EXCHANGE 
 
 REACTIONS BETWEEN POTASSIUM 
 
 AMIDE AND CERTAIN SALTS 
 
 OF NICKEL AND CHROMIUM 
 
 IN LIQUID AMMONIA 
 
 SOLUTION 
 
 A THESIS 
 
 SUBMITTED TO THE DEPARTMENT OF CHEMISTRY AND THE 
 
 COMMITTEE ON GRADUATE STUDY OF THE LELAND 
 
 / STANFORD JUNIOR UNIVERSITY IN PARTIAL 
 
 ' FULFILMENT OF THE REQUIREMENTS 
 
 FOR THE DEGREE OF DOCTOR 
 
 OF PHILOSOPHY 
 
 GEORGE S. BOHART 
 March, 1915 
 
 \ 
 
 UNlVtKS 
 
 Easton, Pa.: 
 EscHENBACH Printing Co. 
 1915 
 
REACTIONS BETWEEN POTASSIUM 
 
 AMIDE AND CERTAIN SALTS 
 
 OF NICKEL AND CHROMIUM 
 
 IN LIQUID AMMONL\ 
 
 SOLUTION 
 
 A THESIS 
 
 SUBMITTED TO THE DEPARTMENT OF CHEMISTRY AND THE 
 
 COMMITTEE ON GRADUATE STUDY OF THE LELAND 
 
 STANFORD JUNIOR UNIVERSITY IN PARTIAL 
 
 FULFILMENT OF THE REQUIREMENTS 
 
 FOR THE DEGREE OF DOCTOR 
 
 OF PHILOSOPHY 
 
 BY 
 
 GEORGE S. BOHART 
 March, 19 15 
 
 Easton, Pa.: 
 EscHENBACH Printing Co. 
 1915 
 
REACTIONS BETWEEN POTASSIUM AMIDE AND 
 CERTAIN SALTS OF CADMIUM, NICKEL 
 AND CHROMIUM IN LIQUID AMMONIA 
 SOLUTION^ 
 
 BY GEORGE S. BOHART 
 
 CONTENTS 
 
 I. INTRODUCTION. I. The Ammonia System of Acids, Bases and Salts. 
 2. Amphoteric Metallic Amides. 3. Object of this Investigation. 
 
 II. MANIPULATION OF WQUID AMMONIA SOLUTIONS AND DESCRIPTION OF 
 APPARATUS USED. 
 
 III. ACTION OF POTASSIUM AMIDE ON CADMIUM SALTS. J. PotaSsium 
 
 Ammonocadmiate, Cd{NHK)i.2NHz. 2. Cadmium Amide, Cd^NH-iji. j. Cad- 
 mium Nitride, Cd^N^. 
 
 IV. ACTION OF POTASSIUM AMIDE ON POTASSIUM CYANONICKELATE. I. 
 
 Preparation of Pure Potassium Cyanonickelate, Ni{CN)iK2. 2. Compound No. i. 
 A Complex Product of the Empirical Formula Ni3N2H2K4{CN)s.8NH3 and its 
 Deammonation Product, Ni3N2H2Ki{CN)&. 3. Compound No. 2. A Mixed 
 Cyanonickelate-ammononickelate of Potassium, K{CN)2NiNHK. 4. Compound 
 No. 3. A Complex Compound of the Empirical Formula Ni3NnH22KT{CN)2. 
 
 V. ACTION OF POTASSIUM AMIDE ON NICKEL SULFOCYANATE. I. Am- 
 
 monated Nickel Sulfocyanate. 2. Nickel Sulfocyanate with Four Molecules of 
 Ammonia, Ni{SCN)2.4NH3. 3. Nickel Sulfocyanate with Three Molecules of 
 Ammonia, Ni{SCN)2:3NHs. 4. Nickel Sulfocyanate with Two Molecules of 
 Ammonia, Ni(SCN)2.2NH3. 5. Nickel Sulfocyanate with Five and a Half Mole- 
 cules of Ammonia, Ni(SCN)2.5^/2NH3. 6. Nickel Sulfocyanate with Eight and 
 a Half Molecules of Ammonia, Ni{SCN)2-8^/2NH3. 7. Potassium Ammono- 
 nickelate, Ni2N3Ki.6NH3. 8. Nickel Amide, NiiNHi) 2. 9. Nickel Nitride, NizN 2- 
 
 VI. ACTION OF POTASSIUM AMIDE ON AMMONIUM CHROMIUM SULFOCYANATE, 
 
 NH4Cr(SCN)4.2NH3. 
 
 VII. SUMMARY. 
 
 I. Introduction 
 
 I. The Ammonia System of Acids, Bases and Salts. — In 
 two important papers,^ Franklin has developed in detail an 
 ammonia system of acids, bases and salts. He has called 
 attention to the fact that the acid amides, the metallic amides 
 and the metallic derivatives of the acid amides are formally- 
 related to ammonia as the familiar oxygen acids, bases and 
 
 1 The author's thesis presented to the Department of Chemistry of the 
 Leiand Stanford Junior University in partial fulfillment of the requirements 
 for the degree of Doctor of Philosophy. 
 
 2 Jour. Am. Chem. Soc., 27, 820 (1905); Am. Chem. Jour., 47, 285 (1912). 
 
 335897 
 
un>KRU 
 
 538 George S. Bohart 
 
 salts are related to water and he has shown that these substances 
 actually exhibit in liquid ammonia the distinctive properties 
 of acids, bases and salts respectively. Acid amides in liquid 
 ammonia solution show an acid reaction toward phenolph- 
 phthalein; they react with certain metals with the evolu- 
 tion of hydrogen and with metallic amides, imides and nitrides 
 in a manner strictly analogous to the action of aqueous solu- 
 tions of oxygen acids on metals, metallic hydroxides and 
 oxides. 
 
 2. Amphoteric Metallic Amides. — A further analogy be- 
 tween the ammonia and water systems is found in the ampho- 
 teric behavior of certain metallic amides which recalls the 
 familiar behavior of zinc, lead and aluminum hydroxides 
 towards acids and strong bases. Fitzgerald^ and Franklin^ 
 have shown that just as zinc hydroxide dissolves in aqueous 
 solutions of potassium hydroxide to form potassium (aquo) 
 zincate in accordance with the equation, 
 
 Zn(0H)2 + 2KOH = Zn(0K)2 + 2H2O, 
 so zinc amide is converted into an ammonozincate of potas- 
 sium by the action of a liquid ammonia solution of potas- 
 sium amide on zinc amide as represented by the equation, 
 Zn(NH2)2 + 2KNH2 = Zn(NHK)2 -f 2NH3. 
 
 An ammonoplumbite of potassium^ corresponding to the aquo 
 plumbite of potassium has also been prepared. 
 
 It has been further found in this laboratory that potas- 
 sium amide in liquid ammonia solution reacts with cuprous 
 imide to form an ammonocuprite, ^ with thaUium nitride to 
 form an ammonothallite^ and with magnesium amide to form 
 an ammonomagnesate,^ three compounds of the ammonia 
 system whose aquo analogs are unknown. 
 
 J. Object of this Investigation. — The work here described 
 
 * Jour. Am. Chem. Soc, 29, 660 (1907). 
 ' Ibid., 29, 1274 (1907). 
 
 ' Jour. Phys. Chem., 15, 509 (191 1). 
 
 ^ Jour. Am. Chem. Soc, 34, 1501 (1912). 
 
 * Jour. Phys. Chem., 16, 682 (1912). 
 
 « Jour. Am. Chem. Soc, 35, 1455 (1913). 
 
Potassium Amide and Salts of Cadmium, Etc. 539 
 
 was undertaken for the purpose of studying the action of 
 Hquid ammonia solutions of potassium amide on certain 
 salts of cadmium, nickel and chromium with the object in 
 view of adding several new metalUc amides, imides or nitrides 
 to the Hmited number of such compounds already known, and 
 to determine whether cadmium, nickel and chromium com- 
 pounds similar to the ammonozincate mentioned above might 
 be prepared. 
 
 II. Manipulation of Liquid Ammonia Solutions and De- 
 scpiption of Apparatus Used 
 
 Since liquid ammonia has a low boiling point, special 
 forms of apparatus must be used to control the high pressures 
 which result at ordinary temperatures. A brief description 
 of the apparatus and manipulation follows : 
 
 A reaction tube of the form shown in Fig. i is connected 
 with a cylinder of liquid ammonia by means of a lead tube (e) 
 and a sealing wax joint (c). The reaction tube is thoroughly 
 dried by heating while a stream of ammonia gas passes through 
 first one branch and then the other. While the gas is still 
 flowing, (a) is corked and the required amount of potassium 
 is inserted at (b) by cutting off portions of the potassium tube^ 
 previously prepared and calibrated. A small amount of 
 platinum black is dried and added, after which the cork is 
 transferred from (a) to (b) and the small tube is sealed off 
 
 ^ In order to purify the potassium which is employed in these reactions 
 a glass tube {bd), Fig. 2, about two centimeters in diameter is drawn down and 
 welded to a long tube having a diameter which will permit its introduction into 
 the reaction tube. The slender tube is fused shut at (a) and a loosely fitting 
 glass plug is introduced at (d), nearly closing the opening. Pieces of potassium 
 are removed from the oil in which they are kept, dried between pieces of ab- 
 sorption paper and dropped into the large tube. Enough is added to fill the 
 slender tube when molten. A one-hole rubber stopper provided with a piece 
 of glass tubing closes the opening at (b) and the apparatus is connected to a 
 suction pump by means of heavy walled tubing. After a good vacuum has been 
 produced the apparatus is heated from (a) to (b) until the potassium is molten. 
 Air is allowed to enter at (b) whereby the pure liquid metal is forced into the 
 slender tube, impurities having been caught by the plug at (d). The tube 
 {ad) is removed and calibrated by weighing a measured length before and after 
 dissolving the potassium in alcohol. 
 
540 
 
 George S. Bohart 
 
 as illustrated at (6), Fig. 3. The gas pressure necessary to 
 give the seal a rounded end of uniform thickness is obtained 
 by momentarily closing the opening at (a) with the finger 
 while the glass is soft. A slender glass tube containing the 
 metalUc salt which is to react with potassium amide is now 
 
 Liquid 
 ammonia 
 reservoir g 
 
 5ucfion i fn 
 
 FIG. 5 
 
 introduced through (a) and its contents forced into the main 
 apparatus with the aid of a fairly snugly fitting glass rod. 
 The opening (a) is then corked at the same time as the key 
 of the stopcock is removed. The leg (a) is sealed off and 
 blown into shape by carefully placing a finger and thumb over 
 the openings left by the removal of the stopper after which 
 the latter is replaced. 
 
Potassium Amide and Salts of Cadmium, Etc. 541 
 
 By placing the reaction tube (Fig. 3) in ice water and 
 opening the valve of the cylinder, ammonia distils over and 
 condenses in both legs. When the liquid first comes in con- 
 tact with the potassium a bright orange or fiery red color is 
 produced which gives place to a deep blue solution with greater 
 dilution. A rapid evolution of hydrogen gas occurs in ac- 
 cordance with the equation: 
 
 2K + 2NH3 = 2KNH2 + H2 
 The platinum black greatly increases the speed of the reaction, 
 reducing the time required for completion from weeks or 
 months down to a half hour or less depending on the amount 
 and efficiency of catalyzer used. When the reaction is com- 
 plete the solution possesses a transparent, pale yellow ap- 
 pearance. 
 
 Upon pouring the potassium amide solution into the 
 solution of the metalUc salt, the action between the amide 
 and the salt may be observed. If the product is relatively 
 insoluble it may be obtained free from other compounds 
 formed in the reaction by repeated washing with pure liquid 
 ammonia. This is accompHshed by placing the leg (a), 
 Fig. 3, in ice water while (6) is immersed in tepid water. 
 Pure ammonia distils over and after stirring and allowing the 
 precipitate to subside the supernatant Uquid is decanted back 
 into (6) . Three or four washings are sufficient for a crystalline 
 product, but it is often necessary to repeat the operation 
 fifteen to twenty times when a flocculent substance is being 
 washed. 
 
 After the precipitate has been thoroughly washed in this 
 manner the stopcock is opened slightly to allow the am- 
 monia to slowly escape. When no more gas escapes the ap- 
 paratus is connected to the ammonia reservoir by means of 
 a "T" tube. Fig. 4, and a slow flow of gas is started to pre- 
 vent any air finding its way into the reaction tube. The stop- 
 cock is now opened wide and the leg containing the washings 
 is sealed off at its upper end. The other leg which contains 
 the product is evacuated and weighed. By placing the nozzle 
 of the stopcock beneath the surface of the solvent to be used 
 
542 George S. Bohart 
 
 and opening the stopper, liquid is drawn in. With the aid 
 of the apparatus shown in Fig. 5 the solution of the compound 
 is drawn into the flask (a). The reservoir (6) is employed 
 for safety. 
 
 The solution is now removed to a calibrated flask and 
 later divided into any desired number of aliquot parts, while 
 the tube is washed first with alcohol, then with ether and 
 finally evacuated and weighed. The difference between the 
 two weighings of this tube is, of course, equal to the weight 
 of the compound. 
 
 III. Action of Potassium Amide on Cadmium Salts 
 I. Potassium Ammonocadmiate , Cd{NHK)i.2NHz. — Con- 
 sidering the fact that cadmium hydroxide is not known to 
 possess amphoteric properties it was somewhat of a surprise 
 to find that a compound represented by the above formula, 
 instead of the amide, imide or nitride, results from the treat- 
 ment of a soluble salt of cadmium with an excess of potas- 
 sium amide in liquid ammonia solution. The behavior of 
 cadmium was found to follow that of zinc in this respect. 
 Cadmium iodide with ammonia of crystallization, 
 .Cdl2.4NH3, is almost insoluble in liquid ammonia, but when 
 crystals of this substance are brought into contact with an 
 excess of potassium amide solution they are gradually re- 
 placed by a light, flocculent mass which subsides very slowly 
 and incompletely. After washing this substance thoroughly 
 to remove the soluble potassium iodide formed in the reaction, 
 it is dissolved in dilute hydrochloric acid and removed from 
 the tube in the manner described earlier in this paper. 
 
 In the preparation of Samples I, II and III, cadmium iodide 
 was used. For preparing Sample IV potassium cyanocad- 
 miate, on account of its ready solubility, was substituted 
 for cadmium iodide. When treated with an excess of potas- 
 sium amide, the double cyanide yields a white precipitate 
 closely resembling that obtained with the use of cadmium 
 iodide. All the preparations were heated to 50° in vacuo 
 before removal from the preparation tube for analysis. 
 Analytical results : 
 
Potassium Amide and Salts of Cadmium, Etc. 543 
 
 I. One-fourth of the specimen which weighed 0.4070 
 g gave 0.0455 S Cd by electrolysis. One-half gave 0.0442 
 gN. 
 
 II. One-fourth of 0.2221 g of substance gave 0.0455 K 
 CdS04; another fourth gave 0.0118 g N; and one-half gave 
 0.0769 g K2SO4. 
 
 III. One-half of 0.2978 g of substance gave 0.1230 g 
 CdS04 and 0.1038 g K2SO4. One-fourth gave 0.0161 g N. 
 
 IV. One-hah of 0.2841 g of substance gave 0.1193 g 
 CdS04 and 0.0959 g K2SO4. One-half gave 0.0305 g N. 
 
 
 Calculated for 
 Cd(NHK)2.2NH3 
 
 Found 
 
 
 I 
 
 II 
 
 III 
 
 IV 
 
 Cd 
 
 N 
 K 
 
 44.1 
 22.0 
 30.7 
 
 44-7 
 21 .7 
 
 44.2 
 21.3 
 311 
 
 44- 5- 
 21 .6 
 
 313 
 
 45-2 
 21.5 
 30.3 
 
 The results of these analyses thus show the empirical 
 formula of the compound to be CdN4H8K2. The compound 
 may be represented by the formulas: Cd(NHK)2.2NH3, 
 Cd(NH2)2..2KNH2, or, after Werner, Cd(NH2)4K2. The re- 
 actions involved are represented by the equations : 
 
 Cdl2 -f 4KNH2 = Cd(NHK)2.2NH3 -\- 2KI 
 K2Cd(CN)4 + 4KNH2 = Cd(NHK)2.2NH3 + 4KCN. 
 
 Potassium ammonocadmiate has been obtained as a white, 
 flocculent material which turns somewhat gray under the 
 influence of light. It is insoluble in liquid ammonia and shows 
 no tendency to assume a crystalline form as does potassium 
 ammonozincate. When brought into contact with water it 
 reacts with the generation of considerable heat and the forma- 
 tion of ammonia, potassium hydroxide and cadmium hydroxide 
 as represented by the equation : 
 
 Cd(NHK)2.2NH3 -\- 4H2O = Cd(0H)2 -\- 2KOH + 4NH3 
 
 2. Cadmium Amide, Cd{NH 2)2- — When either cadmium 
 sulfocyanate or potassium cyanocadmiate in solution in liquid 
 
544 
 
 George S. Bohart 
 
 ammonia is treated with potassium amide in an amount not 
 exceeding one equivalent, a white precipitate forms which 
 settles rather rapidly. After prolonged washing by decanta- 
 tion it begins to disperse throughout the liquid in a colloidal 
 condition. This tendency is probably due to the fact that 
 the concentration of the electrolyte has been reduced almost 
 to zero by the washing process. 
 
 Three of the specimens of cadmium amide analyzed were 
 prepared from cadmium sulfocyanate. Sample IV was ob- 
 tained by the action of potassium amide on potassium cyano- 
 cadmiate. Both of these cadmium salts are abundantly 
 soluble in liquid ammonia. The preparations were heated 
 in vacuo to 80° and then dissolved in dilute hydrochloric 
 acid preparatory to analysis. 
 
 Analytical results : 
 
 I. One-fourth of 0.7520 g of substance gave 0.2728 g 
 CdS04 and another fourth gave 0.0337 g N. 
 
 II. One-half of 0.1365 g gave 0.1004 g CdS04. The other 
 half gave 0.0124 g N. 
 
 III. One-half of 0.1232 g of substance gave 0.0113 g N 
 and the other half gave 0,4910 g Cd by electrolysis. 
 
 IV. One-half of 0.3536 g. gave 0.2549 g CdS04. The other 
 half gave 0.0323 g N. 
 
 
 Calculated for 
 
 Cd(NH2)2 
 
 Found 
 
 
 I 
 
 II 
 
 III 
 
 IV 
 
 Cd 
 
 N 
 
 77.8 
 19.4 
 
 78.2 
 18.0 
 
 79-3 
 18.2 
 
 79-7 
 18.3 
 
 77-7 
 18.2 
 
 It, therefore, appears that cadmium amide is formed 
 by the action of potassium amide on a solution of a salt of 
 cadmium in accordance with reactions represented by the 
 equations : 
 
 Cd(CN)4K2 -h 2KNH2 =' Cd(NH2)2 + 4KCN 
 Cd(SCN)2 + 2KNH2 = Cd(NH2)2 + 2KSCN 
 
Potassium Amide and Salts of Cadmium, Etc. 545 
 
 The fact that nitrogen in the above samples runs distinctly 
 low while cadmium shows a tendency to run high suggests 
 that a small amount of cadmium imide or cadmium nitride 
 may have been present in each specimen. 
 
 When the dry amide of cadmium is exposed to moist air 
 it immediately assumes an orange color which gradually fades 
 to the snow white of cadmium hydroxide. The yellow ap- 
 pearance may be due to the initial formation of cadmium 
 oxide or possibly of a mixed base of the formula HO — Cd — NH2. 
 When pieces of cadmium amide come in contact with water 
 they dance about on the surface of the Uquid much as sodium 
 does but without sufficient rise in temperature to produce 
 incandescence. When heated suddenly to a high tempera- 
 ture, one sample exploded, coating the glass in the heated 
 region with a mirror of metallic cadmium. 
 
 3. Cadmium Nitride, CdzN^} — When cadmium amide 
 is heated to 180° in a vacuum it loses ammonia and is con- 
 verted into cadmium nitride as shown by the following analyses : 
 
 I. One-half of 0.3041 g of substance gave 0.0118 g N 
 and the other half gave 0.2549 g CdS04. 
 
 II. One-half of 0.1064 g o^ substance gave 0.0910 g 
 CdS04. The other half gave 0.00437 g N. 
 
 
 Calculated for 
 CdaNa 
 
 Found 
 
 
 I 
 
 II 
 
 Cd 
 
 N 
 
 92.3 
 
 7-7 
 
 90.4 
 
 7.8 
 
 
 
 92.2 
 8.2 
 
 
 Just as metaUic hydroxides may lose water when heated 
 to form oxides, so cadmium amide undergoes deammonation 
 to form the nitride as represented by the equation : 
 
 3Cd(NH2)2 = CdsNa -f 4NH3 
 
 1 Frantz Fischer and Fritz Schroter [Ber. deutsch. chem. Ges., 43, 1465 
 (1910)] have prepared a black explosive substance, the qualitative analysis of 
 which led them to believe they had cadmium nitride in their hands. 
 
546 George S. Bohart 
 
 Cadmium nitride is a black, apparently amorphous 
 substance which instantly assumes an orange color when ex- 
 posed to moist air. The yellow color later gives place to 
 white due to the formation of cadmium hydroxide, A small 
 sample of the nitride exploded violently when it came in 
 contact with water. Small fragments of the glass container 
 picked up after the explosion were found to be covered on 
 one side with a mirror of metallic cadmium. 
 
 IV. Action of Potassium Amide on Potassium Cyano- 
 
 nickelate 
 
 Attempts to prepare a pure ammono derivative of nickel 
 by treating ammonated nickel iodide with potassium amide 
 resulted in failure. The difficultly soluble, blue crystals of 
 the nickel salt were changed to a red, granular mass but 
 analyses showed the product to be a mixture of two or more 
 compounds which could not be separated. 
 
 A search for a nickel compound which could be obtained 
 in the anhydrous condition and which would be at the same 
 time more soluble in liquid ammonia than nickel iodide, led 
 to the discovery that potassium cyanonickelate could be 
 employed. In order to obtain potassium cyanonickelate free 
 from potassium carbonate, with which it is often contaminated^ 
 the following method was devised : 
 
 I. Preparation of Pure Potassium Cyanonickelate, 
 Ni{CN)4K2. — Nickel sulfate is treated with enough potassium 
 cyanide to form the double cyanide. The mixture of the 
 cyanide and potassium sulfate in solution is then evaporated 
 to dryness and the residue extracted with Hquid ammonia 
 in a vacuum jacketed beaker. Potassium sulfate and any 
 potassium carbonate which may have been present in the 
 potassium cyanide are entirely insoluble, whereas potassium 
 cyanonickelate dissolves in about its own weight of the solvent. 
 After filtering with the aid of a vacuum jacketed funnel and 
 evaporating the ammonia from a Dewar beaker receiver, the 
 salt is obtained pure as a light yellow, crystalline residue. 
 
 The following described compounds have been obtained 
 
Potassium Amide and Salts of Cadmium, Etc. 547 
 
 as the result of treating potassium cyanonickelate in solution 
 in liquid ammonia with potassium amide. 
 
 2. Compound No. i. A Complex Product of the Empirical 
 Formula NizN2HiK^{CN)^.8NHz and its Deammonation Pro- 
 duct, Ni3N<iH2K4{CN)e. — ^When the ammono base potassium 
 amide is added to a large excess of potassium cyanonickelate, 
 a brownish red solution results which after standing fifteen 
 minutes to a half hour yields a crop of rather large, brownish 
 red, prismatic crystals which have been found to have the 
 composition represented by the empirical formula Ni3NioH26- 
 K4(CN)6. The crystals readily lose ammonia and crumble 
 to a light yellow powder having the composition represented 
 by the formula Ni3N2H2K4(CN)6. 
 
 In order to determine the amount of ammonia of crys- 
 tallization thus liberated, each leg of the reaction tube pre- 
 viously described is placed in a bath of Hquid ammonia and 
 after connecting with the apparatus shown in Fig. 4 and 
 opening the stopcock, the leg containing the washings is 
 sealed off. While the leg containing the pure compound is 
 still immersed in the ammonia bath the stopcock is connected 
 to an air pump and ammonia is removed until the liquid phase 
 has disappeared. At the temperature of an open bath of 
 liquid ammonia the vapor tension of the compound 
 Ni3NioH26K4(CN)6 is almost zero. When the manometer 
 shows that a constant low pressure has been reached the stop- 
 cock is closed and the tube is removed to a balance and 
 weighed. It is then connected with the air pump and evacu- 
 ated at 70°. The loss of weight represents the amount of 
 ammonia of crystallization. 
 
 The analysis of the deammonated residue offered some 
 difficulties at first but these were finally overcome by the 
 following procedure: A silver nitrate solution acidified with 
 nitric acid was introduced into the tube containing the sample, 
 whereby the latter was decomposed according to the equation : 
 
 Ni3N2H2K4(CN)6 + 6HNO3 -I- 6AgN03 = 3Ni(N03)2 + 2NH4NO3 -|- 
 4KNO3 + 6AgCN 
 
54^ George S. Bohart 
 
 With the aid of the apparatus described in Fig. 5 the solution 
 containing the silver cyanide in suspension was drawn into 
 a small flask. The silver cyanide was filtered off, dried and 
 weighed. The excess of silver was precipitated from the fil- 
 trate as the chloride and removed by filtration. In order to 
 eliminate nitric acid this filtrate was treated with an excess 
 of sulfuric acid and evaporated until sulfuric acid fumes be- 
 gan to appear. The solution was then diluted and divided 
 into two equal portions. In one half nitrogen was determined; 
 from the other half nickel was precipitated electrol3rtically 
 and potassium determined from the residual solution as 
 potassium sulfate. The same method was successfully ap- 
 plied in the analysis of the two nickel compounds, the de- 
 scription of which is given below. 
 
 In the following analytical data, Nos. Ill and IV were 
 obtained from the analysis of the compound containing am- 
 monia of crystallization, while I, II, V and VI represent 
 analyses of the deammonated salt. The deammonated prod- 
 uct was heated to about 70° in vacuo preparatory to anal- 
 ysis. 
 
 Analytical results : 
 
 I. The specimen which weighed 0.5930 g gave 0.0319 g 
 N and 0.2017 g Ni. 
 
 II. The specimen weighed 0.3906 g and gave 0.4880 g 
 Ag from the decomposition of AgCN. One-half of 0.3906 g 
 gave 0.0664 g Ni. 
 
 III. Dried in vacuo at — 40° the specimen weighed 1.0758 
 g. After heating to 70° the residue weighed 0.8522 g and 
 gave 1.3280 g AgCN. 
 
 IV. Dried at — 40°, the specimen weighed 0.6417 g. The 
 deammonated residue weighed 0.51 15 g and gave 0.0274 S N 
 and 0.1743 g Ni. 
 
 V. The deammonated specimen weighed 0.5528 g and 
 gave 0.8560 g AgCN. One-half of 0.5528 g gave 0.0938 g 
 Ni and 0.1871 g K2SO4. The other half gave 0.0147 S N. 
 
Potassium Amide and Salts of Cadmium, Etc. 549 
 
 NHo 
 
 
 Calculated for 
 
 Ni8N2H2K4(CN)« 
 
 Found 
 
 
 I 
 
 II 
 
 III 
 
 IV 
 
 V 
 
 Ni 
 N 
 K 
 CN 
 
 33-9 
 
 5-4 
 
 30.1 
 
 30.1 
 
 340 
 5-4 
 
 340 
 30.1 
 
 30.2 
 
 34- 1 
 5-4 
 
 33-9 
 
 5-3 
 
 30.4 
 
 30.0 
 
 Calculated for 
 8NH3 
 
 20.8 
 
 20.8 
 
 20.3 
 
 The analyzed preparations of this nickel compound 
 were made up of rather large, brownish red, prismatic crystals 
 of very uniform size and shape. The analytical results 
 clearly indicate their purity. The crystalline substance of 
 the formula Ni3N2H2K4(CN)6.8NH3 shows a very slight 
 vapor tension at — 40° but at ordinary temperature all of the 
 ammonia of crystallization escapes leaving a straw yellow 
 powder of the composition represented by the formula 
 Ni3N2H2K4(CN)6. When the yellow product is brought into 
 contact with water it dissolves with surprising rapidity but 
 without the evolution of a noticeable quantity of heat. When 
 subjected to a temperature of 100° it begins to blacken and 
 decompose. The crystalHne compound is sufficiently soluble 
 in liquid ammonia to give the solution a distinct yellow color. 
 
 While there can be no doubt of the existence of definite 
 compounds of the empirical formulas, Ni3N2H2K4(CN)6.8NHa 
 and Ni3N2H2K4(CN)6, the question of their constitution is 
 a matter which has not been satisfactorily solved. Following 
 are possible formulas: 
 
 K4Ni(CN)6.6NH3.2Ni(NH2)2 and K4Ni(CN)6.2NiNH or 
 K2(CN)3Ni— NH— Ni— NH— Ni(CN)3K2.8NH3and 
 K2(CN)3Ni— NH— Ni— NH— Ni(CN)3K2 
 
 3. Compound No. 2. A Mixed Cyanonickelate-Ammono- 
 nickelate of Potassium, K{CN)2NiNHK. — When potassium 
 amide and potassium cyanonickelate in Uquid ammonia solu- 
 tion are brought together in approximately equimolecular 
 
550 
 
 George S. Bohart 
 
 quantities a bright yellow, curdy precipitate instantly ap- 
 pears. It is necessary to wash this substance very rapidly, 
 because if the amount of the nickel salt is too great the de- 
 sired compound becomes contaminated with the compound 
 described above, whereas if an excess of potassium amide is 
 used the compound No. 3, described below, comes down with 
 the product to be isolated. In spite of the greatest care, 
 small amoimts of these compounds did contaminate samples 
 which were analyzed and the results are somewhat variable 
 on that account. The substance was prepared for analysis 
 by heating in vacuo to 80° and then treating in a manner 
 described for the analysis of the above compound. No. i. 
 Analytical results : 
 
 I. The specimen which weighed 0.3385 g gave 0.3468 g 
 Ag from the decomposition of AgCN. One-half of the speci- 
 men gave 0.0107 g N and the other half gave 0.0472 g Ni. 
 
 II. The specimen weighed 0.1980 g and gave 0.2020 g 
 Ag from AgCN. One-half gave 0.0268 g Ni and 0.0837 S 
 K2SO4. The other half gave 0.00608 g N. 
 
 III. The specimen weighed 0.3130 g and gave 0.4135 g 
 AgCN. One-half gave 0.0427 g Ni. One-fourth gave 0.0672 
 g K2SO4 and another fourth gave 0.00533 g N. 
 
 
 Calculated for 
 
 NiNHK2(CN)2 
 
 Found 
 
 
 I 
 
 II 
 
 III 
 
 Ni 
 
 N 
 
 K 
 
 CN 
 
 28.8 
 
 6.8 
 
 38.4 
 
 25-5 
 
 27.9 
 6.3 
 
 24.7 
 
 27.1 
 
 6.1 
 
 38.0 
 
 24.6 
 
 27-3 
 
 6.8 
 
 38.6 
 
 25.6 
 
 While the analytical data are not as concordant as might 
 be^desired there can be scarcely any doubt that the products 
 analyzed were specimens of a compound having the empirical 
 formula indicated. The constitution of the conpound seems 
 fairly clear. It is potassium cyanonickelate which has, so 
 
Potassium Amide and Salts of Cadmium, Etc. 55 1 
 
 to speak, been half converted into potassium ammononickelate 
 as represented by the equation : 
 
 K(CN)2Ni(CN)2K + 2KNH2 = K(CN)2NiNHK + NH3 + 2KCN 
 
 When first precipitated this compound has a bright, 
 lemon-yellow color and presents a curdy appearance. It 
 settles rather rapidly and after one or two washing crumbles 
 to a finely divided granular material. When brought into 
 contact with water it dissolves with mild sputtering and the 
 evolution of a slight amount of heat. 
 
 4. Compound No. 3. A Complex Compound of the Em- 
 pirical Formula NisNiiH^iK-jiCN)^. — When potassium cyano- 
 nickelate is treated with a large excess of potassium amide 
 an emerald-green solution results which after a few minutes 
 changes to a deep red color. At the end of an hour or so 
 crystals begin to appear on the walls of the tube in which the 
 reaction has taken place and after the lapse of about twelve 
 hours the solution becomes almost colorless while a crop of 
 red crystals adhere to the glass. After washing the crystals 
 were treated for analysis in a manner described under Com- 
 pound No. I. 
 
 Since the analysis of this compound led to such an ex- 
 traordinary formula it was considered advisable to determine 
 carbon and hydrogen in two samples by combustion. The 
 analyses given in III and IV were made by Mr. L. D. KUiott 
 of this laboratory to whom the writer expresses his obligations. 
 Previous to analysis the specimens of this compound were 
 heated to 70° in vacuo. 
 
 Analytical results : 
 
 I. A specimen weighing 0.3750 g gave 0.1565 g AgCN. 
 One-half of the specimen gave 0.0426 g N and the other half 
 gave 0.0476 g Ni and 0.1684 § K2SO4. 
 
 II. A specimen weighing 0.4033 g gave 0.1281 g Ag from 
 the decomposition of AgCN. One-half gave 0.0526 g Ni and 
 0.1803 g K2SO4. The other half gave 0.0468 g N. 
 
 III. A specimen weighing 0.0781 g gave 0.0107 Z CO2 
 and 0.0227 g H2O by combustion. 
 
552 
 
 George S. Bohart 
 
 IV. The specimen weighing 0.1605 g gave 0.0208 g CO2 
 and 0.0454 S H2O by combustion. 
 
 
 
 
 Found 
 
 
 
 Calculated for 
 
 Ni3NiiH22K7(CN)2 
 
 
 
 
 
 
 
 
 
 
 
 
 I 
 
 II 
 
 Ill 
 
 IV 
 
 Ni 
 
 259 
 
 25-4 
 
 26. 1 
 
 — 
 
 — 
 
 N 
 
 22 
 
 7 
 
 22.8 
 
 23.1 
 
 — 
 
 — 
 
 K 
 
 40 
 
 4 
 
 40.3 
 
 40.2 
 
 — 
 
 — 
 
 CN 
 
 7 
 
 7 
 
 8.1 
 
 7-7 
 
 — 
 
 — 
 
 H 
 
 3 
 
 3 
 
 — 
 
 — 
 
 3-2 
 
 31 
 
 C 
 
 3 
 
 5 
 
 — 
 
 — 
 
 3-7 
 
 3-6 
 
 The above concordant analytical results together with 
 the fact that the product was obtained in the form of beautiful 
 crystals must be taken as conclusive ^ proof that a definite 
 compound of the empirical formula indicated has been ob- 
 tained. The reaction involved in its formation is represented 
 by the equation: 
 
 3Ni(CN)4K2 + 11KNH2 = Ni3NuH22K7(CN)2 + loKCN 
 
 Not much can be done in the way of representing the con- 
 stitution of the compound. Of a considerable number of 
 more or less questionable formulas that may be written the 
 formula K(CN)2NiNHK.NH3.K2NNiNKNiNK2.6NH3, is per- 
 haps the most satisfactory in that it represents the substance 
 as a mixed cyanonickelate-ammononickelate, an equimolecular 
 combination of compound No. 2 above ^ and potassium am- 
 mononickelate described below. It may also be represented 
 by the formula Ni(CN)2.2Ni(NH2)2.7KNH2. 
 
 It was found impossible to determine ammonia of crys- 
 tallization in this compound by the usual method of heating 
 and evacuating, on account of the fact that no sharp line of 
 
 1 In view of the fact that all the compounds belonging to the group of 
 which potassium ammonozincate is a typical representative, contain sufficient 
 ammonia to permit their formulation either as salts with ammonia of crystalliza- 
 tion or as molecular compounds of the amides of the two metals, it seems probable 
 that a compound of the formula K(CN)2.NiNHK.NH3 would have been obtained 
 had the preparation No. 2 been dried at low temperature. 
 
Potassium Amide and Salts of Cadmium, Etc. 553 
 
 division exists between the temperature at which pure am- 
 monia comes off and a sHghtly higher temperature at which 
 a mixture of ammonia and another undetermined gas escapes. 
 
 This compound appears as bright, red, skeleton crystals. 
 Through a low power microscope they show evidence of 
 homogeneity. When exposed to moist air they soon become 
 coated with a green film which is probably nickel hydroxide. 
 In contact with water the substance sputters vigorously evolv- 
 ing considerable heat. 
 
 In connection with the three nickel compounds just de- 
 scribed, it is interesting to note the steady decrease in the 
 amount of the cyanide radical as the content of potassium 
 amide increases. 
 
 Compound No. I. 2KCN.2Ni(CN)2.Ni(NHK)2. 
 Compound No. II. 2KCN.Ni(CN)2.Ni(NHK)2. 
 Compound No. III. Ni(CN)2.2Ni(NHK)2.3KHN2.4NH3. 
 
 V. Action of Potassium Amide on Nickel Sulfocyanate 
 
 The remarkable results obtained by the action of potas- 
 sium amide on potassium cyanonickelate led to a search for 
 a soluble nickel salt free from cyanogen in order to avoid the 
 complications encountered in the work above. Finding in 
 ammonated nickel sulfocyanate Ni(SCN)2.4NH3, a readily 
 soluble salt and one which may be easily prepared free from 
 water, it was used in the experiments herewith described. 
 
 1. Ammonated Nickel Sulfocyanate. — -A specimen of a 
 compound which was thought to be tetra- ammonated nickel 
 sulfocyanate, Ni(SCN)2.4NH3,^ was observed to be different 
 from that described by Meizendorff. Upon investigation 
 it was found to have the composition represented by the 
 formula Ni(SCN)2.3NH3. A further search led to the dis- 
 covery of three additional ammonates of nickel thiocyanate; 
 one having two, another five and a half, and a third, eight 
 and a half molecules of ammonia. 
 
 2. Nickel Sulfocyanate with Four Molecules of Ammonia, 
 Ni{SCN)2-4NHz. — Preparation: If a hot saturated solution 
 
 Meizendorff: Pogg. Ann., 56, 63 (1842). 
 
554 
 
 George S. Bohart 
 
 of nickel sulfate is treated with an equivalent amount of 
 ammonium sulfocyanate and enough ammonium hydroxide 
 solution to produce a strong odor of ammonia, the color of 
 the solution changes from green to blue, and upon cooUng a 
 crop of crystals of the compound Ni(SCN)2.4NH3 is deposited. 
 
 Nearly all of the nickel sulfocyanate from the mother 
 liquor may be recovered by evaporating to dryness and ex- 
 tracting with a small amount of concentrated ammonium 
 hydroxide solution. The success of this separation depends 
 upon the fact that nickel sulfocyanate is much more soluble 
 in concentrated ammonium hydroxide solution than is am- 
 monium sulfate. 
 
 If a strong ammonium hydroxide solution of nickel sulfo- 
 cyanate is exposed to the air until the excess of ammonia has 
 escaped, most of the solute is deposited. This is to be ex- 
 pected since nickel sulfocyanate is much more soluble in 
 liquid ammonia than in water. 
 
 Axial ratio : a : b : 
 
 c = 1.398 : ] 
 
 : 0.687. |9 = 
 
 75° 41' 
 
 
 
 Angle 
 
 No. measured 
 
 Measured 
 
 Calculated! 
 
 mm 
 
 (no : ilo) 
 
 19 
 
 107° 7' 
 
 107° 7' 
 
 mm 
 
 (no : iio) 
 
 20 
 
 72° 53' 
 
 
 PP 
 
 (hi : in) 
 
 14 
 
 57° 30' 
 
 
 PP 
 
 (in : in) 
 
 9 
 
 122° 39' 
 
 122" 40' 
 
 mp 
 
 (no : in) 
 
 7 
 
 45° 28' 
 
 45 ° 20' 
 
 my 
 
 (no : 201) 
 
 4 
 
 68° 41' 
 
 69° 7' 
 
 my 
 
 (no : 201) 
 
 4 
 
 ni°3' 
 
 110° 53' 
 
 ma 
 
 (no : 100) 
 
 15 
 
 53° 32' 
 
 52,° 33' 
 
 cp 
 
 (001 : in) 
 
 7 
 
 35° 39' 
 
 36° 12' 
 
 mc 
 
 (iio : 001) 
 
 7 
 
 98° 36' 
 
 98° 28' 
 
 bp 
 
 (010 : in) 
 
 2 
 
 62° 25' 
 
 62° 20'^ 
 
 ay 
 
 (100 : 201) 
 
 7 
 
 52° 47' 
 
 53° 7' 
 
 cy 
 
 (001 : 201) 
 
 10 
 
 5i°32' 
 
 51° 12' 
 
 ap 
 
 (100 : in) 
 
 7 
 
 57° 42' 
 
 57° 46' 
 
 ac 
 
 (100 : 001) 
 
 6 
 
 75° 40' 
 
 75° 41' 
 
 ac (loo : ooij 
 
 7 
 
 104° 19' 
 
 
 1 Dull faces resulting from the instability of ammoniated nickel sulfocy- 
 anate in moist air was responsible for the lack of closer agreement between the 
 measiu'ed and calculated values for the angles. 
 
Potassium Amide and Salts of Cadmium, Etc. 555 
 
 Crystals suitable for measurement on the reflection gon- 
 iometer were obtained by making a saturated solution in a 
 liquid made up of one part of concentrated ammonium hy- 
 droxide solution to four parts of water and exposing to the air 
 for twelve hours. 
 
 Crystallography:^ Crystals of this compound belong to 
 the monoclinic system. Prismatic class. A2.P.(C). 
 
 Listof forms:w(iio),;^(iii),:v(2oi),a(ioo),c(ooi), 6(010). 
 The faces shown in Fig. 6 are those which are found on the 
 typical crystal. Faces a (100) and 6(010) are usually either 
 very narrow or missing. Physical properties: Color, sap- 
 phire-blue. Luster, vitreous. Cleavage, perfect parallel 
 to w(iio) and ^'(201). Solubility, slightly soluble in an aque- 
 ous solution of ammonium sulfocyanate but decomposed by 
 pure water. This behavior may be explained by the equation : 
 Ni(SCN)2 + 2NH4OH ^=i Ni(0H)2 + 2NH4SCN 
 
 The solubility in liquid ammonia is very great. In con- 
 centrated ammonium hydroxide solution about 60 g dissolve 
 
 m 
 
 m \ 
 
 ,<C^ 
 
 \Tn 
 
 FIG. 6 
 NifSCN)^'4NH^ 
 
 riG.J 
 NifSCNL'dNH^ 
 
 in 100 cc at 18°. StabiHty : The crystals are unaltered by dry 
 air but rapidly lose their luster and turn green in moist air. 
 
 3. Nickel Sulfocyanate with Three Molecules of Ammonia, 
 
 1 The crystallography given in this paper was done at the suggestion and 
 tinder the direction of Professor A. F. Rogers of the Leland Stanford Junior 
 University. 
 
556 
 
 George S. Bohart 
 
 Ni{SCN)2.3NH3. — Preparation: If a saturated solution of 
 the above described compound, Ni(SCN)2.4NH3, containing 
 a small amount of ammonium sulfocyanate is exposed to the 
 air for a few hours, crystals having a deeper blue color begin 
 to appear and grow at the expense of the lighter blue com- 
 pound if the latter is present. Analysis showed the deeper 
 blue compound to be a substance represented by the formula 
 Ni(SCN)2.3NH3.^ 
 
 In the absence of ammonium sulfocyanate, nickel sulfo- 
 cyanate with three molecules of ammonia does not form. 
 This behavior results from a reaction expressed by the equation : 
 
 Ni(SCN)2 + 2NH4OH ^^^ Ni(0H)2 -f 2NH4SCN 
 Crystallography: Crystals of this compound belong to the 
 orthorhombic system. Bipyramidal class. 3A2.3P.(C) 
 
 Axial ratio: a : b : c = 0.948 : i : 0.702 
 
 
 Angle 
 
 No. measured 
 
 Measured 
 
 Calculated 
 
 bm 
 
 (010 : no) 
 
 37 
 
 46° 31' 
 
 
 be 
 
 (010 : on) 
 
 14 
 
 54° 55' 
 
 55!"; 
 
 mm 
 
 (no : no) 
 
 23 
 
 86° 57' 
 
 86° 58' 
 
 ce 
 
 (ooi : on) 
 
 20 
 
 34° 49' 
 
 34° 46' 
 
 ma 
 
 (no : 100) 
 
 2 
 
 43° 22' 
 
 43° 29' 
 
 cp 
 
 (001 : in) 
 
 5 
 
 45° 16' 
 
 45° 14' 
 
 mp 
 
 (no : in) 
 
 5 
 
 44° 46' 
 
 
 mz 
 
 (no : 114) 
 
 I 
 
 75° 10' 
 
 75° 49' 
 
 List of forms: a(ioo), fc(oio), c(ooi), ^(01 1), w(iio), ^(iii)^ 
 2(114). The faces shown in Fig. 7 are those which are found 
 on the typical crystal. Forms p{iii) and a(ioo) are of com- 
 paratively rare occurrence and 2(114) was found on one crystal 
 only. 
 
 Physical properties: Color, deep blue. Luster, vitreous. 
 Cleavage, absent. Solubihty and stability, similar to nickel 
 sulfocyanate with four molecules of ammonia. 
 
 Analytical results : 
 
 I. A specimen weighing 0.2329 g gave 0.4824 g BaSO* 
 and 0.0526 g NH3. 
 
 ^ Ber. deutsch. chem. Ges., 41, 3178 (1908). 
 
Potassium Amide and Salts of Cadmium, Etc. 557 
 
 II. A specimen weighing 0.1892 g gave 0.3915 g BaS04 
 and 0.0424 g NH3. 
 
 
 Calculated for 
 Ni(SCN)2.3NH3 
 
 Found 
 
 
 I 
 
 II 
 
 SCN 
 NH3 
 
 514 
 22.6 
 
 515 
 22.6 
 
 
 
 515 
 22.5 
 
 4. Nickel Sulfocyanate with Two Molecules of Ammonia, 
 Ni{SCN)2.2NH3. — Preparation: When an aqueous solution 
 of nickel sulfocyanate in which a considerable amount of 
 ammonium sulfocyanate has been dissolved is left exposed 
 to the air for a few weeks, a compound having the composi- 
 tion represented by the formula Ni(SCN)2.2NH3 is formed. 
 The crystals obtained were not suitable for measurement on 
 the reflection goniometer. 
 
 Physical properties: Color, greenish blue. Luster, vit- 
 reous. SolubiHty and stabiUty, similar to tetra-ammoniated 
 nickel sulfocyanate. 
 
 Analytical results : 
 
 I. A specimen weighing 0.2246 g gave 0.5003 g BaS04 
 and 0.0364 g NH3. 
 
 II. The specimen weighed 0.2214 g and gave 0.4949 g 
 BaS04 and 0.0358 g NH3. 
 
 
 Calculated for 
 
 Ni(SCN)2.2NH, 
 
 Found 
 
 
 I 
 
 II 
 
 SCN 
 NH3 
 
 55-5 
 16.3 
 
 55-5 
 16.2 
 
 
 
 55-6 
 16.2 
 
 5. Nickel Sulfocyanate with Five and a Half Molecules of 
 Ammonia, Ni{SCN)2.5^/2NHz. — Preparation: If a concen- 
 trated ammonium hydroxide solution is sattuated with nickel 
 sulfocyanate at about 20° after which the temperature is 
 lowered a few degrees, a crop of large, beautiful, tabular 
 
558 
 
 George S. Bohart 
 
 crystals appear. On account of their efflorescent nature these 
 crystals were not measured on the reflection goniometer. 
 
 Physical properties: Color, blue with a violet tone. 
 Luster, vitreous. SolubiUty, similar to tetra-ammoniated 
 nickel sulfocyanate. Stability, when removed from the 
 mother liquor and exposed to moist air, crystals of this com- 
 pound instantly lose their luster and begin to lose ammonia. 
 
 Analytical results:^ 
 
 I. A specimen weighing 0.2064 § gave 0.0174 g NH3 
 and 0.3580 g BaS04. 
 
 II. A specimen weighing 0.2715 g gave 0.0940 g NH3 
 and 0.4715 g BaS04. 
 
 III. A specimen weighing 0.1586 g gave 0.0551 g NH3 
 and 0.2751 g BaS04. 
 
 IV. A specimen weighing 0.1946 g gave 0.0679 S NH3. 
 
 
 Calculated for 
 
 Ni(SCN)2.5V2NH8 
 
 Found 
 
 
 I 
 
 II 
 
 III 
 
 IV 
 
 SCN 
 NH3 
 
 43-3 
 34-9 
 
 43-2 
 34-6 
 
 43 I 
 34-6 
 
 • 43-2 
 34-8 
 
 34-9 
 
 6. Nickel Sulfocyanate with Eight and a Half Molecules 
 of Ammonia, Ni{SCN)2.8^/2NH3. — Preparation: When a tube 
 containing a liquid ammonia solution of nickel sulfocyanate 
 is immersed in an open bath of liquid ammonia, and ammonia 
 
 1 Working at o° with a specimen of nickel sulfocyanate weighing in the 
 neighborhood of 0.037 S> Walter Peters [Ber. deutsch. chem. Ges., 41, 3178 
 (1908)] obtained results which led him to believe that he had a compound with 
 the composition represented by the formula Ni(SCN)2.6NH3. 
 
 With the hope of obtaining the same compound, a sample of nickel sul- 
 focyanate weighing o. 1 640 g was dissolved in liquid ammonia and the tube con- 
 taining the solution was connected to a suction pump where it was evacuated 
 at 0°. After the pressure had steadily fallen from 21 cm (the vapor tension of 
 the compound Ni(SCN)2.8V2NH3 at 0°) to nearly zero, the specimen was found 
 to weigh 0.2539 g. The formula of the compound calculated from these re- 
 sults would be Ni(SCN)2.5.6NH3, which agrees closely with the formula of the 
 compound having five and a half molecules of ammonia. In spite of careful 
 observation the pressure in the manometer gave no indication of the existence 
 of an ammoniated nickel sulfocyanate having six molecules of ammonia. 
 
Potassium Amide and Salts of Cadmium, Etc. 559 
 
 is removed from the solution until the Hquid phase has dis- 
 appeared, a crystalline residue is obtained. 
 
 Physical properties: Color, similar to the compound 
 having five and a half molecules of ammonia. Stability, 
 at — 40° the vapor tension is about 7.5 cm. At laboratory 
 temperatiu-e the compound rapidly loses ammonia and goes 
 over to the modification having five and a half molecules of 
 ammonia. 
 
 Analytical results : 
 
 A specimen of Ni(SCN)2.5V2NH3 weighing 0.2520 g 
 was dissolved in liquid ammonia and the tube containing the 
 solution was placed in an open ammonia bath and connected 
 to a suction pump. Ammonia was removed until the pressure 
 became constant at about 7.5 cm, after which the tube with 
 its contents were weighed. This entire procedure was re- 
 peated three times. 
 
 Calculated wt. for 
 
 Ni(SCN)2.8V2NH3 
 
 Found 
 
 II 
 
 III 
 
 0.2999 g 0.2991 g 0.2990 g 0.2986 g 
 
 7. Potassium Ammonickelate, NiJSIzK^.dNHz. — ^When the 
 ammonia-soluble nickel sulfocyanate is treated with a large 
 excess of potassium amide a deep red solution is formed from 
 which a red, crystalline product is slowly deposited. After a 
 few hours the liquid becomes nearly colorless. The crop 
 of crystals may be readily freed from soluble impurities by 
 four or five washings with pure liquid ammonia. For analysis 
 the crystals were heated in vacuo to 50° and dissolved in sul- 
 furic acid. 
 
 Analytical results : 
 
 I. One-half of 0.4967 g of the compound gave 0.0633 
 g Ni and the other half gave 0.0673 g N. 
 
 II. One-half of 0.5684 g of substance gave 0.0724 g Ni 
 and 0.2730 g K2SO4. The other half gave 0.0808 g N. 
 
 III. One-half of 0.3253 g sample gave 0.0425 g Ni and 
 0.1558 g K2SO4. The other half gave 0.0449 Z N. 
 
56o 
 
 George S. Bohart 
 
 
 Calculated for 
 NizNgHisKs 
 
 Found 
 
 
 I 
 
 II 
 
 III 
 
 Ni 
 
 N 
 
 K 
 
 25-7 
 27.6 
 42.8 
 
 25-5 
 27.0 
 
 25-5 
 28.4 
 
 43-1 
 
 26. 1 
 27.7 
 42.9 
 
 The compound, to which either of the formulas 
 K2N— Ni— NK— Ni— NK2.6NH3 or 2Ni(NH2)2.5KNH2 may 
 be given, is obviously a member of the same group of compounds 
 to which potassium ammonocadmiate, described above, be- 
 longs. The reaction whereby potassium ammononickelate 
 is formed may be represented by the equation : 
 
 2Ni(SCN)2 + 9KNH2 = NigNgHisKs + 4KSCN 
 
 This compound is obtained in the form of rather small, 
 red crystals resembhng those of Ni3NiiH22K7(CN)2 in general 
 appearance. It is sufficiently soluble in liquid ammonia to 
 give the solution a pale red color. When brought into con- 
 tact with water it reacts vigorously with the evolution of con- 
 siderable heat. 
 
 8. Nickel Amide, Ni{NHz)2. — With a solution of potas- 
 sium amide nickel sulfocyanate in excess yields a red, floccu- 
 lent precipitate. In order to ensure the purity of this sub- 
 stance it must be thoroughly washed. The analyses of the 
 two following samples were made after heating in vacuo to 
 40° and dissolving in dilute sulfuric acid: 
 
 I. One-half of 0.2242 g of substance gave 0,0730 g Ni 
 and the other half gave 0.0335 S N. 
 
 II. One-half of o. 16 13 g gave 0.0522 g Ni and the other half 
 gave 0.0250 g N. 
 
 
 Calculated for 
 
 Ni(NH2)2 
 
 Found 
 
 
 I 
 
 II 
 
 Ni 
 N 
 
 64.7 
 30.9 
 
 65.1 
 29.9 
 
 
 
 64.7 
 310 
 
Potassium Amide and Salts of Cadmium, Etc. 561 
 
 The formation of nickel amide is expressed by the equation : 
 
 Ni(SCN)2 + 2KNH2 = Ni(NH2)2 + 2KSCN 
 
 It is obtained as an insoluble, flocculent, terra-cotta red 
 substance which settles rather rapidly in liquid ammonia. 
 After long continued washing it shows a tendency to go over 
 into the colloidal condition. It reacts rather mildly with 
 water, forming nickel hydroxide and free ammonia. 
 
 p. Nickel Nitride, NisN^. — When nickel amide is heated 
 to about 120° in vacuo a slow evolution of ammonia occurs. 
 Unfortunately, however, a secondary reaction takes place 
 to a certain extent whereby free nitrogen is liberated. In 
 the analysis of Samples I and II given below it will be seen 
 that nickel runs high while nitrogen runs low. The nitrogen 
 given off in the above-mentioned secondary reaction was 
 measured in Sample III. 
 
 I. One-half of 0.1572 g of substance gave 0.0693 8 Ni. 
 The other half gave 0.00701 g N. 
 
 II. One-half of 0.1682 g gave 0.1981 g NiS04. The other 
 half gave 0.00837 g N. 
 
 III. One-half of 0.0864 § gave 0.00646 g N. The other 
 half gave 0.047 S NiO. The nitrogen gas collected in an eudi- 
 ometer measured 3.75 cc over water at 23° and 760 mm. 
 
 
 Calculated for 
 
 NisNi! 
 
 Found 
 
 
 I 
 
 II 
 
 III 
 
 Ni 
 N 
 
 86.3 
 13-7 
 
 88.2 
 8.9 
 
 89.2 
 lO.O 
 
 85.5 
 
 14.9 
 
 The reactions whereby nickel nitride is formed from the 
 amide is analogous to the formation of nickel oxide from 
 nickel hydroxide and is represented by the equation : 
 
 3Ni(NH2)2 = Ni3N2 + 4NH3 
 
 Nickel nitride is a black, apparently amorphous substance 
 which reacts with water very slowly if at all. It dissolves 
 
562 George S. Bohart 
 
 slowly in dilute acids producing ammonium and nickel salts 
 of the acid used. 
 
 NisNs + 8HC1 = 3NiCl2 + 2NH4CI. 
 At about 120° it slowly decomposes into its constituent ele- 
 ments. 
 
 VI. Action of Potassium Amide on Ammonium Chromium 
 Sulfoeyanate, NH4Cr(SCN)4.2NH3 
 
 Since the double ammonium chromium sulfoeyanate is 
 very soluble in liquid ammonia an attempt was made to de- 
 termine the effect of potassium amide on its solution. Small 
 additions of the ammono base cause the separation of a dense, 
 wine-red, gelatinous substance. With the addition of further 
 quantities of potassium amide the deep red color of the original 
 solution is completely discharged and a beautiful salmon 
 pink, flocculent precipitate appears. If now a shghtly greater 
 amount of potassium amide is added, the flocculent material 
 takes on a dull purple color. With a large excess of the base 
 the flocculent precipitate dissolves, forming a wine-red solution 
 which later yields a crop of small crystals of the same color. 
 A microscopic examination of these crystals indicate the pres- 
 ence of two different compounds. Several analyses showed 
 this material to be composed of ammono chromites. Not- 
 withstanding numerous attempts it has also been found im- 
 possible to prepare either of the flocculent precipitates men- 
 tioned above in a pure condition. 
 
 VII. Summary 
 
 When treated with potassium amide in liquid ammonia 
 solution, cadmium sulfoeyanate and potassium cyanocadmiate 
 yield either cadmium amide, Cd(NH2)2, or potassium am- 
 monocadmiate, Cd(NHK)2.2NH3, depending upon whether 
 the cadmium salt or the ammono base is in excess. When 
 cadmium amide is heated above 180° it is converted into the 
 nitride. 
 
 Potassium cyanonickelate yields three distinct compounds 
 when treated with potassium amide. With the salt in large 
 excess, a brownish red, slightly soluble, crystalline substance 
 is obtained having the formula Ni3N2H2K4(CN)6.8NH3. At 
 
Potassium Amide and Salts of Cadmium, Etc. 563 
 
 ordinary temperature and pressure the eight molecules of 
 ammonia escape, leaving a straw-yellow powder, of the compo- 
 sition represented by the formula Ni3N2H2K4(CN)6. 
 
 When approximately equivalent amounts of potassium 
 cyanonickelate and potassium amide are brought together, 
 a lemon-yellow, curdy precipitate is formed. After a few 
 washings with liquid ammonia this substance crumbles to 
 a heavy powder having the composition K(CN)2 — Ni — NHK. 
 
 If a large excess of potassium amide is used the lemon- 
 yellow product first formed dissolves, forming a deep red solu- 
 tion which upon standing twelve hours or so yields a crop of 
 deep red crystals having the composition Ni3NuH22K7(CN)2. 
 
 By varying the concentration of ammonia in ammonium 
 hydroxide solutions of nickel sulfocyanate, the following crys- 
 talUne modifications of ammoniated nickel sulfocyanate may 
 be prepared: Ni(SCN)2.2NH3; 3NH3; 4NH3; 5V2NH3. A 
 fifth modification having eight and a half molecules of ammonia 
 Ni(SCN)2.8V2NH3 may be prepared by removing the liquid 
 phase from a liquid ammonia solution of nickel sulfocyanate 
 while the temperature is kept at about — 40°. 
 
 A liquid ammonia solution of nickel sulfocyanate gives a 
 precipitate of nickel amide Ni(NH2)2 when treated with an 
 equivalent amount of potassium amide. Nickel amide is 
 soluble in an excess of potassium amide, however, producing 
 a deep red solution, from which a compound having the 
 formula Ni2N3K5.6NH3 crystallizes out. If heated above 
 120° nickel amide is converted to the nitride. 
 
 When ammonium chromium sulfocyanate, NH4Cr(SCN)4.- 
 2NH3 is treated with varying amounts of potassium amide, 
 several different products appear. On account of the difficulty 
 of getting any one of them in a pure condition they have not 
 been isolated. 
 
 This work was done in the chemical laboratory of the 
 Leland Stanford Junior University at the suggestion and under 
 the direction of Professor E. C. Franklin. 
 
 Stanford University 
 
 California 
 
 April I, 1Q14 
 

 M> »«1*** 
 
 
'^^t^'^m^i-^,k 
 
 UNIVERSITY OF CALIFORNIA LIBRARY 
 
 UNIVERSITY OF CALIFORNIA LIBRARY 
 BERKELEY 
 
 THIS BOOK IS DUE ON THE LAST DATE 
 
 STAMPED BELOW 
 
 Books not returned on time are subject to a fine of 
 50c per volume after the third day overdue, increasing 
 to |1.00 per volume after the sixth day. Books not in 
 demand may be renewed if application is made before 
 expiration of loan period. 
 
 16 1917