Chapter 7

Aqueous biphasic hydroformylation

Jrgen Herwig and Richard Fischer

Celanese GmbH, Werk Ruhrchemie. Research & Technology, D-46128 Oberhausen, Germany

7.1 Principles of biphasic reactions in water

7.1.1 Why two-phase catalysis? Scope and Limitations What has to be achieved to combine the major important advantages of homogeneous and heterogeneous catalysis? It seems to be a non-complex problem with a simple solution: find a way of separating the catalyst and the reaction products under mild conditions that is ecologically as well as economically efficient. However, to get there is not that simple. Normally the organometallic catalysts applied in solution are not very stable against high temperatures, water or oxygen. Usually, reaction products, especially bulk chemicals, are separated in two ways. Firstly, by gas phase separation technologies like distillation as applied for e.g. the synthesis of acetic acid by the carbonylation of methanol or the hydroformy lation of alkenes to aldehydes. Secondly, by reaction conditions that keep the reaction partners, starting materials, and the products in the gas phase as e.g. for the gas phase process for the production of phthalic anhydride. In the case of the production of terephthalic acid the product is insoluble in the reaction medium acetic acid, in which the Co and Mn based oxidation catalyst are dissolved. This way it is possible to separate the product from the catalyst solution by simple filtration. However, such beneficial circumstances occur rather seldom, as mixtures of liquid reaction media and liquid products are the rule. Thus it is a major step forward if catalytic reactions can be performed in liquid multi-phase systems with clean and fast separation of
189
P. W.N.M. van Leeuwen and C. Claver (eds.), Rhodium Catalyzed Hydroformylation, 189-202. 2000 Kluwer Academic Publishers. Printed in the Netherlands.

190

Chapter 7

reactant and product phase. See Chapter 8 for an overview of the other separation techniques in use for hydroformylation. In contrast to heterogeneous catalysts, their homogeneous counterparts don't resist high temperatures or harsh work-up conditions mostly due to damage of weak carbon-metal bonds or the loss of ligand stabilization of the central metal atom of the complexes. The avoidance of (traces) of water or air at such conditions is also often a parameter that is critical to the success of the systems. Thus the conditions of the separation of the reaction product from the solution, in which the catalyst has been dissolved, must be as mild as possible. A textbook example of such a modern technology is the application of two- or multi-phase reaction systems. One of the most important developments in 1980-2000 in the area of homogeneous catalysis is the successful development of water-stable as well as highly water-soluble catalyst systems and the consecutive introduction of the aqueous two-phase technology. What limitations have to be accounted for if multi-phase systems are applied is shown in the following listing of possible disadvantages: - low reaction rates and the resulting low space-time yields due to insufficient solubility of the reactants or complex interface processes such as mass-transport limitations, - surfactant effects that might favor the enrichment of the reaction product or side products in the reaction media, - slow or incomplete phase separation, - the need of high reaction volumes, - corrosion effects, - decrease of the chemoselectivity (e.g. pH-controlled side-reactions, etc.). 7.1.2 Concepts for two-phase hydroformylation Basis for two-phase catalysis is the presence of two immiscible phases. These phases can consist of: - water (+ co-solvent) - organic liquid - polar organic liquid - non polar organic liquid - ionic liquid - organic liquid - fluorous organic phase - organic liquid The first concept has been realized in the Ruhrchemie-Rhne Poulenc process for propene and butene hydroformylation. This process will be discussed in detail below. The catalyst is retained in the aqueous phase by applying a water-soluble phosphine ligand. The product constitites the organic solvent. The second concept is realized in the Shell Higher Olefins Process for the oligomerization of ethylene to linear higher alkenes. Butane-1,4-diol is used

7. Aqueous biphasic hydroformylation

191

as the solvent for the Ni based catalyst. The linear higher alkenes are not soluble in the catalyst phase and can thus easily be separated. This concept is not applied in hydro formylation. The third concept is used in an IFP process for the dimerisation of alkenes. Hydroformylation has been successfully conducted in ionic liquids in the laboratory [ 1]. Fluorinated organic liquids, concept four, have not been applied in industry, because of technical and economic disadvantages (expensive ligands and solvents). Hydroformylation has been investigated in fluorinated solvents [2] (see Chapter 10).

7.2 Hydroformylation of propene and butene

7.2.1 Historic overview of two-phase hydroformylation technology In 1982, Rhne Poulenc and Ruhrchemie AG (now part of Celanese AG) joined to develop a continuous aqueous biphasic hydroformylation process. The work was driven by the fact that Rh was much more efficient in hydroformylation of short chain alkenes than cobalt; Rhodium gives lower raw material consumption, lower cost for catalyst recycling, and a higher 1:b ratio. In order to stay competitive Ruhrchemie as a leading oxo producer decided to switch to a Rh based system. Furthermore, a two-phase system could offer several advantages over a homogeneous system e.g. lower catalyst loss, lower heat stress of the catalyst, easy catalyst separation from the products and higher energy efficiency. Following the laboratory results of several, biphasic catalytic reactions (hydroformylation, hydrocyanation, and diene conversion) performed by E. Kuntz and patented by Rhne Poulenc [3] the process work started. In only two years time these laboratory results were transferred to a 100,000 t/a plant in Oberhausen by a team managed by B. Cornils. The first aqueous biphasic propene hydroformylation reactor came on stream in 1984. 7.2.2 Ligand developments The water-soluble ligand used by Kuntz and Cornils is TPPTS [tri(msulfonyl)triphenylphosphine trisodium salt], the properties of which are very similar to the parent compound tpp. Sulfonation of triphenylphosphine, neutralization, followed by pH-controlled selective extraction and reextraction procedures, yields an aqueous solution of the trisodium salt of tris(m-sulfonyl)triphenylphosphine in good yields. The catalyst used is a

192

Chapter 7

rhodium complex of TPPTS, which is highly water-soluble as in the order of 1 kg of the ligand "dissolves" in 1 kg of water. The ligand forms complexes with rhodium that are most likely similar to the ordinary triphenylphosphine complexes (i.e. RhH(CO)(PPh3)3) [7].

Figure 1. The RCH/RP ligand for two-phase hydroformylation

Until today the ligand synthesis is object of permanent improvement and investigation. Described by a quite simple reaction scheme (see below), the sulfonation technology and the consecutive work up procedure require significant know-how [17]. P(C6H5)3 + 3 SO3 P(C6H4SO3H)3

Almost two decades of experience with the technically applied system demonstrated that it is of great importance to reduce by-product formation during the ligand synthesis: not only the improved yields but even more so the elimination of specific reaction impurities bear a large potential for further economic savings. Ligand decomposition can lead to the formation of inactive rhodium species. Ligand decomposition can take place via P-Cbond cleavage, followed by formation of propyldi(sulfonylphenyl)phosphine, which acts as strong electron donor reducing the activivity of the rhodium catalyst. To prevent fast catalyst deactivation [18] it turned out be beneficial to control the P(III)/Rh ratio very carefully and never let it drop below 60/1. In addition it was found that an almost continuously performed addition of fresh ligand solution increases the catalyst life time (catalyst activity and productivity) more than a third compared to the nominal life time [19]. This pseudo-continuous ligand addition mainly compensates the formation of phosphine oxides formed by traces of oxygen brought into the system by the feedstock and the syn gas, which is produced by partial oxidation of heavy oil or natural gas. Besides the TPPTS-system a number of other sulfonated phosphine ligand systems were investigated and tested on a pilot plant scale. Among them are systems which are derived from biphenyl (BISBIS = sulfonated bis(diphenylphosphinomethyl)biphenyl, varying grades of sulfonation of

7. Aqueous biphasic hydroformylation

193

BINAS

Figure 2. Formula of BINAS

BISBI, see Chapter 4) or binaphthyl structures (BINAS = sulfonated NAPHOS, sulfonation grade between six and eight). It turned out that rhodium-BINAS is the most active and selective watersoluble hydroformylation catalyst [20]. Its reactivity is up to 10 times higher than TPPTS, even low Rh/P-ratios (P/Rh = 7/1) give 1:b ratios of 98/2 (TPPTS: 94/6 with P/Rh = 80/1). Concerning the outstanding performance data, BINAS would be clearly the ligand of choice. Compared to TPPTS, these more complicated systems carry higher manufacturing costs. This disadvantage could be compensated for by the lower amount of ligand needed More importantly, BINAS shows higher decomposition rates than TPPTS, which was up to now the overriding reason not to change the established system. As a consequence up to now TPPTS is the best choice for technical application in water based two-phase catalyst systems. 7.2.3 Kinetics and catalyst pre-formation Kinetics. When two-phase technology is applied, phase transfer phenomena can play a dominant role in determining reaction rates and kinetics. The hydroformylation of alkenes in a two-phase water/organic solvent system is defined by the reaction of two gases, carbon monoxide and hydrogen, with gaseous or liquid alkenes like ethene, propene or higher homologues. Dependent on mass transport and reaction conditions (temperature, partial pressures of the gases, stirring) and reactions partners (type of alkene), the reaction of the alkene with CO and H2 takes place at the organic-aqueous interface or in the aqueous bulk phase. No published data are available for propene or butene. One study has been carried out using octene-1, which is easy to handle, thus yielding reliable results for the derivation of kinetic data, although solubilities change with the composition of the system. Using the Rh/TPPTS catalyst in an aqueous solution a rate law was determined for the two-phase

194

Chapter 7

hydroformylation of 1-octene to nonanal. The reaction rate was found to be first order with respect to catalyst concentration. Also the dependency on the alkene concentration is first order. At pressures below 5 bar a positive order in the partial pressure of hydrogen pressure was found, but at pressures above 10 bar this levels off to much lower values. Likewise, at pressures below 5 bar a positive order in partial CO pressure was found, but at pressures above 5 bar the usual retarding effect of increasing pressure was found (Chapter 4). The rate law for the hydroformylation of 1-octene in an aqueous two-phase system was described with the following equation, based on the accepted mechanism of hydroformylation [4]: k K1K2K3 [Rh ][ octene ][ H2 ][ CO ] 1+ [ CO ]

R=

with k = reaction rate constant, K1, K2, K3, K4 = constants. At higher pressures of dihydrogen the graphs seem to indicate a leveling off of the rate increase with hydrogen pressure. If so, the rate-determining step would be similar to the TPP system (Chapter 4). For lower hydrogen partial pressures the following rate expression was developed (hydroformylation of n-octene- 1 with ethanol as co-solvent to increase the octene solubility [5]): Ro= k [ alkene ]o[Rh ][ H2][CO ] (1+KH2[H2])(1+Kco[CO]2)

The mixing of the reaction phases is also important for the reaction rate so as to provide maximum liquid-liquid interfacial area in cases where the reaction occurs mostly at the liquid-liquid interface and not in the bulk aqueous phase. This can be achieved by high agitation speeds, optimized stirrer type and geometry. In aqueous two-phase systems also the hold-up of the aqueous phase plays an important role on the reaction rate. With too high hold-up volumes the reaction rate starts to decrease (phase inversion yielding the dispersed phase as reaction place which would be, in this case, the organic phase). Thus it is recommended to keep the aqueous catalyst solution at an optimized level to provide high productivity with regard to the aldehyde formation. When using co-solvents like lower alcohols, polyethylene glycol or acetonitrile the biphasic character of the system is retained. The rate of the hydroformylation can be enhanced by several times, albeit at the cost of a lower selectivity as a result of e.g. formation of acetals or ethers.

7. Aqueous biphasic hydroformyla tion

195

Naturally, applying aqueous two-phase systems, the pH-value plays a crucial role for catalyst activity. High reaction rates are observed at high pH values (pH = 10, basic conditions). Also here the influence of catalyst concentration on the reaction rate is highly beneficial (up to five times rate increase with only tripling the catalyst concentration with hydroformylation of 1-octene as an example) [6]. Increase of CO pressure under these conditions is positive on the rate, but at too high CO partial pressures the rate drops. At low pH values (7 and lower, acidic conditions) the catalyst activity is lower. Here an increase of catalyst concentration or CO partial pressure can help only little. Thus hydroformylation is faster at higher pH values, but at high pH-values severe side reactions of the reaction product can occur. For example aldol condensation of butyric aldehydes can occur to form different types of aldols and higher condensed systems that lead to an inferior quality of the finished products like C4- aldehydes, alcohols, or 2-EH (2-ethyl-1hexanol). In summary, the kinetics of hydroformylation in an aqueous/organic two phase system applying water soluble Rh-complexes as catalysts are dependent on the type of the Rh/ligand system, the pH-value, as well as on the usage of additives and co-solvents. In particular the kinetics with respect to the influence of the CO partial pressure varies between the different systems. Of course the solubility of the alkene in the aqueous phase carrying the catalyst plays a major role. Therefore, to achieve economically relevant space time yields, higher alkenes require (i) co-solvents, (ii) catalyst binding ligands like TPP, which facilitates the interfacial catalytic reaction by increasing the concentration of the Rh-species at the interfacial area, or (iii) the application of supported aqueous phase catalysis (Chapter 10). Catalyst preparation / preformation Of outstanding importance for the Rh/TPPTS system is the catalyst preparation and pretreatment prior to the continuous use, in order to guarantee a high catalyst lifetime and sufficient performance over the whole runtime. Thus a carefully controlled TPPTS ligand synthesis, especially the work up of the sulfonation mixtures, is a prerequisite to provide optimal conditions during the hydroformy lation reaction. Prior to the conversion of the alkene, the Rh/TPPTS catalyst system has to be pre-formed (formation of the Rh-hydrido-carbonyl complex) applying a specific chemical procedure is necessary to form the active Rh(1) species, mostly starting from Rh(III) precursors. In general this is achieved under hydroformylation conditions. Thus, treatment of the rhodium precursor with syn gas in the absence of alkene for a couple of hours will transform it into a Rh-hydrido-carbonyl complex.

196

Chapter 7

To achieve high catalyst lifetimes it is beneficial to keep the reaction temperature below a certain ceiling value like 130 C to minimize ligand degradation. To stabilize the active catalyst species it turned out to be efficient if ligand is added by a saw-tooth-type dosing during the whole reaction run. This keeps the Rh/P(III) ratio at a level which minimizes the catalyst degradation (see also chapter 9). At the end of a catalyst life marked by significant and constant drop of productivity - the complete aqueous catalyst solution is taken out of the production system and the reactor is refilled with fresh catalyst solution. The old catalyst will be worked up with full Rh-recovery (>90%) and even recovery of the portion of the not degraded or oxidized TPPTS. 7.2.4 Process description The Ruhrchemie Rhne-Poulene-Process applies a catalyst - TPPTS/Rh which is capable to convert both propene and raffinate-II feedstocks to butyraldehyde and valeraldehyde, respectively. The potential of this process is its flexibility to convert propylene or butenes in the same reactor system. The design of the Ruhrehemie/Rhne-Poulene (RCH/RP) reaction system and catalyst recovery method is completely different from that of homogeneous hydroformylation processes. In principle, the two-phase RCH/RP process applies a stirred tank reactor, followed by a phase separator and a strip column. The reactor containing the aqueous catalyst solution (Rh/TPPTS) is fed with propylene and syn gas. The oxo crude product is passed on to the phase separator where gaseous components are removed and the liquids separated into two phases, the aqueous catalyst solution and the organic aldehyde phase. The catalyst solution is recycled to the reactor passing a heat exchanger system, where steam is generated. The organic phase containing the crude aldehydes is stripped with syn gas to remove dissolved propylene that is recycled to the reactor. The crude aldehyde mixture is split into iso- and n-aldehyde in an aldehyde distillation unit. In the past at the Ruhrchemie site, the portion of the not reacted propene, present in the waste gas from the two phase oxo reactors, was converted to butyric aldehyde in a vent reactor, which was based on high pressure cobalt hydroformylation technology. In 1998 the old cobalt technology was eliminated by the implementation of a high-pressure rhodium process giving a more selective process yielding a cleaner reaction product. The rhodium system allows now the possibility to steer the overall 1:b ratio within a comfortable range to serve the market needs. The capacity of this highpressure rhodium-unit is 70,000 tons per year. The TPPTS ligand easily carries all of the used Rh into the water phase by forming Rh/TPPTS complexes analogous to the complexes known from

7. Aqueous biphasic hydroformylation

197

the simple Rh/TPP system. Due to the sufficient strong coordination of the TPPTS-ligand to the Rh metal-center almost no Rh leaching into the organic phase is observed - a performance prerequisite with respect to the desired economy of the process. Even with the low solubility of propene or butene in water, the high reactivity of the used catalyst system is high enough to provide a very efficient hydroformylation rate. To increase the solubility of propylene in the water phase, plant trials were performed with polyethylene glycol with various chain lengths, acting as an auxiliary-agent for better propylene solubility. Good reaction rates (rate increase of more than 10 %) were achieved, demonstrating a simple tool to increase productivity or to switch to milder conditions. In summary, the major advantages of the TPPTS-system are the simple, automatically performed mild catalyst recycle, the comparable low Rh-inventory (rhodium savings), the high 1:b ratio of ~20/1 and the beneficial low energy consumption of the whole process as well as the improved utilization of propylene and synthesis gas (CO/H2). The water-based process is normally operated at a pH-value between 5 and 6 to provide reaction conditions, which depress unwanted side reactions of the formed aldehydes. It turned out that a careful pH-control is beneficial for various operation parameters of the protic reaction media: catalyst reactivity and selectivity are directly influenced by and thus dependent on the H3O+ concentration present in the aqueous phase. 7.2.5 Status of the operated plants With the RCH/RP process it is possible to hydroformylate alkenes from propene to pentene with satisfying space-time yields. Currently the process is solely used for the conversion of propene and butene-1. In the early 1990s developmental work was started to apply the RCH/RP process also to the conversion of alkenes higher than C3. The overall economics looked attractive and the market was prepared. Consequently, a 12,000 tons unit went on-stream at the Oberhausen site in 1995, to convert raffinate-II (a mixture of butene- 1 and butene-2) into C5-aldehydes that are predominantly converted to the corresponding alcohols and acids in further downstream processes at the site. Compared to propene, butenes show a lower solubility in water. Due to this the observed hydroformylation reaction rate of butene mixtures like raffinate-II (in general 40 % butene-1, 40 % butene-2, 20 % butane isomers) is significantly lower than that found for propene. However, a slight increase in the catalyst concentration is sufficient to achieve satisfying space-time yields for the pro-duction of valeric aldehydes. As a very selective catalyst the Rh/TPPTS-system does not catalyze the hydroformylation of butene-2

198

Chapter 7

providing a simple method for the single conversion of the terminal alkene. The catalyst isomerizes a small fraction of butene-1 to butene-2. The latter is not converted under standard reaction conditions. Thus the butanes and the unconverted butene-2 are separated form the reaction products in a stripping column. Advantageously the RCH/RP technology can be applied to the hydroformylation of C3- as well as C4-alkenes in the same production unit; only minor changes (e.g. reaction temperature, partial pressures) are necessary to switch between the different feedstock streams [2 13. This is not the case for alkenes with five or more carbon atoms. Nevertheless the latter feedstocks correspond to approximately 25 % of the worldwide oxobusiness. Today the Ruhrchetme/Rhne-Poulene Process (RCH/RP) is successfully operated at two locations in the world. In the early 1980s a 100,000 ton scale process was developed and commercialized at the Oberhausen site of Ruhrchemie AG (now Celanese AG). The first unit was started up in 1984. In 1988 and in 1999 two additional lanes went on stream. The actual combined capacity of all three units at Oberhausen (combined with the conversion of not reacted propylene in a high-pressure hydroformylation set up) amounts to more than 500,000 tons of C4-products. The 1:b-C4aldehydes produced are converted further to the corresponding C4-alcohols or after aldol condensation and hydrogenation to 2-ethyl- 1 -hexanol (2-EH). With an annual volume of 360,000 tons of 2-EH the Oberhausen facility became the worlds largest 2-EH operating unit. In 1996 the RCH/RP-process was licensed to Hanwha Corp. in South Korea. With a total capacity of 120,000 tons the unit started up successfully in 1997. The aldehydes are dedicated mainly for the manufacturing of butanols. All units are running well without any major problems being observed. The process can be operated for long periods of time at steady and stable conditions. In Table 1 typical process conditions and the composition of the oxo crude product are shown as fifteen years averages. 7.2.6 Economics Three major factors attribute to the reduction in the overall manufacturing costs of the RCH/RP two-phase hydroformylation process: A highly efficient energy recovery system Minimum capital requirements for an RCH/RP oxo unit. A high feedstock usage in combination with a classical reactor that converts not reacted off gas. These are results of the simplified process design, which was only possible by applying a water-soluble catalyst. The catalyst/product

7. Aqueous biphasic hydroformylation

199

separation is achieved by a simple phase separation. Therefore, the exothermal reaction heat of the hydroformylation reaction can easily be recovered by a heat exchanger inside the reactor (high energy efficiency) and, secondly, no further distillation steps and columns are required to separate the catalyst from product (less capital).
Table 1. Reaction conditions and composition of crude product of the RCH/RP process (15 years average) Product [Unit] n-Butanal[%] i-Butanal [%] Heavy Ends [%] 1:b selectivity Selectivity towards C4aldehyde [%] Process conditions Temperature [C] Total pressure [bar] Vol.(water)/Vol.(organics) Heat recovery*) [%] Conversion [%] *) Minus losses by radiation Typical Value 94.5 4.5 0.4 19 99 Variation 95 - 91 4 -8 0.2 - 0.8 13.3 - 32.3 99

120 50 6 >99 95

1 10 - 130 40 - 60 4-9 >99 85 - 99

Both factors together with the reduced fixed costs and the usage of an own technology (no license fees in case of Celanese plants) makes the RCH/RP process ca. 10 % cheaper in manufacturing costs (costs for ligand synthesis already included) compared to classical rhodium process applying a homogeneous phosphine-modified catalyst.

7.3 Reaction of various alkenes

7.3.1 Ethylene to propanal: why not applied? The major advantage of applying two-phase technology is the easy separation of catalyst phase and product phase. Due to the low solubility of butanal, pentanal, etc. in water and vice versa the separation of the reaction and product layer causes no problem in general, phase separation occurs fast and complete. Propanal, however, derived from the hydroformylation of ethylene, bears a significant miscibility with water. This causes two problems: a) the formed propanal is "wet", the water in the aldehyde has to be removed be distillation, which is not trivial due to the formation of azeotropic mixtures,

200

Chapter 7

b) together with the water dissolved in the propanal a significant amount of catalyst is transported out of the reactor, a recovery of this portion of the catalyst is not very efficient. Nevertheless, the hydroformylation of ethylene itself in the water-based Rh/TPPTS system is fast, but the advantage of ethylene, viz. its higher solubility in the water phase, is counterbalanced by the unfavorably high solubility of water in propanal. For this reason the hydroformylation of ethylene is performed in homogeneous Rh/TPP systems to avoid the aforementioned problems. 7.3.2 Long-chain alkenes Concepts for the conversion of long-chain alkenes Long chain alkenes cannot be hydroformylated economically by the classical Ruhrehemic/ Rhne Poulenc process, because of low solubility of alkenes higher than butene in water [7]. Especially for higher alkenes it would be highly desirable to convert them with a two-phase system because of the higher boiling points of the resulting aldehydes. In a homogeneous system these aldehydes have to be stripped from the catalyst solution at a high temperature which leads to formation of more heavy ends and faster catalyst deactivation. Numerous attempts have been undertaken to hydroformylate higher alkenes with a two-phase system. In principle the difference in polarity between the organic and the catalyst phase has to be decreased compared to the classical RCH/RP system. The following principal approaches have been investigated and will be discussed below: 1. increase the lipophilic character of the catalyst phase, 2. increase the solubility of the catalyst in the product phase, 3. immobilize the catalyst on a support. Cosolvents of the classical RCH/RP system Increasing the lipophilic character will increase the solubility of the alkene in the catalyst phase and thus it will increase the rate of hydroformylation. The limits are the solubility of the cosolvents in the lipophilic phase, and leaching of ligand and rhodium into the organic phase. This means, that choosing the right system for the alkene to be converted is crucial. To increase the lipophilic character, cosolvents can be added to the classical RCH/RP system. Solvents like methanol, ethanol, polyethylene glycol strongly increase the hydroformylation rate [8]. The potential disadvantages of these cosolvents are leaching into the organic phase and their reactivity towards acetals. Leaching of catalyst and ligand seems to be no problem, provided that certain concentrations of cosolvents are not

7. Aqueous biphasic hydroformylation

201

exceeded. Another disadvantage is a drop in 1:b selectivity with increasing amount of cosolvent. This implies that the solvent inters with the catalyst. Other solvents than water for the two-phase hydroformylation Because of the increasing lipophilic character of the long chain aldehydes, solvents other than water can be applied in a two-phase system. TPPTS as ligand is almost exclusively soluble in water, but TPPMS (monosulfonated TPP) is soluble in polar organic solvents. Abatjoglou developed a two-phase system with TPPMS as ligand and N-methylpyrrolidone (NMP) as solvent [9]. After the reaction, water is added to obtain phase separation. Prior to recycling of the catalyst phase, the added water is removed by distillation (see Chapter 8). Ionic liquids have been used for the hydroformylation of higher alkenes. Here the lipophilic character of the catalyst phase can be adjusted by proper choice of the ionic liquid. In the recent past good results in hydroformylation of 1-octene have been reported by P. Wasserscheidt and coworkers. Linear:branched ratios up to 16 with special ionic ligands have been achieved [ 10]. Bahrmann discovered that certain amines form ionic liquids with TPPTS. In this instance the ligand itself is the ionic liquid [11]. Detergents in the classical RCH/RP system Anionic detergents like soaps, AOS and alkylsulfonic acids show no increase in hydroformylation rate compared to the classical system. Cationic detergents, on the other hand, can increase the hydroformylation rate up to a factor of 4. This is best explained by the fact that sodium of TPPTS is exchanged by the detergent cation, which increases the solubility of the catalyst in the organic phase. Technical application is hampered by loss of rhodium into the organic phase and foaming [12]. Immobilization of the catalyst Numerous attempts have been made to immobilize rhodium or its complexes on polymeric or on inorganic supports. Immobilization without ligands leads to systems that mainly hydrogenate the alkene [ 13]. Recently van Leeuwen and coworkers described the immobilization of Rh - Xantphos complexes in silicagel [14]. The system can be recycled 8 times without significant loss of Rh or activity, but the reaction rate is very low. Another possibility is the linkage of phosphines to linear polystyrene [15] or other polymers like PEG and polyacrylic acid. These polymers can be separated by membrane filtration or in some cases by thermoregulated phase transfer [16]. In the latter case, the catalyst is soluble at lower temperature and insoluble at higher temperatures. So the catalyst can be separated from the products by phase separation at higher temperatures.

202

Chapter 7

References 1 2 3 4 Chauvin, Y.; Mumann, L.; Olivier, H. Angew. Chem. 1995, 107, 2941. Horvth, T.; Rbai, J. Science, 1994, 266, 72. a)Kuntz, E. D. CHEMTECH, 1987, 17 (9), 570. b) Kuntz, E. G., Fr. Pat. 2,314,910, 1977 (to Rhone-Poulene); Chem. Abstr. 1977, 87, 101944. Bhanage, B.M.; Studies in hydroformylation of olefins using transition metal cataysts, Ph. D. Thesis, University of Pune, 1995: b) Deshpande, R. M.; Chaudhari, R. V. Ind. Eng. Chem. Res. 1988. 27, 1996. Frohning, C. D.; Kohlpaintner C. W. in Aqueous-Phase Organometallic Catalysis (Eds.: B. Cornils, W. A. Herrmnann), VCH, Weinheim, 1998, p. 295. Chaudhari, R. V.; Bhanage B. M. in Aqueous-Phase Organometallic Catalysis (Eds.: Cornils, B.; Herrmann, W. A.; VCH, Weinheim, 1998, p. 290; b) (Cornils, B.; Wiebus, E., EP 0,158,246, 1985 (to Ruhrchemie AG). Horvth, I.; Kastrup, R. V.; Oswald, A. A.; Mozeleski, E. J. Catal. Letters 1989, 2, 85. Cornils, B.; Herrmann, W. A.; Aqueous Phase Organometallic Catalysis, Wiley-VCH Weinheim, 1998, p. 306 ff. Abatjoglou, A. G.; Bryant, D. R. US Patent, 4,731,486, 1988 (to Union Carbide Cooperation). Chem. Abstr. 1988, 109, 57042. Wasserscheid, P.; Salzer, A. XXXIII. Jahrestreffen Deutscher Katalytiker, Tagungsband p. 66ff. Bahrmann, H.; Schulte, M. DE 19756945 (to Celanese). Chem. Abstr. 1999, 131, 75261. Cornils, B.; Herrmann, W. A. Aqueous Phase Organometallic Catalysis, Wiley-VCH Weinheim, 1998, p. 306 ff. Greiner, R.; Burger, B. XXXIII. Jahrestreffen Deutscher Katalytiker, Tagungsband p. 119 ff. Sandee, A. J.; van der Veen, L. A.; Reek J. N. H.; Kamer, P. C. J.; Lutz, M.; Spek, A. L.; van Leeuwen, P. W. N. M. Angew. Chem. Int. Ed. Engl. 1999, 38, 3231. Bayer, E.; Schurig, V. Chemtech 1976, 212. Jin, Z.; Zheng, X.; Fell, B. J. Mol. Catal. A.; Chem. 1997, 116, 55. Grtner, R.; Cornils, B.; Springer, H.; Lappe, P., DE 3.235.030, to Ruhrchemie AG (1982). Kulpe, J. A. Dissertation, 1989, Technische Universitt Munchen; b) Kohlpaintner, C. W. Dissertation, 1992, Techniche Universitt Munchen. Frohning, C. D.; Kohlpaintner, C. W. in Aqueous-Phase Organometallic Catalysis (Eds.: Cormils, B.; Herrmann, W. A. VCH, Weinheim, 1998, p. 304. Herrmann, W. A.; Kohlpaintner, C. W.; Bahrmann, H.; Konkol, W. J. Mol. Catal. 1992, 73, 191; b) Herrmann, W. A.; Kohlpaintner, C. W.; Manetsberger, R. B.; Bahrmann, H. ; Kottmann, H. J. Mol. Catal. 1995, 97, 65; c) Bahrmann, H.; Bach, H.; Frohning, C. D.; Kleiner, H. J.; Lappe, P.; Peters, D.; Regnat, D.; Herrmann, W. A. J. Mol. Catal. 1997, 116,49; d) Herrmann, W. A.; Kohlpaintner, C. W. Angew. Chem. Int. Ed. Engl. 1993, 32, 1524. a) Frohning, C. D.; Kohlpaintner, C. W. in Aqueous-Phase Organometallic Catalysis (Eds.: Cornils, B.; Herrmann, W. A.; VCH, Weinheim, 1998, p. 302; b) Bahrmann, H.; Frohning, C. D.; Heymanns, P.; Kalbfell, H.; Lappe, P.; Peters, D.; Wiebus, E. J. Mol. Catal. 1997, 116, 35.

5 6

7 8 9 10 11 12 13 14 15 16 17 18 19 20

21