T‘HE MICROD ETERMlflAfic-N QF 3ALLM'J-CUM WiTH ALFHA-FUREL DfiGXEME :i'aé‘ifi CYCLQHEXANEMONEDEOXEMg "Ti-ash E‘or *E‘é‘a-z: Oagma z2fM.31 Exfu'CHEGAfi STA-T3 CQLLfiGi Amiga: tiesmM, £52" €949 This is to certify that the thesis entitled TH! MICRODETEBMINATION OF PALLADIUM WITH ALPHAy-FURIL DIOXIME AND CYCIDHEXANEDIONEDIOXIMI presented by Lucien Barnes, Jr. has been accepted towards fulfillment of the requirements for M. So degree in Chemi gt .2? Major profes THE MICRODEI‘ERMINATI ON OF PALLADIUM WITH ALPHA-FURILDI OXIME AND C YC LOHEKAN’ EDI ON EDI OX1 ME 33' Lucien Barnes, Jr. A THEIS Submitted to the School of Graduate Studies of Michigan State College of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of MAST Ed OF SCI ENC E Department of Chemistry 1949 CHEMLSTRY 05m, “7"5 ‘4' 5 6 Q :2» I ACKNOWLEDGMEMP To Professor Elmer Leininger for his able guidance and assistance in this work. *$**#*** #***#* t*** ** * 2-18486 TABLE OF C ONT HITS . Page II~TTROmCTIO}I...COOCOOOOCOOOOOOOC0.0..0....0.................... 1 mPEZIIE§m\:TAI'...C........O......................C............... 5 I. The Microdetermination of Palladium.With Alpha-Furil- dioxime A. Preparation and Standardization of a Palladium Chloride StOCk Solution............................. 5 B. The Reagent Alpha-Furildioxime...................... 8 Co ApparatuB........................................... 8 D. The Effect of Sodium, Sulfate, and Ferric Ions...... 9 E. The Effect of Citrate, Tartrate, and Vereene in Completing Ferric Iron.............................. 10 F. The EEfect of Orthophosphate and Pyrophosphate in Complexing Ferric Iron.............................. 11 G. The Effect of Long Standing and Exposure to Daylight on 8. Solution Of Alpha-Furildioxime................. 12 II. The Microdetermination of Palladiumeith Cyclohexane- dionedioxime A. The Reagent Cyclohexanedionedioximen............... 1-4 B. Application of the Method of Voter, Banks, and Diehl to the Microdetermination of Palladium........ 15 C. The Effect of Increased Standing Time Upon Complete- ness Of Precipitation............................... 19 De The EffOCt Of pH...eeeeeeeeeeeeeeeeeeeeeeeeeoooeeeoe 20 E. The Effect of washing With an Eflectrolyte........... 21 F. Th0 Effect Of Platinum and Ironeeoeee‘eeeoeoeeeeeeeee 23 G. The Effect of Orthophosphate in Complexing Ferric Iron..................O.....................0....... 25 SUT‘GiARYOO00.000.000.000.000000000.0000000000000000.00000000000. 26 BIBLIOGRAPIIY'...‘....O..C..C.‘OCOOO..0.......................... 27 o a g n . a . . . . . v 0 ’ . C ' ‘ C I . O a 9 - s e I I u C 4 o a e t O I C ‘ I ‘ ' O O O . I V . O i - e I o ' ‘ ' ‘ ’ ' g , o I n I D D ' ' l O 0 v . v . . INTRODUCTION With the increased use of organic reagents for the quantitative determination of the elanents, the 1,2-dioximes have received consider- able attention as specific reagents for the determination of nickel and palladium. Of these, dimethylglyoxime is undoubtedly the best known. Other reagents that have been investigated include alpha-mrildioxime, cyclohexanedionedioxime, alpha-benzildioxime, and benzoylmethylglyoxime. Dimethylglyoxime, as a reagent for the determination of palladium and nickel, has three disadvantages that have had an influence on lead- ing investigators to study the possibility of using other reagents. These disadvantages are: l. The limited water solubility of dime‘thylglyoxime makes it necessary to prepare the reagent solution in alcohol or acetone, which, in turn, increases the chance of coprecipi- tation of the reagent. 2. The gravimetric conversion factor with dimethylglyoxime is not as favorable as it is with the other more heavily substituted dioximes. 3. It has been reported (1, 2) that often it is necessary, to allow a long period of standing to insure complete precipitation of palladium with dimethylglyoxime. B. A. Souls (3) in 1925 investigated the use of alpha-furildioxime for the qualitative detection and quantitative determination of nickel. He found it to be much more sensitive than dimethylglyoxime as a quali- tative reagent, and to be soluble in hot water to ‘the extent of about one percent. Souls was also able to use the reagent successfully for the macro-gravimetric determination of nickel. In 1942, D. R. Beas- ecker (4) developed an accurate method for the microdetermination of palladium and nickel using alpha-furildioxime as the precipitating agent, and his procedure was adopted here for subsequent work with the reagent. From a strongly acid solution, palladium yields with alpha-furildioxime an amorphous yellow-orange colored precipitate hav- ing the structure: HC —CH HC —CH Hg\ g—C—i—g EH 0/ g! N \O/ l \ /\ o I OH PdI/Z Cyclohexanedionedioxime, commonly referred to as "nioxime", was first prepared by Wallach (5), who found it to be a more sensitive quali- tative reagent for nickel than dimethylglyoxime, and stated that it had the further advantage of greater water solubility. In 1940 Diehl (6) suggested its use as an analytical agent, but stated that there was no satisfactory method for its synthesis. This difficulty was overcome in 1945 when G. F. Smith (7) and his co-workers successfully developed two methods for the preparation of'nioxime. .One of Smith's methods was later greatly improved by C. C. Each (8), and the reagent may now be purchased from the Bach Chanicalpand Oxygen Company, Ames, Iowa. Johnson and Simmons (9), in 1946, used nioxime in the gravimetric deter- mination of nickel, but reported high results which they attributed to occlusion of excess reagent. In 1948, Voter, Banks, and Diehl (8), successfully used nioxime for the determination of nickel, applying a correction factor for the amount of reagent coprecipitated when the nickel present exceeded fifteen milligrams. Later, these same investigators used the reagent for the macrodetermination of palla- dium (10). The advantages claimed for the use of nioxime were: 1. It is soluble to the extent of 0.8 g/100 ml. in cold water, thus making it convenient to use and minimizing coprecipi- tation of the reagent, and separation of the excess reagent on standing. 2. After a brief period of digestion at 60 degrees centigrade, filtration may be made from the hot solution, thereby shortening the time required for the determination with no loss of accuracy involved. 3. Palladium could be detennined in the presence of platinum in amounts up to one tenth of a gram. 4. The presence of a one hundred and fifty percent excess of reagent does not affect the accuracy of the determination. From solutions of pH range zero to five, palladium forms a bright yellow colored precipitate with nioxime, the structural formula of which is indicated below: H 4R HC Cali-£3 ‘~Pd/2 HC‘ c-eg/ c’ H The object of this study is: 1. To determine the effect of certain foreign ions upon the microdetermination of palladium.using alpha-furil- dioxime as the precipitating agent. 2. To find a complexing agent for ferric iron so that pallad- ium.may be determined quantitatively in its presence with alpha-furildioxime. 3. To develop a micromethod for the determination of pallad- ium using nioxime as the precipitating agent. E‘ZPERIMENTAL I. The Microdetermination of Palladium With alpha-Furildioxime A. Heparation and Standardization if: g Palladium Chloride Stock Solution: A stock solution of palladium dichloride was prepared by dis- solving 1.0 g. of palladium dichloride in seventy-five milliliters of I 1:9 hydrochloric acid. The resulting solution, when made up to a volume of fifteen hundred milliliters, had a concentration of approximately 0.4 mg. of palladium per milliliter, and a pH of 1.3 as measured by the glass electrode. The stock solution was standardized by three methods: 1. A micro- standardization using dimethylglyoxime. 2. A microstandardization using alpha-furildioxime. 3. A macrostandardization using; dimethylglyoxime to serve as a check on the two microstandardizations. The procedure for the macrostandardization of the stock solution with dimethylglyoxime was that of Gilchrist and Wichers (11) and is as follows: Twenty-five milliliters of the stock solution were diluted to one hundred milliliters with distilled water, two milliliters of concentrated hydrochloric acid (12 M) were added, and the palladium was precipitated by slowly adding 3.4 m1. of a one percent solution of'di- methylglyoxime in alcohol. The solution was allowed to stand for three hours, and was then filtered through a previously weighed filtering crucible. The precipitate was washed with dilute hydrochloric acid (1:99), then with hot water, dried at 120°C, and weighed as palladium dimethylglyoxime. The standardization was repeated using fifty milliliter samples. Applying this procedure to the microstandardization, five milliliters of the stock solution were pipetted into previously 'weighed Schwarz-Bergkampf Pyrex filter beakers. Two drops of concen- trated hydrochloric acid (12 M) were added, and the palladium.precipi- tated with 0.7 m1. of a one percent solution of dimethylglyoxime.‘ The solution was allowed to stand for three hours at room temperature. It was then filtered, washed with dilute (1:99) hydrochloric acid, then with hot water, dried for periods of one hour at 120°C, and weighed to constant weight as palladium:dimethylglyoxime which contains 51.62% palladium. For the microstandardization with alpha-furildioxime the method as developed by Beasecker (4) was employed: Five milliliters of the stock solution were pipetted into previously weighed filter beakers, 0.5 ml. of concentrated hydrochloric acid (12 M) was added, the solution was. heated on the water bath, and the palladium.precipitated by the addition of two milliliters of a hot one percent aqueous solution of alpha-furil- dioxime. The solution was agitated to hasten coagulation of the precipi- tate, and allowed to remain on the bath for ten minutes. The precipitate was filtered one hour later, washed three times with warm hydrochloric acid (1:99), and then six times with hot water. It was then dried to constant weight at 120°C, the initial drying period consisting of two hours, and the subsequent drying periods one hour. Palladium furildioxime contains 19.58% palladium. The results of the standardizations are indi- cated in tables I and II. TABLE I macrostandardization of PdClz Stock Solution Using Dimethylglyoxime Mg. Pd / 25 m1. M . Pd / 5 m1. 10.546 iglos 10.578 2.116 10.609 2.122 10.641 2.128 Mg. Pd / 50 ml. 421.210 2.121 21.219 2.122 21.171 2.117 Average: 2.119 TABLE II Microstandardization of PdClz Stock Solution Using Dimethylglyoxime and Alpha-Furildioxime Dimethylglyoxime Alpha-Furildioxime M . Pd ['5 m1. mg. Pd /'5 m1. '-'-J£"E.116 2;115 2.119 2.115 2.117 2.118 2.120 2.116 2.113 2.115 2.116 2.121 fiverage:2.117 Average: 2.11? The mean result of the two microstandardizations is 2.117 mg. Pd /’5 ml., and this figure was taken as being the true concentration of the stock solution. -7.- B. The Eeagemt Alpha-Furildioxime: A one percent solution of alpha-furildioxime‘was prepared by dissolving one gram of the reagent in one hundred milliliters of warm.water. The solution.was then fil- tered four times through number 42 Whatman filter paper to remove traces of a brown sediment that were originally present. Upon cooling, the re- agent crystallized into fine white needles in a colorless solution. It is, therefore, necessary to warm the reagent solution before use. After a period of four or five weeks, when exposed to daylight, the solu- tion takes on an amber color, and small brown colored particles are ob- served on the bottom of the container. Upon warming, these brown colored particles do not dissolve, however if they are filtered off prior to use, the reagent solution still yields accurate results, as will be indicated later in this report. C. Apparatus: One‘type of filter beaker was used throughout this experiment. This type, known as the Schwarz-Bergkampf filter beaker, is made of Pyrex and has a capacity of approximately ten milliliters. The advantage of using this type beaker over the conventional filterstick technique, is that they are decidedly easier to handle. The beakers are cleaned readily, and very few were found to pass precipitate through the sintered glass disc that serves as the filter. All weighings were made in.a constant temperature balance room ('780 F) on an.Ainsworth optical lever microbalance. This balance weighs directly to micrograms. Counter- poise weighing was used exclusively. The procedure used in weighing was the cameras that used by Beas- ecker (4), and is as follows: After removing the sample beakers and counterpoise from the drying oven, they were allowed to cool for exactly thirty minutes, next to the microbalance, in a desiccator over Dehydrite. The counterpoise and sample beaker were then placed upon the balance pans, the approximate weight found, and the exact weight recorded five minutes later. By rigorous adherence to the above time schedule it was usually possible to obtain constant weights, within five micrOgrams, upon the third or fourth weighing. However, during very humid weather, it was exceptionally difficult to obtain constant weights. It was ob- served that constant weights were obtained most readily when the relative humidity was in the neighborhood of fifty to fifty-five percent. D. The Effect of Sodium, Sulfate, and Ferric Ions: The effect of sodium and sulfate ions upon the determination of palladium with alpha- furildioxime was determined by adding an aqueous solution of sodium sul- fate to samples of the stock solution, and carrying out the precipitation in the regular manner. It was noticed, that during the filtration, the normal yellow-orange colored precipitate developed cracks. Therefore it is suggested, that when carrying out the determination in the presence of these ions, not too much time be allowed to elapse between the time that the precipitate is aspirated dry and the subsequent washing. The results in table III indicate that sodium and sulfate ions do not appreciably interfere with the determination. The effect of ferric iron, in the form of ferric chloride is also indicated in table 111. The results are slightly high, and therefore, palladium cannot be accurately determined in the presence of ferric ions unless some suitable complexing agent for the iron is present. The pre- cipitate contaminated with iron is brown in color. -9- TABLE 111 Effect of Sodium, Sulfate, and Ferric Ions Upon the Determination of Palladium With Alpha-Furildioxime ‘mstances Present Mg. Pd Mg. Pd firor firor Substance Mg. Present Found M . % Na2304 5 2.117 2.115 -0.002 -0.10 Nazso4 5 2.117 ' 2.123 +0.006 00.28 Nazso 5 2.117 2.111 -0.006 -0.28 Fe+3 2 2.117 2.125 +0.008 +0.38 Pe*3 2 2.117 2.125 +0.008 +0.33 Pe*3 2 2.117 2.127 +0.010 +0.47 F943 2 2.117 2.126 +0.009 +0.42 E. The Effect _o_f_ Citrate, Tartrate, and Verseng _i__n Complexing Ferric Iron: In the attempt to find a suitable complexing agent for ferric iron, citric acid, tartaric acid and Versene (the sodium salt of ethylenediamine tetra acetic acid) were used. In the cases of cit- ric and tartaric acids, one milliliter of a forty percent solution of the acid was added to the sample of the stock solution containing two milligram of ferric iron. With Versene, six tenths of a milliliter of a thirty-five percent solution was us ed. The results are shown in table IV, and indicate that none of these agents effectively complex the iron. -10- TABLE IV The Effect of Citrate, Tartrate, and Versene in Complexing Ferric Iron Mg. Fe+3 Complexing Agent Mg. Pd Mg. Pd Error Error Pres ent Pr es amt . Pr es ent Found Mg. % 2 Citric Acid 2.117 2.146 +0.029 +1.3? 2 Citric Acid 2.11? 2.134 +0.01? +0.81 2 Citric Acid 2.117 2.144 +0.02? +1.27 2 Tartaric Acid 2.117 2.153 +0.036 +1.70 2 Tartaric Acid 2.117 2.150 +0.033 +1.56 2 Tartaric Acid 2.117 2.147 +0.030 +1.42 1.2 Versene 1.270 ' 1.297 +0.02? +2.12 1.2 Versene ‘1.270 6.529 +5.259 +415 1.2 Versene 1.270 1.288 +0.018 +1.42 F. The Effect of Orthophosphate and Pyrophosphate in Complexing Ferric Iron: The next agent investigated inthe search for a complex- ing agent for ferric iron was orthophosphate. One half milliliter of a forty percent solution of orthophosphoric acid was added to the same ple of the stock solution prior to precipitation. The results listed in table'V indicate that orthophosphoric acid does complex the iron. The precipitate was of the normal yellowhorange color. Similarly, three milliliters of a four percent solution of tetrasodium pyrophos- phate were added to samples of the stock solution containing two milli- grams of ferric iron. The results in.table V show“that pyrophosphate also effectively complexes ferric iron when alpha-furildioxime is used as the precipitating agent for palladium. TABLE V The Effect of Orthophosphate and Pyrophosphate in Complexing Ferric Iron Mg. Fe” Complexing Mg. Pd Mg. Pd B‘ror Error Pres ent Agent Pras ent Found M g. % 2 H3P04 2.117 2.118 +0.001 +0.05 2 1131204 2.117 2.119 +0.002 +0.10 2 H31304 2.117 2.120 +0.00?) +0.14 2 1131304 . 2.117 2.118 +0.001 +0.05 2 H3P04 2.117 2.117 0.000 0.00 2 Na4P207 2.117 2.118 +0.001 +0.05 G. The Effect of Long Standing and Ecposure 1:2 Daylight on e; Solution of Alpha-Furildioxime: As was previously mentioned, the re- agent solution of alpha-furildioxime changes, over a period of weeks, from a colorless solution to amber when exposed to daylight. Brown colored particles may be observed on the bottom of the container which do not dissolve upon warming. A solution that had been exposed to day- light for six weeks was warmed, the insoluble residue filtered off, and the filtered solution used for determining palladium in the usual manner. As can be. seen from table VI, such a solution is capable of yielding accurate results. -12- TABLE VI The Effect of Long Standing and Exposure to Daylight on a Solution of Alpha-Furildioxime Mg. Pd Mg. Pd B‘ror B‘ror Present Found Mg. 7: 2.117 2.118 +0.001 +0.05 2.117 2.116 -0.001 -o.05 2.117 2.117 0.000 0.00 2.117 2.119 +0.002 +0.10 ' -13.. II. The Microdetermination of Palladium With Cyclohexanediozidioxime ' A. The Reagent Cyclohexanedionedioxime: G. F. Smith (7) and his co-workers suggested the name "nicxime" for the reagent cyclohexane- dionedioxime, and after their successful synthesis of the compound, its use in the macrodetermination of palladium was investigated by Voter, Banks, and Diehl (10). They found that quantitative precipitation of the palladium was obtained at pH values from 0.? to 5, but could not be assured at lower pH values. It was found tint the determination could be carried out in the presence of the following ions: chloride, sulfate, nitrate, acetate, tartrate, sulfosalicylate, uranyl, ruthenium, beryl- lium, sodium, potassium,llithium, barium, strontium, calcium, aluminum, lanthanum, zinc, cadmium, and platinum in amounts up'to one tenth of a gram. Aurous ions formed a precipitate with nioxime, so the determina- tion of palladium could not be conducted in the presence of gold. The procedure suggested by these investigators for the macrodeter- mination of palladium with nioxime is as follows: Adjust the volume of the solution, containing from five to twenty milligrams of palladium, to approximately two hundred milliliters. The pH may vary from one to five depaiding upon the other ions present. Heat the solution to 60°C. and add 0.43 ml. of an 0.8% aqueous solution of nioxime for each milli- gram of palladium present. Digest for thirty minutes at 60°C., with occasional stirring. Filter whilehot through a previously weighed fil- tering crucible of medium porosity, wash five times with hot water, and dry for one hour at 110°C. The factor for palladium is 0.2743. -14.. All 0.8% aqueous solution of nioxime was prepared by dissolving 0.8 gram of the reagent in cold water. The solution was filtered three times through number 42 Whatman filter paper in an attempt to remove traces of a milky like suspension that appeared upon agitation of the solution. However, a slight suspension remained even after re- peated filtration. Several series of microdeterminations of palladium using nicxxime as the precipitating agent were conducted, following essentially the procedure outlined above. The description and results of these deter- minations follow: B. Application of the Method gfLVoter, Banks, and Diehl (19) t2 the Microdetermination of Palladium: Initially, two sets of conditions were studied in adapting the macromethod of these investigators to the microdetermination of palladium. These were: 1. The effect of. drying the precipitate at 120°C. 2. The effect of drying the precipitateet 110°C. It was found that when the precipitate was dried at 120°C., it had a tendency to gradually lose weight after each period in the oven, thus making it difficult to obtain constant weights. 0n the other hand, when the precipitate was dried at 11000., no loss of weight was observed with subsequent dryings, and constant weights were readily obtained. The procedure used in the microdeterminations was as follows: Five milliliters of the stock solution were pipetted into previously weighed filter beakers, and these were placed upon a water bath at 60°C. The palladium was precipitated with 1.2 m1. ofan 0.8 3% solution of nioxime, and the solution allowed to digest at 60°C. for one half hour. The -15- bright yellow precipitate that formed did not settle readily, had a tendency to cling to the sides of the beaker, and was considerably lumped together. It was necessary to agitate the solution vigorously to break up these lumps of precipitate. The precipitate was filtered while hot, washed five times with hot water, dried at 120°C. or 110°c., and weighed to constant weight as palladium.cyclohexanedionedioxime. The results of this series of determinations arershown in tables VII and VIII. During the first nine determinations, in.which the precipitate was dried at 120°C., the precipitate was not agitated as thoroughly as it was during all subsequent determinations. This fact might serve to explain some of the high results shown in table VII, but it fails to account for the three abnormally high results shown in table'VIII, be- cause in this latter series of determinations the precipitates were thoroughly broken up during the period of digestion. It should be noted, that while the average of all the determina- tions listed in tables VII and VIII, is near the correct value, the precision of the determinations is poor, some values being abnormally high and others abnormally low; The procedure outlined above is, therefore, in need of modifica- tion before the successful microdetermination of palladium.may be carried out using nioxime as the precipitating agent. -16... TABLE‘VII The Microdetermination of Palladium.With Nioxime, Precipitate Dried at 120°C. Average deviation: 0.008 mg. Pd 1 Mg. Pd ErrOr Error Present Found Mg. % 2.117 2.123 +0.006 +0.28 2.117 2.106 -0.011 -0.52 2.117 2.110 -0.007 -0.33 2.117 2.113 0+0.001 +0.05 2.117 2.100 -0.017 -0.80 2,117 2.118 +0.001 +0.05 2.117 2.124 +0.00? +0.33 2,117 2.125 +0.008 .0.38 2.117 2.126 +0.009 +0.42 2.117 2.129 +0.012 +0.57 2.117 2.106 -0.011 90.52 2,117 2.109 -0.003 -0.38 2.117 2.110 -0.007 -0.33 2.117 2.110 -o.oo7 -0.33 2.117 2.102 -0.015 -0.71 Average: 2.114 ..17- TABLE VIII The Microdetermination of Palladium‘With Nioxime, Precipitate Dried at 110°C. Mg. Pd Mg. Pd Error Error Present Found mg. % 2.117 2.135 +0.018 +0.85 2.117 2.110 -0.007 -0.33 2.117 2.115 -0.002 -0.10 2.117 2.119 +0.002 +0.10 2.117 2.138 +0.021 +0.99 2.117 2.116 -0.001 -0.05 2.117 2.138 +0.021 +0.99 2.117 2.116 -0.001 -0.05 2.117 2.114 -0.003 -0.14 2.117 2.114 -0.003 -0.14 2.117 2.108 -0.009 -0.43 2.117 2.108 -0.009 -0.43 2.117 2.117 0.000 0.00 2.117 2.101 ~0.018 -0.76 2.117 2.106 -0.011 -0.52 Average: 2.11? Average deviation: 0.008 -18- C. The Effect of Increaggd Standing Time Upon Empletenegg p_f_ Precipitation: Reference to tables VII and VIII show that many of the results obtained using nioxime are loan It was thought that a possible remedy to this situation might be to allow the precipitates to stand for one hour at room temperature prior to filtration. The results in table IX show that the errors are still largely negative, and that evi- dently the longer period of standing prior to filtration is not import- ant. TABLE IX Effect of Allowing Solution to Stand at Room Temperature For One Hour Prior to Filtration Mg. Pd Mg. Pd Error Error Pres ent Found Mg. % 2.117 2.114 -0.003 -O.14 2.117 2.112 -0.005 -0.24 2.117 2.120 +0.003 +0.14 2.11? 2.124 +0.00? +0.33 2.117 2.102 -0.015 -0.70 2.117 2.113 -0.004 -0.19 2.117 2.094 -0.023 -l.09 2.117 2.093 -0.024 -1.13 2.117 2.118 +0.001 +0.05 Average: 2.110 Average deviation: 0.009 -19.. D. The Effegji 33 pH: The effect upon quantitative precipitation, of increasing and of decreasing the pH of the stock solution was studied next. A series of determinations was conducted in which the pH of sam- ples of the stock solution was raised to approximately three by the addi- tion of 0.9 m1. of a one molar solution of sodium acetate. Another series was run at a pH of approximately 0.3, by the addition of 0.2 m1. of concentrated hydrochloric acid (12 M) to the samples of the stock solu- tion. The results of these experiments are shown in tables X and XI, and indicate, that within the ranges studied, pH is apparently not a critical factor in the microdetermination of palladium with nioxime. In view of these results, the remaining determinations were made on the unaltered stock solution, the pH of which was 1.3. TABLE X The Effect of Precipitation at pH 3. h1g0 Pd M o Pd E‘I‘OI‘ El‘ror Pr es ent Found Mg . % 2.117 2.125 +0.008 +0.38 2.117 2.122 +0.005 +0.24 2.117 2.119 +0.002 +0.10 Average: 2.122 Average deviat ion: 0.005 -20- TABLE XI The Effect of Precipitation at pH 0.3 Mg. Pd Mg. Pd Ek'ror firor Present Found Mg. 75 20117 2.126 #00009 +0.43 2.117 2.115 -0.002 “0.10 2.117 2.128 +0.011 +0.52 Average: 2.123 Average deviation: 0.007 E. The Effect of Washing the Precipitat_e_ With 32 Electrolyte: It was thought, that perhaps the reason for the low results so far en- countered, might be due to colloid formation, thus causing some of the precipitate to pass through the filter. Therefore, a series of deter- minations was made in which the precipitate was washed, first five times with hot hydrochloric acid (1:200), and then five times with hot water. In this series of washings the total volume of wash liquid used was approximately twelve to fifteen milliliters. In all determinations pre- vious to these thetotal volume of wash liquid used was approximately twenty milliliters. The precipitate of palladium cyclohexanedione- dioxime did not settle as readily as the precipitate of palladium furil- dioxins, and it had a great tendency to cling to the sides of the beaker; consequently it was necessary to use a larger volume of wash liquid, to insure thorough washing, than is ordinarily used in microdeterminations. It was also observed, that in all of the nioxime filtrates, there was -31- present an appreciable amount of a white precipitate. This precipitate was noticeable in the filtrate after the third or fourth washing, and did not appear to dissolve upon warming. The nature of this precipitate is unknown. The results of this series of determinations are listed in table XII, and while still slightly low, they show better precision than.the results obtained under other conditions studied. The improvement in results may be due to the effect of washing with a smaller volume of hot water. TABLE XII Effect of Washing the Precipitate With an.Eflectrolyte Mg. Pd Mg. Pd Error Error Present Found M . 3% 2.117 2.114 -0.003 -O.14 2.117 2.117 0.000 0.00 2.117 2.114 -0.003 -0.14 2.117 2.109 -0.008 10.38 2.117 2.111 -0.006 -0.28 2.117 2.110 -0.007 -0.33 2.117 2.114 -0.003 -0.14 2.117 2.123 +0.006 +0.28 2.117 2.113 -0.004 -0.19 Average: 2.114 Average deviation: 0.004 -22.. The recommended procedure for the microdetermination of palladium with nioxime is, therefore, as follows: Pipette a five milliliter sam- ple of the solution containing the palladium into a previously weighed filter-beaker, adjust the pH to between one and five, according to the other ions presmt, and place on a water bath at 60°C. Add 0.5 - 0.6 m1. of an 0.8% aqueous solution of nioxime for each milligram of pallad— ium present, and digest for one half hour at 60°C., with frequent agita- tion of the solution to break up any lumpy precipitate present. Filter while hot, wash five times with hot 1:200hydrochloric acid, and then five times with hot water. Dry at 11000., and weigh as palladium cyclo- hexanedionedioxime. It is suggested that the first drying period be for two hours, and that each subsequent period be one hour. F. The Effect .93 Platinum and Iron: Voter, Banks, and Diehl (10), were able to determine palladium with nioxime in the presence of up to four times as much platinum. Experiments conducted in this laboratory with equal amounts of palladium and platinum, (2 mg. Pd to 2 mg. Pt) indicate that the results for palladium in the presence of platinum are high. The platinum was added to samples of the stock solution in the form of chloroplatinic acid. Two sets of determinations were run: 1. A series in which the precipitate was allowed to digest thirty minutes at 60°C. prior to filtration. 2. Another series in which the precipitate was allowed to digest forty-five minutes at 60°C. prior to filtration. In both cases the precipitate was green in color, rather than yellow. Apparently, as can be seen from the results in table XIII, the amount of platinum precipitated is a function of time. The precipitate that first forms, in the presence of platinum, is of the normal yellow color, but -23- after a period of about ten to fifteen minutes have elapsed, the precipi- tate grows darker in color. One sample, in an unweighed beaker, was fil- tered after digesting for ten minutes at 60°C. This precipitate was of norml color, but palladium precipitated in the filtrate, therefore it was not cleaned advisable to shorten the period of digestion. The effect of ferric iron, added in the form of ferric chloride, is shown in table XIV. The results are high, and obviously palladium cannot be determined with nioxime in the presence of iron unless some suitable v complexing agent for the iron is present. The palladium precipitate con- taminated with iron was brownish yellow in color. TABLE XIII The Effect of Platinum ’ Mg. Pt Digestion Mg. Pd Mg. Pd Error Error Present Period Present Found Mg. % 2 50 min. 2.117 2.162 +0.045 +2.12 2 30 min. 2.117 2.165 +0.048 +2.26 2 30 min. 2.117 2.160 40.045 .2.03 2 45 min. 2.117 2.230, +0.11: +5.34 2 45 min. 2.117 2.219 +0.102 +4.82 2 45 min. 2.117 2.226 +0.109 +5.15 -24- TABLE XIV The Effect of Ferric Iron Mg. Fe M . Pd Mg. Pd Error Error Pres ent Pres ent Found Mg. % 2 2.117 2.137 +0.020 +0.95 2 2.117 2.138 +0.021 +0.96 2 2.117 20131 #00014 #0066 G. The Effect 9.3. Orthogiosphatg in Complexing Ferric Iron: One half milliliter of a forty percent solution of ‘orthophoephorio acid was added to samples of the stock solution containing two milligrams of ferric iron. The results. are high, as shown in table XV. Therefore, orthophosphate does not effectively complex ferric iron when nioxime is used as the precipitating agent for palladium. TABLE XV The Effect of Orthophosphate in Complexing Ferric Iron Mg. Fe Mg. Pd Mg. Pd Eh‘ror Error Present Present Found M . % 2 2.117 2.134 +0.017 +0.80 2 2.117 2.140 +0.023 +1.09 2 2.117 2.138 +0.021 +0.99 2 2.117 ' 2.135 +0.018 +0.85 2 2.117 2.138 +0.021 +1.09 2 2.117 2.153 +0.016 +0.76 SUI {MARY The effect of sodium and sulfate ions upon the microdetermina- tion of palladium with alpha-furildioxime is negligible. Ferric iron interferes, but may be effectively complexed with ortho or pyrophos- phate. An accurate gravimetric method for the microdetermination of pallad- ium with nioxime has not as yet been worked out. The method as developed here, yields results that can only be described as fair. Platinum and ferric iron cause high results, and as yet no complexing agent for the iron has been found. Nioxime as a precipitating agent for the microdetermination of pal- ladium, offers no advantages over alpha-furildioxime other than solubil- ity in cold water, and the fact that filtration may be made from a hot solution after a brief period of digestion at 60°C. Alpha-furildioxime, on the other hand, possesses the following advantages over nioxime: 1. It is capable of consistently yielding highly accurate results. 2. The conversion factor for palladium is more favorable with alpha-fur- ildioxime (0.1958), than with nioxime (0.2743). 3. Palladium may be determined in the presence of iron, using ortho or pyrophosphate as a complexing agent, with alpha-furildioxime. This cannot be done with nioxime using orthophosphate as the complexing agent. 4. The work of Beasecker (4) indicates that in the presence of platinum, the results with alpha-furildioxime are not as high as they are with nioxime. 5. The precipitate of palladium furildioxime settles readily and has very little tendmcy to cling to the sides of the beaker, thus it is easier to handle than the precipitate formed with nioxime. BIBLIOGRAPHY 1. Scott, W. W., "Standard Methods of Chemical Analysis", Fifth Edition, Vol. I, p. 724, N. Y., D. VanNostrand Co. Inc., 1939. 2. Zschiegner, H. E., Ind. Eng. Chem., .11, 294 (1925). 3. Souls, B. A., J. Am. Chem. $00., 41, 981 (1925). 4. Beasecker, D. R., Master's thesis, "The Microdetermination 0f Palladium And Nickel By Means Of Alpha-Furildioxime“, Michigan State College, 1942. 5. Wallach, 0., Ann., 457, 175 (1924). 6. Diehl, H., "Application Of The Dioximes To Analytical Chemistry", Columbus, Ohio, G. F. Smith ChenicalCo.,_ 1940. 7. Raugh, E. G., Smith, G. F., Banks, C. V., and Diehl, H., J. Org. Chem., E, 199 (1945). 8. Voter, R. C., Banks, C. v., and Diehl, H., Anal. Chem., 20, 458 (1948). ‘— 9. Johnson, w. c., and Simmons, 14., Analyst, _'_I__1_, 554 (1946). 10. Voter,’R. C., Banks, C. V., and Diehl, H., Anal. Chem., 20, 652 (1948). "" 11. Gilchrist, R., and Wichers, E., J. Am. Chen. $00., 57, 2565 (1955). '— T543 218486 3261 Barnes- i‘he microdetermination of palladium with alpha-fur il dioxins and cyclohexanedi- oxime. 3 1293 02446 7270