DERIVATIVES OF PHENYLSTIBINIC AND P-ARSANILIC ACIDS BY IRVIN ALVIN KOTEN A. B., North-Western College, 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 V S ' \ 0 \ %\ UNIVERSITY OF ILLINOIS THE GRADUATE SCHOOL Jul y 29 192 — I HEREBY RECOMMEND THAT THE THESIS PREPARED UNDER MY SUPERVISION BY IRVINALVIN K OTEN. ENTITLED D ERIVATIVE S OF PHENYLSTIBINIC AND P-ARS ANILI C AC IDS . BE ACCEPTED AS FULFILLING THIS PART OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE . Recommendation concurred in* Committee on Final Examination* ^Required for doctor’s degree but not for master’s 476959 1 ' . ■ • ■ . . -.u ACKNOWLEDGMENT . I wish to extend my thanks to Professor Roger Adams for the interest shown, and for the many suggestions offered in the preparation of this thesis. Digitized by the Internet Archive in 2016 https://archive.org/details/derivativesofpheOOkote — 1 - TABLE OF CONTENTS. Chapter I . PHENYLSTIBINIC ACID. I. Historical . 3 II. Theoretical 3 1. Phenylstihinic Acid 3 III. Experimental 9 1. Preparation of Phenylstihinic Acid 9 (a) The Sodium Method 9 (b) The Diazo Method 11 2. Reduction of Phenylstihinic Oxychloride.. 12 3. Reduction of Phenylstihinic Acid . 13 IV . Summary , , , . Chapter II . P-ARSANILIC ACID. I. Historical. * . 14 II. Theoretical 15 1. P-Arsanilic Acid..... 2. Atoxyl. (a) Proof as to the constitution of atoxyl is III. Experimental 19 3 . Atoxyl 19 2. Phenylglycine p-arsenic acid..... 2 0 3. Phenylglycine p-arsenious oxide 20 4. Phenylglycine-p-arsenious chloride. 21 5. P-amino phenyl arsenious chloride 21 6. P-chlor-arsenic acid 21 7. P-amino phenyl arsenious oxide 22 . , . , . . . 1 i : 'I ... * * . . . , ........ . , ..... . I ....... j 8. Attempts at the preparation of Phenylglycine-p-ethyl- arsenic acid. • * 22 9. Attempts a t the preparation of Di^henyl-p-diarsinic acid. 24 (a) Cuprous Chloride Method 24 (b) Ammoniacal Cuprous Oxide Method. 24 (c) Copper Powder Method. * 24 IV . S ummar y . 25 V . Bibliography 26 * • • ... . ... . -3- Chapter I. PHENYLSTIBINIC ACID. I. Historical. Twelve years after Bechamp’s discovery of the first aro- matic arsenical (1860-63), Michaelis began a systematic study of the aromatic derivatives of phosphorus, arsenic, and antimony, establishing first at Karlsruhe and Aachen and then at Rostoch a school of chemistry in this particular branch of organic synthesis. Michaelis with the aid of La Coste, Reese, and others, devised general methods of preparation for the compounds of both series, and prepared the first aromatic antimony derivatives. The syn- thesis of aromatic organo-metalloidal compounds has been further facilitated by the discovery that these compounds are obtainable through the diazo reaction. This process has been applied to aromatic antiraonates by the Chemische Fabrik von Heyden of Dres- den . * II. Theoretical. The method by which Michaelis and Reese first prepared aromatic antimony compounds leads chiefly to the triaryl-stibines . More recently the Grignard reaction has been applied to the pre- paration of aromatic antimony derivatives. Based on the discovery made by the Chemische Fabrik von Heyden that antimonial groups can be introduced into the aromatic nucleus through the agency of the diazo reaction, certain aryl-antimony derivatives containing amino groups, such as the antimony analogues of atoxyl, and salvarsan, have been synthesized. Several processes are available for the synthesis of aromatic antimonials. The sodium process is available, but the .(.-Ll.MQr’.g an -Organic fjuLiprnirifls nf /■. nn^ .Antimony . : • - • ' * r : v • . . , • . - : ! ; - .[ • T'. i r > ■ ’ * 1 ‘ . ' ' , ':'i •>'!•■ ' ■ 1 . ' • ’ •> ‘ ! -i I • :* • •: *r - •jo:. $ J J ;rv ■ r Jr ; • • . • 'Jo • • i ■ * ■ • •; 1 : ' ■ V . . ' ■ -4- reaction is considerably complicated by the greater tendency pos- sessed by antimony of passing from the triadic to the pentadic condition. The following outlines show the steps involved in two of the older methods for the preparation of aromatic antimonials . I. Sodium Method. 1 CgH^Cl, SbCl 3 and Sodium. (C 6 H 5 )3Sn (C 6 H 5 ) 2 SbCl 3 (C 0 H 5 r 3 SbCl 2 heated with SbCl3 C 0 H 5 SbCl 2 I C1 2 C 6 H 5 SbCl 4 i H a 0 C 6 II 5 Sb0(0Il) 2 (C 6 H 5 ) 2 SbCl NaOH HC1 NaOH HC1 (C f >H 5 ) 2 Sb0.0H (C 6 H 5 ) 3 Sb(0H) 2 II. Grignard Reaction For Aryl Antimony Compounds C 6 H 5 MgBr and SbBr 3 or SbCl 3 (C 6 H 5 ) 3 Sb and (CfjH 5 ) 3 SbX 2 » Cl 2 | heating with SbCl 3 (C 6 H 5 ) 3 SbCl 2 (C G H 5 ) 3 Sb(0H) 2 (C 6 H 5 ) 2 SbCl (C 6 H 5 ) 2 SbCl 3 (C 6 H 5 ) 2 Sb0.0H C 6 H 5 ShCl 2 C 6 H 5 SbCl 4 C 6 H 5 Sb0(0H) 2 The modern improvments which have facilitated the pro- duction of aryl -antimony derivatives are two -fold. First we have the application of the Grignard reaction to the production of tri- phenylstibine and its homologues. The second improvement was the discovery made by the Chemische Fabrik von Ileyden that antimony (1) Morgan-Organic Compounds of Arsenic and Antimony. P 294 ■ ; i . :: ' ■? ) V ( • ' • 1 ' • ■ ' • ' •f > l ) ) "7 ■ • 1 ■ ’ 1 ! * ’ i 1 , • - -5- groups could be introduced into the aromatic nucleus through the agency of the diazo reaction. This discovery first led to the syn- thesis of antimony atoxyl . NIL SbO ( OH ) 2 by the following steps _ NHCOCH 3 NHCOCH 3 IIONO SbClo+'NaOH , x NH 2 3 SbO (OH ) 2 HOH nh 2 SbO (OH) 0 The diazo process is a general one and many complex antimony de- rivatives can thus be synthesized. 1. PHENYLSTIBINIC ACID. This acid was first prepared by Michael! s and Reese by what is known as the Sodium Method. (CgHg^Sb 1 is first prepared by adding metallic sodium to a mixture of SbCl 3 and C 3 H 5 CI in CgHg. Besides (CgHg^Sb some (CgH 5 ) 2 SbCl 3 , (CgIl 5 ) 3 SbCl 2 and (CgHg^SbO is formed in the reaction. SbCl 3 + 3C 6 H 5 C1 6 Na > (C 6 H 5 ) 3 Sb + 6 NaCl SbCl 3 -t 3C 6 H 5 C1 + 4Na (C 6 H 5 ) 3 s bC 1 2 -h 4NaCl The crude product is purified by extracting the (CgHg^SbC^ with a solution of fuming IIC1 in C 2 HgOH. By means of chlorine (CgHg^Sb and (CgHg^SbO are converted into (CgHg^SbC^ which is obtained in thin needles. These are dissolved in ammoniacal alcohol and by means of II 2 S, (CgHg)gSb separates at first as a crystalline powder which soon melts to form a colorless oil. The pure product can then be obtained by crystallization from alcohol or ether. Moraophenylchlorstibine 2 is prepared next by heating together (1) Michaelis and Reese, Annalen 233-234, P. 43 (2) Hasenbaumer, Berichte 31, P. 2911-13. • ■:••••*' * . . ' ' r .• 0 ' . ' ' -i * ’ • ' -r * • . ; ' 1 1 - / • .'-'.I - t . ■ . . ■ ■ -G- in a tube (CgHg^Sb, SbClg and some dry xylene. Formerly the desired result was not obtained because the SbCl 3 and (CgHr^gSb were heated without a diluting agent which resulted in the reaction being carried too far. (C G H 5 )oSb + 2SbCl' 3 3C 3 H 5 SbCl 2 The black substance is purified by first removing free SbClo with HC1 and then distilling in a stream of C0 2 . At 270-300 degrees a thick fluid passes over which crystallizes on cooling. C6H5SbCl2 is then converted into C 6 H 5 SbCl 4 by saturating it ether solution with chlorine. The crystals formed are very hygro- scopic, and with water form CgHgSbO (OH)^. The crystals of C 6 \\ 5 SbCl^ are dissolved in dilute NaOH and the filtrate is acidified with dilute ITC1 from which CgHgSbO (Oil ) 2 separates as a white amorphous powder. s C 6 H 5 SbCl 4 + GNaOH — * CgHgSbO (0Na) 2 + 4NaCl H 2 0 . CgHrSbO (0Na) 2 + 2IIC1 — » C G H 5 Sb0(0H) 2 + 2NaCl. The nrocess just described is difficult and does not give very good yields. Due to the remarkable discovery by the Chemischc Fabrik von Heyden, that antimonial groups can be introduced into the aromatic nucleus through the diazo reaction, this acid is o prepared by a much more convenient method. Aniline is first dissolved in H 2 S0_4 and diazotized with NaN0 2 . A solution of NaOH is next added to aqueous SbClg. The solution which is rapidly cooled to 0 degrees when a portion of the sodium antimonate separates is then treated with the diazonium solution. The evolution of nitrogen is favorably influenced by (3) Friedlander 1912-14, P. 1084 -D.R.P. 254,421. • ' ' • ■ ; . ” • I - 't i. 1 ) . * ; ' • ■ • ■ j • . . . . -7- The addition beforehand of copper paste or CUSO 4 . After several hours the excess of NaOH is almost neutralized with dilute II 2 SO 4 and CgHgSbO^H^ is precipitated from the filtrate by means of dilute HC1. The crude product is purified from any Sb 2 03 by dis- solving it in hot concentrated HC1 and saturating the solution with NH 4 CI. The NH 4 CI increases the chlorine ion concentration and thus aids in converting the Sb 2 0 2 into SbC^. Upon cooling C 0 H 5 SbO(OCl )2 separates in leaflets and is dissolved in aqueous NagCC^. The purified CgH 5 Sb 0 ( 0 H )2 is then precipitated by adding HC1 to the alkaline filtrate. NH 2 Q'+ HgS0 4 > NH 2 'H 2 S0 a 4 NH2H2SO4 0 HONO N 2 HSO 4 + 2II 2 0 SbCl 3 + 5N aOII > Na 2 HSb 03 -t 3NaCl + 2H 2 0 n 2 hso 4 0 + Na 2 HSb0 3 + 2NaOH Sb 03 Nao Q + Na 2 S0 4 + 2H 2 0 3b03Na2 0 f h 2 so 4 SbO (OH ) 2 Q) + Na 2 S0 4 SbO (OH) + HC1 SK-Cl O' Cl + 4NaOH „ * 0 Sb-ONa ^ Pj N ONa + 2HC1 8b -0 ^C 1 Cl + 2H 2 0 0 ON a Sb \ ONa -h 2NaCl -f 2H 2 0 Sb - OH N 01-I + 2NaCl t •{ - 8 - Phenylstibinic acid was fitst described by Hasenbaumer , a pupil of Michaelis. It is a white amorphous powder difficultly sol- uble in water, but quite soluble in NH 4 OH, NagCOg, dilute NaOH and hot concentrated IIC1. In the moist state it has a faint but not unpleasant odor. It remains unchanged at 285 degrees. Heated on a platinum foil it darkens and burns with a bright flame. The formula according to Hasenbaumer is CgHgSbOgHg. Schmidt however in 1920 observed that it took less than one mole of alkali to dissolve the acid, and that the solution of the acid, in the meagre amount of alkali, consumed more of the alkali (shown by phenolphthalein) until an end point was reached.' He concluded therefore that phenylstibinic acid is in a polymerized form which breaks up into a monomolecular form when dissolved in alkali. On acidifying it again assumes its polymerized form. According to Schmidt the first neutralization occured with a ratio of one of sodium to three of antimony. It has therefore a trimolecular formula. Structurally it would appear to be- OH ,0H C,H,Sb = 0 C H Sb= 0 6 * '0 <• * '0 C H Sb^ 0 or C, H- Sb'Coj; fr 5 "0 * * V C H Sb 7 - 0 C.H Sb^O ^ b V 0H ^ 5 X 0H ^0 The monomolecular formula would be C. H Sb-OH. It was act- s N 0H ually shown by analysis that the acid exists in a polymerized form. If phenylstibinic acid is repeatedly washed with water^ a turbid filtrate results. This is due to its assuming a colloidal form. That it assumes such a form is also indicated by the fact That it is soluble in solvents used for the solution of high (1) Schmidt Annalen 1920, 421, P. 174 ■ ' :• ■ ■ ■ ' ' D . . . -9- molecular cellulose esters. Its weight also changes with varia- tions in atmospheric pressure, temperature, and humidity. The sodium salt of C 0 H 5 SbO(OH ) 2 has "been prepared by dis- solving 2.5 grams of the acid in 25 cc . water and 10 cc • N-NaOH. The alkali salt is then precipitated from the filtrate by 20fo NaOH. These salts are suitable for therapeutic injections . When m-NOoCgH^SbO (0H ) 2 is reduced with zinc and NH^Cl in an alcohol solution the amino compound is fifcst formed. Further reduction gives the stibinoxy compound Sb-OH rV° H LJno 2 ^0 Sb-OH X 0H IH 2 Sb-0 NH 2 -i- 2K 2 0 It has been found that SnCl 2 will reduce C 0 H 5 SbO(OH ) 2 to C^HgSb=0, while NaHSOg will reduce the acid to < V Sh=Sh < > Sb=0 o - (c 6 H R ) 2 sb 2 o (c6 H 5)3 s, b C 6 H 5 Sb=C r ,II 5 ITI. Exnerimental . 1. Preparation of Phenyls tibinic Acid. (a) The Sodium Method. The preparation of ^he older sodium method v/ as tried. (C^H^gSb was prepared first. 50 grams of metallic sodium were powdered by the usual method of heating one -half inch dubes ©f Na in xylene until the xylene begins to boil and then to pulverize the cubes by means of a stirrer. The apparatus was fitted with a mercury seal After the sodium had been powdered, the xylene was decanted and the sodium washed with benzene. Then 100 cc. of benzene were added. 36.4 grams of SbClg were dissolved in benzene and added through the reflux condenser. 49 cc. of . . ' 1 ' ' ■■ ■•.■ , V 0.1 . •' ' ' i n r, t . ■ . ■ I' Y . . ; ■ j - 11 - (b) The Diazc Method. 1 gram mole of CgH^NHg was dissolved in one liter of water containing 1.5 gramemoles of H2SO4 and diazotized with a solution of 1 mole NaNOg. The temperature was kept between 4 and 7 degrees. A solution of 600 grams NaOH in 3 liters of water was added to aqueous SbClg prepared by dissolving (0.5 mol.) of Sb o 0 3 in 764 grams of HC1 (sp. gr. 1*123) . The solution which was rapidly cooled to 0 degrees when a portion of the NagHSbOg separated was then added slowly to the diazo solution with vigorous stirring. A yellow-brown foam formed very rapidly. Copper paste was then added to hasten the evolution of nitrogen. It was found that a solution of CuSO^ worked as well as copper paste. The solution was now stirred for three hours. At the end of that time it was of a brown color. The excess NaOH was then neutralized with dilute HgSO^, and the mixture filtered. The filtrate was of a clear orange color. The filter was brown and pasty like. When IIC1 was added to the filtrate a white-orange precipi- tate formed. This was allowed to stand for twelve hours. Fil- tration was difficult due to the pasty wax-like precipitate. The crude product was purified from Sb^Og by dissolving it in 600 cc. dilute HC1 and heating to boiling. While hot the sol- ution was saturated with dry NH^Cl and allowed to cool. Red leaf- lets of phenylstibinic oxychloride separated. The phenylstibinic oxychloride was dissolved in dilute aqueous NagCOg which gave a red solution. This solution was then acidified with HC1. A white precipitate formed. After reprecipi- tation a white amorphous powder was obtained. It had a very irrit- ating action on the mucuous membrane of the nose. The yield was ' . . ’ . . • . - 12 - 100 grams 2. Reduction of Phenylstibinic Oxychloride It was attempted to reduce phenyl stilbinic oxychloride hy means of nascent hydrogen to and then to form a condensa- tion product "between the stibine and & three liter flask was set up for the reduction. Passing through the stopper in the flask was also a dropping funnel and a glass tube, which was used to draw off the ether layer. A small mercury seal was fitted to the top of the condenser. All stoppers were always coated with wax to prevent air from leaking into the apparatus . Fifty grams of phenylstibinic oxychloride, made by dissolv- ing phenylstibinic acid in hot concentrated IIC1, and allowing the solution to cool, when the oxychloride would separate, were mix- ed with 300 grams of amalgamated zinc dust, and placed in the flask. Enough ether was then added to cover the mixture to the extent of one half of an inch. Concentrated HC1 was then allowed to drop into the flask at the rate of about ten drops j)er minute. The reaction was allowed to proceed for three days. At the end of that time the ether layer which had a browniHh tinge was forced up into a separatory funnel by means of water poured in at the top of the condenser. The ether solution became a turbid yellow. During this time air was kept away from the ether by means of a stream of COg passed over its surface* The ether solution was then dis- tilled under diminished pressure in the presence of COg* No dis- tillate was obtained in the receiver. Instead a viscous yellow- brown residue remained in the distilling flask, which turned black on standing. Within a few days, on exposure to the air, the i ' l : ' . K > ■ ' ' • : . [ ] i oi ■-in' ; '• >*t' . • • ' ■ • • , -13- black color turned to a yellow-white, and seemed to disappear some what. The residue was insoluble in ether, but when a few drops of HC1 were added it dissolved readily. Because the residue was so small and gummy nothing was done to identify it. Another reduction of the same amount as the preceeding was again made. The ether solution this time was drawn directly into freshly distilled C^H^CHO. No visible reaction took place. Concluding that possibly the reduction was not continued for a long enough period of time the reaction was repeated and allowed to proceed for four days. Distillation of the ether layer however gave no results. It was now thought that a change of solvents might be tried and accordingly CHgOK was used in place of ether. The CHgOPI mix- ture was then subjected to steam distillation in the presence of COg* The distillate contained no reduction products. Dilute CIIgCOOH was now used in place of concentrated HC1, and the reaction allowed to proceed for three days. The reaction mixture turned black and spongy, nearly filling the entire flask. NaOH was then added to neutralize the excess CH^COOH. No results were obtained. The residue was refluxed with equal parts of C2H5OI and C^IIq. Copper powder was added as a catalyst. After the sol- vents were distilled off, a yellow-brown residue remained. When dry it had a peculiar pungent odor and was stable in air. It had no definite melting paint. It was soluble in HCl and was precip- itated by NaOH, in which it was soluble in excess. It had pro- perties which resembled those of Cgll^SbO (0H)2* 3. Reduction of Phenylstibinic Acid. Since no definite positive results could be obtained with . I ■ • * 'I ( ■ ■ •f . 1 ' 1 liiTf ■ ‘ r 1 • ' ' • • . * .. . ' ■’")•? r. ’ i ■ ' ■ ' 0 ' -14- phenyl stibinic oxychloride, phenylstibinic acid w as used in the reductions. Ether and dilute HC1 were now used. During the re- action the ether solution became yellow, and was stable in air. Unon evaporating the ether solution, in a vacuum, very small needle-like crystals separated. These crystals however were not very stable in air and on attempted filtration and purification a gummy mass was obtained. The experiment was repeated with the same results. A change in the kind of reducing agent was now made . Ten grams of aluminium powder were added to a solution of 25 grams of CgHrjSbC^OH^ in NaoCOo. Hydrogen was evolved very readily. The ether solution became slightly yellow. After three days the ether layer was separated, but was found to be stable and to contain no CgHsSbHg. No further attempts at reduction were made. V . Summary . *0 1. Cf^Sb-oiI can be obtained very readily and in good OH yields by means of the diazo reaction. 2. Attempts at the reduction of C0ll5SbO(OCl)2 and C(jH5Sb0(0H|[)2 in an effort to obtain C0H5SbH2 failed. 3. Since CgHgSbC^OH^ has been proved to exist in a poly- merized form in acid solution, this^it is concluded^raay be the cause of the inability to obtain CgllgSbl^ by the reduction of CgHgSb0(0H)2 in acid solution. Chapter II. P-Arsanilic Acid. I. Historical. To Be'champ belongs the honor of having prepared the fifcst aromatic arsenical compound. This work was done between the years - . * - , ; . : * . 0 1860-63. BeXchamp 1 s compound which was supposed to have been an anilide of arsenic acid was tried in therapeutics in or about the year 1902. Thomas and Breine employed the compound in the treat- ment of sleeping sickness. Because of the comparatively non- toxic nature of the drug it was named atoxyl . It nroved to be so satisfactory that Ehrlich and BeTtheim undertook a systematic investigation with the result that atoxyl was proved to be the sodium salt of arsanilic acid. Researches were now started in special laboratories, notably those at the Georg Speyer Hospital in Frankfort and at the Hochst FarBwerke vormals ^eister Lucius und Briinning. In England the chemists of Messrs Burroughs Wellcome & Co, have carried through a considerable amount of re- search on atoxyl and its derivatives. And in America, especially since the World War, greater interest than ever before has been shown in arsenical preparations. A few of these will be discuss- ed in this paper. II. Theoretical. 1. P-Arsanilic Acid. The compound obtained by BeXchamp during the years 1830-63 was a colorless condensation product obtained from the reaction of aniline and arsenic acid. He noticed that the compound reacted as a monobasic acid, giving rise to metallic salts of definite char- acter, and was not hydrolyzed by means of aqueous caustic potash. Because of its acid pr* operties he concluded that it was an anilide and accordingly named it " Phenarsenyl -ammonium” . The formula assigned to it was [(^igHsHgAsO^N] 0 .HO , according to the old notation. The third period in the history of organic arsenic and . ■ -16- antirnony compounds comnenced in 1907 when Ehrlich and Belthei-rn, demonstrated the true constitution of atoxyl* These two men shov/ec that the action of arsenic acid on aniline is comparable to the action of sulphuric acid on the same base. In both cases the NH2 acidic group enters the aromatic nucleus giving rise to Q NHo SOgOH and On respectively. The former is termed sulphanilic X y> U As - OH acid OH an< ^ accordingly the latter arsanilic acid. P-arsanilic acid can be prepared by slowly heating to 170- 200 degrees for two hours, in a vessel fitted with a stirrer, 188 grams of q K 2 mixed with 140 grains of arsenic acid . 1 The product is then mixed with water, made alkaline, and the excess base distilled off with steam. The residue is then cooled, fil- tered, concentrated, and neutralized with IIC1, when crude arsenj ic acid separates. This product is then dissolved in aqueous NaOH The solution which should be only faintly alkaline is boiled with animal charcoal and filtered into alcohol, when the sodium salt (atoxyl) separates in a crystalline form. The by-product di-4- amino diphenylarsenic acid remains dissolved, as its sodium salt is soluble in alcohol. The free p-arsanilic acid is liberated from atoxyl by the addition of dilute IIC1. The acid is sparingly soluble in water or CgHgOH, more so in CH 3 OH, and insoluble in ( 0285 ) 20 , C^Hg or CIICI 3 . It is appreciably amphoteric dissolving in excess mineral acid from which it can be precipitated by 1 ^ 028302 * It can be recrystallia- ed from alcohol and water with two molecules of water of cry- stallization. 2. Atoxyl* Atoxyl the sodium salt of p-arsanilic acid, as (1) Morgan. Organic compounds of Arsenic and Sb. P158 /Ai ' ' - . . ' »-( ■ . -17- has been stated beforehand derives its name from the fact that it is comparatively non-toxic. It has been given various synonyms c such as Arsamin, Soamin, Natrium Ars an ilium. The water content of atoxyl is variable. Ehrlich and Eeitheilh state that it is obtainable with two and six molecules of water of crystallization, according to the solvent used in crystallization of the comrjound. The more hydrated forms lose water by eff lorvesence . Commercial specimens contain from three to five and one half moles of water. The product introduced by Qurroughs Wellcome & Co., under the name of Soamin, is in the form of well-defined colorless crystals containing five moles of water. Dr. Mart indale ' s preparation contains approximately three moles of water. As soon as the therapeutical value of atoxyl was discovered several thought it impossible to sterilize an aqueous solution of the salt at 100 degrees, as a part of the substance would decom- 1 pose . Ehrlich and Beitheirn however claimed atoxyl to be stable. It was shown however by Schmitz that the aqueous solution of sod- ium-arsanilate and other salts of arsanilic acid split out ars- enic acid from the molecule. This was determined quantitatively. A solution of atoxyl was heated for an hour in a Koschsher Ster- ilizer, then for a half hour in an autoclave and then at various temperatures. After cooling, IICl was added to precipitate ars- enic acid. The product remaining in solution was converted into the Resorscin-azo dye. The filtrate was decolorized with animal charcoal and the remaining arsenic acid was precipitated as mag- (l) Schmitz Ber. 1914, 47, 363. ' ' • 't I • . • -18- nesium ammonium arsenate. The adherence of arssinic groups to the aromatic nucleus has been studied by E. Schnitz who heated atoxyl with various pro- portions of aqueous alkali at 100 and 130 degrees. The maximum decomposition was obtained by heating solutions containing one mole of p-arsanilic acid with «r.« . -19- to boiling without the splitting out of aniline. This is only possible by heating under pressure above its boiling point or by fusion with caustic alkali. -Since anilides are easily hydrolyzed and atoxyl is not, the arsenic group is therefore firmly attached to the nucleus. 2. Atoxyl has a primary amine group and shows all the re- actions as such. It can be diazotized and its diazo solution can be combined with amines and phenols to give dyestuffs. It can also be acetj r lated. 3. P-arsanilic acid is an aromatic acid. In ammoniacal solution it gives a calcium and magnesium salt at the boiling point. In the cold no salts of calcium and magnesium are formed. It is easily converted into ^ ^ by HI. I 4. The -NHo group is in the para position, because kP can be obtained as shown in (3). III. Experimental 1. Atoxyl. ^0 As ~0H Q x 0Na NHo This compound was prepared by dissolving 45 grams of 0 //0 As -Oil OH NH. in 10 grams of NaOH dissolved in 80 cc. water. The solution was warmed and allowed to cool. Atoxyl crystallized in clusters of long slender needles. The yield was 49 grams. The crystals were insoluble in CgH^OH but easily soluble in water. On the addition of a few drops of HC1 to an aqueous solution of the salt a white As*8h crystalline precipitate of Q x oiI was obtained. The mother liquor from the cry- stallization of atoxyl was allowed to stand for twelve hours. Large crystals with very defin- ite angles separated. They were very eff lorvescent . When exposed . . - 20 - to the air for about five minutes they became opaque, but kept their shape for an indefinite time. The water of crystallization was determined and found to be five moles of water. A ^/ 0 As 'OH X 0H 108 grams of OH 0 2. Phenylglycine P-Arsinic Acid. 0 NH CH 2 C00H ONa were dissolved in 320 cc. nh 2 of water and warmed gently. Then a solution of 64 grams of C1CII 2 C00H in 80 cc . of water was added. The mixture was heated in a reflux apparatus for 6-8 hours. It was found that heating for seven hours with a low flame gave the best results. It was also found that it was necessary to use atoxyl in the preparation As -qj i of this acid instead of using Q '''oh with a calculated amount NH 2 of NaOII. After refluxing for seven hours the mixture was cooled in ice. A color change was now noticed in the solution as it cooled. When hot it was pink and green when cold. On cooling for some time very minute crystals separated. These were filtered off and recrystallized from hot water. A 69$ yield was obtained. The acid was easily soluble in hot water and in concentrated HC1, but only sparingly soluble in dilute IIC1. It was also easily soluble in alkali hydroxides carbonates and acetates. It did not melt at 295 degrees. The glycine group is firmly attached and resists hydrolytic action. \ S -G 0 NII.CHgCOOH 3. Phenylglycine-p-arsenious oxide. As oh 69.5 grams of oil were dissolved in 1150 cc . NH CH 2 C0GH water and 160 cc. 2-N NaOII and 16 grams of El and 416 cc. dilute II 2 SO 4 (1-5) were added. Sulphur dioxide was then passed into the solution for six hours. The temperature was never allowed to ex- ceed 20 degrees. At the end of that time cold concentrated NH 4 OH ' •’ '1 i > ■ ■ ] > •I • > ■ \ ■ . n ^ ■ ') • . '> •; • • < ' a t ■ • J . • . -:if I - •< : : f ’ . ; i • • S if t-: J • 7 " • 1 ' ■ ' ' V ' M ; \ P * ... •• ; '• » 7 ;• ; • • j; r> . •• . . . * . o .« • ■; 1 ; - *i ■ . ’ 2»n - 21 - was added drop by drop. As the heat of reaction was so great ice was added to keep the temperature below 10 degrees. NH4OH was added until the solution was strongly alkaline. The solution was then filtered and dilute HgSO^ ( 1 - 5 ) was added as long as a pre- cipitate formed. The product was curdy and yellow, which on dry- ing became very hard and brittle. It was quite soluble in alkali. Due to the low yields of this compound no further experimental work was don* 3 3 + - KI in water were added and SO2 passed into the solution which was cooled by a mixture of salt and ice. A white flaky precipitate formed, which was hastened by the addition of CH3COOH. The sol- ution was then aerated to remove excess S02» It was then filtered and the precipitate washed with CH3COOH and finally with absolute (€21*5)20. The compound melted at 120 degrees. It was very sol- uble in water and NaOH with which it formed the sodium salt as those given for Phenyl Glycine P-Arsenious Chloride were fol- lowed. CH3COOH was not added however as the compound separated within two minuted after SO2 had been bubbled through the solution. The compound melted at 138 degrees, and was soluble in H2O, CH3OH and C 2 H 5 0 H. in concentrated IIC 1 in which the latter is very NH -CH2C0CH soluble. A few drops of a concentrated solution of NH-CI^COONa 5 . P-Amino Arsenious Chloride (J) NH 2 «HC1 In the preparation of this compound, the same directions 6. P-Chlor-Arsinic Acid Cl . • i - 22 - As -OH This compound was prepared by dissolving 30 grams of Q N 0H NHo in 45 grams of concentrated H2SO4 and 60 cc. HgO. The sol- ^ ution was then diazotized with 15.3 grams of NaNOg. A cold sol- ution of 21.6 grams of CugClg in 100 cc. concentrated IIC1 was then added. After the solution became green a very finely divided precipitated formed. This precipitate was soluble in warm alcohol, insoluble in cold but soluble in hot HgO. It was therefore recry- stallized from hot water containing a little alcohol. An cooling minute crystals formed. They had a buff color, which could not be removed with bone-black The crystals did not melt, but de- composed at a high temperature. The compound was analyzed for arsenic by the method of Robertson.^ Calculated-31 . 69^ As. Found-31 .48^. As=0 7. P- Amino Arsenious Oxide. Q NH 2 In the preparation of this compound, directions similar to As=0 those given for the preparation of Q were followed. Nil ‘CHpCOOH A white precipitate formed which after it Had been filtered and exposed to the air beeame slightly yellow in color. The crude product was purified by dissolving 27 grams in 220 cc. water and 180 cc. 2N-NaOII. It was shaken with ether. The water layer was filtered off and mixed with 108 cc. 9N-NH/J.C1 when the compound precipitated. The compound was soluble in acetone, alcohol, dilute acids and NaOII and not very soluble in (Cgt^jO and CHClg. 8. Attempts at the Preparation of Phenylglycine-P-Ethyl AS^QTI Arsinic Acid. CgH, 16.8 grams of Q As - C1 NH • CHoCOOII •Cl were dissolved in 80 cc. of NH • CHoCOOII (1) Robertson, J.A.C.S. P.&82 - . . -23- CpH r) OH. This solution was then added slowly to an ice cold sol- ution of 20 cc. lON-NaOH. The solution was kept cold to avoid de- AsC C1 composition of the Q while being added to the NaOH sol- Yh«ch 2 cooh ution. The solution was then filtered from precipitated NaCl. 8 cc . of CrjH^I decolorized by shaking with mercury were added. No heat of reaction was noticed. The solutionwas allowed to stand for twelve hours. It was then diluted with 300 cc . water and the resulting turbidity removed by shaking with ether and separating the aqueous layer. While stirring freshly precipitated AgCl, ob- tained from 25 grams of AgNOg, was added to remove free iodine from the solution. The Agl was filtered off, and the solution accidified with concentrated HC1 until it reacted with congo paper. It was then concentrated on a steam bath. NaCl separated. On evaporating to dryness a black tarry residue remained, which was soluble in alcohol and insoluble in ether. As yet it has not been found whether this resinous material contains some of the expected product . Instead of CgH^I, CgHgBr was used and the use of AgCl omitted. On acidifying with HC1, NaCl would precipitate. No def- As^Jii inite arsenic compound which corresponds to (Y U2H5 has yet been isolated. NH*CH 2 C00H It was also attempted to introduce a -C2H5 group using AsCp} As -oh ^3 , and then using the supposedly prepared f) N c II NHo'HCl Sju 5 and allowing this to react with iNn 2 CICHgCOOH. No definite results however were obtained, with the exception of a product which did not possess the properties of the compound in view. This compound has as yet not been analyzed, as no suitable solvent for purification has thus far been found. -24- Due to so many possibilities of other compounds being form- ed by the methods so far used, this it is thought may explain the inability so far to prepare the compound readily. 9. Attempts at the Preparation of Diphenyl-P-Diarsinic As^OH ^ s r8ll Acid. D N)ii At present experiments are in progress to prepare the a- bove named compound. Three methods for preparing Diphenyl Compounds have so far been tried. (1) Cuprous Chloride Method. A o ^ ASsfjU Cl As?8„ U OH Cl The same proceedure as described under the preparation of A O *0 As^OH was used. The compound expected was but analysis of the compound showed it to be As ?0H \ OH (2) Ammoniacal Cuprous Oxide Method. . A) As -.OH 10 grams of rj were dissolved in 100 cc water and . ml 87 cc. 13 io IIC1 , and IH 2 diazotized with 75 cc. N-NaNOg* CugO in NH^OH, made by reducing an aqueous solution of 37 grams CuSO^ in hhe presence of NH^OH until a white precipitate of CuNH^SOg formed This precipitate was dissolved in 300 cc. water and 25 cc . NH^OH and added to the diazo solution. A yellowish precipitated formed which resembles Cu 2 C^ , but has as yet not been definitely deter- mined . (3) Copper Powder Method As ,M)H OH O v NH* in 40 grams concentrated H^SO^ 31 grams of and 150 grams water were diazotized with 23 grams KaN02* Then 125 cc. 90^ C 2 H 5 OII were added and also 50 grams ■ -25- copper bronze. A brown colored solution resulted. At present work is confined to the study of this solution. IV . Summary . (1) The properties of P-Arsanilic Acid and its sodium salt Atoxyl is described. (2) The structure of Atoxyl is discussed. (3) The preparation and properties of the following deriv- atives of P-Arsanilic Acid are given- 1. Phenylglycine-^-Arsinic Acid. 2. Phenylglycine-P-Arsenious Oxide. 3. Phenylglycine-P-Arsenious Chloride. 4. P-Chlor-Arsenic Acid. 5. P-Amino-Phenyl Arsenious Oxide. 6. P-Amino -Phenyl Arsenious Chloride. (4) An attempt at the preparation of diphenyl-p-diarsonic acid is discussed. (5) Attempts at the preparation of phenylglycine-p-ethyl arsonic acid are discussed. . - . . . * - 26 - BIBLIOGRAPHY. I. Phenylstibinic Acid. 1. Hasenbauraer , J. -Ber. 31, P, 2910-13. 2. Ilasenb'aumer , J. -Ber. 31, 1398, Pt.III, P. 2913. 3. Michaelis & Reese -Annalen 233-234, P. 43. 4. Morgan-Organic Compounds of Arsenic and Antimony . -Intr . 5. Morgan-Ibid. P.294. 6. Fried lander . D.R.P. 254,421 1912-12 P. 1084. 7. Pfeiffer. -Ber. 37, P.4620. 8. Morgan and Mickelthwar t . -Chem. Soc. Trans. 1911, 99, P.2290. 9. Schmidt . -Annalen, 1920, 421, P.174. II. P-Arsanilic Acid and its Derivatives. 1. Ehrlich and Berthein. -Ber. 43, P. 917-19. 2. Bertheim. -Ber . 44, P. 1070. 3. Bertheim. -Ber . 48, P. 350. 4. Adler and Adler. -Ber. 41, P. 932. 5. Benda and Kahn. -Ber. 41, P. 1674. 6. Schmitz . -Ber . 47, P. 363. 7. Bertheim and Ehrlich. -Ber . 40, P. 3292. 8. Morgan. -Organic Compounds of Arsenic and Antimony. P. 158. 9. Ibid. -P.165. 10. Weyl. -P.843. 11. Benda. -Ber. 41, P.2370. 12. Freidlander . - D.R.P. 204,664. Vol. IV P. 1035. D.R.P. 206,057. (1909). D.R.P. 254,187. 13. Vorlander and Meyer . -Annalen, 320, P. 122p P. 134. 14. Gatterman. -Ber . 23, P. 1226. 15. Allman and Forgan.-Ber. 34, P. 3802. ■