Composition of Tar from Low–Temperature Carbonization of Utah

Citation data is made available by participants in CrossRef's Cited-by Linking service. For a more comprehensive list of citations to this article, us...
0 downloads 0 Views 743KB Size
IXDC'STRIAL A N D ENGINEERING CHEMISTRY

392

Vol. 20, No. 4

Composition of Tar from Low-Temperature Carbonization of Utah Coal-I' R. L. Brown2 and B. F. Brantings PITTSBL-RGH EXPERIMENT STATION, U. S. BUREAU OF MINES,PITTSBURGH, PA.

This i s the second in a series of codperative studies of the composition of the tar obtained from Mesa Verde (Utah) coal when carbonized with steam superheated to about 700" C. This investigation covers the examination of the liquid portion (tar I ) , and also total tar soluble zn the aqueous condensate. HE recent study by Karrick of low-temperature

T

carbonization of Utah coal from the Castle Gate bed4 permitted the collection of samples of tar, oil, gas, and other by-products produced under definitely known conditions. The examination of these light oils, aqueous condensate, and the various tar portions is being carried out under a series of cooperative fellowships between the Bureau of Mines, the Carnegie Institute of Technology, and the National Coal Association. The first investigation covered the composition of the light oils in the gas.4 The second investigation covered the examination of the aqueous condensate and of the medium oil called tar I, and the results are reported in this paper. The third stage of the investigation, dealing with heavier portions of the tar, is in progress. The analysis of the coal and the conditions under which it was distilled have been published in detail4 and will be but briefly outlined here. The coal was distilled in a continuous retort 13.5 feet (4 meters) in length and tapered from 5 inches (12.7 cm.) in diameter a t the top to 7 inches (17.8 cm.) a t the bottom. The medium of heat transfer was superheated steam introduced into the retort a t the bottom and the maximum temperature maintained was about 700" C. The products of distillation and the spent steam were collected in order in an air-cooled condenser, a water-cooled condenser, a liquid-air condensing train, and in gas-sampling devices. Table I-Yields

of Distillation Products PERCEXTB Y WT. OF C O A L

Coal used Coke produced Gas produced Gas passed through licluid-air train Aquious condensate, total Total tar Light oils (from 30 642 liters of gas) (liquid-air system) 0 Adjusted to total gas.

324.8kg. 211.8 kg. 36,108liters 30,642liters 506.7 kg. 44.0 kg. 1926 cc.

...

65.2

... ...

13:55

O.4ga

The light oil, as has been d e ~ c r i b e dwas , ~ collected in the liquid-air train. The tar was collected a t two places. The medium oil (called tar portion I) was recovered from the water-cooled condenser. It was a liquid tar. From the air condenser was obtained the larger portion, which was viscous, and of its own accord, while still hot, separated into two portions, called tar portion I1 and tar portion 111. Tar I1 was lighter than water-the condensed steam-and tar portion I11 was heavier. Both portions set solid on cooling. The aqueous condensates from both condensers were mixed and a composite sample taken for examination. 1 Presented before the Division of Gas and Fuel Chemistry at the 74th Meeting of the American Chemical Society, Detroit, Mich., September 5 to 10, 1927. Published with the approval of the Director, U. S. Bureau of Mines, Carnegie Institute of Technology and Mining Advisory Board (not subject to copyright). 2 Organic chemist, Pittsburgh Experiment Station, U. S. Bureau of Mines. 8 Research fellow, Carnegie Institute of Technology, 1926-27. 4 For composition of this coal and other details of distillation, condensation, and collection of products see Brown and Cooper, Ind. Eng. Chem., 19, 26 (1927); also, Frey and \'ant, Ibid., 19, 21,488,492 (1927).

of Tar i n t o Portions Described

Table 11-Distribution

As PER As PERCENT CEKTOF WEIGHT OF COAL TOTAL T.AR

TARPORTION I I1 I11

Medium oil Heavy tar: Sp. gr. < water Sp. gr. > water

Kg. 12.70

3 91

28 8 5

17.75 13 55

5 46 4.17

40 35 30 80

This paper presents the examination of (1) the aqueous condensate and ( 2 ) the medium oil (tar I), as it collected in the water cooled condenser. The literature on low-temperature tar has recently been fully reviewed by Parrish.6 AQUEOUS CONDENSATE

A few investigators have reported on the composition of the aqueous condensate obtained in the process of lowtemperature carbonization. The distillate examined by Wheeler and Jones was shown to contain ammonium chloride and hydrochloric acid. The aqueous condensate from the vacuum tar of Pictet and Bouvier was acidic and contained no ammonia. The water from the tar of Britain, Rowe, and Sinnatt had a faintly acid reaction and was found to contain, besides ammonia, pyridine bases, carboxylic acids, traces of aldehydes and ketones, and phenols or other hydroxy compounds. The aqueous distillates of this investigation were neutral to litmus but contained small amounts of bases, carboxylic acids, and tar acids. The concentrations were small, because the distillation of the coal was accomplished by means of superheated steam-1.5 pounds of steam per pound of coal carbonized. Only the bases and the phenols were studied. Separation of Groups in Aqueous Condensate Ten liters of a composite sample of the aqueous condensate were thoroughly extracted with ether. The volume of the combined ethereal extracts was then reduced by distillation to about 2 liters. The material so extracted was then separated into classes as described later for the tar proper. Table 111-Materials COMPOUND

Present in Aqueous Condensate

Carboxylic acids Phenols Pyridines and ether-soluble bases (HC1 salt) Neutral oil

Grams gcr 10 Liters

3.9 23.0 0.824 1.0

The carboxylic acids were of a tarry, gummy type and apparently of a complex nature. The neutral oil (0.1 gram per liter) was a viscous brown oil which began to crack a t about 200" C. Both were available in very small amounts and mere not studied. Phenols in Aqueous Condensate The phenols, after being dried, were carefully fractionated a t atmospheric pressure; cuts were made a t 180" and a t 290" C., at which point the pressure was reduced to 5 mm. and the distillation continued to 300" C., leaving B resinous residue of 5.5 grams. A second distillation gave the fractions for examination, which are shown in Table II-. 5

Fziel, 6, 435 (1926).

April, 1928

INDUSTRIAL AND ENGINEERING CHEMISTRY Table IV-Distillation of Phenol Fractions AMOUNTBOILINGPOINTDENSITY COMPOSITION Grams c. 2.0 178-185 1 , 0 2 2 Phenol 5.5 185-187 1.0605 Largely phenol; some o-cresol; some pcresol; trace mcresol 4.0 192-211 1,0408 Cresol, mainly para and meta, trace of ortho 1.75 211 up Catechol; xylenols (?) 2.75 Resinous

FRACTIOS B-1 B-2

B-3

B-4 Residue Residue from first distillation

5.5

> 306

(5 mm.)

Resinified residue

EXAMINATION OF PHENOLFRacTIoNs-Fraction B-1. A few drops treated with bromine gave a voluminous white precipitate which, purified, melted a t 92.5' C. A mixture of the precipitate and pure tribromophenol (m. p. 93-93.5' C.) melted a t 92-92.5" C. It is therefore apparent that fraction B-1 is largely phenol. Fraction B-2. The physical properties compared with phenol were, respectively: boiling point, 185.5-186.5" C., 184" C.; di6, 1.0605, 1.0708. It gave a tribromide with the same melting point as B-1. This fraction may, therefore, be said to consist largely of phenol. A portion of this fraction was examined for the cresols by the method of Gluud and Breuer6 with modifications. A sample (1.78 grams) was mixed with a slight excess of monochloroacetic acid and 10 CE. of 25 per cent sodium hydroxide solution and boiled 7 hours under a reflux condenser. It solidified on cooling. It was filtered a t 40" C. (filtrate A). The crystals obtained were dissolved in water, the solution made acid with hydrochloric acid, and the resultant crystalline precipitate filtered. Two recrystallizations brought the melting point to 134.5-135.5' C. (cor.). When mixed with p-methylphenoxyacetic acid (m. p. 135-136" C.) it melted a t 135-136' C. (cor.). Para-cresol was therefore present in this fraction. The alkaline filtrate, A, when acidified, yielded crystals and an oil. The crystals, recrystallized once each from acetone and xylene and twice from benzene, melted at 147.5149.5" C. (cor.). Mixed with pure o-methylphenoxyacetic acid they melted a t 150-152.5" C. (cor.). Ortho-cresol was present in the original fraction. The benzene filtrates from the crystallizations yielded the meta derivative, which, recrystallized from water, melted a t 99.4-101.4' C. (cor.). Mixed with pure meta-cresol (m. p. 101-102" C.) i t melted a t 98.4-100.4' C. Fraction B-2 was therefore considered as being composed chiefly of phenol, together with some para- and ortho-. and a trace of meta-cresol. Fraction B-3 was examined for ortho-, meta-, and paracresols by the method used for B-2, and all three were found. The meta-cresol present as determined by the Raschig' method amounted to about 12 per cent. Fraction B-4. About half of this fraction was a mass of white rhombic crystals. Their melting point was 103.4104.4" C. This crystalline material gave a grass-green coloration with ferric chloride which turned red in sodium carbonate solution with ammonia.* The green solution with ferric chloride gave a violet A solution of the crystals gave a heavy white precipitate with lead acetate. It readily reduced silver nitrate. These are all indications of catechol. When mixed with pure catechol (m. p. 104.5" C.) it melted a t 103.5-104.5' C. Catechol was therefore present in this fraction and constituted a principal part of it. Nothing Ges Abhandl Kenntnis Kohle, 2, 236 (1917). ' 2 . angew. Chem., 13, 769 (1900). Mulliken, "Identification of Pure Organic Compounds," Vol. I, p 94, John Wiley and Sons,Inc., 1908. SBernthsen, "Text-Book of Organic Chemistry," p 443, D . Van Xostrand Co , 1916

393

else was isolated. The remainder probably consisted of cresols and xylenols. Residues. The residues were resinous and were not examined. The coal was rich in resin, present as sizable inclusions as well as resin of constitution. BASES-The total bases, ammoniacal and pyridine, were determined by steam distillation of an alkaline sample of the condensate into hydrochloric acid. The ammonium chloride and the hydrochlorides of the pyridine bases amounted to 1.37 grams per liter of condensate. This is equivalent to about 0.479 gram of the free bases, of which 0.0824 gram was pyridine or pyridine homologs and the remainder (0.396 gram per liter) was ammonia. The bases of the aqueous condensate were therefore composed of about 83 per cent ammonia and 17 per cent pyridines and pyridine homologs. Tests by RiminilO and Simon1' showed the absence of primary and secondary aliphatic amines. Pyridine and its homologs were identified as a class by the characteristic odor and the heavy precipitate given with tannic acid." The hydrochlorides, after treatment with absolute alcohol to remove the pyridine hydrochlorides, analyzed for the proper amount of chlorine for ammonium chloride and the base was ammonia. PORTION I

MEDIUM OIL-TAR

Separation into Classes of Compounds

This oil had the usual appearance of a liquid tar and had a specific gravity a t 20" C. of 0.9337. At atmospheric pressure its distillation characteristics, using a 100-cc. sample, were as given in Table T'. of Tar I TEMPER-DIS-

Table V-Distillation DISTILLED

cc.

2 oil and 3 water

6 over

10 15

TEMPER- DISATURE 0

c.

97 150 180 196

TILLED

cc. 20 25 30 35 40

ATURE 0

c.

206 215 225 233 242

TILLED

cc.

TEMPERATURE

c.

45 50 55 58

251 260 269 275 (cracking began) Remainder, pitch

I n studying the composition of a tar, it is of obvious advantage first to separate into classes of compounds and then to cut up the individual class as best suits the methods to be used. To that end the medium oil, tar portion I, was first separated into the several groups of carboxylic acids, phenols, tar bases, and neutral oils. I n this, advantage has been taken of the experience and work of Brittain, Rowe, and Sinnatt.12 INSOLUBLE MATTER-Thirty-five hundred grams of tar 1 were diluted with approximately 3 liters of ether. The insoluble portion (a), which consisted of coal and coke dust and resinous material, was removed by filtering. It amounted to 1.0 per cent of the tar I. CARBOXYLIC AcIDs-The filtered solution was then extracted twice with a 10 per cent solution of sodium carbonate for the removal of carboxylic acids. The carbonate solution was washed with ether and the washings, which contained some phenols, oils, etc., were added to the main ether solution of tar. The carboxylic acids were liberated from the carbonate solution and extracted with fresh ether, dried, and recovered. They formed a viscous, dark, reddish brown liquid and amounted to only 0.25 per cent of this tar. Pmh'oLs-The ethereal solution of tar from which the carboxylic acid had been removed was similarly treated with 25 per cent excess of 10 per cent sodium hydroxide solution in several portions to remove phenols and any other alkalisoluble material. This solution was washed with ether Mulliken, 00.cii., Vol. 11, p. 35. I b r d . , p. 42. 12 Fuel, 4, 263 (1925). 10

11

I,VDUSTRIAL A N D ENGINEERING CHEMISTRY

394

and the phenolic material freed with very dilute hydrochloric acid (see Table VI), extracted, dried, and weighed. Some of the phenolic material set free was now insoluble in ether and was resinous. B ~ s ~ s - T h e residual ethereal solution was then washed with 2.5 per cent hydrochloric acid for the removal of the tar bases. I n this operation a small amount of solid precipitated. The aqueous solution was washed free from oil and the bases were liberated with sodium hydroxide. Upon being made alkaline, a small amount of solid (1.59 grams) separated from the solution. The bases were then extracted with ether, dried over stick potassium hydroxide, and the ether removed by distillation. NEUTRAL Oms-After being thoroughly washed with water for the removal of mineral acids and other reagents, the residual oil was dried and the ether removed. This oil was then considered to be composed of the neutral oils or hydrocarbons of the original light tar oil. The amount is shown in the analysis given below. Table VI-Analysis MATERIAL

of Tar Portion I (Medium Oil)

Per cent

Dust, solids, carbon, etc. Carboxylic acids (c) Basespyridine, etc. ( d ) Phenols (liquid) Phenols (resinous) (e) Neutral oils Water and working loss

1 0

(a) (b)

0.25 1.7 21.0 2.4 66.0 7.65 100.00

Examination of Classes of Compounds

The individual portions or classes of compounds of like function into which tar I had been separated were then examined in more or less detail. CARBOXYLIC AcrDs-The amount of carboxylic acids reported in various low-temperature tars varies from mere traces to as much as 12 per cent. Most primary tars, however, have been found to contain but small amounts. The small quantity obtained from tar I did not make attempts a t identihation of individual compounds advisable. BAsEs-The quantity and composition of the bases found in low-temperature tar vary considerably with the roals distilled and the method of distillation. The percentage of total bases as determined by several investigators ranges from 0.2 per cent, as found by Pictet13 in his vacuum tar, to 5 per cent in the tar from carbonization by the Piron process, as reported by Crawford14 at the International Conference on Bituminous Coal. Most of the earlier investigators found but comparatively small amounts of bases in their tars. Jones and Wheeler15 found less than 1 per cent of bases in their primary tar, as did Parr and Olin,16who found 0.9 per cent-all of which was in the fraction of tar distilling below 210" C. Parrish points out that in the figures both of Jones and Wheeler and of Parr and Olin the values reported did not necessarily represent the true base content of the tars. I n both cases t h e work was done on the distillates of the tars, which amounted to but 50 to 70 per cent of the original primary tar. Morgan and Soulel7 found 1.9 per cent bases in the fraction of their tar boiling below 326" C. Gluud and Breuer18 found volatile bases amounting to but 0.46 per cent in their tar. Two samples of tar were found by EdwardsIg to contain in the fraction boiling below 310" C., 1.3 and 1.8 per Pictet, Kaiser, and Labouchere, Comfit. rend., 166, 113 (1917). Piron and Crawford, Proc. Intern. Cong. Bituminous Coal, Carnegie Institute of Technology, p. 743 (1927). u J. Chem. SOC.(London), 106, 140,2562 (1914). 16 Eng. Expt. Station, Uniu. Illinois Bull. 60 (1912) and 1 9 (1915). 1) Chcm. Met. Eng., 26,923, 977 (1922). u Ccs. Abhondl. Kcnnlnis Kohic, 8, 227 (1918). 19 Edwards, J . SOC. Chcm. Ind., 43, 143T, 150T (1924). 11 14

VOl. 20, No. 4

cent of bases, respectively. Brittain, Rowe, and Sinnatt,I2 in working with two tars, about 80 per cent of which distilled above 260" c., found 2.45 per cent and 4.1 per cent of bases. A tar examined by Parrish and Rowe20 was found to contain bases to the amount of 2.7 per cent. The crude bases from 3500 grams of tar I amounted to 59.5 grams or 1.70 per cent, possessed a d:' of 0.9848 and an 2: of approximately 1.500. Fractional distillation a t reduced pressure showed the crude bases to contain a little ether and three groups boiling, respectively, a t from 80-95" C. a t 50 mm., a t 118-125" C. a t 30 to 25 mm., and at 130-150" C. at 20 mm. The distillable bases amounted to 61 per cent and the non-distillable residue to 23 per cent. The residue was black, gummy, and almost odorless. The remainder of the crude bases was accounted for in the little ether still present, a trace of oil, and the working losses. The distillate was refractionated at atmospheric pressure. It revealed three clear-cut portions and main fractions: one boiling from 148-167" C., which largely passed a t 155" to 167" C . ; a second, 190-211" C., and a third boiling between 235" and 250" C. The fractionation results and the properties of the fractions are given in Table VII. Table VII-Fractionation FRACTION BI B: B5 Bb Bs

of Bases i n Tar I

"':

"M",4"-;G",f

BOILINGRANGE AMOWNT d:o c. cc. < 148 1.5 1.4810 148-163 6.0 1.4990 0:9323 (mainly 155-163) 8.0 1.4999 . . . . 2.3 1.5332 5.0 1.5271 0.9546 235-248 1.5580 1.0110 239 ... 1.6247 1.093 7 . 0 grams . .,.

{ pEH4

(For quinoline Residue

...

.. .

.

106:7 112.0 136-140 156 128)

...

Fraction BI had the odor of pyridines and gave copious precipitates with tannic acid, methyl iodide, and chloroplatinic acid. This might have been either pyridine or methylpyridine or a trace of dimethylpyridine, but amounted t o only 3 per cent of the bases or 0.05 per cent of the tar I. Fractions Bz and Ba show the proper boiling range, the approximate density, and molecular weights (determined 106.7, 112; theory 107), and responded to class reactions for pyridines, and are therefore considered to be dimethyl pyridines. All the fractions gave negative tests for primary and secondary amines by Rimini's test and the Simons test, respectively. This means that the bases are essentially tertiary in reaction. The fractions give the class reactions of nitrogen in the ring-that is, pyridines and quinolines. An intensive study was precluded by the small amount of material available. However, fraction Bq is evidently largely made up of higher homologs of pyridine. Fraction Bs, in view of properties shorn, when compared with those of pyridine homologs and those of quinoline, might be considered as being a mixture of pyridine homologs and quinoline. The gummy residuals were not examined. PHENOLS-A high content of the homologs of phenol is usually considered a characteristic of a low-temperature tar. However, the quantity and the nature of the phenols, like other classes of compounds in low-temperature tar, apparently vary greatly with the composition of the coal carbonized and the method of carbonization. A tar examined by Morgan and Soule21 was found to contain 14.7 per cent of phenols, while one examined by FischerZ2 contained 45 per cent of this class of compounds. Brittain, Rowe, and Sinnatt,'Z in examining two low-temperature tars, found 26.6 and 30.6 per cent, respectively. 20

11 22

J . SOC.Chcm. I n d . , 46, 99T (1926). Chcm. Met. Eng., 16, 923 (1922). Byennsfof-Chem., 1, 31, 47 (1920).

INDUSTRIAL A N D ENGINEERING CHEMISTRY

April, 1928

The dried phenols, after being freed of ether, were distilled in vacuo. The fractions were carefully refractionated at atmospheric pressure through an 8-inch (20-cm.) bead column. The heat was supplied by a closely controlled elec'tric air bath. Both fractionations were carried out completely in an atmosphere of natural gas to eliminate oxidation and reduce the tendency to resinify (Table VIII). Table VIII-Fractionation of Phenols a t Atmospheric Pressure BOILING BOILING POINT POINT FRACTION (COR.) VOLUME FRACTION (COR.) VOLUME

A B C D

E

c.

cc.

-180 180-187 187-195 195-207 207-215

2 3 7 17.5 15.5

O

F

G H I

c.

cc.

215-228 228-232 232-245 245-265

17 5 10 10

The fractions were liquid up through H, which was as far as distillation a t atmospheric pressure could be carried without evidences of decomposition. Fraction I came from the distillation of fraction 7 in the first distillation. It set to a soft mass, as did fraction 8. From both there was isolated crystalline paraffin, as had been done from the crude phehols by filtration. The persistency of the retention of oils by the alkaline phenolate solutions, despite the thorough backwashing with ether during the separation of the tar into groups, is amply evidenced. This paraffin recrystallized from acetone and petroleum ether yielded wax melting a t 82-85' C. The specific gravity of all the fractions was almost exactly 1.0, except fractions B (d:' 0.9375) and C (d:' 0.9778) both of which contained a little oil. EXAMINATION OB PHENOL FRACTIONS-Fraction A ( < 180" C.) contained a very small amount of oil. The alkalisoluble material yielded tribromophenol which, purified, melted a t 91-92' C. Mixed with pure tribromophenol (m. p. 93" C.) it melted at 92-93' C. The amount of the phenol in the tar is small, being largely in the aqueous condensate previously described. Fraction B. The boiling range (180-187' C.) indicates that it is probably largely phenol (b. p. 184' C.) with possible traces of ortho-cresol (b. p. 191' C.). It yielded large quantities of tribromophenol (m. p. 91-92' C.). The hydroxyacetic acid derivative was prepared and proved to be largely ortho-cresol, containing a trace of the para isomer. They showed the following melting points: ortho, m. p. 150-151" C.; mixed with pure and known material, m. p. 151-152' C.; para, m. p. 133.5'C., and mixedwith pure para, 134-135' C. Fraction C had a boiling range of 187-195' C., and should contain largely ortho-cresol (b. p. 191' C.) together with some small amounts of para and meta (b. p. 202' and 203" C., respectively). It proved to be largely ortho-cresol, but contained a trace of the para, and possibly the meta, isomers of cresol. The evidence for the identifications lies in the melting points (Table IX). Table IX-Melting Points of Hydroxyacetlc Derivatives of Cresol HYDROXYACETIC MATERIAL PURE ISOLATEDDERIVATIVE MIXED DERIVATIVE OF :

c.

Ortho-cresol Para-cresol

149.5 134.5

O

c.

152.5-3 135-6

e

c.

150-151.5 135.6

Fraction D. The study of the hydroxyacetic acid derivatives of this fraction (195207' C.) showed the metaand para-cresols to be present in considerable quantities. Their derivatives melted a t 101-101.5' and 134-135.5' C., respectively, and their identities were established by the method of mixed melting points. Traces of the ortho derivative and of the 1,2,4-xylenolwere present. Fraction E . The examination of the hydroxyacetic derivatives obtained from fraction E (207-215' C.) revealed

395

the presence of three compounds, in appreciable amounts. These were principally meta-cresol and the 1,2-dimethyl4-hydroxybenzene with a small amount of the l,&dimethyl5-hydroxybenzene. The isolations were effected through the hydroxyacetic derivatives. These were first obtained as a mixture of crystals and oil. The crystals, after repeated recrystallization, melted a t 158-162' C. and mixed with the pure derivative from the 1,2,4-xylenol, melted without change. The oil was crystallized from warm water. The material so obtained showed itself to be a mixture (B). It was first washed with a small quantity of warm petroleum ether and then extracted with another small quantity. By reducing the extract to half volume a small crop of crystals melting at 84-6' C. was obtained. Their melting point was not altered by recrystallization. This relative solubility is the expected one and the melting point corresponds to that of the 1,3-dimethyl-5-phenoxyaceticacid derived from the 1,3,5-xylenol found by S ~ h u t z . ~ ~ The mixture (B), by means of an extended series of fractional crystallizations from petroleum ether, was resolved into two main portions. From one of these was obtained the pure derivative from the 1,2,4-xylenol (1,2-dimethyl-P hydroxybenzene) and the other was principally meta-cresol. I n the mother liquors from the crystallizations of the second portion, following an extended series of crystallizations and separations and finally subliming, a further small quantity of material melting a t 85-88' C. was obtained. It is considered to be the derivative from 1,3,5-xylenol (1,3-dimethyl5-hydroxybenzene) . The evidence of the identity of the materials isolated is shown from their melting points as given in Table X. Table X-Melting COMPOUND

Points of Phenoxyacetic A d d s ISOLATED PURE 0

1,2-DimethyI-4-phenoxyacetic acid 1,3-Dimethyl-5-phenoxyaceticacid I-Methyl-3-phenoxyacetic acid

c.

158-161 85-86 97-99

O

c.

MIXED O

c.

161

158-161

101

97l99

85-86**

At all times during the fractionation crops of crystals which melted near 140' and 115' C. were tested by taking mixed melting points with the two derivatives melting at 141' C.-namely, those from the 1,3,4-xylenol (1,3-dimethyl-4-hydroxybenzene) and the 1,3,2-xylenol-and with the 1,4,2-xylenol derivative melting a t 116' C. As all three of these xylenols form their derivatives with chloracetic acid easily under the conditions employed, they were probably not present in this fraction E to any appreciable extent. Fraction F (215-228' C.) was even more difficult to resolve than was E. However, the treatment described above permitted the isolation of the same three compounds as were found in fraction E. As was the case in fraction E, none of the other isomeric xylenols derivatives could be found. Fraction G. Attempts a t isolation by crystallizing out catechol from fraction G (228-232' C.) by cooling and seeding as well as by subliming were unsuccessful. From the oily mixture of hydroxyacetic derivatives none of the catechol derivative could be isolated. Instead, the principal derivative present was that of the 1,2,4-xylenol, melting a t 160-161' C. (mixed melting point 161-162" C.). As obtained, it was contaminated with lower melting derivatives. Their small quantity precluded their identification. Fraction H. The hydroxyacetic derivative obtained from this fraction (232-245' C.) was found to consist almost entirely of that from 1,2-dimethyl-4-hydroxybenzene. Very little of the lower melting derivative was found in this fraction. (From this fraction the yield of the derivatives was low, part of the phenols being resinified. This tendency '8

Ten, Si, 211 (1923).

INDUSTRIAL A N D ENGI NEERING CHEMISTRY

396

was greater in the case of the remaining fractions.) Although no catechol was isolated, it must have been present in traces by reason of its presence in quantities in the aqueous condensate.

OIL PERCENT

Figure 1-Atmospheric Pressure Distillation of Three Petroleum Crudes and Neutral Oils from Tar Portion I from Low-Temperature Distillation of Utah Coal

Fraction I (245-265" C.). The phenols were not identified. They yielded no hydroxyacetic derivative, but were partly converted into tarry material under the influence of the strong alkali alone and very largely so under the conditions employed to bring about the reaction. The paraffin occurring in this fraction has already been discussed. Treated with alkali and washed with ether, the fraction from the evaporated ether yielded a heavy oil saturated to bromine, boiling a t 260" C. and up, and which congealed a t 15" C. Resorcinol is believed not to be present in appreciable quantities in the tar examined. It forms a dihydroxyacetic derivative easily. Residue. The residues were almost entirely soluble in alkali and could be released again with acid. They, like the phenols of fraction I, are undoubtedly complex phenolic bodies. When released by acid from an alkaline aqueous solution, they precipitate as flocculent material characteristic of many resinous acids. NEUTRALOIL-The neutral oil portion amounted to 66 per cent of the tar I. It consisted of oil, paraffin wax, and a little resin. That portion of the oil boiling up to 140' C. amounted to 8 per cent and its d:' was 0.7936. Its other properties were not studied owing to accidental loss. In this first examination only a preliminary study has been completed of the remaining oil. This oil was a dark-brown mobile liquid of di6 0.9061. Subjected to steam distillation under t h e conditions indicated it was cut into 5 parts as shown.

Vol. 20, No. 4

into a saturated picric acid solution, gave no picrates. h few drops of each shaken vigorously with several cubic centimeters of the picric acid solution likewise failed to show the presence of naphthalene. Careful tests for the presence of benzene (C&) made on the light oil as well as those for indene and styrene,26all easily and definitely detectable compounds made on the proper boiling fractions, were all negative. Fraction D. This was a semi-solid mixture a t 25" C. and consisted of 84 per cent of oil and 16 per cent of white paraffin wax melting a t 57" C. The oil was liquid a t 25" C. and semi-solid a t 22" C. and possessed a Saybolt viscosity of 225 seconds a t 38" C. (100" F.). There was no naphthalene in this fraction. Residue. The residue was one-sixth heavy oil, one-third resin, and one-half paraffin wax. After recrystallization the wax melted about 62" C. and was still colored. Analysis. The neutral oil therefore consisted of light oil boiling up to 140" C., 8 per cent; oil boiling 140-275' C., 65 per cent; heavy oil, 12 per cent; paraffin wax, 10 per cent; and resin, 5 per cent.

751

,

1

0.70 0.71 0.72 0.73 0.74 0 75 0.76 0.77 0.78 0.7Y 0.800.81 0.82 0.83 08:=85 DENSITY AT 25 "C.

0.86 0.87 0.88 0 . 8 9 r 9 0

Figure 2-Specific Gravity-Temperature Curves of Three Petrol e u m Crudes and Neutral Oil from Tar Portion I from LowTemperature Distillation of Utah Coal

Summary of Results

That portion of the total tar soluble in the aqueous condensate included: carboxylic acids, 0.39; phenols, 2.25; bases, pyridines, 0.082; ammonia, 0.397; neutral oil, 0.1 grams per liter. The phenols consisted of carbolic acid about 33 per cent; cresols, 18 (mainly para and meta but with some ortho present); catechol, 4; higher phenols, 4; tar-acid resins, 37 per cent. The liquid tar I was about 29 per cent of the total tar, and it was separated into (a) insoluble solids 1.0; (21) carboxylic acids, 0.25; (c) crude bases, 1.7; (d) phenols (1) soluTable XI-Distillation of Neutral Oil-Tar I ble in ether, 21.0, (2) insoluble resin 2.4; (e) neutral oils, TEMPERSP. GR. 66 per cent; (f) water plus working loss, 7.65 per cent. FRACTION ATURE VOLUMEWEICRT AT 25' C. The bases were 60 per cent distillable, and these contained C. Cc. Grams little if any pyridine but were rich in dimethylpyridine and higher homologs. Of the crude phenols 30 per cent were solid and non-distillable substances, 50 per cent were distillable, and the reFractions A, B , and C. These three portions were frac- mainder was made up of contaminations-ether, water, tionated in turn, 10-cc. portions being taken up to 275" C. oils, wax, and fractionation and other working losses. The From the first to the last fraction the color varied from distillable phenols were divided by fractionation as follows: colorless through yellow, greenish with blue-green fluores- phenols, 6; ortho-cresol, 7; para- and meta-cresols, 19.5; cence, to green and then brown; the nzi progressed from xylenol fractions, 36.5; and higher homologs, 31.0 parts. 1.4490 to 1.4955 and d:' from 0.7941 to 0.9023. Figures 1 There were identified phenol, ortho-, meta-, and para-cresols, and 2 show these data compared with those for three typical 1,2,4-xylenol, 1,3,5-xylenol, and catechol isolated above. Of the neutral oil, 8 per cent distilled up to 140" C. and petroleum crudes taken from the l i t e r a t ~ r e . The ~ ~ ~higher ~~ specific gravity of the tar oil is considered as indicative of a 65 per cent between 140" and 275' C. About half of the high naphthenic content. Fractions 12, 13, and 14, boiling remainder distilled with steam. This distillate yielded wax equal to 2 per cent of the neutral oil and 10 per cent of a at 207-14", 214-22", and 2'22-28" C., respectively, were each tested for naphthalene. An air current, passed over a heavy oil-wax mixture with a viscosity of 225 Saybolt seconds sample of each maintained a t 100" C. and passed directly (at 100°F.). The residue contained wax, oil, and resin, equal \ I to about 8, 2, and 5 per cent, respectively, of the neutral oil.

?IT

O

14

Dean, Hill, Smith, and Jacobs, Bur. Mines, Bull. 207 (1922). Kraemer and Calkin, Ibid , Tech. Pager 346 (1925).

21

Brown and Howard, I n d Eng. Chem., 16, 1147 (1933).