US2419247A - Recovery of hemimellitene - Google Patents

Description

Patented Apr. 22,1947

RECOVERY OF HEMIMELLITENE Stanley Francis Birch and Raymond Arthur Lowry, Sunbury-on-Thames, England, assignors to Anglo-Iranian Oil Company Limited, London, England-a British joint-stock corporation No Drawing. Application September 30, 1944, Se-

rial No. 556,672. In Great Britain October 19,

Hemimellitene (1.2.3-trimethylbenzene) one of the three trimethylbenzenes is difiicult to obtain in .a condition of purity. Hemimellitene occurs both in petroleum and in coal-tar, in a small proportion in relation to the other two isomeric trimethyl benzenes namely, mesitylene and pseudocumene, present. Careful fractionation however in highly efficient columns of either petroleum feedstocks, such as a liquid sulphur dioxide extract of heavy Iranian naphtha or a coal-tar distillate, (i. e. solvent naphtha), will yield a narrow boiling range fraction relatively rich in hemimellitene. The hemimellitene is however associated with a number of other hydrocarbons both aromatic and non-aromatic, of the same boiling range, and its separation from them is difficult. Although such a fraction from a product of petroleum origin (unless derived from a cracked or pyrogenic feedstock) would be substantially free from hydrindone, a similar fraction from a coal-tar naphtha would contain appreciable proportions of hydrindene, particularly if a preliminary mild hydrofining treatment to remove such unsaturated hydrocarbons as indene had been given to the coal-tar naphtha. Since the boiling points of hemimellitene and hydridene are very close, the separation by fractionation alone, of hemimellitene from the hyrindene when present is practically impossible.

Known methods for the separation of hemimellitene from concentrates such as the fractions referred to, depend on fractional crystallisation and/or hydrolysis of the sulphonic acids or alkali 'sulphonates obtained by treatment with sulphuric acid. These methods are however tedious and do not lend themselves to the preparation of large quantities of the pure hydrocarbon.

Hemimellitene can be obtained by synthesis, but the most efiective method of synthesis described in the literature is a lengthy operation and yields are small.

The invention consists in the recovery of the hemimellitene under the conditions hereinafter described by reference to the following table:

TABLE Aromatic hydrocarbons boiling between 160 and 185 C. and those of the corresponding hydrogenation products 7 Claims.

Aromatic Hydrocarbon B. P., C. hydrogenation Pseudocumcne 169. 4 140-143. Isobutylbenzene 172.8 170-172. sec. Butylbenzene. 173. 5 178-179. Hemimcllitene 076. l l42-l43:5. 1.2-Diet-l1ylhenzcnc 176. 5 Not given. in Oymene 175. 7 166-167. o-Cyrnenc 177 171. p-Oymenc 177.1 170. Hydrindene 178. 3 cis-166, trans-159. l.3-Diethylbenzel 180. 6 169-173. n-But-ylbcnzene... 183.1 179. o-Propyltoluene-.. .184 175-176. m-Propyltol-uene l 182 170. p-Propyltoluenc. 183 175. lA-Diethyl'bcnzcllen 182 174-176.

It will be observed from the foregoing table in which are listed the boiling points of the aromatic hydrocarbons boiling between 160-'185 C. and the corresponding fully hydrogenated heXa-hy- (ire-aromatics, (namely, the corresponding cyclohexanes) that the difference in boiling point between the trimethylbenzenes and their hydrogenation products is appreciably greater than those of the other aromatic hydrocarbons shown in the table, particularly those boiling in the same range.

It follows that if a narrow fraction boiling for example between 175 C. and 180 C. in which the hemimellitene has been concentrated by efiicient fractionation, is fully hydrogenated by catalysis, re-fractionation oi the hydrogenated product will yield a clear cut separation of the products derived from the trimethylbenzenes from those derived from other associated hydrocarbons 159 respectitvely. Pseudocumene (B. P. 169.4

(3.) Whenpresent is similarly converted into "1.2.4 trimethylcyolohexane which boils at 40 to 143;" 6., and would therefore be found associated with the naphthenes derived from the hemimellitene. If however the fraction derived from the hydrogenated product boiling between and 6.,

that is to say, the fraction containing both the cyclohexanes derived from pseudocumene and hemimellitene is now reconverted to the original aromatic hydrocarbons by catalytic dehydrogenation, a product is obtained containing only pseudocumene and hemimellitene. Since these hystate.

For the first hydrogenation of the aromatics to naphthenes Raney nickel or other known hydrogenation catalysts may be used, provided the feedstock be sulphur free. -Catalysts of the nickel-tungsten sulphide type may however be used if sulphur be present.

For the dehydrogenation reaction the hydrofining catalyst, molybdenum oxide on alumina may be used, but the nickel-tungsten sulphide catalyst could also be used for this operation under more drastic conditions than those employed for the hydrogenation stage of the separation.

With sulphur-free feedstocks dehydrogenation may also be effected in the liquid phase by continued boiling in the presence of Raney nickel alone.

We claim:

1. In a process for the recovery of hemimellitene (1.2.3-trimethylbenzene) (B. P, l76.1 C.) from a narrow boiling range hydrocarbon oil fraction boiling between 175 C. and 180 C. which is relatively rich in hemimellitene and also contains a small proportion of pseudocumene (1.2.4-trimethylbenzene) (B. P. 169.3 C.) and a proportion of hydrindene (B. P. 178.3 C.) whose presence makes separation of hemimellitene by fractional distillation alone practically impossible, the steps of hydrogenating said narrow boiling range fraction by catalysis to convert the hemimellitene into 1.2.3-trimethylcyclohexane (B. P. 142 C.), the pseudocumene into 1.2.4-trimethylcyclohexane (B. P. l40143 C.) and the hydrindene into cisand trans-bicyclononanes (B. P. 166 and 159 C., respectively); fractionally distilling the hydrogenated product to recover an intermediate fraction boiling between 140 and 145 C. containing only the trimethylcyclohexanes; dehydrogenating said intermediate fraction by catalysis to reconvert the cyclohexanes thereof into the original aromatic hydrocarbons to yield a dehydrogenated product containing only hemimellitene (B. P. 176.1 C.) and pseudocumene (B. P. 169.3 C.) readily separably by fractional distillation; and, fractionally distilling said dehydrogenated product to recover hemimellitene in a pure state.

2. The process of claim 1 in which said narrow boiling range fraction also contains m-cymene (B. P. .175.'7 C.) which is converted by the hydrogenation into the corresponding fully saturated cyclohexane boiling at 166-167 C.

3. In a process for the recovery of hemimellitene 1,2,3-trimethylbenzene) (B. P. 176.1 C.) from a narrow boiling range hydrocarbon oil fraction boiling between 175 C. and 180 C. which is relatively rich in hemimellitene and also contains a small proportion of pseudocumene (1,2,4-trimethylbenzene) (B. P, 169.3 C.) as well as a proportion of at least one. other aromatic hydrocarbon so closely related in boiling point to hemimellitene as to make separation of hemimellitene. therefrom by fractional distillation alone practically impossible and hydrogenatable to a fully saturated cyclohexane derivative having a boiling point outside the range C. to C., the steps of hydrogenating said narrow boiling range fraction by catalysis to convert hemimellitene, pseudocumene and said other aromatic hydrocarbon present to their corresponding fully saturated cyclohexane derivatives; fractionally distilling the hydrogenated product to recover an intermediate fraction boiling between 140 C. and 145 C. containing only the relatively low boiling cyclohexane derivatives of said trimethylbenzenes; dehydrogenating said intermediate fraction by catalysis to reconvert the cyclohexanes thereof into the original aromatic hydrocarbons to yield a dehydrogenated product containing only hemimellitene and pseudocumene, readily separable by fractional distillation; and, fractionally distilling said dehydrogenated product to recover hemimellitene in a pure state.

4. The process as specified in claim 1 further characterised in that a hydrogenation catalyst of the nickel-tungsten sulphide type is used when sulphur is present in the feedstock.

5. The process as specified in claim 1 further characterised in that a hydrogenation catalyst of the Raney nickel type is used when the feedstock is sulphur free.

6. The process as specified in claim 1, in which the fraction derived from the hydrogenated product, boiling between 140l45 C., containing both the cyclohexanes derived from pseudocumene (B. P. 169.4), and hemimellitene (B. P. 176) is subjected to catalytic dehydrogenation using molybdenum oxide on alumina under dehydrogenating conditions as catalyst and the fraction thus re-converted into the original aromatic hydrocarbons, to yield the product containing only pseudocumene (B. P. 169.4) and hemimellitene (B. P. 176) readily separable by fractionation.

7. The process as specified in claim 1 in which the fraction derived from the hydrogenated product, boiling between 140145 0., containing both the cyclohexanes derived from pseudocumene (B. P. 169.4), and hemimellitene (B. P. 176) is subjected to catalytic dehydrogenation using as catalyst a nickel-tungsten sulphide type catalyst under dehydrogenating conditions more drastic than the hydrogenating conditions employed whe using the same nickel-sulphide type catalyst in the hydrogenation stage and the fraction thus re-converted into the original aromatic hydrocarbons, to yield the product containing only pseudocumene (B. P. 169.4) and hemimellitene (B. P. 176) readily separable by fractionation.

STANLEY FRANCIS BIRCH. RAYMOND ARTHUR LOWRY.

REFERENCES CITED .The following references are of record in the file of this patent:

UNITED STATES PATENTS Number Name v Date 2,282,231 Mattox Ma 5, 1942 1,965,956 Dunkel et a1 July 10, 1934 Certificate of Correction Patent N 0. 2,419,247. STANLEY FRANCIS BIRCH ET AL.

It is hereby certified that error appears in the printed specification of the above numbered patent requiring correction as follows: Page-1, second column, line 7, in the table, second column thereof, opposite Hemimellitene for 076.1 read 176.1;

and that the said Letters Patent should be read with this correction therein that the same may conform to the record of the case in the Patent Office.

Signed and sealed this 1st day of July, A, D. 1947.

April 22, 1947.

LESLIE FRAZER,

First Assistant Uanum'saioner of Patents.