US2376104A - Process for purification of diolefins - Google Patents

Description

May 15, 1945.

C. E. WELLING PROCESS FOR PURIFICATION OF DIOLEFINS Filed Feb. 19, 1942 c OR HYDROCARBON.

FEED

OVERH EAD VAPORS FRACTIONATOR METHYL FORMATE ENTRAINING AGENT BOTTOM PRODUCT INVENTOR. C. E. WELLING ATTORNEYS Patented May 15, 1945 PROCESS FOR PURIFICATION DIOIEFIN Charles(E. Welling, Bartlesvllle, kla.,' assignor to Phillips Petroleum Company, acorporation of Delaware Application. February 19, 1942, Serial No. 431,599 4 Claims. (cram-42) This invention relates to the concentration of .pentadienes and hexadienes from hydrocarbon mixtures containing the same together with close-boiling hydrocarbons of other types. More particularly this invention relates to the use of methyl formate to render concentration of said diolefins by fractional distillation much more practical and feasible than is fractional distillation of the hydrocarbon mixture alone.

vHydrocarbon mixtures which contain one or more diolefins with five or six carbon atoms in the molecule'together with one or' more other hydrocarbons of the types of mono-oliefinsjparaffins, naphthenes, etc., are available from various sources. One such source is the cracking, on an industrial scale, of petroleum products. Use of diolefins for chemical syntheses and the like is frequently facilitated by having these materials substantially pure and free of other types of hy- I drocarbons. In general, the fiveand six-carbon-atom diolefins have boiling points quite near the boiling points of hydrocarbons of other types containing the same number of carbon atoms. For instance, the range of boiling points of fivecarbon-atom hydrocarbons is from about 50 F. 'to higher than 125 F. and more than forty hydrocarbons are known to boil within this range, which includes the boiling points'of the fivecarbon-atom diolefins, or pentadienes, such as isoprene and piperylene. In the case of isoprene, boiling at 933 F., there are olefins containing five carbon atoms which boil below isoprene at temperatures of 86 and 87.8-88 F. and others boiling above isoprene at about 96.6 and 98-99 F.; normal pentane, boiling atabout 96.9 F. is still another hydrocarbon which has a boiling point quite near to that of isoprene. In the case of the six-carbon atom diolefins, such as hexatypes of hydrocarbons in the presence of a highly selective entrainer, methyl formate, which forms minimum-boiling azeotropic mixtures with individual hydrocarbons in the mixture.- I usually prefer to apply this process to hydrocarbon mix tures of relatively narrow boiling range, such as may be obtained by a preliminary fractionation,

diene-1,3, 2,3-dimethyl butadiene-1,3, etc., the

number of like-boiling non-diolefinic hydrocarbons is even greater.'

. Hydrocarbon mixtures'produced by craclting or other operations which contain fiveor sixcarbon-atom diolefins will in general contain a substantial number of hydrocarbons of other types, at least some of which will have boiling points quite close to the boiling points of the diolefins. Under such circumstances, it is diflicult, and sometimes impossible, to separate in good purity the individual diolefins byv conven- Another A further 7 which would normally contain one or more d1- olefins together with close-boiling hydrocarbons of other types. Such hydrocarbon mixtures should not generally contain more than minor amounts of olefinic, paraffinic, or cyclic compounds having boiling points more than about say 6 F. higher than the boiling point of the diolefin contained in the fraction, although in some cases non-diolefinic hydrocarbons boiling considerably higher than the desired diolefin may be separatedby myprocess. While methyl formate has the effect of widening the spread between distillation temperatures of low-boiling or like-boiling impurities and the diolefin,.the reverse may tend to be true in the case of impurities which boilmuch higher than the diolefin. Thus the azeotrope of a high-boiling hydrocarbon may not be sufliciently volatile to allow its separation overhead from the diolefin. Accordingly, non-diolefinic hydrocarbons which boil considerablyabove the diolefin are ordinarily better separated by conventional fractionation previous to the distillation in the presence of methyl formate. However it should-be noted that the selectivity of methyl formate as between diolefins and non-diolefins is so much greater than other generally known entraining agents that the limitation as to the high-boiling impurities just discussed is not nearly so critical as might be expected, and non-diolefins are frequently taken off overhead ahead of the diolefin by distillation with methyl formate even though their boiling points are as much as 10 to 15 F. higher than that of the diolefin which is to be ,iurified. The extent to which this may be done will depend upon the types and boiling range of the non-diolefinic hydrocarbons involved. No such limitation as the above applies to undesired trates its superiority over hydrocarbons boiling below the particular diolefin in question; a rather wide range of boiling points below that of the diolefln is permissible.

A specific example of my process which illusis as follows:

EXAMPLE TAllI-E I ordinary distillation The course of the dis- *tillation is shown in Table I.

far more efllcient separation of impurities than was accomplished by conventional fractionation. The eflicacy of my process, as embodied in this example, is self-evident.

While I have shown by way of example the separation of isoprene in substantially pure form from admixture with close-boiling non-dioleflnic hydrocarbons, I have found that by the use of methyl formate in accordance with this inven tion, I may also readily efiect the purification of other 5- and 6- carbon-atom dioleflns from complex hydrocarbon mixtures. For example, I have discoveredv that separation of lowand/0r high-boiling piper'ylene from a mixture which also contains the close-boiling unsaturated alicyclic hydrocarbon cyclopentene, to be impossible by' even the most efficient conventional fractionation Conventional fractionation of an isoprene concentrate I Boiling Per cent range Reflux Vol. per gg i Olefin assumed isoprene Fraction t(ot illigected ratio glut of i g preseczliltnitlnu- 71% in fraction 111:)."1 m W t) 84. 6-91. 6 15:1 14 89. 8-00. 7 15:1 13 Z-methyl-l-butene- 66 90. 7-92. 7 15:1 18 do 75 91. 1-92. 3 :1 11 87 91. 4-02. 3 80:1 6 ll 91. 1-93. 2 60:1 5 91 92. 9-93. 2 :1 12 90 8 93.1-93.9 40:1 12 82 Kettle residue+ losses 9 100 Estimates of the purity of the various fractions are based upon the known refractive indices for isoprene and the oleflns. It is seen that in this fractionation about nine per cent'impurities were present in the best isoprene fractions.

A second fractionation was then conducted in the same column but in the presence of methyl formate as an entraining agent. The hydrocarbon mixture contained about 71 per cent isoprene and Z-methyl-l-butene comprised the bulk of the impurity. The refractive index (71. of the hydrocarbon mixture was 1.4090. The liquid volume ratio of methyl formate to hydrocarbon in the charge to the column was 1.35 to 1. The results of the fractionation are shown in Table ent in admixture with the piperylenes, and since its presence in piperylene may. frequently aifect greatly the polymerization properties and the characteristics of a finished product of piperyl ene, it is highly important to flndsome simple method of resolving the afore-mentioned constant-boiling mixtures. I have found thatthismay be done by fractionation in the presence of Tm: II Fractions tion of an isoprene concentm te with methill formate p Boiling vol bent Per cent range cor- Volume Cumul. vol. Olefin assumed isoprenein Fraction rected to 760 2%? per cent per cent of i gg present for purity 11E irsction mai s), of charge charge head calculations (by "3) 1 7:11-7:17 and 1e 10 40.7 I 1.3 84 7s.4 50:1 13.5 20.5 48.1 fl methyl-l-butene. 1.4000 57 3 7B. 3-7114 50:1 11.5 41.0 47.7 r n 1. 4170 so 4 78.8 50:1 10.5 51.5. 48.8 m 1.4204 97 5 7110 50=1 10.5 e: 47.5 1.4211 00 6 78.6 50:1 10.5- 72.5 48.8 n 1.4212 99 7 78.6-84.2 50:1 9.0 81.5 56.0 -..-.do 1.4216 100 8 86.9-88.0 50:1 .5 82 97.0 Loss in distillation and washing mules: 18 1) methyl formate to separate out cyclopentene from thepiperylene if such a constant-boiling mixture is used as hydrocarbon feed to the column. This offers a convenient meansof concentrating piperylene to-any desired extent.

I generally prefer to carry out'my process at temperatures varying from about F-.-to about 120 F. and at pressures from about atmospheric to pressures 01 about two atmospheres, absolute.

It is usually convenient to operate at or slightly above atmospheric pressure. If desired, lower or higher temperatures and pressures may be used.

It is desirable to operate my process at as low a temperature in the kettle and throughout the column as is convenient; methyl formate has mixture is continuously charged to the fractionating column at some point along the column and some other point or points along the column. In such a continuous fractionation the diolefin ordinarily is removed from the kettle of the column and azeotropic mixtures of methyl formate with the other hydrocarbons present are taken ofi the top of the column. The diolefin removed from the column kettle may or may not contain substantial amounts of the entrainer, depending upon the relative amounts of entrainer and hydrocarbon mixture charged to the column, and my process may be operated with any desired ratio of entrainer to hydrocarbon mixture equal to or greater than a minimum ratio which will be determined by the amount of entrainer required to form azeotropic mixtures with the hydrocarbons which are to be distilled away from the diolefin. The many difierent ways of carrying out azeotropic distillations which are known to the art may be applied with suitable modification to carry out the present invention.

In operating my process, I use any convenient method of removing the entrainer from the prod,- ucts of the fractionation, such as water washing. If the diolefin is removed as a kettle product it may contain any high-boiling materials resulting from decomposition or degradation of a part of the methyl formate. Such high-boiling materials may not always be water-soluble and hence would not be removed by water washing of the kettle product as the entrainer is. In such a case it is occasionally necessary first to water-wash the kettle product to remove all water-soluble material, and then to free the diolefln product of high boiling materials by distilling the diolefln overhead in auxiliary equipment.

Y the entrainer may be added at the same point or Since the invention may be practiced otherwise than as specifically described herein, and since many'modifications and variations of it will be obvious to those skilled in the art, it should not be restricted except as specified in the appended claims.

I claim:

1. The process of effecting separation of substantially pure piperylene from admixture with cyclopentene which comprises fractionally distilling the mixture of piperylene and cyclopentene in the presence of methyl formate in amount sufficient to form an azeotropic mixture with all of said cyclopentene and thereby efiecting substantially complete separation of said cyclopentene from said piperylene, said cyclopentene being recovered overhead in the form of an azeotrope with said methyl formate.

2. The process of efiecting separation of piperylene from admixture with close-boiling'Cr, hydrocarbons including a substantial proportion of cyclopentene which comprises fractionally distilling the mixture of piperylene and said Ca hydrocarbons in the presence of methyl formate in amount sufiicient to form an azeotropic mixture with all of said C5 hydrocarbons and removing said azeotropic mixture of methyl formate with said C5 hydrocarbons overhead while leaving said piperylene behind.

3. The process of efiecting separation of piperylene from admixture with close-boiling pentenes including a substantial proportion of cyclopentene which comprises fractionally distilling the mixture of piperylene and said pentenes in the presence of methyl formate in amount suflicient to form an azeotropic mixture with all of said pentenes and removing said azeotropic mixture of methyl formate with said pentenes overhead while leaving said piperylene behind.

4. The process of effecting the resolution of a constant-boiling mixture of piperylene and cyclopentene into piperylene and cyclopentene fractions which comprises fractionally distilling said

Images