The Pennsylvania State College The Graduate School Department of Chemistry I . SOI*® REACTIONS OF POLYHALIDES OF NEOPENTANE I I . MISCELLANEOUS STUDIES A Thesis by John Paul W ilkins Submitted in o a rtia l fu lfillm en t of the requirements fo r the degree of Doctor of Philosophy December, 1942 Approved: r f/ z Department of Chemistry Head of Department ACKNOWLEDGMENT Sincere appreciation and gratitude is expressed to Dean Frank C. Whitmore for suggesting these problems and for his in te re st and guidance throughout the work. TABUS OF CONTENTS I. INTRODUCTION ............................................................................................................. 1 I I . HISTORICAL........................................................................................................ 2 I I I . DISCUSSION............................................................................................................. 15 IV. EXPERIMENTAL ...................................................................................................... 22 Preparation of S tarting M aterials................................................. 22 Action of Sodium on 2 ,2-ditnethyl-l,3-dibrom o- propane in Cyclohexane Solution .................................................. 31 The Action of Sodium on 2 ,2-dir.iethyl-l,3~dibromo- propane in the Absence of aS olvent . . . . . . . 35 The Action of Sodium Ethyl on 2 ,2-dimethy1-1,3- dibromopropane....................................................................................... 36 Run # 1 .................................................................................................. 36 Run ff2 .................................................................................................. 38 Run $ 3 .................................................................. 42 The Action of Sodium Ethyl on P entaerythritol Tetrabromide ............................................................................................. 47 V. BIBLIOGRAPHY......................................................................................................... 51 INTRODUCTION The failu re to produce spiropentane (bicyclopentane) by the action of sine dust on the tetrabrornide of neopentane, pentaery- th rito l tetrabrornide, has increased the in te re st in any other pos­ sible methods of preparation of th is remarkable unknown compound. CH? CHo I " 0 " I ch2^ ^ c h 2 The discovery, in th is laboratory, th at sodium propyl reacts v/ith neopentyl chloride to yield 1 ,1-dinethylcyclopropane suggested th at an attempt be made to prepare spiropentane by th is reaction using the dibromide of neopentane, 2 ,2-dim ethyl-l,3-dibrom opropane, and a sodium alkyl. This preparation also fa ile d . HISTORICAL Several in terestin g reactions of neopentyl halides discovered in th is laboratoi'y led to th is investigation. It was discovered by 1 whitmore and co-workers th at the products of the reaction of one mole of sodium and one mole of neopentyl chloride were 1 ,1-dim ethylcyclo- propane (25/0 > neopentane (36%), 2 ,2,5»5“tetraraethylhexane (137=) an^ a trace of isobutylene. A sim ilar reaction using five moles of neopentyl chloride and one mole of sodium gave 1,1-dim ethylcyclopro- pane (5170 > neopentane (4170 and only a trace of coupled product. These unusual resu lts v/ere explained on the basis of a free radical mechanism. It was postulated th at a molecule of sodium reacted with a molecule of neopentyl chloride producing sodium chloride and a neo­ pentyl free rad ical. The coupled product would resu lt from a union of tv/o such rad icals. On the other hand, if the active position of the free radical collided with a methyl group of another free rad ical, a di-free radical and neopentane would be formed. The bivalent-free radical so formed would be stab ilized as dimethylcyclopropane. If a free radical collided with a molecule of neopentyl chloride and the chlorine subsequently reacted, the products would also be dimethyl­ cyclopropane and neopentane. A sim ilar reaction was performed on the next higher homolog, . neohexyl chloride, l-chloro-2,2-dim ethylbutane (2,3). The product in this case v/ere 2 ,2-dimethylbutane (29/0 • 1,1,2-trim ethylcyclopropane (l37»)j t-butylethylene (87O and a small amount of coupled product. These products were explained in a manner sim ilar to th at described 3 above. However, a m odification is required in th is case to account for the formation of t-butylethylene and 1 ,1 ,2-trim ethylcyclopropane• Work by Smith and Taylor4- on the action of free methyl radicals on hydro­ carbons was cited to show the p o ssib ility of the following changes. ch3 CH- i CH3 - CH2 - G - CH- -» CH3 - CH - C - CH3 + H- CHo X h, ch3 CH, I CH3 - CH2 - C - CH, -+ CH3 - CH - C - CH3 I i CH0 CH, The free radical (I) could be stab ilized by the loss of a hydrogen atom to form t-b u ty lethy len e. It was pointed out that if the free pinacolyl rad ical is formed and if the coupled product is due to dim erization of two free radicals, there should have been some 2 ,2 ,3 »4,5>5“hexamethylhexane present. None of th is compound was found. The principal product from the action of sodium on l-bromo-3,3“ dimethylbutane and with the corresponding chloro derivative was 2,2,7»7“tetram ethyloctane. Only a very small amount of t-butylethylene and neohexane were formed. This is in accordance with the mechanism postulated above. The formation of cyclic compound would require the following change - 4 CH: CH: •CH, - CHo - C - CH, CH3 - CH - C - CH, * i CH, CH, This is unlikely since expulsion of a hydrogen atom could more easily stab ilize the system as t-butylethylene. 5 Whitmore and Zook found th at sodium propyl reacts v/ith neopentyl chloride to yield 1 ,1-dimethylcyclopropane (75/0 j coupled product, 2 ,2-dimethylhexane (4/°), and propane and propene. The major reaction was considered to Tie 1 ,3-dehydrohalogenation of an alkyl chloride by means of a sodium alkyl. The conception of the reaction as a dehydro- halogenation by a strong base has important th eo retical sig n ifican ce. The importance of th is discovery and in terp retatio n is due to the fact th at previously no explanation was ever given fo r the disproportion products of a Wurtz reaction other than one involving free rad icals. The free radical mechanism for the reaction between neopentyl chloride and sodium was reconsidered in lig h t of th is new evidence and explained on the basis of a sodium alkyl dehydrohalogenation of an alkyl halide. This new conception was expanded by Morton and co-workers into a A - . complete explanation for the Wurtz reaction. This work w ill be discus­ sed la te r. Since the compound sought in th is study was spiropentane, a short reviev/ of the history of th is elusive compound w ill be given. The preparation of spiropentane (II) was f ir s t attempted by Gustavson^ in 1896 by means of reduction of p en taery th ritol tetrabrornide ( ill) with zinc dust. The product was thought by Gustavson to be vinyl trimethylene (IV). A controversy occurred over the structure of th is product which lasted u n til 1916 when Philipow? proved i t to be a mixture of methylene cyclobutane s.nd methyl cyclobutene in the proportion of 2:1. A great deal of additional work was done including a stepwise reduction in an attempt to close one ring at a tim e. This attempt gave the same pro­ duct as the original reduction. Er I CH2 CH2 ch2 I ch2 I XC^I Br - CH2 - G - CH2 Br I CH - CH = CH2 ch2 vch2 I CH2 ch2 I Br I I . II I. IV. The reduction was carefully repeated by P. H. Williams® of th is laboratory v/ho found methylene cyclobutane and 2-raethyl-l-butene but no methyl cyclobutene as reported by Philipow?. Philipow reported isopentane from the reduction, without other proof than boiling point, which evidently was 2-m ethyl-l-butane. There was no trace of products other than the two compounds, methylene cyclobutane and 2-m ethyl-l- butene. A complete review of the lite ra tu re pertinent to the reduction of pen tacry th rito l tetrabrornide is contained in the thesis of P. H. Williams®. Gustavson and Popper9 reduced 2,2-dimethyl-l,3-dibromopropane with zinc dust to obtain 1 ,1-dimethylcyclopropane. Sodium and zinc are used 6 in such reactions to form alicy clic rings. The cyclopropane derivatives are the most suitably prepared by th is method. A complete review of the lite ra tu re pertaining to the Wurtz reac­ tion has been made by A. H. Popkin10 and a review of the lite ra tu re concerning sodium alkyls has been made by H. D. Zook11 . Therefore, only the lite ra tu re published since these reviews and th at necessary in discussion of resu lts w ill be included. Y/urtz f ir s t performed the reaction involving an alkyl halide and sodium in 185512* Since then the reaction has been studied extensively because i t is useful in preparing hydrocarbons and has been the center of a controversy concerning its mechanism. This controversy arises from the fact th at the coupling of two organic residue in the Y/urtz reaction may be in terp reted in two d istin c tly differen t ways. One assumes the formation of free rad icals, 2 RX + 2 Na ---------» 2 R» + 2 NaX 2 R. --------------> R - R and the other proposes the interm ediate formation of an a lk a li alkyl. RX + 2 Na ----------* R Na + NaX RNa + RX ------------> R - R + NaX In direct evidence th at sodium alkyls are an interm ediate has been obtained by many in v estig ato rs. The f i r s t to make a c ritic a l in v esti­ gation of the mechanism of the 'Wurtz reaction was Schorigin. From studying the products of the action of a mixture of sodium and alkyl halide, on acid chlorides and esters he showed th at the same type of 7 13 reaction occurred as with the organo magnesium compounds . He postulat­ ed the active agent to be a sodium alkyl. To further te s t th is hypo- 14 thesis - he prepared diphenyl ethyl carbinol from benzophenone, ethyl iodide and sodium; diphenyl isobutyl ethylene from benzophenone, iso ­ butyl bromide and sodium; phenyl isoamyl methyl carbinol from aceto- phenone, isoamyl iodide and sodium. Since the sodium alkyls contained a postulated carbon to metal bond, Schorigin reasoned th at treatment with carbon dioxide should yield an acid. From a mixture of isobutyl bromide and sodium followed by treatm ent with carbon dioxide, he obtain­ ed a good yield of isovaleric acid. If the reaction was run in benzene, a small amount of benzoic acid was formed. These reactions v/ere interpreted as indicated in the following reactions. C^HgBr + 2 Na » C^H^Na + NaBr C^HgNa- + C02 * C^H^COONa and C*H9Na + C6H6 C6H5Na + C02 * C6H5COONa i To further te s t his hypothesis, sodium ethyl was prepared from 1 ? diethyl mercui'y and sodium J . This product gave resu lts sim ilar to those above, i.e ., added to carbonyls, formed acids on treatment with 17 carbon dioxide and exchanged hydrogen with sodium . The behavior of th is known sample of sodium ethyl confirmed his b elief th at the active compound in a mixture of alkyl halide and sodium, was the sodium alkyl. Further proof1^ was given by the following series of reactions. Using benzene as a solvent, carbonation of a mixture of isobutyl bromide