The Determination of Silver in Organic Compounds. The Silver Salts of Pentabromophenol Citation Kemp, Archie Reed (1917) The Determination of Silver in Organic Compounds. The Silver Salts of Pentabromophenol. Bachelor's thesis, California Institute of Technology. doi:10.7907/k6gy-t596. https://resolver.caltech.edu/CaltechTHESIS:07082024-223136719 Abstract No abstract. Item Type: Thesis (Bachelor's thesis) Subject Keywords: (Chemistry) Degree Grantor: California Institute of Technology Division: Chemistry and Chemical Engineering Major Option: Chemistry Thesis Availability: Public (worldwide access) Research Advisor(s): Unknown, Unknown Thesis Committee: Unknown, Unknown Defense Date: 1 January 1917 Record Number: CaltechTHESIS:07082024-223136719 Persistent URL: https://resolver.caltech.edu/CaltechTHESIS:07082024-223136719 DOI: 10.7907/k6gy-t596 Default Usage Policy: No commercial reproduction, distribution, display or performance rights in this work are provided. ID Code: 16541 Collection: CaltechTHESIS Deposited By: Benjamin Perez Deposited On: 11 Jul 2024 19:08 Last Modified: 11 Jul 2024 19:13 Thesis Files PDF - Final Version See Usage Policy. 8MB Repository Staff Only: item control page

THESIS The Dete:nnination of Silver in Organic Compounds. The Silver Salts of Pentabromophenol. Arabie Reed Kemp Olaes of Nineteen Hundred and Seventeen Department of Chemistey THROOP COLLEGE OF TECHNOLOGY Pasadena., .California 191'1 1 PART I ~ DETERMINATION OF SILVER IN OBG.A.NIC COMPOUNDS. Introduction ~ring the last f n years some work has been done both in this country and in GeJ:'!118.ey on the silver salts of halogen aubsti- tuted phenols. thus gin.ng rise to some exceedingly interesting results. A.a moat ot the work done coneerned the lower halogen substi- tuted canpounda, it was thought that an investigation of the ailnr salts ot the higher halogenated phenols might prove valn1>1•; A.s this investigation would necessitate a large number of ana.lyaea, a more satisfactory method for the determination of silver in a.uoh canpounds has been studied. lktview of Methods There have been a large number of methods desori bed in the li terature for the estimation of silver in organic compounds, but in every ca.se they are lacking in one or more of the qualities which det ermine their value, namely, speed, accuracy, and applicability. The universal method whioh is described in all of the text 1 book& covering the subJeot is that of Oariua. This method involves heating the silver compound with taming nitric acid in a long,heavy. sealed glass tube, by means of a specially oonstl'lloted i"lu'naoe, at 1Ber., 3, 697 (18?0) 2 about 200°c for several hours. This treatment decomposes the organic part of the compound, thus freeing the silver, whioh in turn is dissolved by the acid. The contents of the tube are then carefully rinsed out into a beaker, the excess of acid neutralized with pure sodium carbonate and the silver precipitated and weighed as halide. This method, although it is so universally used, has serious defects. The greatest objection to it is its extreme tediousness, (as is revealed in the foregoing description.) Moreover the tubes often explode with great violence, thus making the method a dangerous one. Aside from determining silver in explosive substances like silver oxalate, silver aoetylide, silver picrate, etc., and also in oases of a nwnber of compounds which are not completely decomposed by the treatment within the tube, the Carius method has wide applicability. It is also cap~ble of yielding very exact results if extreme care is exercised in the manipulation. There are many other methods which depend upon decomposition of the silver salts, using either oxidizing or re4uoing agents to accomplish it. The method of Dupont and Freundler, 1 which like the Carius method employs acid for the decomposition, deserves mention. This method consists in heating the silver salt in an open vessel with aqua regia, whereby the silver is converted into silver chloride and determined as such. If bromine is present in the compound, they find that the presence of a small amount of a bromide, which generates hydrobromic acid, greatly assists in the decomposition. 1Manuel Oper~toire de Chemie Organique, P. 80 (1898) - Rindl and Simonis. 3 Dupont's and Freundler's method has found little favor with chemists; · this probably being due to its inapplicability. Datta and Fernandesi have shcmn that aqua regia is a very strong chlorinizing agent and produces chlorination in oases of a very large variety of compounds. It is true also that aqua regia does not ef'feot decompo- sition in oases of some compounds which are not halogenated by it. The only method of WJ¥ significance which does not depend upon the decomposition of the compound is that of Vanino 2• This method is based upon the reducing action of formaldehyde in strong alkaline soluticm.. as such. The silver is reduced to the metal and weighed Due to the fact that only soluble silver salts could be analyzed by this method, lt is therefore of little value. The Cyanide-Sulfide Method. The method of determining silver in organic compound& worked out in this laboratory by the author is based upon the separation of the silver :fran a cyanide solution of the silver compound by its precipitation as a sulfide and its estimation as such. Cyanides au.oh as sodium and potassium cyanides have a de- cided action on silver salts in general; even upon those which have only a veey slight solubility such as the halides. thiocyanate, oxe.late, oyanide, etc., fo:nning the quite stable argentt cyanide ion according to the following typical reaction in whioh A represents the organic part of the silver salt molecule A Ag+ 2 ?le.ON lJ. Am. Chem. Soo., 38, 1809-13. 2Ber., 31, 1763. -- Na Ag(ON) 2 + A Na 4 It is seen fran the above equation that in every case, when this reaction takes place, the sodium or potassium salt of the can• pound is formed aocordlng as sodium or potassium cyanide ls employed. In practically every case either the sodium or potassium salt or both are very soluble, giving at the completion of the rea.otion a solution of the alkali salt and the argenticyanide. It might be thought that the determination of silver in the argenticyanide solution would offer considerable difficulties. If only a slight excess of the cyanide solution were used and some substance added which would combine with the silver given by the dissociation of the complex cyanide fm to to:rm an insoluble compound yielding a still lower concentration of silver ion than is present from the argent1cyan1de, the oanplex would be destroyed with the formation of cyanide ion. As silver sulfide is the most insoluble of all the silver salts, a sulfide would therefore be the best suited for thi• purpose. The reaction is as followa: Ba Ag( Cll) 2 + Na2S == iLg2S + 4liaCN The free cyanide generated in the above reaction would dis• solve an appreciable amount of the silver sulfide if' no precautions were taken to prevent it. If, however, an excess of the sulfide solu- tion is added the concentration of the silver ion coming from the silver sulfide would be so greatly depressed due to the common ion effect that this difficulty would be overcome. In some cases the alkali salt formed in the first reaction might be eydrolized to an insoluble precipitate of the acidic can- pound. \mt if a small amount of an alkali is added, this action will be prevented. On~ in the cases of very insoluble weakly acidic can- 5 pounda would this difficulty be expected. The Cyanide-Sulfide method is carried out as :follows:About 0.3000 g,n. of the sample is weighed out into a 400 com. beaker; two or three drops in excess over the theoretical of a o.25 Nor. solution of sodium cyanide are added. The solution is then warmed to assist the solvent action of the cyanide. After the silver salt is dissolved the solution is diluted to 300 ccna. with distilled water. If a precipitate should appear at this point in the operations due to hydrolysis of the alkali salt, a few drops of a dilute alkali solution an added to dissolve it. About 25 com. ot a o.25 nomal sodium sul- fide solution in excess over the theoretical amount required tor complete precipitation of the silver present, aue added slowly with stirring. The solution is then heated and stirred until the silver sulfide precipitate has coagu.la.ted nicely. The silver sulfide is then filtered from the hot solution into a previously weighed Gooch cru.cible, washed c001pletely free from soluble sulfide with distilled water, and finally washed, first with a little alcohol and then ether. The Gooch cru.oible is then transferred to an oven, dried one-half hour at 100°-110°0, cooled and weighed. It is then dried during 15 min. periods until the weight remains constant. More than two weighings are seldan necessary. The percentage of silver is calculated from the weight of the sample and the silver sulfide by the following formula. log10~ : 1.93982 + log10 (wt. of J.g2S) - log (wt. ot sample). 10 In comparison with the other methods of determining silver in organic canpOW1ds, the Cyanide-Sulfide method is an exceptionally 6 speedy one; the entire determination from the time the weighing of the ~amples is started until the Gooch crucibles are placed in the oven seldom talces more than one hour and a half. Working with silver salts which are very soluble in the cyanide solution the author has made the determination within an hour's time. It might be expected that silver sulfide would oxidize when heated in air at 110°. Vlhen washed with alcohol and ether, however, constant weight is maintained by the sulfide during several hours of heating. The cyanide sulfide method is very easy of manipulati.on, greatly excelling the Cariue method in this respect. leaves little to be desired. Its a.ocuracy Working with a 0.3 gm. sa,nwle, 0.1 mg. of silver sulfide is equivalent to 0.03% of silver. If practical, larger samples can be used thus reducing the error coming from errors in weighing. Another decided advantage of the author's method over those which depend upon decomposition of the compound lies in the fa.ct that by the cyanide-sulfide treatment the compound, from which the silver salt is derived, 1s not destroyed but can easily be recovered by merely acidifying the filtrate from the silver sulfide. If the compound is an insoluble solid, it oan then be separated by filtration. If it is soluble or a liquid, some means of extraction with solvents may be employed. The product recovered in the oase of the insoluble penta.- bromophenol showed after only one crystallization a high degree of purity. When mu.oh time has been spent preparing a compound to be in- vestigated in respect to its silver saltst this refinement of the 7 method can well be appreciated. There remains one more faotor in the method to consider: that is its applicability. It is essentially on this factor that its suooess will depend. There are a great variety of organic silver salts derived from acidic organic compounds. Wherever a hydrogen atom of sufficient acidity exists in an organic compound, there is a possibility of salt formation. The largest group of such compounds is, of course, the oar- boxylic acids. Sulphonic acids, acid amids, imids and phenols are other e.xamples of groups of organic compounds from which silver salts a.re derived. The fact that a large percentage of these silver salts are difficultly soluble crystalline or amorphous bodies is what makes them such a very important class of compounds, for they can easily be investigated in the laboratory. Silver salts have been, from the very beginning of organic chemistry, of great assistance in determining the basicity and molecular weight of organic compounds. Coming back now to the application of the Cyanide-Sulfide method to the analysis of such a large variety of silver salts, there are apparently only two possible limiting factors. In the first place some silver salts may exist which are not soluble in the cyanide solution. This could only be expected in cases where the silver ion con- oentration from the salt when covered with cyanide is so low that the silver cyanide complex cannot form. A case of this kind was discovered with the silver salt of diazoaminobenzene which dissolved only slightly in the cyanide solution. As there are but a comparitively small number of such compounds, the applicability of the method is not seriously 8 affected. Another possible difficulty might arise in the case of silver salts, in which very easily reducible nitro groups are present in the molecule, thus making it possible for the dilute sulfide solution to reduce the nitro body to a difficultly soluble substance which might not form a sodium salt on account of its basic character, but would precipitate out and contaminate the silver sulfide. Due to the very slight reducing action of suop a very dilute alkaline solution of the sulfide which is present and the probability that an insufficient number of amino groups would be formed to render the compound too basic for the formation of the sodimn salt, little difficulty could be expected from this source. The Cyanide-Sulfide Method Experimental The method was first tried out by analyzing some rather impure but homogeneous silver pentabromophenolate with the following results: %Ag Sample I 15.48 %Ag Sample II 15.4'1 The fact that such good checks were obtained gave much encouragement so samples of pure silver salts were prepared and analyzed. Analysis of Silver Oxalate The theoretical amount of a dilute solution of c. P. silver nitrate was added slowly to a dilute solution of c. P. oxalic acid which was kept shaken. The white amorphous precipitate of silver 9 oxalate was filtered and thoroughly washed with distilled water upon a Btlchne~ fwlnel. The product was sucked as dcy as possible upon t he filter, separated caretully from the filter paper and allowed to dry for a day on a watch glass which was set in the desk and pro- t ected from dust. It was then powdered and dried over phosphoNa pentoxide in vacuo for several daNs• As silver oxalate is easily decomposed in contact with sunlight, considerable care was exercised t o avoid this difficulty. fhe produot was very slightly darkened which was no doubt due to reduction. Analysis Sample No. I No. II HQ. III Wt. of Sample Wt. of A82S i J.g . Deter. for c2o4',82 o.2832 o.3893 o.2523 0.2311 71.04 0.3178 0.2065 71.09 '11.12 '11.03 71.06 '11.03 i J.g. Ce.lo. The slight reduction of the silver oxalate ~ have caused t he results to run high. J.nalyeis of Silver Benzoa.te The silver benzoate was prepared by dissolving a fn grams ot benaoio acid, containing 0.1 - o.2% ot combined chlorine in a l arge volume of hot distilled water. !he hot acid solution was then carefully neutralized with a dilute sodium ·: hy dr axUle solution con- t aining a small amount o! halogen. A few drops of a. very dilute solu- t ion of silver nitrate was added and the solution filtered free of the silver halide. A slight excess over the theoretioal of a very dilute solution of O.P. silver nitrate was then added slowly with constant shaking. The large flocoulent precipitate of silver benzoate which 10 separated was dissolved in the solution by boiling and adding the necessary amount of distilled water. After the silver bensoate had dissolved, the solution was filtered. Upon cooling. the tine needler· like cyrstals of silver benzoate which separated were recrystallized twice fr(IJl boiling water. the final crystallization yielding a slightly darkened product. This was first dried in air and then over phosphoras pentoxide in vacuo. The product like silver oxalate had to be protected fran the light. Ana~•1• %Ag. Wt. of l't. of Sample .Ag2S Deter. tor 07H502Ag I o.3187 0.1'128 II o.3263 0.4475 0.1762 0.2429 4'1.18 (47.01)* 4'1.12 4'1.12 47.23 4'1.12 Sample Ill '/, Ag. Calo. *During the analysis of aample II a •few specks of silver sulfide were lost. The tact that the results show a high percentage of silver i s probably due to the slight reduction of the silver benzoate. Analysis of Silver Cinnamate Pllre recrystallized cinnamic aoid m.p. 132.5 was oaretu.lly neutralized with dilute sodium hydroxide solution whioh contained a small amount of halide. This neutral solution was freed from sol- uble halide by adding a fn drops of silver nitrate and filtering. From this point the operations were the same as in the preparation of silver oxalate. The product was a snow white amorphous body. 11 Analysis Sample I II III Wt. of %Ag. Wt. of Sample Ag s 0.2866 0.2,22 0.2734 0.1390 0.1321 0.1327 Deter. 2 42.22 42.25 42.26 i Ag. Ca.le. for C9H702Ag. 42.32 42.32 42.32 .Analysis of Silver Anisate Pure anie1o acid m.p. 185°c was employed in the preparation of the silver salt. The operations were the same as those empl«fed in the preparation of silver benzoa.te. The product after five orystalli- za.tions still exhibited a alight pl.Dk color. Anal.yei1 Sample Wt. of Wt. of Sample Ag2S Deter. Calo. for CaB-70rJ,.g. 41.57 41.61 41.65 41.66 41.66 41.66 I 0.2438 0.1164 III 0.2400 0.2314 0.1147 0.1107 II '/, Ag. '/, Ag. Analysis of the Pink Amorphous Silver Salt of Pentabromophenol 1 Sample I ll HI %Ag. Cale. Wt!t of Wt. of Sample 'I, Ag. Ag2S Deter. tor c6:&r50Ag . 0.3043 o.3475 0.4734 0.0627 0.0716 0.0975 17.94 17.94 18.11 17.91 18.11 18.11 Analysis of the White Amorphous Silver Salt of Pentabromophenol2 Preparation I Sample I II III Wt. of Wt. of %Ag. '/, Ag. Oalo. Sample Ag2S Deter. for 0613r50Ag. 0.3918 0.4131 0.4701 o.oso7 17.93 18.11 18.11 18.11 0.0851 0.0965 17.93 17.88 JThesis Part II, p -zz. ~hesie Part 11, p z3 12 Preparation -II Sample I II %Ag. %Ag. Calo. Ag2S Deter. 0.0523 0.0439 17.91 18.06 18.11 18.11 Wt. of Sample Wt. of 0.2544 0.2118 Analysis of the Lemon Yellow Silver Ammonia Salt of Pentabromophenol. l Sample I II III Wt. of Sample Wt. of Ag 2S %Ag. %Ag. Calo. Deter, for C6Br50Ag(N~) 2 0.2907 0.2715 0.2470 0.0570 Q.0532 0.0487 17.07 17.06 17.17 17.10 17.10 17.10 Analysis of the Pink.~ilver Salt of Pentabromophenol. Sample I II III 2 Wt. of Sample Wt. of Ag 2S %Ag. %Ag Qalo. 0.3457 0.4492 o.2so6 0.0781 0.0935 0.0584 18.10 1s.122 18.12 18.ll 18.ll . Deter. ;Thesis, Pa.rt II, P. 21 Thesis, Part II, P. /?. 3 for c 6Br50Ag. 18.11 13 PART II The Silver Salts of Pentabromophenol. Introduction In 1910 Torrey and Hunter1 , working in the Harvard la.borat ory, discovered upon allowing the red silver salt of 2-4-6 tribromophenol to stand over night on a filter paper that it had turned white. Upon analy- sis of the white product results were obtained indicating that it was isomeric with the red fo~n. This change they thought might have been brought about by the action of the ammonia present in the laboratory. They then proceeded to prepare the red and white in a pure condition for analysis. The red silver salt was prepared by precipitating it with silver nitrate from a solution of its sodium salt. The white salt was found to be best prepared by precipitating it with silver nitrate from a solution of its sodium salt to which ahd been added a few drops of ammonia, thus ma.king use of the fact that ammonia affects the transformation from the pink to the white form. The analysis of the two salts so pre- pared proved their identical composition. Two different forms of the same silver salts were discovered by the above investigators in the cases of 3-4-6 tribromoguaiacol and 2-4-6 tribromoresorcinol monomethyl ether, but only the white salts in each case could be obtained for analysis as the red forms changed too rapidly into the white. No other oases in which two forms of the same silver salt existed were discovered. Torrey and Hunter attempted to discover a difference in the chemioal behavior of the two forms of silver tribromophenolate but l . J. Amer. Chem. soc., 33,194-205. 14 failed to find an:y. ( lla.nzsch¼ ~orking in the same field confirmed Torrey's and Hunter's results and added another case of the kind. He also inves- tigated a number of the silver salts of halogenated phenols and prepared the following table which gives the color and forms of the silver salts which he prepared: Silver Salts of: Chlorophenols Bromophenols Lodophenols o-m-p chlorophenol colorless p bromophenol colorless p iodophenol colorless 2-4 dichlorophenol colorless 2-6 dibromophenol colorless 2-6 diiodophenol colorless 2-6 dichlorocresol •• yellow 2-6 dibromo-p-cresol yellow 2-6 diiodo-p-cresol colorless colorless t.t.trachlorophenol yellow 2-4-6 tribromophenol brick red colorless penta chlorophenol yellow tetrabromophenol colorless **2-4-6 triiodophenol colorless *pentabromophenol colorless ** A yell~ silver salt of 2-4-6 triiodophenol has been prepared by Woollett. • *The author has prepared a pink silver salt of pentabromophenol. Hanzsch made a number of generalizations which in view of the observations of later investigators are open to serious criticism. Ta1ring account of the fact that two cases have already been found which do not correspond with the table of Hanzsoh, a systematic investigation of the silver salts. Of ' halogenated phenols should throw more light upon the existence of other isomeric forms. 4i7~ '\ 1 Ber. 40, 4:552 (1907) ) J . .Amer. Chem. SOO• 38, 2474. Tb.ere is no 15 reason to suppose that the silver salt of tetrabromophenol, for instance, does not exist in isomeric forms as it differs from the silver salts of the tri- and pentabromophenols by one bromine atom. Discussion of Results. Two silver salts of pentabromophenol were described in the Bodrowc:1 obtained literature at the time this research was taken up. a rose colored amorphous precipitate upon adding silver nitrate to a solution of the potassium salt. Hanzsch2 obtained a ·r eddish amorphous precipitate in the same manner as Bodroux but mentioned that the color was probably due to the impurity of his pentabromophenol, which after repeated crystallization retained a reddish color. Hanzsch also de- scribed a colorless amorphous form of silver pentabromophenolate which he obtained by precipitation with alcoholic silver nitrate from a solu~ tion of the ammonium salt in alcohol. Hanzsch himself pointed out that alcohol affects the transformation of the colored into the colorless form in the case of other isomeric silver salts but he evidently failed to realize the possibility of such an action in the case or the silver pentabromophenolate, which he describes as existing only in the colorless form. Although two differently colored silver salts of pentabromophenol have been prepared, no one has shown them to be isomeric. Previous to the authprs investigation the silver salts of 3 2-4-6 tribromophenol described by Torrey and Hunter and the silver 1compt. rend., 126, 1282-1285 (1898). 2-Loc. oi-t-. 1: , . ,. 31 oc. en.• t. y_ q i ).. --I I 16 salts of 2,,.5 di bromo- p-c re sol described by Hanz sch are the only oases of the kind mentioned in the literature where the salts have been shown to be isomeric by analysis. A study of the conditions under which the colored and color~ less forms of silver pentabromophenolate are formed has shown that the pink amorphous form, obtained by adding silver nitrate to an aqueous solution of an alkali phenolate, is readily transformed into the white amorphous salt in the presence of ammonia or alcohol and upon standing in a damp condition. The analysis of. both salts has shown them to be isomeric. As Hanzsch has stated that the color in the reddish silver salt of pentabromophenol may be due to an impurity and since the appearance of a paper by Furcht and Liebenl, who describe a white and yellow levu.linate of silver but ascribe the color in the yellow salt to an impurity, it is worth while to state the reasons for believing that the color in the case of the colored silver salt of pentabromophenol is not due to impurities. In the first place the pentabromo- phenol used in the preparation of the colored salt was of exceptional purity and perfectly colorless. Sooie of the colored and colorless amorphous silver salts were converted into pentabromophenol and from these products the colored salt was obtained just as in the case of the freshly prepared phenol. The analysis of colorless and pink salts agree with the theoretical percentages. Moreover the transformation of the pink into the oolorless form takes place without the appearanoe of any colored by-products. lBull. Soo. Chem. (4) 5, 1069, (1909) CHART Pentabromophenol Alkali Salt Solution X )" -~i? ••i ~"' J, '~ .~ ) 0 I '1 , c3, _ri ·._ .,. C ' ) '--', J ., Pink AfixtU:re ,~ ' dt e Ccystalli h µ~ ('~ 73 (£F:/1.'.l • --N ~?.I' ~ ,.) y-,· Lemon Yellow Solution 17 The author has also prepared two crystalline silver salts of pentabromophenol which contain ammonia of addition. These two salts consist of a colorless and a lemon yellow crystalline form. They are best prepared by adding silver nitrate to an ammonia.eal solution of an alkali salt of pentabromophenol; the precipitation of the different forms depend upon the concentration of the ammonia present. The analysis of the lemon yellow salt has shown it to be an armnonia addition product with the molecular fonnula c Br50Ag.2NH3• 6 Analysis shows the white crystalline salt as possessing the following formula: 0 6Br50NH4 , C6Br50Ag.2NH3.H20 . Two semi-crystalline products have been prepared from the above mentioned white crystalline salt. One of these products is white and is obtained by heating the dry white crystalline salt. The other product is pink and is obtained by drying the moist white crystalline salt over phosphorus pentoxide. These products may be definite com- pounds but every indication points to the conclusion that they are mixtures. The accompanying chart shows in a concise way the numerous changes that the silver salts of pentabromophenol undergo. The details of these changes are discussed more fully in the experimental part. The investigation of the amorphous silver salts of pentabromophenol has thus far indicated no difference in their chemical behavior. These salts therefore are apparently analogous to the sil- ver salts of tribromophenol described by Torrey and Hunter. That the color in the case of the pink amorphous form of silver pentabromophenolate cannot be due to an impurity has already been 18 shown. The only way to account for colored and colorless forms of the same silver salt seems therefore to consider them as differing in structure. To consider the salts as exhibiting a qui noid-benzenoid isomerism is the most logical hypothesis to assume as it best accounts for the chemical and physical properties of the isomers. According to this hypothesis the salts would have the following formulas: fly ~,White Bre,,Pink Br1 I I I I c====c c-- c -- 1 ·/ \ 49. _/ .c-.0, O-C Br--C " ~ ~--c I I .!31""' Br' ~ 15 ,--(or-) / C - c \ ~ c ~ c/ I Br' c- O C>r'-C C \A9 '\ / ·1 131" 81"" Br Solutions of the sodium, potassium and ammonium salts prepared by dissolving the purest pentabromophenol in the bases of these metals exhibit a decided yellow color. to the presence of quinoid salts. This color is undoubtly due There is probably a tantomeric equilibrium between the quinoid and benzenoid fDrms of the anion as follows: .Br' _Br- ' I c -c~ ~c-0- 'pr-Cf' "" c=== c. / I i Br- - C--c I 1---....A9 Br' Upon addition of silver nitrate to the solutions of the alkali salts a mixture of the isomers is probably precipitated. Di- rect evidence to this effect is shown by the fact that the solid yellow silver ammonia salt when treated acetic acid changes to a much redder product than is obtained by direct precipitation. In ascribing a structural formula to the lemon yellow sil- 19 ver ammonia salt the remarkable thing to consider is its color. Hanzsch l observed that the isomeric silver salts of tribromophenol absorb amnonia to form the same colorless ammonia addition product. Recently Bruni and Levi 2 have prepared a large number of ammonia addition products of the silver salts of organic acids. These are all described as colorless products. The absorbtion of ammonia by the isomeric silver salts of pentabromophenol gives the same lemon yellow salt. The absorbtion of ammonia by the white amorphous salt is,however,very much slower than in the case of the pink fonn indicating that there must first be a rearrangement to the quinoid structure before the ammonia oan add on readily. It seems rather remarkable that the isomeric silver salts of pentabromophenol should give a lemon yellow ammonia addition product while the closely allied isomeric tribromo- derivatives give a colorless ammonia sal t. If considered as exhibiting a para quinoid stru.oture,the salt lemon yellow silver amnonia/would have the following formula: 1100. cit. 2Gazz. Chem. ital 46, 11 17-42, 235-46 (1916) u ; r ~ ~ 20 The lemon yellow oolor of the above salt and also the yellow color exhibited by solutions of the alkali salts may be cited as evidence supporting the para quinoid arrangement. The analysis and properties of the white crystalline double silver-ammonium salt indicate that its structural formula is as follows: \:>I"' 1 e,r- I C=--=----C 5(-C ""' C- 0- N \-\~ ~c - - C ✓ / I I BY ~(' ' The avove formulas are only proposed tentatively. Further investigations must be carried out before the question of structure of these salts oan be satisfactorily settled. From an investigation of other metallic salts of pentabromophenol it appears that isomerism is not limited to the silver salts. The salts of mercurous mercury and cadmium show strong evidence of existing in isomeric forms. Summary. Part I 1. A review of the methods used for the estimation of silver in or- ganic compounds. 2. A new method for dete~ining silver in organic silver salts based upon the precipitation of the silver as silver sulfide from cyanide solution of the salt and its estimation as such. Part II 1. A short review of the previous work done on the silver salts of 21 halogenated phenols. 2. The preparation and properties of two isomerio and two ammonia addition silver salts of pentabromophenol. 3. A discussion of the possible structural formulae of silver salts of pentabromophenol. 4. Other metallic salts of pentabromophenol. PART II Experimental Part. Preparation of Pentabromophenol. The pentabromophenol was prepared according to the method of Bodroux1 which consisted in treating phenol with bromine in the presence of anhydrous aluminum bromide. The product after bromination was treated with cold 95% ethyl alcohol to extraot the aluminum bromide and the lower halogenated products. The residue which was mainly pentabromo- phenol was dissolved in boiling 95% ethyl alcohol, the solution boiled for about one half an hour with refined bone charcoal, filtered and allowed to cool, whereby, the pentabromophenol crystallized out in long needles. This final process of boiling with charcoal and crystal- lizing was repeated several times, yielding a perfectly snow-white product. One more crystallization of the snow-white product from pure 95% alcohol was fowid to be suffieient to give a preparation having a constant melting point. Separate batches gave (corrected) melting points as follows: 229°; 229.50; 229.3; 229.4; 230; 229.5; 2290; 230°. 1compt. rend, 126, 1282-85 (1898). 22 1 Previous investigators give 225° as the melting points of their preparations of pentabrQmophenol. In taking the melting points the author used two high quality thermometers. One was nitrogen filled; range 200 to 400°c. The other was an Anohutz normal nitrogen filled; range 200° to 250°. Both ther- mometers were calibrated at 21a.1°c with naphthalene vapor. Preparation of the Pink Amorphous Silver Salt of Pentabromophenol. As the sodium salt of pentabromophenol is the most soluble of the alkali salts, the silver salts were prepared from a solution of it made by dissolving the phenol in an equivalent quantity of normal sodium hydroxide which was practically free from chlorides. The solution thus prepared was yellow and alkaline to phenolphthalein. The . pink amorphous silver salt was prepared by adding a slight excess o:f\0.05 normal silver nitrate to a 0.01 normal solution of the sodium salt with frequent shaking. The precipitate after the addition of the silver nitrate was brown due to the presence of a considerable amount of silver oxide. To dissolve out the silver oxide and give a pure product 0.01 normal acetic acid was added carefully until the liquid showed a slight acid reaction with litmus. The precipitate which at this point appeared pink, was shaken up vigorously to coagulate it and filtered upon a Buchner funnel. After filtration the solid was separated oare:fu.lly from the tilter paper, shaken up with about 500 ccm. of distilled water and refiltered. After washing a few times on the filter, the precipitate was sucked as dry as possible, transferred 1Korner, annalen. 137, 210 Bodroux., Canpt. rend., 126, 1282-85 (1898} A'1%1'1ers, Ber 32, 3596. 23 carefully to a watch glass and allowed to dry in a olean desk for a day. It was then pulverized, spread out thin on a watch glass and dried over phosphor-..is pentoxide for several d~s in vacuo. The pro- duot was analyzed for silver by the cyanide-sulfide method. CL,"'-L-. 0.3043 0.3475 0.4'134 I II III %Ag. Cale. %Ag. Wt. of Sample Sample 0.0627 0.0716 0.0975 Deter. -:J,~ ~ 17.94 17.94 17.93 18.11 18.11 18.11 c6Br50Ag. Some of the pink salt was prepared by heating the lemon yellow ammonia addition product for several hours at 120°. Sample Wt. of Sample Wt. of Ag2S %Ag. Calo. %Ag. Deter, for C6Br5 OAg. ..~ t (1,1-1 0.3457 0.4492 o.2ao6 I II III 0.0781 0.0935 0.0584 18.10 18.12 18.12.. 18.11 18.11 18.11 A oomparison of the above analyses shows that the pink salt prepared by the latter method yields the purer product. Preparation of the White Amorphous Silver Salt of Pentabromophenol. This salt was prepared by adding a slight excess of 0.05 normal silver nitrate to a 0.01 normal solution of the sodium salt of pentabromophenol. In this ease, however, the latter solution con- tained ammonium hydroxide of 0.01 normal which affected the transformation of the pink form into the white and at the same time prevented the formation of silver oxide. The product after washing and drying as in the case of the pink amorphous salt was very light cream.y color. Two separate batches gave the following analyses for silver: 24 Sample Wt. of Sample wt. of Ag S %Ag. Deter. 2 I II III IV V 0.3918 0.4131 0.4701 0.2544 0.2118 r %Ag. Calo. -~ :f,, o nt. C6Br50H. 0.0807 0.0851 0.0965 0.0523 17.93 17.88 17.91 18.11 0.0439 18.06 18.11 17.93 18.11 18.11 18.11 Samples I, II, and III, preparation I Samples IV, and V, preparation II Both the pinlc and white amorphous salts are difficultly soluble in water, alcohol, ether, ligroin, benzene, carbon tetraohloride, bromobenzene, etc. They both dissolve in alcohol to which has been added a little 3.mlllonium hydroxide, yielding a deep lemon yellow solution. Both salts seem to reaot identically with ethyl iodide and in the sunlight this action is analogous to the catalytic decomposition observed by Torrey and Hunter! in the case of silver tribromophenolate and later studied in detail by other investigators2. The color changes in the case of the silver pentabromophenolate were, however, different from those observed by Torrey and Hunter. When covered with ethyl iodide the silver pent abromophenolate first became green, then black and finally yellow; at no time was a color observed in the liquid. The products from this reaction have not been studied. Preparation of the White Crystalline Silver-Ammonium Salt of Pentabramophenol. Instead of making the solution of the sodium salt 0.01 normal in respect to ammonium hydroxide as in the preparation of the white amorphous salt, a much larger concentration up to 0.1 or 0.2 normal is used, a beautifully irridescent white c rystalline precipitate aeparates. l Under Loe cit. 2Hunter, Olson and Daniels, J. Am. Chem. Soc., 38, 1761-71; Woollett, J~ Amer. ()hem. soc., 38, 2474. 25 the microscope the crystals appear as very thin plates whi ch revolve rapidly upon their axes as they migrate through the solution. The " crystals were filtered upon a _i;uchner :fu.n.n.el and. washed with o.~ nor- mal ammonium hydroxide . Tha.t this salt in a moist condition is only stable in the presence of a certain vapor pressure of ammonia is shown from the fact tha t when washed with distilled water or dilute acids and when dried over phosphorus pentoxide in vacuo it turns pink. An analysis of the pi nk product in the latter case has shown tha t it contained no nitrogen. The aqueous tension may al so influence the stability of the salt. One batch was dried in a de~ oator over sodium o.wJ... hydroxide and another in a dessioator over a mixture of anhydrous aei-tt hydrated sodium sulfate. In both oases a little ammonia was intro- duced in the dessicator. The dried products from both dessicators re- tained their color and crystalline structure and were apparently stable upon exposure to air. The analyse s of the white crystalline salt for ~ilver from the two des~oators are: Sample I II III IV V kvvv %Ag. Cale. £er. Wt. of Sample Wt . of %Ag. Ag 2S Deter. 0.5913 o.5760 0.7476 0.3474 0.3203 0.0637 0.0621 0.0796 0.0374 0.0342 9.38 9.35 9.39 9.35 9.27 9.35 9.35 9.35 9.37 9.30 C12H1203N3Br1o,Ag 1 ;, I, II, and III were dried over sodium sulfate IV and V were dried over sodium hydroxide. Two nitrogen and two bromine determinations were made on the white crystalline salt dried over sodium hydroxide . 26 Determination of Nitrogen by the Method of Kjedahl. Sample I II Wt. of Sample 1.065 1.319 ocm. of HCl {0.4942N) used 5.32 6.90 %N2 %N2 Calo. for 3.46 3.64 3.64 Deter. 3.62 C12H1203N3BrlOAg. These results indioate three atoms of nitrogen to one of silver. Some of the white crystalline salt dried over sodium hydroJlide was heated in the oven for several hours at 120° until it lost to constant weight with the following results: Wt. of sample 0.6825 gm; loss in weight 0.0430 g. %loss 6.30; %loss calculated for loss of three molecules of ammonia. and one of water Analysis of the White Crystalline for Bromine by the Method of Carius.l Sample I II Wt. of Sample Wt. of Ag 2S %Br· 0.1720 0.1970 0.1834 0.2100 68.76 68.74 Detet 5avY\ %Br1' Cale ~r C12H1z03N3Br10Ag· 69.31 69.31 These results show ten atoms of bromine to one of silver. Some of the pink product obtained by drying the white crystalline over phosphorus pentoxide was anaiyzed for silver. , ft,,• Sample I II Wt. of Sample Wt. of Ag 2S %Ag %Ag. Cale f-er Deter. C1zHO 2Br10 Ag. 0.3515 0.2938 0.0401 9.93 9.93 9.95 9.95 0.0335 These results indicate that three molecules of ammonia and 1The silver bromide taken from the tube which was mixed with broken glass was analyzed for silver by the cyanide-sulfide method. 27 When boiled with a lcohol the white crystalline s alt is broken up y ieldi ng a residu e of the white amorphous silver sa lt and a filtr ate conta ining pent abr omophenol in solution; approximately e qual quantities of the phenol and silver s a lt are obta ined. The same action talces pl ace when t he white cryst alline salt is warmed with dilute sodium hydroxide whereby the filtrate in t his ca se is a solution of the sodium sa lt. The ammonium salt prepared from the pure phenol exhibits a sli ght pi nk color. When hea ted in the oven at 110° it loses its ammonia and gives a snow-white residue of the pure phenol. From the above data it appears that the white crysta lline silver salt of pentabromophenol is a doub~e salt with the formula C5Br50NH4 , C5Br50Ag•2NH3~H2o and that upon loss of water and ammoni a it bre aks up a ccording to the following reaction yielding either t he white or the pi nk amorphous silver salt depending upon conditions: The loss of ammonia from the a1'Jllllonium radical in the double s a lt when dried over phosphorus pentoxide may be due to a g~eater instability than is shown by the annnoniurn sa lt,\ i-tsei f which does not lose ammonia under these conditions. Prepar ation of the Lemon Yellow Silver .P. Jnmoni a Salt of Pentabromophenol. A soluti on of the sodium salt wa s made about 2 normal in respect to ammonium hydroxide, an excess of silver nitrate over the amount necessary to precipit ate the amorphous salts was added and 28 the white cryst a lline salt which separ ated was dissolved by warming the solution. Upon cooling and standing several hours long, lemon yellow needle like crystals separ ated. The crystals were filtered upon a Buchner funnel, washed with 2 normal ammonium hydroxide and dried in a dessicator containing sodiUln hydroxide and an ammonium salt which generated ammonia. The crystals upon drying retained their color and crystalline structure. The dry salt loses its ammo- nia very slowly a t ordina ry temperatures but when heated in the oven at 1000-110° the ammonia is libera ted rapidly and the pink salt is formed as was previously indicated. Like the other silver salts it dissolves in a mixture of ethyl alcohol and ammonia to give the characteristic lemon yellow solution. {See chart). The iemon yellow salt was analyzed for silver by the cyanide sulfide method with the following results. Sample I II III Wt. of Sample Wt. of Ag 2s %ltg. 0.2907 0.2715 0.2470 0.0570 0.0532 0.0487 17.07 17.06 17.17 Deter. %Ag. ~lo. £or C6Br50Ag.2NH3 17.10 17.10 17.10 A sample of the lemon yellow salt was heated in the oven at 1100 until it had lost ammonia to constant weight. Wt. of sample 1.249 gm; wt lost 0.0684 gm.;% loss 5.46; % 1m3 Cale. in C6Br50Ag.2NH3 5.40. Absorbtion of .Ammonia by the Amorphous Silver Salts of Pentabromophenol. As can be seen from the foregoing chart, both the pin1r and white amorphous silver salts of pentabromophenol absorb aunnonia and form the same lemon yellow ammoni a addition product. 29 Samples of the dry white and pink amorphous salts were weighed out upon watch glasses and placed in the same dessicator containing ammonia generated as previously described. The salts were t aken out at intervals and wei ghed. %NH3 in the pinlr salt %NlI3 in the white salt 4.96% 4.96% 4.96% 0.70 0.88 After 3 hrs. After 6-1/2 hrs. After 22 hrs. After 27-1/2 hrs. 4. 9 6% 1.29 1.29 After the salts would no longer absorb ammonia at room temperature, the dessicator wa s placed in an ice chest beside the ioe. After remaining for three hours the salts were again weighed but no change wa s found. The salts were then transferred to a dessioator containing u much higher partial pressure of ammonia and al+owed to struid 36 hours at room temperature with the result that the percentage of ammonia in the pink salt increased from 4.96 to the theoretical percentage {5.4o%) while that in the white increased from 1.29% to 5.01%. After standing four days in the same dessica- tor, the %of ammoni u in the white increased to 5.09%. • ~1he above data show that the absorbtion of ammonia by the white salt is ver-~ much slower than in the case of the pink. This could not be due to a difference in the physical state as the white salt was light and fluffy whi'.l,e the pink was very compact. The total absorbtion of ammonia by both salts seems to depend upon the partial pressure of the ammonia in the dessicator. Other Salts of Pentabromophenol Various salts were prepared by adding neutral solutions of metallic salts to a dilute solution of the sodium salt of penta- 30 bromophenol. The following is a list of the metallic ions which form difficultly soluble salts with pentabromophenol. Colored .Amorphous Salts Ions Color cu -f- light brown red light pink brown yellow light yellow orange yellow lcu++ co++ Fe-+-r+ Fe++ PbTT 2Hg + Hg*-t White .Amor-_phous Salts Vlhite CrJstalline Salts Ions Ions Ba*+ 3Cd -r-tca+-1'±Ni -1--tMn+-1- K + NH+ Li4,_ !,ig++ zti.,. . . Th..- lDissolves in warm 6N ammonium hydroxide which upon cooling deposits dark brown needles. 2changes to a perfectly colorless salt upon standing exposed to sunlight. 3changes to a deep violet salt upon standing exposed to sunlight. The violet precipitate dissolves in warm 6N. ammonium hydroxide from which lemon yellow needles separate upon cooling. 4Dissolves in warm 6N ammonium hydroxide from which light green needles deposit upon cooling. Either "Baker's Analyzed", or a high grade of Kalbaum•s Chemicals was used throughout this investigation in preparing salts for analysis.