mình xin giới thiệu thêm một phương pháp để tổng hợp ethylacetat.
Steve Colley describes work to develop a new route to make ethyl acetate starting from low grade renewable feedstocks
In 1984, during work to develop the company's new 'oxo-alcohols' process, chemists at Kvaerner Process Technology in Stockton-on-Tees hit a problem. While developing catalysts for the liquid-phase hydrogenation of butanal to n-butanol, KPT chemists found that a significant amount of butyl butanoate ester was produced as a byproduct. They had two challenges: how to convert the ester to n-butanol, and how to do it economically. Hydrogenation of the ester seemed an obvious solution.
For good conversion, the necessary high temperatures and pressures (250°C, 2.7 x 107Pa) would make liquid phase hydrogenation prohibitively expensive. Instead, they investigated the possibility of hydrogenating the ester in the vapour phase. Surprisingly, hydrogenation at low pressure and moderate temperatures over copper-chromium catalysts gave almost quantitative conversion of the ester. Realising the potential of this novel technology, they soon began applying the method to other esters. It was during work to investigate this and the reverse dehydrogenation process that KPT researchers inadvertently hit upon a new process for making ethyl acetate from renewable low grade ethanol starting materials. This process was recognised as a runner up of one of the RSC's Industrial Innovation Team awards last year (Chem. Br., January 2001, p53).
Early development
Ethyl acetate is a chemical in big demand - raising over £600m in world sales (production plus imports) last year. About 60 per cent of this is used in coatings applications, with the remaining 40 per cent being used as solvents for printing inks, organic syntheses and adhesives and cosmetics. Despite restrictions on levels of VOCs, ethyl acetate is a useful substitute for hazardous air pollutants such as methyl ethyl ketone (MEK) and toluene in coatings. In addition, the growth of coatings containing higher levels of solids such as epoxy resins, polyesters and urethanes has increased the demand for polar solvents such as ethyl acetate.
Kvaerner's new route for ethyl acetate synthesis came about as a result of observations during early investigations with our copper-chromium catalysts. In certain conditions, we found that the primary product formed when an alcohol was passed over these catalysts was the corresponding symmetrical ester. Preliminary studies of this dehydrogenation process showed that we could produce ethyl actetate from ethanol by this route (Scheme 1). Nevertheless, development of the process was shelved after work on the other spin-off technologies attracted more interest from customers, and it was held as a 'technology-in-waiting'.
http://www.chemsoc.org/chembytes/ezi...aug01_sch1.gif
http://www.chm.bris.ac.uk/ms/departm...yl-acetate.gif
Scheme 1. Synthesis of ethyl acetate from ethanol
From the start, we realised that the process would not be the universally preferred route to ethyl acetate - typically made by the esterification of acetic acid with ethanol or as a byproduct from the liquid-phase oxidation of n-butane. Rather, this would be an alternative process that uses low purity feed (eg Fischer-Tropsch derived ethanol), being most viable where cheap ethanol was readily available and acetic acid was not (eg Brazil, South Africa). This could be even more important in the future, when pressure to use more renewables could see ethyl acetate being manufactured from bio-derived ethanol, eg corn steep liquor. Even with the market position identified, there was no development beyond the initial four month viability study until a client to whom we could license the technology, Sasol, was brought on board in 1996.
Early that year I was given the task of developing this embryonic alcohol dehydrogenation technology. This work was carried out at KPT Laboratories in Stockton-on-Tees. I was helped by a chemical engineer, Mike Tuck, who was based at our engineering office in Paddington, London. Tuck's overall task was to convert my laboratory data into a preliminary design for a 50,000t per annum plant. The stated target was to synthesise high purity ethyl acetate (>99.8 per cent) in >90 per cent yield from a relatively low grade ethanol byproduct. At the time of the development, the major use for this ethanol was as automotive fuel. Among a plethora of ethers and low boiling hydrocarbons, the ethanol also contained isopropanol (iPA), which is difficult to remove by simple distillation. While iPA and the other contaminants have little significant effect on the ethanol when it is used as a fuel, we had no way of knowing what effect they would have on our dehydrogenation process.
Once we started to investigate the process and plan the testwork, problems began to surface. The first was that of product separation. Ethyl acetate and ethanol form azeotropes - liquids whose boiling point, and hence composition, does not change as vapour escapes on boiling. Being unable to distil the pure product meant that to obtain good yields we needed to convert as much as possible of the ethanol to ethyl acetate. The second problem was the selectivity of the process to ethyl acetate. Preliminary work had shown a relationship between ethanol conversion, selectivity to ethyl acetate and operating pressure - high pressure gave better selectivity but lower conversion and vice versa. With these problems in mind, Tuck set out to devise a method of separating the ethyl acetate product, while I began to optimise the dehydrogenation step.
My initial approach was based on work carried out 12 years earlier, which showed that copper-based materials would be the most effective and least expensive catalysts. To avoid problems of catalyst synthesis and reproducibility, I decided to investigate only catalysts that were manufactured at commercial scale. Heterogeneous catalysis, in which the catalyst and reactants occupy separate phases, is something of a black art when compared with homogeneous catalysis. Small changes in the physical properties of catalysts often have large unexpected effects on conversion and selectivity. I chose the reaction conditions such that the feed material and products would remain in the vapour phase at all times. These restrictions defined an operating envelope that was bounded at one extreme by the need to vaporise the feed and product, and at the other by the upper temperature (250°C) tolerated by the copper catalyst.
Having extensively researched vapour phase hydrogenation, KPT had built up a large library of copper-based catalysts. Our starting assumption was that because dehydrogenation and hydrogenation were taking place in the same reactor, hydrogenation catalysts would be a good starting point for dehydrogenation. From this base, we chose six catalysts, spanning the range of physical properties such as pore volume, surface area, active metal area and so forth. We carried out tests in a small reactor built in-house for the purpose.
I started by testing a catalyst previously used for vapour phase hydrogenation of maleate esters to 1,4-butanediol. At this stage, I used pure ethanol as feed, so that we could subsequently use this as a baseline from which to identify any differences caused by iPA or other contaminants. After collecting and analysing the first product from the dehydrogenation reactor, I was relieved to find that it did contain a significant quantity of ethyl acetate. The sample also contained unwanted byproducts such as diethyl ether, acetaldehyde, acetaldehyde aldol products, higher esters and ketones, notably MEK. Subsequent work to optimise the reaction conditions was disappointing; the best results for selectivity and conversion were far below those required for an economically acceptable process. Screening the range of catalysts gave more or less similar results - there appeared to be some common factor that was producing unacceptably high levels of byproducts.
By mid-1996 it had become clear that our initial thinking was incorrect: catalysts that were good for hydrogenating esters to alcohols were not necessarily good for the reverse reaction. Analysing the mechanism of ethyl acetate and byproduct formation indicated that the catalysts we were using were too active. Instead, we started to look further afield at copper-based catalysts that had been rejected during the initial ester hydrogenation testwork. Unfortunately, many of these catalysts had also been rejected due to high byproduct formation. There remained one catalyst, a pelleted version of Raney copper (an active hydrogenation catalyst) that had not been tested due to its very low surface area and high cost. The expectation was that, because of its low surface area, the catalyst would have low activity. More in hope than expectation, we loaded the catalyst into the reactor for testing. We were pleasantly surprised by the results. Though the overall selectivity for ethyl acetate was low - 65-70 per cent - the forest of byproducts seen with the other catalysts had reduced to one major (diethyl ether) and several minor byproducts. Looking at the catalyst revealed significant concentrations of aluminium at the surface, and it was at these sites that the diethyl ether was being made. It seemed that we were making progress.
Armed with the data from the Raney copper catalyst, we began more concerted testing. In late 1996 a team of three KPT chemists was assigned to the project, and began to carry out the tests on a larger millilitre scale. They started by testing catalysts that had properties similar to the Raney copper. After several cycles of testing and re-evaluation, two promising candidates were chosen. By early 1997, they had increased selectivity to ethyl acetate from 70 per cent to 94 per cent, and optimised conversion at 40-45 per cent. It seemed that if we could solve the problem of purification, the project had a good chance of success.
Meanwhile, Tuck and his engineering team had been investigating a number of possibilities for purifying the ethyl acetate product. Distillation using water had been considered and rejected as too complex and expensive, but a new approach using a technique called pressure swing distillation looked attractive. KPT had previously used this technique for another difficult separation involving ethanol. The technique takes advantage of the fact that the composition of an azeotrope varies with pressure. The composition of the ethanol-ethyl acetate azeotrope swings from largely ethyl acetate at atmospheric pressure to largely ethanol at higher pressure. The practical upshot of this is that if we fed the distillate from the atmospheric pressure column to the high pressure column, the distillate from the high pressure column contains less ethyl acetate. The excess passes out of the bottom of the high pressure column. The distillate from the high pressure column could then be recycled to the low pressure column. It appeared that purification was possible.
Predictions made from simulations of the pressure swing columns carried out by Tuck threw up one final problem. Two of the byproducts, acetaldehyde and MEK, which are present in all dehydrogenation products, were appearing in the final ethyl acetate product. Experience in distilling aldehydes led us to believe that acetaldehyde would form acetals in the overheads of the low pressure column. These acetals would then decompose in the high pressure column, contaminating the ethyl acetate product with acetaldehyde. We predicted that MEK would form an azeotrope with ethyl acetate and ethanol, and contaminate the ethyl acetate product. MEK was seen at a concentration of 0.1-0.2 weight per cent in the crude dehydrogenation products, and at this level the final ethyl acetate product would contain 0.3-0.6 per cent MEK. Commercially available ethyl acetate has a specification of not more than 50ppm aldehydes and ketones. This implies a maximum content of 20ppm MEK in the feed to distillation - a tough target.
Closer examination of the distillation simulations hinted at a solution to this problem. The corresponding alcohol of MEK, 2-butanol, was observed in the crude dehydrogenation product, but was not predicted to form an azeotrope with ethyl acetate. Separation by distillation was therefore possible. This led us to consider selective hydrogenation of MEK to 2-butanol. The approach had the added advantage that if selective hydrogenation of MEK was possible, then acetaldehyde would also be hydrogenated, solving both problems at once.
When thinking about the problem of selective hydrogenation we reconsidered the aldehyde hydrogenation process that had kick-started the entire project. One catalyst used in the hydrogenation of butanal was nickel based, and this seemed to be an obvious starting point. A new reactor was designed and built; the process now called for the product from dehydrogenation to be passed to this new reactor before it was sent for distillation.
Catalyst testing, initially using nickel-based catalysts, proved the concept. A programme of screening other potential selective hydrogenation catalysts led us to identify other, more effective catalysts. The chosen catalyst is highly selective: at moderate pressure and low temperature MEK levels of <10ppm are routinely seen, with losses of ethyl acetate of less than 1 per cent. Needless to say, the more reactive acetaldehyde was also hydrogenated back to ethanol.
Tuck's simulations led to the design of pilot scale distillation columns just 2.5cm in diameter. The dehydrogenation and selective hydrogenation reactors were run continuously and the product collected. Using these columns, KPT distilled ethyl acetate of the required purity in mid-1997.
Proving the process
By early 1998 we had enough information to pull the threads together. So far, testing of the catalysts had been carried out over relatively short times and using pure ethanol. Catalysts for industrial use are expected to have a lifetime of at least 12 months. A continuous run of at least 1000 hours (6 weeks) is needed to gather sufficient data on which to base a prediction. The distillation columns were successful in these initial trials, but over time there is the possibility that a component present in the feed at low concentration could build up and cause problems.
It is usual practice to test catalysts on a fairly small scale (10-100ml) before moving to a traditional pilot unit where large volumes of catalyst are used, often about of 10l or more. This is time-consuming and costly: pilot units often cost a nearly as much as a full scale plant. One way to circumvent the need for the pilot unit is to size reactors so that they represent a 'slice' through what is estimated to be the likely size of the full scale catalyst bed. The aim is to run the reactors in a regime such that mixing and heat transfer is as similar to that of the full scale plant as is practically possible. We were aiming to scale the reactors and distillation units by a factor of ca 70,000, and so the importance of getting this part of the testwork right was very important. The entire process simulation - alcohol dehydrogenation, carbonyl polishing and distillation - was operated continuously for eight weeks.
With process optimisations developed earlier, we were able produce ethyl acetate of 99.95 per cent purity, exceeding our initial targets. The data collected from our 1000 hour run were used to finalise a 'Definition of Technology' (a document that specifies the design of reactors, distillation columns, catalysts etc) and the subsequent full plant design (Fig 1). Process development from bench top to contruction of the first commercial plant had taken less than five years. Sasol began operation of the first 50,000t per annum ethyl acetate from ethanol plant in South Africa in April 2001.
Fig 1. Plant design for ethyl acetate production
Acknowledgements: the following people were involved in developing the ethyl acetate process at the KPT Technology Centre: Gary Bennington, Jay Clarkson, Bill Crallen, Paul Gordon, Bob Harding, Faron Hartley, Richard Humble, Simon Jackson, Iain Milne, Paul Nicholls, Chris Skinner, Anita Sampson, Chowdry Sharif and Paul Willett.
Steve Colley is a research group leader at the KPT Technology Centre, Princeton Drive, Thornaby, Stockton-on-Tees TS17 6PY.
TỔNG HỢP ETHYL ACETATE
Linear Mode
