THE ACTION OF PHENYLARSINE ON ALDEHYDES BY CHARLES SHATTUCK PALMER B. S. University of Illinois, 1917 M. S. University of Illinois, 1920 THESIS Submitted in Partial Fulfillment of the requirements for the Degree of DOCTOR OF PHILOSOPHY IN CHEMISTRY IN THE GRADUATE SCHOOL OF THE UNIVERSITY OF ILLINOIS 1921 Digitized by the Internet Archive in- 2015 ) . L ,. https://archive.org/details/actionofphenylarOOpalm UNIVERSITY OF ILLINOIS THE GRADUATE SCHOOL j.'jov eiiioer j , j 92 1_ I HEREBY RECOMMEND THAT THE THESIS PREPARED UNDER MY SUPERVISION BY CHARLES SHAT T UCK PALME R ENTITLED THE- -A££I Qli-QE -PHBiiYXAESX^SL-QII- ALDEHYDES. ___ BE ACCEPTED AS FULFILLING THIS PART OF THE REQUIREMENTS FOR THE DEGREE OF Doctor of Philosoph y Recommendation concurred in* Committee 0)1 Final Examination* ’Required for doctor’s degree but not for master’s . *■ ; -1 ■ TABLE OF CONTENTS Page I. INTRODUCTION: PRIMARY ARSINES 1 Historical 1 Preparation of arsines 2 Reactions of arsines 2 Sources of difference between arsines and amines 4 II. THEORETICAL 5 Previous aldehyde -arsine condensations 5 Condensation products, C 6 H 5 As(CHOHR) s 6 1,4, 3, 6 Dioxdiarsines ? Reductions, etc, 8 New Mechanism of arsine-arsine oxide reaction 9 III. EXPERIMENTAL 11 General 11 Phenylarsinic acid 11 Phenyl arsine 13 Analytical 14 Condensation products, C 6 H 5 As (CHOHR) s 15 Reactions of C 6 E 5 As (CHOHR) 5 21 1,4, 3, 6 Dioxdiarsines 29 Reactions of 1,4,3, 6 dioxdiarsines 33 Reduction reactions 35 Hydro gen- chlorine substitutions 40 IV. SUMMARY 41 . ♦ . * . m ■ 1 . - 1 - PART 1 INTRODUCTION Primary Arsines . Until comparatively recently it was a fixed belief that arsenic differed from nitrogen and phosphorus in that primary and secondary arsines were incapable of existence. Because of the extreme rapidity with which arsines are oxidized by air, (D many attempts to prepare such substances had resulted in failute. It was not until 1894 that the first secondary arsine, dimethyl- (3) arsine, (CH 3 ) s AsH, was obtained by the reduction of cacodyl chloride. This important discovery was followed seven years (3) later by the isolation of primary arsines. Possibly on account of the inconvenience attending the handling of such compounds, they have not been thoroughly investigated during the succeeding twenty years. The application which arsenicals have found in medicine and in chemical warfare has led to a resumption of activity in this interest ing f ield. The work reported in the following pages consists in reactions which take place when phenylarsine is treated with aldehydes or other substances possessing a similar chemical behavior toward amines. According to the results, striking differences between aniline and its arsenic analogue have been found to exist. ~(1) M ichael is, A .201 , 304 (i860); Michael is, Schulte, B.15, 1953 (1883); Michael is. A . 330 , 276 (1902). (2) Palmer, B.27 T 1378 (1894). (3) Palmer, Dehn, B._34, 3594 (1901) . . . . * . . . . * . - 2 - Theoretically the preparation of phenylarsine closely resembles that of aniline; C 6 H 5 As0 2 + 6H *■ C 6 H s AsH 2 +■ 3 H 2 0 In practice the corresponding arsonic acid, C 6 H 5 As0 3 H 2 , is reduced* The reduction is usually carried out by means of a ( 1 ) ( 3 ) metal and mineral acid, but may be done electrolyt ically with success. All primary arsines are prepared by this general method* Only the first three members of the series of aliphatic arsines are known* The synthesis of aromatic arsines has been confined mainly to such substances as 3-amino“4~hydroxyphenyl- arsine, or p -phenyl glycine arsine, which are comparatively stable in air and are related to the therapeutically important salvarsan, etc. Such arsines have a trypanooidal action about equal to that of the arseno compounds, but are slightly more (3) toxic. The principal reactions of primary arsines are described below; 1. They are rapidly oxidized in air to RAsO, RAs-AsR, (4) or RAs0 3 H 2 2RAsH 2 4 0 g RAs-AsR +• 3H 2 0 RAsH 2 +• 0 2 *- RAsO + H 2 0 RAsH 2 4- 1 -l/30 2 ^RAs0 3 H 2 (1) B.34, 3594 (1901); Am. 33, 101 (1905); D.R.P, 251,571; Sieburg, Ar, 254 . 224 (1916) . ( 2 ) B. 34, 3594 (1901); D.R.P, 267, 083; 270,658. (3) Kahn, Chem* Ztg. 1913 . 1099. (4) B. j34, 3597,3599 *11901); Am.^33, 124,144,149 (1905); Am. 40, 115 (1908) . : . . , ■ , . . ... . . ' . . . ; . . . . : _ . . - 3 - ( 1 ) 2. Inorganic oxidizing agents have a similar action* The following are the most important examples: RAsHg +■ 2Kg0g — RAsO + 3H 2 0. RAsHg + 6HN0 2 - RAs0 3 Hg f 3H 2 0 + 6N0 RAsHg + 3N0 2 ^RAs0 3 Hg + 3N0 5RAsHg +• 6 KMn0 4 •- 3RAs0 3 K 2 + 2RAs0 3 Mn 4 - 4MnO +■ 5H gO 3. With halogens, sulphur, H s S, or compounds containing ( 2 ) active halogen, the hydrogen of the arsine is replaced. RAsHg 4 - ^ 2 RA s 1 2 +- Hg RAsHg + 2S ^RAsS + HgS RAsHg 4 • HgS — — »*RAsS + 2Hg RAsHg + SnCl 4 - *- RAsClg +• 2SnClg RAsHg 4 - SgCl 2 — RAsClg 4- S + HgS 4. When heated with alkyl halides, the arsines yield (3) quaternary arsonium compound si RAsHg * 3CgH 6 I yR(C 2 H 5 ) 3 AsI + 2HI . 5. Primary arsines, condense with RAsO, RAsCl 2 ,RSbClg, K antimonyl tartrate, and As, Sb, Bi halides with formation (4) ol arseno compounds: RAsHg 4- R’AsClg RA s=A sR 1 2 3 4 + 2HC1 . RAsHg + R’AsO RAs=A sR* + H 2 0 (1) Am. 23, 125, 144, 149 (1905); Am. 40, 105 ff (1908) (2) Am. _33, 126, 150 (1905); Am. 40, 105 ff (1908) (3) Am. 33, 128, 145, 152 (1905); Am. 40, 112 (1908) (4) Am. 40, 108 (1908); D.R.P. 254, 187; 269, 744; 269, 745; 270, 259. Ehrlich, Karrer, B. 46, 3564 (1913). • • . . a- * . : ; i } . • • - 4- RAsH 2 4- R«SbCl 2 - j-RAs-SbR' 4- 2HC1 RAsH 2 ¥ SbBr 3 RAe=SbBr + 2HBr RAsH 2 +■ BiBr 3 —+• RAs=BiBr + 2HBr . The only evidence pointing to salt-forming properties of primary arsines is the formation of an addition product (1) between ethylarsine and sulphuric acid, but it is very un- stable and was not analyzed* An extended comparison of amines, phoschines and arsines (3) already exists in the literature. Careful consideration of the foregoing discussion dis- closes three principal sources of difference between the re- actions of arsines and amines: 1. The arsines are strong reducing agents. 2. The basicity of the arsines is practically neg- ligible, 3. The hydrogen bound to arsenic is readily exchanged for halogen or sulphur. These conclusions have been confirmed and strengthened in course of the present investigation. It will be observed that little has been done in comparing the purely- organic reactions of the arsines with the amines. It was with this object in view that the action of aldehydes on phenylarsine has been studied. (1) Am. 33, 144 (1905) (2) Dehn, Am, J33, 101-117 (1905) a - , ...... . ' .: . . : . . . V . . • . w v ••• . . t . , f ) ~ 5- PART II* THEORETICAL Tne first indication in tne .Literature ox an attempt to condense an aiaenyde with a primary ar3ine is tne inc±u3ion by Dehn of formaldehyde in a list of substances wmcn did not U) react with gaseous metnyiarsme • A little later the conden- sation of 3-amino-4-hydroxyphenylarsine with benzaiaenyde-m- (3) su.iphonic acid is described* In tnis case tne ammo-group alone reacted. The aoove are tne only examples where direct condensation has been attempted. Of interest in this connection is the compound obtained oy treating 3-am ino -4- hydro xyphenyl- arsine in dilute hyarocnlorio acid solution with an excess of (3) sodium f ormal deny desulphoSCy late, HO .GH B .0.80Ha at 30®. From tne sulphur content of the product it appears possible that the sulphoxylate has reacted with the -AsH 2 group as with an amine, by splitting out water. It is uniortunate that in- sufficient information is given so that a definite conclusion as to tne constitution cannot be drawn. (4) Recently it has been found that .two molecules of aldehyde react witn one molecule oi phenyl&rsine in the presence of HG1 or ZnOlg as catalyzer according to the following equation:! C 6 H 5 AsHg + 3 RCHO *- G 6 H B A s ( CHOKR) 2 (1) Am. 40, 108 (1908) (2) D.R.P. 373,035 (1913) (3) D.R.P. 378,648 (1814) (4) Am, Soc. 43, 3375 (1920) 6 No exactly analogous derivatives of phosphines or amines are known* Ammonia and benzaidenyde combine at -20® to form the (1) addition compound, NH(uri0HC 6 H s ) 2 , m.p. 45®. Tftis substance decomposes very readily into hydrobenzamide , benzaldehyde and water. When treated with benzoyl chloride in alkaline solution, a little benziiiaine-aibenzamide , C 6 rt 5 Ch(NHu0C 6 h s ) 2 , is formed. Phosphine unites with two molecules of cmcral or butyryl chior- ( 3 ) al. The products are stable solids, which yield di-esters when heated witn acetic or propionic anhydride. Tne condensation products, C 6 h e As(CHOhR) 2 , are colorless oils of high boiling point, or colorless needles. They possess a decided stability toward water, dilute acids and dilute in alkalis. This behavior is/str iking contrast to that of the iso- (3) meric esters of phenylarsenious acid, C 6 H 5 As(0R) 2 . Toward oxidizing agents, halogens, phosphorus pentachloride, dimethyl sulphate, the condensation products act like mixtures of phenylarsine and aldehyde. With phenyl hydrazine , acid chlor- ides or anhydrides (in the cold) reducing agents and dehydrating agents, no action is observed. Highly unstable addition prod- ucts are formed by allowing the aliphatic condensation products to stand with constant boiling halogen acid. No such addition compounds were obtained from the products of the interaction of phenylarsine with aromatic aldehydes. Stable addition (1) Francis, B. 42, 2216 (1909) (2) Girard, C.r. 102,1113; A. ch. (6) 2, 43 (168©)- (3) Michael is, A. J320, 286 (1902) 7 - compounds with HgCl g and chlorplat inic acid were obtained from both aliphatics and aromatics. Attempts to prove hydroxyl groups by means of phenyl isocyanate and the Grignard reagent gave negative results. The former reagent produced decomposi- tion, while magne 3 ium /iodide had no effect whatever. In spite of the lack of very positive proof, the formula C 6 H 5 As (CHOER) 2 seems the safest to advance. When heated with acetyl chloride or acetic anhydride, two molecules of di- -hydroxy tertiary arsines lose two molecules of alcohol with the production of 1,4, 3, 6 dioxdiarsines : Paraformaldehyde and furfural yield these ring compounds directly the first type of condensation product being unstable in these two instances. If agitation is not employed, acetaldehyde and phenylarsine may be condensed mainly to the ring compound, the reaction given above taking place as an intermediate 3 tage. The dioxdiarsines are much higher boiling than the substances from which they are derived, as might be expected. They resemble the latter in most chemical properties, but do not add halogen acid. The addition compounds with HgClg, H g PtCl 6 and CuCl 8 are methyl .O-CHR C e H s As^ x RCH-0 /sC 6 H b + RCE 2 OH stable - ■ . , - * - 8 - The dioxdiarsine s are oxidized by atmospheric oxygen to phenylarsine oxide and the respective aldehyde. This behavior differs from that of the other type of sub stance , from which phenylarsinic acid is formed. The oxidation of dioxdiarsines ( 1 ) may be represented as follows: .O-CKR CgHgAs-s. AsC 6 H s + Og RCH - CT 0 s ^ s A S_ SCH' 0 >AiC-H 0^ 0 * 6 ll & 2 C 6 H 5 AsO t 2 RCHO From analogy with aniline, it might be expected that aldehydes and ketones would condense with phenylarsine, when heated together with dehydrating agents, according to the following equation: ~C~ 0 + H 2 AsC 6 H 5 - ^ =C=AbC 6 H 5 + H g 0 On account of the lack of basicity of the arsines and their strong reducing properties, a quite different reaction takes place : 2=C~0 + 2 CgHgAsHs *~2=CH0H + C 6 H s As=AsC 6 H 6 . This reduction is not held back by hydrochloric acid, since the condensation reactions previously described are entirely prevented at 100© . (l) Engler, Weissberg: "Kritische Studien iiber die Vcrgange der Autoxiiation (1904), 63, Cf . Dehn, Am. 40, 91ff (19©8) ; Steinkopf, Schwen, B. 54, 1440 (1921). . - • r: * . • . • : . : ~ 9 ~ The ease with which halogen is substituted for hydrogen in the arsines is illustrated by the action of chloral on phenyl ars ine » The aldehyde group does not react. The -CCi 3 group attacks the phenylarsine with formation of phenylarsenious chloride, C 6 H 5 AsC1 2 . New Theory of the Reaction between Primary Aryl Arsines and Aryl Arsenious Oxides or Halides. From the previous discussion it appears that water is not split out between the carbonyl group = 0=0 and the primary arsine group, -AsH 2 » The fifth general reaction in Part I indicates the elimination of water between primary arsine and arsine oxide. However, thi3 reaction can be explained more satisfactorily on the basis of the reducing rower cf the arsine . The arsine reduces the arsine oxide to an arseno-compound, and is itself oxidized to another a r3e no-compound in the process: 2R ' AsO + 2 RAsH 2 ^R'As^AsR' + RAs^AsR + 2 H g 0 The rearrangement of two arseno-compounds into the unsymmetrical derivative takes place very smoothly and oractically quant itat- ( 1 ) ively: R'As^AsR’ +■ RAs = AsR *-3 RAs=AsR ! Thus exactly the same product is obtained as by the assumption of direct condensation, and the suggested mechanism is strictly in accord with experimental results. The 3arae reasoning applies to the interaction tof primary arsines with arsine halides, (l)D.R.P • 251104: E .? . 17482: Karrer, B.49, 1648 (1916); D .R.P . 253226; 2702S5. . . * . : ' . ... . . . ... - 10 stibine oxides, bismuth tribromide, etc. 3 R'SbCls + 3 RAsH s *-R'Sb«SbR' + RAs-AsR + 4 HC1 R ' Sb=SbR ' + RA s=A sR *2 RAs=SbR ' 11 - PART III EXPERIMENTAL General . dr Phenylarsinic acid , C 6 H 6 As0 3 H 2 » - The most general method for the preparation of aromatic arsinic acids is Bart’s ( 3 ) reaction : R— N — Cl f As(ONa) 3 ^ RAs0 3 Na 3 + N 2 * NaCl „ N For the production of phenylarsinic acid on the large laboratory acale, a 25~1 . cylindrical copper tank, provided with mechanical stirrer was employed. In the tank were pierced 4 1. of water, 2 kg. Na 2 C0 3 , 1 kg. As 2 0 3 (about 2C$excess), and 45 g. copper sulphate. The stirrer was started and the walls of the tank cooled by several streams of water. As soon as the temperature of the arsenite solution fell to 15», the addition of diazo solution was begun. It was found convenient to prepare the latter in four portions. 186 g. aniline, 400 cc. cone. HC1, 1 1. water and sufficient ice to bring the total volume to 3 1. were placed in a 4-1 • Florence flask. This mixture was diazo tized in the usual manner with a concentrated solution of 140 g. NaNO a . Three hours or more were required for running four of these solutions into the arsenite, the temperature of the latter being maintained at 15© . (1) LaCoste, Michaelis, A. 201, 203, (1880) j Michaelis, Loesner, B. 27, 265 (1894); Bertheim, B. 41, 1855 (1908). (2) D.R.P. 250, 264; 254092. See also J. Ind. Eng. Chem. 11, 825 (1919); Schmidt, A. 421, 159 (1920) . i * . . ' i . ; •* . • . • * . . • . : ' • ‘ . . ■ • ...... . . ' : - 13 - Stirring was continued for 1 hour after all the diazo had been added. The solution was filtered and the filtrate con- centrated to a volume of approximately 5 1. in 16 in. evapor- ating dishes on the steam bath. 100-300 cc . cone. HC1 were added cautiously and the tarry material filtered off. This process was repeated until a clear, pale-yellow solution re- sulted. The phenylarsinic acid was then precip itated by addition of more HC1, avoiding an excess. When the solution had cooled, the product was filtered off and washed with a little distilled water. The average yield of white or cream- colored product exceeded 800 g. (= 50 # of the theory) » Phenyl- arsinic acid softens at 158® and passes into its infusible anhydride, C 6 H 6 AsOs. ( 1 ) Phenylarsine, C 6 H 5 AsH g . 1. Original method of Palmer and Dehn. A 5-1. round-bottom flask was provided with a long bulb reflux condenser. A 3-hole stopper at the upper end of the condenser carried a 3-1. dropping funnel and an outlet tube connected to a mercury -f illed U-tube . In the flask was placed a mixture of 400 g. phenylarsinic acid and 800 g. amalgamated zinc dust, a little water and 1 1 . of ether. 3 1. cone. HC1 were added drop by drop; by proper regulation the apparatus could be allowed to run over night without attention. When (1) B. _34, 3598 (1901); Am. J33, 101 (1905); Chem. Ztg. 1913 , 1099, :1 ; (. . • . . . ■ - 13 the reduction was complete, the condenser was removed and re- placed by a 3-hole stopper carrying a 6 in. funnel and a delivery tube. By pouring water into the funnel, the ether layer was forced through the delivery tube into a C0 2 -filIed, 2-1. sep- aratory funnel. Lumps of anhydrous CaCl 2 were added. When dry, the solution was transferred in portions to a 500 cc.- C0 2 -filled Claisen distillation flask fitted with a receiver of equal size and most of the ether removed by heating on the steam bath, a constant current of C0 2 being passed through the apparatus. The distillation was continued under diminished pressure. By a little practice it was found possible to change receivers without oxidation of phenylarsine, so that use of a Bruhl apparatus is unnecessary. The product boiling around 93®, 70 mm. was collected. When the distillation was completed, the vacuum was shut off, C0 2 admitted to atmospheric pressure and the product sealed in large test-tubes which had been pre- viously constricted near the open end, and filled with carbon dioxide. The best yield obtained was 83%; the average was nearly 60 Phenylarsine is a clear, colorless liquid of dis- agreeable odor. It causes painful blisters when it comes in contact with the skin and is highly irritating to the mucous membrane. Constants not given in the literature are: 25 25 N, 1.6082. Dgg, 1.356 2. Steam distillation method of Kahn. In thi3 method the phenylarsine was removed from the re- . J • ♦ I , - . . . • r *1 - •• . .. « ‘ * ‘ ^ i ■ . . •' . . . . - 14- action mixture by simultaneous streams of C0 2 and steam* The condenser was connected to a large f liter-flask, cooled in a freezing mixture. The outlet of the receiver was joined to a U-tube containing mercury through which C0 2 and excess hydro- gen escaped. Chemicals were employed in the same amounts given in the first method. Although this procedure is comparatively rapid, yields are no better and phenylarsine is so volatile that all of its vapor is not condensed even by use of a glass worm surrounded by salt and ice. Steam distillation is better adapted to the preparation of solid arsines. Several attempts to reduce phenylars inic acid in a manner analogous to the commercial preparation of aniline »by use of iron powder and a small quantity of HC1 - were unsuccessful. This result was anticipated on account of the more ready reduc- ibility of the nitro-group. Thus nitro-arsinic acids may be a) reduced to amino -are inio acids by means of sodium amalgam or ( 2 ) ferrous hydroxide. Analyt ical .- For carbon and hydrogen determination a combustion tube packed with pieces of a sintered mass contain- ing equal weights of copper oxide and lead chromate was used. Ewins' method for the determination of arsenic was given a (4) thorough trial, but is less satisfactory than that of Robertson. All As determinations given below were done according to the directions of the latter author. (1) Bertheim, B.41, 1657(1908); D.R.P. 206344; Bertheim, Benda, B.44, 3299 (1911) (2) Jacobs, Heidelberger . Rolf, Am.Soc. 40, 1580 (1918) (3) Soc, 109, 1356 (1916) (4) Am. Soc. 43, 182 (1921) , - ' i : - . ■ •r . . . . . . ..'ir • ' : : •• • • ' o . ' ,"t . ■ •' . ■ ’ ■> . + .+ t r 'V> U i , . ... . . . : 'i . - 15- Condensation Products. C 6 H B As(CHOHR) 2 . Pi- oc- -hydroxyethylohenvlarsine . A 500 cc. wide-mouth bottle was provided with a mechanical stirrer and placed in an ice-bath. C0 2 was passed into the bottle, 100 g. phenyl arsine and 3 cc. cone. HC1 poured in and the stirrer started.. 75 g. acetaldehyde or paraldehyde (an excess) were added drop by drop. Stirring and the C0 2 -stream were continued for an hour or two after the addition of the aldehyde was completed. The reaction mixture was shaken up with a little fused K 2 C0 3 , which served to remove HC1, phenyl- arsinic acid and moisture, and then distilled under diminished pressure. The yield of redistilled product boiling at 175-176°/ 23 mm. was 137 g. (= 81% of the theory) » Clear, colorless oil, Dgl * 1.252; n 2 ^ = 1.5592. Insoluble in water, soluble in organic solvents. Subst., 0.3637: C0 2 , 0.4743: H 2 0, 0.1430. Subst., 0.1356, 0.1274: 10.2 cc., 9.65 cc. iodine (l cc . = .0041 g. Asi Subst., 0.6072: Benzene, 22.0: f .p • lowering, 0.564°. Calc, for C^qHjqOjjAs: C. 49.59: H, 6.19: As. 30.99: Mol. wt. # 242. Found; C. 49.03: H, 6.02: As. 30.83,31.05: Mol. wt., 246. The aliphatic homologues of this substance resemble it in physical and chemical properties. All of the condensation products were prepared as described above, except that stirring was not used in small runs and in some cases a solvent was em- ployed. Identical yields were obtained by use of dry HC1. . . . - 16 - Pi- oi-hydroxypropylohenylarsine . - 17 g. propionic aide- hyde and 23 g. phenylarsine were condensed in the presence of HC1 « Yield, 70$. B.P., 21 3-3140 /70 mm. ; D§5, 1.176; H 35 , 1.5310. Sub at., 0.2043: 13.8 cc. iodine (1 cc. = .0041 g. As) Calc, for CigHjgOgAs: As, 27.77 Found: As, 27.69. ( 1 ) Di- cl -hydro xybutylphenyla.r sine • * By preparing this sub- stance in the same manner as described for the acetaldehyde condensation product it was obtained in 75% yields. (75 g. from 50g. phenylarsine and 50 g. butyryl aldehyde.) The refractive (l ) index was reported incorrectly in the preliminary paper. P r 25 Corrected constants are: B.P., 187© /10 mm.; Dg£, 1.116; n'"", 1.5271. Subst., 0.2296, 0.2210: 14.0 cc . , 13.5 cc. iodine (l cc. = .0041 g. As) . Calc, for C* 4H 23 0 2 As: As, 25.17 . Found: As, 25.00, 25.05. Di- cx -hydroxvlsovalervlnhenylarsine » - 16 g. isovaleryl aldehyde were allowed to react with 15 g. phenylarsine. Yield, 59$. The product is a rather viscous oil, b .p . 235© /36 mm., P 25> 1*^79; n 25 , 1,5172. A small portion was solidif ied .by immersion in a freezing mixture . After recrystallication from ether, needles were obtained which melted sharply at 62© . (1) Am. Soc. 42, 237 6 (1920) . . . . : ' . ' - 17 - Subst., 0.1980: 11.1 cc . iodine (l cc. = 0041 g. As) Cal c . for CjgHjj^OgAs: As* 23.8? » Found: A 3 , uj2«98» Pi- a: -hydroxvheptylphenyl arsine - ’Mien oenanthol was condensed with phenylarsine, the product distilled with decomp- osition under 20 mm. pressure and formed an agar-like mass when cooled in a freezing mixture. In another experiment, 15 g. oenanthol and 6 g. phenylarsine were condensed. The reaction mixture was dissolved in ether, extracted with NaHS0 3 solution and allowed to stand over night in contact with air in order tc oxidize any free phenylarsine to arsenobenzene , The residue was redissolved in ether, filtered and the purification re- peated. The final ether solution was dried over fused K 2 C0 3 and evaporated. Analysis of the colorless residual oil showed the presence of somewhat impure oo, oc'-dihydroxyisoheptylphenyl- arsine • Subst., 0,1818: 8.1 cc, iodine (l cc. * .0041 g. As) Calc, for CgQHggOjgAa: As, 19.83. Found; As, 16.2? Glucose --phenyl arsine .- All attempts at obtaining any reaction between glucose and phenylarsine failed. A consider- able number of experiments were carried out in which the effects of long-standing, heat ing, solvents and mechanical agitation were observed. In every case the two substances were recovered practically quantitatively. : . * . • 1 ’ . w . ' • • ... : . * - 18 - ( 1 ) Dl- oc -hydroxybenzylphenylarsine - 300 cc. of benzene, 58 g. benzaldehyde, 2 cc . cone. HC1 were placed in a wide -mouth bottle of one liter capacity which was provided with a mechan- ical stirrer. The bottle was immersed in a bath of ice and water, the stirrer started, C0 2 passed in and 40 g. phenylarsine added. Precipitation began almost immediately; after 15-20 minutes the reaction mixture looked like whipped cream. After two hours the product was filtered off, and washed thoroughly with ether. More was obtained by concentrating the combined filtrate and washings to small volume. The whole was recrystal- lized from hot benzene and allowed to stand over night in a dilute NaOH solution to which a little alcohol had been added. The insoluble material was filtered off and washed with water , alcohol and ether. In this way a purer product was obtained than by several recrystallizations since in the latter case a trace of phenylarsinic acid persisted. Yield, 65 g;, colorless, silky needles, m.p. 193©. Sub st., 0.1522, 0.1876; C0 2 , 0.3704, 0.4499; H 2 0, 0.0695, 0.0843. Subst., 0.1740, 0.1505; 8.7, 8.5 cc. iodine (l cc. * .0041 g. As) Subst., 1.3709; naphthalene, 26.73: f.p. lowering, 0.972©. Calc, for C 2 O H 1 90 2 A 8 : C, 65.58; H. 5.19; As, 20.66; Mol. wt . , 366 Found: C, 65770, 65.41; H, 5.11, 5.01; As, 20.47, 20.43: mol. wt . 364 . Pi- cc -hydroxy -p-chlorobenzylphenylarsine . - 9 g. Eastman’s p-chloro- /benzaldehyde were dissolved in 25 cc . of acetone. 5 g. phenyl- arsine and 0.5 cc . cone. HC1 were added and the solution allowed (1) Am. Soc. 42, 2377 (1920) c l - 19 - to stand over night in a tightly stoppered C0 2 -filled flask. A slight heat effect was noticed at first. To isolate the product, the reaction mixture was evaporated to dryness, the residue extracted with alcohol and the insoluble portion re- crystallized from chlorobenzene. Needles similar to the ben- zaldehyde compound, m.p. 160°. Yield, 1-1/2 g. Subst., 0.1057: 4.3 cc. iodine (1 cc. - .0042 g. As) Subst., 0.1411: A gCl , 0.0905. Calc, for C 2 qHj yOgCl 2 A s : As, 17 .24; Clj 16.48 Found; As, 17.08; 01 , 16.65. Pi- oc. -hydroxy -n-methoxybenzyliphenylarsine .- 27 g. anisic aldehyde, 15 g. phenylarsine , 200 cc . ether and 1 cc. cone. HC1 were placed in a 500 cc. wide-mouth bottle and the reaction run as described for the benzaldehyde compound. After standing over night, the solution was extracted with aqueous bisulphite and evaporated to dryness. This purification was repeated until constant composition was obtained. The final product was a yellow syrup. Yield, 33$. Subst., 0.1835, 0.2013: 7.65, 8.5 cc . iodine (l cc, = .0041g. As) 0 al c * for C g gH g gO 4 A s : As, 17. 61 « Found: As, 17.09, 17.31. Pi- cc -hydroxy-o-oxvacet icbenzyl -phenyl s.rsine acid. o-Aldehydophenoxyacet ic acid was prepared by the method of ( 1 ) Rossing. It was purified by fractional precipitation from (1) B. 17 , 3000 (188 4) J. - 20 - solution in dilute NaOH. 30 g. of this aldehyde were condensed with 13 g. phenylarsine in the usual manner, using 100 cc. acetone for solvent and 1 cc. cone. HC1, The solvent was evapo- rated spontaneously and the oil which remained was dissolved in the theoretical amount of dilute sodium hydroxide. After evaporation of the water at room temperature, the sodium salt crystallized in a hard, orange-brown mass. This was dissolved in water and the free acid precipitated in fractions by dilute HC1, the first small fractions being discarded. The remaining product was washed several times by decantation with distilled water and dissolved in just enough concentrated NaOH. On spontaneous evaporation the salt crystallized in purified con- dition. It was transferred to a filter, washed with acetone and dried in vacuo over HgSO*. The final product is yellow powder, readily sol\ible in water. Yield, 8 6 $. Subst., 0.1574, 0.1941: 4.8, 5.8 cc. iodine (l cc.=.0Q446g. A 3 ) Calc, for C g ^ Og A sNag : As, 1 o»4*j. found: A3, 13.59, 13.26. Pi- T ' 2t .JCo . : . • . . ; ■; - 36 - Mercuric chloride . When warmed with aqueous mercuric chloride, the condensation products form highly insoluble solids. The solution was decanted, the solid boiled with alcohol, filtered off and washed with ether. The products are stable white or grayish solids, practically insoluble in water and organic solvents. They decompose without melting. Analysis of the substance obtained from acetaldehyde-phenyl- arsine indicated addition of two molecules of HgCl 2 . Subst., 0.2616: 6.2 cc. iodine (l cc. * .0041 g. As) Calc, for C 1 oH l6 0jAs . 2 EgClgi As, 9. 56$. Found: As, 10,19$ Ethyl iodide . 6 g. butyryl compound and 4 g. ethyl iodide were placed in a 100 oc. round bottom flask connected to a re- flux condenser. After 20 hours of heating on the steam bath, the solution was transferred to a crystallizing dish and con- centrated. Only phenylarsenious oxide, m.p. 122°, could be found. The same result was obtained with the benzaldehyde coiipound, 3 g. acetaldehyde compound and 2 g. ethyl iodide were sealed in a test-tube and allowed to stand for six months. A dark glue was obtained which could not be purified. Phosphorous pehtachloride . 35 g. PG1 S were treated with 20 g. of acetaldehyde compound in a Claisen flask connected to a glass worm condenser. The heat of the reaction was ab- sorbed by an ice-bath so that the distillate came over below . . . . . . . - 2 ? - 60° , It consisted of acetaldehyde mixed with a little ethyli- dene chloride and phosphorous oxychloride, weight 5 g. The distillation was continued in vacuo. A fraction boiling below 60° /80 mm., weight 31 g. consisted mainly of F0C1 3 . The fraction, 153-155° /18 mm., weight 16 g., was identified as phenylarsenious chloride. Subst. 0.2127: 11.0 cc. iodine (l cc. - 0.0064 g. As) Subst., 0.1663: AgCl, 0.2173. Calc, for C 6 H s A3C1 2 : A 3 , 33.63: Cl, 31.84. Found: As, 33.57: Cl, 32.29. C 6 H 6 As(CH0HH) 2 f 2 PClg ^C 6 H 6 AsC1 2 + 2 HC1 + 2 P0C1 3 + 2 RCHO . On account of its low boiling point most of the acetalde- hyde escaped before it could be acted upon by the PC1 6 . Phenylarsenious chloride . 1 g. acetaldehyde compound was placed in a corked test-tube with an equal weight of phenyl arsenious chloride. Slight warming and turbidity were observed. After one week the mixture was practically solid. It was transferred to a filter and washed with ether. The residue was identified as araenobenzene , mixed m.p. 196-199°. The indicated reaction is: C 6 H 6 As(CH0HR) 2 + C 6 H B AsCl s ^C e H 8 As*AsC 6 H 5 4 2 HC1 t 8 RCHO. Acid chlorides . Many attempts were made to prepare esters from all of the simple condensation products by msans of acetyl chloride, acetic anhydride, oxalyl bromide, benzoyl . ♦ 4 * . ... • , * . . . * . ... . . . . - 28 - chloride and p -nit robe nzoyl chloride. In the cold without solvent, in the Schot ten- Baumann reaction and in pyridine solution, the compounds were unaffected. The result of heating certain of the substances with acid chlorides or anhydrides is discussed in the next main section. Dimethyl sulphate . 1 g. acetaldehyde compound and 3 cc. 10$ NaOH were placed in a test-tube . Dimethyl sulphate was aided with constant shaking as long as any action took place. On standing in air the non-aqueous layer crystallized. This product was identified as arsenobenzene, mixed m.p. 196-300®. Apparently the dimethyl sulphate merely splits the aldehyde- arsine product. No attempt was made to determine the action of dimethyl sulphate on the aldehyde. Orignard reagent . The acetaldehyde and butyrvl alde- hyde condensation products were indifferent to the action of a large excess of magnesium methyl iodide in dry ether solution. No gas was evolved after several weeks standing in a fermen- tation tube. By decomposition of the organo -metallic compound with water, ether extraction and vacuum distillation, the original products were recovered nearly quantitatively. Phenyl 1 so cyanate . 5 g. phenyl isocyanate and 5 g. acet- aldehyde compound were sealed in a bomb tube. A pink solution formed which had the same appearance after standing over night. The tube was heated for 30 hours at 150° . On opening a very * ' . . . n . - ' . - * * - 29 - heavy pressure was observed. The reaction mixture was trans- ferred to a filter and washed with ether. The residue was recrystallized from nitrobenzene, weight 5 g. It wa9 identi- fied as diphenylurea, mixed m.p. 335°. The ether filtrate was evaporated, leaving a dark, gluey residue which contained 31.7% of arsenic. It was not investigated further. Phenyl- isocyanate had a similar action on the butyryl compound. Phenylhydrazine . The acetaldehyde compound was unchanged when heated for several minutes in a test-tube with an excess of phenylhydrazine in glacial acetic acid solution. Condensation products, (C 6 H 5 ) gAsgO s (CH) 2 R 2 . - — — — ~ “ ” .O-CH 2 . 3.6 Diphenyl. 1.4. 3. 6 dioxdiarsine , C fi H R A3 v ^ AsC 6 H 5 x CH 2 -0 " 5 g. paraformaldehyde and 12 g. phenylarsine were placed in a C0 2 -filled test-tube and 0.5 cc. cone. HC1 added. The tube was sealed and shaken from time to time for about half-an-hour afterwards. Heat was evolved, the paraformaldehyde dissolved, and a clear, colorless solution resulted. After standing over night or longer, the tube was opened and the contents trans- ferred to a C0 2 “*filled Claisen distillation apparatus. The fraction which came over below 110® /? 43mm. contained water, HC1 and formaldehyde. On continuing the distillation under diminished pressure, the mixture decomposed with loss of a low-boiling substance. No constant boiling distillate was obtained until 21 4®/ 9 mm. was reached. The presence of methyl • '■< . • '• * ' . . . . rf * tjf, ' . * - • • * ■ \ .. -- ..... , i. . . - . .. 1") . . *■ . s a *, r 1 ' . ■ - . - 30- alcohol in the fraction boiling below 214®/ 9 mm, was shown by the preparation of methyl 3, 5 dinitrobenzoate from 3,5 dinit robenzoyl chloride and 0,5 cc, of the liquid. The fraction boiling at 214® -220®/ 9 mm, was collected and immediately re- distilled. This time no loss of volatile substance was noted. The final product is a colorless, viscous oil which rapidly decomposed in the air and must be preserved in C0 2 -filled, sealed test-tubes. Yield, 8 g. B.P. 215-216®/ 9 mm; Dg S , 1*547: n 25 , 1.6522. Several runs were made in an attempt to obtain the condensation product, C 6 H 6 As(CH 2 0H) 2 , but no constant boiling fraction, other than stated above, could be obtained. It seems safe to conclude that dihydroxymethylphenyl-arsine was formed initially, but that on distillation two molecules of methyl alcohol were split out between two of the former substances, a six-membered ring resulting. Sub st., 0.2560: C0 2 , 0.4320: H 2 0, 0.0970. Subst., 0.3953, 0.1617: 39.4,16.1 cc, iodine ( 1 cc ,0043g.As) Subst., 0.7900: 22,0 g. benzene: f .p . lowering, 0.506®. Calc, for C 14 H 14 0 2 As 2 : C, 46.15: H,3.85: As, 41.21: Mol. wt . 364: Found: C, 46,01: H,4.08, As, 41.81, 41.82: Mol. Wt. 355. 2.5 Dimethyl. 3.6 diphenyl. 1,4,3. 6 lioxiiarsine . ^o-ch-ch 3 C 6 H 6 As^ \u c 6 h 6 CH^CH-d 7 110 g. phenylarsine and several cc. cone. HC1 were placed in a C0 2 -filled 500 cc. Erlenmeyer flask. 100 g. paraldehyde (large excess) were added drop by drop simultaneously with the passing of a current of dry HCT1 into the reaction mixture. No - 31 - agitation of any kind was employed * After 1 hour the stream of HC1 was stopped, the flask stoppered tightly and allowed to stand over night. On distillation, 59 g. of di- AsC 6 H 6 CH-0 CHsfOH ,)» 5 g. acetic anhydride and 15 g. di- CC -hydro xybut yip he nyl- arsine were refluxed for 6 hours. The mixture was distilled from an oil-bath using a fractionating column. The fraction 123-140° was insoluble in water, had a strong ester odor and consisted mainly of butyl acetate, although it gave a fuchsin aldehyde test. Apparently the formation of aldehyde was due to heat decomposition. The distillation was continued under diminished pressure and the product which had not distilled below 230° /8 mm. was retained. It was dissolved in dry ether, the solution filtered and the ether evaporated. After a repeti- tion of this treatment, 7 g. of an amber oil were obtained. Subst., 0.1055, 0.1627; 8.7, 13.4 cc. iodine (lcc^.0041 g.As) Calc, for C 2O H 26 0 2 As 2 : As, 33.48 Found: As, 33.81, 33.77. 2,5 Djfurvl . 3.6 diohenyl. 1.4. 3,6 d l oxiiarsine . 28. g. freshly distilled furfural and 1 cc. cone. HC1 were added to 32.5 g. phenylarsine in a C0 2 -filled bottle in an ice-bath. A very violent reaction took place, the mixture turned dark and sufficient heat was evolved to crack the bottle,. On cooling, a hard mass like a piece of coke formed. When powdered, the product resembled zinc dust. It is insoluble in all solvents and when heated on a platinum foil, burns - 33 without melting, leaving no residue* The powdered substance was treated for a few minutes with boiling benzene, filtered off and dried. After a similar extraction with dilute alkali, the compound was analyzed. The yield was quantitative. Subst., 0.1898, 0.1206: 13.1, 8.3 cc . iodine (1 cc .-.00446g.As) Found: As, 3©. 78s 30.33. The same substance was obtained when furfural was added drop by drop to a vigorously stirred benzene solution of phenyl arsine at 20© . Subst., 0.2 5 57: C0 2 , 0.5214; H 2 0, 0.0879. Subst., 0.1477, 0.1470: 7,05,7, 1 cc. iodine (lee .«.0064g. As) Cal c . for C g i g 0 4 A s 2 » 0 , 53.33; H, 3.63: As, 30 *34. Found: C, 55.61: H, 3.85: As, 30.54, 30.90. Reactions of the Pioxdiar3ines . Unless otherwise designated, the technique for the work described below was the same as that employed for the corres- ponding reactions of the compounds, C 6 H s As(CH0HR) 2 . Wate r, dlli H Q1, dll» NaOH . The methyl compound was un- affected by long standing with these rea]gents in the cold. A portion boiled for three hours with alcoholic KOH remained unchanged . Oxidation . The formaldehyde ring compound is very rapidly oxidized by air. 17 g. of the substance were dissolved in 50 cc. of acetone and the solution allowed to stand in the air* The mixture warmed up and gave a very strong aldehyde test* After - 34 - standing over night the acetone was evaporated and the residue, wei^it ,16g. , recry stall ized from alcohol. It was identified as phenylarsine oxide. Subst., 0.1649*, 16.6 cc. iodine (l cc.=.00446g. As) Calc, for C 6 H 5 AsO: As, 44.64. Found: As, 44.89. The methyl and propyl compounds were oxidized in a similar manner by the air, but the action was very slow, 1 g. methyl compound was treated with cone. HNQ 3 . Aldehyde was evolved and phenylarsinic acid obtained by evaporation .of the remaining solution. Reaction: (C 6 H b AsOCHR) s 4- 8 HN0 3 -2 C 6 H 5 As0 3 H 2 + 3 RCHO + 8 JNO 2 4-3H 2 0 Iodine . 1 g, methyl compound was dissolved in 10 cc. of ether and iodine added in small portions. Aldehyde was evolved and a dark heavy oil formed. It was washed with ether by de- cantation. Since it gave free iodine and phenylarsinic acid with cone. HN0 3 , it was considered to be C 6 H 5 AsI 2 . Halogen acids. Constant boiling halogen acids did not form addition products as with the compounds of the first type. Slow decomposition took place, but no attempt was made to identify the products. - 35 - Phosphorus pentachloride , 1 g. methyl compound was treated with excess of phosphorus pentachloride. Vigorous reaction and evolution of aldehyde. The reaction mixture was treated with water cautiously and the precipitated oil identified as phenyl- arsine dichloride by a method previously described. Ethyl iodide , 2 g, methyl compound and a considerable ex- cess of ethyl iodide were refluxed for four hours. On evaporation of the excess ethyl iodide, the original ring compound was recover- ed unchanged and practically quantitatively. Metallic Salts; Addition compounds are formed with HgCl g , CuCl s , HjgPtClg, but not with ZnCl 2 , CdCl 2 , CaCl 2 . To 0,5 g, methyl compound in 15 cc, alcohol was added an excess of chlor^ platinic acid. After shaking a minute or two, the product was precipitated in fractions by water, washed and dried on a plate. White powder, Sub6t., 0,0980: 4,3 cc . iodine ( lcc . s =.00446g. As) Subst., 0.1685: 0,0385 g. Pt . Calc, for C l6 H l8 0 2 As 2 HH s Pt01 6 : As, 18.70; Pt, 34.31. Found: As, 19.55; Pt, 22.85. Reduction Reactions Aldehydes . 7,5 g. phenvlarsine and 5 g. benzaldehyde were dissolved in 25 cc, of glacial acetic acid. The solution was refluxed in a current of C0 2 for two hours. The precipitated solid was filtered off and washed with ether. It was identified as arsanobenzene, weight, 6.75 g. This and many other samples 36 - of arsenobenzene which were obtained were mixed with sufficient phenylarsinic acid to cause them to melt high. After extraction with alkali and one or more recrystallizations from benzene or chlorbenzene, a melting point and mixed melting point of 196-197® was found. Subst. 0.2466; 28.6 cc . iodine (1 cc»=.0042 g. As) Calc, for C 12 H 1 Ci As s ; As, 49.33. Found: As. 49.34, The filtrate from the arsenobenzene had a very strong ester odor. On distillation, acetic acid was recovered. A fraction, which boiled at 200-210® under atmospheric pressure, was refluxed over night with an excess of strong aqueous NaOH. The mixture was distilled, the non-aqueous layer removed and identified as benzyl alcohol by preparation of the p-nitro- ( 1 ) benzoate, m.p. 84-88®. A portion of the residue which remained in the distilling flask was evaporated to dryness and the presence of acetic acid shown by the cacodyl test. 2 C 6 H 6 CH0 4 - 2 C 6 H s AsH 2 + 2 CH 3 C00H 2 C 6 H 5 CH s 00C .CH 3 + 2 HgO + C 6 H s As=AsC 6 H 6 Phenylarsine was oxidized similarly when heated with benzalde- hyde in 1:1 proportion in a sealed, C0 2 -filled tube at 100®. The reduction took place even in the presence of cone. HC1 , the condensation reaction apparently being eiltirely prevented at 100 ® . (1) B. 30, 2288 (1897) . . . • . . 4 • • . * . . ' , fc: ‘ • • . . ' ' . * - 37 - 29 g. benzaldehyde , 20 g. phenylarsine, 75 cc. benzene and a piece of ZnCl s were allowed to stand in a C0 2 -f illed paraff in- sealed flask. After 5 days the flask was opened and the solid material filtered off. From this residue were obtained 16 g. arsenobenzene and 3 g. phenyl arsinic acid. The filtrate was distilled and 10 g. benzaldehyde recovered. The residue from this distillation was recrystallized from alcohol and pressed out between filter papers, weight, 7g., m.p. 58-60®, This corresponds to desoxybenzcin , but the mechanism by which this product could be formed is not clear. 9 g. Eastman's p -chi o robe nz aldehyde and 10 g. phenylarsine were dissolved in 75 cc . benzene, 1 g. fused ZnCl~ added and the mixture allowed to stand for 5 days in a CO 3 -f illed, sealed flask. No condensation took place. The flask was opened and arsenobenzene, dry weight, 10 g., filtered off. On concentration of the filtrate, long prisms were obtained, m.p. 70® -71® , weight, 9 g. This checks for p-chlorobenzyl alcohol • Subst. 0.1216: AgCl, 0.1258. Calc, for C7H7OCI : Cl, 24.84 Found: Cl, 25.59. 2 C1C 6 H 6 .CHO+3 C 6 H b AsH 3 -2 Cl ,C 6 H 4 .CH 3 0H + C gH 5 A s—A sC gH g * . . - 38 - 27 g. anisic aldehyde and 15 g. phenylarsine were treated as described in the previous experiment. After standing six weeks, the mixture was filtered and 12 g. arsenofeenzene ob- tained. The benzene was allowed to evaporate spontaneously. 15 g. anisic aldehyde were recovered. The residue remaining in the distilling flask was dissolved in hot alcohol and the solution filtered. On cooling, white prisms, weight 5 g. were obtained. The melting point, 172-174©, corresponds to hydran- isoin , 4 CH 3 OC 6 H 4 CHO + 2 C 6 H 5 AsHg *2 CE 3 OC 6 H 4 .CHOH .CHOH .C 6 H 4 OCH 3 + C 6 E 5 A s=AsC 6 H b Ketones. On long standing or heating in a sealed, C0 S ~ filled tube at 100©, both acetone and benzophenone oxidized phenylarsine quantitatively to ar senobenzene . This oxidation took place in the same manner in the presence of cone. HC1 or fused ZnCl s » Pyruvic acid, diacetyl and benzil reacted vigorously in the cold, the product being invariably arsenobenzene . Numerous experiments were made with these ketones in an attempt to iso- late an aldehyde-ketone condensation product , but none could be found. The reduction products of the ketone s, formed presum- ably by the action of phenylarsine, were not investigated. • ; : . . . . • • . " * . 39 Nitrobenzene » The action of phenylarsine on this substance was investigated as a possible explanation of the . . - - 40 - Reactions in which Hydrogen Bound to Arseni c is Exchanged for Halogen * Chloral . When chloral or chloral hydrate were allowed to react with phenylarsine a violent action took place. On dis- t illation of the reaction mixture much excess chloral was re- covered, A colorless fraction which boiled at 160-182© /70 mm., was identifed as C 6 H 5 AsClg by methods previously described in this paper. In Another experiment 3.5 g. chloral were placed in a COg-filled Claisen connected to a glass worm surrounded by a freezing mixture. 5 g. phenylarsine were added drop by drop through a cylindrical dropping funnel. The few drops of distillate collected in a well-cooled receiver were identified as acetaldehyde by boiling point and aldehyde test. Some arsenobenzene was obtained despite rigid exclusion of air. Doubtless this was formed by interaction of C 6 H b AsC1 s and CgHgA sRg » 3 C s H 5 AsH 2 f 2 CClg.CHO -2 CH3.CHO 3 C 6 H b AsC1 2 . Benzal chloride. 4 g. phenylarsine, 4 g. benzal chloride, 1 g. anhydrous Na 2 C0 3 , 25 cc. xylene were refluxed for 2 hours in a current of CO-,, The mixture darkened. On cooling, a white precipitate formed. It was filtered off, extracted with aqueous Na 2 C0 3 to remove NaCl and traces of phenylarsinic acid, and identified as phenylarsine oxide. The apparent course of the reaction was formation of C 6 H 5 AsC1 3w which gave phenylarsine (1) oxide by the action of Na s C0 3 . (l) Michaelia, B. 10, 623 (1877) . . . . . . . * . . . . . . - 41 - PART IV SUMMARY 1. In the presence of HC1 as a catalyzer, one molecule of phenyl arsine adds two molecules of aldehyde with the formation of products having the general formula, C 6 H 5 As(CHOHR) s . 3. These aldehyde -arsine compounds are remarkably stable toward water, acid and alkali. With most reagents the sub- stances are indifferent or act like a mixture of phenyl arsine and aldehyde. Only in the formation of addition compounds with halogen acid 3 and certain metallic salts, does this new type of compound preserve its distinctive character, 3. When heated with acetic anhydride, C s H 5 As (CHOHR) E , loses two molecules of alcohol with formation of a 1,4,3, 6 dioxdiarsine ring. Under certain conditions the dioxdiarsines are formed directly from phenyl arsine and aldehyde. The di- oxdiarsines resemble the first type of condensation product in mo3t reactions, but do not add halogen acid, 4. From analogy with aniline it was expected that phenyl- arsine and aldehydes or ketones would give compounds of the type, C 6 H 5 As=CHR, or C 6 H 5 As=CR 2 , when heated with a dehydrating agent. Instead, the phenylarsine reduces the aldehyde or ketone to the corresponding alcohol • 5. This reducing power of phenylarsine has suggested a more correct mechanism for the interaction of aryl primary arsines with arsine oxides, chlorides, etc. . . . . . . * . . . . ' . 1 1 - 42 - 6. In the case of substances containing active halogen, such as chloral, the only reaction is an exchange of hydrogen for chlorine, C 6 H 6 AsC 1 2 being formed* . ■ ACKNOWLEDGEMENT The author wishes to express his sincere appreciation of the direction of Dr. Roger Adams, which has been responsible for the success of this work. Thanks are due to Dr. C. S. Marvel for many helpful aids in experimental detail. . . ... . VITA I, Charles Shattuck Palmer, was born November 11, 1895, at Champaign, Illinois, the son of Arthur William Palmer, (deceased, 1904), Director of the Chemical Laboratory, Univer- sity of Illinois, and his wife, Anna Shattuck Palmer. After passing through the elementary and high school of Urbana, Illinois, I entered the University of Illinois in 1913. Follow- ing graduation in the chemistry course in 1917 with the degree of B. S., I became Professor of Chemistry in Greenville College, which position I resigned in December, 1917, to enter the U. S. Army. My discharge from the army in August, 1919, was followed by a year as Scholar in Chemistry, University of Illinois. The degree of M. S. was conferred in 1920. The year 1930-1921 was spent as Fellow in Chemistry (U. S. Inter- departmental Social Hygiene Board) * University of Illinois.