N-aminopiperidine (NAPP) is an exocyclic hydrazine often used in pharmaceutical industries. We find it in several medication or molecules in progress.

First, NAPP is encountered in molecules which are ligands of cannabinoid receptors. These latters present an interest in the obesity treatment and in the psychiatric and neurological disorders treatments. These molecules are actually studied by many chemical groups such as Sanofi-Aventis [1], Solvay [2-4], Merck [5,6], Bayer [7, 8], Pfizer [9], Astrazeneca [10,11], Bristol-Myers Squibb [12].

Second, we meet this hydrazine in α-substituted benzylnaphtyl and benzylthiophene derivates, used for the treatment of diseases such as osteoporosis, breast cancer, and uterine and endometrial fibrosis. For instance, NAPP is employed as potential intermediate in a Raloxifene^Ò analogue [13].

NAPP also appears in the synthesis of pyridine-2-carboxamide derivatives which are applied in cancer treatment [14].

In view of the great potential of this hydrazine, we have decided to propose a new synthesis strategy by employing the hydroxylamine-O-sulfonic acid (HOSA) reagent. We also defined the extraction steps in order to propose a relevant purification process. In conformity with the particularly restricting pharmaceutical applications, this latter requires then the study of the solid-liquid-liquid and liquid-vapour binary and ternary phase diagrams, in order to optimize the successive unitary extraction operations.

NAPP (C₅H₁₀NNH₂) is obtained by reaction between piperidine (C₅H₁₀NH or PP) and HOSA (NH₂OSO₃H):

Kinetic studies of formation and degradation were carried out in basic medium in order to neutralize the released sulphuric acid [15].

The constant of formation k₁ and oxidation k₂ of NAPP in basic medium at 293 K are as follows:

k₁ = 3.39 ± 0.10 M⁻¹∙min⁻¹

k₂ = 1.17 ± 0.04 M⁻¹∙min⁻¹

Formation and oxidation reactions were found to be of first order.

The influence of the temperature was studied from 283 to 303 K, showing that the variation of k₁ versus temperature obeys to the Arrhénius law.

We have also proposed a kinetic model, taking into account all the reactions of formation and degradation of the mixture:

The resolution of this model shows that the yield of NAPP only depends on two variables. The first one is the p ratio of the initial molar concentrations [PP]₀/[HOSA]₀ and the second one, the ratio of rate constants k₂/k₁.

This reaction model was validated in concentrated medium (1.2 < p < 12; 283 < T < 373 K). Figure 1 shows a modelled kinetic behaviour of a mixture with industrial concentrations.

The optimal conditions for the synthesis were then determined as follows:

-       Aqueous solution of HOSA 32 w% (0.69 M)

-       Aqueous solution of PP 66 w% (5.55 M)

-       Sodium hydroxide (1.39 M)

-       Synthesis temperature: 293 K The reaction is almost ended after 0.35 second and gives NAPP with 96% yield.

A kinetic study of the formation of NAPP was also realized in buffer medium. The synthesis of NAPP in buffer medium implies the two following parallel reactions:

Leading to the global reaction:

The pH is thus controlled by the buffer mixture piperidine/piperidinium sulphate salt. During the synthesis, 3 moles of PP are consumed to give one mole of NAPP and one mole of piperidinium sulphate salt. It involves the theoretical stoiechiometric relations:

The rate constant k₃ of formation of NAPP in buffer medium at 293 K was then determined: at 293 K.

The formation of a piperidinium salt leads to a slow down in the synthesis reaction of NAPP.

Figure 2 shows the evolution of NAPP yield in basic and buffer media as function of the p ratio.

This graph shows that, when p ≥ 8, the NAPP yields are the same. So, it is also possible to implement this synthesis in buffered medium with the optimal conditions obtained in basic medium.

We decided to synthesize NAPP in buffer medium with the following conditions:

-       Aqueous solution of HOSA 32 w% (0.69 M)

-       Aqueous solution of PP 66 w% (5.55 M)

-       Synthesis temperature: 293 K At the end of the synthesis, the crude solutions are complex and diluted. The extraction process was optimized thanks to the study of the phase diagrams involved. The first step of the extraction consists of a demixing generated in-situ by the sodium sulphate released after neutralization of the piperidinium salt by sodium hydroxyde. Four isothermal sections [16] of the solid-liquidliquid H₂O-Na₂SO₄-C₅H₁₀NH ternary diagram were established at 293, 298, 313 and 323 K under atmospheric pressure by Isoplethic Thermal Analysis (ITA) [17]. A polythermic diagram was then designed between 293 and 323 K [18]. Figures 3(a) and 3(b) show respectively the isotherm 293 and 313 K expressed in mass fractions, under atmospheric pressure.

The second step is a demixing induced by addition of sodium hydroxide. This one allows concentrating the organic compounds in order to optimize the next distillation operations. This step required thus the study of the solid-liquid-liquid H₂O-NaOH-C₅H₁₀NH ternary diagram. Three isothermal sections were established at 293, 313 and 323 K by ITA [19]. Figures 4(a) and 4(b) show respectively the isotherm 293 and 323 K expressed in mass fractions, under atmospheric pressure.

The final step of the extraction process consists of a distillation of the recovered organic phase mixture composed of water, PP and NAPP (46.5 w% water; 45.6 w% PP; 7.9 w% NAPP). In order to optimize this step, we studied the liquid-vapour H₂O-C₅H₁₀NH-C₅H₁₀NNH₂ ternary diagram. The three limit binary systems, H₂OC₅H₁₀NH, H₂O-C₅H₁₀NNH₂ and C₅H₁₀NH-C₅H₁₀NNH₂, were established by ebulliometry under atmospheric pre-

ssure. The liquid-vapour H₂O-C₅H₁₀NH-C₅H₁₀NNH₂ ternary diagram is characterized by two distillation domains, linked to the existence of a positive azeotropic mixture in the H₂O-C₅H₁₀NH binary system. Figure 5 shows the H₂O-C₅H₁₀NH-C₅H₁₀NNH₂ polythermal ternary system under atmospheric pressure.

The experimental coordinates of the H₂O-PP azeotropic mixture are:

x(C₅H₁₀NH) = 0.661 T_vap = 364.95 K To secure the process, two measurement sets of mixing enthalpies were realized at 298 K, by mixing calorimetry. The first series concerns the solvation enthalpies of HOSA-water and PP-water binary mixtures. The second one, focused on the measurement of the reaction enthalpy of NAPP according to the dilution of the reagent involved, allowed us to determine the hypothetical reaction enthalpy in anhydrous medium. The infinite dilution enthalpy is equal to −140 kJ/mol and the enthalpy in anhydrous medium is estimated at −340 kJ/mol. The exothermicity of the reaction requires thus the necessity of an impressive cooling of the synthesis reactor in order to manage the temperature of the medium.

We can propose two different processes depending on whether PP is injected anhydrous or as an azeotropic mixture of PP in water (66 w% PP).

In both cases, the first step consists in carrying out the synthesis reaction at 293 K, in a buffer medium and with

a molar ratio of initial concentrations [C₅H₁₀NH]₀/[NH₂OSO₃H]₀ equals to 8. The NAPP yield reaches then 93%.

After the synthesis, the crude mixture is mainly composed of NAPP, piperidinium sulphate salt, PP in excess, water and some by-products (tetrazene and triazanium salt derivatives). In a second operation, we add a 32 w% aqueous solution of sodium hydroxide which generates sodium sulphate by neutralization of the piperidinium salt and induced the salting-out.

G_anh is the mass composition of the crude mixture obtained after the NAPP synthesis by injection of anhydrous PP and neutralization of the salt. G_az is the mass composition of the crude mixture after addition of an azeotropic mixture of PP and water and neutralization of the salt at the end of the NAPP synthesis. These two compositions are the following:

G_anh = 44.7 w% PP; 12.5 w% Na₂SO₄; 42.8 w% H₂O.

G_az = 33.1 w% PP; 9.9 w% Na₂SO₄; 57 w% H₂O.

As shows the isotherm 293 K of the solid-liquid H₂ONa₂SO₄-C₅H₁₀NH ternary diagram (Figure 6), the G_az and G_anh points are located respectively in the L + Na₂SO₄∙10H₂O and the L3+ Na₂SO₄∙10H₂O + Na₂SO₄ domains.

At this temperature, any demixing zone appears.

But at 313 K (Figure 7), G_az and G_anh are respectively located in the demixing zone and in the L1 + L2 + Na₂SO₄ triphasic domains.

It is thus preferable to synthesize NAPP starting from an azeotropic mixture of PP and water and implement the neutralization of the piperidinium salt at 313 K. In these conditions there is no precipitation of anhydrous sodium sulphate. After the neutralization step, the recovered organic liquid M is mainly composed of water (46.5 w%), PP (45.6 w%) and NAPP (7.9 w%).

Looking at the liquid-vapour ternary system, the mixture M is in fact located in the water/azeotrope/NAPP distillation field (Figure 8).

In this configuration, two successive distillation operations are required to get rid of the azeotropic mixture and the water in excess and finally a third distillation, under reduced pressure, permits to obtain the product in conformity with the pharmaceutical applications (purity up to 99 w%).

In order to simplify the distillation process, we preferred to dehydrate partially the organic liquid M thanks to a demixing observed in the solid-liquid-liquid H₂ONaOH-C₅H₁₀NH ternary system.

The aim of this step is to generate a new demixing by addition of an aqueous solution of sodium hydroxide (50 %w). The added quantity of soda was optimized in order to obtain an organic phase with a mass composition very close to those of the azeotrope. This organic liquid M’ is then composed of water (31.8 w%), PP (59.2 w%) and NAPP (9 w%).

This mixture M’ is very near the monoazeotropic distillation curve (Figure 9).