An Investigation of the Synthesis of "Pirylene" Citation Farquhar, John Percival (1941) An Investigation of the Synthesis of "Pirylene". Master's thesis, California Institute of Technology. doi:10.7907/zgmr-qz28. https://resolver.caltech.edu/CaltechTHESIS:02102025-185855030 Abstract No abstract. Item Type: Thesis (Master's thesis) 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:02102025-185855030 Persistent URL: https://resolver.caltech.edu/CaltechTHESIS:02102025-185855030 DOI: 10.7907/zgmr-qz28 Default Usage Policy: No commercial reproduction, distribution, display or performance rights in this work are provided. ID Code: 16989 Collection: CaltechTHESIS Deposited By: Benjamin Perez Deposited On: 11 Feb 2025 16:49 Last Modified: 11 Feb 2025 17:16 Thesis Files PDF - Final Version See Usage Policy. 7MB Repository Staff Only: item control page

.AN INV.E:3TIGATION° OF THE SYNTHESIS OF 11PIRYLffi'B" Tb.esis by John P. Farquhar In Partial fulfillment of the Requirements for the Degree Of Master of Science. California Institute of Technology Pasadena, California. 1941. TABLE O.Il, CONTENTS. Theoretical Part •• • • • • • • • • • • • • • l I. Historical Review • • • l II. Purpose and Outline of the Investigation ••• 3 III. Preparation of the Quaternary Pyrrolidinium .. ....... • • Bromide •• • • • • • • • • • • • • • • • • • IV. 4 Degradation of the Quaternary Bromide to the Isomerio Bases •• • • • • • • • • • • • • • • 5 v. The Nature of the Isomeric Bases • • • • • • • 7 VI. The Separation and Salts of the Isomeric Bases 7 VII. Catalytic Reduction of the Quaternary Salt of VIII. IX. the Stable Base • • • • • • • • • • • • • • • 8 The Formation of Pirylene •• ... ..• • • 9 A suggested Structure of Pirylene. • • • • • 9 Experimental Part • • • • • • • • • • • • • • 11 N-Methylpiperidine • • . • • • • • • • • • • ll Dimethylpiperidinium Iodide • • . . • • • • • 12 Dimethylpiperidinium Bromide 12 "Dimeth.ylpiperidine" • . • • • •••••••• . . . . . . . 12 N,N-Dimeth.yl-~bromomethylpiperidinium Bromide • • • • • • • • • • • • . . . . . . . 13 N,N-Dimeth.yl-i,{-dibromoamylamine Hydrobromide. • . • • • • • . • • . • • . • 13 Preparation of the Mixed Bases c5H7NMe 2, First Method . . • . . . . . • • 14 Second Method •• • • • • • • • • • • • • 14 Third Method 15 • • • • • • • • • • • • • • ,TABLE OF CONTfil.TTS (CONT.) Purifioation and Properties of the C H.,NM3 2 5 Mixture. Diliturates • . • • • • • • . • • . • • • 16 .Meth.iodides ... ... • • ... . . . . . . .. . . . . • 16 . 19 Reduction of the Quaternary Salt of the Stable Base and Itlentifioation of the Product .. .. •• 20 n-Amyltrimethylammonium Bromide and Picrate • • • • 21 Reduct ion of "Trimethylpiperidinium Chloride" • • • 21 Characterization of Quaternary Salts • • • • • • • 22 Pirylene • • . • • • • • • • • • • • • • • • • . • References • • • • • • • • 23 . . . • • . . . • • • • 25 I wish. to tb.a.llk Dr. Edwin R. Buchman for suggesting the problem and also Mr. Herbert Sargent for permission to include certain results obtained in his investigations. AN INVESTIGATION OF THE SYNTHESIS OF "PIRYLEN.H:. 11 Theoretioal Part. (1) I. Historical Review of the Synthes is of Piry1ene. (2) In 1888 Ladenburg obtained by the Hofmann degradation of piperidine the so-called 11 dimethylpiperidine" CH2 CH(CH2)3N(C~)2• Iodine was added to the double bond of this compound and on treatment with silver oxide a tertiary base C5H>,N(CH3 ) 2 oalled dimethylpiperidein was produced. The quaternary salt was formed which, on treatment with alkali, yielded trimethylamine and a hydrocarbon, C5H 6, of unknown structure, which was called pirylene. La.denburg's interpretation of the reactions was as follows: (3) (4) As a result of the investigations of Merling, Roser and (5) Willstatter, it was shown however that the first additi on product (I) of iodine or bromine with "dimethylpiperidine" goes over by rearrangement into a pyrrolidine derivative (II) which by treatment with alkali or silver oxide is converted to the base {III). This last compound on distillation leads to a pro- duct, dimethylpiperidine, with two different possible fornulas: :r-OH2I (Br) 3 H3 I (Br) C~CHCH2CH2If(CH3)2 V III Willstatter obtained salts of (III) 8Jld assumed the ring opening led to the structure (IV). He did not consider the possibility of (V) and paid little attention to pirylene. (l) In 1928 von Braun and Teuffert tried to clarify the relation between dimethylpiperidein and pirylene. The statements of Willstatter in regard to the base (III) and its salt were confirmed for the most part. ~'vidence on the structure of dimethyl- piperidein was obtained by catalytic reduction, 4H being added to produce homogeneous n-anwldimetbyla.mine. As a result dimethyl- piperidein was shown to have a straight chain. (IV) was thus eliminated as a possible structure and the structure (V) accepted since the index of refraction indicated two cuwlative double bonds. (V)- - - - C~CH2CH;aN(OH3)30H (VI) The fornnlation of (VII) as pirylene would follow from (V} but von Braun found that the addition of hydrogen showed two and not tb.ree double bonds. It was therefore assumed that (VII) went over by rearrangement into a five-, four- or three-membered ring. (Cl!__~ (VII) VIII Il, CH,e-<a!:CH 2 CH3 ( ci('CH IX X Cllz--Cll_~ CH3~~C=CHz XII XI Cyclopentadiene (VIII) was elimina t ed, being a knoVJn compound with di ff erent properties than those of piryl ene. Methylcyclobutadiene {X) was considered irrwrobable because it (6) had been shown by Willstatter and Waser , that cyclobutadiene is unstable and decomposes at higher temperatures into acetylene. The molecular refracti on found for pirylene was somewhat higher than calculated for (X) and also (XI) but agreed well with the values for ( I X) and (XII) • (I X) was considered the more likely being a simpler transition from (VII). Only acetic acid and oxalic acid could be obtained from pi:rylene by oxidation which did not help in choosing between (IX) and (XII) since each of these could give rise to both those acids. II. Purpose and Outline of the Investigation. In February 1939, under the direction of Dr. E. R. Buchman, I commenced the synthesis of pi:rylene. Our aim was to clarify aspects of the synthesis which had not been completely investigated or had been ignored by the previous workers, and also, incidentally, to prepare some pirylene f or elect r on di f fraction studies. 4 The synthesis was accomplished following essentially the original procedure of Ladenburg. Certain important modifications were introduced which permitted the obtaining of larger yields of a number of intermediates. Also a careful study of the reaction leading to the formation of dimethylpiperidein revealed that a single compound was not produced here and this finding lei us to investigate the separation and nature of the products formed. III. Preparation of the Quaternary Pyrrolidinium Bromide. (7) Piperidine was methylated by means of formaldehyde and formic acid in the presence of hydrochloric acid. O+ HCH0+HC0~Cl • OCH~+ C02+H20 This reaction gave metb.ylpiperidine accompanied by varying amounts of dipiperidyl methane and in some cases the product was contaminated by unmethylated starting material. It was found that dipiperidyl methane on further treatment with the same methylating agents (HCH0 and HC0OH) went over smoothly to methyl piperidine. It is therefore desirable to insure complete methylation of the piperidine by heating the reaction mixture and adding fresh reactants (HCHO and HCOOH) until CO2 evolution is reduced to the rate from a bla?lk experiment (HCH0-t- HC0(ll alone). Under these conditions a good yield of methylpiperidine can result. Dimethylpiperidinium bromide was formed from methylpiperidine by the action of methyl bromide. OOHFCH3Br---t ~V¾j- _ "---/-..........._ Br CH3 Tb.is reaction has not been described in the literature and has the 5 advantage that the relatively cheap methyl bromide may be substituted for methyl iodide. The reactants were combi ned at -45°, the reaction taking place as the temperature was slowly allowed to rise. Treatment of the quaternary bromide with silver oxide and (9) distillation of the resultant base over KOH gave the so-called 0 ti dimetb.ylpiperidine in good yield. C)iit.4 Br- AgOII > C)liie~ O I I - - - CHz=-CH(CH ) 3NMe • 2 2 The reaction of bromine with dimetb.ylpiperidine goes in two (5) steps to form the product obtained by Willstatter. CH-z-CHCH2CH2CH~ (Me) 2+ Br2 ~ CH2BrCH: BrCH20H2CH~2 rearrangement ~ C"J-.CH2~r mr;~;Br The reaction was carried out in chloroform solution from which the expected quaternary salt precipitates in quite good yield. It was found that in addition to the expected quaternary salt the reacti on gave considerable amounts of a by-product which when purified had a melting point of 158.2 - 158.7°. It could be shown from analytical results and othe r data that this compound is a hydrobromide of the base formed by adding one mole of bromine to the unsaturated base i.e. CH2BrCHBrCH 2CH 2CH~ 2 .HBr. When this compound is liberated as a 'f ree base fr om its salt with alkali it is obtained as a chloroform and ether soluble heavy oil which quickly isomerizeQ to the quaternary bromide. IV. Degradation of the Quaternary Bromide. As mentioned above dimetbylpiperidein (C 5H7NMe 2 ) h.as been 6 Obta ined by various investigators all of whom regarded it as essentially a single compound. In our experiments the degrada- tion of the quaternary bromide was carried out by three methods each producing a mixture of isomeric bases. By the first m3thod the bromide was converted to the quaternary base with silver oxide which also removed bromine from the side chain. The chloride was formed with HCl and the degradation carried out C]onJoH j ~"cH3 05H7NJ1.fe 2 By the second method the bromide in a small arnount of water was over KOH. A KOH By the third method the bromide was treated with t wo moles of silver oxide forming the bromine free unsaturated ammonium hydroxide which was distilled over KOH. CJ-oH2Br] ot(~CH3 a. KOH _ Br 2AgOH ➔ c-J=cHJ ~ ~ H 3 2JOH- ► From the standpoint of yield the second method proved superior. It required fewer and less costly rea,.gents and in our experience can be used equally well with large and small batches. The degradation product, consisting of isomeric C5H7NMe2 bases, was purified by vacuum distillation. A 50-60% yield of material boiling at 54 -650 at 40 mm. was obtained. 7 v. The Nature of the Isomeric Bases. Various facts indicated that we were dealing with a mixture of bases and that these bases have the structure (IV) and (V) (or some structure closely related to these by rearrangement of double bonds). For one of t hese bases, the stable one, fo!'lillla (V) was made probable by showing t hat the ca1•bon at oms in the C5H7 radica l are arranged The other base has little stability in the in a stra ight chain. presence of hydrolytic agents and apparently dec omposes in both acidic and basic soluti ons, particularly on heating, to give dimethylamine. This base presumably has the structure (IV) or (IVa), the double bond o.(l to the nitrogen atom being in accord with the obs erved easy splitting off of dimethylamine. __,,,,. CH3 -cH2Br1 CliF"~CHCH 2CH2NMe 2 Stable (VJ and _ OH~ H3 CHz-- OHCE=::QNlvie 2 (IVa) ( CH3 CH-z-CHOH2~.Me2 1 2 or (IV) HNMez VI. The Separation and Salts of the Isomeric Bases. ( 10) The mixture o f bases was converted to diliturates recrystallized from hot water. which were The diliturates separated into a difficultly soluble fraction and an easily soluble fraction which was also quite soluble in alcohol. Recrystallization of the difficultly water soluble fraction from hot water gave a pure dili turate which was found to be dimethylamine dil i turate. The easily soluble dili turate fractions were found to be 8 substantially the diliturate of the stable base. Recrystalli- zation from an ethyl alcohol-water mixture or from ethyl alcohol alone in every case gave needles of m.p. 159-161°. Regen era- tion of the base from the diliturate with sodium hydroxide gave the free base, boiling point 64-65°/49mm., d~ o.soo, picrate m.p. 100.5-101 0 • The methiodi de of the stable bas·e is formed by treating the pure base, isolated as described above, with methyl iodide. This methiodide, m.p. 247-248°d, in contrast to statements in the literature {von Braun), is quite stable in aqueous or alcoholic solutions. \'/hen pure it can be easily recrystallized from hot water in which it is rather difficultly soluble in the cold• It can also be isolated from the mixed methiodides obtained directly from the mixed bases.* VII. Catalytic Reduction of the Quaternary Salt of the Stable Base. Reduction of this quaternary salt should give normal a.myltrimethylammonium salt provided the stable base has the stra ight chain forllllla {V) postulated. The quaternary iodide and others of its type were found impossible to reduce with the method used• hydrogenation. ta~ iodide ion interfering with catalytic However, conversion to the corresponding chloride *From the mixed methiodides was also isolated the isomeric methiodide presumably corresponding to (IV) or (IVa). The amounts of this substance were however insufficient for further examination. Also present in the quaternary salt mixture from the mixed bases is tetramethylammonium salt derived from (IV) (or (!Va)) via dimethylamine. 9 gave an equally suitable material which could be hydrogenated. The reduced product was shown identical with n-amyltrimethyla.mmonium chloride by conversion to two solid crystalline derivatives, the bromide and picrate, and comparison of these {mixed melting points) with. material of known structure. This result, incidentallr, confirms the worlc of von Braun who, it will be recalled, reduced c51½NMe 2 and demonstrated the identity of the product with n-amyldimethylamine. ~"'vidently von Braun had a C5it,NMa 2 base which was nearly pure (V) (or a closely related substance). It is not surprising that he obtained only the stable isomer since under his conditions the unstable isomer rmy have been destroyed without his having been able to detect its presence. VIII. The Formation of Pirylene. The degradation of c5n7NMe 3I to pirylene was carried out as (2) described by Ladenburg. Pirylene agreed in properties with the (l) literature description. A KOH' IX. A suggested Structure of Pirylene Electron diffraction pictures of pirylene were taken by Schomaker in this laboratory. contains no ring. The results indicate that pirylene It will be recalled that von Brau.n's pre- ferred structures (see P• 3 above) each contain a ring. There was possibly some error in hls hydrogenation of the hydrocarbon. It may well be that pirylene b.~ a straight chain. Comparison 10 (11) Of (1) the physica l constants of methylvinylacetylene and pirylene show close agreements • Pirylene: b. 60°; d l, o.7443; n1~, 1.4505; 1 MR (found)23.ss. l-Methyl-2-vinylacetylene: b16O 59.2°; d ~ 2, o.7401; n 2i, 1.4496; 2 ( found) 23.94 Jacobson and earothers (ll) r eport tb.D,t methylvinylacetylene poly111erized on long standing. We found t hat pirylene polymerizes on long standing (see P• 24). The methYlvinylacetylene structure mi ght arise as follows: Q-cH2Br}llf ---, Cl!z--C=H Cl!2C!!:fJ.le2 ~ CHZ ~ CE.z= ~HCH 2CH~3I ~ CH30::3JCH=CH 2 . In view of the above considerati ons we suggest the identity of methylvinylacetylene and pirylene. ll Experimental Part. (7) (8) N-Methylpiperidiµe. - 557 g. of piperidine were mixed with 1000 g. 37% aqueous formaldehyde in a 5 l~ round bottomed flask in an ice bath. 653 cc. of 12 N. aqueous HCl was added slowly with continued ice cooling and then 1000 g. of 85% aqueous formic acid was added. The flask was fitted with a reflux condensor (ground glass joint) to the top of which was attached a tube by means of which evolved CO2 could be recover ed over water. The refluxing was continued several hours until the gas evolution was slow. The rate of evoluti on wa s compared to the r ate f r om a blank of proporti onate amounts of fol.'maldehyde, formic acid and hydrochloric acid. The mixture was transferred to an erl enmeyer mid while cooling with an ice bath excess satura ted aqueous NaOH was added and then NaOli pellets. The mixture was filter ed through sinter/ed glass with mil d sucti on and the filtrate was separated, to give an aqueous phase and a base phase. The aqueous phase and the precipitate were washed with Et 2o and the washings combined. NaOH pellets were added to the base phase which was then distilled over sodium, collecting the main fraction between 100 and 110°. Yield about The higher boiling residue was remethylated using the same procedure with proportionate a.mounts of the methylating agents. The yield was fu r ther increased by distilling the ether washings. In other runs unmethylated piperidine was found in the main fraction necessitating remethylation. can be tested f or by the addition of cs base but not a tertiary amine. The pres ence of piperidine 2 which precipitates this 12 Dimetb.ylpiperidinium iodide. - .An ether solution of metbylpiperidine was treated with excess CH3I and the quaternary iodide was deposited. After filtering the iodide was washed with Et 2o and recrystallized from 95% ethyl alcohol. Dimeth..ylpiperidinium bromide. - 124 g. of methylpiperidine dis- solved in 1240 oo. of anhydrous ether in a 2 l. erlenmeyer was cooled to -50° in an alcohol-solid carbon dioxide bath and about 170 g, of methyl bromide was added. The temperature was allowed to rise slowly to o0 and the mixture was placed in the cold room and then allowed to stand at room temperature (total length of time at o0 - room temperature at least 4 days). After fi 1tering and drying, 235 g. of the quaternary bromide were obtained. Yield, 96.7% theory. (9) "Dimetl;vlpiperidine." - 200 g. of dimet bylpiperidinium bromide in aqueous solution was treated with about 20% excess AgOH and the mixture was filtered through sintered glass with suction, the precipitate being washed with a little water and the washings combined with the filtrate. The filtrate was transferred to an open dish and evaporated down on a steam bath. The residue was combined with 100 g. KOH and a little water and distilled. The distillate was saturated with NaOH and separated giving 109 g. Of crude base. The base was dried over NaOH pellets and distilled over sodium. There was 5.5 g. of forerun to 112° and a main fraction 112-118° of 94.3 g. and a small residue. Yi eld 81% t heory . 13 (5) N,N...Dimeth.yl-e,,.-bromomethylpyrrolidinium bromide. - 50 g. of "dimethylpiperidine 11 dissolved in 350 cc. of dry (recently distilled) chloroform were placed in a 2 1. three necked flask fitted with a dropping funnel and a stirrer and placed in an ice bath. 70 g. of bromine dissolved in 500 cc. of chloroform were added from the dropping funnel drop by drop with stirring. After the addition of all the bromine the temperature was allowed to rise slowly. Too rapid heating can result in a violent formation of the quaternary salt by rearrangement so the temperature nmst be kept low until the salt is formed. The mix- ture was filtered with suction and the precipitate dried in a vacuum dessicator to give 101.5 g. (84%) of bromide. This salt was recrystallized by dissolving in about 230 cc. hot 90% EtOH and letting stand in the cold room. and drying, 83.3 g. (69% After filtering theory overall) of quaternary bromide of m.p. 228.5-230° were obtained. Samples were recrystallized The. yield can be increased by evaporation down and recrystallizing the alcoholic mother liquors. N.N-Dimeth.vl- & (-dibromoaJILVlamine h.ydrobromide. CH 2BrCHBrQH:a@~~~This substance is a by-product of the synthesis of the above paragraph. The chloroform mother liquor was distilled and the residue recrystallized from 90% EtOR and the crystals pressed on tile and recrystallized. Such a sample was recrystallized from ethyl alcohol using norite and was easily brought to a constant melting point of 158.2-158.7°. 14 Anal. Calcd. for C7H 16NBr3 : C~ 23.75; H, 4.56; N, 3.96; Br. -22.58. Fow1d: c, 24.15; H, 4-62; N, 3.98; Br, -23.32. The addition of s odium picrate to an aqueous solution of the hydrobromide gave a picrate which on recrystallization from ethanol melted at 105-105.3°. Preparation of the Mixed Bases 0&~7-N:Me 2 (so-called dimethylFirst Method. - 20.3 g. of the quaternary bromide was piperidein}. dissolved in water and tre ated with AgOH made from 30 g. AgN03 (about 20% escess). The mixture was shaken a few times, filtered through sinterred glass and the filtrate was neutralized li trrD.1s with HCl, and a,v aporat ed on a steam bath• to The r esulti ng crystall ine deliquescent Viass \'!D,,S t r ansf err ed with a littl e water to a 50 cc. distilli ng fl ask conta i ning 20 g. of KOH pellets and distilled. It was necessary to treat strongly to complete the decompositi on. Sodium hydroxide wa s added to distillate and the phas es were separated with a sepa.ratory funnel. was distilled over sodium. The base phase The fo r e run, mostly 130-135°, was about o.5 g.; the main fraction, 135-144°, 2.0 g.; and a small higher boiling fraction and residue. Yield, 24.3% theory. Second Method. - 91 g. of the quaternary bromide, 56 g. KOH pellets and 100 cc. H20 were c ombined in a 250 cc. distilling flask and distilled using an oil bath. The bath was heated gradually to 265° and kept there until no mo r e oil came over. KOH was added to the distillate which was separated giving 33 g. of crude base. The base was distilled at 40 mm. pressure. 15 The forerun, 4.0 g., being collected to 54°; the main fraction, 25.4 g., 54-65°; and a higher boiling fraction of about lg. to about 10°. The small residue was discarded. Most of the main fraction was from 58° to 64° and in other runs the range for the main fraction was 54-64°. Cloudiness in the main fracti on sug- gested the presence of water so the fraction was redistilled at 40 mm. to give 22.a g. base coming over in the range 56-67°. There was a small forerun of o.5 g. and about lg. free flowing residue. Yield 61.6% theory from the quaternary bromide. The yields for the other runs using essenti a lly this method varied from 50% to 65.8% theory. Third Method. - 20 g. of the quaternary bromide were dis solved in H20 and treated with AgOH (from 27.4 g. AgN03 - 2(ffo excess). The mixture was stirred and filtered through sinterred glass. The filtrate was transfe r red to a distilling flask cont aining 20 g. KOH and distilled until no more base came over. KOH was added to the distillate and the base phase was separated, dried over KOH and distilled. There was 0.4 g. of forerun to 135°; 3.3 g. main fraction , 135-141°; o.5 g. of a fraction, 141-144°; and a residue. Yield 40.5% theory. The mixed bases in a stoppered flask invariably precipitated a dark oily water soluble phase on standing for a few hours. Purification and Properties of the C51!7NMe2 Mixture. Diliturate§, - The diliturates were prepared by adding an equivalent amount or a slight excess of base to a saturated hot water solution of diliturio acid. Diliturates from the mixed bases were recrystallized from hot water 16 and wera separated into a. difficultly soluble fraction and an easily soluble fracti on which was also quite soluble in alcohol. The difficultly soluble fraction was recrysta llized from hot water to a pure diliturate (identified as dimethylamine diliturate). Its solubility coincided exactly with that found for a knovm sample of· dimethylaminediliturate, 36.6 mg./10 cc. of water at 23°. The melting points were above 275° • .Anal. Calcd. Found: C5H10N405: C, 33.03; H, 4.62. C, 33.32; H, 4.75. Treatment of the diliturate with excess alkali and distilling the liberated base i nto a small amount of water resulted in an aqueous dimethylamine soluti on which gave a picrate, m.p. 157-157 .5° • .An authentic sample of dimeth~rlamine picrate melted at t he same point and the mixed melting point was not depressed. The di li turat e fractions easily solubl e i n water, substantially the diliturate of t he stable base, when recrysta llized from ethyl alcohol-water mixture or from .l!.'tOH alone, ga.ve t he same substance. This diliturate crystRllized well in needles m.p. 156-161°. The analysis of the substance was not entirely satisfactory, due perhaps to t he presence of diff icultly r emovable wat er of' crystallization, but there can be no doubt of its essenti a l composition. For re- generati on of the pure base see below (p.19). Methiodides. - The fo rmation of mixed methiodides from mixed bas es was in one case carried out by mixing 22 .s . g. of t he ba ses C5H 7NMe 2 with 50 cc. CH30H in a flask cooled with an ice bath and Some heat was devel 9ped so t he reacti on was 17 controlled by shaking in the ice bath. The mixture was allowed to stand about one hour at room temperature and was then cooled, filtered and the precipitate washed with C~0H to give 20.5 g. dry white crystals • In this and other cases the first crop of crystals deposited was found to be a lmost entirely the methiodide of the stable base together with another substance, presumably a difficultly water soluble quaternary iodide, perhaps partly amorphous since some solutions containing it were difficult to crystallize and giving upon treatment of aqueous soluti ons with sodium picrate, a non-crystalline sticky picrate insoluble in water and alcohol,* Also there is evidence (see below, P• 21) that Me4NI** is present. The separa tion of the by-products from the methiodide of the stable base was not too easily accomplished. Recrystallization several times f rom wate r gave a pure quaternary salt m.p. 259°d, which with sodium picrate gave a pure picrate. Anal. Calcd. f or C8H15IH: C, 37.96; H, 6.37; N, 5.53. Found: C, 38. 33; H, 6.60; N, 5.57. Possibly t he best me thod of obtaini ng t he pure quate rna ry ha lide of the stable base was found by converting the insoluble mixed methiodides to picrates from which the picrate of the 05H7Nl)lfe3 base was easily separated by extracting it with ethyl alcohol and recryst allizing , m.p. 112.5-113°. As menti oned above the picrate *This by-product may be in part the result o f some sort of Di els-Alder reaction of a quaternary i.od.:blie with its elf or its isomer . 18 of the by-product is quite insoluble in alcohol. The purified picrate was treated with halogen acid in aqueous suspension, the pioric acid removed with etbyl acetate and the aqueous solution evaporated to obtain pure quaternary salt. The use of picrates can be applied in the same way to isolating the isomeric methiodide corresponding to the unstable base. This isomeric methiodide is contained for the most part, presumably, in the CH3 0H mother liquors aft er the methiodides have been formed and the first cr op of crystals deposited. The mother liquors also contain (besides small quantities of the insoluble o5H7 methiodide) salts of presumably quaternary bases which give oily insoluble picrates such as were mentioned above. The separation of the isomeric methiodide by recrystallization was a long, tedious procedure with considerable loss, but material was obtained which gave the proper analytical figures. Howev er, the conversion to picrates, separation f rom the alcohol insoluble sticky picrates and regeneration with halogen acid gave satisfactory results. A quaternary iodide was obtained which, on r ecrystalli- zation from ethanol-isopropyl ether using norite and then from ethyl alcohol, had a m.p. 227-228°d. and solubility properties similar to t he other isomer. .An oily picrate was formed which solidified in the icebox and ha~ not been further investi gated because of its l~v melting point and the consequent difficulty of purification. 19 Found: C, 37.59; H, 6.62; N, 5.68 ■ The Pure, Stable Cs1f7NMe 3 Base. - This ba;,se would be prepared from the pure diliturate. Regeneration was easily brought about by treat- ment with sodium hydroxide, steam distilling, extracting with ether, drying with potassum hydroxide and distilling in vacuo. 25 boiled at 64-65°/49 :rmn. and had d 4 o.800 ■ The base Anal. Calcd. for C7H1~: C, 75.60; H, 11.79. Found: C, 75.89; H, 12 ■ 10 ■ '.rhe pure base on standing remains clear. (The dark material precipitated from mixed bases on standing therefo r e owes its origin to the unstable isomer). With ethereal picric acid the base forms a picrate vrhich crystallizes from ethanol-isopropyl et her in well formed cubes 0 m.p. l00.5-101 • Found: C, 46.35; H, 4.86; N. 16.25. With methyl iodide the methiodide was formed. crystallized from hot water, m.p. 247-248°d. It was re- (The difference in melting point between that observed here and that given above was probably .due to a difference in the rate of heating in the melting point apparatus). With sodium picrate the methiodide in aqueous solution, formed a picrate, m■ P• 112.5-113° from ethanol. Tb.us the base has been charact erized by a dilitura te, a picrate, a methiodide and a quaternary picrate. 20 Reduction of the Q;µ.aternary Salt of the Stable Base and Identification of the Broduct. - The catalyst used i n the reducti ons was prepared as follows: Of' 1/2 ml. of 10% PdCl 3 (in H20 HCl) was placed with l gram freshly ignited norite in 100 ml. of distilled H o. 2 The mix- ture was shaken with H2 a~ 30 lbs. pressure for 10 minutes, with mild suction and washed with c.P. methanol. filtered 5 g. of the quaternary iodide and the above catalyst in methyl alcohol were shaken for t wo hours under 37 lbs. pressure of H2 without any absorbtion of hydrogen. The catalyst was filtered out and shaken with 20 g. mesityl oxide in meth.,yl alcohol for 1/2 hour at 37 lbs. pres sure of H2 with no reaction. 1.'l i th the same mix- ture using fresh calalyst the pr essure dropped to 20 lbs. in 15 minutes. Some qua.t ernary chloride was prepared by adding an aqueous solution of the iodide (not entirely pure) to a 30% excess of Agel~ filtering and evaporating the f iltrate on a steam bath. The salt was found to be very hygroscopic. It was dried in a vacuum dessicator and hydrogenated using the same procedure as with the iodide. Roughly 2 moles of hydrogen were added. The catalyst was filtered off and the filtrate evaporated down and dried. The reduced product was identified as a sa.l t of n..C 5H11NMe 3 base by conversion into both the picrate and the bromide and these derivatives were then compared with authentic samples as described below. It was found that each derivative had the same melting point as the corresponding known compou11d and that the mixed melting points were not depressed. The bromide was nade 21 by treating a sample of chloride with excess .Ag20, filteri11g, adding excess HBr and evaporating to dryness. The product was separable into t wo parts by means of a di f ference in solubility in ethanol. The alcohol insoluble portion had no melting point, gave a picrate which melted above 210° and on heating with KOH evolved a volatile base having an odor similar to that of NMe 3 • Analysis of this alcohol insoluble quaternary bromide indicated that it was The alcohol soluble bromide substantially pure Me NBr (see P• 17). 4 after recrystallization from alcohol-isopropyl ether melted constantly a:t 181-181.5 0 • Its aqueous solution treated with sodium picrate gave a picrate m.p. 93.2-94° (mixed m. p . 93.4-94°). n-Alqyltrimethylarmnonium bromide and picrate. - 5 g. n-a.myl bromide and 6.8 g. (25% excess) 36% NMe 3 in benzene were placed in a tube which was then _s ealed and kept in boiling water for 24 hours. Tb,e product was filtered with mild suction, washed with anhydrous ether and dried in a vacuum dessicator. Yield, 4.58 g. which was 65.8% theory. The bromide was crystallized twice from absolute h'tOH to a m.p. of 0 181-181.5 • Its aqueous solution with sodium picrate gave an oily picrate which solidified when placed in the icebox and did not melt again at room temperature. It was washed with water, dried and recrystallized from ethanol-isopropyl ether coming down in prisms Reduction of 11 Trimethylpiperidinium chloride•" - The methiodide was formed by dissolving 11.2 g. of 11 dimethylpiperidine 11 in an equal volume of absolute ~tOH, cooling in an ice bath, and adding 16.32 g. CH 3I (15% excess) slowly in small portions. The p r niuct _was filtered with mild suction, washed with anhydrous Bt20, dried in a 22 vacuum dessicator (yield about 100%) and recrystallized from not absolute EtOH to give 14 g. of iodide m.p. 220-2°. Yield 55.4% theory. The quaternary chloride was formed by treating an aqueous solution of the iodide with a 50% excess of AgCl, filtering and evaporating the f iltrate on a steam bath. to be very hygroscopic. The salt was found On treatment with sodium picrate it gave an oily picrate which became crystalline on standing. Recrystallized from ethanol-isopropyl ether it had m.p. 98.5-99Q. It was f ound that it di d not depress the melting point of the C5H11NMe 3 piorate to any great extent. The maximum depression was obtained with about 10% c5H9NMe3 piorate added to the other (mixed m.p. 90.5-92°). Other examples have been noted in which picrates of similarly constltuted structure did not depress the melting point of each other when mixed. J~st as with c5H7NMe 3 I the quaternary iodide of dimethylpiperidine could not be reduced catalytically, therefore the chloride was reduced, following the procedures given above for The reduced product was converted to the picrate which was compared by means of melting points and mixed melting points and found identical with an authentic sample of Char acterization of Quater nary Salts. - Some att empts were made to wor k out a general p ro cedure for charact er izing quaternary sal ts of the t ype found in t his resear ch. It was diffi cult t o get 23 get easily crystallizable salts having a r eproducible melting point. The iodides are hygroscopic. melt with decomposition and the chlorides Precipitation of the quaternary salts from aqueous solution is not easily done: sodium S-benzylthiosulphate, potassium thiocyanate, sodium styphnate and sodium picrolonate were tried without success. The picrates ar e low melting and quite soluble in mal1y solvents. There were indications in the literature that Reinecke salt might be useful. mxperiments were car ried out which showed t hat thi s agent may be satisfactory. An impure Reinecke salt produced large precipitates wi th aqueous solutions of the quaternary salts but these precipitates were in all cases mixtures being definitely separable into alcohol soluble and alcohol insoluble material. The a lcohol soluble material can be recrystallized from alcohol and in the case of the n-amyltrimethyla.mmonium bromide derivative gave after several recrystallizati ons a substance of m.p. 159.5-160°. The formation of more than on e compound may have been due to impurities in the Reinecke salt used. A highly purified Reinecke salt mi ght form useful derivatives. Pirylene. - 19.5 g. c ~NMe I 3 5 (MeOH insolub l e methiodide mixture) together with 9 . 0 g. KOH dissolved in a small amount of H2o in a 50 co. distilling flask were distilled usi ng an oil bath. The rec e iver was tightly f itted to the end of the condenser, there being an outlet attached to a rubber t ube fr om the top of the 24 receiver and the receiver was in an ice bath. distillate was collected. A two phase The oil bath was carried to 150° for the first fraction and a second fraction was collected when the bath was 150-250°. The first fraction was acidified, and the oil pb.ase separated with a small separatory funnel. fraction had only an aqueous pb.ase. The second Sodium was added to the oil which was put, lightly stoppered, in the cold room. A final distillation was carried out over sodium and l.98 g. of material boiling in the range 54-59° (mostly 58°) was collected. Yield, 39% theory. On standing over a period of months the material polymerized to a gum. 25 REFEilllli"CliS. (1) von Braw1 and Teuffert, Berichtei, fil. (2) Annalen 247 l (3) Berichte 17 2139 (4) ibid. ].2 1601 (1886) (5) ibid. M 365 (1900) (6) ibid•~ 1992 (7) E, R. Buchman, (8) H. T. Clarke, Gillespie and Yieisshaus, J. Am, Chem. Soc. 1092 (1928). (1888) (1884) Dissertation, Frankfurt am :Main 1933 P• 47 55 4571 (1933) (9) , Reference 7, P• 40 (10) Redemann a:nd Niemann, J. Am. Chem. Soc.~, 590 (1940) (11) Jacobson and Carothers, J • .Am. Chem. Soc. M 1622 (1933)