A_V.:':3::) 1'"! THE REACTION OF ITSOBUTYLENE OXIDE WITH SOME ALKYLMAGNESIUM BROMIDE SOLUTIONS riFW;_fffi=A-i:rfl? ~15. g; Thai: for the beam of Ph. D. f MICHIGAN 'STATE come: , ‘ Robert Gem-go Braulf . 1948 THE REACTION OF IéflBUTYLENE OXIDE EITH SURE ALKYLMAGNESIUM BROMIDE SOLUTIONS 3.? Robert George Brault fl TE SIS Submitted to the School of Graduate Studies of Michigan State College of agriculture and Applied bcience in partial fulfiilment of the requirements for the degree of DOCTOR OF EHILOSOFHY Department of Chemistry 1948 5779/{3' W The author wishes to eXprese his sincere appreciation to Dr. R. C. Huston for his in- spiration and patient guidance so generously given during the course of this work. "\ v-— . y . _ . n k . A‘ -\ -. a 5 (1‘? 2w : r e, TABLE 93 comsms Introduction Historical Theoretical Discussion EXperimental I. II. III. IV. V. Materials Apparatus Reaction of isobutylene oxide with one~half, one and two moles of alsylmagnsium bromide solution A. Preparation of Grignard reagent B. Reaction of isobutylene oxide C. Analysis of the reaction product Reaction of isobutylene oxide with dialkyimagnesium solutions A. Preparation of dialkylmagnesium B. Reaction of isobutylene oxide with dialkylmafinesium Reaction of isobutylene bromohydrin With diethylmagnesium. A. Preparation of isobutylene bromo- hydrin 3. Addition of one mole diethyl- magnesium to two moles isobutylene bromohydrin Page 13 19 19 19 20 20 22 23 25 25 26 27 27 29 hm _W‘i ti 3.2:? .- * * a: 3. Addition or. two Iona ethyl-anon“ _ _ braid. to one mole isobatylm “in 58 VIII. Sentry _ 31. fl. tables X. Reference: INTRODUITION Work in this laboratory has revealed that the reaction between Grignard reagents and epoxy compounds is not a simple general reaction as is usually indicated. Henry (1) has shown that in the reaction between isobutylene oxide and methyimagnesium bromide the product formed is that product which would be formed it isobutylene oxide rearranges to isobutyraldehyde and methylmagnesium bromide then adds to it. Norton and Bass (2), on the other hand, have shown that when diethylmagnesium solution is allowed to react with isobutylene oxide the expected product, dimethyl propyl carbinol, is formed. When isobutylene oxide was made available to us, we decided to investigate these reactions. i .. fiat—351.1% .Haxlllwnql..nllt . . I uni l1 new [In 1907 Henry (I) reported that when isobutylene oxide is allowed to react with methylmagnesium.bromide, Sqmethy1~2-butanol is the principdl product formed. This is the product which would be formed if the iso- butylene oxide isomerizes to isobutyraldehyde and then reacts by the regular addition with methylmagnesium bromide. Henry suggested that this isomerism.is influenced by the presence of magnesium. In 1936 Norton and Hess (2) reported that when isobutylene oxide is allowed to react with diethyl~ magnesium, the proouct is 2Qhethyl~2~pentanol. This is the product which would be formed by breaking the bond between the oxygen and the primary carbon as in the reaction of ethylene oxide with Grignard reagents. In 1902 Blaisse (3) allowed ethylene oxide to react with ethylmagnesium bromide. The principle product after hydrolysis was ethylene bromohydrin. Grignard (4) studied this reaction and found that when the reaction mixture was heated by distilling off some of the ether, a violent reaction took place. After this "second phase" good yields of n—butyl alcohol were obtained. 3. ,Meieenhelner (5) in 192a analyzed the precipitate ‘which is formed.when ethylmagnesiun.bronide and ethylene oxide are allowed to react at -21°c. This analysis cerr38ponded to the molecular addition of the Grignard reagent to the ethylene oxide. He assigned the following formula to this product: caz\ (0235 ,0235 \ /o----Ms—-~-9 CH3 ‘Br can5 Huston and Agett (6) have shown that ethylene bromohydrin is the product formed by hydrolysis of the intermediate formed in the reaction between ethylene oxide and an alkylmagnesium halide. They showed that when dialkylmagnesium he added to the magnesium.elco- holate of ethylene bromchydrin, a primary alcohol is formed. Huston and Bostsick (7) have shown that the predom- inating product in the reaction tetween one mole of prepylene oxide and one mole of elhylmagnesium bromide is the magnesium alcoholate of proyylene bromohydrin. With equal nolecular quantities of these reagents the addition product is quite LtfiblO, giving small yields of the alcohols. When two moles of epoxide were used however, much better yields of the alcohols were obtained. They were able to Show analytical evidence for the presence of a product similar to the addition product siown below: l“4“||1.IHuWIv\MUr :. ,1. . .u .. .. . ldvuulmiflflll . t r... {I ll ‘c care— 933 + “33‘?“9533!‘ +- caa- cn~desa 035‘ on ones In his studies or the reaction or oyclohexene oxide with Grignard reagents which gives alkyl oyclo~ pentane carbinols, Bedos (8) has shown that cyclo- pentane aldehyde can be isolated in good yields and is probably an intermediate in the reaction. Cottle and Hbllyday (9) suggest that the primary alcohols usually formed in the reaction between Grignard reagents and ethylene oxide can be explained by the nucleophylic attack by the carbanion of the Grignard reagent upon one of the epoxide carbon atoms: c4739 + CH2.\JCH2—’ c4ngcszcazo °or by the neucleOphylic attack of the carbonion upon the bromohydrin: 0411 "+ Bronzcnao'fl c a CH CH 0" + Br- 9 d 9 2 2 However, neither of these reactions explain the forma- tion of 2-hexanol which is also formed in the reaction. When Cottle and Hollyday heated a mixture of ethylene oxide and magnesium bromide, a violent re- action took place and an acetaldehyde resin was formed. They reason that since a higher yield of 2~hexanol 5. is obtained when the reaction mixture is heated, acet- aldehyde is the intermediate which makes the formation of this alcohol possible. Huston and Bostwick (7) refluxed a mixture of propylene oxide and various alkylmagnesium bromide etherate solutions with benzene. These reagents, which give the expected secondary alcohols at room temperatures, gave large amounts of tertiary alcohols when heated. These are the alcohols which would be formed if propylene oxide rearranged to acetone and the Grignard acted on it. Even when this reaction Was run at room temperature and none of the tertiary alcohol was recovered, acetone and polymers of acetone were identified among the reaction products. Huston and Bostwick have proposed that‘O-gH-CH; is the intermediate which rearranges to acetone through an intermediate protonized double bond. \. .4 EEEQRETIGAL In the reaction between isohutylene oxide and diethylmagneeium annaddition product is formed which hydrolyzes to 2—methyl-2-pentanol. This addition product could be formed by going through an_inter— mediate addition compound by the electrOphylic attack of the magnesium on the epoxide oxygen: 033 CH3 c—cnz + (02H5)2Mg~—-> CH3 0%,032 Ms(0235)3 As a result of this electrophylio attack on the oxygen atom, one of the carbon-oxygen bonds is so weakened that the nucleophylic carbanion which is formed is able to add to the electrophylic, primary carbon atom: $35 + :' («‘33 engorgcsz ncng— I¢H 3032 0235 __> wig.- cnzczn5 nucznyz 241g 0 H5 _ ng-czas This is a slow reaction taxing six days or longer to be completed. It is suggested that the reaction of isobutylene oxide with magnesium bromide to form an addition product which can be hydrolyzed to isobutylene bromo- ’9 hydrin takes place through the same type of iatemdiate products: .~ 05. on » ca _ — ~ A ‘ - - + {u 0 sale 9'. lGHg-rEng-z—a ens (Kg-(“He —-> (:33 3 deg 31;. Mg Bra fig Br 0H5.§§§E; ‘5?= “—-> 035,§§3H2$?= e-Br file-BI? This reaction is very rapid, giving best results if the reaction is kept cooled with salt and ice and hydrolyzed after two hours. The difference in reactivity could be due to the fact that the magnesium-bromine bond ionizes more easily t;an the alkyl-msgnesium bond. The magnesium alcoholate was prepared by two different methods. Isobutylene oxide was allowed to react with magnesium bromide: 3 CH5 3% .. CH3~C -— 0512+ ng Bra —-> on ~ ,— (3- Mg ‘0’ cram- 2 It was also prepared by drorping dimethylmagnesium into isobutylene bromohydrin: 9H5 ( ) figso CH‘C—CHBr+ CH hg—+ CEr — E Mg son 2 2521 écngarz The precipitate formed in each case was analyzed, and the halogen content use 10? indicating that the 3? addition products were not pure. Each of these predueta gave isobutylene bremohydrin as the.mein product when hydrolyzed. However, isobutyraldehyde and isobutyl alcohol are also found in the hydrol- yeie mixture. These products indicate that the addition compound is unstable and undergoes re- arrangement: . 935 - CH3- CH5- - 0- Mg ___,...). 2 CHg‘Q" 0 +MgBr2 CfigBlj 2 . 032+ 93 J on 033- 0 ~90 —‘* 0115- c gone cm 2% 2 __ - , 0H3 + This is similar to the transformation of O-CH-CHz 0 to CH3‘6-CHS which Huston and Bostwiok (7) Show as one of the interneaiates in the decomyoeition of a similar magnesium dibrouonydrin olcoholate. They progose that this transfornation requires the inter- mediate formation of a grobonizod double bond. It should be noted, LoneVor‘tnat the ion (‘fi3)§Q-O CH§* ' gas must undergo a preliminary rearranzement to CH5'9+ H90- H before the hydrogen can ehift to its position on the branched carbon. . ... ||...i. ...I. Illalfllfifl-nhill: :...|..... .|.. .l .. n. . .... . ....L. .h. ”.n...:.n....m.m.l{ . .. filial». .Ililvll 9. The addition of isobutylene bromohydrin to diethyl- magnesium gives, on hydrolysis, 2-methyl-3-pentanol. This alcohol is also formed 1: diethylmagnesium is allowed to react with isobutyraldehyde. It appears reasonable to conclude, therefore, that the intermediate 033 01133—01123:- Ms 0 CHg¢— anger on 3 CH .4.— will dissociate to form CEE-Q-gCHz in the absence of At 0— added CgfisMgBr of MgBra. The reaction between one mole of ethylmagnesium bromide and either one-half or one mole of isobutylene oxide produces an addition product which on hydrolysis gives 2-methyl-3-pentanol. CHE 3H8 CHgH ca3 q; 32+ 02H5MgBr——>~ CHBH— HO +02H5MgBr was)? (502% MgBr This is a comparatively fast reaction, being completed in twelve hours or less. We have shown above that the magnesium alcoholate or isobutylene bromohydrin rearranges Spontaneously to isobutyraldehyde. This rearrangement must take place in the reaction between Grignard reagents and . 1 ..v. . . . .. .r- .... . . 1.1111..hl1.l.l|.1 ill“?! ...-.1! . I ..n flitting lO. isobutylene oxide, since iSQbutyraldehyde was found among the reaction products in every case. The following series of reactions shows the action of ethylmagaesium bromide on isobutylene oxide to produce 2~methyl—5~pentanol: Q3: 6235‘ big-3:2 93 93 01137033 c212 ~____, cas— —c~—a 02:33:12 C285 " M ug“ BI". M3C2H5 CH5 c3313 — 0- ~24 éC=O+CZH5MgBr—> Hag—— ggeBr 2 5 11 CH5 H CH3 9— c~ gBr + H20 —> 035- ci C~0H + LigBr(0H) H C ‘ V CH3“ 03"; 235 i 235 ll. The above reaction takes place any time that isobutylene oxide is allowed to react with an alkyl- magnesium bromide solution. The products formed when oneuhalr male or one mole of isobutylene oxide is allowed to react with one mole of Grignerd reagent are 2-methy1-3—pentanol, isobutylene bromohydrin and isobutyraldehyde. 'Besides these three products, when two moles of isobutylene oxide are allowed to react with one mole of Grignard reagent, 2-methy1~2-pentanol and the trimer of isobutyraldehyde are formed. The Zemethyl- 2-pentanol is formed because the excess epoxide present allows the reaction shown in the following equation to compete with the normal reaction. 033 cn5c—cazar 3H3 9 . __ I A. 2 CH3 ‘O’CHB + CZHSLgBr-——+ gg CH (3— on c 3033 2 2H5 One molecule of the epoxide reacts with the magnesium- bromine bond while a second molecule reacts with the ethylmagnesium bond. The formation of the trimer of isobutyraldehyde may be attributed to the accumulation or the aldehyde by rearrangement of the ion (CH3)2-q-Cfi; (See reactions Q. DISCUSSION The formation of a coordination compound is generally regarded as the first step in all Grignard reactions (10). For simplicity we have omitted the other when we show the complex in step I. This com- plex is undoubtedly the same type as that preposed by Heisenheimer (5) for ethylene oxide which includes ether: fl egg; ,czfis (593 CI ‘? :Mg : O ‘4 5 2c” \ Br bass The electrophylic attack of magnesium on the oxygen upsets the electronic equilibrium so that one of the carbon-oxygen bonds is weakened. If an alkylmagnesium bromide is the reagent, the bromide, being partially diaplaced from the ma nesium, attacks the primary carbon nucLeOphilically. CH 9H3 -+ QH w 3 3 Cn--C-.~CH .__,~. one—c; cs :7 ....) CH.-c— CH Br CZHSMgBr CBHSMg hgczfis The magnesium alcoholate of the bromohydrin is thus formed as shown in step III. If this bromohydrin alcoholate is stable under 7 the conditions of the reaction, as i\ the case with T Li. W ethylene brmehydrin alcohols“ (c) and the propylene bromohydrin alcoholato (7), the bronohydrin is the product formed. Isobutylenc bromohydrin a1- coholate is not stable even at room temperature. It undergoes rearrangement as shown in steps III and 1113. This is true whether the reaction is carried out with molu'equivalents, or whether an excess of either Grignard reagent or epoxide is present. This is indicated by the alcohols which are formed and by the isobutyraldehyde found in every reaction. It is even true when no Grignard reagent is present as evidenced by the isobutyraldehyde and isobutyl alcohol found when the magnesium alcoholate of isobutylene bromohydrin is hydrolyzed. Consider— ing this evidence we can say that the rearrangement is catalyzed by the magnesiumubromine bond even when this is formed by the decomposition of the magnesium alcoholate. We obtained the trimer of isobutyraldehyde as the main product when we heated the ether solution of the magnesium alcoholate of isobutylene bromo— hydrin before hydrolysis. This indicates that heat promotes rearrangement. Cottle and Hollyday (9) have preposed the fol- lowing mechanism for the aldehyde formation in the 15. reaction between ethylene oxide and butylmagnesium bromide: ' (arcnzo)"+ R0-—-? Ron + (Bx-0212030): (31033030): —-> Br" —r (cnzcaof' (angelic)? ROH ——-—> 021301-10 + Ro" After preposing this mechanism, they suggested that the removal of the second proton from ethylene bro- .mohydrin would only be accomplished with such dif- ficulty that such a meoheniam seems unlikely. The fact that isobutylene bromohydrin rearranges to an aldehyde also makes this seem unlikely as a general mechanism since there is no hydrogen on the carbon atom to which the oxygen is attached as in ethylene bromohydrin. + - The transformation of CHZ‘Q-O to CH3’C=O sug- 01:3 CH3 gested by Huston and hostwick (7) can be apgliod to the transformation of isobutylene oxide to isobutyr- aldehyde, as we have shown in reaction Illa. Step IV is the normal addition of the Grignard reagent to the aldehyde, and step V is the hydrolysis. Eeaction A shows how both 2—metlyl—5-p.ntanol and 2-methyl~2-pentanol can be formed in the reaction between one mole of ethylmagnesium bromide and two moles of isobutylene oxide. If the bronohydrin glo- 16. chalets rearranges as we have shown before, lsobutyr- aldehyde is formed. Since the reaction.between the epoxide and the ethyl magnesium.bond is a slew react- ion, some is available to react with the aldehyde and form Zemethyl~3-pentanol. The reactions, when the ethylmagnesinm bromide was drapped into the epoxide, show that an addition product like that in reaction A is formed only if excess epoxide is used. Hhen one mole of the Grignard reagent was drOpped into one.sole of isobutylene oxide, Semethyl-b-pen- tanol and some polymer were formed. Ehen one mole of the Grignard reagent is drOpped into two moles of isobntylene oxide, on the other hand, isohutyl alcohol, Remethyl-Z—yentanol and polymer are formed. No 2-methyl-S-pentanol was isolated. The formation of polymer in the former case as contrasted with no polymer in the normal addition indicates that, although the rearrangement is Spontaneous and does not need Grignard reagent to catalyze it, if insuffi- cient alkyl group is present, the aldehyde polymerizes. An addition product similar to that in reaction A must be formed when two K0183 of epoxide were used. This is indicated by the tertiary alcohol formed. 17. The bronohydrin alcoholate rearranged and polymerized since there was no alkyl group present to react with it. ~ . Isobutyl alcohol was present every time polymer was found. This indicates that it much isobutyrale dehyde is present, reduction takes place. Conant and Blatt (12) have shown that reduction is a common reaction between branched aldehydes and Grignard reagents. The polymer of isobutyraldehyde which is formed in these reactions is of uniform composition, probably the trimer. It distills at the boiling point of the trimer of isobutyraldehyde until the distilling flask is dry. When heated with a drop of sulfuric acid, it decomposes into isobutyraldehyde. Gaurizchi and Garzino (11) have shown the struc- ture of isobutylene bromohydrin by preparing the sodium sulfate and comparing its melting point with that of (CH3)20(OH)CH2803Na whose structure was known. We prepared a bromohydrin by the action of methylmagnesium bromide on bromoacetone. The 5,5-dinitrobenzoate of this preparation gave no melting point depression when compared with the same derivative of the bromo- hydrin prepared by the action of magnesium bromide on isobutylene oxide. n . ul- u-lrlt ....- 18. The yields of alcohol are in good agreement with the results expected from consideration of steric effects; the yields are lower as the alkyl groups become more branched. The action or diisoprOpylmagne- sium on the epoxide is the major exception to this rule, This reaction was carried out three times and no 2,4- dimethyl-z—pentanol was detected. When tert—butyl— magnesium bromide was allowed to react with the epoxide, some polymer was formed but no alcohol was found. This was the only case that polymer was formed when one-half mole of epoxide was added to one mole of Grignard reagent. The fact that polymer was formed indicates that the alkyl group was not reactive enough to add to the aldehyde before the aldehyde polymerized. The polymer of isobutyraldehyde is formed only when the magnesiumpbromine bond is present to catalyze the rearrangement to isobutyraldehyde, and then only when there is not enough magnesium-alkyl bond to react with the aldehyde. The only exception to this rule is the reaction of tert-butylmagnesium bromide with iso- butylene oxide. In this case one must assume that steric considerations slow the addition of the alkyl- magnesium to the carbonyl to the extent that none of the expected alcohol is formed. I. HAEERIALS: Isobutylcne oxide furnished by the Shell Chemical Company was purified before use by redistillation. Alkyl bromides were all dried over anhydrous calcium chloride and fractionated before use. Magnesium turnings were dried overnight in an oven at 110°C. Bromine was dried by shaking over concen~ trated sulfuric acid. Anhydrous diethyl ether was rsdried over metallic sodium before being used. Dioxane was purified by refluxing over dilute hydrochloric acid, drying, refluxing over sodium and distilling. II. PPARATUS: The apparatus described below was used for all of the reactions. The Grignard reagents and the magnesium bromide etherate solutions were also prepared in it. A one liter three necked, round bottom flask was fitted with a reflux condenser, III. a glycerino sealed stirrer, and a Eiruhberg dropping-tunnel. All or the.etherate solutions were protected from atmospheric moisture and carbon dioxide with calcium chloride and soda— lime filled tubes. REACTION OF ISGBUKYLE 13E 32(IDE fiITfl ONE-HELF, ONE all!) TV. '0 MOLES 0F ALn’YilmG—NHCIUD BROMIDE SOLUTIONb A1 Preparation 2;.Grigggrd reagents. One and oneutenth moles (26.5 g.) of magne- sium turnings were placed in the flask. Approx- imately ten milliliters of a mixture of one mole of the alkyl halide and an equal volume of anhy~ drous other were drotjed into the flask and the mixture stirred until the reaction was started. Four hundred milliliters of anhydrous ether were then added to the flask and the remainder of the alkyl halide solution was added dronise. After the alkyl halide addition was complete, the stirring was continued for two hours. The primary alkyl Grignard reagents were prepared by adding the alkyl halide at such a rate that a gentle reflux was maintained. The best yield of isoprOpylmaunesium bromide was obtained when the leap mp5l bromide .;as added 31. slowly enough at room tmmture that the reaction mixture did not warm up enough to cause the ether to reflux. About twentyufour hours were taken to prepare one mole of isopropyl- magnesius bromide. Tert-butylmagnesium bromide was most success- fully prepared when the reaction flask was cooled to - 0°c-with ice and salt and the tert~butyl bromide was diluted with twice as much ether as usual. The addition of the magnesiumrether slurry was made over a twenty-four hour period. Methylmagnesium bromide was prepared in essentially the same ap aratus. The dropping funnel was replaced by a tube leading from the tank of methyl bromide gas to below the surface of the ether. All of the ether was placed over the magnesium, the stirrer started and the methyl bromide was bubbled into the other at such a rate that a gentle reflu: was maintained. The Grignard solutions were decanted from the excess magnesium into a graduated cylinder. Two milliliter aliquots were removed and analyzed by Gilman's titration in the following way: The aliquot was hydrolyzed with water, an excess 33: or standard hydrochloric acid added and hook titrated with standard sodium hydroxide solution to a phenolphthalein endpoint. B. Reedtion.with isobutxlene agigg, The Grignard reagent which had been pre- pared and analyzed as described above was placed in a clean dry apparatus._ The flask was cooled by surrounding it with crushed ice and salt. The calculated amount of isobutylene oxide was mixed with an equal volume of anyhdrous ether, placed in the dropping funnel and allowed to drop slowly into the cooled Grignard solution. When the addition was complete, the temperature was raised to room temperature. The mixture was hydrolyzed after standing twenty-four hours. In the case of Grignard solutions, use of Michler's ketone showed that no Grignard reagent was left after twelve hours. The reaction was hydrolyzed by cooling the mixture with ice and then adding 175 to 225 milliliters of saturated ammonium bromide solu- tion dropwise. The other solution was decanted from the basic precipitate and dried over an- hydrous sodium sulfate. Analzsir g£.thg reacgigg products. 1. Decomposition and analysis or the bromohydria. The dried products of hydrolysis were refluxed over a mixture of forty grams of sodium hydroxide and two hundred milliliters of water for two hours with vigorous stirring. The mixture was cooled, the layers were separated, and the aqueous layer was ex- tracted three times with twenty-five milli- liters of ether. The ether extracts were re-dried over anhydrous sodium sulfate. The aqueous layer was diluted to one liter. Five milliliter aliquot portions were removed and titrated for bromide ion by the Volhard method. 2. Alcohol distillation. The dried other solution containing the alcoholic products was distilled at atmosPheric pressure. The results of these distillations are collected in Tables I and II. 3. Alcohol derivatives. The 3,5-dinitrobenzoates were prepared using 3,5—dinitrobenzoyl chloride and pyridine (13). Alpha-naphthylurethano and phenyl- urethans were prepared from the corresponding isoeyanatee using a drop or 5% trimethyl- amine-ether solution as catalyst (14). The acid phthalates were prepared by refluxing the alcohol with phthalic an- hydride in pyridine. (15). Isobutyraldehyde was identified in the fraction boiling between 55 and 65°C. of every reaction. This was shown by preparing the 2,4-dinitrophenylhydrazone and comparing it with the same derivative of pure isobutyral- dehyde. I‘:'£.p. 185°C. (16). In the reaction between one-half mole of tert-butylmagnesium bromide and one mole of isobutylene oxide no 2,4,4 trimethyl-E-pentanol was found. About eight grams of polymer were recovered. No polvmer was found in any of the other reactions between one-half mole of epoxide and one mole of Grignard reagent. then two moles of ieobutylene oxide were allowed to react with one mole of a Grignard reagent, isobutyl alcohol and a polymer were isolated. The isobutyl alcohol was identified by its boiling point; 106-108°0 and by preparing as. its 3,5wdinitrcbenzoate, which gave no depression in melting point when mixed with the same deriv- ative prepared from known iscbutyl alcohol. The polymer was found to contain no halogen. When heated with a drOp of sulfuric acid, it slowly depolymerized into isobutyraldehyde. Its physical properties, B.p.17mm127°C.-; n801.4370 indicate that it is the trimer of isobutyraldehyde. n%71.4529 (17); B.p.105-1oe°c (29) IV. REACTION OF ISOBUTYLENE OXIDE WITH DIALKYL- MAGNESIUM SOLUTIONS. A. Preparation gg'gialgylmagnesium. One mole of Grignard reagent was prepared in the manner described in IIIA. A mixture of 110 milliliters of dioxane and 100 milliliters of anhydrous ether was slowly added to the Grignard reagent to precipitate the bromine- containing com ounds. This mixture was allowed to stand from three to five days before the dioxane precipitate was separated from the dialkylmagnesium by centrifuging. The dialkyl concentration was determined by the Gilman methOdo 26. 3 Reaction 2; isobutylene oxide with dielgzlv . magnesium. The dialkylmagnesium was cooled and a mixture of the calculated amount of isobutylene oxide in anhydrous ether was slowly dropped into the cooled dielkylmagnesium. Since this was a much slower reaction than those involving the Grignard reagent, these reactions were allowed to stand from one to three weeks before they were hydrolyzed with saturated ammonium bromide solution in the same manner as reaction III B. The other solution of the products was dried over anhydrous sodium sulfate. The dried mixture was distilled, and the results are shown in table I. Ditertiarybutylmagnesium was prepared by using tert-butyl chloride because better yields of tert-butylmagnesium chloride are obtained than of tert-butylmagnesium bromide. Dimethylmagnesium was found to be so pyro- phoric that it was not possible to handle it in the Same manner used with all the others. The dimethylmagnesium was separated from the dioxane precipitate by adding an excess of ether, allowing the precipitate to settle and decanting the super- nattnt ether solution. As this dialkylmagnesium 89. m not analysed, no quantitative information is available for the reaction between dimethylnegnesiun and iscbutylene oxide. Y. REAGTIONS 0F ISOBUTYLENE BROEOHYDRIN WITH DIETEYLg MAGNESIUM A. Preparation 9; isobutylene bromohydrin. The apparatus for the reaction was the same as that used to prepare the Grignard reagents except that all connections were ground glass Joints and a mercury sealed stirrer was used. One and one—tenth moles (26.5 g.) of magnesium turnings were placed in the flask with 500 milli- liters of anhydrous ether. One mole (26 ml.) of bromine was slowly dropped into the flask. After the bromine addition was cowplete , the solution was refluxed for an hour and allowed to stand over night. The mixture was decanted from the unreacted magnesium into a graduated cylinder and an aliquot was titrated for bromine by the Volhard method. The other solution was placed into another clean, dry flask and then cooled in a mixture of salt and ice. The calculated amount of isobutylene oxide in an equal volume of ether was.dropped slowly into the magnesium bromide solution. .... Ifuun ... ..ul..u|_r_u.lnr. .. .... . .Ilnl. 38. The solution was kept cool and stirred for two hours after which it was deopqposed in ice. The basic salts were dissolved in dilute hydro- chloric acid. The ether layer was separated, shaken with five percent sodium bicarbonate solution and then with water and dried over anhydrous sodium sulfate. The ether fraction was removed until the boiling point rose to fifty degrees, after which reduced pressure was used and the isobutylene bromohydrin fractionated at 14 to 16 millimeters. Fifty-six grams were recovered between 58 and 60°C. This represents a 36.6% yield. A precipitate forms in this reaction, In one case this precipitate was removed and washed with aniydrous other by centrifuging and dried in a vacuum desiccator. This precipitate was analyzed for bromine and magnesium. 1. Analysis of the precigitate. Magnesium was determined by decomposition of'weighed samples with Maker burners and weighed as age. The percentage magnesium Was calculated from the formula: Wt. of ago x 0.6032 x 100 . “ st of samfifgtrtr* % magnesium 29. Bromine was determined by the Valhard method. doighed samples were hydrolyzed in fifteen milliliters of water and enough nitric acid was added to dissolve the precipitate. Seven grams of sodium hydroxide were added and the samples were refluxed for one hour. The samples were then cooled and diluted with one hundred and fifty milliliters of water, made acidic with dilute nitric acid and titrated with one-tenth normal silver nitrate and potassium thiocyanate, using the ferric alum indicator. Calc'd for 08H1602MgBr : Mg, 7 31; Br, 48.78 Found : lag, 7. 67; Br, 42.30 In another experiment the reaction was re— fluxed over night before hydrolysis. When this reaction was hydrolyzed, dried and distilled as usual, the only product found was the polymer of isobutyraldehyde. B. nddition of one mole diethylmagnesium_ to two moles isobutylene bromohydrin. One-fourth mole of diethylmagnesium was prepared as described in IIIA. It was analyzed and the calculated amount of isobutylene bronc- hydrin in an equal volume of ether was placed in 30. a 500 milliliter reaction flask. The diethyl- magnesium was placed in the drapping funnel. The flask was cooled with salt and ice and the addition started. A precipitate formed immediately. When the addition was complete, the solution was stirred for two hours. Some of the precipitate was removed and washed with other by centrifuging. The precipitate was dried in a vacuum desiccator and analyzed for.magnesium and for bromine. The solution was hydrolyzed with saturated ammonium bromide solution as usual, separated from the precipitate and dried over anhydrous sodium sulfate. Calc'd for 08H1602Mg8r2 : Mg, 7.51; Br, 48.78 Found : ng, 9.40; Br, 42.91 1. Analysis of the hydrolyzed solution. The dried solution was fractionated. Isobutyraldehyde, isobutyl alcohol and iso- butylene bromohydrin were identrifiad. C. nddition.9£.one mole isobutylene bromohydrin 32.0ne mole diethylmagnesium. One-fourth mole of diethylmagnesium was prepared as a ova, analyzed and placed in the reaction flask. The calculated amount of iso- butylene bromohydrin was placed in the drOpping funnel and drOpped slowly into the cooled diethyl- magnesium. A precipitate formed which was sepa- rated and analyzed as above. v1; Calo’d far 0 H’ GgMg Fennel 1'2 2'6 Mg, 10.73 .‘I .0 PREPARATION OF ISOBUTYLENE BRGHOHYDRIH Mg, 9.64; 51', 5.95 Isobutylene bremohydrin was prepared by acetone. Bromoeoetone was prepared by Levene's procedure (18): the action of methylmegnesium-bremide on bromo- A three liter, three-necked, round- 31. bottom flask was provided with an efficient mechanicer'stirrer, a reflux condenser, a thermomemer, and e 25 milliliter se sratory funnel, the stem of which reac nearly to the bottom of the flask. Th «7: oyjaretus ass pl 30526. in 8 ::"'3‘I}er 135th. bed is 800 milliliters of water, 250 milliliters of acetone and 166 milliliters of n the P30 acetic acid were gleced The stirrer was started and the of the water bath raised to 70— 174 milliliters (3.6 moles) of were carefully added through tb funnel; A ter the bromine additi temnerature I. 80°C. bromine Then e separatory '5 331 WU; S completed and the solution was colorless, it was diluted with 400 mill ll ters of 38; can water, cooled to 10°C, made neutral ' to conga rod with solid anhydrous sodiu- oarbonato. The oil which separated was fractionated and the traction boiling at 40-42% was taken. Yield 119 grams (25$). Ono-half mole of methylmagnesium bromide was prepared and analyzed as described in IIIA. One—half mole (68.3 g.) of bromoacetone was placed with an equal volume of other in a three liter, three-necked, round bottom flask and cooled to 0°C with salt and ice. The Grignard reagent was added drOpwise from a separatory funnel. The reaction mixture was hydrolyzed immediately after the addition was completed with saturated ammonium bromide solution. The other solution was dried over anhydrous Sodium sulfate and fractionated. Twenty milliliters of isobutylene bromohydrin was obtained. B.p.15mm47°C. The dinitrobenzoate of this alcohol was prepared and it was shown by the mixed melting point method to be the same as the broachydrin obtained by the action of magnesium bromide on isobutylene oxide. M.:. 120°C. 33. VII. aggggIOfi 0T ETHIEKAGHESIUH BRGKIES T0 ISOBflTILERB Ono mole of ethylmagnesium bromide was prepared and analyzed as in IIIA. A mixture of the calculated amount of lsobutylcne oxide in an equal volume of anhydrous other was placed in the flask and cooled with salt and ice. The ethylmagneslum bromide was placed in a separatory funnel and slowly dropped into the flask. After the addition was completed, the reaction was stirred for two hours. It was hydrolyzed and dried as usual. The bromohydrln was decomposed as in section III and the bromine titrated. Yield: 27.26% bromohydrln. The dried other solution was distilled. Yield: 13.7% 2-mothyl-3-pentanol; 3 g. polymer. B. dealtuuaei_oaezmflsLe12$hreosod¥o.hnodhnz £2,339.colee.lechusxleos.calds. One mole of ethylmegneslum bromide was prepared, analyzed and allowed to drop into a mixture of two moles of leobutylene oxide in an equal volume of other as above. The reaction products were analyzed similarly. Results: 63.15% brcmohydrln; 5.8% isobutyl alcohol; 8.5% 2—methyl-2-pentanol; 11 g. polymer. No 2-methy1-5—pentanol could be detected. w- l. 2. 3. SUMMARY Isobutyraldehyde is an intermediate in the formation of ieopropyl alkyl carbinol by the action or alkylmegnesium bromide on isobutyl» one oxide. The magnesium alcoholate of isobutylene bromo- hydrin undergoes rearrangement to isobutyraldehyde o to a slight extent at O C and to a greater extent when heated. Isobutylene oxide does not rearrange to lsobutyr— aldehyde when treated with dialkylmagnesium. The magnesium—bromine bond is necessary to catalyze the rearrangement of isobutylene oxide to iso- butyraldehyde. The addition of two moles of alkylmagnesium bromide to one mole of isobutylene oxide produces both the alcohol formed by the rearrangement of to ilebutyreldohydo and the alcohol rem-ed by the direct addition or on eikylnegnoliun to the opoxido. G. The structure or isobutylene brenohydrin is shown to be (023)30taEJOEéBr by its preparation by the action or methylmagnesiun bromide on bronc- acetone. .oanuaae>¢ ma eowpaeoomsa 0>fipdpfl#fidfid OS 95.9 «G0 “(Pdoflwfluysmwvfl LOH UOQHWPQO mam? HO£OOHN nwflwwOHMMSm * a.» 0.8 9mm 3253 Regions mi 5.3 :6; 8263 Havana 9.00 m.H~ ~.mm oewaoan ahaoumUOww m.m~ mi: 0.3. 8 2.3 8.2 «gm afimm amen 830.3 fleeting o.mm hm.Hm 4.0: 0H m.~H ©H.nH mm.Hm :.N: -.ww_ cowaoup Hanum m ...... 93 «.3 825.5 one»: Honooae Homeono cane»: heahaoa Hozooae Honooae catch: Hocooam sauna: hhuafluo¢ .oou oceans madam .vnop .oom lagoon humonooom teach» «avenger? “use ”amoxlowmmo m ~on H+ ~=o\W/oneo N Home 3 «Minamoto H and vegan nae . 95 new ma enema?" nudges cwuomnoeohm oeoampspoeH one mHoaooa< we named» nuanceonum H wands Table II Physical Constanta of the Alcohol: Alcohol use B.P. QC 1-8:amo~2~methy1-2-propanol 1.4710 49.5 (16mm) Sauethyl—Z-butenol 1.3973 110-112 (745mm) BéMethyl-z-butanol 1.4020 102 (740mm) z-Methyl-S-pentanol 1.4168 127-128 (740mm) 2~Methy1-2-pentanol 1.4125 117-118 (740mm) Z-Methy1~5-hexanol 1.4178 142-145 (740mm) ZéMethyl-2~hexanol 1.4175 139-140 (740nm) 2,4—Dimethy1-3—pentanol 1.4250 137-138 (740mm) z-Methyl-fi-heptanol 1.4259 165-167 (740mm) B-Methyl-Z-heptenol 1.4248 65 (15mm) 2,4,4-Trimethy1—2-pentanol 1.4038 55 (15mm) Ref. 19 20 21 22 23 24 25 22 22 26 26 Table III Derivatives of the Alcohols mp. °c Rot. l-Bromo-z-methyl-2-pr0penol 3,5-dinitrebenzoate 120 S-Methyl-z-butanol 3,5-dinitrobenzoate 163 Z-fiethyl-Z—butanol 3.5-dinitrobenzoete 116 28 z-Methyl-S-pentanol 5,5-dinitrobenzoate 85 28 3-nitrophtha1ate 150.5 22 2-Methy1-2-pentanol 3,5-dinitrobenzoate 72 28 benzoate 182-183 23 z-Methy1-3-hexanol 3,5-dinitrobenzoate 59-60 acid phthalate 59-60 20 Z-Methyl-2-hexanol 3,5-dinitrobenzoate 141-142“ phenylurethane 44-45 27 2,4-Dimethy1-5-pentanol 3,5-dinitrobenzoate 75 phenylurethane 94-94.5 12 Z-Methyl-S-heytenol 3,5-dinitrobenzoate 53-54: acid phthalate 47-48 20 Table IV Analyses or flow Derivatives for iitrogen by Seminioro KJeldahl (30) Alcohol i N Galo'd g B Found l-Bromo-z-methyl-2-propanol 3,5-dinitrobenzoate 8.04 8.11 s-Methy1-2-butanol 3,5-dinitrobenzoete 9.93 10.13 2-Methy1-5-hexanol 3,5-dinitrobenzoate 9.80 9.60 Z-Methyl-Z-hexanol 3,5-dinitrobenzoate 9.03 8.68 2,4-Dimethyl-3—pentanol 3,5-dinitrobenzoate 9.03 9.27 Z-Methyl-S-heptanol 3,5-dinitrobenzoate 8.64 8.44 Z-Methyl-Z-heptanol 3,5-dinitrobenzoate 8.64 8.40 l. 23 9. 10. ll. 12. 13. REFERENCES Henry, 1a, Compt. rend., 145, 21 (1907) Norton, F.H. and Haas, H.B., J. Am. Chem. Soc. 58, 2157 (1936) Blaise, M.,rCompt. rend., 134, 552 (1902) Grignard, v., Bull. Soc. Chim., [3] g, 944 (1903) Meieenheimer,.3., Ann., 442, 180 (1925) Huston, R.C. and Agett, A.A., J. Org. Chem., 6” 123 (1941) Huston, 3.0. and Bostwick, 0.0., J. Org. Chem., “ig, 331 (1948) . Bedos, F. an Buyer, 3., Comgt. rend., 188, 962 (1929) Bedos, 13., Cora-pt. rend., 189, 2.55 (19297“ Cottle, D.L. and Hollyday, 3.0., J. Org. Chem. ‘12, 510 (1947) Gilman, H., "Organic Chemistry, an Advanced Treatise" and. Ed., 3.509, John Eiley and Sons, Inc., New York, 1943. Gaurischi, J. and Garzino, L., Ann. Chim. Farmicoligie, g2, 110 (1887) Garzino, L., Jahrbrichte uber dioFortschritte der Chemie, 1529 (1889) Conant, 3.5. and Blatt, A.H., J. Am. Chem. 800., 51. 1255 (1929) Cheronie, N. and Entrikin, itative Organic Lnelysis", 00., New York, 1947 J.B., "Semimicro Qual- P.225, Thomas Y. Crowell .I III“! ..n..........|.....a . . .....n .L. I .. I. ...: . u . .ll .: ..u... 311......Ili: ...u: . .... . )“ (FEEEIL’Ik-IILI‘LLIIIHW'F it? L). ‘ 14. 15. 16. 17. 18. 19. 20 21. 22. 23. 24. 25 26. 27. 28. 29. 50. Shriner, R.L. and Fuson, 3.0., ”Identification of Organic Compounds" 3rd Ed., P.163, John Riley and Sons, New York, 1948 Elving, P.J. and Warsho sky, B.J., Anal. Chem., 1g, 1006 (1947) Bryant, w.:.:.n., J. Am. Chem. 800., a, 3758 (19:52:) Quelet, P. and Sicard, G., Compt. rend., 212, 162(1932) Blatt, A.H., ”Organic Synthesis" Collected Volume II, P.88, John Kiley and Sons, Inc., New York, 1943 Suter, 0.M. and Zook, H.D., J. Am. Chem.Soc., §§_, 738 (1944) Pickerd, R. H. and Kenyon, J., J. Chem. Soc., 101, 633 (1912) Timmermans, J. and Hennault-Roland, J. Chim. Phys., 33, 529 (1952) Pukirea, A., Chem. Centr., 1942 II, 389 Deschamps, A., J. Am. Chem. 5300., 3.3, 2672 (1989) Henry, L., Compt. rend., 143, 103 (1905) Whitmore F.C. and Weedburn, H.M., J. Am. Chem. 800., _5_§. 361 (1935) Huston, R.C., Guile, R.L., Sculati, J.J., and Wesson, W.N., J. Org. Chem.,‘§, 252 (1941) Jones, E.R.H. and McCombie, J.T., J. Chem. 800., 261(1943) Butter, P., Helv. Chem. Acta., 21, 1266 (1939) Franka, A. and fiozelka, 8., Chem. Centr., 1912 II,322 Clark, E.P., "Semimicro Quantitative Organic Analysis", P.37, Academic Press, Inc., New York, 1943