US3833650A - Preparation of nitrilotriacetic acid - Google Patents

Abstract

AN IMPROVED PROCESS IS PROVIDED FOR PREPARING NITRILOTRIACETIC ACID (NTA) AND ITS SALTS FROM TRIETHANOLAMINE (TEA) IN THE PRESENCE OF ALKALI METAL HYDROXIDES AND CADMIUM SALTS THAT IS CHARACTERIZED BY EMPLOYMENT OF HIGH TEMPERATURES AND VERY SHORT REACTION TIMES.

Description

United States Patent 3,833,650 PREPARATION OF NITRILOTRIACETIC ACID Heinz Schulze and Edward Thomas Marquis, Austin, Tex., assignors to Jefferson Chemical Company, Inc., Houston, Tex. No Drawing. Filed Aug. 19, 1971, Ser. No. 173,315 Int. Cl. C07c 99/00, 101/20 U.S. Cl. 260531 C 4 Claims ABSTRACT OF THE DISCLOSURE An improved process is provided for preparing nitrilotriacetic acid (NTA) and its salts from triethanolamine (TEA) in the presence of alkali metal hydroxides and cadmium salts that is characterized by employment of high temperatures and very short reaction times.

This invention relates to an improved and novel process for preparing NTA and its salts.

The preparation of tertiary aminocarboxylic acids, such as NTA, from tertiary aminoalkanols, such as TEA, in the presence of alkali metal hydroxides, has heretofore been extensively described. The use of metal catalysts, such as those containing cadmium, copper, nickel, zinc, silver, and the like, has also been suggested for converting TEA to NTA. U.S. Pat. Nos. 2,384,816; 2,384,817 and 2,384,818 are basically representative of such knowledge.

With the recent advent of environment conservation and ecology awareness. renewed interest has developed in the use of NTA salts in detergent formulations. Accordingly, scientists have endeavored to improve upon these known processes for preparing NTA. Several patents have recently issued that demonstrate these efforts. U.S. Pat. Nos. 3,535,373; 3,535,374 and 3,535,375, hereinafter 3,535,373 et al., are exemplary and describe the use of alcohols, parafiins and aryl carbinols as promoters for the preparation of NTA from TEA.

In like manner, U.S. Pat. No. 3,578,709 describes a process for preparing nitrilotriacetic and alkali metal salts from triethanolamine in the presence of alkali metal hydroxides and cadmium catalysts wherein mixtures of alkali metal hydroxides are employed.

These recent methods for preparing NTA have been extolled, such as in the above-identified patents, for their ability to provide improved yields of NTA. In this regard, the basic process art, such as represented by U.S. Pat. Nos. 2,384,816, 2,384,817 and 2,384,818, has been the subject of substantial criticism.

These recently issued patents, however, correctly point out the fact that although U.S. Pat. No. 2,384,817 describes the use of cadmium catalysts for the preparation of NTA from TEA, in the presence of alkali metal hydroxides, no examples were actually presented. They also further emphasize the fact that the only example specific to TEA was that reported as Example V of both U.S. Pat. Nos. 2,384,816 and 2,384,818. Although said Example V did not employ a cadmium catalyst, it did, like the other examples of this patent series, base its reported yields upon the quantity of reaction gas that was developed during the reaction and/ or upon the chelating eifect of the reaction product.

The danger in using chelating effects and gas development as the criteria for yield determination was correctly demonstrated and reported in U.S. Pat. Nos. 3,535,373 et al., particularly in 3,535,375, at column 2, lines 41 to 45. In German Pat. No. 1,809,263 the same conclusion was reached and reported, i.e., yields based on gas development are generally misleading. Our work, as will be hereinafter described, further confirmed these conclusions.

3,833,650 Patented Sept. 3, 1974 In spite of the foregoing facts, if yields of NTA are calculated on the basis of evolved hydrogen as reported in said Example V of U.S. Pat. No. 2,384,816 et al., they still represent a very low yield of NTA. In this regard see U.S. Pat. Nos. 3,578,709; 3,535,373 et al.; and German Pat. No. 1,809,263.

It is also noteworthy that in said Example V of U.S. Pat. Nos. 2,384,816 and 2,384,818 it is reported that the free amino acid could not be readily precipitated from the aqueous solution because of its high solubility in water. Contrary to this disclosure, NTA is quite insoluble in water and readily obtained by acidification of the alkaline solution. This fact emphasizes the belief that little, if any, NTA was actually prepared in said Example V. Further, we endeavored to duplicate said Example V of U.S. Pat. Nos. 2,384,816 and 2,384,818 but failed to produce any NTA even though we employed reaction conditions and reactor facilities that were suggested as preferred embodiments of said patents.

As herein mentioned, there is described in U.S. Pat. No. 2,384,816, at page 2, column 2, line 69 to page 3, column 1, line 3, a copending application, now U.S. Pat. No. 2,3 84,817, which teaches broadly the use of catalysts, including cadmium catalysts, for preparing NTA from TEA.

Although no specific examples directed to the preparation of NTA from TEA were described in U.S. Pat. No. 2,384,817, it was summarily stated therein that this reaction could be carried out at lower temperatures and with greater facility by the aid of certain metals, or their compounds, as catalysts; and that the temperatures in every case are lower than the temperatures required for the same oxidation in the absence of the catalyst, frequently by as much as 50 C. to C. Applying this teaching to Example V of U.S. Pat. No. 2,384,816 the cadmium catalyzed reaction should be conducted at a temperature in the range of about C. to C.

Our efforts, as will be hereinafter described, employing cadmium oxide catalysts according to these teachings resulted in relatively low unacceptable yields of NTA.

Likewise, U.S. Pat. Nos. 3,353,373 et al., describe their unsuccessful efforts (8.8% yield NTA) to duplicate the cadmium catalyzed preparation of NTA from TEA according to the teachings of U.S. Pat. No. 2,384,817. In like manner, reference to Example IV of U.S. Pat. No. 3,578,709 shows that a yield of only 11.6% NTA was obtained using a cadmium catalyst in such a process.

It is evident from such recent art, as corroborated by our work, that the obtainable yields of NTA are unacceptably poor when the teachings of U.S. Pat. Nos. 2,384,816; 2,384,817 and 2,384,818 are followed.

Although recent improvements in the art of NTA production have been made, such as represented by the art herein described, it is also apparent that although increased yields of NTA have, in some instances, been obtained, the described processes do not provide a comparatively simple process such as one that could facilitate a commercially feasible operation.

For example, even with the use of the alkanol, paraffin and aryl carbinol promoters of U.S. Pat. No. 3,535,373 et al., the production of NTA from TEA in the presence of cadmium catalysts in alkaline solution is unduly long. It is described therein as an extended reaction requiring about 8 to 70 hours.

Likewise, when the reaction is conducted in the presence of mixtures of alkali metal hydroxides, according to the teachings of U.S. Pat. No. 3,578,709, reaction times of six hours or more are employed.

It is, therefore, a clear fact that artisans in this field still believe that reaction times on the order of several hours to several days are necessary for the conversion of TEA to ice NTA in the presence of alkali metal hydroxides and cadmium catalysts.

Another prevalent fact in evidence throughout the pertinent art, and of general importance to this background discussion, is the belief that high reaction temperatures are to be avoided because of the thermal instability of the aminocarboxylic acids, and their salts.

For example, in U.S. Pat. No. 2,384,816 the practitioner is cautioned against the use of high temperatures when oxidizing aminoalcohols by heating in the presence of alkali metal hydroxides because of the general thermal instability of aminocarboxylic acids and their salts under such conditions. It is noted that Example V of U.S. Pat. No. 2,384,- 816 employs a temperature of 210 C. to 220 C. and, as hereinbefore mentioned, temperatures of about 50 C. to 75 C. lower than this are taught when a metal catalyzed reaction according to U.S. Pat. No. 2,384,817 is employed.

Adherence to this fact is also evidenced by Dwyer & Mellor, Chelating Agents and Metal Chelates, Academic Press, New York, 1964, where at page 287 it is stated:

The main difficulty in the reaction (of aminoalcohols to yield aminocarboxylic acids) is in preventing the oxidation of the strongly reactive amino groups which are susceptible to attack by alkalis and oxidizing agents The avoidance of high temperatures is likewise taught and practiced in U.S. Pat. No. 3,578,709 and in U.S. Pat. Nos. 3,535,373 et a1.

For example, although U.S. Pat. No. 3,535,373 et a1. generally state that temperatures of about 150 C. to 260 C. can be employed, they state that temperatures from 190 C. to 240 C. are preferred. Reference to any of their examples clearly shows that temperatures higher than 240 C. were strictly avoided.

In like fashion, broad temperatures of about 150 C. to 300 C. are generally described in U.S. Pat. No. 3,57 8,709, but all of the representative examples were conducted essentially within the preferred temperature ranges described above, i.e., 190 C. to 240 C.

In summary, the prior art recognizes and practices, without known exception, two important beliefs:

(a) The avoidance of high temperatures such as those greater than about 240 C., and

(b) The necessity of long reaction times on the order of hours or more.

In spite of the above knowledge and beliefs, we have surprisingly discovered the NTA, and its salts, can be prepared in high yields from TEA in the presence of alkali metal hydroxides and cadmium salts at high temperatures and at reaction times, seemingly impossible according to prior art standards.

Accordingly, an important aspect of our process was the discovery that NTA salts were unexpectedly stable at high temperature even in the presence of aqueous alkali metal hydroxides and cadmium salts.

Further, because of our discovery NTA and its salts can now be prepared at reaction conditions that enable an economically feasible application to a commercial process.

Therefore, in accordance with our process, NTA, and its salts, are prepared by heating TEA in the presence of an alkali metal hydroxide and a cadmium salt at a temperature in the range from about 250 C. to 375 C., preferably from about 300 C. to 350 C., for a time in the range from about 1 to 45 minutes, preferably less than 30 minutes and most preferably from about 1 to 15 minutes.

Cadmium salts, such as the acetate, propionate, butyrate, oxide, chloride, sulfate, admixtures thereof, and the like, are representative catalysts and can be suitably employed in amounts efiectively determined by the skilled artisan. Generally, an efiective amount is within the practical range of about .4 to grams of cadmium salt per mol of TEA employed.

The above-described reaction is conducted in the presence of alkali metal hydroxides, admixtures thereof, such as sodium hydroxide, potassium hydroxide, and the like, which are generally employed in amounts to provide a stoichiometric quantity relative to the triethanolamine, although excess alkali metal hydroxides are usually provided.

Sufiicient water is employed to essentially maintain the components in the reaction medium essentially in solution and sufficient pressure is usually employed to essentially maintain the water in the liquid phase.

The NTA salts produced by our process are candidates for large scale use in detergent formulations. The free acid can, of course, be obtained by acidification, and be used in other reactions characteristic of carboxylic acids. Accordingly, other water soluble salts of nitrilotriacetic acid can be prepared, if desired, such as with ammonia or with substituted ammonia radicals, such as triethanolamine, and the like. The trisodium salt of nitrilotriacetic acid is the generally preferred water soluble salt.

To illustrate the foregoing discussion and description and not to be interpreted as a limitation on the scope thereof, or on the materials herein employed, the following examples are presented.

EXAMPLE I In a copper-lined autoclave were placed triethanolamine (119 grams, 0.8 mol), potassium hydroxide (200 grams, 3.12 mols), and water (32 grams). The autoclave WaS purged with nitrogen and the mixture rocked 16 hours at 210 C. to 220 C. The final pressure in the autoclave was about 950 p.s.i. After cooling water was added to the reaction product until the total weight of the mixture was 826 grams, part of the solution (785 grams) was acidified with 350 ml. concentrated hydrochloric acid until a pH meter with glass electrodes indicated a pH of 0.35. No nitrilotriacetic acid was precipitated from the solution.

The foregoing example represents a concentrated eifort to repeat Example V of U.S. Pat No. 2,384,816. Although we even employed a copper-lined vessel to further facilitate the reaction, no NTA was produced.

EXAMPLE H In a similar run, we placed in a l-liter copper-clad stirred autoclave, provided with a Monel liner, 0.8 mol triethanolamine, 4.0 mols sodium hydroxide and 164 grams of water. The autoclave was purged with nitrogen and heated to 200 C. with stirring. At this temperature the stirrer was turned off and turned on about every 5 minutes for about 1 or 2 seconds to read and adjust the temperature. When the reaction temperature reached about 250 C. the heater was turned off. By intermittent cooling, a temperature of 250 C. to 270 C. was maintained in the autoclave for 40 minutes. At this time the pressure in the reactor had risen to 2,400 p.s.i. After cooling, the reaction product was removed and the autoclave rinsed with water. The mixture of reaction product and wash liquid was filtered. The residue was mainly sodium oxalate as indicated by infrared analysis. Acidification of the filtrate as in Example I failed to produce any NTA.

The foregoing example effectively demonstrates that gas development during the reaction does not necessarily indicate the formation of NTA.

EXAMPLE III In a copper-lined autoclave were placed 0.8 mol triethanolamine, 3.2 mols 90% potassium hydroxide, 32 grams water and 6.0 grams (.047 mol) cadmium oxide. The autoclave was purged with nitrogen and rocked for 16 hours at 170 C. to 175 C. The final pressure in the autoclave was 2,200 p.s.i. Water was added to the reaction product and the turbid solution (800 grams) filtered. The filtrate was acidified as in Example I, heated to C. and cooled. Precipitated NTA was filtered and washed with ice water until chloride-free, then washed with methanol and dried at C. The NTA yield was about 53%.

EXAMPLE IV In a 1-liter stirred copper-clad autoclave were placed .8 mol triethanolamine, 2.62 mols sodium hydroxide, 148.5 grams water and 6.0 grams cadmium oxide. The autoclave was purged with nitrogen and the mixture heated three hours at 220 C. to 230 C. Maximum pressure was about 3,300 p.s.i. The reaction product was worked up as in the preceding Example III and the NTA yield was about 46%.

The foregoing Examples III and IV, although demonstrating preparation of NTA, are typical prior art processes wherein the reaction times are long.

EXAMPLE V In a 1-liter stirred copper-clad autoclave were placed 4.0 mols sodium hydroxide, 180 ml. water, 6.4 grams cadmium oxide. After purging with nitrogen the mixture was heated to 330 C. with stirring and 0.8 mol triethanolamine was added over 95 seconds while a pressure of 1,200 p.s.i. was maintained on the reaction mixture. At the end of the triethanolamine addition period the temperature of the mixture had dropped to 286 C. and a vigorous gas development occurred. Within another two and one-half minutes the temperature was increased to 320 C., gas development ceased and the mixture was diluted with 300 ml. water, cooled and worked up as hereinbefore described. Although only about four minutes reaction time was involved, the recovered NTA yield was surprisingly about 84%.

EXAMPLE VI Added to a Hastelloy B autoclave were 0.8 mol triethanolamine, 148.5 grams of water, 3.0 mols sodium hydroxide and 6 grams of cadmium oxide. Under hydrogen the mixture was stirred and heated for 3 minutes at 260 C. The autoclave was cooled, pressure vented, and another 6.0 grams cadmium oxide added and the mixture reheated under hydrogen at 260 C. for another 3 minutes. The maximum pressure developed during the first heating period was 3,600 p.s.i.g. and during the second period 460 p.s.i.g. The reaction mixture was diluted with water and worked up as hereinbefore stated. The yield of NTA recovered was 82.4%.

EXAMPLE VII A mixture of 119 grams (0.8 mol) triethanolamine, 123.0 mols sodium hydroxide, 148.5 grams water and 6.0 grams cadmium oxide were heated under hydrogen with stirring in a Hastelloy B autoclave at 270 C. for 3 minutes. The mixture was cooled and worked up as in the previous example, and the NTA yield was 77% EXAMPLE VIII In a similar fashion, to a 1-liter stirred copper-clad autoclave were placed 0.8 mol triethanolamine, 4.0 mols sodium hydroxide, 164 grams water and 3.2 grams cadmium oxide. The autoclave was purged with nitrogen and the mixture was heated 20 minutes at 250 C. to 270 C. with intermittent stirring. The reaction mixture was cooled and worked up as in the preceding examples. The NTA yield was 72.5%.

The foregoing Examples V-VIII are demonstrative of our invention and represent seemingly impossible achievements according to prior art standards and beliefs.

EXAMPLE IX To a 1-liter stirred copper-clad autoclave were placed 0.8 mol trisodium nitrilotriacetate monohydrate, 1.6 mols sodium hydroxide, grams water and 3.2 grams cadmium oxide. The autoclave was purged with nitrogen and heated 15 minutes at 320 C. to 330 C. The reaction product was then diluted with water, filtered, acidified with concentrated hydrochloric acid and worked up as in the previous examples. The yield of NTA recovered was 81%.

Contrary to earlier beliefs, the foregoing example effectively demonstrates the unexpected thermal stability of NTA salts in the presence of alkali metal hydroxides. It should be further noted that when the reaction mixture of Example IX was stirred for an additional 45 minutes the yield of recovered NTA was still 65.6%.

The preceding examples can be repeated with similar success by substituting the generically and specifically described reactants and conditions of this invention for those employed in the examples. As will be evident to those skilled in the art, various modifications of this invention can be made or followed in light of the discussion and disclosure herein set forth without departing from the spirit or the scope of our invention.

We claim:

1. In the aqueous process for preparing nitrilotriacetic acid salts by heating triethanolamine in the presence of excess or stoichiometric quantities of alkali metal hydroxides relative to the triethanolamine and cadmium salts in the liquid phase, the improvement comprising conducting said heating at a temperature within the range of about 250 C. to 375 C. for a time within the range of about 1 to 20 minutes.

2. The process according to Claim 1 wherein said heating is conducted at a temperature Within the range of about 300 C. to 350 C.

3. The process according to Claim 2 wherein said heating is conducted for a time in the range of 1 to 15 minutes.

4. The process according to Claim 1 wherein said heating is onducted for a time in the range of 1 to 15 minutes.

References Cited UNITED STATES PATENTS 3,708,533 1/1973 Bechara et al 260531 C VIVIAN GARNER, Primary Examiner