Method for preparing phosphorus-containing a-aminonitriles

Abstract

The present invention relates primarily to processes conducted in a continuously operated reactor for preparing particular phosphorus-containing α-aminonitriles of the formulae (Ia) and (Ib) defined hereinafter from corresponding phosphorus-containing cyanohydrin esters and to the use thereof for preparation of glufosinate or of glufosinate salts. The present invention further relates to a process for producing glufosinate/glufosinate salts.

Claims

1. A process for preparing a mixture comprising at least one compound of the formula (Ia) and at least one compound of the formula (Ib) ##STR00007## wherein one or more compounds of the formula (II) ##STR00008## where in each case: R.sup.2 is (C.sub.1-C.sub.12)-alkyl, (C.sub.1-C.sub.12)-haloalkyl, (C.sub.6-C.sub.10)-aryl, (C.sub.6-C.sub.10)-haloaryl, (C.sub.7-C.sub.10)-aralkyl, (C.sub.7-C.sub.10)-haloaralkyl, (C.sub.4-C.sub.10)-cycloalkyl or (C.sub.4-C.sub.10)-halocycloalkyl, R.sup.5 is (C.sub.1-C.sub.12)-alkyl, (C.sub.1-C.sub.12)-haloalkyl, (C.sub.6-C.sub.10)-aryl, (C.sub.6-C.sub.10)-haloaryl, (C.sub.7-C.sub.10)-aralkyl, (C.sub.7-C.sub.10)-haloaralkyl, (C.sub.4-C.sub.10)-cycloalkyl or (C.sub.4-C.sub.10)-halocycloalkyl, are converted in a continuously operated reactor while mixing with NH.sub.3 in liquid or supercritical form, where the mixing of the compound(s) of the formula (II) and NH.sub.3 is effected under the following parameters: a mixing time for attainment of a coefficient of variation of 0.10 or less (CoV≤10.sup.−1) is less than 30 seconds.

2. The process according to claim 1, wherein R.sup.2 is (C.sub.3-C.sub.6)-alkyl, R.sup.5 is (C.sub.1-C.sub.4)-alkyl, (C.sub.6-C.sub.8)-aryl or (C.sub.5-C.sub.8)-cycloalkyl.

3. The process according to claim 1 or 2, wherein R.sup.2 is (C.sub.4-C.sub.5)-alkyl, R.sup.5 is methyl, ethyl or isopropyl.

4. The process according to claim 1, wherein the mixing is effected under the following parameters: the mixing time for attainment of a coefficient of variation of 0.05 or less (CoV≤5*10.sup.−2) is less than 15 seconds.

5. The process according to claim 1, wherein the continuously operated reactor is a tubular reactor having a length of more than 100 times the characteristic length.

6. The process according to claim 1, wherein a total amount of 2.0 to 3.6 molar equivalents of NH.sub.3 is used.

7. The process according to claim 1, wherein the NH.sub.3 used is essentially anhydrous, wherein a water content in the NH.sub.3 used is not more than 0.5% by weight.

8. The process according to claim 1, wherein one or more compounds of the formula (II) and NH.sub.3 each in liquid form are mixed in the reactor.

9. The process according to claim 1, wherein the reaction is effected at a temperature in the range from 10 to 80° C.

10. The process according to claim 1, wherein the reaction is effected at a temperature in the range from 20 to 70° C.

11. The process according to claim 1, wherein the reaction is effected at an absolute pressure of not more than 120 bar.

12. A process for preparing glufosinate ##STR00009## or glufosinate a salt, wherein, in this process, a mixture comprising at least one compound of the formula (Ia) and at least one compound of the formula (Ib) is used ##STR00010## where R.sup.2 is (C.sub.1-C.sub.12)-alkyl, (C.sub.1-C.sub.12)-haloalkyl, (C.sub.6-C.sub.10)-aryl, (C.sub.6-C.sub.10)-haloaryl, (C.sub.7-C.sub.10)-aralkyl, (C.sub.7-C.sub.10)-haloaralkyl, (C.sub.4-C.sub.10)-cycloalkyl or (C.sub.4-C.sub.10)-halocycloalkyl, are converted in a continuously operated reactor while mixing with NH.sub.3 in liquid or supercritical form, where the mixing of the compound(s) of the formula (II) and NH.sub.3 is effected under the following parameters: a mixing time for attainment of a coefficient of variation of 0.10 or less (CoV≤10.sup.−1) is less than 30 seconds.

13. A process for preparing glufosinate or a glufosinate salt comprising steps (a) and (b): (a) preparing a mixture prepared by a process defined in claim 1 and comprising at least one compound of the formula (Ia) and at least one compound of the formula (Ib), and (b) converting the mixture prepared in step (a) and comprising at least one compound of the formula (Ia) and at least one compound of the formula (Ib) to glufosinate or to glufosinate salts, or (b) using the mixture prepared in step (a) and comprising at least one compound of the formula (Ia) and at least one compound of the formula (Ib) for preparation of glufosinate or of glufosinate salts.

14. The process according to claim 13, wherein, in step (b), an acidic hydrolysis of the nitrile group and the phosphinic ester group to give compounds of formula (Ia) and an acidic hydrolysis of the nitrile group to give compounds of the formula (Ib) are effected.

15. A mixture prepared by a process defined in claim 1 and comprising at least one compound of the formula (Ia) and at least one compound of the formula (Ib) for use in preparing glufosinate or glufosinate salts.

16. The process according to claim 4, wherein the mixing time for attainment of the coefficient of variation of 0.05 or less (CoV≤5*10.sup.−2) is less than 4 seconds.

17. The process according to claim 5, wherein the continuously operated tubular reactor has a length of more than 1000 times the characteristic length.

18. The process according to claim 6, wherein 2.2 to 3.2 molar equivalents is used, each based on the amount of compounds of the formula (II) used.

19. The process according to claim 12, wherein the glufosinate salt is glufosinate sodium, glufosinate hydrochloride, or glufosinate ammonium.

20. The process of claim 13 wherein the mineral acid comprises aqueous HCl (hydrochloric acid).

Description

EXAMPLES

[0118] In the course of the in-house studies, experiments were conducted in order to enable direct comparability of batchwise and continuous process regime on the laboratory scale. In addition, the products from the aminolysis were converted to glufosinate-ammonium in order to be able to better detect differences in the yield of glufosinate-ammonium on the basis of in situ yields than would have been possible on the basis of the phosphorus-containing α-amino nitriles of the formulae (Ia) and (Ib). Furthermore, the experimental results were ascertained by uniform analysis.

[0119] All data are based on weight unless otherwise stated.

[0120] The water content in the NH.sub.3 used was never more than 0.25% by weight, and usually not more than 0.1% by weight.

ABBREVIATIONS USED

[0121] ACM: 3-[n-butoxy(methyl)phosphoryl]-1-cyanopropyl acetate, compound of the formula (II) [0122] AMN:n-butyl (3-amino-3-cyanopropyl)methylphosphinate, compound of the formula (Ia) [0123] GFA: glufosinate-ammonium [0124] GFA-is: in-situ yield of glufosinate-ammonium [0125] conti-reactor: reactor operated by means of a continuous process regime

Example A1: Fully Continuous Process Regime

[0126] The reaction system was composed of the following reactor sections arranged in series: [0127] i.) micromixer, as main mixer for the ACM and liquid ammonia reactants (cascade mixer 2, Ehrfeld Mikrotechnik BTS GmbH); [0128] ii.) MIPROWA reactor with three comb sections in a crossed arrangement per channel in the interior (comb elements in 700 μm, comb element separations 1500 μm, comb element angle) 45°, as heat exchanger and post-mixer (MIPROWA LAB reactor A4, Ehrfeld Mikrotechnik BTS GmbH); [0129] iii.) flow tube with internal diameter 1.58 mm and variable length: 0 m (i.e. not used), 10 m, 20 m or 40 m (according to volume required, i.e. further reaction time).

[0130] The mixing time to attain a coefficient of variation of 0.05 or less (CoV≤5*10.sup.−2) was less than 4 seconds, and usually less than 1 second.

[0131] All reactor sections except i.) were operated under temperature control. For details of temperature control of reactor sections ii) and iii), reference is made to table 1 below. The micromixer (reactor section i)) was operated at ambient temperature (about 20° C.); the volume of the micromixer was < 1/100 of the total reactor volume.

[0132] Before each experiment, the entire reaction system was purged with ACM and the desired ACM flow rate was established. Ammonia gas was liquefied under pressure and pumped constantly in the direction of the reaction system by means of an HPLC metering pump, the flow rate of which was regulated by means of a mini-Coriolis mass flow meter. The ammonia conduit to the reactor was purged with liquid NH.sub.3 prior to the start of the experiment in order to lead off residues of inert gas and ammonia gas by bypass. Subsequently, in the bypass position, a constant ammonia pressure and flow rate was built up.

[0133] The technical grade ACM (the ACM content was 90% by weight) was conveyed continuously into the reactor from a reservoir vessel. Subsequently, the ammonia stream was switched from bypass to reactor and hence the aminolysis was started. The reactant streams were chosen such that the corresponding residence times in the continuously operated reactor were attained. The mass flow rates were typically in the range of 3.2-10.9 g/min (based on the technical grade ACM) or 0.5-1.8 g/min (based on liquid NH.sub.3). As well as the adjustment of the mass flow rates or volume flow rates and of the temperatures, the volumes of the reactor were also varied in order to establish different residence time. The molar NH.sub.3/ACM ratio was always 2.77.

[0134] Table 1 shows an overview of the different residence times and reaction temperatures in the case of fully continuous aminolysis.

[0135] In order to prevent degassing of as yet unconverted ammonia under reaction conditions, the pressure at the end (the outlet) of the continuous reactor was always set to 21 bar. At this minimum pressure in the plant, pure ammonia reaches its boiling point at 51° C. Outgassing of unconverted ammonia was never observed at the reactor outlet under reaction conditions, not even at the higher temperatures.

[0136] After passing through the reactor sections, the product mixture obtained was quenched directly. For this purpose, the liquid stream obtained was introduced into a stirred initial charge of hydrochloric acid solution without immersion of the introduction pipe into the hydrochloric acid. The heat of neutralization released was removed by cooling in an ice bath, and the temperature did not exceed 40° C.

Example A2: Predominantly Continuous Process Regime

[0137] The procedure for experiments with predominantly continuous process regime corresponded largely to that of example A1. In a departure from the fully continuous process regime described in example A1, in this case, the product stream exiting from the continuous reactor was collected unquenched in a discontinuously operated receiver, where it reacted further. This process regime was referred to as predominantly continuous since the majority of the ACM conversion takes place in the continuously operated part of the reactor under the reaction conditions used.

[0138] In this variant, the reaction system was composed of the following reactor sections connected in series: [0139] i.) micromixer, as main mixer for the ACM and liquid ammonia reactants (cascade mixer 2, Ehrfeld Mikrotechnik BTS GmbH); [0140] ii.) MIPROWA reactor with three comb sections in a crossed arrangement per channel in the interior (comb elements in 700 μm, comb element separations 1500 μm, comb element angle) 45°, as heat exchanger and post-mixer (MIPROWA LAB reactor A4, Ehrfeld Mikrotechnik BTS GmbH); [0141] iii.) flow tube with internal diameter 1.58 mm and variable length: 0 m (i.e. not used), 10 m, 20 m or 40 m (according to volume required, i.e. further reaction time); [0142] iv.) autoclave (series 5500, Parr) with total volume 300 ml.

[0143] The mixing time to attain a coefficient of variation of 0.05 or less (CoV≤5*10.sup.−2) was less than 4 seconds, and usually less than 1 second.

[0144] All reactor sections except i.) were operated under temperature control. For details of temperature control of reactor sections ii) and iii), reference is made to table 2 below. The micromixer (reactor section i)) was operated at ambient temperature (about 20° C.); the volume of the micromixer was < 1/100 of the total reactor volume.

[0145] In a departure from the procedure according to the above example A1, after establishment of equilibrium in the continuous reactor, the product mixture from the continuous reactor was introduced directly into an autoclave (batch reactor) that was under an argon pressure of 21 bar (absolute) and had already been preheated (the initial products obtained beforehand were discarded) and stirred therein. During the filling of the autoclave, the plant pressure rose since the system was being operated in a closed manner. In the autoclave, a distinction was made between two times: the filling time for the introduction of the product stream (t.sub.2F), and the further reaction time (t.sub.2N) in which the mixture can react further.

[0146] Table 2 shows an overview of the different residence times, filling times and further reaction times and of the reaction temperatures in a predominantly continuous process regime for the aminolysis.

[0147] At the end of the chosen further reaction time, the product mixture (under pressure) was introduced through the sample tube of the autoclave rapidly into a stirred, ice bath-cooled reservoir containing hydrochloric acid under temperature control (max. 40° C.). This brings about immediate stoppage of the reaction (quenching). The autoclave was subsequently rinsed out with a calculated amount of hydrochloric acid and this mixture was combined with the already quenched product.

Comparative Example C1: Batchwise, Metering-Controlled Process Regime

[0148] The batchwise aminolysis of ACM was conducted in a 300 ml autoclave (series 5500, Parr). The autoclave was initially charged with 164.3 g (566 mmol or 150 ml) of technical grade ACM (the ACM content was 90% by weight), preheated and charged with argon. Subsequently, liquid ammonia, while stirring constantly (600 rpm) and while cooling, was metered in below the liquid surface. In this way, 26.7 g of NH.sub.3 were introduced (1.57 mol) and hence a molar NH.sub.3/ACM ratio of 2.77 was obtained. A metering time of four hours was chosen, one reason being to permit comparability with configuration variants on the industrial scale. The ammonia metering time was followed by a further reaction time for completion of the conversion, which was either 10 min or 60 min.

[0149] Table 3 shows an overview of the respective times and temperatures.

[0150] After the end of the further reaction time, as described in example A2, the product was introduced into hydrochloric acid and the same procedure was followed.

[0151] Hydrolysis, neutralization, content determination and reproduction of the reaction products from examples A1, A2 and C1

[0152] After leaving the reactor in example A1 or after the end of the further reaction time in example A2 or of comparative example C1, the reaction was stopped immediately by discharge into an initial charge of hydrochloric acid. The subsequent workup was effected by hydrolysis, which was conducted with 4.6 molar equivalents of HCl (per mole of ACM).

[0153] A previously weighed three-neck flask (500 or 1000 ml) with magnetic stirrer was initially charged with a previously weighed amount of 32% hydrochloric acid. Either the product of the aminolysis directly from the experimental plant of example A1 (100 g) or the entire contents of the autoclave from example A2 or C1 (190 g of product) via the sample tube thereof was guided directly in a dropwise manner onto the acid in the flask. The contents of the flask were cooled in an ice-water bath and mixed by magnetic stirrer (700 rpm). The addition was rapid; the temperature was kept at <40° C. An orange/yellow solution was obtained. After the dropwise addition or quantitative conversion, the total weight of the solution was determined. The mixture was stored in a refrigerator overnight.

[0154] The next day, the three-neck flask was equipped with a reflux condenser and dropping funnel with pressure equalization (dropping funnel between flask and condenser). The entire apparatus was purged with argon, stirred (600 rpm) and then boiled at reflux for 7 h. During the heating, a distillate with an organic upper phase and aqueous lower phase condensed out in the dropping funnel connected, and a portion of the lower phase was repeatedly discharged back into the flask. Organic coproducts, conversion products and by-products accumulate in the upper phase. After the hydrolysis for 7 hours, about 200 ml of distillate were removed, consisting of upper phase and lower phase. The upper phase was discarded. The mixture in the flask was cooled down to room temperature while stirring and left to stand overnight.

[0155] Solid substrate precipitated out overnight. The crystal cake was first stirred up with a glass rod and then stirred thoroughly by magnetic stirrer. The mixture was neutralized dropwise to pH 6.5 with 25% aqueous ammonia solution while cooling with ice water and stirring. The pH was determined by means of a pH electrode. Subsequently, a sufficient amount of water was added to dissolve all solids and there was a clear, orange solution at room temperature. The weight of the overall solution was determined (about 700 g). Accurately weighed samples thereof were taken, for the purpose of determination of GFA content in situ by HPLC.

[0156] All experiments were conducted more than once; therefore, the yields of GFA stated are averages of the yields of GFA in situ, both from multiple experiments and from analyses conducted repeatedly.

[0157] The yields of GFA in situ (identified as GFA-is) are shown in tables 1-3 below. In addition, tables 1 to 3 give an overview of the various residence times, filling times and further reaction times and of the reaction temperatures in a continuous, predominantly continuous and batchwise reaction regime for the aminolysis.

[0158] Elucidation of the abbreviations used in the tables:

TABLE-US-00001 T.sub.1M Temperature in the MIPROWA reactor τ.sub.1M Average hydrodynamic residence time in the MIPROWA reactor T.sub.1S Temperature in the flow tube τ.sub.1S Average hydrodynamic residence time in the flow tube τ.sub.1_total Average hydrodynamic residence time in the overall continuous reactor (contribution from micromixer negligible) T.sub.2 Temperature in the autoclave 0.5 * t.sub.2F Half the filling time of the autoclave t.sub.2N Further reaction time in the autoclave after filling GFA-is In situ yield of glufosinate-ammonium (GFA)

TABLE-US-00002 TABLE 1 Reaction conditions and results from example A1 MIPROWA reactor Flow tube Experiment T.sub.1M τ.sub.1M T.sub.1S τ.sub.1S τ.sub.1_total GFA-is No. [° C.] [min] [° C.] [min] [min] [%] A1.1 30 19.6 40 39.4 59.0 97-98 A1.2 30 19.6 50 39.4 59.0 97-99 A1.3 50 5.0 50 5.0 10.0 94-95 A1.4 50 6.6 50 13.4 20.0 96-97 A1.5 60 5.0 60 5.0 10.0 97-98 A1.6 60 6.6 60 13.4 20.0 95-96

TABLE-US-00003 TABLE 2 Reaction conditions and results from example A2 MIPROWA reactor Flow tube Autoclave Experiment T.sub.1M τ.sub.1M T.sub.1S τ.sub.1S τ.sub.1_total T.sub.2 0.5 * t.sub.2F t.sub.2N • GFA-is No. [° C.] [min] [° C.] [min] [min] [° C.] [min] [min] [min] R [%] A2.1 30 5.0 — — 5.0 50 6.6 10 21.6 0.30 94-95 A2.2 30 6.6 30 13.4 20.0 50 8.9 10 38.9 1.06 95-96 A2.3 40 3.4 40 1.6 5.0 50 4.5 10 19.5 0.34 96-97 A2.4 40 5.0 40 5.0 10.0 50 6.7 10 26.7 0.60 96-97 A2.5 50 5.0 50 5.0 10.0 50 6.7 10 26.7 0.60 97-98 Sum • = τ.sub.1_total + 0.5 * t.sub.2F + t.sub.2N Ratio R = τ.sub.1_total/(0.5 * t.sub.2F + t.sub.2N)

TABLE-US-00004 TABLE 3 Reaction conditions and results from comparative example C1 NH.sub.3 Further Experiment T metering time reaction time GFA-is No. [° C.] [min] [min] [%] Cl 30 240 60 93-94 C2 35 240 10 91-92 C3 35 240 60 94-95 C4 40 240 10 92-93 C5 40 240 60 93-94