[1]United States Patent Office 3)472,906 3,472,906 HYDROGENATION PROCESS Hiepke Boerma and Johan van der Drift, Vlaardingen, Netherlands, assignors to Lever Brothers Company, New York, N.Y., a corporation of Maine No Drawing. Continuation of application Ser. lszo. 559,115, June 21, 1966. This application Aug. 29, 1968, Ser. No. 756,737 Claims priority, appheation Great Britain, June 29, 1965, 27,461/65 Int. Cl. C07c 5116 U.S. Cl. 260-666 8 Claims ABSTRACT OF THE DISCLOSURE This disclosure is concemed with selectively hydrogenating non-conjugated polyenes to the corresponding cyclic monoolefins with a copper catalyst without any other metal showing a higher activity than copper. The reaction medium has hydrogen without a solvent. This application is a continuation of an application Serial No. 559,115 filed June 21, 1966, now abandoned. This invention relates to hydrogenation and especially to the hydrogenation of cyclic compounds containing two or more ethylenic bonds in themolecule. The invention provides a method whereby cyclic polyenichydrocarbons, especially those containing 8 to 24 carbon atoms in the ring, can be selectively hydrogenated to yield cyclic monoenic hydrocarbons. The process of the invention may be applied to the hydrogenation both of polyenic compounds in which the double bonds are conjugated (which are usually relatively easy to hydrogenate) and of those in which the double bonds are in isolated position. Cyclic polyenic hydrocarbons, i.e. cyclic hydrocarbons containing two or more carbon to carbon double bonds, of the kind mentioned above include cyclododeca-1,5,9triene, cyclododeca-1,5-diene, cycloocta-14-diene, cycloocta-1,3-diene, cycloocta-l-diene and cyclohexadeca1,5,9,13-tetraene, and their substitution products, especiary those in which the substituent is a saturated or aromatic group. Also mixtures of these starting materials, in particular of various cis-trans isomers which are often found together, can be used. In such products the substituent is preferably a saturated hydrocarbon radicai, but other substituents, especially those of relatively low polarity such as ester or ether groups, also come into consideration. The presence of more polar groups, for instance hydroxy, carbonyl and carboxyl groups rnay render cyclic compounds in which they occur less suitable. Many of such substances can be obtained by oligomerisation reactions of butadiene, isoprene, piperylene, whereas butyns and/or other unsaturated compounds can be used in co-oligomerisation reactions. The term "oligomerisation" used below is to be understood to include cooH.-Omerisation. The cyclic monoenic compounds obtained according to the invention may be used in the preparation of useful compounds such as 1,10-decanedicarboxylic acid, 11aminodecane-l-carboxylic acid lactam, cyclododecanol, cyclododecanone or corresponding compounds containing more or less carbon atoms. In thepreparation of these compounds it is usually desirable to use as a starting material only those cyclic hydrocarbons containing but one reactive double bond. Hydro-,enation of cyclic polyene compounds by previously available methods usually results in a complex reaction mixture. Thus, for instance, @bydrogenation by such methods of cyclododeca-1,5,9-triene will usually result in reaction mixtures containing cyclododecane and Patented Oct. 14, 1969 2 the various isomeric forms of cyclododecene, cyclododecadiene and of the trienic starting material. As the separation of the compounds of a specific unsaturation in the molecule is difficult and as in particular the presence in substantial amount of polyenic cyclic compounds having a plurality of vulnerable double bonds is objectionable in the course of further reactions, it is desirable to provide a relatively simple method by means of which cyclopolyenic compounds can behydrogenated selectively 10 into monoenic compounds of good quality. Processes purporting to effect some degree of selectivehydrogenation of such cyclic compounds by the use of a partially poisoned normally rather unselective catalyst in the presence of a so-called displacement solvent (also 15 sometimes referred to as a hydrogen donor) have been proposed. The partially poisoned catalysts, however, for instance sulphided nickel catalysts, bave the drawback that they are relatively inactive and so must be used in comparatively high quantities at high temperatures and 20 for long reaction periods. Moreover, nickel sulphide catalysts show pronounced isomerising properties leading among other things to the formation of objectionable amounts of unstable polyenic compounds. The presence of these proportions of unstable @compounds affects the 25 stability, especially the colour, of the filtered product formed especially upon storage. The so-called displacement solvents proposed were solvents which, by virtue of their affinity for the catalyst, were expected to function by displacing the monoenic 30 compound by desorption from the catalyst surface so that the monoenic compound would not be hydrogenated to the fully saturated compound. The present invention provides a method for the selective hydrogenation of cyclic polyenic compounds to cyclic 35 monoenic compounds using a copper-containing catalyst. An important feature of the -catalyst that has been found most successful is that the metal part of the active material consists of more than 95% of the element copper. The chemical state of the element copper in the catalyst 40 can be symbolized by Cux+, in which x may have a value ranging from 0 to slightly below 2. A wide variety of such catalysts is available. The active material may be dispersed on carriers such as, for example, silica, magnesium oxide, zinc oxide, diatomaceous earth, activated 45 clay, activated carbon, aluminum oxide, chromium oxide, asbestos, iron oxide, titanium oxide or the like or combinations of these. Suitable catalysts are: CU/MgO/SiO2, Cu/AI203, Cu/MgO, Cu/diatomaceous earth, Cu-formate, Cu/Cr2O3 and Cu/Cr?03/BaO. It is 50 desirable that the active copper-containing material is dispersed on a carrier, but the carrier should not be of such a nature as to impair the catalytic activity of the copper. Other metals, e.g. Ni, Pt, Pd, Fe and Co, pa-rticularly those showing higher catalytic activity than cop.55 per, should preferably not exceed some 5% of the metal part of the active material of the catalyst, since the presence of those metals in substantial proportions has been found to decrease the selectivity. Catalysts used according to the invention may be obtained by precipitating copper 60 compounds from solutions by means of an alkaline reagent with or without coprecipitation of other compounds, optionally in the presence of a carrier, after which the precipitate is dried. With certain types of catalysts a subsequent heat treatment may be of advantage 6,3 (for instance thermal decomposition without substantial sintering may be effected). Another method is to impregnate a suitable support, for instance diatomaceous earth, silica gel or charcoal, with an aqueous -solution of a copper salt and to dry 70 the impregnated material. Yet another method is to mix the dry copper compound with a carrier and to heat the mixture. Sometimes, depending on the selected reaction
[2]-31472 906 3 temperature, it is advanta.-eous to pre-reduce the catalyst, but tisually this occurs in the first sta.-e of the hydrogenation. The amount of catalyst used is not very critical and is mainly dependent on the activity of the catalyst, but other factors, such as the presence of contaminants in 5 th.- material to be hydrogenated, which may have a poisoning effect, are of importance. The amount tised usually varies between 0 and 20% by weight of the catalyst calculated on the weight of the substance to be 10 hydrogena',ed. Preferably the amount of catalyst will vary between 1 and 5% calculated as indicated above. The reaction temperatures employed also depend on the activity of the catalyst employed, as a low activity can be partly compensated for by employing higher tem- I 5 peratures, and is further affected by the nat,,lre of the substance to be hydrogenated in particular by the thermal stability. Reaction temperatures ran.-ing 'between 20 and 300' C., preferably between 150 and 220' C., have beeri employed. 20 According to the invention the hydrogenation may be carried out at widely diverging hydrogen pressures, for instance between 1 and 200 atm. Pressures below 2 atai. have the drawback that they will cause the reaction to proceed only at a slow rate unless high arrounts ()f 25 catalyst are used. Consequently it is preferred to carry out the reactions in an autoclave at pressures between 2 and 100 atm., and especially between 2 and 30 atm. Thus in the case of the selective hydrogenation of cyclododeca-1,5,9-triene it was found that optimal results 30 were obtained with 2.0% of a CU/M.@O/SiO2 catalyst containing about 40% of copper at a reactio@n temperature of 185' C. and an average hydrogen pressure of 10 atm., for which a reaction time of 30 minutes was needed. The process according to the invention provides a 35 method for selective hydrogenation among the advantages of which are the following: (1) High yield (80 to 100%) of monoenic compounds depending on the nature of the starting material used. Generally the yields decrease slightly when the molec- 40 ular weight of the startin- material increases. .1 (2) High intrinsic catalytic activity resulting in short reaction times (I to 60 minutes) in combination v,,ith low catalyst concentrations (1 to 5%); (3) The process can be carried out at pressures below 30 4,3 atmospheres at relatively low catalyst concentrations whilst maintaining a relatively hi.-h reaction rate and a high selectivity. (4) The presence of a solve-nt is not required owing to the high selectivity; 50 (5) The products obtained according to the invention are more stable than those dbtainable und,-r comparable -conditions usin.- nickel sulphide catalyst and can be used without further purification; (6) The catalyst can be easily prepared and is - relatively a-5 cheap. As to the startin-. materials employed, the various technical grades of the cyclopolyenic hydrocarbons can be used, provided that they do not contain appreciable 60 amounts of contaminants such as polymer;sation inhibitors impairing the activity of the catalyst. Thus it has sometimes been found advanta.-eous to subject the starling material to a purification step comprising treatment with an absorbent such as activated alumina, or to dis1 65 tillatioli. The technical grades cf the polyenic - compouncis usually consist of mixtures of various isomeric forms, blit of course th-- various isomeric forms may be selectively hydrogenated according to the invention. Although there is in g--neral little advantage in the use of a solvent, an inert solvent may b.- used. 70 Hydrogen addition is preferably stopped when an iodine value has been reached of approximately the theoretical valtie of the monoenic compolind. As refractive index and degree of unsaturation are closely related, it is convenient to record the refractive index and stop 75 4' at the desired value. Su'bsequently the catalyst is removed by filtration. As indicated above an important application of the invention is the production from cyclic polyenic compounds especially hydrocarbops of the corrcsronding cyclic monoenic compounds uncortaminated by substantial proportions of polyenir, compounds and saturated compounds. Ideally such polyenic @coiiipounds should be completely absent but small proportidns amounting for instance to some 5 to 10 (jreven 12.5,Yo of the fnal reaction iiilxture may not be objectionable, especially when composed mainly of cyclic unconjugated, hydrocarbons. As indicated above the copper-containing catalysts emploved show a hi-h degree of s.-Tectivity in the reduction of cyclic polyenes to cyclic monoelies in that they catalyse the reduction of the cycli@ polyenes more rapidly than that of the cyclic monoenes. This makes it a simple matter to control the nature of the product by stopping the hydrogenation at an appropriate point, for instance when most of the initial polyene has been converted into monoer.-e and before much of this has been converted into corresponding satlirated cyclic compounds. Thus fc)r instance the hydrogenation can be stopped when the polyene content of the product has been diminished to 12.5% or less, e.g. 10 to 5 or even 5 to 0%, but before the monoene content has fallen below 70 or 75% and preferably while it is considerably hi.-her e..-. 80 to 90% or even 90 to 100%. The progress of the reduction can be followed by determinations of refractive index and, as shown in Example 1, preliminary trials will indicate for particular conditions an appropriate refractive index at which to stop. The invention will be illustrated with reference primarily to the hydrogenation of cyclododeca-1,5,9-triene since this gives products of particular interest, for instance for the preparation of monomers for use in polycondensation reactions such as are involved in the production of polyamides. The same method can however be used with advanta.-e in hydrogenating other cyclic polyenes including those specified above. EYAMPLE 1 This example relates to the hydrogenation of cycloocta1,5-diene to cyciooctene. A 1-litre rocking autoclave was charged with 200 cy clo-1,5-octadiene purified over a column filled with alumina (nD20 1.4934; I.V. 469) and 5 g. of a copper chromite catalyst (ex. Degussa, sold as a 2 CuO/Cr2O3 catalyst activated with Ba2CrO4, copper content about 40%). Hydrogen gas was added to the autoclave to a pressure of 5 atm. and the system was heated to 1851 C. Subsequently the autoclave was washed with hydrogen three times in order to remove any water liberated during theactivation of the catalyst. After that hydrogen was introduced to a pressure of 30 atm. The hydrogen uptake took place rather fast. When a pressure of 20 atm. was reached, additional hydrogen was pressed into the autoclave. This had to be repeated several times; the amount of hydrogen added was recorded. Consequently hydrogen ressure varied durin- the reaction between 20 and 30 p atm. Owin.- to the reaction heat set free the temperature of the mixture increased to 215' C. Samples were drawn from the autoclave at various intervals after the hydrogen was introduced to a pressure of 30 atm. for the first time. The refractive index of the samples was measured. The samples were subsequently filtered and their contents of cyclooctane, cyclooetene and cyclooctadiene were analysed by means of -asliquid chromatography. -
[3]31472)906 5 The results are given in the following table: Tinic of Hy&ogen Cycloreaction in uptal-,e 'n Refi-active Cyclo- Cyclo- octaminutes @@tM. index n,)20 octane octene diene 3 ------------- 60 1.4740 4 76. @ 20 5 ------------- 72 1.4713 5 86 , 8 ------------- 82 1.4696 7 91.5 1 ' 5 20.5 ----------- 90 1.4686 10.5 89.5 ---------- 65.5 ----------- 100 1.4668 28 72 ---------- From the above data it was calculated that in the case of theoretical ideal selective hydrogenation to cyclooctene the hydrogen uptake would have amounted to 88 atm. By plotting the figures of the table inentioned abov6 it was estimated that under the prevailing reaction conditions a maximum yield of 92% cyclooctene could be obtained at a hydrogen uptake of 85 atm. and that this could be attained in a reaction time of 9 minutes. It was further established that in order to obtain the maximum yield mentioned above the reaction should be stopped when flie refractive index at 20' C. of the reaction mixture is 1.46@0. EYAMPLE 2 A I-litre stirred autoclave provided with a cooling coil was charged with 250 g. of the same purified cyclooeta1,5-diene used in Example I and 6 .-. of a non-activated CU/MgO/SiO2 catalyst containin.- 40% copper (prepared by co-precipitating a solution of copper sulphate, magnesium sulphate and water-glass with sodium carbonate at 100' C.). The autoclave was washed three times with hydrogen and subsequently hydrogen was introduced at a pressure of 20 atm. The autoclave was heated to 150' C. under constant stirring and a,-ain washed three times with hydro,-en, after which hydrogen was introduced to a pressure of 30 atm.; the temperature. quickly rose to 1900 C. Whenever the hydro.-en pressure had fallen to 20 atm., additional hydrogen was introduced to a pressure of 30 atm. The heat of reaction evolved durin@ the hydrogenation was removed by means of a cooling coil in the autoclave throu.-h which a mixture of glycerol and water was circulated. After 10 min. the hydrogen uptake slowed down and the hydrogenation was interrupted after 16 min. The contents were withdrawn, filtered and analysed by means of gas-liquid chromato,-raphy. The results were as follows: nD 25 1.4671; the product contained 6% cyclooetane and 94% cyclooctene. EXAMPLE 3 In an experiment similar to Exam-ple 2, 250 g. cyclooctadiene and 5 g. catalyst ivere heated under a pressure of 5 atm. to 170' C., the autoclave was washed three times with hydrogen with the hydrogenation carried out at 205' C. and 6-4 atm. After 35 min. the hydro,-enation slowed down and was interrupted after 45 min. The product (nD25 1.4651) consisted of: 3% cyclooctane and 97% cyclooctene. EXAMPLE 4 In an expbriment sifiiil@r- to Example 2 with 250 9. cyclooctadiene and 5 g. activated Cu/MgO/SiO2 catalyst, the activation was carried out by heating the catalyst in hydro,-en at 200' C. for I hour at a pressure of 5 atm. and 185' C. the following results were obtained: hydrogenation time 32 min., nD25 1.4642 and composition 2% cyclooctane and 98% cyclooetene. EXAMPLE 5 A 14itre rocking autoclave was charged with 250 g. purified cyclododeca1,5,9-triene (nD201.5063; consisting of 97% cis, trans, trans-iscimer and 3% trans, trans, trans-isomer; I.V. 464 Wijs method) and 2.5 g. of a non-activated Cu/M,-O/SiO2 catalyst (prepared by coprecipitatin- a solution of copper sulphate, magnesium sulphate and water-.-lass -with Na2CO3 at 100' C.) containing about 40% copper. 6 After having washed the autoclave,several times with nitrogen and hydro.-en, the system was heated at a hydrogen pressure of 10 atm. to a temperature of 185' C. After 15 minutes the autoclave was washed three times to remove any water liberated during the activation of the copper catalyst. Hydrogenation was carried out at about 185' C. and under an average hyd-rogen pressure of 25 atm. After the uptake of 125 atm. of hydrogen several samples were drawn and the corresponding uptake 10 of hydrogen was recorded. The refractive indexes were measured and the samples analysed by gas-liquid chromatographyafter filtering off the catalyst. The results are given below: 15 Cycl odoTim e of Hyd rogen Cycl oCycl odec adiene plus reac tion in upta ke dod eedod eecycl odomin utes in atm. nD O5 ane ene dec atriene 42 ------ ------- 128. 51.4692 6 75 20 49 ------ ------- 1351. 4676 ii. 5 81 7.5 20 56 ------ ------- 1411. 4667 19 74 7 Under the pievailing reaction conditions 81% cyclododecene could be obtained within a reaction time of 25 about 50 minutes. The purity and stability of the final product obtained according to this example were compared with those of the product obtained under similar conditions but using a nickel sulphate catalyst and applying a corresponding 30 higher pressure. These products were examined spectroscopically and the extinction of the product at 2790 A. was determined in a 1 cm. cell, which is considered an index for the relative stability of the crude monoenic compound; also a sample of 25 ml. of each of the crude 35 products was heated to 1000 C. and kept at that temperature while blowing air (5 litres/hour) through the liquid and the products were examined before and after this treatment in the Lovibond colorimeter using a I" cell. The extinction at 2790 angstrom units was also 40 measured after the heat treatment. The results of these tests, and the refractive indices of the various products are shown below: Lovi bond 45 Sain ple nD" E27 90 A. colo ur-yellow NiS prodnct: Untr eated ------ ---------------- 1,46 68 5.5 0.1 Treated (25 ml.) ---------------- 1.4678 18.0 0. 6 Cu product: Untreated --------------------- 1.4667 1.1 <0. I 50 Treated (25 ml.) ---------------- 1.4667 1.4 <0. I EXAMPLE 6 Example 5 was repeated, except that t'@ic autoclave was 55 charged with 250 g. of the same purified cyclododeca1,5,9-triene of Example 5 and 5 g. of a CU/MgO/SiO2 catalyst (the same as in Example 5) which contained 40% of copper. The results obtained are tabulated below: 60 Tiiiie of Hyd roge . CycloCycl odon dycl odec adiene plus reac tion in upta ke dod ecdod ee- ey lodomin utes in atm. nDO 5 ane ene dec actriene 32 -- ------- 116 1.46 87 6.5 77.5 16.5 65 39 -- ----------- 123 1.46 74 10 79 11 57 -- ----------- 136 1.46 41 27.5 70 2.5 In this way a yield of 79% cyclododecene could be obtained in a reaction time of 39 minutes. 70 EYAMPLE 7 Example 6 was repeated, using a I-litre stirred autoclave as described in Example 2, 5 g. of the same Cu/MgO/SiO2 catalyst as in Example 6 containing about 75 40% copper and 250 g. purified cyclododecatriene of
[4]374721906 7 8 Example 6. Experiments were carried o-at under dieerent lyst (of Example 5) at 200' C. and at an avera.-e pressure conditions and the results are given in thei following table: of 25 atm. hydrogen. Pres- Hydro- CycloCyclo- Cyclo- CycloHydrogenation sure genation diddecane , dodecene, dienes, trienes, temp., I C. atm'. time, min. 'ADBS percent percent percent percent 190 -------------- 50 10 1.4668 12.5 79 7 1.5 190 -------------- 25 19 1.4668 11.5 81 6.5 1 190 -------------- 10 81 1.4665 13.0 83.5 3.5 ---------- 190 -------------- 5 310 1.4682 12 81 7 ---------- EXAMPLE 8 The table below gives the results obtained: The experiments of Example 7 were repeated in the Cyclodosame autoclave with 250 g. of the same purified cyclo15 Time of Hydrogen Cyclo- Cyclodecadii3ne plus dodecatriene but with activated Cu%MgO/SiO2 catalyst, reaction in uptake dodee- dodeecyclodoactivated in hydrogen by heating for 1 hour at 200' C. minutes in atM. nDeS ane ene decatriene The results of these experiments are shown in the fol10.5 ----------- 145 1 4691 9.5 64 26.5 lowing table: 12 ------------- 153 1:4675 13 79 9.5 Pres- Cata- Hydrogencyclo- Cyclo- Cyclo- Cyi,,Iosure, lyst, ation tiiiae, dodecane, dodecene, dienes, trienes, Temp., I C. atm. 9. MM. 'nDeS percent percent percent percent 190 -------------- 5 10 44 1.4663 13 83.5 3 0.6 225 -------------- 6 5 42 1.4691 9.5 82.5 8 ---------- 205 -------------- 5 5 43 1.4674 13 78 8 1 205 -------------- 10 5 17 1.4670 11 83 5 1 205 -------------- 25 5 6 1.4671 10.5 81.5 7 1 190 -------------- 5 20 35 1.4669 11 86.5 2.5 ---------- EXAMPLE 9 The experiments of Example 8 were repeated, using a 0.3 liter stirred autoclave with adjustable stirring speed 35 and with cooling coil. Now 100 g. purified cyclododeca1,5,9-t@nene and 2 g. activated CU/MgO/SiO2 catalyst were used. The washing with hydrogen was carried out in the same way as described in Example 2. Hydrogenations were carried out at 190' C. and a pressure of 6-4 40 atm. The results are shown in the followin.- table: Thus within a reaction time of 12 minutes a 79% yield of cyclododecene could be obtained. EXAMPLE 12 Example 5 was repeated, except that the autoclave was charged with 250 g. purified cycIododeca-1,5,9-triene of Example 5 and 2.5 g. of a Cu/CrOs catalyst (ex. The Harshaw Chem. Co. U.S.A., Code No. Cu-0202P) containing 83% of copper. Hydrogen- CycloCyclo- Cyclo- CycloStirring speed, ation tiine, dodecane, dodecene, diones, tri enes, r.p.m. min. n,66 percent p ercent perceiit perceiit 11500 -------------- 75 1.4666 8.5 84.6 B. 5 1.5 2,000 -------------- 58 1.4662 7.5 86.0 5. 5 1.0 2,500 -------------- 72 1.4660 8@o 86.0 5. 0 1.0 EXAMPLE 10 The results obtained are tabulated belgw: T'he experiments of Example 8 were repeated with 250 pyclodog. pudfled cyclo-dodedeca-1,9-triene and 5 g. activated Thne of Hydrogen Cyclo- Cyclodecadiene plus Ca/MgO/SiO,2 catalyst containing 1% Ni, Pd or Pt, 55 reaction in uptake dodee- dodeecycl@dorespectively. For comparison an experiment with a cat, niinutes in atm. 'aD65 ane ene decatriene alyst without additions was carried out. The hydrogena150 ------------ 128 1.4703 6.5 74 19.5 170 ------------ 138 1.4682 12.5 75 12.5 tions were carried out at 190' C. and 6-4 atm. The re195 ------------ 147 1.4668 17.5 75.5 7 sults are shown in the following table: Hydrogen- CYCIOcycl(@- Cyclo- Cyclaation time, dodecane, dodecene, dienes, trienes, Catalyst atMM. 'AI,65 percent percent ppreent p ercent 100% Cu ---------- 117 1.4668 13 80 7 -- ---------- 99% Cu, 1% Pd --- 42 1.4662 17.5 72 8.5 2 99% Cu, 1% Pt --- 50 1.4667 12 79.5 7 1. 5 99% Cu, 1% Ni --- 33 1.4665 16.5 73 8.5 z 5 Here a maximum yield of about 75% cyclododecene EXAMPLE 11 70 could be obtained in a reaction time of 195 minutes. Example 5 was repeated, except that the, autoclave was EY-AMPLE 13 charged with 250 g. purified cyclododeca1,5,9-triene (IID20 1.5043; I.V. 465.4 consisting of 51.5% trans, trans, Example 2 was repeated in substance, bitt two moditrans-isomer, 39% cis, trans, trans-isomer and 9.5% of fications were made. The autoclave was charged with unknown constituents) and 5 g. of a Cu/MgO/SiO2 cata75 250 g. purified cyclododeca-1,5,9-triene (same as in Ex-
[5]3)4721906 9 ample 5) and 2.5 g. of a Cu/Ba/Cr2O3 catalyst (same as used in Example 1). The avera,-e hydrogen pressure was 25 atm. at a reaction temperature of 185' C. After the uptake of 95 atm. hydro.en (after 166 minutes), a quantity of 2.5 9. fresh catalyst was added and th-- reaction was continued after activation of the catalyst. The results are tabulat-@d below: Cyclo doTime of Hydrogen Cyclo- Cyclodecadiene plu reaction in uptake dodec- dodo-.- cyclodo" minutes in atm. nDG5 ane ene decatriene 216 ------------ 128 1.4692 5.5 75 19.5 246 ------------ 135 1.4682 9 80.5 10.5 H,-re a maximum yield of about 80% of cyclododecene was attained, but the reaction time of 246 minutes is certainly not optimal, because the original amount of catalyst was evidently too low. With an initial addition of 5 g. catalyst the duration could be shortened appreciably. EXAMPLE 14 A 1-litre stirred autoclave provided with a cooling coil was charged with 250 -. distilled technical cyclododeca1,5,9-triene (97% cis, trans, trans and 3% trans, trans, transisomer) and 5 g. of a CU/MgO/SiO2 catallyst. The autoclave was washed three times with hydrogen and subsequ,-,ntly hydro.-en was introduced at a pressure of 20 atm. The autoclave was heated to 190' C. under cotistant stirrin@. The autoclave was washed again three times with hydro.-en, after which hydrogen was introduced at a pressure of 30 atm. Whenever the hydrogen pressure had fallen to 20 atm., additional hydrogen wasintroduced at a pressure of 30 atm. The reaction heat evolved during the hydrogenation was removed by means of a cooling coil in the autoclave through which a mixtiire of glycerol water was circulated. After the uptake of about 141 atm. of hydro,-en a sample was drawn which was analysed by means of gas-liquid chromato.araphy. Subsequently the reaction was continued for some minutes and the procedure repeated. The results were as follows: Time of Hydrogen Cyclo- Cycloreaction in uptake dodec- dodeeCyclodoicainutes in atm. 'nDG5 alie ene decadione - 16 ------------- 135 1.4675 S. 5 83. 0 8.5 About 18.5 ---- 141 1.4664 11.5 83. 0 5.5 In the polyene fraction no cyclododecatriene could be identified. EXAMPLE 15 The procedure of Example 8 was repeated under the same conditions, but now 250 g. undistilled technical grade cyclododeca-1,5,9-triene was used. Within about 40 minutes an uptake of only 36 atm. hydrogen took place and the rate of reaction slowed down. Therefore, another 5 g. of the above catalyst were added and the same steps were repeated. A sample was drawn at a hydro,-en uptake of 146 atm. and analysed by gasliquid chromotography. The following results were obtained: Time of Hydrogen Cycloreaction in uptake dodeec= ininutes in atm. NDOS ane ene de 151 ------------ 146 1.4667 14.5 76.5 9.0 - - From the Examples 14 and 15 it is evident that the technical grade starting material was contaminated by a constituent impairing the activity of the catalyst. 10 EY-AMPLE 16 A stiff ed autoclave provided with a cooling and heating coil was charged with 6.5 kg. technical grade cyclododeca1,5,9-triene (containing 97% of the cis, t@rans, trans-isomer and 3% of the trans, trans, trans-isomer) and 260 g. nonactivated CU/MgO/SiO2 catalyst containing 40% copper. The reactor was heated with steam to 180' C. under a hydrogen pressure of 5 atm. and kept at 180' C. for 20 10 min. while venting hydro.-en gas to activate the catalyst. After 20 min. the temperature increased to 200' C. and the hydrogen pressure was maintained at 5 atm. After 70 min. at 200' C. the hydro.-en pressure was rel@eased and the contents of the vessel quickly cooled to 100' C. 1,5 and filtered free from catalyst. The product had the following characteristics: nD 65 1.4668, 16.5% cyclododecane, 76 ' 5% cyclododecene, 6% dienes and 1% trienes (by gasliquid chromatography). EXAMPLE 17 20 A cascade of 5 continuous flow stirred tank reactors with a capacity of 6.5 kg. each was fed with technical grade cyclododeca-1,5,9-triene, as described in Example 16 at a rate of 17.5 kg./hour. The cyclododecatriene was 25 mixed with 4% by weight non-activated CU/MgO/SiO2 catalyst (Example 16). The temperature of the reaction mixture in the cascade was brought at 200' C. and hydrogen was supplied at a pressure of 5 atm. and vented at a rate of 4 Nm.3/h. After some time a product emerged 30 from the fifth reactor that had the following characteristics: nT)65 1.4664, 21.5% cyclododecane, 70% cyclododecene, 7.5 % dienes and 1 % trienes. What is
