US2470712A - Catalytic dealkylation of alkylated compounds - Google Patents

Description

Patented May 17, 1949 CATALYTIC DEALKYLATION OF ALKYLATED COMPOUNDS Charles W. Montgomery and William A. Home, Oakmont, Pa., assignors to Gulf Research a: Development Company, Pittsburgh, Pa., a corporation of Delaware No Drawing.

Application September 18, 1946,

Serial No. 697,826

4 Claims.

This invention relates to catalytic dealkylation of alkylated compounds, including alkylated aromatic and paraflinic compounds.

Many compounds in the past have been subjected to dealkylation but this dealkylation becomes increasingly difllcult in removing the short chain alkyl groups, such as methyl and ethyl groups, and a method for selectively removing methyl groups from a compound containing a multiplicity of methyl groups is particularly desirable.

In straight-run gasolines there is normally present a considerably larger amount of C8, C9 and. higher aromatics than the more useful lower boiling homologues, such as benzene and toluene. Thus, in a typical East Texas-straight-run naphtha, it has been found that the total aromatic portion consists roughly of 14 per cent benzene and toluene, 26 per cent xylenes and the balance high boiling aromatics (to 210 0.). Also it is found that in the operations of thermal cracking,

catalytic cracking, hydroiorming, etc., relatively large quantities of higher boiling aromatics are produced. These aromatics boiling above 140 C., while possessing good rich mixture sensitivity and high octane number, as do benzene, toluene and the xylenes, nevertheless, on account of their higher boiling points, can not be tolerated in substantial quantities as aviation gasoline components.

Accordingly, dealkylation may take place by passing a high boiling alkyl aromatic or a mixture, for example, an aromatic-containing petroleum cut, containing such high boiling alkyl aromatic, in vapor phase over and into contact with an activated catalyst in the presence of hydrogen. It has been found that under these conditions the alkylated aromatics can be dealkylated, with substantial conversion of, for example. xylene, pseudocumene, etc., into benzene, toluene, etc. Aside from these lower boiling aromatics. the principal product appears to be methane. The process is characterized by the following equation:

' or, in the illustrative case of xylene:

vide a process for dealkylation'and particularly.

demethylatlon of either aromatic or aliphatic compounds.

It is a further object of this invention to provide a process for dealkylation and particularly demethylation of either aromatic or aliphatic compounds with such control as to selectively remove a single alkyl group.

' These and other objects of this invention are attained by treating the compound to be dealkylated with hydrogen in the presence of 'a metallic catalyst of the eighth group of the periodic table at a temperature in the range 250 to 400 C.

The temperature'range 250 to 400 C. embodies the operable temperature range for the dealkylation but peak efflciency is obtained in the temperature range 275,to 300 C. In commercial operation this latter temperature range is most desirable. In the case of the aromatic compounds, temperatures below 250 C. tend to cause saturation of the ring while at temperatures above 400 C. there is a greater amountlof cracking than is desired. 3

The catalyst comprises'a metaliof the eighth periodic group usually supported on an inert support, suchas alumina or kieselguhr." Nickel x 3 is the most preferable metal. and this may be best used in such a manner that there is 60 to 75 per cent oi nickel deposited on the catalyst support.

Operable pressures will depend upon the degree of dealkylation but will usually tall within the range of from atmospheric to about 500 pound per square inch.

The compounds to be treated include aromatics. either monocyclic or polycyclic and aliphatic: such as parail'ins and oleiins. Aromatic compounds include xylene, pseudocumene. mesitylene, diethylbenrene. triethylbenzene, etc. Examples of the paraiiinic compounds include neohexane and 2,3,3-trimethylpentane. While the invention is directed to dealkylation generally, it is particularly directed to the dealkylation of the shorter chain alkyl groups and most particularly to the removal of a methyl group from a hydrocarboncompound.

We have found that hydrogen pretreatment of the catalyst'has a markedly deleterious eflect on the catalytic dealkylation of higher aromatics and aliphatics. During the initial period of retion being balanced by methane formation. When methane-producing side reactions of the decomposition type take place (see Reaction B, supra), however, they result in a pronounced gas volume contraction. Moreover, gradually increasing contraction may take place upon long continued runs, even following hydrocarbon pretreatment according to this invention. This increase in contraction appears to be paralleled by increase in catalyst temperature.

Thus, in a continuous process, a check should be made of gas volume contraction or catalyst temperature, and the catalyst should be treated by the method of the invention when required. The method involves a pretreatment of the catalyst with an inert deactivating fluid including nitrogen-rich gases, other hydrocarbons, or the charge itself.

The following tables show the results of runs on various high boiling alkyl aromatics. The catalysts used were nickel supported or kieselguhr and containing from 60 to 70 per cent by weight of nickel. The runs were at substantially atmospheric pressure. Other details of operation are action following hydrogen pretreatment. an ingiven in the tables. controllable temperature rise occurs throughout the catalyst mass. amounting in some cases to Table! 60-75 C. The total liquid recovery is appreciably lowered, extensive dealkylation to benzene Run Number 1 2 :4 occurs in the case of aromatics, the methane content of the exit gases increases greatly, sintering. Charg xylene xylene xylene and permanent impairment of the catalyst ac- L%{2 f}: f1; tivity may occur. The overall effect of such re- AV.T'1.C. 000 200 212 action may be represented, in the exemplary case Tho-H8513; P ts 21.4 21.0 .0 of xylene by the :quaflon iiivlznguratiogitminfi 61 en 600 9! Clef-I0 CH:)2+IIH2- 8CII'I4 (B rmliauctspmdi is: t Y 5 6 7 5 It appears t at as a result of pro onged hydro- 'r 1 0. 0. 24.0 gen pretreatment of the catalyst, the catalyst 1353? WY. 1 0:20 3? 10.0 133 surface is initially covered with a high hydrogen 1 concentration which would be expected to pro- 00.0 0011 03.0 duce extremely rapid reaction at the start of a 3 f g run. This rapid reaction rate may in turn produce a catalyst temperature rise on account of Table 1] Run Number 4 5 6 7 8 9 10 Charge diethyldiethyi trictbyicumene cumene cumene mesitylene 0100B 2.0 2.4 2.3 1.0 400 ve'gm 0.44 0.41 0.40 0.44 .4v.'r '0 211 200 200 200 Mani" 201 200 310 2.00 Throughput 1. 0.00 0.11 0.00 0.00 Run 1 120 00 120 so HiPlflJ/Hl' 21.0 21.3 21.0 21.1 21.0 21.0 Wt. Per Cent Liq. Recovery 01.0 01.4 01.2 00.1 001 10.0 00.4 Products, Wt. Per Cent:

Benzene trace trace trace 2. 5 4. 0 2. 0 Toluene 1.0 3.0 5.6 7.5 9.0 8.6 x 0.0 0.0 1.0 0.0 0.0 1.0 10.0 10.0

000 04.0 00.0 02.0 21.0 20.0 20.4 20.1 204 20.1 20.1 01.2 00.1 01.0 01.0 10.4 30.0 01.2 11.4 10.1 10.1 11.0 21.4 004 11.0 amw... 1.4 1.2 1.0 1.2 22 1.0 1.0 vol. Br/Vol. Exit Gas 1.02 1. 0a 1.03 1. 00 1.00 1.10 1.

1 Contains 2% of material boiling between dietbyl and trietbyl benzene.

excessive heat of reaction. The combination of high catalyst temperature and high surface con- Tcentration of hydrogen is probably responsible for the complete'decomposition reaction.

Reaction A above shows that no gas volume change takes place therein, hydrogen consump- 1s The above tables illustrate that a high degree of demethylation is secured by the process.

The following table illustrates the conditions and results obtained on a xylene charge wherein the catalyst is cobalt-thoria-kieselguhr in a weight ratio of -18-100:

. 6 Table III 4.24.1851, in which the catalyst (same as that R mm: 1 descr bed in Table I) was pretreated with xylene Charge; i g g alone before the first run (liquid throughpu Moles Trig/mole charge 1 9 0.12). Cumulative throughputs (i. e. volumes of space velocity 5 liquid xylene per volume oi catalyst) are tabulat- A T cc ed and the data under the throughput entries re- M -fer to the average yields and temperature for the L 2 run periods shown. Between successive runs, the Throughput 1.39 catalyst was maintained in a hydrogen atmos- H2 parts/hr 27.5 10 phere at 305 C. This hydrogen treatment was Run duration, min. 105 in all cases or such duration as would normally Weight percent liq. recovery 95.3 produce a highly exothermic initial reaction.

Table 1' Run Number 17 18 19 20 21 22 2'1 Currmilatizro l'lliroughrautr clrgsrvals o s. o x- Liquid Matt"; 1."%.."....i. i?.- 88% 21 2 e: illi 253 592 $2. 5 Per cent Methane in Exit Gas.. a. 4 1. o 1. 2 20.1 m. 2 :51. e 231 c Benzene, Vol. Per cent trace 0. 5 3.3 4. 5 ii. 5 9. 0 10.0 Toluene, Vol. Per cent 16.0 10.3 at 2 35.0 35. 0 1m. 7 '37. 5 tigi fiiig afirag" 0 iii 0 3&3 o it? 1% Av. ea. Tem 312 312 '312 '3" "3i? "3T8 "313 Products, prod. basis: Run Number 11 I While catalyst activity is somewhat reduced by Benzene, we ght percent 2.0 25 the hydrocarbon pretreatment, normal activity Toluene, weight percent 8.5 is restored after a liquid throughput of about 2.0, Xylene, weight percent 89.5 and in no case does the exothermic reaction pro- Exlt gas parts/hr 26.1 duce a temperature rise of more than 15 C. The Hydrogen 86.1 trend toward increased temperature may be coun- Methane 11.8 terbalanced by a periodic hydrocarbon treatment Ethane+ 2.1 of the catalyst. Vol. Hz/vol. exit gas 1.05 By a comparison of Tables IV and V, it will In the following table are shown results of three runs (12, 13, 14) of xylene dealkylation of the usual, initial highly exothermic type of reaction with hydrogen treated catalyst and for comparison two runs (15, 16) on the subsequent stage of the reaction after the catalyst temperature had subsided. These runs were carried out with an initial catalyst temperature of 305 C. and a hydrogen to xylene ratio of 4.2-4.4:1. The catalyst was nickel supported on kieselguhr.

Table IV Run Number 12 13 14 15 15 Average Temperature, C 337 326 339 312 313 Maximum tern erature, "'0 371 374 390 316 316 Liquid Space eloeity 0. 37 0.36 0.38 0.36 0. 36 Throughput, Vols/vol. cat 0.37 0.42 0.35 1. 14 1.26 L We r v1 iqu ecovery, o per cent 68. 4 71.4 66.9 93.9 90.9 Benzene, Vol., per cent. 26.5 4.5 4. 1 Toluene, Vol., per cent. 39. 5 28.5 31.9 Hydrogen Charged, parta- 21. 5 24. 9 19.5 67. 7 74.8 Exit Gas, parts l8. 9 22.7 17.4 68. 1 74.9 Hydrogen, per cent. 39. 8 6. 5 87.2 Methane, per cent. 57.9 91. 2 10. 5 Ethane, per cent--- 2. 39 2.3 2.2 Eflieiency, per cent. 80 80 80 98 98 Three runs (12, 13, and 14) combined for distillation analysis.

From the above runs, it is evident that while a high degree of demethylation is secured following hydrogen pretreatment, a high ultimate recycle yield could not be obtained under these conditions, because of the low liquid recovery.

Various other runs which were conducted illus-. trate that the higher hydrogen to hydrocarbon mole ratios give higher conversions. Thus while hydrogen-to-hydrocarbon mole ratios in the range of 1:1 and 5:1 are desirable, a ratio of about 4:1 is preferable.

The following table gives the results of a series of runs carried out at an initial temperature 01 be seen that an increased liquid recovery is attained with the xylene pretreatment of the catalyst.

Table VI shows results obtained with the alternative expedient 01 following hydrogen reduction with pretreatment of the catalyst (same as in preceding table) in a nitrogen stream. In the three examples, the catalyst was pretreated at 305 C. for 18 hoursin a nitrogen stream. Between each run, 1. e. between 24 and 25, and 25 and 26, sufllcient hydrogen was passed over the catalyst normally to produce an initial decomposition reaction. The hydrogen-xylene ratios were as in the precedin table.

Table VI Run Number 24 25 26 Average Temperature, C 311 312 314 Maximum Temperature, G. 315 314 316 Liquid Space Velocit 0. 36 0. 37 0. 36 Throughput, VoL/Vo Cat 1. 27 1. 28 1. 16

Liquid Product:

Liquid Recovery, Vol. per cent 94. 5 92. 1 90. 8 Benzene, Vol. per cent.-. 3. 5 3. 5 5. 5 Toluene, Vol 23. 5 29. 5 33. 5 Hydrogen Charge 74. 8 75. 1 69. 5 Exit Gas, parts 74. 8 75. 4 70. 1 5y gen" 90. 5 87. 7 85.2 at 8. 0 10.0 12. 7 Ethane 1. 5 2.3 1.9 Eiiieiency, per con 98 98 98 It is manifest from the two preceding tables that where the catalyst has been pretreated, for example, with hydrocarbon charge alone or with nitrogen, there is no' appreciable gas volume contraction. However, subsequent prolonged treatment or the catalyst with hydrogen will again produce a contraction.

Following run 26, the catalyst was subjected to a hydrogen treatment 23 times that normally required to produce the exothermic reaction. The

305 C. and with hydrogen-xylene ratios 0! results areshown in the following table:

- homologue which comprises contacting a mixture 7 Table VII Run Number 27 Average temperature, "C 344 Maximum temperature, C 985 liquid space veocity 0.88 Throughput, vols/vol. cat 0.50 Liquid product:

Liquid recovery, vol. ,per cent 62.1 Benzene, vol. per cent 23.6 Toluene, vol. per cent 34.3 Hydrogen charges, parts 28.7 Exit gas, par 24.7 Hydrogen 25.0 Methane 72.3 Ethane+ 2.7 Eiliciency ..-per cent..- 71

The hydrogen-xylene ratio was the same as in Table V. A considerable contraction in gas volume, and a marked rise in temperature are evident.

In the above Tables IV, VI and VII are given per cent efliciencies. These are calculated as the moles of liquid product divided by the moles of charge times 100. During the highly exothermic reaction shown in runs 12, 13 and 14 efficiency was relatively low. Run 27 has a very low efllciency since it is an example where the cata lyst was purposely pretreated /tvith hydrogen.

When this highly exothermic phase does not occur, the efllciency is quite high. Where the catalyst had been previously flushed with nitrogen, as in runs 24, 25 and 26, efllciency was 98 per cent. Similarly, the series of runs shown in Table V had an average efficiency of 95 per cent. In' runs of relatively long duration after the exothermic reaction has run its course, high efliciency may subsequently be attained as in runs 15 and 16 where the average efflciency was 98 per cent.

What we claim is:

1. In a processfor catalytically dealkylating a high boiling alkyl aromatic hydrocarbon into a lower boiling homologue while in contact with hydrogen, a metallic catalyst of the eighth group of the periodictable, and at a temperature in the range 250 to 400 C., the step which comprises treating the catalyst in the absence or hydrogen with said high boiling alkyl aromatic hydrocarbon.

containing hydrogen and the said hydrocarbon with an activated nickel catalyst at a temperature of 250-400 C., and periodically eliminating the hydrogen from the mixture while continuing contact between the catalyst and the alkyl aromatic hydrocarbon.

3. The method or converting a high boiling alkyl aromatic hydrocarbon into a lower boiling homologue which comprises passing a mixture of hydrogen and the said hydrocarbon over an activated nickel-on-kieselguhr catalyst at a temperature of 250 to 400 C., and periodically replacing the said mixture by a current of nitrogen 88.8.

4. The method of dealkylating a high boiling alkyi aromatic hydrocarbon into a lower boiling homologue which comprises contacting a mixture containing hydrogen and the said hydrocarbon with an activated dealkylation metal catalyst of the eighth group of the periodic system at a temperature of 250 to 400 C., periodically testing the catalyst temperature, and periodically treating the catalyst in the absence of hydrogen with a fluid which will remove hydrogen from the surface of the catalyst and thus reduce its activity whenever said catalyst temperature reaches a predetermined maximum.

' CHARLES W. MONTGOMERY.

WILLIAM A. HORNE.

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

UNITED STATES PATENTS Number Name Date 2,295,672 Meharg et a1 Sept. 15, 1942 2,344,449 Ogorzaly Mar. 14, 1944 2,381,677 Matuszak Aug. 7, 1945 2,422,673 Haensel et a1 June 24, 1947 2,422,674 Haensel et al June 24, 1947 FOREIGN PATENTS Number Country Date 127,690 Switzerland Sept. 1, 1928 Certificate of Correction Patent No. 2,470,712. 'May 17, 1949.

CHARLES W. MONTGOMERY ET AL. V It is hereby certified that errors appear in the printed specification of the above numbered patent requiring correction asfollows:

Column 4, line 21, for the words supported or read supported on column 5, Table IV, under the heading 13?, next to the last line, for 2.39 read 2.3 column 7, line 5, Table VII, for veocity read velocity;

and that the said Letters Patent should be read with these corrections therein that the same may conform to the recordof the case in the Patent Office.

Signed and sealed this 11th day of October, A. D. 1949.

THOMAS F. MURPHY,

Assistant Oommz'esz'oner of Patents.