Patent Application
20100041919
The present invention relates to recovery of carboxylic acids by hydrolysis of esters, and in particular to recovery of acetic acid by hydrolysis of methyl acetate containing mixtures from polyvinyl alcohol production processes.
Several industrial processes produce streams containing one or more esters of carboxylic acids, RO.CO.R′, where R and R′ are selected from among hydrocarbyl radicals or substituted hydrocarbyl radicals. Examples of such industrial processes include those for production of vinyl alcohol polymers or terephthalic acid. Carboxylic acid esters can be hydrolyzed to generate the corresponding acid and alcohol, as shown in Equation 1. Equation 1 is an equilibrium reaction and requires an excess of water to drive the reaction well to the right hand side.
RO.CO.R′+H2OROH+HO.CO.R′ [1]
Processes for production of polyvinyl alcohol (PVA) and its derivatives are described by Marten in “Vinyl Alcohol Polymers” in Kirk-Othmer Encyclopedia of Technology, John Wiley & Sons, Inc. A variety of vinyl ester monomers can be polymerized to form a polymer, of which polyvinyl acetate (PVAc) is preferred. PVAc then is further reacted to manufacture PVA. Commonly, PVAc is reacted with methanol (MeOH) to form PVA and methyl acetate (MeOAc). The components of the polymerization reaction mixture are continuously separated. Unreacted monomer can be stripped from the reactor using, for example, methanol vapor. The overhead fraction from the stripper comprises a mixture of vinyl ester monomer and at least one solvent such as methanol. The vinyl ester monomer is then extracted for recycle to the polymerization reactor. In the production of polyvinyl alcohol (PVA) MeOAc is produced as a by-product at a ratio of 1.68 tons of MeOAc per ton of PVA.
One outlet stream typically comprises a mixture including MeOAc, MeOH and a small amount of water. The weight ratio of these components varies over a range of relative concentrations, among which a typical composition is approximately 75% MeOAc, 23% MeOH and 2% water. Among these components MeOH and water have relatively low value when compared to the values of MeOAc and acetic acid (HOAc).
The MeOH and MeOAc can be distilled off and, at the same time, water can be added in order to obtain an aqueous PVA solution. However, there are disadvantages to this approach. The resulting PVA suspension is fine, difficult to filter, and so the process is uneconomical. Further, this approach requires time-consuming, energy intensive and hence expensive distillation of large amounts of solvents requiring a plurality of distillation columns. Several approaches have been undertaken to improve the chemical efficiency and economics of processes for production of PVA. In particular, efforts have been directed to recycle of solvents and processing of the outlet streams to recover valuable by-products.
Kowaka et al. in U.S. Pat. No. 6,743,859 issued in 2004 describe a method for production of high-strength PVA with a high degree of saponification. The apparatus for the process of '859 includes an outlet line for recovery of MeOH and MeOAc identified in
Bauer et al. in U.S. Pat. No. 6,576,720 issued in 2003 describe an alternative approach in which a liquid phase comprising MeOH, MeOAc and HOAc is recycled for use in further polyvinyl ester transesterification. The process of '720 can be used for other alcohols and esters. The mixture of alcohol and corresponding ester is recycled for use as the polymerization reaction medium. Make-up comprising one or both of MeOH and HOAc is provided to maintain the composition of the recycle mixture. The PVA is produced and isolated using saponification with KOH and then neutralization, preferably with a strong acid such as HCl. Thus water, less than 1% by weight in the initial reaction mixture, is produced by both the reaction of MeOH and HOAc and the neutralization process.
MeOAc may be sold or further hydrolysed to recover HOAc.
Kim et al. in U.S. Pat. No. 5,770,770 issued in 1998 describe a reactive distillation process for the well known equilibrium reaction 2 for the recovery of MeOH and HOAc from catalytic hydrolysis of MeOAc. Reaction 2 is a specific example of the type of reaction shown in Equation 1.
MeOAc+H2OMeOH+HOAc [2]
It has long been recognized that this reaction could be used to recover HOAc from MeOAc from a PVA manufacturing process as described, by example, by Adelman et al. in U.S. Pat. No. 4,352,940 issued in 1982. It also was recognized in '940 that it was necessary to minimize the amount of water used in the process to reduce the costs of recovery and re-use of the products from the reaction in the PVA manufacture process. However, when a minimum amount of water is used, the equilibrium reaction 2 lies to the left hand side. Reaction 2 can be driven to the right hand side only by continuous removal of at least one of the products.
In an alternative approach for treatment of the PVA manufacture outlet stream, MeOAc can be separated from the mixture using extractive distillation. One example of this method is described by Xiao et al. in Chemical Engineering Science, volume 56, pages 6553-6562 (2001). In the first column, water is added to the liquid stream from the PVA plant. The volatiles from the first column are then hydrolyzed in a fixed bed reactor containing a bed of an acidic catalyst which catalyzes the hydrolysis of MeOAc to MeOH and HOAc. The effluent stream from the fixed bed reactor is distilled in a second distillation column to provide a volatiles stream and a bottoms stream. The volatiles from the second column are recycled for mixing with further MeOAc feed to the first extractive distillation column. The bottoms from the second column are separated into a water rich stream and a HOAc rich stream by distillation in a third column. The bottoms from the first column are separated by distillation in a fourth column into a water rich stream and a MeOH rich stream. Thus the overall process for recovery of HOAc requires four distillation columns and a fixed bed reactor. Further, to drive well to the right hand side the well known catalytic MeOAc hydrolysis equilibrium reaction shown as Equation 2, it is necessary to use a large amount of water. Thus the process is energy intensive as that water must be volatolized in both the second and third columns.
Each of the above processes requires use of a plurality of columns and reactors to react, separate and recover the components of the stream from the PVA manufacturing reactor. Consequently, capital and operating costs are high. Further, when water is added, either as reagent or for extractive distillation, that water must also be separated, which is a costly and time consuming feature.
Hoyme et al. in U.S. Pat. No. 6,518,465 issued in 2003 describe another concept based process, derived from simulations using the commercial available program Aspen Plus, in which the stream containing MeOAc from PVA manufacture is reacted in a reactive distillation column to produce DME and HOAc. Water was added to hydrolyze MeOAc and thereby generate HOAc which is recovered. The molar ratio of water in the process stream is between 0.05% and 20%, and preferably is between 0.3% and 3%. In this process it is recognized that methanol also may react to generate dimethyl ether (DME) and water in the also well known acid catalyzed equilibrium reaction shown in Equation 3. The process of Hoyme et al. in '465 is basically hydrolysis of MeOAc to HOAc by addition of water.
2 MeOHMe2O+H2O [3]
What is needed is an apparatus and process for separation and recovery of HOAc produced by hydrolysis of MeOAc in a stream from a reactor in which PVA is produced.
The present invention is a process for recovery of HOAc by catalytic distillation of streams comprising MeOAc, MeOH and lesser amounts of water wherein the apparatus comprises only two columns: the catalytic distillation column and a distillation column for HOAc recovery, the process requiring no added water. Reactive distillation is used for conversion of a mixed MeOAc-MeOH feed to HOAc and DME, through a two-step consecutive reaction process. The first reaction involves the equilibrium dehydration of MeOH to form DME and water. The second reaction employs this water to react with MeOAc in the hydrolysis reaction that produces MeOH and HOAc. The MeOH produced during hydrolysis of MeOAc is then able to participate further in MeOH dehydration to produce more water. In effect, the two reactions are “self-feeding” and can be visually represented as follows:
The benefit of this reactive distillation process is that no additional water is added as feed and the process continues as long as there is water and/or MeOH in the column to react. Once the MeOAc-MeOH azeotrope has been broken, any further separation of MeOAc and HOAc becomes a matter of simple distillation.
The present invention relates to a process for recovery of carboxylic acids from mixtures containing a corresponding ester and an alcohol through hydrolysis in a catalytic distillation column containing an acidic catalyst. The alcohol is dehydrated to produce the corresponding ether and water, and this water is used to hydrolyze the ester to the corresponding alcohol and acid. A bottoms stream containing the acid exits the catalytic distillation column and is sent to a distillation column from which the substantially pure acid is recovered as a bottom stream. There is no requirement for water to be added to the feed for the process. An alternative embodiment of the apparatus includes means for addition of supplemental amounts of methanol and/or water to maintain optimum operation of the process independent of the composition of the ester containing mixture from which acetic acid is to be recovered.
For a more complete understanding of the present invention and for further objects and advantages thereof, reference is made to the following description taken in conjunction with the accompanying drawings.
The following description comprises data obtained through laboratory experiments and simulations using ASPEN PLUS. However, in order to perform the simulations and thereby correlate experimental data with theory, it was necessary to obtain data on the physical and chemical equilibria occurring in the respective reaction mixtures. The data for individual reactions 2 and 3 are available in the open literature. However, no physical and chemical kinetic data were available for mixtures comprising all five components involved. To obtain accurate simulation results, experiments were performed to obtain kinetic data for reactions among combinations of all five components. The following brief descriptions provide background data available and the measures taken to determine the equilibria parameters.
As a consequence of having experimental data on physical and chemical equilibria that were hitherto unavailable to those skilled in the art, the construction, experimental demonstration and modeling of the process comprising the invention described herein have higher reliability than those described elsewhere. For the first time, we show accurately a process comprising the two concurrent continuous reversible reactions occurring in a single reactor, one generating water and DME from MeOH and another consuming said water by reaction with MeOAc to generate both MeOH and HOAc. Water generated in one reaction is consumed in the other, and MeOH generated in the latter reaction is consumed in the former.
We are for the first time able to compute fully the roles and impacts of each of the components in the reaction mixture, experimentally verify the computed data, and so establish operating parameters so that:
there is essentially no water generated in the overall process, and substantially all water is consumed when there is excess MeOAc to undergo hydrolysis;
DME is recovered as a volatile in the overhead;
a stream that is richer in HOAc is generated for transfer to a distillation column for recovery of essentially anhydrous HOAc, separately from recovery of DME; and
the process has high overall energy efficiency, as there is no requirement for consumption of energy to remove added high concentrations of water.
The process of the present invention therefore offers advantages over those described in the prior art.
We will show that components of the bottoms from the single reactor in which the two equilibrium reactions are occurring include a much higher concentration of HOAc, along with MeOAc and smaller amounts of additional components. For the overall process to operate with high energy efficiency and generate substantially pure HOAc, it is necessary that the bottoms from the reactor are transferred to a separate distillation column. In a modified embodiment of the process, all methyl species are converted when DME is completely removed by catalytic distillation with the consequence that the bottoms stream is substantially pure HOAc.
In a further modification, using either a batch reactor or a continuous reaction system, if the feed composition is adjusted through addition of stoichiometric amounts of methanol and/or water, substantially pure HOAc can be produced in a single column because DME is very volatile and can be completely removed from the reaction mixture.
Physical Equilibria
An important consideration in any reactive distillation simulation is the choice of physical equilibrium model and the ability to reliably predict multicomponent vapour-liquid equilibria (VLE) and liquid-liquid equilibria (LLE). Reliable VLE and LLE are needed to establish distillation boundaries and to determine if and where azeotropes and phase separations occur. There exists an abundance of patchwork/piecemeal phase equilibrium data in the open literature on the multicomponent mixture of MeOAc-MeOH-water-HOAc-DME and the respective subsystems. The various equations used to model these systems take into account the non-ideality of the vapor phase (due to dimerization of HOAc).
Chemical Equilibria:
The two concurrent reversible reactions considered in the process are the hydrolysis of MeOAc and the dehydration of MeOH. The hydrolysis of MeOAc with water over an acid catalyst produces HOAc and MeOH (Equation 2). Reaction 2 is equilibrium limited with a reported equilibrium constant of 0.13 at 25° C. (Ge et al., “Kinetics of Heterogeneous Hydrolysis of Methyl Acetate”, Chemical Reaction and Engineering and Technology (Chinese Journal), Vol. 14, No. 2, 1998, pp 138-144). The dehydration of methanol, consumes two moles of methanol to produce one mole of DME and one mole of water (Equation 3).
The equilibrium constant has been reported to range from 68.4 to 52.3 in the temperature range from 85° C. to 115° C. (Nisoli et al., “Attainable Regions for Reaction with Separation” in AIChE Journal, Vol. 43, No. 3 (2), 1997, pp 374-387). The overall reaction is given by:
MeOAc+MeOHDME+HOAc (4)
Because water is both consumed and produced stoichiometrically, it drops out of the overall reaction equation.
The Invention:
An exemplary embodiment of the invention, which is non-limiting, will now be described with reference to
Referring to
Catalytic distillation column 12 has a top portion (“rectifying zone”) 18, a middle portion (“reactive zone”) 20 and a bottom portion (“stripping zone”) 22. A fixed bed 24 containing an acidic catalyst 26 is situated within middle portion 16 of catalytic distillation column 12. When, for example, a feed 30 is an HOAc rich stream from a chemicals manufacturing facility (not illustrated) such as a PVA manufacturing facility, a feed line 28 carries feed 30 to a location 32 part way up fixed bed 24. A first condenser 34 is used to condense a portion of vapors in top portion 18, from which a fraction is recycled to distillation column 12 and another fraction is recovered as first volatiles product 44 via a first volatiles outlet line 36. A first reboiler 38 is used to reboil a portion of first liquid bottoms 40 exiting bottom portion 22 for recycle to catalytic distillation column 12, and another portion of first liquid bottoms 40 is removed via a first liquid bottoms outlet line 42.
Acidic catalyst 26 can be one or more of several different acidic catalysts. It has been found through experiments that any one of several acidic resin catalysts is suitable for use in the catalytic hydration of MeOAc according to Equation 2, including but not limited to use of the acidic forms of commercially available resins, Amberlyst 15, 35 or 37.
Distillation column 14 has a top portion 50, a middle portion 52, a bottom portion 54, a second condenser 56 and a second reboiler 58. The another portion of first liquid bottoms 40 from catalytic distillation column 12 is fed via first liquid bottoms outlet line 42 to a position 60 approximately midway up middle portion 52 of distillation column 14. The optimum position 60 for connection of first liquid bottoms outlet line 42 and distillation column 14 has been determined experimentally using modeling and predicted operating performance data have been confirmed experimentally, as described in the EXAMPLES below.
Second condenser 56 condenses a portion of vapors in top portion 50 for recycle to distillation column 14, and another portion is recovered as a second volatiles product 61 via a second volatiles outlet line 62. Second reboiler 58 reboils a portion of a second liquid bottoms 64 in bottom portion 54 of distillation column 14, and another portion of second liquid bottoms 64 is recovered via a second liquid bottoms outlet line 66.
When the chemicals processing facility manufactures one or more of PVA and derivatives of PVA feed 40 is a stream from the manufacturing facility typically comprising MeOAc, MeOH and water. It is desirable to recover acetic acid, which may be recycled for use in manufacture of vinyl acetate which in turn can be polymerized for manufacture of PVA. The ether formed from dehydration of MeOH is DME and it is recovered as first volatiles product 44. DME can be recovered for sale or for other use.
The second volatiles product 61 from distillation column 14 is much richer in MeOAc than was feed 40, and this stream can be recovered for sale or recycled to distillation column 12.
The method of use of apparatus 10 will now be illustrated with reference to
The composition of feed 30 from a PVA manufacturing facility is shown in Table 4. Also in Table 4 are the compositions, for one exemplary set of operating parameters, of that portion of first volatile products 44 recovered via first volatiles outlet line 36 and that portion of first liquid bottoms 40 removed via first liquid bottoms outlet line 42.
Laboratory experiments, described in EXAMPLE 1 below, showed that the rate of dehydration of MeOH to form DME (Equation 3) is affected by the concentrations of both water and MeOH in a reaction mixture initially comprising MeOH-water-HOAc-MeOAc. Thus, to maintain efficient operation, it is desirable that the mixture fed to catalytic distillation column 12 has an optimum composition profile.
An alternative embodiment of the present invention enables more efficient operation of the process independent of variations that may occur from time to time in the composition of feed 30 before it is fed via feed line 28 into catalytic distillation column 12, illustrated in
A benefit from use of the second embodiment of the present invention is that the apparatus is more versatile. The capability to provide amounts of one or both of methanol and water in addition to feed 30 can be used for recovery of HOAc (or other carboxylic acids) from a variety of different industrial streams containing hydrolysable esters.
A benefit arising from use of either apparatus 10 or the second embodiment of the present invention is that the process can be operated so as to produce a stream of high purity DME and HOAc as products. In this modification of either embodiment, DME is continuously removed from apparatus 10 so that substantially all methyl species are converted to DME and so are removed from the system, with the consequence that first liquid bottoms 40 comprises substantially pure HOAc.
The following EXAMPLES will illustrate application of apparatus and method for recovery of HOAc. Each set of conditions has been modeled using ASPEN and results from modeling have been confirmed through experiment.
Laboratory experiments were conducted to confirm literature data and to obtain additional insights into the two catalytic reactions for the formation of DME and HOAc, using an apparatus 100 illustrated in
Using the system demonstrated in these laboratory experiments and illustrated in
To examine the reliability and effectiveness of the physical and reaction equilibrium models, a batch catalytic distillation simulation was performed and compared to experimental data collected in the lab.
For the experimental test, a batch Parr reactor was loaded with a 150 g mixture of MeOAc/MeOH feed in a 75:25 weight ratio, along with 30 g of previously dried acid catalyst (Amberlyst 35). The top of the batch reactor was modified with a stainless steel condenser, which allowed the most volatile component (i.e., DME) to escape. The condenser was cooled with cold water; GC analysis of the off-gas from the condenser confirmed that >99% DME was leaving the reactor. The reactor was operated at 120° C. and 150 psig over a period of 13 hours until equilibrium had been reached. It was determined that equilibrium was established based on vapor and liquid samples from the reactor which showed constant concentration over several samples. Vapor and liquid samples were measured on an HP 6890 GC equipped with capillary column and TCD. A two-stage equilibrium reactor with distillation column was sufficient to model the batch distillation process in Aspen Plus. Table 1 shows the liquid composition for the batch catalytic distillation simulation and compares it to experimental batch catalytic distillation results.
Comparison of data from the simulation and experimental results confirms that the predictions from our model are accurate in terms of reaction rate and equilibrium constants.
The continuous catalytic distillation process consists of a 20 stage CD column with total condenser and partial reboiler. The parameters for the base case simulation are given in Table 2. All feed streams enter at 25° C.
Operating pressure is one of the key elements in the design of a CD column. The choice of operating pressure for a catalytic distillation column depends on many considerations such as the overhead temperature, bottom temperature, reaction temperature and relative volatilities of the components in the system. The column pressure sets the lower and upper bounds of the temperature within the column. For the present process the operating pressure is chosen within a range such that water can be used as a coolant for the overhead condenser and steam can be used as a heating medium for the reboiler. Within this range, the reaction temperature mainly determines the operating pressure. Because the catalytic reactions take place within the liquid phase, the reaction temperature is close to the boiling point of liquid flowing around the catalyst. As a result the reaction temperature increases with column pressure.
The amount of water in the feed is an important variable in the CD process because water is a necessary reactant in the hydrolysis reaction. However, water is also produced via MeOH dehydration. Therefore, there should be an optimum amount of water that will allow both reactions to proceed readily without accumulating a large amount of water in the reaction mixture. The optimum process achieves maximum conversion of reactants while at the same time completely removing water from any of the product streams.
Conversely, when no water enters in the feed, only the water that is produced through MeOH dehydration is available for the hydrolysis reaction. In this case DME is again the distillate product, and a mixture of HOAc and MeOAc is now the bottoms product. All of the water and MeOH is reacted away. The resulting profile of concentrations of all CD column reaction mixture components is shown in
Based on the overall reaction equation given by equation (3), one might predict that the there should be very little influence on the process arising from changes in the ratio of MeOAc to MeOH. Increasing either MeOH or MeOAc in the feed should favour DME and HOAc product formation. However, based on the stoichiometry of the individual reactions, this is not necessarily true. The effect of increasing the MeOAc/MeOH feed ratio results in complete conversion of water and methanol in the CD column with a profile similar to that shown in
Simulation results show that column parameters can be varied to completely remove both MeOAc and MeOH from the column mixture, and a mixture of water and HOAc remains as the bottoms product.
Catalytic distillation column 14 has 20 stages. Feed 30 comprises a mixture containing about 75% MeOAc, 23% MeOH and 2% water, and is fed at 100 kg.h−1 into stage 12 into catalytic distillation column 12 at 100° C. The pressure under which feed 30 is supplied is 10 atm. The column operates at 5 atm. and so feed 30 is supplied via a back-pressure regulator (not illustrated). The reflux ratio is 2 and the distillate-to-feed ratio is 0.45. At steady state, the temperature at stage 3, at the top of the reaction zone is 31° C. and at the bottom of the reaction zone, stage 16, is 127° C.
First volatile products 44 exiting top portion 18 of catalytic distillation column 12 comprises almost entirely DME (over 99%; Table 3). First liquid bottoms 40 exiting bottom portion 22 of catalytic distillation column 12 comprises over 92% HOAc and the balance is less than 8% MeOAc, with only trace amounts of MeOH, water and DME (Table 4).
First volatiles product 44 comprises close to 100% DME.
First liquid bottoms 40 comprise a mixture of about 91% HOAc, 8.9% MeOAc, 0.04% water, and traces of MeOH and DME.
Distillation column 14 has 10 stages and has a reflux ratio of 7 and a distillate-to-feed ratio of 0.07. First liquid bottoms 40 from catalytic distillation column 12 is fed at stage 5 into distillation column 14 at a temperature of 170.5° C. and a pressure of 5 atm.
Second liquid bottoms 64 from distillation column 14 is substantially pure HOAc (see Table 4). Second volatiles product 60 is rich in MeOAc and is recyclable.
Ge, X., Wang, Z, Yong, J., Yang D., “Kinetics of Heterogeneous Hydrolysis of Methyl Acetate” Chemical Reaction and Engineering and Technology (Chinese Journal), Vol. 14, No. 2, 1998, pp 138-144.
Gmehling, J., Bölts, R., “Azeotropic Data for Binary and Ternary Systems at Moderate Pressures” Journal of Chemical Engineering Data, Vol. 41, No. 2, 1996, pp 202-209.
Marten, F. L., “Vinyl Alcohol Polymers” in Kirk-Othmer Encyclopedia of Chemical Technology, John Wiley & Sons, Inc.
Nisoli, A., Malone, M. F., Doherty, M. F., “Attainable Regions for Reaction with Separation” American Institute for Chemical Engineering Journal, Vol. 43, No. 3 (2), 1997, pp 374-387.
Teshima, T., Hiyoshi, S., Matsuda, H., Monma, S., and Iwaba, S., “Extractive distillation. II. Vapor-liquid equilibrium relation of methyl acetate-methanol-water system” Vol. 55, 1952, pp 801-803).
Xiao, J., Liu, J., Juntai, L., Jiang X., and Zhang, Z., “Increase in MeOAc conversion in PVA production by replacing the fixed bed reactor with a catalytic distillation column” Chemical Engineering Science, 56, 6553-6562 (2001).
| Number | Date | Country | |
|---|---|---|---|
| 61088468 | Aug 2008 | US |
