Jaiiuary, 1930
lNDUSTRIAL A N D ENGINEERING CHEMISTRY
which require a definite quantity of negative ions for neutralization in excess of that furnished by the clays and other inaterials found in normal ivaters. These charges are therefljrc tieUtra1iZed by thc sulfates present in the precipitating c::ectrolytc and produce a solution link. Above neutrality t1ici.c is sufficient alkalinity to produce complete coagulation. 111 tlic study n i t h iron and lime ( 4 ) it was shown that invompicte oxidation will accoulit for t,he formation of thc ;olutioli link. With chlorinated copperas solutions below iicutrality, incomplete ncutra1izatio:i of the positive charges on t,lx Fc ion accounts for the sulfate adsorption.
81
Literature Cited Baylis,
J.
A m . il'ater
Assocll,,
303 (,323),
(2) Enslow, itfunic. H e w s Ii'o/rr ri.s7i;,, 76,227 (1020). (3) Hedgepeth and Olsen. J. A n i . 1C'c:rr Il'orks .-lssorn., 20. 467 (1928).
it;
~ ~ ~ N ~ ~ ~ ~f s ~r ~ G ~ l 'IOB ~~ , ~(1924). ~ ) ; .2 , ; ~ ~ ; n. !6) Krause, Roczniki Ciicm,, 6, (7) Krause, z. anorg. al/ge,n. C h e m , 174, 146 (1928). ( 8 ) hliller, U.S. Pub. Health Service, Pub.Heall/: Regts. 40, 1413 (1925)
(::i" , : ~ a h ~ ~ : n , ~ ~ " , ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ e ; ~ : " . . 4 b 9 0 ~ : ~ ~ (11) lveiscr, r , f l y d r o u s
p,
(1926).
(12) WVo!nian antl Ifannari, Chem. M e t . Eng., 24, 728 (1921).
Cracking of High-Boiling Coal-Tar Acids' C. E. Senseman
a
High-boiling coal-tar acids were fractionated to o v e n . T h e g a s e s are led remove all material boiling under 207' C. The pora w a y a n d condensed, and scarcli here rcportccl tion of the acids boiling above this temperature was tlie tar acids are removed by w a s t o c o n v e r t the run through a cracking apparatus to determine whether tlie usual processes. These 11 i gli -boi 1 i n g phcnols, frcor not any would be converted into the more useful a c i d s a r c somewhat more cliiciitly rcferrcd t o as conllower boiling phenols. Following each run the conconiples than the domestic tar acids, d i i c h occur abundensate was fractionated until all material boiling o n e s , b u t they were more d a n t l y i n t a r s rcceivcd below 207" C. was removed. Water, benzene, xylenes, readily availnble at thc time tlirougli the low-tempcrature cresols, and much smaller quantities of phenol and thc investigation was started. coking of coals, to tl:c more coumarone derivatives were identified as products KOanalysis was inadc in this useful Ion--boiling pl3cnols. formed in the process. laboratory to dctcrminc the For t h c s c m o r e y o l a t i l e Studies were made of the effect of temperature, plicnols t l m e is already a constitucnts, but Warnes (3) charging rate, and pressure. Tables are given to show great dcnmnd in the innking aiid othcrs report tlic presthe influence of the first two variables. Pressure was of mir.s, disinfectants, and cncc of a small amount of not found to be a significant factor in producing the tiicrcsyl pliosphatc. L a r g e p h c no1 , crcsol s , xylenols, lower boiling acids. quantitics of t l x high-boiling pseuclocumcnol, n n p h t h o l s , Tests were made o n the cresols formed to determine acids are annually imported and prohab!y phcnol cthcrs. their usefulness in the manufacture of resins. The ~)!ivsicalDromrtics of the by the Gnited States. It lias long been k n o w n mntiriils uset~ ncrc as folthat tile quality, as well as the quantity, of any conl tar lows: s2ccific grayity a t 26' C., 1.021; phcnols boiling rincler 207" C., 32 per cciit; boiling range of rcsiduc, 207-290" C. J L tiependent upon a large number of factors, such as tlx As the principal objective in this rwcarch was to protlucc source of tlie coal, the temperature of coking, the rate of licating, tlie compactness of the charge, the shape and size from tlie higher acids those plicnols which could bcst be used in the manufacture of resin.. . any phenol or cresols present in of the retort, aiid the method of removing the volatile matter. Of tlie 200 or more compounds (3) present in the crude had first to bc rcmoved. 7'llis was done by fractar Irom 1;igli-temperature coking processes it is generally tionating in a Henipel column pac~licdwith glass beads antl n g m t l that ninny arc formed through the decomposition, well insulated from draft uiitil 3 teinpcrature of 207" C. was crr ci,ncl;ing, of compounds formed during low-temperature reached. The portion boiling above this temperature was h i t n.hich subsequently undergo this change before used for cracliing purposes. S o practical method for the t1.c rcturt during the high-temperature coking proc- separation of these various constituents is known. Hence tlie €59. In fact, Parr ( 2 ) quotes Gentry as defining tcmpera- group separation through fractional distillation. turc c8::rbonization as "the destructive distillation of coal a t Apparatus or I;clon. the clacking temperature of the hydrocarbons in primary tar." Thc apparatus used (Figure 1) is basically the same as that This linowledge, together with tlic results on thc cracking of vxi.ims hydrocarbons, both aliphatic and aromatic, ob- used by Dean and Jacobs ( I ) . Some modifications of the tained tluring tlie last two decades, led to a belief that tlie original set-up m r e made, and others m r e introduced as the clacking of the high-boiling acids from lowtemperature coking investigation progressed. Briefly, the apparatus consists of a iiiiglit be effected so as to yield more useful materials, par- reservoir for holding tlie tar acids, which is equipped with a pressure equalizing tube, a needle valve, and a t the base, a ticularly for the making of resins. sight feed, permitting a regulated flow of material through the drip tube to the center of the top of the cracking chamber. Material Used The cracking chamber consists of an electrically-heated 1.25T1:c work here reported was done on coal-tar acids pro- inch extra heavy steel pipe 30 inches long. The lower end of iluccd i n Scotland from blast furnace gases. The splint coal the clianiber has a clean-out plug, iirinicdiately above which o f Scotland servcs very well for blast furnace purposes, tlie a condenser tiii~rl01 0.75-inrti extra heavy steel pipe takes off. upper part of the furnace serving as a low-temperature coke Tlie loner end of tlle condenser is supplied with a drain cock, which liad to be kept closed when a pressure greater than atICcceived September 1.1, 1920.
HE purpose of the re-
I
.
82
INDUSTRIAL A N D ENGINEERING CHEMISTRY
Vol. 22, No. 1
means of a thermocouple and millivolt meter, the thermocouple well being placed midway of the chamber. While it cannot be claimed that such location of the thermocouple gives a true record of the temperature throughout the apparatus, it was felt that it was the best method available without using additional couples and that all runs would be made under the same relative conditions. As originally built, a steel chain, so long that when curled about i t would fill the space from the thermocouple well to the top of the heating coil, was suspended from the top of the cracking chamber. This soon became so thoroughly coated with a hard carbon deposit that it was necessary to drill out both chain and carbon. Thereafter four steel plates (Figure 1) were placed,in the tube to run its entire length. Whenever the plates and the walls of the tube became coated with carbon to the point of choking the flow of gas a stream of air was blown through while the apparatus was hot and continued until all but a thin layer on the various surfaces had been burned out. Experimental Procedure
\IN UI
Figure l--Cracking Apparatus
mospheric was used. The cracking chamber is centered in a 3.50-inch extra heavy steel pipe made gas-tight a t the bottom by being clamped to a flange shrunk on the inner tube and a t the top by means of a packing box, which takes care of the unequal expansion of the two tubes. When pressure was desired nitrogen was used, the tank being connected to the cracking and equalizing chambers. Both chambers are equipped with pressure gages. Heat is s u p plied to the reaction chamber by means of two coils made of No. 17 nichrome wire, wrapped on mica, and covered with a layer of alundum cement and magnesia pipe covering. Either or both of these coils may be used, depending upon temperature requirements. The four ends of the resistance wire are led out through the walls of the equalizing chamber and insulated therefrom by means of cones of soapstone, tightened down by packing nuts. The temperature of the cracking chamber is measured by
As already stated, the tar acids were first fractionated to a temperature of 207" C. for the removal of water and cresols, little or no phenol being present in any of the acids received. A portion of the acids boiling above 207 O C. was placed in the receiving reservoir a t the top of the apparatus. Meanwhile the reaction chamber was being brought to the temperature at which it was desired to make the run. This temperature being reached, the needle valve was opened far enough to give the desired flow, as determined by counting the drops per minute through the sight feed. When pressure was used nitrogen was first let into both chambers until the gages registered the desired pressures. In these experiments the cock a t the bottom of the condenser had to be kept closed and the condensate allowed to collect in the lower part of the condenser. As would be expected, very close watch on the apparatus was necessary, frequent adjustments being called for in the pressure lines as well as in the rheostats which control the current supplied for the heab ing Also the needle valve required occasional readjustments because of the gradual heating of that part of the apparatus through conduction of heat from the reaction chamber. I n the runs in which atmospheric pressure was not exceeded the drain cock was left open during the course of the experiment. At the completion of each run, when all material had drained from the apparatus, the liquid products were fractionated in the same manner as the original crude, everything boiling under 207" C. being removed. The figure representing that portion boiling between 190" and 207" C., which was found to be completely soluble in alkali, was used in calculating the percentage yield of acids. This range does not cover the boiling temperature of phenol. The various distillations of the cracked acids never showed more than traces of this compound, of which no quantitative determination was attempted. To obtain data on the further cracking of material put through two or more times, the portion boiling above 207" C. was distilled from an ordinary distilling flask in order to separate i t from the tarry matter. Temperatures a t which the runs were made covered the range from 525" to 850" C. Pressures used ran as high as 400 pounds per square inch (28 kg. per sq. cm.). Experimental Results
Early in the investigation it became evident that the carbon deposited on the exposed surfaces played a very important part in increasing tht: yields of the more volatile products and that, as such deposit was unavoidable, the other factors would
INDUSTRIAL AND ENGINEERING CHEMISTRY
January, 1930
always have to be studied with this condition present. During the course of the work a number of heating coils burned out. I n replacing them most of the carbon deposit was dislodged. Upon continuance of the work after repairs had been made, enough material to insure a heavy deposit was fed, thus obtaining comparable results. The data collected through the earlier runs indicated that such pressures as could be safely used with this apparatus, considering the temperatures involved, would not be a material factor in increasing the yields of the low-boiling phenols. The earlier results also pointed to an optimum temperature of 650" C. or higher. Below this the amount of cracked material boiling under 207" C. was small. If this temperature was greatly exceeded the condensable material was cut considerably in yield, although the percentage boiling under 207" C. was relatively high. As temperatures were increased there was an attendant increase in the volume of gas produced, until at 750" C. or higher great clouds of dense brown fumes were given off. The portion of the cracked material boiling under 207" C. consisted chiefly of cresols, with water, benzene, and smaller quant,ities of xylene, coumarone derivatives, and phenol. Owing to several factors, particularly the size of the apparatus, the amount of carbon deposit, and the variation in composition of the crude acids received, it was very difficult to obtain check runs and data which could be considered a t all quantitative. A great number of runs were made to determine the effect of temperature, pressures, and charging rate. The results of several runs showing the influence of two of these variables upon the yields of the desired materia.ls are given in Tables I and 11. The charging rates reported in these tables were, whenever possible, determined by counting the drops per minute through the sight feed. The result reported in each case was determined by dividing the number of cubic centimeters used by the minutes required for the run. of Tem erature on Yield of Acids Boiling below 207' (Atmospheric Pressure) YIELDS OF ACIDSBOILING CHARGING 190-207O C 5 TEMPERATURE RATE RECOVERY Used Recovered c. Cc./min. Per cent Pn cent Per cent b b b b 525 and 550 91 14.6 16.1 600 3.33 30.0 83 25.3 3.33 650 31.8 25.0 78.5 700 3.33 21.0 33.8 62 750 3.33 35.0 40 14.0 800 3.33 a Based on material. b Yield very low, regardless of pressure and charging rate.
Table I-Effect
d:
As previously stated, there was much evidence to prove that the carbon present in the cracking chamber functioned in a way which aided the cracking of further material passing over it. Either the reaction itself is a contact-surface phenomenon, in which case the carbon affords vastly more surface for aiding the reaction, or else i t serves as a catalyst for increasing the yields of cracked material. Runs were made a t 650" to 700" C., which gave yields of 40 per cent and more of cracked acids. At such times, however, the reaction chamber was almost filled with a carbon deposit. After cleaning this out thoroughly there was always a marked decrease in yields. of Charging on the Yield of Acids Boiling under 207O C. (Atmospheric Pressure) YIELDS OF ACIDSBOILINQ CHARGING 190-207' C." TEMPERATURERATS RECOVERY Used Recovered c. Cc./min. Pn cent Per cent Pn cent 91 14.6 16.1 600 3.33 17.0 17.5 600 6.66 97 25.3 30.0 650 3.33 83 20.5 23.7 650 6.66 86.5 88 16.5 18.7 650 9 25.0 31.8 78.5 700 3.83 26 31.2 83 700 8.66 89 22 27.1 700 12.50 " Based on material.
Table 11-Effect
The results in Table I1 were obtained at various charging rates and at temperatures which had previously given the
83
highest yields a t approximately 3.33 cc. per minute. These results agree with reports of other investigators that as the charging rate increases the percentage of recovery increases, although the percentage of material actually attacked may decrease because of insufficient exposure to either temperature or contact, or to both. A series of experiments was carried out with pressures of 125 to 400 pounds per square inch, and at temperatures ranging from 525" to 700" C. The charging rate was approximately 3.33 cc. per minute. Under none of these conditions was i t indicated that elevated pressures favorably affected the production of the lower boiling phenols. By repeating the cracking process several times under the most favorable conditions (650" C. and a charging rate of 3.33 cc. per minute), using each time the higher boiling fraction obtained in the previous cracking, a total yield of acids boiling under 207" C. was equivalent to 36 per cent of the original material. Condensation Tests
The condensation tests on the cracked acids with formaldehyde were carried out by the Material and Process Engineering Department of the Westinghouse Electric and Manufacturing Company of Pittsburgh. Two samples were submitted to t h e m - o n e which had been prepared by distillation methods only, containing first, second, and third crackings; and the other containing only first crackings from which traces of coumarone derivatives had been carefully removed. The reports received showed that the two samples did not react differently toward condensing agents, and that a resin can be made from these acids, although some polymerizing agent, such as hexamethylenetetramine, is required to give a product sufficiently reactive for molding. The use of such an agent requires added operations which are undesirable from the standpoint of time and expense. Many of the commercial grades of U. S. P. m-p-cresols likewise will not produce highly reactive resins without the use of condensing or polymerizing agents. It is thought that possibly with less refinement, permitting the presence of one or more of the higher boiling phenols, the acids would give a more quickly reactive material. Conclusions
Investigation of the cracking of high-boiling coal-tar acids shows that phenols of lower boiling points can be produced to the extent of 25 per cent in a single cracking and 36 per cent by recycling. The data collected indicate that this yield of 25 per cent may be obtained by operating at 650" C. with a charging rate of 3.33 cc. per minute, or a t 700" C. with a charging rate of 6.66 cc. per minute. The results further indicate that pressures above atmospheric do not favorably affect the production of lower boiling phenols. Condensation tests on the cracked material for the production of resins show the need of some polymerizing agent, such as hexamethylenetetramine. Acknowledgment
The author wishes to acknowledge the support given to this investigation by H. D. Weihe of the Color and Farm Waste Division, the Material and Process Engineering D e partment of the Westinghouse Electric and Manufacturing Company, and R. M. Baker of the Chemical Engineering Division of the Bureau of Chemistry and Soils, for his drawing of the apparatus. Literature Cited (1) Dean and Jacobs, U. S. Bur. Mines, Tech. P a p n 468. (2) Parr, IND. ENG.CHEM.,41, 164 (1929). (3) Warnes, "Coal Tar Distillation and Working Up of Tar Products." Benn, 1917.
