Extended Abstract. Preliminary project of two pilot units destined to investigation and production of catalysts.

Transcription

1 Extended Abstract Preliminary project of two pilot units destined to investigation and production of catalysts. Manuel Ângelo de Brito Lopes Sintra Delgado Instituto Superior Técnico, Departamento de Engenharia Química e Biológica, Lisbon, Portugal October 2008 Abstract Throughout this Extended Abstract, production processes of catalysts destined to polyol, polyurethane and other products synthesis were summarized. These products are important for the Construction and Automotive Industries, among many others. The catalysts researched were Stannous Octoate, Cobalt Octoate organometallic and a bimetallic catalyst, Zinc Hexacyanocobaltate complex. After the processes research, two pilot units preliminary projects were made, based on two patents previously chosen and with the state of the art loop reactor technology, for the investigation and production of the desired catalysts. These preliminary projects were included in an application form, for the Incentives System for Research and Technological Development, between firms in co-promotion, under the guardianship of Quadro de Referência Estratégico Nacional (QREN) of Portugal, with a total of investment of about Finally, two laboratory synthesis of the bimetallic catalyst were made. Their granulometry was compared with the patent chosen and with a sample of the same catalyst from a Chinese manufacturer, using a laser analysis apparatus Eyetech. The average diameter particles size obtained were above the patent figures, but 1

2 below the ones of the Chinese sample, meaning that the process chosen should be the proper one. In the near future, an analysis with a device that allows a previous dilution of samples in an adequate solvent, may improve the results. The installation of the pilot units, with the innovative technology of the loop reactor, will help to develop the investigated catalysts, as well as discover others which can contribute to scientific and technological progress. Keywords: catalysts, polyols, polyurethanes, pilot units, loop reactor, granulometry 1. Introduction Nowadays, the most used catalysts in the catalysis of polymerization reaction of polyurethane foams are organometallic, in which there is a clear link between carbon and metal. These are made of a metal atom, with organic compounds connected to it, called ligands. These monometallic catalysts are particularly intended to obtain the referred foams, elastomers, adhesives and sealants of polyurethanes, because they are highly selective for the reaction that occurs between polyols and isocyanates. Two known examples of such catalysts are dibutyl tin dilaurate (DBTDL) and tin octoate or Stannous Octoate (SnOct). The first is more offen used to obtain adhesives, sealants and elastomers, while the second is more appropriate for flexible foams of polyether polyol, molded or in blocks, besides more rigid foams (e.g. soles of shoes) and clusters of cork. Other known catalyst, Cobalt Octoate (CoOct), is more offen used to obtain polyester resins, which are introduced into concrete, used in the construction industry, to increase their strength and durability; it has, also, other applications, such as obtaining reinforced fiberglass, adhesives or as additives for paints. The bimetallic catalysts are an example of heterogeneous catalysts that, increasingly, are seen as an alternative to monometallic catalysts. Despite being under-used, they have advantages of performance in relation to the other mentioned catalysts due to the synergy between the two active metals. As these catalysts have active centers with two or more metallic elements, there is a cooperation" in order to increase the activity and/or selectivity in relation to the catalysts showing only a metallic element. Generally, the bimetallic catalysts have a catalytic performance superior to monometallic, in terms of activity, selectivity and even in durability. One example of this kind of catalysts is the Zinc 2

3 Hexacyanocobaltate complex (Zn 3 [Co(CN) 6 ] 2 ), a DMC catalyst (Double Metal Cyanide catalyst). The "DMC" catalysts are highly active and therefore contribute to a high speed of reaction of polymerization. They are active enough to allow its use at very low concentrations, in the order of 25 ppm or less [1]. Accordingly, the catalyst can be left in the polyol obtained without affecting the quality of the product, thus avoiding the step of removing the catalyst, which increases the cost of the production process [1]. This aspect reveals a big advantage in using this type of bimetallic catalysts to obtain polyols and, therefore, high quality polyurethanes. 2. Catalysts production processes 2.1 Bimetallic catalyst production processes The majority of the production processes studied for the zinc hexacianocobaltate complex catalyst showed the use of a very well known reactor technology, the Continuously Stirred Tank Reactor (CSTR). All of these processes contain very similar operational conditions (temperatures, pressures, agitation velocities, etc.) and equipment (pressure filters, vacuum filters, vacuum ovens, etc.), varying only in a few details, like the initial mixture of the principal reagents Zinc Chloride (ZnCl 2 ) and Potassium Hexacyanocobaltate (K 3 [Co(CN) 6 ]) - or the way they are mixed with each other. In the last few years the main focus in the scientists investigation was the combination of ligands, the complexing agents of the bimetallic catalyst. Tert-butyl alcohol (TBA) is the most used ligand, because the catalysts obtained with it showed a superior stability and activity when compared with previous ones. Therefore, the combination of TBA with another suitable organic complexing agent causes the obtaining of even more stable, durable and active bimetallic catalysts. Briefly explaining, the production process for this catalyst consists of combining the principal, above mentioned, reagents in a CSTR, at a certain temperature and at a certain agitation speed, for a period of time. These are combined with the ligands individually, or after reacting with each other, to finally obtain a catalytic dispersion. Then, the dispersion is filtered, using a membrane filter press, or a vacuum suction filter, with the objective of obtaining a filter cake. The cake is washed with a solution composed by the mixture of ligands and then filtered again. This process is repeated at least one more time, so that the final catalyst product is free of impurities, which can affect its activity and durability, and also to enhance even further the mentioned activity. Finally the cake is dried in an oven, preferably under vacuum, for a certain period of time. 3

4 Only recently, a few authors innovated in a completely new production process for the referred catalyst, with a different reactor technology [2]. This contemplates the use of a loop reactor which comprises a nozzle jet disperser; it consists of a reagents blender, which mixes them through its injection to a nozzle (mixing chamber), at a certain pressure. This compendium of the loop reactor is prepared by the patent authors so that a short period of residence occurs, in order to not exist a crystal`s formation in the interior of the mixing chamber, just in the flow of exit from it. The diameter of the inside nozzles of this device is also prepared with the intention of accelerating the input current of reagents, for that a proper homogenization of the mixture takes place. The advantages of this type of reactor in relation to the CSTR are, in the first place, to enable energy savings, because the use of blades of large dimensions in the CSTR, for the reagents mixing, causes a greater expenditure of energy. In addition, CSTR has the great disadvantage of the formation of foams during the vigorous shaking of these solutions, which results in reduction of yield and activity of the catalysts obtained. Also, the savings in equipment is considerable, as the CSTR is significantly more expensive than a loop reactor [2]. This latest reactor technology was the chosen one to carry out the preliminary pilot unit s projects, due to the mentioned advantages. These preliminary projects were included under the QREN (Quadro de Referência Estratégico Nacional), a governmental entity that applies an existing system of incentives on projects with companies in copromotion, in order to encourage the research and technological development in Portugal. These preliminary projects are described further in the corresponding chapter. 2.2 Organometallic catalyst production processes The production processes studied for Stannous Octoate (SnOct) and Cobalt Octoate (CoOct) are all also very similar, like it was said for the bimetallic catalyst. This is because these production processes are being widely used today oxidation processes - and these catalysts are also widely used in the production of polyurethanes and other important products. So, the patent chosen for the production process of the referred organometallic catalysts [3], contains one of the latest advances in obtaining them, while the others studied are older and do not have any kind of breakthrough or technological innovation. Briefly explaining, this process consists of adding to a CSTR reactor the principal metal reagents for obtaining the SnOct and CoOct catalysts, in the form of tin powder (or cobalt powder, because the process is the same, only the metal reagent changes in the beginning) and tin shot (a kind of granules) or cobalt shot, 4

5 depending on the final desired catalyst. These are introduced into the CSTR with 2- ethylhexanoic acid, which acts as a carrier of the reaction promoter, 4-tertbutylcatechol (a kind of a catalyst of this reaction). It is also used as one more carrier of the promoter, dipropylene glycol, so that it better promotes the reaction of oxidation. This synergy of carriers is the main breakthrough of the process chosen, and it is used in order to obtain organometallic catalysts more quickly and with greater quality and activity, without the recurrence of many and varied methods of separation and purification, which greatly increases the costs of production of these catalysts. After hours of oxidation, first with air (10 hours), and then with nitrogen (2 hours), the reaction mass obtained is then decanted to remove the tin or the cobalt not reacted, and then filtered as to remove small quantities of tin or cobalt oxide, which could affect the catalyst. Finally, the reaction mass is subjected to a vacuum stripping - removal of excess acid not reacted - to get the final catalyst. This process can be performed in the same type of reactor used in the patent of the bimetallic catalyst - loop reactor - but, because it is an oxidation reaction, it must be performed in a stainless steel reactor, for the reason that is a gas-liquid reaction with corrosive components. In the patent of the "DMC" catalyst, the kind of reaction is liquid-liquid, so it is feasible in a loop reactor built in glass. The advantages of using the loop reactor to the detriment of the CSTR reactor have been described previously, and it was chosen to use this innovative reactor in the production of the catalysts studied, which is described below in the body of the preliminary projects. 3. Preliminary projects of the two pilot units 3.1 Bimetallic catalyst pilot unit Based on the production process chosen [2], briefly explained on section 2.1, it was idealized a pilot unit for the production of zinc hexacyanocobaltate complex catalyst. This can be visualized in the following figure: 5

6 710,4 g Distilled TBA water 50 L 40 L 640 g K3[Co(CN)6] 18 L Cholic acid sodium salt 5,5 L 6,2 L Distilled water 20 L 9,5 L H2O DMC dispersion H2O 8,7 L Jet Loop Reactor (JLR) 875 g Filter press ZnCl2 Drying of the final filter cake in a vacuum oven, at 100ºC, for 5 hours Washing of the filter cake in JLR and new filtration (2X) catalyst. Figure 1. Layout of the pilot unit for the production of the bimetallic As it can be seen in the schematic figure, this pilot unit consists of several storage tanks for the principal reagents of this reaction. The distilled water is pumped by dosing pumps from its tank to the ZnCl 2 and K 3 [Co(CN) 6 ] storage tanks, which contain mechanic agitators. Here, the aqueous solutions are homogenized before being introduced to the loop reactor (a 50 L pilot glass reactor). After the first reaction, the TBA aqueous solution is made from blending a certain quantity of it with distilled water in a pipe conduct, before being also introduced in the reactor, where another reaction takes place. Finally, occurs a similar blending of an aqueous solution of cholic acid sodium salt (another ligand), from the storage tank, where it is homogenized with another mechanical agitator, with TBA, in the same pipe conduct as before, and this mixture is also introduced in the reactor to finally form a catalytic dispersion. This dispersion is stored temporarily in a tank, before being filtered in a filter press. Then, the moist catalytic cake obtained is pumped into the reactor where it is washed by circulating it in the loop reactor with aqueous solutions of the mentioned ligands. This process of filtering and washing the catalytic cake is made two more times. In the end, the catalytic cake is dried in a vacuum oven for 5 hours. The product obtained is a very thin white powder. The mass balance calculations made for this process where based on one of the 6

7 examples of the patent chosen, with a base of 1 Kg of final product obtained. The results are shown in the following table: Table 1. Total quantity of each reagent necessary, in Kg, to obtain 1 Kg of the bimetallic catalyst. Reagents used on the catalyst (Kg) Zinc Chloride 0,875 Potassium Hexacyanocobaltate 0,710 Tert-Butanol (TBA) 37,367 Cholic acid sodium salt 0,638 Distilled water 47,372 It is important to note that this process, besides the time necessary to dry the final catalyst, has a total time expenditure of about 10 hours, which enables the possibility of producing at least 2 Kg of the bimetallic catalyst per day, in two different shifts, in a 50 L glass loop reactor. 3.2 Organometallic catalyst pilot unit This process was based on what was explained on section 2.2, and it is destined to produce the two desired organometallic catalysts, SnOct and CoOct. This can be visualized in the following figure: Tin or Co shot Vacuum pump Dipropylene glycol 4-tert-butyl catechol Stripping Decanter 2- ethylhexanoic acid Air or N2 H2O Filter press H2O Final product Jet Loop Reactor (JLR) Tin or Co powder Figure 2. Layout of the pilot unit for the production of the organometallic catalysts, SnOct and CoOct. 7

8 As can be seen in the schematic figure, the 2-ethylhexanoic acid is pumped from its storage tank to Tin or Cobalt powder and shot tanks, where the respective solutions are homogenized through mechanical agitators. Then, these solutions are pumped into the loop reactor, alongside dipropylene glycol, which is isolated from the others (small quantity compared with the other reagents), where the first oxidation reaction occurs with air. Afterwards, 4-tert-butilcatechol is introduced into the reaction and air is substituted by nitrogen, and the second oxidation reaction takes place. The reaction mass obtained is then decanted to remove the tin or the cobalt not reacted, and then filtered as to remove small quantities of tin or cobalt oxide, which could affect the catalyst. Finally, the reaction mass is subjected to a vacuum stripping - removal of excess acid not reacted - to get the final catalyst. Due to the high efficiency of the loop reactor technology, it is expected a reduction in the total production time of the catalyst to about 11 hours, instead of the more than 12 hours expected with a CSTR. The mass balance calculations made for this process where based on one of the examples of the patent chosen, with a base of 1 Kg of final product obtained. The results are shown in the following tables: Table 2. Total quantity of each reagent necessary, in g, to obtain 1 Kg of Tin Octoate catalyst. Reagents needed mol g Tin (Sn) 9, ,64 2-ethylhexanoic acid 11, ,12 4-tert-butylcatechol 0,03 5,51 Dipropylene glycol 0,06 8,15 Air d (g/cm 3-100ºC) 0,94 Nitrogen d (g/cm 3-155ºC) 0,80 Volume (L) 0,58 0,68 Table 3. Total quantity of each reagent necessary, in g, to obtain 1 Kg of Cobalt Octoate catalyst. Reagents needed mol g Cobalt (Co) 18, ,54 2-ethylhexanoic acid 11, ,44 4-tert-butylcatechol 0,03 5,52 Dipropylene glycol 0,06 8,16 Air d (g/cm 3-100ºC) 0,94 Nitrogen d (g/cm 3-155ºC) 0,80 Volume (L) 0,58 0,69 8

9 As it can be seen on tables 2 and 3, the quantity needed to produce 1 Kg of each catalyst is very similar, so the loop reactor utilized is the same, being a 50 L stainless steel loop reactor. The CSTR used in the patent chosen is a 6L reactor, so with this 50 L reactor here will be an expected the production of at least 15 Kg of each catalyst per day, on two different shifts. As a final note on this chapter, it is important to mention that these preliminary pilot units projects were put in an application form, for firms in Co- Promotion, destined to gain governmental incentives for research and technological development in Portugal. After all the calculations made for all the equipments, reagents, personnel and many other important technical and economical factors involved in a two and a half years research project, the total investment presented was of about Laboratory synthesis of the bimetallic catalyst After the preliminary projects were made, two laboratory synthesis of the zinc hexacyanocobaltate complex catalyst were made, following one of the examples (with a CSTR) of the chosen patent for its production process. Their granulometry was compared with the results of the patent and with a Chinese sample of the same product. It was obtained using a laser image analysis apparatus- Eyetech. The particles average diameters obtained were of µm, µm and µm for the Chinese, first and second synthesis, respectively. As the standard deviation for the average diameter of these samples analyses were between 60,7 µm and 65,8 µm - too high -, this means that this laser analysis apparatus cannot give an accurate comparison with the values of the patent, which presents an average diameter of particles analyzed of 4,51 µm. A reason for this situation is that, when the samples were introduced inside of the apparatus for evaluation, there was an agglomeration of particles, which can mask the actual individual size of them. A future analysis in a device that allows a previous dilution of samples in an adequate solvent could reveal an even closer or similar result to that obtained by the authors of the chosen patent. As a final note, it can be said that the lowest mean diameter values obtained for the bimetallic catalyst synthesized in the laboratory, in relation to the Chinese catalyst, demonstrate that the process chosen should be the proper one. 5. Conclusions As it was written, throughout this Extended Abstract preliminary projects of two pilot units for the production of 2 organometallic and 1 bimetallic catalysts were 9

10 made, based on patents search of their production processes. The search focused primarily on patents for three catalysts, two well-known and widely used on an industrial scale, Tin or Stannous Octoate and Cobalt Octoate - organometallic catalysts - and other, less common, a bimetallic catalyst - the zinc hexacyanocobaltate complex catalyst. The latter was the catalyst most researched, because the information on it was scarce. It is an innovative product that is being currently, increasingly, used in obtaining polyols and polyurethanes, due to the advantages that shows on the level of synergy of the two metal atoms that it presents; it shows to catalyze the polyols synthesis reactions, one of the reagents to obtain polyurethanes, faster and more efficiently, with higher yield and quality, when compared with the monometallic catalysts. In relation to reactor technologies, the loop reactor described allows higher yields than the CSTR, with a high utilization of all reagents, resulting in fewer waste products and a final product of better quality. Thus, the expense of raw materials is lower and this kind of technology also permits a saving of energy in relation to the CSTR. For all these advantages and because this innovative technology is adaptable to oxidation reactions, it was decided to use the same reactor technology for the production of organometallic catalysts. It is considered that the importance of this work will be revealed in the progress of the small and medium enterprises involved, with the production of better quality and innovative products, with great use in the industries branch of Construction, Automotive, among others. The laboratory synthesis of the bimetallic catalyst revealed that the catalyst obtained has a very similar look to the one from China, in the form of a very thin white powder, but with lower average diameter size of particles. So, the operational conditions used, based on the patent chosen, produce the same product, but with better results in terms of size, demonstrating that the process chosen should be the proper one. In the future, with the approval of the application delivered to QREN, it is possible to proceed with the installation of the pilot units, after having dimensioning their equipment and studied the process in more detail, with the innovative technology of the loop reactor. Then, it will be possible to proceed in the research and production of the catalysts studied. With this work, the development of products of increasing quality and the discovery of new and more efficient catalysts, also with the patenting of new production processes, with higher yields, less energy expenditure and little waste; it is hoped that it will contribute to the ultimate objectives of scientific and technological progress. 10

11 6. References [1]. ELEVELD, Michiel Barend, et al. - DMC Complex Catalyst and process for its preparation. United States Patent US 6,699,961 B2, 2 Mar [2]. HOFMANN, Jörg, et al. Method for producing Double Metal Cyanide (DMC) Catalysts. United States Patent US 6,835,687 B2, 28 Dec [3]. KNEZEVIC, Vasilije, et al. Direct Synthesis of Tin (II) Carboxylates and Tin (IV) Carboxylates from Elemental Tin or Tin Oxides. United States Patent US 6,303,808 B1, 16 Oct [4]. LE-KHAC, Bi; WANG, Wei - Double-Metal Cyanide Catalysts which can be used to prepare Polyols and the processes related thereto. United States Patent US 7,223,832 B2, 29 May 2007 [5]. SUZUKI, Chitoshi, et al. Double- Metal Cyanide Complex Catalyst, its production process and its utilization. United States Patent US 7,169,956 B2, 30 Jan [6]. LE-KHAC, Bi, et al. Highly active Double- Metal Cyanide Complex Catalysts. United States Patent US 5,712,216, 27 Jan [7]. HOFMANN, Jörg, et al.- Zinc/Metal Hexacyanocobaltate complex compounds, a process for their preparation, and their use in a process for the production of Polyether Polyols. United States Patent US 5, 998, 327, 7 Dec [8]. LE-KHAC, Bi Process of making an Epoxide Polymer using highly active Double-Metal Cyanide Catalysts. United States Patent US 6,211,330 B1, 3 Abr [9]. HOFMANN, Jörg, et al.- Process for producing DMC Catalysts. United States Patent 6,780,813 B1, 24 Aug [10]. ELEVELD, Michiel Barend, et al. Process for the preparation of Polyether Polyols. United States Patent US 7,300,993 B2, 27 Nov [11]. HOFMANN, Jörg, et al. - Double-Metal Cyanide Catalysts for preparing Cyanide Catalysts Polyols. United States Patent US 6,833,431 B2, 21 Dec [12]. HOFMANN, Jörg, et al. Method for producing Double Metal Cyanide (DMC) Catalysts. United States Patent US 6,835,687 B2, 28 Dec [13]. HOFMANN, Jörg, et al.- Process for producing DMC Catalysts. United States Patent 6,780,813 B1, 24 Aug [14]. KNEZEVIC, Vasilije, et al. Direct Synthesis of Tin (II) Carboxylates and Tin (IV) Carboxylates from Elemental Tin or Tin Oxides. United States Patent US 6,303,808 B1, 16 Oct [15]. KOVSMAN, E.P., et al.- Synthesis of Metal 2-Ethylhexanoates. United States Patent 6,033,551, 7 Mar