THE SEPARATION OF LANTHANUM AND PRASEODYMIUM By JOHN WIERDA A.B. Hope College, 1921 THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE IN CHEMISTRY IN THE GRADUATE SCHOOL OF THE UNIVERSITY OF ILLINOIS, 1922 URBANA, ILLINOIS . , . 1322 W 6 3 UNIVERSITY OF ILLINOIS THE GRADUATE SCHOOL July 29 1922 1 HEREBY RECOMMEND THAT THE THESIS PREPARED UNDER MY SUPERVISION BY John WIer&a ENTITLED The Separation of Lanthanum and Prase odymium. BE ACCEPTED AS FULFILLING THIS PART OF THE REQUIREMENTS FOR THE DEGREE OF ScLene© In Charge of Thesis Acting Head of Department Recommendation concurred in* •o 1 vn Committee on Final Examination* Required for doctor’s degree but not for master’s Digitized by the Internet Archive in 2016 https://archive.org/details/separationoflantOOwier CONTENTS Page ACKNOWLEDGMENT I. INTRODUCTION 1 II. HISTORICAL 1 III. METHODS USED 3 IV. EXPERIMENTAL 3 Magnesium Oxide Separation 3 Precipitation by NH 4 OH in Presence of Ammonium 7 Salts Partial Decomposition of the Nitrate 8 V. COMPARISON OF METHODS 8 VI. SUMMARY 10 VII. BIBLIOGRAPHY 11 ACKNOWLEDGMENT The writer wishes to express his appreciation to Dr. H. C. Kremers under whose direction this in- vestigation was carried out. 1 THE SEPARATION OF LANTHANUM AND PRAESODYMIUM I. INTRODUCTION. Lanthanum and praseodymium are two of the so-called rare earth elements occupying adjacent positions in the periodic table and re- sembling each other very closely in chemical properties. On account of this resemblance it is not only impossible to effect a quantita- tive separation of the two but it is very difficult to obtain, even, the major portion of either, free from the other. The object of this investigation was to compare several of the reported methods in order to determine which one should prove the most satisfactory for the separation of these elements, one of which is to be used in a subsequent investigation. II. HISTORICAL. Methods used for the separation of the rare earths are divided into the following four classes: I. Separation by fractional crystallization of salts. II. Separation by fractional precipitation. III. Separation by means of processes of oxidation. IV. Separation by physical methods. The salts that can be used best for the separation of lanthanum and praseodymium by fractional crystallization are the sulphates, the double ammonium nitrates, the double magnesium nitrates, and the ethyl sulphates. Mosander 1 separated didymium (praseo- and neody- mium) from lanthanum by use of the former. Mendeleeff 2 , and later 2 von Wellsbach 3 , used the double ammonium nitrates for the same pur- pose, the former using an aqueous and the latter a nitric acid solu- tion. Drossbach" introduced the use of double magnesium nitrates. Spencer 5 states that ethyl sulphates can be used for f ractionating a lanthanum-didymium mixture, when neodymium concentrates in the first fraction and lanthanum in the last. Of the above methods the one most commonly used is the second. The process is slow and tedious. Spencer 0 in "The Metals of the Rare Earths" says: "Despite the large amount of material used by von Wellsbach, and the large number of recrystallizations carried out, it is extremely doubtful whether he ever got praseodymium free from it lanthanum. The double magnesium nitrates answer well for the sepa- ration of praseo- and neodymium but are not nearly as efficient for effecting a separation of praseodymium and lanthanum. The separation by fractional precipitation includes the partial decomposition of the nitrates and precipitation methods employing magnesium oxide, oxalic acid or oxalates, ammonia, caustic alkalies, organic bases, sodium sulphite, and. ammonium or alkali carbonates. A method recently devised by Prandtl and. Rauchenberger 7 consists in precipitating the didymium with ammonia in the presence of ammonium salts which cause the lanthanum to remain in solution. Oxidational processes can only be used for separating cerium and thorium from the rare earths. They cannot be used to separate the rare earths themselves, due to the fact that the latter consistently display a valence of three in all their salts. ' - <* . . . . . - 3 The physical methods of separation consist of absorption, dis- e CL tillation, and electrolysis. Hofman and Kriiss found that solution of rare earths with an equivalent weight of R 111 = 116.8 after shak- ing with charcoal and filtering, possessed an equivalent weight of Riii = 134.4. The amount of charcoal necessary, however, is large since the amount of material absorbed is small in comparison to the amount used. Investigation of distillation methods has failed to reveal anything promising except for the separation of scandium and thorium. Dennis and Lemon 9 by ten repetitions, - that is redissolv- ing the precipitated hydroxides and resubmitting them to electrolysis , succeeded in separating pure lanthanum from didymium. III. METHODS USED. The methods tried out during this investigation we re (1) the partial decomposition of the nitrates, (2) fractional precipitation by means of magnesium oxide and (3) fractional precipitation by means of ammonium hydroxide in the presence of ammonium salts. IV. EXPERIMENTAL • Magnesium Oxide Separation. For a preliminary trial a quantity of praseodymium oxide contain- ing some neodymium and lanthanum was dissolved in nitric acid and the solution filtered. The boiling solution was then treated with mag- nesium oxide in the powdered form. The d.ry oxide forms pasty glo- bules when added to the solution and these adhere together firmly being broken up only by long continued boiling. Thereafter a thin paste was made by add_ing water slowly to the oxide and stirring. - . 4 Afterwards, however, it was noted that if the water were added rapid- ly and the material stirred vigorously at the same time the oxide remained in suspension without forming any pasty material, whatso- ever. This could then he stirred and poured into the rare earth solution at any rate desired. Enough magnesium oxide was added to precipitate part of the rare earths, v/hich were then filtered off and washed. The precipitate was rather granular and filtered easily. Before and after the first precipitation the solution was tested spectroscopically. Praseodymium absorption lines showed very plainly each time. Neodymium lines were distinct but not very heavy before precipitation. In the filtrate after the first precipitation they were very faintly visible. (It may be advisable to say here that all spectroscopic examinations were made by examining the solution in a quartz box of 4 cm. by 1 cm. internal dimensions). The solution was then treated with about the same amount of mag- nesium oxide. More precipitate was obtained than in the former pre- cipitation, due, evidently, to the fact that the original solution was slightly acid and part of the oxide was used in neutralizing this acidity. After precipitation there were no absorption lines visible in the spectrum, showing the absence of both praseodymium and neodymium. The solution was then acidified with dilute nitric acid and tested for lanthanum by adding oxalic acid solution. No precipitate was obtained, indicating that the rare earths had all been precipitated. The first precipitate obtained was dried and weighed. (All weighings were made on platform balances and are only approximately 5 correct). It amounted to 48 grams. Some lanthanum oxalate was then reduced to the oxide by ignition in an electric furnace. The pra- $’ c,/ "' seodymium precipitate and/ the lanthanum residue was then dissolved in nitric acid and the solution filtered. Spectroscopic analysis indicated neodymium lines when viewed thru 4 cm. of solution, none thru 1 cm. The excess acidity was neutralized by adding carefully magnesium oxide to neutral reaction to litmus. 6 grams magnesium oxide were then added. After filtering and drying 25 grams of pre- cipitate were obtained. The color of it was light green. Neodymium lines were still present in the filtrate but very light. The precip- itation was repeated with the same amount of magnesium oxide. 24 grams of precipitate were obtained. The color was the same as the previous precipitate. Neodymium lines were absent from the fil- trate. Two precipitations were then made using 4 grams of the oxide for each. The color of the first precipitate amounting to 21 grams, was still greenish but lighter than the first two. The second pre- cipitate, amounting to 19 grams, was white. Praseodymium lines were evident in the filtrate after the first precipitation with 4 grams of the oxide but not after the second. Upon concentrating the solu- tion, however, they reappeared. Thereupon 1 gram magnesium oxide v/ as added. 5 grams of precipitate, were obtained. No prase odynium lines were visible until the solution was concentrated still further, and then only the line at 483. The line at 444 was entirely absent. The solution v/as then acidified with nitric acid and the lanthanum precipitated with oxalic acid. 50 grams of dried lanthanum oxalate were obtained showing that there v/as still considerable lanthanum in solution. f 6 To try out the method on a somewhat larger scale and to deter- mine whether all the praseodymium could he removed, by precipitation with magnesium oxide the upper fractions of a double magnesium ni- trate series were used# Both contained lanthanum and praseodymium but no neodymium. The saturated liquid of the two fractions was filtered and the crystallized, part dissolved in water and filtered. The total filtrate was then heated on a steam bath and treated in- termittently with small amounts of magnesium oxide suspended in water. After a considerable precipitate had formed it was filtered, dried and weighed and the process repeated. After a couple precipi- tations the praseodymium lines became weaker and the amount of the precipitate was decreased. After five precipitations of 138, 108, 20, 23, and 31 grams respectively, the praseodymium lines were only faintly visible. To obtain an idea of hov/ much lanthanum w as left in solution the remaining rare earth material was precipitated with oxalic acid and the oxalate dried and weighed. 510 grams of dried oxalate were obtained. The oxalate was reduced by ignition and dis- solved in nitric acid. The amount of solution was thus decreased considerably and the praseodymium lines showed plainly in the abeorp tion spectrum. Thereupon magnesium oxide was added to neutralize the excess, acidity and then precipitations were made by treating with 2 grams of magnesium oxide. After 11 further precipitations the praseodymium lines were invisible in the filtrate. Upon con- centrating the solution, however, till crystallization set in on cooling, the praseodymium line at 483 was visible. During these last 11 precipitations 120 grams of precipitate were obtained. • • 7 Precipitation by NH 4 OH in Presence of Ammonium Salts, Investigations by Prandtl and Rauchenberger 7 showed lanthanum hydroxide to be appreciably more soluble in the presence of ammonium salts than the hydroxides of praseodymium or neodymium. Experiments showed that the presence of magnesium or zinc increased the differ- ence in solubilities. Accordingly lanthanum and praseodymium should be readily separated by precipitation with ammonium hydroxide, using the double magnesium nitrates. For the experimental work with this method 500 cc. of fraction -3 of a double magnesium nitrate series was used. The fraction con- tained only a trace of neodymium. It was diluted to 1000 cc., 50 cc, of 4 times normal ammonium nitrate added, and the solution heated to boiling. Thereupon an equal volume mixture of 4 normal ammonium hydroxide and 4 normal ammonium nitrate v/as added drop by drop from a dropping funnel and the solution stirred, continually with a mechan- ical stirrer. After running in the precipitating solution at the rate of about 25 drops per minute for 40 minutes the solution was filtered .and the precipitate dried and weighed. 7 grams of precipi- tate were obtained which possessed a faint green color. The pre- cipitation v/as repeated and the amount of the precipitate increased somev/hat. After 7 precipitations totalling 88 grams of precipitate, there we re no praseodymium lines visible in the absorption spectrum. Upon concentrating the solution to crystallization a very faint ap- pearance of line -483 resulted. It was weaker than the same line in the absorption spectrum of the final filtrate from the magnesium oxide separation. The solution v/as then diluted, acid.ified and 8 treated with oxalic acid. 66 grams of dried oxalate, which gave 36 grams of residue after ignition, were obtained. Partial Decomposition of the Nitrates. For the nitrate decomposition some material containing only lanthanum and praseodymium was used. The solution contained about 100 grams of the oxides of these metals and an approximate spectro- scopic estimation showed about 10 /° praseodymium. The solution was evaporated carefully by heating in a casserole above a Bunsen flame. The heating was then continued till brown fumes of N0 2 were given off. After a certain amount of decomposition had taken place the heating was stopped, the fused mass allowed to cool and then treated with water. The residue was thoroly disintegrated by boiling after which it was allowed to cool and filtered, and the precipitate dried and weighed. Thirteen decompositions were made averaging slightly more than 3 grams each of precipitate. After the 13th pre- cipitate had been filtered off the filtrate still possessed a slight green color and showed praseodymium absorption lines thru 1 cm. of solution. The precipitates were not fractionally decomposed since the main purpose of the determination v/as to compare the speed of separation with the other methods used. V. COMPARISON OF METHODS. Of the methods tried the method that showed the most promising results for the separation of lanthanum and praseodymium to obtain pure lanthanum the precipitation method of Prandtl and Rauchenberger , - * 9 seems the most promising tho in this investigation pure lanthanum was not obtained thereby. The magnesium oxide precipitate shows signs of giving pure lan- thanum after sufficient precipitations but the yield would evidently be comparatively small. From the foregoing results it is evident that for obtaining pure lanthanum the method is not very efficient. According to Spencer 3 - 0 , Muthman and Roelig effected the separation of didymium and lanthanum by a single operation. It is doubtful whether they obtained pure lanthanum. The first precipitate thrown down by the magnesium oxide separa- tion has a greener appearance than, the corresponding precipitate of the Prandtl and Rauchenberger method. This may be due, however, to the fact that it may have a different chemical composition. The former is probably a basic nitrate while the latter is a simple hydroxide. One gram of each, precipitated from solutions of similar praseodymium content and dissolved in the same amount of solution gave absorption bands very nearly similar in intensity. (This weight was taken on an accurate balance). The precipitate obtained by magnesium oxide is more easily fil- tered than the other. The latter while not gelatinous when thrown down in the manner described above, does clog the paper somewhat and filters with more difficulty. What may be the better plan in trying to effect a rapid separation of lanthanum and praseodymium is to use magnesium oxide to precipitate the greater part of the praseodymium and then use the other method for removing the remainder. Whether 10 the last traces of praseodymium can be removed by it cannot be said at this writing. The nitrate decomposition should work well if heated gradually in an electric furnace where the heat would be uniform but as worked out in the open casserole, heated with a Bunsen flame, evidently too much local action takes place, due to superheating at the contact of the residue with the dish. It may work better for smaller amounts even as v/orked out in this investigation but for the amount used it fails to produce a very rapid separation. VI. SUMMARY. 1. Three methods for the separation of lanthanum and praseody- mium were tried out with the following results: (a) The partial decomposition of the nitrates as carried out in an open casserole over a Bunsen flame is not satis- factory. (b) The magnesium oxide precipitation effects a rapid separation up to a certain praseodymium concentration after which it removes the praseodymium only very slowly. (c) The method of Prandtl and Rauchenberger consisting of the precipitation by ammonia in the presence of ammonia salts apparently gives the best results for obtaining pure lanthanum. 2. A comparison of these methods was- made. 11 VII. BIBLIOGRAPHY. 1. Mosander. Annalen 48 210-23 (1843); J. pr. Gherri .50 276-92 (1843); Pogg. Ann. 60 307-15 (1843); Phil. Mag. 23 241-54 (1843). 2. Mendeleeff. J. Russ. Chem. Ges.j5 119-30 (1873); Liebig's Ann. 168 45-63 (1873); Ber.6 558 (1873); St. Petersb. Acad. Sci. Bull . 16 45-51 (1871). 3. von Wellsbach. Monatech.6 477-91 (1885); Sitzber. K. Akad. Wiss. Wien. 92 II. 317-31 (1885). 4. Brossbach. Ber.35. 2826-31 (1902). 5. Spencer. The Metals of the Rare Earths, pp. 31-32 (1919). 6. Spencer. The Metals of the Rare Earths, p.28 (1919). 7. Prandtl and Rauchenberger . Zeit. Anorg. Chem. 120 120-8 (1922); Berichte 53 759-69 (1920); Chem. Abs.L4 2305 (1920); Chem. Abs.16 1189-90 (1922). 8. Hofmann and Krftss. Zeit. Anorg. Chem, 5 89-91 (1893). 9. Dennis and Lemon. J. Amer. Chem. Soc.-39 151-37 (1915). 10. Spencer. The Metals of the Rare Earths, p.35 (1919). m