example, gravels, sands, silicates, graphite, silica, talc, mica etc. in the particle size distributions as common used in this area. Suitable additives comprise e.g. antioxidants, UV absorbers/light stabilizers, metal deactivators, antistatic agents, reinforcing agents, fillers, biocides, lubricants, emulsifiers, colorants, pigments, rheology additives, mold release agents, catalysts or accelerators, flow-control agents, optical brighteners, flame retardants, antidripping agents and blowing agents.

An overview of additives, adjuvants and curing agents, which may be used in curable compositions is given in Ullmann's Encyclopedia of Industrial Chemistry, Wiley VCH Verlag GmbH, Elec- tronic Edition 2005. Chapter "Epoxy Resins", in "Epoxy Resins, Chemistry and Technology", of C. A. May, Marcel Dekker Inc, New York, 1988 , Chapter 3 and in "Epoxy Resins, Curing Agents, Compounds and Modifiers" by E. W. Flick, Noyes Publications, Park Ridge, 1987.

The weight ratio of epoxy resins and the amine curing agent compositions as well as additives and adjuvants may be varied to achieve and refine the desired application properties of the final cured epoxy and can be routinely determined by a person skilled in the art, e.g. the amine curing agent may be contained in the composition in such an amount that a molar ratio of epoxy groups of the epoxy resin to active hydrogen atoms of the amine curing agent ranges from 0.5 to 1 .5, preferably 0.7 to 1 .3 and more preferably from 0.8 to 1 .2.

Therefore, the present invention also relates to curable composition comprising an amine curing agent composition selected from the group of amine curing agent compositions I, II, III, IV and V and

one or more epoxy resins, wherein the molar ratio of epoxy groups of the epoxy resin to active hydrogen atoms of the amine curing agent ranges from 0.5 to 1.5, preferably 0.7 to 1.3 and more preferably from 0.8 to 1 .2.

The present invention also relates to a

method for the production of a curable composition according to the present invention by mixing an amine curing agent composition selected from the group of amine curing agent compositions I, II, III, IV and V with at least one epoxy resin.

The process of mixing amine curing agents with one or more epoxy resins is well known to a person skilled in the arts. Generally mixing is effected by a mixing apparatus. The mixing appa- ratus can be of any type that can produce a highly homogeneous mixture of the epoxy resin and amine curing agent composition (and any optional components that are also mixed in at this time). Mechanical mixers and stirrers of various types may be used. Two preferred types of mixers are static mixers and impingement mixers. Mixing can be conducted batch-wise, semi- continuously or in a continuous fashion.

The epoxy resin and amine curing agent are generally separately heated to above room temperature prior to mixing them together, so that a curable composition is formed immediately upon mixing them. The epoxy resin and amine curing agent may each be heated to a temperature of 25°C, preferably 50°C, more preferably 80°C, or higher prior to mixing.

Other additives, such as the ones mentioned above, may be mixed with the amine curing agent or the epoxy resins prior to mixing the amine curing agent with the epoxy resin. It is also possible to mix other additives with the curable composition at the same time the amine curing agent and the epoxy resin are mixed, or afterwards.

After mixing, the thus obtained curable composition is typically transferred to a suitable mold (structural applications) or applied to a surface (coating applications), e.g. by spraying the curable composition on a surface, to obtain a cured epoxy resin. Therefore, the present invention also relates to the use of amine curing agent compositions I, II, III, IV or V for epoxy resins and the production of epoxy resins.

Accordingly, the present invention further relates to a method of producing a cured epoxy resin by transferring the curable compositions according to the present invention to a mold or applying said curable compositions to a surface.

Appropriate processing technologies are known to a person skilled in the art and can be found e.g. in B. Ellis, "Chemistry and Technology of Epoxy Resins", Kluwer Academic Publishers (Februar 1993).

Generally, the cured epoxy resins are obtained by allowing the curable composition to set after mixing and transfer to a mold or after application to a surface. During setting, the amine curing agents undergo a reaction with the epoxy resins present in the curable composition.

In a preferable embodiment of the invention, mixing and transfer of the curable composition is performed in one step, e.g. by reaction injection molding. The epoxy resin and amine curing agent composition (and optionally other components which are mixed in at this time) are pumped under pressure into a mixing head where they are rapidly

mixed together. Operating pressures in high pressure machines may typically range from 7 to 14 MPa, although operation at lower pressures is also possible. The resulting curable composition is then preferably passed through a static mixing device to provide further additional mixing, and then transferred into the mold cavity.

In other embodiments, the curable composition is prepared by mixing as described before, and then applied to a surface, in particular by spraying the curable composition into a mold.

The mold is typically a metal mold, but it may be ceramic or a polymer composite, provided that the mold is capable of withstanding the pressure and temperature conditions of the molding process. The mold usually contains one or more inlets through which the reaction mixture is introduced. The mold may contain vents to allow gases to escape as the reaction mixture is injected. The mold is typically held in a press or other apparatus which allows it to be opened and closed, and which can apply pressure on the mold to keep it closed during the filling and curing operations. The mold or press is provided with means by which heat can be provided.

In a preferred embodiment of the present invention, the curable composition is applied to a rein- forcing agent and cured in the presence of the reinforcing agent to form reinforced composites.

Reinforcement agents may be, for example, blended with the epoxy resin or the amine curing agent (or both), prior to mixing to obtain the curable compositions. Alternatively, the reinforcing agents may be added io the curable compostions at the same time as the epoxy resin and the amine curing agent are mixed, or afterward but prior to introducing the curable composition into the mold or applying the curable composition to a surface, e.g. by spraying the curable composition into a mold. Suitable reinforcement agents are fibrous materials or non-fibrous materials.

Fibrous materials include glass, quartz, polyamide resins, boron, carbon and gel-spun polyethylene fibers.

Non-fibrous reinforcing agents include glass flakes, aramid particles, carbon black, carbon nanotubes, various clays such as montmorillonite, and other mineral fillers such as wollastonite, talc, mica, titanium dioxide, barium sulfate, calcium carbonate, calcium silicate, flint powder, carborundum, molybdenum silicate, sand, and the like.

Non fibrous reinforcing agents may also include conductive materials, such as aluminum and copper, and carbon black, carbon nanotubes, carbon fibers, graphite and the like.

The reinforcement agent can take any of several forms, such as a fiber preform, continuous fiber rovings, cut fibers or chopped fibers.

Preferably the reinforcement agent is in form of a fiber preform, i.e., a web or mat of fibers. The fiber preform can be made up of continuous filament mats, in which the continuous filaments are woven, entangled or adhered together to form a preform that approximates the size and shape of the finished composite article (or portion thereof that requires reinforcement). Alternatively, shorter fibers can be formed into a preform through entanglement or adhesive methods. Mats of continuous or shorter fibers can be stacked and pressed together to form preforms of various thicknesses, if required.

Fiber preforms are typically placed into the mold prior to introducing the curable composition. The curable composition can be introduced into a closed mold that contains the preform, by injecting the curable composition into the mold, where the curable composition penetrates between the fibers in the preform and then cures to form a cured epoxy resin. Reaction injection molding and/or resin transfer molding equipment is suitable in such cases. Alternatively, the preform can be deposited into an open mold, and the curable composition can be sprayed onto the preform and into the mold. After the mold is filled in this manner, the mold is closed and the curable composition is cured. In either approach, the mold and the preform are preferably heated prior to contacting them with the curable composition.

Short fibers can be used instead or in addition to a fiber preform. Short fibers (up to about 20 cm in length, preferably up to 5 cm in length, more preferably up to about 2 cm in length can be blended with the curable composition and injected into the mold together with the curable composition. Such short fibers may be, for example, blended with the epoxy resin or the amine curing agent (or both), prior to mixing to obtain the curable compositions. Alternatively, the short fibers may be added into the curable compostions at the same time as the epoxy resin and the amine curing agent are

mixed, or afterward but prior to introducing the curable composition into the mold.

Short fibers can be sprayed into a mold. In such cases, the curable composition can also be sprayed into the mold, at the same time or after the short fibers are sprayed in. When the fibers and the curable composition are sprayed simultaneously, they can be mixed together prior to spraying. Alternatively, the fibers and the curable composition can be sprayed into the mold separately but simultaneously.

Various processes for the production of reinforced composites, which are well-known by a person skilled in the arts, may be used such as RTM, VARTM, RFI and SCRIMP. In these pro- cesses, a reinforcement agent in form of woven or matted fiber preform is inserted into a mold cavity. The mold is closed, and the resin is injected into the mold. The resin hardens in the mold to form a composite, and is then demolded. Reinforced composites may also be produced by pultrusion processes.

Pultrusion processes use continuous fibers that are oriented parallel to each other, in the direction of extrusion. Pultrusion processes are operated in a manner analogous to molding processes, the main difference being that the hot reaction mixture is delivered into a resin bath rather than into a mold. The resin bath is a reservoir filled with the curable composition, through which the continuous fibers are pulled. Once the fibers are wetted with the hot reaction mixture, they are pulled through one or more dies, in which the fibers are consolidated and formed into the desired cross-sectional shape.

Unexpectedly it has been found that cured epoxy resins in which the amine compositions I through VI of the present invention are used as amine curing agents have a improved cold processing and cold curing properties compared to "commercially available TETA" or "TETA 85".

Therefore the present invention also relates to the use of amine compositions I, II, III, IV, V or VI as a curing agent for epoxy at temperatures of below 20°C.

The present invention further relates to the use of amine curing agent compositions I, II, III, IV or V as a curing agent for epoxy at temperatures of below 20°C.

Cold curing and cold processing conditions denote conditions in which the ambient temperature is in the range below 20°C, preferably below 15°C and more preferably below 10°C. Such con- ditions are encountered in construction applications, e.g. flooring applications or adhesive applications, outside the summer season and/or in some geographical locations. Under these conditions it is particularly important that the amine composition as a low initial viscosity and remains processable for a longer time. On the other hand it is important that the curing process, in particular the gel times are reduced compared to conventional products. Under cold conditions, it is important that the curing process is as short as possible because normally construction work cannot resume until the epoxy resin in which the amine composition is used is fully cured.

Unexpectedly it was also found that in addition to the excellent cold curing and cold processing properties that the amine compositions of the present invention are an excellent drop-in product for "commercially available TETA" achieving similar or sometimes even better properties, e.g mechanical properties, than "commercially available TETA" in many applications, especially epoxy applications.

The amine compositions I, II, III, IV, V or VI of the present invention may also advantageously be used for the production of reactive polyamide resins.

Therefore the present invention also refers to reactive polyamide resin, obtainable from the reaction of an amine composition selected from the group consisting of amine composition I, amine composition II, amine composition III, amine composition IV, amine composition V and amine composition VI with dimer fatty acids.

Reactive polyamides are lower-molecular-weight (1 ,000-2,000 g/mol) products from the con- densation of dimer fatty acid and one of the higher ethyleneamines (diethylenetriamine, triethyl- enetetramine and others). Reactive polyamides are mainly used as curing agents in two- component epoxy systems for industrial and marine maintenance coatings, thermosetting adhesive systems, electronics encapsulation and flooring grouts and trowel coatings. Their amine groups provide reactive sites for cross-linking interactions with epoxy resin molecules.

Reactive polyamides are usually produced in a batch condensation process. The reactants (dimer fatty acid and the amine compositions l-VI according the present invention are generally heated to 150-250°C. By-product water is usually removed by vacuum distillation. The resulting polyamide is then

removed and converted to forms suitable for shipping.

Dimer fatty acids are most frequently obtained by the polymerization of monocarboxylic acids containing ethyleneic unsaturation. The monocarboxylic unsaturated acids generally contain from about 16 to 26 carbon atoms and include, for example, oleic acid, linoleic acid, eleostearic acid and similar singly or doubly unsaturated acids. To obtain the preferred dimer acids 2 mols of the unsaturated monocarboxylic acid are reacted, i.e., dimerized. Oleic acid, linoleic acid and linolenic acid are generally used as unsaturated fatty acids. The dimer acids, obtained in this manner, can subsequently be

hydrogenated.

Reactive polyamide resins obtained from the reaction of dimer fatty acids and amine compositions l-VI according to the invention have following advantages: The mixture of the reactive pol- yamide resin (dissolved in xylene) and the epoxy resin has a lower initial viscosity at room temperature leading to an improved workability. Furthermore, the through cure time is significantly decreased while gel and tack free time remain identical. Thus, the coating can be applied in the same time period and will be hardened in a shorter time period. This offers efficiency advantages for the coating industry.

Reactive polyaminde resins according to the present invention also be used for expoxy resins and adhesives.

Therefore, the present invention also relates to the use of reactive polyamide resins obtained from the reaction of dimer fatty acids and amine compositions l-VI for coatings, epoxy resins and adhesives.

Preferred embodiments of the production process for the amine compositions of the present invention are detailed with reference to the appended drawings.

Figure 1 shows a preferred embodiment of step A of the preparation of the amine compositions of the present invention from EDA (1 ) and FACH (5).

The preferred process parameters can be inferred from the above description. First, EDA (1 ) is mixed with water (2) in a mixer (I) to give an aqueous EDA stream (3). The mixing of EDA with water releases heat of hydration, which is led off in a heat exchanger (II). An FACH-containing stream (5) is mixed with toluene (6). The toluene-containing FACH stream is mixed with the aqueous EDA solution (3) at a mixing point and introduced into an adiabatic tubular reactor (III). At the outlet of the tubular reactor (III), the exiting reaction mixture (7) is decompressed in a decompression valve and introduced into a flash evaporator (IV). The gaseous phase (8) comprising water, toluene and low-boiling compounds which forms is condensed in a condenser (V). Uncondensed constituents (9 or LS), such as ammonia, HCN, methanol or CO2, are discharged from the process. The condensate (10) condensed in the condenser (V) is introduced into a phase separation vessel (VI) and separated into an aqueous phase (14) and a toluene- containing phase (1 1 ).

The aqueous phase (14) from the phase separation vessel (VI) can be recycled into the process, for example to produce an aqueous EDA solution in the mixer (I), or introduced into a biological waste water treatment (not shown). The aqueous phase (14) can also be introduced into a column K2 (VIII) in which water as a bottom product (16) is removed from low boilers (15). The low boilers (15), for example solvents having a lower boiling point than water or low-boiling water azeotropes or HCN, can be conducted directly to the condenser (V), in which the gaseous phase from the flash evaporation is also condensed. Uncondensable constituents are discharged from the process as stream (9).

The toluene-containing phase (1 1 ) can be recycled into the process as an organic solvent and mixed with the FACH-containing stream from the FACH preparation. Losses of toluene can optionally be replaced by a toluene addition. However, the toluene-containing phase (1 1 ) can preferably be introduced into a column K1 (VII) together with the liquid phase (12) from the flash vessel (IV).

The phase (12) remaining in liquid form in the flash evaporation is conducted out of the flash vessel (IV) and likewise to the top of column K1 (VII), optionally together with the toluene- containing phase (1 1 ), in order to deplete water.

In column K1 (VII), a gaseous, essentially aqueous top product is drawn off and is conducted directly to the condenser (V) and passed into the phase separation vessel (VI). In the phase separation vessel, as described above, aqueous phase (15) which forms can be discharged, passed into the mixer (I) or supplied to column K2 (VIII).

At the bottom (17) of column K1 , a mixture comprising EDDN and toluene is drawn off.

The mixture (17) comprising toluene and EDDN is diluted with THF (18) and treated in an adsorber (IX) with adsorbent, preferably with a solid acidic adsorbent. A mixture comprising EDDN and toluene and THF (20) obtained from the adsorber comprises only small amounts of water. The EDDN comprising mixture can be passed into a hydrogenation. Figure 2 shows shows another preferred embodiment of step A of the preparation of the amine compositions of the present invention from FA (1 ), EDA (2) and HCN (5), wherein FA (1 ) and EDA (2) are first converted to EDFA (4), and the latter then reacts with HCN (5) to give EDDN. The preferred process parameters can be inferred from the above description. First, FA (1 ) with EDA (2) is mixed into the loop of a loop reactor (I). In the loop reactor, FA (1 ) is reacted with EDA (2) to give EDFA. A portion of the reactor content of the loop reactor is discharged (3) and passed into a tubular reactor (II). The effluent (4) from the tubular reactor (II) is mixed with HCN (5) and toluene (6) at a mixing point at the inlet of a tubular reactor (III) and passed through the tubular reactor (III).

At the outlet of the tubular reactor (III), the exiting reaction mixture (7) is decompressed in a decompression valve and introduced into a flash evaporator (IV). The gaseous phase (8) comprising predominantly water and toluene which forms is condensed in a condenser (V). Uncon- densed constituents (9), such as ammonia, HCN, methanol or CO2, are discharged from the process. The condensate (10) condensed in the condenser (V) is introduced into a phase separation vessel (VI) and separated into an aqueous phase (14) and a toluene-containing phase (1 1 ).

The aqueous phase (14) from the phase separation vessel (VI) can be recycled into the pro- cess, for example to produce an aqueous EDA solution in the mixer (I), or introduced into a biological waste water treatment (not shown). The aqueous phase (14) can also be introduced into a column K2 (VIII) in which water as a bottom product (16) is removed from low boilers (15). The low boilers (15), for example solvents having a lower boiling point than water or low-boiling water azeotropes or HCN, can be conducted directly to the condenser (V). Uncondensable con- stituents are discharged from the process as stream (9).

The toluene-containing phase (1 1 ) can be recycled into the process as an organic solvent and mixed with the EDFA-containing stream from the EDFA preparation. Losses of toluene can optionally be replaced by a toluene addition. However, the toluene-containing phase (1 1 ) can be introduced into a column K1 (VII) together with the liquid phase (12) from the flash vessel (IV).

The phase (12) remaining in liquid form in the flash evaporation is conducted out of the flash vessel (IV) and likewise to the top of column K1 (VII), optionally together with the toluene- containing phase (1 1 ), in order to deplete water.

In column K1 (VII), a gaseous, essentially aqueous top product is conducted directly to the condenser (V) and passed into the phase separation vessel (VI), where the aqueous phase (15), as described above, can be discharged, passed into the mixer (I) or supplied to column K2 (VIII). At the bottom (17) of column K1 , a mixture comprising EDDN and toluene is obtained.

The mixture comprising toluene and EDDN (17) is diluted with THF (18) and treated in an adsorber (IX) with adsorbent, preferably with a solid acidic adsorbent. A mixture comprising EDDN and toluene and THF obtained from the adsorber comprises only small amounts of water. The EDDN comprising mixture can be passed into a hydrogenation.

Figure 3 shows a preferred embodiment in whicht the reaction mixture obtained from Step A is purified in an adsorber before hydrogenation (Step B).

The preferred process parameters can be inferred from the above description.

In figure 3, "crude" EDDN refers to an amine compositions of the present invention, which can be prepared by converting FA, HCN and EDA according to one of the options a) to d) cited in the description, and which has been worked up, preferably by i) removal of low boilers, for ex- ample by stripping, flash evaporation or distillation, and ii) distillative removal of water, preferably in the presence of an organic solvent which has a boiling point between water and EDDN under the conditions of the water removal or which forms a low-boiling azeotrope with water. Such a "crude" EDDN is mixed with THF (18) and treated in an adsorber with adsorbent, preferably solid acidic adsorbent. The stream (1 ) which leaves the adsorber is passed into a hydro- genation reactor (I) in which the EDDN "purified" by adsorption is hydrogenated in the presence of hydrogen (2).

Figure 4 shows a preferred embodiment of the hydrogenation (Step B) with subsequent workup (Step C).

The preferred process parameters can be inferred from the above description.

The amine compositions according to the invention can be prepared by conversion of FA, HCN and EDA according to one of the options a) to d) specified in the description. The workup is effected preferably by i) removal of low boilers, for example by stripping, flash evaporation or dis- tillation, and ii) dehydration, preferably in the presence of an organic solvent which has a boiling point between water and EDDN under the conditions of the water removal, or which forms a low-boiling azeotrope with water.

The dehydrated reaction mixture is preferably mixed with THF and with adsorbent, preferably solid acidic adsorbent. The mixture (1 ) is hydrogenated in a hydrogenation reactor (I) in the presence of supplied hydrogen (2). The reaction effluent from the hydrogenation (3) is decompressed into a flash vessel (II). The gaseous constituents (4), such as hydrogen, portions of the THF, HCN, methanol or methylamine, can be discharged from the process or recovered partly or fully. The phase (5) remaining in liquid form after the decompression is passed into a column K1 having a stripping section and a rectifying section. At the top of the column, a low-boiling THF/water azeotrope (6) is drawn off and condensed. The condensed stream is mixed with toluene (7) in a phase separation vessel. In the phase separation vessel, an aqueous phase (8) and a

THF/toluene phase (9) form, the latter being recycled into column K1.

From the bottom of column K1 , a stream (10) is drawn off which comprises TETA, DETA, THF, toluene and organic compounds such as PIP, AEPIP and TEPA. This stream (10) is passed into a column K2, in which THF is drawn off as the top product (1 1 ). This THF (1 1 ) can be recycled directly into the process, preferably into the adsorption step. Before being introduced into the adsorber stage, the THF (1 1 ) can be contacted with a molecular sieve for further depletion of water.

At the bottom of column K2, a stream (12) is drawn off which comprises TETA, DETA, toluene and organic compounds such as PIP, AEPIP, Me-TETA and TEPA.

This stream (12) is introduced into a column K3, in which toluene is drawn off at the top (13). For dewatering of THF, the toluene (13) drawn off can be passed via line (7) into a phase sepa- ration vessel in which it is combined with the condensate (6) from column K1. The toluene (13) drawn off can also be discharged from the process via line (14) or preferably used as a solvent in the process.

The bottom product of column K3 (16) comprises TETA, DETA, toluene and organic com- pounds such as PIP, AEPIP and TEPA.

This mixture is then fed into an evaporator (E) operated at a pressure in the range of 5 to 500 mbar at a temperature of 10 to 250°C. Part of the bottom product of column K3 is evaporated and removed as gaseous phase (17) from the evaporator.

The amine composition according to the present invention (termed "TETA" in Fig.4) is obtained by condensing the gaseous phase (17) at a condenser.

Abbreviations:

Ethylenediamine (EDA)

Ethylenediamine-formaldehyde bisadduct (EDFA)

Ethylenediamine-formaldehyde monoadduct (EDMFA)

Ethylenediaminediacetonitrile (EDDN)

Ethylenediaminemonoacetonitrile (EDMN)

Diethylenetriamine (DETA)

Triethylenetetramine (TETA)

Tetraethylenepentamine (TEPA)

Formaldehyde (FA)

Formaldehyde cyanohydrin (FACH)

Piperazine (PIP)

Aminoethylpiperazine (AEPIP)

Mixture of formaldehyde and hydrogen cyanide (GFB)

2- and 3-methyltetrahydrofuran (MeTHF)

Aminoacetonitrile (AAN)

The present invention is exemplified by following examples:

Example 1 : Synthesis of an amine composition according to the present invention. 700g/h FA and 209g/h EDA were mixed together in a 138 ml. loop reactor operated at 45°C. The output from the loop reactor was passed through a tubular reactor (volume =19ml_) and mixed with 1 ,1 kg/h toluene. The resulting mixture was warmed to 40°C, mixed with 205g/h 90% HCN (10% water content ) and passed through a tubular reactor (Volume =18 ml_). Tempera- ture at the tubular reactor outlet was 75°C. The reactor effluent was decompressed to 150 mbar at the top of distillation column VII, resulting in an instantaneously temperature drop of the stream to 43°C. Destination column VII consisted of a stripping section with 960 mm high sheet metal packings from Montz A3-500 type. The vapours at the top of VII were condensed and passed into a separation vessel, where the aqueous phase (630g/h) was separated from the toluene containing phase. The aqueous phase was completely discharged. The toluene- containing organic phase was recycled into the process: 1 ,1 kg/h were recycled to the cooling of the EDDN synthesis reactor III, 3,9 kg/h were recycled to the top of column VII. Bottom temperature of column VII was set at 70°C. 541 g/h crude EDDN were drawn off at the bottom of VII.

Bottom effluent had following composition: 14.0% toluene, 5.9% EDMN, 73.4% EDDN, 4.8% BCMI (biscyanomethylimidazolidine) and 0.95% EdTriN (ethylendiamine-triacetonitril). This crude EDDN was mixed with 2.8kg/h THF. The resulting mixture was passed through adsorption column. The adsorption Column was filled with 7.5 kg silica gel in form of 2 to 3 mm beads Sorbead WS. The mixture comprising EDDN, toluene and THF obtained from the adsorber was fed into the hydrogenation reactor which pressure was maintained at 240 bar via continuous hydrogen supply. The reactor was operated at 120°C. The hydrogenation reactor had a volume of 2.6 L and was filled with 500 g Raney-Cobalt. Diameter of the catalyst particles was between 25 m and 55 μηι. The catalyst had following composition: 91 .7% Co, 3.3% Al, 2.6% Ni, 2.1 % Cr and 0.3% Fe. The catalyst was separated from the reaction effluent using continuous filtration. A sintered metal filter with a pore diameter of 500 nm and a surface of 700 mm² was used as filter element. Pressure loss across the filter was 3 bar. The hydrogenation effluent was flashed down to a pressure of 30 bar into a flash vessel. After flashing, the resulting gas partly consisting of previously dissolved hydrogen was fed into a partial condenser and cooled down to 5°C. The condensate was returned to the vessel. The remaining liquid phase from flashing had following composition (without solvent, online GC): 6.0% DETA, 1 .8% N-Me-DETAs, 6.2% AEPIP, 76.7% TETA, 7.9% N-Me-TETAs. This phase was fed at the top of a column K2. This column contained 1440 mm Montz A3-500 internals and was run under atmospheric pressure. Top temperature was 62°C and bottom temperature 1 17°C. 300g/h toluene from the top of the next distillation column K3 were returned to the bottoms of the column K2. At the top of the column K2, a stream comprising 98.3% THF was drawn off. From the bottoms of the column K2, ca. 1 ,1 kg/h of a stream with following composition was drawn off: 0.3% THF, 65.4 % toluene, 2.1 % DETA, 0.7 % N-Me-DETAs, 2.1 % AEPIP, 26.1 % TETA, 1 .5 % M-Me-TETAs. The rest was unknown side products. The bottom stream of K2 was introduced into a column K3, in which toluene was drawn off at the top. K3 consisted of a rectifying section with 1000 mm Sul- zer DYM internals and a stripping section with 1000 mm Sulzer DYM internals. Column K3 was operated at 80 mbar and top temperature was 42°C.The vapors at the top of column K3 were condensed and the ca. 600 g/h stream comprising 98,8% toluene fed into a storage vessel. A 200g/h stream out of this vessel was returned on the top of distillation column K3. From the bot- toms of the column K3, a stream ( ca.500g/h) with following composition was drawn off: 2.8 % toluene, 5.7% DETA, 1 .9% N-Me-DETAs, 5.5 % AEPIP, 74.8 % TETA, 6.7 % N-Me-TETAs and 0.3 % TEPA. The rest was unknown side products.

The bottom stream of K3 was fed into an evaporator operated at a pressure of 20 mbar and bottom temperature of 165°C. Part of the bottoms product of column K3 was evaporated and removed from the evaporator as gaseous phase. After condensation of the gaseous phase, the amine composition according to the present invention was obtained. The amine composition (ca. 480g/h) consisted of

75 % by weight of TETA,

7 % by weight of N-Me-TETAs,

6% by weight of DETA,

2% by weight of N-Me-DETAs,

5.8 % by weight of AEPIP,

2.9% by weight of toluene,

The rest (1 ,3% by weight) was unknown side products.

Example 2: Use of the amine composition obtained in example 1 in epoxy formulations The epoxy formulations were prepared by mixing of stoichiometric amounts of an amine composition (see Table 1 ) with a liquid epoxy resin based on bisphenol-A-diglycidyl ether (EEW 182).

Table 1 : Different amine compositions used in the epoxy formulations in weigth percent

^*organic side products, solvent and water The rheological measurements for investigating the reactivity profile of the amines with epoxy resins were carried out at a shear stress controlled plate-plate rheometer (MCR 301 , Anton Paar) with a plate diameter of 15mm and a gap distance of 0.25 mm at different temperatures. Analysis 1 a)

Comparison of the time to reach a viscosity of 10000 mPa^*s at a defined temperatur: The measurement was carried out using the before mentioned rheometer at different temperatures (10°C, 23°C, 40°C, 75°C) (rotating). The results for the different amine compositions are shown in Table 2.

Table 2: Isothermal viscosity increase until 10.000 mPa^*s

The epoxy resin prepared from an amine composition according to the present invention (TETA 80) has a lower initial viscosity and a longer processing time at 10°C.

Analysis 1 b)

Comparison of rheological gel times: The measurement was carried out using the before mentioned rheometer at different temperatures (10°C, 23°C, 40°C, 75°C) (rotating-oscillating). The rheological gel time is given by the insection of loss modulus (G") and storage modulus (G'). The results are shown in Table 3.

Table 3: Isothermal rheological gel times

The epoxy resin prepared from an amine composition according to the present invention (TETA 80) shows a shorter hardening time at 10°C, a slightly shorter hardening time at 23°C and comparable hardening times at 40°C and 75°C. Analysis 1 c) Comparison of pot life: 100 g of the reaction resin molding material were mixed in a paper cup, equipped with a temperature sensor and stored at 23°C. The temperature is monitored as a function of time. The pot life is the time to reach the maximum temperature. The time to reach 50°C is also monitored. The results are depicted in Table 4.

Table 4: pot life at 23°C

The epoxy resin prepared from an amine composition according to the present invention (TETA 80) shows a comparable pot life compared to the TETA compositions known in the state of the art.

Example 3: Exothermic profile of reaction resin molding material and glas temperatures of the cured thermosets

The DSC-measurements of the curing reaction of the amines with a liquid epoxy resin based on bisphenol-A-diglycidyl ether (EEW 182) for determination of onset temperature (T₀), exotherm (ΔΕ) as well as glas temperature (T_g) was carried out according to ASTM D 3418.

Analysis 2a) Temperature program fort he DSC-measurements: 0°C -> 5K/min 180°C -> 30min 180°C -» 20K/min 0°C -» 20K/min 220°C.

The exothermic reaction profile parameters are shown in Table 5.

Table 5: exothermic profile and glas temperature

The epoxy resin prepared from an amine composition according to the present invention (TETA 80) shows a higher glas temperature than conventional TETA for the fresh sample, a slightly higher exotherm and a comparable onset.

Example 4: Mechanical testing of the cured thermosets

For investigation of the mechanical properties of the cured thermosets from the amines with a liquid epoxy resin based on bisphenol-A-diglycidyl ether (EEW 182) both components were mixed (speed mixer 1 min at 2000 rpm), degassed (vacuum 1 mbar at 23°C) and sample strips were prepared. Hardening conditions: 2h 80 °C, 3h 125 °C. The mechanical testing were carried out according to ISO 527-2:1993 und ISO 178:2006. The results are shown in Table 6. Table 6: mechanical testing of cured thermosets

The epoxy resin prepared from an amine composition according to the present invention (TETA 80) shows a slightly better or comparable mechanical data compared to amine compositions of prior art.

Example 5: Variation of mixing ratio for maximum Tg

The measurements were carried out as in example 2 using different mixing ratios resin/hardener. The stoichiometric ratio with an AHEW of 24,4 is 1 :1 (100 g resin + 13,41 g amine). The dependency of the glass temperature T_g on the mixing ratio of hardener to resin is shown in Figure 5.

Maximum Tg can be achieved with lower amount of amine when an amine composition according to the present invention (TETA 80) is used compared to conventional TETA compositions of prior art (100:14,4 instead of 100:14,75).

Example 6: Preparation of a reactive polyamide resin:

A dimer acid based on C18 fatty acid was charged into the reaction vessel, then an amine curing composition according to Table 7 and monosodium phosphate were added. The mixture was stirred well for 30 minutes and then heated to 204°C under nitrogen. About 6-8% of the batch was set free as water and collected in a receiver during the heat-up. Then the receiver was drained and the drain valve closed. The mixture was held under vacuum (50 mmHg) between 193-202°C for 1 hour, then the vacuum was broken with nitrogen. The obtained polyamide was diluted with xylene (70% solids) and analyzed for amine value, IA AA ratio (The IA AA ratio is defined as imidazoline to amidoamine functional group ratio in the polymer measured by FTIR), viscosity and color.

Table 7: Composition of the amine compositions used for the polyamide resins

Table 8: Physical parameters of polyamide resins based on different amine compositions

Example 7: Use of the reactive polymamide resin in an epoxy application

105 parts of the polyamide curing agent obtained in example 6 were mixed with 100 parts of an epoxy resin (diglycidylether of bisphenol A). A mixture in a pot was used to measure the viscosity and gel time. Additionally, a film was prepared to measure the tack free time and through cure time after an induction time of 30 minutes.

Gel time: Gel time is defined as the time it takes until the stirrer stops after mixing the two components.

Tack free time: The tack free condition is reached and measured using mechanical recorders when the film surface has dried or cured. In this test method the tack free time is achieved when the continuous track in the film ceases and stylus starts to tear the film or leave a ragged/sharp- edged grove as it first begins to climb over the film.

Through cure time: Through cure time is reached when the film has solidified so completely that the stylus on the drier no longer leaves any visible mark on the film.

The different parameters obtained by using the polyamide resins from example 6 are shown in Table 9.

Polyamined based on amine compositions according to the present inventtion (TETA 80) show a lower initial viscosity and a faster through cure time.

Claims

Patent Claims

Amine composition comprising

60 to 95% by weight of linear TETA of formula I,

H

1 to 10% by weight of a methyl-substituted TETA,

1 to 12% by weight of diethylenetriamine (DETA),

0.1 to 5% by weight of methyl-substituted DETA, and

1 to 20% by weight of N-(2-aminoethyl)piperazine (AEPIP).

Amine composition according to claim 1 comprising

70 to 90% by weight of linear TETA of formula I,

2 to 8% by weight of a methyl-substituted TETA ,

2 to 12% by weight of diethylenetriamine (DETA),

0.5 to 4% by weight of methyl-substituted DETA, and

2 to 15% by weight of N-(2-aminoethyl)piperazine (AEPIP).

Amine composition according to claim 2 comprising

75 to 85% by weight of linear TETA of formula I ,

2 to 8% by weight of a methyl-substituted TETA,

3 to 10% by weight of diethylenetriamine (DETA),

1 to 3% by weight of methyl-substituted DETA, and

2 to 10% by weight of N-(2-aminoethyl)piperazine (AEPIP).

Amine composition according to claim 1 consisting of

60 to 95% by weight of linear TETA of formula I;

1 to 10% by weight of a methyl-substituted TETA,

1 to 12% by weight of diethylenetriamine (DETA),

0.1 to 5% by weight of methyl-substituted DETA,

1 to 20% by weight of N-(2-aminoethyl)piperazine (AEPIP), and

0.1 to 5% by weight of organic side products, organic solvent and water.

Amine composition according to claim 2 consisting of

70 to 90% by weight of linear TETA of formula I,

2 to 8% by weight of a methyl-substituted TETA , 2 to 12% by weight of diethylenetriamine (DETA),

0.5 to 4% by weight of methyl-substituted DETA, and

2 to 15% by weight of N-(2-aminoethyl)piperazine (AEPIP), and

0.2 to 4% by weight of organic side products, organic solvent and water.

Amine composition according to claim 3 consisting of by weight of linear TETA of formula I ,

by weight of a methyl-substituted TETA,

by weight of diethylenetriamine (DETA),

by weight of methyl-substituted DETA, and

by weight of N-(2-aminoethyl)piperazine (AEPIP, and

by weight of organic side products, organic solvent and water

Method for the production of an amine composition according to any one of claims 1 to 6 comprising

A) conversion of formaldehyde (FA), hydrogen cyanide (HCN) and ethylenediamine (EDA);

B) hydrogenation of the reaction mixture obtained in step A); and

C) purification of the hydrogenated reaction mixture obtained in step B), wherein the purification step C) comprises

1 ) removal of hydrogen;

2) removal of organic solvent and water;

3) passing the reaction mixture obtained in step 2) to an evaporator operated at a pressure in the range of 5 to 500 mbar and a temperature in the range of 100 to 250 °C;

4) removal of the gaseous phase from the evaporator; and

5) condensation of the gaseous phase.

8. Method according to claim 7, wherein the pressure is in the range of 5 to 100 mbar.

9. Method according to claim 8, wherein the pressure is in the range of 5 to 20 mbar.

10. Method according to any of claims 7 to 9, wherein the temperature is in the range of 125 to 200°C.

1 1 . Method according to claim 10, wherein the temperature is in the range of 125 to 180°C.

12. Use of an amine composition according to any one of claims 1 to 6 as an amine curing agent.

13. Amine curing agent composition comprising of 10 to 100% by weight of an amine composition according to one of claims 1 to 6 and 0 to 90% of other amine curing agents. Amine curing agent composition according to claim 13, comprising of 50 to 100% by weight of an amine composition according to one of claims 1 to 6 and 0 to 50% of other amine curing agents.

Amine curing agent composition according to claim 14, comprising of 90 to 100% by weight of an amine composition according to one of claims 1 to 6 and 0 to 10% of other amine curing agents.

Amine curing agent composition according to claim 13, consisting of an amine composition according to one of claims 1 to 6.

Use of an amine curing agent composition according to any of claims 13 to 16 for the production of epoxy resins.

Curable composition comprising an amine curing agent composition according to claims 13 to 16 and one or more epoxy resins, wherein the molar ratio of epoxy groups of the epoxy resin to active hydrogen atoms of the amine curing agent ranges from 0.5 to 1.3.

Method for the production of a curable composition according to claim 18 by mixing an amine curing agent composition according to any one of claims 13 to 16 with at least one epoxy resin.

Method of producing a cured epoxy resin by transferring the curable compositions according to claim 18 to a mold or applying said curable compositions to a surface.

Use of amine compositions according to any one of claims 1 to 6 for curing epoxy resins at temperatures below 20°C.

Reactive polyamide resin, obtainable from the reaction of an amine composition according to any of claims 1 to 6 with dimer fatty acids.

Use of amine compositions according to any one of claims 1 to 6 for the production of reactive polyamide resins.

Method for the production of reactive polyamide resins by condensing an amine compositions according to any one of claims 1 to 6 with a dimer fatty acid and removing water formed as a by-product from the condensation reaction.

Use of a reactive polyamide resin according to claim 22 for coatings, epoxy resins and adhesives.