Technical Field

This invention pertains to aryl amides, in particular to an improved process for synthesizing N- (4-anilino phenyl) Amides and related compounds.

Background

U S Patent 3,907,893 (referred to herein after as the

Parker Patent) describes the basic process for synthesizing aromatic antioxidant amides, an improvement of which is the subject of this application.

U S Patent 3,538,159 describes the use of alkali metal alkoxides as catalysts in the production of di-N-alkyl- amides.

The basic process comprises reacting an aliphatic or aro- matic ester, either saturated or unsaturated, having formula

(A)

0

R ¹ - C - 0 - R ² (A) with an aromatic amine having the general formula (B)

in the presence of (C , a base selected from the group con¬ sisting of alkali metal amides and alkali metal alkoxides to yield an intermediate metallic salt and an alchol. R is selected from the group consisting of alkyl radicals having from one to 20 carbon atoms, aryl radicals having from six to 12 carbon atoms, cycloalkyl radicals having from five to 12 carbon atoms, alkenyl radicals having from two to 10

2 carbon atoms, R is selected from the group consisting of alkyl radicals having from one to 10 carbon atoms, and alkenyl radicals having from two to 10 carbon atoms and Rp- is selected from the group consisting of alkyl radicals having from one to 20 carbon atoms, cycloalkyl radicals having from five to 12 carbon atoms, aryl radicals having from β to 12 carbon atoms and aralkyl radicals having from seven to 13 carbon atoms.

OMPI

This reaction is performed in an aromatic solvent. The conversion in this first reaction is not complete, ranging generally from 50 to 90 . Because .of this and because the base and the aromatic amine are sometimes charged in excess of the stσichiometric amount, the reaction product contains significant quantities of unreacted base and unr.eacted amine.

The next step-in the Parker process is removal of the intermediate metal salt from the reaction mixture by some physical separation means, such as filtration. The excess of raw materials ^' referred to above makes this separation difficult and time consuming because the particles in the reaction slurry are very fine (some being less than one micron in diameter). Alkali metal alkoxides, such as sodium methoxide, have the appearance of talcum powder, and thus the very small particle size of this material contributes to the filtration problem. The reaction slurry has the appearance of a milk shake, and the filter cake is a slimy material which tends to blind the filter medium. The filter cake is invariably saturated with solvent, and removal from the filter for eventual hydrolysis, becomes a very messy operatio In the basic process, the filter cake is next washed with an organic solvent such as.xylene to remove unreacted ^' aromatic amine (the presence of which must be minimized in the final product). This may be followed by subsequent washes with the same or another solvent such as hexane or toluene. ^" The filte cake is then reslurried in dilute, aqueous acid which hydrolyzes the metallic salt to the desired product.(D).

H H O

I 1 I! _* . R ³ - N-/o^-N - C - R ¹ ^

Prior to hydrolysis the filter cake may be dried. The solid product is then removed by filtration or another appropriate unit operation and is dried.

Some reactions yield intermediate metallic salts which are liquids, not susceptible to filtration but which can be separated from the solvent by extraction and/or decantation. The Parker process is applicable, with variation of separatio techniques, to these various intermediates.

UUREA

OMPI

Because of the poor quality of the first filter cake, the particle size of the product upon hydrolysis is often very small, making the final filtration also difficult.

It is therefore highly desirable to alleviate the problems 5 associated with the hydrolysis and filtration steps.

An article by- DeFeoand and Strickler, "An Improved Method of Synthesis of ^'" Secondary Amides from Carboxylic Esters" in J.O.C. , Volume 28 p. 2915-2917 (1963), discusses the synthesis of amides by reaction between an ester and an amine 10 using sodium methoxide. This article teaches the hydrolysis of the reaction mixture in dilute .mineral acid (10% HC1 used in example).

A separate improvement in the Parker process achieved through adding an alcohol to the first reaction is the subject - 15 of .another U.S. patent application, serial number 878,199, currently pending.

Summary of the Invention

The Parker or ester process has been improved by replacing 20 three of the finishing steps (1. removing the alkali metal salt of the first reaction by filtration or other means, 2. washing the salt (filter cake) with an organic solvent, and 3. hydrolyzing the salt with dilute acid). These, three steps are . replaced by the single step of hydrolyzing the salt in-situ by 25 mixing the reaction product with water at a temperature at which most of the reaction product is suspended in the water phase and most of the unreacted aromatic amine is dissolved . in the reaction_solvent phase.

By following this technique, a number of advantages are 30 realized:

1. ^" When sodium methoxide is used for the base, it is transferred to the aqueous phase where: it reacts with water forming sodium hydroxide and methanol, thus alleviating the fine "talcum

35 powder" solid phase which caused problems dur¬ ing filtration, since sodium hydroxide and methanol are miscible with water.

2. Both the first filtration ^' step and the ing wash step are eliminated.

3. The slurry resulting after hydrolysis is easy to filter and relatively free of organic solvent when water washed and air dried.

4. Decreased processing time has.been realized through increased rates of-hydrolysis and filtration.

5. A raw material savings.is realized in the elimin¬ ation of the organic solvent used in washing the filter cake. 6. Another advantage realized through this process improvement has been a decrease in water pollutio In the old process, there is not complete recover of the reaction solvent as filtrate in the first filtration because a significant quantity of sol- vent remains with the filter caJsre. In the filtration following hydrolysis with the dilute acid, this residual reaction solvent is transferre to the aqueous filtrate and wash water, being a pollutant therein. On the other hand, because of "the improved filtration achieved through this improvement, very little reactiQn solvent remains on the filter cake which is more in the nature of granules or pellets rather than fine particles or

.slime. Almost all of the reaction solvent is the recovered by a simple decantation of the filtrate, leaving, very little in the aqueous phase. The slurry resulting after in-situ hydrolysis does not have ^• the slimy appearance of the old reaction slurry. This could be due.to the increased particle size of the product over tha which formerly occurred. It also could be due to the absence of the "talcum powder" unreacted base. ^' This results in a muc improved filtration rate and also gives a cake which is easie to wash with water.

The in-situ hydrolysis has been carried out with water alon and with water plus one or two organic solvents, such as hexane and xylene. When hexane or hexane and xylene were use together with water, an increase in the crude yield of the product was observed. It is believed that impurities (side reaction products) which would normally remain in the orgs

phase, were transferred to the aqueous phase when hexane was added to the solvent system. Although these side products added to the apparent or crude yield, they did not increase the true yield of N-(4-anilinophenyl) -amide. In fact,, the amide produced by solvent-aided in-situ hydrolysis was some¬ what lower in purity and performance than those made by in- situ hydrolysis- ith water only.

The amides (D) described herein are useful as antioxidants in elastomers. They can be- incorporated into elastomers by conventional techniques well known to those in the art, such as by addition to polymer latices or to solid polymers in an internal mixer or on a mill.

Amides formed from an unsaturated ester such as methyl methacrylate are copolymerizable with conventional unsatur- - ated monomers such as those used in the preparation of unsaturated synthetic elastomers. When copolymerized with these monomers, the amide products become a part of the elastomer molecules, are non-volatile and are not extractable. These built-in antioxidants are advantageous for long term resistance in polymers. An amide produced by in-situ hydro¬ lysis has been successfully copolymerized.

Description of the Preferred Embodiments

In-situ hydrolysis with water only, gives the best product with the highest purity, and the best polymerization rate and conversion for the copolymerizable amides. The use of water only has another advantage over the solvent-aided hydrolysis. After filtration of the hydrolyzed reaction product, the aqueous phase of the filtrate.can be easily decanted, and the organic phase can be recycled as reaction solvent. Filtrate from the solvent aided hydrolysis requires a separation of hexane (or other solvent) from the reaction solvent before the organic phase can be considered for recycling. The investigation which led to this improvement was directed particularly toward improving the process for synthesis of N-(4-anilinophenyl) methacrylamide (E) through reaction of methyl methacrylate with p-aminodiphenylamine in the presence of sodium methoxide.

.5 Preferred conditions for the in-situ hydrolysis of the inter¬ mediate sodium salt in this case are 21-24°C. and a 1:1 volume ratio of water to the reaction mixture to which it is added.

The improvement of this invention will be further clarified

10 by a consideration of the following examples, which are in¬ tended to be purely exemplary. The following definitions apply to terminology used in the data reported:

% Crude Yield , ("* ^crude ? ^rod ^- ^~ Ψ ^~ ^ _x Λ ^ot - ^slur *T ** V I

*w slurry sample ' ^theoretical yld

15 "Slurry" is the reaction mixture after the first reaction (be fore addition of water). "Crude product" is the dried pro¬ duct after hydrolysis (no analysis for purity).

20 Filtration Rate= (fflis. product slurry/454)x ^β θCsT?

(0.0664 f ^~ ) x (filtration time in in.) • "Product Slurry" = 2 phase mixture obtained after hydrolysis. 454 = grams per pound.

Melting point is indicative of product purity, a range of 104-105°C. corresponding to almost pure product.

25

Also, in the data which follows, the term "standard hydro¬ lysis" refers to the old method described in the background section.

-~ Example I

Table 1 shows the data generated from various hydrolysis experiments utilizing the sodium salt slurry ^' from a pilot plant product run of N-(4-anilinophenyl) methacrylamide. Most of the batches were made with an amount of water equal

35 to the amount of reaction slurry, and the two phases were agitated with mild agitation (250 rp , half-moon teflon agitator in one liter round bottom flask) at 21-24°C. for 30 minutes. Teflon is the registered trademark of E.I. Dupont de Nemours. The resulting product slurry was then filtere

washed with water, and the cake was dried in an oven at 60°C. These conditions gave the best quality product with satis¬ factory yield. A few variations ^• in temperature and amount of water were tried, but only adverse effects either in filtration or quality were noticed.

It was apparent from the very beginning that the solubili¬ ties of p-aminodiphenylamine, impurities, and the product in the mixed xylene solvent being used would play an important role in determining the right conditions for hydrolysis. Increasing the temperature increases the solubility of the p-aminodiphenylamine in the mixed xylenes, but it also in¬ creases the solubility of the product, resulting in less recoverable yield _* Decreasing the temperature reverses this trend and the p-aminodiphenylamine or other impurities pre- cipitate out affecting the quality of the product.

Table 1

Laboratory ' Mixed Hydrolysis Filtration+ %

Run Slurry Water Xylenes Hexane Temp Rate Crude # gms gms gms gms °C. lbs/hr-ft^ Yield Comment

88-1 655 Standard hydrolysis 21-24 66.8 Slurry filtered a

(pH=12.5) washed with mixed xylenes and hexan arid then reslurri in water and filt again.

89-4 500 500 375 21-24 77.0

89-6 200 100 100 100 21-24 995.0 74.4

90-7 200 200 100 100 21 898.0 74.4

90-8 200 200 21-24 1592.3 54.7

91-9 200 200 150 21-24 1095.0 74.4

91-10 200 200 50 21 896.0 63.4

91-11 400 Standard hydrolysis 21-24 710.0 73.3 Hydrolyzed Produc

(pH=6.0) wouldn't filter a all at pH of 12.6.

92-12 200 200 35 —— 21-24 1731.6 54.7 Mixed xylenes use for rinsing the flas .

Example II

The new hydrolysis technique has also been tried in pilot plant scale, using a 500 gallon reactor for the first re¬ action. Table 2 shows melting point and purity data on the final product of 9 pilot plant runs. After formation of the metallic salt in the first reaction, half of batch 38 was ^' filtered, washed- and hydrolyzed by the standard hydrolysis technique in two ^" splits because of pilot plant equipment size limitations. The filter used in all the runs was a horizontal plate type pressure filter. Filter papers (grade 8A) were used in place of filter cloths to prevent blinding and improve filtration rate. Filtration, washing and nitro- ^• gen blowing through the filter cake were quite satisfactory compared to former runs using filter cloths. Filter cake resistance (i.e. pressure drop through the cake) was about 30 psig (207 kPa).

Since the removal of the filter cake from the filter would have meant the removal of plates and solvent exposure to the operator, hydrolysis in the filter by water washing was planned. This hydrolysis was quite slow. It took almost 12 hours to bring the ph of the effluent down to 9.5 from 12.6. It was desired to bring ph down to 7.0 to assure total absence of alkaline material from the cake. The water washing of the cake was stopped and the cake was air dried by blowing high pressure (90 psig, 620 kPa) air through the cake for one hour. The cake, when taken out of the filter, was quite wet, (38% solids) and had a strong odor of xylenes. As shown in Table- 2, the final product melting point and purity were slightly lower than desired in both splits. The product from the second split had a dark gray color suspected to be due to the impurities left behind by solvent evaporation.

Due to the long time involved and the product quality, a decision was made to try in-situ hydrolysis on the remaining slurry from batch 38. The slurry was cooled to 24°C, and an amount of water equal to the slurry was added to the reactor. The water-organic mixture was then agitated for one hour and was recirculated through the _. filter for one hour.

The filtrate was milky in color due to a suspected emulsion of xylene in water. The pressure drop through the cake was only 10 psig (68.9 kPa) even after one hour. The filtrate was then directed to a filtrate collection (decant) tank. Th filter cake was blown with nitrogen and washed with water. The cake was then dried by air blowing for one hour and was removed from the--filter. The cake was wet (48 solids) but was relatively free of solvent. The product, when dried, had a light green color and had a melting point of 104-.105°C. The polymerization of this product was also quite satisfactor For comparison, the first half of Batch 38 required 32 hour to filter and hydrolyze; whereas, the second half required only 8 hours.

Batches 39 to 45 were finished in two splits, but each spli was in-situ hydrolyzed as the third split from batch 38 above. At the end of the first reaction, half of the slurry (about 1400 lbs or 636Kg) was transferred from the 500 gallon (1895 liters) reactor to a 400 gallon,. (1516 liters) weight tank and was held there with agitation and under a nitrogen blanket. The other half of the slurry was then in-situ hydro lyzed, filtered, water washed and the cake placed in a -tray drier. The slurry, from the 400 gallon tank was then trans- - ferred to the 500 gallon reactor and finished as above. Batc 46 was a half-size batch and was filtered in one split. The filter cake wetness was reduced by properly positioning the plates in the filter to prevent the cake from being erode by incoming slurry and water wash. Batches 41, 42, and 45 ha 74%, 76%, and 64% solids respectively. The filter cake from Batch 42 was vacuum dried in a laboratory vacuum drier and th condensate collected. Material balance showed that the filte cake contained 76% solids, 20.4% water and 3.6% mixed xylenes.

Product quality of all the in-situ hydrolyzed product was excellent, except Batch 43 which had a faint purple cast to i and Batch 44 (split one) which had a dark gray color and had low melting point. It is believed that these two cases were due to operator error in either hydrolysis (not enough water added) or filtration. Both batches were tested for polymeri¬ zation and were found satisfactory.

Table 2

Purity (by

Batch Split Yield % of Melting Point Gas Chroma-

# ^• # Theoretical* +°C tography)

-38 1 60.5 102 - 86.3

2 - 101 85.1

3 104-5 96.7

-39 1 60.5 ^' 104 95.01

2 104 96.38

-40 1 60.5 105 94.96

2 104.5 96.23

-41 1 53.5 105

2 104 95.0

-42 1 51.7 104 .95.8

2 104 95.4

-43 1 50.4 104 96.2

2 105 97.5

-44 1 57.8 102 91.3

2 104 96.2

-45 1 62.3 104 83.9

^' 2 103 81.7

-46 1 50.2 105 96.2

Batch Charge Quantities: Mixed Xylenes 2,160 lbs (980 kg) p- minodiphenyl amine (PADPA) 460 lbs (209 kg)

Methyl methacryl¬ ate 354 lbs (161 kg)

Sodium methoxide ^' 150 lbs ( 68 kg).

^■ ^Theoretical yield based on p-aminodiphenylamineϊ

charered = 460 lbs „ ^{• w} —,— —PADPA kg)

**Expected yield based on 60% conversion of p-aminodiphenyl amine=378 lbs

(171 kg)

÷Melting Point of N- (4-anilinophenyl) methacrylamide standard (recrystallized, 99.9% pure ) = 106° C .

Batches 38, 39 and 40 had identical yield of 60.5%, but Batches 41, 42, 43 and 46 had substantially lower yield, it is believed that lower yields were a result of handling losses and losses in the filtrate due to holes in the filter paper. When the filter cake from the first split of Batch 43 was re¬ moved from the filter, there were quite a few holes in the paper on the top ^' plate, and there was no cake built-up in those areas. A substantial amount of product was found in the filtrate tank. It is believed that paper was accidentally cu while assembling the filter. Batch 46 was reacted with the normal amount _. mixed xylenes but with half the amount of the reactants (to finish off sodium methoxide on hand). The low yield from this batch is attributed to the lower concentratio of reactants. It is believed that a better separation of the hydrolyzed product can be obtained by use of a centrifuge rather than a filter.

Other embodiments of this invention will be apparent to thos skilled in the art from a consideration of this specification or practice of the invention disclosed herein. It is intended that the specification and examples be considered as exemplar only, with the true scope and spirit of the invention'being indicated by the following.claims.

0MPI