V PREPARATION OF CERTAIN DERIVATIVES OF PHENYL ARSINE, AND THEIR REACTION WITH ALDEHYDES BY SAMUEL MARION McELVAIN B. S., Washington University, 1920 THESIS Submitted in Partial Fulfillment of the Requirements for the Degree of MASTER OF SCIENCE IN CHEMISTRY IN THE GRADUATE SCHOOL OF THE UNIVERSITY OF ILLINOIS 1921 N\\-^ UNIVERSITY OF ILLINOIS THE GRADUATE SCHOOL Ma y 27 J I HEREBY RECOMMEND THAT THE THESIS PREPARED UNDER MY BE ACCEPTED AS FULFILLING THIS PART OF THE REQUIREMENTS FOR THE DEGREE OF ^>iiAtLCEH-CLb’ BCIMHCS Recommendation concurred in* Committee on Final Examination* *Required for doctor’s degree but not for master’s Digitized by the Internet Archive in 2016 https://archive.org/details/preparationofcerOOmcel ' •, -.a'.’-,. ' :■ . ;r ■,' 'f * • ' ■ ' ■''■ *S‘’' I p.? ' ;■'. - -^. ".>■ ' •' t-'-'i" 3 1 ' A *1 E JK f 5ll!P t,T ' k--. f . V • A * 4 :'.^- 4 , 5 ; .,-■ ^ ' 0 TO/r’-‘vi" ., - kL it, ■‘“#™ . ‘ t-i . ' , 's..-'V-''i ra » j t>. ,; ,,,’A:t(,'. . . „,,:ji - 1 - It has heen recently found that phenyl arsine can he con- densed with aldehydes in the presence of a small amount of hydro- chloric acid to give definite and relatively stable addition products. The reaction is between one mol of the arsine and two mols of the aldehyde and goes probably according to the follow- ing equation: CsHsAsHg 2RCH0 > GeHsAsCCHOHR) ^ By this reaction condensation products between phenyl arsine ( 1 ) and butyraldehyde and benzaldehyde have been prepared. The discovery of this new series of organic arsenic com- pounds immediately suggested the possible therapeutic value of some of them. Consequently it was the purpose of the work with which this paper deals to prepare an alkali soluble arsine and to attempt condensations of it with various aldehydes. Such condensat ion ..products would have water soluble salts and there- fore be capable of therapeutic testing. p-Carboxyphenyl arsine was chosen as the arsine to be employed. This arsine should give a series of alkali soluble addition products of the general f ormula p-H 00 CC 6 H 4 As(CH 0 HR) 2 . It was found, however, that condensations could not be brought about to any appreciable extent between this arsine and acet- aldehyde or benzaldehyde, when the same conditions under which phenyl arsine condenses were enployed. In the case of benzalde- hyde a small amount of what appeared to be an addition product ( 1 ^ Adams" and Palmer,’ J . ■Ame'r. Chem, Soc., 42, 2375 (1920)' was isolated, but' it differed from the compound expected in that there was no free carboxyl group present. The compound also appeared to be unstable, for its arsenic content consistently decreased after each crystallization from hot acetone. EXT^ERTMENTAL . Preparation of n-Benzarsonic Acid. HOOC-CfiHAAsOr>Hr>{'D) . This acid was prepared by two different methods, (a) the oxida- tion of p-tolylarsonic acid and (b) from p-amino benzoic acid by diazo reaction. Sieburg^lias reported quantitative yields of p -benzarsonic acid from the oxidation of p-tolylaraonic acid, with nitric acid (d*1.3) in a sealed tube at 170® C. Such a method would not, however, be very practical for the preparation of fairly large quantities of the acid. Instead of nitrio acid alkaline perman- ganate was substituted as an oxidizing agent. PotassiuM' perman- ganate (28g.), p-tolylarsonic acid (20g.) and sodium hydroxide (I2g.) were dissolved in about 1500 cc. of water and the oxi- dation allowed to proceed at room temperature. After about two weeks the colorless solution was filtered. The filtrate was acidified (to Congo red) with hydrochloric acid and evaporated to about 100 cc. iihen, upon cooling to 5®C., crystals of p-benz- arsonic acid separated out. This acid is readily separated from any p-tolylarsonic acid by the fact that the sodium salt of the ' latter is soluble in alcohol while the corresponding salt of the (l) Sieburg, Arch. Pharm., 254, 224 (1916). - 3 - benzarsonic acid is insoluble. The yield by this procedure is about 18 g. (60^ of theory) . For the preparation of p-benzar sonic acid from p -amino benzoic (1) (2i acid a modification of Barts reaction was employed. One mol of p-amino benzoic acid was dissolved in a liter of water con- taining l.S mol 3 of hydrochloric acid. This solution was cooled to about 0°C. and a saturated solution of one mol of sodium nitrite was slowly run in from a funnel. The solution was stirred vigorously by a motor during diazot ization. The diazo solution thus prepared was then slowly siphoned into an arsenite solution, made by dissolving one mol of arsenic trioxide and one mol of sodium carbonate in about 800 cc. of hot water to which 10 g. of CUS 04 . 5 H 20 was added as a catalyst. During this reaction the solution was stirred vigorously and the temperature kept be- tween 12o-15©C. with cracked ice. After all the diazo solution had bean added the reaction mixture was placed on a steam cone and, with continued stirring, heated at a temperature bf 60® C. for one hour. UlThile still hot the reacEtion mixture was acidified to Congo red with hydrochloric acid and filtered. A dark red by product is removed by this filtration leaving a clear straw colored filtrate. It is important that the reaction mixture be acidified at least to congo red, otherwise the filtrate will contain some of the red by product which is very hard to separate from the benzarsonic acid. (ll Ind. Eng. Chem. 11, 824 (1919) (2) D. R. P. 250624. /■r f V l#.»,. I V*l.v Vf , 't ,v . w." ■ :V; 1 1 %x-' ■:■ .' i '■,* V ;: ’: i '■ 'V :4 ^■"' it .. t* • • ' ' ' if- • ' ' ' , n' ' r :■,» m , - V’ •::i V ••Iv. 5T W* ’I'l: f.' '"'" 'Dr '•■■' * ' ■■ v^’. "■,Z .'■ ‘i . ^ ■.) i ■■ 'j I • f >:■. Vi-:!: ' ffl/J > ■ ■ ■': ■■ ./S' ■• , ■ ■ ' "j f';' "■■; .; "• L ■> ■ i " %4v Jf,:V ■ -. r.v , j 'V , ■: ' , :'i; ’V> ti Cl'-:' ^j'r :'43 ■ ‘OJ a' / , f # ,- n - "r r , t.fr- :N: - L --' i.;'] - . ’ ... '.t' V ... •■••<■. O i J O *'4 • * ':C :; lvv>v=; ■■m ' . ■ , . 'i. •' i;;l. - . j| - 4 The filtrate was then concentrated to about 50Ooc. and cooled. The benzarsonic acid separated out in yellow colored scales which were decolorized by boiling their alkaline solution with bone black. The yields by this method were between 55-60^ oi theory. Schmidt has reported good yields, of the various arsonic acids using sodium hydroxide instead of sodium carbonate in the arsenite solution. Several rans were made using sodium hydrox- ide in order to compare the relative merits of the two alkalies. The yields obtained were much poorer than those given by sodium carbonate. They varied from 10'^ to 35*^ of theory, and, due to the increased formation of the red by product, the benzarsonic acid was much harder to purify. This arsine has been prepared by Sieburg by the active reduction of the corresponding arsonic acid in methyl alochcl with zinc and hydrochloric acid. Both ether and methyl alcohol ?;ere tried in this work as solvents for the acid with the result that methyl alochol was found to be much the better. The main diffi- culties with the ether appeared to be its volatility at the temperature of the reaction and the fact that it was not a very good solvent for the arsonic acid. ( 1 ) Preparation of r-Carboxvphenvl (1) Schmidt, A., 421, 159(1920) (2) Sieburg, Arch. Pharm. 254,224 (1916) a,/ ■;-.'<^5 ■ . y'V\- 't. , . ••* ■ % ■ ' "^1 .-•/O >5 ^ 'L..-.,.f\ J2 1; .-'c . .. v-» * - >4 ^ .«1 *■ .* I . «; A - - - .('. , ■: ';v^ feJ 7 - - ’ ■:.;. ; Vfl ’ 1 ■i ■ • c. ,J i ■■ C M >? JC ^.v . . :^COi? "t* . 4 1 ..a ;pj )£ rj.'i p-Benzareonic acid (50g.) and zinc dust, smalgamatad with 3 g. of mercuric chloride (400g.) were placed in a 3 liter round bottom flask fitted with a stopper carrying a dropping funnel and a reflux condenser. The upper outlet of the reflux conden- ser carried a mercury trap. Methyl alcohol (300 cc.) was run into the flask thru the funnel and the contents of the flask were warmed until the arsonic acid went into solution. Concentrated hydrochloric acid (l 1.) was then slowly dropped into the flask at a sufficient rate to keep “up a fairly vigorous evolution of the gas. The time required for the addition of the acid was usually about 8 hours. After the acid had been added, the contents of the flask were diluted with an equal volujne of water and the arsine removed. Two procedures were available for the extraction of the arsine, viz, (a) steam distillation and (b) extraction from the reaction mixture with ether. Owing to the extreme sensitivity, of these arsines to oxygen, these procedures must be carried out in an atmosphere of carbon dioxide. Steam distillation was found to have a distinct s^dvantage over ether extraction of the reaction mixture because the ether dissolved substances the other., than /arsine . These substances were probably other re- duction products of the arsonic acid. ^hen the reduction mixture was subjected to steam distilla- tion, a white crystalline condensate separated out in the con- denser tube. This condensate together with the aqueous 9 fX distillate was received in a 3 1 , filter flask containing 300 cc. of ether. The distillation was kept Tsp with carbon dioxide passing thru the system until no more of the arsine appeared in the condensing tube. Fnen this stage was reached the distillate usually amounted to about one liter. The ether layer containing the arsine was separated from the aqeueous portion and the ether, after drying over anhydrous sodium sulfate, removed in a vacuum desiccator. Yield 10-15g. (35-37'^ of theory) . p -Garbo xyphenyl arsine crystallizes from ether in short white prisms, which melt at 79-80<>C. in an atmosphere of carbon dioxide. It cannot be distilled under 10 mm. pressure without decomposition into metallic arsenic. This arsine behaves quite different on exposure to air than phenyl arsine. The latter compound is oxidized in the air to a mixture of arseno benzene and phenyl arsinic acid. If this kind of change were to take place with p-carbox^qjhenyl arsine an alkali sol- uble product should result. Such is not the case, however, for on exposure to air p-carboxyl phenyl arsine quickly changes to an amorphous yellow compound which does not melt under 300® and which is insoluble even in boiling sodium hydroxide. The oxidation is evidently accompanied by some sort of conden- sation by which the identity of the carboxyl group is destroyed. v: / •\ ■ V »r € V » I ) ,■ . . r> ir *|. • ' - ^ • • r ■< 7 CONDE^JSATION OF p-CARBOrYPHENYL ARSINE WITH ALDEHYDES. Acetal dehvde 'j p -Garbo xyp he nyl arsine (15g.) was dis- solved in 100 cc. of ether and this solution transferred to a 200 cc . Erlenmeyer flask. This flask was fitted with a stopper carrying an inlet and outlet tube for carbon dioxide together with a small dropping funnel. A few drops of concentrated hydrochlcr icracid were added as a catalyst to the ether solu- tion and then paraldehyde (7g.) was rum in from the funnel. During this pmcedure a continuous stream of carbon dioxide was kept passing thru the flask. Almost immediately after the addition of the paraldehyde to the arsine solution a yellow compact precipitate appeared. This precipitate had the appear- ance of the compound resulting from the oxidation of the arsine in air, in that it was insoluble in alkali and did not melt under 300<>C. Similar results to these were obtained on each attempt to condense the arsine with paraldehyde. From these facts it appears that acetaldehyde instead of condensing, exerts an oxidizing effect rpon the arsine. Benzaldehyde . With this aldehyde the same procedure was followed as in the previous case, except that benzaldehyde (l5g.) was subsitututed for the paraldehyde. There was no immediate precipitate, but on standing over night a white voluminous precipitate settled out. An excess of benzaldehyde --if'. ' I. - /v; . . ■ ..’ * : . ( J’ y . - . V /■ jr.*^ ••* k i. c y ; ■ "iJl - •' ;.ei . •• \^r. i K ' : ,, r.,6' '*y \ (xt 6 - f'Kt. ; ^ 4 . ' : .. ^' ., ■ ■■ ' ' ■ .. .u-A < 4- i ( \ 8 ever the theoretical amount should be avoided, for it was found that the precipitate was quite soluble in benzaldehyde-ether mixtures. The largest yield of this compound obtained was about 3 g. from runs in which 15 g. of the arsine was used. This condensation product was very bulky and crystallized from hot acetone in large cotton-like masses. It began to shrivel and darken at 234oC. and melted with the evolution of a gas at 236-239oC. Its weight was unchanged after drying at 110®Q«for two hours. It was insoluble even in boiling 5'^ sod- ium hydroxide. A most peciiliar property of the substance, in contrast to its apparent stability when heated to 110© C., was its consistent loss of arsenic upon repeated crystallizations from hot acetone. The arsenic content was determined by the the arsenic content was 14.4^ (theoretical value for HOOCCgHsAs (CHOHCsHg) 3 - 18. After each of two more crystallizations of the same product from acetone, the arsenic content dropped to 11.2*^ and respectively. From these results it would appear that p-carboxy phenyl arsine does not condenserwith benzaldehyde in the same way that phenyl arsine does. Undoubtedly a condensation product of some kind is formed, but in very insignif iO;ant yields. This condensation product appeared to be unstsabl%, losing ( 1 ) method of Ewins A-fter two crystallizations from acetone (1) Ewins, J. Chem. Soc., 109, 1356 (1916) • V V ... '■ m • -v; r c■:^^; ' i':. ' t. iii '1 ■‘' ' -v-4 . < n: -.7 ,v ^ {:-i V - ^ V -. ;• I .i. * 1 ;'^ «r' .> ^ / i'w.rir*'. '■ ,' n '■ ' ''■ - y-^i.:t.ck>r4i u:^ '>■•' ■ ., . > r“ f ■*: ^- -'I 5- , » ,r: .:C'i ;y;v \i • . a ' ' ... t.- r» t ». — I '■ * ' *f 'V - - ' 'VC.! ■ri^ f '■ • 1 ■ •■ ’ . . •; , t ■ ■ T "• .--I*' . • \ I ' .<* , > .j - * ,y ' .' . •; V.M'Vl i. /-.r, *' /J '■ ?.,V ''"i "’ w ■. •I, ■ ' ». .* ^ A K‘' 1. J .'fc' k<^t • -> V / ' ■ : 1 y . . .■ • : F-J- : i ■ \ : - 9 - arsenic in some way, when boiled with acetone. The results of the attempt to prepare snluble members of the series, C6H5 As(CH 0HR) 2, using p-carboxy phenylarsine as the arsine to be condensed have been negative. The reactivity of the AsEg group in these condensations may be effected by the presence of a carboxyl group in the ring. Such a difficulty mi^t be easily overcome by using an arsine of phenyl acetic acid, such as p-H2AsCsH4CH2C00H. The effect of the CHgCOOH groiJp in the ring should approach the effect of a CHg groijp, in any event the effect should be markedly different than that produced by a -COOH directly attached to the ring. In case the presence of a carboxyl group in the arsine should give such a conden- sation product the tendency to form more complex condensations thru the carboxyl group, and thereby have its identity de- stroyed, the problem mi^t he approached from another angle, i.e., using an aldehyde carrying a salt forming group. With such an aldehyde as p-hydroxy benzaldehyde , phenyl arsine might readily condense to give a product soluble in alkalies. SUMMARY AND CONCLUSIONS. 1. p-Garboxyphenyl arsine on air oxidation is not changed simply to an arsdno compound or an arsinic acid, but the change is more complex, involing the carboxyl group in such a way that its identity is destroyed. J 1 « 10 - 3. Acetaldehy^de does not condense with this arsine, but instead seems to exert an oxidizing effect similar to exposure to air. 3. Benzaldehyde with the arsine forms small amounts of a condensation product, but the condensation is not to simplify to the formation of H00CC6H4As(CH0HCsH5) The product formed is insoluble in alkalies and unstable in boiling acetone. A • ’if, > ... i t\\’ t- » . j \ \ s ^ wJ'l •, 1