Chemical Refining of Aromatic Hydrocarbons from Petroleum

up to thetime of World War II, toluene treating has generally ... With the advent of World War II and the unprecedented de- ..... in connection with t...
1 downloads 0 Views 2MB Size
Ehtractive,Uistillatiori Lnit for Production of Nitration-Grade Toluene

Chemical Refining of Aromatic Hydrocarbons from Petroleum -

J

TOLUENE TREATING P. L. B U N D T , R. J. LEE, AND F. T. WADSWORTH Pan American. Rejining Corporation, Texas C i t y , Tex.

URING the past several decades it has been coiiiinoii practice to subject crude or raw grades of benzene, toluene, xylene, and various solvent naphthas t o sulfuric acid refining in the process of producing “pure” or refined grades of these aromatic hydrocarbons. Such treatment has been used to render the hydrocarbons suitable for nitration purposes arid for certain solvent and chemical applications. I n the case of aroniaiic hydrocarbons produced for subsequent nitration, particulnrly toluene for trinitrotoluene numufacture, the purpose of suliaric acid treating has been t o remove unsaturated hydrocarbons which, if not removed, might lead to the production of an unstable nitrated product or to possible hazards in the nitration operations. The: e are, however, practically nn published data dealing with the basic considerations of toluene tre:ttiiig, asidc from a few generalized statements (1, 13, 15) concerning the removal of olefins or unsaturates from toluene by acid trcating. It appears t h a t the practice has evolved in the coal t a r industry, and, up t o the time of Korld TYar 11, toluene treating has generally been carried out in large batch agitators (15) in what appears to

have been a rather inefficient mariner, from the standpoint of acid requireinenis arid t,olueiie losses. K i t h the advent of Korld War I1 and the unprecedented di:. rnand for nitration-grade toluene, ivhich was predominantly sui)plied for the first time by the petroleum industry, many problenih were encountered in producing toluene meeting nitration-gratlt, acid-wash color specifications. This was particularly true o i t rwptxct. Also, it was not possible to devise other tolueric treating nietliods which \vould be more satisfactory than sulfuric. acid treating. (:onscrluently, investigat,ions n-ere undertaken i u dctermine the types of compounds responsible for poor aci;lwash color characteristics, and to devise more effective metliods for treating toluene and similar aromatic hydrocarbons.

1010

August 1947

INDUSTRIAL AND ENGINEERING CHEMISTRY

1011

A C I D - U A S H COLOR

Of the many specifications for nit,ration-grade toluene, the acidwash color specification has been one of the most difficult to meet so far as many petroleum refiners have been concerned. This has been a consequence of the lack of inforniation on the causes of poor acid-wash color and the delayed n:tture of the problem arising from instability in storage. The acid-wash color test (4, 7 , 2 f ) , which \\as devised by the coal tar industry, was intended t o indicate whether the content of undesirable unsaturated hydrocarbons in the t o l a n e was suficiently Ion. for nitration purposes. This estreniely s;.nsitive test comprises shaking the aromatic hydrocarh 111 (21 nil.) and sulfuric acid (7 ml.) together for 20 swontls ut 70' * 2' 1'. under specified conditions. After a 15-minute standing follo slight sirirling of the bottle, the c o l ~ rof thc s.ilfLiric ac (the acid Ivash) is compared with a set of B.irt.ett c.ol?r ita!idit,ds (21 i. I n the c:csc of nitrxtioii-gr:itic tolwnr, :t miiiim.tni acidwash color of 2 (light amber) was establislic~l:tq represeiitin)~tolJene of satisfactory purity from the stanc1p:)itit of limiting the content of uns:tturated hydrocarbons and possib!y other acid-rewtive components. Hon.cver, this enipiricil test sets ni e m c t limitation o n the allowable pcrceritage of unsaturated hytlrocarb'xis. atid very little has been known concerning tho type o r types of components particularly delisterious Tvith respect to acid-.msh color. lloreover, no definite specification (such a~ hiomine number) has been set t o limit the c xitent of unwiarated hydrocarbons, other than those limitations inherently imposed by the specific gravity specification (0.871 0.032 at 69/63" F.)and the acid-wash color test. These limitation. hsve been more thzn adequate in the case of toluenc refined by treating n-ith sulfuric acid. since practically complete olefin reinoval was nxesszry in the acid-treating step in ortler to nieilt the acid-:vash color spccifieation. Ho\\evcr, it Iias boeri found t h s t diol sponsible for acid-\rash color difficultic.p, whc have little effect. Hence. n-hen using the improved t,rcating procedures designed t o remove diolefins in )).trticdar, as described later, it may be desirable to establish a limiting mono-olefin content to supplement the acid-wash color s;)ecificntion. The exact magnitude of this specification is a matter vihich should he evaluated in terms of the effect of mono-olefins on nitration of toluene and is heyorid the scope of this paper. H o w v e r , a n unnecessarily rigid limitation in this regard should not he imposed, since it would place an undue burden on the rcfiner and limit toluene supplies in time of emergency. Other acid-wash color specifications have been established by 'The Barrett Company and oihprs for various refined grades of benzene, toluene, xylene, and high-solvency naphthas (7, f3). Kenney (1s) stat,es in this connection that the acid-mash test is a criterion of the degree of purification that a wlvent has received in its manufacture. The impurities removable by acid treating include substances Ivhich give rise t o gummy bodies on aging, heating, or exposure to air. If present i n large amount, the gummy bodies may cause darkening in color. The acid-wash test is, according to Iienney (I.?), of importance to the paint and varnish industry as a measure of the removal of these impurities tn an extent that ensures satisfactory behavior of the solvent.

+

TABLE I. EFFECTOF Time o n Test. HI. 0 0 0 7 2.1 .3 , 1 4 5

S.0 12 0

16.0

T h e problems connected with the chemical refining of toluene from petroleum sources were studied, and the application of the information to the refining of various other aromatic hydrocarbons is indicated. Investigations with specific hydrocarbons showed that the presence of diolefins in toluene is harmful in the acid-wash color test, e\en at low concentrations (0.005-0.01%). Reactions with maleic anhydride confirmed these conclurions. A toluenetreating process using 0.005-3.01 q~maleic anhydride was shown to produce toluene of exceptional acid-wash color characteristics. Vapor-phase treating with U.O.P. phosphohc acid polymerization catalyst was demonstrated as a n effective method of toluene refining. Data are presented on the variables in conventional sulfuric acid treating and o n accelerated test methods to compare the efficiency of diflerent treating conditions and to predict the arid-wash color stability of the product.

It is the belief of the a,.tthws t h l t tho cs'us::~of [ Y J J ~acid-wish color and stsbility, aiid the impr.ived ntcthods of toluciic trezting describ.1 in this paper, xi11 be wizlely a;)plico ble to variuJs other arotmtic h y d r x i r b J n s useJ in the sJlvent fizld. Hence, although the p r e s x t work \\as c o n f i x d to r i i t r ~ ~ t i ~ ) t i toluene, -~r:~~~, its wider i!iipiic.%ti:rn s h o d d ii.)l be ove~looki?d. ACID-W.ASI1 COLOR STABILITY

In addition to the initial acid-wash color of Ircslily produced t,oiiiene. the acid-.xash c a l x sttbility or storago sta!)iiity of the toluene must tx: seriously considered in ev:iluzting the merits of different nesting processes. LIaiiy cases were encountered in the present work, for e.;s;n?Ie, w h e r e freshly p r o d l e d t7luene of petrolcilni origin \vas found t7 li!~vt:sstisfactory initial acid-\yash color, h'Jt its :rcid-wash c a l x d u p t d e d rapidly to an unsatisfactory level in a nistter of a fe;v dsys. To pred ct the stability of toluene \vith respect to acid-wash color d>terioration, studies were iiiade of storage chsracteri~ticsus. t\yo .accelerated aging tests. One such accelerjted test developed by the Humble Oil & Refining Conipsny and the Bzytown 0rdn:mc.e K o r k s (B.O.W.) nhen properly correlated for t d x n c from a given s3ilrce. waa found to enable fairly relirble prediction of the actual stability of the toluene during iiormal storage and shipping periods (at, least 1-3 months). Another test, the Pari hmerican rapid ultraviolet test developed in the course of this work, was found to he a valuable test for routine plant control.

The U.O.\V. accelerated test developed by the Humble Oil I% Refining Company (8)is carried o u t as fo1bn.s: Approximately GOO mi. of the sample to be tested are placed in a flisk equipped with a refiux ondenser and iinmersed in a water bath maintained a t a temperature of 195' * 5' F. Purified air (sulfuric acid scrubber aiid caustic tower) is psssed through a heating coil also immersed in the heating bath and thence into the sample through a tube of 4-5 mm. inside diameter a t a rate such that there is rapid bgbbling. Ground-glass joints are used throughout; joint lilbricsnts should be avoided. Samples are taken initially and a t &hour intervals thereafter for the determination of acid-wash color. The test is run for 24 hilurs, or until the acidvalue in excess of No. 6 standard. Typical AIR A X D OXYGES o s B.O.W. TOLIJENB data are shown i n Table I. The test rcsu1t.s are reported as the acid-wash color after 24 hours-for example, 3-4 A.1V.C. after STABILITY TEST 24 hours-or the time required t o reach 6 i- calor-for example, A.n'.C. on B.0.W. Test 6 + A.W.C. in 16 hours. I t is important t h a t the test equipWith air With 0 2 Heat only ment be set up where i t will not be exposed to direct silnlight or 0-1 0-1 0-1 bright reflected light. Every precaution must also be observed ,.. 9+ 2-3 regarding cleanliness in sampling and avoidtlnce of exposure 3-4 3-4 3-4 ... 3-4 3-4 of the samples to sunlight even for a matter of minutes, since 4 3-4 4 sunlight cause3 extremely rapid degradation of acid-wash color. .. ... 4-5 The importance of this precaution on regular acid-wash color .. 5 .. . 5. . 6+ determinations in routine plant, sampling operations as well cannot be too strongly stressed.

1012

INDUSTRIAL AND ENGINEERING CHEMISTRY

Vol. 39, No. 8

PROCEDURE. A sample of 200 nil. a t 80-90" F. is placed in a 250-ml. beaker under the center of the window area of the lamp, and the wholt, apparatus is covered with a cloth or 6 housed in a closed box. The di+ tance from the quartz tube of the lamp t o t h e surface of the toluencb 5 is 6-7 cm. Samples (25 ml.) are witha di-awn a t 20- or 40-minute interJ vals, and the acid-wash color is deg4 termined until i t has reached 6I iusually 2 hours or less). Since the FIGURE 1 thickness of the toluene layer ip ACID - W A S H COLOR S T A B I L I T Y gradually decreased by this procedurr 130 as the samples are removed, i t would OF TYPICAL S A M P L E S OF 4 appear t h a t the rate of deterioration N I T R A T I O N -GRADE T O L U E N E might vary. It was found in prac2 PAN AMERICAN tice, however, t h a t exposing the total AS M E A S U R E D B Y ULTRA-VIOLET T E S T 200 nil. sample for 2 hours gave practically the same results as were T H E B.O.W. AND PAN AMERICAN I I obtained on the material remaining ACCELERATED TESTS a t the end of 2 hours when 25 ml. samples had been removed a t 20COMPOSITE DATA FROM 9 SAMPLES 1 I minute intervals. eo 40 60 80 100 120 dLTERS.4TE PROCEDURE. For plant EXPOSURE T I M E T O U L T R A - V I O L E T LIGHT- MINUTES control purposes the following method was found- t o be useful as-a rapid check on the quality of production: Thirty milliliters of the sample in a 5O-ml. beaker arc placed directly under the center of the window The effect of bubbling air 01'oxygen through the sample in the area of the lamp. The whole apparatus is then covered with a B.O.W. test was found to have little effect on the results obtained, heavy cloth. The sample is exposed for 40 or 60 minutes, and thc as compared with the accelerating action of heat alone. This is acid-wash color determined. T h e results are reported as acid-wash shoTm for a typical commercial nitration-grade toluene sample in color after 40 or 60 minutes of ultraviolet exposure. Table I, with the temperature at 200" F. However, the original It has been our general wperience in plant control work t,liat, if test procedure using air as described was folloved in all work the toluene s h o w a n acid-wash color of 3 or less after 40-minute reported here. exposure, the stability will be satisfactory. Unstable and unsatBased on extensive experience ryith toluene produced by hydroisfactory toluene will show a n acid-wash color of 4-6 in 40 minutrc forming followed by extractive distillation (phenol system) and or less by this test, although the initial acid-wash color may be mild sulfuric acid treating (3-5 pounds per barrel), the B.O.X. completely satisfactory (0-2). -4comparison of the results obtest results are interpreted qualitatively as follows: excellent and ra,incd by the B.O.K. and ultravielet tests is shown in Figure 1. unusual stability if the toluene sample showed an acid-wash color This figure should not, be construed to be a general correlatioii (A.W.C.) of 1-2 after 24 hours on test; good stability for an chart between the two methods, since different samples lvere ob.1.K.C. of 3-6 aft,er 24 hours; satisfactory if A.7T.C. is 6 after served t o show some divergence when tested by the tn-0 methods. 12-20 hours; and questionable or unsatisfactory stability if The correlation between the ultraviolet and B.O.W. tests showii A.W.C. is 6 in less than 12 hours by this accelerated method. in Figure 1 holds fairly well for the first 30-10 minutes of ultraA second accelerated test for acid-wash color stabilit'y evaluaviolet exposure. Thereafter the increase in acid-wash color may tion-the P a n American ultzaviolet test-was developed based be slower on some samples on the ultraviolet test than would lx* upon the accelerating effect of ultraviolet light on acid-wash color predicted from Figure 1. deterioration. The advantage of the ultraviolet t i s t over the B.O.K. procedure lies in the fact that a sample can be evaluated in 30-120 minutes as compared to 8-24 hours am5 0.100 0.125 a50 -0.175 CALCULATED B R p HO. I by the B.O.K. test,. For this reason 1 I 5 5 the ultraviolet method is most used I I ful in plant control. However, the 2 METHYL PEHTADIENE B.O.K. test is considerably more sensitive to slight differences in I stability and is usually relied upon in connection wkh tank-car shipI VOLUME m 2-METHYL. P E N T A D I E N E 1 ments.. Because of this greater sensitivity, the B.O.W. method was used as the control test in all investigations reported here regarding improved treating methods and in studies of the factors influencing acid treating and poor acid-wash color. The P a n American test is carried out in the following manner: 27r

4

B.O.W.

ACCELERATED 12

TEST-

H O U R S ON 16

TEST

20

24

s

APPARATUS.Mineralite ultraviolet lamp (Model V41, manufactured by Ultra-Violet Products, Inc.) u-as obtained from the Braun Corporation as No. 361 Mineralightlamp. T h e equipment consisted 'of a 250-ml. beaker, a 50-ml. beaker, and a 25-ml. pipet.

P

I ' Figure 2.

I

[VOLUME

GIGLOPENTIDIEWE

I

I

J

Effect of Diolefins on Acid-Wash Color of Toluene-Diolefin Blends Made Up in Olefin-Free Toluene

OF TABLE 11. EFFECT

h l O X 0 - A S D DIOLEFINS Oh' COLOROF TOLUEKE

ACID-FF-ASH

.I.W.C. of Toluene with Concn. of Added Compound, W

1013

INDUSTRIAL AND E N G I N E E R I N G CHEMISTRY

August 1947

Mixed Cs mono-olefins

2-Methylpentadiene

-

Cyclopentadiene

0

0.000 0.005 0.001 0,002

0

0

...

...

...

0-1 2

0-1

3+

0.003 0.004 0.006 0.008

...

1

0.016 0.032 0.064 0.167

...

b'

0.67 0.79 1.00 1.96 3.85

1 2+ 3 4+ 5-6

5.66

6 f

...

...

, . .

1

4+

...

3 3

6' '

4 i 5 6

.. ...

,..

... ...

...

...

...

...

... ...

. .

...

...

, . .

...

COWPON EKTS RESPONSIBLE FOR U S SAT1 S FACTORY R E S U L T S

The cause of unsatisfactory acid-wash color in the case of refined toluene has been attributed in a general way t o unsat'urat'ed hydrocarbons, principally olefins (1;). However, eshaustive st,udies by Brooks and Humphrey ( 2 ) on the refining of cracked distillates shon-ed the importance of removing reactive tar- and resin-forming diolefins for the production of satisfactory motor gasoline. Similar data by numerous other investigators (1, 6 , 16, 17) also indicated the deleterious nature of diolefins in gasoline with respect t o color (not acid-wash color) and gum formation. B?; implicat,ion, therefore, it was suspected that diolefins might be responsible for unsatisfactory acid-wash color characteristics in the case of toluene. KO reports were found in the literature, however, of attempts to determine the effect of the various classes of unsaturates on the acid-imsh color of hluene, so that the more deleterious class could be singled out. The problems of finding the causes of unsatisfactory acid-wash color of toluene was attacked from two standpoinh: ( a ) an investigation to determine the type or types of compounds which would lead to poor acid-wash color and the concentration at m-hich they would be detrimental, and ( b ) a study involving the treatment of unrefined toluene of poor ( 6 f ) acid-wash color with reagents and under conditions known t,o remove specifically or react with various types of diolefins. EFFECTOF SELECTEDl l o s o - AND DIOLEFISS.I n order t o determine the effect of olefins and diolefins on acid-xvash color and acid-xash color stability, blends of several individual olefins and diolefins were made with a highly purified sample of toluene. The components of t,he blends n-ere prepared as falloffs: 1. Olefin- and diolefin-free toluene: Best quality nitrationgrade toluene was further exhaustively treated with 98% sulfuric acid, caustic-washed, and distilled slowly from a flask containing maleic anhydride. The distillate y a s finally washed with aqueous caustic and water. 2. Isomeric eight-carbon branched-chain mono-olefins (codimer) : A sample of codimer produced b y t h e polymerization of mixed butenes was fractionated through a 35-plate laboratory column. T h e fraction boiling 225-235" F. was separated for this work and treated with 5% maleic anhydride for 2 hours a t 200" F. This material possessed a bromine number of 133; this showed i t t o contain virtually 1 0 0 ~ olefins. o 3. Aliphatic diolefin (2-methylpentadiene) : This compound, purchased from Commercial Solvents Corporation, was a mixture of t'he difficultlyseparable P-methyl-l,3-pentadiene and Z-met,hyl2,4-pentadiene. The sample was redistilled prior t o use. 4. Cyclic diolefin (cyclopentadiene) : This material was freshly repared from t,he dimer, dicyclopentadiene (Eastman Kodak eompany) , by depolymerization over stainless steel turnings a t 350" F. followed b y fractionation.

I n preparing the blends of the diolefins, stock solutions of 1 c; 2-methylpentadiene and cyclopentadiene, respectively, were madt, with the toluene (olefin- and diolefin-free) and aliquot portions made up with additional toluene t o give t,he desired concentration. The acid-wash color and accelerated acid-wash color stability were determined on t,he blends as previously described. It XTas soon evident that extremely small quitntities of diolefius had a pronounced effect on the acid-wash color. After orientation as t,othe concent,ration range required t o keep t,he blends Tithin the range of the acid-ivash color standards, a series of blend. were prepared t o show the effect of concentra,tion of the nionoand diolefins on the acid-wash color. These data are shown in Table I1 and graphically in Figure 2. Very low concentrations ( O . O O 1 - O . O 1 ~ o ) O F diolefins haw a serious effect on the acid-%ash color of toluene, whereas CSmonoolefins, as represented by codimer, are relatively ineffectual even in concentrations as high as 0.7% corresponding to a bromine number of 1.1, Since diolefins are detrimentid in estreinely low concentrations, and since mono-olefins appear to have practically no effect on acid-xT-ash color a t t,he concentra,tions normally m- ' countered, bromine number (or other comparable' measure of olefin content) has no significance in predicting acid-wash color or color stability. Furthermore, the more reactive cyclic diolefin (cyclopentadiene) shows a pronounced effect at a considerably lower concentration (0.001-0.002~0)than the aliphatic diolefin (2-methyl pentadiene). This fact Till be referred to again in connect,ion with maleic anhydride treatment of raw toluene. I n hypot'hetical refining of cyclopent,adiene blends from a 6 acid-wash color (0.008% cyclopentadiene) to a O + color, a bromine number reduction of only about 0.05 unit is iiivolved: xhereas with the aliphatic diolefin the same reduction in acidwash color would be equiva1en.t t o a reduction of 0.2 bromine number. A comparable effect was noticed in the catalytic trcating of a sample of commercial toluene. For example, in this csatalytic treatment, of raiT toluene the acid-wash color was reduced from 6+ to 1 while the bromine number reduction amounted to only about 0.054.1 unit,. This observation is advanced as further confirmatory evidence that reactive diolefins, probably cyclic in nature, are the main cause of unsatisfartorl: acid-m-ash color.

ACID-TT-.ISH COLORSTABILITY L'S. JIOSO-OI~EFIN 4x11DIOLWIS COSTEST. The authors' data on the effect of mono- and diolefins on the acid-wash color stability are not so complete as would be desired. H o m v e r , a number of blends of both types of unsaturat,es with toluene were prepared and the stability measured by the B.0.K. accelerated test, so that, the trend is quite apparent. The data are shown in Table 111. The diolefins also contribute to poor stability of the tolueiie, w-hereas mono-olefins seem to have little efiect. I n the latter case the acid-xvash color remains very close to the initial level after accelerated aging. For comparison the data for three tvpi-

OF bloxoTABLE111. EFFECT

DIOLEFINS ON ACID-WASE

AND

COLOR ST.4BILITY

O F TOLC13NE

B . 0 . K . Accelerated A.W.('.

Added Compound Sone CII mono-olefins CBmono-olefins 2-Methvlnantadiene P-i&thyl&ntadiene 2-hlethylpentadjene CBmono-olefins Typical T a n k

Car Shipment

Vol. yo Added

0.00 0.17 1.00 0 006 0.10

g:y;5}

Bromine No.

h.KC. of Blend

0 0 3 2 6' 3

Initial A.W.C.

___ Stability Test

24

4

8

12

hr.

hr.

hr.

16 hr.

20 hr.

hr

o+

Of 0-1 3-4 2

O t 0-1

0-1

0-1

0-1 3-4 2

0-1 3 4 2-3

0-1 3-4 2-3

4-5

6+

3 i

4

5-6

6f

3 3-4 ~~

...

. . . . . . . . . . . . . . . . . .

___ A . R . C . Stability 4 hr.

8 hr.

12 hr.

... ~-

16 h r

I N D U S T R I A L A N D E N G I N E E R I N G CHEMISTRY

1014

cal toluene samples from commercial production are included. These \\-ere selected to substantiate a previous statement that bromine number is no index of the acid-wash color stability of the sample. Sample A is definitely of questionable stability, although i t had the lowest bromine number of the group. I n the light of these data and the fact that toluene produced by hydroforming normally contains only relatively small amounts of diolefins as compared to mono-olefins, which are normally present in concentrations of 1% or less, the most important function of acid treating appears to be the removal of small quantities of diolefins (0.170 or less). It is evident, therefore, that intimate and thorough contacting of sulfuric acid and toluene arc nccesszry for the successful and efficient application of acid treating, cspecially if small amounts of treating acid (that is, 3 pounds per barrel of toluene) are used. E F F E C T OF MALEIC ANHYDRIDE T R E A T l l E N T

,

From the data just presented on the effects of mono- and diolefins on the acid-Tvash color of toluene, it appears that conjugated diolefins can be maiuly responsible for poor acid-wash color and unsatisfactory acid-wash color stability. Since maleic anhydride is fairly specific for the removal of conjugated diolefins (9, .?I, treatment with maleic anhydride should result in a product of improved acid-wash c o l x and improvcd stability, if conjug:itcd diolefins are present in the toluene and if they are the compounds mainly responsible for bad acid-wash color. Furthermore, the type of diolefin (cyclic or aliphatic) influences the rate and tcmperature a t which reaction with maleic anhydride occurs ( I , 3). For this reason a study of the effect of time and temperature in the treatment of raw toluene should give additional information useful in explaining the c:iuses and peculiarities of poor acid-wash color. 5 I A L E I C h - H Y D R I D E T R E A T l f E N T O F R.l\\- TOLCESE ASIJ TOLUENE OF POOR ACID-KASHCOLORSTABILI,TY.A sample of

plant acid-treated toluene of poor stability was trcatcd with 0.28 weight yo of maleic anhydride for 2 hours at 200' F. Unreactcd maleic anhydride and reaction products were removed by washing with 15% potassium hydroxide, and the treated toluenc !vas distilled. A control sample was given identical treatment cscept that no maleic anhydride was used. Table IV shows the surprisingly good stability of the toluene treated ivith maleic anhydride a.s compared with the poor stability of the control sample.

TABLE Iv.

l\IALEIC .4SHY'DRIDE 200" F. FOR

TREATMEST O F TOLUEXE AT 2 HOURS B.O.W. Accelerated A.W.C.

Srtiiiple Treated

.

hlnleic Anhydride 4 Lsed, Yo Initial hr.

Stahility Test 8 12 16 hr. hr. h r .

6f

O+

Of

..,

.

24 lir.

h+

1

... ,.

,

Acid-treated toluene Same, X1.A.-treated

?*'one 0.28

Of

R a w toluene

Same, M.d.-treated Same, hl..i.-treated

None 0.20 1.0

6-t 2 1

2 3-4 1--2 1-2

4-5 1-2

5-6 2-

%+ 3+

Typical tank rnr shipment of satisfactory toluene

h-one

0-1

2-3

5-6

6 i

..

Of

3-4

'0 hr.

, ,

4-5

O+ ,

0-

.. ...

,

.,

.,

I n view of the effectiveness of maleic anhydride treatment in stabilizing acid-treated toluene, some work on the treatment of raw toluene (not previously acid-treated) seemed in order. Samples of raw tolucnc of 6+ acid-wash color n-ere'treatcd in one h of maleic anhycase with 1% and in another example ~ i t 0.2% dride for 2 hours a t 200" F. Caustic Rashing folloncd; no distillation step was used. Other data in Table I V show that treatment with maleic anhydride alone produced a color-stable toluene n-itho u t sulfuric acid treatment; this is convincing evidence that con-

Vol. 39, No. 8

jugated diolefins are responsible for the unsatisfactory acid-wash color and color instability of incompletely refined toluene. The evidence of the presence of a reactive conjugated dioleki in ram toluene was just presented. If this type of diolefin is present, maleic anhydride treatment a t room temperaturo (8090" F.) rather than a t 200" F. should be effective in reducing the acid-bvash color and stabilizing the toluene. This proved to be the case. For example, raw toluene of 6+ A.W.C. was trcatcd n-ith 1% maleic anhydride for 20 minutes a t 80" F. The treated tolueno was caustic-mashed and filtercd; acid-wash color and color stability n w e thcn detcrmincd. The acid-wash color after treatment was 2-3, and a 3+ .4.K.C. was obtained after 24 hours on the B.O.K. accelerated stability test. Removal of the reactive diakfiiis is sccn to improve the initial acid-wash color from G I to2-3. Of evc~igr.caterinlerr.jtis tlie fact that the unsntxates, somc of which are probably relatively unrcactivc diolefins, remaining altcr this maleic anhydride treatment do not contribut: appreciably to poor acid-wash color stal ility. I t sccms thcn thbt if the most reactive diolefius are removed from the r A w toluene, even though the acid-wash color is not reduced to 0-1, the toluane is cstremoly stable agzinst further dctcrioration. This fart is confirrned in the work on treatment of raw toluene n i t h phosphoric acid polyincrization catalyst ticscribed later. RESTOR.ATIOSO F .icrD-\V.isrI COLOROF FINISHED TOLVENE DETERIORATEIJ TO G + -4.K.C. The mechanism of the reaction which accounts for the deterioration of acid-wash color is obscure. The nature of the dcgrsdation products formed h y this rciletion are a h unknown. To clarify this situation at least partially, a sample of refined toluene (initial A.lYC. = 0 + ) was deteriorated to G + A.W.C. by means of the B.0.W. acccleratcd t e s t by heating at 200" F for 16 hours. The resultant toluene sample, now of 6+ AL\V.C.,was given the following successive maleic anhydride treatments: llaleic .\nhydride, "0

Sone 1 3 3

Tiine of Reaction. Hr.

Temp..

.i,\V.C. a f t e r Trea t n i r n t d h o n 11

...

6+

O ' i

200 200

200

3+ 3

2 0 5.0

F.

4t

I t \vould appear that the degradation products causing the deterioration in acid-ivash color are not entirely compounds reactive with maleic anhydridc, since even the comparatively drastic succession of maleic anhydride treatments described did not reduce t h a t color to the original value (O+). Rloreover, the compounds present in unrefined t,oluene which are responsible for the high ALW.C. are not the same as the compounds present in refined t'oluene xhich has deteriorated in acid-wash color; the former can be practically completely removed Ivith maleic anhydride treatment, wlicrem the latter cannot. This suggests that h.1V.C.deterioration is caused, a t least in part, by the polymerization of conjugated diolefins t o higher polymers of nonconjugated structure. EFFECT O F TREATMEST TINE: \\-ITH 1 I A L P : I C LkNHYDRIDC. It was shon-n in the prcccding discussion that the acid-wash color of unrefined toluene could be reduccd from 6+ to 1 by means of treatment solely with 1% maleic anhydride. The stability of the tolucnc so trcatcd wns found to be much superior to that of toluene generally produced by commercial acid treating. The effect of tinie of maleic anhydride treatment on the reduction in acid-wash color of raw toluene (6+ initial A.W.C.) was studied nest. h sample of raw toluene was heatcd to 200' F. a t vvhich t,ime 1% by ncight of maleic anhydride was added and vigorously agitated; 25-ml. ssmples were withdrawn a t regular time interval3 and immediately shaken with 15% potassium hydroxide to quench the reaction. The toluene layer was separated and filtered to remove suspended water; after t h a t the acid-wash color was determined, The treated toluene was not distilled, since distillation was shown t o be unnecessary with maleic anhydride treatment. The data, shown in the following

INDUSTRIAL AND ENGINEERING CHEMISTRY

August 1947

table. rrere obtained for raw toluene (bromine number 0.31) treated Kith 1% by weight of maleic anhydride a t 200' F.: Time of Treating. Iliii.

Acid-Wash Color

0 1 2 3 5

6+ 3T 3+ 3 3

Time of Trea'ing. Min. 7 10 20

Acid-Wash Color 2 1 2 1

120

The initial rate of removal of the undesirable diolefins is rapid since the acid-wash color decreased frum 6 + to 3+ during 1 niinuw of treatment, whereas 23 minutes v x r e required for a further decrease to 2 A.1Y.C. and 100 additional minutes for 1 4.W.C. From the rate of reaction observed, there is evidence of the prcsence Of two or more types of conjugated diolefins-that is, a very reactive diolefin, probably cyclic, which reacts Tvith maleic anhydride dniost instantaneously, and a less reactive type which reacts o n l ~ ,very slon.ly-. Similar conclusions \vcw ren.ched by Birch and Scott ( f ) in studies concerned irith the identification ol' the diolefins in the low-boiling fractions of vapor phase cracked gasoline. It is also of possihle interest that Iiranier and Spillm (14) isolattxl thc rtvictivr c ~ - c l o ~ ~ c n t a i l ifrom i ~ n c coke-oven twnzene. STIHILlZh'l'lON OF AC;II)-TREATEL) TOLCl