A STUDY OF THE RARE EARTH ACETYLACETONATES Thesis for the Degree of Ph. D. MICHIGAN STATE UNWERSITY James A. Kennelley, Jr. ’ 1956 ¥__4 4 THEQGS This is to certify that the thesis entitled A Study of the Rare Earth Acetylacetonates presented by James A. Kennelley, Jr. has been accepted towards fulfillment of the requirements for Major pro essor \ Date May 18) 1956 H LiIRARY Michigan State University PLACE IN RETURN Box to remove this checkout from your record. TO AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE 6/01 c:/CIFiC/DateDue.p65-p.15 A STUDY OF THE RARE EARTH ACETYLACETONATES BY James A. Kennelley, Jr. A THESIS Submitted to the School for Advanced Graduate Studies of Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 1956 Q r L O ACKNOWLEDGMENT The author wishes to express his sincere appreciation to Dr. L. L. Quill, Professor and Head of the Department of Chemistry, for his advice, counsel, and friendship, without which this work.wou1d have been impossible. I. II. III. Iv. TABLE OF CONTENTS Introduction . . . . . . . . Historical . . . . . . . . . . Experimental . A. Average Atomic Weight Determinations B. Purity of Rare Earths C. Neodymium Acetate . . . . . . . 1. Preparation and Analysis 2. Solubility Determinations . . D. Rare Earth Acetylacetonates . 1. Purity of Acetylacetone . 2. Preparation and Analysis . . 3. Drying Studies . . . . . 4. Melting Points . . . . . 5. Solvent Extraction Studies 6. Absorption Spectra . . . . . a. Ultraviolet Region . . . b. Visible Region . . . . . c. Infrared Region . . . . . 7. Solubility Studies 8. Structural Considerations 9. X-ray Diffraction Studies Summary . . . . . . . . O . 21 . 21 . 22 . 23 23 . 30 . 33 . 3b . 37 . 43 50 51 . 57 . 57 62 6h 77 . 84 . 88 . 91 INTRODUCTIQE ‘—.-. One aim of every rare earth chemist is to find a simple, rapid, and economical method for the separation of the group of inner transition elements which are termed the rare earths. The great chemical and physical similarities among these elements have made separations of the pure compounds very difficult. Various investigators have tried fractional crystallization, fractional sublimation, homogeneous phase precipitation, ion exchange, extraction, and other methods to achieve this separation (2, 22, #3, 52). Most of these methods are time consuming and require careful attention or repetition of tedious procedures. In recent times ion exchange has been used most extensively for obtaining pure compounds of the rare earths, but dilute solutions must be employed and as yet commercial applications have been limited (63). Separations making use of differences in oxidation states are valuable for isolating several members of the series (#0). Recently Peppard gt_gl, (#9) have used solvent extraction techniques with success. If volatile compounds of the rare earths could be prepared, separation of the series into its individual members might be vastly simplified. This fact has appealed to investigators for many years, but as yet no usable, volatile compounds of the rare earths have been prepared. -2- The alkyls seem likely to be volatile, but Quill and Seifter (65) were not successful in attempts to prepare such compounds. It might be that covalent rare earth compounds would have desirable properties for separations. The rare earths are primarily trivalent and have six coordination positions. To form a covalent compound with a given rare earth a monovalent, bidentate chelating reagent is required. The metal organic compounds formed by 2,h-pentanedione, which hereafter will be called acetylacetone, are examples of this type of compound. In the enol form acetylacetone has one replaceable hydrogen and a carbonyl oxygen suitable for coordinate covalent linkage. Since this carbonyl oxygen is on the carbon atom beta to the carbon to which the hydroxy group is attached, a stable six memberedring«ie possible. The above may be pictured as follows: CHB-C-CHZ-Dc-CHB CH3I~C30H~C~CH3 ll ii I II 0 O O O H Keto Form Enol Form CH -C=CH-C-CH 3 A g 3 \../ M Chelate Ring -3- In the above diagram M is any di, tri, or tetravalent metal which is capable of chelation with the coordinating ligand. Such compounds with acetylacetone as the organic constituent have been prepared for a variety of metals. Many of these compounds are soluble in organic solvents and some are found to be volatile. Morgan and Moss (b7), finding that scandium acetylacetonate sublimes without decomposition from 2100 to 260° C., accomplished a separation of scandium from thorium by taking advantage of the volatility of the scandium compound. Chelate compounds have been used to effect the separa- tion of metals by extraction (66, 67), but it is believed that for the rare earths the potentialities have never been fully exploited. It would seem logical that a chelats compound might be used for this type of separation, since for this purpose the following requirements should be met: 1. The substance should be easily formed and should have definite stoichiometry. 2. The chelating agent should be readily available and of moderate or low cost. 3. The compound should be soluble to an appreciable extent in organic solvents. Metal acetylacetonates seem to fulfill these qualifi- cations. Acetylacetone forms compounds with over sixty metals (66) to give sufficiently stable compounds so that formation occurs at low metal ion concentration, which suggests the possibility of a systematic scheme for the -h- separation of the metals by means of extraction. The rare earth acetylacetonates are stoichiomstric compounds, and the reagent is commercially available in practical or pure form at reasonable cost. Statements in the literature concerning the properties of the rare earth acetylacetonates are not too clear. Ephraim (20) states that the acetylacetonates of the rare earths are particularily valuable as they volatilize unchanged, are soluble in many organic solvents, and may be used for the determination of molecular weights as well as for the separation of the elements from one another. In a later edition the statement is similar, but the last phrase states that this method is not satisfactory for the separation of these elements. According to Little (3?). the rare earth acetylacetonates may not be sublimed without decomposition. Little also states that rare earth acetylacetonates are sparingly soluble in water, but may be crystallized from aqueous alcohol or organic solvents such as benzene or chloroform. No quantitative data on the solubilities of rare earth acetylacetonates are recorded in the literature, but several very broad and general state- ments may be found (8, 33, us, 74, 77). In short, these rare earth compounds have never been completely characterized. In this work solubilities, ultraviolet, visible and infrared absorption data, Xpray data, extraction curves, and various other properties of the rare earth acetylacetonates will be presented. HISTORICAL Combes in 1887 (lb) was the first person to synthesize and work with acetylacetone. The following equations describe the method he used for the preparation: o N 6 033-0-01 + A12C16 5; 4 H01 + C12H1h06A12C18 CleiuAlzcls see CH _c-cs -c_cs 3:12:13 0 0 He discovered that one of the hydrogen atoms of this compound was quite acidic and behaved like the hydrogen of a carboxyl group. Combes also found that this hydrogen was easily replaceable by sodium or potassium. In addition to these two salts he also prepared the magnesium, aluminum, copper (II), iron (III), and lead derivatives of acetylacetone (1b, 15). These preparations were accomplished by allowing the acetylacetone to react with an aqueous solution of a 702)n was assigned to these compounds with M as a metal and n as its salt of the metal. The empirical formula M(05H oxidation state. Combes reasoned that since one hydrogen was replaceable, the second hydrogen atom of the methylene group, which is in the alpha position to both carbonyl groups, should also be replaceable. Hence compounds of the following type might be made: -6... where M is a divalent metal. It is apparent that Combes was not aware of the keto-enol tautomerism existing in this compound. He prepared 3-methy1 2,h—pentanedione and found it to have very much the same properties as acetylacetone. The melting point of aluminum acetylacetonate is given as 193° 6., and Combes stated that this compound may be distilled without decomposition at 310° C. (16). Analysis for aluminum gave a percentage of 8.h3 as compared to the theoretical value of 8.47. Vapor density measurements were also made, and it was concluded that the compound was monomeric. Combes made the observation that these metal derivatives of acetylacetone are soluble in organic solvents. Claisen and Erhardt in 1889 (13) prepared acetylacetone by condensing acetone with ethyl acetate in the presence of sodium ethoxide as the condensing agent. Later Claisen (12) ‘prepared acetylacetone by the same reaction using sodium as the condensing agent. Sprague, Beckham, and Adkins (64) have given a summary of the various methods of synthesizing 1,3 diketones. It was not until 1901 that a metal compound was actually thought to contain a ring structure. Werner (79) prepared potassium dichloro 2,#—pentanediono platinate (II) and proposed a structure containing two covalent oxygen to metal bonds. The aluminum, manganese (III), cobalt (III), chromium (III), and platinum (II) salts of acetylacetone were also prepared by Warner. These compounds were said -7- to be soluble in chloroform, benzene, and other organic solvents. The platinum (II) compound was analyzed and found to contain 30.08 per cent 0, 4.31 per cent H, 49.57 per cent Pt as compared to the theoretical values of 30.5, 30.58, and 49.62 respectively. Werner found the molecular weight of this compound by ebullioscopic methods to be 403. Since the theoretical molecular weight for the monomer is 393, he concluded that the platinum complex was monomeric. Ley (36) advanced the work of Werner and cited as evidence for the ring structure several properties of chelate compounds. Among these properties were the absence of (a) the normal reactivity of the metal ion bound in the complex, (b) the difference in the reactivity towards metals of alpha, beta, and gamma amino acids, and (c) the low electrical conductivity of solutions of these compounds. Thorium acetylacetonate was first prepared by Urbain in 1896 (74). Thorium was precipitated as the hydrous oxide with dilute ammonia, washed with water by decantation, and made into an alcohol slurry to which acetylacetone was added. The mixture was heated, and the hydrous oxide underwent a conversion to the acetylacetonate which was recovered by cooling the reaction mixture in ice. Urbain reported a melting point of 171-1720 C. and analysis gave 37.09 per cent Th, 37.75 per cent 6, ."h.‘54per cent H, and 20.62 per cent 0 as compared to the theoretical values of 36.98, 38.19, 4.45, and 20.38 respectively. Cryoscopic measurements -8.. on this compound using ethylene dibromide-as the solvent gave a value of 630 for the molecular weight. The calculated molecular weight is 628, and Urbain accordingly concluded that the compound did not dimerize. The substance was found to be soluble in organic solvents such as chloroform and benzene, but less soluble in ether and ethylene dibromidc. urbain also reported that thorium acetylacetonate sublimes at low pressures. A In 1897 Urbain and Budischovsky (76) attempted to separate the rare earths by the fractional dissolution of the acetylacetonates. The solvents used were alcohol and benzene. Mixtures of rare earths were used and separations were followed by average atomic weight determinations. It was discovered that the rare earths of higher atomic weights were dissolved first and concentrated in the mother liquor. The conclusion was that, although there was some fractiona- tion, this method of separation was inferior to other methods available at that time. During the course of their work, Urbain and Budischovsky noted positive temperature co- efficients of solubility for these compounds. Later Urbain and Debierne (77) showed that the acetylacetonates of iron (III), manganese (III), cobalt (III), chromium (III), aluminum, and nickel (II) were monomers. They noted that iron (III) acetylacetonate was very soluble in chloroform, benzene, alcohol, acetone, and ethYI acetate; less soluble in ether and essence of turpentine; poorly soluble in water. -9- Biltz in 1904 (8) prepared acetylacetonates of lanthanum, cerium, praseodymium, neodymium, samarium, and didymium. These compounds were prepared by mixing rare earth nitrate solutions with ammoniacal acetylacetone in an aqueous medium. Lanthanum was found to form a dihydrate, La(CSH702)3.2H20. The percentage of lanthanum found was 29.2 as compared to the theoretical of 29.3. This substance became anhydrous after recrystallization from alcohol and the anhydrous material was reported to melt at 185° C. Cerium (III) was found to form a trihydrate, Ce(05H702)3.3H20, and after air drying the analyses gave values of 29.16 and 29.24 per cent cerium as compared to the theoretical value of 28.15. The melting point was reported as 131-1320 C. The compound was easily soluble in alcohol and difficultly soluble in ether. Biltz also determined molecular weights by ebullioscopic methods. These data given by Biltz are listed in Tables I and II. -10- TABLE I Solubility Observations, Melting Points, and Analyses of Rare Earth Acetylacetonates Compound % Rare Earth % Rare Earth Melting Found Theoretical Point 00. prtcss7oz)3 31.98 32.0u 146° Nd(CsH702)3 31.8 32.6 144-460 sm(C5H702)3 146-470 Pr(CSH702)3-Soluble in carbon disulphide, ether, chloroform, benzene, and alcohol. Sm(CSH702)3-Easily soluble in alcohol, soluble in carbon tetrachloride slowly soluble in carbon disulphide, (These data are from Biltz.) TABLE II Molecular Weight Determinations Compound Mol. Wt. M01. Wt. Solvent Dimer Found Di(C5H702)3 878 33% carbon tetrachloride 853 864 Pr(CSH702)3 874 820 carbon disulphide NdR(CSI-i702)3 882 868 diethyl sulphide Sm(CSH702)3 894 831 carbon disulphide Di is a mixture of the_1ighter rare earths. -11- The various values found for the molecular weight of didymium acetylacetonate are due to varying concentrations, the smaller values resulting from the more dilute solutions. From these measurements Biltz concluded that, although an equilibrium must exist between the monomer and the dimer, the rare earth acetylacetonates are primarily dimeric in solution. Biltz also found that, upon heating, solvolysis of these compounds occurred in alcoholic solutions. Ce(C5H702)3 could not be crystallized from alcohol since cerium dioxide was formed. Other investigatbrs who have prepared these compounds have used either the method of Urbain or Biltz, with certain modifications. Jantsch and Meyer (32) prepared Gd(05H702)3-2H20 using Biltz's method and reported Gd(CSH702)3~H20 when recrystal- lized from absolute alcohol. The melting point of the monohydrate was given as 142° C., and this compound was said to become anhydrous when crystallized from dry benzene or dried over phosphorus pentoxide for twenty-four hours. Freed, Weissman, and Fortess (23) studied the absorption spectrum of Eu(CSR702)3. In carbon tetrachloride or benzene solutions, the number of lines in the visible spectrum was greatly increased relative to the normal europium spectrum. The intensities of these lines varied with time and after three weeks an apparent equilibrium was reached. These shifts in intensities were regarded as equilibrium phenomena between the monomeric and the dimeric forms of this compound. -]_2- From the number of possible transitions, Freed gt_gl. postulated the existence of four distinct kinds of triposi- tive europium. The crystals dissolve to give primarily dimer and require several weeks to reach equilibrium. Lines with shifting intensities were assigned to either the monomer or dimer. Two species which are due to the equilibrium are the active and inactive forms of the dimer. Since this leaves two species unaccounted for, Freed gt_gl. postulated that the monomer must have two isomeric forms which would be a kind of isomerism unsuspected to that time. Quill, Stites, and McCarty (51) studied extensively the preparations of rare earth acetylacetonates. They found that the highest yields may be obtained using rare earth chloride solutions rather than nitrates. To the chloride solution an ammoniacal solution of acetylacetone is added and the pH is maintained at a value Just below the pH necessary for the precipitation of the rare earth hydroxide. Wold (80), in further studies on preparations, reports still higher yields if the rare earth chloride solution is added to the acetylacetone solution maintained at the pH described by Quill gt_§1. Wold also prepared the propionylacetonates of lanthanum, praseodymium, neodymium, and samarium, These compounds on air drying were monohydrated but could not be dried by heating or desiccation without decomposition. Analyses gave theoretical values for the percentage of rare earth present -13- in the monohydrates. Wold repeated the ebullioscopic measurements of Biltz and obtained comparable results. Data on melting points and hydration of rare earth acetylacetonates have previously been mentioned and are summarized in Table III. Scandium and yttrium are included because their chemistry is similar to that of the rare earths. -]_Zi- TABLE III Reported Hydrates and Melting Points of Rare Earth Acetylacetonates Compound Reported Melting Reference Point La(CSH702)3 150—510 C. 77 151 31 185 7 142-43 63 188-90 63 La(05H702)3-2H20 7 158-60 77 C6(C5H702)3 144-45 66 131-32 7 165‘ 73 ( ) 145 31 CS C H 02 4°11 H O 32 C8(C§H;02)3-3H202 1&5 7 138-39 63 Nd(C5H702)3 150-52 38 144-46 7, 31, 66 143-44 63 Sm(C H o ) 146-47 7. 31. 66 5 7 2 3 143-44 63 Eu(csH702)3 144-45 63 Eu(05H702)3-3320 136-37 58 Gd(C5H702)3 143-44 63 Gd‘05H702)3°3320 58 Gd(05H702)3-2H20 31 Gd(CsH702)3'H20 142 31 1(0 H o ) 131 31 5 7 2 3 114 38 129-30 63 Sc(CSH702)3 177-8? 46 187-87.5 46 -15- It can easily be seen that these data are not con- sistent. The large variations are probably caused by varying degrees of hydration of which the investigators were unaware. No quantitative data on the solubility of acetylace- tonates appeared in the literature until the work of Hatch and Sutherland in 1948 (28). These authors prepared the acetylacetonates of sodium, potassium, magnesium, beryllium, and aluminum. The melting points, ultraviolet absorption spectra, densities, and solubilities in benzene, n-hexans, and cyclohexane were determined. For the solubility work Hatch and Sutherland used the cloud point method first introduced by Aleerew (1). Miller and Dittmar (41) determined the solubility of urea in water using this technique and found that their results compared very well with results reported in the literature. Hoerr £3 51, (29) have measured the solubility of aliphatic nitriles in organic solvents, and Rhodes and Eisenhauer (56) used cloud points for determining the solubility of napthalene in aromatic hydrocarbons. A good review of the method is given by Zimmerman (81). Some crystal structure work has been done on the acetylacetonates of trivalent metals other than the rare earths. Morgan and Drew (46) examined microscopically the acetylacetonates of aluminum, gallium, indium, scandium, and iron. Astbury (4) did x-ray work on scandium, cobalt, and chromium acetylacetonates in addition to those compounds -]_6- studied by Morgan and Drew. The results of these investi- gators agreed very well. An isotrimorphism was found to exist among these compounds. Astbury lists crystal class, crystal structure, and densities. He assumed octahedral symmetry within the molecule and observed an overall dimension of about 8.2 A. for these molecules. Solvent extraction and the use of non-aqueous solvents for separation of the lanthanons has received much attention. Hopkins and Quill (30) investigated the solubilities of rare earth chlorides in a number of organic solvents and achieved a partial separation of neodymium and praseodymium utilizing solubility differences. Fischer, Dietz, and Jubermann (22) suggested extraction 'of rare earth chlorides with ethers, alcohols, and ketones. They claimed the existence of 50 per cent extractibility differences between adjacent rare earths, but none of the actual data were published. It is probable that duplication of these results was not possible. Appleton and Selwood (2) found rare earth thiocyanates to be soluble in n-butyl alcohol and obtained a partial separation of lanthanum and neodymium by extracting aqueous thiocyanates. The enrichment factor for neodymium was only 1.06, but they expressed hope for the possibility of a con- tinuous extraction procedure. Templeton (71) worked with rare earth nitrates utilizing extractions with higher alcohols or ketones. The extractibility of the rare earths -17- was found to increase with atomic number. Templeton and Peterson (72) studied extractions of rare earth nitrates, and in the separation of lanthanum and neodymium by extraction with n-hexyl alcohol found the enrichment in neodymium to be about 1.50 for a single stage operation. However, a three stage extraction gave an enrichment of 2.14. ' Suttle (69, 70) has studied the chelation of lanthanum and cerium by thenoyltrifluoroacetone using extraction techniques. Lanthanum and cerium in tracer quantities were extracted with benzene solutions of the chelating agent from aqueous solutions at various pHs. Keenan and Suttle (34) made similar studies on praseodymium. Broidio (10) found that scandium may be separated from the rare earths in a single extraction with a benzene solution of thenoyltri- fluoroacetone at a pH of 1.5. Moeller and Jackson (44) determined extraction curves for neodymium and europium by extractions of aqueous solutions at varying pHs with chloroform solutions of 5, 7-dichloro-8-quinolinol. Recently Peppard g£_gl. (49) developed what appears to be a very promising extraction method for the separation of the lanthanons. The aqueous phase consists of rare earth nitrates dissolved in concentrated nitric acid. The organic phase is tri-n-butyl phosphate in the pure form or diluted with an inert hydrocarbon. Extractibility was found to increase with atomic number, but this could be reversed in -]_8- dilute (0.3 N) nitric acid. Scandium could be separated almost completely from yttrium and the rare earths in a single operation. Peppard gt_§1, state that it is theo- retically possible and economically feasible to separate individual rare earths in high states of purity. Weaver, Kapplemann, and Topp (78) utilized this technique to isolate a kilogram of gadolinium, 95 to 98 per cent pure. Varsol, a commercial organic petroleum fraction, and tri-n-butyl phosphate, 40 and 60 per cent respectively, composed the organic phase. In a three stage process the separation factor between gadolinium and samarium was about nineteen and between gadolinium and terbium about seven. The authors state that optimum conditions were not obtained and that better separations would be possible if better techniques were used. Peppard.§t_§1. (48) studied the extraction of thorium from pitchblends with tri-n—butyl phosphate, and Peppard, Gray, and Markus (50) examined the behavior of the actinides using the same reagents. In a very recent paper Schweitzer and Scott (59) determined the extraction of yttrium into acetylacetone from aqueous solutions at various pHs. Steinbach and Freiser (66, 67) made similar studies for various other metals, but no work of this type has been reported for the rare earths. Little has been said concerning the volatility of rare earth acetylacetonates. Ephraim altered his opinion -19- as has already been pointed out. Marsh (39) reported that lanthanum acetylacetonate was slightly volatile but gave no further details. Marsh drew his conclusions from analytical work to determine the percentage lanthanum by a direct ignition to the oxide. It is likely that Marsh lost a small amount of his sample mechanically by too fast a temperature rise during ignition. No other work has been reported. A large number of papers concerning the absorption spectra of acetylacetonates may be found in the literature. Several groups of investigators (6, 9, 60) have determined the spectrum of acetylacetone and some (3, 24, 55) discuss the effects of keto-enol tautomerism and shifts due to hydrogen bonding. Others (18, 35, 45) have examined the infrared spectra of several metal acetylacetonates and have found that the absorption bands for each are very similar. Haszeldine §£,§1. (27) determined the spectrum of mono, tri, and hexa fluorinated acetylacetonates of copper, uranium, and iron. Radoichic (52) studied the effect of forty solvents on the spectrum of neodymium acetylacetonate and concluded that variations were irregular. Radoichic also (53, 54) found that the spectra of neodymium and samarium acetylacetonates are considerably different in various solvents, but these variations were .independent of dipole moment and dielectric constant of the solvent. Howe and Herbert (31) determined the -20- absorption spectra of praseodymium and neodymium acetylace- tonates at 120° K. Wold (80) recorded the infrared absorption of cerium (IV) acetylacetonate. Spectra of these compounds will not be mentioned at length here, but will be considered in a later portion of this dissertation. With the above background it is possible to proceed to the experimental portion of this work. -21- EXPERIMENTAL AVERAGE ATOMIC WEIGHT DETERMINATIONS Quantitative work on the rare earths often necessitates average atomic weight determinations of the particular samples of rare earths being used. It is not possible to do this by a simple ignition of the oxalate because rare earth oxalates do not easily become anhydrous. The method utilized and known as the oxalate to oxide ratio method (21) is based on the following hypothetical equation. R2(0204)3°XH20 $3.203 + 30203 + XH20 The average atomic weight is then calculated as follows: wt. oxide obtained wt. oxalate ignited x wt. oxalate titrated x30203 2R+3 0 == N of KMnOu x ml. KMnOu x meq. wt. of 0203 PROCEDURE Rare earth oxide is dissolved in dilute nitric acid, and the solution is filtered to remove any insoluble material. The rare earth is precipitated as the oxalate with a solution of oxalic acid, and the precipitate is washed by decantation with hot water. The oxalate is then filtered and dried at 1100 C. for eight to twelve hours. The oxalate does not become anhydrous, but it does become constant in composition. Duplicate samples are weighed into tared crucibles and ignited to the oxide at 9000 C. in a muffle. Two other samples of oxalate are dissolved -22- in 10 N sulfuric acid, the solutions are diluted with distilled water and titrated with standard potassium permanganate. The data from average atomic determinations are listed in Table IV. TABLE IV Average Atomic Weight Determinations Rare Ht. Wt. Wt. M . N Average Earth Oxalate Oxide Oxalate Khnou KMnOu Atomic Ignited Obtained Titrated Weight Nd 0.2584 0.1442 0.3009 27.17 0.1099 144.5 .2522 .1408 .2205 19.93 .1099 144.3 Eu .0962 .0534 .1071 9.98 .1017 151.6 -.1250 .0692 .1802 16.68 .1017 152.7 La .2569 .1416 .2172 21.65 .1017 139.3 .1853 .1021 .1909 19.10 .1017 138.6 This method gave satisfactory results, but difficulty was encountered in weighing the dried oxalate previous to ignition. This partially dried oxalate tends to pick up water very quickly, hence it must be weighed by difference from a capped weighing bottle which is open for as short a time as possible. PURITY OF RARE EARTHS The rare earths used for this research were presented by Dr. Quill. The average atomic weight of the lanthanum used was 138.9. Ignition of lanthanum oxalate samples gave a perfectly white oxide. The presence of small amounts of praseodymium would cause the oxide to be dark colored. Tests for cerium were negative, and the absorption bands -23- of other rare earths were not observed in the lanthanum solutions. The neodymium had an average atomic weight of 144.6. The freshly ignited oxide was very pale blue in color, and solutions of neodymium did not show absorption bands of praseodymium. Tests for cerium were negative. The average atomic weight of the europium used was 152.2. Solutions of europium did not show the absorption bands of other rare earths, and the freshly ignited oxide was pure white. NEODYMIUM ACETATE The usefulness of neodymium acetate for solvent extraction had never been explored. This compound was prepared and solubilities in various solvents were determined. Preparation and Analysis Neodymium acetate was prepared by dissolving the oxide in acetic acid, filtering to remove any insoluble material, and then slowly evaporating until crystals appeared. The compound could be recrystallized from water by slow evaporation at 60° C. The percentage neodymium was determined by direct ignition of the acetate to the oxide at 950° C. The results of these analyses are given in Table V. -214»... TABLE V Neodymium in Neodymium Acetate Acetate Oxide Neodymium Ignited Obtained Found g. g. ’4 0.2318 0.1143 42.5 .1855 .0919 42.5 The theoretical compositions of neodymium acetate and its hydrates are Formula % Neodymium Nd(02H302)3 45.0 Nd(C2H302)3.H20 42.6 Nd(C2H302)3.2H20 140.4 It was concluded that the product of this preparation is a monohydrate, which is in agreement with the work of Thomas (73), who used the same method of preparation. Microscopic examination revealed small, homogeneous, needle-like crystals. When dissolved in glacial acetic acid acetates behave as bases and may be titrated as such. One method for determining sodium acetate involves a perchloric acid titration in glacial acetic acid (5). The acetate in Nd(CZH302)3~H20 could theoretically be determined by this technique. Data for the standardization of the perchloric acid are given in Table VI. -25- TABLE VI Standardization of Perchloric Acid in Glacial Acetic Acid Potassium Acid Perchloric Calculated Phthalate g. Acid m1. Normality 0.7260 34.97 0.1017 .5913 28.43 .1019 .8853 42.64 .1017 Crystal violet was used as the indicator, and no difficulty was encountered in observing the end point for these sodium acetate perchloric acid titrations. Attempts were made to titrate neodymium acetate in the same manner. The end point color change was so gradual that no definite end point could be detected. Acetic anhydride was added to remove the water, which was present from the glacial acetic acid and the hydrated salt, but no improve- ment was noted. A potentiometric titration was attempted using a quinhydrone—silver chloride electrode system. A Cenco titration pH meter was used to measure the emf., and the solution was stirred manually. These data are listed in Table VII. -26.. TABLE VII Potentiometric Titration of Nd(CZH302)3-H20 with H0104 in Glacial Acetic Acid Sample H0104 emf Sample HGlOg emf a. 0.1017 N V} g. 0.1017 N v. m1. ml. 0.2636 0.0 0.228 0.2688 6.00 0.207 1.00 .261 7.00 .208 2.20 .263 7.50 .209 2.78 .267 8.00 .211 4.00 .269 8.50 .214 4.94 .272 8.80 .215 6.00 .275 9.00 .216 7.00 .279 9.50 .220 15.70 .301 10.10 .221 10.50 .222 0.2688 0.0 0.212 11.00 .230 1.00 .216 15.00 .237 2.00 .208 17.00' .243 3.00 .210 20.00 .260 4.00 .207 25.00 .320 5.00 .207 An examination of these data and a plot of these points reveals that no clear cut end point may be recognized, hence with the electrode system used potentiometric measurements will not detect the end point. A high frequency titration was utilized in an attempt to accurately determine the end point. An instrument of the Hall type was used (25). These data are listed in Table VIII. -27- TABLE VIII High Frequency Titration of 0.2941 g. Nd(CZH302)3.H20 with 0.1017 N H010“ in Glacial Acetic Acid H010“ Plate H0104 Plate m1. Voltage ml. Voltage 0.00 2.82 10.00 0.33 1.00 1.37 11.00 .35 2.00 0.97 12.00 .38 3.00 .67 13.00 .40 4.00 .55 14.00 .42 5.00 .43 15.00 .43 6.00 .37 17.00 .48 7.00 .36 19.00 .52 8.00 .34 24.00 .57 9.00 .33 These data are given in graphical form in Figure l. FIGURE | HIGH FREQUENCY TITRATIONOF Nd(C2H302)3H20 WITH 01017 N H0104 IN GLACIAL ACETIC ACID PLATE VOLTAGE 3.0 2.6 2.2 |.8 |.4 0.6 0.2 y— -29- The end point should come after the addition of 25.5 ml. of acid. Although there is no sharp break in the curve, the midpoint of the dip comes at about one half of the theoretical value. To determine whether neodymium acetate is completely ionized, a freezing point depression was run on this compound. Water was used as the solvent. The standard technique using a Beckmann thermometer and a manually Operated loop stirrer was used. These data are listed in Table IX. TABLE IX Freezing Point Depression of water by Nd(CzH302)3 g. Nd(CgH302)3-H20 1.38 g. anhydrous Nd(CZH302)3 1.31 8. water 49.9 Freezing reading for water 3.680o Freezing reading for solution 3.300o Freezing point lowering 0.380o From these data the apparent percentage ionization may be calculated. The value found is 50 per cent. This value for the per cent ionization is not to be unexpected for rare earth salts. Dutt (19) determined the dissociation of rare earth thiosulfstes and found it to be of the same order of magnitude. This partial dissociation is probably the reason why the titrations attempted did not work. If this compound is only 50 per cent ionized in water, the basicity probably would not be high even in glacial acetic acid. Perhaps if more time was allowed between additions of perchloric acid -30- the end point would be more sharp. The other method for determining acetate in neodymium acetate which seemed feasible was to dissolve the sample in syrupy phosphoric acid and distill the acetate off as acetic acid, which.then could be titrated. Thomas (73) reported very poor results with this technique so it was not tried. No good method of determining acetate in the presence of neodymium was found. An attempt was made to dry Nd(CzH3021§H20 by placing it in an oven at 110° C. After twelve hours the weight loss amounted to 5.98 per cent and after 36 hours, 6.22 per cent. Since the monohydrate contains 5.30 per cent water, it appears that the neodymium acetate decomposes at this temperature. Solubility Determinations The solubility of Nd(02H302)3oH20 in a number of organic solvents was determined in a qualitative manner. Samples of the compound were placed in test tubes and shaken for a few minutes with a number of organic solvents. The solutions were allowed to settle overnight, and were then viewed with a hand spectroscope. Since the neodymium absorption bands for the various solvents were not observed, it was concluded that Nd(CZH302)3-H20 is not soluble to any appreciable extent in any of the solvents tested.- Those solvents used are as follows: absolute methanol, absolute ethanol, n-hexl "-' ‘ -31- alcohol, chloroform, chlorobenzene, isopropyl ether, xylene, n-heptane, nitrobenzene, bromobenzene, n-hexane, acetophenone, n-amyl alcohol, and bromoethane. For the solvents listed below solutions were prepared by shaking the solvent and Nd(CzH302)3-Hzo for thirty—six hours on a mechanical shaker. The optical density was determined with a Beckmann Model DU Spectrophotometer. The wave length was kept constant at 5220 A., and the solvent being used was the reference for each reading. Data are listed in Table VII. TABLE X Absorption of Various Nd(CzH302)3.H20 Solutions at 5220 A. Solvent % T. D. g.Nd./1. M of Nd of Soln. methanol 92.0 0.032 1.2 8.0 x 10-3 hexone 98.5 .007 0.27 1.9 x 10-3 chloroform 98.1 .008 .31 2.1 x 10-3 nitrobenzene 99.3 .003 .12 8.3 x 10‘“ heptane ‘ A 98.0 .008 .31 2.1 x 10-3 ethylene dichloride 99.0 .000 .15 1.0 x 10-3 acetophenone 100.0 .000 .00 0.0 ethyl acetate 98.0 .008 .31 2.1 x 10-3 n-hexyl alcohol 96.7 .015 .57 3.9 x 10-3 methyl n-amyl ketone 97.8 .010 .38 2.6 x 10"3 hexane 99.1 .004 .15 1.0 x 10"3 isopropyl ether 100.0 .000 .00 0.0 aeetonitrile 95.8 .018 .69 “.8 x 10"3 chlorobenzene 99.7 .002 .08 5.3 x 10‘: n-amyl alcohol 99.3 .003 .12 8.3 x 10- n-butanol 98.7 .007 .27 1.9 x 10-3 Rodden (57), and Moeller and Brantley (#3) outlined absorption spectra methods for quantitatively estimating the rare earth elements. The peaks and absorption coefficients listed by Moeller and Brantley are -32- A. k a NdCl3 5218 0.0300 Nd(0104)3 5218 .0298 Nd(N03)3 5218 .0269 Nd(CzH302)3 5230 .0263, where k : log IQZI , and Nd+3 in g./. a 2, c is the concentra- cl k tion of the absorbing species, and 1 is the cell width. One centimeter, matched cells were used in this work. It may be observed that the peak for Nd(02H302)3 does not have the same value as the other salts, but the absorption coefficient, k , is‘of the same order of magnitude. Since this work was done in non-aqueous solvents, the peak may be shifted slightly and the k value may differ slightly. The calculations for the above table are based on the above equations given by Moeller and Brantley and the k value is assumed to be 0.026. These numbers are intended to give only the order of magnitude of the solubility. It may be concluded from these data that Nd(023302)3.H20 is not very soluble in the solvents which were investigated. The solubilities are so low that it would not be practical to use this compound for extraction work with any of the above solvents. -33- RARE EARTH ACETYLACETONATES Many chelate compounds are known to show appreciable solubility in organic solvents; rare earth chelates have been prepared and are reported to show similar solubilities. Although the rare earths are chelated by a number of different reagents, only the 2,h—pentanedione derivatives may be easily isolated in the pure form. Because of these facts the acetylacetonates were chosen as the compounds to be investigated. Purity of Acetylacetone Acetylacetone may be purchased from several sources. It has been said that acetic acid in considerable quantities is an impurity in certain acetylacetones. A check for the presence of acetic acid was made by titrating with sodium hydroxide. Since acetic acid is a strong acid relative to acetylacetone, it should give an extra break in the curve if it was present. Two different commercial samples were titrated. TABLE XI Titration of Eastman Acetylacetone with 1N NaOH '6 m UNHHOWmmVflwm-C'NOV-PQW? NtWOUWfiOVHI-‘NNWHVH'QQV mmmmmflflflflflflfiflfifl O\U\ F-P'U m1. ’6 :11 O “FADE-‘00“) ocx‘lxlmmmmmu Hoocooxoxooovr-onI-onooxo :6? \O\O\O\O\O\O\O ooooooocoooooooooooooo m1. pH IO FIGURE 2 TITRATION OF EASTMAN 05H.02 WITH IN Nc OH l J l i 0L. 4 8 l2 I6 m2 No 0H -35- It is seen from Figure 2 that there is only a very slight suggestion of a plateau in the region expected for acetic acid. Paragon acetylacetone showed a definite break in the curve at a pH between four and six, hence the Eastman product seemed to be purer. For the quantitative determination of acetylacetone the method developed by Smith and Duke was employed (62). Cerium (IV) quantitatively oxidizes acetylacetone in a u M perchloric acid solution. The equation for this reaction is O5H802 + 6Ce+u +3H20~——9 HCOOH + 20H3000H + 6Ce+3 + 6H+ from which it is noted that six equivalents of cerium (IV) are required per mole of acetylacetone. The procedure of Smith and Duke consists of adding a 25 to #0 per cent excess of cerium (IV) reagent to acetylace- tone in a b M perchloric acid solution. A ten minute interval is allowed for the oxidation. The excess of cerium (IV) reagent is determined by a back titration with standard sodium oxalate solution using nitro-ferroin indicator. The cerium (IV) reagent is h M in perchloric acid, the sample to be analyzed is dissolved in b M perchloric acid and the sodium oxalate solution is 0.1 M in perchloric acid. For this work a 0.“ M solution of cerium (IV) reagent was prepared according to the procedure given by Smith (61). Ceric ammonium nitrate (c. F. Smith) was the starting -37- material. The procedure given by Smith is for the prepara- tion of a 0.1 N solution, and when the more concentrated solution was prepared the separation of a voluminous white solid was observed. As more perchloric acid was added, the mass of this material increased. It was finally reasoned that this solid was slightly soluble ammonium perchlorate, which was filtered out of the solution. Care was taken not to allow it to become dry. It was dissolved in water and the solution discarded. Eastman acetylacetone was weighed in glass ampules which were broken under the surface of the cerium (IV) reagent. The results of these analyses are given in Table XII. TABLE XII Analyses of Eastman Acetylacetone mg Sample m1. Ce(IV) ‘ml. Sodium Oxlate mg Acetylacetone Found 17u.7 28.ooé.hh78 N) 11.60é.1778 N 17u.7 129.8 25.00 .uu78 N) 19.u8 .1778 N 129.0 61.3 50.00(.1158 N) 12.o7(.23u6 N) 61.2 It may be seen that the Eastman acetylacetone is very pure, and this acetylacetone was used exclusively throughout the course of this work. Preparation and Analysis Two generally different methods are available for the preparation of rare earth acetylacetonates. One is the method of Urbain (7h) who prepared such compounds by allowing -38- acetylacetone to react with the hydrous oxide of the metal. Another is the method of Biltz (8) which involves mixing solutions of the metal nitrates and ammoniacal acetylacetone. The latter method has been thoroughly studied and revised (51). The method of Urbain does not account for the possi- bility of an incomplete reaction, and the result could be a product contaminated with hydroxide, which would lead to high percentages of rare earth when analyzed.. Marsh prepared such compounds using this method, and after analysis concluded that the products were basic (39). Analytical data given in the introduction also seem to verify this observation. The procedure of Biltz involves only one solid, which is the product. If the pH of the mixture during preparation becomes too high, the product could still become basic. Quill (51) g§_gl, prepared acetylacetonates using solutions of rare earth chlorides and ammoniacal acetylacetone with carefully controlled pH to avoid the formation of basic substances. The latter method was used for all acetylace— tonate preparations in this work, and a typical preparation may be described as follows: Five grams of Nd203 was dissolved in concentrated HCl and evaporated to dryness on a steam bath. The resulting chloride was dissolved in water and the solution was filtered to remove any insoluble material. The pH was adjusted to a value of 5.0 to 5.5 as measured by the glass electrode. Ten mls. of concentrated ammonia was added to 15 mls. of -39... acetylacetone. Enough water was added to dissolve the ammonium acetylacetonate, and this solution was added slowly by means of a dropping funnel to the rare earth chloride solution. The pH was carefully observed and never allowed to exceed a value of 6.5. After all the reagent was added, the pH was adjusted to 6.5 with dilute ammonia. The mixture was stirred for five hours, the product was filtered, washed with distilled water, and air dried. The product of such a preparation is a very finely divided crystalline material which has the consistency of talcum powder. It is not easily wet by water. The acetylacetonates of lanthanum and europium were prepared in a similar manner, and the maximum pH values during preparation were 7.8 and 6.5, respectively. For all of these preparations a 50 per cent excess of acetylacetone was used. During the course of many such preparations it was discovered that large crystals of rare earth acetylacetonates resulted if the preparative procedure was altered slightly. The entire amount of ammoniacal acetylacetone is added very slowly and the pH is kept low enough (5.0-5.5) so that no precipitate forms. 0n standing for several hours needle-like crystals of the acetylacetonate appear. These crystals grow slowly and for maximum yields the mixture should be allowed to stand overnight. The formation of such crystals could be duplicated with ease for lanthanum, neodymium, and europium, No seeding was necessary. -110.- The percentage of rare earth in these acetylacetonates could theoretically be obtained by a direct ignition to the oxide at 950° C. This temperature insures the complete decomposition of any rare earth carbonates which might be formed during ignition. To verify this, two duplicate sets of samples of neodymium acetylacetonate were weighed; one set of these samples was ignited, and the other was dissolved. in dilute acid, the neodymium precipitated as the oxalate, the solution was filtered, and the oxalate was ignited to the oxide. The results are given below. TABLE XIII Analysis of Nd(05H702)3 for Nd by Two Methods Direct Ignition Grams Sample Grams Oxide % Nd 0.1h84 0.050” 29.1 0.2322 0.0791 29.2 Oxalate Precipitation 0.19h6 0.0665 29.3 0.1730 0.0589 29.2 0.1681 0.0573 29.2 The results from the two methods are nearly identical, hence the rare earth in these compounds may be determined by direct ignition. It also may be concluded that under the conditions in a muffle up to 950° 0., neodymium acetylacetonate is not volatile. Acetylacetone in these compounds was determined by the method of Smith and Duke (62) described previously. The rare earth acetylacetonate was dissolved in h M -41.. perchloric acid to liberate the acetylacetone, which was titrated as before. Data are compiled in Table XIV. TABLE XIV Analysis of Neodymium Acetylacetonate for Acetylacetone Sample g. Cerium Reagent Sodium Oxalate % Acetylacetone 0.uu78 N, mls. 0.1778 N, mls. 0.0803 10.00 8.90 60.1 0.1027 10.00 17.00 60.0 The theoretical compositions of the various hydrates of Nd(°5H702)3 are Compound, % Nd %.Acetylacetone % H20 Nd(csH702)3 32.7 67.3 0.0 Nd(CSH702)3-Hzo 31.5 - 6u.6 9.0 Nd(05H702)3.2H20 30.3 62.8 7.5 It may be readily observed from these data and the data previously given that the neodymium acetylacetonate is not“ anhydrous. A test for water was made by titration with the Karl Fischer reagent. Values of 10.0 and 10.7 per cent water were obtained, but the end point seemed to be very vague and tended to fade rapidly. Precision was of the order indicated by the two values given. A sample of Nd(CSH702)3oxH20 was sent to the Clark Mioroanalytical Laboratory for carbon and hydrogen analysis. The results were as follows: -22- Found Theoretical for Nd(CSH702)3.3H20 1 C % H % c % H 36.47 5.29 36.32 5.99 36.h6 5.22 The calculated carbon to neodymium atom ratio is 15 to one, and from these data it is exactly that. The hydrogen to neodymium ratio for the trihydrate should be 27 to one, and from these data it is 26 to one. It was concluded that the product of these preparations is a trihydrate. In Table XV it is shown that europium forms a similar trihydrate and lanthanum a dihydrate. TABLE XV Analysis of Eu and La Acetylacetonate .rydrates Compound % Rare Earth % Acetylacetone Found Theo. Found Theo. Eu(C H O ) o3H O 30.3 30.23 59.1 59-03 5 7 2 3 2 30.1 59.2 30.2 La(CSH702)3o2H20 29.n 29.93 62.2 62.93 29.5 62.2 29.7 After sixty days standing in a loosely capped bottle, Nd(CsH702)3-3H20 gave the same analysis as it did Just after preparation. Hence, the hydrates are air stable for fairly long periods of time. -h3- Drying Studies Since these compounds were shown to be hydrated, several attempts were made to remove the water. Analytical data given in the introduction for compounds prepared by the method of Biltz tend to indicate that all of the water of hydration was not removed by the particular drying techniques used by the various investigators. The most straight forward technique is to dry the material in an oven. This was tried at two different temperatures, and the results are tabulated below. TABLE XVI Oven Drying of Nd(C5H702)3-3H20 Nd(C5H702)3-3H20 Temp. Time Weight LOSS WeightgLoss 8o g. 0.6279 110°C 3 hrs. 0.1236 19.7 0.8080 80° 3 .0651 8.05 9 .0686 8.48 23 .0730 9.03 #8 .0815 10.1 72 .0982 12.1 lhh .1225 15.2 Theoretical 10.9 In three hours at 11000., the weight loss was more than the theoretical value. After seventy-two hours at80°C., the weight loss exceeded the theoretical. These data indicate decomposition at these two temperatures and it is apparent that these compounds may not be dried by heating in air. -44- Nd(05H702)3-3H20 was dried in a desiccator over four mesh, anhydrous calcium chloride, and over magnesium perchlorate. Both large crystals and finely divided samples were studied. The data are presented in tabular and graphi- cal forms as follows: TABLE XVII Drying of Nd(05H702)3-3H20 over 08.012 Large Crystals Sample Time Weight % Weight g. hrs. Loss Loss 0.2386 0 0.00 0.0 5 .0010 .42 28 .0089 3.73 54 .0152 6.37 76 .0173 7.25 108 .0181 7.59 144 .0184 7.71 175 .0184 7.71 268 .0184 7.71 460 .0185 7.75 580 .0185 7.75 604 .0185 7.75 Finely Divided Crystals 0.1165 0 0.0 0.0 5 .0022 1.89 28 .0096 8.24 54 .0123 10.6 76 .0123 10.6 108 .0126 10.8 144 .0129 11.1 175 .0129 11.1 268 .0129 11.1 460 .0131 11.2 580 .0133 11.4 604 .0133 11.4 -45- TABLE XVIII Drying of Nd(05H702)3'3H20 over Mg(Cth)2 Large Crystals Sample Time Weight 5 Weight 3. hrs. Loss Loss 0.1122 0.0 0.00 0.00 5 .0006 .54 28 ‘ .0054 4.8 54 .0077 6.9 76 .0084 7.5 108 .0085 7.6 144 .0087 7.8 175 .0086 7.8 268 .0085 7.6 460 .0088 7.8 580 .0087 7.8 7.8 604 .0088 Finely Divided Crystals 0.1030 0.0 0.0 0.0 5 .0021 2.0 28 .0099 9.6 54 .0105 10.2 76 .0107 10.4 108 .0108 10.5 144 .0111 10.8 175 .0110 10.8 268 .0111 10.8 460 .0115 11.2 580 .0115 11 2 604 .0115 11:2 Om. «mac: 2. ml; 00. 03 ON. OO. 00 q - i cm 0.? ON d _ q 4 d - madam-E..- omoS-o Suzi o «.60 .8- 6: 5:6 corms-scar.- » mane-u m4mo mac...’ 6 3 v2 .10 02:10 o. N. 8801 .LH9I3M X ago-- 2- m2: CO. CO. OV. ON. OO. 00 OO O¢ ON O 1 W a m 4 J a . a 1 ‘1?) |||l‘)| I)? lb) lb) m4cu Own—.20 >4uz: O 3325 32.3 a «do so 55 oaxm.n.~o-:so- oz no oz-Eo t 3.30.... O. N. SSO'I 1H9|3M % -48.. After 144 hours all samples had reached constant weight. There were slight variations after this time, but an examina— tion of the weight loss column in the preceeding tables reveals that these variations are probably due to experimental errors. They are caused by differences of a few tenths of a milligram in weighings, which could be caused by variations in the humidity of the room at the time of weighing. A 10.9 per cent weight loss represents the loss of three molecules of water. Both finely divided samples have come to equilibrium after losing this amount of water. They have become anhydrous. A 7.5 per cent weight loss represents the loss of two molecules of water. Both samples of large crystals have lost this amount of water at equilibrium, and do not become anhydrous, but remain as a material which ' would analyze as a monohydrate. Attempts to dry large amounts of finely divided material by desiccation were not successful. After seven days the material was still not anhydrous. vacuum desiccation over P205 was also tried. The vacuum desiccator was opened twice per day to break the surface of the P205,and the pressure was kept at 10"3 mm. or less. Some Nd(05H702)3o3H20 was placed in the desiccator and after 72 hours analyzed 32.2 per cent neodymium as compared to the theoretical amount for the anhydrous compound, 32.7. It looked as if the water had not been completely removed so more time was allowed. After 48 additional hours of drying the per cent neodymium was 33.0, and after a III 'e<" 90..“ -49- further 24 hours, 33.8. La(05H702)3~2H20 was dried in the same manner. The theoretical per cent lanthanum in La(C5H702) is 31.8, which may be compared with a value of 31.8 per cent lanthanum after 72 hours in the vacuum desiccator. .Twenty-four hours later the per cent La rose to 32.2. It was concluded that it is possible to obtain an anhydrous material in this manner, but it is difficult to know when to cease the drying. Decomposition results if the compound is kept at low pressures for too long a period. This method is not to be recommended for drying rare earth acetylacetonates. Reports in the literature indicate that rare earth acetylacetonates may be recrystallized from alcohol (8). It was thought that the product of such a recrystallization might be anhydrous. Methyl alcohol appeared to be a good solvent and recrystallization was attempted. Neodymium acetylacetonate trihydrate was dissolved in absolute methanol and solvent was evaporated with an infrared lamp until crystals appeared. At no time was the solution allowed to boil. The product was dried for seven days over calcium chloride to remove the alcohol. The neodymium was determined by direct ignition to the oxide and the percentages found were 37.7 and 37.9. These values are higher than the theoretical value (32.7) for the anhydrous compound; it was concluded that hydrolysis or solvolysis occurred. -50- To determine whether'hydrolysis occurred when neodymium acetylacetonate trihydrate was in contact with water, a mixture of the compound and distilled water was shaken for twenty-four hours at 25°C. The mixture was filtered and the solid was air dried. Analysis for neodymium showed that there was no change in composition. The same experi- ment was run with 0.1N sodium hydroxide in place of the water. The neodymium was determined via an oxalate pre- cipitation and ignition to the oxide. The percentages found were 50.0 and 45.6, which are considerably higher than the theoretical value of 32.7 per cent. The product was concluded to be a heterogeneous, basic mixture. Melting Points It has been previously pointed out that considerable confusion exists concerning the melting points of rare earth acetylacetonates. Probable reasons are the use of impure materials and varying states of hydration. Stites (68) found that if the pH was allowed to become too high during preparation, products with inconsistent melting points resulted. Investigators who used Urbain's method of preparation did not obtain pure substances, thus the melting points cannot be considered accurate. The rare earth acetylacetonates do not melt sharply and melting is always accompanied by decomposition. All melting points in this work were determined using capillary -51- tubes and a mineral oil bath which was slowly heated with a micro burner and stirred mechanically. La(C5H702)3-2H20 underwent a slight change at 85-91°C. It appeared to soften slightly, but did not melt. At 147-151°C. another change took place and a gas appeared to be given off. At 180°C. the substance turned brown and it actually melted at l88—190°C. La(05H702)3 softened slightly at 85°C. and melted at l49-151°C. Nd(CSH702)3-3H20 melted with decomposition at l44-146°C. The anhydrous neodymium compound melted at the same tempera- turO. Eu(058702)3.3820 melted with decomposition at 145-14700. Eu(CSH702)3 turned brown above 150°C. but did not melt. At 190°C. the material remaining was black. Melting points could not be observed on a melting point block. Only decomposition could be observed at temperatures far above the melting points determined by the capillary tube method. Solvent Extraction Studies Qualitative solubility determinations disclosed that rare earth acetylactonates are readily soluble in chloroform, hhnce this organic solvent was chosen for these studies. Reagent grade chloroform with no further purification was used. Since Stites (68) has shown that the formation of rare earth acetylacetonates is pH dependent, it was expected that 3&1 I.) -52- pH would be the most important variable in these extractions. In the first attempts at extraction the rare earth was present as the chloride in the aqueous phase. However, difficulty was encountered when the pH of this phase was adjusted. The rare earth precipitated as the hydrous oxide when dilute ammonia was added. This was avoided by having the rare earth initially present in the chloroform phase. The procedure adopted for determing the extractability of neodymium acetylacetonate is described in the following section. A solution of neodymium acetylacetonate in water saturated chloroform was prepared. Chloroform saturated water was adjusted to various pH values with dilute ammonia or hydrochloric acid. Twenty-five milliliters of each phase were placed in a mixing cylinder and the mixture was agitated for one hour on a mechanical shaker. The mixture was allowed to stand in an upright position for one hour to assure complete separation of the phases. Samples of the chloroform phase were removed by placing the finger over the top end of a pipet and quickly immersing it through the aqueous phase into the organic layer, and then pipeting the sample in the normal manner. After removal from the mixture the pipet was wiped with a clean cloth to remove any drops of the aqueous phase which might have adhered. The aliquot of the chloroform phase was placed in a platinum crucible and evaporated to dryness with an infrared lamp. It was -53- then ignited at 950°C. to convert the solid material to neodymium oxide. The oxide was weighed, and since the concentration of neodymium in the initial chloroform solution was known, the percentage neodymium remaining in the organic phase could be calculated. Data gathered in this manner for neodymium acetylacetonate are listed in Table XIX. TABLE XIX The Distribution of Nd(C5H702)3 Between H20 and CHCl3 pH % Rare Earth g. Oxide from . oxide from in Organic Phase 5 mls. CHCl3 Phase’ mls. Initial CHC13 Solution 1.10 0.0 0.0000 0.0216 1.68 0.0 .0000 .0132 2.02 0.0 .0000 .0132 6.18 0.0 .0000 .0132 6.60 3.5 .0008 .0226 7.15 26.1 .0059 .0226 7.16 10.6 .0014 .0132 7.44 58.3 .0126 .0216 7.59 50.9 .0110 .0216 7.60 64.8 .0140 .0216 7.69 71.7 .0155 .0216 8.18 87.9 .0116 .0132 8.18 87.9 .0116 .0132 8.22 90.3 .0195 .0216 8.34 95.0 .0205 .0216 8.47 94.0 .0124 .0132 9.74 97.9 .0129 .0132 9.82 98.5 .0130 .0132 ”The equilibrium values listed are average values when exact duplication was not achieved on analysis. -54- To prove that one hour was sufficient time for equi- librium to be attained, samples were withdrawn from one run after one and two hours shaking time. The analytical results in each case were identical. To check for weight accountability samples from both phases of one run were analyzed. The neodymium in the aqueous phase was precipitated as the oxalate and ignited to the oxide. The results are listed in Table XX. TABLE XX Weight Accountability Check pH Oxide in Oxide in Total Oxide Total Oxide 5 mls. H20 5 mls. CHCl3 Found Initial Phase g. Phase g. 7.59 0.0106 0.0111 0.0217 0.0216 0.0104 0.0109 0.0213 0.0216 These data show good accountability, and that analysis of one phase is sufficient to attain the desired results. To observe the effect of rare earth concentration, three different concentrations of neodymium acetylacetonate in chloroform were used. From the graphical presentation in Figure 5 it may be seen that all the points fall on a smooth curve, indicating that the distribution of neodymium is independent of concentration of the metal over these ranges. Since radioactive lanthanum and europium were available, tracer concentrations were used to determine the extracti- bility of the compounds of these two elements. Similar -55- extraction techniques were used, and analyses were made by counting five milliliter portions of each phase. Only enough extractions were made to establish the positions of the lanthanum and europium relative to neodymium. TABLE XXI The Distribution of Eu(05H702)3 Between H20 and CHCl3 pH % Eu in Activity in Activity in Organic Phase Organic Phase Aqueous Phase 2.50 0.89 468 52.309 6.40 2.63 481 19.890 6.50 4.00 ‘463 11.463 6.50 1.55 252 15-998 6.80 15.3 1150 6.395 7.55 59.2 962 663 TABLE XXII The Distribution of La(05H702)3 Between H20 and CH013 pH % La in Activity in Activity in Organic Phase Organic Phase Aqueous Phase 1.63 0.43 82 18.804 3.07 .34 72 20.820 4.42 1.10 208 18.499 5.15 .98 179 18.023 6.53 3.54 403 10.991 6.95 8.38 421 4.602 7.03 18.4 423 1.872 7.48 20.3 222 874 96 RARE EARTH ACETYLACETONATE IN ORGANIC PHASE IOO 90 80 7O 60 50 4O 30 20 l 0 FIGURE 5 DISTRIBUTION or RARE EARTH ACETYLACETONATES BETWEEN H20 AND CI~ICII3 -57... There is little difference in the positions of the extraction curves for these three compounds. Separations of the rare earths by this type of extraction would not be very efficient. Absorption Spectra Complex formation is known to cause changes in the absorption spectra of the metal and the chelating agent. It was decided to investigate the absorption spectrum of at least one rare earth acetylacetonate in the ultraviolet, visible, and infrared regions. (a) Ultraviolet Region Since acetylacetone absorbs in the ultraviolet region, known concentrations of acetylacetone were prepared using a weight burst, and the spectra in the ultraviolet region was determined with a Beckman Model DU Spectrophotometer. More dilute solutions were prepared by adding solvent. Chloroform was used as solvent and reference in the density measurements. _58- TABLE XXIII Absorption Spectrum of Acetylacetone in CHCl3 Solutions M or CSHBOZ D 10.2. M of CSHBOZ D m.1_1_. 9.51x10-5 0.656 260 3.80x10-5 0.365 270 .800 265 .373 275 .892 270 .347 280 .915 272 1.52x10-5 .148 270 .920 274 .152 275 .919 275 .142 280 .909 276 5.90x10-6 .060 270 .886 278 .062 275 .850 280 .057 280 .779 283 .713 285 .523 290 These data (first column) are presented graphically in Figure 6. Observation of these data indicate that acetyl- acetone does not follow Beer's Law in the concentrations 'studied at 270, 275, and 280 mg, . EIGURE s s A, _. ' II 0" s.sIIr-_I0"III ACETYLACE ' III. eIICL,‘ soLIITIcN ‘ - 0.95 0.9 - .0 m I 0.75 - OPTICAL DENSITY O ~I , 0.65b ° 0.6- l I I I I I I 260 265 270 275 280 285 290 WAVE LENGTH IN mu. -60- The absorption spectrum of neodymium acetylacetonate in the ultraviolet region was also determined. Concentrations of the solutions were determined by evaporation of the solvent and ignition to the oxide. The starting material for these solutions was the trihydrate. TABLE XXIV Absorption Spectra of Nd(CSH702)3 in CHCl3 Solutions 2.75x10-5 0.489 260 2.75x10-5 0.670 282 .662 268 .598 286 .692 270 .500 290 .712 272 .408 294 .726 276 .268 300 .718 278 4.30x10-5 .998 275 .722 277 2.15x10‘5 .542 275 .727 274 1.08x10-5 .287 275 .728 275 5.40x10-6 .150 275 .697 280 Data for various concentrations of neodymium acetylace- tonate at 275 mg, the absorption maximum in this region, are included in the above table. It may be seen that Beer's Law is not obeyed for this compound in the concentrations studied at 275 mg. This is not unexpected since the absorption is due to the acetylacetone of the complex. The absorption peak is shown in Figure 7. From Figures 6 and 7, one observes that the peak for acetylacetone is shifted slightly to a longer wave length when this substance is chelated with neodymium. This behavior might be expected if the chelating agent donates electrons to OPTICAL DENSITY 0.8 0.7 0.6 0.5 FIGURE 7 ABSORPTION OF 2.75xIO"NI Nd (c,II,O,),-3H,O IN cues, SOLUTION l 1 IA 1. l I. 260 265 270 275 280 285 29 mm LENGTH IN mu. -62- the metal, resulting in the production of lower excited states and shifts in absorption to the red. (b) Visible Region A qualitative determination of the spectrum of neodymium acetylacetonate in the Visible region was made using a Beckman Double Beam Recording Spectrophotometer with chloroform as the solvent and reference. The spectrum is shown in Figure 8. In the spectrum of aqueous neodymium (III) there are bands at 510, 520, 740, and 800 m.u. with a broad band at 574 m.u., all of which are of the same relative order of intensity. However, for neodymium acetylacetonate in chloro- form solution, it may be seen that the band in the 574 m.u. region has been resolved into three sharp peaks at 573, 576, and 582 m.u. These peaks are much more intense than the others which may be observed at 514, 525, and 748 m.u. Since Beer's Law studies on other neodymium chelates have been made, no quantitative studies were made (44). A O 4 . . 6‘ O ‘ I I . . . . Ihb In.” strcIIIIRIII..I-- I . IIIRL. IILII :F .I 7e TRANSMISSION 90- GOP 00- 50b 40a FIGURE 8 ABSORPTION SPECTRUM OF Nd(c,H,O,),-3H,O IN SHOE, SOLUTION l l L l l l 500 550 600 650 700 WAVE LENGTH IN m.u. 000 900 -614- (0) Infrared Region Qualitative observations of the absorption spectra between 2 and 15 u were made with a Perkin Elmer Double Beam Infrared Recording Spectrophotometer. Spectra of solid materials were determined by preparing Nujol mulls and placing a small amount of the mull between sodium chloride plates. For the solutions cells with sodium chloride windows were used and chloroform was the solvent and refer- ence. An examination of the spectra given in Figures 9 through 14 shows that all the compounds studied have very similar absorption.’ The spectra of Nujol and chloroform are given for reference purposes. Since the bonds within the organic part of the chelate molecule are responsible for the absorption in this region, the spectrum Of acetylacetone was determined and it will be considered first. The spectrum of pure acetylacetone is recorded in Figure 14, and Figure 15 is the spectrum of a chloroform solution of acetylacetone. Small peaks in the vicinity of 3 u may be assigned to carbon-hydrogen and hydroxide absorption. The peak at 5.76 u and the more intense peak at 5.85 u are due to carbonyl absorption in the keto form Of acetylacetone. A large number of carbonyl compounds are known to show absorption in the 5.8 u region (7). The very intense, broad band at 6.1 to 6.3 u is carbonyl absorption in the enol form Of acetylacetone. -65- This large shift in carbonyl absorption is attributed by Rasmussen g§_§l, (55) to conjugate chelation within the molecule.' The carbon to carbon double bond absorption is either obscured by carbonyl absorption or shifted. Bands in the 6.5 to 9 u region may be caused by CH, CH2, and CH3 groups (7). Bands beyond this region are difficult to assign and are possibly due to skeletal vibrations. Except for those peaks coincidental with intense solvent absorption, the two Spectra are very nearly identical. However, when acetylacetone is chelated with a metal, several differences in the spectra are apparent. The spectra of the rare earth acetylacetonates are the same either as mulls or in solution, except for bands due to or obscured by the chloroform or the Nujol. Absorption bands in the 3 u region are due to carbon-hydrogen absorptions; the peak at 3.4 u is due to Nujol as may be seen from Figure 15. The absence of absorption in the 5.7 and 5.8 u region, which might be expected as the carbonyl absorption of the keto form of acetylacetone, indicates that chelation occurs with the enol form only. The peak at 6.1 to 6.3 u is due to carbonyl absorption of the enol form of acetylacetone, which has been shifted by chelation. A new band appears at 6.6 u which may be assigned to a.perturbed carbon to carbon double bond (35). Since carbon hydrogen absorption has been shown to Occur at about 7.2 u, the band at 7.2 u may be due to this linkage. In the latter region, it may be seen that the -66- bands in acetylacetone have been shifted and the intensities have changed, due probably to changes in the number of contributing groups. The assignment of the various absorption bands may be summarized as follows: Observations of the absorption spectra for the different rare earth acetylacetonates emphasize the fact that it is not possible to follow the change of concentration of rare earth ions during a chelation reaction by changes in rare earth absorption. as LII, ‘ III II... I P 1‘1 I!“ 30.: 2. 35.020; 8% .. . . -. . ., -... .a... . ...a- .a.. :1. 1 .-. 83 Tie craze-scum“- .S-E .882 a. ma onn.n.moammo.am no samaommm mmmamazH may m gag“... “ZOCQZ 2. §J§<>> O. ON 8% or s a SS 2m 8 e. I; Z : co- ?io 52339- N m m a m mums momma 4 me 0 mm. . o m o.cq so samaommm nmmamazH mma 0.... 55.0.?“ - -.-..-.o ommn.n.moammuvez so zmmaum HH Emame Agni 40652 d m4 mm QHNNMEZH HEB “Edge; 2. 35.2; 1 80¢ -L£0 EOZRKE¢A $9530.: 2. :52; V W0 2': 2 V. — o1Ie004.lu-Iliemssss_ Isa... I, .. aids awn» an» ”muw «pm .Mm_ “a M_.. . ,4. 1H » M-L &u H. “rt-1 H M- “LL%~_MM .M MN“ ”MAIN. H H.” Mam—L «La LL-Tw. L“; I . n. .._. “if, .34. .uan- , .. 1.. n-, . .1” , .H ...,m-:._.I. ..amA-...s..§n-,...N-N.,....m.£.. _ _ r: n. i _ .. I .P l J O. Iir _ .I: 0... M I o: u- I H La! 0.4. M “.4 I an I I c-I II. ..-. -.m ctr. do. I an arm. “H 0.0 . h , 6.0 I .. Man a... so so so. so so . .me _ _ . - I. . a f» as no No no “PW Re no .4 ON _ _. a 0» .. . I. + ONO U. so so so . o co .. . _ ._ . l _ l: . . . . .resx .... . .. s - crc 0. .- n-n nso F . H one _ n. 1 . h . . M . Z 1 . L - I g . ; J. . , a 078 to 13 so 10 MN so 4 so a w H a l a w . .p . . . n . _ . l .1 . .< _ . . no no . no, n no 4 Jo on - . , . l . * N I . . . w - . . . . .Il - . - \- . s: an M ~cK M so \ ~o an. ac m . - . . . . . . . . . . . a .r . [Oh IKII. r _ III \ . CI! _ . . {L M _ o . \ . , AIM-o .o . JIIXII/rI-rI O. - .0 - . .1. 7941 h. fkv‘a‘l — q r a m f _ ._ . .. I _ . . O. _ . .- . .. I . . ... . . . .. ”H. -.. . _, — . . . M I .. . a. . .. . _. . s a a. . . . s .. -. - I. . . .s . .a > I I t pr p A a IIL .- 4 I » rebel . I . F a l I _ _ uOIIPIs PL P AH V a acre. p r H] r {(0th I > p 00 L s . w P y 00—)» y p p > A r k H oo- 1 4 4 4 q 4 4 d a 4 4 a d 1 a a q q a a q Tie 8233:“- N n N a m onapqom maomo 2H 0 mm. A o m o.am mo zpmaumdm nmmemazH are NH H580?“ 39: 2. Egg N m m a m 20Ha=aom maomo 2H 0 mm. . o m 6162 so homeommm ammamazH may m... MESH... 7:0 5230mm“- 2 es on x $2080.: 2. 3kg; Tie 5230mm“- mzoemojwamo4 ho vsmaommm mama-"EH mm...- :H numb-bunk OZOKU.‘ 2. Egg . 1...: 11...... ..-..Co.0?: T... L: ECHO-I. :: :.... 218.11... -:.:.... .1. TIE :I-fi:I-. .1. I..- .Z. CL. 2.11:: 2:... Z O -..n--.. .4... .1“ -. u...- ...M -...:.s. .-._A A...“ _n_..- ..-.. I _.-n-. - - -. _ .w... . .I .... ... . .- . .. .. ,. . ... -. . . . .. . ... 4..“ .b..... . v. - _ _ I. - - .. -.h .> - .. . . . H”. 0.. 1..“0: -II- II.” .Io-QHH HHJILI: :19...- .:. LICLIHHO-Hfl. H..- I I «c... H . I -91....”a O. t I . 1.6,. .. fl 2 I L: .g ..€ ; II . t. . .8- .- oao . Jo . . o-o . - oao . .. . . co no a. a I no - In.o . 5.0 I I K.o . . . NI . . . Mo . no . . ON 1 . T . . - - i .1. or. . 0.0 ”.00 06 x/ . \ . w.0 . 00 H .. H." . . 1 I I .1 . 1L. y . ._f- a. f H. _ .. .. _. 21-1 Tanya n+0 A . ._n-o. I- . s a.W NIT} .. mac N. J . H . . I Jo Jo . .1 I. .2 a s I. . . - fi. H MN H . P It 8 . wo «o F 12.. fr. .. so to h w . a r H . , H. H M g a H . . .f . . . . . .. - . _ “N. no no ~ Allfr/ no N . . no .U. no. no. . . -H on I N . _- .TOO H-.- 07-”- I»-.-- oo oo . 00. I . 00 .I H g“. Tic craze-semen- ZOHquom maomo 2H mzoemoamwamod ao ammeomdm mamamazH may m." Hubbub... (C 333/ 710 52333 ‘8th .mo gmeommm Gaga—AH mma w.” HEHUHE $2080.: 2. 13.09342; o o. loo co co. v. _w a II.“ TJUI MM .nMUI...I M._I I.I. Hmu. I M m. .,n__.. I .nmi . . m2: _ H 0.. it .6. r I H I I.» .III we, a > . I», H I _ who JM. . Jo . N I— ”. P I _ n . .I. . . a _ I . I I w I F N W I I I I o..r co 0. 0.0 a . a m r. m . H # _ I” _ W I we FIE w m} I N . I 3 H I I I . . _ I I N I . _ I . I + v0 r. A a»... + I + . I «Iv «r . I T I II I w ~ I I . w I . b _ I. I I _ .I I x I II I I I m a _ Ia n...I.. no M no _ L w I. I. an. _I 3%. . I. _ I I .I r p . .IL” I I _ I, I I r. I . a I I _ I. & I I _I . I I . II _ _ we ~o ~II . u . I. o .. I _ . . : _._ . + I I I .. PI . _ ._ ” fl _v\ J \I V J I . F .o o w _ I \x) . I, I I , FM _ .C . 5 F. . I (I. . . I I I I m I M _ u _ _ I I I O»? 0.0 0,0 I H M 00 M. a q . q 4 «J 4 DON. maomo ho samaommm nmm