LOCAL ANESTHETICS BY FRANK LOUIS ROMAN B.S. University of Illinois, 1911 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, 1922 URBANA, ILLINOIS Digitized by the Internet Archive in 2016 https://archive.org/details/localanestheticsOOroma I wish to extend my sincere thanks to Professor Koger Adams for his many suggestions and kind interest during the investigations which are here described. TABLE OF CONTESTS Part I Preparation of Compounds of the Type 0 R l H p-nh 2 -c 6 h 4 -c -o-9 - ch 2 - n -tc 2 H g ) 2 1 Rg 01 Theoretical Part 1 Experimental Part... ••••••••••• 3 Preparation of diethylaminoacetone.... ..3 " " diethylaminodimethylethylcarbinol..... .4 " " the nitro ester 5 Reduction of the nitro ester. •••••• 5 Preparation of the monoohlorhydrate 5 rt ” diethylaminotertiarybutylparaaminobenzoate- mono chi or hydrate 6 " " monochlor tertiary alcohols 7 Part II The Structure of the Compounds Produced fromOlefines and Mercuric Salts. Mercurated Benzofurans ..••..•••.8 Theoretical Part 8 Experimental Part 11 Preparation of Phenyl Allyl Ether............ 11 Rearrangement of phenyl allyl ether to ortho allyl phenol. .11 Preparation of chloromercuri-l-methylbenzofuran 11 " ** bromomercuri-l-methylbenzofuran 12 M ** mercuric sulphate derivatives of ortho allyl phenol. 13 Reduction of Chloromercuri-l-methylbenxofuran to dibenzofuryl-l-methylnercury 15 Arsenic chloride - ortho allyl phenol addition products. ... 17 Bibliography •••••••••••• •••••• ••••••..••19 Part I Preparation of Compounds of the Type 9 ft ? p-HH2 -C6H 4 - C - 0 - 9 -CH 2 -N(C 2 H 5 )2 % Cl Theoretical Part, Because of their value as local anesthetics, the compounds of the 0 H type p~HH2-C6H4-C-0(CH2) x -N=E2 have been the subject of considerable study /i o) during the last few years. The best known of these compounds is novocaine' J ■» c, 9 ? or procaine, its chemical constitution being NH2-C5H4-C-O-CH2-CH2 -N*(C2Hg)2* C 1 diethylaminoethylparaaminobenzoatemonochlorhydrate. The corresponding dinormalbutylaminopropylparaaminobenzoatemonochlorhydr&te: 0 H ira 2"°6 H 4 ” G - 0 -C H 2 ~C H 2 -CH 2 - is known as butyn^J* The compounds of the two types: 0 H KH 2 -C 6 H 4 -C-0-CH2-CH 2 -N=B2 Cl 9 H nh 2 - c 6 h 4 - c-o-ch 2 -ch 2 -ch 2 -h=h 2 were the subject of extensive studies by Jenkins^), Burnett^, and Peet.^ To complete these studies it was necessary that the corresponding compounds containing forked side chains between the oxygen and the nitrogen be investigated, and it was the object of the following researches to prepare compounds of the type H 1 0 ?1 kh 2 -c 6 %-c-o-c -ch 2 -n=(0 2 h 5 ) 2 Cl JL O V. ' j* i l . : y j . i. v 2 Attempts -were made to prepare first the simplest compounds 0 CH 3 H m z -c 6 h 4 -c -o -9 -ch 2 - ]jr=(c 2 H 5 ) 2 ch 3 Cl 9 92H 5 ¥ -°6 H 4- C -0 -C - CH 2 - K=(C 2 H 5 ) 2 ch 3 Cl and as the results were unsatisfactory, the preparation of the more complex products was not undertaken. This investigation deals therefore only with the preparation of the above two compounds. There are several methods for preparing the novocaines^» 2 ) , the methods differing in general in the order in which the various intermediates are combined. The method chosen for this work comprised the following steps: 1. Preparation of the diethylaminoacetone^ 7 ) from monochloracetone and diethylamine. ch 3 -co-ch 2 oi+hn(c 2 h 5 ) 2 — 3 * ch 3 -co-ch 2 — n(c 2 h 5 ) 2 2. Preparation of diethylaminotrimethylcarbinol and diethylaminodi- methylethylcarbinol from diethylaminoacetone and Grignard reagent. on hydrolysis R CH 3 — C0-CH 2 -K ( C 2 H 5 ) 2 +R Mg Br ►ho -6 — ch 2 -n(c 2 h 5 ) 2 ch 3 3. Preparation of nitro esters from paranitrobenzoylchloride and the above tertiary alcohols. R P R H R° 2 -C 6 H 4 -C -Cl + HO -C — CH 2 -N(C 2 H 5 ) 2 •iT0 2 -C 6 H 4 -C-0-C — CH 2 -N=(C 2 H 5 ) 2 CR 6 h 3 Cl 4. Reduction of the nitroesters Pe 0 R II I OR H N° 2 -C 6 H4-C-0-C — CHg-N (C 2 H 5 ) 2 i -^-^ini 2 ^ 6 H 4 -C-0-C^CH -N(C 2 H 5 ) 2 CE 3 Cl CH 3 5. Preparation of the hydrochloride by neutralizing with alcoholic HC1. 3 9 ? HCl 0 R H HH 2 -C6B4-C- 0 -C-CH2-N(C 2 H5) 2 — > HHg-CgH.-C -0-C -CE, - N (C.HJ. oh 3 ch 3 Cl 2 The following modification of the above method was also tried: 1* Preparation of the monochlortrialkyl carbinol from monochloracetone and Grignard reagent. CHg-CO-C^-Cl + RMgBr — HO -C — CH 2 C1 ch 3 2. Preparation of the diethylamine derivative of the above product R R HO - 6 — CH -Cl ♦ N(C 2 H 5 ) 2 ^ HO - C — CHg -N(C £ H 5 ) 2 ch 3 ch 3 Prom this point, the steps were the same as outlined for the first method. Fairly good yields of the diethylaminotertiary alcohols were obtained, but esterification of these alcohols with paranitrobenzoyl chloride gave poor yields. These esters decomposed readily, even on slight warming in benzene, and on reduction and neutralization of the resulting amino esters syrupy pro- ducts were formed. Attempts to crystallize these products from absolute alcohol yielded only a few crystals. Experimental Part Preparation of diethylamino acetone ^ 7 ^ Thirty-four grams of chloracetone were added slowly to 54 grams of diethylamine in 200 cc of ether. The mixture was then heated under reflux for three hours, a heavy crop of crystals being formed. After standing for 24 hours the mixture was filtered, and the crystals of diethylamino hydrochloride on filter were washed with ether. Yield of diethylaminohydrochloride - 32 grams of snow white, large, flaky crystals. The filtrate and washings were distilled under reduced pressure, the portion distilling at 55 to 70®C at 14 - 25 mm being collected. This fraction . . ... • ., r . 1 ■ ■ ff . • , ' T was redistilled, the portion boiling at 58 to 65°C at 14 - 20 nun being collected Yield, 34 grains (72$) of light yellow liquid, with properties corresponding to those of die thy laainoace tone. Mg turnings were placed in a 1 liter flask and 200cc of dry ether added. The flask was connected with a reflux condenser provided with a calcium chloride tube. The flask was placed in a pan of cold water and 29 grams of ethylbromide were added through the condenser. The reaction was slow and the water bath we-s warmed gently. When the magnesium had nearly all dissolved, the ether mixture was boiled for 2 hours. to prepare the diethyl amino dimethyl ethyl carbinol. The flask was cooled, the water bath being replaced by an ice bath. 34 grams of diethylamino acetone were then added very slowly, froma dropping funnel, through the condenser. Each drop caused a very violent reaction at first, and constant cooling was necessary. A yellow precipitate formed after about one-half of the diethyl amino acetone had been added. The dropping funnel and condenser were finally washed with ether and the mixture in flask boiled for one hour. The mixture was then allowed to stand until the next day. The Grignard reagent obtained in the ether solution was used as such The addition product formed was hydrolysed with chips of ice A concentrated solution of fifty grams of ammonium chloride was added to dis- so ve the magnesium hydroxide formed. The addition product was sticky and dissloved only slowly. . O' . .. - ' I . . . . V . j • L' . ' - .. ' . . I . . I .. . . I • . ■: ■ . , ■ ■ c • • . i ■ • - • v.- : : . : i - r. . — . — — ; • f : ' ' •» * . ■ 5 The ether layer was separated from the water solution and dried over night over anhydrous sodium sulphate. The ether was then distilled off and the residue distilled under diminished pressure. The distillate which came over at 76 to 86 # C at 14 to 20 mm was taken as diethyl amino dimethyl- ethyl carbinol. Yield 17.5 grams (42$). Preparation of the Nitro Ester. ( 2 ^ Seven grams of diethylamlnodimethylethylcarbinol and 8.5 grams of para nitrobenzoylchloride were dissolved separately in the smallest possible quantities of benzene. The solutions were brought to room temperature and mixed within a few minutes with constant stirring. The mixture was kept on ice and the white precipitate which had formed was filtered off within twenty minutes. Cooling and filtration within a short time were found to be essential, as decomposition would otherwise occur, yielding a dark precipitate and finally a tarry mass. Yield of nitro ester, 5+ grams, or 30+$ of theory. Attempts were made to recrystallize the diethylaminodimethylethylcarbinolparanitrobenzoate- monoohl or hydrate thus obtained but without success, as even slight warming in benzene, acetone or ethyl acetate caused decomposition of the nitro ester. The original product obtained was white, however, and of apparently good quality. It melted with decomposition at about 160*0. Reduct ion of the Nitro Ester^ 4 ^ and Preparation of the Monoohlorhydrate The reduction of the nitro ester was carried on by means of iron dust and water, the HC1 in combination with the nitro ester being sufficient to produce the reaction. Only enough water was used to produce a thick paste and the weight of iron dust used was about three times that of the nitro ester. , . . . • . , , . , ] i ■ ■ . ... , . ■ 6 The reaction evolved considerable heat, and the container was cooled with ice, or chips of ice were added to the mixture whenever necessary in order to keep the temperature below boiling. When evolution of heat had ceased, the mixture was heated on a water bath for one-half hour to assure completion. Two methods were tried in extracting the amino ester from the mixture. In the first method, the mixture was made alkaline with NaOH and the ester extracted with ether. In the second method, enough tartaric acid in water was added to make the mixture strongly acid, and the mixture filtered at once. The iron dust on filter was washed first with water and then with IfaOH solution. The filtrate was made al- kaline and extracted 4 times with ether. The ether was evaporated off, leaving the amino ester. Considerable decomposition took place during this reduction, 10 grams of the diethylaminodimethylethycarbinolparanitrobenzoatemonochlorhydlrate yield- ing only about 2 grams (20$) of diethylaminodimethylethylcarbinolparaaminobenz- ate. This amino ester was a viscous oil with strong characteristic amine odor. To prepare the monochlorhydrate , the amino ester was dissolved in N absolute alcohol and titrated with alcoholic HC1. The resulting diethyl- 4 aminodimethylethylparaaminobenzoatemonochlorhydrate did not crystallize out on evaporation of the alcohol. Attempts were made to recrystallize this residue by redissolving it in the smallest possible quantity of a hot mixture of equal parts of absolute alcohol and ethyl acetate. Only a few crystals were obtained on cooling, however. Preparation of Diethylaminotertiarybutylparaaminobenzoatemonochlorhydratei ^^ Attempts were made to prepare this product by methods similar to those described in the preparation of diethylaminodimethylethylcarbinolpara- aminobenzoatemonochlorhydrate. The steps involved included: 1. The preparation of diethylaminoacetone as already given. • \ / • -J ■. -i •• “ i r » .-■* . . . . ' t r : . . ,i . : r . ’> ,« i ... 7 2. The preparation of the diethylaminotrimethylcarbinol' ' from diethylaminoacetone and Grignard reagent CH 3 Hgl. 5.6 grams magnesium, 33,1 grams methyl iodide and 30 grams of diethylaminoacetone yielded 8.5 grams of diethylaminotrimethylcarbinol, or about 25$ of the theoretical quantity. This product was nearly colorless and distilled at 55 - 60 # C at 15 to 20 mnupressure. 3. The preparation of diethylaminotert iarybutylparanitrobenzoate- hydrochloride by the method described under "Preparation of Nitro aster." 15 grams of diethylaminotert iarybutylalcohol yielded 8 grams of the nitro ester (27$ of theoretical quantity.) The product was white but some- what sticky and decomposed on heating. 4. Seduction of the nitro ester and preparation of the hydrochloride, diethylaminotert iarybutylparaaminobenzoatemonochlorhydrate. 8 grams of the nitro ester yielded 3 grams of viscous amino ester of light yellow color. The monochlorhydrate produced by titrating the amino N ester with --- alcoholic JTC1 until neutral to litmus, was of light yellow color, somewhat sticky and could not be crystallized. Preparation of Monochlor Tertiary Alcohols . Attempts were made to prepare (CH 3 ) 2 COH CH 2 Cl< 13 ’ 14) and C 2 H 5 CH 3 COH CH 2 Cl {14) , which on treatment with diethylamine , would yield the corresponding diethylamino alcohols Monochlor ace tone was treated with the Grignard reagents, CH 3 Mgl and C 2 HgMgBr being used for the respective monochlor tertiary alcohols. Addition with the Grignard reagent appeared to take place satisfactorily, but on hydrolysis of the addition products, only a few drops of the corresponding monochlortertiary alcohols were obtained. - . . . . . . . . . ■ . i • ' > ' I 8 Part II The Structure Of the Compounds Produced from Olefines and Mercuric Salts. Mercurated Benzofurans Theoretical Part. The compounds produced from olefines and mercuric salts have been the subject of controversy since they were first prepared. Probably the most generally accepted view is that mercuric salts form true additions to the ethylene double bond, while the opposite view is that molecular addition products are formed. Hoffman and Sand^ 5 ^ first discovered that ethylene and mercuric salts would react, and to the two products obtained in aqueous solution, they gave the formulas HO-C^-CHg-Hg-X and X-Hg-CHg-C^-O-C^-C^'Hg-X, because when treated with iodine, these two substances gave HO-C^-CHg-I and I-CH 2 " CHg" 0- -CH 2 -CH 2 -I respectively, the structures of which are apparently established. They noted, however, that the compound HO-C^-C^-Hg-X did not react in the same way as the ordinary alkyl mercuric halides when treated with alkyl iodides. Ordinary alkyl mercuric halides are not affected by boiling with alkyl halides while the ethylene-mercuric salt reaction products give the reaction: H0-CH 2 -CH 2 -Hg-X ♦ SI s. 0 £ % + Hg-X-I + ROH In addition, it was found that the ethylene-mercuric salt reaction products were decomposed readily by mineral acids such as HC1, even when a dilute solution of the acid was used. A typical decomposition is represented by the equation: H0-CH 2 -CH2-Hg-Cl + HC1 ^Hg Cl 2 ♦ HgO + C 2 H 4 This is quite different from the action of mineral acids on the simple alkyl mercuric halides, which are affected only by boiling with concentrated . . . , ' • V. ' • ; . . . • : , ■ ■ 9 acid. To explain these reactions. Sand assumed that the product HO CHg CHg H g X could take the tautomeric form CHp « CH 2 ^HgtOHlX. Mancbot(^) held that the products obtained by the action of mercuric salts on olefines -were not double bond addition products and that molecular type products such as °2 H 4 .Hg (0H)X and 2 C 2 H 4 .HgO.HgXg were actually formed. It has been pointed out, however, that the action of acids and alkyl halides on the ethylene-mercuric salts reaction products might be due merely to a special reactivity of the grouping in the true double bond addition product. Con- siderable evidence has been obtained in the following investigation to show that this is actually the case. Some insight in the nature and structure of these olefine-mercuric salt reaction produots could be obtained if products could be prepared which would condense intermolecularly to give new produots the structure of which might be readily determined. Orthoallylphenol gives with the nercuric salts condensa- tion products of this type, the mercurated benzofurans. CH - CH 2 + %x 2 *► O CH 2 -CHX-CH 2 HgX OH *» CHoCH CHgHgX / +HX 0 These reactions take place readily in aqueous solutions at room tempera- ture, the chloride, bromide and sulphate having been prepared during this investi- gation, while the iodide and acetate have also been prepared by other investiga- tors. ^ 17 ^ As might be expected, the initial addition product is unstable, but instead of hydrolysing as the usual ethylene-mercuric salts addition products according to the equation: H 2 0 CH 2 =CH 2 + HgX 2 > XCH 2 CH 2 HgX HO CHg CHg HgX +HX the orthoallylphenol- mercuric salt addition product condenses intermolecularly, splitting off HX. That this mechanism is correct is shown by the fact that the _ _ . . i — — *■ ' • . 10 above reaction takes place in absolute alcohol solution as well as in water. (I 7 ) In addition, the mercurated benzofurans are formed in neutral and acid solutions, while the preparation of the usual ethylene -mercuric salt addition products re- quires the addition of alkali to neutralize the acid formed. Sperry( 17 ) found that the chloromerouri-l-methylbenzofuran was not decomposed by 20$ HC1 after 20 hours at room temperature. Products which are thought to have been the primary ortho allyl phenol- mercuric salt addition products were found in small quantities in some of the preparations. They could not be isolated in the pure state, but when redissolved in hot alcohol, they yielded the corresponding mercurated benzofuran, which crystallized out on cooling. Further proof of the structure of the ortho allyl phenol - mercuric salt addition compounds was obtained by the reduction of chloro-mercuri-l-methyl benzofuran with Ha amalgam and treatment of the resulting product with mercuric iodide, yielding the iodomercuri-l-methyl benzofuran. It would be difficult to explain these reactions if the molecular type compound was assumed to be formed by the action of mercuric chloride on ortho allyl phenol. The mechanism of the reaction is very probably that shown by the following equations: Ha amalgam 2B HgCl — E Hg H ♦ 2 HaCl R Hg R + Hg Ig > 2 E Hg I Attempts were made to obtain additional proof of the structure of these ethylene addition compounds by preparing the arsenic chloride - ortho allyl phenol addition product. These attempts were not successful, but this was not entirely unexpected as Green and Priced®) have found that no reaction re- sulted when ethylene was passed in arsenic trichloride under various conditions of temperature and pressure. I t . ! . • . ... -i C » i .... c . ; j i • . • . _ .1 l i . ( : t ■ . . ■ . • $ ■ ' 1 11 Experimental Part Preparation of Phenyl Allyl Ether Phenyl allyl ether was prepared according to the method of Claisen^ 19 ) CgHgOH ♦ CgHgBr ♦ KgCOg s* CgHgO C S H 5 + KBr + KH COg 150 grams of phenol, 200 grams allyl bromide, 225 grams potassium carbonate and 150 grams of acetone (as solvent) were placed in & round bottom flask fitted with reflux condenser, and the mixture heated on the steam bath for 10 hours. To follow the progress of the reaction, samples were taken from time to time, shaken with petroleum ether and 1C >$ NaOH solution. The sodium hydroxide solu- tion was separated and acidified to determine if phenol would still be thrown out. Only a trace of phenol was found to be present at the end of ten hours. Water and petroleum ether were then added to the mixture in the flask, the layers separated and the petroleum ether solution thoroughly shaken with sodium hydroxide solution, After washing the ether layer with water, it was dried over anhydrous potassium carbonate. The ether was evaporated off and the residue distilled under diminished pressure. A yield of 176 grams (84$) of phenylallylether , distilling at 83 # C, under 20 mm pressure, was obtained. Rearrangement of Phenyl Allyl Ether to Ortho Allyl Phenol^ 2(9,21 ^ The phenyl allyl ether was refluxed until the temperature, which registered 188*C at first, reached a constant point of 218*0. The rearrangement was nearly quantitative, the product being practically pure ortho allyl phenol and requiring only a distillation under reduced pressure for complete purifica- tion. It distilled at 109-110°C at 22 mm pressure. f 17 \ Preparation of Chloromercuri- l- methylbenzofuran v 1 20 grams of mercurlcchloride were dissolved in 250cc of water in a 500oc round bottom flask provided with a stirring device. 10 grams of ortho - . . . . . . ■ ' • V . - • ■ • •"«' » . . . . . • ;•* 1 ' . . . . . 1 ; 1 • . . - c . • . • ■ . • . 12 allyl phenol were added very slowly from a dropping funnel, the time required for addition being about two hours, A voluminous white precipitate, partly crystalline, had formed, and the mixture obtained was stirred for an additional half hour, and allowed to stand over night. The precipitate was filtered off and dissolved in hot alcohol, from which it crystallized on cooling. The chloromercuri-l-methylbenzofuran thus obtained appeared to be a pure product, melting sharply at 136 # C (uncorrected). Yield 23 grams (84$ of theoretical quantity.) In subsequent experiments it was found that the chloromercuri-1- methylbenzofuran was also obtained in good yields when the mercuric chloride solution was made somewhat acid with HC1 (to repress HgClg hydrolysis and facilitate its solution) before addition of orthoallylphenol. In one experiment in which the mercuric chloride solution was made acid with HC1 (about 2$) and heated on water bath during the addition of orthoallyl phenol, the precipitate consisted in part of a pink product which was less soluble in the hot solution than the chlorcmereuri-l-methylbenzofuran. This pink product was a powder when dry and yielded chloromercuri-l-methylbenzofuran. It appeared to consist, therefore, of the initial addition product of ortho allyl phenol and mercuric chloride according to the equation: //Xnn ch 2 -ch =CH 2 * OH +HgCl 2 \CH 2 -CH - CH 2 HgCl 01 'OH I CH-CH-CHpHgCl 2 +HC1 Preparation of Bromomercuri- l- methylbenzofuran 3 grams of the mercuric acetate derivative of ortho allyl phenol were added to 200 cc of an aqueous solution containing 3 grams potassium bromide. The mixture was warmed on the water bath for one hour and allowed to stand over night . ■ - ;i i . . . . i . 1 .i . .... . i 1 • . . j • t. . , . * !■■■!) I I 1 ■ : ' * 13 The crystalline precipitate formed was filtered off and dissolved in hot alcohol from which it crystallized on cooling. The pure white shiny crystals obtained were filtered off and dried. Weight 2.7+ grains or about 90$ of the theoretical yield of the bromide CH 2 -GH -CHg-Hg-Br. These crystals melted sharply at 122 # C (uncorrected). The analysis of these crystals for mercury content gave the following results: Weight of sample in grams Stand. Ag NOg solution in cc. Stand. Na CH '» " " N.P. of Ag N6g solution 0.02523 N.P. of JJaCN " 0.02599 Mercury in per cent Average Theoretical I 0.4595 7.1 50.0 II 0.4040 6.3 44.0 48.77 48.86 48.81$ Eg 48.50$ Hg Preparation of Mercuric Sulphate Derivatives of Ortho Allyl Phenol At least 3 distinct products appear to have been obtained by the action of mercuric sulphate on ortho allyl phenol. The normal sulphate cor- responding to the formula [f) CHgCH-CHjHg Lv -0 J SO, was obtained as follows: 7 grams of mercuric oxide were heated with 5 grams cone. HgSO^ to form the mercuric sulphate. When the reaction had gone to completion as shown by a uniform yellow product, 200 cc of water were added. The mixture was heated to boiling and just enough I^SO^ added to cause complete solution of the mercuric sulphate. 9 grams of ortho allyl phenol were added drop by drop, and with vigorous stirring, to the mercuric sulphate solution. A solid light gray pro- duct settled in the bottom of the flask. Several attempts were made to purify the product, the best results being obtained by dissolving in hot solutions of acetic acid or dilute HgSO^. 1 . •>< . ' i Si 14 In general, the solid formed on cooling was an amorphous product, hut a solution containing about 5 $ acetic and 10 $ sulphuric acid which had been heated to boiling with the impure mercuric sulphate-rortho allyl phenol addition compound yielded on very slow cooling nodules of a white product* These nodules were filtered off, washed repeatedly with water and dried, a total of about one gram of pure white material being obtained. It did not melt on heating, but decom- posed suddenly at 123*0. It hydrolysed only slowly in concentrated hydrochloric acid, yielding ortho allyl phenol. Its analysis for mercury content gave the following results: Weight of sample in grams 0.3030 Standard Ag NOg solution in cc. 9.8 " JffaCN solution in cc. 40.0 N.F. of Ag NO 3 solution 0.02523 " NaCN " 0.02599 Mercury in per cent 52.45 Calculated $ of mercury 52.56 During the attempts at purifying the original addition product obtained, some of the material was dissolved in a boiling 10 $ solution of H 2 SO 4 , and a pink product of apparently uniform composition separated out on cooling. It was filtered off, washed repeatedly with water and dried, about one gram of amorphous powder being obtained. This powder yielded on hydrolysis a red oil which was not identified. This powder also decomposed without melting at about 180 # C. analysis gave the following results for meroury content: Weight of sample in grams 0.3675 Ag NOg solution in cc. 6.8 NaCN " " 50.0 N.F. of Ag NOg solution 0.02523 " NaCN " 0.02599 Mercury in per cent 61.56 A third mercuric sulphate - ortho allyl phenol addition product pre- pared by another investigator was a light gray powder somewhat similar to the impure product obtained in the first preparation described. It decomposed without melting at about 137* C. . . 15 Reduction of Chloromercurl- l- methylbenzofuran to form Dlbenzofuryl- l- methylmercury The mercuric chloride - ortho allyl phenol addition compound was re- duced with sodium amalgam, according to the equation: Ha amalgam 2R Hg 01 ^ R Hg R + 2HaCl 15 grams of the recrystallized chloromercuri-l-methylbenzofuran were placed in a 200 oc round bottom flask and 100 cc absolute alcohol added. The flask was connected with a reflux condenser provided with a CaCl 2 tube. 40 grams freshly prepared amalgam, containing 4$ sodium, were added, a few pieces at a time, through the condenser. The flask was heated very gently to start the reaction and to keep the mixture boiling for one-half hour. The crys- tals of the mercuric chloride- ortho allyl phenol addition compound had all disappeared at the end of that time and a light gray precipitate had formed in the bottom of the flask. When the flask was cooled on ice, no chloro mercuri-1- methylbenzofuran crystallized out, which indicated complete reduction of the above product. The precipitate was filtered off, and found to consist of sodium chloride with a trace of impurities. Its weight was 2.5 grams, or nearly exactly that of the theoretical yield of Ha Cl. Evaporation of the alcoholic solution on the water bath yielded a viscous residue, which crystallized only with difficulty. Spontaneous evapora- tion of the alcohol yielded, however, nodules of white crystals together with a viscous sticky residue. These crystals were dissolved in ether, a large por- tion of the impurities remaining undissolved. On spontaneous evaporation of the ether somewhat impure crystals again separated out. These crystals melted at 88 - 90°C. They were again separated twice from impurities by redissolving in 16 ether and recrystallizing* Pure white crystals were finally obtained, melting sharply at 93*C (uncorrected). Yield: 6 gram3 of pure product and about 1 gram impure material or about 50$ of theoretical yield. Analyses for mercury content gave the following results: I II Weight of sample in grams 0.4700 0.41! A g UOg solution, in cc. 4.0 2.4 HaCH " " 42.55 36.0 N.F. of Ag NOg solution 0.02523 " HqCU " 0.02599 Mercury in per cent 42.68 42.76 Average 42.82$ Hg Theoretical 42.98$ Hg 3 grams of the impure dibenzofuryl-l-methylmercury, obtained by evaporating the alcoholic solution on the water bath, were dissolved in 50cc alcohol. 3 grams red mercuric iodide, dissolved in 100 cc alcohol were added to the above solution and the mixture heated on the water bath, the flask being provided with a reflux condenser. The heating was continued for about one hour. Very fine white crystals formed at first, but larger crystals formed after some time, a heavy growth settling to the bottom of the flask after a few hours. 3.5 grams of somewhat impure crystals of iodomercuri-l-methylbenzofuran were ob- tained. M.P. 112*0. The above results would seem to indicate the reaction: fi 2 Hg + Hg I 2 ► 2E Hg I but when attempts were made to repeat the experiment using pure dibenzofuryl- l-methylmercury, no lodomercuri-l-methylbenzofuran crystallized out. It is thought possible that the impure product which was used in the first experiment contained enough ortho allyl phenol so that its reaction with mercuric iodide furnished enough H I to hydrolyse the dibenzofuryl-l-methylmercury in solution. The fact that the reduction product usually obtained from chloromercuri-l-methyl benzofuran contains appreciable quantities of ortho allyl phenol would tend to confirm this assumption. It has also been found that hydrolysis of dlbenzofuryl- . . . . . . . “ • . /< - ; ■ . • i , - 17 1-methylmercury produces some ortho allyl phenol, although strong acid is required at ordinary temperature. Arsenic Chloride - Ortho Allyl Phenol Addit ion Product s In the initial experiments anhydrous arsenic chloride and dry ortho allyl phenol were mixed in various proportions and kept at various temperatures, but no apparent reaction took place, and the arsenic chloride and ortho allyl phenol were recovered on distillation. Sdlutions in non-reactive sol vents, such as dry benzene, gave the same results. Catalysts such as anhydrous ferric chlor- ide and anhydrous aluminum chloride reacted with the phenol group and could not be used. An attempt was then made to replace the Hg Cl group in chloromercuri- 1-methylbenzofuran by As Cl 2 ^ 22 ^ as follows: 15 grams of recrystallized chloromercuri-l-methylbenzofuran and 50 grams of anhydrous arsenic chloride were placed in a 100 cc flask provided with a stopper and calcium chloride tube. The chloromercuri-l-methylbenzofuran crystal were quickly replaced, at room temperature, by a white amorphous precipitate, the solution taking a lavendar color which darkened on standing. The precipitate was filtered off and identified as mercuric chloride. Yield, 10 grams, or practically theoretical quantity for complete replacement of Hg Cl by As Cl 2 * The filtrate was distilled at a pressure of 20 mm to remove the excess of arsenic trichloride. The product remaining was a dark liquid, rather viscous, which de- composed when attempts were made to distill it under a pressure of 20 mm. In a second experiment, the residue remaining after distilling off the excess of arsenic trichloride was refluxed for 3 hours with 5 % HaOH. The solution was then neutralized with acetic acid to precipitate any of the arsenic compound O CHg -CH -CH 2 -As« 0 which had formed. A very viscous yellow liquid was _ ... > . - * 'V ■ • . r . • : ’ : • ' . . ' # . • ■ • - • . 1 .i . , 1 •* IWR > • , . ;• : • • ‘ • : ‘ " "" ' 1 ' ■ • , . . ' .v . 18 obtained. This product becams semi-solid on standing, but could not be re- crystallized. 19 BIBLIOGRAPHY (1) Einhorn - Annalen 371. 131. German Patents: 172568 189335 179627 194365 180291 194748 180292 (2) Adams and Karan - J.A.C.S. 42, 1030. Cham. Abstracts L5, 412. U. 3. Patent 1,358,750. (3) Kamm, Adams and Volwiler - U.S .Patent 1,358,751. (4) Jenkins - Thesis, Univ. of Illinois, 1921. 1 5) Burnett - " " " " " (6) Peet " " " " " (7) Stoermer and Dzimski - Ber.28j 11,2226. (8) Riedel - Friedlander 8^ 1030. German Patent 169819. (9) Grignard - Comptes Rendus 130. 1322. 132. 835. (10) Ullmann and Munzhuber - Ber. 36.404. (11) Einhorn - Eriedlander 8_, 995. German Patent 179627 U.S .Patent 812554 (12) Einhorn - Friedlander 8, 994. (13) Henry - Comptes Rendus 145. 24. (14) Tiffeneau - Comptes Rendus 134. 775. (15) Sand - Ber. 33, 2692. " 34, 1385 Ann. 329. 135. (16) Manchot - Ber. 53, 984. Ann. 420. 170. . . . . V . ■ . , , . ~ ; ■ , , . . . ' » . . . • . • ' t V* I & ■ ■ 20 (17) Sperry - Thesis, Univ. of Illinois, 1922. (18) Green and Price - J.Chem.Soc. 119. 448. (19) Claisen - Ann. 401. 29. (20) Jacobs - J.Amer.Chem.Soc. 39_, 2202. (21) Adams and Hindfnsz - J.Amer.Chem.Soc. 41, 654. (22) Boeder and Bias! - Ber. 4£, 2750.