I. The Isomeric 2,3-Epoxypentanes and 2-Pentanes. The Extent to Which Mixtures of Diastereomers are Formed in Reactions of Some Pentane Compounds. II. The Preparation of Cyclopropane. III. The Thermal Decomposition of Trans-1,2-Cyclobutane-Bis(Trimethylammonium) Hydroxide Citation Schlatter, Maurice Jay (1941) I. The Isomeric 2,3-Epoxypentanes and 2-Pentanes. The Extent to Which Mixtures of Diastereomers are Formed in Reactions of Some Pentane Compounds. II. The Preparation of Cyclopropane. III. The Thermal Decomposition of Trans-1,2-Cyclobutane-Bis(Trimethylammonium) Hydroxide. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/9txa-0p49. https://resolver.caltech.edu/CaltechTHESIS:02282025-211013383 Abstract No abstract. Item Type: Thesis (Dissertation (Ph.D.)) Subject Keywords: (Chemistry) Degree Grantor: California Institute of Technology Division: Chemistry and Chemical Engineering Major Option: Chemistry Thesis Availability: Public (worldwide access) Research Advisor(s): Buchman, Edwin Raphael Thesis Committee: Unknown, Unknown Defense Date: 1 January 1941 Record Number: CaltechTHESIS:02282025-211013383 Persistent URL: https://resolver.caltech.edu/CaltechTHESIS:02282025-211013383 DOI: 10.7907/9txa-0p49 Default Usage Policy: No commercial reproduction, distribution, display or performance rights in this work are provided. ID Code: 17036 Collection: CaltechTHESIS Deposited By: Benjamin Perez Deposited On: 03 Mar 2025 22:41 Last Modified: 03 Mar 2025 22:48 Thesis Files PDF - Final Version See Usage Policy. 18MB Repository Staff Only: item control page

I. The Isomeric 2,3-Epo,cy:pentanes and 2-Pentanes. The Extent to which Mixtures of Diastereomers are Formed in Rea ctions of Some Pentane Compounds. II. III. The Preparation of Cyclopropene. The Thermal Decomposition of Trans-1,2-cyclobutane. bis-(trimethylammonium) Hydroxide. Thesis by Maurice Jay Schlatter In Partial Fulfilment of the Requirements for the Degree of Doctor of Philosophy California Institute of Technology Pasadena, California Table of Contents Page I. The Isomeric 2,3-Epoxypentanes and 2-Pente.nes. The Extent to ilnich Mixtures of Diastereomers are Formed in Reactions II. of Some Pentane Compounds - - 1 Experimental - - - - - - -- - - 4 Summary - - - - - - - - - - - 7 The Preparation of Cyclopropene - g · E:xperimental - - Acknowledgment - - - - 11 - - - - - - - - - - - - - -- Summa.ry - - - Bibliography III. 22 23 - - - - - - - - - - - 24 The Thermal Decomposition of Trans-1,2-cyclobutane bis-(trimethylammonium) Hydroxide - 25 Experimental 27 Acknowledgment 37 Summary - - - 38 Bibliography 39 Propositions - - - - - - - - - - - - - - - - - - - - - - - - - - 4o The Isomeric 2,3-lilpoxypentanes and 2-Pentenes. The Extent to which Mixtures of Diastereomers are Formed in Reactions of Some Pentane Compounds. The Preparation of Cyclopropene 8 Cyclopropene was synthesized in 1922 by Demjanow and Dojarenkola and our entire knowledge regarding this interesting compound is due to the researches 1 carried out by these Russian scientists. In order to make it available in amounts sufficient for an extended investigation of its physical and chemical properties, the conditions atteniing its formation were studiea. in some detail. The steps employed in the synthesis are outlined in the accompanying chart. The yields indicated are conservative, reproducible values obtained on a preparative scale as the result of four to thirty runs at each step. The overall yield based on these figures and choosing the :path (bold.-facea. arrows) found to be most convenient is 3.3%. Taking into account the quaternary iodide recovered from the dimethylcyclopropylamine formed in the last step, this value becomes 4.7%. From a detailed study of the decomposition of the quaternary base on platinized asbestos it was shown that the optimum temperature for cyclopropene formation is 320-330°. Dimethylcyclopropylamine and methylacetylene which had previously been reported la were also obtained. A more detailed discussion of the thermal decomposition is given in the experimental part. Pure cyclopropene was obtained by distilling the hydrocarbons It distilled at -36° to -35° at 744 mm. 1 In confirmation of Demjanow a.nd Dojarenko through a Podbielnia.k column. and analyzed for C3 H.4,. it was found that the olefine reacts energetically with bromine and polymerizes readily. It could not be conveniently stored, even at -78°. 9 HOCH2 CH 2 CH2 0H 55% t HC1 2 ClCH 2 CH 2 CH2 0H 95% l l Pllra 3 BrCH 2 CR2 CH 2 Br ! ClCH2 CH 2 CH 2 Br 46% 68% BrCH 2 CH 2 CH2 CN NaCN a 60% ClCH2CH2 CH 2 CN + BrCH 2 CH 2 CH2 CN 60% NaCN b NaNH 2 C t NaNH c 2 6cN _ _ _ _ _ _ _ _ __, I HCl + C2H50H 7 Q'5 <f 0 t- 95% l H20 8 LlcooH d I PCl3 9b ' io t •/\' I\ c:oc2Hs ~ ., 93% \: NH.HCl ~COCl / \I Heat 7 ~CONH2 91% I 9 NHa 25% - 68i NaOCH 3 + Br2 10 NHCOOCH 3 I 6NH 76% 80,1 H2 0lO 2 1 CH I la ~ 3 N(CH 3 )aI ~ I AgOHla • - /\ 320-330° a 98% I 6 N(CH3) CHalla 2 + g The method of Allen4 gives ~-chlorobutyronitrile in the yield indicated. A modification giving a similar yield of a mixture of 10 r-chloro and Y-bromobutyronitrile (which is equally useful in this synthesis) was employed to facilitate the handling of large batches. b The method used ~~s a modification of the procelure of Gabriel. 6 C The method used was essentially a large scale modification of the method of Cloke et al.3 This procedure was equally satisfactory with the pure r'-halogen nitriles or mixtures of them. d The yields of this acid by decarboxylation of cyclopropane-1, 1dicarboxylic acid are unsatisfactory. The reaction is being investigated. e f This step has not been previously described in the liter-~ture. The conversion of cyclobutanecar·boxylic acia. directly to the corresponding amine has been reportet recently by Heisig. 11 This yield is based on dibrooocyclopropane isolated. 11 The la statement that cyclopropene can be regenerated from dibromo- cyclopropane by treatment with zinc and alcohol was not confirmed. This procedure gave a product consisting to a large extent of cyclopropane. The problem of the reconversion of the stable dibromide to cyciopropene is of importance in connection with the handling of large amounts of the latter and is the subject of further studies. Experimental * Trimethylene Chlorohydrin. was prepared by the methods of Marvel and Calvery 2 and of Hultman, Davis and Clarke 12 . The treatment of trimethylene glycol with the theoretical amount of phosphorus trichloride gave the less satisfactory yield of 40%. An attempt to use thionyl chloride, either alone or in conjunction with pyridine, gave, in ad.di tion to chlorohydrin, other products not easily separated by fractional distillatfon. These interfered with subsequent steps. Trimethylene Chlorobromide was prepared. in twenty mole batches by the method of Cloke et a1 3 and was used without purifying by distillation. r"-Chloro and r"-J3romobutyronitrile from Trimethylene Chlorobromide.--To 765 g. (15.6 moles) of technical sodium cyanide eggs in 1350 ml. of warm water in a 12-liter round bottom flask surrounded by a water bath and equipped with a large capacity reflux condenser, was added 3500 ml. of 96% alcohol and 2360 g. (15 moles) of trimethylene chlorobromide. The mixture was warmed and the ensuing vigorous reaction controlled by addition of ice to the bath and by playing a stream of Melting points reported in this paper are corrected. 12 water on the upper pa.rt of the flask. After the reaction had subsided, refluxing was continued for an additional hour and then j-1/2 liters of alcohol rapidly distilled off. The oil was decanted from the aqueous and solid phases in the flask, these latter extracted thoroughly with chloroform and the extract combined with the oil. The resulting solution was washed with half saturated calcium chloride and with water, the solvent removed and the residue fractionated at reduced pressure; 1193 g. of product (b.p. 89-115°/26 mm.) was obtained. Analysis of the mixture by fractionation showed it to consist of 67.5% of ~-chlorobutyronitrile (b.p. ~-bromobutyronitrile (b.p. 108°/26 mm.). 94°/26 mm.) and 32.5% of This corresponds to a combined yield of 69%. ~-Bromobutyronitrile from Trimethylene Dibromide.•--To 808 g. (4 moles) of trimethylene dibromide in 3200 ml. of 96% alcohol was added 160 g. (3.14 moles) of sodium cyanide in 300 ml. of water. The mixture was shaken thoroughly for five minutes, permitted to react at room temperature for one hour and then refluxed for 45 minutes. After cooling, the mixture was poured into 9 liters of water, and extracted repeatedly with chloroform. Fractionation at reduced pressure gave 268 g. of recovered trimethylene dibromide and 188 g. of ~-bromobutyronitrile (b.p. 104-107°/23 mm.)(47% on the basis of trimethylene dibromide consumed) • • The following procedure was worked out in this laborator:r by Messrs. Raymond Clinton and Thurston Skei. 13 Cyclopropyl Cyanide from the ~-Halogenbutyronitrile Mixture.-- One liter of liquid ammonia and 0.5 g. of hydrated ferric nitrate were placed in a two-liter three-neck flask equipped with mechanical stirrer and dry ice reflux condenser protected by a soda lime tower; 92 g. (4 moles) of sodium shavings was added over a period of three quarters of an hour. The mixture was stirred until the blue color had disappeared (one or two hours). In a similarly equipped three-liter three-neck flask (reaction flask) were placed 1-1/2 liters of liquid ammonia and 4.25 moles of r'-chloro- or f-halogenbutyronitrile mixture (pure I"" -bromobutyroni trile can be substituted for the mixture). The flasks were connected by a 10 mm. glass tube reaching to the bottom of the sodamide flask and extending 1 cm. through a stopper in one neck of the other. By means of positive air pressure, controlled by a finger over an air by-pass, the sodamide suspension was forced over into the reaction flask at such a rate that about one hour was required for the addition; constant stirring ~~s maintained in both flasks. The sodamide flask was rinsed out with 300 ml. of liquid ammonia, and the washing added to the reaction mixture. Stirring was continued for four hours*, during the first two hours of which the mixture refluxed gently. During the second two hours no more dry ice was added to the traps and slow evaporation of ammonia took place. . The sodamide aa_di tion tube was replaced with a dropping funnel and one liter of anhydrous ether slowly added to the solution. The reaction mixture With pure r-chlorobutyronitrile and l""-bromobutyronitrile, the time of stirring should be two and six hours respectively (time of refluxing zero and 4 hours). 14 was rapidly filtered through a sintered glass funnel and the filter cake washed with anhydrous ether. The c>.mmonia and ether were removed on a water bath•, the residue filtered** and then fractionated at 80 mm.; 14-9 g. of cyclopropyl cyanide (o.p. 69-70°/80 mm.) (bl% yield on the oasis of r'-halogenoutyroni trile actually used) and 0.6 mole of ~-halogenbutyronitrile were obtained. A rosin-like residue was left in tne distilling flask. Cyclopropanecarooxylic 4,cid.--To 336 g. (6 moles) of potassium hyct.roxide in 1200 ml. of water, was added 268 g. (4 moles) of cyclopropyl cyanide and the mixture refluxed. After initiation of the reaction it was found necessa.r;r to cool until the vigorous refluxing subsided. The solution was heated for six hours, cooled in an ice bath and 1500 ml. (4. 5 moles) of 6 ! sulfuric acid slowly addect. Continuous extra.ct ion with ether and fractionation of the ether extract yieldecl 332 g. (96%) of cyclopropanecarboxylic acid (b.p. 80-81°/13 mm.). Cyclopropanecarboxamide from the Acid Chloride.--The acid chlorice*** was prepared by the method described by Lipp et al lO The amide was prepared by saturating a solution of 171 g. (1.6~- moles) of the acid In one experiment using ~ure bromonitrile, crystals of organic material separated out at this point. These crystals were filtered off and recrystallized from absolute alcohol, m.p. 134°0. Anal.Ca.led. for C4 H N2 Br: C, 29.11; H, 5.50; N, 16.98. 9 Found: C, 29.39; E, 5-79; N, 16.23. This compound d.issolved readily in water and ga.ve an immecUate test for bromide ion. The substance is probably cyclopropylcarbamid.ine hydro bromide, /:::::. C(lUI2 )+Br-. , *"' With pure r"-chlorobutyroni trile, no solid separates at this point. *** Commercial thionyl chloride gave none of the desired product; cf. Skraup and Binder.13 15 chlori t e in 1500 ml. of anhyd.rous ether with ammonia at 0° unc:er anhydrous cond_itions. Solvent \·Jas removed in vacuo a.nd the resicLua.1 fine white powder extracted with chloroform in a Soxhlet extractor. Large colorless prisms separated out; yield 126.5 g. Cyclopropanecarboxamide from Cyclopro:pyl Cyanide.--Ethyl imidoc;rclopropane carboxylate hydrochloride was prepared from 67 g. (1 mole) of cyclopropyl cyanide by the method7 of Cloke and coworkers. Without isolation of the imido ester salt, volatile materials were removed at reducecl pressure, ancl the soli d residue hea,ted in the same vessel in an oil bath at 130-150°; a stream of ethyl chloride was smoothly given off. To complete the decomposition, the temperature was raised. to 165° and the pressure reduced to 20 mm. cyclopropane carboxylate was trapped out. A small a.mount of ethyl The crystalline residue on recrystallization from ethanol gave 75 g. (87%) of cyclopropanecarboxamide. Cyclopropylamine from Cyclopropanecarboxamide.--The urethane was prepared in 68% yield accord.ing to Lipp et al 10 It is important that the material be completely worked up within a short time after the methanol has been di stilled off.• From this, cyclopropylamine ·was Distillation of the residue at 2 mm. yielded two solids boiling at 120-125°/2 mm. and 155-160°/2 mm. The low boiling material proved to be unchanged cyclopropanecarboxamide (analysis, mixed m.p.). Recrystallization from alcohol of the higher boiling material gave a compound melting at 100°. Anal. Calca_. for 04 H6 CN: C, 57.12; H, 7.19; N, 16.66. Found: C, 57.16; H, 7.17; N, 16.63. The compound. is probably N-cyclopropyl-N-carboxycyclopropyl urea ~NHCONHCO~ ; cf. Hofmann, Ber., 1:.::, 2725 (1881); .t.2• 754 (1882). 16 obtainedlO in 76% yield. Cyclopropylamine from Cyclopropanecarboxylic Acid.--In a one-liter three-neck flask equipped with dropping funnel, mechanical stirrer, gas evolution indicator, and thermometer <:Lipping into the liq_uid was placed 130 ml. of concentrated sulfuric acid. First 62 g. (0.72 mole) of cyclopropanecarboxylic acid was slowly added, keeping the temperature below 4o 0 ; then a solution of hydrazoic acia. 14 in chloroform (from 75 g. of sodium azide and found. by titration to contain 34.4 g. (0.80 mole) of hydrazoic acid) was a,c.ded over a period of two hours, maintaining the temperature between 35-4o 0 • Stirring and heating were continued. another five hours I a.t the end of which time gas evolution had almost stopped. The reaction mixture was poured on 200 g. of ice and the resulting mixture steam distilled to remove chloroform and excess hydrazoic acid. Then 250 g. of sodium hydroxide dissolved in a small amo1.mt of water was ac!.ded slowly while the liberated ba se distilled. over into 250 ml. of 3 ! C.P. hydrochloric acid. Evaporation of the hydrochloric acid solution gave 21.6 g. of solid containing ammonium chloride in addition to cyclopropylamine hydrochloride. This solid vias dissolved in the minimum amount of water and add.eel through a dropping funnel to 30 g. of potassium hyclroxide and 5 ml. of water in e. distilling fla.sk. The amine which distilled was dried over barium oxide and. fraetionat~d; 10.3 g. (25% of theory) of cyclopropylamine (b.p. 49-50°/750 mm.) was obtained.• were preparea.:' The following derivatives phenyl thiourea 9 , colorless plates from aaueous ethanol, * A similar experiment with 20 g. of isobutyric acicl yielded 80% of lfhe theoretical amount of isopropyla,mine. 17 m.p. 125.5°; benzamiae 9 , white needles from aq_ueous methanol, m.p. 99°; picrate, orange needles from ethanol and petroleum ether, m.p. 149 0. Ide!;.tical derivatives were p repared from the amine obtained from the urethane. T'rimethylcyclopropylar:unonium Iodide from Cyclouropylamine.-- To 68 g. (0.73 moles) of cyclopropylamine hydrochloride in 150 ml. of methanol and. 4oo g. ( 2.82 moles) of methyl iodide was aa_ded 160 g. of pote.ssium hydroxide in 800 ml. of methanol with stirring over a period. of three hours, kee:ping the temperature at 30-40° C. The mixture was refluxed for an additional two hours; evaporation to dryness in vacuo at 4o° C. gave 551 g. of soli d . This yielded. 165 g. (83%) of trimethylcyclopropylammonium iodide on extraction with chloroform for ·ninety-six hours in a SoY.hlet extractor. Trir.1ethylcyclopropylammonium Iodide from Dimethylcyclopropylamine.-To l+o g . (0~t47 mole) of a.imethylcJrclopropylamine ( see below) in 4oo ml. of absolut e alcohol was added 100 g. (0.704 mole) of freshly distilled methyl iodicl.e with stirring over a period of fifteen minutes. mixture was allowed to stand overnight. separ~ted out; yield 104 g. C, A large crop of white crystals (96%) m.p. 274°. Anal. Calcd. for C6 H14NI: Found: The C, 31.71; H, 6.23; N, 6.17; I, 55.90. 31.79; H, 6.24; N, 6.20; I, 55.99. Thermal Decomposition of the ~ternary Base.--Trimethylcyclopropyla.mmonium hydroxide w·as prepare a from 22. 7 g. (0 .1 mole) of 1 q_u.aternary iod.ide by Demjanow' s a procec.ure. The solution was con- centrated in vacuo at 40°, taking care to avoid carbonate formation by 18 the use of nitrogen-filled ap:'.)aratus. The concentrated base (approx. 20 ml.) was bro't-mish and. cloudy, and was used in this form. The pyrolysis tube was made by blowing a test tube bottom on a piece of j0 mm. pyrex tubing 12 cm. in length with neck constricted to hold a small hro-hole rubber stopper. A condenser ma.de from B mm. tubing was sealed to the tube 8 cm. from the bottom. The rubber sto,per carried a gas inlet tube (for S\\-eeping the system with carbon dioxide) and a short stem dropping funnel which projected 1/2 cm. through the stopper.• A hydrostatic head was lll8.intained on the liq_uid to be added_ by connecting the sidearm of a filte.r flask to the top of the dropping funnel. The filter fle.sk was fit tec1 with a rubber stopper carrying a funnel with a 30 cm. stem; this reached the bottom of the flask and was filled with water. The bulb of the pyrolysis tube was linecl with a layer of 20% platinized asbestos, 3 mm. thick. The condenser attached to the pyrolysis tube was connected to a 100 ml. receiver which was followed by a 100 ml. spire.l gas wash bottle containing 3 _!i! hydrochloric acid (to absorb ttimethylamine and any dimethylcyclopropyla.mine which did not condense in the receiver). This, in turn, was connected to a gasometer containing sa.turated sodium chloride. n After sweeping the It is important that this dropping tube does not project into the hot zone of the pyrolysis tube. ** The gasometer was made from t,..,o 2-1/2 liter bottles. The rubber stopper of the first carried a gas inlet tube and differential pressure manometer flush with the stopper and a 12 mm. tube reaching to the bottom of the bottle. This wa.s connected by rubber tubing to a similar tube reaching to the bottom of the seconct bottle. The gasometer was calibrated. Gas could be drawn into or displaced from the first bottle 'by lowering: or raising the secona. bottle. 19 air from the apparatus with carbon dioxicle, the pyrolysis tube was heated to the desired temperature by a metal bath ancl the concentrated solution of base added over a period of 10-12 min. This yielded 1600-1300 ml. of gas at the optimum temperature (320-330°). The apparatus was flushed out with 800 ml. of ca.rbon dioxide which was combined with gas in the gasometer. Bromination (see below) of gas obtained under optimUJn conditions gave 8-9. 5 g. of 1, 2-a_i bromocyclopropane (material boiling below 50° /25 mm.) and 1. 5-2 g. of higher boiling resia_ue. The cont ents of the hyc!.rochloric acid wash bottle were poured into the receiver and the combined solutions evaporated to dryness in ~acuo. 'fhe mixed_ amine hycrochlorides were trer. ted with alkali a.nd steam cUstilled. Drying of the amines over potassium hyclroxide and subseq_uent fractionation gave dimethylcyclopropylamine, b.p. 60 .1 ° /743 mn.; nD20 1.3999; a. 25 4 0. 7607; picrate, yellow needles (from ..ethanol), m.p. 196. 5° a_. A series of more than twenty clecomposi tions, ee.ch of one-tenth mole of quaternary ioa_ide, was carried out at temper8.ture intervals within the range 250° to 450°. It was founa that of temoeratures below 320° the ratio of hydrocarbons to dimethylcyclopropylamine formed decreased rapidly while there was very little change in the ratio of cyclopropene to meth;rlacetylene. At 250° the total hydrocarbon formed was less than one-third that obtained at the optimum temperature. As the temperature of tne decomposition was increased above 330°, the total volume of hydrocarbon obta.ined decreased. somewhat and 20 the ratio of cyclopropene to methylacetylene obtained fell off rapiaiy. At the optimum temperature (320° to 330°) this ratio was bout 10 : 1, while at 450° it dropped to 1 : 1.5. The condition of the catalyst improved with use for the first three or four times, but after t!,is was somewhat erratic in its behavior. From t11is it is seen that the ratio of hyd.rocarbons to amine a:ppea,rs to be affected by the condition of the catalyst, though the ratio of cyclopropene to methylacetylene depena.s primarily on the temp~rature. A variation in the ratio of hydrocarbon to amine is not seriously disadvantageous as the amine may be quantitatively reconverted la to the quaternary salt. In some experiments, products isolated from a.ecom:position of the quaternary base account for 85 to 90% of the quaternary iodide. Examination of Hydrocarbon Pyrolysis Products.--The presence of an acetylenic hyci.rocarbon was confirmed by testing samples of gas with cuprous chloride dissolved in ethanolamine; a yellow precipitate was obtained. The relative amounts of cyclopropene and methylacetylene in the mixture were determined by passing further samples from the gasometer (corresponc!.ing to 0.1 mole quaternary iodide) into excess bromine at 0° over a period of one and one-half hours. The bromine and bromide were then treated with sodium bisulfite to remove excess oromine, washed with water, dried with calcium chloriae and. distilled.. The portion boiling to 50° /25 mm. was collected and founa. to be nearly pure dibromocyclopropane; the weight of the residue was taken as an index of the amount of methylacetylene present. 21 One hunclred. a.nd fifty g. of the combined bromine,tion product from a series of pyrolyses at various temperatures, on careful fractiona.tion, yielded 117 g. of 1, 2-c i bromocyclopropane (b.p. 57-58~ / 50 mm.; m.~. - 1 to +1 9 ; n 20 n 1.5360; d~5 2.0838), 11.9 g. of 1,1,2,2-tetra.bromopropane (b.p. 110-112° /10 mm., n 20 D 1.6162; d~ 5 2.668), 6.1 g. of 1,1,3,3- tetrabromopropane (b.O. l22-123°/l0 1:1u .; n 20 D 1.6198; d~5 2.683), end small intermediate fractions. Isolat ion of Cyclopropene. - -Three liters of hyd.rocarbon :pyrolysis products, dried with magnesium perchlorate anci freed of carbon dioxide with a scarite W<',s fractionated at atmospheric :pressure (fl-4 mm.) in a sr:1all Podbielniak column.* Appro :x:i ma tely 700 ml. of gas distilled over at - 36 to -35°; the resi due, 3 to 4 ml. of slightly yellow somewhat viscous oil gave no further distillate on warming to room temperature. As the original sa.nple cont 2. ined methylacet:rlene as well as cyclopropene, this seems to indicate that cyclopropene can comb i:1e with meth;irlacetylene as well as with itself. The ratio of hydrogen to carbon ·wa s determined on three 10 ml. samples of pure hydrocarbon gas by burnin€: in a semi--rnicro combustion furnace. Anal. 0[4,lcd. for C3 H4 : H/C rat:l.o, .1.33. Found: 1.332, 1.3l.t6, 1.320. The author 1-, ishe s to a cknowledge the assistance of Dr. B. H. Sage anct Mr. Robert Dourson, of the Ar.1ericDn Petroleum I nstitute Laborato ries at this I nst itute , in carr~ring out t h is distillation. 22 Acknowledgment The author wishes to acknowledge his ina_ebtedness to Dr. E. R. Buchman for suggesting the problem and to him and Professor H. J. Lucas for guidance during the course of the investigation. Thanks are also due to Professor J.B. Cloke for suggestions relating to the preparation of cyclopropyl cyanide and to Mr. Alf O. Reims and to Mr. Herbert Sargent for aid with some of the preparations. 23 Summary The preparotion of cyclopropene by the thermal decomposition of trimethylcyclopropylammohium hydroxide on a platinum catalyst was studied in aetail. ~ne optimum temperature for cyclopropene formation wes fotmd to be 330°. Pure cyclopropene was obtained by fractional distillation of the hydrocarbons obtained by pyrolysis of the quaternary base. It boils at -36° at 744 mm. Cyclopropylamine was prepared from trimethJrlene glycol by several paths. The ina.ividual steps were investigated from a preparative point of view. 24 Bibliography 1. a. Dem~anow and Dojarenko, Ber • .2..§, 2200 (1923); Bull. acad. sci. Russ., L6) ~97 (1922); b. Bull. acad. sci. U.S. S.R., [7Tb53 (1929); [ C.A., ~. 1348 (1930)] . 2. Marvel and Calvery in ttOrganic Syntheses", Coll. Vol. I, John Wiley and Sons, Inc., Mew York, New York, ~ . p. 519. 3. Cloke , Anderson, Lachmann and Smith, J. Am. Chem. Soc., .22• 2794 (1931). 4. Allen, "Organic Syntheses 11 , Coll. Vol. I, John Wiley and Sons, Inc., New York, N.Y., ~ . P• 150. 5. E'.amm and Marvel in "Organic Syntheses 11 , Coll. Vol. I, John Wiley and_ Sons, Inc., New York, N. Y., ~ . p. 28. 6. Gabriel, Ber., 22, 3336 (1889); see Derick and Hess, J . .Am. Chem. Soc., 4o, ....... 546 (1918).--- 7. Cloke, Knowles and Anderson, J. Am. Chem. Soc.,~. 2547 (1936). 8. Henry, Bull. classe sci. Acad. roy. llelg., [3] Jl, 17 (1899) [ Chem. Zentr ♦- )2, I, 975 {1899) J • 9. Kishner, a.) J. Russ. hys. chem. Soc., .,:U, 377 (1901); b.) ibid. 2.,, 304 (1905 Chem. Zentr., ]_g, II, 579 (1901); I, 17(5If"(1905) 10. Lipp, Buchkremer and Seeles, Ann.,~. 13 (1932). 11. Reisig, J. Phys. Chem.,~. 1210 (1939). 12. Hultman, Davis and Clarke, J. Am. Chem. Soc.,~. 369 (1921). 13. Skraup and Binder, Ber., 62, 1130 (1929). 14. von l3raun, Ann., ~_20, 125 (1931). J.&, - - J. The Thermal Decomposition of Trans-1,2-cyclobutaneb is- ( t rime thylammoni um) Hyd.roxi ci.e 25 The possibility of synthesizing cyclobutadiene is being investigated on a broad basis in this laboratory under the direction of Dr. E. R. Buchman. One line of attack on this problem is a study of the thermal decomposition of the four possible cyclobutane di-quaternary ammonium bases. The results obtained from the pyrolysis of the most easily obtained member of the series, trans-1,2-cyclobutane-bis-(trimethylammonium) hydroxide, are reported in this paper. O[lo_xr_ OH H Chart: Reactions which might occur during the thermal decomposition of trans-1,2-cyclobutane-bis-(trimethylammonium) hydroxide without breaking the four-ring. 26 The first step of the thermal decomposition would be - by analogy with the pyrolysis of trimethylcyclopropylammonium hydroxide - one or more of the reactions, I, II or III, shown on the chart. assuming retention of the four-ring structure. This is Hurd and Drake 1 found that the pyrolysis of 1,2-butane-bis-(trimethylammonium) hydroxide gave 44% of ethylacetylene and 56% of methylallene. Pyrolysis of 2,3- butane- bis-(trimethylammonium) hydroxide gave approximately 45% of 1,3-butadiene and 55% of methylallene or a mixture of methylallene and dimethylacetylene. On the basis of these data, both reactions II and III might be expected. In the case of trans-1,2-cyclobutane-bis-(trimethylammonium) hydroxide, reaction III, if it takes place at all, occurs to a ve-r-J small extent. This conclusion is based on the probability that l-dimethyl- aminocyclobutene-2 would be stable by analogy with other ~- r-unsaturated tertiary amines 2 and therefore would have been isolated in the scheme given in the experimental part if it had been present in any considerable amount. It is not possible to say definitely that re~ctions I and II both take place, as the ~.~-unsaturated dimethylamine might be formed by either path. That reaction I occurs, however, is shown definitely by the isolation and characterization of trans-1,2-bis-(dimethylamine) cyclobutane as tne picrate and methiodide. Although the fact that 1-dimethylaminocyclobutene-l (V, VI) is formed as an intermediate was not sho,..m by its i ·s olation or by the isolation of ad_ditive derivatives formed from it, it is clear that this must be the case from the 27 isolation of considerable amounts of cyclobutanone (IX) (characterized as the semicarbazone, phenylhydrazone and 2,4-dinitrophenylhydrazone), the condensation products X and XI (characterized as the 2,4-dinitrophenylhydrazones) and dimethylamine formed in reaction IX (characterized as the benzenesulfonamide). The tendency of vinyl tertiary a.mines to give the corresponding aldehyde or ketone and corresponding secondary 2 amine, in the presence of water, has been reported by Sou. The polymerization product (X) resulted from the well known aldol condensation, not previously reported for cyclobutanone, and the compound (XI) by loss of one molecule of water from X. Experimental Thermal Decomposition of the Di-quaternary Base.-- Trans- 1,2cyclobutane-bis-(trimethylammonium) hydroxide was prepared from 42.6 g. (O.l mole) of the di-quaternary iodi de by shaking it in aqueous solution with an excess of freshly prepared silver oxide• until the colloidal particles in the supernatant liquid had coagulated. The mixture was then filtered rapid.ly with suction and the filter cake washed with distilled water. The filtrate and washings were concentrated in vacuo at 4o 0 , ta.king care to avoid carbonate formation by the use of nitrogen filled apparatus. The evaporation was stopped when the volume had been reduced to about 40 ml. and the gray to brownish, cloudy solution used for pyrolysis in this form • .-The silver oxide was prepared from 68 g. (0.4 mole) of silver nitrate and 26 g. (o.465 mole) of potassium hydroxide. It was washed with ten changes of distilled water by decantation in order to remove free alkali. 28 ·1·he apparatus for the thermal decomposition was the same as that described for the pyrolysis of trimethylcyclopropylammonium hydroxide.• The spiral gas wash bottle contained 100 ml. of 3 _! hydrochloric acid. The air was displaced from tne apparatus with carbon dioxide and the concentrated solution of base pyrolyzed by dropping it on platinized asbestos at 350-360° over a period of 15 minutes. No gas collected in the gasometer and very little went into the hyd.rochloric acid wash bottle.** The pyrolysis distillate was made acid with a small excess (25 ml.) of 6 N hydrochloric acid and the organic phase collected in ether. The aqueous phase was shaken with several portions of etner; the extracts combined, dried with sodium sulfate and most of the etner removed through an efficient total reflux column. The residue will be referred to as the "neutral fraction". The above aqueous solution was made basic by the addition of 30 ml. of 6 N sodium hydroxide solution and continuously extracted with ether for four hours (further extraction gave a negligible amount of basic material). Most of the ether was removed from the extract through an efficient total reflu..~ column. the 11 higher amine 11 fraction. The residue will be referred to as The volatile amines were lost in this case but were investigated in another experiment. This thesis, 11 The Preparation of Cyclopropene 11 , pg.lg In one pyrolysis carried out at 410-420°, about 300 ml. of gas was collected in the gasometer. Fifty ml. of this condensed in a dry ice trap to give a mobile liquid with pleasant, unsaturated odor. This analyzed for an H/C molal ratio of 1.94 when burned in a semimicro combustion apparatus. 29 Investigation of the Neutral Fraction.--The concentrate from the ether extraction of the acidic solution was transferred to a small distilling flask with a capillary and surrounded by a water bath. The flask was connected to a Dry Ice trap and everything which would distill below a bath temperature of 100° and pressure of 2 mm. was taken over into the trap. The material from the trap was distilled at atmospheric pressure through a small Young-type column made from a 10 cm. section of 7 mm. glass tubing. This gave 1.84 g. of material boiling from 80-100° and a small amount (0.23 g.) of an intermediate fraction boiling from 100°/760 rnrn.-57°/18 mm. The residue was combined with the residue from the first distillation giving 2.15 g. of higher boiling neutral fraction. The material boiling from 80-100° (I) had an odor similar to that of cyclopentanone (on standing it acquired a different, very penetrating odor). Two drops of (I) in one ml. of 001 4 rapidly decolorized 20 drops of 3% bromine in carbon tetrachloride with evolution of hydrogen bromide (fumes and acid reaction to moist litmus), further decolorization occurred very slowly. (In a parallel experiment, 2 drops of allyl alcohol decolorized 120 drops of 3% bromine). characteristic of the carbonyl group. (I) gave derivatives The .• semicarbazone of (I) was prepared by adding just enough alcohol to 0.3 ml. of water and 0.10 ml. of (I), in a centrifuge tube, to give a solution and then stirring in 0.10 g. of semicarbazide hydrochloride and 0.12 g. of hydrated sodium acetate. In a few minutes ,mite crystals began to separate and in 30 minutes the mixture was centrifuged. After two recrystallizations from water the semicarbazone was obtained as rosettes of white needles, 30 M.P. 212-212.5°. Anal. Calcd. for Found: c?9N 0: C, 47.23; H, 7.14; N, 33.05. 3 C, 47.36; H, 7.25; N, 32.61. Curtius3 reported the semicarbazone of cyclobutanone, M.P. 211-212° (dee). The phenylhydrazone of (I) was prepared by adding just enough alcohol to 1 ml. of water and 0.10 ml. of (I) to give a solution and then ad.ding 0.2 ml. of phenylhydrazine. An oil separated in 2-3 minutes. The oil was induced to crystallize by removing all of the aqueous phase, adding a few drops of absolute alcohol and .s cratching. On recrystallizing three times from water-alcohol mixtures pale yellow needles were obtained, M.P. 98-98.5°. days. The phenylhydrazone decomposed ~fter standing for three Curtius reported the phenylhydrazone of cyclobuta,none, M.P. 95-96°. The 2,4-dinitrophenylhydrazone of (I) was prepared by adding 0.1 ml. of (I) to a boiling mixture of 0.2 g. 2,4-dinitrophenylhydrazine, 5 ml. of 50% ~lcohol and 1 drop of concentrated hydrochloric acid. An immediate color change was obtained, the mixture was heated at 80° for 10 minutes, enough alcohol added to ca.use complete solution of the derivative and the mixture set aside to crystallize. Two recrystallizations from absolute alcohol gave orange-red needles, M.P. 147.1-147.3°. Found: C, 48.49; H, 4.29; N, 22.53. Lipp, Buchkremer and Seeles 4 reported the 2,4-dinitrophenylhydrazone of cyclobutanone as orange-red needles, M.P. 132.3° and an acceptable nitrogen analysis for the derivative. On the basis of the 0, Hand N analysis of our derivative and the agreement between the melting points of the semicarbazone and phenylhydrazone with those reported by Curtius, it seems reasonable to believe that the melting point of the 2,4dinitrophenylhydrazone of cyclobutanone is 147°. Kishner5 reported that cyclobutanone forms an addition COlllJ)ound with sodium bisulfite and has a b.p. 98.5-99°. (I) formed such an addition compound (somewhat less readily than cyclopentanone) and .had a b.p. approximating that reported by Kishner for cyclobutanone (the amounts available to us were too small for convenient fractionation). The higher boiling neutral fraction was transferred to a special micro still having a very short vapor path. On distillation, 1.36 g. of material boiling at 60-70°/5 mm. (II) was obtained. A forerun of 0.29 g. was collected and a small residue left in the still. The material boiling from 60-70°/5 mm. had a very aromatic terpene-like odor similar to that of green hemlock. (II)gave a phenylhydrazone which could not be induced to crystallize. The semicabazone formed very slowly in low yield, and was not investif;ated further. (II) formed an addition colllJ)ound with saturated sodium bisulfite. On re- crystallization from saturated sodium bisulfite solution it separated in white needles. Ohromatography of the 2,4-Dinitrophenylhydrazones Prepared from (II).-The 2,4-dinitrophenylhydrazone mixture was prepared by adding O.l ml. of (II) to a hot mixture of 0.2 g. of 2,4-dinitronhenylhydrazine and 3 ml. of glacial acetic acid. The color changed slightly and the hydrazine went into solution. The mixture was heated 20 minutes at 80° and then the acetic acid evaporated in a stream of nitrogen. The last 32 traces of acetic acid were removed from the red gummy residue at 1 mm. and 80°. The residue was taken up in benzene and chromatographed on alumina,• Three bands appeared on the column, a thin dark •• brown very strongly absorbed band on the top of the column (III), a moderately strongly absorbed yellow band (IV) and a very weakly absorbed orange band which was generally washed completely into the filtrate (V). The top band (III) was eluted with alcohol and evaporated to dryness in a stream of nitrogen (10-20 mg.). dark and had a reddish brown cast. The residue was very It dissolved very readily in alcohol and Wc'\S precipitated by the addition of benzene or isopropyl ether, in which it was insoluble. It was not investigated further. The yellow band (IV), which corresponds to the main constituent in the mixture, was eluted with';;$ acetic acid in benzene. The filtrate from the elution was evaporated to dryness (230 mg.) in an inert atmosphere and the residue crystallized from benzene. This gave clusters of orange needles, M.P. 186-187°.••• Anal. Ca.led. for C14H16N4 05 : Found: C, 52.49; H, 5.04; N, 17.50. C, 52.96; H, 5.30; N, 17.93. * For a descri~tion of the methods of chromatography see Zechmeister and CholnokyO. ** Chromatographic absorption on talc, alumina, aluminum phosphate and other absorbents from benzene, ligroin and chlorinated hydrocarbon solvents has been applied to the separation of 2,4dinitrophenylhydrazones by Strain7. *** This is the temperature at which a fresh sample will melt when placed in the bath for 10 seconds. 33 The column filtrate containing the orange band (V) which was washed throug.11. the column was evaporated to dryness (4o mg.) in an inert atmosphere. The residue crystallized from benzene and isopropyl ether in clusters of scarlet blades, M.P. 184-187° (Sinters)*. Sample deepens in color several degrees below this. Found: C, 55.99; H, 5.0l; N, lS.93. Investigation of the Higher Amine Fraction.--The "higher amine 11 concentrate was obtained from the ether extra.ct of the alkaline pyrolysis distillate after removal of the neutral components from the acidified mixture. It was transferred to a small distilling flask equipped with a capillary. This was connected to a Dry Ice trap and ever--Jthing that would distill below a bath temperature of 90° and a pressure of 2 mm. was taken into the trap. This was distilled through a small Young-ty-pe column; 0.2 ml. of material boiling from 90-110°/760 mm. (VI) and 0.3 ml. of material which would distill below a bath temperature of 150° at 25 mm. (VII). A small residue remained in the flask. One ml. of methyl iodide was added to (VI). White crystals immediately separated, but turned to a milky oil in about 30 seconds. Absolute ether was added and the methyl iodide-ether solution poured off. This was shaken with semicarbazide hydrochloride and sodium acetate in a small amount of water. A small amount of semicarbazone was obtained which was found by a mixed melting point to be identical with the cyclobutanone derivative. This is the temperature at which a fresh sample will melt when placed in the bath for 10 secona.s. 34 The oily material was taken up in a small amount of water and treated with saturated sodium picrate. A precipitate was obtained which crJstallized in yellow needles from alcohol and water, M.P. 269-272° (dee). Anal. Ca.led. for Cz2H2gNg014: Found: C, 42.05; H, 4.50; N, 17.s3. C, 41.83; H, 4.73; N, 16.74. In spite of the low nitrogen analysis, consideration of the possibilities leads to the conclusion that the compound is probably tra.ns-1, 2cyclobutane-bis-( trimethylammonium) picrate. This compound would be formed if(VI) should contain any trans-1,2-bis-(dimethylamino) cyclobutane. Treatment of a few drops of ~II) with an ethanol solution of picric acid gave a precipitate. This crystallized from alcohol and water in yellow needles; M.P., dee. sharply at 239°. The identity of this compound was established by an approximate agreement of the d.ecomposition point and exact similarity of crystal form as seen by microscopic comparison with an authentic sample of the picrate of trans-1,2-bis- (dimethylamino) cyclobutane. In one pyrolysis carried out at 350-4o0°, the contents of the hydrochloric acid wash bottle were co~bined with the pyrolysis distillate and the neutral material removed with ether. The acid solution was then evaporated to dryness in vacuo, taken up in a small amount of water and slowly added to potassium hydaroxide pellets in an apparatus arranged for the absorption of volatile amines in hydarochloric acid. All of the amines which would distill from saturated :potassium hydroxide solution below a bath temperature of 60° were carried over. An oily layer remained in 35 the distilling flask. This changed to a brownish amorphous solid on standing over concentrated base for 24 hours. Steam distillation removed no more basic material from the residue. It was thoroughly wa shed with water and dried in a vacuum desiccator; weight of residue, 1. 7 g. The amorphous chocolate-brown residue (VIII) was very slightly soluble in water, very soluble in dilute hydrochloric acid, insoluble in dilute base, soluble in alcohol, pyridine and nitrobenzene; insoluble in ether and isopropyl ether. Fractional precipitation from hydrochloric acid with dilute base, showed that the solid was not homogeneous-the last flocculent precipitate obtained w-a s a pa.le yellow color. Chromatographic absorption of (VIII) from alcohol on alumina showecl it to contain at least three components. One component was strongly absorbed at the top of the column as a brown layer with a weak orange fluorescence under ultraviolet light. Another more weakly absorbed orange yellow band with a lemon yellow fluorescence under ultraviolet light and a filtrate having a pale yellow color and greenish fluorescence under ultraviolet light were also obtained. The mixture was not investigated further. Investigation of the Volatile Amine s .--The acid solution containing the volatile a.mines was evaporated to dryness in vacuo at 100°, giving 12. 5 g. of a mixture of volatile amine hydrochlorides. These were dissolved in water and the volume made up to 100 ml. One tourth of the solution was added slowly with shaking to 4-0 ml. of 20% potassium hydroxi de in an ampoule chilled in an ice se.lt freezing mixture. .36 Then 2.l~ ml. of benzenesulfonyl chloria.e was added and the ampoule sea.lea.. It ·was allowect to warm to room temperature while shaking-- large white crystals separat,ed. The ampoule was heated on a water bath at 60° for 20 minutes (the crystals turned to an oil), cooled, opened and extni.cted with three 20 ml. portions of ether. On evaporation of the ether 0.81 g. of solid was obtained, M.P. 47° after recrystallization from isopropyl ether. This is the melting point reported for the 8 dimethylbenzenesulfonamic1e . Identical control experiments carried out with dimethylamine and trimethylamine hydrochloride mixtures of known composition indicate that the original mixture contained approximately one-fifth a.imethylamine. The presence of trimethylamine was established by elimination, from the fact that the volatile amine mixture distilled completely between 0° and 10°. The presence of other amines which might form benzenesulfonamicles is excluded from the fact that the low melting dimethylamine derivative was obtainecl as a solid without recrystallization. 37 Acknowledgment I wish to thank Dr. E. R. Buchman for suggesting the problem and Mr. Alf O. Reims for preparing the starting material for the research. 38 SUilU!l8, ry The thermal decomposition of trans-1, 2-cyclobuta.ne-bis( trimethylammonium) hyd.roxid.e on a platinum catalyst at 350-360° w·a.s studied in detail. No evidence was obtained which would indicate that any cyclobuta.diene or l-dimethylaminocyclobutene-2 were formed <:luring the pyrolysis. Trans-1,2-bis-(dimethyla.mino) cyclobuta.ne, cyclobutanone, aldol condensation products of cyclobutanone, dimethylamine and trimethylamine were shown to be present in the pyrolysis distillate. 1-Dimethylaminocyclobutene-l was postulated as an unstable intermediate, reacting with water under the cond.i tions of the experiment to give cyclobutanone and dimethylamine. 39 Bibliography ~61, 1943 (1939). 1. Hurd and Drake, J. Am. Chem. Soc., 2. Sou, Bull. faculte sci. univ. fra.nco-chinoise Peiping, 2• 3. Curtius, J. prakt. Chem., (2) 4. Lipp, Buch.l{remer and Seeles, Ann., ~ . 20 ~!-, 22, 4463 (1936). ~. 262 Kishner, J. Russ. Phys. Chem. Soc., 22, ]-i,0§ I, 123. 6. (1916). (1932). 924 (1907); Chem. Zentr., C\.v,J Zechmeister and Cholnoky, "Principles ¢' Practice of Chromatography" Translation by Bacharach anc" Robinson, Chapman and Hall, Ltcl., Lonclon. 1941 Strain, J. Am. Chem. Soc., ;j], 8. 1 (1935); 758 (1935). Shriner ana_ Fuson, 11 Identification of Org2.nic Compounds 11 , John Wiley and Sons, Inc., New York,~. p. 119. 40 Propositions 1. The series of reactions: NaI 601 ~ is less likely to give cyclopropene than the reaction Dc1 2. KOH g An explanation is offered for the following data obtained when 10 g. of 1,2-dibromocyclopropane was treated with zinc dust and 80% alcohol at 80°. Time (min.) Volume of gas evolved (cc.) 0 16 0 200 1100 1200 20 25 55 1250 3. Another mechanism is proposed to explain the following configurational changes - (1) retention with S001 2 ; (2) partial inversion with S001 2 and pyridine; (3) partial inversion with PC1 3 ; (4) complete inversion with PC1 3 and pyrilline - in the halogenation of optically active phenylmethylcarbinol. 4. (Kenyon, Phillips and Taylor, J. Chem. Soc. 1931, 382) Ethyl carbonate is proposed for use as a solvent in many sodio malonic ester and similar types of condensations in ord.er to avoid some of the difficulties which arise when ether, benirene or alcohol are used in this ,_,._,ay. 5. Mereshkowski 1 s claim (Zent. 1914, I, 2161) to have synthesized methylcyclopropene is doubtful. 6. The strong tendency for the following reaction to take place, even in the presence of very small amounts of alcohol (Cope and McElvain, J.Am.Chem.Soc. 2!, 4311 (1932)), BrCH=CHC(Et)(COOEt) 2 + EtOH - - - BrCH 2 CH=C(Et)COOEt + Et 2 C0 3 7. The use of camphor-an.re in the Bast method for determining molecular weights is a useful refinement. 41 8. A modification of the method for carrying out sod.io malonic ester and similar condensations is proposed for use in the following cases. i) Preparation of 1,1-dicarbethoxy-ring compounds from dihalides, ii) Preparation of the monocondensation products with compound.s possessing two active groups, iii) Preparation of mono-substituted malonic esters from compounds having a high tendency toward di-substitution. 9. The following method is proposed for preparing certain difficultly obtainable substituted malonic esters. The cyanacetic ester is treated with hydrogen chloride and. alcohol to give the imino-ether hydrochloride and then hydrolyzed to the substituted malonic ester. 10. The freezing point depression and analysis for certain inorganic constituents in the blood serum of cambarus clarkii (crayfish) reported by Lienemann (J. Cell. and Comp. Physiol. 11, 149 (1938)) is selfcontradictory. 11. The inhibition of the skeletal muscles of the crustaceans is not closely related to inhibition in the vertebrates. 12. A semimicro analytical laboratory should be set up to supplement the micro analytical department. 13. A scheme of research direction involving the following points is suggested. 1) Course work for the first year includes a :paper on some llllnited field within the sphere of interest of one of the professors in the department. Later a seminar is given by the new student on some phase of the same subject, with emphasis on unsolved 42 problems in the field and projected schemes for attack of these problems. 2) A new man, on arrival at the school, is assigned to a senior graduate student and, after careful discussion with the professor and senior student, begins work on some little problem of immediate interest both to the professor and to the student w;i.th whom the new man is to work.