A New Type of Semimicro Fractionating Column SHERMAN D. LESESNE AND H. L. LOCHTE University of Texas, Austin, Texas
A
LARGE number of fractionating devices for use on 1 to
10 cc. of liquid have been described in the last decade. Among the simpler and more practicable may be mentioned those of Cooper and Fasce ( I ) , Weston (4),and Craig ( 2 ) . I n a n attempt to devise a column showing high efficiency, large throughput per hour, and low tendency to slugging, a number of columns using rotating members as packing were tested. The form finally adopted has a metal band rotating a t about 1000 r. p. m. in place of the usual packing or indentations. The complete apparatus for use at atmospheric pressure on solutions of 1 to 10 cc. is shown in Figure 1. The length of the condenser was 10 cm. (4 inches), that of the column proper 37.5 cm. (15 inches). The column was a length of Pyrex tubing of 6-mm. inner diameter, to which the heating jacket was sealed. Nichrome wire wrapped around the jacket was used to provide approximately adiabatic conditions in the column. The boiler was placed on an asbestos board provided with a hole slightly smaller than the bulb and heated with a microburner. Asbestos cord was used to provide partial insulation of the portion of the boiler above the asbestos board. The assembled apparatus was tested with a 2.52 mole per cent solution of carbon tetrachloride in benzene and with a mixture of 1 cc. of methanol and 1 cc. of water. Table I shows the results obtained with the band rotating a t 1000 r. p. m. Using the refractive index-mole per cent data and curves of Zawidzki (6) and the -Electric methods of calculaStirr e r tion outlined by Walker, Lewis, and M c A d a m s ( S ) , bet w e e n 15 a n d 1 6 theoretical plates for -Wafer the 37.5-cm. (&inch) Tondenser column and still head Nlihfoq Wire combined mere ohtained. Since t h e empty column with same still head Wafer showed only 4 plates, the still head evidently could not have added more than 2 to 3 plates and the spinning column must have shown at least 13 plates for its 37.5 cm. (15 inches) of I n ? ' length or almost one 4mm. L plate per inch under Nkhrome Hea f e r %c Ke f total reflux. An interesting and somewhat unexpected fact developed was that the column did not reach equilibrium with this mixture in less than 1 hour of total reflux. Equally unexpected was the fact that its tendency to slugging is so low that the 6-mm. inner FIGURE1. APP.4RATUS
Ribbons
diameter column could be operated with a reflux above 2 drops per second, although, as the results show, the efficiency !vas lowered somewhat by such a high reflux rate.
l.333C X W V
E
1.3292 91
.-> 4-U
D * L
d 1.326C
Volume FIGURE2. DISTILL.4TIOs COHOL IT
of D i s t i l l a t e OF
10 CC.
hfETHYL AI.-
OF
10 CC. OF WATER
Figure 2 shows that the relatively simple methanol-water mixture can be cut in 80 minutes to yield only 1 cut of 0.06 cc. that is not practically pure methanol or water. TABLE I. TESTS FOR EFFICIENCY (lOOyoreflux for 80 minutes)
Reflux Diops/sec.
1 2 1
Rotator R. p. m . 1000 1000 0
Mole Per Cent CCL In In residue distillate 1.96
2.01
2 04
26.6 14.2 4 82
No. of The+ retical Plates
H. E. T. P. Inches
1.5
1 0
4
3 8
9.5
1.5
Since the power required to rotate the band is very low, it should be possible to use a very small steel shaft and packing box and operate under vacuum, provided the slow air leak at the stuffing box is not injurious. Such a column is now in use in fractionation of esters of petroleum acids, but no tests of efficiency have been run.
Literature Cited (1) Cooper and Fasce, IND.EXG.CHEM.,20, 420 (1928). (2) Craig, Ibid., Anal. Ed., 9, 441 (1937). (3) Walker, Lewis, and McAdams, "Principles of Chemical Engineering," 2nd ed., pp. 596-627, New York, McGraw-Hill Publishing
Co., 1927. (4) Weston,
IND. EXG.CHEM., Anal. Ed., 5, 179 (1901).
(1933).
(5) Zawidaki, 2. p h y s i k . Chena , 35, 145
RECEIVED May 7, 1938. Presented before the Microchemical Section at the 95th Meeting of the American Chemical Society, Dallas, Texas, April 18 t o 22, 1938.
