January 1948
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
saturation in order to operate in the metastable solubility zone, and changes of crystsl form under various conditions, should be investigated further. Elucidation of the relatione among variables such 88 these would aid materially in advancing the knowledge of crystalliaation. The research described in certain articles indicates new a p proaches or new techniques in crystallization processes which appear worthy of further exploration. Berlaga (6) shows that, with adequate cooling to remove the heat due to vibrations, application of a supersonic field t o undercooled salol (phenyl salicylate) increases the rate of crystrtlliaation. Furthermore, he notes an increase in the number of centers of crystallization. Boerboom (7) has studied crystallization of supersaturated solutions by means of an electric current. He states that when supersaturated solutions of electrolytes are electrolyzed, such a high concentration results from the transference of ions at one of the electrodes that crystallization nuclei are formed spontaneously and grow in the supersaturated solution. Gorbunova and Dankov (17) discuss elementary electrocrystallization processes theoretically, by taking as an example the isolated growth of a face of a single crystal forming part of the electrode in an electrolytic cell. They believe that the growth of a crystal from a mixture of gases and solutions is subject to the same relations as exist in electrocrystallization. Schutz ($6) and Bader and Schutz ( 8 ) describe the method and principles of fractionation by adsorption and crystallizatian on foam. West (44, 46) has patented an apparatus for growing single crystals of uniaxial material, such as sodium nitrate. This depends on the use of an added orienting surface such as mica. ACKNOWLEDGMENT
The authors acknowledge, with thanks, the help of W. L. McCabe of the Flintkote Company, East Rutherford, N. J., for supplying a portion of t,he literature references. LITERATURE CITED
(1) Aktieselskaget Krystal., Norwegian Patent 62,374 (1940). (2) Bader, R., and Schutz, F., Trans. Faraday SOC.,42, 571 (1948). (3) Badger, W. L. and Seavoy, G. E., “Heat Transfer and Crystallization,” Harvey, Ill., Swenson Evaporator Co. (1946). (4) Barkhuysen, F. H. C., Chem. Weekblad.,43,234 (1947). (5) Becker, R., Am. Physik., 32,128 (1938).
13
(6) Berlaga, R. Ya., J. Exptl. Theoret. Phy8. (U.S.S.R.), 16, 647 (1946). (7) Boerboom, A. J. H., Nature, 159,230 (1947). (8) Caldwell, R. B., C h m . I d . , 61, No. 3, 418 (1947). (9) Coalstad, 8. E.,Intern. Sugar J., 48,319 (1946). (10) Coalstad, S. E., J . Soc. Chem. Ind., 65,206 (1946). (1 1) Corn Products Refining Co.. Brit. Patent 596,951 (1947). (12) Dankov, P. D., J . Phys. C h m . (U.S.S.R.), 20, 853 (1946). (13) Davis, C. P., U. S. Patent 2,396,689 (1946). (14) Erikson, A. M. and Ryan, J. D., ZbM., 2,416,682 (1947). (15) Foster, G. H., and Williams, E. F., ZbM., 2,395,856-60 (1946). (16) Frenkel, Ya. I., J.Phys. U.S.S.R.,9,392 (1945). (17) Gorbunova, K. M., and Dankov, P. D., Compt. rend. acad. 815. U.R.S.S.,48,15 (1945). (18) Grove, C. S., Jr., Montillon, G. H., and Mann, C. A., Ph.D. Thesis, Univ. of Minn. (1941). (19) Haas, W. O., Jr., U. S. Patent 2,424,273 (1947). (20) Harbury, Lawrence, J. Phys. & Colloid Chem., 51, 382 (1947). (21) Havelock, J. R., Parr-Burman, B. A., Elkington, Frank, and Gilpin, W. C., Brit. Patent 587,054 (1947). (22) Hey, M. H., Nature, 158,584 (1946). (23) Holden, A. N., Phvs. Rev., 68,283 (1945). (24) Hughes, J. S., U. S. Patent 2,384,747 (1945). (25) Jang, J. A., Montillon, G. H., and Mann, C. A., Ph.D. Thesis, Univ. of Minn. (1943). (26) Jeremiasaen, Finn, U. S. Patent 2,375,922 (1946). (27) Kokatnur, V. R., Ibid., 2,393,108 (1946). (28) Kolb, €1.J., and Comer, J. J., J . Am. Chem. SOC.,68, 719 (1946). (29) Laurent, Pierre, Rev. d t . , 42,22 (1945). (30) McCabe, W. L., IND. ENG.CHEM.,38,18 (1946). (31) Morgan, R.A., and Walker, R. D.,Zbid., 37,1186 (1945). (32) Mukherjee, N. R., J. Imp. Coll. Chem. Eng. Soc., 2, 68 (1946). (33) Nelson, J. B., Bull. Brit. Coal Utilisation Research Assoc., 10, 257 (1946). (34) Ramberg, Haus, Geol. Fbren. i Stockholm Fbrh., 69, 189 (1947). (35) Ross, E. T., Pacific Chem. Met. Inds., 2 , No. 3 , 9 (1938). (36) Schuta, F., TrUn8. Faraday Soc., 42,437 (1946). (37) Semenchenko, V. K., J. Phya. Chem. (U.S.S.R.), 19,298 (1945). (38) Stuckenbruck, L. C., Osburn, J. O., and Grove, C. S., Jr., M.S. Thesis, State Univ. of Iowa (1947). (39) Til’mans, Yu. Ya., J. Gen. Chem. (U.S.S.R.), 16, 3 (1946). (40) VanHook, Andrew, IND. ENQ.CEEM.,38,60(1946). Am. Sugar Beet Technal., 4,559 (1946). (41) VanHook, Andrew, PTOC. (42) Wells, A. F., Phil. Mag., 37,184 (1946). (43) -Ibid.. 37.217 11946). (44) West,C.’D.,U.S.Patent 2,414,679 (1947). (45) Ibid., 2,414,680 (1947). (46) Zil’berman, Ya. I., and Ivanov, P. T., J . Gen. Chem. (U.S.S.R.), 16,1589 (1946). \--I
RECEIvlDD
November 24,1947.
HIGH TEMPERATURE DISTILLATION Bg T. J. WALSH,
CASE INSTITUTE
OF TECHNOLOGY, CLEVELAND, OHIO
HE future student of distillation will probably find that the year from November 1, 1946,to November 1, 1947,contains more interesting and useful data than any preceding year. Sparked by two symposia, one on distillation in general sponsored by the Division of Industrial and Engineering Chemistry of the AMERICAN CHEMICAL SOCIETYand the other on analytical distillation sponsored by the American Petroleum Institute, the number of articles in this field was double that of 1946. The subjects covered the range from microcolumns with a charge of 10 ml. to industrial bubble-cap stills several feet in diameter; pressures from 0.10 mm. absolute to the critical pressures of the distilling mixtures; from highly theoretical mathematical analyses of the differential equations involved to operational
comment on how a particular problem has been solved. A major trend has developed toward a return to the study of column calculations and how they can best be solved, with less attention directed toward the improvement of column capacities or efficiencies, The revelation of data has continued at a somewhat slower rate than previously. Principles of Distillation. Several reviews of the general principles of distillation have appeared during the year. Reed (479, in introducing the A.P.I. symposium, defines many distillation terms and describes the types into which analytical columns, column heads, and jackets may be classified. Edmister (17) and White (66) describe distillation behavior and outline the standard techniques for calculating column height. Cica-
14
INDUSTRIAL AND ENGINEERING CHEMISTRY
lese et al. (11) review tower operation and the solution of operational problems. The sequence of distillation opeiations has been considered from an economic viewpoint by Lockhart (36). As it is normal to think of highest pressure operation first, he calls “direct series” the operations of de-ethanizing-depropanizing and deisobutanizing-debutanizing. “Inverted series” refers to the reverse sequence-that is, depropanizing-de-ethanizingor debutanizingdeisobutanizing. While his calculations are based on incremental cost rather than on absolute cost, they show that each case should be considered as a separate problem. I n general, the relative amounts of products seems to have a greater influence on the cost than does the pressure. Rectification and desorption viere compared ,in the same Raschig ring-packed column, 6 inches in diameter and 47 inches long, by Deed et al. (14). There appears to be an agreement in the height of a transfer unit for the two operations. Distillation Calculations. The calculation techniques for distillation problems were the subject of many articles. Two nomographs have been proposed for the solution of the Rayleigh equation for simple differential distillations. That of Nord (42) is applicable to binary mixtures and that of Stanton (50) to either binary or multicomponent mixtures. Frisch (21)presents another nomograph which permits solution of the Carey formula for the allowable vapor velocity through bubble-cap plates. This article also includes a good derivation of the Carey formula from hydraulic considerations. Flash vaporization curves from relative volatility data are the subject of Smith and Wilson (49). If the relative volatility of any two components of a mixture is constant over the desired temperature range, it is possible to calculate flash vaporization curves quickly using this method. Underwood (63,6 4 ) has rederived his famous equation for minimum reflux ratio and extended it to ternary and multicomponent distillations. The use of the equation is illustrated with numerical examples. Stoppel (61) has developed an expression for the number of theoretical plates in a column by extending the operating lines and the equilibrium line of a McCabe-Thiele diagram beyond the normal limits of the chart until they intersect. The result is a simple power function of the two points of intersection. Applicable to nonideal as well as ideal systems is the method of Eshaya (18) for determining the number of theoretical plates. Unfortunately this method requires the expression of equilibrium data as a power series in x, and most data are not available in this form. The use of log-log plots for McCabe-Thiele calculations is recalled by Fowler (20). This is a technique that can be very useful in some problems and is often overlooked. A graphical solution of multicomponent distillation problems has been indicated by Harbert (29). This should serve to simplify the trial-and-error problem connected with plate-to-plate calculations. An entirely new approach to distillation in packed columns has been suggested by Bowman and Briant (6). With a refreshingly original concept, they set up the differential equations that must apply in a packed column. I n order to integrate these equations it was assumed that the interphase transfer rate of a component is proportional to the amount of that component that would have to be transferred from one phase to the other to bring the phases into equilibrium. Where comparison is possible, the numerical results obtained in this manner and by classical formulas are nearly identical. Plate tower calculations should deal with finite difference equations rather than differential equations. Methods of handling these equations through matrices are demonstrated by Amundson ( I ) . Laboratory Columns. This section includes all of the in-
Vol. 40, No. 1
formation on packed columns, even though the capacity of some of these columns is rather large for most laboratories. The most noteworthy laboratory column has been that of the Kational Bureau of Standards described by Willingham et al. (60) as a rotary concentric-tube distilling column. This column has a packed length of 58.4 cm. and an annulus of 1.09 mm. When rotating a t 4000 r.p.m. it showed an H.E.T.P. (height equivalent to a theoretical plate) of 1.2 cm. and a throughput of 2000 ml. (liquid) per hour. This combination of capacity and efficiency is quite remarkable in any column. Another concentric-tube column is reported by Donne11 and Kennedy (15). This is a microcolumn capable of fractionating efficiently charges as small as 15 to 30 ml. The column has an annular spacing of 0.75 mm. and is 100 cm. long. H.E.T.P. values range from 0.6 to 0.8 cm. The construction and testing of this column is given in detail. The results of several tests have been compared with the efficiencies predicted from opentube column theory. Although the shape of the curves for efficiency us. throughput follows the predicted shape, the actual values of H.E.T.P. are about four times that expected. It is believed that the discrepancy is caused by uneven flow of reflux on the wetted surfaces. Willingham and Rossini (59)have described the high efficiency columns used by the Bureau of Standards on American Petroleum Institute Research Project 6. These columns have between 100 and 200 theoretical plates a t total reflux, and charge capacities from 1/2 t o 15 liters. Operated a t reflux ratios between 125:l and 200:1, they yield product a t rates of 2 to 12.5 ml. per hour. McMahon packing (97) was introduced to fill a definite place in the production of oxygen from liquid air. This packing is essentiaily a wire saddle. When formed from ‘j4-inch squares of 100-mesh brass wire, the packing had 95,3y0 empty space and gave an H.E.T.P. of 1.5 inches. Here 450,000 pieces were packed into 1 cubic foot which weighed 25.2 pounds. Performance tests were made on this packing by Forsythe et al. (19) in a 6-inch column packed to heights between 5 and 9 feet. They report H.E.T.P. values ranging from 2 to 4 inches but observed a low pressure drop for the packing. This feature should be useful in vacuum columns. They also observed that increasing the packing height decreases the packing efficiency. Fractionation in a ‘jrinch Ewell column is described by Reed (48). Three types of insulating jackets, a heater jacket, a slip vacuum jacket, and a vacuum jacket, are compared. H.E.T.P. values range from 0.9 to 1.1 inches in a column 36 inches long. The charge is from 150 to 200 ml. and the boil-up rate between 150 to 300 ml. per hour. To divide a 200 ml. charge into 1 % cuts, Hall and Jonach (88) used 3/~-inchhelices. I n a vacuum-jacketed column 22 inches long and 11 mm. in diameter, these authors report a t least 80 theoretical plates a t total reflux. Care in column packing was the secret of their success. I n a carelessly packed column the efficiency might go as low as 20 plates. I n such a case the column was repacked before use. Holdup was 8 ml. of liquid, reflux ratio was 50:1, and 43 hours were required for the average distillation. Brandt et al. (8) describe the operating characteristics of a 25-mm. Hypercal column 36 inches long. Eight hours are required for this column to reach equilibrium. The efficiency varied from 92 theoretical plates a t 50 ml. per hour boil-up to 48 plates at 2000 ml. per hour boil-up. I n spite of this variation the column s e e m to have a good combination of efficiency and tkroughput. Larger columns were the subject of the articles by Borns et aE. (5) and Marschner and Cropper (88). The former describes a 3-inch Stedman packed column having a length of 12 feet. Operating a t pressures up to 100 pounds per square inch this column was designed to have 100 theoretical plates. During service the efficiency of the column decreased until, after 3 years of operation,
January 1948
INDUSTRIAL AND ENGINEERING CHEMISTRY
the column actually tested 40 plates at 20 liters per hour boil-up. From 10 to 14 hours are necessary for this column to reach equilibrium a t the start of a run. The columns of Marschner and Cropper are packed with l/3inch stainless steel helices. This packing has 88% free space and a static holdup of 4oj0. A 2-gallon-capacity still column is 1inch in diameter and 20 feet long. The 5-gallon still column is 1.6 inches in diameter and 8 feet long. The usual material balance in these stills is only 97%. A suggestion is made that, in futuro columns of this scale, provision be made for operating the columns at the condenser level to save holdup in the take-off lines. The authors also note that temperature recording to 0.1O F. is desirable. Zenz (61) considers the pressure drop through packed towers as a continuous function rather than as three straight lines depending on the operating point in the system. On this basis he is able to evaluate the flow constants for the gas-phase continuous state from orifice-type equations. When the liquid phase is continuous the relations seem to break down, and it is necessary to evaluate the constants empirically. Pressure drop is also considered by Leva (34). A general expression for pressure drop through packed tubes accurate to 8% is given. The shape of the packing is included through use of a shape factor in the equation. “Plate equivalents” is a new term used to express the efficiency of a column by Collins and Lantz (12)and by Goldsbarry and Askevold (84). Plate equivalents are defined aa the number of theoretical platcs a t total reflux which would give the same separation as is obtained when the column is operated a t a finite reflux. Columns evaluated in this manner are an Oldershaw plate, the Heli-Grid column, and l / & m h helices. The use of intermittent or continuous take-off is the subject of the article by Oldroyd and Goldblatt (45). In some cases of high efficiency columns it appears that intermittent take-off may result in sharper fractionation than the more common continuous operation. A new analytical technique is multiisothermal distillation described by Echols and Gelus (18). Isothermal distillation is fractional evaporation at constant temperature. Multiisothermal distillation is the use of the technique at several controlled temperatures. With a sample of limited composition range, such as a butane cut in a petroleum refinery, this may be used as a quick method of analysis. Samples as small as 2 ml. of gas may be analyzed with an accuracy of 1%. Necessary to any column is a still head or reflux divider. The standard procedures are to provide for intermittant operation on a cycle timer (8, 27); to split the liquid return from the condenser (38,67); or to divide the vapor stream to the condenser. The development of this latter technique, known &B CORAD heads (Constant RAtio heaD) is described by Lloyd and Hornbacker
(366). Industrial Columns. The derivation of an expression for calculating the hydraulic gradient across a bubble-cap tray is presented by Davies (13). This amounts to considering the caps as a dam across the tray. Although the equation is based on data from experiments on open trays, tower data seems to bear out the accuracy of the calculated gradients. A superfractionation operation waa carried out by Ward et al. (66) in fractionating 100 gallons of a 12% naphtha cut from an Oklahoma City crude. Data on 0.5% fractions of the naphtha are reported. Tower control is reported by Boyd (7)’ Governale ($66)’ Hutchinson (&?), and Mattix (39). Most of the control systems are specific to particular problems. The control of a de-ethanizer (Hutchinson) is perhaps the most interesting, although all of these articles are valuable aids in any control problem. A distillation-dr.ying operation for a light hydrocarbon mixture is described by Gester (23). The operation is successful in spi@
15
of extremely poor plate efficiencies (5%) obtained for this type of separation. Henry’s law WBS used in the design of the equipment. Plate efficiency for benzene-toluene separations as a function of operating variables was studied by Griswold and Stewart (86). A maximum efficiency of 55% was determined. No significant trend of efficiency with liquid composition, reflux ratio, or vapor rate was noticed. The system benzene-toluene-xylene was studied by Nord (43). This worker found a gradation of efficiency with composition. Maximum efficiency of 70% occurred at a benzene concentration of 60%. Extractive and Azeotropic Distillation. Information on the development and commercial operation of a furfural extractive distillation plant for the separation of the CChydrocarbons is presented by Buell and Boatright (IO). Solvent characteristics, process flow arrangement, performance data, process controls, and the degree of separation obtained are discmsed. This article also reviews the behavior of the materials of construction during three years of plant operation. Data on binary mixtures of furfural and hydrocarbon are given by Mertes and Colburn (41). These are extended to ternary mixtures by Gerster et al. (28). Azeotrope formers suitable for the purification of styrene are compared by Berg et al. (4). Optimum entrainers are 1nitropropane and isobutanol. In all cases it is necessary to make a choice between capacity of the column and quality of the desired styrene, That azeotropic distillation may not always be desirable is the conclusion of Berg and Harrison (3). The still heat load is increased and the capacity reduced from that of the operation without the azeotrope former. Distillation Data. Vapor pressures of organic and inorganic compound are compiled by Stull (62). These are basic to most distillation calculations. The existence of azeotropes in distillation systems and the effect of pressure on the azeotrope where one exists are considered by Horsley and co-workers (9, 30,31, 44). Data are presenbd on 6200 binary systems and 300 ternary systems. Graphical methods of predicting azeotropes are also included. A method for estimating the relative volatility of a system from the boiling points of the pure components is developed by Melpolder and Headington (40). The equation is valid only if the mixture obeys Raoult’s law. A graphical solution of Fenske’s equation for the minimum number of plates in a column operating at total reflux is included with this article. Three systems have been recommended for testing the e 5 ciency of columns operating under vacuum. Data on mixtures of di-n-butyl phthalate and di-n-butyl aeelate at 1 mm. pressure are given by Williams (68). This mixture is satisfactory for columns having up to 7 )theoretical plates. Perry and Fuguitt (48) present data on the systems di-2-ethylhexyl phthalate-di-2ethylhexyl sebacate suitable for 5 theoretical plates and di-2ethylhexyl sebacate-di-n-octyl phthalate suitable for 30 plates a t 0.10 mm. pressure. Analysis of all three systems is by refractive index. LITERATURE CITED (1) Amundson, N., Trans. Am. Inst. Chem. Ewr8,, 42, 939 (1946). (2) Bartieson, J. D., Conrad, A. L., and Fay, P. S., IND. ENQ. CHEM., ANAL.ED., 18,724 (1946). (3) Berg, L., and Harrison, J. M., C h m . Eng. Progress, 43, 487 (1947).
(4) Berg, L., Harrison, J. M., and Montgomery, C. W., IND.ENQ. CHEM.,38, 1149 (1946).
(5) Borns, W. J., Coffey, B.L., and Garrard, L. G., Proc. Am. Pe trolsum Inst., 26, III, 32 (19M). (6) Bowman, J. R., and Briant, R. C., IND.ENO.CHEM., 39, 745 (1947). (7) Boyd, D. M., Jr., Petroleum Re$ner, 26, 83 (1947).
(8) Brandt, P. L., Perkins, R. B., and Halveraon, L. K., Proc. Am. Pdrobum Z l z e t . , 26, 111, 57 (1946).
INDUSTRIAL A N D ENGINEERING CHEMISTRY Britton, E. C..Nutting, H. S., and Horsley, L. H., Anal. Chem., 19. 602 (1947).
Buell, C. K., and Boatright, R. G., IND.ENQ.CHEM.,39, 695 (1947).
Cicalese, J. J., Dsvies, J. A., Harrington, P. J., Houghland, G. S., Hutchinson, A. J. L., and Walsh, T. J., Proc. Am. Petroleum Inst., 26,111, 180 (1946).
Collins, F. C., and Lantz, V., Ibid., 26, 111, 72 (1946); IND. ENQ.CHEM.,ANALED., 18,673 (1946). Davies, J. A.,IND. ENG.CHEM.,39,774 (1947). Deed, D. W., Schuts, P. W., and Drew, T. B., Ibid., 39, 766 (1947).
Donnell, C. K., and Kennedy, R. M., Proc. Am. Petroleum Inst., 26,111, 23 (1946).
Echols, L. S., Jr., and Gelus, E., Anal. Chem., 19,668 (1947). Edmister, W. C., Pdroleum Engr., 18,No. 8,156 et seq. (1947). Eshaya, A,M., Chem. Eng. Progress, 43,555 (1947). Forsythe, W. L., Jr., Stack, T. G., Wolf, J. E., and Conn, A. L., IND.ENG.CHEM.,39,714 (194’7). Fowler, F. C., Petroleum Engr., 18,No. 13,57 (1947). Frisch. W.C., Petroleum Refiner, 26,598 (1947). Gerster, J. A., Mertes, T. S., and Colburn, A. P., IKD.ENQ. CHZM.,39,797 (1947). Gester, G. C., Jr., Chem. Eng. Progress, 43,117 (1947). Goldsbarry, A. W., and Askevold, R. J., Proc. Am. Petroleum Inst., 26,111, 18 (1946).
Governale, L. J., Chem. Eng., 54,No. 4, 126 (1947). Griswold, J., and Stewart, P. E.,IND.ENG.CHEM.,39, 753 (1947).
Grossberg, A. L.,and Roebuck, J. M., Chem. Eng., 54, No. 1, 132 (1947).
Hall, H. J., and Jonach, F. L., Proc. Am. Petroleum Inst., 26, 111, 48 (1946).
Harbert, W. D., Petroleum Processing, 2, No.8,610 (1947). Horsley, L. H., Anal. Chem., 19,508 (1947). Hnrslev ~. L. H.. ___._._I, , Ibid.. 19.603 (1947). Howard, F. L., Ibid.,’ 19,’144 (1947). Hutchinson, A. J. L., Petroleum Processing, 2,792 (1947). Leva, M., Chem. Eng. Progress, 43,549 (1947).
Vol. 40, No. I
(35) Lloyd, L. E., and Hornbacher, H. G I Anal. Chem., 19, 120 (1947). (36) Lockhart,, F. J., Petroleum Eng., 18,No. 12, 111 (1947). CHFM.,39,712 (1947). (37) McMahon, H. O.,IND.ENGI. (38) Marschner, R. F., and Cropper, W. P., Proc. Am. Petroleum Inst., 26,111, 41 (1946). (39) Mattix, E. D., Petroleum Refiner, 26, No.9,714 (1947). (40) Melpolder, F. W.,and Headington, C. E., IND.ENG.CREM., 39, 763 (1947). (41) Mertes, T.S..and Colburn, A. P., Ibid., 39,787 (1947). (42) Nord, M.,Ibid., 39,232 (1947). (43) Nord, M.,Trans. Am. Inst. Chem. Engrs., 42, 863 (1946). (44) Nutting, H. S.,and Horsley, L. H., Ana?. Chem., 19,602 (1947). ENG.CHEM.,ANAL. (45) Oldroyd, D.M., and Goldblatt, L. A., IND. ED., 18,761 (1946). (46) Perry, E. S., and Fuguitt, R. E., IND.ENG.CHEW.,39, 782 (1947). (47) Reed, C’. R., Proc. Am. Petroleum Inst., 26, 111, 10 (1946). (48) Reed, C. R., Ibid., 26, 111, 14 (1946). (49) Smith, D. A., and Wilson, G. W., Trans. Am. Inst. C h m . Engrs., 42,927 (1946). (50) Stanton, W.H., IND. ENG.CHEM.,39, 1042 (1947). (51) Stoppel, A. E., Ibid., 38, 1271 (1946). (52) Stull, D. R., Ibid., 39, 517 (1947). (53j Underwood, A. J. V., J.Inst. Petroleum, 32,598 (1946). (54) Underwood, A. J. V., Ibid., 32, 614 (1946). (55) Ward, C. C., Gooding, R. M., and Eccleston, R. H., IXD.ENG. CHEM.,39, 105 (1947). (56) White, R. IT., Petroleum Processing, 1, 151 et seq. (1946). ENG.CHEM.,ANAL. (57) Wilkinson, W. R., and Beatty, H. A., IND. ED., 18,725 (1946). (58) Williams, F. E., IND.ENG.CHEM.,39,779 (1947). (59) Willingham, C. B., and Rossini, F. D., Proc. Am. Petroleum Inst., 26, 111, 63 (1946). (60) Willingham, C. B., Sedlak, V. A., Rossini, F. D , and Westhaver, J. W.,IND. ENG.CHEM.,39,706 (1947). (61) Zenz, F. A., Chem. Eng. Progress, 43,415 (1947). RECEIVED October 31, 1947
HIGH VACUUM DISTIL PRODUCTS, INC., ROCHESTER 13, N. Y.
URING the past few years the interest in molecular distillation has shown signs of extending into molecular fractionation. At the same time, vacuum column rectification has advanced toward the high vacuum region so that it is not always convenient to treat molecular distillation as a separate art, sharply defined. This year’s review will attempt to cover the wider field and, in particular, to make a quantitative comparison between molecular and near-molecular types. The usual number of variations of laboratory short-path stills has appeared. A pot still (23) used in research on organic sulfur compounds warrants attention because of the total immersion in the heating bath, which ensures all the condensate being deposited on the internal condenser. A multiple micropot still assembly has been devised for the rapid analysis of vitamin E in fats and oils (21). It has been found that the four tocopherols are evaporated quantitatively from vegetable oils by 1/2-hour heating a t 170’ C., and the condensate is suitable for analysis by the Emmerie-Engel procedure. The falling-film cyclic batch stiM for laboratory use has appeared in yet another modification (24)?and similar concentric surface stills (19) continue to attract attention industrially, although not necessarily for use at molecular pressures. The quest for rapid distillation continues, with emphasis on extreme agitation of the evaporating surface. I n addition to the use of centrifugal force, this has been accomplished in the fallingfilm still by a traveling brush (16). A spiral centrifugal still (19) has becn described which allows a longer travel for the distilland
than is available with a simple rotating cone or plate. The distilling element is like a broad clock spring, the inside surface of which accommodates the distilland. The center of the spring is attached to the driving shaft, and the outside end revolves within a stationary gutter into which it discharges the spent distilland. The condensate collects on the walls of the container. An attempt has been made to secure countercurrent heat conservation in the molecular still by circulating a stable heat transfer liquidfor instance, near the center and periphery of a rotating evaporator (1.4)in a closed cycle operated by a pump traveling with the rotor. Thought has been given to the use of low pressure inert gas as a vehicle (5)for vapor in the high vacuum still. Commercial molecular distillatian has advanced through the duplication of centrifugal units with rotors 5 feet in diameter. A fairly complete disclosure (10) of operating conditions has been made, both for simple distillation and fractionation with batteries of stills. The volume of material distilled industrially is believed to have doubled during the year. High Vacuum. Passing now to molecular fractionation (multiple redistillation) and rectification under high vacuum in column stills, perhaps the simplest laboratory device that has appeared is the cascade fractionator (3)devised for the separation of mercury isotopes. Here ten evaporators are arranged in a row in such fashion that the condensates fall from the condensers, one cell upstream in series, while the distillands overflow, each one cell downstream. The still is operated 2-5 days on a charge and gives separations equivalent to 6-7 plates, each cell being
