INDUSTRIAL AND ENGINEERING CHEMISTRY
422
facturing process upon the vita& G content of these fish meals was so pronounced, i t is of practical importance to use that method which best preserves the vitamin content, since the special sources of this vitamin used in feeding poultry are expensive. If this is done, a fish meal of highest nutritive value will be obtained, since the protein efficiency will also be superior.
Acknowledgment Acknowledgment is hereby made of the cooperation of the Grange League Federation Exchange, Inc., of Ithaca, N. Y., which made this investigation possible by establishing a temporary investigatorship a t Cornell Unhersity.
Literature Cited J. Nutrition, 5,503 (1932). (2) Daniel, E. P.,and McCollum, E. V., U. S. Bur. Fisheries, Investigationel Rept. 2 (1931). (1) Curtis, P. B., Hauge, S. M., and Kraybill, H. R.,
VOL. 27, NO. 4
(3) Fisher, R. A., "Statistical Method for Research Workera,"
w.,
(3A) Harrison, p. 99,Edinburgh, R. and Oliver associates, and Boyd, to be1930. published. (4) Inmaldsen, T.,can.Chm. M ~ . 13, , 97 (1929). (5) Ibid., 13, 129 (1929). (6) Maynard, L. A., Bender, R. C.. and McCay, C. M., J. Aut. ,-, Research, 44,591 (1932). (0 Maynard, L. A., and Tunison, A. V., IND.ENO. CHEM.,24, 1168 (1932). (8) Norris, L. C., Heuser, G . F., Ringrose, A. T., Wilgus, H. S., Jr., and Heiman, V., Atti V congr. mondiale pollicoltura, 2, 512 (1934). (9) Norris, L. C., Heuser, G . F., Wilgus, H. S., Jr., and Ringrose, A. T., Poultry Sci., 10,93(1931). (io) Record, P. R., Bethke, R. M., and Wilder, 0. H. M., J. Aut. Research, 49,715 (1934). Schneider, B., Ibid., 44,723 (1932). Wilder, 0.H. M., Bethke, R. M., and Record, P. R., Ibid., 49,723 (1934). (13) Wilgus, H. S., Jr., Norris, L. C., and Ringrose, R. C., Atti V congr. mondiale pollicoltura, 2, 541 (1934).
RECEIVED November 30, 1934.
Technical Extraction of Protoactinium
@A
PROCESS for purity to the extent of 0.1 per A detailed description of a technical the large-scale cent. The protoactinium conprocess for the extraction of the radioextraction of tent was, on the average, 300 active protoactinium (the longest lived protoactinium, the longest lived mg. of PalOr per metric ton (a isotope of element 91) from radium residues isotope of element 91, is here concentration of 1:3,000,000), is given. The protoactinium concentration described. This is a continuawhereas the richest pitchblendes tion of the technical extraction contain (5) about 200 mg. per is raised from 1:3,000,000 in the residues to process developed by one of the ton. 1:5000 in the plant product, with a yield authors' a t the I. G. FarbeninProtoactinium has exactly the of 80-90 per cent of the theoretical. dustrie's factory in Ludwigschemical properties we would hafen, Germany, in 1928. expect from a higher homolog T h r o u g h the c o u r t e s y of of tantalum; besides the natural A. V. GROSSE AND M. S. AGRUSS Lindsay Light and Chemical analogy with the latter, it exUniversity of Chicago, Company, operating space was hibits a number of individual Chicago, Ill. rented a t their factory in West properties (2, 3). A characterChicago, Ill., where a pilot plant istic reactidn' iS its c o m p l e t e . was &stalled for the conversion of the radium residues. precipitation with zirconium phosphate. Protoactinium occurs in nature in smaller concentrations The process finally adopted for the large-scale extraction of than radium. Every natural uranium mineral contains 8 protoactinium is shown in Figure 1, The three main steps are: grams of protoactinium for every 10 grams of radium ( 4 ) . 1. Hydrochloric acid leach of the starting material. Fortunately, during the process for radium extraction from 2. Sodium hydroxide fusion of the residue from the leach. pitchblende by Mme. Curie's method, the protoactinium 3. Water leach of the fusion and elimination of silica. concentrates in the residues and gives a better starting material than is found in nature. The starting material in the Plant Process present experiments was the final residue, the so-called The treatment with hot 25 per cent hydrochloric acid disRiickriiclcstande, from the Czechoslovakian State Mining solves all of the iron, the more basic oxides, and most of the Factory in Joachimstal, Bohemia. lead, leaving a protoactinium concentrate consisting chiefly The pitchblende Riickriickstitnde are in the form of a fine of silica with small quantities of zirconium, titanium, and powder, reddish in color, and very gritty, containing smaller other less basic oxides. The residue from the hydrochloric or larger quantities of foreign material (like bricks, stones, acid leach was fused with flake caustic soda to convert the etc.). The average percentage composition is as follows: silica to sodium silicate, the melt was leached with water, and SiOa 60 Ah01 5 CaO 0.6 the sodium silicate extract was iiltered. This treatment Fer08 22 MnO 1 MgO 0.6 PbO 8 eliminated most of the silica and all of the lead. The residue, containing most of the protoactinium, was dissolved in acid, It contained also small amounts of titanium (0.3 per cent), thereby causing the precipitation of the remaining eilica zirconium (0.1 per cent), and hafnium together with many other elements. Graphite occurred as an occasional im- (from insoluble silicates). This precipitate was coagulated with steam and removed by filtration. The protoactinium 1 A complete description of this process was left in the hands of 0. Hahn in the acid filtrate was recovered by precipitating it in the and later briefly described by 0. Erbacher in Ullman's Enzyklopaedie der cold, together with zirconium phosphate (termed ZrP2OT I), technisohen Chernle, 2nd ed., Vol. 8, p. 641 (1931). See also Grosse (8.
424
INDUSTRIAL AND ENGINEERING CHEMISTRY
VOL. 27, NO. 4
Sodium Hydroxide Fusion
operations. It must be borne in mind that, when the caustic soda leach is washed by decantation, a perfectly clear supernatant liquid cannot be obtained unless a day is allowed for each settling. Therefore, the small loss due to a few fine particles in the supernatant liquid was completely counterbalanced by the saving of time and labor involved.
The second operation of the process consisted of a 670 KO. caustic soda fusion of the dried residue o b t a i n e d 18,000 m e n s from the hydrochloric acid NaOH N S l O Y AND WATER UACW leach. The f u s i o n was OF MELT (YLCH4YICAL LOSS 6 % ) c a r r i e d o u t i n a steel c y l i n d r i c a l pot set in a s t e e l t r i p o d o v e r a gas burner. Fire brick and sheet asbestos surrounded SOLUTIOd I N 20% Kl ANQ CMOVUITION OF S I L I C A the pot to eliminate heat losses. Precipitation of ReThe residue from the maining Silica a c i d l e a c h (10 kg.) and flake caustic soda (30 kg.) The residue f r o m t h e were thoroughly mixed. caustic s o d a l e a c h was Half of this mixture was dissolved in 20 per cent placed i n t h e p o t a n d hydrochloric acid in a 60heated until the reaction gallon (227-liter) cr o c k , ceased; the remainder of and the remaining silica the mixture was a d d e d was coagulated by passFl LTRATE gradually to the molten ing steam into the acid somaterial. When all the lution u n t i l i t r e a c h e d mixture had melted and 95" C. The coagulated the reaction ceased, the silica takes with it 80 per liquid was s t i r r e d a n d cent or more of the protoheated until it reached a actinium in the r e s i d u e temperature of about 700' from the caustic soda melt, to 800" C. in order to reas well as 90 per cent of duce completely the small the zirconium hydroxide amount of lead still present in t h e s a m e r e s i d u e . to the m e t a l l i c s t a t e . Table I11 shows the protoThe metal was free from actinium content of the FIGURE1. PROCESS FOR LARGE-SCALE EXTRACTION OF protoactinium; its reduccoagulated silica. PROTOACTINIUM tion was probably due to T o t h e cold acid filthe carbon present in the trate, obtained from the original ore. silica, was added an exEach fusion required 2 hours for melting and heating to the cess (250 to 300 cc.) of 30 per cent hydrogen peroxide (to oxidize the titanium to pertitanic acid and prevent its co-prefinal temperature, and consumed a total of about 600 cubic feet (16.8 cubic meters) of gas. The very light and fluffy residue from the first operation caused some loss (about TABLEI. PROTOACTINIUM CONTENT OF RESIDUE FROM WATER 6 per cent) during the process of fusion, even though the LEACH OF MELTS melting pot was covered. This loss has, however, been Residue from Per Cent Pal06 greatly decreased by deflecting the upward rush of air from of Initial Content, Yield Leach, the burner. It can also be eliminated by moistening the Run No. Kg. Ore Mg.6 Per Ceht melting mixture with a very concentrated (70 per cent) solu7.7 12.8 16.3 90,s 6 ulus 7 ~- . ~. . 8 plus 9 7.3 12.1 15.4 85.5 tion of sodium hydroxide before placing it in the melting pot. 7.3 12.1 10 plus 11 15.9 88.3
-
i4
Leaching of Melts The third step in the process was a water leach of the cooled caustic soda melt. Eight such melts were combined and leached with 2600 liters of hot, soft water in a 3200-liter steel tank. The residue was washed by decantation five or six times before it was separated from the supernatant liquid by a plate and frame filter press. The material in the filter press was washed with hot, soft water until free from alkali and was dried with compressed air. The average dry weight of the residue from the caustic soda leach was 3.6 kg. from 30 kg. of starting material, and the yield of protoactinium a t this point was between 85 and 90 per cent. Table I shows the protoactinium content of the residue obtained from the water leach of the melts and Table I1 shows the loss of protoactinium during this operation. The data in Table IT may give the impression that large losses of protoactinium occurred during the melting and leaching
a
12 plus 13 6.1 10.2 10.0 Average Pa205 content of initial material is 18 mg.
TABLE11. Loss
OF
88.8
PROTOACTINIUM DURING LEACHING OF MELTS
Run No.
Water for Leach, Liters
PmOa Loss, Per Cent of Initial
1 2 12 plus 13 14
1280 1280 2560 1280
3.0
5.0 1.2 5.1
cipitation) and an excess (250 to 300 cc.) of 20 per cent phosphoric acid. The precipitated zirconium phosphate took with it the protoactinium (usually 2 mg. or less from 240 kg. of starting material) still contained in the acid filtrate and was termed ZrP207I. Originally it was intended that this operation should consist of dissolving the whole residue from the caustic soda
INDIJSTRI4L AND 13YGINEEHING CI'IEMISTHY
Al'lUl~, 1935
leach in acid arid precipitating all the zirconium as a phosphate f r m r this acid solution. Quite nnexpectedly this proved to be a troublesome operation under the conditions of these experiments. When the residue from the leach was dissolved in 20 per cent acid (sulfuric or hydrochloric), it thick gelatinous mixture was obtained because the starting material contained undecomposable, water-insoluble silicates which n'ere decomposed by the acid and gave free silicic acid. Reinciting of the residue did not completely eliminate the silica arising from these nudecomposable silicates. Ilowever, it was noticed that if live steam was passed into the gelatinous acid solution until it reached 95" C., the silica senarated out in ._ elohules which settled aoicklv . _ and filtered easily. This precipitation was later uscd to advantage in that it provided a concentrate from which the protoactinium could he easily separated and allowed to minimize the amount of zirconium oxide added during the first operation of the process.
425
(potassium hydroxide-melt), contained most of the protoactinium in a concentration of 1:5000, whereas the s t a r t ing material contained the protoactiiiium in a concentration of 1 : 3,000,000. Tahle IV shows the yield of protoactinium in the plant product. Since the plant product contained the protoactinium in so high a concentration, it was advisable to out the further purification in the laboratory. To date 1200 kg. of original residnes have been processed and 250 mg. Pa& have been obtained i n a concentration of 1 :5000 or higher. One hundred milligrams of this protoactinium have been isolated in the
'I'rcntment of Precipitated Silica The silica, collected from sixteen caustic sods melts (corresponding to 240 kg. of original starting material), was treated in an iron tank aitli a hot, 20 per cent sodium hydroxide solution. This treatment dissolved most of the silica and the small amount t h a t r e m a i n e d could be easily handled in subsequent operations. The residue from this treatment was washed by decantation and separated from the supernatant liquid by means of a plate and frame filter press. The moist residue, consisting of a small amount of - ..... silica and basic oxides easilv soluble in acid. was dissolved in 20 per cent h;drochloric acid; the solution filtered free from graphite, and the clear filtrate treated at room temperature with an excess of 30 per cent hydrogen peroxide and 20 per cent phosphoric acid as previoosly described for ZrP,O, I. The airconium phosphate thus obtained contained the protoactinium which remained with the silica and was termed ZrPAOI11.
1,250 3,284 570 4.835 20,339
2 12 plus 13
15 18
20-23,~ n r l ~ ~ i v e
TABLEIv.
20-27 28-33 38-43 44-51
230 225 222 225
Y I E L D 08
S5.6 75.5 84.3 90.0
1.8 1.0 1.2 2.0
7.8 16.5 7.6 8.0 38.i
PHoTOACTLNlUM IN
765.4 761.5 9O1.i 5W.0
48.2 32.R 49.7 49.3
86.3 91.7 85.0 80.0 83.2
PLANT PRODUCT
250.0 5.0 208.7 20.0 4.0 4W.0 130.0 4.5
55.0
53.8 54.8 55.8
82.6 83.9 SP.3 82.0
Plant Product The ZrP& I and 11, together with the graphite, cornprised t,he plant product. Each of these three products still contained substantial quantities (about one-third) of silica and was worked separately. The ZrP,O, I1 contained, on the average, 70 per cent of the protoactinium in t.he star!.ing material, the Z:rP20iI contained an average of 3 per cent., and the paphite an average of 15 per cent, giving an average yield of 85 to 90 per cent for the plant product. The plant product, after eliminating the phospborio acid and silica
jiurc state as the pentoxide (6), ParOr. Iarger quantities will be extracted in the future. 11%order to follow the protoactinium during the process, it is hnperative to run control analyses, if possible, after each operation. Owing to the fact that protoactinium is a radioactive element emitting alpha particles, it ca.n be easily traced through the subsequent steps of its Concentration by means of an ordinary Rutherford alpha-electroscope, Since other radioactive elementsare also present in the residues, the protoactinium has to be chemioally separated from them before its alpha activity is measured. (For analytical procedure see citation 4.)
Discussion 'The present process d8em from the first process used by of the authors in that it uses silica besides zirconium phosphate as a carrier for protoactinium. Since the silica can he easily separated from the protoactinium by means of sodium hydroxide, it simplifies matters greatly. The same process, with appropriate adjustiiients, could be used for residues from any pitchblendes. For carnotites or other material the same principle 4. e., bringing the protoactinium in acid solution and precipitating with zirconium phosphate and silica--F;hould also prove successfnl. Ibcently this process for extract,ion of protoactinium has been successfully used in Germany hy Grane and B d i n g ( 1 ) who were able to extract from 5 tons of the same residues 0.7 gra,m PazOa. The only changes these authors made were the reversing of sequence-i. e., first melting the original material in caustic soda and then making a hydrochloric acid leach with the residue obtained from the melt. This is a disadvantage because larger amounts of silica are left in the residue from the melt and cause difficulties in the hydrochloric acid extraction. These difficulties induced Graue and VIE
INDUSTRIAL AND ENGINEERING CHEMISTRY
426
Kfiding to eliminate large quantities of silica by fuming with hydrofluoric and sulfuric acids, a troublesome operation on a technical scale. Furthermore these workers used exceedingly large amounts of zirconium oxide (25 kg.) which causes great difficulties in filtration and handling and which are absolutely unnecessary. One-tenth of this quantity would be sufficient, as can be seen from the present results.
Acknowledgment The funds necessary to accomplish the work here described were generously donated by Hiram J. Halle of New York, K. Y., to whom the writers wish to express their sincere thanks. They are indebted to A. C. Ratchesky, U. S. Minister to Czechoslovakia, and to F. Novotny, of the U. S. Legation in Prague, for their efficient support in obtaining the raw material from the Czechoslovakian Government. The authors are also indebted to Julius Stieglitz for his personal interest and helpful advice during the course of this work. They wish to thank the Lindsay Light and Chemical Company, as well as M. W. Eichelberger and C. W. Stabenau, for their cooperation and assistance in carrying out the process a t their plant. Due thanks are given to
VOL. 27, NO. 4
Herbert N. McCoy for his technical advice and to D. R. Sperry, of D. R. Sperry and Company, for his coijperation in the use of a filter press. The authors are also grateful to the management of the Universal Oil Products Company and to Gustav Egloff for their continuous and many-sided help. Acknowledgment is due F. Benson who carried on most of the operations a t the plant very efficiently. At the present time the work is being supported by a grant from the R. A. F. Penrose Fund of the American Philosophical Society.
Literature Cited Graue, G., and KLding, H., Nutunoissenschaften,22, 386 (1934); 2. ungew. Chem., 47, 650 (1934). Grosse, A. V., Ber., 61, 233 (1928). Grosse, A. V.,J. Am. Chem. SOC.,52, 1742 (1930). (4) Grosse, A. V., Phgs. Rev., 42, 565 (1932). (5) Grosse, A. V., Science, 80, 514, Table 3 (1934). (6) Grosse, A. V., and Agruss, M. S., J. Am. C h m . SOC.,56, 2200 (1934). RECEIVED December 27, 1934. This paper is taken partly from a thesis submitted by M. S.Agruss to the faculty of the Division of Physical Sciences, University of Chicago, in partial fulfilment of the requirement8 for the degree of doctor of philoeophy.
Behavior of Oxidizing Agents with Activated Carbon A. S. BEHRMAN AND H. GUSTAFSON, International Filter Company, Chicago, Ill.
W
ITHIN the past few years the water purification world has become well acquainted with activated carbon and is now employing this new reagent in an ever-increasing number of applications. Generally speaking, activated carbon is utilized in water purification in two ways. In the first, the carbon is employed for removing from the water by direct adsorption in the carbon those substances which give to the water undesirable taste, odor, color, or other objectionable characteristics. Among the taste- and odor-producing substances commonly removed in this fashion are the phenols and chlorophenols, as well as certain tastes and odors arising from a variety of organic sources, such as the decomposition products of algae and other microorganisms. In the second method the carbon is employed primarily for the removal of free chlorine. This free chlorine may be either the relatively small residual amount (ordinarily a small fraction of a part per million) intentionally left in a community water supply as a protective measure t o assure continuous sterility; or it may be the considerably higher amount-frequently as much as 1 or 2 parts per million, and sometimes more-following a heavy dosage of chlorine applied usually for the purpose of effecting the oxidation of organic matter responsible for the objectionable taste, odor, color, or other of the undesirable characteristics mentioned. I n this second method of utilization the carbon functions not simply as an adsorbent but as a chemical reagent as well, acting as a reducing agent to convert the applied free chlorine to the chloride ion. Theoretically, 1 part by weight of pure carbon, completely reactive, could thus convert about 12 parts of free chlorine to the chloride ion. It is not the purpose of this paper to discuss in detail the I
application of these known methods of water purification employing activated carbon. Several very full discussions of this sort have already appeared in the literature (1). The object is to point out that the chemical reaction of the carbon and the free chlorine, employed in the second method described, is not an isolated phenomenon but is in agreement with a theory of behavior that was worked out several years ago in this laboratory; that is, activated carbon exhibits a marked adsorption tendency toward oxidizing agents in aqueous solution in general, and the adsorbate may or may not react with the carbon, depending partly on concomitant conditions and principally on the nature of the oxidizing agent itself. There is an important and practical development of this theory: In those cases in which the adsorbed oxidizing agent is retained in the carbon, it is possible to treat other fluidsboth liquid and gaseous-for a variety of purposes merely by bringing them in contact with the carbon containing the adsorbed oxidizing agent. Under proper operating conditions, the adsorbed oxidizing agent is retained by the carbon and is released in appreciable amount only as required by the presence of some reactive substance in the fluid being treated. For convenience in this discussion we may divide into three types of behavior the action of various oxidizing agents with activated carbon: 1. Adsorption and catalytic decomposition of the oxidizing agent. 2. Adsorption of the oxidizing agent accompanied by its reaction with the carbon: a. With release of the reaction products from the carbon. b. With retention of at least one reaction product in the
carbon.
3. Adso tion and retention of the oxidizing agent unaccompa-
nieTby chemical reaction with the carbon.
