7/29/2019 ja01493a046.pdf

1/4

April 20, 1960 AN ABNORMALHAPMANEARRANGEMENT 2035[CONTRIBUTION FROMTHE STERLINQ-WINTHROP RE SEA RCHNSTITUTE]

An Abnormal Chapman RearrangementBY OHN W. SCHULENBERGND S. ARCHER

R E C E I V E DULY 28, 1959Pyrolysis of the imino-ether V derived fr om benzanilimino chloride and methyl o-hydroxyphenylacetate gave, in additionT he structure of this abnormal producto the expected C hapman product, 3-benzani l i mno- 2(3H) - benzof uranone(XVI) .

was deduced from hydrolyses and proved by synthesis.We were interested in developing a generalmethod for the preparation of dihydrodibenzazepi-nones (111),since these could undoubtedly be con-verted into various derivatives of dibenzazepinel(IV) of potential pharmacological interest. Manskeand Ledingham have prepared the oxepinone corre-sponding to I11by ring closureof o-phenoxyphenyl-acetyl chloride; reduction and dehydration thenyielded dibenzoxepine.2 Accordingly, the synthe-sis of the benzoyl derivative I of o-anilinophenyl-acetic acid was undertaken.

Hc=oI=oICsHs CsHsI , R =H I 11 IV11, R CHIThe best route to 11, a benzoyldiphenylamine,appeared to be via Chapman rearrangement of theappropriate imino-ether. This reaction had pro-ceeded normally, usually in high yield, in all casesfound in the li terat~re.~ he required interme-diate V was therefore prepared by the reaction be-tween methyl o-hydroxyphenylacetate and benzan-ilimino chloride. That the product was indeed

the desired imino-ether was proved by its hydro-genolysis to N-phenylbenzylamine and its acid-catalyzed hydrolysis to aniline and VI.CHZCOOCH3 + CG H5C=NCGH 5NaoCHJIO O H c1-HZCOOCH3 H+OC=NCaHsI I/c a s v 6 VI

Heating the imino-ether V at 268 for 45minutesyielded a mixture from which three products wereisolated. Two of these were obtained in yields ofapproximately 25%; the third was formed in morevariable amounts, a maximum of 7% being ob-tained. Reasonable analytical values could not beobtained for this minor product and its structurehas not been determined.

(I ) (a) Australian patent applications 32,722/57 and 32,733/57;(b) P. N. Cr aig, C . A. L eonard, B. M . Lester and D. H. Tedeschi, Ab-stracts of Papers, Second Regional Meeting Delaware Valley, Feh. 5.1958(2) R. H. F. Manske and A. E. Ledingham, THIS OURNAL, 78,4797(1950).(3) (a) A . W . Chapman, J . Chcm. SOC.,147,1992(1925); (b) A. WChapman, ibid., 1743 (1927).(4) P. Moller in M ethoden Der Organische Chemie, Band XI/ I ,Vlerte Autlage, Georg Thieme Verlag, Stuttgart, pp. 910-913.

One of the major products was a colorless sub-stance which, on the basis of analyses and spectra,appeared to be the expected Chapman product 11.This structure was confirmed by hydrolysis withexcess alkali followed by treatment with acid togive N-phenyloxindole (V II) Partial hydrolysisof I1gave the desired N-benzoyl acid I . When thiswas treated with thionyl chloride, then aluminumchloride, ring closure to I11 did not occur. In-stead, the benzoyl group was displaced and N-phenyloxindole (V II) resulted. This result was notunexpected since Astill and Boekelheide had dis-covered that treatment of VI11 with phosphoruspentachloride, then aluminum chloride in benzenegave high yields of N-phenylpyrrolidinone andphenyl $-tolyl sulfone.6 Later, Proctor and Thom-son showed that the acid chloride of VI11 could notbe prepared, treatment with thionyl chloride, phos-phorus oxychloride or phosphorus pentachloridegiving directly the pyrrolidinone and p-toluenesul-fonyl chloride.6 I n contrast, the reaction betweenI and thionyl chloride did give the desired acidchloride, since treatment with methanol regener-ated the ester 11. The oxindole formation there-fore occurred during the reaction with aluminumchloride. Thus, as we had hoped, the additionalbenzene ring in I (as compared with VIII) did de-crease the ease of displacement on nitrogen, al-though not sufficiently to permit formation of theseven-membered ring.

excess11zchloride

The second major product from the pyrolysis ofV was a yellow compound which showed strong ab-sorption at 375mp, this being characteristic of theC,&N= chromophore. The molecular formulaCz1H16N02 suggested that methanol was alsoformed in the pyrolysis, and this was confirmed ina subsequent run. Basic hydrolysis of this ab-normal Chapman product yielded aniline and asomewhat unstable hydroxy-ketone, C14HI 2O2.l-though readily extracted from aqueous sodium hy-droxide with ether, this compound was a weaklyacidic phenol and therefore probably o-substituted.Treatment with acid caused dehydration to a neu-tral substance possessing neither the hydroxyl norketonic functions.It seemed likely that the hydroxy-ketone was thethen unknown 2f-hydroxydesoxybenzoin (IX)(5) B. D.Astill and V. Boekelheide,THI SJ OURNAL,^ 4079(1955).(6) G. R. Proctor and R . H.Thornson, J . Chem.Soc., 2302 (1957).

7/29/2019 ja01493a046.pdf

2/4

2036 JOHN W. SCHULENBERGND S. ARCHER Vol. 82Previous routes which could have given this ma-terial had, instead, led to %phenylbenzofuran(X).7,9 Dilthey and Quint have prepared the ben-zoate X I of I X by cleavage of the pyrylium saltderived from salicylaldehyde and deso~ybenzoin.~Basic hydrolysis of XI , followed by acidification,then gave X. \Ye repeated Dilthey's preparationand hydrolysis of X I and found that extraction ofthe basic hydrolysate gave material identical withour hydroxyketone. Furthermore, our dehydra-tion product was found to be identical with authen-tic 2-phenylbenzofuran (X), while treatment of ourhydroxy-ketone with benzoyl chloride yielded ma-terial identical with Dilthey's benzoate XI . Afterour work had been completed, an independentsynthesis of I X was reported, the compound beingprepared in low yield from isocoumaranone (XII ):md phenylmagnesium br~mide.~

0

0 XIAcid-catalyzed hydrolysis of the abnormal Chap-man product gave, in addition to Z-phenylbenzofu-ran (X), an acid, C15H1003. This was found to beidentical with an authentic sample of 2-phenylben-zofuran-3-carboxylic acid (XV) sent to us by Dr.J . N. Chatterjea.'o Although it was conceivablethat this acid was a primary hydrolysis product, i t

seemed more likely that i t had resulted from acid-catalyzed rearrangement of 3-benzoylisocoumara-none (XI I I ). This compound would be expectedto give readily the unstable acid XIV which couldthen either decarboxylate to I X (followed by ringclosure to X) or cyclize to XV. Chatterjea, how-ever, has prepared XIII'O and. boiled it with acid;%phenylbenzofuran (X) was the only product iso-lated." In several similar cases, however, Chat-terjea was able to effect the rearrangement to ben-0

XI11-- 1 ,

xv XIV(7 ) W. Dilthey and F. Quint, J . prakl. Chem., 131, 1 (1931).(8) P. Y ates, THIS OURNAL, 74 , 537G (1952).(9) A . Spetz, A d a Chem. Scand., 10, 1422 (1956).(10) J . N. Chatterjea, J . Zndion Chem. Sor.. 33 , 175 (1956). Weare grateful to Dr. Chatterjea for sending us samples o several benzo-furan derivatives.(11) J . N. Chatterjea, ibid., 34, 299 (1957).

zofurancarboxylic acids." We therefore repeatedChatterjea's synthesis of XI11 and subjectcd thecompound to the conditions of our acid hydrolysis.Conversion into XV was readily accomplished. 'On the basis of the above hydrolysis products,only two structures, XVI and XVI I (or tautomericforms of these), appeared reasonable for the ab-normal pyrolysis product. Both of these wouldyield 3-benzoylisocoumaranone (XI I I ) and anilineas primary hydrolysis products. However, whilethe formation of XVI I could be rationalized, thefollowing route to XVI appeared much more reason-able. 3 c, 15C=NCaH$/CHCOOCH3

v - fJ -\ + (h. i \HCOOl .H,JqVa O- - ( ' xN ( ' 11 VI)1 ' hHr., 10The imino-ether V, itself a weak base, could beconverted into its anion Va, this then reacting in-tramolecularly to yield the phenolate anion Vb.Attack at the ester carbonyl with displacement ofmethoxide ion would then give XVI.The abnormal Chapman product was then provedto be XVI by direct synthesis. When isocoumara-none (XI I) and benzanilimino chloride were heatedto 1 8 5 O without a catalyst, HC1was given off andXVI obtained. This was identical in all respectswith the pyrolysis product.Finally, since a base-catalyzed mechanism hasbeen suggested for the formation of XVI under py-

rolysis conditions, it seemed that a strong baseshould convert V into XVI at a much lower tem-perature, The imino-ether was therefore heatedin benzene with sodium methoxide; XVI was ob-tained in 56y0yield, thus providing strong supportfor the above mechanism.After our work had been completed, Chatterjeareported the synthesisof XVI from isocoumaranoneand N,N'-diphenylbenzamidine. By mild basichydrolysis he obtained 2-phenylbenzofuran (X) and3-benzoylisocoumaranone (X I I I), while acid hy-drolysis gaveX .14Experimental 4a

Methyl o-Hydroxyphenylacetate and Isocoumaranone(XII).-The former was prepared in four steps from o-methoxybenzaldehyde essentially according to the l iteraturemethod.15 However, our crude o-hydroxypherlylacetic acid

(1 2) For examples of this type of rearrangement in other hetero-cycles seeF. Korte, Angew. Chem., 71, 135 (1959).(13) The mechanism of the normal Chapman rearrangement hasbeen studied by K . B. Wiberg and B. I. Rowland, THIS OURNAL, 77 ,2205 (1955).(14) J . N. Chatterjea, J . Zndion Chem. Soc., 36, 69 (1959).(14a) A nalyses were carried out under the supervision of M r . K . L).Fleischer. Spectra were run under Dr. F. C. Nachod's supervision.Ultraviolet spectra were determined in 95% ethanol, infrared spectrain potassium bromide.(15) J . Levine, T. E. Eble and H . Fischhach, THIS OURNAL, 70 ,1930 (1948).

7/29/2019 ja01493a046.pdf

3/4

April 20, 1960 A N ABNORMAL HAPMANEARRANGEMENT 2037was esterified directly and the product distilled, b.p. 90-94" (0.1mm.).Isocoumaranone was obtained by heating the crude o-hydroxyphenylacetic acid at atmospheric pressure until theevolution of water ceased; the residue was then vacuumdistilled. From 2 moles of o-methoxybenzaldehyde, 152 g.(57%) of isocoumaranone, b.p. 131-134' (22 mm.), was ob-tained. Recrystallization from ether-hexane gave whiteprisms, m.p. 47.5-49.5".Methyl o-hydroxyphenylacetate could also be prepareduia isocoumaranone, 20.1 g. (0.15 mole) of the lactone be-ing refluxed for 5 hours with 200 ml. of methanolic HCl.Solvent removal and recrystallization from ether-pentanegave 20.8 g. (84y0) of white solid, m.p. 71-73'.Benzanilhino Chloride.1-A mixture of 296 g. of benz-anilide and 312 g. of phosphorus pentachloride was slowlyheated. After the initial, vigorous reaction had ended, thedark liquid was refluxed for 2 hours, then distilled, 305 g.(94%) of lioght yellow product, b.p. 115-120" (0.3 mm.),m.p. 38-40 , being obtained.N-Phenylbenzimino o-(Carbomet hoxymet hy1) - phenylEther (V).-The reaction was run with stirring in anitrogenatmosphere. To a cold solution of 27 g. (0.5 mole) of so-dium methoxide in 600 ml. of methanol was added 83 g.(0.5 mole) of methyl o-hydroxyphenylacetate. As soon asthe phenol had dissolved, a solution of 108g. (0.5mole) ofbenzanilimino chloride in 150 ml. of anhydrous ether wasadded quickly and the mixture stirred at room temperatureovernight. After most of the solvent had been removedon the water-pump, the residue was poured into water. Theresulting solid was air-dried, then recrystallized from iso-propyl alcohol to give 128.4 g. (74%) of white prisms, m.p.102.5-104.5'. .4n additional recrystallization from thesame solvent gave the analytical sample, m.p. 103-105".The infrared spectrum had bands at 5.80 and 5.98 p whilethe ultraviolet spectrum showed Am,, 275 mp ( e 5,170) and295mp ( e 3,700).Anal. Calcd. for CnzHlsNOs: C, 76.50; H, 5.55; N,4.06. Found: C,76.93; H, 5.66; N, 3.81.Hydrogenation of the imino-ether was carried out in ab-solute ethanol with a 10% palladium-on-charcoal catalyst.The reduction was stopped shortly after 2 moles of hydrogenwas absorbed. Removal of catalyst and solvent left an oilwhich yielded, with ethanol hydrogen chloride, N-phenyl-benzylamine hydrochloride, m.p. 212-215" (lit. m.p. 214-216').o-Benzoyl oxyphenyl acet i c Acid (VI).-A portion of theimino-ether V was boiled for 10minutes with concentratedaqueous HC1. A gummy solid resulted which was recrys-tallized twice from chloroform-hexane to give fine, whiteneedles, m.p. 141-143' 142'). The infrared spectrumshowed bands at 5.76 and 5.88p.Anal. Calcd. for C16H1204: C, 70.30; H, 4.72. Found:C, 70.05; H, 4.62.Pyrolysisof the Imino-ether V.-The reaction was carriedout in a vessel suspended over a bath of boiling bibenzyl,the temperature of the reaction mixture thus being main-tained at 288". Wi th stirring and a slow stream of nitrogen,25 g. (0.073 mole) of imino-ether was heated for 45 minutes.The resulting very dark oil was diluted with 75 ml. of ab-solute ethanol, cooled to room temperature, and the productfiltered and washed with ethanol. A 29% yield (6.5 g.) ofyellow solid, m.p. 202-210", resulted; this was J -benzanil-imino-Z(3H)-benzofuranone, the abnormal Chapman prod-uct XVI. Recrystallization from chloroform-ethyl acetategave an 88% recovery of material melting at 209-212".Two more recrystallizations, one from absolute ethanol andthe other from ethyl acetate, gave the analytical sample asslightly greenish-yellow needles, m.p. 210-213'. Theultravioiet spectrum had A, 253 mp (e 15,900) and 375mp ( e 21,800). The infrared spectrum showed a singlecarbonyl peak at 5.89 p .Anal. Calcd. for C21H1&02: C, 80.49; H, 4.83; N,4.47; 0, 10.21; mol. wt., 313. Found: C, 80.27; H,5.03; N, 4.50; 0,10.20; mol. wt.,299.Attempts to reduce the abnormal Chapman product withhydrogen and palladium or platinum or with sodium boro-hydride were not successful. An oxime could not be pre-pared even under forcing conditions.

( 16) 0. Wallach, Ann. , 184, 79 (1877).(17) W. Di lthey and W. Hoschen, J . puakf. Chem., 198, 42 ( 1933) .

The normal Chapman product methyl o-(N-benzoylani-1ino)-phenylacetate (11) was obtained by cooling the eth-anolic mother liquors to give 6.4 g. (29%) of tan solid,m.p. 126-131'. Three recrystallizations from absoluteethanol gave the analytical sample as white prisms, m.p.132.5-134.5'. The ultraviolet spectrum had Amax 234nip ( e 15,500) and 271mp e 7,900). Infrared bands werepresent at 5.78 and 6.04 p.Anal. Calcd. for CZZH~~N O~:, 76.50; 11, 5.55; N,4.06. Found: C, 76.14; H, 5.65; hT,.05.When the pyrolysis was carried out at 282" for 20 minutes,much lower yields of both products were olitained, but at250' for 3 hours approximately the sameyields as above re-sulted.I n onerun, the gaseous product was condensed and char-acterized as methanol by i ts reaction with p-nitrobenzoylchloride, the resulting ester being compared by m.p. andmixed m.p. with an authentic sample.The separation of products was complicated by a thirdmaterial which was always present in the crude abnormalChapman product. In one run, this minor product couldbe partially separated by hand, since it crystallized as red-brown prisms while the major component came down as agreenish powder. However, it was best obtained by basichydrolysis of the mixture, the abnormal Chapman productbeing completely destroyed (see below) while the minorproduct remained unchanged. I n this way, i t was deter-mined that the amount of this material varied from less than10 to about 25y0 of the initial ethanol-insoluble material.

Three recrystallizations from ethyl acetate gave fine, yellow,needles, m.p. 223.5-225.0'. Strong ultraviolet absorptionoccurred at 243 and at 374 mp, while the infrared spectrumshowed bands at 5.84 and 5.75 p (both weak), and at 6.07and 6.12 p. No hydroxyl band was present.Anal. Found: C, 82.32, 82.37, 82.41; H, 5.21, 5.06,5.02; N, 6.98, 6.86; 0, 5.55; mol. wt., 345.In contrast to XVI, this material gave a positive test withferric chloride in chloroform and this test was used to deter-mine the presence of this minor product in XVI. Acidhydroysis gave aniline, but the other products were not de-termined. I t is possible that this product is actually a mix-ture; however, neither recrystallizations nor chromatog-raphy on alumina affected either the m.p. or the analyticalvalues.N-Phenyloxindole (V II) -Complete hydrolysis of thenormal Chapman product I1was accomplished by refluxinga solution of 500 mg. of I1and3g. of potassium hydroxidein 7 ml. of water and 20 ml. of methanol for 3hours. The

solution was then diluted with water and acidified withhydrochloric acid followed by brief heating on the steam-bath to cyclize the resulting o-anilinophenylacetic acid.The mixture w a s then made alkaline with excess sodiumhydroxide and cooled to give leaflets of VI I . Recrystalli-zation from aqueous ethanol gave material melting at 115-118". This did not depress the m.p. of authentic N-phenyloxindole18o-( N- Benzoyl ani 1i no)- phenyl aceti cAcid (I )-A solutionof 6.9 g. (0.02 mole) of I1and 1.08 g. (0.02 mole) of sodiummethoxide in 50 ml. of methanol and 20 ml. of water wasrefluxed for 2 hours. After cooling, the solution was di-luted with water, washed with ether and acidified. A gumresulted which was induced to crystallize, then recrystal-lized from ethyl acetate-hexane to give 3.8 g. (58%) ofwhite acid, m.p. 154.5-156.5'. A second recrystallizationfrom the same solvent pair gave the analytical sample, m.p.155.5-157'. The ultraviolet spectrum showed A 236mp (e 14,500) and 271mp ( e 7,650). The product absorbedin the infrared at 5.75, 6.05 and 6.15 p.Anal. Calcd. for C21H17N03: C, 76.12; H, 5.17; N,4.23; neut. equiv., 331. Found: C, 76.13; H, 5.22; N,4.22; neut. equiv., 334.Attempted Cyclization to 111.-To a solution of 1.66 g.(0.005 mole) of I in 5 ml. of chloroform was added 5 ml. ofthionyl chloride. After 3 hours of refluxing, the solventand excess reagent were removed and the red, gummy, resi-due dissolved in 10ml. of nitrobenzene. This solution wascooled and added in portions to a cold mixture of 2 g. ofaluminum chloride and5ml. of nitrobenzene. After stand-ing at room temperature, the mixture was decomposed withice, then hydrochloric acid, and the organic material was( 18) R. StollC, Chem. Bcr., 47, 2120 (1914) .

7/29/2019 ja01493a046.pdf

4/4

2038 JOHN W. SCHULENBERGND S.ARCHER Vol. 82extracted into chloroform. The extracts were washed withaqueous hydrochloric acid, sodium bicarbonate and water,then dried and the solvent removed to leave a red gum whichcould not he crystallized. This was therefore chromato-graphed on silica ge to give, by ether elution, N-phenyl-oxindole (VI I ) as the only crystalline product. This mate-rial had m.p. 118-121.5' and did not depress the m.p. ofauthentic V I1 preparedviathe Stollb procedure.I8Re-esterification of I to 11-A solution of 662 mg. of I , 5ml. of chloroform and 3 ml. of thionyl chloride was refluxedfor 30 minutes. The resulting gummy acid chloride wasthen boiled with 15ml. of methanol for 30 minutes. Dilu-tion with water then gave 450 mg. (65%) of almost whitesolid, m.p. 129-132". This material strongly depressedthe m.p. of N-phenyloxindole but did not depress that of thenormal Chapman product 11.Basic Hydrolysisof XVI; 2'-Hydroxydesoxybenzoin (IX).--A mixture of 3 g. of crude XVI, 4 g. of potassium hydrox-ide, 15 ml. of water and 45 ml. of methanol was refluxed forone hour, most of the solid soon dissolving. The majorabnormal Chapman product XVI was thus hydrolyzedleaving the minor Chapman product untouched. This wasfi ltered froin the cooled solution and the filtrate extractedseveral times with ether. The ether solution was then ex-tracted with aqueous hydrochloric acid to yield the basicfraction from which aniline was obtained and characterizedas its hydrochloride, m.p. 197-199'. The non-basic frac-tion was dried and the ether removed to leave a solid whichcould be recrystallized from ethyl acetate-hexane. Almostwhite hydroxy-ketone I X resulted. Due to the instabil ityof this compound, the m.p. was sometimes raised, sometimeslowered on recrystallization. The highest m.p. we obtainedwas 127.5-131.5'. The highest yield of material meltingover 120" was about 50%. Higher yields could be obtainedby acidifying the reaction mixture and immediately extract-ing with ether, but the product was less pure than that ob-tained directly from the basic solution. When the acidifiedsolution was allowed to stand overnight, only 2-phenylbenzo-furan (X ) was obtained. The product was soluble in 10%aqueous sodium hydroxide, but not in carbonate solution.I t gave a positive ferric chloride test. The ultraviolet spec-trum exhibited two maxima, at 242 mp ( e 13,000) and at275mp ( e 3,500). These data are in good agreement withthose of Spetz, who has prepared IX via alternate route^^The infrared spectrum had bands at 2.95 and 5.97p .Anal. Calcd. for C14H120n: C, 79.22; H, 5.70; 0,15.08. Found: C, 79.14; H, 5.93; 0, 15.25.A 2, 4- di ni t rophenyl hydrazonewas prepared in methanoland recrystallized twice from absolute ethanol to yieldthick, orange-red needles, m.p. 197-198.5'.Anal. Calcd. for C~oH l~N~O~:, 61.22; H, 4.11; N,14.28. Found: C, 61.11; H, 4.19; N, 14.15.The hydroxydesoxybenzoin I X was converted into itsmethyl ether by refluxing one hour with methyl iodide andsodium methoxide in ethanol. After phenolic materialhad been removed with sodium hydroxide, the gummyether was converted into its 2, 4- di ni t r ophenyl hydr azoneand this recrystallized twice from absolute ethanol togive adark orange solid, m.p. 177.5-179'.Anal. Calcd. for Ct1H18N40b: C, 62.06; H, 4.46.Found: C, 61.70; H, 4.62.Authentic hydroxy-ketone IX was synthesized by re-fluxing a solution of 1g. of its benzoate (X I),' 1g. of po-tassium hydroxide, 3 ml. of water and 10 ml. methanol for3hours. The solution was cooled, diluted with water andextracted with ether. The extracts were washed withwater, dried and the solvent removed to leave a gummysolid which was recrystallized from ethyl acetate-hexane.Tan prisms, m.p. 122-127', resulted. This productstrongly depressed them.p. of 2-phenylbenzofuran, but didnot depress that of hydroxy-ketone obtained by basic hy-drolysis of the abnormal Chapman product. I ts ultravioletand infrared spectra were virtually identical with those ofthe hydrolysis product, confirming their identity.2' - Benzoyl oxydesoxybenzoi n (XI).-To a mixture of 212mg. of 2'-hydroxydesoxybenzoin (I X ) (from basic hydroly-sis of XVI) and 5 ml. of 10% aqueous sodium hydroxidewas added 0.15 ml. of benzoyl chloride. Shaking soonproduced a solid which was recrystall ized from absoluteethanol. This product, which had m.p. 103-105", wasfound to be identical, by mixed m.p., ultraviolet and infra-red spectra, with authentic XI prepared via Dilthey's

route.' The ultraviolet spectrum had Xmax 230 mp ( e24,800), 275mp ( E 3,300) and 281mp (e 2,700).2-Phenylbenzofuran (X).-A few drops of hydrochloricacid was added to a solution of 100 mg. of I X (from hy-drolysis of XVI) and 5 ml. of acetic acid. The solutionwas heated 10minutesonthe steam-bath, then diluted withwater to yield a white solid. This was recryspllized from95% ethanol to give product of m.p. 119.5-121 . This wasidentical, by mixed m.p. determination, with authentic Xprepared by hydrolyzing Dilthey's benzoate and allowingthe acidified solution to sit overnight at room temperat~re.~The ultraviolet spectrum of our material exhibited themaxima reported by Yates for 2-phenylbenzof~ran.~Acid Hydrolysis of the Abnormal Chapman Product XVI.-One gram of XVI was refluxed for 5hours with a solutionof 25 ml. of 48% hydrobromic acid and 25 ml. of ethanol.That the gas evolved was carbon dioxide was proved by itsconversion into barium carbonate. Almost all the originalsolid dissolved during the rdux period, at the end of whichthe remaining insoluble material was filtered off and thefil trate cooled to yield a white solid. This was recrystal-lized three times from heptane to give a product meltingat 194.5-197'. A mixed m.p. determination with authentic2-phenylbenzofuran-3-carboxylic acidlo (X V) showed thatthe two materials were identical. Their infrared spectrawere identical, both absorbing at 6.02p with a shoulder at5.95 p. The ultraviolet spectrum had Xmnx 222 mp ( 624,000) and 303mp E 25,000).In a second run, XVI was hydrolyzed by refluxing 2.5hours with a solution of 25 ml. of concentrated hydro-chloric acid and 25 ml. of ethanol. The resulting mixturewas diluted with water, extracted with ether and the abovecarboxylic acid XV extracted into aqueous sodium bicar-bonate. 2-Phenylbenzofuran (X ), identical with that ob-tainedoiabasic hydrolysis of XVI and subsequent acidifica-tion and via the benzoate X I was then obtained from theethereal solution.Rearrangement of XI11 to XV.-3-Benzoyl-P(3H)-benzo-furanone (X I I I ) was prepared by Chatterjea's rnethod.'OThe ultraviolet spectrum had Xma, 232 mp ( e 11,600), 252mp ( e 13,200), and 339 mp ( e 11,000). The infrared spec-trum showed bands at 5.92 and at 6.16 p. When 1g. ofXI11 was refluxed for 1hour with 35 ml. of ethanol and 25ml. of concentrated hydrochloric acid a cloudy mixture re-sulted from which a solid crystallized on cooling. One re-crystallization from heptane gave XV, m . ~. 93-195",identical by mixed m.p. and infrared analysls with the acidobtained by hydrolysis of XVI and with the authentic 2-phenylbenzofuran-3-carboxylic acid of Chatterjea.'O

3-Benzanilimino-2(3H)-benzofuranone (XVI) -A mix-ture of 20.1 g. (0.15 mole) of isocoumaranone (X I I ) and32.4 g. (0.15 mole) of benzanilimino chloride was heatedslowly with stirring under nitrogen. At 185', the evolu-tion of hydrogen chloride began. The mixture was slowlyheated to 240", then cooled toasolid black mass. This wasdissolved in boiling chloroform and the solution diluted withabsolute ethanol to give 28.0 g . (60%) of crude, brownXVI, m.p. 204-208'. Several recrystallizations fromethanol, one with charcoal, gave a poor recovery of orange-yellow XVI, m.p. 209-212'. This was identical, by ultra-violet and infrared spectra and by mixed m.p., with theabnormal Chapman product. The analytical figures weresatisfactory.In similar runs with added potassium carbonate, noneofthe desired product was obtained, presumably due to self-condensation of the isocoumaranone.Our alternate preparation of XVI was carried out by dis-solving 3.45 g. (0.01mole) of the imino-ether V in 50 ml. ofdry benzene, flushing the system with nitrogen and adding,with stirring, 1g. of sodium methoxide. The mixture wasthen refluxed, with stirring and nitrogen, for 1.5 hours,during which period solvent was continually distilled off andreplaced from time to time with fresh, dry, benzene. Themixture was then cooled, diluted with water, acidified andthe aqueous layer extracted with benzene. The yellowbenzene solution was dried and the solvent removed. Onerecrystallization from absolute ethanol gave 1.75 g. (56%)of bright yellow solid, m.p. 211-212'. This material wasidentical, by mixed m.p. and ultraviolet spectra, with XVIprepared from isocoumaranone (above) and with that pre-pared by pyrolysis of the imino-ether.RENSSELAER,7, .

Found: C, 80.22; H, 4.75; N, 4.43.