5i2 PERKIN: THE CIS AKD TRANS-MODIFICATLOSS XLVII.--T12R Cis- and Trans-Jrlodzjkations of 1 : 2- Tetraiizetlzyteizedicarbox~lic acid, and 1 ; 2-Peizta- rnethylenedica.r.bo;l~c acid. By W. H. PERKIN, Juii. IN the course of his interesting researches on the reduction of phthalic acid (Annulen, 258, 145 ; 269,145), Baeyer has shown that hexa- hydrophthalic acid exists in two modiEcations, which differ very widely in their properties, and which he distinguishes by the prefixes “ fumaroxd ’’ or trans ” end ‘I maleinoid ” or ‘I cis.” Graphically, these two acids may be represented thm- Trans- or fumsroid Cis- or mnleinold modification. modification. but it is not easy to clearly understand the different arrangement of the groups of atoms in these two formulae unless models are used.The properties of the trans- and cis-hexahydrophthalic acids which hare a direct bearing on the results described in this paper are the following. The trans-acid (m. p. 21.5’) distils unchanged if quickly heated, but when kept at a temperature slightly above its melting point, water is eliminated, and the acid is partly converted into the anbpdride of the cis-acid. The cis-acid (m. p. 192”) is gradually decomposed at its melting point, yielding the anhydride, which melts at 32”, and when heated with hydrochloric acid at 180°, it is converted into the trans-acid.OF 1 : ~-TETBAMETHYLEXEDICARBOXYLIC ACID, Em. 573 One of the most interesting results obtained in the examination of these acids is the obserration that the trans-acid, when heated with acetyl chloride, also yields an anhydride melting a t 140", which, at a temperature of 210-220", is gradually converted into the anhydride of the cis-acid.BaeFer points out the great similarity there is between these acids and the dimethylsuccinic acids, which also exist in two modifica- tions- ( p 3 yH3 H*Y*COOH H.7.C 0 0 H COOH*V*H H*Q*COOH Trans-modi fieation. Cis-modification. CH, CH, These two modifications are convertible the one into the other, by the same methods as those used in the case of the hexahydrophthalic acids ; both yield anhydrides, melting at 87" and 38" respectively, the former being converted into the latter by heating. It is quite easy to understamd from an examination of the models that the cis-acids should, like male'ic and succinic acids, yield an- hydrides, but i t is somewhat more difficult to explain the existence of the trans-anhydrides, especially when the fact is taken into account that it has not been found possible to produce an anhydride of fumaric acid.However, as Baeyer carefally points out,, the study of the molecule models of these trans-acids at once reveals the fact that the relative directions of the affinities holding the carboxyl groups are not by any means the same--the affinities in the case of fnmaric acid being inclined at an angle of 180", whereas in the case of the trans-hexa- hydrophthalic acid and trans-dimethylsnccinic acid the angle is approximately 109". It may be assumed, therefore, that the reason why famaric acid does not give an anhydride is because the distance between the two carboxyl groups is too great ; in the case of trans-hexahydrophthalic acid and trans-dimethylsuccinic acid, the distance is far less, and the strain necessary to bring the two carboxyl groups together is not sufficient to prevent the formation of an anhydride ; nevertheless, as the strain is still much greater than in the anhydrides of the cis-acids, it may be assumed that there is a tendency in the trans-anhydrides to change into the more stable cis-forms.From considerations such as these, Baeyer concludes that the trans- modificaticn of 1 : 2-pentamethylenedicarboxylic acid must also yield an anhydride, because the conditions of strain in this acid are almost exactly the same as in traizs-hexahydropEthalic and trans-dimethyl-5 i 4 PERKIN : TEE CIS- AND TRANSMODIFICATIOKS snccinic acids; he also states that it is clear, from the examination of the models, that, trans-trimethylenedicarboxylic acid, COOH H \/ CH2<6 C /\ cannot yield an anhydride, whereas in the case of trans-tetramethyI- H COOH COOH H \/ FH,.S: enedicarboxylic acid, , it is doubtful whether an an- CH2*C /\ H COOH hydride could be formed or not.These theoretical deductions agree with the known facts as far as trimethylenedicarboxylic acid is con- cerned ; the trans-modification of this acid does not give an auhydride, although the cis-modification does. With regard to the intermediate tetramethylene- and pentamethyl- enedicarboxylic acids, the cis- and tram-modifications of these acids had not at that time been prepared; and, indeed, some preliminary experiments which I made with the object of obtaining two modifi- cations of 1 : 2-tetramethylcnedicarboxylic acid gave results which, I thought, pointed to the existence of only one form of this acid (compare Ber., 26, 2245).This want of success was doubtless due to the very slight differerce in the melting points of the two tetramethylene acids, as some time since, on again carefully experimenting on the subject, I found that both tetramethylene- and pentarnethylene-dicarboxylic acids exist in well defined cis- and trans-modifications. Tetramethylenedicarboxylic acid (m. p. 138")" has been described in a previous paper (Trans., 1890, 57, 18). It was prepared by the following series of reactions, which leare no doubt, as to its constitu- tion.When ethylene dibromide acts on the sodium compound of ethylic malonate, ethylic trimethylenedicarboxSlate is formed, together with small quantities of ethylic butanetetracarbox-ylate, the latter result- ing: from the direct action of 1 mol. of ethylene bromide on 2 mols. of the sodium compound, thus- 2(COOC,H5)2CHNa + BrCH2*CH2Br = (CO 0 C2H5)2CH*C H2CH2*CH (COO C2H5) + 2NaBr. This ethereal salt yields a disodium compound, which, when * The melting point, 130°, previously giren, is too low.OF 1 : 2-TETRAMETHYLENEDICARBOXYLXC ACID, ETC. 575 treated with bromine, is converted into ethylic tetramethylenetetra- carbox ylate. On hydrolysis, this ethereal salt yields the corresponding tetra- carboxylic acid, which, a t a temperature of ZOO", is rapidly decom- posed with formation of 1 : 2-tetramethylenedicarboqlic acid or its anhydride.~H2*~(COOH)z $lHZ*C;H*COOEI 3- eco, - CH:,.C(COOH), CH,*CHCOOH The yield of ethylic butanetetmcarbosylate obtained by this method is so small (about :3 per cent.) that it mas found impossible, at that time, to submit the tetramethylene derivatives obtained to a, detailed examination ; since then, however, by substituting ethylene chloride for ethylene bromide, and introducing other modifications in the method of preparation, as fully described in this paper, the yield has been so much improved that considerable quantities of material can now be obtained with comparative ease. The first fact that was established in re-examining tetramethylene- dicarboxylic acid was that the acid obtained in the above synthesis under the conditions described in this paper, and which melts at 138O, is the cis-modification; when heated with acetyl chloride it is con- verted into an anhydride (m.p. '75"), which dissolves readily in boil- ing water, with regeneration of the same acid. T?-am-tetramethylenedicarboxylic acid is prepared by heating the cis-acid with concentrated hydrochloric acid at 190' ; itl melts at 131", or only 7" lower than the cis-acid, and, like the latter, it is very readily soluble in water; to this great similarity in properties must be attributed the fact that in the first experiments the existence of the second modification was overlooked. Generally speaking, when an acid esists in two forms, the trans- melts a t a higher temperature than the cis-modification, notable exceptions being the ax-dimethylglutaric acids, and the hexnhydroisophthalic acids ; to these must now be added the tetramethy lenedicarboxylic acids, as in this case the ti-ans- melts lower than the cis-modification.The determination of the dissociation constants for the electric con- d uctivity of the two isomeric tetramethylenedicarboxylic acids, which Dr. Walker kindly undertook, has gi-i-en very interesting results, the values found being the following. C 0 OH* C H (CH,) C H2.C H (C H,) *C 0 0 H, T?-am-tetramethylenedicarboxjlic acid.. . . . . K = 0.0028 Cis- 7 ) , :, . . .. .. K = 0.0066376 PERKE: TEE CIS- AXD TRBXS-MODIFICATIONS Here, again, these two acids show an unusual behaviour, as, in nearly all cases which have been inrestigated, the trans-acid has a, higher constant than the cis; for example- Trans-dimethylsuccinic acid ......Trans-hexahydrophthalic acid.. .... R = 0-0191 H = 0.0062 Cis- 11 ,? ,, ...... K = 0.0123 Cis- 9 , Y f ,, ...... K = 0.0044 But this is not always so, and, as an exception, the case of the diethylsnccinic acids may be cited. Trans-diethylsuccinic acid. ........ K = 0.0245 cis- 1, 1 , ........ K = 0.0343 Although the two tetmmethylenedicarboxylic acids possess, in many ways, very similar properties, they are very sharply characterised by their behaviour towards acetyl chloride. The cis-acid, when digested for a short time with this reagent, is completely converted into its anhydride, but the trans-acid may be heated at 160-170" with acetyl chloride for 1$ hours without any perceptible change taking place ; again, the cis-acid decomposes on distillation, with formation of the anhydride, but the trans-acid, when rapidly heated in small quantities, distils almost unchanged, and it is only when repeatedly distilled that water is eliminated, and then with formation of the anhydride of the cis-acid, the conversion being, apparently, very incomplete, even after four distillations.In this behaviour, the cis- and trans-tetramethylenedicarboxylic acids show great similarity to the corresponding hexahydrophthalic acids, with the exception, however, that trans-tetramethylenedicub- oxylic acid does not yield a, distinct anhydride, whereas trans-hexa- methylenedicarboxylic acid does, this being in strict accordance with the properties which Baeyer predicted for these acids.1 : 2-Pentamethylenedicarboxylic acid has already been prepared (Tmns., 1887, 51, 244) from ethylic pentanetetiwarboxylate, (COO C*H,),CH*CH,*CH2'CH,*CH (COGC2H5)2, by a series of reactions exactly analogous to those employed in the synthesis of tetramethylenedicarbox~lic acid, as described above. The acid thus obtained melts at 160": and is trans-pentamethylene- dicarboxylic acid, whereas, in the corresponding synthesis of the tetramethylenedicarboxylic acid, the cis-modification is produced. Trans-pentamethylenedicarboxy Iic acid is attacked by acetyl chloride with difEcnlty; it is, for example, not appreciably acted on when heated with this reagent a t 100" for three hours, but at 140-150" it is decomposed, with formation of the anhydride of the cis-acid.This same change takes place when the acid is distilled, the conrer-sion, after two or three distilhtions, behg apparently complete; in this respect it differs f porn trans-tetramethylenedicarboxylic acid, which is mush more stable, and only partially converted into the anhydride of the GiS-acid on distillation. It is remarkable that, in spite of numerous experiments, I have not been able to prepare an anhydride corresponding to fruns-penta- methylenedicarboxylic acid, although, according to Baeyer's theory, such an anhydride should be formed with about the same ease as the anhydride of trans-hexahydrophthalic acid. Possibly, under other conditions? mch an anhrdride might be obtained, but this seems imp roba bl e.Cis-pentame thylenedicarboxylic acid is readily obtained by dissolr- ing the anhydride, prepared by distilling the trans-acid, in dilute potash. It melts at 141", and is much more readily soluble in water than the trans-acid ; when heated with concent rated hydrochloric acid at lW", i t is converted, apparently quantitatively, into the trans- acid (m. p. 160"). The dissociation constants for the two pentamethylenedicarboxjlic acids were determined by Dr. Walker, with the following results. K = 0.0120 Trans-pentameth;~lenedicarboxylic acid.. . . . . Cis- 9 7 7, ,, ...... K = 0.0158 Here again the cis-acid has the higher constant, as was also observed in the case of the tetramethylenedicarboxylic acids. It is very remarkable that the constants of the pentamethylenedi- carboxylic acids should be so wry much larger than the constants either of the tetramethylene- or hexamethylene-dicarboxylic acids, ns the following comparison sliows. Trans-.Cis - Tetramet hylenedicarboxylic acid . . 0.0028 0.0066 Pentamethylenedicarboxylic acid . . 0*0120 0.01 58 Hexamethylenedicarboxylic acid . . 0.0062 0.0044 (Hexab jdrophthalic acid) In this, and in many other respects, which mill be a t once noticed in comparing the properties of these acids, there are sharp differences in their behaviour. There is no gradual change in properties in passing from the tetramethylenc to the hexamethylene derivatives, such as might have been expected. Obviously, then, these acids, although tbey are analogously constituted, and differ from each other by CHz, cannot be considered as true members of a homologous series; tbey behave rather as if they belonged to three distinct series, and a similar deduction may be drawn from a consideration of the pro- perties of the other derivatives of tetra-, penta-, and hexa-methylene. I propose to return to this point in a subsequent paper.578 PEELKIN : TEE CIS- AND TRANS-MODIFXCATIOSS The results of this investigation show once again that it is not safe to attempt to deduce the configuration of stereoisomeric acids from their melting points or dissociation constants ; the only sure method is to investigate the behaviour of the isomers when treated with acetyl chloride and with hydrochloric acid at a high temperature.Preparation of EtR ylic Bu f anetetracarboxyla f e , (COOC2H,)2CH*CH,*CH~*CH (COOCZH,)Z. This substance is best obtained, as explained in the Introduction t o this paper, by treating the sodium compound of ethylic malonate with ethylene chloride, the reaction proceeding thus- 2(COOC,HJ2CHNa + C1CH2.CH,Cl = (C 0 0 C2H,),C H*CH,*CH,*CH (C 00 C2H5) + 2XaC1.The details of the preparation are as follows. 9.2 grams of sodium is dissolved i n 120 C.C. of absolute alcohol, and the solution, when quite cold, transferred to a soda-water bottle, 64 grams of ethylic malonate and 21 grams of etliylene chloride added, and the whole mixed as thoroughly as possible by shaking, an opera- tion rendered somewhat difficult on account of the separation of solid particles of the sodium compound of ethylic malonate, which some- times cause the liquid to set to a jelly-like masfi. The soda-water bottle is securely corked and tied down, and heated in a water bath at 100" for eight hours ; usually four to six such bottles were heated at the same time.When cold, the alcoholic solution is carefully decanted from the cake of sodium chloride which has separated, and the alcohol distilled off on a water bath ; the residue, together with that remaining in the bottles, is then mixed with sufficient water to dissolve the salt, and extracted three times with ether. The ethereal solution is well washed with water, dried over calcium chloride, the ether distilled off, and the brownish, oily residue, which from four bottles weighs always about 220 grams, is fractioned under reduced pressures (40 mm.> until the thermometer, immersed in the boiling liquid, rises to 150".The colourless oil (125 grams) which passes over consists c.f a mixture of ethylic trimethylenedicarboxylate with unchanged ethylic malon,tte ; the very dark coloured residue (70 grams) which remains in the distilling flask contains ethylic butanetetracarboxylate. The distillate of low boiling point is now fractioned under the ordinary pressure, and the fraction 180-225" collected and used again in pre- * In this and in subsequent distillations, when the amount of liquid to be fractioned was not less than 5 0 c.c., the distiliation flasks recommended by Claisen (Annalen, 277, 177) were found to be very useful, as, with these, there is much less risk of frothing over than when the ordinary fractioning flasks are used.OF 1 : 2-TETUMETBYLEXEDICARBOXYLIC ACID, ETC.579 paring further quantities of ethylic bntanetetracarboxylate, the opera- tion being conducted in the same way as before, except that, aR this fraction is assumed to contain only 50 per cent. of ethylic malonate, the relative quantities employed are 128 grams of oil, 9.2 grams of sodium, and 21 grams of ethylene chloride ; the isolation and separa- tion of the product into two fractions is carried out as before. It is very remarkable that the yield of crude ethylic butanetetra- carboxylate is now very much larger than in the first series of operations ; from four botkles (in each of which 128 grams of recovered oil had been treated), no less than 440 grams of oil, boiling above 150" (40 mm.), was obtained, or at least double that resulting from the treatment of pure ethylic malonate ; and i t is also noticeable that the product obtained in the second preparation is not nearly SO dark coloured, and is much more easily purified.The crude ethylic butanetetracarboxylate is now submitted to fmc- tional distillation under reduced pressure (40 mm.), not more than 100 grams being distilled at once ; a small quantity of oil passes over below 200°, biit the temperature rises rapidly to 2Y0°, between which and 250" the principal portion distils as an almost colourless oil, a small quantity of a tarry mass remaining in the flask. The distillate is once more fractioned, and the fraction 230-2.50" (40 mm.) collected for use in subsequent experiments ; the amount obtained from 500 grams of crude oil boiling above 150" (40 mm.) is about 300-330 grams.That this oil is nearly pure ethylic butane- tetracarboxylate is shown by the following analysis. Theory. 55-49 per cent. Found. C16H2608. <: . . . . . . . H .. .. .. .. 7-67 ,, 7.51 ,, 55-21 per cent. Pure ethylic butanetetracarboxylate boils at 240-24.5" (50 mm.) . The oil from the second preparation, boiling below 150" (40 mm.), is fractioned as before under ordinary pressure, and the portion boiling at 180-225" mixed with one-third of its weight of ethylic rnalonate, and again treated with sodium ethoxide and ethylene chloride, the quantities to bc employed being calculated on the assumption that the mixture now contains 50 per cent. of ethylic malonate. As the result of a number of experiments it was found that this method could be continued apparently indefinitely ; after the first preparation, the yield of ethylic bntanetetracarboxylate is always very good.This reaction is at present being carefully investigated with a view to affording some explanation of the remarkable series of decompositions which evidently take place.580 PERKIX: THD CIS- AND TRANS-MODIFICATIOSS FH2-$l(COOH)2 CH2* C (C 0 0 H) ?' 1 : 1 : 2 : 2-Tetranzethy7enetetracarbozylic acid, This acid has already been obtained by the hydrolysis of ethjlic tetramethylenetetracarboxylate by means of alcoholic potash, the crude product being purified by conversion into the lead salt, (Trans., 1887, 51, el), but, in preparing large quantities of this acid, this method proved too laborious, and after many experiments the follow- ing was adopted as being the most convenient and yielding the purest product.In the first place, ethylic butanetetracarboxylate was converted into ethylic tetramethylenetetracarboxylate i n the same way as before, with this slight difference, that in all cases a slight excess of sodium and of bromine were employed, so as to ensure the product being free from unchanged ethylic butanetetracarboxylate, as even small quantities of the latter render the subsequent purification of the tetramethylene derivatives a matter of considerable difficulty. The quantities used in each operation were, 35 grams of ethylic butane- tetracarboxylate, 5 grams of sodium, and 18 grams of bromine ; tlie crude ethereal salt was isohted as before, and at once cooverted into tetnmethylenetetracarboxylic acid. For this purpose, the crude ethereal salt was digested on a water bath with 18 times the calculated quantit!y of a strong, hot solution of pure barium hydroxide ; hydrolysis took place very rapidly, the liquid becoming almost solid, owing to the separation of insoluble barium tetramethylenetetracarboxylate.After heating for two hours, the mix- ture was vigorously boiled on a sand bath for half an hour, and the pre- cipitated saiidy barium salt collected by aid of the pump, and mashed well with hot water. The barium salt was then stirred up with a quantity of boiling water and exactly decomposed by dilute sulphuric acid, great care being taken that no trace of sulphuric acid was left in the product, as other--ise charring would take place during the subsequent puri6 ca tion of the anhydride of t e trame th ylenedicarb oxjlic acid (p.582) by distillation. The filtrate from the barium sulphate was evaporated to a small bulk(on a water bath and allowed to stand, when, after some days, beautiful colourless crystals of nearly pure tetramethylenetetracarb- oxylic acid separated. These were collected by aid of the pump, drained on a porous tile, dissolved in a small quantity of water, and the eolution allowed to concentrate over sulphuric acid in a vacuum. The magnificent glistening crjstals which formed were collected and exposed to the air for some days ; they then contained two mole- cules of water of crystallisation.Found. Theory.I. C8€TsOB + 2H,O. 13.43 -2 H,O ........ 13.50 13.58 These crptals lose their water of crystallisation rapidly a t loo', arid more slowly in a desiccator over snlphuric acid in a vacuum, becoming quite opaque. For analysis the substance was dried a t 100". Theory. Found. CSHROS- C .......... 41.38 per cent. 41.38 per cent. H .......... 3.60 ,, 3-45 ,, When heated rapidly in a capillary tube, tetramethylenetetracarb- oxylic acid decomposes at about 198-203" with evolution of carbonic anhydride, and formation of tetramethylenedicarboxylic acid (or its anhydride), but small quantities of impurity and also the rapidity of heating very mwh influence the observed decomposing point ; the temperature 145-150", preriously given (loc. &t., p. 22), is certainly much too low.It dissolres very readily in water, alcohol, and ether, and behares in all respects like a saturated acid. Its solution in sodium carbonate does not decolorise permanpnate, even on long standing. The siher salt of tetramethylenetetracarboxylic acid was obtained as a white, amorphous precipitate on adding silver nitrate to a faintly alkaline solution of the ammonium salt; it was collected, washed well with water, and dried over snlphuric acid in a vacuum. This salt decomposes very suddenly when heated, leaving a voluminous mass of spongy silver. For combustion it was intimately mixed with finely dirided copper oxide, and the silver was determined in the wet Kay by Carins' method. Theory. Found. CsH,A&O,. C .......... 14.36 per cent. 14.54 per cent.H .......... 0-82 ,, 0.60 ,, Ag ........ 65.40 ,, 65.45 ,, The neutral solution of the ammonium salts shows the foIIowing behaviour with reagents :-Lead acefate, a white, amorphous precipi- tate ; barium nitrate, n white, gelatinous precipitate ; calcium chloricle, no precipitate ; copper sulphate, no immediate precipitate ; but OR standing the copper salt gradually separates as a light blue pre- cipitate- QH2*v E€*C 0 CH~.CH.CO>~* Cis- Te t ram e t h y 1 e nedic ar b ox y 1 ic anhydride , This may be readily obtained from the crude tetramethylenetetm- For this purpose the carbosylic acid prepared as described above.382 PERKIN: THE CIS- AND TRANSMODIFICATIONS filtrate from the barium sulphste is evaporated to dryness, and the syrupy mass thus obfained is heated in a.n oil bath at 200" until the evolution of carbonic anhydride has ceased ; the dark brown residue is then digested with three times its volume of acetyl chloride for two hours, the excess of acetyl chloride and acetic acid distilled off, and the crude tetramethylenedicarboxylic anhydride purified by frac- tionation under reduced pressure (160 mm.). If the product is free from adipic acid (arising from the decomposition of butanetetracarb- oxylic acid present in the crude tetramethylenetetracarboxylic acid, 2oc.cit., p. 20) the whole distils, after twice fractioning, between 210-212" as a colourless oil, which solidifies completely on cooling, and c0nsist.s of pure tetrametbylenedicarboxylic anhydride. This substance is difficult to burn, and requires a very hot tube, the com- bustion being conducted rery slomly.The folloming results were obtained. Found. Theory. 57.14 per cent. I. r----3. C,H,O,. C . . . . . . . . , . 57.15 57.10 per cent. H.. . . . . . . . . S.03 4.89 ,, 4.76 ,, I t is di5cult to decide what the correct melting point of this an- hydride is. Using the products from different preparations, it was ob- served on several occasions, that when the melted substance is stirred with a thermometer as t,he mass gradually solidifies, the temperature remains constant at 71". If, however, a strong solution of this an- hydride in acetyl chloride is allowed to slowly evaporate over potash, beautiful, colourless crystals are deposited, which melt at about 77", t h i s then probably represents the correct melting point of the sub- stance, and agrees closely n-ith that previously found (76-78").When melted on a watch glass, the anhydride gives off vapours which are very irritating to the throat and produce violent coughing; it distils, under ordinary pressures, almost without decomposition, at, The pure anhydride dissolves only very slowly in cold water, but readily on warming, with formation of' cis-tetrametbylenedicarboxylic acid. about 270-273'. It is readily soluble in alcohol and ether. H COOH \/ yH'2v Cis- Tetramethylenedicarboz ylic acid, CH$C /'\ H COOH When tetramethylenedicarboxylic anhydride is dissolved in a little hoiling water, and the solution allowed to concentrate over sulphuricOF 1 : 2 4 X T R A M E T H Y A R B O X p L I C ACID, ETC. 583 acid in a vacuum, cis-tetramethylenedicarboxylic acid gradually crystallises in magnificent transparent plates.These were freed as far as possible from the thick mother liquor by filtration on a pump, drained on a porous tile, and recrystallised from a small quantity of water, or from hydrochloric acid. For analysis, the substance was dried at 100". Found. FJ----7 Theory. I. XI. C4H6 (COOH),. C .......... 49.91 5905 per cent. 50.00 per cent. H .......... 5-62 5.70 ,, 3-56 This acid melts a t 137-138" and not at 130" as previously stated; it is very readily soluble in water, but much more ~ p m - ingly in concentrated hydrochloric acid, from which i t crystallises well. The dissociation constant for the electric conductivity of this acid was kindly determined by Dr. Walker, who obtained the result K = 0.0066; this value is identical with that for succinic acid, and very much lower than that of the corresponding cis-pentamethylene- dicarboxyh acid (K = 0.0158).Tetramethylenedicarboxylic acid has st11 the properties of a satu- rated acid ; its solution in sodium carbonate does not decolorise per- zanganate, even on long standing, and, on boiling, action only takes place very slowly. When treated with bromine and amorphous phosphorus it readily yields dibromotetramethylenedicarboxylic acid, thus- YH,*YBr.COOH + 2Br2 = + 2HBr, CH,-CBr.COOH YH2* H COO H C H2. CH* C 0 OH a decomposition which is at the present time being carefully in- vestigated. Methyl ic Salt of Cis- Tefranzeth y lened icarbox ylic acid .-In preparing iarge quantities of cis-tetrameth~lenedicarboxylic acid, the method usually employed mas t o heat the crude tetrabasic acid at 200" until carbonic anhydride ceased to be erolred, and then to dissolve the dark-brown residue in methyl alcohol and by the addition of sul- phuric acid, or by saturating with hydrogen chloride, to convert the whole into t h e Iuethylic salt.After two days, the product was poured into water, the oil which separated extracted with ether, the ethereal solution washed well with water and dilnte sodium carbonate solu- tion, dried over calcium chloride, and the ether distilled off. The residual oil on fractionation under reduced pressure (80-100 mm.) distilled faiyly constant'ly a t about 18.5". Theory. Found. C4H6 (COOCHJp C .......... 55.48 per cent.55.81 per cent. H.. ........ 7.11 ,, 6-98 ,,554 PERKIN : THE CIS- AND TRANS-XODIFICATIOSS Methylic cis-tetramethylenedicarbxylate is a colourless, pleasant smelling oil, which distils with very slight decomposition under ordinary pressures at about 225"; the corresponding ethylic salt has already been described (Trans., 1887, 51, 23), and boils a t 238-242" (720 mm.). Cis-tetramethylenedicarboxglic acid is prepared from the methylic salt by digesting it with an excess of alcoholic potash for one hour ; the product is dissolved in xater, evaporated till free from alcohol, acidi- fied, and the clear solution extracted three times with pure cther. The ethereal solution, after drying over calcium chloride, deposits the acid as a colourless oil, which solidifies completely on cooling, and may then be purified by recrystallisation from hydrochloric acid. yHz*vH*CONH2 This CH2*CH-CONH2' Diam.ide of T e t r a r n e t hyle ned icarboz yl ic a cX, substance is very readily prepared by leaving the methSlic salt of the acid in contact with concentrated aqueons ammonia: after standing for 24 hours, the aqueous layer is decanted from the crys- talline cake which has formed, and the latter is purified by recrystal- lisation from water.The diamide separates from its hot, concentrated, aqueons solution, on cooling, in magnificent, colourless, transparent prisms, which are readily soluble in hot water and alcohol, but only sparingly in these solvents i n the cold ; it melts at about 228", and when strongly heated, give& off ammonia, and a crystalline substance distils which is probably the corresponding imide.An a1 y sis. Theory. Fmnd. C,KIoO??C', N.. ........ 19.79 per cent. 19-70 per cent. Phen y limide of Tetra nt e th y lenedica rb0.T y 1 ic acid, yH2*QH*C0 C H2* CH* C 0 > N* C6H5, is readily prepared by heating a mixture of the anhydride of the acid with excess of pure aniline for about 10 minu?es to boiling. The product is poured into a large volume of dilute hydrochloric acid, to remore excess of aniline, and the solid mass which separates is collected, washed with water, and recrjstallised first from 50 per cent. alcohol, and then from methyl alcohol. The magnificent glistening needles thus obtained were dried at 100" and analysed with the following result. Found. r--A- 7 Theory.I. 11. CJl, 1 s 0,. C..... ..... 71-74 - per c x t . 71*6& per cent. H .......... 5.75 - .. 5.4i ,, N .......... 6.45 6-77 ,, 6 Y G ,,OF 1 : %TETWETHYLENEDICARBOXYLIC ACID, ETC. 585 This substance is formed according to the equation >N*CsH, + H20. H2* QH*CO yH2*?H*CO >O + C6H5*NH2 = CH**CH*CO CHz. CH-CO It melts at 127", and crystallises so readily that it 6erves as a valuable means of identifying tetramethylenedicnrboxylic acid It is easily soluble in methyl and ethyl alcohol and in benzene, very sparingly in light petroleum and cold water; hot water, however, dissolves it more readily, and, on cooling, deposits it in magnificent glistening needles. It also crptallises well from a mixture of benzene and light petroleum, or from 50 per cent.alcohol. When heated in small quantities in a test-tube i t distils without decomposition. HOOC H \/ Trans- Tetrameth ylenedicarboxylic acid, yB2*v CH2*C A H COOH This interesting acid was obtained by heating cis-tetramethylene- dicarboxylic acid with concentrated hydrochloric acid for 3-4 hours K t 190" ; the product, which contained ciytals, and also some specks of carbon, was heated t o boiling to dissolve the crystals, filtered, and the filtrate evapornted to a small bulk, and allowed to stand for 24 hours ; the crystals, which had separated, were then collected and recrystallised 3 or 4 times from hydrochloric acid with the addition of animal ckarcoal. In this way, the pure trms-acid was obtained in beautiful, colonrless needles, which, after drying at loo", gave the followiog results on analysis.C . . . . , . . . . . 49-12 per cent. H.. . . .. .. . , 5-57 ,, 3-56 ,, Found. Theory. 50.00 per cent. Trans-tetrnmethylenedicarboxylic acid melts at 131", or about 7" lower than the cis-modificativn, from which it differs in a marked manner i n many important particulars. The cis-acid is very readily converted into an anhydride. It is only necessary to boil it with acetyl chloride for five minutes to completely convert it into its anhydride, which is deposited in beautiful crptals, melting at about 75", when the excess of acetyl chloride is allomcd to evaporate over potash in a vacuum desiccator. Under exactly similar conditions, the trmzs-acid remains quite unacted on, and, even when heated at 160" with excess of acetyl chloride, in a sealed tube for one hour, no anhydride appears to be formed, as mas ~liown by the fact that the crystals which separated as the acetyl chloride evaporated melted at 127-129", and were very VOL. LXV.2 s586 PERHZN: THE CIS- AXiD TRANS-MO11LFICATIOXS readily soluble in water. consisted of unchanged acid. An analysis showed that these crystals Theory. 50.00 per cent. Found. C,H,(COOH):. C . . . . . . . . H.. .. .... 5.72 ,, 5-56 ,, 49.61 per cent. When, however, trans- tetramethylenedicarboxylic acid is repeatedly distilled under ordinary pressures, elimination of water does gradually take place, and, after four distillations, the product has the appear- ance of the anhydride of the cis-acid ; it melts for the most part a t 70°, but a small quantity is always left which does not melt until a higher temperature ; this experiment therefore seems to shorn that trans-tetramethylenedicarboxylic acid on distillation is conrerted partially into the anhydride of the cis-acid.Dr. Walker also found that the cis- and tmns-acids had very different dissociation constants, for the cis-acid (m. p. 138") K = 0.0066, whereas in the case of the tram-acid (m. p. 130-1S1") the constant found was E = 0.0028. Tyans- Pentamethy lenedicarbox y lie mid, C H, < CH,*C /\ H CCOH In investigating this acid, it was found that the method of prepar- ing it previously adopted (Trans., 1887, 51, 244) was not altogether satisfactory, the purification of the crude product by conversion into the ethylic salt as there described being attended with unnecessaq loss of time and material.After many experiments, the best yield of acid was obtained by modifying the process in the following way. The crude ethylic pentamethylenetetracarboxylate, obtained by the action of bromine on the disodium compound of ethylic pentanetetracarboxylate, was dissolved in glacial acetic acid (2-3 vols.), concentrated sulphuric acid (1 vol.) and water (1 ~01.) added, and the whole heated in a reflux apparatus for about two days, until hydrolysis was complete, as shown by the fact that a drop of the liquid dissolved in much water form a clear solution. A rapid current of steam mas then passed into the boiling, brownish liquid until the odour of acetic acid was no longer ap- parent, and the product allowed to remain for two days in a cool place.At the end of this time, the brownish, crystalline crust which had separated was collected, mashed with a little water, and pui-ified by crystallisatmion from water with the aid of animal charcoal. PureOF I : 2-TETRAMETHYLENEDIChFtBOXYLIC ACID, ETC. 587 trans-pentamethylenedicarboxylic acid was thus readily obtained as a colonrless, sandy, crystalline powder, which, after drying at 100°, melted at 160" and gave the following results on analysis. Theory. 53-16 per cent. Found. C,H,(COOH),. C . . . . . . . . H.. . . . . . . 6-37 ,, 6.33 ,, 53-10 per cent. The dark-coloured mother liquors separated from the crystalline acid, if extracted with ether, yield a further quantity of crude acid, which is, however, somewhat difficult t o purify, and is best employed in the preparation of the anhydride of the cis-acid (see below) .The hydrolysis of ethylic pentamethylenetetracarboxylate by the above method is attended with the loss of 2 mols. of carbonic anhydride, and the direct formation of trans-pentamethylenedicarb- oxylic acid. CH,.? (CO 0 C,H,) , C H2.y H* C 00 H HZ < C EX2* C H C 0 0 H + 4H20 = cHz<C E2*C (CO 0 C,H,) , + 2CO2 + 4C,H,*OH. The yield of acid obtained in this m-ay is certainly better than that previously obtained, and, as the operation takes less time, it was adopt,ed in preparing the acid required for this research. The properties of ZrLrns-pentamethylenedicarboxylic acid have already been given (Trans., 1857, 51, 246) ; in addition, it should be mentioned that the solution of the acid in sodium carbonate does not decolorise permanganate in the cold, and only very slowly on boiling ; the behaviour of the acid when treated with bromine and phosphorus is being investigated.Anhydride of Cis-Pentamethy knedicarboxylic acid, > 0. CH,*YH*CO CH,*CHCO CH, < In view of Baeyer's suggestion ( A m a l e n , 257, 179), that trans- pentamethplenedicarboxylic acid should yield an anhydride isomeric with, and readily convertible into, the anhydride of the &-acid, very careful experiment,s on the action of acetyl chloride on this acid mere made with the following results. The acid was somewhat sparingly soluble in cold acetyl chloride, but dissolved readily 03 warming ; the salotion was heated t o boiling in a reflux apparatus for half an hour, but 110 formation of anhydride took place, and on distilling off the bulk of the acetyl ohloride and allowing the solution to stand over potash under reduced pressure, 2 s f L5ss PERKIS : THE CIS- AND TRANS-MODIFICATIONS the unchanged acid crystallised out in nodular masses melting at A second experiment, in which the solution of the acid in acetyl chloride was heated at 100" in a sealed tube for one hour, gave a similar result, as is shown by an analpis of the crystals melting at 154--156P, which were obtained by allowing the product t o evaporate over potash in a desiccator.155-158". Theory. Found. CjH,(COOH)Z. C ........ 53-51 per cent. 53.16 per cent. H.. ...... 6-45 .. 6.33 ,, I n the next experiment, the acid was hested in a sealed tube with a large excess of acetyl chloride at 140" for about one hour.On opening the tube, considerable pressure was observed, and the almost colourless liquid, on evaporation, deposited a thick oil ; this on standing over potash for a few days, deposited a quantity of colourless crystals which were collected by the aid of the pump, and freed from adhering =other liquor by placing them on a porous tile. They melted at 70-72", and consisted of the anhydride of the cis-acid (see helow). I n order to be sure that the isomeric anhydride of the trans-acid was not contained in the mother liquors from these crystals, the thick, somewhat dark-coloured, filtrate, was dissolved in warm dilute potash, the soiution filtered, acidi6ed, and allowed to stand in a cool place. The beautiful colourless c r j s tals which separated melted at 140-141°, and consisted of pure cis-pentamethylenedicarboxylic acid, no trace of the !runs-acid being present.It appears certain, there- fore, that no anhydride of the trans-acid is formed under the condi- tions observed in the abot-e experiments. The best method of preparing the anhydride of cis-pentamethylene- dicarboxylic acid in quantity is to digest the crude trans-acid, obtained as explained on p. 587, with acetic anhydride €or two hours, and then to fraction the product under reduced pressure (160 mm.). After two distillations, the whole passes over at about 220" as an almost colourless oil, which, on cooling, sets to a semi-solid gelatinous mass ; this, after a time, gradually becomes crystalline. Although this substance distils at a constant temperature, it is not quite pure, but, contains, apparently, small quantities of unchanged acid, as the following analpis shows.TheorF. 7 r--- Fonnd . C,H,C,03. CSHS(C0OH)Z. C ........ 58.91 per cent. 60.00 53.16 per cent. H ........ G.22 .. 5.71 6.33 ,, If, however, it is dissolved in twice its volume of acetic anhydride,OF 1 : 2-TETRAMETffPLEXEDICARBOSTLIC ACID, ETC. 589 and the solvent allowed to evaporate gradually over potash under reduced pressure, the pure anhydride separates in magnificent, hard, colonrless, tabnlnr prisms ; these were collected, washed with a little acetic anhydride, and analysed. Found . r--- 7 Theory. I. 11. CjHsC203. C .......... 59.95 59.96 per cent. 60.00 per cent. H.. ........ 5-98 6-12 ,, 5.71 ,, The anhydride of cis-pentamethylenedicarboxylic acid melts a t about 73" to a colonrless oil, which, on cooling, sets to a semi-trans- parent gelatinous mass of about the consistency of pa-affin wax.When rubbed with a glass rod, or touched with a crystal of the anhydride, it gradually crystallises in a very characteristic manner, and it is interesting to note that the anhydride of cis-hexahydro- phthalic acid behaves i n a similar way when melted and allowed to solidify (Bacyer, AnnaleTL, 258, 219). The anhydride of cis-pentamethylenedicmboxylic acid is insoluble in and only rery sloxly attacked by cold water, but it dissolves readily on boiling, and on cooling, if the solution be sufficiently con- centrated, the cis-acid separates in beautiful, glistening ne2dles.The anhydride is readily soluble in ether, alcohol, or benzene, but only sparingly in light petroleum. In the previous paper on pentamethylenedicarboxylic acid (Trans , 1S87, 51, 249), it was nieiitioned that this acid, if heated at 300" and subsequently distilled, yielded an anhydride melting at 64-67". This observation is correct, but the substance, after recrystallisation from acetic anhydride, melts at 72-73'. GO Phenylimide of PentametliylenedicarBoleylic acid, C5H8<CO>NC,H5. In order to prepare this substance, the auhydride of cis-penta- methylenedicarboxjlic acid was heated to boiling with excess of aniline for about 15 minutes, the product poured into water, excess of dilute hjdrochloric acid added, and the whole well stirred until the oily drops had completely solidified.The brownish product was readily purified by recrystallisation once from dilute methyl alcohol, and twice from a mixture of benzene and light petroleum. Found. C,3H],SOp Theorj. N .......... 6.90 per cent. 6.51 per cent. This phenylimide is sparingly soluble in cold water, more readily in hot, crystallising from the solution a'; it cools in magnificent, glistening needles, which melt at 89". It dissolves readily in alcohol,590 XODIFICATIOSS OF TETR~THTLENEDIChRBOXPLIC ACID. benzene, and acetic acid, but is only sparingly soluble in light petroleum. When rapidly heated in small quantities, it distils with- out decomposition, the distillate solidifying to hard, transparent balls, which remain in this condition sometimes for days, but crystal- lise at once when touched with a crystal of the substance. H COOH \/ CHfF Cis- Pentam et hylenedicarboxy Eic acid, C H, < CH2*C /\ H COOH I n order to prepare this acid, the anhydride just described was dissolved in dilute potash, and the concentrated solution acidified and allowed to stand ; the long, needle-shaped crystals which sepa- rated were then collected, and recrystallised twice from water. The analysis of the pure substance, dried at loo", gave the follow- ing result's. Found. +- Theory. I. IT. CjH, (COOH)? C .......... 53.04 53-20 per cent. 53.16 per cent. H.. ........ 6.42 6-40 ,, 6-33 ,, Cis-pentamethylenedicarboxylic acid melts at 140", and is, at this temperature, gradually converted into its anhydride with loss of water, the change taking place very rapidly at 150-160". It is much more readily soluble in water than the trans-modifica- tion, and crystallises from the hot concentrated solution in beautiful, colourless needles, whereas the trans-acid is deposited in the form of a sandy, crystalline powder, the difference iu appearance being, in fact, very much the same as in the case of male% and fumaric acids. The &+acid is readily soluble in acetyl chloride, and, on heating the solution at loo", it seems to be quantitatively converted into its anhydride, which crystallises in plates when t h e product is allowed to evaporate over potash in a vacuum. The cis-acid may he readily converted into the trans-acid by heating it with hydrochloric acid at 180" for two hours ; the sealed tube in which the experiment mas conducted was found to be filled with hard crystals, somewhat discoloured by specks o€ charcoal. The contents of the tube were warmed with water until the crystals had completely dissolved, and the solution filtered ; on standing, the pure traozs-acid was deposited as a sandy powder, which, after recrystallisation, melted at lGJ", and, on analysis, gave the f o l l o ~ ~ i n g results. Theory. Found. C,H,( C OOH) 2. C ........ 53.21 per cent. 53.16 per cent. H . . ...... 6.36 ,, 6.33 ,,ACTION OF HExpIlLETHyLENB: DIBROMIDE ON SOD~UM, ETC. 591 As far as could be judged from a small experiment, the conversion of the cis- into the trans-modification by heating with hydrochloric acid was complete. CltemicaZ Luborutory, Owens College, Munchester.