J. BISHOP T I S C L E .\XI) E K S G S T E. CORSLINE.
I874
high current, hut ma\-he made with a low current, if formic acid is the elect rol \- te. This study is, of course, o n l y a fragnient of \vhat remains to be done. ~ ~ N I V E R S I TO YF P E N X S T L V A S I A , PHILADELPIIIA, P A .
INVESTIGATION OF THE CLAISEN CONDENSATION. m. FURTHER CONTRIBUTIONS TOWARDS THE ELUCIDATION OF THE MECHANISM OF THE REACTION.’ I ~ YJ . BISHOP T I N G L EA N I )
EKXEST E. GORSLINE.
Keceired October 7 , 1908.
The Yarious current theories regarding the inechariisni of the ethyl acetoacetate condensation w r e discussed i t i our second paper. I t is, therefore, only necessar!. i t i the present coiiiiiiuriicatic,ri to state that, broadly, the question is nhetlier the sodium. or other condcnsing agent, acts directly on the ethyl w e t a t e , or whether it reacts lirst with a trace of alcohol, the actual condensation being brought about b!- the sodium ethylate so produced. I t is e\.iderit that valuable evidence on the questioii at issue might be obtained l)!, at1 esaiiiinatioti of the behavior of ethyl acetate free from alcohol. ;lccordingl!-, o u r early eirorts were directed towards the purification of this ester. The product which we evetituall!r obtained was probably the purest specimeii of this compound hitherto prepared. I t reacted with sodiuni, to form ethyl acetoacetate, as reulily as the ordinary “pure” material. We think that this elfectively disposes of the idea that the action is caused by the presence of a trace of alcohol in the ester, unless, of course, ~ to some the statenierit be made t h a t ethyl acetate is c i l x t r ~ , .dissociated infinitesimal extent into alcohol, or into some hypothetical ’ ‘ethylitiene,” or “active variety of ethylene,” whatever these m a y he. Statements of this nature, though tlie). are made not infrequentl!., iire outside discussion, because they assume the \.cry point at issue. RIuch work has beeti done b y A. Michael, Nef and others on the purilication of ethyl acetate ant1 also 011 the qtiaiitity of hydrogen evolved from sodium and eth!.l acetate. This quantity is not equivalent to t h a t of the s certain of the reacting sodium. Ai portion of the gas a l n ~ ~ yattacks other substances which are present. \I’e ha\-e carried out a very large number of experiments with the object of eliminating as far ;is possible the efrect of this absorbed gas. \Ye added to the reacting materials a iiumher of easily reducible s u b stances in the hope that the hydrogen would attack thein in preference to reacting with the ethyl -acetate derivatives. Our results were not very encouraging. Our previous papers of this series appeared in the A m . Chem. I., 37, 483 (1907) ; 4% 46 (1908)-
INVESTIGATION OF CLAISEN CONDENSATION.
1875
I n continuation of our work on the action of sodium on esters,' we have confirmed our observation that ethyl malonate reacts with two atomic proportions of sodium, hydrogen being evolved only during the earlier part of the experiment, when the first half of the metal is reacting. Ethyl dimethyl malonate and ethyl chlormalonate react with four and one atomic proportions of sodium, respectively. These results appear to prove that diethll malonate has a different structure from the other two esters. As mentioned in our earlier paper, its formula is probably, C,H,OCOCH : C(OH)OC,H,. The monosodium derivative is, doubtless, C,H,OCOCH : C(ONa)OC,H,, and, from analogy with ethyl benzoate, it is most likely t h a t the second atom of metal attacks the ethoxy group giving C,H,OCOCH : C(ONa)C(ONa) : CI1CO,C2H,, although, of course, the carbethoxy group may be involved, and the two complexes which result may be linked together by the CO groups. We hope to offer some evidence on this point later. Ethyl dimethylnialonate behaves, mutatzs wmiantlzr, like ethyl benzoate, and, therefore, the sodium derivative CONa must be formulated as Me$ . The possibility t h a t it may (JoNa have the double formula is sufficiently obvious. The original ester agrees in behavior with the current formula, (CH,),C(CO,C?H),. Ethyl chlormalonate acts as though a limited portion of i t were in the hydroxy form ; more work will be required before a satisfactory explanation of its behavior can be given. Since the publication of our earlier paper A. Wahl has communicated some very interesting results obtained b y the study of the action of sodium on a variety of esters.2 I n the light of our knowledge t h a t ether and tertiary bases act catalytically in promoting the Claisen condensation we have investigated the more important reactions which Claisen studied. We made parallel experiments, in the one series excluding ether, whereas, in the other i t was added to the reacting substances. I n each case it was found that the compounds react with such great rapidity t h a t i t was not possible to observe any acceleration caused by the ether. It is, doubtless, for this reason t h a t Chisen himself failed to observe the catalyzing influence of this substance. Included in the same section we give a brief description of some additional work on the condensation of camphor with certain esters. The results demonstrate t h a t the method of work and the conditions employed may require considerable variation for different esters. This is particularly true, for example, with ethyl benzoate and ethyl phthalate. Am. C k m . J., 40, 46 (1908).
Compt. rend., 147,72.
1876
J. BISHOP TIXCLE: A S D ERSEST E. GORSLINE.
U'e also describe t he preparation of a condensation conipound from benzophenone and ethyl oxalate. It vas obtained as a n oil and has not yet been studied. I t gives. in alcoholic solution, an intense coloration with an aqueous solution of ferric cliloritlc. \Ye h i r e to resen-e the study of this interesting coinpound froni 1)eiizoI)lieiiorie. After the conclusion of our work we noticed the publication of ;L pai)er l)!. \\-. 1)iecknianri :iiid A . Kron' in ivhich the!, also shon.cd that tertiary krtones. R,CF-ICO, niay undergo the Claiseii co~idensatioii. 'l'liis fact is entirely incompatible with the Chisen-Nef theory of the reactioii. whereas it inight well be anticipated from A. Michael's explanation of tlic nieclianisni of the coildensation. A t the end of the present paper \ve give a brief record of a feu- esperirnents 011 the interaction of sodium and berizophenorie.* The product consists of a deep blue po\vder which yields a mixture of benzhytirol and beiizpinacone when treated Ivith ivater. .l colorless crystallinc deri\-ative is obtained 1)y the action of iiietaiiitro1,enzoyl chloride c)n tlie blue powder. 11-e made iiuinerous esperinieiits, under var\,ing conditions, with the object of condensing acetaldehyde with ethyl oxalate, ethyl formate and camphor, respectively. In 110 case were ive successful because the sodium, or more probably some sodium compound, caused the aldehyde to polynierize with ver!- great rapidity. 11-e describe one of the experiments only, because it lccl to the production of a rather curious substance.
Experimental. I. EXPERIMENTS W T I ~ G m u r , X c r ~ ~ a ~ ~ . - - I ~ a h l l ) abest i i ~ i i ethyl ~s acetate, which was used in all our cxperinients. was dried over calcium chloride. A little sodium wire vas now added, the liquid slinken for a few moments, theri filtered quickly and redistilled. The product \vas divided into two portions, each of \vliicli was treated with sodium ( I atom : 3 mol. of ester) in the usual manner, but to one portion 2 5 drops of absolute ether were added. The reaction was complete after two hours at the room temperature. The !.ield of ethJ.1 acetoacetate appeared t o he distinctly larger from the experiment in which ether \vas used. I n a . W C ~ J Hx~r~i c s of experinients the ethyl acetate \vas shaken \\it11 fused sodiuni acetate ant1 then distilled from it. Tlle distillate was shaken first Jvith calcium chloride solution, theii wit11 i w t e r , I t was now dried over calcium chloride and distillctl. I t 1)oilcd a t 7;' and failed to react with litnitis paper. 7'112'rrl.i'c~ic~--'I'he product from the second series of experiments \vas allo\recl to remain o\.er pliosphoric aiihydritle during 2 4 hours. then hoiletl with it for two hours. and finall!. distilled ofi. I t boiled at 7;' and hat1 BcY.,41, 1260 (1903). Am. Chenz. ,I., 29, 604 (1903)
* Acree:
INVESTIGATION OF CLAISEN CONDENSATION.
1877
an acid reaction; the anhydride became dark. The distillate was neutralized with a dilute aqueous solution of sodium carbonate, washed with water, dried over calcium chloride and filtered. Fozwth .Series.-The product from the third series of experiments was boiled in contact with phosphoric anhydride for three hours and then distilled. It boiled at 77' and was neutral to litnius. The ester from the second, third and fourth series, TI hen treated with sodium, behaves like that obtained b y the first method of purification. Fiji12 Series.---The ester as obtained from Kahlbaum was dried over calcium chloride, distilled, and then boiled during tlirec working tinys with calcium turnings, which remained entirely untarnished. The ester was then distilled. Sirrtlz Series.---Il:ster from the fourth series was treated as in the fifth series. The ester from the fifth and sixth series reacted quickly with sodium, the experiments and results being similar to those described in connection with the first series. I n order to show how sensitive calcium is to alcohol, some of the metallic turnings were covered with absolute alcohol (Kahlhaum) in a flask protected from moisture. Reaction soon commenced without heating and after 3 days all the metal had disappeared. In order to see if the state of division of the sodium had any material influence on the results, some of the ester from the fifth series was treated with sodium chips instead of wire. The action was slower, but i t was complete after three hours at the room temperature and three hours' boiling. Duplicates were made of all the condensation experiments described above, except t h a t the ratio of ester to sodium was z mol. : I atom. The results were entirely similar. 11. SYNTHESIS OF ETHYL ACETOACETL4TE K'SIN(; E T H E R 4s s0LVEYT.Ethyl acetate ( I 7.6 grams, 2 mols.), sodium (2.3 grams, I atom) and absolute ether (2j cc.), when mixed, reacted vigorously. After being boiled for five hours the reaction was complete. The product was a dark brown liquid, b u t after standing over-night it became a red solid. It was acidified with dilute acetic acid ( I : I ) , poured into saturated brine and extracted four times with ether. After being dried, the ether was distilled, the temperature being raised to 9 j '. The residue boiled a t 100' ( I I O mm.) and at 175-17;' under the ordinary pressure. It gave an intense coloration with aqueous solution of ferric chloride and is, doubtless, ethyl acetoacetate. A large number of experiments were made with the object of causing the hydrogen, which is produced during the synthesis of ethyl acetoacetate, to react with some readily reducible substance that was added to the ethyl acetate and sodium. The results of this work were not very decisive. The experiments described above show, we think, t h a t ethyl acetate, entirely free from alcohol, reacts with sodium as readily as do those
1878
J. BISIIOI' TIXGLIl .\SI) T:IINI o I the reacticxi is I I C J rv duced hy the presence of etlicr. I1 I . I < i ; . ~ c ~I~ii:~SC o JI)II.LI s \v1'171 CI:I~I%.\IX 1 ~ ~ , 1 ~ 1 ~ 1 ~ ~ester , ~ . - ' 1io. ' 1 Ii egrarri mol.) \vas tnisetl irith "liesane" ( 2 0 o cc'., 1). 11. 00') a n d heated with sodium wire. Divt/i,yl i u ( ~ / o u u k , ,\vhich \viis dissolved in pctroleuni etlier ( ' 0 0 cc., b. p. 7 0 ' ) instead of hexane, reacted quickly with the metal, hydrogen being evolved briskly at first. 'I'he mixture \vas boiled for a1)otit ,;o hours. .I white powder formed which gradually turned !-ello\v. 'I'\Yogram atomic proportions of sodium reacted. iirith ( / Z d h j d ciiloi.iii~rion~it(. the evolution of hydrogen \vas slight. .\fter five hours a t the room temperature tlic liquid began to turn !,ellow, mid after 48 hours most of the wire had disappeared. After 7 0 Iiours' boiling only one atomic proportion oE sodiuni hat! reacted. I ) i & j i tii~ric'tli~iliiicilc,llc2tc , under similar conditions, reacted ivith oiie atomic proportion of sodium after two hours a t the room temperaturc, the liquid becoming dark red. 1;iI.c hours later ;I second atomic proportion of sodium had dissolvetl, a n d ;I tliirtl oiie alter ten hours' boiling. IYlieii the boiling \vas coiitinued for about ;o hours longcr the greater part of a fourth proportion of sodium reacted. 'l'lie product consisted of a dark red solid which surrounded the metal a n d greatly hiiitlercd its contact with the remaining liquid. 'I'hcrc was n u aI)preciable eyolutioii of ]I!.drogen during this espcrimenr. 'Hie hearing of these results 011 the constitution of the respective esters is discussed i l l the theoretical part o f this paper. 11.. C..w.u,Ync L:I:FI~CI' 01: ~ ~ T I I I ; I ZO S n i i i C I , . ~ c ~ l s I ~ I ~ s s . \ ~ r- .I o s . 1he following pairs of substances were coiidensed, in Iietroleum ether (b. 1). 70') solution, by means of sorliiiiii \\.ire: ethyl benzoate and acetone : ethyl oxalate and acetophenone : ethyl lmizoate and acetophenone. Ihplicate experiments irere made i n which petroleurri ether containing j per cent. of absolute ether was einployed. No differences could be 01)served either in the behavior of the materials or in the results of the eslwrinients. I n all cases the reactions proceeded very yigoroLisl>.. I S . - -111 our earlier Lvork \ r e V. A\I'Iv.IC.~l'I:.II,.\TI< WITH C,iRwr,.--Se\-eral years ago Bishop 'l'ingle and ( )'Hyrnc described some preliminary esperi metits' on the condensation of eth!.l oxalate with various phenols, i n presence of sodium. They stated t h a t carvol gave a resinous compound which had an odor of violets. *It the end of last year Air. 1:. 11. Slocuin informed the senior author t h a t he had repeated this work several times and had been unable to obtain the odoriferous substance. \\'e therefore carried out a number of experiments with carvol and ethyl oxalate, b u t we also failed to obtain any substance with the property in cluestioii. \Ve employed carvol froni Kahlbaum ant1 distilled it hefore use. Alr. Slocum's carvol was also highly purified. I t appears certain, therefore, t h a t the fragrant smelling compound is not,formed froni carvol. Bishop Tingle and O'Ryrne's results were worked o u t with great care, under the strict supervision of the senior author, ioho has a distinct recollrc tion of the powerful and unexpected odor of the substance in question. They used Kahlbaum's carvol as received, without further purification. I n view of the discrepancy mentioned above ant1 of tlie fact that, in tlic ' Am. ( ' h p t n . ,I., 2 5 , i o r ( 1 ~ 1 ) . ~
INVESTIGATIOK O F CLAISEN CONDENSATION.
1881
paper cited, the description of the experiments is very brief, it may be well to give fuller details taken from the laboratory note-book. ‘Carvol, ( I S grams) was treated with ethyl oxalate ( 1 1 . 2 grams), light petroleum ( 2 5 0 cc.) and sodium wire ( 2 . 3 grams). After the metal had dissolved the product was allowed to remain over-night and was then poured into water and the light petroleum solution remoxred. The aqueous portion of liquid was estracted with ether ( A ) . The aqueous residue was then acidified and again extracted with ether (E). A and B were distilled and the residues, after being examined, were combined with that from the petroleum. This latter solution was also distilled, the temperature rising to 2 3 5 O , up to uhich point two-thirds passed over. The residues were treated a t the ordinary temperature with aqueous-alcoholic sodium hydroxide, the liquid was acidified, made alkaline with solution of sodium The residue obhydrogen carbonate and extracted twice with ether. tained after distilling this ethereal solution gave a deep red coloration with an alcoholic solution of ferric chloride. I t was amorphous and had a strong fragrant odor of violets.’
Summary. Ethyl acetate, highly purified by means of phosphoric anhydride and calcium, reacts with sodium as readily as that purified in the ordinary manner. Consequently the effect on the sodium cannot be due to traces of alcohol in the ester, as Nef and his students have stated. 2 . Contrary t o various statements which have been made, ethyl acetoacetate is formed without difficulty from sodium and ethyl acetate, in presence of ether. 3. Because of their varying behavior with sodium, diethyl malonate and diethyl diniethylnialonate appear to have different constitutions. Diethyl chlornialonate reacts as though it were intermediate between the other two. 4. The catalytic effect of ether and of tertiary bases in promoting the Claisen condensation is only apparent if the velocity of the reaction is relatively small. 5. The Claisen reaction has been studied under various conditions and a condensation compound has been obtained from benzophenone and ethyl oxalate. The fact that benzophenone, which contains no CH,COor RCH,CO-, reacts in this manner shows the entire error of the ClaisenNef hypothesis regarding the mechanism of the condensation, 6. -4ttempts to condense acetaldehyde with various esters were unsuccessful. The aldehyde foi med complex polymerization or condensation products under the influence of the sodium. 7 . Benzophenone combines with two atomic proportions of sodium. After treatment of the sodium derivative with water, the product consists chiefly of benzhydrol and benzpinacone. I.
I 882
J. BISHOP TINGLE AX13 11. I:.
ROLKER.
The experimental work dcscrilml in this paper was carried out by the junior author at the Uiiivcl-sit? of Rochester, N.J-,, a i i t l a t tlic Johns I-Topkins Universitx, duriiig tlie sessioii I (p); s. MCMASTEKU N I V I S R S I T Y ,TORONTO, CANAI)A, September, xgoS. -.
-. . . .
..
__
INTRAMOLECULAR REARRANGEMENT OF PHTHALAMIDIC ACIDS. 111. B Y J . BISHOP T I N G L E A N D H. F. ROLKER
Received October 7 , ~ y o h .
The previous papers' on this subject described tlie forniation of phthal. . phenylimide,
F0)NC&
C,HZ
from
phthalphenylamidic (phthal-
co
anilic) acid, a t a temperature about 100' below its melting point, in presence of aniline. The fact that tertiary bases behave like aniline, led to the suggestion being made that the dehyclration is due to salt formation. It was also shown t h a t certain other phthalamidic acids behave in a similar manner, b u t not in an identical one. I11 the 1)resent coiiimuiiicatioti we give an account of further work ivhich \ve have done on this subject during the past year. We have employed the following acids : phthalphenyla~~iidic, meta- and paratolylamidic, CBH,NHCOC,H,CO,H ; phthalortho-, CI-I,C,H,NHCOC,H,CO,H; phthalmeta- nrid paranitroplienylamidic, O,NC,H,NI~COC,H,CO,~-I; and phthal-ci- and -P-naplithxlaniidic, C,,H,NHCOC,,H,CO,H. The two nitro acids and the iiieta- and paratolyl acids have not been obtained previously. --is we required relatively large quantities of these acids we have worked out the details of their preparation and purification with care. ( ~ I I ' method is described later iii this paper (1). 1%~). I t is very rnuch simpler and more expeditious than the older process. but does not differ from i t in general principles. In the experiments recorded in the earlier papers the acid arid base were generally heated 2 l / , hours, or longer. I n this later work we L I S L I ~ ~ ! limited the time to 35 minutes because we wished to avoid the complete transformation of the acid into the imide. I n this n-ay \ye liave heeii able to isolate certain intermediate products and have also succcctled, in some cases, in determining tlie relative velocity with whicli the various imides are formed. LiTe have investigated the effect of the following amines on the acids mentioned above: Pyridine, quinoline, ethylaniline, diphenylamine, aniline, u- and ,3-iiaphth\.laiiiiiie aiitl, in most cases, of benzylamine. I n Bishop Tingle and Cram, A m . Chem. Lovelace, Ibid., 38, 642 (1907).
,I,,
37, 596 (1307); Bishop Tingle and
