heptane over the range of coverage investigated tion of toluene is very much greater than the and the molar integral entropy of the adsorbed n- calculated value for the gaseous model. I t is heptane decreases steadily with increasing coverage probable that the toluene is not freely mobile but it in a normal manner. The results for benzene are is not possible to suggest here what fraction of the similar. The energy of adsorption of toluene is loss of 51.3 e.u. is attributable to restriction of constant up to a coverage of 0.018 X 10l6 niole- translation and what fraction to restriction of rotacules/cni.2. I t then falls by some A kcal./niole as tion (of the benzene ring and of the methyl group). the coverage is increased to 0.030 X lo1&moleIn this discussion we have ignored the small gain cules fcni.2,and there 1s a correspoiidmg rise in the in entropy associated with the new vibrational molar entropy of the adsorbed material. Keinball degrees of freedom which must replace the lost and Kidealj suggested that a t high coverages the degrees of translational and rotational freedom . :idsorbed molecules were undergoing a change in SAV.4L LlEDICAL RESEARCH INSTITUTE orientation from “flat” to “vertical” adsorption BETHESDA,MARYLAND, AND and. this was confirmed by the discovery of a DEPARTMENT OF PHYSICAL CHEMISTRY UNIVERSITY m a l l change in surface potential.lo We believe CAMBRIDGE ENGLAND that the fall in the energy and rise in the entropy CAMBRIDGE, .____ of adsorption is consequent on this change of orientation. I t is not likely to be due to the formation of a second layer because the area per molecule a t Fischer Indole Syntheses with Polyphosphoric Acid whde the the ccmmiencenicnt oi the change is 3(i B Y H. If. KISSMAY, I ) . 1 % ’ . FARNSWORTH AND B. WITKOP area of a toluene tiio~eculejis only about :;T -4.-1lthough the range I? in Figs. S arid 4 is small, we RECEIVED MARCH10, 1952 can say something about the limiting values of The successful application of polyphosphoric acid ES and ss as I‘ -+ 0 and r + (assuming that in cyclodehydration reactions’J and Beckmann reI‘ actually does approach infinity as x 1, as one expects below the critical temperature of the arrangementsa prompted us to try i t in the Fischer indole synthesis especially in connection with those adsorbate on a non-porous adsorbent). As r phenylhydrazones which are known to require a ; as 0, EL - ES -j finite and ss - si, -7 drastic conditions for the elimination of ammonia r-+m,Et-Es--OOndSsSt--+(). The three substances show an interesting varia- and for indole ring c10sure.~ I t was found that the use of polyphosphoric acid tion in freedom in the adsorbed state. Taking the vapor a t one atmosphere pressure as the reference gave 2-phenylindole in good yields not only from state, the experimental entropies of adsorption’’ acetophenone phenylhydrazone but also from a mixture of phenylhydrazine and acetophenone. A of toluene, benzene and n-heptane are -51.3, -40.3 and -29.4 e a . a t coverages of 0.0135, 1-igorous exothermic reaction and a color change 0.02 and 0.02 X IOi6 molecules/cm.2, respectively from orange to dark brown took place when such a (roughly “corresponding states” judging from the homogeneous mixture of phenylhydrazine, the keenergy and entropy curves). It is instructive to tone and polyphosphoric acid was heated to 120’. calculate the theoretical loss in entropy assuming The indole could be isolated in fairly pure form by that the niolecules are forming a two-dimensional adding water to the reaction product. .In extension of this method to the preparation of gas on the surface (riegIecting interactions for simplicity) with free areas corresponding to the other indole derivatives showed that it had fairly differences between the areas per molecule and the wide applicability. The aryl- and alkylsubstituted c‘o-areasof the molecules as determined by Keinball indoles shown in Table I mere prepared from the ant1 Rideal.5 -1sthe co-areas are 37, 35 and 38 A.?, corresponding ketones, phenylhydrazine and polyrespectively, the free areas are 37, 13 and 17 .kS2. phosphoric acid. The reaction involving N-methUsing the usual equation for the three-dimensional ylphenylhydrazine and acetophenone was carried translational entropv and the equation given by out in order to see whether polyphosphoric acid Iiembal16 for the two-dimensional translational could be used for the synthesis of N-substituted inentropy, the calculatctl losses are 19.6, 21.2 arid doles. As expected, 1-methyl-2-phenylindole was 21.3 e.u., respecti\-elv. The fact that the experi- obtained in good yield from this reaction. The use mental loss in entropy for n-heptane is only X.2 of isobutyrophenone and phenylhydrazine led to the e.u. greater than the \ ~ i l u ecalculated on the model isolation of 3,3-dimethyl-2-phenylindolenineshowo f the two-dimensional gas indicates that n-heptane ing that the polyphosphoric acid procedure could has considerable freedoin on the surface. KembaW also be applied to the synthesis of indolenines. In order to avoid tar formation it was very imporsuggested that the adsorbed benzene might retain I‘reeclom to rotate in the plane of the ring but mould tant, in all cases, to cool the reaction mixture imrnediately after the reaction had started. The maxilose the other two degrew of rotational freedom. This model correspcinds to a further loss of 16.9 mum temperatures used in these reactions are e.u. making a total calculated loss of 38.1 e.u., shown in Table I. It is probable that the yields of which is to be compared with the experimental some of these indole derivatives could be improved value of 40.3 e.u. The loss in entropy on adsorp- by changing these temperature conditions. The
a.
Q)
-
--f
+
-
(10) C . Kemball, Proc. Roy. Soc. (London), ZOIA, 377 (1950). ( 1 i ) The values of &.,(l atm 1 i& at T 3lO.lOK. Dire 24.9, 23.3 and 24 4 e u respectively. J W l > r e n s r i .and I I l I i l 1 J C u e m Phri 11. 776 (>we)*
-
.
(1) 1%. R. Snyder and F. X.Werber, TEISJOURNAL, 72,2862 (1950). (2) E. C. Horning, J. Koo and G. N. Walker, ibid., T I , 5826 (1951). (3) E. C. Horning and V. L. Stromberg, i b M , 74, 2880 (1952). (4)
R
(1942).
I
i:ir~ner
Vs’ C i h L l e y a n d Ls Welch 0 . u
Sqirlhrrrr
ZP. 9 6
NOTES
Aug. 5, 1952
3949
TABLE I Ketone
Hydrazine
Indole
Reaction temp., Yield,@ OC.
% 73d 76' 63A 5sk
Ivf.p.,b
'C.,cor.
Previous m.p., OC.
1-Methyl-2-phenyllooc 100-101 1oo-10la 2-Phenyl180 187-188 188-189g 2,4'-Diphenyl188 293-295' 297-298' 3-Methyl-2-phenyl170 90- 92 91- 92' 2-Methyl203 60k 56- 59 BOrn 2,3-Dimethyl-" 230" 68' 100-102 10V 155c 45 ...... ...... 3,3-Dimethyl-2-phenyl indolenine' Sui>Based on amount of ketone used. * All melting points are corrected (Kofler hot-stage). Bath temperature. limed in oacuo and recrystallized from cyclohexane. 0 J. Degen, Ann., 236,155 (1886). f Sublimed i n vacuo. " See ref. (4) Sublimed i n vacuo and recrystallized from large amouiits of ethyl acetate. * Microscopic slide, in sealed tube, m.p. 322326'. i S. Buu-Hoi, S. Honn, and R. Royer, Bull. soc. chim. France, [5] 17, 489 (1950). Sublimed twice in vacuo. A. Collet, Buli. soc. chinz. F I ~ Z C [3] P , 17, $4 (1897). E. Fischer, Ann., 236, 127 (1886). Best yields were obtained when phenylhydrazine and ketone were heated together on the steam-bath for a few minutes before the polyphosphoric acid was added. 0 The reaction is vigorous in spite of cooling but the mixture did not remain at this temperature for more than a few minutes. Sublimed twice in zlacuo and recrystallized from cyclohexane. g E. Fischer, Ann., 236, 129 (1886), obtained this meltitig point only after converting the indole to its S-nitroso derivative and reducing the latter after several recrystallizations. T See experimental. Acetophenone rlcetophenoc e p-Phenylacetophenone Propiophenone -4cetone Methyl ethyl ketone Isobutyrophenone
a-Methyl-a-phenylPhenylPhenylPhenylPhenylPhenylPhenyl-
@
use of temperatures much lower than those indicated (110-120°) led to the isolation of small quantities of bisic by-products in the case of "phenylindole and 3-methyl-2-phenylindole; these substances have thus far not been identified. The two benzyl ketones, desoxybenzoin and dibenzyl ketone, when allowed t o react with polyphosphoric acid and phenylhydrazine, yielded intractable, highly colored, polymeric substances. These results could not be changed by using the separately prepared ketone phenylhydrazones or by varying the temperature conditions. Similar negative results were observed with phenylacetaldehyde and butyraldehyde, both as free aldehydes with phenylhydrazine or as phenylhydrazones. Therefore, i t seems probable that polyphosphoric acid will be useless for the preparation of 2-unsubstituted indoles.
dolenine. I t melted a t 194-195' after three recrystallizations from ethanol. Anal. Calcd. for CrsHuNCIO1: C, 59.72; H, 5.01; S , 4.53. Found: C, a9.65; H, 5.10; N, 4.45. NATIONAL HEART INSTITUTE NATIONAL I N S T I T UOF ~ EHEALTH S BETHESDA, MARYLAND
Heats of Fusion of Aliphatic Polyesters' BY L. MANDELKERN, R. R. GARRETT AND P. J. FLORY RECEIVED APRIL 3, 1952
An investigation,* the results of which were published several years ago, yielded discordant values for the heats of fusion of two linear polyesters, poly-(decamethylene adipate) and poly-(decamethylene sebacate). Heats of fusion calculated from the depressions of the melting points T m caused by Experimental introductiun of copolymerized units in the chains General Method.-To a mixture of 0.05 mole of the ke- were substantially lower than those deduced from tone and 5 ml. of phenylhydrazine was added approximately the dependence of T m on chain length, and from 20 g. of polyphosphoric acid. The mixture was stirred with t h a t obtained in one instance from the lowering of a thermometer and warmed gently on a steam- or oil-bath Tm by diluents2 In more recent investigation^^,^ until a sudden rise of temperature indicated beginning of the reaction. By cooling the mixture externally with water the latter method has been chosen in preference to a t this point, the temperatures were maintained at the level measurements on copolymers principally because shown in Table I. After the end of the reaction, 100 nil. of of the better definition of the melting points of the cold water was added to the cooled mixture which was theti extracted thoroughly with ether. The combined, dried, polymer-diluent mixtures, and also because of its ether extracts yielded the corresponding indoles which were wider applicability to polymers including those for purified by the methods indicated in Table I. Caution which random alteration of some of the units, as by must be used in large scale runs because of the large amount. copolymerization, is impractical. Furthermore, . of heat given off by the reaction mixture. 3,3-Dimethyl-2-phenylindolenine.-A homogeneous mix- the use of dilatometric methods permitting gradual melting over periods of several days yields more reture of 14.2 g. of isobutyrophenone, 10.8 g. of phenylhydrazine and 30 g. of polyphosphoric acid was heated with liable melting points than the earlier micro method vigoraus stirring for a few minutes in an oil-bath a t 158". involving comparatively rapid heating of a small The color of the mixture changed from light yellow to dark brown. After cooling, there was added 150 ml. of ice- sample on the stage of a low power microscope. In the light of these advances it seemed desirable water and enough solid sodium carbonate to make the solution basic. Extraction with ether and removal of the sol- to repeat the earlier determination of the heat of fuvent after drying yielded an oil which was distilled t o give sion of an aliphatic linear polyester by the diluent
9.5 g. (45%) of a slightly yellow liquid; b.p. 141-145" (1.5 mm.). A picrat@ melted a t 153-155" after two recrystallizations from ethanol. A perchlorate was prepared by the addition of 60% aqueous perchloric acid t o a cold ethanolic solution of the in( 3 ) H. Leuchs, A. €Idler and A. IToffman, Ber., 61, 877 (1929), used zinc chloride for the preparation of this indolenine which they isolated by means of the picrate, m.p. 158-160°.
(1) This work was carried out a t Cornell University as part of a research program supported by the Alleganp Ballistics Laboratory, Cumberland, Md., an establishment owned by the United States Navy and operated by the Hercules Powder Company under Contract NOrd 10431. (2) R.D. Evans. H. R . Mightou and P. J Flory, THISj o w w m . 7 2 , 2018 (1950). (3) P. J. Flory, L.Ifandelkern and H. K. Hall. ibid., 73, 2532 (1IIJ1). ( 4 ) L. Mandelkern and P. J. Flory, i b i d , , 1 S , 3206 (1951).
