7/28/2019 ja01211a082

1/2

1368 FRANK, LEGRN,EARBORN,YERSN D WOODWARD Vol. 68in which [ q] is intrinsic viscosity; t o is flow time for ben-

TABLE I1POLYMERS ROM SUBSTITUTEDTYRENES

Soften--Analyses of polymer- ingCalcd. Found pep.Monomer %C % H %C %H C. [V I

3-Methoxystyrene 80. 56 7. 51 80.19 7. 53 235-240 0.6 94-Methoxystyrene 80.5 6 ,7 .5 1 80.08 7. 42 124-129 0. 414-Methoxy-3-methyl-5-Isopropyl-4-meth-oxy-2-methyl-3.4-Dimethoxysty-4-Phenoxystprene 85. 68 6. 16 85.7 4 6. 17 14C-i45 1.11

styrene 81. 05 8. 14 80.22 8. 02 137-145 1.4 6

styrene 82. 06 9.5 4 82.0t3 9. 38 153-157 0. 15rene 73.17 7 .32 72.31 7. 31 185--188 0. 41

zene; t is flow time for polymer solution; c is conccntra-tion of polymcr in grams per 100 ml. of benzene.Summary

1. Syntheses are described for 3-methoxy-styrene, 4-methoxystyrene, 4-methoxy-3-methyl-styrene, 6-methoxy-3-methylstyrene, 5-isopropyl-4-methoxy-2-methylstyrene, 3,4-dimethoxysty-rene and 4-phenoxystyrene.The formation of substituted styrenes bychloroethylation is discussed.An unusual reduction occurs in the cata-lytic dehydration of 1-(4-phenoxypheny1)-ethanolby means of alumina to give 4-phenoxyphenyl-ethane as a by-product.

2.3.

URBANA, ILLwoIs RECEIVED ECEMBER, 194.5

[CONTRIBUTION FROM THE NOYES CHEMICAL LABORATORY,N I V E R S I T Y O F ILLINOIS]Pyridines. I, Aldehyde-collidine and 5-Ehyl-2 -vinylpyridine

B Y ROBERTL. FRANK,AMES R. BLEGEN,R ~ B E R T. DEARBORN,OBERTL. MYERSAND FRED .WOODA RDThe condensation of acetaldehyde or paralde-hyde with ammonia to form 5-ethyl-2-methyl-pyridine (aldehyde-collidine) has been describedby a number of investigators2 but low yields havebeen generally obtained due to the formation ofhigher-boiling pyridines and large amounts of tar.On the assumption that the tarry by-products aremost likely the result of polymerization of acet-aldehyde with itself or with the substi tuted pyri-

dines formed, it was thought that the use of alarge excess of ammonia might increase the ratioof aldehyde-collidine to higher-boiling pyridinesand tar, and this has now been investigated.Procedure.-Mixtures containing from 0.87 to 8.0 molesof ammonia (1.17 to 10.7 times the theoretical amount)as 28% ammonium hydroxide and 0.044 mole of am -monium acetate per mole of paraldehyde (U . S. p.) wereheated in a steel autoclave to a temperature of about 200',whereupon sti rring was begun and the temperature furtherraised to 250" over a period of an hour. (The reaction isexothermic.) The autoclave was then cooled by meansof its internal cooling coil and the two layers of the reac-tion mixture were separated. To the non-aqueous layerchloroform was added (500 ml. per 24 moles of paralde-hyde), causing separation of water which was combined

with the main aqueous portion. The aqueous layer wasthen extracted with three portions of chloroform (400 ml.per 24 moles of paraldehyde), and the extracts were com-bined with the main organic solution. Fractional distilla-tion through a 30-cm. Fenske-type columna gave nearlypure aldehyde-collidine, b. p. 68-75" (20 mm.). Redis-tillation gave pure material, b. p. 68" (18 mm.), nZDD1.4971. Itspicratemeltedat 164" reportedref. 4,164").(1) This nvestigation wa s carried ou t undcr the sponsorship of th eOffice of Rubber Reserve, Reconstruction Finance Corporation, inconnection with the Government Synthetic Rubber Program.(2) For instance, Durkopf, Ea., O, 444 (1887); Chichibabin,B u l l . SOC. ch im . , [5 ] 4, 1831 (1937); Graf, J . grokt. Chem. , 160, 153(1938).(3) Fenske, Tongberg and Quiggle, I n d . Eng. Chcm. , 46, 1169(4) Auerbach, Ecr. , 16, 3486 (1892).(1934);.

ResultsThe yields of aldehyde-collidine increased from31y0when 1.17 times the theoretical amount ofammonia (1.00 mole of ammonia is equivalentto 1.33 moles of paraldehyde) to about 70% when10.7 times the theoretical amount was used, butit was found convenient to sacrifice some yield toreduce the volume of the reaction mixture;therefore a ratio of 3.7 was adopted. This ratio,

using 24 moles of paraldehyde, gave yields inseveral runs of 52 to 57% of nearly pure aldehyde-collidine.Refractionation of t he fore-run from the dis-tillation gave 45.6 g. of a-picoline, b. p. 127-128', n% 1.4882, identified by its picrate, m. p.183-164' (reported,6 164'), and its derivativewith mercuric chloride and hydrogen chloride,m. p. 151-152' (reported,6 154'). No other pyri-dines were found in the fore-run.Reaction of Aldehyde-collidine with Formalde-hyde.-The condensation of aldehyde-collidinewith formaldehyde or paraformaldehyde t o forma mixture of 5-ethyl-2-(2-hydroxyethyl)pyridineand 5-ethyl-2-vinylpyridine, originally reportedby Pra us& ~,~as then studied.5-Ethyl-2-(2-hydroxyethyl)-pyridine.-Using the sameautoclave as above, a mixture of 2700 g. (22.3 moles) ofaldehyde-collidine, 670 g. (22.1 moles) of paraldehyde,27.0 g. of potassium persulfate, 215 g. of ethanol as solventand 0.25 g. of t-butylcateehol was heated as quickly aspossible (three hours) to 220" and then cooled as above soth at the time of heating above 150" was only one hour,thus avoiding polymerization of the 5-ethyl-2-vinylpyiidinef6rmed during the reaction. The t-butylcatechol alsohelped in this respect.(5) Constam and White, A m . C hcm . J., 9, 38 (1903).(6 ) Ladenburg, A n n . , 247, 6 (1888).(7 ) (a) Prausnitz, Bw.5 33, 2725 (1890) ; (b) Prausnitz. i b i d . , 46 ,Pa04 (1892).

7/28/2019 ja01211a082

2/2

July, 1940 PREPARATION OF SOME ALIPHATICMIDES 1309The contents were rinsed from the bomb with chloro-form and fractionally distilled under reduced pressurethrough a 30-cm. Fenske-type columns using a trace oftrinitrobenzene to inhibit polymerization. Products were1137 g. of aldehyde-collidine, b. p. 73-75" (20 mm.), ~ O D1.4970, 314 g. of 5-ethyl-2-vinylpyridine, b. p. 64-65' (3mm.) , n a o D 1.5383, and 712 g. of 5-ethyl-2-(2-hydroxy-

ethyl)-pyridine, b. p. 103" (1 mm.), 125" (3 mm.), ~ * O D1.5370. Combined yield, accounting for the recoveredaldehyde-collidine, was 73%.The picrate of 5-ethyl-2-(2-hydroxyethyl)-pyridine,crystallized from ethanol, melted at 101-102 '.AnaZ.8 Calcd. for CtaHlsN408: C, 47.37; H, 4.24.Found: C,47.55; H,4.18.The chloroplatinate was prepared by adding aqueousplatinum chloride solution to an alcoholic solution of th;product. Pink crystals were obtained, m. p. 165-167(dec.). Prausnitz7. reported the chloroplatinate to be ared crystalline compound, m. p. 159' (dec.).In t he preparation paraformaldehyde (1 mole per moleof aldehyde-collidine is optimal) gives bet ter results thanformalin. The use of ethanol as a solvent improves theyields, and it is most interesting th at the addition of atrace of potassium persulfate (1% of the aldehyde-colli-

dine) does so materially. Ammonium acetate, zinc chlo-ride, sodium hydroxide, silver and potassium ferricyanideshowed no such catalytic effect.S-Ethyl-2-vinylpyridine.-The conversion of 5-ethyl-2-(2-hydroxyethy1)-pyridine o 5-ethyl-2-vinylpyridine wasachieved in good yield by catalytic dehydration, eitherusing alumina at 40 0420 ' or, preferably, molten potas-sium hydroxide a t 175".A. By Dehydration over Alumina.-A ver tical 19-mm.Pyrex tube heated by an electric furnace for a length oftwelve inches was packed with &mesh activated alumina(Hydralo) and preheated a t 450' to drive off moisture.From a separatory funnel attached directly to th e top of(8 ) hlicroaxialyses were car ried ou t by Mr. Howard Clark.

the tube was passed 170 g. (1.12 moles) of 5-ethyl-2-(2-hydroxyethy1)-pyridine a t a r ate of one drop per second.The tFmperature during this time was'maintained a t 400-420' . An equal volume of chloroform was then added tothe product, the water layer separated, and the organiclayer dried over potassium carbonate. Fractional distilla-tion, after the addition of a trace of trinitrobenzene as apolymerization inhibitor, gave 88g. (6675, or 71%account-ing for recovery of 12.6 g. of st arting mater ial ) of 5-ethyl-2-vinylpyridine, b. p. 95-97' (18 mni. ); ~ * O D 1.5380.B. By Dehydration with Potassium Hydroxide.-Theapparatus and procedure was that described previously forthe preparation of 3m et ho ~y st yr en e. ~ rom 160 g. (1.3zmoles) of 5-ethyl-2-(2-hydroxyethyl)-pyridine t 175and 17-18 mm. pressure was obtained 116.3 g. of crude dis-tillate. Redistillation through a 30-cm. Fenske-typec o l ~ m n , ~sing a small amount (1:1000) of trinitrobenzeneas a polymerization inhibi tor, gave 88.8 g. (63%) of 5-cthyl-2-viriylpyridine, b. p. 95-97' (18 mm. ) ; nzoD 1.5383.The picrate of the product, crystallized from water,melted a t 129.5-130.5'.Anal. Calcd. for CISHIINIO,: C, 49.72; H, 3.90.Foun d: C, 49.73; H, 3.90. 4

SummaryIn the condensation' of ammonia with paralde-hyde t he use of a large excess of ammonia leadsto 50-70y0 yields of aldehyde-collidine.The presence of a trace of potassium persulfateimprove9 the yields of 5-ethyl-2-(2-hydroxy-ethyl)-pyridine in the condensation of aldehyde-collidine with paraformaldehyde.(9) Frank , Adams, Allen, Gander an d Sm i t h , THISO U R N A L , 68,

1365 (1916).URBANA,LLINOIS RECEIVED ARCH, 1946

[ C O N ' I R I R U T I O S FROM 'I'IIE STERLING-M/'INl"ROP RESEARCHNSTITUTE]Studies on the Willgerodt Reaction. 111. The Use of Simple Aliphatic Compounds

B Y JOHN A . KINGND FREEMAN. MCMILLANPrevious communications from these labora-tories have shown that the Willgerodt reactionis by no means restricted to aryl alkyl ketones1and have suggested a mechanism2 for the reac-tion. Although Willgerodt3 reported, in 1888,that if oenanthol were heated "a long time"at 300" with yellow ammonium sulfide the prod-uct was oenanthylamide, no other successful ap-plication of the reaction to an aliphatic com-pound had been published prior to last October.A t that time it was demonstrated by Cavalieri,Pattison and Carmack' that the reaction can becarried out successfully on purely aliphaticketones. Experiments conducted in these labora-tories similarly have shown that the reaction isapplicable to aliphatic ketones, alcohols, thiolsand olefins. The present paper describes some ofour results and indicates how all data thus far

( I ) King and McMillan, THISO U R N A L , 68, 525 (1946).(2) King and McMillan, i b i d . , 68, 632 (1946).(3 ) Willgerodt, Bcr., 11, 534 (1888).(4) Cavalieri, Pattison and Carmack, THISO U R N A L , 67, 1783

(1846).

obtained are in accord with our previously pro-posed reaction mechanism.It seemed desirable to select as a typical ali-phatic ketone one which contained no tertiaryalkyl group, in order that the similarity to anaromatic aliphatic ketone would be lessened.Methyl isobutyl ketone, to which we first appliedthe reaction, gave an 88% yield of isocaproamide.In order to determine whether the reactionwould proceed satisfactorily when neither of thealkyl groups of the ketone contained a branchedchain, the reaction was applied to methyl nonylketone. The product was undecanoamide.Willgerodt,s in his initial paper on the treat-ment of carbonyl compounds with yellow am-monium sulfide, found that a t room temperatureacetone and ammonium polysulfide formed "du-plodithioacetone," of unknown constitution. Itwas easily verified that when acetone and am-monium polysulfide were sealed in a pressuretube the separation of a heavy red oil occurred;( 5 ) Willgerodt, Rr r L 90, 2467 (1887).