Synthesis of new macrocycles. Part I. Monomeric and dimerico-phthalate esters

摘要

2148 J.C.S. Perkin ISynthesis of New Macrocycles. Part 1. Monomeric and Dimerico-Phthalate EstersBy S. E. Drewes and P. C. Coleman, Department of Chemistry, University of Natal, Pietermaritzburg, SouthAfricaCondensation of the dipotassium salt of o-phthalic acid with a series of alkyl dibromides of increasing chain lengthafforded cyclic esters ranging in size from a ten- to a thirty-four-membered ring. Dibromides bearing an unevennumber of carbon atoms generally yielded dimeric compounds, while the even numbered dibromides gave mono-meric compounds. The dimeric esters are highly crystalline and appear to exist in fixed conformations.IN a preliminary communication the synthesis of ten-,eleven-, sixteen-, and eighteen-membered cyclic estersbased on o-phthalic acid was described.The presentinvestigation extends this work and has led to the isola-tion and characterisation of a number of new macro-cyclic esters.Few syntheses of cyclic esters of o-phthalic acid areknown. Spanagel describes the synthesis of cyclicethylene ester of o-phthalic acid (1) as well as that of thecorresponding dimeric ester (2). Yoneda, Yoshida, andFukui3 have outlined methods for preparing esters ofdicarboxylic acids with dimethylformamide, and weinitially used conditions similar to theirs. Ehrhart?using thermolytic depolymeri~ation,~ was able to preparea series of macrocycles from phthalic anhydride andglycols (e.g., pentane-l,5-diol, hexane-l,6-01). Yieldsof up to 90 were obtained. The largest ring esterobtained was compound (2), which was the only dimericcompound to be isolated.The yields of compound (2)and the monomeric analogue (1) were less than 5 andEhrhart suggests that no other dimeric esters wereisolated since they are too involatile to distil under theconditions employed (235"). The ' poor ' yield of com-pound ( l ) , an eightmembered ring, is ascribed toinherent strain in the molecule since the two carbonylgroups are twisted well out of the plane of the benzenering.(crown compounds) obtained by PedersenRelated compounds are the macrocyclic ethers 6,from the(41 n = 3( 5 1 n = 7( 6 ) n = 9( 7 ) n : 11( 8 ) n = 4( 9 1 n - 5(10) n z 6( 1 l ) n = 8(12) n 2 10(13) n = 12condensation of catechol and a suitable difunctionalreagent. Thus cyclic structures containing up to 60carbon atoms in the ring were obtained.These macro-cyclic ethers form stable complexes with the salts ofL. M. R. Crawford, S. E. Drewes, and D. A. Sutton, Chem. W. H. Carothers and E. W. Spanagel, J . Amer. Chem. SOC.,E. W. Spanagel, U.S.P. 2,092,031/1937. A. Liittringhaus and K. Ziegler, Annalen, 193?, 528, 155.S. Yoneda, 2. Yoshida, and K. Fukui, Kogyo h'agaku ' K. Ziegler, A. Luttringhaus, and K. Wohlgemuth, Annalen,W. A. Ehrhart, J . Org. Chem., 1968, 33, 2930.and Ind., 1970, 1351. 1936, 57, 929.Zasshi, 1966, 69, 641 (Chem. Abs., 1967, 88, 10,637). 1937, 528, 162.8 C. J. Pedersen, J . Amer. Chem. SOC., 1967, 89, 70171972alkali metals. One of these ethers, dicyclohexyl-18-crown-6-cyclic polyether (3) has been used in mechan-istic studies which involved removal of potassium ionthrough complexation with the crown ether.Presentindications are that the phthalate macrocyclic esterscomplex only weakly or not a t all with alkali metalsalts.The eleven compounds (2) and(4)-( 13) were preparedbythe same procedure, the only variation being an increasein the reaction time with increasing chain length of thedihalide. The yields quoted are thus not necessarilymaximum yields (see the Table). A slight molar excessCorn- Di-pound bromide(2) 1,2-(4) 1,3-(5) 1,7-(6) 1,9-(7) 1 , l l -(8) 1,4-(9) 116-(12) 1 , l O -(13) 1,12(10) 1,6-(11) 1,8-Ring341618Reactiontime (h)48489614416243672120168Yield M.p.10.1 1899.4 1936.5 16314113913.2 10616.6 1017.8 642.62.83.6() ("C)4.2 140-2.7 138-(10) of dihalide over phthalic acid salt was usedthroughout.N.m.r.and Mass Spectra.-Characterisation of thecyclic esters was mainly by n.m.r. and mass spectro-metry. The n.m.r. spectra of the dimeric compoundsdiffered in certain important respects from the mono-meric compounds. For all compounds the protons aregrouped in three areas at ca. z 2.35 (symmetrical peakof the aromatic protons), 5.60 (sharp triplet or multipletdue to the ' terminal ' methylene protons, i.e. the methyl-ene groups adjacent to the carbonyl group), and 7.5-9.0(broad multiplet or multiplet ending in a sharp peak)(q., see Figure and Experimental section).From the physical and n.m.r.data, several observ-ations can be made: (a) Use of an alkyl dibromidebearing an uneven number of carbon atoms invariablyleads to the formation of dimeric compounds, e.g.(4)-(7). An exception is compound (9) obtained from1,5-dibromopentane. The even-numbered dibromidesyield monomers except in the case of 1,2-dibromoethane.(b) In the n.m.r. spectra the triplet due to the terminalmethylene protons is always sharp if a dimeric ester isinvolved. For compound (2) it is a sharp singlet. Forthe monomeric esters this group of signals is much lessclearly defined. For all the compounds the signalsdue to the terminal methylene protons collapsedto a singlet by appropriate irradiation of the upfield( 7 7.5-9-0) methylene protons. It was also clear fromthe n.m.r.spectra of the three largest monomers com-pounds (1 1)-( 13) that impurities, absent from all other9 J. N. Roitman and D. J. Cram, J . Amer. Chem. SOC., 1971,lo H. Budzikiewicz, C. Djerassi, and D. H. Williams, 'Mass93, 2231.compounds, were present. Contamination was small incompound (11) but was detectable as additional peaks a t7 6.3 and 7-15 in compounds (12) and (13). Redistilla-tion of these particular fractions effected no markedpurification. (c) The esters are generally crystalline atroom temperature but the dimeric compounds havem.p.s well above those of their monomeric counterparts.Isolation of the dimers was a simple procedure whereasthe larger monomers required special techniques.Mass spectrometry was most useful, since the n.m.r.data alone exhibit little difference between a monomerand its corresponding dimer.For example, the inte-gration would be identical and the coupling patterns(apart from the one difference mentioned before) couldbe very similar.The mass spectra show a small number of characteris-tic peaks but for the rest are relatively featureless. AllAromatic protons met hyleneprotonsI__ 1-1-2 4 6 8TN.m.r. spectrum of compound (6)compounds show a molecular ion peak, from a 1 rela-tive abundance for compound (2) and 4 for compound(7) to 31.6 for compound (9). Fragment peaks norm-ally associated with phthalatesl* appear in all thespectra: ions at m/e 149 (invariably the base peak,protonated anhydride ion), 104, and 76.In all thedimeric compounds, a very diagnostic peak is located a t(M+*/2 + 1). This is a prominent peak and varies inintensity between 14 and lOOyo relative abundance. Apossible fragmentation pattern is shown in Scheme 1 forcompound (4).The monomeric esters (8)-(13) do not have the(M+'/2 + 1) peak. The molecular ion peak in thelower members of the series is fairly prominent (up to32) but declines in the larger-ring compounds to avalue of 1.7 in compound (13). Impurities in com-pounds (11) and (12) already apparent in the n.m.r.spectra, are probably responsible for the appearance ofa prominent peak at m/e 161.1.r. Spectra.-All the esters showed a characteristiccarbonyl absorption at 1739- 1715 cm-l.Two addi-tional bands (C-0 stretching) were located at 1290-1280 and 1135-1129 cm-l.Ease of Ring Formation.-It is instructive to considerthe ease of cyclisation by comparing the present findingswith conclusions drawn by Carothers l1 and by Flory.12Spectrometry of Organic- Compounds,' Holden-Day, San l1 W. H. Carothers, Chem. Rev., 1931, 8, 853.Francisco, 1967, p. 203. P. J. Flory, Chem. Rev., 1946, 39, 1372150 J.C.S. Perkin IFlory states that if a ring containing five carbon atomscan be formed, this will form exclusively, whereas six-and seven-membered rings are formed in addition tochain compounds, when the structures are suitable.4 .A/+: 412 ( 1 ‘/o) I aC-FissionOH JI 0 Im / e 1 04 (27 O/O)m i e 76 (18°/olii0 Ic=oI0m / e 149 (l0OVe)SCHEME 1 Mass fragmentation pattern to show formationof major peaks in dimeric cyclic esterLarger rings are formed only under special conditions(such as high dilution) which favour intra- over inter-molecular condensation.However, even under these13 K. Ziegler in Houben-Weyl ‘Methoden der OrganischenChemie,’ Verlag Georg Thieme, Stuttgart, 1966, vol. 4/2, p. 729.14 W. Baker, J. F. W. McOmie, and W. D. Ollis, J . Chern.SOC., 1951, 200.l5 S. Smolinski and J. Jamrozik, Tetrahedron, 1971, 27, 4977.conditions poor yields of eight- to twelve-memberedrings are obtained.Isolation of the ten-twelve-membered compounds(8)-(10) as major products, contrasts with these con-clusions. Incorporation of oxygen into the cyclic systemseems to favour cyclisation and Ziegler, Liittringhaus,and Wohlgemuth have demonstrated that yields ofmonomeric alkyl ethers of catechol were undepressedfor ten- and eleven-membered ring compounds. Theisolation of compounds (8)-( 10) together with the othercyclic esters indicates that steric factors must operateduring the difunctional condensation. Carothers andSpanage15 have shown that in the intermediate range,odd-membered rings are more difficult to produce thaneven-membered ones. They suggest that interferencesto ring closure are still operative after cyclisation andsince these interferences are greater for the odd numbers,this leads to instability.The present findings are in agreement with the general-isations of Carothers and S~anagel.~ Thus, the uneven-numbered dibromides (1,3-, 1,7-, 1,9-, and 1 , l l - ) do notgive the odd-membered esters, which would contain9, 13, 15 and 17 atoms respectively, but instead producethe corresponding even-numbered dimeric compoundsas major product.The exception to the rule is com-pound (9; n = 5) which should be dimeric.It has been re-emphasisedl that the steric factorwhich controls the formation of macrocycles must be thatof incorporation of a structure leaving at least fourcarbon atoms lying in one plane with angles of ca. 120”.o-Phthalic acid fit these requirements. Other accept-able compounds are for example, catechol, benzene-l,2-dicarbonitrile, o-phenylenediamine, cyclohex-l-ene-l,2-dicarboxylic acid, and maleic acid.This concept is ofcourse part of the more generalised ‘ principle of rigidgroups ’ stated explicitly by Ziegler l3 following thework of Baker et aL1* An example of recent work em-ploying rigid groups is that of Smolinski and Jamrozik,15who condensed 1,2- and 1 ,8-dihydroxynaphthalenes withthe tetrabromohydrin derivative of pentaerythritol.The findings of Pedersen8 afford information onfactors affecting ring closure. Pedersen warns, how-ever, that his yields of compounds do not necessarilyrepresent maximum yields, but it is evident that certainpolyether rings are formed more readily than others.The preferred rings are those containing five to sixoxygen atoms, each separated from the next by twocarbon atoms, for example, compound (14), which wasobtained in 80 yield, and also compound (15), obtainedin the lowest yield of all the compounds listed (2).These two compounds are identical except that one con-tains a rigid group whereas the other does not.Bycontrast, Yoshino and his co-workers16 report the syn-thesis in high yield (20) of the cyclic amide (16) fromcondensation of 4,4‘-methylenedibenzoyl chlorideand 4,4’-methylenedibenzylamine, where there cannotbe any assistance from rigid groups.l6 Y . Urushigawa, T. Inazu, and T. Yoshino, Bull. Chem. SOC.Japan, 1971, 44, 26461972 2151The mechanism of cyclisation in the o-phthalateesters is probably closely akin to that advanced by1 1 4 ) ( 1 5 )'-a,'( 1 7 )Looker and Sondheimer l7 for the formation of thetwenty-membered cyclic tetra-acetylene (17).Byanalogy, the sixteen-membered dimeric compound (2)could form as shown in Scheme 2. The product (a) inr 1L I n termediateIntermolecularcondensation I ( 6 1Dimer -SCHEME 2 Mechanism of the formation of cyclic estersthis Scheme has not been isolated, but Ehrhart 4 wasable to obtain it by the thermolytic technique. While(b) is the major product @yo), numerous other products,not isolated at present, are formed. This is demon-strated readily by examination of the products by t.1.c.l7 B. E. Looker and F. Sondheimer, Tetrahedron, 1971, 27,2567.(Kieselgel ; chloroform). There is some evidence thatthe intermediate (c), of low RF, is present. Brown andTyman,18 in synthesising a dimeric product (inter-molecular reaction) similar to the o-phthalate esters byDieckmann cyclisation, ascribe the absence of intra-molecular product to an unfavourable ' divergent 'rather than parallel orientation of the side chains of thestarting material, the diester (18).Conformation of Rings.-Predictions regarding possibleconformations of cyclic o-phthalate esters with more thanfour methylene groups in the central ring must be specu-lative since the large number of methylene groupsresults in an extremely flexible molecule.At the outsetit will be stressed that the presence of two rigid endgroups (as in the dibenzo-compounds) will impose re-strictions on the flexibility of structures. The n.m.r.I I AI19 1 TOP SIDE( 2 0 1spectrum of compound (2) shows four methylene groupsresonating as a sharp singlet a t 7 5.35 at 25".Thissinglet remains unaltered a t -20". The equivalenceof the four methylene protons implies an extremelysymmetrical molecule in which these eight protonsexperience the same shielding. Dreiding models indicatethat a possible conformation in accord with the aboveobservation is the one shown schematically, (19). Inthis conformation the two carbonyl groups, the twoester oxygen atoms, and the two methylene carbonatoms composing one half of the macrocyclic ring, alllie in one plane. The corresponding atoms, makingup the other half of the ring are similarly in one plane.These two planes are not quite parallel to one another.The rigid portion of the molecule, i.e.the benzene rings,are inclined at an angle of ca. 65" to one another. Thetwo benzene rings can be regarded as ' conformationalanchors' which will ' restrict' the molecule to theG. R. Brown and J. H. P. Tyman, Chem. and Ind., 1970,4362152 J.C.S. Perkin Iparticular conformation. For the larger dimeric com-pounds, related conformations are probable. In allthese compounds the eight ' terminal ' methyleneprotons are again equivalent and appear as a sharptriplet a t T 5-55-5.72 ( J 6.0 Hz). A similar triplet isobserved for a dimeric sixteen-membered ring (20) de-rived from catech01.l~In the monomeric esters the signal due to the four' terminal ' methylene protons is still centred round T5.60 but generally it is a multiplet or a partially-resolvedtriplet. I t is suggested that the monomeric esters areless rigid than their dimeric counterparts and do notreside in any particular conformation.The ' central' methylene protons for all the cyclicesters lie in the range z 7.8-8.8.The signal is amultiplet but generally one sharp peak stands outprominently, particularly in the dimeric esters. Thisobservation is again taken to reflect a greater degreeof symmetry in the latter compounds.Choice of Solvent.-The solvent used throughout theinvestigation was redistilled dimethylformamide.Initially conditions approaching those employed in thehigh dilution technique were used, but it was subse-quently found that these conditions did not improveyields.Other solvents (benzene, xylene, and n-butanol)were used without success.The reaction of dipotassium phthalate with alkylbromide proceeds via an SN2 mechanism (Scheme 3).1L CH2Br JSCHEME 3In dipolar aprotic solvents such as dimethylfonnamidethis type of bimolecular substitution is much faster,since solvation of the cation is promoted whereas thereactant anion (RC0,-) is less solvated (thus unen-cumbered and highly reactive) than in protic solventssuch as water or methanol.20EXPERIMENTALN.m.r. spectra were recorded on a Varian T60 instrumentand mass spectra were obtained on an A.E.I. M S 9 instru-ment. M.p.s were determined with a Kofler micro-hot-stage apparatus.The synthesis and isolation was similar for all the cyclicesters, except that the reflux time was increased for thelarger esters, and for compounds (1 1)-( 13) distillationunder high vacuum (220" at 6 x lo-, mmHg) was necessaryfor purification.7,8,17,18-Tetrahydrodibenzof,n 1,4,9,12tetraoxacyclo-hexadecene-5,10,15,20-tetrone, Cyclic Dimeric Ethylene Esterof Phthalic Acid.(2).-Dipotassium phthalate (10 g,41 mmol) and 1,2-dibromoethane (8.46 g, 45 mmol) in di-methylformamide (100 ml) were heated under reflux for 4 h.The dimethylformamide was then distilled off under re-duced pressure (50"), the residue was poured into coldwater (300 ml), and made basic with sodium hydrogencarbonate. Generally a gummy precipitate separated andthis was kept at 0" for 12 h. The gum was separated offand then dissolved in hot ethanol.On cooling, crystallinematerial precipitated out. This was filtered off and washedwith ice-cold ethanol to give white needles (800 mg), m.p.189" (from ethanol) (Found: C, 62.2; H, 4.0. Calc. forC2,H1608: C, 62.5; H, 4*2), M', 384, 7 (CDCl,) 2-38 (8H,m, ArH) and 5-35 (8H, s, 4 x CH,).All other esters were prepared and isolated thus, unlessdescribed later (see the Table).8,9,19,20;Tetrahydro-7H, 18H-dibenzog,p 1,5,10,14tetra-oxacyclo-octadecene-5,11,16,22-tetrone (4) .-This product wasobtained as white needles (9-4y0), m.p. 193" (from ethanol)(Found: C, 63.9; H, 4-8. C,,H,,O, requires C, 64.1; H,4-9), M', 412, T (CDC1,) 2.32 (8H, m, ArH), 5.54 (8H, t,4 x CH,), and 7.82 (4H, m, 2 x CH,).8,9,10,11,12,13,23,24,25,26,27,28-Dodecahydro-7H,22H-dibenzoc,p 1,6,14,19tetraoxacyclohexacosene-5,15,20,30-tetrone (5).-This was obtained as white needles (6.5y0), n1.p.163" (from ethanol) (Found: C, 68.7; H, 7.0.C,,H,,08requires C, 68.7; H, 6-9), M', 524, T (CDCl,) 2.35 (8H, m,ArH), 5.70 (8H, t, 4 x CH,), and 8-57 (20H, m, 10 x CH,).8,9,10,11,12,13,14,15,25,26,27,28,29,30,31,32-Hexadeca-hydro-7H,24H-dibenzoc,r 1,6,16,2 ltetraoxacyclotviaconl-ene-5,17,22,34-tetrone (6) .-This was obtained as whiteneedles (4-2), m.p. 140-141" (from ethanol) (Found : C,69.9; H, 7.7. C,$H4@, requires C, 70-3; H, 7.6), M+,580, z (CDCl,) 2.35 (8H, m, ArH), 5-70 (SH, t, 4 x CH,), and8-63 (28H, m, 14 x CH,).Eicosahydro-7H, 26H-dibenzoc, t 1,6,18,23tetmoxacycZo-tetratriacontene-5,19,24,38-tetrone (7) .-This was obtained aswhite needles (2.7), m.p. 138-139" (from ethanol) (Found:C, 71.3; H, 8.3.C,,H,,O, requires C, 71.7; H, 8.2),M+, 636, T (CHCl,) 2.35 (8H, m, ArH), 5.72 (8H, t, 4 xCH,), and 8.70 (36H, m, 18 x CH,).3,4,5,6-Tetrahydro-2,7-benzodioxacycZodecene- l,dione (8).-The compound was obtained as white needles in highyield (13-2), m.p. 106" (from ethanol) (Found: C, 65.2;H, 5.5. C,,H,,O, requires C, 65.5; H, 5.5), M+, 220,z (CDCI,) 2-35 (4H, m, ArH), 5.63 (4H, m, 2 x CH,), and8.07 (4H, m, 2 x CH,).4,5,6,7-Tetrahydro-3H-2,8-benzodioxacycloundecene- 1,9-di-one (9).-Of all the cyclic esters this one was obtained inhighest yield (15-5y0), as needles, m.p. 101" (lit.,4 100-8-102.3") (from ethanol) (Found: C, 66.5; H, G e l .Calc. forC1,H140$: C, 66.6; H, 6-Oy0), M+, 234, T (CDCI,) 2.37 (4H,in, ArH), 5.68 (4H, m, 2 x CH,), and 8.24 (6H, m, 3 x CH,).3,4,5,6,7,8-Hexahydro-2,9-benzodioxacyclododecene- 1,lO-di-one (lo).-This was obtained as white needles (7.89/,),m.p. 64" (lit.,$ 63-7-64.5") (from ethanol) (Found: C, 67.7;H, 6.5. Calc. for C14Hl,O$: C, 67.7; H, 6.5), M+, 248,T (CDC1,) 2.42 (4H, m, ArH), 5.62 (4H, t, 2 x CH,), and 8.29(SH, m, 4 x CH,).Cyclic Octamethylene, Decamethylene, and DodecamethyleneEsters of Phthalic Acid, ( 1 1)-( 13).-These three esters wereobtained in low yield (2-6, 2.7, and 3-6y0, respectively),were non-crystalline, and generally difficult to isolate. Pre-8,9,10,11,12,13,14,15,16,17,27,28,29,30,31,32,33,34,35,36-1s A. W. Archer and P. A. Claret, Chem. and Ind., 1969, 171.*O A. J . Parker, Chem. Rev., 1969, 69, 11972 2153cipitation from water yielded a gum in each case and thiswas still contaminated with the dibromide. Purificationwas achieved by distillation under high vacuum (200-220"a t 6.3 x mmHg) but C and H analyses were unsatis-3,4,5,6,7,8,9,10,11,12,13,14dodecahydro-2,15-benzodioxa-cyclo-octadecene-1, le-dione (13), M', 332, T (CDC1,) 2.38 (4H,m, ArH), 5.75 (4H, m, 2 x CH,), and 8.77 (20H, m, 10 xCH,).factory : 3,4,5,6,7,8,9,10-oclahydro- 2,1 l-benzodioxacyclotetra-decene-1,12-dzone (ll), M+, 276, T (CDC1,) 2-38 (4H, m, ArH), We thank Professor Dm ** Sutton for suggesting the5.60 (4H, m, 2 CH,); project, Mrs. L. M. R. Crawford for initiating the work, and3,4,5,6,7,8,9,10,11 , 12-decahydro-2,13-benzodioxacyclohexa- the South African for Scientific and Industrialdecene- 1,14-dione (1 2), Mf, 304, T (CDCl,) 2-45 (4H, m, ArH), Research forCH,), and 8-57 (12H, m, 6assistance*5.80 (4H, m, 2 x CH,), and 8-80 (16H, m, 8 x CH,); 2/600 Received, 14th March, 1972