I X D USTRIAL AND ELVGI.VEERING CIIEMISTRY
832
Vol. 16, s o . 8
Slow Gelation in Pressure-Still Tar',2 By S.E. Sheppard and L. W.Eberlin E A S ~ M AKODAS N Co., R O C W L S ~N. P , Y.
4 TIZE course of derelopment of colloidal fuels3 a considerable quantity of a petroleum residue, so-called "pressureitill" tar or oil, from Burton stills, was obtaincd and used. At the time of this work (1917-18) this residue, although of relat.ively low viscosity, was ill a.dapted for marine fuel oil: or in fact any purpose where it had to stand, owing to the behavior of its asphalt and free-carbon content. Kat only did this tend to flock out at relatively short notice, and under different conditions of contact with other substancc, but prolonged standing the whole mass tended to conged or consolidate, owing to formation of a carbonaceous gel in many respects reseNlbhg coke. At the timc, two methods mere investigated, primarily to prevent the flocculation and settling, and secondly to arrest or delay the slow gelation. Tlie pressure-still tar contained both asphalt and frec carbon. Analyses according to the procedure of \"/'eiss4 gave, for tlie sample in question, about 4.6 per cent asphalt (beiizme-soluble) and about 1.0 per cent, free carbon. It was not possible to remove the latter by filtration, and only part,ly so by running through a highspeed Sharples centrifuge, at 18,000 r. p. m. The asphaltic material no doubt acted as a protective colloid, keeping the carbon dispersed. On adding 5 per cent naphthalene the viscosity was considerably reduced, and centrifuging at 18,000 r. p. 111. removed about half the carbon.
I
Pel- cent Pressure still oil with 5
Asphalt [Cerlxm Asphalt arbon
cenfiiii'King
I
per ceiii naphthalene
4.7
1 . I15 4.5
o,c2
With 10 per cent naphthaleiie, up to two-thirds of the carbon could bc removed in this nianner; tire viscosity was still further reduced, and although the flash point was lowered from 143" C. ('LSOO F.) (open cup) to 127" C. ('260'P.j (open cup), the material was greatly improved as a fuel oil. The naphthalene t.reatmerit. soon reached a point, however, at which it. becnme unecorromical. T*a,,.!
I Alter 3 Months
Amount of Pirateui Fer cent 0.0
0.1
0.2 0 .5 1.0 Sfd0
viscosity
Engler at
20" C.
rt start
7.3 7.3 7.6
Top layer
Asphaltum Per cent
.60
;i
5.80
5.30
8.S 10.0
5.35
7.3
5.02
5.44
Crrbon 1'er cent 0.30 0.53 0.li.l 0.7X 1.22 1.21
The other method used was to treat the oil with certain amounts of the "6xateur" developed for suspending pulverized coal in fuel oil. The protective colloids developed for this purpose were lime and otlier soaps, chief of which was a lime-rosin greasc, prepared under conditions giving the maximum protective action.a Different amounts of this, up tu 1 per cent of lime resinate, were added to the pressure-still tar, and uniformly mixed by heating. The treated samples were placed in large sedimentation vessels, of about 2 gallons capacity, and the upper layers analyzed after 3 months. In 1 Presented under the titlc "Note OD the Formation oi a Carbon Gel io Tar;," before the Divisbo of Physical and Inor%aoicChemirtry at the 65th Meeting oi the American Chemical Society, New Haven, Conn., April 2 to 7. 1923. Received April 11, 1924. Communication N O . 204 froin the Research Laboratory of the Eastman Kodak Company. a Sheppard. THIS Joosw~r, IS, 37 (1921). Ibid., S, 19.5 (1913).
'
this time the bottom layer of the untreated sample had already begun to thicken considerably, and showed a noticeable increase in free carbon. This was confirmed by analysis of the top layer. The bottom layer of untreated tar v a s found to contain 4.0 per cent asphaltum and 2.3 per cent carbon. It was evident from these experiments that up to 1per cent of iixateur-and this not of the highest grade-was sufficient to hold off flocking and sludge formation for a period of 3 to 6 months. Considerable quantities of pressure-still tar were used under these conditions in compounding colloidal fuels in largescale trials at Brooklyn in 1918, with quite satisfactory results. As regards the slow gelation, samples t,hat had been standing since April, 1918, were examined in April, 1921, aft,er a period of 3 years. All the samples showed congelation, t,his being most pronounced, however, in the straight tar containing no fixatcur, and least in the ones with 0.5 and I per cent fixateur. The gel structure was easily broken up hy stirring with separation of the fluid; there had formed a continuous but easily friable sponge tliroughout the mass, but, beginning at the walls of tho container, as was evident in the samples where its format.ion was retarded by fixateur. The lowered dispersity of the precipitate was evident to the touch, the siibstnnce being quite gritty. One of the glass sample jars was cut through endways, and the appearance of the coke deposit on the wall is shoim in Fig. 1. It seems possible that the free carbon resulting from the cracking process in the pressure still retains the property of coalcsciiig to the solid state, probably in graphitic form. The faculty of carbon atoms of building up chains and networks, postulated by T2aiigmuir i n regard to adsorption to charcoal, seems well exemplified in this slow coking process. Iirc. 1 A similar type or gel is apparent,ly formed in tlie alcoholic so1 of cadmiurn, described by Tbe Svedberg,S the gel being immediately desdroyed on shaking. The nature of the coking process, because of its importance for the blast furitace, is a matter of great technical significance. In a recent paper by Bone and collaborators6 on the coking propensities of coal, it is concluded, from fractionation with solvents, that while the "binding principle" of coking coals lies chiefly in the chloroform-soluble portion of their pyridine extracts, it is not necessarily resinic in origin. "The coking propcnsitirs of coals are principally due to the presence, or formation in them bg heat' of such nonresinous (humic) substances of cellulosic origin, whose fusion temperatures are below those at which tliey undergo rapid decomposition." The formation of a pyrogel, resembling petroleum coke, by slow after-action, from the pyrosol (tar) may be capable of throwing light oil tlie related problems of coke formation from coal. * Faradw Society Symposium on Colloids, 1911. Bone. Pehrwan, Sinkinson. and Stockingr, P r o r Roy. So
