WO2008067112A2

WO2008067112A2 - Oxidation of alkylarenes in expanded liquid reaction mixture

Info

Publication number: WO2008067112A2
Application number: PCT/US2007/083425
Authority: WO
Inventors: Peter D. Metelski; Ian C. Zaenger
Original assignee: Bp Corporation North America Inc.
Priority date: 2006-11-14
Filing date: 2007-11-02
Publication date: 2008-06-05
Also published as: WO2008067112A3

Abstract

A C8-C2O alkylarene is converted to an arene carboxylic acid in an oxidation reactor by reacting the alkylarene with oxygen under oxidation conditions in the presence of an MC catalyst or bromine-free oxidation catalyst in a volumetrically expanded liquid reaction mixture in which a preferred solvent comprises a homogeneous phase of a monocarboxylic acid and carbon dioxide under elevated pressure, carbon dioxide partial pressure and temperature conditions.

Description

Oxidation Of Alkylarenes In Expanded Liquid Reaction Mixture

Field of the Invention

This invention relates to catalytic oxidation of alkylarenes in an expanded liquid reaction mixture and more particularly relates to oxidation of methyl arenes such as xylenes using a cobalt / manganese / bromine catalyst system in a liquid reaction mixture in which a solvent comprises a homogeneous phase of acetic acid and carbon dioxide under oxidation conditions and a carbon dioxide partial pressure effective to volumetrically expand the liquid reaction mixture.

Background of the Invention Oxidation of lower alkyl arenes to corresponding carboxylic acids is well known and is used to produce many commercially-important chemical products. For example, para-xylene, meta-xylene, pseudocumene, and dimethylnaphthalene are oxidized to form terephthalic acid (TA) isophthalic, (IPA), trimellitic acid (TMAc), and naphthalene dicarboxylic acid (NDA), respectfully. Each of these products is an important chemical commodity. As an example, TA is polymerized with ethylene glycol to form polyethylene terephthalate (PET), which is commonly referred to as "polyester", and is used to produce fibers, films, containers such as bottles and other packaging products.

The current commercial oxidation technology to form these products uses a cobalt/manganese/bromine catalyst system in an acetic acid solvent initially described in U.S. Patent 2,833,816, assigned to the Mid-Century Corporation, and thus called the "MC" process. A description of the MC process and MC catalysts is provided in W. Partenheimer, "Methodology and scope of metal/bromide autoxidation of hydrocarbons," Catalyst Today, vol. 23 (1995), pp. 69-158, incorporated by reference herein. In a typical MC oxidation process, a methyl-substituted aromatic hydrocarbon in acetic acid solution is oxidized using a cobalt/manganese/bromine catalyst system in the liquid phase by adding an oxygen-containing gas to the solution. Typical process conditions include operating under pressure of up to about 35 bar at temperatures of about 100 to 250 ⁰C. Current MC oxidation technology is complicated by several factors. Among these is a minimum ratio of acetic acid solvent to oxidizable alkylaromatic suitable for maintaining a liquid reaction mixture and dispersing reactants and catalyst within the liquid mixture for practical production rates. Other factors include the limited solubility of oxygen in the solvent and solvent loss due to burning (oxidation) of the solvent.

Another aspect complicating MC oxidations is the formation of by-products, some of which can cause off-color in the product and may inhibit polymerization to polyester. Production of terephthalic acid suitable for direct polymerization with a glycol to form a polyester typically requires a separate purification step to form purified terephthalic acid (PTA). Such a purification step removes certain partial oxidation by-products such as 4-carboxybenzaldehyde (4-CBA) and para-toluic acid.

An oxidation process which minimizes such impurities represents a significant process improvement.

In an oxidation process, an alkylarene (e.g., an alkylaromatic hydrocarbon such as a xylene) is oxidized to a corresponding carboxylic acid with concurrent formation of water. Such water formed in the oxidation reaction typically is removed, for example, using large distillation towers based on the relative boiling points of water and solvent (e.g., acetic acid). A process in which the water/solvent separation is easier represents a substantial improvement.

Recently, there have been proposals to practice so-called "green chemistry" by using carbon dioxide in a supercritical state (scCO₂) as solvent in various organic chemical conversion reactions that conventionally use organic solvents. However for many reactions, maintaining a system in a supercritical state requires extraordinarily high pressures in the range of hundreds of bars, which makes use of ScCO₂ processes commercially impractical. The use of "CO₂-expanded" organic solvents has also been described for certain chemical reactions. Generally see, M. Wing et al., "CO₂-Expanded Solvents: Unique and Versatile Media for Performing Homogeneous Catalytic Oxidations", J. Am. Chem. Soc, Vol. 124 (2002) pp. 2513- 2517, and U.S. Patent 6,448,454, all incorporated by reference herein.

U.S. Patents 6,448,454 and 6,740,785 describe use of carbon dioxide- expanded liquid reaction mixtures that include certain organic solvents for oxidation reactions but do not describe oxidation of alkyl arenes using MC catalysts in an expanded reaction mixture in which an acetic acid solvent is present.

U.S. Patents 6,180,822 and 6,194,607 describe production of aromatic carboxylic acid products using an MC-type catalyst system in which the feed gas contains oxygen and carbon dioxide. However, only a relatively low partial pressure of carbon dioxide is disclosed, and data presented show a commercially unacceptable conversion of aromatic feed to carboxylic acid product.

Summary of the Invention We have found that operating in an expanded liquid reaction mixture for alkylarene oxidation provides conditions in which one or more of superior product, high total conversion rates, reduced solvent monocarboxylic acid consumption and reduced burning can be attained.

In one embodiment of the invention, a C₈-C2o alkylarene is converted to an arene carboxylic acid in an oxidation reactor by reacting the alkylarene with oxygen in the presence of an MC catalyst in a volumetrically expanded liquid reaction mixture, in which a solvent comprises a homogeneous phase of a monocarboxylic acid and carbon dioxide, under oxidation conditions including a carbon dioxide partial pressure such that the ratio of the volume of the volumetrically expanded liquid reaction mixture under such conditions (V_e) to the volume of the liquid reaction mixture cooled and depressurized to normal conditions (V₀) is more than 1.25.

In another embodiment, the invention provides a process for the production of arene carboxylic acids, and especially terephthalic acid, comprising reacting alkyl arene, one or more partially oxidized derivative thereof or combinations thereof, and especially para-xylene when producing terephthalic acid, in an oxidation reactor with oxygen in the presence of an oxidation catalyst comprising a heavy metal component and bromine in a liquid reaction body in which a solvent for the alkyl arene, partially oxidized derivative or combination and the oxidation catalyst comprises a homogeneous liquid phase of acetic acid, carbon dioxide and up to 35 % water, by weight of the acetic acid, under oxidation conditions with a carbon dioxide partial pressure of at least 30 bar. In another embodiment, a process for the production of arene carboxylic acids, and especially terephthalic acid, comprises reacting alkyl arene, one or more partially oxidized derivative thereof or combinations thereof, and especially para-xylene when producing terephthalic acid, in an oxidation reactor with oxygen in the presence of a catalyst composition which is active in the absence of reactive bromine for oxidation of the alkyl arene or its derivatives or combinations thereof to arene carboxylic acids wherein the reaction is conducted in a liquid reaction body that comprises water and, preferably, in which a solvent for the alkyl arene, partially oxidized derivative or combination comprises a homogeneous liquid phase of acetic acid, carbon dioxide and up to 35 % water, by weight of the acetic acid, under oxidation conditions with a carbon dioxide partial pressure of at least 30 bar.

Another embodiment of the invention provides a volumetrically expanded solvent composition comprising a homogeneous phase of a monocarboxylic acid and carbon dioxide under elevated pressure and temperature conditions, wherein the ratio of the volume of the expanded solvent (V_e) under such conditions to the volume of monocarboxylic acid under normal conditions (V₀) is more than 1.25. Such a composition is suitable for use as a reaction solvent for liquid phase oxidation of alkyl arenes to their corresponding arene carboxylic acids.

Brief Description of the Drawing

Figure 1 is a graph showing the amount of an 80/20 carbon dioxide/oxygen gas mixture as a function of pressure with and without the presence of acetic acid.

Description of the Invention

Our invention in some of its embodiments provides a process to oxidize alkylarenes to corresponding carboxylic acids using MC catalyst systems in a volumetrically expanded liquid reaction body in which a solvent for the reaction comprises a homogeneous phase of monocarboxylic acid and carbon dioxide. The practice of this process lessens the use of organic-based solvents. Reduction of organic solvents lowers potential organic solvent emissions and organic solvent losses due to combustion, and results in a more economically efficient process.

As described herein, an MC oxidation process uses an oxidation metal component (e.g. Co, Mn or, preferably, a combination thereof or including both) in combination with a bromine-containing component in a monocarboxylic acid (e.g. acetic acid) liquid solvent system. As explained in the Partenheimer article, the solvent plays an important role in the catalytic reaction. However, some of solvent

(e.g., acetic acid) also typically is consumed by oxidation, which results in a less efficient overall process. Thus, there is an incentive to lessen the amount of solvent used in the process. In the current invention, the oxidation is conducted in a liquid phase reaction mixture in which a solvent for the reaction is a liquid homogeneous phase comprising a combination of a monocarboxylic acid, and especially acetic acid, carbon dioxide and up to 35 wt% (by weight of the monocarboxylic acid) water and under pressure and temperature conditions such that the volume of the liquid reaction mixture is expanded over that of the liquid mixture cooled and depressurized to normal pressure and temperature conditions. For purposes hereof, normal pressure and temperature conditions are 1 bar pressure and 20⁰C temperature. The "expanded" liquid reaction mixture is distinct from merely dissolving minor amounts of a gas (e.g., CO₂) in a liquid in that the volume of the expanded liquid mixture is appreciably greater than the volume of the liquid at atmospheric pressure.

A characteristic of such an expanded phase useful in this invention is a measurement of the expansion ratio of the volume of the expanded liquid reaction mixture (Ve) to the volume of the liquid mixture cooled and depressurized to normal conditions (Vo). The expansion ratio as so determined is designated herein as Ve/Vo. As used in this invention, a suitable expanded liquid reaction mixture has a Ve/Vo ratio typically above 1.25, preferably above 2 and more preferably above 2.5. There is a practical upper limit to the extent of expansion based on the pressure used and the mole ratio of carbon dioxide to solvent used in a particular system. Typically, for a CO₂/acetic acid system with 100% CO₂ maintained over liquid acetic acid under pressure of 60 bar, the extent of expansion ratio may be up to about 5, and usually is below about 4. The volume ratio VeΛ/o may be determined from Vo and Ve. Vo is measured, especially in continuous systems, by withdrawing as a sample a known volume of the liquid reaction mixture (which may also include solids such as precipitated product arene carboxylic acid) from the reactor under operating temperature and pressure, preferably with removal of entrained gas bubbles, depressurizing and cooling the sample to normal conditions, filtering any solids from the sample, and measuring the volume of the resulting liquid as Vo. Ve is determined by reducing the known volume of the sample withdrawn under operating conditions by the volume of any filtered solids. As illustrated in the examples herein, values for Ve and Vo also may be determined by volume calculations from, or correlations to, levels of liquid reaction mixture in a reactor under operating conditions and under normal conditions, respectively. Another characteristic of a suitable expanded liquid reaction mixture is the pressure under which the system is maintained. The carbon dioxide partial pressure should be sufficient to maintain the expanded reaction mixture at a desired expansion level and temperature. For a CO₂/acetic acid system, which may also include up to about 35% water by weight of the acetic acid, such a CO₂ partial pressure is above 30 bar, preferably above 35 bar, and typically above 40 bar. Such pressure may be up to 100 bar, typically below 80 bar, and usually is between 55 and 70 bar. A maximum practical pressure is that at which further expansion of the liquid reaction body is negligible. Total pressure is determined by the amount of gas in the closed system and the temperature of the system, and generally is at least 35 bar, preferably at least 40 bar, but not greater than 120 bar and preferably no more than 95 bar. A preferred pressure range is 65 to 85 bar. Partial pressure of carbon dioxide is determined (using the Ideal Gas Law) by the mole fraction of CO₂ in the total gas. Operating temperature is selected to maximize catalytic activity for optimum yield and solubility of the reactants and catalysts in the liquid reaction body. In a typical system useful in the practice of this invention, a reaction system is maintained above about 50⁰C, usually above 100°C and preferably above 150⁰C. Practical operating temperatures may range up to 300⁰C and usually are below 250 ⁰C. A suitable continuous oxidation process of this invention may be operated in stages in which process conditions such as temperature, pressure and catalyst composition, may be different in each stage. Typically, initial stages operated at a higher temperature than a later stage can be beneficial for increasing yields by conversion of partially oxidized alkyl arene derivatives to desired arene carboxylic acid products. The useful range of mole ratios of carbon dioxide gas to solvent typically is determined by the CO₂ partial pressure and system temperature used. In order to maximize the effect of CO₂ for expansion of the liquid body, presence of inert gases such as nitrogen should be minimized in the system. Thus, use of an enriched oxygen stream as an oxygen source for the process is preferred over use of air as an oxygen source.

The expanded liquid reaction body in a typical system will contain solvent (e.g., acetic acid/water), expansion gas (e.g., CO₂), dissolved catalyst (e.g., Co/Mn/Br), oxidizable substituted arene (e.g., para-xylene, its partially oxidized derivatives or combinations thereof (including combinations of derivatives or combinations of para-xylene and one or more derivatives)), oxidant (e.g., O₂). Other components commonly present in the liquid reaction mixture, especially in continuous processes, will include arene carboxylic acid (e.g., terephthalic acid when oxidizable substituted arene includes para-xylene or its partially oxidized derivatives), which may be present in solution, as suspended solids and, most commonly, both; intermediate oxidation products of the oxidizable arene, again in solution, as solids or in both forms (e.g., 4-carboxybenxaldehyde and para-toluic acid when oxidizable substituted arene is para-xylene); water, both in liquid form and as entrained gas, generated as by-product of the oxidation; and by-products resulting from oxidation of the solvent monocarboxylic acid and reaction of such by-products (e.g., methanol and methyl acetate when the solvent monocarboxylic acid is acetic acid), which can be present in liquid form, entrained gases or both.

The operation of an oxidation process of this invention is enhanced due to the increased solubility of oxygen in the expanded liquid reaction body in contrast to a conventional reaction system with acetic acid solvent. The increased solubility of oxygen in the expanded solvent system may be 100 times or more than the solubility in acetic acid. Increased mobility of oxygen in the expanded liquid reaction body may decrease the need for high agitation in the oxidation reactor. Further, reduced density of the expanded liquid phase as compared to conventional acetic acid solvent-based systems can reduce power inputs needed for agitation at a given rate, other things being equal.

In conventional MC oxidation processes, typically there is about three to five times the amount by weight of acetic acid solvent to oxidizable substituted arene. The invented process permits operation such that the weight ratio of acetic acid or other solvent monocarboxylic acid to arene (e.g., para-xylene) may be less than 2 and preferably 1 or less, which is more efficient use of organic solvent components.

Suitable catalytic materials for use in this invention are based on "MC" catalysts. These catalysts comprise a metal component, which has autoxidative catalytic activity, together with a bromine-containing component. Suitable metal components are typically selected from Groups 4-12 (IUPAC nomenclature) and the lanthanides, which alone or in combination with other metal components demonstrate autoxidative activity under the process conditions used. Preferably, a suitable metal catalyst component contains a Group 7-10 metal species, and more preferably cobalt, manganese, or nickel, and also may contain one or more additional metals.

The preferred metal catalyst component contains cobalt and most preferably is a combination of cobalt and manganese in autoxidative catalytically active oxidation states. Typically, these metal species are used as salts such as acetates. In a typical manganese/cobalt containing metal component, the Mn/Co atom ratio is at least 0.3, preferably above 0.5 and more preferably more than 0.7 and may range up to 15, preferably below 10, and typically below 3. Because cobalt is more expensive than manganese, there is an incentive to minimize the amount of cobalt used in the metal component.

Additional metal species other than manganese and cobalt may be added to the metal component. Examples of additional metal species which may be present in the metal catalyst component include zirconium, titanium, hafnium, cerium, calcium, silicon, nickel, niobium, iron, rhodium, rhenium copper, zinc, and cadmium. Typically these metal species may be added up to about 10 atom percent of total metal component.

Sources of the bromine component of MC catalysts used according to the invented process include molecular bromine (Br₂), ionic bromide (such as HBr, NaBr, KBr, NH₄ Br, etc.), and organic bromides that can provide bromide ions at the operating temperature of the oxidation (e.g., bromoanthracene, dibromoanthracene, bromobenzenes, benzylbromide, mono- and di-bromoacetic acid, bromoacetyl bromide, tetrabromoethane, ethylene-dibromide, and the like). The total elemental bromine in molecular bromine and ionic bromide is used to determine the bromine-to- total metal component (such as cobalt and manganese) atom ratio. Bromine ions released from the organic bromides at the oxidation conditions may be determined by known analytical means. For example, tetrabromoethane has been found to yield about three effective gram atoms of elemental bromine per gram mole, at operating temperatures of 170 to 225°C. The atom ratio of bromine component to total metal component used in a suitable MC catalyst typically is at least 0.1 , and preferably is at least 0.2, and may range up to 1.5 and preferably is below 1.0. A typically preferable range is between 0.3 and 0.8. A typical Co/Mn/Br atom ratio is 1/1/1. Typically, each of the metal components (e.g., cobalt and manganese components) may be provided in any ionic or combined forms which provide soluble forms in the reaction solvent. For example, cobalt and/or manganese carbonate, acetate, acetate tetrahydrate, and/or bromide can be used. Preferably cobalt and manganese acetate tetrahydrates are used. The preferred bromine source is hydrobromic acid.

MC catalysts are preferred catalysts for use in the invented process; however, catalyst compositions that are active in the absence of bromine for oxidation of alkyl arenes and their partially oxidized derivatives with selectivity to arene carboxylic acids may be employed instead of MC catalysts according to embodiments of the invention. Bromine-free compositions having suitable activity and selectivity may comprise palladium, platinum, vanadium, titanium, tin, antimony, bismuth, molybdenum, and preferably combinations of two or more thereof. Homogeneous forms of such catalyst that are soluble in the expanded liquid reaction mixture, as well as heterogeneous forms and combinations thereof, can be used. For use in homogeneous form, acetate, acetyl acetonate and other soluble salts of the catalyst metals are conveniently used. When heterogeneous forms of catalyst or catalyst components are employed, they may be used in unsupported form or they may be supported on a particulate support material such as particulate carbons, high strength, acid stable silicon carbides and various metal oxides such as silicas, aluminas, titanias and zirconia.

In one embodiment, such a bromine-free catalyst composition comprises palladium, an element of Group 15 of the Periodic Table of the Elements selected from antimony, bismuth and combinations thereof, and at least one metal or metalloid of Group 4, 5, 6 or 14 of the Periodic Table of the Elements. More specific examples of such catalysts are bromine-free compositions comprising palladium, antimony and one or more of tin, vanadium and molybdenum. In another embodiment, a bromine-free catalyst composition comprises (A) at least one of platinum, palladium, titanium or vanadium and (B) at least one Group 5, 6, 14 or 15 metal or metalloid, provided that (A) and (B) do not each consist only of vanadium. More specific examples of such compositions include combinations of vanadium and titanium and of platinum, titanium and vanadium. Atom ratios of metals and metalloids in such compositions generally range from about 1 :1000 to about 1000:1 and more preferably about 1 :10 to about 10:1. Groups of the Periodic Table of the Elements referred to in this paragraph correspond to "New Notation" designations according to the Periodic Table of the Elements as found, for example, in the Handbook of Chemistry and Physics, 78th Edition, CRC Press, 1997. The volumetrically expanded liquid reaction mixture in the process of this invention includes water in the liquid mixture and preferably a solvent system comprising an aqueous organic component phase, which is able to solubilize the alkyl arene feed material and homogenous forms of catalyst and form a carbon dioxide-expanded liquid oxidation reaction mixture, including the solubilized components under the process conditions used. The typical organic solvent component is a C₂-C₈ monocarboxylic acid such as acetic acid, propionic acid, n- butyric acid, benzoic acid, isobutyric acid, n-valeric acid, trimethylacetic acid and caproic acid or a combination thereof. Preferably, the monocarboxylic acid solvent component is acetic acid. Because water is formed during the oxidation reaction, water is present in the solvent mixture. Water must be removed from the reaction mixture during a continuous oxidation process. Typically, this is accomplished by removing an oxidation off-gas or vent stream containing water and monocarboxylic acid solvent (e.g., acetic acid) that vaporize from the liquid reaction mixture due to the exothermic oxidation reaction and carbon dioxide. The carbon dioxide is easily separated due to the large difference in boiling points between carbon dioxide and either water or acetic acid. Preferably, a portion of separated CO₂ is recycled back to the reactor in this process. Because the amount of water remaining with the acetic acid from the vent stream is proportionally larger than would occur in a conventional process (which does not use an expanded liquid reaction mixture), water separation to a desired level is easier. Thus, there is less need for large dehydration towers characteristic of the conventional process. Typically, the level of liquid water in the reaction mixture is maintained at a level of at least 1 wt% (typically above 3 wt% and preferably at least 5 wt%) up to 35 wt% or more (preferably less than 25 wt%) based on the amount of monocarboxylic acid solvent component. A typical reaction mixture contains 10 to 20 wt% water based on the amount of monocarboxylic acid (e.g., acetic acid) solvent component.

The process of this invention converts Cs-C_2O alkylarenes to corresponding arene carboxylic acids by an oxidation chemical reaction. Typically, a suitable alkylarene useful in this process contains at least one aromatic ring which is substituted with at least two groups which are capable of being oxidized to carboxylic acid groups. The most typical of such groups is a

alkyl group (most preferably methyl) although partially-oxidized groups such as acyls may be used. For the purposes of this invention, an alkylarene includes arenes substituted with one or more partially-oxidized alkyl groups, which may be the same or different, in addition to or instead of alkyl substituent groups, such as acylalkylarenes, acylarenes, acylarene carboxylic acids, hydroxyalkylarenes, hydroxyl arene carboxylic acids, alkyl arene carboxylates and combinations thereof. The most typical alkylarene used in this invention contains one or two aromatic rings such as a benzene and naphthalene. Typical substituted benzene derivatives include the xylenes (e.g., o- xylene, m-xylene, and p-xylene), trimethylbenzenes (e.g., pseudocumene and 1 ,3,5- trimethyl benzene), and the dimethylnaphthalenes (e.g. 2,6-dimethylnaphthalene).

In commercial practice, MC oxidation technology is most used to produce terephthalic acid from p-xylene, isophthalic acid from m-xylene, 2,6-naphthalene dicarboxylic acid from 2,6-dimethylnaphthalene, and trimellitic acid (which may be dehydrated to trimellitic anhydride (TMA)) from pseudocumene (1 ,2,4- trimethylbenzene).

Typical uses of terephthalic acid include use as a monomer component for making polyesters such as polyethylene terephthalate (PET) and polybutene terephthalate (PBT). PET is used primarily in fiber, film, container and other packaging applications. Isophthalic acid is often used as a comonomer with terephthalic acid in polyesters for such applications and as a component in polyester resin compositions useful in coatings, inks, reinforced plastics, and packaging applications. 2,6-Naphthalene dicarboxylic acid and its derivatives are useful as monomer components in high performance polyesters used in film applications.

Trimellitic acid and its derivates (e.g., TMA) are used as components in plasticizers and coatings applications.

In an embodiment of an oxidation process according to this invention, a homogeneous liquid solution of CU₂-expanded acetic acid and water is formed in a suitable reactor vessel into which is added an MC catalyst system including a metals component (e.g., Co/Mn) and a bromine-containing component or a suitable bromine-free catalyst composition. The reaction mixture includes an oxidizable alkylarene (e.g., p-xylene) and molecular oxygen is used as the oxidizing agent.

The process of this invention may be conducted either in a batch, continuous, or semicontinuous mode. Preferably, the process is conducted in a continuous mode in which oxidizable alkylarene, catalyst, solvent, and oxygen-containing gas are continuously introduced into the oxidation reactor maintained under temperature and total pressure as described above and carbon dioxide partial pressure effective to volumetrically expand the liquid reaction body, and a product stream containing the resulting crude acid oxidation product and catalyst components dissolved in the solvent is withdrawn from the reactor.

The reactor vessels and associated equipment used for the oxidation reaction should be constructed of suitably inert materials to withstand the corrosivity of the reaction medium at the process temperatures used. Typically useful reactors and associated piping and equipment subject to corrosive conditions are constructed from titanium or titanium-clad materials. Typically, the reactor vessel contains an agitation devise used to promote contact of oxygen with the oxidizable substrate and catalyst.

The oxidizing agent used in the process of this invention typically is gaseous molecular oxygen (O₂). Although air is useful as a source of oxygen, preferably the oxygen source for the invented process is oxygen enriched such that at least 50 (preferably 75) mole percent of the gas is oxygen. The most preferred oxidizing gas contains at least 95 mole percent (and more preferably at least 99 mole percent) molecular oxygen. In a continuous process, oxygen is fed to the reaction at a rate sufficient to substantially react with the oxidizable substituents of the alkyl arene substrate (e.g. methyl groups). In a batch reaction system, oxygen is loaded batchwise to the reactor. In order to avoid the formation of explosive mixtures, the oxygen-containing gas fed to the reactor should provide an exhaust gas-vapor mixture containing no more than about 8 (preferably 0.5 to 4) volume percent oxygen (measured on a solvent-free basis). Typically, 1.5 to 2.8 moles of oxygen are fed per mole of alkyl arene starting material. The oxidation reaction temperature suitably is about 100 ⁰C to about 250 ⁰C.

The temperature range within the oxidation reactor generally is from about 120 ⁰C, preferably from about 150 ⁰C, to about 240 ⁰C, preferably to about 230 ⁰C. In the continuous oxidations, a suitable temperature range is about 175 ⁰C, to about 235 °C preferably about 190 ⁰C, to about 210 ⁰C. If a continuous oxidation process is conducted in stages, the first stage temperature suitably is about 190 °C, to about 210 ⁰C, and the second or at least one subsequent stage temperature is about 170 ⁰C, to about 195 ⁰C.

The oxidation reaction is conducted under total pressure effective to maintain a liquid phase reaction mixture at the operating temperature. Suitably, total pressure is at least about 35 bar, and preferably about 40 bar to about 120 bar and preferably 95 bar. More preferably, total pressure is about 65 to about 85 bar. Carbon dioxide partial pressure is maintained at levels effective for expanding the volume of the liquid reaction mixture. Under oxidation temperature and total pressure conditions described above, CO₂ partial pressures in the range of 30 to 100 bar are suitable for expanding the liquid phase reaction mixture such that V_e/V₀ is 1.25 or greater. The residence time in the oxidation reactor generally is from about 5 to about

150 minutes and preferably from about 10 to 90 minutes.

In a continuous system, some of the solvent and water generated as byproduct of the oxidation reaction is withdrawn from the reactor as a vapor, which is then condensed and recycled to the reactor. In preferred embodiments of the invention, solvent monocarboxlic acid and water present in the withdrawn vapor are separated. Some solvent also is withdrawn from the reactor as a liquid in the product stream. After separation of the crude acid product from the product stream, at least a portion of the liquid reaction mixture that remains after separation of crude acid product from the product stream, commonly referred to as mother liquor, is generally recycled to the reactor. Carboxylic acid product solids typically are removed from the reactor and separated from the solvent system by filtration or centrifuging.

Heat generated in the highly exothermic liquid-phase oxidation may be dissipated at least partially by vaporization of a portion of the liquid reaction mixture in the oxidation reactor. The resulting vapor phase comprises solvent monocarboxylic acid, water and carbon dioxide and may also include lighter byproducts of the oxidation reaction and inert gases that may be present from the oxygen source. According to another aspect of the invention, a portion of the homogeneous liquid solvent phase is removed from the reactor, and carbon dioxide is evaporated therefrom. The evaporated CO₂ may be compressed, externally cooled, and recycled back to the reactor to provide process cooling.

Reaction conditions (e.g., temperature, pressure, residence time, catalyst and reactant concentrations) preferably are used to convert substantially all of the oxidizable substrate to carboxylic acid. In order for a process to be commercially practicable, conversions (i.e., yield in terms of mole product per mole of feed) should be above 95 mole %, preferably above 98 mole %, and most preferably above 99 mole % based on the initial oxidizable substrate. For conversion of para-xylene to terephthalic acid, conversions above 99.5% are typical; however, MC oxidation conversion of substrates that are not xylenes or their partially oxidized derivatives typically is lower. For example, conversions of dimethylnaphthalene to naphthalene dicarboxylic acid and pseudocumene to trimellitic acid may achieve above 80% and typically above 90% in well optimized systems. Use of the process of this invention typically reduces the amount of carboxylic acid (e.g. acetic acid) solvent which is oxidized (i.e., burned) during the oxidation process. In conventional processes, solvent burning is monitored by measuring the amount of carbon dioxide present in the system. In the process of this invention, carbon dioxide gas present in the system includes not only that resulting from burning of the solvent monocarboxylic acid but also that pressured to the reaction system for expansion of the liquid reaction body, such that carbon dioxide may not be a reliable indicator of solvent burning. However, measurement of carbon monoxide has been found to reflect acetic acid burning. Use of a carbon monoxide measurement as a reflection of solvent burning was confirmed through experiments in which a known amount of carbon monoxide was added to an oxidation. Subsequent measurements showed that carbon monoxide was not oxidized to CO₂ in the reaction.

Depending on the oxidation reaction conditions used, the aromatic feed compound selected, the oxidation catalysts, and the levels of catalyst selected, the reaction product mixture contains impurities and reaction by-products in addition to the desired aromatic carboxylic acid. For example, impurities formed in terephthalic acid production include 4-carboxybenzaldehyde (4-CBA), tolualdehyde, p- hydroxymethyl benzoic acid and p-toluic acid, which are intermediate products in the oxidation of para-xylene. Unidentified color-forming precursors and color bodies, possibly of the benzil, fluorenone or anthraquinone structure, often are also present.

Typically, a carboxylic acid product produced using MC oxidation is purified to reduce impurities levels in order to produce "polymer" grade product that is preferred for direct condensation polymerization with a glycol (such as ethylene glycol) to form a polyester. Carboxylic acid with reduced levels of impurities can be made by staged or further oxidizations of crude oxidation products, typically at one or more, progressively lower temperatures and oxygen levels, and during crystallization to recover products of the oxidation, for conversion of partial oxidation products to the desired acid product, for example as described in U.S. 4,877,900, 4,772,748 and 4,286,101. Preferred pure forms of terephthalic acid and other aromatic carboxylic acids with lower impurities contents, such as purified terephthalic acid or "PTA", can be made by catalytically hydrogenating less pure forms of the acids, such as crude product comprising aromatic carboxylic acid and by-products generated by oxidation of alkyl arenes according to the invented process, in solution at elevated temperature and pressure using a noble metal catalyst. For example, U.S. 3,584,039 and 4,892,972, incorporated by reference herein, describe purification methods using hydrogenation in an aqueous or acetic acid medium at an elevated temperature solvent over a noble metal catalyst to convert various color bodies present in the relatively impure dicarboxylic acid to colorless products without substantial hydrogenation of the aromatic ring of the aromatic dicarboxylic acid. The conventional hydrogenation catalyst is palladium on a carbon support, as described in U.S. 3,726,915, incorporated by reference herein. Another purification-by- hydrogenation process of aromatic polycarboxylic acids produced by liquid phase catalyst oxidation of polyalkyl aromatic hydrocarbons is described in U.S. 4,405,809, incorporated by reference herein.

As described in these patent documents, catalysts suitable for use in a purification by hydrogenation are insoluble under the conditions employed therein and typically contain at least one supported or unsupported Group VIII noble metal, including palladium, rhodium, ruthenium, osmium, iridium, and platinum. Preferably, the noble metal is at least one member of the group consisting of palladium and rhodium. A typical catalyst of palladium on a support comprises from about 0.01 weight percent to about 2 weight percent of palladium, based on total weight of catalyst and calculated as elemental metal. The support or carrier for the palladium is porous and inert, and preferably is active carbon having a surface area of about 600 m²/g to about 1 ,500 m²/g. Other suitable supports are described in U.S. 5,362,908, incorporated by reference herein. The hydrogenation purification step for producing, for example, purified terephthalic acid for use in combination with the process of the present invention is suitably conducted at an elevated temperature and pressure in a fixed catalyst bed. Crude arene carboxylic acid, e.g., terephthalic acid, to be purified is dissolved in water or a like solvent. Although water is a preferred solvent, other suitable solvents include relatively lower molecular weight aliphatic carboxylic acids, alone or admixed with water. Suitable reactor temperatures used in this purification step typically range from about 100 to about 350 ⁰C. Preferably, temperatures used in the purification step range from about 275 to about 300 ⁰C.

The concentration of arene carboxylic acid in the solution to be purified by hydrogenation can vary over a relatively wide range. Concentration can be as low as about 5 percent by weight or as high as about 35 percent by weight, based on the weight of the solution. Preferably, the solution concentration in purification of crude terephthalic acid products is in a range of from about 10 to about 30 % by weight.

Pressure used in the purification step depends primarily upon the temperature. Inasmuch as the temperatures at which practical amounts of the impure aromatic dicarboxylic acid may be dissolved in a typically useful solvent are substantially above the normal boiling point of the solvent, the process pressures are necessarily above atmospheric pressure to maintain the purification reaction solution in a liquid phase. In general, the reactor pressure during hydrogenation can range from about 1.5 to about 10 mPa, and usually range from about 6 to about 8.5 mPa.

In general, the amount hydrogen supplied to the purification reactor under reaction conditions is sufficient to effect the desired purification by hydrogenation.

Space velocity, reported as weight of crude arene carboxylic acid solution per weight of catalyst per hour, in the purification step ranges from about 5 hours^"1 to about 25 hours^"1, preferably from about 10 hours^"1 to about 15 hours^"1. Residence time of the solution in the catalyst bed varies depending upon activity of catalysts present. Recovery of a purified arene carboxylic acid in such a purification step typically involves crystallizing the desired product acid, e.g., terephthalic acid, from the purification reaction mixture resulting from the hydrogenation step, for example by releasing pressure on the mixture for cooling by evaporation of the reaction solvent. Crystallized solids can be recovered from the remaining liquid reaction mixture by solid-liquid separation steps such as filtration and centrifuging.

Our invention is illustrated, but not limited, by the following examples.

A series of oxidations of p-xylene using an MC catalyst was conducted in a volumetrically expanded liquid reaction mixture in a reaction vessel pressurized with carbon dioxide. The oxidations were conducted in a 78-milliliter liter titanium batch reactor fitted a valve for introduction of gas and equipped with an external shaker for agitation and an internal thermocouple for monitoring temperature of the reactor contents. The thermocouple also aided in evaluating expansion of the liquid oxidation reaction mixtures in the reactor because solid particles, primarily of terephthalic acid reaction product, from the reaction mixtures coated or clung to the thermocouple over the course of the oxidations such that observations of the initial liquid levels in the reactor at the start of the oxidations and of reactor volumes corresponding to heights of particles on the thermocouple at the end of the oxidations could be used to calculate expansion of the liquid body. The reactor was charged initially with catalyst solution in acetic acid and p-xylene feed and pressurized with the desired gas mixture. The reactor was placed in the shaker and immersed in a sand bath to bring the internal temperature to the desired reaction temperature within 1.5 minutes with shaking at a rate of 340 cycles per minute. After the desired oxidation time (typically 20 minutes), the reactor was cooled to room temperature with water directed into contact with its exterior surfaces from jets. The total product from the reactor was analyzed for carboxylic acid product content, as well as content of major intermediates (4-CBA and p-toluic acid). Off-gas was analyzed for CO, methyl bromide and oxygen.

Preliminary to running oxidation experiments, the reactor was pressured with a 80/20 by volume mixture of CO2 and O₂ with and without the presence of a known, constant volume of liquid acetic acid, in all cases at a temperature of 23 ⁰C, and the reactor was weighed to determine the mass of the gas in the reactor with and without the acetic acid. Presence of an expanded volume of acetic acid in the reactor and greater solubility of oxygen in the expanded acetic acid are consistent with the plot of gas mass against pressure with and without acetic acid in Figure 1 , from which it is seen that at given pressures, the mass of the gas mixture added to the reactor was greater when acetic acid was present in the reactor than when acetic acid was not present.

Example 1

A stock catalyst solution containing cobalt (as cobalt acetate, 21 mM), manganese (as manganese acetate, 21 mM), and hydrogen bromide (21 mM) was prepared in 95% aqueous acetic acid. Catalyst solution (5 g) and p-xylene (0.5 g) were added to the reactor followed by pressurization at room temperature with 80/20 by volume CO₂/O₂ to 3000 kPa, then further pressurized to 4930 kPa with CO₂, such that the partial pressure of CO₂ was 4330 kPa. The result was a 91/9 by volume CO₂/O₂ gas mixture with the concentration of Co, Mn, and HBr in the liquid mixture of 20 millimolar (mM) each. An oxidation reaction was conducted at 195 ⁰C for 20 minutes with agitation, after which the reactor was cooled, depressured, opened, and product recovered by filtration. From the initial liquid level in the reactor and reactor volume corresponding to height of reaction product particles present on the thermocouple, the volume of the expanded liquid was calculated to correspond to an expansion ratio (V_e/V₀) equal to 2.9. Oxidation of acetic acid was estimated from the amount of carbon monoxide produced as reported as mole CO/mole p-xylene. Yields of terephthalic acid and intermediate oxidation products, expressed in mole % based on feed, were determined by liquid chromatographic (LC) analysis of dry filter cake. The terephthalic acid recovered showed morphology similar to that observed in conventional preparations. Results are shown in Table 1. Example 2

An oxidation was conducted in a manner similar to that described for Example

1 except that the reactor was initially pressurized to 2310 kPa with 80/20 by volume

CO₂/O₂ and then further pressurized to 4930 kPa with CO₂ providing a 93/7 CO₂/O₂ mixture, carbon dioxide partial pressure of 4585 kPa and an estimated Ve/Vo of about 3.2. Results are shown in Table 1. Example 3

An oxidation was conducted in a manner similar to that described for Example

1 except that the reactor was initially pressurized to 3550 kPa with 80/20 by volume

CO₂/O₂ and then further pressurized to 4930 kPa with CO₂ providing a 89/11 CO₂/O₂ mixture, carbon dioxide partial pressure of 4390 kPa and an estimated Ve/Vo of about 3. Results are shown in Table 1.

Example 4

An oxidation was conducted in a manner similar to that described for Example 1 except that the oxidation reaction was conducted for 10 minutes. Results are shown in Table 1.

Example 5

An oxidation was conducted in a manner similar to that described for Example 1 except that 2.5 g of catalyst stock solution and 2.5 g of 95% acetic acid were used, which yielded 10 mM concentrations of Co, Mn, and HBr. Results are shown in Table 1.

Comparative Runs A1-A8

A series of eight comparative oxidation runs using the catalyst and p-xylene concentrations indicated in Example 1 was conducted without pressuring with CO₂ using 95% aqueous acetic acid as solvent and different 1/1/1 Co/MnBr catalyst concentrations of 5 to 35 mM. The total pressure was 3550 kPa using air as the oxidizing gas. Results from these runs including averaged data for baseline comparisons are reported in Table 1.

Table 1

The amount of methyl bromide produced in Examples 1-5 was comparable to that produced in the comparative runs. Example 6

An oxidation was conducted in a manner similar to that described for Example 1 except that m-xylene (m-X) (0.5 g) was used as the oxidizable alkyl arene substrate. The product was isophthalic acid. Results are shown in Table 2. Example 7

An oxidation was conducted in a manner similar to that described for Example 1 except that o-xylene (o-X) (0.5 g) was used as the oxidizable alkyl arene substrate. The product was phthalic acid. Results are shown in Table 2.

Example 8 An oxidation was conducted in a manner similar to that in Example 1 but using

2,6-dimethylnaphthalene (DMN) (0.5 g) as the oxidizable alkyl arene substrate. The product was 2,6-napthalene dicarboxylic acid. Results are shown in Table 2.

Example 9

An oxidation was conducted in a manner similar to that described in Example 1 except that pseudocumene (pC) (0.5 g) was used as the oxidizable alkyl arene substrate. The product was trimellitic acid. Results are shown in Table 2.

Table 2

Comparative Run B The oxidation reaction procedure of Example 5 was repeated except that the reactor was pressurized with air (80/20 by volume N₂/O₂) to 3550 kPa and then to 4930 kPa with nitrogen. The resulting gas mixture was 85/15 by volume N₂/O₂. The resulting mole CO/mole pX, which reflected solvent burning, was 0.092, the total intermediates was 0.76 wt%, and the yield was 99.2 mole%.

Examples 10-16

A series of oxidations was conducted similar to that described for Example 1 , except that the concentration of catalyst was varied, while the molar ratio of Co/Mn/Br catalyst components was fixed at 1/1/1. The results are shown in Table 3.

Table 3

Examples 17-20

A series of oxidations was conducted in a manner similar to that described for Example 1 except the amount of water in the solvent (percent based on acetic acid) was varied. Ve/Vo was estimated to be about 2.9 in Examples 17 and 18, about 2.6 in Example 19 and about 2.0 in Example 20. Results are shown in Table 4.

Table 4

Examples 21-24

A series of oxidations was conducted in a manner similar to that described for Example 1 at varying reaction temperatures. Also in these examples, 2 grams of a catalyst stock solution containing 21 mM concentrations of the Co/Mn/Br catalyst components were used with 2 grams of glacial (>99.5%) acetic acid and 0.5 gram deionized water. The resulting catalyst concentration was 10 mM of each component and 10 wt% water. The results are shown in Table 5.

Table 5

Claims

We claim:

1. A process to convert a C₈-C₂0 alkylarene to an arene carboxylic acid in an oxidation reactor comprising reacting the alkylarene with oxygen in the presence of an MC catalyst in a volumetrically-expanded liquid phase reaction mixture in which a solvent comprises a homogeneous phase of a monocarboxylic acid and carbon dioxide wherein the reaction is conducted under oxidation conditions including a carbon dioxide partial pressure such that the ratio of the volume of the volumetrically expanded liquid reaction mixture under such oxidation conditions (V_e) to the volume of the liquid reaction mixture cooled and depressurized to normal conditions (V₀) is more than 1.25.

2. The process of claim 1 wherein the volumetrically expanded liquid reaction mixture contains water.

3. The process of claim 1 wherein the solvent contains at least 1 weight percent of water based on the amount of monocarboxylic acid.

4. The process of claim 1 wherein the alkyl arene is a benzene substituted with two or three methyl groups.

5. The process of claim 1 wherein the alkyl arene is p-xylene, o-xylene, m- xylene, a dimethylnaphthalene, or pseudocumene.

6. The process of claim 1 wherein the alkyl arene is p-xylene.

7. The process of claim 1 wherein the monocarboxylic acid is acetic acid.

8. The process of claim 7 wherein the weight ratio of acetic acid to alkylarene in the volumetrically expanded liquid reaction mixture is less than 2.

9. The process of claim 7 wherein the partial pressure of carbon dioxide in the oxidation reactor is greater than 35 bar.

10. The process of claim 1 wherein the MC catalyst comprises cobalt, manganese, and bromine.

1 1. The process of claim 1 wherein the partial pressure of carbon dioxide in the oxidation reactor is greater than 30 bar.

12. The process of claim 1 wherein the oxidation temperature is between 100 and 250⁰C.

13. The process of claim 1 wherein gas in the reactor contains less than 5 wt% nitrogen.

14. The process of claim 1 wherein at least 95 mole% of alkylarene is converted to an arene carboxylic acid.

15. The process of claim 1 wherein at least 99 mole% of alkylarene is converted to an arene carboxylic acid.

16. The process of claim 1 in which a vent stream comprising carbon dioxide, water, and acetic acid is withdrawn from the reactor; water is separated from that stream; and carbon dioxide and acetic acid are recycled to the reactor.

17. A process to convert para-xylene, a partially oxidized derivative thereof or a combination thereof to terephthalic acid in an oxidation reactor comprising reacting the para-xylene, partially oxidized derivative thereof or combination thereof with oxygen in the presence of an oxidation catalyst comprising a heavy metal component and bromine or a bromine-free catalyst in a volumetrically expanded liquid reaction body in which a solvent for the para-xylene comprises a homogeneous phase of acetic acid, carbon dioxide and up to 35% water by weight of the acetic acid wherein the reaction is conducted under oxidation conditions including a carbon dioxide partial pressure of at least 30 bar.

18. A solvent composition comprising a volumetrically expanded homogeneous phase of a monocarboxylic acid and carbon dioxide under elevated pressure and temperature conditions, including a carbon dioxide partial pressure such that the ratio of the volume of the volumetrically expanded homogeneous phase under such conditions (V_e) to the volume of monocarboxylic acid under normal conditions (V₀) is more than 1.25.

19. The composition of claim 18 containing water.

20. The composition of claim 19 wherein the monocarboxylic acid is acetic acid.

21. The composition of claim 20 having an arene carboxylic acid dissolved therein or slurried in solid form therein.